MAP SHOWING SITES

N
M

4HE

E?tQp.p PIC)ACHAEOLOGcL I)ECE

c_
(

\
. tNDCTfN

SETTLEMENTS

WERE

/L

E \

A
:

c;ns

f+Ih

.
\
.*.. ' .+
\ d,

0 .
4 .

. - -

\- a ' \

?tCQ
\

\ \
I: - b

--..

'
k

, _

k \
\ \ b
A

'

JI

,x ' '

, ' ',

N
\ ,+ -
\

;___ w
"\' - ,
\\

"
'\ k

\\

4 -- --v-v]

TH - cNTT

-
A

TRV

N1
A

:'I

\'

a'

1 +

I I

, 7 '%

-, I *

' p - 1

k ?_ -- a .t_ I I - I --

' a N,

+ a

4; I I )4

* + 4 4' ' A

DRAVID IAN ZETTLENEITS IN CETI)N AND TEE BEGI1NI1GS OF THE KIIDOM OF JAFFNA

by Karthigesu Indrapala

Thesis subziitted for the Degree of

Doctor of Philosophy University of London

1965

2
ABSTRACT This thesis is a study of the settlements founded by Dravidian-speakers from South India, chiefly the Tamile, in Ceylon before the end of the thirteenth century. Although any notable Dravidian settlement was not established in the island until after the conquest of the Caa at the turn of the tenth century, we have included in this study the sporadic and scattered settlements of earlier times as well. The first chapter deals with these earliest settlements and analyses some of the theories put forward by earlier writers on the subject. The main section of the thesis, comprising the second, third, fourth and fifth capters, deals with the settlements established in the northern and north-eastern parts of Ceylon in the period between the beginning of the eleventh and the end of the thirteenth century. This study ends with an examination of the circumstances under which an independent kingdom, controlled by Dravidians, emerged in northern Ceylon. The sixth chapter deals with the events of the first half of the thirteenth century which directly led to the foundation of the kingdom, while the last chapter is concerned with the establishment of the dynaety of Iryacakravartins, from South In ia, ho consolidated the position of the new kingdo

3
ACKNOWLEDGENENS

I wish to express my sincere gratitude to Dr.J.G.de Cas aris, Reader in the History of South an South-east Asia at the School of Oriental and African Studies, who has su ervised the whole of this work and given me invaluable advice and guidance. I am thankful to Nr.W.J.F.LaBrooy, Reader in History, University of Ceylon, for his as i tance in choosin the subject and in obtaining budy leave from the University of Ceylon, which enabled me to undertake this work. I owe a particular debt to Nrs.Indranee Kan iah for her help in re aring the p. My thanks are also due to Mr.H.Somadasa, Librarian, University of Ceylon Library, Mr.Lyn de Fonseka, Librari n, National Mu eum Library, do bo and Nr.S.Tha biaii, Librarian, Jaffna College Library, V ddukoddai (Ceylon) as well as the staffs of the S. .A. . Library, ritish Museum Library and enate ou e Library for th ir help in connection with this work.

4
COTES

Page Abstract . . 2 3 . . . . . 5 6 25

Acknowledgements . Abbreviations Introduction Chapter I Chapter II

The Beginnings of Dravidian Settlements Settlements in the Period of Ca Occupation .

8k

Chapter III

Settlements in the Late Eleventh and the Twelfth Century 133

Chapter IV

Settlements in the Thirteenth Century I - The Jaffna District 26

Chapter V

Settlements in the Thirteenth Century II - Vanni Districts . 306

Chapter TI

The Beginnings of the Kingdom of Jaffna-I 399

Chapter VII

The Beginnings of the Kingdom of Jaffna-'II Lf77

Conclusion

542
549

A Select Bibliography Map .

endpocket.

5
ABBREVLMIONS - Annual Bibliography of Indian Archaeology, Leyden. A.B.I.A. - Ancient India (Bulletin of the Arch. Survey of India). A.I. A.S.C.A.R. - Anchaeological Survey of Ceylon Annual Report. - Ceylon Antiquary an Literary Register, Colombo. C.A.L.R. C.J.Sc. (G) - Ceylon Journal of Science, Section G, Colombo. - Cekarca-ckara-mlai Cc in. C.H.J. - Cey1n Historical Journal, Colombo. Cv. - Cilavasa Dv. - DTpavasa B.C. - Epigraphia Carnatica - Epigraphia Indica Elu-ay . - ju-attanagalvai - Epigr p ia Zeyalanica Gk. - Greek Hvv. - Hatthavanagalla-vi .ra-vasa l.A. - In ian Antiquary J.A.S. - Journ 1 of ian tudies J.R.A.S.(C.B.)Journal of the Royal Asiatic Society (Ceylon Branch) N.S. - New Series Bk. car-kalveu - Km. - KaiIyamlai L. - Latin M.E.R. - Madras Epi raphical Reports (Annual Report on epigraphy Southern Circle,adras Government. Mm. - M4akk4appu-zn iyam - Nahva N.I.A. - New In ian Antiquary ks. - Nikya-saipgrahaya Port. - Portuguese - Pjvalipa Py . Rv. - Rjvaliya 5.1.1. South In ian Inscriptions Sixth. - Sinhalese Skt. - Sanskrit T.A.S. - Trava core Arc ae logical Series - ak*ia- i1ca-purain - Tirikcala-purrn _. U.C. .C. University of Ceylon History of Ceylon U.C.R. - University of Ceylon Review - Vaiyp1 VT. - verses TYrn. - ia-V ipava-nlai

6
INIRODUCT ION

In this work we have attempted a study of the early settlements established by the Dravidians, notably the Tamila, in Ceylon and of the beginnings of the Tamil kingdom in the northern part of the island. This aspect of the history of Ceylon has been neglected for a long time. The seriousness of this gap could be appreciated by anyone who reads the comprehensive history of the island recently published by the University of Ceylon Our subject has not been dealt with at all in this authoritative work. The chapter on the northern kingdom begins abruptly with the rule of the Xryacakravartins and nothing is stated about the beginnings of this kingdom. Until about the thirteenth century A.D., the history of Ceylon was the history of the Sinhalese people. But from about the middle of the thirteenth century, it has been the history of the SinhaleBe and Tamil people in the island. From that time for over three centuries, the majority of the Tamils were concentrated in a kingdom of their own in the northern part of the island. In 1620, the last of the Taml-1 rulers was executed by the Portuguese conquerors who brought the Tamil areas

1. University of Ceylon History of Ceylon, editor-in-chief R.C.Ray, I , pt.l

(1959), pt.2 (1960), Colombo.

7
under their direct rule. Like the Sinhalese in the maritime provinces of southern Ceylon, the Tamils passed through a period of colonial rule, first under the Portuguese and then under the Dutch. Under these two European powers the Tamil areas of northern Ceylon were administered separately from the other areas. In the nineteenth century, after the British took over from the Dutch, the whole country was politically unified and the administration was centralised. This enabled the Tainils and Sinhalese to work together in the national politica and government. During the period of British rule a further wave of Tamil immigrants went to the island as workers in the newly-opened plantations. The descendants of these recent immigrants, whose numbers exceed that of the descendants of earlier Tamil settlers, play a vita], role in the economy of modern Ceylon. These

Tmils are

officially designated Indian

Tamils while the descendants of earlier settlers are called Ceylon Tainils. The Tamils, who comprise nearly twenty-five per cent of the island's population, are now concentrated mainly in the Northern, Eastern and Central Provinces. The chronology and early history of the T-m-i1s of Ceylon have not yet been systematically and scientifically studied, A few works have been written, mainly in Tamil, on the history of the Taniil kingdom u.t many of these could hardly be described as scientific histories. Among the earliest writings

8
on this subject is Simon Casie Chetty's paper' 'On the History of Jaffna, from the Earliest Period to the Dutch conquest', read at a meeting of the Ceylon Branch of the Royal Asiatic Society in

18k7 and

published in the journal of that society

In this short paper, the author has based his account of the early history of Jaffna on the references to gas in the Mahvasa and on the Tenr(1 chronicles, Kailyaxnlai and a-vaipava-

mlai. It is by no means a critical work. This was foUowed by a few Tamil works, almost entirely based on the T-mi1 chronicles and floating traditions. The earliest of these is S.Jobn's pa-carittiram (History of Jaffna), published in In

l879

188k, V.V.Cataciva

Pillai published his

a-vaipavam

( A Narrative of Events in Jaffna) It was followed by A.Muttutampi-pillai's pa-carittiram (History of Jaffna), which

was published in 19l2 These, too, are uncritical narratives embodying almost the entire contents of the Tamil chronicles, with all their mythical and legendary elements. The sections dealing with the period of British rule are useful as source materials for that period since these are contemporary and near-

1. J.R.A.S. (C.B.), I, No.3,

l8k7-k8, pp . 69-79. 1879. Second

edition

2. American Ceylon Mission Press, Jafmna Revised edition 3. Madras,

1882.

1929.

188k.

4. Jaffna, 1912.

9
contemporary accounts. In this respect, K.Veluppillai's compilation, p$a_vaipavakaunniti, is a valuable work Its sections on the administration of Jaffna by British Government agents and on the leading families of Jaffna in the nineteenth century are useful sources for the modern period of the history of Jafmna, Equally important is the section on the place-names of Jaffna, in which the Sithalese origins of over a tbouand names are dealt with This section, despite the fact that it is not&scientifiC study of the place-names, is a useful contribution to the topographic study of Jaffna, which is of utmost value for a work like ours. Almost all the works mentioned above are concerned with the history of Jaffna after the establishment of the Tamil kingdom and do not deal with the history of the Tmila who were settled outside the Jaffna kingdom or with the early Tamil settlements. This is chiefly due to the fact that they are narratives based on the Jaffna chronicles, which deal with the history of the Tamil kingdom only. Nudaliyar C . Rasanayagani' a Ancient Jaffna, published in

1926, marks

the first attempt at a critical history of Jafmna.

1. Vasvir, Jaffna 1918. Two sections of this work have been written by S,Kumaracuvami and S .Katiraverpillai. 2. S.Kamaracuvami, 'Taa-nikatt$a Cila Ia Peyark4i Varalu'.

10
Unlike the earlier works, Ancient Jaffna is the result of an attempt to trace the history of the

Tmi1

of Ceylon from the

earliest times to the sixteenth century. It has been based on a wider variety of sources and much effort has gone into it. For the first time the Sixthalese sources as well as the South Indian inscriptions were consulted. It marks a leap forward in the research into the history of the Tamils of Ceylon. But despite its distinct merits,Rasahayagam's work suffers front several serious defects. The work has been marred by an earnest attempt to prove the thesis that the Tamils were settled in Ceylon from pre-Christian times and that there was an independent1 in northern Ceylon which existed front about the fifteenth century B.C. to the seventeenth century A.D. In his attempt to prove this thesis, Rasanayagam has used methods which are questionable and materials that are totally unrelated to the history of the Tamils in Ceylon. These have been briefly pointed put in our work. A more critical and, in many respects, a better work on the history of Jaffna is Fr.Gnanapragasar's 1a-

Yaip ava-viinarca (A Critical History of Jaffna), published in

l928

It stands in great contrast to the disappointing articles

of the same author published posthumously in the Tarnil Culture,

1. Accuvli, Jaffna

1928.

11
in l952 It may be reckoned as the most valuable study of the

early history of Jaffna in Tmil. Besides the South Indian inscriptions and the Sinhalese sources Ganapragasar has made use of place-name materials also for his study. The same author's Kings of Jaffna, published after the Taml-1 work, deals exclusively with the history of the sixteenth and seventeenth centuries. Among the other works published around the same time and a little later are K.Matiyaparanazn's vaipavam2and Civanantan's a-p1rv!ka

__am These two &lso

fall under the category of the earlier uncritical works. Two other works of recent years, which also fall under the same category, are K. Kanapathi Pillai' s Ilakai-v-Tami Vara] and C.S.Navaratnani's Tamils and Ceylon The University of Ceylon History of Ceylon has & chapter on the kingdom og Jaffna, entitled 'The Northern Kingdom', by S.Natesan As mentioned earlier, this section

1. 'Ceylon Originally a Land of Dravidians', T.C., I, No.1, Feb.1952; 'The Taniils turn Sinhalese', T.C., I, No.2, June 1952; 'Beginnings of Tamil Rule in Ceylon', T.C., I, No.3, Sept. 1952. 2. K.Matiyaparanam, 3. Kuala Lumpur, 1933. k. Peradeniya, 1956. 5. Jaffna 1958. 6. U.C.H.CI, I, pt. 2, pp. 691-702. pp;a-prvjka-vaipavam, Jaffna 1927.

12
begins abruptly with the reign of the iryacakravartins. It is stated that the earlier part of the chapter was deleted by the editor. As a result, this chapter falls outside the period and subject matter dealt with in our work. The latest and among the most critical of the contributions to the history of the Jaffna kingdom is the article by S.Paranavitana in the Journal of the Royal Asiatic Society, (Ceylon Branch), l96l in which the theories put forward by Rasan&yagam and other writers are analysed. In this penetrating an @ysis th. author has made use of new evidence as well. But some of his arguments have been marred by his attempt to give a Jvaka origin to the founders of the Jaffna kingdom. We have discussed these arguments in our work. Only the history of the Jaffna kingdom forms the subject of almost all the works mentioned above. These do not deal with the history of the Tamil settlements that preceded the foundation of the kingdom. This reason, above all others, has recommended itself to us for undertaking a study of this nature. The major part of our work, five out of the seven chapters, is exclusively devoted to a study of the Dravidian settlements that were founded in Ceylon between about the ninth and the end of the thirteenth century. This section serves as a background

1. J.R.A.S. (C.B.), N.S., VII, pt. 2, 1961, pp. 17k-22k.

13
to the rise of the Tamil kingdom of Jaffna, in northern Ceylon, which forms the subject of the last two chapters. The term Dravidian is used in this work to mean the different communities of South India speaking the Draviian family of languages, chiefly Tamil, Telugu, Kannada and Nalayalam. Although the only Dravidian-speakixig community to be found in the island now are the Tamils, there were settlers from the Kannada, Kerala and Telugu countries, who were ultimately assimilated into the major Dravidian group or into the Sinhalese population. The kingdom of Jaffna in this work refers to the Tamil kingdom of northern Ceylon which was founded in the middle of the thirteenth century and ceased to exist in 1620. A historical study of the early Dravidian settlements in Ceylon, like that of early settlements in any country, presents a number of problems that cannot be s4ved pirely with the help of such materials as chronicles and inscriptions. Other branches of studies such as archaeology, physical anthropology, historical geography and historical linguistics have an impoitant part to play in the solution of these problems. These problems would include among others the determining of the original home of the settlers, the causes of their migration, the routes of migration, the areas of settlement and the extent

14
of the survival of earlier inhabitants. The evidence of archaeology is very helpful in tracing the routes of migration and locating the areas of settlement. The historical linguist has an important contribution to make by his analysis of the place-name evidence, which helpd a good deal in the understanding of the social conditions under which the settlements took place and the institutional ties which first bound the settlers together as well as in the location of early habitation sites Place-names also help to an extent in the inquiry into the survival of earlier inhabitants. The historical geographer could help in the understanding of the influence of such factors as physique and defence on the location, and sometimes on the form, of the settlements. Sometimes the contribution of physical anthropologists is also valuable. In Britain attempts have been made, though not with much success, to use the evidence derived from cephalic indices and tables of nigrescence in the study of the Anglo-Saxon settlements. In this study of the Dravidian settlements, the use of evidence from sources other than inscriptions and literary

1. In this respect the place-name evidence in Britain has been of immense help to the historians of(Anglo-Saxon settlement. See A.wer and F.M.Stenton, An Introduction to the Study of English Place-names, pt.l, (Camb. Englaiid),

1929.

15
works has been rendered difficult. Even the inscriptions aiid literary works that we have used have proved to be inadequate in the reconstruction of a satisfactory history of the settlemerits and in the solution of many iiiportant problems. 1hile the Pli and Sinhalese chronicles of the island provide very reliable, fairly adequate and surprisingly continuous information regarding the political, and to an extent the religious, history of Ceylon, their contribution to our inquiry is very little. The activities of the Dravidians in Ceylon find mention in the chronicles only when these affected the political and religious affairs of the country. No evidence relating to the Dravidian settlements is available in these sources. We have made use of these sources mainly in the discussion of the circumstances under which the settlements were established arid the northern kingdom emerged into existence. On the Tamil side, the chronicles that are extant are those written nea1ly three centuries after the foundation of the Tamil kingdom. These are the ICailyamlai, Vaiypal, Vaiy, a-vaipava-mlai and the Maakk4appu-mmiyam.

The chronicle Ircamura, mentioned in the Ciappuppyiraiu (preface) of the pa-vaipava-mlai, is not extant now.

With the possible exception of the the other works cannot be dated exactly. But, as we shall see presently, certain references in these works make it clear that

16
these were all writtem after the fifteenth century. The Vaiypal is probably the earliest of these chronicles The mention of Parafiki (Portu uese) as well as the occurrence of certain Portuguese words in this work suggest that it was composed after the arrival of the Portuguese in the island (A.D. i5O5) Only one clumsy manuscript of this work, full of orthographic mistakes, has survived. 1n its present state, it is very confused and at times unintelligible. Fortunately, an old prose rendering of this chronicle has survived and it is with the help of this work that we are in a position to understand the Vaiyal. This paraphrase is known as Vaiy The early part of this chronicle is based on the Pxnyaa, the popular versions of the Viaya legend and on the popular etymology of some of the place-names of Jaffna. The sections dealing with the Dravidian settlements of the thirteenth century and later appear to have been based on certain genuine traditions which were current in a confuaed

1. Vaiypuri iyar, Vaiypal, ed. J.W.Arutpirakacaia, Jaffna 1921.

2. Parafdd. 1= L. Franci) is the Timi1 name for the Portuguese. The occurrence of the names Parafiki (v.3k) and Piippa (Philip

v.58) and

such words as ayutanti (Port. adjutante) and

puravar3taiyar (Port. provedor) may not all be due to interpolations.

3. Vaiy, ed. Z.Gnanapragasar, Jaffna 1921.

1?
form when this work was written. According to tradition, Vaiypuri

Aiyar, the author of this chronicle, was the court poet of one of the kings of Jaffna who bore the consecration name Cekarcack.ara As Gnanapragasar is inclined to believe, 'it would seem that the Vaiy was composed during the times of the last Jaffna kings' The contents of this work have been critically analysed and used with caution in our work. The Kailyamlai, a chronicle of the Kailyantar temple in Jaffua, contains an eulogistic account of the kings o Jaffna and. appears to have been composed at the beginning of the seventeenth century. It mentions the letupatis of 1?Rmnad, the first of whom began his rule around l6O 1 ^ Some sections of this work have been based on the Vaiypal. Perhaps the most
useful section is that dealing with the settlement in Jaffna

of certain families from the Tamil country. The Y pa-vaipava-mlai is a prose chronicle of
the Jaffna kingdom and was written, as stated in its preface,

when the Dutch Conimsindant Ian Maccara (Mkka.1a) was administering Jaffna (A.D. 1736). As admitted in the preface, the author has

1.

title page.

2. 'Sources for the Study of the History of Jaffna', T.C., II,

Nos. 3&k, p. 3]Lf, fn.18. 3. Mutturca Kavircar, Kailyamlai, ed, C.V. Jampulinkazu Pillai,,.< C.Rasanaagam'a ForewGrd, ibid., p. lf ; J. .A.S. (C.B.), N.S., VII, pt. 2, p. 176.
S

t1f%mts I93i

18
based his work on the Vaiypal, Kailyamlai and the two non-extant works Ircamur and Pararca-ckaraul The sections dealing with the period beore the Portuguese rule reproduce almost entirely the contents pf the Kailyam1ai and the Vaiypal. The 4akaappu-mmiyam is a chronicle of the Batticaloa district of the Eastern Province of Ceylon Its existence is not known to many writers on Ceylon history. This prose chronicle, in its present form, appears to have been written in the eighteenth century, for, it deals with the Dutch rule in Ceylon. Though a late work, it embodies many genuine traditions of earlier times which are remarkably corroborated by the P].i and Sizihalese sources. It is the only Tamil chronicle that preserves any memory of the very early times. It is also the only Tamil chronicle that mentions Ngha by name and des with his activities in Ceylo4. We have discussed these merits in our work. It main use for our work has been in the reconstruction of the history of Dravidian settlements in the Eastern Province and the rise of Vanni chieftaincies there. The traditional historical poems relating to Batticaloa, appended to the Mafakk4appu-mmiyam, have
P1o.yslv.iQb

ed. LCapanatan, Colombo 1 953 ; Eng. tr. C,Brito, Colombo 1879. 2. Maaidc4appu_mniyam, ed. F.X.CNataraca, Coloinbo 1962.

19
also been useful in this respect. Besides these chronicles, a few other Tamil works of Ceylon containing valuable historical information have also remained extant. Among these, the T ir-kc ala-puram Kc ar-kalve Cekarc a-ckar a-nilai1 and the Cekarca-ckaram have been of some use in rour work. The first three are chronicles of the temple of E5ttvaram, in Trincomalee. The exact date of these works cannot be erxnined. The Takiia-kailca-puram, written in the reign of a king of Jaffna who bore the consecration name Cekarca-ckata is probably a work of the fifteenth or sixteenth century. The Kcar-kalvetu and the Tiri-kcala-purtam are later works. The Cekarca-ckara-mlai is an astrological work composed in the time of an Iryacakravartin named Vartaya who had the consecration name Cekarca-ckara According to the vaipava-mlai, this ruler was the father of ?rttaa whom Paranavitana has identified with ?rttam Peruxn j un of the ia-

1. Ed. P.P.Vayittilinka Tecikar, Point Pedro, Jaffua, 1916. 2. Ed. A.Canmukarattina Aiyar, Jaffna 1909.

4. Ed. I.S.Irakunata Aiyar, Kokkuvil, Jaffna 1942. 3. Appended to Takia-kailca-puriam. . , 7:116.

6. Ccm., v.158.

20
Ndav4a inscription, dated in the third year of VikramabThu III

(l36o) If the identification is correct and if the statement

of the is to be accepted, the Cekarca-

ckara-nilai may be dated to the first half of the fourteenth century. Some are inclined to date this to the fifteenth century The Cekarca-ckaram is a medical work, the date of which cannot be determined easily The historical information, relevant to our studs, contained in all these works is incidental and very meagre. None of the Tamil works mentioned above contains any reliable information concerning the Dravidian settlernena in Ceylon prior to the thirteenth century. These have not been, therefore, made use of in the major part ofour work. Even for the period after the thirteenth century, these sources are full of legendary material that it has to make much use of their evidence. For the major part of our work, we have depended mainly on epigraphic and archaeological materials. Though the evidence of these materials has been far more encouraging than that of the literary sources, it has been by no means adequate. b L been difficult

1. Yvm., p . 37 ; S .Paranavitana, 'The rya Kingdom in North Ceylon', J.P.A.S. (C.B.), N.S., VII, pt. 2, p. 197. 2. U.C.H.C., I, pt. 2, p. 691 ; C.A.L.R., V, p. 175 ; J. .A.S. (c. .), N.S., VIII, pt.2, p. 372. 3. The verses of this work used here are those quoted in the introduction to the Ccm.

21
Excavation work is still an undeveloped branch of archaeological research in Ceylon. As long as excavation work remains undone, much that is relevant to our study will be wanting. For the period prior to about the third century A.D., we may reasonably expect a few sepulchral and other remains, which are invaluable for a study of settlements, to be brought to light. The only burials relating to Dravidian settlements in the island, namely those of Pomparippu, were discovered by chance and today, nearly forty years after the discovery, the sire still awaits a proper scientific excavation. For the period after the third century, it is aiva and Vaiava temples and icons as well as Tanill inscriptions that will help us in our inquiry. Here, too, owing to the lack of excavation work, we have to depend solely on surface finds. Archaeologists have not helped us so far to know something of the earliest aiva temples, such as the Tiru-ktTvaram temple at Mahtittha, referred to in the literary sources. No surface remains of these exist now and only an excavation of the sites is likely to yield something of value. The fer ancient temples so far unearthed have been f immense value in locating and dating some of the earliest settlements of the Dravidians. Of greater value for our work are the Tamul inscriptions. More than a hundred of them have been discovered in the island and nearly half of them are unpublished. These epigraphic records1 most of them belonging to the eleventh and twelfth centuries, have helped us a good deal in the location, dating and the

22
determination of the nature of some of the settlements. The material for the first three chapters is mainly derived from these inscriptions. We have not been able to make use of the evidence of physical anthropology, historical geography and place-name studies for the simple reason that no work has been done in these fields so as to be of any help to us. A preliminary survey of the place-name material shows that much valuable information could be gleaned from it for our study. For instance, the earlier Sinhalese occupation of the Jaffna peninsula, the long survival of the Sinhalese there and the Taniil occupation of the North-central Province before the Sinhalese resettled there are unmistakably indicated by the place-names. The collection and analysis of these toponyms require a proper linguistic training Besides, the establishment of sound-pecligrees with the help of earlier forms and the analysis of sound and word substitution and Sinhalese-Tamil compounds are beyond the scope of our work. But wherever possible, place-name material has also been used though never as an independent evidence. Some attempt have been made by certain physical anthropologists to analyse the physical characteristics of the people of Ceylon. Their surveys are neither exhaustive nor

23
complete and the results are not of any help to us Perhaps not much could be expected from physical anthropologists even in the future owing to the complex nature of the problem. It ma be difficult to contend that differences between human communities are easily recognizable in differences of physical structure. Distinctions based on physical characteristics may be unreliable in the present state of knowledge. Even if these were reliable neither the Sinhalese nor the Tamils of Ceylon can be regarded, in view of their previous history, as a sufficiently homogeneous group to enable any superficial distinctions to be used with confidence in their difUrentiation. In view of these limitations and difficulties, while we may claim to have added something to our knowledge of the history of the Tanmils of Ceylon, the account presented here is inevitably incomplete and not always definite. We have often been led to state our conclusions in hypothetical terms. As one Indologist has remarked, 'they are better than no conclusions at all or than categorical assertions based on inadequate evidence'.

1. N.D.Wijesekera, People of Ceylon, Colombo 1951. P.I.Chanmugamn, 'Anthropometry of Sinhalese and Ceylon T(1s', C.J.Sc. (G), IV, pp. 1-18 ; Marret and Wijesekera conducted an ethnological survey of Ceylon, the materials of which are in the National Museum, Colombo and at the University of Harvard.

24
This is especially so regarding the beginnings of the kingdom of Jaffna, where the gains from this research, valuable though they are, have not increased our knowledge of the origins of the kingdom. With the progress of archaeological research and place-name studies, we hope these limitations could be overcome to a great extent. As we have stated earlier, the materia]. used in the first part of our study dealing with the Dravidian settlements is mainly derived from sources hitherto untapped. These include nearly a hundred Tamil inscriptions, about half of which are unpublished, and the Tamil chronicles. Most of the Tamil inscriptions and. the Tainil chronicle Maak4appu-mmiyam have been used here for the first time. In this sense, a substantial section of the first five chapters forms an original contribution to our subject. In the transliteration of Tamil names and. words we have adopted the system used in the Madras Tamil Lexicon. However, in the case of more familiar names, we have used the transcription that is familiar to all Indologists (j., agam for Cafikani and P4ya for Piya). We have usually broken up the longer compounds with hyphens and simplified the junction of words ao as to facilitate the understanding of their meaning.

25
CHAPTER I

TEE BEG INNIS OF DRAVIDIAN SETTLEIEIITS Front the earliest times to the end of the tenth century A.D.

No appreciable light is thrown by either tradition or archaeology on the darkness in which the history of the earliest Dravidian settlements in Ceylon is shrouded. The archaeological finds so far have not been very encouraging and few definite conclusions can be drawn from the little that has been discovered. The Pli and Sinhalese chronicles furnish some evidence regarding the political relations between the Dravidian kingdoms and Ceylon, but contain little information on Dravidian settlements in the island. The late Tamil chronicles of Ceylon, on the other hand, hardly preserve any memory of the very early times. Under these circumstances, one has to piece together the hopelessly meagre evidence in the above sources to determine the chronology and nature of the early Dravidian settlements in Ceylon. It has been claimed by certain writers on the history of Jaffna that the people of northern Ceylon at the time of the earliest Indo-Aryan settlements, called gas in the chronicles, were T,n11s Some others have claimed that these gas were Tantil

1. S. Gnanapragasar, 'Ceylon originally a land of Dravidiana',

LQ .

I, No.1, pp.27 If.

26
in culture and language, although ethnically they were not Dravidian These conclusions, as we shall see presently, are based on the legendary accounts of the gas in the P1i chronicles and the Tamil Buddhist epic Maini!kalai as well as on the erroneous identification of some of the place-names mentioned in early Tami]. literature. Gaaaapragasar, a leading proponent of the theory that the Ngas of the Pli chronicles were TRndls, has put forward four n'xi{n arguments in support of it In the first place, he baa argued that the island of Ceylon as well as the language spoken there were known in ancient times as 1ain and that the name of the language was later corrupted to Eu. These factors, in his opinion, 'should lead one to conclude prima facie that, at the earliest times, am was occupied, at least in the main, by a Tamil-apeaking people' This argument is far from logical. Presumably it rests on the fact that am is now used only in Tamil as a name for Ceylon. But the origin of this name, far from indicating that the island was occupied by Taniil-speaking people in ancient times, shows that the people from whose name ain is derived were Sinhalese. The earliest occurrence of this name is in the Brhxri! inscriptions of South India. In these

1. C.Rasanayagarn, Ancient Jaffna, pp. 13 if. 2. S.Gnanapragasar,'ceylon originally a land of Dravidians', pp.27 ff 3. Ibid., p.

30.

27
inscriptions, from Tirupparafkuam and Sittaaval, occurs the Prit form of this name, namely Evidently it is from

this Prkrit form that the Tamil Tam is derived. It could be shown that a is derived from Si *h4a through the Pli STh4a, or more probably through another Prkrit form Sihi.a. The name Sih4a has two elements, s14 a and a. The Sanskrit siria becomes sha in Pli STha becomes sihi and a! (the consonant Ii is dropped when its position is between two similar vowels and the two vowels coalesce) in Sinhalese Si*h4a could, therefore, have become Sihija and later SIa in early Sirthalese, aJr4. probably even in other Prakritic languages, although no record of such a form has survived in Sinhalese It is not difficult

1. C.Narayana Rao,'The BrThini Inscriptions of South India', N.I.A., I, pp. 367, 368, 375.
2. ., 6:10.

3. S.Paranavitana, Sigiri Graffiti, I, p.xci. becomes Si-gin in Sinhaleae.

Siiha- in

I.

Cf., Sihila (Sih4a) and Sihilaka (Si*h4aka) in a aroh inscription from Loriyn Tangai, in Wsat Pakistan, belonging to about the second century LD. S.Konow, Kharoth! Inscriptions, p. 110; U.C.H.C, I, pt.l, p. 90. j., also Sielediba in The Christian Topography of Cosmas Indicopleustes, ed. F.O.Winstedt, p. 250.

28
to derive the forms

Tja and

Iam from Sja. In the early period,

when Sanskrit and Prkrit words were borrowed into Tamil, those with the initial letter a often dropped that consonant The name Sja, when used in the Tamil country, would, therefore, have become a, as indeed it occurs in the pre-Christian Brhm* inscriptions there Since the a in Prkrit and the a in Tamil are interchangeable a would become a and the final

form am is derived front this by the addition of the consonant in, which too is in keeping with the rules governing the form of borrowed Sanskrit and Prkrit words ending with the vowel a, 14. especially the neuter nouns or those designating inanimate objects. Thus, am could be derived from the name Sih4a and would, therefore, mean the land of the Sinhalese rather than indicate that Ceylon was originally settled by Tamils. Guanapragasar's arguments, on this score, will become groundless. The derivation

l.., Pkt. sipp (Skt. sukti)>

Tiinil ippi

ippi in Maimkalai,

XXVII, 1.6k; Cittnta-cikmai, 23:5 ; Madras Tamil Lexicon,I,p.297; cf. also Pkt. slsa>Tamil jyant ; Skt. san hi> Tamil anti. This probably occurred in the Prakritic languages,too. Cf. Skt. Sih4a> Sinh. He or Heja > E.n and Skt. Si* 4a seems to have become Ia in Ia-nga (My . ,35:l5). 2. C.NarayanaRao, 3. i. . p. 375.

., Damia> Tamiar ; Co'a >

C1ar.

Skt. mafig4a ) mafik4am ; Skt. si*ha , cifikam

29
of !2am from Sihaa is accepted by leading Tamil scholars Secondly, Gnanapragaar has argued that the original inhabitants of Ceylon came from South India and that these pre-Aryaii aborigines were Dravidians who seem to have spoken a Tamil dialect He bases this on the assumption that the preAryan inhabitants of India represeit an earlier wave of immigrants from the Mediterranean area and that no trace og any langtage other than Tamil is found in India till the arrival of the Indo-Aryans. Although the pre-historic relations between India and Ceylon are undeniable the rest of his arguments are based on mere assumptions. It is not true to say that all the non-Aryan inhabitants of India were necessarily Dravidian. There were others as well, chief among whom were the Mu-speaking people The chronology of the Dravidian migration to India is itself an unsettled question There is no evidence to suggest that Tamil was the only language spoken in India in pre-Aryan times.

1. S.Vaiyapuri Pillai, Madras Tamil Lexic n, I, p. 382;

S.Krishnaswaniy Aiyangat in the Preface to S.Rasanayagam's Ancient Jaffna, p. v.

2. S.Gnanapragasar,'Ceylon originallyland of Dravidians', p.30. 3. U.C.H. ., I, pt. 1, pp. 75, 79. If. K.A.Nilknta Sastri, History of South In ia, p . 59.
5. Cf., C.von FUhrer Haimendorf,'New Aspects of the Dravidian

Problem', T.C., II, No.2, p .131. The author dates the Dravidian migration to the first millenium B.C.

30
His third argument is that 'hundreds of Taniil placenames in Ceylon are pre-Sinhalese' He has given a few examples of ele ents of present-day Sinhalese place-names and what have been considered by him to be their Tamil origins. It is clear that this argument is based on superficial similarities and not on any historical study of the development or evolution of these names. This could be seen in the two sets of elements as well as from their phonological development. He has clai ed, for instance, that the Sinhalese element dea, meaning 'low-lying land or valley', is derived from Tamil ti, meaning corn But dea and its more common variant de are derived from

Sans1rit droi (=valley), through the Phi doi and medieval Sinhalese doa and The fourth argument that Sinhalese is based on Tamil and that, therefore, 'the original inhabitants of Ceylon' spoke Tamil is unconvincing Gnaaapragasar arrives at this conclusion by adopting unscientific methods in his linguistic research. One can only quote the views of Wilhelm Geiger on this matter:Gnanapragasar's methods are not at all Indian; they are simply a relapse into the old practice of comparing two or more words of the most distant languages merely on the basis of similar sounds without any consideration for

1. S.Gnanapragaaar,'Ceylon originally a land of Dravidians', p.31. 2. Ibid. 3.

Also, If.

Skt. Jambu-droi ) Phi Jambu-doi ) Sixth. Daba-dei and Daba-deiya. Cv., 81:15;
ip c nsciipven s

Pv., p . 119;

u.c.riJ

., p. k5.

S.Gnanapragasar,'Ceylon originally a land of Dravidians', p.31 ff.

31
chronology, for phonological principles, or for the historical development of words and. forms. 1 Similarly the attempts of Raaanayagam to show that the Ngas of Ceylon referred to in the Pli chronicles were Tamil in culture and language are based on the erroneous identification of some place-names in the Tamil Safigam texts, without any consideration for chr.ogy or for known historical facts An analysis of these early Tamil poems shows that the geography of their account is mainly confined to the Tam.tl country of their time, which was bounded in the north by the Vkaam (Vgaam) hills, in the south by Kuzriari (Cape Comorin) and on the west and east by the seai There is no indication in any of the poems that chieftains and rulers from outside these limits were eulogised by the Tamil poets. A notable exception is the Arya king Pirakatta, (Brasta) who is mentioned in the colophon of Kapilar's ifici-pu Despite this factor,

Rasanaya,gam has tried to identify -ilaAk8i, }ntai and ICutirai of the aigam poems with Ceylon, Mahtittha (in Ceylon) and Kutirai-malai (in Ceylon) respectively The

)-ilMciii

of

1. W.Geiger, A Grammar of the Sinhalese Language, p. vii. 2. C.Rasanaragaii, . cit., pp.

13_h14.

3. Tolkppiyam, Pyiram, 11. 1-2 ; Cilappatikram, VIII, U. 1-2.

k.

ifici-pu, Pattu-pu , ed. U.V.Cuvmfnata Ayyar, p. . cit., pp.

380.

5. C.Rasanaagam,

19-26.

32
the Tamil poems cannot be identified as Ceylon. It was a chieftaincy in the Tamil country when these poems were composed and has been identified as a region in the North Arcot district It is held to be the same as the Uttara-lk! of the Ca inscriptions There were also other places in the Tamil country with ilaikai as the chief element of their names which find mention in the agam poems. We hear of To-n-ilaikai, KIm-ila.fLkai and Nau-n fl-ni-ilai3.kai Of , To-m-ileAki

is considered by some to represent Cey1on This may or mar not be correet, for there is no evidence in the Tamil poems to identify it properly. However, Ceylon was not the only place known to the Tamils as I].akai. In the earliest literature and inscriptions of the Tamils Ceylon is referred to as a or In later times, the names Ci.kajam and
Ilafikai

were also used.

But when Ilakai was ued to denote Ceylon, it was usual to qualify it with some epithet so as to distinguish it from the

1. K.A.Nilakiinta Sastri, The Cas, pp. k35, kk2 fn. 83 V.Kanakasabhai, Tamils Eighteen Hundred Years Ago, pp. 27, 29 ; J.R.Narr identifies it as a place near Dharmapuri, The Eight Tarnil Anthologies with special reference to and

Patiuppattn, thesis submitted to the University of London, 1958. 2. See infra, p. 3. u-p uppaai, 11. 119-120. . cit., p.27 ; J.R.Narr,
p.375 ;

V.Kanakasabhai,

. cit.

5.

C.Narayana Rao, op.

applai, 1. 191.

33
other I1a..kais. The Cilappatikram refers to Ceylon as 'kaa1 Ilaiikai' (Iiakai surrounded by the sea) 1 while the Main1kalai cUs it Ila3kti:pam (Skt. Laik-dvIpa) The most common epithet
was tea, meaning south, thereby denoting that it was the Lk

in the south (Te-i1kii) By about the Ca period when Ceylon became the La.k par excellence such epithets bece.xne unnecessary. The place named ?ntai in the early Tami]. poems is also different from }Iahtittha, which is now known as }ntai. Intai is a recent name for the ancient port of Mahtittba. The name does not occur in any of the early works. In the Sixihalese inscriptions arid literature, Nahtittha is referred to as Ntoa, Mahavoi, Mahapuu, Mahavun, havautoa and Mhapaana ' In the Tami]. poems of about the seventh, eighth and. ninth centuries and in the Ca inscriptions of the eleventh century, the Tamilised form tam has been used By about the seventeenth century, the variant jam was in use In Ptolemy's map this

1. Ci1appatikram,

p.636.

2. M4inka1ai, XXVIII, 1. 107. 3. See infra, p . ++&.

k. C.W.Nicholas, 'Historical Topography of Ancient and Medieval

Ceylon', J.R.A.S. (C.B.), N.S., VI, 1959, pp. 75-81. r Tvra Tiruppatikafik4, Tirumuai, 2, Patikam 2k3 and Tirumutai Pabi1t&-381, pp. 518, &; S. 1.1., IV, Nos. ].k12, ]I].k.

3,

6. Cf., Mantotte in Memoirs of Rijckloff van Goens 1665, Tr. S.Pieters

p . 106.

34
port is named }Iodouttou The present name of 1ntai is evidently an abbreviation of I.ntam There is no evidence whatsoever to identify Jntai of the agam poems with Nahtittha. Similarly, the identification of Kutirai and Nutirarn with Kutiraimalai

in

Ceylon is untenable. Not only was there a

place by the name of Kutirai in the Tinii1 country there is also no reason to suppose that the name Kutirainialai for the place on the north-western coast of Ceylon was in use at the time of the agam poems. It is clear from the references in the Tm(1 poems that Kutirai and !4utiram were chieftaincies

in

the Cra

kingdom The argument that Ptolemy's Hipporos (Gk. hippos = horse, oros = mountain) is identifiable as Kutiraimalai (Tamil kntirai = horse, malal = mountain) and that since Hipporos is a direct translation of Kutiraimalai, the Tami]. name was

in uBe

in Ptolemy's

time, is also not convincing. It is possible that the presentday Tamil name is itself a translation of an earlier Sinhalese name. We find that there is still a place called Avagiri (Skt. ava = horse, girl = mountain) very close to utiraimalai. Possibly Avagiri was the earlier name, covering a larger area,

1. C.W.Nicbolas,

p. 75; U.C.H.C., I,
. cit., pp. 113, 118.

pt. 1, map facing p.8.

2. Cf., Taflcavr > Taflcai. 3. V.Kanakasabhai,

I.

v.168 . J.R.Marr, in the work mentioned above,

locates Mutiram near Udamalpet.

35
and Kutiraimalai may be a Tamil rendering of later times However, the identification of Hipporos with Kutiraimalai is itself in doubt. It is admitted by critical scholars that the legendary accounts in the Pli chronicles about the Igaa are quite unreliable Even if there were a people called 1gas, there is no evidence to suggest that they were Tamfl- in language and culture. There were persons with the name Nga all over India. Even to this day we find a people called Ngas living in Northeast India. The Ngas of the chronicles, like those of many Pli and Sanskrit works, seem to be superhiinn beings Rasanayagam' s arguments for the existence of Taniil settlements in Ceylon in pre-Christian times, therefore, are wholly unacceptable1 Ceylon's geographical proximity to and close contacts with the Tami]. country and. early conquests by TRnr1 adventurers have been often used as the basis for the assumption that Tamils were settled in the island in the early centuries of its history.

1. There are other place-names on the north-western coast of 6eylon which are Tamil renderings of Binhalese names. Cf., Si13.h. J4agult o a-mune >Tamil Kaliy;a-tuai-mukam. 2. S.Paranavitana, 'The irya Kingdom in North Ceylon', pp. U.C. C., I, pt.l, p. 3. See infra, p.
11oIt.

180-183;

95.

k.

See infra, pp.

36
Our sources undoubtedly indicate that Tamils had established contacts with Ceylon by about the second century B.C., if not earlier. There is reliable data in our sources relating to the commercial, cultural, political and religious connections between South India and Ceylon in the early centuries of the island's history. One of the earliest references to such contacts occurs in the Akitti taka This taka story alludes to the intercourse between Kvripattinam, in the C1a country,and K.radipa, near NgadIpa. Ngadipa is identifiable with the Jaffna district which was known by that name in the pre-Christian and early Christian centuries Kradipa appears to be the island of Kraitivu, about two miles west of the Jaffna peninsula The Dlpavai1lsa and the Mahvaisa refer to the two Tamil usurpers, Sena and. Gutta.ka, who ruled from Anurdhapura in the second century B.0 They appear to have been connected with the horse-trade in the island. According to the I4ahvasa, their father was an a sa-nvika or ship's captain dealing with horses Sena and Guttaka were followed by the Tamil poli. tical adventurers, Era, Pulahattha, BAhiya, Panay nira, Piaymra

1. The Jtaka, IV, ed. E.B.Cowell, tr. W.K.D.Rouse, p. 150. 2. G.P.Malalaaekera, Dictionary of Phi Proper N nies II, p.k2. 3. C.W.Nicbolas, . cit., p. 8k ; Malalasekera has i entified . cit., I, p. 570.

it as 'an island in the Dami4a country',

k. Dv., l8:k7 ; Mv., 21:10. 5. Mv., 21:10.

37
and Dhika, who ruled at nurdhapura for short periods in the second and first centuries B.0 Among the paramours of Queen Anul were two Tamils, who also rpled. at Anurdhapura for some time in the first century B.0 In the first century A.D., 4anga

(33_li 3) went over to South India and took mercenaries to win

back his throne These merceflarie8 were probably supplied by some chief or ruler, apparently not unconditionally. For, we find that his son, Candukhasiva, was married to a Tamil lady who came to be known as Dami-dev Paranavitana is of the opinion that 'this alliance of his son with a TRmil princess was, perhaps, a part of the price which Ianga had to pay when he obtained military aid. from South India against his adversaries' But it is also possible that it was the result of a friendly alliance between a Tamil chief or ruler and Ianga. Thus, the

1. Mv., 21:13 ft.,

33:39 ft. ; Dv., 18:k9, 20:16-18. The relations

between South India and Ceylon during this period have been dealt with in great detail by W.M.K.Wijetunge in his thesis, The Rise and Decline of Ca Power in Ceylon, submitted to the University of London in
2. Nv.,

1962.

31f:l9, 26 ; Dv., 20:27, 29.

3.

., 35:26, 27.

11. Ibid., 35:k8.

5. U.C.H.C., I, pt.]., p. 176.

38
evidence of the Pli chronicles shows that from about the second century B.C. the Tamils of South India had established contacts with the island. The earliest literature of the Tamils, belonging to about the second and thud centuries A.D., does not contain information on this point. But there is a solitary reference in one work, the Pa applai, to trade relations

with Ceylon. It mentions the vessels laden with food-stuffs from 1am (Ceylon) among those that called at the port of K'vri-pat'iam, in the Ca country In the agam anthology, there are some poems attributed to Pta-tvar, a Temil poet from Ceylon But it is from the seventh century that we get any direct reference to Ceylon in the literature of the Tsmils. The evidence of the literary sources is confirmed by a few inscriptions as well. There are three pre-Christian Brhm inscriptions in Ceylon which attest to the presence of
Ts,w1R in Ceylon. One of these, from Axiurdhapuraz popularly

known as the Tamil Householders' Terrace inscription, records the building of a prsda (terrace), probably used as an assembly hail, by some Tami1s On one of the sides of the terrace are found inscribed the following names: Kubira, Tia, Kubira ujata, aga, Naata and Krava the navika (ship's captain). The last-

3. S.Paranayitana, 'Tsmi1 Householders' Terrace - Anurdhapura',

A.B.I.A., XIII, pp. 13-1k. 1. Paplai, 1. 191. 2 _________ ; Kuuntokai, v 3L-3 ; Nariai, . fo

39
mentioned person seems to have occupied. the highest position among them, judging fro; the height of the terrace he occupied. As Paranavitana has pointed out, it is interesting to note that the person who occupied the highest seat was a ship's captain. This may mean that the community of Tmfls who insed this terrace was a mercantile community, possibly organized into a guild. The two other Brhmi inscriptions, from Periya-pi4iyan.k4am in the Vavuniy district, mention a Tm11 trader Lamed Vikha, who owned a cave in that place In South India, at Tirupparakuam and Sittanaval, there are at least three Brhmi inscriptions of about the second century B.C. mentioning householders from Ceylon (a) The establishment of religious contacts with the Andiira country as early as the second century A.D. is attested to by inscriptions at Ngrjuikoa, which refer to the foundation of a monastery called the Sh4a-vihra by monks from Ceylon Probably Telugus from the ndhra country were in Ceylon, too t at this time. The Mahvaipsa mentions a 'Dam.13a' named Vauka, 'a city-carpenter in the capital', among the paramours of Queen Anul to be raised to the throne in the 1. V, pt.2, p. 2k2.

pp. 367, 368, 3755. . 3. J.Ph.Vogel, 'Ngrjuikoa Inscriptions', E.I., XX, pp. 22, 23.

2. C.Narayana Rao,

40
first century B.0 Although the Mahvai1sa and the Dipavaipsa refer to him as a Daini4a, his name suggests that he was a Telugu, for Vauka is a term that was applied to the Telugus by the Tamila. Probably Vauka was an artisan from ndhra-dega. The archaeological sources provide valuable data regarding the cultural relations between ndhradea and Ceylon in the early centuries of the Christian era. It baa been pointed out by Paranavitana that the majority of the early sculptures of Ceylon bear a striking similarity to those of the ndiira school A number of portable marble reliefs and statues, which by their material and style belong to the ndhra school, have been discovered in the northern parts of Ceylon, in places like Z4aha-illuppallama, Sjgiriya, Hingurakgama, Naradnki dawe la, Pemadu and Kuccaveli Commenting on these finds, Paranavitana says:' The evidence of the influence of Indhra art on that of early Ceylon is so overwhelming that it may be suggested that a branch of that school was established in Ceylon and that the sculpture on the frontispieces of the ancient ______ are the work of sculptors from the Kiatna valley or local artists trained by them. k

1. Mv., 3k:20.

2. S.Para.navitana,'Examplea of ndhra art recently found in Ceylon', A.B.I.A., XI, pp. 15-18. 3. A.S.C.A.R. for 1952, p.2k ; A.S.C.A.R for 195k, P.5: A.S.C.A.R. for 1956, p. 11; A.S.C.A. for 1957, p . 2k; A.S.C.A.R. for 1955, pp. 10,11,29 i. 'Evidence of earliest ihhalese art', c'lon Observer, k.2.195o,p.6,

41
It is clear from the evidence that has been briefly adduced above that before the third century A.D. close contacts had been established between Ceylon and the Ta.xnil and Telugu countries. But this evidence does not necessarily suggest that there were settlements of Dravidians in the island at this time. The question to which we have to seek an answer is whether these early contacts between South India and Ceylon led to the rise of permanent and widespread settlements of the Dravidiama in the idland. The evidence outlined above reveals that commercial interests, political adventure and the prospect of military employment had led Tamils and possibly some Telugus to go to Ceylon

in

the early centuries of the island's history. Tamil

traders possibly established temporary settlements in the ports and main towns. But there is no reliable evidence in our literary or epigraphic sources to c&nclude that there were notable settlements of Dravidians in the island before the third century A.D. The Maivaisa and the late chronicle Rjvaliya contain some references to the migration of people from the

Tmi1 country

to Ceylon before the third century A.D. In the account of Vijaya, the )ahvasa refers to the arrival of a princess, seven hundred maidens and 'craftsmen and a thousand families' from the Pya country This statement does not inspire any confidence in us.

1.

;55 ff.

42
It is as unreliable as the many other elements that have grown, in the course of the centuries, around the traditiofl of the original Indo-Aryan settlements in Ceylon. It is significant to note that the earlier chronicle, Djpavaisa, has no semblance
of this

tale in its account of Vijaya. It seems to have been

included later in order to enhance the prestige of the founders of the Sinhalese kingdom. The Pjvaliya would have us believe that GajabThu I

(U.k-136) settled twelve thousand Tamil prisoners

in the districts

of Alutktruva, Srasiyapattuva, Yainuvara, Uunu'vara, Tumpan , v5ha, Pansiyapattuva, Egoatiha and Megoatiha in the central highlands The cycle of Gajabhu legends in the literature and tradition of the Sinhalese has been discussed by scholars in some detail and it is now agreed that, although there seems to be some kernel of truth in the accounts regarding GajabThu's visit to South India, many of the details are highly incredible and improbable That GajabThu visited South India is confirmed by the Cilappatikrani Around the tradition connected with this event, several legends seem to have grown in the course of the centuries. The account of the Tami]. settlements in the valiya

1.

., p. 35.
. cit.

2. LC.H.C., I, pt.]., pp. 182-185 ; W.N.K.Wijetunge, 3. Cilappatikram, pp.

18, 636.

43
may form part of the later details added to the original tradition. It is also possible that it is based on some minor Tamil settlements that were established in the island in the second centuryy or later. The second century A.D. appears to have been a period of expansion for the

Ca

country. The several accounts of Karikla

Ca's activities reveal that Tarnil settlements were established in the newly-cleared territories north of the Ca country, namely in Toaimaalam It is possible that the expanding population of the Ca country went in search of new lands and some of them settled in the western regions of Ceylon, where even now the few Tamil-speaking Sinhalese claim descent from those who are supposed to have been settled by Gajabhu. These events, or more probably later Tamil settlements, may have
given rise to the legend of the twelve thousand prisoners in

later times. With the evidence that we have now, it is not possible to verify the account in the RLTvali y a. As it stands, however, it is difficult to accept it as reliable. Although the literary and epigraphic sources are not helpful in our inquiry regarding the Dravidian settlements of the earliest period, the evidence of archaeology has been of much value. The earliest and perhaps the most defiaite evidence

1. V.Kankabhai,

. cit., pp. 27-29 ; Pafapplai, 11. 280-28k.

44
concerning any Dravidian settlement in the island prior to the third century A.D.is provided by the megalithic urn burials from Pontparippu, on the north-western coast of Ceylon. Partial excavations at this site at different times during the last four decades have uncovered several urn-burials, which have rightly been related to the megalithic culture-complex of southern India The niegaliths of the peninsular Indian region have generally been associated with the Dravidian-speakers, who are believed to have occupied the area in the course of the first millenium B.C. This theory is held by most modern scholars, though there are several points of controversy which have not been satisfactorily solved Although the urn-burials at Pomparippu have been associated with the South Indian complex, they have not yet been systematically excavated, and it will be difficult to express anything conclusive till such ah excavation is completed and the finds thoroughly eximined. It was in 1925 that one of the pots from the burial site was exnii,ed for the first time by the Archaeological Department But it was not until 1956 that a systematic, though

1. C.J.Sc. (G), I, pt. 2, pp. 51-52 ; A.S.C.A.R. for 1 957, pp.

11-17,

30-31.
2. K.R.Srinivas..and N.R.Banerjee, 'Survey of South Indian Nega].itbs', Ancient India, 9, pp. 113-114. 3. C.J.Sc. (G), I, pt. 2, p. 51.

45
by no means extensive, excavation was carried out there. In 1956 more than a dozen jars were discovered and in and around these were smRller pots which contained skulls and other hnm,4n bones, some of which were post-cremation remaine In the next year, nearly fourteen urns were unearthed and these, too, contained human bones, skulls, food and personal belongings These burials were either fractional or secondary. Of the metal artefacts, four are of bronze and one of iron. Some of these artefatta are ainrilar to those discovered at the megalithic sites at Brahmagiri, in the Kannada areas of South India Deraniyagala, who was in charge of the 1956 excavations, has compared these with the finds of the fourth quarter phase of the Bronze Age in the Deccan, datable to about the third century B.C1 An examination of the material from Pomparippu shows that it is not to the material from Brahmagiri and Chandravalli that the Ceylonese artefacts bear the closest affinity, but to those from the sites in the Tamil country, such as 4dichchanalltlr. The Pomparippu site differs in one important respect from those of Mysore and Ker4a, in that its interments belong to a class called urn-burials and have no litbic appendage either in the

1. A.S.C.A.R. for 1 956 , p.i 2. A.S.C.A.R. for 1957, pp. 11-17, 3. Ibid., pp. 16-17.

30-31.

k. Ibid., p. 17.

46
form of a bounding circle or dolmens and cists. Even the absence of sarcophagi is conspicuous. Such burials have been found. in large numbers at icIichchanal].r, in the Tinnevelly district and are peculiar to the extreme south of the peninsula The Pomparippu site lies closer to dichchanallr in respect of the large contents of bronze ware, than to the sites of Mysore. But it has all the common features that makes it representative of the megalithic culture, namely iron implements, the wheel-turned Black-and-Red ware and the post-excarnation fragmentary and collective burials The large and. pyriform urns are similar to those from Adichchanallr and Brahmagiri. Probably the people responsible for these burials were Tamils from the neighbouring Tinnevelly district, the area which is closest to Poniparippu. The common prevalence of such urn-burials among the Tamils of early times is evidenced by the aam literature as well In the light of this evidence, the Poniparippu region could be taken as one of the earliest settlement sites of the Dravidians, probably Tamils, in Ceylon. The problem lies not so much in the identification of the authors of these burials as in the determination of their date. The South Indian sites have been dated variously from

1. K,R,Sriniyasan and N.R.Banerjee,

p. 110 ;

A.Raa, Catalogue of Prehistoric Antiquities from Adichchanallur and Perumbair. 2. K.R.Srjnjy asan and N.R.Banerjee, p. 115. 3. K.R.Srinivaaan,'The Megalithic Burials and Urn-fields of South India in the light of Trnnil literature and tradition', icient Thdia t.L,.. bp . q .

47
the seventh century B.C. to the first and second centuries AD. From the evidence of the agam literature we find that such urn-burials were in vogue in the Tamil country as late as the second and. third centuries A.D By a closer comparison of our artefacts with their opposite numbers in the South Indian sites as well as on the basis of stratigraphy it is possible to arrive at a specific date for the Pomparippu burials. But unfortunately, the excavations at Pomparippu have not been systematically completed nor has a comparative study been undtaken. Till these things are done it is impossible for a non-archaeologist to pronounce a judgment on this vital question. The Adichchanallr and Perumbair sites in the Tinnevelly district can be dated to about the third century E.0 Considering the fact that our artefacts bear the closest similarity to those of the latter sites, it may not be wrong to assign them to about the same period. Al]. that could be said for the present is that the Poinparippu site is earlier than the third century L.D. and. is one of the earliest settlement sites of the Dravidiana in Ceylon. T p king into consideration the location of the site, near the mouth of the Kafl Oya, close to the pearl bnkR and only a few miles south of the ancient,

1. K.R.Sriuivasan,

. cit., pp. 9 U. . ., p. 113.

2. K.R.Srinivasan and N.R.Banerjee,

48
though lesser known, ports of Kutirimlai and Pa].lugatuai, where ancient ruins are still to be seen, it is possible that this originated as a settlement of traders as well as pearl-divers and fishermen from the opposite coast. It is diZficult to say whether these Dravidians continued to survive as a distinct group till later times when Pomparippu definitely becomes k.nown to us as a Tstmil area, or whether they were assimilated to the local Sinhalese population before long. The proximity to as well as the continuous relations with South India may have helped them to maintain their ethnic identity for a long time. But these are matters of speculation. Another possible megalithic site is to be found in Katiraveji, on the north-eastern coast of the island. Some years back, Paranavitana discovered here several rude slabs of stone, cut to some size aDid shape, scattered. about the place, but not without some order. These stones 'lie in groups of four or five; and there are nnm1takable signs tbat some of them may have been set up on the ground. There is one group which still shows the original structure' Paranavitana also found 'other relics of hunmn occupation' On the basis of the description of such a structure in the Paramatta-jotik, be

1. C.J.Sc.(, II, pp. 91_95. 2. Ibid., p. 95.

49
surmised that these could be 'connected with a akha If these structures served the purpose of warship, as Paranavitana is inclined to believe, it is unlikely that several of them were erected in one particular site. It seems more probable that these were sepulchral structures, similar to those found in several parts of South India. Among the many different types of megaliths found in that peninsula, dolmenoid cista form one claas These are either made of dressed slabs of stone and coed by a capstone or are constructed with rough unhewn boulders. Such cists are found in places like Tiruvlag4u in Andhra Pradesh and Ariyr in Nadras But almost all these have portholes, whereas the dolmenoid cists in Cochin do not have this features It is possible that the cists at Katirav4i belong to the latter class. In fact, Paranavitana states that according to his guide there was at least one structure which had four side-slabs and another slab at the top, only a few years before he visited the site. Two of the side-slabs bad fallen down and the top slab had been removed for building a temple in the

1. C.J..Sc. (G), II, p. 95. 2. K.R.Srinivasan and N.R.Banerjee, p. 105. 3. Ibid., p. 106. if. Ibid., p. 106.

50
vicinity This means that all the other groups of stones at this site may have originally stood in the form of dolmenoid cists. Further, the name given to this site by the villagers, who are Tami].s, is Kura u-paai-eutta-vnrpu ( (The region of) the margosa tree under which the monkeys mustered). This name seems to connect these structures with the South Indian do].menoid date. For, in the Tamil country,the megalithic structures are known by a remarkably similar name, Kuraiku-paaai, a corruption of the name Kurakkuppaai, meaning 'a sepulchre or tomb lowered into the earth' The villagers of Katirav4i, like those of South India, believe that these stone structures mark the site where the monkeys of Rma's army encamped before the battle with Rvaa This is a case of popular etymlogy based on the element kuraiku (=monkey), the corruption of kurakku It is possible that the later Tamil settlers in the Katirav4i region, having seen the remarkable similarity between the megaliths of their South Indian homeland and these structures, used the name Kurakku-patai or Kuraku-pataai which later became Kuraz5.ku-paai. As no excavation was carried out at this site, it is not known whether burials exist here, and, therefore, it is not possible to say

1. C.J.Sc. (a), II, p. 95.

2. K.LSrinivaaan,

. cit., p. 9.

3. Ibid.; C.J.Sc. (G), 11, p. 95.

51
anything definite on this matter. Since the sire is on the coastal area not far from the ancient port of Gokya, it is not impossible that the people who erected these were traders from the Cochin area, the dolmenoid cists of which place bear the closest similarity to our cists. Perhaps the Katirav4i area bad a small settlement of Dravidians some time between the third century B.C. and the first century A.D., the period normally assigned to most of the South Indian megaliths. It is not impossible, however, that these structures are independent of the South Indian complex. But this is unlikely on account of their isolated character, which goes agint their association with some other culturecomplex. Until the ninth century, with the exception of the megalithic remains of Pomparippu and the possible exception of those of Katirav4i, there is no definite evidence regarding any Dravidian settlement in the island. The P].i chronicles, South Indian literary works, and Cey].onese and South Indian inscriptions attest to the continuous relations between Ceylon and South India. Between the third and the ninth century, there were two South Indian invasions of Ceylon. The first was in A.D.529 which resulted in the rule of six Tamils at Anurdhapura for twenty-six years The second took place in the reigm of Sena I

(833-853). On this occasion, the Pya ruler ri Na

Vallabha

1. Cv., 38:11 ff. ; W.M.LWijetunge,

. cit.

52
defeated the Sinhalese ruler and returned with a large booty In the same period, at least on nine occasions, Sinhalese aspirants to the throne went over to South India and took mercenaries to achieve their enda There were also close relig&us relations between the two regions. The Pli chronicles refer to Buddhist monks
from

South India going to Ceylon and vice versa. Scholars like

Buddhadatta and Mahynists like Sagharnitta went from the Ca country. Monks from CLAn went to the ndhra country and from there helped to spread Buddbism In the time of the aiva revival
in

the Tamil country (sixth to the ninth century), monks from

Ceylon are said to have gone there and participated in public disputations An interesting information regarding South Indian Buddhists in Ceylon is found in some late Telugu Jam works as well as in two Kannada inscriptions of about the twelfth century. The Telugu works, such as the RLj vali-kathe, Akalka-carita and the kalfika-stotra, refer to an eighth-century Jam teacher, Aka1ka by name, from ravaa Belgola in !rsore, as having disputed with the Buddhists of 1flci and defeated them These Buddhists,

1.
2. Mv.,

50:12 ft. ; W.M.K.Wijetunge,

. cit.

36: 115, ft.,

119;

Iv,, p. 1f9 ; Cv., 11.Lf:?l,1.0S, 125, 152; k5:18; . cit. ; U.C.H.C., I, pp.3O95J

117:33

k6 ft. ; W.M.K.Wijetunge,

3. J.Ph.Vogel, .2, cit., pp. 22, 23.

lf .-iru viai 5. H.R.Wilson,

a1-pnam, pp.. W .Taylov, Mackenzie Collection, I, p..Lv ;

1L
., II, pp.k5, k6,a

53
we are told, were in consequence banished to Ceylon. The substance of these accounts seems true, for, two Kannada inscriptions of earlier dates also refer to the same incident.
Au

inscription

Akal.ka's defeat of the Buddhists, while another from Sravaa Belgola celebrates Aka1fika or his victory at flci over the 2 Buddhists who were in consequence banished to the island of Ceylon'. Since more than one literary work and two Lt1Gr&pions record this traditionand since the details of the account are not intrinsically impossible, it may be allowed qualified credence. In the cultural sphere, too, there is evidence of close relations between South India and Ceylon in this period. sew'w.f The influence of Pallava art and architecture onhe buildings and sculptures of the island between the sixth and the ninth
century bears testimony to this. The land' Geig stands out as a unique monument of Pallava architecture in Ceylon The well-

from Tirumak4lu-NarasIpr t1k, of A.D. 1183, alludes to 1 S

known Man-and-Horse's-Head and the bas-relief from Isurumuniya as well as the dvrap1a statues at Tiriyy and the bodhisattva figtthes from Situlpavva and Kurukk4-maam exhibit unmistakable

1. E.C., III, Inscr. No.105 from 1iruniaklu-Naras!ptr t].nk '-tl.

2. E.C., II, Inscr. No.5k from ravaa Belgola, p. k. 3. A.S.C.A.R. for 1910/11, pp. 112-50.

54
influence of the Pallava school of scu1pture The use of the Grantha script of the Pallavas in the Sanskrit inscriptions at Tiriyy and the influence of this script on the Sinhalese script in the seventh and eighth centuries bear further testimony to the expansion of South Indian influence into Ceylon in the later Anurdhapura per iod In the context of all these relations between Ceylon and South India a certain amount of two-way traffic in population may be expected, too. According to some traditions in Kerala, there was a migration from Ceylon to that part of the subcontinent in early times and the descendants of these Sinhalese are said to be the caste of people still known as 1avar (Sin.halese or Ceylonese) Probably there were some settlers from Ceylon in

1. A.S.C.A.R. for 1936, pp. 16-19 ; Artibus Asiae, XIX, pp. 165ff.,
335 ff. ; Indian Arts and Letters, XI, p. 28 ; U.C.H.C.,I, pt.2,

p. 403. 2. A.S.C.A.R. for 1953, pp. 21, 26 ; P.E.E.Fernando, 'Palaeographical development of the Brhm! script in Ceylon', U.C.R., VII,pp.300-301. 3. V.Nagam Aiya, Travancore State Manual, II, pp. 398-402; T.K.Veluppillai, Travancore State Manual, II, (19 40), pp. 14-15; C.A.)4enon, Cochin State Manual, pp. 33, 203.

55
South India in early times. In Ceylon, South Indian traders probably established temporary settlements in the ports. The fact that the two earliest and most renowned iva temples of Ceylon are to be found in the ancient ports of Mahtittha and Gokaa may point to the establishment of South Indian settlements in these ports at an early date. The antiquity of these shrines can be traced to about the third century A.D. The aiva Tmi-1 works of later times, prominent among these being the Takia-kailcapuram and the T a-vaipava-rlai, trace their origin to

pre-Christian times Much of the material in these works, relating to the early period, falls outside the realm of historical probability and one has to turn to other sources for reliable information concerning this quwation. The iva temple at Nahtittha namely Tiru-kttvaram, appears to have been noti g e1 in the Dhvaisa. According to this work, there was a temple of god at the port of Mahtittha in the ninth year of Kitti Sin Megha (A.D.3lO) The existence of a iva temple at Goka4a in the time of }1ahaena

(27k-3 01) is vouched for by the Naiivasa, which mentions the

construction by Mahsena of a vihra at Goka after the destruction of a temple of god there The Laattha pp

inI, the commentary

1.

., Tirnmalai Carukkam ; Yvm., p.

6.

2. Dahavatsa, ed. and tr. B.C.Law, p. k2.

3.

., 37:kl.

56

on the 1fahvasa, states that this temple of god was a ivaliga teniple Probably it was the predecessor of the Kvaram temple, about which we bear from the seventh century onwards. It

i8

first

mentioned in the hymns of Taa-campantar, the aiva hymnodist of the seventh century A.D He has also sung a hymn on Tiru-ktivarazn, the diva temple of Mahtittba These iva temples of the island, situated at the major ports of the Anurdhapura period, were presumably built by South Indian aiva traders. Probably there were temporary settlements of South Indian mercantile communities at these places from the early centuries of the Christian era. But it is not tin the ninth century that we get any definite evidence of any Dravidian settlement in the island. Considering the number of occasions when South Indian mercenaries were enlisted, it appears bhat before the ninth century more South Indians went to Ceylon as hired soldiers than as traders. Most of the mercenaries went to the island in the seventh century, when Sinhalese aspirants to the throne enlisted them on no less than seven occasions. There is no positive evidence to suggest that these South Indians remained behind in the island

1. Laatthana1sin!, II, (P.T,S.), p. 685. 2. Tiru-fia-canipantar Tvra Tiruppatikak4, pp. 810-812. 3. Ibid., pp. 518-520.

57
and established permanent settlements. Probably they, or most of them, stayed behind permanently. The situation created by the increasing numbers of Ker4a and Taniil mercenaries in the seventh century and later is comparable with that caused by the Teutonic federates in Britain and on the Rhine and the Danube frontiers of the Roman empire in the fifth century A.D. The British parallel is striking in this respect. We find that a British king employed Saxon mercenaries from the mainland to repel the invasions of his ememies and granted land in the eastern parts of his kingdom for their settlement. Eventually the federates created trouble over payment, plundered the country and asserted their power Although the situation in Ceylon was not similar in magnitude, it is in a similar ninner that the South Indian mercenaries appear to have behaved on several occasions between the seventh and the tenth century. The Cflavasa refers to instances when the mercenaries showed no desire of returning to their homelands, resisted to being expelled by the Simhalese rulers, created trouble over payments, plundered the country and at times took over power at the capital. For instance, immediately after the death of Kassapa II (650-659), his nephew

].. R.G.Collingwood and JN.L.Myres, Roman Britain and the English Settlements, pp. 358-359. 2. Ibid., p. 359.

58
1na 'had the Damias expel1ed' But they resisted this and
1banded themselves with the resolve: With that resolve 'they
seiied we

will drive him out'

the town' and it was only by

making a mock treaty with them that ?.na was able to regain power This uneasy truce did not last long. Soon after this a Sinhalese aspirant to the throne, Hatthadha, returned to the island with a Tamil force and the Tamile who were already in the island 'arose and joined him on the way as he approached' 'Hattbadha who had won over the party of the Dain4as for himself, occupied the royal city' and ruled for some time On an earlier occasion, T'mi-1 mercenaries of Dhopatissa I (639-650) resorted to plunder and destruction. 'The canoes
in

the Mahpli Hall he left to the Damias; (and)they burned

In

down the royal palace together with the Relic Teniple'

the

reign of Sena V (972-982), the Tamil mercenaries were again in power. 'The Dam4as now plundered the whole country like devils and pillaging, seized the property of its inhabitants'?

1.

45:11.

2. Ibid., 45:12. 3. Ibid., 45:13-16. 4. Ibid., 45:19. 5. Ibid., 45:21. 6. Ibid., 44:134. 7. Ibid., 54:5-6.

59
Not long after this, in the tenth year of Mahinda V

(992),

'the Ker4as who got no pay planted themselves one with another

at the door of the royal palace, determined on force, bow in hand, armed with swords and (other) weapons, (with the cry) "So long as there is no pay he shall not eat" When the king fled to Iohaa, 'Ker4as, SIh4as and K$aa carried on the government as they pleased' Just as the Saxon mercenaries founded Teutonic settlements in places like Kent, it appears that the Dravidian mercenaries, namely the DamIas, Ker4as and Kaaae, founded small settlements in Rjaraha which formed the nuclei of later settlements. As we shall see in the sequel, the Sinhalese rulers seem to have granted lands for the settlementsof mercenaries. The inscriptions of the tenth century refer to Tamil allotments and lands, which, according to Paranavitana, seem 'to have been set apart for the maintenance of the Tamil soldiers in the king's service' But it appears that there were Tamil allotments, lands and villages which were not necessarily set apart for the maintenance of TamiJ. soldiers but were places where Tamila were living There is also some indirect evidence in the Clavasa which points

1.

, 55:5-6.

2. Ibid., 55:12.

3. L

III, p. 273.

k. Se

infra, p. 71.

to the existence of minor Tamil settlements in Rjaraha in the seventh and ninth centuries That Tamils were living scattered here and there is hinted at in a reference in the account of Eatthadha (68k). It is stated that when Ratthadha went to Ceylon with an army of mercenriea from South India and marched towards Anurdhapura, presumably from Mahtittha, 'all the Damias who dwelt here arose and joined him on the way as he approacbed' Evidently this is a reference to the Tamils who lived in the areas between the port and the capital. Another reference is found in the account of the Piya invasion during the reign of Sena I

(833-853). When

the Pya ruler ri

a r1 Vallabba invaded

many Dam4as who 2 . / dwelt scatteredi here and there, went over to his side'. Probably there were minor settlements of mercenary and other Dravidians
in some parts of Rjaraha from about the seventh century. A reference in the ClaTaWsa seems to imply that many of the Tamils

the island and encamped at Mahtlitagma, '

in the island in the eighth century were soldiers. While recounting the meritorious deeds of Nahinda II (777-797), the chronicle states that he gave horses to the Dami.as 'as they would not take cattle' This probably refers to the Tamils in the capital city,

1.

k5:19. 5O:lE. 48:lk5.

2. Ibid,,
3. Ibid.,

61
for, it is unlikely that Nahinda II distributed horses to the Taxnils living in all parts of the kingdom. That these Tamils refused cattle and acdepted horses may mean that they were not a settled peasantry but mercenaries who had more use for horses than for cattle. But this, however, is a flimsy evidence and the Tamils who received horses were probably a few mercenary leaders. It is in the thIih and tenth centuries that we again get any definite epigraphic and archaeological evidence, though meagre, pointing to Dravidia.n settlemOnts. For the first time in these centuries, Tamil inscriptions come to light and Sinhalese inscriptions refer to Tamil lands and villages. The earliest of the ruins of iva temples are also datable to the same period. Several Saiva ruins, aptly termed the Tamil Ruins, have been discovered in a section of the norbhern quarter of Anurdhapura These ruins consist of temples and residences for priests, with some lesser buildings scattered here and there. Some of these are ivali.ga temples while some others are dedicated to the mother goddess. Several stone lifigas, too, have been unearthed in this area. Al]. the shrines are of ome design, which is simple and reminiscent of the style of early Dravidian temples. These have a vestibule (antaria), a middleroom (ardhama4) and a sanctum (garbha-gha), and were all built of brick
1. S.C.A.R. for 1892, p. 5 ; A.S.C.A.R. for 1893, p. 5.

62
basentents These aiva ruins of Anurdhapura, according to ParaxLavitana, belong to the 'latest period of that city's history' The style of these temples, which is in marked contrast to the embellished granite temples of the Ca and later ?eriods, is undoubtedly pre-Ca and, therefore, belongs to about the ninth century, if not earlier. This date for these ruins, or at least for most of them, has also been cnfirmed by the Tamil inscriptions found among them, these being the earliest known Tamil records in the island. Two of them are dated in regna]. years of Ciiicaka-pti rya (Skt. Sri Sagbabodhi Mhrja) who has been identified as Aggabodhi III (629-639) by Krishna Sastri This identification rests on the consecration name, Sababodhi, and on the script of the inscriptions. He seems ts have been guided mainly by the consecration name or 'throne name', judging from his statement: 'The writing employed in the records is sufficiently archaic to be referred to the time of Aggabodhi III, who according to the Ceylonese chronicle Mahvasa, was surnamed Sin Sahabodhi' Apparently, Sastri was not aware of the fact 1. A.S.C.A.R.for 1893, p. 5. 2. U.C.H.C., I, pt. 1, p. 386. 3, S.I.I., IV, Nos, 1k03, lkOk. li. M.E.R. for 1913, p. 103. 5. Ibid.

63
that Sin S4ghabodhi and Sil.megha were borne alternately by Sinhalese kings as consecration names in much the same way as Rjakari and Parakari were used by the Ca rulers. The name rI Saghabodhi was used by several rulers from the time of Aggabodhi II and it is not easy to identify the ruler of our inscriptioriswith any one of them. But it is possible to date the inscriptions on other grounds. The occurrence of the terms kumrakaam and akkcu in these inscriptions is of some help in this respect. The term kuxnrakaiam, referring to a group or a corporation in the position of a board of managers or trustees of single shnines does not occur in axy of the

Tmi1

inscriptions of South India before the ninth century It appears

1. U.C.H.C., I, pt. 1, p. 365. 2. See _____ 3. K.Kanapathi Pjllaj, A Study of the Language of the Tamil Inscriptions of the Seventh and Eighth Centuries A.D., thesis submitted to the University of London, 1936. Even the two early as of South India, namely the u1ka4am and the amtagaa find mention in the inscriptions only from te time of Nandivarman III (8kk-866) and Aparjita (879-897) respectively, C.Minaksbi, Administration and Social Life under the Pallavas, pp. 130, 132.

64
to have been an institution of the early Ca period. The term (Ceylon money), referring to a particular type of cin of Ceylon, occurs for the first time in the inscriptions of Parntaka I (907-955) and not earlier It is unlikely that this term had come into use in the time of Aggabodhi III, nearly three centuries earlier. The king mentioned in our inscriptions is also given the title of rya (Mahrja). This is an attribute given to Mahinda IV,(956-97a), along with the name Sri Sagbabodhi,
in

the Vessagiri inscription But since the use of the title

}rya may have been indiscriminate, it is not possible to date these inscriptions to the reign of Nahinda IV on this evidence alone. Besides, this date may be somewhat late considering the script of the records. Probably these belong to the ninth century. The aiva ruins amidst which these epigrapha were found may also be dated to the same time. These Tamil inscriptions from Anurdhapura clearly attest to the existence of corporate organizations among the Tamila of Anurdhapura around the ninth century. One of the records, dated in the fifth year of Cii-cafika-pti }Trya,, registers

1. A.Velup:pillai, A

Study of the Language of the Tamil Inscriptions

of the Ninth arid Tenth Centuries, thesis submitted to the University of Oxford, 2. I, p. 3.

196k.

65
the grant of money, amounting to

thirty akkcus, for the daily

offerings and the burning of the perpetual lamp, evidently to one of the Siva temples in the area, by the members of a kumrakaam (kunrakaattu prrm), from the money loaned by Cki1 Cei Caka The other inscription, dated in the seventh year of the same king, records the gift of the same amount of money, for the identical purpose, by the same group, from the money given by Ckki Ceai The phrase kumrakaattu prrm was misunderstood

by Krishna Sastri when he rendered it as 'residents of KumrakaatttaPr1r' Kurakaattu-prir is certainly not the name of a village. Prrm literally means 'we the residents of the big village' and stands for the members of the village assemblies or corporations in the same way as (residents of the district) and

nakarattr (residents of the town) means members of the district assemblies and mercantile guilds of the towns respectively Kumrakaam is a term which occurs in contemporary South Indian inscriptions and stands for a group or corporation holding trusteeship of single shrines refers to those villagers

1. S.1.I., IV, No. 1k03 2. Ibid., No.lkOk. 3. M.E.R. for 1913, p . 103.

k C.Minakshi, .2g. cit., p. 122. 5. K.A.Ni].knta Sastri, The cZ, p.

66
or citizens who were members of the kunrakaam. Since these groups were not mercantile guilds, it cannot be said that they bad extra-territorial interests. This would mean that the Iun1rakaam of our inscriptions was a local body without any kind of relationship with a South Indian body. The important fact to be noticed Ia that the Tamils settled in Anurdhapura in this time bad organizations and institutions similar to those of their kinsmen on the mainland and used Tamil, presumably for the first time in Ceylon, in their donative records. A third Tami]. inscription from the same ruins throws further interesting light on the Tamils who lived in that area This long but badly weathered epigraph records the building of a Buddhist vihra by the Nku Nu Tamia ('The Tamils of the Four Countries'). It is dated in the reign of Seavarma. Since palaeographically the epigraph may be said to belong to the ninth century, this Seavarma could be either Sena I or II (833-853 and 853-887). The 11ku Iu Tamiar of this inscription also refer to themselves as Nku (We of the Four Countries).

It appears that they were a single body rather than a group of Tamils from four different countries. The evidence of some of the Kannada inscriptions shows that it is so. These inscriptions are those left by the mercantileAcalled the AiMMTuvar and their

1. S.I.I., IV, No. 1k05.

67
associates and belong to about the twelfth and thirteenth centuries In these inscriptions, we find references to a community called the lku }u (Four Countries), who were among the associates of the Aififlh1uvar. They were probably a trading community like the lu Nakarattr (Those of the Four Cities) bit there is no evidence on this point. The Nku ku Nu Tamiar o Tm11- of the

Nu,

mentioned in our inscription from Anurdhapura, seeni

to have been members of the same community as the lku

I4u of

the Kannada inscriptions. It is interesting to note that this community of Tamila erected a Buddhist temple at Anurdhapura some time in the ninth century and named it )kktai-pai. }kktai is an epithet that refers to the Cra or Keraja king The fact that the Buddhist
WLS

or vihra built by the Tamils

of the Nku I tu suests that they may have hailed from Kerala. , On the basis of the mounds of tile fragments and potsherds met with all over the area of the Tamil Ruins, H.C.P.Bei]. has surmised that the TRmII community relegated to this quarter would appear to be the caste of the potters Tt would, however, seem rather difficult to ascertain the profession of the community

1.

., VIII, p . 89 of the text ; see infra p.-o

2. M.LR. for 1916, No.130 of 1916.

3. T.A.S., V,

p. 100 fa.

4. A.S.C.A.R. for

1893, p. 5.

68
that lived in this quarter on the basie of these mounds. It is somewhat far-fetched to suppose that the Tamils at Anurdhapura were assigned different quarters of the city on the basis of their castes. Further, the evidence of the above inscriptions, revealing the presence of the ku Iu who were possibly traders,

goes against this conclusion. This area, where the ruins are all of a religious nature, appears to have had aiva as well as Buddhist temples which were common places of worship for the Draidians who lived in and near the city. Although the Tami]. Ruins are concentrated in the area between the path from Jetavaxirnia to Vijayrina and the path to Pa.k4iya from Kuctam PokuQa scattered remains of aiva monuments have been discovered here and there even outside these limits, but almost all in the northern part of the city. For instance,

in the

area north of the Basavak4ant

tank some stone ligas were dizcovered Near the sluice of the same tank was discovered a stone-based Piaiyr teniple Ih the Citadel area, the figure of a small na.ndi and the argha of a _____ were unearthed A quarter of a mile north of the Thprma

1. LS.C.A.R. for

1892, p .

5.

2. A.S.C.A.R. for 1890, p. 2. 3. Ibid., p. 3

k.

A.S.C.A.R. for 1898,

p. 3.

69
a small Hindu temple similar to those in the Tamil Ruins was also excavated In Vihra No.1 at Paki4iya, there are three inscriptions in Tamil and Grantha scripts Some of these remains may belong to later times but generally several of them seem to belong to the period before the Ca occupation. The consensus of evidence from all these finds should lead us to conclude that there was a Dravidian settlement in the northern part of Anurdhapura. By the time of Kassapa IV (898-91k) we get in the Sirihalese inscriptions definite references to Tamil villages and lands. There are three significant terms which occur in this connection in these inscriptions. They are Dem4-kblla, Dem4at-vlademin and Dem4-am-bim, which have been translated as 'Tamil allotment' ,'Tamil lands' and 'Tamil villages and lands' respectively As pointed out earlier, Paranavitana has interpreted

1. A.S.C.A.R. for 1898, p. 5. 2. A.S.C.A.R. for 1892, p. k ; 5.1.1., IV, Nos. 1399, lkOO.
3. D.Lde Z.Wickramasinghe, 'Anurdhapura Slab Inscription of Naheridra IV', E.Z., I, p. 117 ; S.Paranavitana, 'Colonibo iseum Pillar Inscription of Kassapa IV', E.Z., III, pp. 272,

273;

'Polonxiaruya Council Chamber Inscription of Abhaya Sa1mevan',

IV, p. 36 ; ' Girital Pillar Inscription of Udaya III', III, p. 1k3.

70
the term Dem4-kblla to mean 'an allotment of land in a village, set apart for the Tamils'. In his o4nion, they seem 'to have been set apart for the maintenance of Tamil soldiers in the king's service and must have been administered by royal officers' On an examination of the different occurrences of this term in the published inscriptions, it appears that the abpve interpretation does not always yield a satisfactory meaning. It is difficult to arrive at the exact meaning of this term; it appears to be an allotment of land enjoying privileges different from those of lands classified as paniunu But certainly it is not always an allotment from the royal household. For instance, in the Polonnaruva Council Chamber inscription, a Tamil allotment occurs as the private property of an individual In this record the allotment was granted immunities as a pamunu on condition of paying annual'y oie pla of dried ginger to a hospital. There is no reference in this record, or for that matter in any of the records where the term Dem4-kbfla occurs, to any share of the revenue being allocated for the maintenance of the Tamil soldiers.

., III, p. 273. 2. A pamunu was 'an estate possessed in perpetuity by a family in hereditary succession, or by an institution like a monastery pr a hospital', U.C.H.C.., I, pt. 1, p.

375.

3.

., IV, p. 36.

71
It is, therefore, clear that a Dernej-kbUa did not always denote an allotment from the royal household nor was it necessarily set apart for the maintenance of Tamil soldiers. It could only mean an allotment in a village where Tamils lived, presumably separated from the others. Some other references in the inscriptions seem to lend support to this interpretation. In the Rjarnigva inscription of }lahinda IV (956-972) it is recorded that certain immunities were granted to the village of Kiigama The piralkkam, who appear to have been a class of officials, were granted certain privileges in Dem4-kiigam but not in Kiigarna. It is clear from the context that Dem4-kiigam was not far away from Kiigama. Deme.-kiigam (Tamil Kiigam) appears to have been a

Tmil sector

which was originally part of the village of

Kiigama. This probably is an example of a Deme-kb].la. The Colombo Museum Pillar inscription refers to an officier called Deme-adhikra, who was presumably in charge of matters concerning Tamils, or more probably Tamil mercenaries, for, as Paranavitana has remarked, it is. when the edicts are concerned with the Tamil allotments that this official takes a part in the promulgation of edicts It is unlikely that it is one of the titles that were

1.

II, I,56.

2. E.Z., III, pp. 272, 27k ; U.C.H.C., I, pt. 1, p. 372.

72
conferred on certain officiala of the kingdom. In the time of ParkraniabJiu I (1153-1186) we come across at least two officials who were known as Damiadhikrins. The ciava4sa refers to

Dmf-.

dhikrin Rakkha who was a comm8nder in ParkramabThu's army The Galapta-vihra rock inscription, which is sometimes held to be of the time of Parkramabhu II but appears to belong to the reign of Parramabhu I, mentions Dem4a-adhikra Kahaibalk4u J4indaln, held to be probably identical. with Nagaragiri or Nagaraga.11a Nahinda of the Clavasa, who was one of the military comm'nders o Parkramabhu I From the last two occurrences of this title or designation, Dem4a-adhikra or Damidhikrin seems to have been a term applied to a military officer. Probably be was in charge of the Tamil mercenary forces and was, therefore, known as Dem4a-adhikra (Tamil official or authority). Probably the Dem4a-adhikra mentioned in the tenth century Colombo Museum Pillar inscription was also an official commanding the Tamil mercenary forces. His participation in the promulgation of edicts concerning Tamil allotments was probably due to the reason that these Tamil allotments were places where Tamil mercenaries had settled down. The presence of Tamil settlers in some of the villages of Rjaraha in the ninth and tenth centuries is also evidenced by the term Dem4e-kuli which occurs in some of the Sinhalese

1. 2.

75:20, 69, 74.

2. LZ., IV, p. 208 ; tJ.C.H.C., I, pt. 2, p. 488.

73
inscriptions of that time. This term occurs always with the term B4e-ki4i. Wickrainasinghe translated the terms as 'Tamil coolies' and 'Sinhalese coolies' respectively but Paranavitana has rightly rendered them as two types of imposts levied from Tamils and Sinhalese respectively It seems clear from the context that these refer to some kind of tax and not to people. It is a very probable conjecture that the foreign settlers had to pay imposts different from those paid by the Sinhalese. It is not always that these two terms occur when a reference is made to imposta. For example, in the case of the immunities granted in respect of the village of Kiigaina, mentioned earlier, the term k$i (impost) occurs without the epithet Dem4e (Tamil) or

H4e (Sinhalese) We have seen earlier that there was another

village called Dem4-kiigam, which was probably a Tarnil allotment in Kiigama Since Dem4-ld4igazn was treated as a separate village, there was apparently no need to qualify the term ku with Dem4e and Here in respect of the immunities granted to KiQigama. This may suggest that the two distinct types of k4! were mentioned in the immunity grants only in regard to villages where both Tanhils and Sinhalese were living. On the basis of this

1.

. ,I, pp. 170, 175.

2.LZ., IV, p. 511,fn.

6.

3. !'' II, p. 6. k. See supr,p.'I.

.74
body of indirect evidence, it may not be wrong to conclude that in the ninth and tenth centuries there were Tamils living in separate allotments in some Sinhalese villages and that such an allotment was known as a Deme-kbtlla. There seems t-e- he to have been such allotments in royal as well as pmivate villages. Similarly, the term Demea-v!lademin (lands enjoyed by Tm11s) 1 also appears to refer to the lands that were owned by Tm11s This phrase occurs in the Giritale Pillar inscription of Udaya III (935-936). The Anurdhapura Slab inscription of Mahinda IV (956-972) lays down that '(the produce) of trees and shrubs which exist ...... in the Tamil villages and lands (Dem4 gam-bim) (situated) in the four directions shall be appropriated in accordance with former custom' Bere the phrase 'Tamil villages and lands' evidently refers to the villages and lands where Tamils had settled. As mentioned before, the Clava111sa also refers to Tanii].s living here and there in Rjaraffha The placename evidence relating to Dravidian settlements in this period is negligible. Besides Dem4-ki4igam, there is another place-name with the element Deme occurring in the Ayitigvva inscription

1. I am indebted to R.A.L.H.Gunawardena for explaining this term to me. 2. E.Z., III, pp. 139, lk3.

3. LZ . , I, p. 117.
k. See supra, p. co

75
of Kassapa IV (898-9lk) The Ku ur_,1'an-d m ina Pillar inscription of the same monarch refers to a place called Kera].gama which may have been a place where Ker4a settlers were found The foregoing evidence of the Sinhalese inscriptions and the Clavasa is far too scant; and vague that it is difficult to arrive at definite conclusions regarding Dravidian settlements outside Anurdhapura in the ninth and. tenth centuries. These evidences certainly point to the presence of some Tamil settlers in the villages not far from Anurdhapura. It is interesting to note that Tamil inscriptions of the eleventh century have been discovered within a few miles of the Dem4-kbllas and Deme-gam-bim mentioned in the Sinhalese inscriptions This fact may not be purely coincidental but may be a pointer in the same direction, namely that these allotments and lands had settlements of Tamils. These settlements were probably sm'11 and embryonic. To sum up the evidence so far discussed, we have in the first place references in the chronicles to the presence of Tamil traders, invaders and mercenaries in the island from abont the second century B.C. There is no evidence, however, to

].. E.Z., II, p, 38.

IL

S ) te t.. !j..

fiti..y.'

2. Ibid., pp. 22-23. The reading of this name is tentative for the inscription is damaged at this point. 3. See map at the end of the thesis.

76
suggest that there were Dravidian settlements either in the pre-Christian period or in the early centuries of the Christian era, On the contrary, the general impression given by the Pli chronicles is that the Tamils were foreign to Ceylon. Their usurpations and unpleasant intrusions are not always dealt with favourably. We have the evidence of three Brhm! cave inscriptions, datable to the second or first century B.C., for the presence of Tamils, presumably traders, in the island. But here, too, the impression given by the inscriptions is that these Tamils were foreigners. Although the inscriptions were set up by Taniils, whose names are mentioned in them, the language of these records is Proto-Sinhalese as in the case of all the other inscriptions of the island at this time. But more important than this is that the recorders have described themselves as Tamils, which would indicate that they considered themselves to be distinct from, if not alien to, the general population, just as much as the Sixihalese donors in the pre-Christian cave inscriptions of the Tami]. country made known the fact that they were Sinhalese householders kuunipikan = Skt. Sifih4a kuumbikrini) In later times, too, we get instances of Tamils, who made grants to temples outside the Tamil country, recording them in the language of the arbut

1.

Mu., 25:110; Cv., 38:35-37. . cit., pp. 367, 368, 375.

2. C.Narayana Rao,

77
niking mention
of the fact that they were Tamils There is,

therefore, no epigraphic evidence suggesting the existence of Tainil or other Dravidian settlements in Ceylon in the period before the ninth century. It is only the archaeological evidence that points to the existence of a Dravidian settlement at Pomparippu and possibly another at Katirav4i, between about the second century B.C. and the third century A.D.After this there is a long gap till we reach the seventh century, when we get some flimsy evidence that points to possible Tamil settlements in the island. According to the Pli chronicle, bands of Tamil mercenaries were taken to the island at least on seven occasions in the seventh century. It also contains vague references to Taniils living in some parts of Rja.taha. Certain prominent Tamils, in possession of villages and tnk, also find meution In the contemporary Tamil works of South India, there are references to iva temples at the ports of Gokaa and Mahtittha which were venerated by Taniils. However, it could not be claimed that there is any definite evidence relating to Tamil settlements in the seventh century. It is only in the ninth and tenth centuries that we get such evidence in the Sinhalese and Tainil inscriptions

in the

archaeological sources and to an extent

in the

Pli

1. E. g ,, LE.P. for 189k, No. 18k of 1893. 2. Cv., 146:l92k.

7r
chronicle. That by the tenth century permanent Dravidian settle ents had begun in the island is fairly clearly borne out by these SOurces. On the basis of this meagre evidence that is available, we have to conclude that there were no notable Dravidian settlements of a widespread nature before the ninth century. The settlement at Pomparippu and the possible settle ent at Katirav4i have to be treated as isolated earlier settlements. These are comparable to the earliest Saxon settlements in Britain, at places like Dorchester, where the Teutonic artefacts are so early that they are not sometimes considered to belong to the period of Saxon settlement at all The burials at Pomparippu apart, the evidence as a whole does not warrant the assumption of a date earlier than the ninth century for the beginning of permanent and distinct Dravidian settlements in Ceylon. Before that century, there was intercourse between South India and Ceylon in the commercial, political, cultural and religious spheres in the wake of which some Dravidians went over to the island and possibly settled down there. Probably there were some mercenary settlers, too, from about the seventh century. Many of them may have been assimilated to the Sinhalese population before long. Besides the absence of positive evidence, there are also other considerations which lead us to think that Dravidian settlements worthy of the name were not founded before the ninth

1. R.G.Co].lingwood and J.N.L.Myres,

. cit., p. 39k.

79
century. As we shall see later, the evidence of the literary and epigraphic surces indicates that the present-day Tamil areas were then settled by Sinhalese people. The evidence of place-names, too, supportithis conclusion. A number of Simhalese inscriptions of this period have been discovered in the Maar, Vavuniy, Trimoomalee and Batticaloa districts, where Dravidian settlements were found in the thirteenth century. Some of these inscriptions provide us with the earlier Sinhalese names of villages and tanks which now bear Taniil names. For instance, the Mar Kaccri Pillar inscription of about the ninth century mentions the villages of 'Pepodatua, Kumbalhala,and Tumpokon, situated in the Kuakadavuk division of the Northern Coast', presumably close to Kahtittha, where the record was found The Sinhalese name of Alaricca for the raperiyak4am tank occurs in an inscription of GajabThu I (1l11. _136), from the same place The Sinhalese name for Kurunta-k$am, in the Northern Province, a appears as Kurugama in an inscription of l4ahinda III (801-80k) from that vi1lage In this last name, the derivation of the Tm11 form from the Sinhalese is clearly evident. Besides,these considerations, it is also worth noting that the Tamils of South

1.

., III, p. 105.

$e1?' k3.

2. A.S.C.A.R. for 1905, p. 3. Ibid.

80
India did not consider Ceylon as a Ta.mil-speaking region till very late times. In their early demarcation of the 'good Tamilspeaking world' (Tamiu-nal-ulakam), the omission of Ceylon is conspicuous We may, therefore, conclude that evidence for extensive or permanent Dravidi-an settlements bearing the signs of a date earlier than the ninth century is definitely absent. Permanent settlements of the Dravidians probably began by about the ninth century. Before the eleventh century these were by no means extensive. There were Tamils and possibly Ker4as and Karas settled in the northern quarter of Anurdh pure. after the ninth century. Outside the capital city

Tmi1,

and probably other Dravidian, settlers were found scattered in some of the villages of Rjaraha. It is not possible to locate all these villages with the evidence at our disposal. The Dem4-k1b11a referred to in the Colombo Museum Pillar inscription is stated in that record to have been situated in 'Gaagami, a revenue(village) of Valvii in the Northern Province' Unfortunately, neither Gaagami nor Valvii admits of any identification.Since the provenance of the record is also unknown, not even a rough location is possible. Since the village was in the Northern Province of the kingdom, it is to be located somewhere north of Anurdhapura. The Deme-k lla mentioned in the Polonnaruva

1. See supra,

2.

., III, p. 276.

81
Council Chamber inscription is identifiable. This al].Qn1ent, according to the inscription, was in the village of Kogm, in the district adjoining Mahara, in the province of Ginvaiunn-danaviya Nicholas has identified this village with the present Kogan-vela, in }tal Eaet In the Giritale Pillar inscription of Udaya III some Tamil lands are stated to have been situated in the Panisk4iya district of the Eastern Province This Parisk$iya district, according to the identification of Nicholas, extended over the Giritale areas Dem4inheihaya of the Ayitigvva inscription has been identified by Nicholas with the present Ayit1gv!va, in the flurulu division of the Anurdhapura distnict It has not been possible to identify Dem4-ki4igam mentioned in the R jmigva inscription of Nahinda IV. According to the inscription, this place was in the Eastern Province Probably it was situated somewhere in the region east of Anurdhapura. There is a KiLigama to the southeast of Anurdhapura but it is rather difficult to identify this

1. E.Z., IV, p. 36. 2. C.W.Nicholas, . cit., p. 3k.

3. E.Z., III, p . 139. 1, C.W.Nicholas, p. 18k.

5. E.Z., II, p. 38 ;
6. E.Z., Ii i p. 56.

C..Nicholas, p. 168.

82
b4 with the Xi4igani of the above inscription, for, the mo&ern Kiigama seems to fail outside the limits of the ancient Eastern Province. The village named Kerellgama in the Kuurumahan-dmna Pillar inscription was in the district of Valapu, in the Western Province of the Anurdhapura 1cingdoni It baa not been possible to identify this village exactly. Since the provenance of the inscription is 14al1iinau, in the Vilpattu National Park, which lies in the area of the ancient Western Province, we have to seek the ancient Kerelgama soniewere in that region. The ripixuiiyva and RabUva inscriptions of the time of Sena II (853-887) refer to the impost, Dem4e-ku2, in connection with the villages of Posonavull and Gliduru-gomaala, which have been identified as !ripinniyva and RaMblva respectively The Vihrgama Pillar inscription of Kassapa IV also refers to Deme in connegtion

with another village, the name of which is not preserved Another inscription of Kassapa IV, mentioning Dem4e-kulI comes from Sgiriya The nature of these possible Tamil settlements and the strength of the Tamil population in the island caxu.ot be determined with the help of the meagre evidence available to us.

1. E.Z., II , pp. 22-23. 2. E.Z., 1 pp. 1 67, 175 ; C.W.Nicholas, p. 3. E.Z, IV, p . 52.

169.

k.

A.S.C.A.R. for 1911/12, p. 108.

83
It appears that the Tainil settlers were found scattered in different villages and probably there was no single area which was peopled entirely by Dravidians. Many of the settlers may have been mercenaries who were taken to the island froa time to time by
Sinha].eae

princes. All that could be said with some

amount of certainty is that the ninth and tenth centuries saw the beginnings of the Dravidian settlements which covered several parts of the northern half of the island in the eleventh, twelfth and thirteenth centuries.

84
CHAPTER II

SL'rTLENTS IN THE PERIOD OF CLL CCC TJPAT ION (c 993-1070)

The period of some two centuries that lies between the fall of Anurdhapura and the collapse of Polonnaruva has long been recognized as one of very close political, cultural and social intercourse between South India and Ceylon. The events of this period, it may not be wrong to claim, led to some of the far-reaching changes that took place in the history of the island in the thirtbenth century. These changes determined the course of the future history of the island in many ways. But the two main results were undoubtedly the drift of Sinhalese political power from Rjara to the south-west and the rise of Tm{1

power in the northernmost regions of the country. The events leading to these dramatic changes may be said to begin at the turn of the tenth century with the Ca occupation For the first time a large part of Ceylon became a province of a TRm41 empire, and this naturally drew the island into the arena of South Indian politics and opened the way to the influence of South India and the influx of the Dravidian people into Ceylon.

1. See W.LK.Wijetunge, The Rise and Decline of Ca Power in Ceylon, thesis submitted to the University of London, 1962.

85
The history of the Ca occupation of Ceylon has been critically examined by W.X.K.Wijetunge in a thesis submitted to the University of London in

1962. It

is, therefore, not our

intention to deal here with the Ca conquest which was begun by Rjarja I in c.992/993 and completed by Rjndra I in 1017. In this chapter, we shall confine ourselves to a discussion of the Dravidian settlements that were established in the period of the CN1a occupation. Although it is possible to argue that the transformation of northern and a part of eastern Ceylon into Tamil-speRking areas must have been well under way by the time of the foundation of the independent Tamil kingdom of Jaffna in the thirteenth century and that this process must have begun at least a century or two before the latter event, it is not so easy to trace the course of the Dravidian occupation of these areas. The settlements of the Dravidians in the eleventh and twelfth centuries cannot be told as a na*rative with the materials at our disposal. We can only attempt to seek an answer to some of the important questions concerning their migration and settlement. Was there any large-scale migration of Dravidians in the period of Ca rule 2 What was the nature and extent of some of the settlements indicated by the inscriptional and archaeological materials 9 It may not be possible to set out on our inquiry with the hope of arriving at the whole truth, but at least we may be able to

86
arrive at more than what has been known so far. The first problem that confronts us in examining the course of the Dravidian settlements in this period is the question of whether there was a migration of South Indians into the island in the wake of the C 1a conquest. Of the different

kinds of evidence that lie before us, that of the literary sources is not of much help to opr inquiry. The only literary works that contain any notable references to the Ca conquest are the PV.i ClavaU1sa and the Sinha.lese Pjvaliya and RLvaliya.
No notice of the occupation of the island is found in any of

the contenrporary Tamil works of South India, apart from the incidental allusions to the conquest in such works as the Kalikattu-parai The Tami]. chronicles of Ceylon, written in much later times, strangely enough do not preserve even the memory of the Ca conquest of the eleventh century. The names of such C 1a conquerors as Rjarja and Rljndra are not even mentioned in these sources. Such works have little claim on our confidence for the history of Tam11 settlements in the eleventh or earlier centuries. The account of the C!lavqsa is by far the most important literary source for the history of the period of CN1a

1. Xalikattu-parai, v. 6k.

87
rule. Four of its chapters have been devoted to the events of this period and these have been written not very long after the time of the foreign occupation But despite this distinct value, it is of little use in our inquiry. The author of this section of the Pli chronicle, while relating the untold dpmages wrought by the C]as and denouncing their wickedness, does not interest himself in the affairs of the Ca administration or in those of the Tamils and Sinhalese in the Ca domains. The subject of his history is the resistance organized by some Rohaa princes. Of these princes, VijayabThu, the final liberator of the country from the C 1a yoke, is chosen as the hero of this

section of the chronicle. The conquest of the island and the desecration of the monasteries by the invaders are dismissed in a dozen verses in the chapter entitled 'The Pillage of _____ After these, any reference to the Caa is made only in connection with the resistance that was carried on against them. Repeated references are made to the hordes of Taini]. invaders who were taken to the island to suppress rebeliions In short, it is an account of the miseries wrought by the Cas and of the bitter

1. Cv.,

55-58. 55:13-25. 58:25.

2. Ibid.,

3. Ibid., 55:25 ; 58:1k ;

88
struggle that went on between the patriotic Sinhalese rebels and the ruthless oeign invaders. The traditions concerning the areas uhder foreign rule may have been considered irrelevant to the purpose of the author. But it is more likely that the author was depending on records which were preserved in the south of the island and which, therefore, did not contain any information regarding the goings-on in the districts contrlled by the Cas. All that we can positively gather from the P12i chronicle is that Tamil armies were sent to Ceylon at frequent intervals and that they were stationed in different parts of Ceylon. Whether there were Dravidians, other than these soldiers, who went over to the island at this time is a question that cannot be answered with the the help of the Ctflavqisa. In the Clavaisa account of the final campaign of Vijayabhu against the Caa, some of their strongholds in D2kkhiadesa and

in the eastern part of the island are named.

It was after the subjugation of these places that the Sinhalese commanders sent word to VijayabThu to join them at Poloxuiaruva The strongholds in D2kk(iadesa are given as Muhunnaru (Ruvar2k1), Badalatthala (Batalagola), Vpinagara (VEziaru), Tilagulla (Talagall.1a), Nalilgalla (Ngalla or Nikavrai), NaagaUa (Mahamaagalla)

1. .2z. 58:k6.

89
and Buddhag.ma (I nikdea) Of thse in the east only Chagrna (Sikxnain) is mentioned by name It is not known from the Clavasa whether there were any Tamil settlements in these

Ca

strongholds. Evidently there were many Ca troops stationed

at these places and possibly some of them settled down there.

Although the evidence of the C1tTlavasa is rather flimsy for

such a speculation, there are other considerations which support it. The discovery of Tamil inscriptions of the twelfth century and the occurrence of place-names denoting Taniil settlements in or not far from most of the C]a strongholds mentioned above suggest that there may have been Tamil settlers in and around the Ca strongholds in the eleventh century The Cflavaisa claims that 'all the warlike, valiant Co.as who were to be found here and there, gathered together in Pulatthinagara' on the e f the final debacle, and that the army of Vijayabhu, when it triumpahntly entered the city, 'at once exterminated a].]. the Damias root and branch' The statement that all the Te mils who lived in Polonnaruva during the C]a rule were annihilated is obviously an exaggeration.

1.

58:k2-k5. .$.C.

2. Ibid., 3.See infra,

'I.

58:51, 56.

90
That Vijayablhu did not have any animosity against the Tamils but was only fighting the C

1as is

borne out by ample evidence.

The employment of the Vaikkras, some of whom. may have been 2 erstwhile mercenaries of the Cas, 1 his patronage of Saiva temples, and his political and matrimonial alliance witb the PIyas3 show that Vijayabhu did not harbour any grievances against the Tamils. The evidence of the Abagamuva inscription that he 'drove away the whole darkness of the Dam4a forces' appears to be closer to the truth than the Clavasa statements The Pfli
C.PiCtPVi

chronicle has, therefore, no valuable inforniation,<Tam.il settlers in the island during the Ca occupation. The and the fljvaliya, while mentioning

the Ca occupation, give no details regarding the period of foreign rule. They continue the chronicle of the Sinhalese kings withoub a break by filling in the period of Ca rule with the account of the Rohaa rulers. They repEat with greater brevity the story of the destruction wrought by the foreigners The Nikya-sagrahaya and. the Saddharma-ratnkaraya also contain

].. See infra, 2, See infra, p. r$

; U.C.H.C., I, pt. 2, pp. k33-kkk.

; U.C.E.C., I, pt. 2 , p. 563. k29.

3, U.C.H.C., I, pt. 2, p.

k. D.Lde Z.Wickramasinghe, 'Abagamuva Inscription of VijayabThu I',

II, p. 216.

5. Pv., p.

10k;

., p. k2.

91
notices of this period. The former refers to the presence of 'the great multitude of Tamils in the villages, market towns and all over the kingdom' ( gam niyafigam rjadini pur un Dem4a maha senaga) 1 but it goes a step further than the Clavasa by stating that all these Tamils were destroyed by Vijayablhu. The reference to the presence of Tamils in the market towns is notable, for we learn from the Tamil inscriptions of the period that there were is (money-lenders) and traders of Tamil origin in some

of the places, presumably market towns, outside Polonnaruva The surprising brevity with which the Sinhalese chronicles deal with the C]a rule may be partly due to the paucity of records relating to that period. The above isolated and vague statements in the literary sources provide no sure guidance to the nature and extent of the migration and settlements of South Indians in the island. These sources have omitted much that is wanting and, in the absehce of any valuable guidance from them, we are thrown back upon the evidence of other sources. The

Tmi1-

inscriptions and the aiva and Vaiava archaeelogical remains provide better and more reliable information, though it is by no means adequate for our purpose. For the first time an unusually

1,

p. 20.

2. See infra, p.1oc

92
large number of Tamil inscriptions, more than three dozen compared with only three for the period before the tenth century, were set up in different parts of northern Ceylon during the Ca rule. Their sudden appearance could be explained easily if they are official records. But the interesting fact is that not a single one Is official, although several of them appear to have been set up by Ca officials in their private capacity. The sudden appearance of so many Tamil epigraphs presupposes the presence of more Tamils in Rjaraha than before. Such an impression seems to be confirmed by the internal evidence of the inscriptions as well. Unfortunately these epigraphe, almost all of which register private grants to temples, do not, by their very nature, contribute very much to our inquiry. Some are extremely brief while some others are badly damaged. However, they indicate the probable areas of settlement and, in some cases, the nature of the settlement. They range from the time of Rjarja I (985_bill.) to that of Adhirjndra I

(1067/68-1070) and, therefore, cover the whole period of Ca

rule. Tearly a third of these inscriptions comes from Pobonnaruva, which was renamed JaanItha-mag4am by the Cas 1. It is not possible to give the exact figure of the inscriptions as the find spots of a number of these in the museums of CoLombo and Anurdhapura are not known, although, judging from the contents, some of them appear to have come from Pobonnaruwa.

93
Most of these are from the Siva Dvles Nos.II and V and from the VaadEge. Of these, at least two belong to the reign of RJndra I, but unfortunately only a portion of the historical introduction (praasti) has survived in these Two inscriptions of the time of AdhrJndra I, beginning with the familiar historical introduction of Tik4r malarutu, are in a better state of preservation. They are both inscribed on the walls of iva Dvile No.11. One of these, dated in the third year of the king (1070), is a long record registering the grant of a perpetual lamp and some money for its maintenance by ...... Cra Ti...ya alias Eta.....koa Cappa1lavaraiya, a Ve

4a

of Mank4appi in Vir.pectu-nu, in the kfam (district)

of Tak....... in

maiani The title Eta......koa (Victor

of Eta......) and the name Cappallavaraiya suggest that the donor was an official in the Ca administration who had distinguished himself in battle by taking (koa) some place. The grant was made to the temple of VIava-w1tvi-jharam, the present Siva Dvle No.11. The names of nearly twelve temple officials, including those of the officiating Brhmaa, are given in the inscription. These officials and their successors

1. S.I.I., IV, Nos. 1389 and 139k 2. Ibid., No. 1388.

94
as well as the pariyclrakar, the supervisors of the parnayvarar, the nr and the tvar4iyr are held reBpOflsible for the maintenance of this gift. The terms pariycrakar and yvarar

refer to the temple attendants and the company of aiva devotees respectively is a term used in the South Indian inscriptions

of this period to refer to the members of the district assembly (nu). Its occurrence here seems to reveal the organization of the local assemblies on the lines of the South Indian institutions. The term tvaraiyr also occurs in the contemporary South Indian inscriptions as well as in another Tamil inscription in Ceylon and refers to temple dancers, commonly known as dvadsis The institution of the temple dancers appears to have been introduced into the island by the the
Tami1

The evidence of

inscriptions from Polonnaruva, therefore, shows that

at least some of the aiva temples in Ceylon were organized in much the same way as the temples of South India during the period of Ca occupation. The other inscription of the time of MhirIjndra, from Polonnaruva, contains the whole of a praasti and registers the grant of a lamp,
ULt the

donor's

n p me

and the regnal year of

1. Both occur in contemporary Ca e igraphs of South India. 2. E.Z., IV, p.

195. klk.

3. Cf., U.C.H.C., I, pt. 2, p.

95
the king are missing All the other inscriptions from Polonnaruva are undated and have to be assigned to the Ca period purely on palaeographical grounds. Two of these, which are fragmentary, refer to the gods and Aakiya

Maav1ar (Viu) Four others inscribed on the pillars of iva Dv1e No.V contain the following names: (a) TiruppvaaUaiy of Zkair, (b) Tillaikkaracu Tiyka-cintznai vnta-vi, (c) Karpakam, daughter of Mukari-ntv and (d) Paflca-neti-va, Uaiy of Nallr These persons seem to have been responsible in some way for the building of the temple now known as iva Dvle No.V. Such titles as }vnta-v, v and TJtaiy, borne by some of these persons, occur in the South Indian inscriptions as the titles of C 1a officials This indicates that all the persons mentioned above except the woman Karpakam, were officials. It is not certain whether the village kar and Naillir, which were assigned to two of them, were in Ceylon or in South India. Since the officials were serving in Ceylon, these may have been Ceylonese villages. Na11r is a common village name

in the

Tamil country. There are at

least four places in Ceylon with that name. One is in the Jaffna

1. S.I.I., IV, No.1392.

2. Ibid., Nos. 1390 and 1391. 3. Ibid., No. 1393.

k. K.A.Nilakanta Sastri, p. k6k.

96
peninsula and three are in the Kurungala district Pafica-neti-. mwntioned in the above inscription, may have been the uaiyg of any one of these places, probably one of the villages in the Kurungala district, for, according to the C'flavqmsa, the Cas had several strongholds in that district Of the other C,a inscriptions from Polonnaruva, there are two gragmentary inscriptions which record the gift of a large number of cows. One is from Siva Dvle No.11. It records the gift o one hundred and fourteen cows to the temple of the 'Lord of Jaanitha-puram alias Pulainari' (Vava-mtviivaram) The other is found engraved on the flight of steps at the Vaadge. It registers the gift of a certain measure of ghee and thirty cows Unfortunately the names of the donors are not preserved in these two epigrapbs. An interesting aspect of the iva Dvfle inscription is the occurrence of the toponymn Pula.jnarj. This shows that the Tamilised form of the Sinhalese Polonnaru was used side by side with the new C3a name of Jaantha-puram or Jaantha-mag4am, as in the case of other like Nahtittha and Velgama where Tamils were living

1. There was also a place called Vikrama-p.iya Nal].r, see infra,p.2,, 2. See supra, p. gf. 3. A.S.C.A.R. for 14 SI, IV, No.

1909, P . 1395.

27.

5.

See infra, p.

97
A few other short epigraphe of little or no value also come from the city and the vicinity of Polonnaruva. One such inscription is engraved on a bell found in Siva DvIle No.VI and has the name of SrT 4piai Perum alias Ton ......, the donor. Palaeographically, it has been assigned to the Ca period Another, registering the grant of Adhikaraia Craa,, a 'Vaikkra of the )u-kai division, comes from Ga]. Oya, near Polonnaruva The title Adhikaraa may suggest that the donor was an administrative officer among the Vaikkras. An analysis of these inscriptions from Polonnaruva and its surroundings reveals that almost all those in which the donors' names are preserved are grants by persons who may have been C,a officials. This perhaps explains the occurrence of many Tam!]. inscriptions in this region. Since Polonnaruva was the capital of the island under the Cas, several officilas from the C]a country were presumably stationed there. The occurrence of several Tam!]. inscriptions here may not necessarily indicate the presence of many Tamil settlers. The absence of grants by traders is rather surprising, for one would normally expect them to figure prominently among the donors of grants to temples. There is no evidence in our inscriptions of the existence

1. A.S.C.A.R. for

1908, p. 15.

2. S.I.I,, IV, No. 1398.

98
of a 6trong civilian population of South Indian extraction in and around Polonnaruva. Nevertheless, the organization of some of the Ca temples at Polonnaruva on the lines of those of South India suggests that these temples catered for the interests of more than a handful of Ca officials and some troops. There may have been peaceful Tamil settlers, too, in the city during the Ca occupation. Outside Polonnaruva, Periyak4am in the Trincomalee district has yielded the largest nuber of C1a inscriptions. More than a dozen Tamil inscriptions of this period have been found at the site of the well-known Rjarija-perum-p4i or Velgzn-vehera at Periyak4am. The Rjarja-perum-p4.i is an interesting example, and perhaps the only one, of a Sinhalese Buddhist vihra being converted into a Tamil Buddhist after the C 1a conquest. The existence of a Buddhist vihra

at this site as early as the second century A.D. is known from an inscription of the time of Bhika Tissa in one of the caves near the present The old Sinhalese name of this

pfl

was Velgain-vehera, which is also given in the Tami]. inscriptions along with the Tamil na*e of Rjarja-perum-pai.

1. A.S.C.A.R. for 195k, p. 1k; the occurrence of a brick with BrhmT letters at this site seems to place the origina.]. foundation of the stipa here in the pre-Christian times (ibid., p.13:

99
The fragmentary nature of moat of the inscriptions from this site deprives them of much value. The little that is recoverable front them, however, seems to indicate the presence of Tamil settlers of Buddhist faith in the Periyak4am region. The conversion of the old vihra into a Tamil

pfli apparently

took place in the reign of Rjarja I (985-101k) or immediately after that, for the has been named after this monarch.

The Zinhalese origin of the vihra is clearly indicated by the use of the Sinhalese name along with the Tamil name in the inscriptions. The absence of Tamil inscriptions of a date prior to the eleventh century and the occurrence of Sinhalese inscriptions oZ the tenth century at this site strehgthen the argument that the conversion of the vihra into a took place early in

the period of Ca ru1e Almost all the Tamil epigrapha from this place belong to the Ca period. But there is a slab inscription buried in the foundation of one of the image-houses which contains the name of Jayab-tva,, inscribed in the T2mil 2 scrpt of about the twelfth century. This Jayaba-tevaa is presumably JayabThu I (1110-1111). This may mean that Tamil patronage of this continued even after the C2a period,

which is to be expected if there were Tamil Buddhists living

1. A.S.C.A.R. for 2. Unpublished.

1953, p. 9.

100
at Periyak4am. With this possible exception, no

Tinil inscription

of the period after the eleventh century is known to have been discovered at this site. But the shrine continued to be venerated by Sizthalese pilgrims down to modern times The absence of Tamil inscriptions after the eleventh century may be due to the possible conversion of the Tamils of this region to Saivism. The inscriptions are a].]. donative records and register the gift of cows , buffaloes and perpetual lamps. Most of the records are damaged and are only partly decipherable. At least three of them are dated in regna]. years of Rjndra I (1012-lO'44) One at least of the donors appears to have been a C]a official. This person, 4titta-pr-araiya of Pa1avaputu-kui, gifted thirty-five cows and a perpetual lamp The element pr-araiya (the great chief) and the gift of a large number of cows suggest that he was an important personality1 Another person who gifted forty heads of cattle may also have been an official Most of the other donors appear to have been humble peasants or traders whose grants were lamps or small

1. A.S.C.A.R. for 1953, p. 27 ; E.Z., II, p. 178 ; Nanrpota, p.6. 2. Unpublished - Nos.I 776 A, 776
B

and 775

of the epigraphical

list in the Archaeological Dept., Ceylon.

3. No. I 776 B

of

the above list.

k. Pr-araiya is a common element in the names of Ca officials. 5. Unpublished - No. I 775 A.

101
amounts of money The occurrence of these Taniil inscriptions at Periyak4ant clearly suggests the presence of Tanii]. settlers in that area in the Ca period. It is not possible to say whether there were any Tamil settlers in this region before the Ca period. Presumably the Tamil settlement of Periyakuain originated during the period of Ca rule. Four Ca inscriptions have been found on the north-western littoral. Three of them are from Nahtittha (ntai). One is a fragmentary record and only the praasti of 2jndra I is preserved in it The second inscription is the longest of those belonging to this period, running into more than ninetyfour short lines with the first and the last few lines missing It is a grant by one Ti Kumara, the headmen of

Citu-1ra-nall1!r in Vr -ntu, in the K atr iya4 i chmai-v4anu of Ca-maalam. The grant was made to the temple of Rjarvaram at ttam alias RjarIjapurain (Mahltittba), which was built by the donor himself. Certain provisions made by him for the seven-day celebration of the festival of Vickam (Vi]tha) as well as the grant of a plot of tax-free land. and the assignment

1. Unpublished - Nos. 775 B, 776 B, 357 etc. 2. S.I.I., IV, lklk A. 3. Ibid., No, 1k12.

102
of certain taxes for the coat of the daily offerings are recorded in this inscription. Though a headman of a village in South India, Ti Kumara appears to have held an important post in the island, as is suggested by the powers he had of assigning portions of the public revenue for the upkeep of a temple built by him. The epigraph provides some useful information about the revenue system and the temple rituals in the Ca period. The name of RjarIja looms large in the local nomenclature. Not only was the temple of }tam named after him (Rjarjvaram), but the town itself and a. main street (perun-teru) were named Rjarjapuram and Rjarja-perun-teru respectively. Besides this information, the inscription does not give any details about the Tamils settled in Mahtittha. Only one Tamil settler, Kua, Ema, who was a citizen (kui) of ZtVtam owning a mansion (ikai), a house (v!u) and a garden (tam), is referred to in the record. The third inscription from Mahtittha records the arrangement made for the buraing of a street lamp outside the Tiru-irmivaram temple at }ttam, by ..... Tva, the utaiy of Ciu-k4att'ttr and an official of the Peruntaam of Rjendra Ca It is not possible to say whether Ciu-k$attr was a place in Ceylon or South India. Probably it was in South India, for

1. S.I.I., IV, No. lklkB ; see infra, p. rj,.

103
a place of that name ia mentioned in some South Indian epigraphs It is stated in our inscription that the money for the purpose of burning a street lamp was deposited with the cakara-piy'r, the ri1ai-viyar and the aikkf-viyar, all of )ttam. Cakara-ptiyr is a term that occurs in the contemporary South Indian inscriptions as well
An

examination of these occurrences

shows that the cafkara- p tiy r were a group of people who had 'duties connected with the maintenance of lamps and in prticular the supply of oil' in a temple Two records imply, moreover, that they were a corporation of oil-niongers! It appears that sometimes families of cakara-piyir were settled in special quarters close to the temples in order to maintain the burning of the temple lamps. For instance, an inacriptio4 of the second year of Ku]Zttuiiga I (1071) from Tiruvlafigu, refers to the settlement of twenty-five families of cakara-piyr on land belonging to the TiruvIlagu temple The settlement was named 1. LE.R. for 1912, Nos. 160 and 236 of 1912. The South Indian village was in Poyyi-kLam, in Te-karai-nu in Camaalam. 2. M.E.R. for 1897, No.80 of )LE.L for 1921/22, No.
1897 ;

M.E.R. for

1898, No.78

of

1898;

5k7

of 1920; M.E.R. for 1925, No. 395 of pp.

k89,

1925 ; K.LNilakanta Lastri, The 3. K.A.Nilakanta Sastri, The Cas, p.

4. Ibid.

516.

516, tn.

12.

5. S.I.I., III, p. 136.

104
ndra-ca-pi and the cakra-ptyr were asked to supply the oil required for fifteen perpetual lamps and to be in charge of lighting them in the temple of I4ahdeva at TiruvIagu It seems clear, therefre, that the cafkara-piyr of )t?am, referred to in our inscription, were there to perform a similar function in respect of the temple of Tiru-irmvaram and were probably settled there by the temple trustees. The i-

viyar were a community of people who sold betel leaves, as their name implies. The term ilai-viiyar (leaf-sellers), a variant form of ilai-viyar, occurs commonly in South Indian

inscriptions, especially of the Vijayanagara period. The 1ikky-viiyar, as their name implies, were sellers of plantains (bananas). Probably these two communities were expected to supply the betel leaves and the plantains required for the daily offerings in the temple. It is not possible to say whether they, like the cazkara-piyr, were settled near temples for this purpose. Probably they set up their business on their own accord near temples. The fact that the money for the maintenance of the street lamp at Nahtittha was deposited with these communities shows that they were organized as guilds or corporations rather than as loose groups.

1. S.I.I., III, p. 136.

105
The above Ca inscriptions of Mahtittha, therefore, provide us with some information about that port in the time of the Ca occupation. We find that it was renamed by the Cas as Rjarjapurarn. There were at least two Saiva temples, one of which was built L tb period and named RIjarijvaram, after Rjarja Ca. There were at least a few Tamil trading communities who were associated with the temples. Probably there were also other Tamil settlers at Mahtittha during the Ca period. Tanii]. inscriptions of the Ca period have also been discovered in the Hurulu and Nuvaragam divisions of the Tamankauva district in the North-central Province, Most of these are too brief or badly weathered to be of any use to us. In the Hurulu division, the inscriptions are mainly concentrated in Padaviya. Some of them date back to the time of Rjarja I. There are more than rn, half a dozen of these. Most of these have been found among the ruins of Siva temples. One of them, dated in the twenty-seventh year of Rjarja I (loll), appears to be a record of a mercantile community for it contains the names of a number of cet'is (money-lenders or traders) Another, also dated in the reign of Rjarja I, registers a number of gifts to a temple which appears to have been named after Rjarja.

1. Unpublished - No. I 3k0 of the epigraphica]. list in the Archaeological Dept., Ceylon ; A.S.C.A. . for 1891, p. 64.

106
The gifts were made by several individuals who may have been members of some mercantile or other body since they have all recorded their gifts in one inecription Padaviya seems to have been a coznmercia]. centre in the eleventh and twelfth centuries, for, Tamil inscriptions of the twelfth century also attest to the presence of Southnlndian traders, especially the well-known
VIra-valafijiyar,

at this place The Tamil 8ettlement here in

the Ca period may have been mainly mercantile in character. A Tamil inscription assignable to this period, from MahI-kachcha-. koi, nearly fifteen miles west of Padaviya, also mentions a number of ceffis Probably it was also set up by a mercantile community from the Tamil country. In the Nuvaragam division, Anurdhapurai Saftgilikanadarva and Atkaa have yielded a few inscriptions of this period, which are, however, disappointingly short and of little value to us. Of the inscriptions at the Pak$iya-vihra at Anurdhapura, one records the gift of a certain Kcari Araci while the other two mention two persons, who probably had donated something to that establishments They were apparently Buddhists. A short inscription on a pillar within the precincts

1. Unpublished - No. I

34.
pp. 36, 50.

2. Unpublished - see infra, p. 3. Unpublished ; A.S.C.L.. for 1 9 05,

k. 3.1.1.,

IV, Nos. 1399, 1 14.00, 114.01.

107
of the RuvanvJ.isya gives the name Jaatppa-kaa Perum-p44 (the Great Temple of the Victor of Jagat3ppa), evidently referring to the same stpa The surprisingly few and unimportant Ca inscriptions from Anurdhapura shows that the former capital city did not remain an important centre under the There

are no Ca inscriptions indicating the presence of South Indian officials in that city. Even the aiva temples there apparently did not enjoy the patronage of the Ca ruling class in the island, But Anurdhapura seems to have attracted the attention of a few Tamil Buddhists who presumably lived there or weht there on pilgrimage. The inscription from Safxgili-kanadarva registers the grant of land and the deposit of some money on interest by the army chief Jayamui-n apparently to a temple,

the name of which is not preserved The title Spatik4 as well as the eI.ement in his name suggest that he was

a military official in charge of some Ca troops in the Nuvaragam region The inscription is dated in the reign of Rjndra I. The epigraph of Atkaa, which is dated in the twenty-eighth year of a ruler whose name is not given, records

1. S.I.I., IV, No. lkO2. 2. Ibid., No. lko8. 3. See supra, p.qc

108
the gift a 1_' by a certain Araka Irina of land in Kallaiyil-teliyal-peu, twenty beads of cattle and fifty coconuts to the Uttama-ca-ivaram teniple The identity of Kallaiyil-teliyal-peu is not known. Probably this place and the temple of Uttama-ca-ivarani were both in the region of Saxgili-kanadarva. Uttama Ca was not only the name of the immediate predecessor of Rjarja I on the C 1a throne, but was

also used as a title by members of the Ca royal family in the time Rjndra II (lO5k-lO63) It is not likely that the temple referred to in our inscription was named after King Uttama Ca, for Ceylon was not under the Cas in his time. It is possible that it was named after a member of the Ca royal family, with the title Uttama Ca, in the reign of Rjndra ii. Only one Ca inscription has been discovered in the North-western Province. This record comes from Attaragalla in the Puttalam district and is dated in the ninth year of Rjndra Ca, who may be the second of that name. It is badly damaged and seems to record the building of an ambalam (iim) The Central Province has also yielded one Ca inscription.

1. 3.1.1., IV, No. 11111. 2. Y.A.Nilakanta Sastri, The Cas, P. 261.

3. S.I.I.,

IV, No.

1k15.

109
This inscription found in Diyavinna is written in Tami]. and Grantha characters and 'states that a person called Virabhavaaa DahaIbbha ?1al1ai arrived at this place' This is rather interesting, for it is the only Tamil inscription of this period which has beeh discovered so far south, outside the limits of the area which was under the actual control of the Cas. The personality referred to here rna..y have been one of the Ca soldiers or army chiefs who were operating against the Rohaa princes in the central highlands. The names VIrabhavaa and Malla suggest that he was a warrior, but it is not possible to draw any conclusions on the basis of this stray epigraph, which may have been set up by an adventurous personality totally unconnected with the C 1a wars. The occurrence of this single inscription cannot be taken to indicate any Tamil settlement in that area in the C1a period. The contemporary Tamil inscriptions of South India, while referring to the frequent battles fought in the island, have little information about axy South Indian migration or about the Tami].s of Ceylon. They confirm the statement of the Clavaisa regarding the large armies sent by the Caa to suppress 2 uprisings in Ceylon. One of these inscriptions from Tirumukk1Ial,

1. Unpublished - No. 580 ; S.Paranavitana, 'Epigraphical Summary',

C.J.Sc. (G), II, p. 191.

2. UVC.H.CI,, I, PT. 2, p.

k25.

110
of the year 1067, mentions a personality named Kuriakulattaraiya as one of the commanders who fought on the side of the Sinhalese prince VijayabThu His name suggests that be was a Tamil. Tamil inscriptions of the Pya country, belonging to the thirteenth century, refer to a high official called Kurukulattaraiya It is not known whether the Kurukulattaraiyars belonged to a noble family. The first element of the name, kurukula (Skt. gurukula),
"S

reminds,(of the Kurukula caste, who are of South Indian origin, living in the western districts of Ceylon. The large majority of them speak Tami]. as their mother tongue. Kuruku1attaraiya (Chief of the Kurukula) is referred to in our inscription as a feudatory 'who wore a golden enk1et' It has been claimed that he was a chief of the Kurukula community in the is1and If it is true, it would mean that as early as the eleventh century the Kurukula community was in Ceylon and that its chieftain paid allegiance to the Sinhalese ruler. But the evidence of one name is far too flimsy to be the basis for such a conclusion. In this period, officials in outhnIndia often adopted names ending in

1. LV.Subrmiinya Ayyar, 'The Tirumukka]. Inscription of Virarjndra', E.I., XXI, p.

2k3.

2. M.E.R. for 1923, No. 5kk of 192 ; 5.1.1., VIII, p. 212 ; K.A.NilkRnta Sastri, The Pyan Kingdom, pp. 13k-.155.

3.

J. ,

XXI, p.

2k3.

k. M.D.Raghavan, The Karva of Ceylon, p. 5 ff.

111
-araiya, as, for instance, Pallavaraiya (Chief of the Pallavas) and Kalifdcattaraiya (Chief of the Kligas). Kurukulattaraiya, may also have been just a title of a mercenary leader in the army of VijayabThu. He may have been mistaken by the Cas for a feudatory chief.
An inscription of Rjarja I from Tanjore registers

the grant of land in five villages of Ceylon to the temple of Tanjore. These villages are said to have been located in }ppicuzxzpu Koiyram alias Rjarja-v4anu and in Kaakka Koiyram alias Vikkirama-c5a-v4anu Koiyrazn is a territorial
division

in the Trincomalee district, still known by that name,

which is a Tamilieed form of the SirLhalese Kotasara (Pii Kohaera). Although several Ca inscriptions have been discovered in the Trincomalee district, none comes from Koftiyram. Since the above grant was made by a Ca ruler and not by the citizens of the five villages, it is not possible to aay whether there were Tamils living in Koiyram in this period. But there are at least two other South Indian epigra hs which record the grants of some Ceylonese citizens to South In ian temples. One of them is in Kuttflam, in the Tinnevelly district and registers 'the grant of land by residents of Viandai alias Vikrama-pyanalltr, a village in KUr-nAu, which was a sub-division of

1. S.I.I., II, p.

k26.

112
a-ma4alam, to the temple of KuttLam' There are several interesting points to be noted in this record. Who were thesc. residents of Viandai 7 Evidently they were Saivas. Though generally at this time the aivas who were in the island were Tamils or Ker4as, there may have been some aivas

originating
.

from other parts of India or even from among the Sinhalese, It seems, however, unlikely that a group of Sinhalese or other non-Tamil Saivas from a particular village in Ceylon evinced an interest in the affairs of a temple in a South Indian

village, unless they were in some way connected with that villge. It seems more likely that they were Tamil settlers frok the Kuttlam area who still showed an interest in the affairs of their former village and temples. Such an interest can be seen even now among Saiva settlers from India and Ceylon in places like Malaya, who send gifts to the temples formerly frequented by them. Moreover, the name of the Ceylonese village was changed from Viandai to Vikrama-pya-nallr and. the district, too, was given the Tamil name of tr-nu. As pointed out earlier, in Tamil many Sinhalese villages were arbitrarily renamed/by the C]as and the occurrence of a Tam.il toponym does not necessarily point to

Tmi1 occupation

of the area designated by it. But such names

1. LE.R. for 1917/18, No. k5k of 1917.

113
1 were always derived from the names ofroya1ties. Viandai is the only place in the island known to have been named after a Pp4ya prince. Probably some settlers from the Pya country were responsible for this change of name. Kuttlam was in the Pp 4ya country and the residents of Viandai who made the above grant may have hailed from Kutt1ani or from some other place near this village. Perhaps Viandai was renamed after Vikrama Pya who took refuge in Ceylon after his defeat at the hands of the Cas It has not been possible to identify this place. The above account practically completes the total of our epigraphical knowledge as far as the period of Ca rule is concerned. As we have seen, almost all the inscriptions are donative records and Lea]. with matters that are of little help to our inquiry. They help us to trace the areas of Tamil settlement, but they are not alw&ys a sure guide in this respect. The records 1. Rjarja-puram for Nahtittha, Jaantha-niag4ain for Polonnaruva, jarja-vaanu for a division in Kotasara, Vii ama-ca-v4anu for another division in Koasara as well as Nikarili-c1a-v4an u, Rjndr a-c inka-v4antu and }Aummuli C malam are from the names and titles of Rjarja I and Rjndra I. 2. K.A.Nilajcanta Sastri, The Caa, pp. 250-251.

114
set up by officials do not necessarily indicate the presence of Taxnil settlers in their areas. But the place-name material and the evidence regarding the existence of aiva and Vaiava temples found in these inscriptions are often useful in finding an anawer to our questions. This evidence has to be compared with that of other aources before it is used to draw conclusions. Now we have to turn to the archaeological material that is available for this period. Unfortunately there is no positive material which could help us to trace the settlements of the Dravidians in Ceylon, similar, for instance, to the Saxon cemeteries of England which have greatly helped to map

u4 the earliest English settlements. The only archaeological

remains of considerable importance that have been left behind by the Dravidians are their religious monuments and sculptures. As the Hindu temples of this period were normally erected in areas where there were aiva and Vaiava Tamils, the remains of these structures could indicate in some way the regions of Tami]. settlement. It could be argued that there may have been aivas and Vaiavas among the Sinhalese and, therefore, the presence of aiva-Vaiava remains may not be a sure guide to the location of the Dravidian settlements. The presence of aivas and Vaiavas among the Sinhalese is only a theoretical possibility. There is no evidence to suggest that there were Sinhalese who
were aivas or Vaiavas in the island during this period.

115
Moreover, it is not difficult to identify the monuments of the Taniils in this period. The architectural style, the style of the sculptural finds and the occurrence of Tamil inscriptions are factors which help to reveal the identity of the temples built by Tamila. As the archaeological exploration and discoveries in the northern and eastern parts of the islan&, which are the traditional Tamil areas in Ceylon, are by no means complete, the material available to us is inevitably limited. To make
S

this position worse, the Saiva and Vaiava remains so far unearthed have not been properly dated. But many of these can, however, be dated on the basis of their architectural style and the inscriptions found among their ruins. The ruins of no
less than thirty aiva-Vaiava temples, belonging to the period

between the beginning of the eleventh and the end of the thirteenth century, have been discovered in the island and. all these are
in the northern and eastern parts. In the case of the well-preserved

temples, the architectural style serves as a guide to the dating. The occurrence of inscriptions in some other foundations helps to determine their age. A few are not sufficiently well preserved or they yield no inscriptions and. cannot, therefore, be dated with any certainty, although some of the remaining architectural members provide a rough guide to their age.

116
About ten Siva Dv1es, five Viu temples and one

KI

temple have come to light in Polonnaruva Of these, iva

Dvles Nos. II and V, and probably No. V1 belong to the period of C]a rule. iva Dvle No.11 is the only aiva temple of this period which has been completely preserved. It is considered to be an outstanding exampi. of the C 1a style of Dravidian

architecture All the three iva temples yield inscriptions of this period. iva Dvle No.1 is in the Pya style of architecture and belongd to either the twelfth or the thirteenth century' Almost all the other temples at Polonnaruva appear to belong to the thirteenth century Of the Ca temples at Polonnaruva, the name of iva Dvle No.11 alone is known from the inscriptions. It was called Vava-mtvi-Dvaram, after the chief queen of Rjarja I. A nu*ber of bronze images, representing aiva saints and deities, were discovered in these temples. Some of them have been acclaimed as masterpieces of Hindu sculpture

1. A. .C.A.L for 1902,

pp.7-8 ;A$'C14ror

1908, pp. 3-10 ;

A.S.C.A.R. for 1909, p. 17; A. .C.A.R. for 193k, pp. 2. A.S.C.A. . for 1911-12, p. Uk.

16-17.

3. S.Paranavitana, Art and Architecture of Ceylon - Polonnaruva Period, p.

31.

k.

Ibid.

5. See infra, L1.j 6. A.LCoomaraswamy, Bronzes from Ceylon, p.

9.

117
Some remains of Ca temples have been discovered outside Polonnaruva, too. At Norago1a, near Padaviya, were unearthed the remains of three temples Tamil inscriptions, some dating to the time of Rjarja I, have been found in these ruins. The occurrence of nandi, lii!iga and yoni figures shows that these structures belonged to aivism The names of these temples are not known, but one of them appears to have been named after Rjarja Besides these Structures, three other

aiva temples of the C3,a period are known from contemporary inscriptions, but their remains have not been unearthed so far. These are the Tirn-irnff g varam and Rjarja-ivaram temples of Mahtittha and the Uttama-ca-varam of Xtkaa Among the ruins at Nahtittha (!ntai), remains of some buildings of the eleventh century as weLLas a nandi, a lifga and a Gaa image were found Some of these may be the ruins of C1a temples. Saiva temples of the Polonnaruva period have also been discovered

1. A.S.C.A.R. for 1891, p. 10; A.S.C.A.R. for 1961/62, p. 67. 2. Ibid, 3. See supra, p.

oc
1'

k. See supra, 5. A.S.C.A.R.

Papep e,

for 1908, pp. 28, 30; K.Vaithianathan, Thirpketheesvaran

p. 13.

118
at Nalla-t ai-iakkam, Buddhannehe la, Maha-kanadarva, Ptak.0 and P4amai Some of them may date back to the period of Ca occupation. But there is no evidence with which we can date these to the Ca period with certainty. The remains of only one Ca Buddhist temple }Jave been found in the island, viz., at Periyak4am, in the Trincomalee district. As we have mentioned earlier, the temple was originally a Sinhalese Buddhist institution by the name of Velgam-vehera It was rebuilt in the eleventh century in the style of a Tamil
4i

and was renamed Rjarja-perum-p4ji. The architectural

style of the temple 'differs from that of the Anurdhapura Buddhist shrines and is akin to the Tamil Hindu shrines at Polonnaruva' The discovery of bronze and stone sculptures in South Indian style and of temples built in Ca style seems to indicate the presence of sculptors and stone-masons from South India. The style of & building is not always a sure guide to the racial or communal origins of the masons and architects

1. A.S.C.A.R. for 1907, p. 27 ; A.S.C.A.R. for 1891, pp. II, 30; A.S.C.A.R. for 1 9611 62 , P . 59 ; A.S.C.A.R.

for 1933, p. 18;

cJ.Sc. (G), II, pp. 156-157.

2. See supra, p.
c1g

3. S.Paranavitana in the Encyclopaedia of Buddhis - Volume of ecimen Articles, p.

83.

119
responsible for its erection. It is always possible for the artisans of one country to learn the architecture of another neighbouring country. It would be of interest, therefore, to find, out whether Dravidian craftsmen were employed to execute the Saiva and Vaiava monuments at Polonnaruva and elsewhere. The opinio4 of archaeologists is divided on thia matter. Godakumbure points out that the Sinhalese people were 'experts in architecture, sculpture and painting' and claims that 'the Cholas who brought the Sinhalese under subjection at the end of the tenth century employed these Sinhalese craftsmen to build temples for their gods, and make sculptures of thent' On the other hand, Paranavitana's opinion is that the 'Sinhalese sculptors and painters had no opportunity to practise their arts, for their patrons - royalty, nobility and the Buddhist Church - had ceased to exist under the Cho].a rule' If we turn to see the monuments in the C1a style in the island, we find that only the Aaiva-Vaiava temples at Polonnaruva and Noragoa and the Buddhist Rjarja-perum-p4i at Periyak4am, all of which enjoyed the patronage of Tamils as revealed by their inscriptions, fall under this category. No Sinhalese monument

1. C.E.Godakumbure, 'Bronzes from Polonnarnva, JR.A.S. (.B.), N.S. , VII, pt.2, p. 2k).
2. S.Paranavitana, Art and Architecture in Ceylon - Polonnarua

Period, p. 22.

120
of the tenth century was built in the Ca style. The Sinhalese raftsnien would have been unfamiliar with the Ca style of architecture and, therefore, would not have been in a position to execute at the very beginning of (Zla rule such an outstanding example of Ca architecture as the iva Dvle No.11. Only artisans skilled in the architecture of the South Indian temples could have accomplished this task. There is some evidence to show that there were Tamil masons in Ceylon in the Polonnaruva period who were employed by Sinhalese monarchs to build Buddhist monuments. Tami]. letters have been used as mason's marks in certain constructions dating from the time of ParkramabThu I at Polonnaruva and Padaviya One of the Tamil inscriptions from Budumuttva attests to the presence of a community of blacksmiths, identifiable as Tamils, in the KurunRgala area in the regin of Vijayabhu I These considerations lead us to think that the Ca authorities, who introduced from South India such communities as the cafikara-p tiy r for the maintenance of aiva temples would ha'e invited craftsmen from the mainland to build temples in the style of the South Indian examples. Further, from the

1. See infra, p . irl 2. , III, P. 305.

3. See supra, p.D3

121
practice of later times we see that Tamil or South Indian artisans were engaged for the building of Dravidian-style structures in the island. Inscriptions of the reign f Bhuvanekabhu IV

(l3kl-].351)

declare that the Lktilaka and Gaa]1dei shrines near Gampola, the two well-known Dravidian-style temples of the fourteenth century, were the creations of Sthapatiryara and Gaevarcrya respectively The names of these architects clearly suggest that they were of South Indian origin. It is, therefore, probable that South Indian craftsnien were invited to Ceylon to build the Zaiva-Vaiava temples of the Ca period, although it is not impossible that local cra!tsmen were also engaged to do this work. There is also a controversy regarding the sculptors who were responsible for casting the many Saiva and Vaiaa bronzes that have been discovered in the C 1a temples at Polonnaruva.

Godakunibure claims that these bronzes have certain 'distinctive features' which mark them as products of Sinhalese sculptors, but fails to explain what these distinctive features are Bell's opinion is that they were 'doubtless cast in India' Paranavitana

1. U.C.H.C., I, pt. 2, p. 782 ; U.C.R., XVIII, Nos.l&2, p . 11. 2. C.E.Godakumbure, 3. A.S.C.A.R. for

cit., pp. 2k3 , p.

ft.

122
supports him by stating that 'these bronzes have certainly been imported from South India and belong to the history of art of that region. ' 10n the other hand, Basham feels that an 'important school of bronze-casting existed in Ceylon, and produced works similar in style to those of South India' It is difficult to determine who cast these bronzes purely on the basis of their style. In the first place, there is so much in common between the plastic arts of Ceylon and Soi.thh India in this period that slight variations in form do not always indicate a difference in origin. Secondly, it is difficult to make a comparison between the styles of the two regions. Iconography in Ceylon was mainly represented by Buddhist images while in South India it was generally represented by Saiva and Vaiava images. In the absence of a tradition of casting aia and Vaiava icons, the Sinhalese craftsmen, when employed to cast such images, would have evidently turned to South India for the style. l4oulds may have been brought from the mainland and the bronzes cast in the island. Since there would have been little difficulty in transporting bronzes across the narrow straits, some may have been im orted from South India. It is, therefore, not possible to determine who cast these bronzes. All that we can say is that they belong to the South Indian school

1. U.C.H.C., I, pt.2, p. 6 9. 2. A.L.Basham, he onder that wa India, p. 376.

123
of sculpture. In the case of the aiva temples, however, it is possible to conjecture that they are the work of South In ian stone-masons who had gone to the island. Some of the Tamil masons employed in the time of Parkramabhu I fr the building of Buddhist structures may have been descendants of the C 1a masons

The evidence of place-names for the period prior to the thirteenth century is almost negligible. Only a few recorded forms of Tamil toponyms which show some Dravidian association are available for the Ca period. Even these forms have to be used with extreme caution owing to several reasons. Under normal circumstances the occurrence of Tamil or Tamilised toponyma would indicate the presence of Tamil settlers in the places represented by them. But the Tamil and. Tamilised place-names of the Ca empire outside the Tamil country do not necessarily indicate Tamil settlement. Tamil names were often arbitrarily given to places by the Ca administrators. Nost of these were frequently altered to commemorate personalities and events. As Nilakanta Sastri puts it, 'the subordinate divisions evidently underwent numerous reshufflings, and their names were changed so often as to justify the complaint that 'Ca geo raphy came to suffer as much from the plague of homonyms as the kings themselves' ' The following Tnmil names of places in Ceylon

1. See infra, p. IS'f 2. I.A.Nilakanta Sastri, The Cas, p. k65.

124
are known from the Ca inscriptions: Jaantha-mag4ani Rj ar ja-pur a4 R jar ja-perunt-t eru R j ar j a-v4anu' Vikrama-. cia-v$antu Arumoi-tva-v4anu Parakc ari-v4anu' Nikarili-ca-v4an Rjndra-c inka-v4a.nt ? t1r-nu, are

and Vikrama-p tiya-nallTh? All these, except

names derived from those of roya1tiea All of them, with the possible exception of the last, have evidently been given by the Ca administrators. Such names, therefore, do not always reflect the existence of Tainil settlements in those places. There are a few other Tamil toponyms occurring in the CVa epigraphs of the island, such as Zkar, Nallr, kari-nund Palava-

1. 6.1.1., IV, No. 1388. 2. Ibid., No. 1k12. 3. Ibid. 4. 6.1.1., II, p. k26. 5. Ibid.

6. 6.1.1., IV, No. lkl2.

7. Unpublished - No. I 775. 8. A.S.C.A.L for 1909, p. 27. 9. Unpublished - No. I 357. 10. M.E.R. for 1917/18, p. 1k3. 11. Ibid. 12. See supra, p.gj2,.

13. S.I.I.,

IV, No.

1393.

125
putu-ci$i But it is not possible to determine whether these are place-names of Ceylon or South India. Some of these appear as 1!vita villages held by CVa officials while some others occur as the places of origin of certain donors. As villages in South India were sometimes assigned to officials serving in Ceylon, as in the case of Ti Kumara, who was assigned the village of Ciu-ka-nallr in Cnmsalam although he was serving at Mahtittba and as many of the donors mentioned in the inscriptions may have come from South India, we cannot be certain that these villages were in Ceylon. Moreover, some of these names occur both in the Tamil country and. in Ceylon, so that it is difficult to identify them in any particular region. NaUUr, for instance, is a very common name in South India. In Ceylon, too, there at least four places of that name Putu-kui also occurs in both regions It seems probable that most of these places mentioned in our inscriptions were in South India. There are a few Tamilised forms of Sinhalese placenames which occur in the contemporary inscriptions, namely Ntffam5 (Sinh. ?toa, Pli Mahtittha), Pulainari 6 (Sinii. Polonnaru,

1. Unpublished - No. I 776. 2. See suDra, p. iO 3. See

k.

p. qrr. Pudukkuiya in Anurdhapura district and Putukki4i in

Mafig4anu in South India, M.E. for 1917/1 , p. 89.

5. S.I.I., IV, No.l1fl ; see supra, p.1o1.

6. A.S.C.A.R. for 1909, p. 27.

126
Pli Pulattbinagara), Koiy'rain1 (Sixth. ICoasara, P].i ICoffhasra) and. Velakmam (Sixth. Velagama, Phi Velag.nia). The occurrence of Tamilised forms of toponyms in the Ca inscriptions may not always suggest Tamil occupation of the areas denoted by them. The context in which they occur is equally important. The long list of South-east Asian place-names in the inscriptions of RjThdra I, for instance, are Tamilise d forms of Malay and other names They may have been Tamilised earlier by the Tainil traders who frequented these places. But what is important is that they occur here as places invaded by the Ca navy, and when their names came to be recorded in Tamil the original forms could not inevitably be retained. The mere fact that the kames of these places are Tamilised may point to Tamil association with those areas but does not necessarily indicate T mil settlement there. The situation is different in the case of the above plpcenames in Ceylon. They were probably Tamilised as a result of Tamils living in those areas. lttam is a name that occurs in Tanzil literature as early as the seventh entury There is some evidence for the presence of Tamils in this pI&ce in the Anurdhapura period Tamil inscriptions of the Ca period have been found here
1. S.I.I., II, p. 1426 ; see supra, p .'(. la. Unpublished - No.1 357.

2. LA.Nilakanta Sastri, The Cas, pp. 215-218. 3. See supra, p.33

14. 5.

See supra, See supra, p. Iol

121
The present name of this lace is a variant of this Taniil form. It still continues to be part of the Tamil areas of Ceylon. These considerations would lead u to infer that the Tamilisatiozi of this place-name was the result of Tamil settle ent in the area. Similarly, Kotiyram is a place-name still in use. The area which it denotes is settled by Tamils. Velskmm and Pulainari are not in use today. But Taniil inscriptions as well as Dravidian ruins of the C1a period have been found in the places denoted by them It is, therefore, very probable that these Tamilised forms of Sinhalese place-names are the result of Tamil settlement. Thus, the evidence of toponyms have to be used with caution. By itself, it cannot form the basis of any important conclusion, It could, however, be used to strengthen arguments based on other more reliable evidence. The evidence of the four different types of sources that we have just analysed generally corroborate and supplement each other. While some of them provide more information than the others, the general conclusions that ,could be drawn from them remain basically the same. In the first place, while these sources reflect the presence of a larger number of Tamila than before, they do not at the same time point to any great Dravidian migration having taken place in the wake of the Ca conquest.

1. See supra,

128
They reveal the presence of Tamils in certain regions but no particular area o considerable size seems to have been completely settled by them. This becomes clear if we try to trace the places where evidence of their settlement is available. According to the literary sources, Tamils were found scattered in the villages and market towns, all over the kingdom. There were C1a strongholds in the Kurungala district, the northern region of the ?tale district and in some parts of the Batticaloa district. The evidence of the other sources confirm this. The find spots of the inscriptions show that the Tamils were living scattered in the northern parts of the island. With the exception of the inscriptions from Attaragalla and Diyavinna, all the other inscriptions could be grouped under five regions, namely the Burulu and Nuvaragain divisions of the Nuvarakalviya district (North-central Province), the Siih4a Pattu division of the Tamanic.aIuva district (North-central Province), )ntai in the Nanr district (Northern Province) and Periyakuam in the Trincomalee distriet (Eastern Province). The inscription from Nahkachchatkoi could be grouped with those of the Hurulu division, for its provenace lies only about three miles outside this divisiox Dravidian archaeological remains have been

1. a) Rurulu - inscriptions from FAoragoa, Padaviya, Paragiyaviya; b) Nuvaragam - Anurdhapura, Sagili-kanadarv , Atkaa; c) Sih4a Pattu - Polonnaruva, Giritale, Gal Oya.

1)
.1.

discovered. in the Eurulu division, Sih4a Pattu division and Periyak4am. Tamilised forms of Sinhalese place-names occur in Nntai, the Trincomalee district and the Siih4a Pattu division. There is no archaeological, inscriptional or place-name evidence of the Ca period in the Kurungala district, ?tale and Batticaloa where, according to the Clavawa, Ca strongholds had been established. But there is evidence of this type, belonging to the twelfth century, which indicate the presence of Tarnil aettlers in those areas during that century Possibly these settlements of the twelfth century, or at least some of them, had their origins in the Ca period. Thus, there is a general agreement between the different types of evidence relating to the settlements of the period of C 1a rule. We may not be wrong, therefore, in concluding

that there were Tamil settlements in this period in ntai and. in some parts of the Nuvaragam, Hurulu and Si4a Pattu divisions and in Periyak$am. This conclusion is further strengthened by the fact that these are the very areas where tie presence of Tamil settlers

is indicated

by the sources of the Anurdhapura

period Although each of the arguments to assume Taniil settlements in these areas in the Ca period may not be very strong, they have a certain cumulative strength which is enough to justify us

1. See infra, p. I&. 2. See supra,

p4.1.

130
in provisionally marking those areas as occupied by Tamil settlers. There is, however, no sufficient evidente to warrant the conclusion that there was a large-scale migration of Dravidians into the island in this perio&. The available evidence shows a strengthening of the earlier settlements, which appears to have been mainly due to the arrival of more traders, mercenaries artisans and officials, The history of the subsequent period clearly shows that the Sinhalese were in control of the northern and eastern parts of Ceylon, which were to fall later into the hands of Tamils and Ker4as, for the next century and a half. The Sinhalese chroniclers, who refer to the abandonment of the northern regions by the Sinhalese and their occupation by the foreigners, give no such comment for the C]a period Furthermore, there seems to have been no reason for a. mass exodus from South India in this period. The Cas provided political security for that region and tke general impression given by the inscriptions is one of prosperity everywhere. These sources, of course, do not reflect the true economic condition of the common people. However, only four instances of famine in the Tamil country are known from the

1. 2z 80: 63-78, 81:1-10.

131
1 inscriptions. All these famines occurred after the eleventh century and were confined to small areas There was, of course, the need for new land, as appears from the numemous references to the reclamation of forest and waste land and to the efforts 'to increase the area under the ploh and the inducements offered to encourage such efforts on the part of the people' There ' t5 I14t.VtY , no evi ence of any migration to foreign lands due to political or economic reasons other than perhaps commercial motives. In this period, colonisation did not follow imperial expansion, as in modern times, and there is no evidence that any policy of establi hing settlements was followed by the C1a rulers. But there is, however, some evidence regarding the settlement of certain castes or communities, for the performance of particular duties, in different parts of the Iamil country by rulers and local assemblies. The Brhamaas formed one such community. Nany BrThmaa villages, called agrahras, mag4ms or catur-vdi-mafig4anis, were created by royal grants and 'new colonies of pious and learned Brahmins were settled in the different parts of the couhtry' Another community, which was ettled close

1.

.P. for 1 99, No.33; LE.R. for 1911, No.29;

M.E. for 191k, No.17 ; N.E.R. for 1935, No.1k.

2. See infra, p. jqo 3. M.E.R. for 1902, Nos. 4 85, 506 ; Y.E.R. for 19 3, No.385 ;

N.E.R. for 1911, No.2 7, etc. ; K.A.Nilakanta Sastri,

The C j .as, pp. 584-5 5. , 4. K.A.Nilakanta Sastri, The Cas, pp.k92-493.

132
to temples for the performance of services to them, was that of the ckara-piyr Some records of the thirteenth century attest to the creation of mercantile settlements by chiefs and local assembliea in South India It is not known whether such a practice existed in the eleventh century. The Brhmaaa of Polonnaruva and the cakra-piyr of Mahtittha may have been settled in those places by Ca authorities. The caturvdi-maxg4am,at Kanta, of which we know from an inscription of the time of VijayabThu I ( L055- l].l0 ) may have similarly

originated in the Ca period. But these instances are different from a policy of settling people in conquered lands. In a Tamil epigraph of Ku]Zttuga I (1070-1120) from Mulbagal, in }&ysore, there occurs the phrase i tou nilaik4 iar4i ' It has been

rendered aa:'who was pleased to establish settlements of people on all sides (in the conquered country)' The implication is that Ku1ttuiga opened up settlements, presumably of people from his e pire, in newly conquered territories. But the translation is not accurate, for nilaik :j. doe8 not mean settlements of people but stations and what was mean evidently were military posts rather than settlements.

1. Lee supra, p. 103. 2. M.E.R. for 1935/3 , Nos. 150, 196.

3. E.Z., IV, pp. 19k-195.

i. E.C., X, No. k2b.

5. Ibid., p. 81.

133
CHAPTER Ill

STLE1'TS IN THE LATE ELEVENTH AND THE TWELFTH CE!TURY

The slow migration and settlement of the Dravidians

ttss4' Mt

in the island seem to have continued,Awith greater vigour,in the period after the Ca rule. A number of
Tami-1

inscriptions

belonging to the late eleventh and the twelfth century have been found in the northern and eastern areas of the island. The evidence of the literary sources, too, suggests a growth in the strength and influence of the South Indian element in the country in the period between the end of Ca rule and the invasion of Zgha (lO7O-l25). In this period, the reign of Parkramabhu I (1153-1186) may be said to mark the heyday of Sinhalese power and glory. This monarch, while he succeeded in controlling the growing influence of the South Indians in the island and in preventing any inroad from the mainland, 'left the country in a state of exhaustion at the end of his rule' The political confusion that ensued the reign of Parkramabhu I greatly helped the increase of Kalifiga and Dravidian influence on an unprecedented scale and, in some ways, prepared the ground for the rise of an independent Tamil kingdom in northern

1. U.C.H.C., I, pt.2, pp. 5k7,

716.

134
Ceylon This period, therefore, sees the culmination of that had begun to influence the course of the island's history towards the end of the &nuridhapura period. From about the middle of the the SwaLi.ec VitLf thirteenth century&had to contend with a permanent enemy in the northern regions of the island, instead of the traditional enemy from the mainland. The period under review was conspicuously free from foreign inroads. On the contrary, there were Sinhalese invasions of South India, and the island was involved in South Indian politics to a greater extent than ever before. The very close relations between South India and Ceylon in the political, cultural and economic spheres brought into the island more and more mercantile communities 1 mercenaries, artisans and BrThmas from the Dravidian kingdoms. The invitation of foreign mercenaries by aspirants to the Sithaleae throne, a common feature in the Anurdhapura period, was absent in this period. But there were strong and influential bodies of foreign mercenaries in the island in this period, many of whom, as we shal]. see later, seem to have gone there along with some of the mercantile communities.

1. The political confusion that followed the reign of ParlkramabThu I and the rise of Kaliga influence have been dealt with by A.Liyanagamage in his thesis, The Decline of Polonnaruva and the Rise of DaMbade, University of London, 1963.

135
The most important feature of tbi8 period, in regard to Dravidian settlements in the island,

is the

presence

of a number of mercenaries, traders, artisans and Brhmas from a].]. parts of South India. The most important among these Dravidian communities were the mercantile bodies known as the Aififi1!ruvar,
Valafljiyar

and the Nnidis as well as the mercenary

forces called the V aikk!ras. An analysis of the activities of the mercantile communities reveals that they may have been responsible for the migration of several traders, artisans and mercenaries into the island in the eleventh and twelfth centuries. More than seventy inscriptions, in Tamil, Telugu, Knnada and Malayalam, which refer to the activities of the mercantile communities mentioned above, have been discovered in South India. These range between the eighth and the seventeenth century. Outside the South Indian peninsula, there are seven in Ceylon one in Sumatra2and another in Burma which refer

1. i) S.I.i., IV, No. 11+05, from Anuridhapura; ii) E.Z., I,

p. 181, from

P4iyafiku!ama ('I) ; XVIII,

iii) E.Z., II, p. 236 , from

naulundiva; iv) El.,

pp. 330-338, from Polonnaruva;

v) Unpublished - from Vhlkaa; vi) Unpublished - from Padaviya; vii) Unpublished - from VihrThTnna. 2. M.E.R. for 1 89 1 -1892, p. 11.

3. E.I., VII, p. 197.

136
to them or to their associate bodies. These range between the nita.nd the twelfth century. The widespread activities of these mercantile communities have long been recognized by South Indian historians. They have often been referred to as mercantile gu.ilds and autonmous corporations a! merchants But the actual nature and organization of these 'guild as well as the relationship between them and the large number of professional bodies often associated with them have not been fully analysed. There has been considerable confusion in the names used to describe these 'guilds'. In Ceylon, writers on the history of this period have made passing remarks on these communities and have collectively referred t them as Valafljiyar Some of these, like the Nakarattr, have sometimes been mistaken for divisions of the VaikkIrae To understand the presence and the activities of these communities in Ceylon, it is necessary to nlyse briefly their organization and activities in South India. The Ai1iuvar, NLidis, Va1ajiyar and the Nakarattir appear together in the majority of their inscriptions

1. K.A.Rilkanta Sastri, The

p. 595 ;

T.V.I4ahalingam, South Indian Polity, pp. 389 If. 2. U.C.H.C., I, pt.2, p. 550. 3. 5 Paranavitana, 'The Polonnaruva Ins cript ion ol Vi jayabThu I', E.I.., XVIII, 1926, pp. 33k, 335.

137
in South India. The moat prominent among these were the Aifruvar, who were also known as the Ayylv4e, Vra Balafijiyar and Banaflju Dharmaara as well as by numerous other variants of these names Most of their records are preambled by a long eulogy or praasti,
giving an account of their origin and achievements. By far the

largest number of these, nearly forty, are in Kairnada while there are about a dozen each in Tamil and Telugu and a handful in Malayalam. The earliest of these goes back to the eighth century The latest is dated aka 1602 (A.D. i68o) At present, therefore, the origin of this community can be traced back only to the eighth century. The earliest inscription of the AifitTuvar is found at Aihoe, the ancient Ayvoe, which they claim as their seat of power. Since the earliest available of their records comes from Aihoe, it is possible that it belongs to the early years of their history. The origin of this community, may,

1. M.E.R. for 1 9 1 6, No.97 of 1915 - Ayyappoa1 ; ., VIII, p. 89 of the text - Ayyvajeya Aynrvva SvImig4u ; i.E.R. for 1918, No. 18 of 1917-18 A - Samayins of

Ayyav4i;

XI, p. 181 of the text - Aiv4e )thzkhyarada

2. JT.Fleet, 'Sanskrit and Canarese Inscriptions', l.A., VI, Nay 1877,

p . 138.

3. ?,E.R. for

1918, No. 18

of 1917-18 in Appendix A.

138
therefore, date back to about the eighth century. There is little doubt that Ayyvoe was the place of their origin, for not only did they call themselves the yIvoe or the Five Hundred of Ayyvoe but they also named many of the places where they had established themselvea in later times as Southern Ayyvoes (Teftka Ayvoe) In most of their inscriptions of the twelfth and thirteenth centuries, the members of this community claim to have come from Ahichchatra in North India It is not possible to say how far this claim is true. But it is not impossible that some of the wandering traders from North India were responsible for the founding of a mercantile body which later grew into this powerful community. The name of this community has been the subject of various interpretations by different scholars. In most of the Kannada inscriptions, it appears as the Five Hundred Svznina of Ayyv4e (Ayyv4eya Ayn!rvva Svmig4u) Several variants of this name occur in the Tamil epigraphe, such as tci-

ticai-1yirattu-airruvar (live Hundred of the Thousand Directions of the Several Countriea)

1. 2. 3.

V, p. 325 of the text. .E.R. for 1906, No. 180 of 1905. VIII, p. 89 of the text.
Ptvo%.

$A-NiL..tc..Ii4* S.c4ri, '4

G1'J.

k.,( Tiidschrift Voor Indische Taal-, Land- en Volkenkunde, LXXII, 1932, p. 318.

139
(Five Hundred of the Thousand Directions of the Four (quarters) Ticai-yirattu-aififiuvar (Five Hundred of the Thousand Directions)2 and Aififih1uvar (Five Kundred) Sometimes they are just referred to as Aiv4e, Ayyav4e or Ayyappoal. The use of the name Aiflffiiuvar has led R.C.Majumdar to think that this 'organization' consisted of five hundred members T.V.!.halingam, on the other hand, feels that 'their extra-territorial organization was managed by an executive committee of five hundred members' A,Appadorai, too, holds a similar view when he states that the 'most important personages were constituted into a board called the Five Hundred Svmis of Ayyvoe'? L.D.Barnett's opinion is that this 'corporation' had their central body at AyyIvoe, which .as the seat of their Board of Directors, consisting of a council of five hundred members All these opinions are based

1. K.A.Nilakanta Sastri, A T; \ ir 2. Ibid. 3. K.A.Nilakanta Sastri, The C5as, p. 596. k. M.E.R. for 1919, No.223 of 1918;
.'

E.C., II, (Revised ed.),

p. 78 of the text; M.E.R. for 1916, No. 97 of 1915. 5. R.C.}lajumdar, Corporate Life in Ancient India, P. 88. 6. T .V. Mahalingam, . cit., P . 392.

7. A.Appadorai, Economic Conditions in Southern India, II, p. 39k. 8. L.D.Barnett, The Ancient Tamil Township and Village, Unpublished, quoted in B.A.Saletore's Social and Political Life in the Viiay anaara Empire, LI, p.'I8 j.i.

140
on the interpretations of the term AiZfiuvar and deserve closer exeini-nation. Majumdar's view that the organization consisted of five hundred members is not tenable. The activities of this body was never limited to any one area or century. On the contrary, its records have been found in several districts of South India and. in Ceylon and. Sumatra. These range over a period of nearly ten centuries. It is, therefore, improbable that the organization had a constant number of members throughout this period and ail over the vast area covered by t their activities. The meaning of the term Aiuvar has to be expiLined differently. The opinions of Appadorai, Barnett and Nahalingam are based on the assumption that the Aifflh1uvar were a single unified body with their headquarters at Ayivo.e and that all the members mentioned in their inscriptions all over South India owed allegiance to a central body. It seems rather too much to expect the Aiflfluvar to have been such a unified body with a continuous history of nearly ten centuries and with branches all over South India and even overseas. Such an organization is too modern a concept and may not be applicable to this period of South Indian history. A careful analysis of the inscriptions reveals that there is no justification for the view that the AiZffI1ruvar represented a central body of the corporation or for the theory that they had their headquarters at Ayyv4e. In the first place, it is

141
not necessary to assume that the membership of the corporation or its central body, if ever there was one, was limited to five hundred. In this instance, the number five hundred has to be taken as a conventional number from which this mercantile community baa derived its name. This may be explained in several ways. The number five hundred may preserve the memory of the number of people who banded themselves together and originated this mercantile community. This number may have been a figure close to five hundred and may have been rounded off to five buntred. Naming guilds and other bodies after what may have been considered their numerical strngth is not something unusual in Indian history. In the Jtakaa we get references to carpenters and robbers organized in guilda of five hundred Certain other guilds had a thousand men in each of theme Even in later times there were such bodies in South India. An inscription from Travancore refers to a body of six hundred while another from Bijpur refers to a body of BI'Ihniaas called the Five Hundred 14ahjanas (}lahljanag4 Aynflrvvara)

1. The Jtakas, IV, ed. E.B.Cowell, Tr. W.H.D.Rouse, (1901),

p. 268 - Sattigumba Jtaka.

2. Ibid., p. 99 - Samudda-vIija Jtaka.

3. E.I., V,

p. 47.

142
There was also a community of BrIbmas in Cidambaram (Tillai) who were known as the Three Thousand of Tillai (Tillai NvIyiravar) In modern times we get the example of the Syrian Christians of Ker4a being divided into the Seven Hundred (Eunhrukar) and the Five Hundred (LfiffUrukar) Th. name Five Hundred is, therefore, no indication of the numerical strngth of the AififT!nvar in the later centuries of their history. There is also another possibility of explaining this name. The AiZfit!ruvar claim in their inscriptions that they migrated from Ahichchatra. This may mean that the community originated in South India with the arrival of some wandering traders from North India who already belonged to a mercantile corporation known as the Five Hundred, for it is quite possible that such corporations continued to exist in North India from the time of the JItzka. It is equally untenable to claim that the Aififlh1ruvar had their headquarters at Aihoe or Ayylvoje. The fact that only one, and that the earliest, of their records has been discovered at Aiho.e is not without any significance. If the Aiflffuvar had their headquarters in this town, it is strange

1. M.E.R. for 1922/23, No. 395 of 1922. V 2. T.A.S., II, p. 75.

v7t+

143
that out of over seventy records left by them not a single one belonging to the period after the eighth century has been found at Aiho. The fact that they continued to call themselves the Aififfaruvar of Ayyv4e or Ayyvoe does not necessarily mean that they had their headquarters in that town. There is no etidence to suggest that they had any connection with Ayyvoe alter the eighth century. As we have pointed out earlier, .Ayylv4e appears to have been the town where this mercantile community had its origins. It seems possible, therefore, that the Aifi!flTuvar were originally known as the Aiffuvar of AyyIvoe. Nembers of their community seem to have used this name in the later centuries, too, when they had established themselves in other parts of South India. The occurrence of Ayyvo.e in their name seems , therefre, to be a reminder of their place of origin rather than a reference to their headquarters. Their practice of calling many of the new places where they established themselves as Southern Ayyvoes (TerTh or Dakia Ayylv4e) also shows that they considered Ayyvoe to be their original home Further there is no evidence

1. Cf., The Ki BrI1' maas and the Tillai BrThmaas called themselves so even when they were settled in places other than Ki and Tillai respectively.

144
to suggest that there was any kind of communication between AyyIvoe and the other places where the Aififf1ruvar were found. Ayylv4e seems t have declined as a centre of their activity before the tenth century, when we begin to get their records in other places. There is also no evidence to support the hypothesis that the AiIflUruvar were a single unified corporation of merchants. The application to the Aiflflh!uvar of the terms 'corporation' and 'guild' appears to be rather UfljflBtified. It seems more appropriate to cal]. them a community of merchants with common origin, interests and beliefs. They were all bound together by the Banafiju Dharma3 which they claimed to follow There is no evidence of any other fora of bond, administrative or otherwise, between them. There is little evidence regarding any definite organization in their community. An official called Paaa-svimi is met with in most of their inscriptions. But be appears as the head of a mercantile town who took part in the meetings of the Aiflffruvar as well as of other mercantile bodies It is, therefore, difficult to decide whther he was an official of the Aiflfi71uvar or the head of a town acknowledged

I. EC., VIII, p. 89

of the text.

2. E.C., VII, Inscription No. 9k from Shikarpur TV.uq.

145
by all the professional bodies in the town. The term ff aa -svmi (lord of the town) itself sug est hat he was the g neral head of a town rather than an official of any particular body. The only other official term that occurs in the records of the
Aiflfl!ruvar is

igeya, which is of doubtful meaning' Some have

suggested that it could mean a passport department But it is difficult to determine the real meaning of the term. The Aiflflh!uvar may have been a loosely organized body because of their community of interests but sufficient evidence is lacking to call them a corporation or a trading guild. The fact that some of their associate bodies like the Valafijiyar and the
Nakarattlr,

who are conunonly referred to as mercantile corporations

and guilds have survived to this da as mercantile castes in South IndiaL4 should serve as a corrective to the impression that the
Lifluvar

and other trading communities of this

period were organized as corporations

L.

II, (Revised ed.), p. 90 of the Kannada text.

2. Ibid., p. 78 of the translation.

3. A.Appadorai, . ., p. 378.

4, Iperia1 5. Even

Gazetteer of India, XVIII, pp. 188-189.

in modern times the business community of Ceis have

their own fiscal year, stick to their own system of book-keeping and follow their own type of business practices and customs. But they could hardly be called a corporation on these grounds.

146
The AifffhIruvar were primarily traders in various types of merchandise as they themselves claim in their inscriptionai They call themselves wandering traders and claim to have visited a large number of countries, many of which were in North India and some outside the subcontinent But none of their inscriptions has been found in any part of North India and this claim may not be altogether true. But the fact that their records have been left in Ceylon and in Sumatra shows that they were an adventurous community whose members went to far off lands in pursuit of their profession. Apart from their fundtion as traders, they seem to have occupied a leading position among a larger number of occupational groups in the towns, exercising much power and influence over tbem In many of their inscriptions, we get as

1. E.G., VII, p. 159 of the text. 2. E.C., VII, No. 118 from Shikarpur T.1uq - The following countries are mentioned:- 'Chra, Cba, PIya, Nagadha, Kausala, SaurItra, Dhanustra, Kurumbha, Imbboja, Gaufla, La, Barvvara, Praaa, Npja, kapda, Lambakara, StrT-rljya, Ghola-mukha emba nIndafigaum'. 3. Ibid., Noe. 118 and 119 from Shiknipur T.luq, pp. 158-163 of the text.

147
many as forty-six such bodies associated with the Aififih1uvar and their Banaflju Dharma3 These include the NLdie, Baajigas or Valafljiyar and. the Naaratt1r. The exact relationship between the AififUruvar and these bodies is not easy to determine. Nany of the inscriptions give us , however, some idea of the nature of this relationship. Certain writers have expressed the view that the Aif1Zuvar were a federation of all these bodies and not a community by themselves T.V.Mahalingam considers th associate bodies to be sub-divisions of the Ain1mznvar But these views are again based on the assumption

1. E.C., VII, pp. 158-159 of the text ; M.E.R. for 191 8 , p. 17k. The following communities are mentioned:Gavaras, Gatrigas, Seis, Seiguttas, Mtgakras, Settiputras, Bra-vaigaa, NnIdis, IIu, Nagara, Baafijigas, Etivraa, Nuai-vras, I flcika-vTraa, Koga-vas, Kaalis, Bhadrakae, Gav$aevmis, igam, iu-puli, Valattnkk.i, Variyas, Paradis, Svadis, Va1afLgai-taiyar, Niyyattir, BTras, Ga4igae, GavuQ4ae, Murnmuri-da4as, 4vaQa-klras, V!rakoi, Vyavahri1cas, Pa.flchflaa, KumbhalflcaR, Tantuvayins, Vastrabhedakae, Tila Ghak, Kurantak s, Vastra-rakk, Dvagas, Parikeliti, G-rakakas, Kfras, Rjk and Zauras. 2.. S.Cbandrasekhara Sastri,'Economic Conditions under the Hoysalas', Half-Yearly Journal of the Mysore Univeristy, II, 2, July,( p.223. 3 TV. Mahalingam, . cit., p 390.

145
that all these bodies were organized trade guilda. There is no doubt that the Aiflruvar were a community d.etinct from *1]. others mentioned with them. They are referred to as such in their records, where the Aiflffaruvar, NnIdie, the NrattIr, the eighteen samayas and other communities are distinctly listed as different bodiee But in all these instances, the leading position of the AiflfiIruvar over all the others is clearly brought out. There are many instances to show that the Aifiruvar presided over several meetings where affairs of the other communities were aettled The prahsti appearing in many of their records is called the AynUrvvara praheti and is clearly in praise of the This community seems, therefore, to have enjoyed considerable power and influence over the many pafeboiona-]. groups in the towns and trading centres of South India. They, as well as the other major mercantile communities like the Valajiyar and the Nakarattr, were conceded a share of the administrative duties of the state. We find in the inscriptions that they had

1. LE.R. for 1925/26, No. 131 of 1926 ; M.E.R. for 1919, No. 216 of 1918. 2. E.G., VII, p. 159 of the text. 3. Ibid.

149
a share in the collection of tolls, taxes and rates 1 and had the power of declaring certain towns as i-v!ra-paa4as and Southern
Aryv4es

They also reserved for themselves the

power to grant trading privileges in certain articles to individual tradera They were great benefactors of temples to which they granted part of the tolls and rates collected by theme The communities associated with the
Aiflflh1ruvar

were not all mercantile in character. Many were other occupational groups which later evolved into castes. Examples of such comnn(ities are the Paflc3las (the five classes of smiths), K.unibhallkzae (potters) and the Kauras (barbers) who were among the eighteen samayas. Even the Valafijiyar and the Nakarattr later evolved into castes There were also some communities which were given to martial pursuits, such as the Ei-vras, Muai-vTras, Ificii'lkz-vTraa, Koga-vI.as and the Mummuri-daQ4as? Their names

1. E.G., VII, p. 159 of the text; M.E.R. for 1919, No. 9 of 1918-19, Appendix A , No. 216 of 1918 ; M.E.R. for 1912, No. 377 of 1911. 2. M.E.R. for 1913, No. 3k2 of 1912; E.C., VIII, p. 89 of the text. 3. LE.P. for 1919, Nos. 10 and 11 of 1918-19 Appendix A.

k. M.E.R. for 1912, No. 377 of 1911.

5. The Imperial Gazt.teer of India, XVIII, p. 198. 6. Ibid. 7. See infra,

150
suggest the nature of the work they were doing. As mentioned earlier, there are seven inscriptions in Ceylon which attest to the presence of the AiZ.ffuvar and. their associates in the island although Ceylon has been omitted in the list of countries covered by their activities, furnished in their inscriptions The earliest of these inscriptions comes from Anurdhapura and its contents have already been discussed The inscription is datable to the ninth century. The Nku-nItu (Four Countries), a community identifiable with the Nlku-nu of the Kannada inscriptions, was responsible for setting up this record. The Nilku-n4u of the Kannada inscriptions are found associated with the Aiflffruvar and were probably a trading community like the ].u-nakarattr (Those of the Four Towns) This may mean that some associates of the Ai1Iflh1uvar, and probably the lattei too, were in the island in the ninth century. There is definite evidence regarding the presence of the Valafliyar and the Nakarattr in Ceylon in the twelfth

1. See eupra, p. 2. See supra, p.

3. E.C., VII,

p. 310 of the text ; E.C., VIII, p. 89 of the text;

LE.R. for 1917, No. 130 of 1916.

151
century. The source of our information is the fla l]drTra inscription from Polonnaruva, which appears to have been set up not long after the death of Vij&yabThu I (iUO) The Valafijiyar and the Nakarattr are referred to here as those closely associated with the Vjikkaa. The exact relationship is couched in the following lines of the inscription:)-tantirattm ki ekajukku nittaikai ua Valaflceyaraiyuni eau kt!i varum Nakarattir uiraiyum_ki ..... 2 Paranavitana has rendered it in English as follows:We of the I4ahtantra, having called together the Valafijiyar who are our leaders, and the Naka.rattr and others, who always accompany us... 3
The Valaceyar of our inscription,were, of course, the

Valafijiyar who are sometimes referred to as Baafijigas in the Kannada inscriptions. In the first place, these lines inform us that the Vala.jiyar and the Nakarattr were in Ceylon along with the VjaikkIras in the twelfth century and possib].y in the eleventh, too, for there is evidence for the presence of the Vaikkras in the eleventh century But more important than this is the light thrown on the nature of the relationship

1. S.Paranavitana, 'The Polonnaruva Inscription of VijayabThu I', 2 i ' pp. 330-338. 2. Ibid., p. 337. 3. Ibid., p. 338.

k. See supra, p.c7.

152
between the Valafljiyar and the Vaikkras.Commenting on this Paranavitana writes that 'it seems from our inscription as if the three divisions or 'hands' to which the V.ai-kkras were divided consisted of the Mahitantra, the Va1ajiyar and the Nagarattr', and adds that 'as the Valafijiyars are said to have been the leaders (nfddai) of the V i-1ckra troops, it might be conjectured that the latter migrated to Ceylon with the Valafjiyar whom they served' But, as Ni].knta Sastri has pointed out, there is no reason to assume that the Valafljiyar and the Nakarattr formed two of the three divisions of the aikkras

The lines quoted above refer to the Valafijiyar as the nt!ttaik4 of the tja ikk ras. Paranavitana has translated the word nttitaikaj as leaders. Nilakanta Sastri has the following comment to make on this translation:The translation of n!tdaig4 into 'leaders' is not quite accurate; the word literally means 'grandfathers', and what is meant cannot be physical descent when it is one corporation claiming this relation to another, and must imply some kind of spiritual, or constitutional. relation. 3 Although nfjtaik4 means 'grandfathers' or 'ancestors', it could also be taken to mean elders In this context, it is not

1. S.Paranavitana, 'The Polonnaruva Inscription of VijayabThn I',

2. K.A.Nilakanta Sastri, 'VijayabThu, Th. Liberator of Ceylon',

p.335

J.LA.S. (c.B.), N.S., IV, 195k, p. 70.

3. Ibid.
k. E.Z., II, p. 25k. D.}Lde Z.Wickrn2Ringhe has given this

translation.

153
possible to take the first two meanings, for the Valafijiyar being a mercantile community and the V t1dc!pa being a mercenary body we cannot say that one is descended from the other. The meaning 'elders' seems to be more appropriate. But here, too, the Valafijiyar cannot be taken to be the elders in the physical sense. They appear to have been regarded as the leaders of the T!af1dc!ras, as Paranavitana. baa rendered, and seem to have been elders in the social sense. This relationship becomes clear if we look at the social structure in South India in this period. In the eleventh and the twelfth century, and in fact till recent times, the various castes of the Dravidian areas were divided into two major sections called the Iei (Left Rand) and the Valakai (night Hand) The Vjatkk!ra inscription under discussion attests to the presence of the members of these two sections in Ceylon in the twelfth century Certain mercantile communities were considered to be the heads or leaders of these sections. In the case of the Valki 1 the N.udia and the Valafijiyar were among the leaders while the Iikai had

1. B.A.Saletore, Social and Political Life in the Vijayanagara Empire, II, p. 68 ff. 2. S.Paranavitana, 'The Polonnaruva Inscription of VijayabThu I', p. 337.

the 1a1r,rattir as one of their leaders The V1?ri- were often considered to be of a higher social atatus than the IaI%kRi Although the ffaikk ra forces in Ceylon were drawn from both sections, as they admit in our inscription, the Nabltantra division, who alone call the Valafijiyar their mtaik4, appears to have been drawn from the Valkii. This assumption is further strengthened by the fact that the 'Nakaratt.r

4ir'

are referred to as those who accompany the atantra The phrase Nakarattr baa been translated by Paranavitana

as the 'Nakarattir and others'f' But literally uiffr means

'those included', from uflitu meaning 'include'. The phrase would, therefore, mean 'those included (in a group) with the Nakarattir'. This seems to be a reference to the Iafi1tai leaders. The reason why this group is mentioned as those who accompany the Z4ahitantra must be the subordinate position held by them
in the presence of the ValAksi. The leaders of both sections

were invited for the meeting of the

aik1ra.g obviously

because the latter were drawn from both seetions of the Dravidians.

1. E.G., XI, p. 61 of the text ; The Imperial Gazetteer of India,

XVIII, pp. 198-199.

2. B.A.Saletore, 2 cit., p. 68 ff.

3. See supra, p. 151

k. See

supra, p. s5S

155
It seems, therefore, reasonable to assume that the mercenary forces called the Vfl.a ikkrae went to the island along with the mercantile communities who, as we have seen earlier, had a number of martial communities associated with them in South India. Some of the Vjaikkras, however, may have gone there independently and later acknowledged the leadership of the mercantile communities. The Valafijiyar, Nakarattr and some of the mercenary bodies were only a few of the associates of the Aififfauvar who were in Ceylon in the twelfth century. There are at least three Tamil inscriptions in the island which refer to several others. These records, which are unpublished, are found at
Vihalkaa,

Vihrh!nna and Padaviya and contain the praasti at the beginning Unfortunately all are

of the

Aififii1ruvar

badly damaged and the actual purport of the inscriptions cannot be ascertained. These could be assigned to the twelfth century on pa].aeographica]. grounds. The script of these inscriptions is very similar to that of the Tamil inscriptions of the time of GajabThu II ( 1132-1153). Apart from these Tamil epigraphe, there is another record of the Aifffih1ruvar in the Sinhalese script of the twelfth or the thirteenth century. Unfortunately only

1. See supra,

p.i3n'

153
three lines of this inscription, containin( part of the paasti, have been preserved Since the praasti is in Sanskrit, it is not known whether the rest of the inscription was in Sinhalese or not. The prag asti in the Tamil inscriptions is a shorter version of that appearing in the Kannada records and begins with the words Samaatha bhuvanraya pafica ata v!ra sisana. As in the Kolar inscription of ii8O the AiflfiUuvar describe themselves here as the 'children of the Goddess of the City of Ayyappoil'
(&,yappo]4pura Paramvarikku makkaj). The praasti is followed by a list of members of the different communities who were

associated with the Aiflflttuvar in the island. The number of communities is not as large as in the South Indian inscriptions. Among those mentioned in the preserved portions of the inscriptions are the Ceis, Ceiputras, Nndis, Valafijiyar, V!rakkoti,
Valaf&kai, A2gIckiras, Xvaa 1dc ras, Iaflc
I A1m

and the Koga-vias.

The Cettis were traders as well as money lenders or bankers In our inscriptions they are sometimes referred to as the 'Ceis of the countries of the eighteen worlds' ( patifl!.
Several meznbezs of this community are

-pThniuceftk4).

1. D.Lde Z.Wickramasinghe, 'Polonnaruva: Anaulundva Slab-Inscription' E.Z.,

II, p. 236.

2. E.G., X, No.170 from Kolar -' Ayyv 4epura Paramevariya makk4'.

3. A.Appadorai,

. cit., pp. 379-380.

151
named in the8e inscriptions. In the VThalka4a inscription, these Ceis and the V!rakkoiyar are recorded to have done something in order that a certain town may not be destroyed. It is not clear what the nature of their work was and which town was protected in that manner. There are some place names in the North-central and North-western ProviAces with cei fl eir first element. These may date from the twelfth and thirteenth centuries and may indicate the presence of Cefti settlers in those places The Ceiputrae may also have been traders, bat we have no information regarding the nature of their activities. The ndia, as we have already noted, were a community of traders like the Valafijiyar. In many of the Kannada inscriptions they are referred to as the Ubhaya Indis (both ndis) It is not known who these two classes of Nndis were. In an inscription from Bangalore, there occurs the phrase svad!a t'arada nindam (local and foreign Nnida may be that this mercantile community was divided into a It

..

c,,i,,,, I?

?) 1 , c-

2. E.G., IX, p. 83 of the text ; E.I., XIX, p. 25. 3. E.G., IX, p. 1 of the text.

158
local and a foreign group. But this phrase could also mean 'those of the svad&a, parad!a and nThda (communities),' for we come across a body known as the Paradis in another inacription It is difficult to settle this question without ez.i m-i- n ing some of the unpublished inscriptions which remain inaccessible at the moment. It is also not known whether the Nndis were actually a community of merchants from different countries, as the name implies, or whether they were so called because they were descended from foreign merchants who had established their business in the Kannada country, where the community seems to have originated. Whatever their origin ma have been, they were among those who travelled to distant lands in pursuit of their trade, possibly with the AiZflturuvar. Their presence in Pagan, Burma, is known from the Teinil inscription found there In Ceylon, apart from the evidence of the Tamil inscriptions, there is also a Sinhalese epigraph of the time of queen Lt]1vat (1197-1200, 1209-1210, 1211-1212) which mentions - them. According to this record, in the reign of LT]1vat the lndia (Nnd!i vyp ray an) had an alms house at Anurdhapura

1. M.E.R. for 1932/33, No. 173 of 1932/33. 2. ., VII, p. 198. -

3. D.M.de Z.Wickramasinghe, 'The Slab-Inscription Marked of queen LT]ivat!', E.Z., I, p.

179.

159
As in South India, these mercantile co n"'fl 'nitiea wereengaged
in providing charitable services and in patronising religious institutions. The VT.rakkoi or Vrakkofiyar were another mercantile community found in Ceylon in about the twelfth century. They are recorded in the VIha 1kaa inscription to

have associated themselves with the Cetis in tkirg certain steps to protect a town They are mentioned in a few South Indian inscriptions, too, but do not seem to have been a prominent trading comnnznity The Akkkras and the va zkk ras appear to have been two of the non-mercantile Dravidian communities that were in the island in this period. The
mentioned

1Akak1ras are

frequently

in the

records of the AiffflXuvar but it ha not been

possible to find out the nature of their profession. An akakk.ra,

in T41,

is a dandy or a masquerader and is derived from the

Sanskrit word a.figa (=body) Perhaps they were professional entertainers who specialised in masque. The Avaa kr rae do not find mention in the South Indian inscriptions but are referred

1. See supra, p. JS7. 2. M.E.R. for 1910, No. U of 1910.

3. Nadras Tamil Lexicon, I, p. 18.

160
to in the V from
1kala

and VihirhThna epigraphe. The

meaning market

name

is

derived

am (Skt. and other

or

bazaar Perhaps the

4va4 akkras

were those who were responsible for the maintenance

public places.

of markets

The Valkai of our inscriptions are the eighteen

castes of South India who were categorized under this name. The term occurs in the Jaik1cra inscription as well, along

with the term It.Aki This clearly shows that the South Indian caste system was maintained b.mong the Dravidiana in Ceylon, too. The Valafijiyar and the } ndEis as well as some of the laikkras
were Valakai comnmnities But no information is available

regarding the number and names

island.

of

the Va1a.kai castes

in the

There is no information at all regarding the activities of the Koga-vias

4aciAkams

(young lions). It is

not possible

to conjecture from the name the nature of their profession. The appear to have been a community given to

military

pursuita The name means 'swordsmen of Koigu'. They may have been a class of sword fighters who were among the mercenary communities

1. Madras Tamil

Lexicon, I, P. 21+9.

2. See supra, p.

I3Sumatra', p. 319.

3. See enpra, p. I^31+.

E.A.Nilk2nta Sastri, 'A Tamil Merchant-guild in

161
who accompanied the mercantile bodies. The Vhalkaa inscription has a list of several military personalities who were associated with the Ceis, VTrakkoiyar and the Valafljiyar. They have such titles as C!pati (army chief), mallan (wrestler) and (victor). A certain Citta, is referred to as the Valaftceyar Cpati, apparently because he was in the service of the Valafijiyar. Mercantile communities may have employed mercenary forces to protect their endowments as well as to safeguard their trust properties. This perhaps explains the presence of Munimuridaas ( a class of mercenaries) on many occasions when grants were made by the mercantile communities Some at least of the mercenary communities from which the Vai1dras were drawn seem to have gone to the island along with the mercantile communities. The claim of the }1ahtantras, a section of the rja1kkras, that the Valafijiyar were their elders and the invitation extended t the Valafljiyar and the Nakarattr to attend an important meeting of the aikkaa

at Polonnaruva show that these mercenary bodies were closely associated with the leading mercantile communities from South Indja

1.

VII, p. 159 of the text ; E.G., IX, p. 83 of the text; E.G., XI, p1 126 of the text.

162
As pointed out earlier, the period under review was one during which the practice of inviting foreign mercenaries to the island was absent. But despite this, the Dravidian mercenaries formed one of the significant sections of the army
of

the Sinhalese kings. Of these, the V!.aikk r aa iLndoubtedly

formed the most important troops. Since much has been written on the origin and history of these mercenary forces, it may be necessary to point out some of the misconeeptions regarding them among writers on Ceylonese history and to attempt a better understanding of the subject The best source of our information on this subject is undoubted]$ the Polonnaruva inscription of the aikkrae. Be8ides this, there are at least two other

Tamil inscriptions

in the

island, from Ga]. Oya and P4aniDai, q1si

referring to the Vjaikkras and the notices in the

Nilakanta Sastri has explained that the word vaikkra is derived from the word v!ai (= time, occasion, moment) and that it stands for the 'time or occasion indicated in an oath by the soldier who binds himself by the oath to lay down his life in certain contingencies4 Sastri has also given the alternate

1. K.L.Nilakanta Sastri, 'Vijayabhu I, The Liberator of Ceylon', pp. 58-60, 67-71; The Cas, pp. 315, 316, 'ik, 455; S.Paranavitana, 'The Polonnaruva Inscription of V.jayabhu I', p P. 333-335; T.V.!halingam, . cit., pp.

258-260. p.68

2. LA.Nil2kpnta Sastri, 'Vijayabhu I, The Liberator of Ceylon',

163
interpretation that 'their designation implies that they were ever ready to defend the king and his cause with their lives when occasion (v!ai) arose' It is not possible to determine whether the element vai is the sane as the Tamil word meaning time or occasion. It is quite possible that it is derived from some other word now unknown to ue References to the V!aikkras occur in Taniil inscriptions and literature from about the eleventh century. These mercenaries were not all in the employ of kings. Many were employed by village assemblies and other institutions They seem to have been divided into different types according to the nature of the duty performed by them. We meet with the following types, for instance, in the sources mentioned above:- a) P-VaikkLar (Tami]. p = flowers)t b) K4a-fffr

dr.ra

(Tamil k4jaz' = robbers) c) Rkasa-V. dcrar

(skt. rkasa = giant) d) Tacca-V.aikkrar (Tainil Taccar=carpenters)?

1. LA.Nil2k1,ta Sastri, The Cas, p. k5k. 2. Cf., Indoneasian word b&la, 'to defend'. 3. B.A..Saletore, I, . cit., p. 3k8.

4. K.LNilakanta Sastri, 'VijayabThu I, The Liberator of Ceylon', p.67.

5.

. . R. for 1914, No. 368 of 1914.

6. M. .R. for 1921, No. 393 of 1921. 7. LR. for 192k, No. 194 of 192k.

164
e) Tiru-ciila-aikkrar (Taniil tiru = sacred, = spear) and 1) Tiru-ciampala-Va1 krar (ciampalam = hal]. of wisdozn) The exact functions of these different V,ik1r.ras is not clear from their names. PU-V!.aikkrar may have been those who guarded the flower gardens in a temple. Kaa-Vjafkkirar may have been employed to guard a place against thieves. Rkasa Vaikkrar may have got this na4e because of their size. Tiru-cfla-Vaikkrar may have been guards of temples who were armed with spears. We can only speculate on their functions from the names they bore. A number of divisions of VaikkIras, probably in the service of kings, were named after kings and princes. Among them were the (a) Nittavinta ffaUdcirar (b) Jaantaterifica V.aikkrar1'
Cc)

k]akiya Ca-teriflca Va1 kkrar

(d) Aridurga-lngbana-t erint a Valafkai Vaikkrar (e) Candra Parlkrama-t erint a Va].afikai aikkrar' and (f) Iaiya-rja-terint a

1. M.E.P. for 1925, No. 188 of 1925. 2. Ibid., No.

2k3 of

1925.

3. LE.R. for 1927, No. 282 of 1927.

k. M.E.R. for 1921, No. 393 of 1921.

5. 3.1.1., II, Introduction, p. 9. 6. Ibid. 7. Ibid.

165
Va1Aki VjaUdcrar The epithet Va1a.fkai or Ijatkai denoted their caste group. These mercenaries were not a 'warlike tribe or a clan or a military community' as Geiger thought? but 'a type of troops bounds by specific oaths of loyalty which they were bound to keep at the risk of their own lives' They yore drawn from different castes and were probably organized as a military guild. R.C.Majumdar takes them as 'a good example of Katriya The assembly of the Vaikkra community at Polonnaruva as well as the organized manner in which they sometimes revolted against the Sinhalese rulers may support this contention. But apart from these, there is no substantial evidence to prove this conclusively. Vai1dcra mercenaries were employed in Ceylon in the time of Vijayabhu I (1055-1110) and possibly even earlier, under the Cas. The Colombo Museum Pillar inscription of Kassapa IV (898-91 k ) has a reference to a Vekk who was a body-guard Judging from his profession, this person may have

1. 8.1.1., II, Introduction, p. 9. 2. W.Geiger, Culture of Ceylon in !iedieval Times, p. 152. 3. K.A.Nilakanta Sastri, 'VijayabThu I, The Liberatthr of Ceylon' ,p.58.

k. See supra, p. 1S3.

5. R.C.Majumdar, . p. 31.

6. EL III, p. 276.

166
been a aikkra (Pii V.akkr& may have become Vekk in ), but it is not certain
as

Sinhalese by the omission of the final

whether there were tenth

VakkTras in the island

early as the

century. A Tamil inscription from Gal Oya, near Polonnaruva,

records a grant by a certain Atikaraa, C&Aa4a, a MIu-kai Tiru-Jaikkra Although on palaeographical grounds this epigraph may be dated to the eleventh century, it is difficult
to say whther it is a

Ca record. It may well belong to the say whether there were V.a1-kkiras
in

time of Vijayabhu I. In the present state of our knowledge, it is not possible to

the island before the time og Vijayabhu I. The Cilavaisa makes a few references to the

influence exerted by the Jaikkras in the island

e^tLtt ttf4tLi

the

centuries. It

is clear from these that

the

Sinhalese rulers

from

the time of Vijayabhu I depended

to a great extent on these mercenaries for the defence of their kingdom It is not necessary to assume that all these mercenaries came from South India. As Nilknta Sastri has pointed out, some
of the

Vlaikkras may

have been enlisted from

among the Dravidians

who were settled in the

island

This is

also

perhaps one of

1.
2.

I.I.,

IV, No.

1398.

.,6O:36, 63:2k, 29

and 77:kk.

3. K.A.Nila.kanta Sastri, 'VijayabThu I, The Liberator of Ceylon', p.60.

167
the reasons why we do not hear often of mercenaries being enlisted from the mainland in this period. From the Polonnaruva inscription we learn that there were several sections among the Vaikkraa in the island. The Mahtantra appears to have been the leading group among them. According to this inscription, it was the Nahitantras who first met and invited the Va]Jjiyar, the Nakarattir and others for the assembly of the V.aikk!ras at Polonnaruva. They seem to have followed a code of conduct or rules called the Mahtantra for, at the end of the inscription, there is an imprecation to the effect that those who violate the Mahtantra will go to hell We agree With Nile 1 nta Lastri that the interpretations that they belonged to some sort of Mahyna 2 or Saiva sect seem to be very unlikely. In the South Indian inscriptions, the term Mahtantra occurs as the name of a military community or claas Nilknta Sastri is right in suggesting that this term may relate to some school of militarism in South India -

1. E.I., xVIII, P . 337

2. K.A.Nilakanta Saetri,'Vijayabihu I, The Liberator of Ceylon', p.71. 3. M.E.R. for 1917, No. k33 of 1916. 4. K.A.Nilakanta Sastri,'Vi4yabThu I, The Liberator of Ceylon', p.71.

168
Among

the other sections of the Va{k-kraa the

Polonnaruva inscription mentions those of the Valakai, I1kRi, Ciu-taam, Piaik4-taam, Vatukar, J!a1ayIar and Parivrakntam. This mixed composition of the VTi

{kkiras

clearly

shows that they were not members of one military caste or community but were organized more like a military guild. Of these different sections, the Vaikkras of the Valfika{ and Ttafikai. were obviously tho8e drawn from the two categories of Dravidian castes known as Vale lflc ei and Iaikai. The ValAkai

laikkras

appear to have been further sub-divided into

various sections. In the P4anilai inscription of the fortysecond year of VijayabThu I (1097), a Vaikkra of the Vi
kid

rama-ca].mka-terinta Val _b1 division is mentioned

This division was apparently named after VikramabThu I (1111-1132), the son of VijayabThu I and bearer of the consecration name Ca1mka (Pii, Si1megba) The naming of a division of the army after a ruler indicates that the Sinhalese rulers were following a South Indian practive

1. S.Paranavitana, 'A Tamil Slab Inscription from P4amftai', IV, p. 19k. 2. As Vikramabhu considered himself to be the legal successor of Vijayabhn I, who bore the consecration name of Siri-saghabodhi1 be would have adopted the consecration name of Si1-megha-vaa. 3. See supra, p.

169
The meanings of the terms CiEu-taarn and Piaflc4ta 1am are still obscure. Ciu-taam and P, un-taam occur in a number of South Indian inscriptions of this period. The term 4u-ta has been variously interpreted as (a) private

treasure (b) minor or small treasury (c) the followers of the king during his minority Cd) one of the 'purely honorary titles conferred on officers as well as private individuals according to the status held by them in official position or societyk and (e) a c1as of subordinate officials The first two interpretations are based on the assumption that the element is derived from the word dhana, meaning treasure or wealth. The term occurs in connection with certain officials whose position could be described as military. We get, for instance, the phrases u-taattu Vafuka kvalar (the Vatuka,

i.e. Telugu, guards of the Ciu-taam) and Ciu-taattu Valafikai Vaikkra_paaik4 (the Valakai Vaikkra troops of the Ciu-taaxn) in two South Indian inscriptiona In another

1. Madras Tamil Lexicon, III, p. lk6o. 2. S.Paranavitana,'The Polonnaru'va Inscription of VijayabThu I',p. 336 S.I.I., II, p. 9. 3. S.Paranavitana,'Tlie Polon.naruva Inscription of VijayabThu I',p.336. i. Ibid. 5. S.I.I., II, Intro., p. 11; K.A.Nilknta Sastri, The Cas, p.k63.

6. E.I., XVIII, p. 336.

170
inscription, an official who had military personnel under him is described as Citu-tanam Perun-taam ri:ya (The Great chief or Lord of the Citu-taam and Perun-taam) In yet another record, certain }alay3a officers of Kulttufiga I are stated to have 2 been attached to the Perun-taam and the Ciu-taani. A Ca record in Ceylon refers to an official of the Perun-taani of Rjndra I (RjThdra C1a Tvar Perun-taattu paimaka) serving in the island While in Ceylon Dravidian mercenaries were in the Citu-taain and the Pi.aik4-taijam, we find that Telugu and Nalaya persons were attached to the Ciu-taam and the Perun-taana in the Ca country. The above occurrehces of the terms u-tan, perun-tan and aik4 am show

that these stand either for departments of the state or for divisiona of the army. It is not possible to explain u-tan

as private treasure for it does not suit the context in the above instances. It is used, however, in this sense or in the sense of treasury in some of the C3a inscriptions of the time of Rjarja I ' The explanation that the ciu-ta and

1. N.E.R. for 1913, No. 14]. of 1912. 2. M.E.R. for 1938/39, No. 130 of l98/39.

3. 5.1.1., IV,

No.lLflk.

4.. 5.1.1., II, P . 3.

171
perun-taxi were the 'minor' and 'major' treasuries respectively, though plausible, leaves the term inexplicable.

It is not known whether the royal treasury was divided into three different departments to which were attached three sections of the army. This seems unlikely. It is also unlikely that the guards and soldiers who protected a king during his minority belonged to a section of the army different from that of those who protected him in his later life. It is also not possible to explain these terms as mere honorary titles,for, in the instances quoted above as well as in the Polonnaruva inscription, such an interpretation does not seem to suit the context. Hultsech baa sometimes referred to the Ciu-taam as a class of subordinate officials and Ni].akanta Lastri, too, is of the opinion that this was a flower grade of official nobility' The latter opinion is the result of confusing the terms perun-tan. and perun-tarani as referring to the same institution But an examination of the occurrences of these two terms in the south Indian inscriptions reveals that the two are different in their meaning. Peruntaram appears as an honorific indicating high official

1. S.I.I., II, Intro., p. 11. 2. LA.Nilakanta Sastri, The Cas, p. k63. 3. Ibid. ; peruntarani - perum (= high or big) + taram (=status).

172
status and was used for individual officers, whereas perun-ta seems to have been a department o a body to which were often attached a number of soldiers. Perun-taram is often used for single individuals as, for instance, ar RLjirlja Dvar Peruntaram which occurs as the title of a s!npati in a Tanjore inscription Another inscription from G5vindaputtflr describes a person called VikramCa Nahirja, as a peruntaram of Mumnnii Ca But whenever individual officers are mentioned in connection with the Perun-taam, they are referred to

as those

attached

to the Perun-taam as, for instance, Perun-taattu-paimakap (servant of the Perun-taam) The occurrence of the phrase Ciu-ta.attu peun-taram (perun-taram of the Ciu-taam) in some of the inscriptions not only shows clearly that the terms perun-taram and perun-tan are different but also demonstrates that the latter stands for someone attached to a larger body which is the Ciu-taani Nilkanta Sastri explains this, however, as implying an intermediate status between the perun-ts and the u-tanam in the official nobility This is not correct
FYvntY

1. s.I.I., II, p. 161. 2. i. .R. for 1928/29, No.168 of 1928/29. 3. S.I.I.,IV, No.lk])i.
i. s.i.i., II, p. 56. 5. K.A.NilaJcanta Sastri, The Caa, p. 1163.

173
for the term perun-taram in such instances stands for individuals rather than for a clase. None of the foregoing explanations for the terms u-ta, perun-tan and fl.aik4-tan,

therefore, seems to be wholly satisfactory. The ju-ta occurring in the sense of treasury in some of the Tanjore inscriptions of Rjarja I appears to be different from the ciu-tan of the other inscriptions. In the latter inscriptions we may have to take all the three terms to stand for certain sections of the army which might have bad different functions in the admtnistration during this period. Wherever the nature of the profession of those individuals and groups associated with these terms is indicated, we find that it was military. The derivation of the element that it is related to is not clear. It is unlikely

i (army). It may be related to

(army; Nalayalam t4am ; Kannada d4am) The Vaukar and the Malayjar were, of course, the mercenaries from the Telugu and Ker4a countries. This shows that the recruitment to the V

i1dcra army

was not

confined to the Tamils alone but included other Dravidians as well. The literary sources, too, contain frequent references to

1. The Ja of t4am may have interchanged with va and later became a. Cf., Kannada baaftjiga - Tamil v4afijiyar.

the Ker4a and ica4ia mercenaries in the employ of the Sinhalese rulers in this period The Telugus are, however, not mentioned in these sources. Keraja and Telugu mercenaries seem to have been numerous in the Ca country in this period, while Kannaa mercenaries went as far north as Bengal in pursuit of their profession The Cavaa makes a distinction between the Ier4aa and the ffaikkras which might mean that only a section of the Ier4a mercenaries were included in the VN.aikkra army The Parivra-kntam of the Polonn2rTa inscription is not known from the South Indian inscriptions. It is, therefore, difficult to say whether it was just another division of the Vflaikkra army or a military community included in that army. It has been suggested that it may stand for the spearmen in the king's procession ( p arivL,a) A division of the Ca army was known as the pariirattr and a number of such divisions are named in the inscriptions We also come across a troop of

1. Cv., 69:18 ; 70:230 ; 7:144. 2. D.C.Sircar, 'Karta5 outside Karta', J.N.Banvrjea Volume, p. 211. 3. Cv., 7k:k4.

k. D.M.de Z.Wickramasinghe,'Polonnarnva: Slab Inscription of the

Vaikkrae', E.Z., II, p. 25k. 5. .I.I., II, Intro., p. 9.

175
body-guards known as parivra-meykpparka In modern Mysore there is a caste called the Parivira Bant, which is claimed to have been originally a military class The Parivra-kntam of our inscription appears to have been a similar military body which was perhaps associated with the royal procession. There has been some difference of opinion among scholars regarding the interpretation of the phrase M1u-kai airk'ra,. This occurs in the Polonxiaruva inscription as well as in another Tamil record from Gal Oya Mu-kai has been generally taken to refer to three divisions in the Vaiickra army. 'It seems from our inscription as if the three divisions or 'hands' to which the ffa{k k'ras were divided , consisted of the Mahtantra, the Valafljiyar and the Nagarattir', is the comment of Paranavitana on this phraae Wickramasinghe baa observed: 'Whether the term u-kai refers to the triple

principle, namely, giva-$akti-Au or Pati-Pacu-Pica corresponding to the trika of Cashmere aivism, or it is only an epithet of the VaUckras due possibly to their army being composed of

1. S.I.I., II, p. 96.

2. J.Sturrock, Manual of South Kanara, I, pp. 156-157.

3. S.I.I., IV, No. 1398. k. S.Paranavitana, 'The Polonnaruva Inscription of VijayabThu I',
p. 33k.

1(0
three wings, we are unable at present to say' Nilknta Zastri is inclined to think that refers to the 'traditional

three arms left after the chariots went out of use, viz., elephant corps, cavalry and i.nfantry' As we have seen earlier, the Valafijiyar and the Nakarattr were mercantile communities and. not divisions of the V!aik1 ra army. There is no evidence suggesting any connection between the triple principle of Saivism and !u-kai. Although Nilakanta Sastri's suggestion seems to be plausible, there are certain difficulties in accepting it. In the first place, there are some inscriptions in which an individual member of the Ia ikkira army is referred to as Mu-kai VWikk raa It is difficult to assume tat

some of the Vaikk.ras belonged to all three divisions of the army, if their army was divided into three different sections. Secondly, it is not likely that the V.aikkras called themselves Nu-kai Vaikkras because their army was divided into three divisions, for, such a division was not a distinctive feature of their army alone. It is more likely that the epithet

1. D.M.de Z.Wicla'amasinghe, 'Polonnaruva: Slab Inscription of the Vaikkrae', p. 251. 2. K.A.Ni].akanta Sastri,'VijayabThu I, The Liberator of Ceylon', p.69.

3. S.I.I., IV,

No.1398.

177
u-kai has some other significance which we are not in a position to grasp clearly. does not occur only in connection

with the Vs1&ras. In an inscription from frrmdvi, we are informed of the existence of a regiment called the Mu-kai Mahsai, who were also known as the Paai-piitta-pafl1yiravar (The Many Thousands who are armed as a Troop) 4occura

here in a similar sense as Valkii in the names Valakai Nahsai and Valakai 1a1ickrar It seems likely that NUu-kai was the name of a group of castes like the Valki and the Iakai or of a community given to military pursuits This name might have been chosen for some reason unknown to us. Although the Vaikkraa appear to have been the most prominent of the Dravidian mercenary forces in the island during this period, there were other Dravidian troops, too. Of these the Agampai troops deserve mention. It is in the reign of Parkramabu I that we first hear of the Agampais. The Ni1ya-sarahaya mentions their army among the forces despatched by Par.kramabhu on his foreign expeditions. This Agampai

1. Y.E.R. for 1905, No.120 of 1905 ; LA.Nilakanta Sastri, 'Vijayabhu I, The Liberator of Ceylon', p. 69. 2. 1.E. . for 1911, No. 116k of 1911; see supra, p.. 3. u-kai may be a variant of }m-kai (the third hand), the Valkai and the I1'IkRi being the other two 'hands' (kai), and may denote a third group of castes, possibly a minor one.

178
army is said to have been 2, 1125, 0 00 (sic) strong, which is undoubtedly an exaggeration The Aganrpais are again referred
to in the Polonnaruva Council Chamber inscription of Nia.ka

NaUa The literary works and inscriptions of the thirteenth, fourteenth and fifteenth centuries contain several references to the Agampai troops of the later Sinhalese ru1ers Four classes of Agampais, namely the ja-, Nuhukala-, Netti- and Bla-Aganipais are mentioned in these sources. In the South Indian inscriptions, the Aganipai or Akampaiy.r find mention from about the time of Kulttufiga I (l070-ll2O) They often served under minor chieftains and their leaders or chiefs were known as Akapai-muta1is The Akanrpaia are mentioned in the inscriptions of the Tamil country and of Ker4a and appear to have been a military caste or community rather than a mixed force like the 1a1ic1ras? Even the women

1.

p. 18.

2. C.J.Sc, (G), II, p. 137.

3. Dabadei-asna, p.3 ; Mayura-sanda, v. 157; C.J.Sc. (G), II,

p. 139 - Niya.gampya inscription ; E.Z., III, p. 2k0 - )aavala

rock inscription of Parkramabhu VI. 1i. M.B.Ariyapala, Society in Medieval Ceylon, p. 162. 5. M.E.R. for 1926, No. 72 of 1926. 6. M.E.R. for 1913, No. 506 of 1912.

7. T.A.S., V, p. 1k?.

179
of thths community (apai-peuk4) found service in the inner apartments of the palace and in the teniples The name AkanrpaiyIr is a compound of the Taml-1 words akampu (inside or inner apartment) and 4iyr (servants) and this community may have originated as a class of servants in the inner apartments of the palace and the temples, and evolved into a caste. This caste has survived to this day in Arcot, Pudukktai and Nadural districts and is variously known as Akampaiyr, Akaniui and AkamufiyL In Ceylon, too, this caste was existent in the Tamil areas till very recent times As in some parts of South India, the members of this caste seem to have gradually mixed with the Vear and given rise to the saying that 'the K4.ar, Maavar and the staunch Akaiup4iy.r have gradually become Vetar' (K4ar Maavar kaatta Akampatiyr niella india Vear
ki

This saying is prevalent in South India

as well as in Ceylon Some sections of the Jkampati caste in the Madurai district are 'regarded as a more civilized section

l.11.E.R. for 1913, No, 506 of 1912. 2. A.F.Cox, }'Ianual of North Arcot, I, P. 211; N.Thiagarajan, A Manual of the Pudukai State, pp. 202-203.
3. K.Velupillai, a-vaipava-kaumuti, p. 108;

M.B.Ariyapala, 4. A.F.Cox, ,

. cit., P. 162. . cit, p. 211.

180
of the southern Maavars' In Ceylon, too, certain writers

consider the Agampai to have been South Indian Maavar who were taken to the island as znercenaries There is, however, no evidence on this point. Whatever their origin may have been, it seems certain that by the twelfth or the thirteenth century they had become an exclusive caste and that several of their members bad gone to Ceylon as mercenaries. After the twelfth century, our sources record the presence of some other Dravidian mercenary communities serving under $inha].ese rulers. Of these, the Nukkuvas and the Kurukulas, who in modern times are among the major castes in the
Tanhil

regions of Ceylon, appear prominently But it is not known whether they bad already begun their migration to the island in the twelfth century, The Dabadei-aana gives the earliest reference to the Mukkuvae. It is recorded here that they formed part of the troops employed by ParlkramabThu II (l236-l27O) The Kurukulas may have been in the island as early as the time of the Ca occupation. As pointed out earlier, the earliest reference to the presence of the Kurukulas in Ceylon may be said to be found

l.A.P.Cox,

. cit., p. 211.

2. Pufiflarataha Thera, Lakv! Pur' Tattvaya, p. 95. 3. Dabadei-asna, p.

I. ;

Nukkuva-haana, p. 175 ff.

k. DaMbadepi-asna, p. .

181
in a Ca record from Tirumuicka1, of the year 1067, if the person named Kuruku1attaraiya in it is taken to be a chief of the Kuruku1as They are mentioned in later literary sources as having been in the service of ParkramabIhu VI (lkl2].k67) But no evidence is available regarding the presence of these two communities in the twelfth century. The Mukkuvas, as we shall see later, were Ker4ae and the Xuruku].as appear to have been from the Tamil country. It is possible that members of these two mercenary communities were also found among the many Ker4as and 'Dam4as' in the island during the twe]Lfth century. Apart front the mercenary troops who were

in the

regular service of the Sinhalese kings, there were also other South Indian soldiers who were taken to the island as prisoners of war by the generals of Parkramabhu I. It haa been. the practise of Sinhalese generals even in earlier times to capture prisoners during their South Indian campaigns, which were, however, rare, and send them over to Ceylon Lakpura, the

1. See sura, p.110 ; LD.Ragbavan, The Karva of Ceylon, p.9. 2. M.D.Raghavan, pp. 8-55. I.43. 3.

182
general of Parkramabhu I who conducted campaigns against the C2a and PIya rulers, was under special instructions from hi8 king to send prisoners from South India for a particular purpose. In the words of the Clavaisa, some of the defeated Tamil armies, 'at the commend of the ruler of LaAk who thought to have all the cetiyas formerly destroyed by the Dmijas rebuilt by them, he [LsAkpuraJ had brought to Lk and the

work of restoration begun on the Ratanaviluka cetiya. In another place, the Clavaisa further states that Lakpura, after having made over the government of the Pi4ya kingdom to VTra Pya, 'sent with speed to STh4a the many horses, men and elephants captured from the Coa country and from the Pap4u 2 land'. In addition to the restoration of ruined Buddhist monuments, the Tamil prisoners were engaged in the task of building new and ambitious structures as well. It is stated in the Ctflavaisa that ParkramabThu 'also had the Mahthpa erected which bore the name of Damia thpa because it had been built by the Damiaa who had been brought hither after the conquest of the Pa4u kingdom' It is confirmed by the contemporary Ca records

1. Cv., 76:103-10k. 2. Ibid., 77:103. . Ibid., 78:76-77.

183
that Lafikpura actually won many successes in his initial campaigns in South India We cannot, therefore, cast doubt on these statements of the P1i chronicle. The gigantic Daxn4a-tbUpa stands at Polonnaruva to this day, preserting the memory of the Tamil prisoners. It was intended to sprpass al]. other monuments of its type in Ceylon and. its circumference at the base is given in the Chronicle as 1300 cnbits Unfortunately it has not been possible to ascertain the dimensions of the base from the ruins at Polonnarua, as it has not been completely excavated. The dome stands at about fifty feet from the ground. forming an extensive circular plateau at the top ParkramabThu must have commanded a large force of South Indian prisoners to undertake the building of such a stpa and the repair of other buildings. have These prisoners must, in the course of time,/mingled with the Tamil population of the island. Some of them may have been employed in vihras, as on an earlier occasion when Tami]. prisoners were enslaved and. given over to vih.ras An inscription from the Ga1apta vihra, near Bentoa, dated in the thirtieth

1. K.A.Nilakanta Sastri, 'ParkramabThu and South India', C.H.J., IV,

pp. k6-48 ; U.C.H.C., I, pt.2, pp. k82-k83.

2. Cv., 78:77 ; S.Paranavitana, The tpa in Ceylon, p. 10. 3. S.Paranavita.na , The Stpa in Ceylon, p. 10.

- P - c.!..1 44 '

184
year of a ParkramabThu, who is probably the first of that name, mentions some Tamils among the slaves attached t that vihira It is possible that some of them were prisoners from South India. In addition to the Tam1l prisoners, Tamil artisans also seem to have been employed in the time of Parkramabhu I for the erection of Buddhist edifices. According to the C1flavasa, there was a dearth of stone-masons in the time pf Parkramabhu and, as a result, members of other occupational groups were employed for the work of stone-carving As Paranvitana has pointed out, in this period, 'when there was a demand for their art, it is likely that skilled workers came to Ceylon from the neighbouring continent, where they 'ust have been quite numerous at that time' As evidence of this, Tamil letters have been found as mason's marks on the stones of some of the buildings dating from the time of ParkramabThu I. These letters have been found not only at Polonnaruva, in such constructions as the Lotue Bath, but also in some momuments at Padaviya The building of aiva and Vaiava temples in the Dravidian style

1. S.Paranavitana, 'Galapta Vihra Rock Inscription', E..Z., IV, p.211. 2. Cv., 68:25-26. 3. U.C.R.C.., I, pt.2 , p. 592. i. Ibid. ; A.S.C.A.R. for 195k, p. 20.

185
of architecture would have also led to the employment of South Indian artisans in Ceylon. As mentioned before, a number of aiva and Vaiqava temples, roughly datable to the Po].onnaruva
period, have been found at Polonnaruva and. elsewhere With the possible exception of the Ca temples, it is difficult to determine which of the others were built before the thirteenth century. Nan; of them, as we shall see later, appear to have been constructed in the thriteenth century The Tanii]. inscriptions of this period attest to the patronage extended by some of the Sinhalese monarchs, especially VijayabThu I, Gajabhu II

and Vikramabhu I, to Saiviai In fact, two of these epigraphe

mention two temples named after VijayabThu I and VikaramabThu I. One i the temple of Vijayarja-Tvarazn at Kant4y and the other is the Vik1cirama-ca]imka-varam at kal (1galla) These temples were built apparently in the reigns of VijayabThu I

and VikramabThu I respectively. Some of the iva and Vin

temples at Polonnaruva may also have been built in the reigns

of these monarchs. It is, therefore, possible that several

1. See supra, p.117. 2. See infra, f411.^.

3. E.Z., IV,

p. 191 fl. ; Ibid., III, p. 302 If. ; S.I.I., IV, No.1397;

unpublished inscription No. 1 359 of the epigraphica]. list in the Archaeological Department, Ceylon. 1 E.Z., III, p. 302 ; ibid., IV, p. 191.

186
South Indian artisans found employment in the island in this time, as in the later periods. BrThmaaa were among the South Indian communities in the island in this period. The Polonnaruva period was one of increasing Bria influence in Ceylon. The services of Brhmaaa were enlisted for the performance of various rites in the royal court a.td palace. This was especially so in the time of Parkramabhu I when Brahmanic rites, we are told, were performed at every important occurrence in his life. The Ct!lavaisa refers to the sacrifices performed by BrIhznaas and to the alma offered to them by rulers like ParkramabThu and nbharaa The Tamil and Sinhalese inscriptions, too, furnish evidence on this matter.
The Sinhalese inscriptions of N1a1ika Nalla mention Brhmaas among those to whom that monarch offered Two Tamil

inscriptions from P 4am t ai and Nahakirinda refer to two

Brhmaa settlements named after VijayabThu and JayaAkoa Calmka (probably VikramabThu I), namely the Vijayarjacat urv!di-mag4am at Kant ajy and the Jayai.ko a-c alImkacaturvdi-magg4am at Nahk4rinda An unpublished Taniil inscription

1. Cv., 62:33, k2, k6 ; 6k:l6; 67:9k; 77:105. 2. II, P. 17k.

3. E.Z., IV, p. 19k; unpublished inscription from Nahakirinda, No. I 29k of the epigraphica]. list in the Archaeological

Department, Ceylon.

187
from 1Cantay associates Gajabhu 11 with a sacrifice held at

the brahinadeya

that place The same monarch is credited in

the T ia-]cailca-puram with the patronage of Brhmaaa attached to the temple of cvaram The patronage extended to them by the Sinhalese rulers as well as the need for them in the new temples of the island may have been responsible for the migration of BrIhmaas from South India. Although it is possible that there were in Ceylon BrThmaas from other parts of the subcontinent, there is little doubt that the catur-v!climafga!ams were settlements of South Indian BrThmaas. The occurrence of Tamil inscriptions in these places lends support to this hypothesis. Even the little internal evidence that we get in these inscriptions regarding the BrIhmaas points in the same direction. The P4aniftai inscription, for instance, furnishes us with the names of a BrThmaa couple from the Vijayarija-caturvdi-mag4am at Kant4y. They are KrLnpacceu Tajfia K.iramavitta and NaAkaiccii These persona were South Indians, probably of Telugu origin. In the contemporary inscriptions

1. No. I 359 of the epigraphical list in the Archaeological

Department, Ceylon.

2.

., VII, vv.95-97.

3. S.Paranavitana, 'A Tpm4l Slab Inscription from Paaxzai',

., IV, p. 195. The reading NJcaiccai for NaAkeicclzi is wrong.

188
of South India, especially in those from the South Arcot district,

Kiramavitta occurs as a name among the BrThmaas Kr.zzTpaccetcu, or more correctly, Krmpicceu, occurs as the name of a village somewhere in Uayrgui, in the South Arcot district Natkaicci, too, occurs in these inscriptions as a name of Brihmaa 1adies The element ci in this name is a Telugu word signifying women and often applied to names of married women as a mark of respects Numerous names of BrThmaia women with the element ci occur in the inscriptions of South Arcot and Guntr districts The BrThmaa couple of our inscription were, therefore, probably Telugus who came from South Li-cot or Guntr district.

1. M.E.R. for

1921, NO.556 of 1920 - The

name Tajfia Kramavitta

occurs in this ; M.E.R. for 1922/23,

No. 380 of 1922 - A

certain

mai Nagai4i, wife of Yajfla Kramavitta is mentioned here; ibid., Noe.369,

371, 37k and 382 of 1922 refer

to several persons

named Kramavitt an,. 2. M.E.R. for 3. M.E.R. for

1921, No.603 of 1920. 1922, No.380 of 1922.

k. M.E.R. for 1921, pp . 92-93. 5. M.E.P. for 1922, Nos. 380, 55k, 558, 57k and 585 of 1922.
Such names as NaAgai4Ii, AniarIeIi, Ayita41i, and Pr1-ii occur in these records.

189
As we have seen, our literary and. epigra hic sources rovide information, though by no means adequate, regarding the presence of bouth 'ndian mercantile, mercenary, artisan and Brhmaa communities only. There is hardly any evidence regarding the migration of peasant settlers during this eriod. ihe absence of any evidence, however, does not necessarily ean that no such migration took place. It is possible that side by side with the migration of the above occupational groups there were igrations of peasants, too. Such migrations and settlements may have occurred in the northernmost regions as well as in the north-western and north-eastern ].ittorals of the island which lay close to South India and where the power of the Sinhalese ruler at Folonnaruva does not seem to have been felt effectively. The areas of Taniil settle ent in this period, as indicated by the presence of inscriptions and archaeological finds, lie ostly in the vicinity of ancient irrigation works The majority of these settlement sites are far removed from the capital city and the known provincial towns, where such co unities as merchants and mercenaries would have norzzaUy lived. these considerations lead us to think that there may

1. See infra, p. )'1.

190
have been a Blow and unnoticed migration of small groups of peasants from the iami]. country into the island in this period. These are, however, matters of conjecture and in the absence of any evidence nothing definite can be concluded. The South Indian 6ources, while providing fairly substantial evidence regarding the migration of mercantile and mercenary bodies from the Dravidian regions to outside areas, are silent on the question of peaceful peasant migrations As mentioned earlier, there were at least three .2 famines in the Tam.il country in this period. These were not widespread but confined to certain regions only. We are informed by one inscription that there was a famine in Arakaanal].r, in the Squth Arcot district, in A.D. 1131 and that 'people moved after selling their lands' It is reasonable to think that such movements of people during times of famine were confined to South India, although it is pos ible that a few went to Ceylon, too. Like famine, excessive

1. See supra, .,3r. 2. See supra, p. g3 3. M.E.R. for 1931+/35, Inscription o.151 of K.A..Nilakanta Sastri, The Cas,

193k/35;

p.562.

191
taxation or inability to pay taxes also forced villagers to abandon their homes and migrate to other places. Under the thoroughly centralised revenue administration of the Cas, people who, for three years, failed to pay taxes due on the lands owned by them forfeited their lands, which were then sold by the village assembly+ Inscriptions of the reign of Ku]Zttithga I furnish instances of assessments not being paid regularly and the lands of tenants who defaulted payment being sold in consequence. Some BrThmaa tenants of Vavai-mtvicaturvdi-magaam, for instance, being unable to pay the assessments, left the village Again, in the forty-ninth year of Ku]Zttuga I (A.D. 1118), tenants deserted the village of K!ri-rAjapuram as they could not pay the taxes There are several examples of such desertions in the later But it is not possible to say whether there were many such instances during this period. Except for a few scattered examples, there is hardly any evidence regarding migrations.

1.

.E.R. for 1897,

2. !'.E.R. for 1910, Inscription No.98 of 1910. 3. Ibid. 1 Inscription No.6 Zf7 of 1909. k. Cf., B.A.Saletore, Social and Political Life in the Vijayanagar Empire, II, pp.l97-198.

192
Even these minor movements of people would have been confined to the Tamil country only. There is, however, a late tradition recorded in the Ca-prva-paftayam which refers to an overseas emigration of a hundred families from Trichinopoly in the time of one Vikramditya But the tradition loses its value by several discrepancies. According to this account, in the time of Vikramditya, 1SlivThana and his Sainaia troops lay siege to Trichinopoly. Thiring the siege a pariah named Ve and

hundred others with their families escaped, went to the seashore whence proceeded to some island' It is not clear which Vikramditya is referred to here. It is possible that the ruler was Vikramditya VI (io7, - ( Il-c. ) of the Western Ch.lukyas for, of the many Chlukya and Ba rulers of this name, it was Vikramditya VI who made successful inroads into the doniinions of the Cas. SlivThana is a variant of tavhana and its occurrence here is apparently the result of mixing up different legends regarding early invasions. It is hardly possible that this legend preserves any memory of the Stavhana invasions of the period prior to the third century A.D. The Samaa troops, according to Taylor, are in fact Yavana or 'uslim troops

1. C -prva-paayam, No.165 of the Lackenzie Manuscripts in 1.adras, quoted in the Analysis of the }ackenzie Nanu cripta, W.Taylor, p .1i., 60. 2. W.Taylor, . cit., p.5k.

3. ;_4.

193
The whole account seems to be of late origin, based on different traditions of earlier invasions, and has hardly a claim to any credence. Perhaps it refers to an iinlcriown migration, but this is a matter of speculation. It is possible that small groups of peaceful settlers from South India trickled into the island

in this

period. But the increase of the Dravidian element in the local population seems to have been steadily maintained mainly by the migration of mercenaries, mercantile communities and artisans. It was only after the downfall of Polonnaruva, as we shall see later, that many peaceful settlers from the Taniil country migrated to the northern regions of the island. The areas of Dravidian settlement in the island in the latter part of the eleventh and the twelfth century, as in the period of Ca rule, have to be traced primarily with the help of Tamil inscriptions and Saiva archaeolo ical remains. The meagre evidence of the literary sources and toponynis is useful in supplementing the testimony of the epigraphic and archaeological material. Inscriptions do not present much difficulty in the attempt of tracing some of the probable areas of settlement. Normally the provenance as well as the internal

194
evidence of the epigraphs help us to a great extent in locating the settlements. In most cases, the regnal years of kings which they provide help to date them almost accurately. Even in the case of inscriptions which do not carry dates, their palaeography aids us to date them roughly. On the other hand, the archaeological materials present several difficulties. The limited materials that are so far available to us are in the form of ruined aiva and Vaiava temples and icons. The presence of these temples and icons in any area almost certainly indicates South Indian settlements, often those of Tamils. A major difficulty which besets any attempt to trace the areas of settlement by such archaeological evidence is the doubt whether the date of an image or of a temple can be determined with sufficient accuracy to be of use to us. There are some temples which contain datable inscriptions. There are others, more numerous, which on the basis of the architectural style as well as the iconographic style of their images could be roughly assigned to the eleventh and twelfth centuries. But some of these may well belong to the early part of the thirteenth century. Therefore, only those few which could be assigned to the latter part of the eleventh and the twelfth century with a reasonable degree of likelihood are considered here in our discussion. The rest have been taken to belong to the

195
thirteenth century. There is room, however, for a small margin of error on either side. The other difficulties in the use of archaeological materials are mainly those concerning the identification of the aiva and Vaiava temples of this period. This is due to the transformation of some of these temples into Buddhist institutions and vice versa in the course of time. In areas where Tamil settlers have been assimilated to the Sinhalese population or where resettlement by the Sinhalese took place after the sites had been abandoned by Tamils, hiva and Vair&va temples have often been converted to Buddhist dv.les. This appears from the architectural style of the buildings or from the distinctively aiva and Vaiava finds in them A similar conversion, perhaps on a greater scale, has occurred in the areas where Tamils have established permanent settlements. As we shall see later, by far the largest number of such structures belong to the period 2 after the twelfth century. The date of such converted temples can Only be ascertained roughly with the help of earlier Sinhalese inscriptions found on their architectural parts or of the style of the buildings. In some cases stones and pillars

1. Cf., A.S.C.A.R. for 1911/12, p. 2. See infra,

8.

196
from ruined Buddhist buildings have been used to build aiva temples and such temples, too, cannot easily be dated. In areas where there has been a continuous Tainil settlement from this period or earlier, old temples have been renovated Q3ularly and kept in a good state of repair. Among these, it is difficult to identify those of this period unless inscriptions or other datable finds are available. As a result o such difficulties it is not possible to use afl. the archaeological evidence in a work of this nature until a thorough survey of all these temple sites has been completed. For the present, we have to rely on the evidence of those few which could be assigned to this period with certain degree of confidence. A few early forms of Tamilised Sinhalese place names are available from the inscriptions of this period. These

as well

as the little evidence of the literary sources can be

used to confirm and supplement the evidence of the epigraphic and archaeological materials. The P2i and Sirthalese chronicles have hardly any information to offer in this respect. But for the first time we begin to get fairly reliable traditions relating to this period in the

Tmi1 chronicles

of the inland.

The Takia-kiiI!c a-puram, T ii' i-kc ala-puram and the car-kalveu, all of which are chronicles of the temple of

197
Kvaram at Trincomalee, contain traditions of the time of Gajabhu II, probably preserved originally in the annals of Kvaram. With the help of all these sources it is possible to locate several, if not all, of the Tamil settlement sites of this period. Two regions have yielded the majority of the Tamil inscriptions and aiva archaeological materials for this period. One is the north-eastern littoral, forming the northern part of the present eastern Province, from the Kokkuly Lagoon down to Verukal with a width of about twenty-five miles from the sea to the interior. The other is the northern part of ancient Dakkhiadesa, now comprising largely the southern regions of the North-western Province. In the former region, Tamil inscriptions have been discovered at Padaviya, h1ka, Kanta]iy, Paaznai and MafLkai. aiva archaeological remains datable to this period are found at Kumpakaia-malai, Kandasmi-malai, horagoa, Kantaly, P4anzftai, Pta-ku and Tampalakmam. In the literary sources we get traditions pointing to the presence of Tamils at TrincomaJ.ee, Tampalak niin , KantalAy and Verukal. In the previous chapters we have seen that there is a certain amount of evidence which points to Tainil settlements at Padaviya, Moragoa, Paragiyaiya, Periyak4am and Trincomalee

1. See supra,

198
The evidence relating to this period not only points in the same direction but also indicates an extension of these

settle ents.

The area around the anciant port of Gokaa, the modern Trincomalee, seems to have had a fairly strong Tamil element in its population in this period. The Tainil settlements of this region seem to have extended from Trincomalee to Periyak$azn and Nakai in the north, Kant aily arid Pta-k4u in the south-west and possibly Verukal in the south. Kantaly, Pta-ku and P4an1ai, lying within three miles f each other, have yielded three Tamil inscriptions and some aiva remains dating back to this period. Two Taniil inscriptions of the time of Gajabhu II (1132-1153) come from Kantaly One records the setting-up of a boundary stone at a sacrificial ground (ii naanta bhmi) in the brahmadeya of Kantaly, by Laikevara, Gajabahu Dvar. The other epigraph also records the setting-up of a boundary stone by one K4ivai Apinia nip, who bore the title of Laflai Vijaya Ceaviruttar (the victorious commander of Lafk), at the request of Lafikvara Gajabhu Dvar, at Kanta1y The Tamil inscription from Pa.amai,

1. .1.1., IV, No.1397; the other inscription is unpublished and is listed as No. I 359 of the e igraphical collection in the Department of Archaeology, Ceylon. 2. Unpublished - Inscription No.1 359.

3. 3.1.1., IV, 10.1397.

199
of the time of Vijayabhu I (A.D. 1097) refers to Kantaly as Vi jayar ja-c at urvdi-niaiig4am (Kant alya Vij ayarj a-c aturvdimak4attu) and mentions the existence of a aiva temple in that place called Te Kailsam r! Vijayarja-varam This inscription was found among the debris of a ruined aiva temple. The stone pillars of tbs temple are in the style of those of the aiva temples which could be dated to the Polonnaruva period A fragmentary image of Prvat was also found among the debris Since the inscription recording donations to the Vijayarja!varam has been found at this site, the ruined temple may be identified as the Vijayarja-varam. In the twelfth century, Paamtai must have formed part of Kanta1y. This temple seems to have been built in the time of Vijayabhu I (1055-1110), for it bears the name of that monarch. It is interesting to note that it was also known as Te Kai1sam (Southern Kaila), for this name is given only to the temple of Kvaram temple at Trincomalee in the Tamil chronicles Irom the P4anhai

1. 3. aranavitana, 'A Ta ii Slab Inscription from Paavai', E.Z., IV, p.19k. 2. A.S.C.A.R. for 1933, p.18. 3. Ibid.

1 Cf.,

7:28, p.68.

200
inscription we find that the Vija,.yarja-varain was organized more or less on the same lines as any contemporary shrine in South India. The gifts to the temple sometimes took the form of cash deposits, the interests of which were used for the maintenance of various services in the temple. The institution of devadss was a feature of the organization of this temple for it is stated that seven girls were branded on their foreheads and given over to the temple as tvar-atiyr (Skt. devada&Ts). As ln . many of the South Indian temples, the endowment was placed in the trust of a a1-kkra, of the

Vikkirama-caI.m!ka-terinta Valakai division. Near P4aznai and Kantaly is the village of PVta-ku where, too, were found a stone image of Viu and the ruins of a aiva shrine. The style of the image belongs to the period between the tenth and the thirteenth century while the temple is believed to date back to the time of Vijayabhu I The foregoing evidence clearly indicates that South Indians, especially BrThxnaas, were settled in Kantaly and the surroundin villages from at least the time of VijayabThu I.

1. E.Z., IV, p.

193.

2. A.S.C.A.R. for 1 933, p.1S; C.J.Sc. (G), II, pp.156-157.

201
The Kanta].y Stone Seat inscription of

Niai"k

Malla (1187-1196)

attests to the continued existence of the BrThmaa settlement (caturveda-brahaapura) in that plaoe during the twelfth century and refers to a aiva establishment called Prvat1-satra, probably built in the time of Niaka Ma11a The Tamil chronicles, too, refer to Kanta1y as a place of importance to aivas and associate Gajabhu II with that p1mce This is confirmed not only by the Tamil inscriptions found there but also by the C1avarjisa, according to which 'Gajabhu betook himself to Gafig-taka (Kanta1y), made it his residence and dwelt there happily' and 'died during his sojolLrn there' That Gajabhu patronised non-Buddhists is implied in the C1avaisa where it is stated that 'he had fetched nobles of heretical faith from abroad and had thus filed Rjarafha with the briers (of heresy)' The toponym Kanta1y supports the conclusion that the place it represents was settled by Tamils or Tamil-speakers in the twelfth century. It is in tke Pa.amtai inscription

1. D.M.de Z.bckremasinghe, 'KantaIi Ga1-sarxa Inscription of Kitti Nissaka Malla', E.Z., II,, p.286.

2.

., 7 : 87-106; Kk., p.20; ;

Cv.,

., pp.170-178.

3. See supra, p. 19 k. Cv., 70:53, 3k.

71:1,25.

202
that we get the earliest occurrence of the name Kantaly. After that it occurs in the two inscriptions of GajabThu II mentioned above. Kantaily is the Tami].ised form of the Sinhalese name Gagata]. (Pli Gagtaka) This Tamilisation seems to have taken place during this period as a result of the Taznil settlement. The Tamil form has remained in usage to this day and the Sinhalese origin has been completely forgotten. It is interesting to note that today, when they have begun to recolonise this place, the Sinhalese call it Kantail after the Taniilised name. As early as the fifteenth century, folk etymology among the Tamila has attempted to explain the origin of the name Kantaily in a different way. The name was split into two elements, ka (eye) and t4ai (to grow), and a story was woven round it. It is said that Gajabhu II regained his lost eye-sight at this place and hence the name Kat4ai (where the eye grew) Such an explanation is typical of the folk etyqiology that one finds in the Tamil speaking areas where the origin of a large number of Sinhalese toponyms has been puzzling the new settlers.

1.

.,

, p.116; Cv.,

71:2.

2. Cf.,

7:63.

203
About four miles north of Trincomalee is }1ai where a Tamil inscription of the time of Gajabhu II has been found The inscription records the grant by Gajabhu of land to one Minta Koa, who is designated Superintendent of the Pa].anquin Bearers (tiru-p41i-civikaiyri1_kaki) . The land was granted as a jvita evidently for services rendered by the donee. On another side of the slab on which this inscription is indited, there is another Tamil epigraph, the purport of which is not quite clear It states that nbharaa paraia Tvar) sent a letter (tirumukam) approving the deed (ceya].) of Gajabhu (Gajabmu T!var) and caused a stone inscription (ci1-l!kam) to be set up. There was only one Mnbharaa who was contemporaneous with GajabThu II, namely the one who ruled in Rohaa in the middle of the twelfth century, and, therefore, the }nbharaa of our inscription must be the same person. It is not clear whether the transaction referred to in the inscription had anything to do with the grant of Gajabhu recorded on the same slab. But this seems unlikely for Inbharaa never had any authority over Rjaratha,

1. K.Kana athi Pillai, '1akai Inscription of Gajabhu II', U.C. , XX, No.1, April 1962, pp.12-1 2. Ibid.

204
where this inscription is found. It seems more likely that it concerns about one of the many transactions that took place between the two rulers during their wars with ParkramabThu I Within about three miles of Mafikai is Periyaku.am where, as we have alyeady seen, the presence of Tami]. settlers in the eleventh century is indicated by many epigraphs No Tamil inscription of the twelfth century has been found here, yet it may not be wrong to sa that there must have been Tamils in the twelfth century, too. At Trinconialee, only a fragmentary Tamul inscription of about the twelfth century has come to light The paucity of Tamil inscriptions in this place may be explained by the fact that the original temple of K5varam, where one would have normally expected to find any inscription, was completely destroyed by the Portuguese in the seventeenth century. The materials of the destroyed temple were used by the Portuguese to build a fortress at Trinconialee. Inscriptions of about the twelfth, thirteenth and sixteenth centuries have been found here on the bricks

].. Cf., Cv., 71:1-5. 2. See supra, 3. Unpublished.

205
and door-jambs or on fragments of atone in the fortress It is said that more inscriptions were found on the stones of Portuguese buildings demolished in the area in the last century A number of bronze images of iva and PrvatT have been unearthed in different parts of the present temple precincts within Fort Fredericl It has been surmised that these were buried by the temple priests at the time of the Portuguese attack in l62. Most of these have been assigned, on grounds of style, to a date between the eleventh and the thirteenth century' some of these may belong to the period of Ca occupation and some may date to the thirteenth century when a prince named Ku.akka appears to have carried out renovations to the temp1e It is not possible to say whether
many

of these belong

to the twelfth century. These finds, however, indicate that

1. H.1.Codrington, 'The Inscription at Fort Frederick, Trinconialee', 3. .A.S. (C.B.), No.80, p. 1fk8; S.Paranavitana, 'A Fragmentar Sanskrit Inscription from Trincomalee', E.Z., V, p.173; A.S.C.A. . for 1957, p.8. 2. A.Sriskantaraca, 'Tirukamalai V aralLYUla.k4', Tirukkcar Xlaya Kump!pika alar, 1 963, p.95. 3. .Balendra, 'Trincomalee Bronzes', Tamul Culture, II, No.2, April 1953, pp.176-198; A.S.C.A.R. for 1950, p.32.
14

.Balendra,

. cit., p.190.

5. A.S.C.A. . for 1950, p.32. 6. See infra, p. 327

206
around the twelfth century the temple of K!varam was a flourishing institution. The ancient port of Gokaa may have had a notable Tamil settlement in this period. The C1I].avaipsa states that there were Ker4a and V.aikkAra mercenaries dwelling at Kotha6ra in the time of ParkramabThu Kothasra is the district of Koiyram, around the port of Gokapa. We have seen earlier that five villages in this region contributed in money and in kind towards the maintenance of the temple of Rjarjvaram at Tanjore in the time of Rjarja I This fact as *l as the re-nami-ng of two districts in this region after Ca princes during the period of C 1a occupation seem to suggest that Kohasra was one of the region4ihere Ca rule was effectively felt Taniil settlements may have been established here under the Cas. The presence of Ker4a and V.aikkra mercenaries would have further strengthened the Dravidian element in the population of this area in the twelfth century. The T !a-kail5ca-irilai also refers to the presence

1. !. 7k:kk. 2. See supra, p.111 3. Ibid.

207
of Saivaa in the region of Trincomalee in the time of GajabThu II and to the antagonism between them and the Sinhalese Buddhists of the area The event described in this Tamil chronicle to illustrate this antagonism seems to take us back to a time when the Sinhalese of the Trincomalee district were being gradually ousted by or assimilated to the Tamil population. It is stated in this work that opposite the rock on which the temple of Kvaram stood the Bud hists built a temple of the Buddha and harassed the Saiva devotees who took flowers to (varam. This led to quarrels between the Buddhists and the Pcupatar (aivas). The latter triumphed over the Buddhists and pushed some of them down the rock into the sea. The matter was reported to GajabThu II, who tried to take revenge on the aivas by attempting to destroy the Cvaram temple. But through divine intervention he realised his folly, became converted to aivisxn and made generous benefactions to the temple and to the Brhmaas there Although the details of this account may not be wholly acceptable, it is not altogether untrustworthy. GajabThu is the only Sinhalese monarch who finds mention in the Takiakailca-mlai. In this chronicle he is said to have taken the

1.

E2' 7:89-96.

2. Ibid.

208
consecration name of Ciica..ka-pti (Siri-sa4ha-bodhi) and this is corroborated by the other sources He is associated with Kantaly and is credited with the patronage of Brhinaias and aivism. This is confirmed by Tainil inscriptions and is implied 2 in the Clavaisa. Perhaps the contention that GajabThu was converted to aivisxn is an exaggeration but the gist of the account cannot be doubted. Further north of Trincomalee and Kantaly, in the coastal region east and south-east of Kokkuily Lagoon, Tami]. inscriptions and aiva remains have come to light in several sites. At Padaviya and Vhalkaa, two villages in this area, were found two inscriptions of the Aiffuvar community and their associates The Vhalkaa inscription was set up to record certain steps taken by the and the Vrakkotis so that a certain town 'may not be destroyed's This town was presumably a market town somwwhere in the VThalkaa region. Padaviya, as we have pointed out earlier, also appears to have been a market town of considerable

1. The predecessor of Gajabhu had the consecration name of Salnrvan (Calmka) and, therefore, GajabThu's consecration name rust have been Siri-saga-b. Ap arently ParkramabThu I did not recognise the unconsecrated Gajabhu and also took the consecration na e of Siri-saga-b. 2. See supra, p4';aoo .

3. See supra, p..r.

k. See supra,

209
importance for a number of inscriptions of the ceis belonging 1 to the Ca period were found here. The inscription of the Aiftfiuvar found at Padaviya contains the names of some members of the AififiUuvar, Valafljiyar and the Cei communities Another short inscription from the same place records some deed by a certain Iaicc'ri Iamaiyr of the Sr! Vijayarcaa ..pa.... priamaiyr Sr! Vijayarca seems to form the first part of the name of a place or temple, the second part of which is undecipherable. SrL Vijayarja in the Pa.amai inscription of the time of Vijayabhu I occurs as the name, evidently after the same monarch,of a catur-vdi-mag4am and of a aiva temple The temple o place bearing the name of ri Vijayarca in our inscription from Padaviya was also presumably named after Vijayabhu I. Pri.amaiyr is a term met with in the contemporary South Indian inscriptions as well but its exact connotation is not known. Nilakanta Sastri takes this to stand for a body connected with the temple, the duties of which are obscure

1. See supra, p. IOC. 2. Among the names are Azanta Araka, Uttama Ca .....ce1i, .t iyparaia Valafkaiya, Tcamata Vraia. 3. .1.1., IV, No.3)409,

k, See upra, p. I7 5, K.A.Nilakanta Sastri, The C.as, p.k89.

210
ut from its occurrences in the inscriptions, p!riamaiyr appears to have stood for a local body responsible for the
S 1 - sabha of a Brhma village. As a nu ber of Siva temples of the

C5a period have been unearthed at Padaviya, it is possible that there was a Brhmaa settlement at this place in this period The pri.amaiyr of our inscription was probably one of the bodies of a sabh of such a settlement. As mentioned earlier, in this period Tamil artisans seem to have been employed at Padaviya as at Polonnaruva for the building of Buddhist structures, for Tamil mason's marks can be seen on some of the Buddhist ruins here? In three sites to the north and north-east of Padaviya, at Kumpakana-malai, Kandasmi-ma1ai and Budd.hanagehe].a,
S S

there are ruins of Saiva temples datable to this period. Kumpakaa-malai is about eight miles to the north of Padaviya. The old temple at this place has bricks with Taxnil letters inscribed on theme Almost midway between Kumpakana-malai and

1. M..R. for 1923, p.lOI. 2. See supra, p. i7 3. See supra, p. t.

k. A.S.C.A.

for 1905, p.35.

211
Padaviya is Buddhanaghe1a where exists a ruined aiva shrine in a cave with broken representations of a lifga, yoni and Ga4ea. On one of the pillars used as door-jambe of the shrine is a Sinhalese inscription of Kassapa IV (898-91k) and this evidently points to the shrine having been built with materials from a Buddiiist structure, either abandoned by the Buddhists 1 L or destroyed by the baivas. About twelve miles east of this site, on the western shore of Kokkuly Lagoon, stands another small temple 'of excellent stone-work similar to that at Polonnaruva' (iva Dv1) These aiva ruins are clear indications of Tamil settlement in the region around Kokkuly Lagoon during this period. Thus we see that the north-eastern hinterland between Trincomalee and Kokkuly had Taini]. settlements by about the twelfth century. In these settlements there is unmistakable evidence of the presence of BrThinaas and mercantile communities. In a region where there were two well-known ports at this time, namely Gkaa and Pallavavki, the presence of mercantile communities is only to be expected. But this does not explain the existence of several

Tmi1 settlements

in this region.

1. A.S.C.A.R. for 1891, p.11; E.Z., I, p.191. 2. A.S.G.A. . for 1 90 5,

pp.36-37.

212
It is not probable that all these were mercantile settlements. If we locate these settlement sites on a map, we find that most of these are near ancient reservoirs or at the mouth of rivers. Padaviya, Ioragoa, VhalkaIa, Kantaily, Ptaku, Pa.am t ai , liakai, Parazgiyav4iya, Periyakt4am and Kumpakaa-malai are all situated close to ancient irrigation works, while a place like Kandasmi-malai is at the iouth of a river. It seems possible that there was a slow infiltration of peasant settlers from South India which was responsible for at least some of these settlements. The process which culminated in the transformation of this region into a T'nil-speaking area appears to have been well wider way by about the twelfth century. The other area which has yielded considerable epigraph*c material relating to Tamil settle exits of this period is, as we have stated earlier, In the north-western part of the island and could be said to compr6.a roughly the southern districts of the North-western Province. This area stretches from the coast of Chilaw for about fifty miles into the interior, as far east as }nikdexa, and from )'iahananneriya in the north to Kurungala in the south.

213
We have already noticed in the last chapter that there were several C]a strongholds in this area which had to be first controlled before Vijayabhu could march on Polonnaruva We also surmised that there may have been Tainil settlers, especially mercenaries, in these strongholds during the Ca occupation. The Tamil inscriptions of this period seem to confirm this supposition. The Ca strongholds in this region, as given in the C!lavaisa, were Nuhunnaru (Nuvarakl), Badalatthala (Batalagoa), Vpnagara (Vnaru), Tilagulla (Talagall-la), Nahgalla (}galla or Nikavr4i), Naagalla (Nahamaagala) and Buddhagma (nikdea) Only one Tamil inscription of the Ca period was discovered in this area. This was at Eriyva, nearly- eight miles north-west of Nahainaagalla But the number of Tamil inscriptions of the tweLfth century coming from this area is These are from }iahananneriya,

Mahakirinda, Budumuttva, Pauvasnuvara and VihrThinna, which are all within a few miles of the Ca stronghol s mentioned above. In fact, the stronghold of Mahga1la is specifically referred to in one of the Budumuttva inscriptions as a place where there was a iva temple in the time of Gaabhu II

1. Cv., 58:k2-k5. 2. .1.1., IV, No.1k15.

3. S.Paranavitana, 'Two Tanill Inscriptions from Budumuttva', E.Z.,II, p . 311.

214
We also learn that the site of the inscription, the present Budumuttva, was part of Mahgalla in the twelfth century. In this epigraph Nahgalla appears in its Taniilised form of }ka]. and its other name is given as Vikkirama-calinka-puram, evidently the same as Vikkamapura of the Clavaisa which has eluded identification by scholars This new name seems to have been given after Vikramabhu I who would have bad the consecration name of Calnika (Sinh. Salmvan) The iva temple of Iahgalla was also evidently named after Vikramabhu fr it was known as Vikkirama-calmka-ivarain. Perhaps it was built in the reign of Vikramabhu. The existence of this temple points unmistakably to the presence of Tamil settlers in this area. The settlement may have originated in the time of the Ca occupation. It is of interest to note that our inscription was set up to record certain gifts to the iva temple by Cuntamlliyvr (Cuttamaliyvr 3 ), the daughter of Ku]Zttufiga I and wife of Virapperuni., a Paya prince. No remains of the temple have come to light in the area. The present inscription was found

1. Cv., 72:1k7. 2. See upr , p. j3. This is the for in which the name ap ears in the South Indian inscriptions; cf., M.E.R.for 1931/32, No.67 of 193]J32.

215
inscribed on one of the pillars of a Buddhist temple which, in the opinion of Paranavitana, was built in the Kandyan period (sixteenth to the eighteenth century) with the materials of an earlier building This earlier building was evidently the Vikkirama-calinka-ivaram, which must have been abandoned or destroyed after the aiva population of the area ceased to exist, probably as a result of assimilation to the Sinhalese Buddhist population. The present site of the inscription, which is only a mile north-west of modern Ngalla, must have formed part of the ancient Mahgalla. Another Tamil inscription, 2 dated A.D.11lS, comes from the same site. This epigraph records the settlement of a dispute between the blacksmiths and the washermen of the area over certain privileges. The dispute was inquired into and settled by the paca-pradhnis of VirabThu, the dipda of Dakkhiadesa. Among these officers were also two Taniils, !kk]. j Mcm Kaxavati and Vijayparaa. The important fact is that the settlement is recorded in Taniil. It was obviously meant for the benefit of the disputing communities. It is reasonable to assume , therefore, that the members of these

1.

III, .302.

2. Ibid., pp.305-306.

216
communities were Tamils, for it does not seem robable that the pafca-pradhnis of a Sinhalese dipda set up the record of a settlement in Tamil when the contending parties were Sinhalese. This inscription could, therefore, be taken to confirm further the testimony of the earlier record regarding the presehce of Tamil settlers in rAgalla. There is also a third Tamil inscription from BudumuttRva, but it is much weathered to admit of its being deciphered About six miles north of }galla, at I1ahakirinda, has been discovered another Tamil inscription dated A.D.113k The contents of this inscription, too, points to South Indian settlements in the area. The purport of the epigraph is to record the grant of certain lands to the Brhmatas of Jaya.koac alma-c aturvdimaig4am. This BrThmaa Settlement was presumably named after either Jayabhu I or Vikramabhu I, who bore the consecration name of Salm!van (Calmka), and is to be located in the region of Iiahakirinda. Nearly eighteen miles south of flgalla, at Pa4uvasnuvara, was found another Tamil epigraph dated in the fifth year of ia.ka }alla (A.D.1192) It records the

1. i.Z., III, p.3O2. 2. Unpublished - Inscription No.29k of the epi raphical list in the Archaeolo ica.l Department, Ceylon. 3. X.Kana athi Pilai, 'A Ta il Inscription from Pauvasnuvara' U.C.R., XVIII, I os.3e k , July-Oct., 1960, pp.157-162.

211
bull jug of a Buddhist pirivena at rI-pura by a general of Niafxka Lalla called Natimapaftcara, which name suggests that be was a Tamil. Perhaps he was in command of one of the mercenary troops stationed at rI-pura which was the ca ita). of Dakkhiadesa. Besides, Pauvasnuvara is only two miles east of Iuhunnaru, one of the Ca strongholds of the eleventh century. Two other Tamil inscriptions of this period come from Mahananneriya1 and Vihrhinna 2 which are nearly twelve miles north and twenty-four miles east of J.ahakirinda respectively. Vihrhinna is closer to two of the Ca strongholds, namely Iahamaagala and Buddhagma, which lie about ei ht miles away. The inscription from VihArThinna, as already mentioned, is a record of the Aififfuvar conimunity This shows that members of this mercantile community were active in Dakkhiadesa, too, in the twelfth century. No archaeological remains of Dravidian origin, with the possible exception of the pillars at the Budumuttva

1. Unpublished - Inscription No.1 980 of the epigra hical list in the be artment of Lrchaeolog Ceylon. Unfortunately, it has not been possible to obtain a photograph of the estam age of this inscription and hence the contents are unknown to us. 2. Unpublished - see supra, p. )3 3. Ibid.

218
te pie, have been definitely identified in this region so far. The only ancient iva temple in the area is the well-known Muvaram shrine, near Chilaw. The origins of this temple are unknown, though the Tanii]. puras trace its beginnings to hoary antiquity The Tami]. inscriptions in this temple belong to about the fourteenth and fifteenth centuries Several finds in this place, such as brass lamps, camphor-burners and a candelabrum, have been described by Ananda Cooxnaraswaxny as medieval and may well date back to this period As Tamil literature and tradition in the island have venerated this temple along with Tiruktivaram and. K!varam as a place of special sanctity dating from early times, it is possible that this temple was in existence in the twelfth century, with TanLi]. settlers around it as now. Its location close to the pQo.l banks of Chilaw suggests that it may have originated as p place of worship for pearl divers from South India. The svara-mmiyam, tkie

chronicle of this temple, gives a detailed account of the settlement of the Muvaram district with people from the

1. Cf., Musvara-mmiyam, in the Sri Vaivmpik-sata aintasvmi Tvastam Kyarccaai 1alar, Cob bo, 19&l, p.3ff. 2. Unpublished. 3. !emoirs of the Colom o Museum, Series A, No.1, Colombo, 1914,

pp. 28-29.

219
Tamil country in the Kali year 512 (2590 B.c.) by the Ca prince Kuakk 1a As we shall see in the next chapter, these traditions may reflect the events of a later period and may not go back to very early times The prince Kua1ka,who is associated with varam in the other Tamil chronicles,

seems to have lived in. the thirteenth century We cannot, therefore, be certain about the origin of the Tamil settlement around Mu&varam. The CUlavaisa contains a reference to the presence of Tainil mercenaries in Dakkhiadesa in this period. According to this reference, there was a Dami3a army stationed in the district called Raktakra in the reign of Gajabhu II. flattakara has been identified with Ratkaravva, nearly four miles north-west of Kurunga1a and close to the Ca stronghold of Vpiriagara In addition to the above epigraphic and other material indicating settlements of the Tamils in the northern regions of anciant Dakkhiadesa, there are also some place names

.lf

1. Mu

svara-mmiyam,

. vit., p.8.

2. See infra, p32l 3. See infra, p. 317

k.

Cv.,

69:6.

C.W.Nicholas, 'istorical Topogr p y of Ancient an Yedieval eylon', J. . . . (C. .), N.S., V1 1 1959, p.90.

220
in the area which point in the same direction. The major difficulty in the way of using this toponymic evidence for our purposes is one of establishing the date of their origin. Unfortunately early records of these toponyms are not available to us. However, the presence of a number of Tamil place names or Sinhalese place names indicating Tamil settlement in an area now largely occupied by Sinhalese speakers suggests that the names are not of recent origin. It is by no means justifiable to assign the origin of all these names to this period. But it may not be wrong to assume that some of them at least originated at this time. any of the Sinhalese place names with the element dem4a may have originated in this period for they occur close to the places where Tarn!]. inscriptions have been found or where the C1as had established their strongholds. Dem4a-divullva, for instance, is about three miles east of Jahananneriya and about six miles north-west of Eriyva, places where Tamil inscriptions of the elventh and twelfth centuries have been found. Dema.a-srakku.ama is about eight miles west of Mahakirinda and. udumuttva, where, too, Tamil inscriptions were discovered. Similarly, Dema.a-dora is about six miles south-west of Nuhunnaru and about eight miles south-west of Pauvasnuvara. Dema4a-nina is about ten miles south of VihrThinna and Demaussa is about six miles north of Vpinagara,

221
a Ca stronghold. The bulk of the Tainil place namea in this area are probably of later origin, possibly of the fourteenth and fifteenth centuries when, according to some Sinhalese works, several Tamils were settled in this area In the light of all the above strands of evidence, we have to loeate a number of Tamil settlements of this period in the northern regions of Dakkhiadeaa. With the evidence that we have it is not possible to determine the nature or strength of these settlements. They may have arisen partly as a result of the establishment of C 1a fortresses in this region to protect Rjaraha from the attacks of the Rohaa princes. Perhaps some of them were natural extensions of possible settlements of South Indian pearl fishers along the Chilaw coast. In the absence of any evidence to this effect, one has to be contented with mere conjectures. In the western littoral north of Dakk4adesa only one place has yielded a Tainil inscription which may be assigned to this perio This place is Virandagoa, about ten diiles south-east of Pomparippu. Even at }1.ntai, where inscriptions of the

Za

period have been found, no Tamil epigraph

of this period has come to ii ht. Since there were Tamil

1. Unpublished - Inscription

916 of the epigraphica]. list

in the Archaeological De artment, Ceylon.

222
settlements here in the period of Ca rule, it is reasonable to assume that such settlements continued to be there in the twelfth century, too. Next in importance to the two coastal regions dealt with above are the two interior regions around Anurdhapura and. Polonnaruva. It may be recollected that by virtue of the fact that Anurdhapura was the capital of the Sinhalese b '" kingdom, there weremercenary and mercantile settlements in that city in the ninth and tenth cneturies and. possibly even after that For similar reasons, there were Tamil settlements at Polonnaruva and the surrounding areas under the Cas. In this period we see that these settlements continued to exist in these places, especially in the areas around the citie8. But the evidence is certainly not sifficient to warrant the

conclusion that such settlements were numerous. Although Anurdhapura has failed to yield any Tamil inscription or aiva artefacts datable to this period, such finds have come to light at l'ioragahavela, V!ragala, }iahakanadarva and KanadayAva, which are situated close to Anurdhapura.

1. See supra, k-L

223
From Moragahavela comes a Tamil inscription of A.D.fl38 recording the gift of a piece of land at PatL1ya to a Buddhist temple by one UakakkittaD. Patlya is stated to have been received as a jvita by the donor. This may mean that the donor was an official in the service of Gajabhu II, in whose reign this inscription was set up. The occurrence of th&s Tamil inscription at !4oragahavela, therefore, may not necessarily indicate the presence of Tamil settlers there. But since this site is within about fifteen miles of nuriIura, V'!ragala and Safigili-kanadarva, where Tamil inscriptions and Saiva remains of this and earlier periods have been found it is possible that there were Tamil settlers at Noragahavela in this period. The Tamil inscription from Kanadarva is unfortunately fragmentary and only the na e of Sri Ca!Lkabodhi-(val3xuar alias Cakravatti SrI ParkramabThu Tva, who was probably the first ruler of that na e, has been decipherable It is unlikely that the ruler mentioned here is Parkramabhu II for the site of our inscription is outside his de facto realm.

1. 13.1.1., IV, No.1k06; K.Kanapathi Pil].ai, 'APil].ar Inscription from Moragahavela', U.C. ., XVIII, Jan.-April, 1960,

p.k6 ff.
2. See supra,

3.

.1.1., IV, No.]Ji07.

224
About a mile away from Kanadarva is Mahakanadarva where a bas-relief of the goddess of a ruined dv]

Cmu

was unearthed in the vicinity

Several statuettes of the 3aptamtk goddesses

were also discovered in the same village At Vragala, about

ei ht miles north-west of ahakanadarva, a unique bronze

image of iva in the Ardhanrivara form was discovered It is possible that this image was originally housed in a iva temple in that area. Though these aiva finds from Ziahakanadarva and Vragala cannot be precisely dated, they could be roughly assigned to this period on grounds of style. Their presence may be taken to indioate Tamil aiva settlements in these places near Anurdhapura. The evidence of inscriptions and archaeological remains discovered in and around Polonnaruva points to South Indian mercenary and mercantile settle ents in this region. The V!].aikkra inscription from Polonnaruva and the Clavaisa attest to the presence of V.aikkras, Ker4as and other mercenary forces in the capital in this period Probably some members of

1. A. .C.A.I?. for 2. Ibid.

1961/62, p.59. 56, p.k.

3. A.S.CA.R. for 1

k. See

supra, pq.-gJ..

225
the Valafijiyar and the Nakarattr communities were also living there, as is implied in the V.aikkra inscription An inscription of the AififftTuvar, containing only a part of their Sanakrit praasti, was found about three miles north of Polonnaruva, at Anaulundva As Polonnaruva was the capital city, it is reasonable to assume that atich mercantile communities were living in and near the city. These communities may have been responsible for the building of some of the iva temples in the area. Of the two dozen aiva and Vaiava temples to be found here in different stages of disre air, some belong to the C]a period as we have already seen. Some others belong to the thirteenth century. It is possible that a few were built in the twelfth century under the patronage of such patrons of aivism as Gajabhu II and VikramabThu I. The identification of the temples of this period is, however, rendered difficult by the problem of dating all the temples with any degree of accuracy. The same problem applie to the large number of aiva and Vaiava bronzes discovered recently

1. See sura, p. gc 2. .1.de Z.Wickremasinghe, 'Polonnaruva: An ulundva SlabInscription', .., II,

.235.

22
at Polonnaruva As a result we are not in a position to use the archaeological material confidently in the determination of the Dravidian settlements in the region of Polonnaruva. As mentioned earlier, it appears that ParkramabThu I, and possibly some of his successors, may have int&ted artisans aia& stone masons from South India to help the Sinhalese craftsmen in the building of Buddhist establishments in Polonnaruva These artisans as well as the South Indian prisoners of war employed in repairing and building Buddhist monuments must have strengthened the Tamil element in the population of Polonnaruva in the twelfth century. In addition, it appears that Tamil officials, or at least many of them, who served under the Cas when the island formed part of the South Indian empire, were retained by Vijayabhu I. This is the impression given by a statement in the Pakauva inscription of this ruler. In this copper plate inscription there is a reference to a register of Tamil clerks (Dem4a lea aru pota) maintained by a special keeper For a se arate register of the Tamil

1. C. .Godakumbura, 'Bronzes fro Polonnaruva', J. .A.S.

(c. .),

N.S., VII,
2. See supra,

pt.2,

1961, p.239

ff.

p.I.

3.S.Paranavitana, 'Paka1uva_Copper-Plate Charter of Vijayabhu 1, ., V, p.27.

227
clerks to be maintained separately there must have been several of them in the service of the Sinhalese king. It has been pointed out earlier that there is no etidence to sug est that Vijayabhu harboured any grievances against the Tamils His battles were directed against an empire that had annexed his country, but once the country was freed he appears to have treated his Tamil subjects with favour. The employment of Taniil mercenaries and clerks and the patronage extended to aiva establishments at Kantaly, as implied by the P4amtai inscription, undoubtedly attest to the tolerant policy adopted by him towards Tamils. Such a policy would have encouraged the Tamila in Polonnaruva to stay behind the Cas were defeated in 1070. Presumably

the Tamils who were in Polonnaruva during the Ca period continued to be there in the reign of Vijayabhu I. The South Indian population of Polonnaruva consisted not only of Tamils but also of Ker4as, Kannaas (Pli, Kaas) and Telugus. In. the aikkra inscription we are told that the 1aikkra mercenaries at Polonnaruva consisted of Telugus (Vaukar) and Ker4as(Ma1ay.ar) among others The Va1ajiyar and the

1. See supra,

p.7o.

2. See supra, p. 173.

228
Nakarattr were probably of Kannaa origin as these mercantile communities originated in the Kannaa country According to the Ct1lavasa, there were Ker4as and Kaas at Polonnaruva serving under GajabThu II Nidka Nalla claims in his Kantaly Gal sana inscription that his queens from Karnna and 1ellru (in the Telugu country) brought with them large retinues of elephants and cavalry Though this may be a vain boast, it is possible that the matrimonial alliances contracted by Nia.ka Ma].la led to the arrival of Kannaa and Telugu courtiers and soldiers with the princesses as was the custom in those days. Such groups, however, would not have considerably added to the Dravidian element in the city. The Kannaa or Telugu element in the Dravidian population of Polorinaruva, or 1 for that matter, of the island, does not seem to have been strong. It is notable that there is no Kannaa or Telugu inscription in the island belonging to this or any other period. At Dimbu].gala, about ten miles south-east of Polonnaruva, is a rock inscription of Sundaramandevj, the chief queen of VikramabThu I, which refers to a Dema-pTh (Pii, Dami4a-psda) in the area We have already seen that

1. See supra, p. 2. Cv., 70:230. 3. II, Ibid.,

ILt3

p.289.

k.

p.195.

229
there were a Dem4-veher at Vva1kiya in the ninth century1 and a Dam.i4a-thpa at Polonnaruva in the twelfth century Now we find that there was a Dema.U-pliu1 i Dimbu1ga1a in the time of VikraniabThu I (1111-1132). Wickramasinghe conjectures that this was probably erected in the time of Vijayabhu I It is interesting to know the exact significance of the element De or Dem4a in these names. In the case of the Daxni.a-thUpa we are specifically told in the Clavaisa that the stupa got this name by virtue of the fact that it was erected by Tamil prisoners from South Indiat In the case of the Deni4-veher it is possible that it got its name dueZa similar reason or because it was a residence for Tamil Buddhist monks. The Dema-ph may also have got its name on account of one of these reasons. This prsda was already there in the time of Vi]amabhu I who reigned twenty years before ParkramabThu I, iii whose time prisoners from South India were taken to Ceylon for the purpose of repairing and building Buddhist monuments. It is not known whether VijayabThu I also had such prisoners captured in his

1. S4e supra,p
2. See supra, p.j3.
3. E.Z.,

EI ,

II, p.l&7.
See

If.

23
wars against the Cas and had them similarly engaged. This seems unlikely, for Vijayabhu adopted a different attitude towards the Tamils of his kingdom and may not have, therefore, enslaved Ca soldiers. This prsda may not have been the work of Tamil prisoners. It not have been an exclusive

residence of Tamil Buddhist monks either, for we know from other sources that Dimbulgala was a renowned forest dwelling of this time where there were five hundred Sinhalese monks 1 in residence. Perhaps a pious group of Tami]. Buddhists from some nearby area paid their reverence to these learned monks by building them a prsda and hence the name Dema4.U-ph. It is not easy to decide between the different possibilities except by mere conjecture. The occurrence of the name Dema-phli does not, however, add to our knowledge of the Tamil settlements even if we take it to indicate the presence of Tamil Buddhist monks. At the most it may only suggest the presence of among the Tamils of the island. A Tamil inscription dated in the reign of ParkramabThu I has been discovered in the Jaumna district This is the Buddhists

1.

.G. .C., I, pt.2,

p.566. p.63 ff.

2. Llndrapala, 'The aitIvu amil Inscription of Parkramabhu 1, U.C.R., XXI, No.1, A ru 1963,

231
earliest Tainil inscription so far discovered in this district which now has the hi best concentration of Tamils in the island and once formed the main territory of the Jaffna kingdom. The record is at present in the island of Naitivu (Pli Igadipa) and is a proclamation of certain regulations concerning trading vessels wrecked off the port of rttu ,ai (Sinh. Vrtoa, Pli Skaratittha, now known in English as Kayts). The fact that the record is in Tamil may only mean that most of the traders in this region being possibly Tamils from South India their language was preferred to Sinhalese. But the manner in which the proclamation is worded shows that it was addressed to the officials at the port rather than to the traders of the wrecked vessels. The preserved portion of the record statethat the foreigners who came to the port of Vrttuai 'should be protected', that if vessels bringing elephants and horses got wrecked 'a fourth (share of the cargo) should be taken by the Treasury and the (other) three arts should be left to the owner' and that 'if vessels (laden) with (other) erchan ise get wrecked an exact half should be taken by the Treasury and (the other) exact half should be left to the owner' This may mean that the officers in this port to whom the proclamation wasaddressed

1. K.Indrapala,

. cit., p.70.

232
were Tamils. Presumably this part of the island was settled by Tamils. The toponyniic evidence of our inscription also lends support to this hy othesis. 1rttuai, which occurs in this record, is one of the earliest recorded forms of Tamilised place names available to us in the Jaffna district. It is derived from the Sinhalese name urtota by the substitution of the second element toa th its Tamil equivalent ai, a

phencnmenon commonly met with in a number of iamilised Sinhalese toponyms. This Tamilised place name also occurs in another contemporary Tamil inscription, namely the Tiruvlagu inscription of Ca Rjdhirja II This South Indian epigraph also contains two other Taniilised Sinha].ese place na es of Jaffna. These are Vallikmam (modern Valikmam = Sinh. Vliligama) and

ia l iv1 (modern Iafluvil = 5mb. Iauvil) The occurrence of

these names further supports our hypothesis that there were Tamil settlements in the Jaffna district in the twelfth century. The Tamilisation of Sinhalese place names was evidently the result of such settlements. As we shall see later, it is quite possible that the Tamil settlement of the Ja'fna district

1. V.,Venkata ubba Aiyyar, 'Tiruvlagu In cription of Pj hirja II',

L!.
2. Ibid.

XXII, p.86

233
proceeded slowly after the Ca conquest, although any attempt at large-scale settlement of people from South India does not seem to have taken place before the foundation of the Jaffna kingdom in the thirteenth century. No definite evidence regarding any significant Tamil settlement in the Batticaloa district of the Eastern Province, which is now a predominantly Taniil area, or in other parts of southern Ceylonhas so far come to light. It is possible that jhere were some Tamil settlers in the atticaloa district for 1 from the thirteenth century onwards,we get archaeological, epigraphic and literary evidence pointing to Tami]. settlements in that area The Cas had a stronghold at Chagma in this district. Not far from this place, which is now known by the Tamilised form of Skmam, is the Tirukvil Siva temple, built in the P4ya style of architecture and held to be of the same 2 date as the Siva Dvl No.1 at Polonnaruva. Although it is possible that this temple was built in the twelfth century, it seems probable that it is a construction of the thirteenth century In all probability significant Tamil settlements were

1. See infra, Pf-37-3 2. C.J.Sc. (G) , II, 3. See infra, p. 37

pp.l60-l6l.

234
not established in the Batticaloa district before the thirteenth century. As for the other areas of southern Ceylon, it is not very likely that there were Tamil settlers in this period, except perhaps some mercantile communities in the ports along the southern coast. Such communities were found in these ports in the fourteenth and fifteenth centuries A rock inscription in Sinhalese from Galap'ta refers to some Tamil slaves attached to the Galapta-vihra in the time of ParkramabThu. Unfortunately it has not been possible to,identify this monarch definitely. It has been surmised that he may be either the first or the second of that name but probably the former Galapta is near Bentota in the outhern Province and, if the inscription belongs to the time of Par.kramabhu I, it may seem that some of the South Indian prisoners of that monarch were sent to the Galapta-vihra as slaves. Such a stray instance, however, is no evidence of any !amil settlement in that area. Thus, we see that the period between 1070 and the end of the twelfth century was a time when Dravidian settlements were established slowly but steadily in the northeastern re ion and in the southern parts of the North-western Province. These two areas had a reater concentration of Tanifls

1. j., U.C.H.C., I, pt.2,

p.?68. p.198.

2. S.Paranavitana, 'Gala ta Vihra ock-Inscription', .Z., IV,

235
than. possibly any other area. The areas around Anurldhapura and Polonnaruva, where Tamil settlers were present in earlier periods, continued to be regions with scattered Tami]. settleinents. In the Jaffna district, for the first time, we get evidence of Tamil settlements in this period. The thirteenth century saw the steady grorth of these settlements and the beginninga of the transformation of the northern. and eastern parts of the island into areas permanently occu ied by Tamil speakers.

23U
CHAPTER IV

STTLEHENTS IN THE THIR.EEITH CEITURI - I : THE JAYFNA DISTRICT

The death of Ni

p i3ka Malla in 1196 marked the

end of an era of comparative security, beginning from 1070, during which the island was not plagued by foreign invasions. Internal dissensions created by rival aspirants to power and foreign aspirations for control of the affairs of the island came to a head almost immediately after the demise of Nia.ka Malla. Princes from the Ca, Pya and Kalifiga countries exploited the weakness of the Sinhalese rulers, awopped on the island at quick intervals and succeeded in holding power for short periods. To add to this chaotic 8tate, petty king-makers were active at Po].onnaruva enthroning and dethroning their Zavourites. The rapid deterioration of the political situation culminated in the onslaught of !gha in 1215, the like of which was perhaps not known earlier. The impact of Ngha's occu ation of Rjara'ha was tremendous. The Sinhalese rulers were 'ist

permanently ousted from the northern parts of the island. A oonsiderable proportion of the Sinhalese people, too, began the slow abandonment of that area. The occupation of Rjaatha by Tamil and Ker4a elements became more marked and permanent.

237
For nearly seven decades this part of the island became an arena for the contest for power among different foreign contenders, chiefly }gha and his associates, the Pyas and the Jvakas, until at last the P4yas settled the contest to their advantaEe and paved the way for the rise of a dynasty from the Tamil country in the newly founded kingdom of that region. While these events in the island made the situation favourable for the settlement of the South Indians in northern Ceylon, events in South India soon provided some of the causes for the migration of these people. The downfall of the C1as in the middle of the thirteenth century and the decline of the Pyas at the turn of the century were followed by the invasions of the Muslims. The resulting insecurity and disorders seem to have led several Tamils to migrate to Ceylon where they found a welcome hand in the South Indian dynasty that had established itself in Jaffna. The rulers of this dynasty a pear not only to have welcomed such migrants but al6o3en to the extent of inviting settlers to the new kingdom. The thirteenth and the early part of the fourteenth century were, therefore, a period of Tamil immigration for the purpose of settlement. All these were set in motion by the events that took place between 1196 and 1215, es ecially the invasion of Igha which may justifiably be called a land-mark in the history of the Tamils of Ceylon.

Lvc.

238
The first important feature of this period is the renewal of foreign invasions almost immediately after the death of NMaka Malla. Within the short span of twenty years beginning from 1196 there were at least ci ht invasions of the island, most of which were led or inspired by the Cas. These are referred to in the Somth Indian and Ceylonese inscriptions and in the literary works. In the reigns of ParkramabThu I and Niaka Malla there were Ceylonese invasions of South India and, possibly, counterLnvasions from the main1and Sinhalese troops were supplied to Pya princes in their wars against the Cas and the Ca-supported rivals. Although the Sinhalese forces won initial successes, in the end they seem to have lost to the Cas One such victory over the Sinhalese is claimed by Ku1ttui!xga iii in the ninth year of his reign (A.D.1l87) From his tenth year (1188) this monarch claims the conquest of Ceylon in his inscriptions The ruler of Ceylon in 1188 was Niaka Malla, who would then have been on the throne for hardly a year. It seems likely that there was a C1a invasion of the island at this time, if we take the vague and fragmentary

1.

f., S.Wickremasinghe, The Age of ParkraxnabThu I, thesis submitted to the Univerity of London in 1958; A.Liyanagaivage, The Decline of Polonnaruva and the ise of Dabade, thesis submitted to the University of London in 1963.

2. Ibid. 3. S.I.I., III, p.86. k.A.Butterworth and Venu opal hetty, Nellore Inscriptions, Inscription No. N 85.

239
statement in the Galpota inscription of Ni6inka Malla, namely

'Lsnk in times gone by ..... thinking it was dangerous and

annoyed' (Lakdiva pera davaa ...... napur yl dh va ..) , to refer to some foreign intervention at the beginning of his reign Kulttuga III may have taken advantage of the confusion that ensued the death of Parkramabhu I and invaded the island. That the conquest of Ceylon was not effected in 1188 is admitted in a C]a inscription of 119k, in which KulVttufxga is recorded to have ordered his troops to conquer the island in that year The claim in the inscription of 1188 is, therefore, an exaggeration based robably on a futile invasion of the island. It is not possible to think that any sccess in this direction was achieved by the Ca ruler before 1196, the year of Niaf&ka Malla's demise. Ku1ttithga's next claim of victory over Ceylon is made in his inscription of 1199 from Tirumikku.i The ruler of Ceylon in 1199 was Queen Li1vatI, whose rule was chiefly guided by the able general Kitti. Kulttu.ga(s claim in 1199 does not appear to be altogether unfounded. In a Sinhalese poem

1. 2. 3.

., II, p. ia. ; U.C.H.C., I, pt.2, p.523. .E. . for 1907, Inscription No.288 of 1907 from Tiruviaimarudr. .1., VII, p.17k ;

.1.1., III, p.205.

240
called Sasa vata, composed in the time of LilvatI, there is an allusion to three invasions from the Ca country which were Successfully checked by the general Kitti The commentary (sann!) on. this poem gives certain details of these invasions. It is stated that on two of these occasions the invading armies landed at 1vau (Nahtittha) and proceeded as far as Anurdhapura before they were defeated. On the third occasion, they proceeded from Salvat (Chilaw) as far as rIpura in Dakkhiadesa. They .2.. were all defeated by Kitti. Kitti was a general, presumably in the armies of Par.kramabhu I and Niaka Malla, who ousted C1agaga from the throne and enthroned Lilvati in ]. l97 e was ousted from power in l2OO ' The three invasions alluded to in the Sinhalese poem must have, therefore, taken place before 12 0 and probably after 1197. It is,possible, however, that one or two of the invasions took place in the period before
1197

when Kitti was probably one of the generals of Nia.ka

Ialla. The claim of the cja king in 11 9 was very probably

1. asadvata Sann!, p.5 ; J. .A.S. (C.B.), XXXI, No.82, pp.38k-385. 2. Ibid. 3. U.C.H.C., I, pt.2, p.516.
14

Ibid., p.517.

241
based on one of these invasions, Nilkiita Sastri doubts the validity of the Sinhalese account on the ground that the sann - on the Sasadavata is of a later date. 1 But even if we dismiss the details provided by the sann!, the evidence of the Sasadvata, a work contemporaneous with the alleged invasions, cannot easily be set aside. The next and the last claim of KuIttuiga is made in an inscription of 1212 from Pudukki where the conquest of Ceylon is referred to as already accomplished This claim is perhaps based on the successful invasion of the island in 1209, which appears to have been Ca-inspired. It is referred to in the Ct!lavaisa and in the contemporary Sinhalese inscriptions. In the Bpiiya inscription of queen Kalyavati (1202-1208), it is recorded that the queen had to leave i.ip on account o a Tamil invasion This Tamil invasion is also mentioned in the iip inscription of the same queen, dated in her eighth regnal years According to this

1. K.A.Nilakanta Sastri, The Cas, p. kl2, note 76. 2. Pu ukktai Inscriptions (Text), No.166 ; K.A.Ni1kuita Sastri, e CN.as, p.382. 3. D. . e Z. Wickremasinghe, 'Bpii Slab Inscription of

Kalavati',

..,

II,

.19 -192. .Z., V, p.157-158.

S.Paranavitana, 'Ki4ip Slab In cription'9

242
record, the general Iti repelled this invasion but lost his life. On the other hand, it is stated in the Cflavaisa that in 1209 the Iiandipda Anikafiga 'came at the head of a great army from the Co.a zn9LIcyt, slew the ruler in Pulatthinagara, Prince Dhanmisoka, together with the general yasmant and reigned seventeen days' The invasion referred to in the Miiip! inscription and the invasion of Anikaga have been treated as identical and rightly so The claim of conquest in the Pudukktai inscription may be, therefore, a reference to this short-lived conquest achieved by G1a troops with Anikafiga at their head. But it is also possible that it refers to the Ca invasion in the reign of Lokec. (1210-1211), alluded to in the KoV.ge inscription After 1212 neither the South Indian nor the Ceylonese sources mention any Ca invasion of the island. Apart front the Ca invasions mentioned above, there were two other invasions front South India which occurred shortly before the onslaught of Ngha. One was led by a Sinhalese aspirant Lokevara, who brought 'a great Damila army from the opposite shore, brought the whole of Laz3.kA under his sway and
.!' 80:k3-k'+.

1.

2. Cf., A.Liyanagamage,
3. C.J.Sc, (G), II, p.1 7.

. cit. ; E.Z., V, p . 159 U.

243
reigned, dwelling in Pulatthinagara, nine months' It has been surmised that this Damia army could not have been from the Ca country, for Lokevara was no particular friend of the Cas, as is shown by the subsequent Ca invasion of the island during his reign Bit it is possible that this Tamil army was only a mercenary fotice and may have come from either the Ca or the Pya country. The invasion does not necessarily postulate an allja,nce with any of the South Indian rulers. The last South Indian invasion before that of }gha is claimed to have been led by a P4ya prince Parkrania, who succeeded in capturing power and ruling for three years The quick series of invasions which began after the death of Niaka YAalla culminated in that of gha, who has been described in the chronicles as a Kliga and sometimes as a Tamil1 The identity of this ruler has remained a matter of much controversy. For the present we shall confine ourselves to the invasion and its results in so far as the Dravidian settle ents in the island are concerned.

1. Cv., 80:k7-k8. 2. E.Z., IV, p.88; U.C. .G., I, pt.2, p. 520. 3. Cv., 80:52-53. k. Cf., V., 80:58;
3:15.

244
The conquest of northern Ceylon by gha and his troops is one of the most dramatic events in the history of the island, with far-reaching results in the lives of the Sixthalese and the Tamils. For the Sinhalese this was a tragic event and its memories were preserved in fairly genuine traditions which came very early to be incorporated in the Sinhalese and P].i chronicles. For the Tamils it was an event which widely opened the doors to the occu ation and colonisation of northern and eastern Ceylon amidst the instability and the turbulence that characterised the history of the old Rjaraha in the thirteenth ceatury. At a time like this no genuine traditions of the events were preserved by them until a stable kingdom was established there. %Jhen genuine traditions failed, others, based partly on later events, were supplied to meet the needs of a later period. In the chronicles of Jaffna, these traditions centre round the personality of Vicaya K'lakai Cakkaravartti, who, as we shall later, was in all probability no bbher than gha or Vijaya 1liga Cakravartti In the chronicles of Batticaloa more genuine traditions seem to have been preserved and the invasion of gha Ok) occupies an important place in these.

1. See infra, p.

JiL

245
It is the accounts in the Sinhalese and Pli chronicles, especially those of the Clavaisa and the P1Tjva]4ya, that form the basis of the study of the conquest. Though unsatisfactory in some waye, the first familiar statements of what happened comes from them. The most important of these accounts are those of the C11laaisa and the Pjvaliya. The Pjvaliya account is of exceptional value as it was written within half a century after the invasion. The nature and value of these two accounts have formed the subject of a lengthy and. critical discussion by A.Liyanagamae Suffice it to say here that much of these accounts is devoted to denouncing the wickedness of the invaders and. bemoaning the damage dome to the Buddhist Order. Despite the bitter tone of the accounts, there is no doubt that they are based on genuine traditions as is confirmed by the archaeolth ica]. evidence and by one of the Tamil chronicles and are very valukble to our study. Hence our account of the invasion and of the subsequent occupation of northern Ceylon is to be primarily based on these sources.

1. A. Liyanagamage, 2. See infra, p.

. cit.

3'i.

24G
The CUavaida account of the invasion begins with the following trophes: 'But since in consequence of the enormously accu u].ated, various evil deeds of the dwellers in Lak, the devats who were everywhere entrusted with the protection of Laiik, failed to carry out this protection, there landed a man who held to a false creed, whose heart rejoiced in bad statesmanship, who was a forest fire for the burning down of bushes in the forest of the good, - that is of generosity and the like - who was a sun whose action closed the rows of night lotus flowers - that is the good doctrine - and a moon for destroying the grace of the groups of the day lotuses - that is of peace (a man) by name )gha, an unjust king sprung from the Klitga line, in whom reflection was fooled by his great delusion, landed as leader of four and twenty thousand warriors from the Klifiga country and conquered the island of Lafik.l In these preliminary strophes we are told of the character and lineage of gha and of the numerical strength as well as the country of origin of the army he led. By describing Ngha as a man who held to a false faith, the author informs us that he was a non-Buddhist. This is confirmed by all the literary sources, including the Tamil M4 akk4appu-mmiyam He is described here as a king of the Kliga line. This is generally re eated in the other Pli and Sinhalese works and in the Maakk4appu-nmiyam. But it is contradicted in the C1lavaisa itself in another place where be is called a Dami3a king

1. ! 80:5k-59. 2. 1m., p.53. 3. i!'' 83:15.

247
This is t0t the only instance of a state ent in the above account being contradicted elsewhere in the C'Lllavaxpaa, as we shall see resently. The identity of Ngha, as stated earlier, has formed the subject of an important controversy anon.g scholars in the recent past. The controversy centres mainly round the identification of the Kliga home of Igha, which has been variously identified as Kaliga in Eastern India, Kaliiga in the Malay archipelago and as Jaffna Pavnavitana, as we shall see later, has recently adduced evidence from some unpublished documents in support of the identification of Kalitiga in South-east Asia These documents, if their authenticity is established, should settle this problem to the satisfaction of all concerned. In the light of this new development we have to await the publication of these documents before we discuss this question further. The statement in the Glavaisa that gha's 'four and twenty thousand warriors' who are later described as fortyfour thousand strong, both incre ible numbers, came from Kali!xga is not consistently maintained throu hout the chronicle.

1. See infra, L7i 2. See infra, P:&.

248
Almost inimediatel after the strophes quoted above, there appears the following strophe: thus his great warriors oppressed the people, boasting cruelly everywhere: "We are Ker4a warriors",.. 1 This contradicts the earlier statement that they came from Kalifiga. This is not all. Again in strophe 70 of the same chapter they are referred to as 'Dand3a warriors' In the whole C11lavaisa account, ha's soldiers are described in three places as Keraas in eight places as Daniask and in one place as Ker4as and Damias (Ker4a Dam.i4a) This confusion regarding the identity of ! ha and his soldiers is not confined to the C11lavaisa alone. It is found to the same extent in the PUjva].iya as well. In fact, the accounts of 1gha's invasion and occupation in these two works are remarkably similar so mich so that one is inclined to think that one is based on the other or that both are based on a common source. As the P1jvaliya is almost

1. Cv., 80:61. 2. Ibid., 80:70.

3. Ibi ., 80:61, 76; 81:3.

il. Ibid., 80:70; 81:1k; 82:6, 26; 83:12, 1k, 2k; 87:25.
5. Ibid., 83:20.

249
contemporaneous with the occupation of Egha it is unlikely that this work is based on another sOurce, unless we take the source to be monastic records. It seems likely that the Ci11avaisa account is based on that of the P1valiy or om the records used by the author of the latter work. The iporant point, however, is that the discrepancies in the two works are identical. The Pjvaliya, too t calls gha a 1C1i!iga king (Kaligu-raja) at first and Dravia king (Dravia-raja) and Tamil king (Dem4a-raja) later on Similarly, his soldiers are called }lalala (Ker4a) at first, Dema.a (Tamil) in several other places and Dema3a and Nalala (Dem4a Nalala maha senaga) in one place That Parkramabhu II, in his campaigns against gha, fought the lala1as and the Dravias (Tamils) is maintained not only in the P1jvaliya but also in other Sinhalese works such as the Dabadei-katikvata, NiJcyasafigrahaa and the Saddharnia-ratnkaraya There are several ways of explaining this confusion in our sources. Firstly,

1. Pv.,

pp.108, 114, 116.

2. Ibid., pp.108, 116. 3. Ibid., pp.117-118; Katikvat-sa1gar, ed. D.B.Jayatilaka, p.8; Nks., p. 8; Saddhar a-r tnkaraya, p.314.

250
when the author of the Clavaisa states that gha 'landed as leader of four and twentythousand warriors from the K5J.iixga country', he may be failing to be precise in his statement rather than mRking a factual mistake. For, it is possible that what the author is saying is that }gha landed from the Kliiga country and conquered Lak with twenty-four thousand soldiers, who may have been recruited in South India. This is quite probable for even on earlier occasions princes from KV.iiga, like AnIkaga and. Lokevara, captured the throne of Polonnaruva with the help of Tamil mercenaries from South India It is nowhere recorded that there were mercenaries from the Klifiga country in the island at this time. Secondly, it may be that }gha's army consisted of Ker4a and Dami3a mercenaries who were already in Kliga in search of employment or in the service of This is not impossible for we know that in this period there were Ker4a and Karia mercenaries in the employ of
11*

not only the Tamil kings of South India but alsoin far-off places like Bengal The Nanahali plate of hadanap.la, for instance,

1. See supra, p.24Z ; U.C.H.C., I, pt.2, p. 520. 2. See supra, p.7%- ; cf., M.E. . for 1 9, No.315 of 1909.

251
includes Karas and. Choas among the mercenaries employed by the Pla riler in the twelfth century It is, therefore, possible that M.gba(s er4a and Damia mercenaries went to Ceylon from Kaliga itself. But if we are to belive the Sinhalese and Pli chronicles that the army of ?gha was very large, then it is unlikely that all the soldiers came from far-off Kali!xga and we may have to accept the first possibility, namely that }gha recruited them, or at least most of them, in South India. The confusion between Dam4as and Kera3as also could be resolved without much difficulty. It has been suggested that the 'Damia Ker4a' of the C!flavaisa, translated as Dand3as and Ker4as, should be rendered as Ker4a Dami4as, like So]! Dema.un in Sinhalese meaning C1a Tamils The implication is that like the Cas and the Pyas the Kera3as were also treated as Damias and, therefore, there is actually no discrepancy in the C].avaisa account regarding the identity of Mgha's soldiers. Though this is a plausible explantion, it appears that 1gha's army did not consist of only Kera4as but also Dami3as. Perhaps at the beginning there were many Ker4as but once }TAgha had

1. D.C.Sircar, 'Karas outside Kara', J.N.Banerjea Volume, p.2].l. 2. A.Liyanagamage, . cit.

252
established himself at Polonnaruva, more soldiers would. have been recruited from among the Tamils who were resident in the northern parts of the island. This erhaps is the reason why both the Pjvaliya and the C1flavasa refer to them as Keraaa and Dami.as. There seem to have been several South Indian leaders who were united under 1gha and led different contingents of mercenaries This may be the reason why the Hatthavanagallavihra-vaisa states that there were 'many thousands of enemy forces with their kings, the Coas, Ker4as and the like, who had destroyed the world and the ssana and were living in 2 Pulatthipura', The statement that there were forty-four thousand Ker4as and Dami3as at the time of their final dbcle as opposed to the twenty-four thousand at the tine of the invasion also seems to support the view that more Dami.as 8Welled the ranks of the invaders after their initi.al victories In this period, the invaders from Kaliga and the Malay Peninsula appear to have solved the problem of transporting soldiers from their home countries by hiring mercenaries from the near-by Tamil and Ker4a countries. Once they landed in tiae island

1. See infra, ::&, 2 Hat thavanagalla-vihra-vaisa, p 32.

3. Cf.,

gv.,

80:59; 83:20.

253
they may have enlisted further mercenaries resident there. Paranavitana's contention that } ha's soldiers were I,alays does not seem convincing It is rather difficult to accept that an army that wrought much destruction and held forth in a number of fortresses in the northern parts of the island could ha.e consisted entirely of soldiers froril far-off Jalay Peninsula. While the C1Ilavaisa specifically mentions that there were Jvakaa in the army of the Nalay invader Candrabhnu, who invaded the island in the middle of the thirteenth century, there is no mention of

Jvks

in connection with 1gha There

is little doubt that Ngha depended on South Indian mercenaries for his success. The invasion of )1gha, though in many ways similar to the earlier invasions of the island, stands out prominently in respect of the results it produced. As in the case of the earlier invasions, there was much destruction wrught in Rjaraha, especially in the capital city. The account of the Clava1sa and the Pjvaliya may be somewhat exaggerated but

1. S.Paranavitana, 'Ceylon and }alaysia in Medieval Times', J. .A.S. (C. .), N.S., VII, pt. 1,
2. Cv., 83:36, 37.

254
there is no gainsaying the fact that destruction was caused by the armies of gha. Even in the Tamil traditions this as ect of Ngha's rule has been preserved. In the Naakaappumiyam it is explicitly stated that Ik (I4gha) 'caused all the Buddhist vihras and Buddhist temples at Tppvai (Polonnaruva) to be destroyed and sought all the Buddhist 1 monks and imprisoned them'. But unlike in the time of the earlier invasions, this time there seems to have been much appropriation of land and property by the invaders. They are stated to have taken away all the possessions of rich people It is claimed that 'villages and fields, houses and gardens, slaves, cattle, buffaloes and whatever else belonged to the Sih4as he Ogha) had delivered up to the Ker4as' Even in yratha, DaQoia warriors 'dwelt as they pleased in the single villages and houses4 Thus, the soldiers of 1gha appear to have seized villages, fields and houses in Ijaraha and 1yraha. We may make an allowance for possible exaggeration but we cannot reject these statements

1. Im., p.53, 'Tppvaiyil u3a putta vikrai puttlayaik4 ellm

iippittu putta kurukk4ai elJm ti pitittu_ciaipp utti vaittu'1 2. Cv., 80:6k. 3. Ibid.,

80:76.

k. Ibid., 81:1k.

25
wholly. These allegations are made in the Pjvaliya as well and repeated in most other works Considering the fact that the northern regions of the island slipped away from the hands of the Sinhalese with the conquest of Zgha and the fact that the slow migration of the Sinhalese people from Rjaraha to the south-western parts started around this time, we cannot rule out the possibility of confiication of lands by the invaders. The migration of the Sinhalese population, or the bulk of it, from Rjaraha to the south-western region of the i].and has formed the subject of much study by scho1ars It is generally agreed that the weakness of the successors of ParkramabThu I, the incessant invasions of the island and the consequent break-down of the administrative machinery which was so vital for the upkeep of the irrigation system were among the more important causes for the abandonment of Rjaraha in the thirteenth century. While it is true that the break-down of the administrative system was greatly responsible for the abandonment of Rjaraha, one cannot underestimate the

1. Pv., pp.108-109. 2. A.Liyanaganiage, . cit. ; U.C.Ii.C., I, pt.2, pp. 713-719;

R.Murphey, 'The uin of Ancient Ceylon', Journal of Asian Stu ies, XVI, pp.l81-200.

25G
importance of the foreign invasions, especially that of Igha. This latter factr was in some ways responsible for the break-down of the administrative system as well as for the shift of Sinhalese power to the south-west. It is important to note that the severity of the rule of 1gha and the confiscation of lands by the Ker4a and Damija warriors would certainly have led to the flight of the official class, which more than any other factor is held to have been the cause of the break-down of the irrigation system and the subsequent abandonment and depopulation of Rjaaha. The fact that even after Polonnaruva was regained by the Sinhalese the seat of government was not shifted to Rjaraha shows that conditions were not quite normal in that region. The break-down of the administrative machinery was not the only reason for this. More iportant than this is the fact that the enemy had not been quite got rid of. On earlier occasions when the capital city was regained from the invaders, they were completely ousted from the island. But,in this instance,the enemy had only been driven further north. Moreover, new enemies, namely the Jvakas, ap eared on the scene and took the place of the earlier enemy. After the Jvakaa the P4ya feudatories,called the ryacakravartins, took their place. Thus, there was a succession of enemies in northern Qyl0 and neither Polonnaruva nor any other place in Rjaraha was

257
quite safe for re-occupation. This factor, as much as the breakdown of the administrative machinery, was responsible for the depopulation of Rjaraha and the failure to shift the Sinhalese seat of government there. Not only the Sinhalese but even the Taniils found Rjaraha unsafe. Whereas in the case of the Sinhalese the fertile regions of the south-west afforded new homes, the arid peninsula of Jaffna became the seat of Tamil power and provided homes for many of the new settlers from South India. Why was it that the new dynasty chose the arid tvcc&bc. in the northern region of the island which had neither irrigation works worthy of the name nor sufficient rainfall to enable easy cultivation ? Compared with that peninsula, the north-central parts of the island lying north of Polonnaruva, even after the break-down of the irrigation system, would have been a better place. These regions were never completely abandoned by either the Sinhalese or the Tamils. We shall see in the sequel that small numbers of Sinhalese and Tamils continued to live in those areas under the rule of petty chieftains called Vannis or Vaiyr, who changed alliances between the Tamil ruler of northern Ceylon and the

inhalese ruler

of the south, accordin

to the political climate of the times. We shall also see later that the occurrence of a large number of Tamil to onyms in places which had Sinhalese names in the period before the thirteenth century certainly suggests that the area was occupied by Tamils

258
and that the majority of the Sinhalese peo le of these re ions were either ousted by or, less probab]4, assimilated to the .1 Tam3.l population. Even in many areas where Sinhalese re-occupation took place and where Sinhalese live at present Taniil place names occur in considerable numbers, thereby showing that such areas were settled by Tamils at the time of the Sinhalese re-occupation. The area lying between the Jaffna kingdom and Iyraha,

generally known as the Vanni from the thirteenth century until recent times, ap ears to have formed some sort of a buffer between the warring Sinhalese and Tainil kingdoms. The rulers of the two kingdoms appear to have found it more convenient to leave this area under the rule of petty chiefs who paid nominal allegiance to either of them. The northern and eastern parts of the Vanni were in the hands of the Tamil chiefs while the southern parts bordering on the Sinhale e kingdom proper were in the hands of the Sinhalese. It is, therefore, necessary to ap reciate the significance of the Tamil occupation in the abandonment of Rjaraha by the Sinhalese. It appears that already in the eleventh and twelfth centuries the Sinhalese were being slowly pushed out of the northern regions where Tamil settlements were numerous, especially from the north-eastern

1. See infra, .4y%

259
littoral This process was, therefore, expedited by the ruthless occupation of gha and his troops in the thirteenth century. Our literary sources attribute the cause of Sinhalese migrations from Rjaraha at this time solely to the foreign occupation. We get the following statement in the Clavaa on this point: During this alien rule several virtuous people had founded on divers of the most inaccessible mountains a charming town (or) a village and dwelling here and there protected the laity and the Order 60 that they were in peace. 2 After this some of these new towns and villages are enumerated An echo of the Ct!lavaisa statement is found in the Hattbavanagalla-vihra-vaisa, where it is said that when the enemy iorces oppressed them , the ministers and such other rnportant personages and the people left their villages and their townships in thousands in search of places of protection in the rocky LI. mountains and forest strongholds. Of those who remained behind, many came under the rule of the new Vanni chieftains. The permanent dislodgement of Sinbalese power from Ijaraha, the confiscation of lands and properties by the Ker4a and Dami.a soldiers and the consequent migration of the official class and several common people to the south-west

1. See in4-a, p.L,. 2. Cv.,

81:1-2.

3. I i ., 81:3-9.
k. Hatthavanaga11a-vihra-vaisa, p. 30.

260
were among the more iportant results of the invasion and occupation of )gha. These directly helped the transformation of the northern and eastern parts of the island into areas predominantly settled by Tamils. The invasion of ?4gha may, therefore, be considered to be the most important factor that helped the establishment of more Tamil settlements in the island in the thirteenth century. After the invasion of )gha and before the rise of a dynasty from the Tamil country in northern Ceylon at the turn of the thirteenth century there were more than five foreign invasions of the island. All except one were undertaken with the help of South Indian troops, thas bringing in more South Indians to the island. The first o these was the invasion of the Jvaka ruler Candrabhnu in l2k7 This expedition of the Jvaka ruler was undertaken with Jvaka troops from his kingdom, according to the Clavaiisa and the Pjva].iya The next invasion was that of Javarma Sundara P4y4some time before 1258. This does not find mention in the Clavaisa but some inscriptions of Sundara Pya dating from 1258 claim that he exacted

1. UC. .C., I, pt.2, pp. 622-625 ; A.tLiyana amage, 2. Cv., 83:36-37 ; Pv., p. 117.

. cit.

261
1 tribute from the Ceylonese ruler. A second P4ya invasion appears to have taken place in or about 1262. This, too, is not mentioned in the C1avaisa and is known to us only from the inscriptions of Javarma; Vira P4ya I (acc. 1253) About the same time, the ,Jvaka ruler Candrabhnu led a second invasion of the Sinhalese kingdom, on this occasion with the help of 'many Dami.a soldiers, representing a great force' whom he recruited in 'the countries of the Paua and Coas and e1sewhere' It has been claimed that the second Pya invasion was undertaken to help the Sinhalese ruler combat the forces of Candrabhnut After 1263 there appears to have occurred a few minor invasions of the island under the leadership of Pfra feudatories like Ka1igaryar and Co.agagadeva. These are referred to in the Clavaisa as having taken place immediately before the accession of Bhuvanekabhu I (A.D. l272)

1. 1. .R. for 1 9k, Inscription No. 166 of 189k; K.A.Nilakanta Sastri, The Pyan King om, p. 162. 2. N.E. . for 1917, Inscription No.588 of 1916; K.A.Nilakanta Sastri, The Pyan Kingdom, p.1?6. 3. Cv., 88:62-63.

k. U.C.H.C., I, pt. 2 , p. 621. 5. .!' 90:32.

262
In Paranavitana's opinion, 'all these events appear to have happened soon after the accession of }avarma Kulakhara Pya'(A.D. l268) The next invasion was led by the P4ya feudatory named L'yaccakkaravartti (Lryacakravin) about l28k We shall discuss these invasions fully in connection with the foundation of the kingdom of Jaffna. For the present,
it is sufficient to note that these frequent invasions brought

into the island further bands of mercenaries and other soldiers, many of whom may have stayed behind and found new homes either in the new kingdom in northern Ceylon or in the Vanni chieftaincies. As on earlier occasions, these invasions added to the strength of the Tamil element in the island and were, therefore, an important factor in the establishment of Tamil settlements in tbe thirteenth century. Apart from the mercenary elements that went as invaders, there would have been migrations of mercantile communities, artisans and other peaceful settlers as well. But, unlike in
earlier times, very little information regarding such settlers is available to us. No inscription of the thirteenth century

referring to any mercantile community has come to light so far.

1. U.C.R.C., I, pt. 2, p. 685.

2. Cv., 90:kk.

263
As for monuments, only one Dravidian-style temple has survived without much damage. Several others appear to have been erected in the thirteenth century but they are almost all in ruins and are not by any means significant buildings It may be that owing to the unsettled conditions that obtained in the island in this period there was not much building activity. The chronicles give hardly any information in this respect. The Tamil chronicle, Vaiyipal, refers to the migration, around this time, of the Krnaiyar who were a mercantile community from the Telugu 2 country. This work also refers to the settlement of such artisans as Taccar (carpenters), Tatr (goldsmiths), Kar (braziers) and Kollar (blacksmiths) As we shall see later, the authenticity of these statements can be questioned. It seems possible that the author of the Vaiypal sometimes based his statements on the conditions obtaining in his time, that is to say he was just enumerating the castes of Jaffna in his time as having migrated in the thirteenth century. In the records of South India, too, there is little or no evidence regarding the migration of peaceful settlers.

1. See infra, p ?. 3 I ff.

2. Xu.

v.41

3. Ibid., m

264
One could suppose that the numerous internecine wars that characterised the decline of the Cas would have led to the flight of some of the defeated to places like Ceylon. The Muslim invasions would have certainly led to such flights, but South India was hardly affected by them in the thirteenth century. We hear very little about fimines or any other forms of distress that might have led to the migration of people. In the inscriptions of the time of Kulttuga III there are references to fRminea in two areas of the Ca kingdom. One of the inscriptions from Tiruppmburam, dated in the twenty-third year of KuJZttuhga III (A.D. 1301), refers to the distressing circumstances that prevailed in that village and to the sad incident o a and his

two daughters selling themselves to the local temple to be saved from starvation In another inscription from TaAv1r, of the year 1305, there is an allusion to similar distress being suffered by the villagers for a long time But we are not in a position to say whether such conditions were widespread in the Ca country during these declining years of the empire.

1. LE.R. for 1911, Inscription No.86 of 1911, p.7k. 2. LE.P. for 191k, Inscription o.k58 of 1913, p. 91.

26
Perhaps such conditions led to the migration of some Tamila to plames like Ceylon, but there is no clear evidence on this point. Despite the absence of any evidence, we may not be wrong in saying that migrations of small groupd of peaceful settlers from South India to Ceylon would have gone on in this period as in the earlier centuries. The South Indian invasions were, therefore, still the most predominant factor that helped to strengthen the Dravidian element in the local population. There were more than thirteen invasions from the mainland in the thirteenth century and one of them at least brought in a large contingent of mercenary forces. The chaotic conditions that prevailed in Rjaraha afforded ample opportunities for these mercenary elements to appropriate land and seize property. Under such circumstances it is doubtful whether many would have liked to return to the subcontinent. As is alleged in the Pli and Sinhalese sources, a large number of the mercenaries must have found new homes in Rjaraha The Tamil sources, however, seem to prefer to treat them as peaceful settlers who went to the island in response to invitations from the Tamil rulers

1. See supra,

p.9S

2. See infra, pp.rt

ft.

26G
It may be recollected that several writers on the history of Jaffna, basing their studies on the traditional legends found in the late Tainil chronicles, have put forward certain thorics claiming the establishment of Tmil settlements - 1 in Jaffna in the period of the Anuradhapura rulers. These theories are not accepted by serious students of istory as they are not based on trustworthy data. Nany of these have been convincingly dismissed by scholars in recent years It is, therefore, not our intention to analyse these theories and take serious notice of writings which at best culd be described as popular. In the main we shall confine ourselves to the sources on which these writings have been based. The story of the Tamil settlements in the Jalfna peninsula ha8 been told in the Tamil writings of a period at least three centuries later than the time of the events. These works are the Vaiypal and its paraphrase Vaiy, the Kailyamlai and the As pointed out in the

introduction, these works have much historical data mixed with

1. See supra, pp. .L

ff
VII, pt. 2, 19&]., pp.17Lf_22k.

2.

., S.Paranavitana, 'The Arya Kingdom in North Ceylon',

J.R.A.S. (C.B.), N.S.,

3. See supra, pp.

ic-2.o.

261
legendary material, some of which are based partly on popular etymology and partly on Sinhalese legends. The chronology is hopelessly arranged and one has to exercise great caution in using these chronicles as source materials. Their value for the period prior to the twelfth century is almost nil. Hence, we have to rely almost entirely on the more trustworthy Linhalese and Pli works and on the meagre archaeological material for any satisfactory reconstruction of what happened in the Jaffna peninsula before the thirteenth century. Although our purpose in this chapter is to deal with the Dravidian settlements of the thirteenth century, it is necessary to analyse briefly the history of the Jaffna peninsula before our period in order to clear certain common but important misconceptions. By way of this analysis we will be able to show how unfounded many of the arguments of popular writers are. We have already shown that there is no case for arguing that Jaffna was settled by Tamils in the pre-Christian 1 centuries or even in the early Christian centuries. On the contrary, there is some evidence in our sources which points to the occupation of Sinhalese in the area in the eetty ivs centuries. The meagre evidence in the Mahvaisa regarding the
OP

1. See supra,

268
Jaffna peninsula does not help us to know anything about the identity of the people who lived there in the pre-Christian centuries. The Pli chronicle informs us that the port of Jambukola (Camputtuai), on the eastern coast of the peninsula, was the main port of embarkation to Tmralipti in Eastern India from at least the time of Devnampiya Tissa (JD _LIO B.C.). The two embassies from the island to the court of Aoka embarked on their voyage from Jambukola Saghamitt arrived with the Bo-sapling at this porte The Samudda-paa-sl, commemorating the arrival of the Bo-sapling, and the Jambukolavihra were built there by Devnampiya Tissa These facts only reveal that the northernmost part of the island was under the suzerainty of the Anurdhapura king in the third century B.C. and that Buddhism had begun to spread by that time in that part of the island as in the other parts. But it is in the second century A.D. that we get some evidence regarding the people living there. The language of the gDld-plate inscription from Vallipurain, the earliest epigraphic record diecovered in the Jafmna peninsula, is the early form of Sinhalese, in which

1. Mv., 11:23.

2. Ibid., 19:23. 3. Ibid., 19:27; 20:25.

269
inscriptions of the time in other parts of the island were 1 written. This may suggest that the Sinhalese were settled in the Jaffna peninsula, or in some parts of at least, in the second century A.D. There were perhaps Tainil traders in the port of Jambukola but there is no evidence that points to Tamil settlements in the peninsula. That Jaffna was peopled by Buddhists during the first millenium A.D. is borne out by the meagre evidence of the Mahvasa and the Clavaisa as well as by the exidence of the few archaeological and epigraphic materials found in that peninsula. We find that in the second century A.D. }'ahallaka ga built the Sli-pabbata-vihra in Ngadpa 2 (modern Jaffna district). In the same century, Kaniha Tissa had a temple repaired in that area In the third century, Vohrika Tissa built walls round the Tissa-vihra in the same region Aggabodhi II (6Of-6])f) is recorded to have 'presented the Ua1omaghara temple to the Rjyatanadhtu (vihra) as well as an umbrella for the Amlacetiya'

1. 2. 3. Mi,., 3:' i. Ibid., ':3 5. x k2:62.

270
It is not possible to identify the sites of these Buddhist establishments, but they are stated to have been in gadIpa. These references in the Mahvaisa and the Clavai1sa not only show that there were Buddhists in the Jaffna peninsula in the Anurdhapura period but also indicate that it continued to be under the suzerainty of the Anurdhapura rulers. The gold plate from Vallipuram reveals that there were Buddhists in that part of the peninsula in the 1 second century A.D. At the site of this inscription the foundations of a Buddhist vihra were uncovered. These foundations are in the premises of a modern Viu temples There is little doubt that the Viu temple was the original Buddhist monument, converted inttoa Vaiava eStablishnient at a later date when Tami.ls settled in the area. Such conversion of Buddhist establishments into aiva and Vaiava temples seems to have been a common phenomenon in the peninsula after it was settled by Dravidians. In the premises of another Viu temple at Moolai were discovered some 'vestiges of ancient remains of walls' and a broken sedent Buddha image Again,

E.Z.,IV 1. - - rr 31 2. Ibid., ?2

ff

3. A.S.C.A.P. for 19k9, p. 28.

271
in the aiva temple at Nahiyappitti a Buddha image was found under a stone step in the temple tank A lime-stone Buddha image and the remains of an ancient dgba were unearthed at Ni1varai, in Navakri Among the debris were two sculptured fragments of shaped coral atneB with a stthne-railing design. According to D.T.Devendra, who conducted the excavation at this site, the dgba can be dated at least to the tenth century Near these ruins are the foundations of an ancient building and in the middle of these is a modern iva temple. It has been conjectured, and rightly so, that the old fotindations are those of the vihra attached to the ancient dba Buddha images have also been discovered in Uuvil, Kantarai and Jaffna town Kantarai has yislded very important Buddhist finds which prove the existence of an important Buddhist

1. P.E.Peiris, 'adipa and Buddhist Remains in Jaffna', J.P.A.S. (C.B.), OCVI, No.70, 1 9 1 7, p. 26. 2. A.S.C.A.R. for 195k, p. 32 ; ibid. for 1955, pp.17-19. 3. A.S.C.A.R. for 1955, p. 19.

k. Ibid. 5. S.Kumaracuvami, 'V4a _attua Cila Iappeyark4 Vara1',

in the a-vaipava-kaumuti by ILVeluppilai, Jaffna, 1918, . cii., p.25 ff.; J.P.A.S. (C.B.), XXVI,

p.1k ; P.E.Peiria,

No.70, p. 11.3, C.A.Z.P., II, pt. 2, p.96.

1 establishment in the region in early times. Such artefacts as the glazed tiles and the circular discs discovered here have helped to connect the finds with those of Anurdhapura The Sinhalese Nampota, dated in its present form to the fourteenth or fifteenth century, preserves the names of some of the places of Buddhist worship in the Jaffna peninsula. Kantarai is mentioned among these places. The others are Ngakvila (karkvil), Telipola (Tellipp4ai), Nallgama (}1allkam), Minuvagomu Vihraya (VLnazjkmm ), Taidivayina Taa-t!vu or Kayts), Ngadivayina (katIvu or Nayit!vu), Puvafgudivayina (Puz.ki4u-tIvu) and Kradivayina (KraitIvu) Of the Buddhist establishments in these places, only the vihra and dgba at Nkatvu have survived to this day. It is justifiable to assume that the Iampota list dates back to a time when the Buddhist establishments of these places were well-known centres of worship. This was probably before the thirteenth century for after this date the people of the Jaffna peninsula were mainly Saivas.

272

1. P.E.Peiris,

. ' pp.26..28.

2. Me oirs of the A.S.C., II, pp. 3. Nampota,

5, 12.

273
The foregoing evidence points to the inevitable conclusion that in the Anurdhapura period, and possibly till about the twelfth century, there were Buddhists in the Jaffna peninsula. Although it may appear reasonable to presume that these Buddhists were Sibhalese like those in the other parts of the island, some have tried to argue that they were Tamils. While it is true that there were Tamil Buddhists in South India and Ceylon before the twelfth century and possibly even later, there is evidence to show that the Buddhists who occupied the Jaffna peninsula in the Anurdhapura period were Sinhalese. We refer to the topon;niic evidence which unmistakably points to the presence of Sinhalese settlers in the peninsula before Tamils settled there. In an area of only about nine hundred square miles covered by this peninsula, there occur over a thousand Sinhalese place names which have survived in a Tamil garb. The study of these names has not yet been systematically undertaken. One serious that besets a valuable study

of these names is the absence of records of early forms. Except for a handful recorded in the Vallipuram plate, the Naintjvu inscription, the Tiruvlafg4u inscription of Rjdhirja II, the Mhvasa, C1!lavaWsa, P!jvaliya and the Nampota the early forms of the bulk of the place names are not recorded

274
anywhere. This poses a serious problem in the establishment of a reliable etymology of the toponyms before their language could be definitely identified and their evidence used for historical writing. But in spite of this, it has not been difficult to separate the Sinhalese names from the Tamil. The difficulty of finding any men{ng in Tamil for the component elements, the ease with which meanings could be found for theni in Sinhalese and the commonness of the final elements with those of the present-day Sinhalese names in the southern parts of the island are factors that help us in the identification of the Sinhalese place names of the Jaffna peninsula. The early forms of the few names that are available to us point in the same direction, for, they, too, reveal their Sinhalese origin very distinctly In the whole of the peninsula more than

1.

ttuai (Kayts) is derived from Sinh. rtota or Ertota (Pli Skaratittha). The second element toa (=port) has been substituted by the Tamil synonym a feature

common in the Tamilised Sinhalese place-names of Ceylon. The earliest occurrence of this name is in the inscriptions of the twelfth century. See K.Indrapala, 'The Naitvu Tamil Inscription of ParkramabThu I', U.C.R., XXI, No.1, Ap. 1963,p.68. Other names for which early forms are available Vali1mam,

Maffuvil, Pufkkuu-tivu, Campu-tuai, Mallkam, Tellipp4ai,

Krai-tIy u, etc. See infra, pp. 143o

275
a thousand Tamilised Sinhalese names of villages, fields and estates have been collected. As early as the beginning of the 1 century their significance was recognised. But the main difficulty of using the evidence of these names - a difficulty that sterna from the fact that early forms are not available - is that of establishing the date of their origin. Despite this, however, they help us to draw the irresistible conclusion that the Sirijialese were settled in Jaffna before the Tamils. For how else could one explain the occurrence of such a large number of Tamilised Sinhalese toponynis in that small area 2 If we are to learn from the experience of scholars who studied the place names in Britain and in the Scandinavian countries, the survival of Sinhalese elements in the local nomenclature of a region now occupied by Tamuls will indicate to us certain important points In the first place, just as in the case of English place names where Celtic elements revdal earlier Celtic occupation, the occurrence of Sinhalese elements in the place names of Jaffna shows that the area was originally occupied by Sinhalese speakers

1. S.Kumaracuvami,

. cit.,

2. Cf., The publications of the English Place-name Society.

S

27
who were responsible for giving Sinhalese names for villages, fields and estates The Y a-vaipava-inlai, the Tamil chronicle

of Jaffna, confirms this when it states that there were Sinhalese people in Jaffna at the time of the first Tamil colonisation of the area Secondly, the survival of Sinhalese elements in the local nomenclature indicates a slow and peaceful penetration of Tamils in the area rather than a violent occupation. This is in contrast with the evidence of the place names in the Northcentral Province, where Sinhalese names have been largely replaced by Tami]. naxnes The large percentage of Sinhalese element and the occurrence of Sinhalese and Tsmil conipounde in the places names of Jaffna point to a long survival of the Sinhalese population and an intimate intercourse between them and the Tamila This is also,borne out by the retention of some territorial names, like Valikmam (Sixth. VUligama) and Marcci (Nracci-rata), which points to the retention of the

]..

E.Ekwal]., 'The Celtic Element', in The Introduction to the Study of English Place-names, ed. A.Mawer and F.LStenton, I, pt. 1, Cambridge 192 4 , pp. 17

If

2.

pp.9, 24.
. cit., pp. 17,27,28 and 31.

3. See infra, p.3! 4. Cf., E.Ekwall,

277
old territorial divisions and tell strongly against wholesale extermination or displacement of the Sinhalese population We are in no position to confirm with the aid of other evidence that the penetration and settlement of Tamils in the Jaffna peninsula was peaceful and slow, though it seems implicit in the Tamil chronicles. According to these, Jaffna was settled from time to time by Tamils, invited from South India, after the establishment of a separate kingdom there That the relations between the new-comers and the Sinhalese were sometimes not too cordial is reflected in these works But there is no evidence to suggest that the Sinhalese were completely ousted from the peninsula. In fact, we are in a position to conclude from the evidence of the place names that the Sinhalese p0 ulation survived there long. In the later and more reliable sections of the a-vaipava-

mlai, dealing with the rule of the Taniil kings of Jaffna, there are references to the frequent clashes between Sinhalese and Tamils as late as the middle of the sixteenth century. The following statement

in the

is

1. Of the seven territorial divisions of Jaffna three bear the name of Valikmam, namely Va1ikmam North, East and West; two bear the name of Marcci, namely V4a-Marcci and Tealarcci. See infra, p. 31 2. See infra,
3. Cf., Yvm., pp.2k, 36, k5.

278
worthy of note in this respect: Ke (C"ilci 1i, 1519-1565) caused all the Buddhist temples that were found in several places in Jaffna to be destroyed and completely ousted the Sinhalese subjects. None of the Siiihalese subjects remained (behind after that). 1 The Vaiypa1 refers to the Sinhalese as one of the communities
living in Jaffna in the time of the Tamil kings It is perhaps

not reasonable to assume that all the Sinhalese were ousted from the peninsula by the sixteenth century. In all probability several of them were asimi1ated to the Tamil population. Some are inclined to believe that the Sinhalese element is represented in at least one of the castes of Jaffna, namely the Kviyar. No such caste exists in the social structure of the Tamils of South India and, what is more, the name Kviyar appears to be a Tamilised form of the Sinha].ese Goviya (peasant caste). This caste, whose occupation it was to serve the Ve43as (the peasant caste of the Tamils), was often referred to in the past as Kviyacciai (Kviya prisoners), which has been taken to mean Sinha].ese Goviyas who were taken prisoners by the Tamil Vefl4as

1.
2.

p.59.
., v.q- ,

3. Cf.,

fri.1

279
While it i8 possible to show to some extent at least that the Jaffna peninsula was first occupied by Sinhalese settlers, that the Tamil penetration was probably slow and peaceful and that the Sinhalese long survived there before they were either assimilated to the Tamil population or ousted from the peninsula,
it is

not easy to determine the period when the Tamil occupation

began. The late chronicles of Jaffna are unanimous in their assertion that the settlement of the Tamils in the peninsula began under the Tainil rulers of that area. The X a-vaipava-

mlai refers to the presence of Nukkuvas in Jaffna before the foundation of the kingdom but distinguishes them from the Tamils The Nukkuvas, as we shall see later, were a Ker4a caste who went to the island in the centuries after the C1a occupation The evidence of the Tamil chronicles is, however, not entirely r1iable. The ! ppa-vaipava-rulai refers to two different attempts at settling Tamils in Jaffna. The first attempt was made in the ninth century but it ended in a failure as the settlers returned to South India after some time This account is based on the Ypi legend which , as we shall show later, has no historical basis and. is a fabrication based on popular

1. Yvm.,

p.

2. See supra, p.I&V ; see infra, 3. See infra, p.

280
etymology The account of the second settleient is based. on that of the Kailyamlai and is similar to the account in the aiyThis account appears to be somewhat reliable an fits into the story that could be reconstructed from the Sinhalese sources. In the main, it places the Tamil settlement of Jaffna after the foundation of the kingdom, which event took place in the thirteenth century These accounts of the Tamil chronicles, despite their late date and their obvious errors, cannot be altogether brushed aside. As we have seen above, the Sinhalese and P2i chronicles, too, claim that the northern parts of the island were settled by Tamils in the reign of Igha during the thirteenth century The Jaffna peninsula was perhaps no exception to this. But althogh it is possible that Tamil immigrants from Soith India went to settle down in Jafmna in the time of )gha, it may be wrong to ascribe all the settlements referred to in the Tamil chronicles to the time of )gha. After all Ngha's reign was not a peaceful one. It is difficult to believe that many peaceful settlers would have gone over to the island under the

1. See infra,?p.2ff. 2. See nra, 3. See supra,

281
turbulent conditions that obtained in northern Ceylon at that time. It is true that the large number of South Indian mercenaries in gha's army would have found their homes in different parts of I?jaraba, including the Jaffna peninsula. But the kind of peaceful settlements that the Tamil chronicles mention may have taken place towards the end of the thirteenth or at the beginning of the fourteenth century, after the establishment of the iryacakravartin dynasty. The Sinhalese literary sources of about the thirteenth and fourteenth centuries generally consider the northern areas - the region north of Salgal-kadura, to be precise - as Tami]. areas In view of these considerations, therefore, it is reasonable to demarcate the thirteenth century as the lower limit for our date of the Tamil colonisation of Jaffna. The earliest evidence regarding the presence of Tamils in the Jaffna peninsula is possibly the Taniil inscription of Parkramabhu I (1153-1186) from Naitvu We have seen earlier that till about the ninth century our evidence points to minor settlements of Tamils in such important ports as Mahtittha and Goka as well as in Anurdhapura, where there was a considerable number of mercenary soldiers. In the ninth

1. See infra, p.c37 ; cf., Pv., p.111 ; Tri-sihal-kaai -sahavitti, ed. A.J.W.Narambe, (Kandy 2. See supra, p.74.

1926), p.21.

282
and tenth century some villages in Rjaraha seem to have accommodated Tamil settlers but these were by no means numerous. It seems unlikely that there were many Tajn.il settlers in the Jaffna peninsula or in any part of the island other than the major ports and the capital city before the tenth century. As we stated earlier, there were perhaps some Tamil traders in the ports of Jambukola and rtoa, in the Jaffna peninsula. But we have no evidence on this point. It is possible that after the Ca occupation of the island in 1017 there were Tamil settlers

in Jaffna.

Apparently no strongholds were established in that

region and there is no evidence pointing to the presence of Tamils. Perhaps the Cas used the port of rtoa for their commercial and naval activities. The toponym V4avar-k-pa4am perhapsof the Vajava king, i.e. Ca king) perhaps preserves the memory of some C 1a association As claimed by some writers, the place name C4i-puram may be a Tamilised form of the Sinhalese So!-pura (Ca town) and not of S4u-vra (Small Vihra) But these are matters of speculation and cannot be confirmed with the available evidence. We can only say that since there were Tamil settlers in some parts of Rjaraha in the eleventh century, some may have been in the Jaffna peninsula, too.

1. Cf., S . Rasanayagarn,

. cit., p

; V4ava is commonly used

in Tamil to refer to the Cas but it is possible that here it iifers to a Vallabha ruler. 2. Cf., S .Kumaracuvami, j. cit., p. tg

283
We may not be wrong in placing the upper limit of the date of the Tamil settlements in Jaffna in the eleventh century. As we have already seen, the occurrence of a TanLi]. inscription and of three Tamilised forms of Sinhalese toponynis in the becords of the twelfth century may point to the existence of Tamil settlements in the Jaffna peninsula in the twelfth century It is, therefore, justifiable to place the Tamil pezetration into Jaffna and the beginnings of the gradual absorption and displacement of the Sinhalese there between the eleventh and the end of the thirteenth century. For the study of the Tamil occupation of Jaffna in. the thirteenth century we have hardly any archaeological remains or epigraphic material. The absence of archaeological material may be explained in several ways. In the first place, no archaeological excavations worthy f the name have been conducted in that region, except for some preliminary diggings in places like Kantarai and Nilvarai In the second place, the litbology of the peninsula in partly responsible for the absence of early monuments. In the North-central Province and the other southern parts of the island the natural outcrops of granite rocks afforded

1. See supra, P.273 2. See supra, pp.27o-17L

284
materials for the erection of lasting monuments and for inscribing records. In contrast to this, the Jaffna peninsula, the adjoining islands and the north-western coastal strip from Kalpiiya to MullaitIvu, all lying in the Tamil areas, are covered with a layer of sedimentary limestone of the Miocene and. the later ages This limestone, with its high degree of solubility, has not ptoved to be a good medium for the expression of the arts of architecture and sculpture. It is possible that many of the structures of our period were reduced to mere rubble in the course of time and used by villagers for building their semipermanent and modest houses. Thirdly, the kings of Jaffna would not have had the necessary ecomod.c basis for the undertaking of ambitious building activities. The temples of the early period, as now, would have been limestone structures of modest proportions These buildings have been kept in constant repair, as no part of the peninsula was abandoned in the centuries after our period, and only a proper archaeological survey will help us to identify them. The secular buildings at the capital

1. U.C.H.C., I, pt. 1, p.14. 2. Cf., F.de Queyroz, The Temporal and Siritua]. Conquest of Ceylon, I, Tr. S.G.Perera, p.50 - 'They never had any other city save Nelur (Nallr) ,......... Nor is there in that place anything else worth recording save some tnk, almost devoid of water....'

285
as well as some of the important temples, according to our sources, were destroyed in the time of the Portuguese occupation and the materials used for the building of the fort at Jaffna town The discovery of stone bricks and steps with Tamil inscriptions in the Jaffna fort and in some of the old houses in Parkf Teru (Portuguese Street) confirms this We have, therefore, to depend on the evidence of the Tamil chronicles and examine the extent to whic reliable information could be gleaned from them. As mentioned earlier, the story of the Tami]. settlements in the Jaffna peninsula is found in four chronicles, namely the Vaiypal, Vaiy, Kailyamlai and the Thppavaipava-nilai. The Vaiypal and its paraphrase, the Vaiy, both datable to about the sixteenth century, purport to relate the story of the Tanhil settlements in the Jaffna peninsula as well as in other parts of the Northern Province. In these accounts, as pointed out earlier, there is much hietorica]. data mixed with legendary material and their chronthiogy is highly unreliable and faulty. Their value for our study is

1. !.!_.

pp.78-79. p.73.

2. AJuttuttampi Pillai, Yppa-carittiram, Jaffna, 1912,

It has not been possible to trace these inscriptions since.

28G
much depreciated as a result of these serious defects The Kai].yam1ai, a work datable to the early part of the seventeenth century, on the other hand, confines itself to the story of the settlement of prominent Tamil families in the different villages of Jaffna under the iryacakravartina. This account seems to have been based on traditions preserved among the important families in the kingdom at the time of the author. The Ya-vaipavam1ai, written in the eighteenth century, admittedly bases its account of the settlements on the KailyaniIaii, Vaiypal and on the two non-extant works Pararca-ckara-u]. and the - .2 Irac a-mu. The account of the kingdom of Jafmna and of the Taniil settlements there begin with the legend of the (lutist) in all the four chronicles. The Vaiypal, which is the earliest of the four works, places the foundation of the Jaffna kingdom in the Kali year 3000 (= 102 B.C.) The legend of the is presented briefly in a confused mRnner and it is only with the help of the Vaiy that any sense could be made out of it. There is no reference )re to the invitation

1. See supra, pq..cff2. Yvm., Ciappuppyiram, pp. 1-2. 3. ., vv.13-lk.

287
of Tamil settlers from South India by the The laconic and confused statements in the Vaiypal are expanded in the Vaiy!, where it is stated more clearly that the after

having obtained the tract of land known as Maarial (present Jaffna) from the king of Ceylon, invited a thoudand families from India, settled them in Naar14al and them persuaded an 1 Indian prince to rule over them. The incomprehensible statements in the Vaiypal and their clear elaboration in the Vaiy remind one of the mnemonic verses in the DTpavaisa and their expansion in the later Pli chronicles. The Kailyainlai, while mentioning the establishment of the Jaffna kingdom by Yppa (Tpi), does not refer to any settlement of Tanils in his time The Y ppa-vaipava-mlai, on the other hand, states that fl2ppta, after founding the kingdom in the ninth century, invited some Tamil families from South India and settled them in Jaffna. But this settlement did not last long as the immigrants went back to South India after the death of I ppa As we shall see later, the legend of the in these

chronicles has no historical basis and is based on popular

1. Vaiy, ed. S.Gnana ragasar, Jaffna, 1921, p.12 ff. 2. Km., p. k. 3. Yvm., p. 2k.

288
etymology which attempts to explain the origin of the name YL.ppam. Once the origin of Yppam was attributed to the legendary za or , it became necessary to

include stories of Tamil colonisation in his time in order to give substance to the tale of his founding of the kingdom A second Taxnil settlement is described in these chronicles to have taken place in the reign of the first Lya king of Jaffna, who is called K].aAkai Iriya in the Vaiypal, Vicaya IGlakai Cakkaravartti in the Vaiy, Ciz.kai Iriya in the Kailyanilai and Ciikai Ariya alias Vicaya Kakai Cakkaravartti

in the

a-vaipava-mlai We shall see later

that Cifkai Ariya and Vicaya Klaikai Cakkaravartti were probably two different personalities whom Tamil tradition has identified as one. The identification of Vicaya XTflakai Cakkaravartti withKliga Vijayabhu (Vijaya Klifiga Cakravartin) is very plausible and Gifikai iriya, may be identified with the first of the ryacakravartin rulers of Jaffna? What is important for the present is the unity of the Tamil traditions in claiming that the second Tainil co].onisation of Jafmna took place in the reign of the founder of the first proper royal

1. See infra, p.

2.

v.57; Vaiya,

p.6; Km., p.6; Yvm., p.30.

3. See infra, p.441

289
dynasty in the Jaffna kingdom. They disagree only in regard to the identity of this personality. The Vaiy calla him Vicaya K1'i Cakkaravartti while the Kailyanlai refers to ha as Cii3ki iriya. In the Y ia-vaipava-rnlai the t o names are

given to the same person. It appears that Tamil tradition resolved the disagreement among the earlier works by identifying Vicaya Kakai Cakkaravartti with Cikai riya. An examination of the accounts in our chronicles shows that the two names may refer to different personalities and that the Tamil settlements ascribed to their reigns are also those of different periods. The Vaiypa]. and the Vaiy refer to the settlements established

by the Tamils as well as by some Kannaas, Telugus and Ker4as

in the Jaffna peninsula and in the Vanni areas of northern and eastern Ceylon. The Kailyamlai and the ,a-vaipava-mlai

refer only to the settlement of certain prominent families of Jaffna. These were mainly the official class invited by the first Aryacakravartin to organize the administration of the new kingdom. The accounts of the Vaiypal and the Vaiy seem to relate to the time of Vicaya Kafikai who may be identified as Kli a ViayabThu alias I1gha. These accounts appear to preserve some memories of the events of the thirteenth century, but these are hopelessly enmeshed with traditions of later events that it is not always possible to separate the

290
earlier traditions from the later. The Vaiy'pal account rune as follows.

Ci-fdc, the

ruler of Aazkpau, in the Vanni

regions, during the time of Kt!a 1c i, sought the hand of the daughter of the king of Nadurai and with that P4ya princess want sixty Vaiyar or Vanni chieftains. These Vaiyar were asked by Cifka, to rule kakparu, presumably on his behalf. Having accepted the offer, the Vaniyars invited several people of the eighteen castes from South India to settle down in their new dominions. These colonists were invited from Naturai (Nadura), Marithkr, Tirnccilppai (Trichinopoly), Malaiyam (Ker4a), Tu.uvai-nu (north-western }ysore), Toaimaalam (To4aima4alani), Vata-kiri-ntu and Kvarpati. They went and settled in different parts of northern and eastern Ceylon. The Kaikki4ar, Cnr, Kuyavar, Valaiyar, C!ar, K'rar, Timilar, Paravar, Maikkualr, Nauvar, ?mavar, Akamp4iyr, Malaiyakam, KThnaftiyir, Kaar, Cifdca4avar, Taccar, Tar, Kar and Kollar and. those who were exceedingly compassionate '1 every caste was happily living in unity in Thlppam (Jaffna). 'The }'Ialaiyakatt (Ker4as) and the Kanr (braziers), along

with the Kaai Klika (Short KalifLga), lived in. Kaacy; the woman Telli with the intimate (friends 'i) R5viyar resided

1.

y., v. ki.

291
in the town called Pajai; the confident Cvakar (Jvks) lived in their Cri (i.e. Cvaka-cri), the AkampaiyI!, Kucavar, Kollar, Otiyar and }lukkiyar (Mukkuvas) lived in Pnakari' 'Natuvra-mauvarya and the (other) Mauvarya, who governs the beautiful land, lived in Yppam (Jaffna) along with the king' 'Villavaryar lived in Nallr; the Maapp4is, who are held in high esteem by the great, lived in ippy; the

Kavarar, KThnaiyr and the Tillai-mThryirattr lived in Varai-uu' This account of the Vaiyptal is slightly altered in the extended VaiyL In the latter it is stated that the Vaiyar sent messengers to laturai, Toaimaalam, rukt!r, Tiruccirppa.].i, KUta1r and Kra{1ckl in order to invite as many settlers as possible toom among the Vefl.ar, Pirn'ar (Brhmaas), Ce$is, Cakkiliyar, Akampais, 11alaiyakam, Timilar, Kuyavar and other such castes, both the higher and the lower, as well as the personalities called I.aficifika-nippa, Nallavku-tva,, Atti-mppa, and Karutta-vku-ci

ik-mp

1. i2

v. k5.

2. Ibid., V. 73. 3. Ibid., v. 7k. 4. Valy3,p. 26.

9Q9
Those who went to the island in response to the invitation were Atti-mppa, M4uvarya,, Ticai-vi1afiku-mauvarya, Ctuvant a-mauvarya,, Karut ta-vku, C iki -mppa, Ira-c iz.kav mppa, Iafic ika-nippa,, Nallavku-niey-tva, V!ra-ct aiya, Tit a-v!ra.c ika-nippa, Anrc apuri VTra-mauvarya, Ki4aikttava,, Nui-kttava, Ciika-vku, Yppaiyir, M1kkaiyir, Kppaiyir, Thnaicciyar, Tovvi-car, Tic ai-ve, Iac ka-v.ku-tva, , Taat tia-kirpa, Vkkia-mayit t afl,, Karut tavarya-c ifika-kumra, Nut iyi, AiMcac ifika,, Kfica-kat ta.iya, Klifika, Tillai-mvyiravar, Cuva-t i a-rya, K k i-v4aK'vri-ataitt, Mu1lai-maappa3i, Kumra-maapp4.i, Caku-mat app4i, Caruku-ma app4i , Akampa iy.r, and the BrhMaas of the Ariya-vkiam (Irya-vaisa) . They crossed the sea in boats, arrived and stayed in

Thppm in

I1ki-nu

(Ceylon)' Of these some later went to the Vanni and settled there. 'Of the four named Yppaiyir, Kppaiyir, xnaicciyar ,and Telli, the last mentioned went and ruled in Thcpa-ntu and hence the name Teuipp4ai' (for one of the villages there).

1. Vaiy, p. 27 ff. 2. Ibid., p. 30.

2 J1
'Attimpp4a, and Na1uvariya became lords (atipati, Skt. adhipati) of Iyppam (Jaffna). ViUavarya resided at Nallr. Kaaiyar-Klifikan resided at Cvukaccri (Cvakacri). Vekacalam Virutufii of the Tillai-mUvyiravar resided at Varai-nu. The ?4ukkiya (Nukkuva) named Tiruvca Vefiyaraca, became lord of Pnri (Pt!akari). The sixty Cakamar and the V?arnaycurar resided at Keruvil. The Cnr, Valaiyar, Timi].ar, Karaiyr, Paar, N4avar, .Akampai, 1'ia1aiyakam, K'viyar, Naappa3.i, Puravartayar (Portuguese, provedor), Cintu-ntr (Those of the Sindhu country), Kaikk4ar, Maavar, Paravar, Muaittvar, Kollar, Kar , Nvitar , Va4r , Tar and the Paaiyar went and lived in. the sixty-four districts of Iyppam'

1. The Kaikk4ar (weavers) are the same as the Kaikklars mentioned in the inscriptions of South India. In. the C 1a period, they also served as royal troops. Cf., K.A.Nilakanta Sastri, The Cas, p. 2. Vaiy, p. 30 ff.
1+57.

294
The accounts in the Vaiy!pal and the Vaiy are thus basically identical, but the Vaiy provides more details regarding some of the important colonists. It appears that the author of the latter work obtained this supplementary information from other traditions prevalent in his time. The problem now is to examine the extent to which these accounts can be relied upon for the history of our period. There is no doubt that these are not wholly acceptable as information relating to the thirteenth century. We shall see later that the story of
ka

or Vara-

rca-cifik is an unfounded myth based on the Vijaya legend It has, therefore, no relevance to the story of the Tamil settlement in Jaffna. The contention that Tamil settlers were invited to the island by Vanni chieftains in the time of Vicaya Kulet!ikai may not be wrong. It is in the time of }gha that we hear for the first time about the rule of Varmi chieftains from the Pli and Sinhalese sources We have also noticed that the settlement of Tamils and Ker4as in a number of villages was actively pursued in the time of }1gha, according to the P'jvaJ4y and the Ctflavqsa. It is possible that }gha himself was not directly responsible for this but that his subordinates, probably

1. See infra, p.
2. .,g:il

L1D7

; Pv.,p.1l

295
including the Vanni chiefs, who were in control of different parts of northern Ceylon pursued a policy of settling colonists front South India in the newly acquired dontinions. There may, therefore, be some truth in what the above accounts have to say. Some of the communities mentioned in the list of colonists were already in Ceylon in the twelfth century. These are the Malaiyakam (Ker4as), Kaar (Kar as) and Akampat iyr They had gone to the island mRinly as mercenaries. The Ker4as are also mentioned in the Clava1sa and the Pjvaliya among the soldiers of }Igha Similarl; the reference to the Jvaka settlement in Cvakacc!ri seems to have been based on reliable traditions. As we know, it was in the thirteenth century that the Jvakas under Candrabhnu occupied the northern regions pf the island and possibly settled in places like Cvakaccri which preserve (Jvaka-cri ),Aand C q Afi.k al

(Jvk-kffai),

their memory in their names It is possible, however, that the reference to te Jvaka settlement is based on the place name and not on any genuine tradition, but this seems unlikely.

1. See supra, p. 2. See supra, p 14r 3. S.Paranavitana, 'The Lya Kingdom of North Ceylon', p. 19k ; see infra,
pr.CJ1 . . I

. cit.,

S . .

29
The Naavaryar (variant: Mauvaryar) or Maava chieftains are mentioned among the more important colonists. It is very probable that certain Naava chieftains were among those who led the mercenary forces of }gha. The Naavar (variants: M4apar, Majepar) were chiefs of certain hill-tribes in the Kara and Taini]. areas of South India. Their warlike habits led to their employment as mercenaries in the armies of the South Indian rulers. In fact, their recruitment for such employment is specially recommended in the Sanskrit work E.mandak!ya In the latter part of the twelfth century and

in the

thirteenth century

the Maavars of the Tamil country became prominent as feudatories of the P4ya rulers and played a leading role

in their

wars.

Many Maavaryars find mention in the Pya records of the thirteenth century It is, therefore, probable that some of these warlike chiefs provided mercenaries for ?igha and accompanied him to Ceylon. After the conquest of Rjarattha they may have been given certain villages in Jaffna and

in other

parts of northern Ceylon. But it is also possible that they went to the island with the Pya armies that invaded Ceylon

1. Cf., J.D.LDerrett, The Hoysalas, Madras, 1 957, 2. LE.R. for and 573 of

pp.7-9.

1926, Nos.k3, 50, 178, 180, 181, 536, 557,

1926.

297
in the latter hail of the thirteenth century. The author of the Vaiypal may have confused the 4 lrazt4tions relating to these later events with those of the time of gha. There are certain other statements in the above accounts which are totally unfounded. The mention of C!ar (Chinese) among the settlers of Jaffna is surprising. This is improbable unless some Malaya who went with Cangrabhnu were mistaken for Chinese and some traditions regarding them had survived. It is not likely that there were Chinese soldiers among the forces of Candrabhnu. Probably this reference is the result of sheer imagination on the part of the author. Similarly, certain other statements are based on popular etymology. This is betrayed in the derivation of Tellipp4ai from a woman named Telli who had settled in that village. Tellipp4ai is actually a Tamilised form of the Sinhalese name Telipola. This Sinhalese form occurs in the Nampota Pa3ai is clearly a Sinhalese element (poa = market place) which occurs commonly in the Tamilised place names The list of castes seems to have been based on the social conditions obtaining in the Jaffna peninsula in the time of the authors of the Vaiypal and the VaiyL it is possible

1. Nampota,

p.5.

2. Cf., Gam-poa = Taniil, Kamp4ai.

29
that all these castes were represented in the Tamil population of Jaffna in the thirteenth century, but it is unlikely that genuine traditions about al]. of them were preserved. The author of the Vaiy, when expanding the list in the Vaiypal, was obviously not depending on traditions. This follows fro g his inclusion of the Puravartayar. Puravartayar is a term derived from the Portu&uese provedor, meaning supervisor Such names as IZkkaiyir (The (Long) Nosed), Thnaicciyar (The Dumb), Yppaiyir (t;-e J ' ) and Eppaiyir

are obviously not names of communities. While the Vaiypal mentions the MaappaflJ.s without referring to their various divisions, the Vaiy elaborates this by listing the different sections, namely the Mullai-, Kumra-, Cahku-,and Caruku- Naapp4is. The Maappagis, as Gnanapragasar has pointed out, appear to have been people who went to Jaffna from Matappa3i in the Kaliga countx7 From the ppa-vaipava-niLai we know that they were members

of the royal family of Jaffna They may have gone originally with Jgha. it is unlikely that at the time of their arrival

1. S.Gnanapragasar, 2. Ibid., p. 3. p.

pa-vaipava-vimarcan, p.kk, note.

lk8
.

____

299
in Jaffna they were divided into several groups. As Gnanapragasar is inclined to think, such a division must have occurred at a later date, for it is hard to aasume that a group of people who were called h4appa4is because they came from a place named Naappafl.i would have been divided into several sub-sections even before their arrival The Vaiypa1 seems, therefore, to preserve a genuine tradition when it refers to the Naappais without mentioning their sub-sections. The author of the Vaiy, on the other hand, has expanded the original version on the basis of the conditions prevalent in his time. It is likely that some of the other castes, too, have been similarly included

in the

VaiyL Some of the prominent colonists mentioned in the

Vaiy may be fictitious personalities. But such personalities as the Nauvaryars and KlifLkas were probably leaders of mercenary forces under }gha. It is possible that persons like NaUa-vku-tva and Kautta-vku-tva, were also true persons for there are some place names in Jaffna which are possessive names with the element vku-t!v. Examples of such names are Vku-tva-c!m (in VTniaikniam), Ceya-vku-tva-c!m (in Teilippa3ai) and Vicaya-vku-tva-cTh (in Nallkam)

1. S.Gnanapragasar, !a-va ipava-vimaraan, p.1kB.

2. S.Kumaracuvami, 2' .2i.t P !

c4

300
These were evidently named after some of the leading occupants of those areas at the time of the Tamil settlements and may well go back to the thirteenth century. The interesting fact about these toponyma is that their final element, namely cima is a word in Sinhalese (s!m) denoting boundary. This may suggest that these names came into existence at a time when Tamils began to settle in the midst of Sinhalese people. On the other hand, sni is also found in Malaylam. If the element cTm in our place names is derived from the Malaylam word, then it may indicate that there were Kera3as among those who settled in the Jaffna peninsula. In fact, there are other place names which clearly suggest this, Toponyms such as }alaiyaka-kaavai (in Pulli, Malaiya-pyi i and Malaiyaa-v4avu (in Acoeu), (in Accuvli), Malaiya-ollai in

(in Uuvil) and Nalaiyakaa-v4avu (in rvli) preserve the memory of the Keraa aettlement The settlements described in the Kailyamlai and the Y a-vaipava-mlai are probably those of the time

o the first Aryacakravartin,as stated in the former work. The account in these two works is the same, except for minor variations

in names

and certain other details, and is confined

1. S.Kuxnaracuvami, 2 cit., p.333.

301
to only the settlements of certain leading families in the new kingdom. The ,a-vaipava-mlai follows the tradition in

the Vaiypal and the Vaiy when it says that the Tamil colonists were invited from the Tamil countries by the ruler of Jaffna. The Kailyanilai is silent on this score. The rest of the account runs as 1 llowe. Puvakavku (BhuvanekabThu), the chief minister who came from ?4aturai (Nadura), was made to reside at Nallr, the capital. Pti-ma1avaj, of the Pak!rati-kulani (Bhag!ratha kula ), from Popariyr, his brother, his brotherin-law Cepaka-maava and the latter's brother were settled in Tirunelvli. According to the Ya-vaipava-mlai, Pinzalvan also took with him five more families. Nara-cika-vkutvaa, the Tu.uva at Mayilii. The who came from KviriyI1r, was settled a-vaipava-mlai adds that he was the

eldest son of Pravalti-tva. Capaka ppa from Vli-nakar (Yvm., Vvi-nakar), his relative Cantira-cra-mppa and another Kaakarya were settled at Tellipp4ai. Pryirava (Yvm., Pryiram-u t aiy settled in Iuvil. The Y ) from Kvapati (Yvm., Kva]i1r) was a-vaipava-rlai adds that since

this village was found to be unsatisfactory he moved to a village further west. lakaa, a ve from Kaccllr, and his four

brothers were settled in Paccilaipp4ai. Kaaka-maava from Cikari-mnakar (Yvm., Cikara-ninakar) and his four brothers were settled in Pulli. KUpakrntira (Yvm., K'pakryntira)

from Kpakanu and Pu.ya-inakIpla-ppa (Yvm., were settled in Tolpuram. Tvarcntira, from Puflr was settled in Kvilkkai. u-koa-mutali from Totai-nu

(Yvm., Toai-niaalam) was settled in Iruplai. Iru-kulaniunit uyya-t ainyaka (Yvm., Iru-niarapum-tuyya) from Cyr (Yvm., Ceyyr) was settled in the island of Netuntivu (Delft). Pa].lava, and two other chiefs from Vafici were settled in Vei-ntu (Yvm., Veli-natu alias Pallavaraya-kattu). The variations in the lpaa-vaipava-m1ai are not too significant and may be due to either a different version of the Kailyamlai used by the author or the author's own corrections on the basis of boa other sources. There is no doubt that the a-vaipava-mlai account is based on that

of the Kailyanilai. This is admitted by the author in the prefatory verse It is difficult to say how far this account is reliable. As this account, unlike those of the Vaiypal and the Vaiy, relates only to certain important families, there i every possibility of it having been based on genuine traditions and genealogies maintained in those families. Even now there are a number of families which claim descent from

1. Km., 11., lk9-l99; Yvm., pp. 27-29. 2. Tim., p.1

30J
one or other of these early colonists It is, therefore, quite possible that the author of the Kailyani]. was bang his account on reliable traditions. As mentioned before, the l44avar chieftains had attained prominence in the Pya country in the twelfth and thirteenth centuries as feudatories of the Pyas. As the first ryacakravartin came from the Pya country, it is possible that he took with him or invited some liaavars to be his administrators. According to our account, a personage called aava was sebtled in Tirunelvli. In this village

there is still an estate called P i-ma1avarya-va4avu This may mean that was one of the early Tamil

colonists in that village and may confirm the statement in the KaiIyamlai. It is not impossible, however, that the author of this work was depending on such place names for some of his statements. This seems unlikely. Another place name with the personal element Naavariya,, namely Maavarya-kuricci, occurs in Vaa-marcci A family in this place claims descent from one Kaaka Ma1ava, who is said to have settled there in the time of the first ryacakravartin Kaaka }Iaava is not mentioned

1. K.Velu pillai, 2. S.Kumaracuvami, 3.

14

, cit.,

p 2o2

. cit., p. oP

K.Velup illai, .2 cit., p.

301
in our sources. In view of such traditions, it may be reasonable to hold that some at least of the colonists mentioned in the Kailyaiu].ai are true personalities. It appears that Puvankavku,who is referred to in our sources as a minister of the first Lyacakravartin, is a later personage. He has been identified with Prince Sapumal Kurnray who conquered Jaffna in the midd].e of the fifteenth century Some later traditions seem to have been confused with earlier ones in our chronicles. Some of the persons mentioned in these accounts may very well be later colonists. The foregoing account of the Tamil chronicles seem to contain some historical information in spite of their obvious errors. We may be justified in placing some reliance on their general story, There is hardly any epigraphic or archaeological evidence to confirm o supplement the above account. The only information outside the Tamil chronicles about the Tamil occupation of Jaffna in the thirteenth century comes from the P1jAvaliya and the Clavaisa . This relates to the Keraja and Dami.a garrisons maintained by gha and his associate JayabThu in V1ikg.ma (Valikmam) and Skaratittha (rtota ) The tenance of garrisons in these two places, -'-

in addition

to the

1. Cf., S.Paranavitana, 'The irya Kingdom of North Ceylon', p. 193 ; see infra, p..5.2

. cit.,

2. Cv., 83:17 ;

p. )I

3O
many outside the pmninaula, shows that Jaffna had been subdued Mgha. Several of his Ker4a and Dami.a mercenaries were evidently given lands in the peninsula. The reference in the Tamil chronicles to the Nalaiyakas or Ker4as may be based on traditions regarding these Ker4a soldiers of Ngha. The subsequent P4ya invasions would have added to the Tamil element in the population of the peninsula. The thirteenth century appears, therefore, to have witnessed a marked increase in the occupation of the Jaffna peninsula by Dravidian settlers, chiefly Tamils. Much of the t traditions recorded in the Tamil chronicles may date back to the thirteenth century. The confused character of these sources and the absence of other evidence prevent us from getting a better picture of the settlements established in the thirteenth century. The evidence of toponyms, however, suggest that the character of the Dravidian settlement of the Jaffna peninsula was different from that of the major part of the Vanni districts. The settlement in the peninsula appears to have been more peaceful and slower. The vilence that characterised the occupation of Polonnaruva and the surrounding regions by }1gha' s forces appears to have been absent in the Jaffna peninsula when Tamila and Kera.as occupied it.

3OGo.

Province, the chieftaincies of the Batticaloa and Trincomalee

Please read 306 before 306a.

The archaeology and history of the Vanni is still an unexplored field, although the jungles of that region are fast vanishing in the face of government-sponsored colonisation schemes. Much of our knowledge is confined to a few writings of
some

British civil servants who evinced a keen interest in the

archaeology and history of this region. Among these, the 1 writings of H.Parker and J.P.Lewis deserve special mention.

1. Henry Parker, 'Irrigation in the Northern Province', Papers Laid Before the Legislative Council of Ceylon, NoI, 1886, pp.105-116; J.P.Lewis, Manual of the Vanni Districts; 'The Archaeology of the Vanni', J.R.A.S. (C.B.), No.'45, 1691. Anonymous, 'Historical Sketch of the Vanni', The Monthly Literar egister and Notes and Queries for Ceylon, I, No.1, Jan. 1893,

pp. 1-7; Feb. 1893, pp.25-30

3O
CHAPTER V SEiTLE!4ENTS IN THE THI1EE1TB CE1IT1JR! : VAMNI DISTRICTS

The borders of the Jaffna kingdom proper and the Vanni chieftaincies that owed allegiance to it cannot be ascertained with any degree of certainty. The peninsula of Jaffna, including the neighbouring islands, was undoubtedly under the direct ule of the kings of Jaffna as we know from the Tamil chronicles Beyond the peninsula there appear to have been some parts, especially in the present )1annr district, which came under the direct rule of the kings of Jaffna. But the rest of the present Northern and Eastern Provinces as well as some of the northern parts of the North-central Province were in the hands of chieftains often loosely referred to as the Vannis or Vaniyar. The area that came under their rule was also referred to as the Vanni. The extent of the Vanrii lands has varied from time to time. In the Sinhalese chronicles of the thirteenth and forteenth centuries the depopulated jungle area that separated the Sinhalese kingdom from the Tamil kingdom was generally referred to as the Vanni. In the chronicles of Jaffna the name was mainly used to describe the chieftaincies of the Northern Province. In the chronicles of the eastern

1. See infra,

p-k,

307
In l9kl Geiger published an interesting article on the Vanni based almost entirely on the Pli chronicle In recent years was published another work which sadly lacks a proper scientific analysia There is much difference of opinion among all these writers on the important problem of the origin and spread of the Vannis or Vaiyars. Who were the Vannis who emerge into limelight in the thirteenth century amidst the confusion that followed gha'a invasion ? This is a question which i not easy of solution with the evidence at our disposal. The derivation of the name itself presents lot of difficulty. Paranavitana has the following to Bay about the Vannis:The government of the districts away from the capital was carried on by a class of chieftains referred to as Varini who someties defied the authority of the ruler at the capital. The people who lived in the ancient Rjaraha , which in our period (thirteenth to the fifteenth century) was being steadily encroached by forests, were under chieftains called vanni, some of whom were of Tamil race, and who transferred their allegiance to the Sinhalese king, or the ruler in. Jaffna, as the exigencies of the changing political situation dictated.

1. W.Geiger, 'Die Vannis', Sitzungsberichte der Bayerischen Aka ernie dci' Wissenachaften, II, Heft 1f, Juni

l9kl, 1'lUnchen, pp.3-li1

2. C.S.Navaratnam, Vanni and the Vanniyars, Jaffna, 1960.

308
The word vanni is generally derived from Skt. or P. vana, 'forest', and 18 taken to have been borne by these chieftains because they ruled tracts of territory mostly in forest. The number of vanni and their territories is sometimes given as eighteen, and sometimes as three hundred and sixty-four. Two classes of vannis are also mentioned, namely maha-vanni 'great vannis'and siri-vanni, 'saaller vanni' , Perhaps the eighteen were the maha-vanni and the three hundred and sixty-four the siri-vanni. ]. According to this, the Vannis were only a class of chieftains who derived their name from vana because of the nature of the tracts that came under their authority. While agreeing with the derivation of the name, Geiger has a different opinion to express on the identity of the Vannis:Der Name der Vannis (mod. Sgh. vanniy , Pali vanni oder Va!i!la ) jet in seiner Bildung nicht vollig kiar, abel' es ist kaum zu bezweifeln, da,er mit vana ,,Wald" zusammenhtngt. Wir konnen ihn passend mit ,,Waldleute" oder ,,Waldsiedler" wiedergeben. Weiterhin jet es sehr bemerkenswert, dadas Wort vanni oder vafla niemals allein vorkommt, sondern immer in Verbindungen wie vanni-rajnO und dergleichen, an 3_Stellen (83.10; 87.52; 90.33) uberdies mit dem Zusatz sih4a. Es ergabe aber em schifes Bild, woilte man das mit ,,Vannikonige" ubersetzen und nur auf die Anfuhrer und Hauptlinge der Vanuis beziehen. Nein, es war das Name der Gesamtheit. Das Wort rjan hat in Ceylon eine allgenieinere Bedeutung angenommen, die den Sk. katriya entspricht. Die vannirajno beanepruchen also, ein adeliger Clan zu scm, genau so wie der in Vesli herrschende Adeleclan der Liccbavi in eingha1esichen uel1en (vgl. z. B. Saddharmaratnva1iya, ed. D.B.Jayatilaka, p.298) ala licchavirajjuruv, wtl. ,,Licchavikonige"
bezeichnt wird. Wenn sich aber die Vanuis aus drucklich selber shaa nennen, so stellen sic sich dainit ale

1. UC.H.C., I, pt.2, pp. 736-737.

309
Arier in bewu ten Gegensatz zu den Darni3as (3k. dravia) wie zu den Vddaa. Wir sehen also, da(5 schom im 13 Jahrhundert die Vannis ebenso, wie dies ihre heutigen Nachfahren tun, den Anepruch auf arieche Abkunft und vornehme Kaste erhoben, und davon dem Chronisten der zu.. Anfang des 1k Jahrhunderts sein Werk Verfate, also ale Zeitgenosse gelten darf, dieser Anepruch of fenbar ale durchaus berechtigt anerkannt wurde. 1 We shall presently,(that while Geiger is partly right in applying the name Vanni to a whole community or caste rather than to a group of chieftains, he is wrong in claiming that they were all Sinhalese and consequently of Aryan descent. But before we come to that, let us consider the various derivations that have been suggested for the name Vanni. Tennent mentions two possible derivations, namely 'one significant of the forest (vanam)zhich it (the Vanni region) covers to a great extent, the other of the intense heat which characterisee the region' (vanni = fire Some have tried to derive it from the TaIL1 val, 'hard', denoting the hardness of the soil Still others have suggested a derivation from Baniy or merchant These are all fanciful derivations based on the similarity of their sounds with that of vanni.

1. W.Geiger, 'Die Vannis', 2.

. cit., pp.k-5.

.Tennent, Ceylon, II, (ktb edition),

p.508.

3. J.LA.S. (C.B.), )I" i. Ibid.

, No. 115, 1 89k , p. 151, note.

310
The derivation from vana appears to be plausible but unusual. The Pli form vaflfa does not seem to have been derived from vana. No tradition has been preserved in Ceylon regarding the derivation or origin of the name, but in South India where, too, we hear of Vannis or Vaiyar in this period and later, there are certain traditions regarding their origin which throw some light on our problem. The Tainil work entitled Cilai-eupatu, probably composed in the period of the Vijayaxiagara empire though ascribed to Kainpa, who lived in the Ca period, is a panegyric on the Vaiyars According to this work, the Vaiyars belonged to the Agni-kula and were descended from a certain Sambhu-niuni. Gnanapragasar is inclined to think that this association with the Agni-kula is a theory borm of the similarity between vahni (=fire) and vanni In fact, there is a le end among the Vaiyar caste of North Arcot which illustrates the derivation of their name from valini. H.F.Cox has recorded this legend in the following manner:In the olden times two giants named Vata i and 1ahi worshipped Brahma with such devotion that they obtained froi him immunity from death from every cause save fire, which element tbe bad carelessly omitted to

1. S.GnanaDragaaar, 2. Ibid.

a-vaipava-vimarcana, p. kO.

311
include in their enumeration. Protected thus they harried the country, and Vatapi went to the length of swallowing Vayu, the god of the winds, while Mahi devoured the sun. The earth was therefore enveloped in perpetual darkness and stillness, a condition of affairs which struck terror into the minds of the devatas and led them to ap eal to Brahma. He, recollecting the omission made by the giants, directed his supplicants to desire the rishi Jambava Munlmuni to perform a yagam or sacrifice by fire. The order having been obeyed, armed horeemen sprung from the flames, who undertook twelve expeditions a ainst Vatapi and Mahi, whom they first destroyed and afterwards released Vayu and the sun from their bodies. Their leader then assumed the government of the country under the name of Rudra Vanniya Maharaja, who had five sons, the ancestors of the Vanniya caste. 1 This is one of the many Vtpi legends current in South India and has no special historical significance. Perhaps it may be preserving some memory of their origin as a warrior caste. But its importance lies in the fact that it is meant to illustrate their origin from fire and the derivation of their name from vahni Thus we find in the literature and tradition of South India the origin of the Vaniyar being associated with fire or the Agni-kula. The derivation of their name from vahni, therefore, seems to be plausible but not very convincing. As Gnanapragasar has suggested, this association may represent a later attempt to derive the name from vahni Even if we allow the association

1. H.F.Cox, A Manual of Iorth Arcot, I, (Revised by H.A.Stuart, }iadras, 2. Ibid. 3. S.Gnanapragasar, a-vaipava-vimarcaam, p. kO.

1895), p.236.

312
with the Agni-kula as plausible, it is difficult to explain why their name was derived from a rarer word like vahni instead of agni. Vanni being a caste name in modern India, the early occupation of the Vaiyar may provide a clue to the origin of the name, for almost all caste names are based on the occupations followed by the different castes. The modern Vaiyar caste of South India follows the profession of cultivation like the Ve.as The Vaiyars of the Vijayagagara period, too, seem to have been engaged in the same occupation, for they appear in inscriptions of the time as tenants of BrThmaa and Ve33a landlords and paid a special tax called the in iya-vari But in the earlier centuries they appear to have been warriors. The Cilai-eupatu praises their skill in the art of archery and gives the bow as their emblem The Kallam refers to them as aai-vaflLiyar (Vaiyars of the four-fold army) which

shows that they were also warriors employed in the four-fold army of the states The evidence in the KalIam agrees with the attributes showered on them in the Cilai-eupatu. It appears, therefore, that in times past the
Vaiyars

were a community

1. H.F.Cox,

. cit., p. 236.

2. I. .R. for 1913, Inscriptions o. 223 of 1912 and Nos. 30 and 3F of 1913.
3. S
pv&asar v.7 ,

k. Kalltam,

ippa-vaipava-vimarc auam p. ko. .t1.Aics p 3oJJ (4 Yv.0

313
of warriors or tribesmen who were noted for their skill in archery and employed as soldiers in the armies of chiefs and kings. Gradually they must have begun to lead a settled life and taken to agriculture. Traditions relating to the Vijayanagara period refer to them as a 'forest race, a tribe of low cultivators' They may have lived originally in the forest regions. If their
C.,nP1QC-4-iofl

name has anytiag t-e---do-- with their original habitat, then it

may be derived from Skt. vanya (='wild, savage or existing in the forest' 2 ). Vanya becomes Tamil, in Tami]. (ef., Skt. =

) and takes the suffix -r (ir) as a persona]. plural

noun. The P.li form vafifia also suits this derivation (cf., Skt. = P. pufifia ). As the name is not of Tamil derivation,

it is possible that this caste or tribe originated

in the

Telugu or Kannaa areas, where Sanskrit caste names are not uncommon. Indeed the Vaiyar caste is still found in the North Arcot district which borders on the Telugu regions. There is no evidence regarding the date of the origin of this caste. There is no reference to the Vaiyar

in the

early

1. W.Tay].or, Examination and Analysis of the Mackenzie Manuscripts, Madras 1838, p. 78. 2. M.nier-1il1iams, Sanskrit-English Dictionary, p. 919

314
sources. It is, therefore, difficult to say when this caste originated We are inclined to think that the name Vanni originated in India and not in Ceylon. In the first place, it occurs in the South Indian sources earlier than in the Ceylonese works Secondly, it is unlikely that a Sinhaleae caste with the name Vanni migrated to South India or that the term vanni was introduced from Ceylon to designate a caste in South India. But the converse is possible. Further, the absence of traditions in the island regarding the origins of the Vannis and their prevalence in South India may also point in the same direction. Finally, the Tamil chronicles of Ceylon refer to the migration of the Vannis from the Tamil country to Ceylon. It seems, therefore, not justifiable to say that the name was applied to a class of chieftains or a group of Sinhalese in Ceylon because they were living in. the forest regions. It appears that the term Vanni became current for chieftains in the abandoned regions of Rjaraha and in. the forest tracts ofsewhere after Vanni chiefs from South India established themselves in the northern parts of 1. In the Kallam the Vaiyars are said to have been created as a result of a miraculous conversion of twelve boars into hiimin beings. Some take this to indicate their origin as subordinates under the Chlukyas whose emblem was the boar. Cf., Hindu Organ, Jaffna, and S.Gnanapragasar, Y ppa-vaipava-vi arcane, p.kl. This is mere speculation. 2. See infra, .

8.1.23

315
of Ceylon. It is even possible that the term was introduced into the island before the Vanni chiefs went there, in the same manner as South Indian administrative terms came to be introduced But this seems unlikely since vanni is not a term used in esimilar sense but rather a name that was applied to a caste or comzminity. The earliest occurrence of the term is

perhaps in the inscription No.556 of 1919, which appears to belong to the time of Rjarja Ca The basis of this surmise

is the reference to one Pottappicca in this inscription. Presumably he is the same as the Pottappicca who figures in other records of the time of Rjarja I The term that occurs in our inscription is vanniyapparru, meaning the area or region of the Vaiyars. A more definite occurrence of the term
is in einscription of Rjendra I. The reference here is to a
o-p If

certain Vaiya Rva (Rva the Vaiya). After this a number of persons with the name Vaiya are mentioned in the epigraphs of the time of Rjdhirja II, Ku1ttufLga III, and avarma

1. Cf., meykppar, m1si, mutten, etc. 2. }.E.P. for 192 , No.556 of 1919. 3. Cf., LA.NilRknta Sastri, The Cas, p. 505. If. LE.R. for 1 98, p. 2.

31
Kulaekhara P4ya as well as in the inscriptions of the Vijayanagara period 1ost of the persons mentioned in the Ca and P4ya records bear the title of Vaiya-nya or Vaiyar-nya (Lord of the Vaiyars) and appear to have been Vaiya chiefs. Prominent among thea id one Vaiya-xiya Cuut who figures in as many as fifteen records of the time of Rjdhirja II. He is described in these as a Ialaiynp chief, with the fulsome epithets Malaiyam Iaiyr Periya Uaiy CuUut Vaiya-nya, Rjarja Cdiyarja One of the Ceylonese works of this period, the Upsakajanlaikra, also refers to a Vanni feudatory of the P4ya ruler. (Pau-bhThnaale yo'bh vafiflo smanta bhThnipo) On the basis of these references we may venture to suggest that tabout the

twelfth century some Vanni chiefs had risen to prominence as feudatories of the Cas and the Pyas. Perhaps they were able to wield much influence as the suppliers of Vaiya soldiers to these South Indian rulers.

1. ?.E. . for

1903, Nos.

5k6 and 558 of

1902; N.E.I.
for

for 1910,

Nos. 215 and

?.E.R. for

13k of 1910; M.E.R. for 1913, Nos.30 and

34 of

1913;

1920, No .556 of 1919; F.E.P.

1922, No.352 of 1922;

.E.R. for 193 4/35, Nos. 122, 143-149, 154-159, 126, 162, 177,
215 and 189 of 1934/35.

2. }.E.R. forI3f3c,

No.lofI1344M

3. Upsaka-janlakra, p.157

317
In Ceylon, the earliest work in which the name Vanni occurs is the P1!jvaliya. In connection with the occupation of RAjaraba by Igha, this work refers to the ]ahavanni areas 1 and the chiefs of those regions who lived in. fear of }TAgha. This would mean that by about the latter half of the tbriteenth century the term vanni bad come to be used in Ceylon to designate minor chieftaincies in the areas of Rjaraha where the authority of the Sinhalese ruler was not felt any more. The PUjvaliya and the Clavaisa frequently refer to the Vannie. The period to which these references relate is what may be called the post-Polonnaruva period (after 1215). Geiger is, however, of the opinion that there is a notice in the CUavaisa regarding the Vannis of the twelfth century, although they are not mentioned by that name here. E hat Vannis oline Zweifel auch schon im l2.Jahrhundert gegeben, denn auch in der Beschreibung der Zustande, wie ale durch die daxnaligen Burgerkriege geworden waren, findet sich }hvs. 61.62 die Notiz : ,,Leute von vornebmer Abkunft (kulina) bielten sich, bier und dort an geeigneten Platzen (phasutthanesu) verstreut, verborgen und nahmen ihren Wohnsitz daselbst". 2 This claim of Geiger is based on his assumption that the Vannis were Sirihalese of noble descent who sought refuge in the forest regions in times o distress and later came to be called Vannis,

1. Pv., p. 109. 2. W.Gei.ger, 'Die Vannis', . cit., p.

8.

318
signifying 'jungle settlers' (Waldsiedler). But this interpretation, as we shall see presently, is unacceptable. Paranavitana, too, feels that the Vanni chiefs appear to have been in Ceylon 'from early days' (earlier than the thirteenth centur) Thi8 opinion is based on certain statements in the Nikya-sahrahaya and the Eu-attanag4uvaisa . The latter works alludes to certain Siri-vannis in the Attanag4u region who disregarded the authority of Nugalan ) who was ruling at Anurdhapura

This monarch may be any one of the three Moggallnas who ruled between the fifth and the seventh century. The source of our information is a work of the post-Polonnaruva period and, therefore, the reference to Vannis in the period before the eighth century does not seem to be authentic. Paranavitana himself has cast doubt on this reference by saying that 'we cannot be certain that the author of this text was not attributing to the past conditions which were normal in his day' The Nika-sagrahay-a, too, has a similar reference. According to this work, ParkramabThu I conquered three hundred and sixty-four Vanni territories. This
is,

however, not mentioned in the

1. U.C.H.C., I, pt. 2,
2. E.u-ay . ,

P.

738.

p. ki.

3. U.C.H.C., I, pt. 2, p. 737. Li..

NIL gi&n,, p.

20.

319
C!1avasa. Though at first sight these statements may appear inauthentic, they are actually not so. This could be explained easily if we analyse the exact use of the term vanni in our medieval Sinhalese and Pli sources. In these sources, vanni is applied to chieftaincies and chiefs in Ijaraha and in other forest tracts. Jhile in the Sinhalese sources vanni occurs alone to mean either a chieftaincy or a chief in the C'alavaisa - 2 _3 it occurs always in compounds, namely vanni-rajattam, vanni-raja, vanni-r jno vanni-mahIpla maha-vafifia-r ja!fia 6 (variant: niahavanya-rjafi?1a)and vanni-rjhi all of which stand for 'Vanni e.xce kings'. Perhaps the on]$ e-u%.,w.ic-e is the occurrence in the personal name Vanni BhuvanekabThu, but here,too, it is part of that name Geiger's contention that the whole compound vanni-rjno refers to a noble clan (adeliger Clan) of the Sinhalese in the same way as 'Licchavi-rajjuruv'stood for the Licchavi clan is not convincing. We are inclined to take these compounds to mean 'kings of the Vanni'. Geiger's argument that the word vanni never occurs alone but always

in a

conipund is based purely on

the Clavaisa. Paranavitana' a statement that vannin were a 1. f., Pv., p.109; Rv., pp. kk, 65, 66; Nk.s*n., p.20; Girtsand.a, v.128.
2. Cv., 81:11.

3. Ibid., 83:10. 6. Ibid., 88:88.

k. Ibid., 87:26, 52. 7. Ibid., 89:51.

5. Ibid., 88:87. 8. Ibid., c101Oc

320
class of chieftains is right in so far as the Sinhalese and P.li sources are concerned. In these sources the term is used to denote chiefs and chieftaincies in the areas that did not come under the direct rule of the Sinhalese king. When the authors of the Eu-attanag4u-vapsa and the Nikba-sarahaya refer to vanni chieftaincies of earlier centuries, they were only using a term that came to be applied to those chieftaincies in the thirteenth century and later. These references need hot be taken to imply the presence of a clan of people called Vinnia in those times. When Geiger referred to the Vannis as a noble clan of the Sinhalese who took refuge in the jungles in the time of }Agha, he was only referring to those Sinhalese who set themselves up as minor chiefs in the abandoned areas of Rjaraha which came to be known as the Vanni. e was basing his statement on solely on the Pli chronicle and did not take into account the evidence of the Tamil sources regarding the Vauiyars. He is wrong, as we shall see, in calling the present-day Vanni caste of the Northern and North-central Provinces as the descendants of the Sinhalese Vanni-rjno of our period. In the Sinhalese and Pli works of the thirteenth and fourteenth centuries, therefore, the name Van.ni has been applied to the chieftaincies of Rjaraha and other forest tracts. As Paranavitana has pointed out, there seem to have two classes

321
of Vannis, namely the }iaha-vanni and the Siri-vanni. In some works the number of Vanni chieftaincies is given as eighteen 1 and in some others as three hundred and 8ixty fotar. These could hardly be taken seriously. Eighteen is a conventional number often met with in Indian literature. In fact, some Sanskrit works refer to the existence of eighteen forest kingdoms (avikar
jy a)

Since the Vanni chieftaincies were also forest kinglets,

the Sinhalese authors may have referred to them as eighteen in number, following the Indian practice. In South India, too, there are references to the Vamiyars of the eighteen (districts) It is possible that traditionally it was considered that there were eighteen Vanni chieftaincies. In the Tamil chronicles, however, the number of such chieftaincies in the island is given as seven Probably this referred to the major chieftaincies that were feudatory to the Jaffna kingdom. As in the Sinhalese sources, the name Vanni is applied in. the Tamil chronicles of the island to the chieftaincies
_$%

511.I'P.

of

northernkCeYlon.

But the name Vaniyar

is applied to a caste

1 a. E.u-a y ., p. kl.

1. U.C.H.C., I, pt. 2, p. 737.

2. Cf.,
4. D.C .

Vsv1'1bv
rcQ.1,CCc. 1s7,?c

-+o'y .e..ik QivI1i.J.i 1 T,

3. A.S.S.I., IV, (J.Burgess, Tamil and Sanskrit Inscriptions), p. 120.

k. Yvm., p. 38.

322
of South Indian Tamila whose leaders were the chiefs of the Vanni districts These Tamil sources preserve traditions relating to the migration of this caste to Ceylon, which event appears to have taken place in the early part of the thirteenth century. In the present day, with the opening up of several colonisation schemes in the Vanni, the Vaiyar caste has almost become integrated into the Sinhalese and Tamil population. But in the nineteenth century when the Vanni was being opened up for the first time the Vanniyar formed a distinct caste and followed their age-old occupation of bunting and occasional cultivation. Not all the people of the Vanni areas belonged to the Vaiyar caste. In fact only a small percentage of the people of the Vanni were Vaiyars. The following observation of Fowler is worthy of note in this context. These people are the Wanniaha and are entirely dependent on hunting and occasional chena cultivation. They have no money and cannot buy land. These Wanniaha are a distinct caste, of which these men are the only representatives in the province. (There are five or six villages in the North-Central Province, I believe). They still use the primitive bow and. arrow and. are well acquainted with the most remote jungles through which they wander in search of honey and game. There are some peculiarities in their dialect, which with their mode of life, suggest relationship with the Veddah, but they alto ether repudiate the idea. 2

1. See infra, pp . 3

fl-

2. S .Fowler, Diary of 3rd ?iay 1 87, quoted in the lionthly egister and Noted and Queries for Ceylon, II, No.

5, May

189k , p. 98.

323
The Vaiyars of the nineteenth century were divided into two different communal groups. Those who lived in the Vavuniy and Mullaitivu districts were Tamil speakers while those in the Nuvarakalviya district were miinly Sinhalese speakers. There are reasons to believe that these Sinhalese-speaking Vannis were in fact descended from Tamil Vaiyars who had become assimilated to the Sinhalese population after the Nuvarakalviya district was re-colonised by the Sinhalese. It was traditionally believed by these people that they were descendents of Tamils. A.Brodie, basing his account on certain traditions preserved among them, wrote in 1856:There is one (caste) here not general over the Island and which is superior to that which is elsewhere considered the highest. I mean the Wanni caste. These persons are descendants of certain Tamils who came over from the continent in the time of Raja Zen, who granted to each extensive tracts of land. 1 There were other aspects of their life which revealed their close affinity to the Tarnils. Another o# the nineteenth-century writers makes the following observations on these Sinhalese of the Vanni region;They have adopted the T mu system of personal names, thus a man has his father's name prefixed to his own and does not take his name from the village or family he belongs to or the land he owns, as is the common Sinhalese custom elsewhere.

1. J. .A.S. (C. .), iii

, 1&56, p. 1k9.

324
Many of their names, too, are Tamil in a Sinha].ese shape. The older generation have taken to weari*g earrings, but this practice has been discouraged by the present Sinhalese headmen. The Sinhalese villagers have as much faith in the Hindu god Pillaiyar (Ganes&') as have the Tamil villagers whose favourite god he is......... As regards dress the Linhalese keep generally to their own customs, but they often wear the Jaffna cloth (chayaveddi) and fasten the handkerchief on their heads after the Taniil manner. 1 The foregoing observations of nineteenth-century writers reveal certain facts about the so-called Sinhalese Vanniyas. In the first place, we find that they were a caste distinct from the rest of the Sinhalese. Secondly, traces of Tamil descent could be found in their traditions, customs and nlRnners. Thirdly, they considered themselves to be superior to all other castes in the Vanni. This feeling of superiority was evidently due to the fact that they were at one time the ruling caste in the Vanni. In the light of these considerations it n.s difficult to accept the view of Geiger that the Vannis were a degenerate group of Sinhalese. It appears that the Sinhalese Vanniyas who lived as a separate caste in the North-central Province were descendants of Tamil Vaniyars who migrated to the island in the thirteenth century. It is in the same century, as we have noticed earlier, that we first hear of Vannia in the Sinhalese sources. It is unlikely that a Sinh ieee caste called the Vannis

1. Anonymous, 'The_Vanni', The lionthly Literary egi ter and Notes and ueriea for eylon, II, No.5, Iay l89+, p . 98-99.

32
came into existence independently in the present North-central Province at a time when a community of Tamil Vaiyars settled in the adjoining districts. It seems more plausible to assume that the Vanni people of the North-central Province in the thirteenth century were settlers from South India like the Va D iyars of the Vavuniy and IlullaitTvu districts, and that their descendants became assimilated, to the Sinhalese 0 ulation when Sinhalese re-colonisation took place in those areas at a later date. This is clearly suggested by the evidence of place names in the North-central Province. By far the majority of the names of Sinhalese villages in this province in the nineteenth century was of Tami]. on in. These villages, as we know from the inscriptions and literary sources, bore Sinhalese names before the thirteenth century' What led to this change of local nomenclature? The explanation is not far to seek. Some time in or after the thirteenth century these villages were occupied by Tamils who gave Tamil names to them. When Sinhalese re-colonisation took place the Tamil settlers seem to have been gradually assimilated to the Sinhalese population. This would explain the retention of Tamil place names by the Sinhalese as well as the presence of Tamil castes like the Vanniyas, Cliyas and Bairis speaking Sinhalese but still retaining traces of Tamil descent. It appears, therefore,

1. See infra, p.3c

32
reasonable to assume that the latter-day Sinhalese Vanniyas were descendants of Tamil settlers from South India and were related to the Vaiyars of the Tamil areas. The term Vanni was also used to refer to the chieftains of the Vanni 'egiona who may or may not have been of the VRnniyJ eaata.The term vannirjno of the Clavaisa does not refer to the Vanniya settlers but only to the chieftains of the Vnniregions who were both Sinhalese and Tamil. Traditions regarding the migration of the Vaniyars from South India are preserved in the Tamil chronicles. In the Takia-kailc a-pur and the Tiri-k;c ala-pur K!c ar-kalveu

a-vaipava-xnlai this migration is connected whose identity and activities

with a personage named Kua

have formed the subject of some amount of controversy. He is said to have invited Vanni chieftains from the mainland and entrusted them with the care of the Kvaram temple in Trincomalee and its lands. The identity of K4akka, has not been easy to establish. Recently Paranavitana identified him with a C4a-gaga prince who went to Ceylon in 1223 , presumably from the KaliAga country It cannot, however, be claimed that be has settled the problem once and for all.

1. S.Paranavitana, 'The Irya Kingdom of North Ceylon', p.

cit.,

179.

321
In the sources mentioned above Ku.aka is described as a aiva prince from the C5a country who went to Ceylon on a pilgrimage, tarried at Trincomalee and effected repairs to the ruined temple of X5!varam. TheTTakia-kai1capurarn calls him K4akk alias C.akakai, the son of

Mau-n!ti-kaa Ca The Tiri-kcala-puram refers to him - as the son of Vararamateva C akn'ik of the Ca country.2 The K car-kalvetu agrees with this statement but does not The a-vaipava-

give Vararmatva's suename as C

mlai follows the Takia-kailca-piiram and refers to Kuakkffa's father as }au-nti-k a-cia 1au-n!ti-kaac]a is a mythical ruler reputed for his benevolence and compassion who finds a place in the legendary geneaoogy of the Cas It is hard to explain how he caine to be associated with Kuakkta. As far as we know no Ca ruler by the name of Vararmatva CakazUca or Vararniatva ever existed. The Ca descent attributed to,Kuaic.ka in the Tainil sources appear to be

1. 2.

., Pyiram, v.

8.

3.

., p. 1. yy , p. 8.
XV, p. 4-

5. Li.

323
rather unreliable. Even the name K4adca, does not seem to have been the real name of the person. It means 'He of Tanks and Temples' and is very probajIy a sobriquet he earned after his tank- and temple-building activities. As Paranavitana is inclined to believe, Cakaka may have been the real surname of the prince. While the Tak1a-kailca-puram, the oldest of the above Tamil sources, calls him Caiu1 k ij, the Tincala-puram calls his father

C1akaxtka,. C

akaka seems

to have been the family name of the prince, as it was in the case of the Eastern Gagaa (Co.agaga or Cagaga). We are, therefore, inclined to agree with Paranavitana that the real name of this prince was Cakaka. The identification of this Cakaika, is no easy matter for the name was commonly used by the Eastern Gaigas as well as by princes and feudatories in the Ca, Pya and Karta countries in the twelfth and thirteenth centuries. Noreover, our sources reveal that a number of Indian princes of this name had been in Ceylon in these centuries, As a result we are up against several possibilities. Our prince may have been (a) CoagaAga-kumra who lived in the court of Gajabhu II (b) Coagaga who invaded Ceylon and captured power in U96

1. X7 :V3.
2. Ibid.,

329
(c) Co4agai a of the Trincomalee Sanskrit inscription who landed in Ceylon in 1223, Cd) C4agagadeva who invaded Ceylon some time before l28k or (e) any one of the princes of the Eastern GafLga, Western Gafiga or Ca family or one of the feudatories of the Cas or Pyas. To consider the first possibility, we find that Co.a agakum.ra was a Ia34a prince who lived in the court of Gajabhu II around 1153. The only evidence which may be used in support of the identification of our prince, Kuakka, with Coagafigakumra is to be found in the Tamil chronicles. In the Takita-kai]c a-puram, T iri-kc a1a-puram and the K'c arkalve tt u , Gajabhu and

K4a

are closely azsociated with

the K!varam tenxple The Ma akk4appu-mnmiyam, which refers to as Makc!na (Nahsen, states that this

prince married a Ka1iga princess who was an adopted daughter of Gajabhu hese-tr&i-t4ou inaj kre-eeTrvc omo memory of aand Ga j abnu Is it likely that GajabThu was closely associated with two different lAd-i-an- princes named Cbagaga or are we to treat them as one ?

1. 2.

, V , p. t73 ,

3.

., 7: 8
p. 2

ff. ;

., Kayavkuppaa.1am, p. 170 ff.; Kk., p. 20.

k. Mm.,

330
The evidence of the Tamil chronifles is not strong enough for such an identification. Though the traditions concerning the Kvaram temple were preserved by the temple authorities for a long time, it may be difficult to base our conclusion on the evidence of the late works which embody these traditions. It is not impossible that Gajabhu and Kua&a lived in &Lffereut periods, as indeed the Tkia-kailca-puramand the Tiri-kcalapurazn treat them, but were brought together by late tradition owing to their close association with the Kvaram temple. Coagaftgakumra who lived in the court of GajabThu II may, therefore, be different from K4a.kta. The Kaliga prince Coagafga who seized power in

1196 is said to have been a nephew of Nii3ki Na1la it is not

stated in our sources whether he invaded the island in 1196 or whether he had gone there some time before that date and captured power in 1196. If he had gone to Ceylon in 1196, it is unlikely that he is the same as the C4agafxga who effected repairs to the K3!varam temple and settled Vaiyars in Ceylon, for he was ousted from the throne within a year and it is hard to think that under very insecure conditions he would have undertaken the task of re airing temples and settling peo le from South India. Moreover, if he was a nephew of NLaka halla

1. 2x.

331
and aspired to the kingship of the island be may have been a Buddhist and not a aiva. It seems unlikely that he is the Co.agaga whom we are seeking to identify. The Sanskrit inscription from Trincomalee, discovered among the ruins of the Kvaram temple, refers to a personage 1 named Coagaga who went to Ceylon in 1223. Paranavitana has i identified this person with Ku.ak1Za. The inscription is fragmentary and is engraved on a part of a stone door jamb. Among the decipherable words is the name Gokara, the ancient name of Trincomalee and the root from which the name of the temple is derived (Gokarevara). Since the epigraph is inscribed. on a part of a building, Paranavitana feels that it ' may reasonably be assumed to have recorded the building of the monument of which it formed a part' He therefore argues that 'it is very unlikely that there were two Coagaftgas who both came from a foreign country, landed at Trincomalee and busied themselves effecting improvements to the Saiva shrine there' He adds further that the date of CoagaAga's arrival being 1223, it 'agrees with the statement of the Yvm. that this prince had dealings with chieftains known as Vanniyars,

1. E.Z., V ,

p.fl3
. cit., p. 179.

2. S.Paranavitana, 'The Arya Kingdom of North 3. Ibid.

332
for it is only from the thirteenth century that Vanniyars or
Vannis

are mentioned in the contemporary writings! Paranavitana

is also,of the opinion that this Coagaga is an Eastern GafLga. There is, however, no evidence for such an assumption. But there is no evidence eo the contrary either. Gokar.vara was the favourite deity of the Eastern Gagas The fact that a CoagaAga from outside the island interested himself in the affairs of a temple of Gokarevara in Ceylon may indicate that he was an Eastern Gaiga. Probably Paranavitana is right in identifying him as an Eastern Gaga prince. Paranavitana's arguments for the identification of this Coagaga with Ku.aka seem quite tenable. But let us consider the other possibilities, too, before we arrive at a satisfactory conclusion. The CUavaipsa has also a reference to another C 4agagadeva who is stated to have invaded Ceylon some time before 128k. From the manner in which this event is introduded to in the chronicle it does not appear to have been a major invasion. It is said that BhuvanekabThu I 'drove back all the Damia , like K'ligaryara, C4agazgadeva and the

rest who had landed from the opposite coast' Apparently these

1. S.Paranavitana, 'The Xrya King om of North Ceylon', . cit., p.180. 2. ee-iiira1---. g.I i 3. Cv., 90:32. , I

333
enemies had led punitive raids which were not of much significance and in time Bhuvanekabhu got rid of them. The nature of the expedition of Kuaic.ka also seems to have been similar, according to the Tamil sources which state that he had an army with him but did not effect any conquest But there is one miin difficulty in identifying C4agagadeva with Kuakka,. The former's invasion took place not long before

128k, the

year of Bhuvanekablhu's death. If we are to accept the testimony of the Tamil sources that K4akka introduced Vaiyars into the island, this event must be places before 1270 when we first begin to hear of Vannis in the literary sources Unless we take that Kuai&ta only introduced a further band of Vaiyars in to the island, it may not be possible to identify K4aa, with Co.agagadeva. Cc4agagadeva may be

different from Kuacka. It is possible that K4aa, is different from any of the four Coagaigas known to have been in Ceylon in. the twelfth and thirteenth centuries. He may have been a C1a, Eastern Gaga or Western Gaga prince or a South Indian feudatory ruler whose visit to the island is not recorded in the Sinhalese and

1. Cf., Tv ., p. 10. 2. Cf., Pv., p. 109.

334
Pli sources The Ya-vaipava-mlai, Takia-kailc a-puriam, Tiri-kc a1a-puram, Kc ar-kalve u and the Munn vara-mnm.iyam maintain that Kuak was a Ca prince The last mentioned

workgives Crya Ku]Zttka (SUrya Kuittuga) as one of the titles of the father of K$a1c.kffa. It is unlikely that this is a reference to KulttufLga Ca . Of the three Ca rulers named Ku]Zttuga, only Kulttuga I is known to have had a son named Coagaga (or Coagaga) This prince 'vanished into obscurity after his viceroyalty at V!gi' What happened to this prince after his viceroyalty at VgI ? Did he go on a pilgrimage to Ceylon and effected repairs to the Kvaram temple ? We can only speculate on this point. There were also other G1a princes named CoagafLga. We know of at least one, whose other name was iadhurntaka, figuring in one of the i4acriptions from Bangalore Taluq. There were also several Ca-Pya feudatories named Co.agaga. One of them figures in about five inscriptions dated in regnal years of Rjarja III and Kulttuga III, between 1210 and l222 Another feudatory named Periy Aakiyapperiim,
if

1. Yvm., p. 8; Kk., p. 1;

7:23;

., Varar atvar varu patalam, v. 4;

vara-m iyam , p.

2. K.A.Nilakanta Sastri, The Cas, p.3'l 3. Ibd.

if.

., IX, Tamil section, p. 17.

5. F. . . for 1913, Nos. 535, 546, 549, 556, 557 and 559 of 1912.

335
alias Co.agafiga whds mention in a number of records of the time -. 1 of Rjaraja III. Certain other C4agagas are known from P4ya inscriptions The Upsaka-jan1a.kra refers to a P4ya feudatory named C4agazga who was himself a vafifia (Vaaiya& This may mean that some of the Vaiyar chiefs were themselves known as C4agafiga and one such chief may be Ki4akka who, after establishing himself in the eastern part of Ceylon, settled there some of his kinsmen or Vaiya subjects. It was from South India that gha obtained a large part of hi8 troops. Probably there were Vaniyars among his mercenary leaders. In the folk traditions included in the aakk4appu-mmiyam it is stated that the Mukkuva Vaiyars (Mukkuva chiefs) of the Batticaloa area went there as leaders of 1gha's troops and that they were granted nipams (chieftaincies) by

?gba, who is often referred to in this work as Klika (Kalixga) The first mention of the Vannis in the time of gha5 and the
tradition associating him with the Vanni chiefs of Batticaloa aug

est that there may have been some connection between }gha

1. ?.E.L for 1908, Nos. 202 and 205 of 1908; Y.E.R. for 1926,

19k of 1926.
2. Cf., LE.R. for 1921, No.lkO of 1921; Z._.R. for 1922, No, 203 of 1922; M.E.R. for 1915, Nos. k09, kb, k13 of 191k. 3. Upsaka-jan1azkra, p. 9
k. }., p. 10k. 5. See supra,

33t
and the Vannis. Vaniyars were probably among the mercenary leaders in the army of Igha. It is possible that a Vaiyar chieftain named Coagaga was among them. }gha may have granted him a chieftaincy in the Trincomalee area where he became as Ki4akka. All these are, however, matters of speculation and in the absence of any real evidence no certain concluions can be drawn. It is, therefore, a difficult task to identify K$aka Cakazka with any degree of certainty. As the traditions concerning him are very strong one cannot doubt his historicity. The chances are that he is the same as Coagaftga of the Trincomalee inscription. In the first palce, as Paranavitana has argued, Coagafxga of this inscription is associated with the Kvaram temple like K4akkZta. Secondly, the dates of their activities in Ceylon also seem to agree. Ki4akka appears to have been in the island in the thirteenth century when we first hear of Vannis. Coagaga, according to the inscription, was in the island around 1223. It is probable that Ku is the same as this prince. He was presumably

a scion of the Eastern Gaiga family. The reference

in the

Tamil

chronicles to him as a Ca prince may be a confusion resulting from the name Coagaga. In the Tamil chronicles he is credited

337
with not only the renovation of aiva temples but alsO the repair of irrigation works such as the Kantay, Allai and Vearasa tanks This accounts for his sobriquet 'K4akkta (Builder of TnkR and Temples). The personality of K4akk3a has been obscured in Tainil tradition by several factors. His fame as a repairer of tnkR, for instance, has led to a confusion of traditions relating to him with those of the earlier and better known tank-builder Nahsena. In fact, the Maakk4appu-nimiyani refers to K4akkta as Makc Although Kuakka seems to have taken an interest

only in the repair of the major irrigation works in his principality, Taniil tradition has credited him with the building of those tanks. The Maffakk4appu-nnmya also refers to him as a Vaitulliya CaivaA (Vaitulya aiva) an obvious confusion with Nahsena who, according to the c1flvaisa, was a Vaitulya Buddhist Similarly, while Nahaena is recorded to have destroyed Brhmanic temples in Trincornalee, Ku.akka is stated to have destroyed Buddhist structures in the same place Evidently

1. Cf., 2. Nm., p. 3. Ibid.,

, Tiruppai cey- p4alam, v.

65.

32.

p. 33. !v., 36:111.

Nm.,

5. Ibid., 37:kl ;

p.3k.

33:
these traditions are due to a confusion between Mahsena and K4akk t a in the eyes of later chroniclers and story-tellers.

There has also been a confusion between K4akkta and Ukkiracinkaa (Ugra Sifxha), a legendary king mentioned in the chronicles of Jaffna Traditions relating to Ku.akktta, and. those about ?gha also appear to have been confused in the Tamil works As a result of these confusions the personality of K4akka is shrouded in mystery and his identification &eis rendered difficult. Divested of all these legends that have surrounded his persoxality, K4akkta appears as a powerful chieftain of the Trincomalee principality in the thrteenth century, probably in the reign of ?igha. He was probably an associate o Igha. Ganapragasar has attempted to identify him with Jayablhu, who, according to the Clavasa, was a powerful associate of Ngha It is difficult to say whether K4akktta was known to the Linhalese chroniclers as Jayabhu. There is, however, no person by the name of JayabThu in the Tamil works. All that we can say is that Kujakka was probably a owerful chieftain of the Trincomalee principality under gha and earned his fame by renovating the KTvarani temple and repairing some of the tnk in his chieftaincy.

1. See infra, 2. See infra, p9.3714. 3. S.Gnanapragasar, -vaipava-vimarca, p. 6k.

33)
In the light of the evidence that emerges out of the confused sources at our disposal we have to assign the migration of the Vaiyars who settled in northern Ceylon to about the first quarter of the thirteenth century. We are inclined to believe that they were among the ercenaries who went to Ceylon with Ngha or with some of his associates, chief among whom ap ears to have been the personalit T known to us as K4akka. After the conquest of northern and eastern Ceylon was effected by the invaders, the present Vavuniy, Mullaitvu, Trincornalee and Batticaloa districts were probably divided into several chieftaincies and granted to Vaiya and other mercenary leaders. K4ackaa seems to have been res onsible for the creation of such chieftaincies in the Trincomalee and Vavuniya districts. Igha, according to the Batticaloa traditions, created similar vaipams in the Batticaloa district It was in this manner that the Tami]. Vanni chieftaincies a pear to have emerged. Once the term vanni became current in Rjaraba, it was probably applied to similar chieftaincies in the depopulated Sinhalese areas of Rjaraha as well and came to stand for any jungle chieftaincy. Those Vanni people who were settled in the areas which later came to be re-colonised by Siniialese gradually evolved into the Sinhale e Vanniya caste while those

1. See infra, p.379

340
of the Tamil regions remained Tamil Vaiyars. This is how we have to reconstruct the story of the Vaniyars with the meagre evidence at our die osal. The picture may change when further evidence comrs to light but the general outline is likely to r remain almost the same. The Tamil chronicles refer to seven Vanni chieftaincies in the island. These corresponded roughly to the present Vavuniy, Nullaitivu, Trincomalee and Batticaloa districts, and possibly included some parts of the North-central Province. By about the nineteenth century only the Vavuni and NullaitYvu districts continued to be known as the Vanni. The Tamil chronicles do not mention the Sinhalese Vanni chieftaincies that covered the major part of the North-central and Northwestern Provinces. There is evidence of these regions having been settled by Tamils in the thirteenth and fourteenth centuries Even now two of the revenue divisions in these provinces is are known as Vanni Hatpattu and Dema.a Hatpattu (Seven Tamil Divisions). Like Jaffna, the Tamil Vanni districts of Vavuniy, Mullaitivu, Trincomalee and Batticaloa districts were settled by Sinhalese before the conquest of the Cas. E igra hical, archaeological and place-name evidences bear testimony to the

1. See infra,

p.37

31
Sinhalese settlements that once covered these regions. ProtoSinhalese and Sinhalese inscriptions ranging from about the third century B.C. to the tenth century A.D. have been found in these regions Almost the whole of the area is spotted with ruins of early Buddhist structures. The present Batticaloa district and parts of the Trincomalee district were included in the kingdom of Rohaa and it is needless to say that these were peopled by Sinhalese before Tamils settled there. The inscriptions of these districts preserve the Sinhalese names of many of the places which now bear purely Tamil names or Tamilised. forms of earlier Sinhalese names Only a few Sinhalese inscriptions of the eleventh and twelfth centuries have been found here and after the twelfth century we do not come across any. Tamil inscriptions occur in these regions from the eleventh century. With the Ca occupation a slow and not too visible displacement of the Sinhalese by the Tamils seems to have begun. Ca

1. Cf., U.C.H.C., I, pt. 1, pigraphic map opposite p.l, A.S.C.A.R. for

1905, pp.k2-k3 ; 2k0 ff.;

for

A.S.C.A.R. for

19 7, p. 29; 1933, 19k -4, p. 39;

E.Z., V,
2. p.

pt.2, pp.

E.Z., I, p.70 ; A.S.C.A.R. for 10 ; A.S.C.A.R. for

1k; A.S.C.A.R. ,

1935, p.

A.S.C.A.R. for 1953, pp. 21, 2 ; A.S.C.A.R. for

195k, pp. 29, 36.

2.

GagataI (Kantay) , A.S.C.A.R. for

1937, p. 1 ; p. 8.

P.iagamu (Pak m), E.Z., I p. 39; C.W.Nicholas, p. 81. Velaka or Velagama (Velakmani), A.S.C.A.R. for 193k,

342
inscriptions and temples have not been discovered in the Batticaloa district. In the Vavuuiy district, it is only in the regions bordering on the Trincoxnalee district and the North-central province that we get a few Tamil inscriptions and aiva temples .1 dating to the eleventh and twelfth centuries. But in the Trinconialee district several Tamil inscriptions and Saiva temples dating from the Ca period have been found This district seems, therefore, to have attracted Tamil settlers earlier than the other three districts. It is in the thirteenth century , with the invasion of Igha, that widespread Tamil and Ker4a settlements appear to have been established in these districts. The establishment of Dravidian settlements in the Vanni districts is dealt with in the Kcar-kalvettu, Tinkcala-puram, Vaiy, Vaiyal and the Ya-vaipavazrilai. The car-kalveu and the Tini-kcala .punam are

both chronicles of the K;varam temple and therelore embody the same tradition. The account in these two works may be summarised in the following manner. The prince K4akZaQ, after effecting repairs to the K!va.ram temple, decided to invite some families from South India and entrust to them the task of maintaining the different services in the temple.

1. See supra, 2. See supra, p c I

343
According to the Kcar-kalveu, Ku.akka went in person to invite the famileee while the other chrdinicle states that he sent his ministers. In response to the invitation thirty families went from JarufLkiir. They were of the V4ava caste. They were assigned the duties of attr and were settled in

Trincomalee. Twenty families went from Krai (K'raikkl), They were conferred the title of Parattr and assigned various duties. Pa3ave.i, in the Trincomalee district was granted as to them for settlement. A nobleman of the Krar family was invited from Tirunelvli (Tinnevelly and conferred the title of mutanm (chief). He was assigned duties concerning the conduct of festivals and was granted the villages of Kauk4attr and Nilv4i. A minstrel from Kfici was assigned the duty of singing hymns at the temple *nd was settled at Campr in the Koiyr division of the Trinconialee district. The prose section of the K!car-kalvettu adds that five cris (master craftsmen) were invited from the Ca country and were settled in Trincomalee. When all these people were assigned different duties and were settled in and around Trincomalee a nobleman from ladurai was invited and was a ointed as their chieftain).

1. Kk., pp.2, 3, 36, 37;

., pp. 131-135.

344
In the La-vaipava-nilai this account has r atly been greatly modified. According to this chronicle of Jaffna, the fince Ku.akkan, after having completed the renovation of the temple, assigned fields and estates in seven districts to the temple. lie then invited and settled Vaiyars in those districts and entrusted them with the task of cultivating the temple lands. The seven districts in course of time became the seven Vanni chieftaincies The Vaiy and the Vaiypal contain a different version of the Tamil settlement in the Vanni districts. The confused account in these works may be summed up in the following manner When Vararcac ika, (Vara Rja Sifha), a son of Ukkiracika (Ugra Zifxha) and a king of Ceylon married oa e .. from the P4ya country, she brought with her a retinue of sixty Vaiyars. One of the Vaiyars stayed behind at the capital of the kingdom and the rest were asked to take over the chiefships of the Aa.kparu region, which corresponds rou hly to the present Vavuniy district. These Vaiyars then invited prom South India a number of people belonging to the eighteen castes. They were settled in different parts of au as well as in the peninsula of Jaffna, as mentioned earlier. Ladurai, IaruAkir, Tiruccir a3i (Tric inopoly)

1.

1!! . '

PP . 11-12.

345
Malaiyakam, T4uvai, Toaimatalam, V4akiri-nu and Kvarati were among the laces from which iznnii rants went to settle in Aak au. After this follows a long and confused list of places in the Vanni districts where various castes and prominent personalities went and settled. These places are Muimnakar (Nujiyav4ai), Kakki, Taikkal, Kiakku-m'L i, Ntku-mlai,
It

Karaipparu, Karuv u-ki, Kat ukk4a-pau (Ka4ukki4am Division), Tirukkai (Trincomalee), Veruka]., Tanrpalakmam, Ko'iyrani (Kiyr), Ilaivayal, Varpp4ai, Tuukkyr (Tuukky),

Itt imau, Ne tu.ki , Noccimtai, Pulv4i, Akkaraip aru,

Tiriyy, Varavecti, Ceikk4am,and PaaikThnin. All these places are in the mo&ern Vavuniy, Nullaitvu, Trincoma].ee and Batticaloa districts. Among the castes and communities mentioned are Cr,(oil-mongers), Paaiyar (drummers), A)mpatiyr (Agampai mercenaries), Kaliz.kar (KalifL as), Malaiyakam (Ker4ad) and Vaiyar It is interestin to note that the Vaiypal mentions 'Pflpla Vaimai, K La and others' among the prominent people who went and settled in Tiriyy and Kaukk4am

1.

., vv. 29-81 ; Vaiy, p. 26 ff.

2.12.

v.73.

34i
The name of the Vanni chief appointed by Ki4akka is given in the K!car-kalveu and the ri-cala-purarn as PtTpla,. Ppla Vanimai of the Vaiypal appears to be the same person as Pp].a Vaniya. Like the chronicles of Trincomalee, the VaiypIal mentions Kaukk4am, Trinconialee and Koiyr among the places where Immigrants were settled and NarukUr among the places from where settlers Ident to Ceylon. Some of the traditions in the Vaiypal may have been based on those of Trincomalee. The Maakk4appu-mmiyam deals with only the origin of the castes of Batticaloa. The creation of these castes as well as the assignment of duties to them are attributed to lgha. Except in the case of a few, it is not stated whether these castes migrated to Batticaloa in the time of gha or earlier. The Mukkuva Vaniyar are stated to have gone from K3ikaam (unidentified). They belonged to the Paaiyci (military caste) and it was the X1ika (Kalix' a ruler) who chose the best among them (ek4flattrai) and took them to the island as commanders of his army The Kuru-ntar (Skt. uru Nthas) similarly went to Ceylon with the Klika Those of the

1. Nm., p. 10k. 2. Ibid., p. 105.

347
Ppla Kttiram (Bh' la Gotra) and the Pvaciyar (a mercantile community) also went to the island with the ____

It is difficult to reconclie these different versions and separate the historical sections from tile re8t. As we have already noted, the chronicles of Trincomalee and Batticaloa seem to preserve a more reliable tradition than those of Jaffna. An analysis of the above versions reveals certain important points. In the first place, it becomes doubtless clear that there has been a confusion of traditions relating to ?1.gha, K4aa and possibly other prominent personalities

connected with either the Tamil settlements or the creation of petty chieftaincies in or about the thirteenth century. Shorn of their details, the accounts of K4akka and ZrAgha appear very similar. In the akk4appu-m.flmiyam, the account

of }igha has four main strands which are similar to those of the account of Kuakka, in the chronicles of (varam. Firstly, }gha is described as an ardent Saiva who was intolerant of Buddhism and even the Vaiiava faith K4akktta,, too, is stated to have been a very devoted aiva although there is nothing in the Trincomalee chronicles to indicate that he

was a bigot. It is in the Maakk4ppu-mmiyam, where he is

1. Mm., pp. 105-106. 2. Ibid., pp. 53, 70.

348
called Na1aTh, that he is said to have destroyed Buddhist temples in Trinconia1ee Secondly, ?gha is associated with the building of the Tirukkvil temple and its tanks and with the invitation of priests to perform service in that temple

K4akka is credited with the renovation of R!varam t

temple and with the building of tanks. Thirdly, Ngha is stated to have assigned various duties for the different castes of Batticaloa This account is remarkably similar to the assignment of duties by Ku.a cta, to the various castes he invited from South India for the performance of services at the Kvaram temple. Finally, while K4a kDa is said to have created the chieftaincy of Trincomalee, the foundation of chieftaincies in the Batticaloa, Trinconialee, Mannr and Jaffna districts is attributed to gha In the second place, there has been a confusion of the traditions relating to K4akkUa with those about UkkiracifLkaIj, who may not have been a historical personality. Ukkiracifika's association with Jaffna is in some ways similar to Ku ka's association with Trincomalee. The story of

1. L. P . 3k.

4. Ibid., p. 77.
3. Ibid., pp. 70, 71, 95-97.

If. Ibi ., pp. 7k, 75, bk.

34J
Ukkiracika, , as it ap ears in the Tainil chronicles, is basically a different version of the Vijaya le end Sinhalese traditions as well as some South Indian legendary material have gone into the creation of this story which forms the starting point of the history of the Jaffna kingdom in the chronicles of Jaffna. Traditions of the Rohaa kingdom, which once included parts of the present rincoma1ee district, also appear to have helped the growth of the story of Ck1&6TA!. This is seen in the story of 1akacavuntari, the queen of Ki4aka in some of these accounts. In this story, traditions similar to those about Vihradev1, the mother of Di4hagmai, are to be found The confusion of many of these traditions seems to have been the result of a belated attempt on the part of the later Tamil chroniclers to reconcile the different floating traditions in the Tamil regions and to give these a historical sequence. In the story of Ukkiraci.ka,, for instance, we see an atte pt to reconcile the stories of Vijaya, K4akkta and possibly a third personality associated with the kingdom of Jaffna In this story, as we shall discuss later, we could see the

1. See infra, 2. See infra, .Ck.i' infra, 3. See ___ p. ckvs

350
character Siha and SihabThu of the Vijaya legend in the personalities named Ukkiraci-ka, and VararacacitLka r s ectively. Vararcacika combines the characters of SiiabAhu and Vijaya. The matrimonial mission sent by him to the P4ya court and the arrival of the Pi4ya princess with a large retinue, as mentioned in the

Vaiypal, are

both based on the Vijaya

legend. The coming of the Vaiyars and the invitation of the different castes from places like rui.kr in South India, their settlement in parts of the Trincoinalee 4istrict and the arrival of the chief called Ppla Vaimai are clearly based on the story of Ki4akka,. The rest is an elaboration of these main aspects possibly baeed,(other traditions and on the conditions that obtained in the time of the writing of the chronicle, that is, from the knowledge of the different castes that were found in the Vanni districts. The later chronicle a-vaipava-mlai attempts to reconcile the discre ancies seen in the traditions of the Jaffna and Trincornalee chronicles. The invitation of the Vaiyars is, therefore, attributed to Kuak]a, and. the invitation of Tauiil settlers by the blind minstrel Y]pi is included in the Ukkiracifikag story. Thus it has only helped to confuse the traditions further. Thirdly, we fin that some of the traditions in the above accounts of the settlement in the Vanzii districts are drawn from popular etymology o place names. Such names

351
as Kantajy, Pakai and Carnpl!r have formed the basis of these traditions. We have already seen how the author of the Tinkicala-puram has incor orated a tradition which attempts to derive the name Kantay (variant: Kant4ai) from the Taniil words ka (eye) and ta.ai (to grow), weaving a story round it,
(oJso

whereas it is actually derived from the Sinhalese name GataJ1va (Pli, Gagitaka) through the later form GaAtal Similarly,

the Kcar-kalvettu, Naakk4appu-mmiyam and the Tini-calapuram contain a story that has been woven round the place name Pamai, in the Batticaloa district According to this story, Pamai was the place where the Kaliga infant princess 4akacavuntart drifted ashore in her wooden cradle. Since the baby was found here smiling, the place was named Plar-nakai ( Plar = baby or young one; nakai = smile) which later became Pakai and eventually Pamai. But in fact Pamai is derived from the Sinhalese name Pnama. In the Linhalese inscriptions of the period between the fifth and the seventh century found in this place, its ancient name occurs as Pnava The final

1. See supra, p.34.I,J,1,i. 2. Kk., p. 33 ; 3. C.W.Nicholas,

. ,

Paalam XII, v.12 ;

Mm., p. 28.

. cit., p. 22 ; C.J.Sc. (G), II, pp. 113, 11k.

352
was replaced by ma1 in course of time and, when it became Tamiliaed after the Tamil settlement there, the final a was substituted with ai In this way some of the traditions preserved in the Tamil chronicles can be traced ultimately to the work of popular etymologists. Thus we see that a number of unreliable traditions have got enmeshed in the story of the Tamil settlement narrated in the Tanill chronicles. As things are, it is very difficult to extract from this anything more than a bare sequence of events. Compared with the chronicles of Jaffna, those of Trincomalee and Batticaloa are less confused. Of the later chronicles, that of Batticaloa, namely the aakk4appumiya , is certainly more reliable. It is the only Tamil chronicle which contains a number of episodes from the history of the Sinhalese before the thirteenth century, many of which tally with the accounts in the Mahvaisa and the Ct!lavasa. Further, the Maakk4appu-mmiyam is the only Tamil chronicle which mentions MAgha by that name and deals with his activities
in a manner that compares favourably with the Sinhalese accounts.

The miraculous and legendary elements which mar the accounts

1. The interchange of va and ma is common in Sinhalese, , navaya

2. This is

namaya .
with Tainil practice, Gamp4a Kamp4ai,

in keeping

Polonnaruva

Poloauvai.

353
in the chronicles of Trincomalee and Jaffna are found to a lesser extent in the Batticaloa chronicle. These qualities do not, however, entitle the account of the 144 akk4appu-mmiyam to be wholly acce ted. By a comparison of this and other Tarnil accounts with those of the Sinhalese and Pli chronicles we may be able to arrive at some of the basic facts, decide which statements are acceptable and te leave aside the dubious details that have to be treated with some amount 01 scepticism. As we have already seen, the Sinhalese and Pli chronicles leave us in no doubt that the invasion of gha resulted in the occupation of several parts of Rjaraha by Tamil and Ker4a soldiers and in the dislodgement o many Sinhalese from that area. Under the tyrannical rule of ?gha Bud hist institutions were destroyed and what is called a 'false faith' 1 was propagated. 'Villages and. fields, houses and gardens' were 'delivered up to the Ker4as' Dami3a warriors were 'settled here and there in the country' Even in

1yr4tha there

were

'Daniia warriors who dwelt as they pleased in the sin le villages and houses's The forces of 1gha and JayabThu had

1. Cv., 8O;75. 2. Ibid., 80.76.

3. Ibid., 83:12. k. Ibid., 81:1k.

351
set up fortifications in several places in Rjaratha. These included Po].onnaruva, Kohasra (Koiyram), Gafigtaka (Kantay), KkIlaya (Kokkiy), Kavt4vi4u (Ka12kk4am Pattu), Pad! (Padaviya), Kurund! (Kuruntar in Karik aumlai South), )inmatta (possibly near Giant's Tank), I4ahtittha (tam), Mannra (Mar), Goa district (Trincoma].ee district), VLikagma (Valikmam), Skaratittha (ttuai or Kayts), Gonusu district (K1avpi region), Madhupdapatittha (possibly Iluppaikkaavai), Pulacceri and Deb rap4an The last two laces have not been identified. It is not known whether Debarapaan is a variant of Dem4apaanama, by which name Jaffna was sometimes called, as evidenced by the Naxnpota On the other hand, Debarapaan and Pulacceri may well be places in the Eastern province. The element ceri in the latter name suggests that this is a Tamil name. In the light of this account in the Sinhalese and Phi sources we may accept some of the statements in the Tamil chronicles. The information in the Mafta_4appu-nfrniyam that Ngha bad in his army }ukkuva mercenaries who were given

1.

v., 83:15-17 ; Pv., p.116 ;

p. 3. For the
. cit.,

identification of the p]. ce names, see C.W.Nicholas,

pp. k5, k6, 1, 8 1f, 86. Kklaya is sometimes i entified with

Kavuv4u.
2. ampota, p. 5.

35
va (chieftaincies) may be based on some genuine tradition. The Nukkuvas, as we shall setin the se uel, were from Ker4a. The fortifications in the Trincomalee district, Kantay and Koiyram, all in the Eastern Province, were doubtless in the hands of mercenary leaders who probably became chieftains of those regions. Parts of the present Battica.loa district may have been occupied in this manner by }1gha(s mercenaries, among whom there may have been ?lukkuvas. The creation of the of Trincomalee may be related to the establishment of the fortification in hhat district and may not be an independent event connected with the varam temple. Later tradition

may have separated it from the general story of the establishm nt of chieftaincies all over northern and. eastern Ceylon and treated it with special si nificance owing to the connections of the chief of Trincomalee with the temple of varam.

Similarly, some of the statements in the Vaiy, Vaiypal and the i regarding the settlement of the

Vaiyars and other castes may be accepted. The list of fortifications established by } ha1 s soldiers clearly indicate their control of the areas which later became the Tamil Vanni istrict . Many of the new Tamil settlement sites mentioned in the V i7 and the Vaiypal are to be located in the istricts mentioned in the Sinhale e sources as the areas where lgha's

Ker4a and Tamil soldiers had established fortifications. Mujiyav4ai, Kaiukki, Taikkal, Vapp4ai and Kruviikki are in the Kurundi region. Kiakkumalai and Noccimai are in the Pad! region. Tiriyy and Katukku.am Pattu are in the Kklaya-Kavuvu.0 regions. Tiru-kamalai, Vervki1, Tampalcmm and Koiy.ram are in the Goa-Kothasra regions. These areas form a major part of the Vanni districts where, according to the a-vaipava-mlai, Kuakkta, settled Vaiyars.

If we discount the details provided in the Vaiy and the Vaiypal, which we are in no position to confirm except to say that the settler-castes enumerated in these works were 1 pbab' found in those places in later times, we may not be wrong in concluding that several parts of the Vanni districta, especially those along the north-eastern coast from Kurund! in the north to Kohasra in the south, were occupied by the soldiers of 1Zgha and his associates. These soldiers established fortifications in these places and settled there. Their leaders probably invited more settlers from among thtir kith and kin on the mainland. The Vaniyars, Mukkuvas and other mercenary leaders appear to have become chieftains of these new settlements. Presumably they were appointed by Igha and his associates.

1. J.P.Lewis, A Manual of the anni Di tricts, pp. TO-

357
Probably Ki4akkan and possibly P la Vaiya, were among these associates. We know from the Sizihalese sources that Jayabhu was definitely one of them. All the settlements described in the Vaiy and the Vaiyp!tal may not have been established in the thirteenth century. The process of settlement which began in that century probably lasted till the fourteenth century. However, among the miin settlemntss established in the first half of the thirteenth century were those along the northeastern coast, namely in Kurund! (Kurunta11r), Eklaya (Kokkiy) Kavuvu.0 (Kaukk4am Pattu), Pad! (Padaviya), Goa (Trincomalee) and Kohasra (Kottiyram). Tamils had begun to settle in :1. most of these places in the eleveith century. The new immigrants would have helped to strengthen the older settlements and to establish the semi-independent Tamil chieftaincies. This much could be gleaned from the literary sources. The archaeological and place-name materials not only confirm this as far as the Tamil settlements are concerned but also point to the sudden occupation of the major part of the Nar and Vavuniy districts by Taniils. The majority of the place names in these districts are Tamil - a feature which is in contrast with the place names of Jaffna. The number of Tamilised

1. See supra, .ii

358
Sinhalese names is very small. These names are mainly confined to the coastal regions where peaceful penetration of Tamil settlers had begun earlier than the thirteenth century and, therefore the retention of Sinhalese elements in the place names could be explained. In the Vavuniy district, for instance, nearly eighty-two per cent are in Tamil. Three percent are Tamilised forms of Sinhalese names. About four per cent are Tamil-Sinhalese compounds. Less than two per cent are Sinhaieee names. Nearly nine per cent are of doubtful origin, where the constituent elements in the majority of the cases are common to both Sinhalese and Taniil. Less than one per cent of the names have the elements ikka and pulavu indicating Ker4a

association. The remarkable feature of the place names of the Vanni districts is the large number with the final element k4am, meaning tank or reservoir. In the Vavuniy district, for instance, nearly sixty-three per cent of the names end in k4arn This may mean that by far the majority of the small

1. The percentages refer only to village names. These have been calculated on the basis of the list of village names found in the following works : a) azetteer No.k9: Ceylon, Official Standard Names A proved by the U.S. Board on Geographie Names, Office of Geo raphy, Dept. of the Interior, ashin ton, D.C., 1960; b) J.P.Lewis, Manual of the Vanni istricts, .

35
settle ents of Dravidians that spread over the major part of ara$ha after the fall of Polonnaruva originated as peasant settle ents around the many tanks that were built during the time of the S&nhalese rulers. It also seems to indicate the original home of many of the new settlers in these areas, for the element ku.am is more common in the place names of ICer4a than in those of the Tamil country. In fact a large number of the names with the element k4ani occurring in Ceylon are to be found in Ker4a as we1l This may su gest that several of the newwsettlers in the Vanni regions hailed from Ker4a, The Sinhalese sources, as we have seen, repeatedly state that the lCer4a soldiers of Ngha played a prominent part in the confiscation of lands and the establishment of settlements. The Tamil

1. E.g., Periyak4am, Vppak$am, Kalk$am, Karuk4am, Kollak4am, !k4am, etc. There are also other names, without the element kt4am, which are common to both regions. A. comparative study of the place-names of Ker4a and the Northern and North-central Provinces is likely to yield much interesting information rel ting to Ker4a settlements in Ceylon.

3G0
chronicles, too, mention Nalaiyam or }alaiyakam (Kera.a) among the places from which settlers went to the is1axid 1taru.k1r, in Ker4a, is mentioned in the Vaiy, Vaiy5pa1, car-kalveu and the Tiri-kcal puram as the home of some of the settlers in the Vanni districs The tradition that they came from Marufk'tr was current among the Vaniyars even in the nineteenth century. J.PLLewis records this tradition in the followin manner:The Tamil Vaiyas are descendants of Vai chiefs. The local exp'anation of the origin of this caste is that they are descendants of the chiefs (Palaya aus) who ca e over from Murithkr in India, and became rulers of the Vai. 3 The na e 1uru.kUr is evidently a corruption of ruikr. Further, the Mukkuva mercenaries who figure prominently in the Na akkajappu-m iyam a ong the soldiers of Ngha were from Ker4a, as we shall see presently. Apart from the indication that the place names provide about the original home of the new settlers, the absence of Sinhalese elements in the majority of the names may mean that the occu ation of the Vanni regions of northern

1. See upra, p. Zio 2. See upra, ,'. 2.D 3. J.P.Lewis, Nanual of the Van i Di tricts, p. 7.

361
Ceylon by the Dravidians was not as peaceful as it a ears to have been in the case of the Jaffna peninsula. In these res ects the p1 ce na es of the Vanni areas preserve valuable information that may well o a long way in confirming the statem nts in the Sinhalese sources that 1gha's army consited of many Ker4as, that these mercenaries occupied several villages and were settled here and there in Rjaratha and that their occupation was far from peaceful. The caution that we have to exercise in the use of to onymic materials which still await a proper examination prevents us from drawing any definite conclusions. But it should be stated that the evidence of these names and that of the literary sources point in the same direction. The archaeolo ical material, thou h disappointingly small, also seems to confirm some of the above points. In the Tamil Vanni districts only a few Dravidian-style aiva temples of the thirteenth century have been foun , Among these the temples at Tirukkvil, Kapuralla,and Nallatai-iakk m and the - 1 Saiva remains at Uruttirapurani and Kuruntanr are notable. These certainly indicate the existence of Tamul settlements in those places in the thirteenth century. But monumental

1. S.Paranavitana, 'Archaeolo ical Summary', C.J.Sc. II, p.160-i 1; A.S.C.A. . for 1933, p. 19 ; A.S.C.A.R. for 1907, p. 27 ; A.S.C.A. . for 195 , p. 0.

U')", reiriains of a different type attest to the destruction wrou lit by the invaders and the conversion of Buddhist institutions into places of aiva worship, effected by the new settlers, thus confirming thee of the statements in the Sinhalese sources. The many scattered ruins of Bud hist monateries and temples all over the Vanni region preserve the memory of the Sinhalese Buddhist settlements that once covere these parts. Several of the pilimag!s (image houses) attached to monasteries in places
like Kvilkdu, Iikai, mant al, Kaakarya-k4am, Ircnt frar-

ki4am, Ciappvaraca.k4am and I'iaukanda were converted into aiva temples, often dedicated to Gaea Bud ha images or inscribed slabs from the Buddhist structures were used to make the Gaea statues A number of small aiva shrines have been found in association with Buddhist reniins The destruction of several of the Buddhist edifices and the conversion of pi1imags into aiva temples may have begun in the time of IAgba. In the North-central Province, too, we find evidence of such activities. On the Ninnrya Road, close to Polonnaruva, were discovered a few aiva edifices which were built of materials from Buddhist

1. J.P.L.ewia, Manual of the Vanni istricts, pp. 297, 30 -3 6, 311. 2. Ibid. , pp. 297, 303, 3. I id.

306.

363
. 1 structures. A door-jamb from one of the aiva shrines there was found to bear part of an inscription of Parkramabhu I A broken pillar sha't with Sinhalese writing of the tenth century was recovered from the enclosing wall of 4$other shrLne In one of the Viu temples of Polonnaruva, fragments of Niaksi Nalla's stone inscriptions were foundtl In the same place, two fragments of a broken pillar with Sinhalese writing of about the tenth century served as steps to one of the Vaiava shrines A pillar in the ma4 of iva D!vl! No.5 at Polonnaruva as diecdwered with a Sinhalese inscription of the eleventh century on it In iva Dv1 No.7 a square stone sana with an inscription of Niaka Nafl.a was used as a base for a ].izga? Another of the Saiva shrines unearthed at Polonnaruva yielded a pillar with a Sinhalese inscription of Jayabhu I These examples leave us in no doubt that materials from Buddhist structures were used in the building of aiva and Vaiava temples. The date of most

1. A. .C.A. . for 19 2, pp. 7-13. 2. Ibid., p. 7. 3. Ibid., p. 11. . A.S.C.A. . for 19 8, p. 9. 5. A.S.C.A. . for 19 7, p. 8. 6. Ibid., p. 5. 7. A. .C.A. for 19 , p. 11. 8. A. . .A.I. for 193k, p. 16.

364
of the inscriptions found on the pillars anti slabs is the twelfth century. The date of the construction of these aiva and Vaiava shrines is cert inly later than that. It is not possible to surmise that these were built before 1212, when Buddhist rulers were on the thorne. Nor is it possible to date them after lgha was ousted from Polonnaruva, for with that ecent this city appears to have been abandoned by the 1.ligas, Tamils and Ker4as althoIi it is possible that some of the settlers continued to be there even after that. In all probability the destruction of Buddhist edifices and the construction of several at least of the aiva and Vaiava shrines took place in the time of 1ha. In fact, this is the testimony of the Sinhalese and Tamil chronicles, too In the light of the examples at Polonnaruva we may not be wrong in surmising that some at least of the Saiva shrines found in association with Buddhi t remains in the Vavuniy district were the work of the invaders and new settlers in the time of 1 ha. It is possible that some were built of materials from an abandoned or ruined Buddhist structure at a later date. Some may have been converted into aiva temples at a time when the Buddhist po ulafion of the area ceased to exist due t either slow mi ration or assimilation

1. See sura,

pf...ci

363
1 to the Tamil population, as in Jaffna. The absence of Sinh lese lemente in the names of many of the laces where such temples have been found, which speaks against a ion survival of the Sinhalese o ulation in tho e places, may preclu e t e last possibilityin most of the cases. However, to some extent at least, the archaeological evidence may be said to confirm the information in the Sinhalese sources ab ut the occu ation of the Tainil knd Ker4a mercenaries in the time of Ngha. The main sequence of events that emerges from the different types of evidence that we have discussed may be summarised now. Till about the tenth century A.D. the Vanni regions of Vavuniy, Trincomalee, Nul1aitvu and Batticaloa were almost entirely populated by Sinhalese. By about the beginning of the elventh century Tainil settlements were established along the eastern coast nei hbouring Vavuniy district. In the twelfth century there were notable Tamil settlements in the area from Kurund! (Kuruntar) in the north to Trinco lee in the south as far west as Padaviya and Kantay. These were, however, scattered settlements. In the thirteenth century, with the invasion of }gha, T mil and Ker4a mercenaries occupied several parts of these istricts, particularly alon the eastern

1. See upra,

36
coast. Vaniy rs and Mukkuvas a ear to have been prominent a ong these mercenaries. There seems to have been a visible dislodgement of the Sinhalese po ulation from the Vavuniy and }ullaitvu districts from this time. Some uddhist structures were probably destroyed and aiva temples built in their places. The conquered parts of northern and eastern Ceylon were probably controlled by mercenary leaders. Presumably they were appointed as chiefs by ha and his associates. These principalities were

the Vanni chieftaincies which later owed allegiance to either the Sinhalese ruler in the south or the Tamil king of Jaffna. These Vanni chiefs appear to have invited settlers from South India and strengthened the Ker4a and Tamil elements in the local population. This process of settlement may have gone on well into the fourteenth century. In these Vanni districts, the areas of Dravidian settlement in the thirteenth century seem to have been mainly confined to the Vavuniy, NullaitTvu and Trincomalee districts. Neither the chronicles of Jaffna nor the Makkaappu-mmiyam refer to extensive Tamil-Ker4a settlements in the Batticaloa district or in the Nar district in the thirteenth century. As we have seen earlier, there were a few Ca stron holds in 1 the Batticaloa district in the eleventh century. Among these

1. See supra, p.j

36/
Chaggxna (Skmam)finda mention in the C1lavaisa The others are nt named. No inscription in Tami]. belonging to the Ca period or, for that matter, to the twelfth century, has been discovered here. Of the fortifications set up by the soldiers of Ngha and Jayabhu, none is to be located in the Batticaloa district. There are at least two of these fortifications which have not been identified, namely Pulacceri and Debara atan, and possibly these are to be located in the Batticaloa region. A iva temple in the P4ya style of architecture has been at Tirukkvil, a few miles from Sk.mam. On grounds of style this temple has been dated o the thirteenth cezitury In the traditions preserved in the Naakk4appu-miyam this temple is associated with 1gha Probably it was built in the reign of this ruler. There is another iva temple at ICa uralla datable to about the same period Except for these, no significant archaeological materials or inscriptions indicating DraTidian settlements in the Batticaloa district in the thirte nth century have been found. All the Tainil inscriptions of the Batticaloa

1. See supra, p. 2. S.Paranavitana, 'Archaeolo ical Summary,' C.J.Sc. (G), II, pp. 160-161.

3 Lee
k. A.S.C.A.

------- ii' r 7c for 1933,

p. 19.

368
area are of a later date. The place names here are largely Tamilised forms of earlier Sinhalese names, as in Jaffna. But unlike in the latter place, the earlier forms are readily recognizable in these names. This seems to indicate the relatively late date when the Tamilisation occurred, It would appear that the Batticaloa district was not extensively settled by Dravidians in the thirteenth century although }lAgha's mercenaries seem to have occupied the area and becoae its chieftains. Extensive Dravidian settlements here were probably established after the thirteenth century. In the Maakk4appu-nfrniyam the Mukkuvas figure prominently among the mercenaries who were given chieftaincies in Batticaloa by Zgha The Mukkuvas are an influential and strong caste among the Taniils of Batticaloa in the present day. Members of the Mukkuva caste are also found in the Jaffna, Vavuniy, MullaitTu, Mar and Puttalaa districts. n analysis of the social organization and legal institutions of this caste has shown that there exists a close affinity between these Nukkuvaa and the Ker4as of South India. The Mukkuva law, which forms a separate code in the traditional law of the Tamils of Ceylon, is largely based on the Narumakkattyam law of the Ker4

1. See supra,

p.3'C

2. Cf., K. .Thambiah, The Laws and ustoms of the Tamils of Jaffria, pp. 8-12 ;C.Brito, Mukkuva Law, Colombo, 1872.

369
The Ker4a origin of this caste is further confirmed by the fact that the only area in South India where we find the Mukkuva caste now is the Na].ayam-speRking westora]. littoral The name Mukkuva, too, is of Malayjam origin, as we bhafl see in the sequel. According to some traditions in Ker4a, the Nukkuvas, like the Tiyars and avars of Kera3a, originated imihigrated there from Ceylon But there are some other traditions which claim that the Mukkuvas are the only indigenous people of Ker4a The Mukkuvas, being a fishing caste, may have maintained close and continuous cont cts with the coastal areas of Ceylon and this may have given rise to a tradition in later times that they migrated from Ceylon. The traditions among the Mukkuvaa of Ceylon regarding their and date of their migration are rather late and are clouded by attempts to enhance their social position among the Tamils. One of these is the attempt to relate their ancestry to Kuga, the ferryman who ap ears in the Rnyaa as a friend of Rma. This is based on the final syllable of the name Mukkukar (a variant of I4ukkuvar), namely

1. C.A.Menon, Cochin State Manual, (1911), p.10k; J.Sturrock, Manual of South Canara, (189k), p . 169-170. 2. C.A.Menon, . cit., p. 20k.

L.Yore, Nalabar Law an Custom, p. ]..

370
kukar (equated with Kuka = Skt. Kuga) Such attempts at claiming descent from personalities appearing in the epics are co among some castes of South India and Cey1on The name Mukkuvar is to be derived from the Malayam nkkuka, 'to dive' (Tamil Ma1ay4am zm4ukuka>
Ofl

rnukkuka) Mukkukan literally means a divert Probably the Mukkuvar were a caste pearl-divers who later took to fishing. Some of the early foreign notices of this community confirm their maritime profession. The Italianatravel].er Vartheina (1510) and the Portuguese writers Correa (1525) and Barros (1552) refer to the
Nukkuvar of

Ker4a as fishermen But it appears

that sometimes the term mukkuvar had a general application

1.

S.Casie

Chetty, Manual of the Putta].am District, p.2k. , the example of the Kurukulas who claim

2. See supra, p. tic>

descent from the Kurus of the Mahbhrata. 3. k. . urrow and LB.Emeneau, Dravidian Ety ologic 1 Dictionary, (1961), pp. 337-338. H.Yule,and A.C. urnell, Hobson-Jobson, London 1903, p. 592. 5. L.Varthema, The Travels of Varthema, Tr. T.W.Jones, (1863), p.1k2& 'The fourth class are called Yiechua, and these are fishermen'; E.J.Stanley, Thre Voyages of Vasco a Gama and is Viceroyalty, fro the Len as In i of spar Correa, (1869), p. :

'Nacuas which are fishermen'; J. de Barros, 'Decadas de Asia, etc., Lisbon 1778 : 'ucuaria, a fisherman's village', quoted in Hobson-Jbsoh, . cit., p. 592.

371
meanjn mariners or boatmen As a sea-faring community they were considered to be a low caste There is little reliable information in our sources regarding the migration of the Mukkuvar from South India. They do not find mention in our sources before the thirteenth century. The earliest known reference is in the Dabadei-asna, where the Mukkuva mercenaries of the time of Parkramabhu II are mentioned The next references to them are to be found in a number of Sinhalese and Tamil works of later times. The most important among these are the Mukkaru-hatana, Vanni-upata, RLiv4i-kathva and the Uarata-vitti and the iamil chronicles of Jaffna and Batticaloa. In the Sinhalese works, there is an account of an invasion by a people called the Kka Mukkaru, identifiable with the Mukkuvar, in the time of Bhtika Tissa
(lkO-l6k)

According to this account, Kka Mukkar was a Tamil

1. Pyrard de Lava]., Discours du Voyage des Francai aux Indes Orientales, I, (1887),

p.

3]Jf;

'These mariners are called Noucois';

2. A.T.Pringle, The iary and Consultation Book of the Agent, Governor an Council of Fort St.George, 1st Series, III, p. 131: 'Naquas or boatmen'. 2. I4ahuan (A.D.].k09): 'The liukkuvas, the lowest and poorest of all', quoted by V.Nagam Aiya, Travancore State !anua1, I, (1906), p.280. 3. Daffi a ei-asna, p. 1i.
k.

Cf., Vanni-upata, (Cob bo Yiuseum }ianuscript), p. 15.

372
who , with his chief N4a Nuda].iy and a host of Tamils, appropriated the territory between the Ka1 Oya and the I Oya. These Tamils are also called the Kka !kkaru. The above area was peopled by Tamils, chief among whom were the }ukkuvar, in the time of Portuguese rule. The account of the invasion of the Kka Mukkaru is not found in any of the earlier Sinhalese chronicles. It appears to have been based on a later event connected with the settlement of the 1Iukkuvar in the northwestern coast. In the !a-vaipava-mlai, 'the fishers caUed Mukkukar' are stated to have been settled in the ports and coastal regions (karai-tu .aik4) of the Jaffna peninsula in the time of Pau (k33-k38). On account of their disrespect for and defilement of aiva temples, it is said, they were punished by Pa4u and driven away to Batticaloa where they settled in places likePPakai (PThama) and Valaiyiavu Ucuma, and Cnta,, are mentioned as two of the Nukkuva chiefs who were punished in this manner and who later established Nu.kkuva settlements in Batticaloa. It is also stated that the places Ucuma-tuLai and Cnta-k4am in Jaffna were two of the )iukkuva settlements that were abandoned This story in the Y ia-vaip ava- n1 ]. i

has certain similarities with the story of two Mukkuva chiefs

1. Yvm., pp. 9-10. 2. Ibid., P . 10.

373
narrated in the Vaiy and the Vaiypal. According to this story, a Pya ruler, desirous of obtaining a Nga diamong from Ceylon for the enklet of Kaaki (sic), the heroine of the epic Cilappatikrant, sent a chief of the Karaiyr conuuunity called ?1ma to Ceylon. The latter defeated two Mukkiava chiefs, V4i-araca, and r-araca, and obtained the Iga diamond. The defeated Mukkuva chief went to Batticalo'and settled there while the other chief settled in Vi$attal-tvu, also in the Batticaloa district The account in the is apparently a later fabrication based mn the story in the other two chronicles. The author of the I a-vaipav -mlai

has carefully discarded the story of the Pya ruler obtaining a diamond for the p nklet of Kaaki, an obvious contradiction with the account in the Cilappatikram. Instead of the

P4a

ruler and Nkma, he has introduced the Ceylonese ruler Pa4ii into his story. In place of Veti-araca and ?1r-araca, the two pirates whose memory is still preserved in the folk tradition of De].ft, Puttalani and Batticaloa, he has introduced two other Nukkuva chiefs called Ucunian 2 and Cnta. These two characters, as Gnanapragasar has an ested, are probably creations of folk-etymolo ists based on the place names Ucnma-tuai and

1.

.,

vv.53-55.

2. Ucuzna seems to be a Tainilised form of the Muslim name Usman and many be associated with Muslim tra era at this port, i.e. Ucunia-tuaj (Port of Ucuma).

374
Cnta-k4am Furtherri the pa-v ipava-nilai states that it was Pau's queen who sent troops to prevent K$akkzfta from building a temple at Kvaraii& But the Trincomalee chronicles have it that Iakacavuntari, who later married Kuakka, was the queen who attempted to prevent the work of K4aia at Trincomalee Presumably the author of the ppa-

vaipava-nlai was drawing from several sources when f bricating the story of the Mukkuva settlements in Jaffna and Batticaloa. This account has little claim on our confidence. The account in the Vaiy and the Vaiypal is clearly based on the popular tradition about the l4ukkuva pirate Jr-araca, sometimes known as Veti-araca, and is probably related to a later peri.od. In the N kka4appu-nmiyam , the Mukkuvar are

referred as those of the Kuka ku]A, the origin of which tradition i Th*'r we have aire dy explained.. '-lire said to have been military leaders under 1gha. Their place of on in is given as Kikaam, which may be a distorted form of Kikktu (Calicut) in Kera.a. This tradition seems to preserve some element of truth in it, for in the thirteenth century we hear from other sources of the

1. S.Gnanapragasar, 2. Iv., p. 10.

_vaipava_ka11ni.r,a, p. 5.

3. 4.

., 'Tiruk$a.kaa Patalam', v.2 ff. -.--

Dr s-.... ,y.109.

375
presence of Ilukkuva mercenaries in Ceylon The Mukkuvaa being Ker4as were probably among the Ker4a mercenaries of }ga. As we have already suggested, this tradition may well be accepted as true. There are also other traditions among the }u.kkuvas of Ceylon regarding their migration from India. There is one recorded by Casie Chetty about the Nukkuva settlement under Vei-araca This relates to the western and will be discussed later. Veti-araca may have been a historical personality whose memory has been perpetuated in the folk traditions of the Nukkuvas He probably belonged to a period later than the thriteenth century. Front the foregoing discussion it ap ears that the }ukkuva settlement of Batticaloa began in the thirteenth century. Other Ker4a and Taniil mercenaries of ?gha and other invaders may also have settled there at this time. But it does not appear that the Battialo" 1district bad extensive Dravidian settlements before the fourteenth century. Probab]4 widespread settlements of Tamils and Ker4as in this district took place in the fourteenth

1. Dabadei-asn_, p. 2. See infra, p.33D ; S.Casie Chetty, Ceylon Gazetteer, p. 278. 3. S .Gnauapra,gasar, -vaipava-vimarcan, p. k.

37i
and fifteenth centuries, when for the first time Tarnil inscriptions become available there. There is a dearth of evideace regarding the Dravidian settlements on the western coast in the thirteenth century. The Taniil-speaking region from Mar in the north to Chilaw in the south has no independent chronicle similar to those of Jaffna, Batticaloa and Trincomalee. The ancient temples of Tiru-ktvarani and Nuvarain on this coast do not seem to have possessed any early chronilce or pura. The Muvaram temple has a chronicle entitled i vara-nmiyam which is

of recent origiJ Some of the earlier sections of this work are based on the traditions of the K&varam temple. Although no formal chronicle of events has been preserved in this region, there are still folk traditions concerning the Ker4a and Tamil settlements there. Some of these have been collected and recorded by British civil servants in the nineteenth century As in the case of most folk traditions, these lack a pro er chronology. This has resulted in a confusion early and later traditions. and it is difficult to extricate the genuine traditions from

Z. F.Modder, A Manual of the Puttalam District, Colombo F.)dder, azetteer of the Putta].am District, Colombo 1908; S.Casie Chetty, Ceylon Gazetteer, Colombo, 183k.

37'
the rest. Consequently we are not in a position to reconstruct a satisfactory account of the Dravidian settlements on the western littoral in the thirteenth century. The island of Mar and the coastal districts of ?ntai and Puttalam were formed into Vanni chieftaincies in the time of the Jaffna kingd. These owed allegiance to the rulers of Jaffna. According to ueyroz, Putela (Puttalam) and Nantota (Itoa or }ntai) were two of the kinglete into which the territories outside the beginning of the kin dom were djvide& at the period (beginning of the fifteenth century)

He mentions Puttalam among the smaller principalities rled by the Vane.z (Vaiyrs) in the sixteenth century This is confirmed by a copper sannasa of Bhuvanekabhu VII dated in the Saka year 111.69 (A.D. 1511 .7) accoHing to which a Mukkava chieftain called Navaratna Vanniyl was ruling in the Puttalam region with his residence at Lunuvila He owed allegiance to the ruler at and not to the ruler of Jaffna. However, it

appears that it was generally considered that on the western coast Chil&6 (Chilaw) was the southern-most point of the 'lands of the Vani (Vanni , which belong to the Kingdom of Jafanapata'

1. F.de Queyroz, The Temporal an Spiritual Conquest of Ceylon, Tr. S.G.Perera, I ,( l 930, p. 2. Ibid.
3. Casie

32.

Chetty, Ceylon Gazetteer, pp. 190-191.

373
(Jaffnapatam, now Jaffria) In the Portuguese period, 'from

Nigumbo [Negonibo) to Jafanapata they speak the TanLil language better' To this day this holds good to some extent. There are still Tamil speakers in Chilaw and Negombo, many of whom also speak Sinhalese. As we have already seen , the settle ent of the Dravidians on the western coast began at a very early ate There were sporadic settlements at places like Pomparippu and Mahtittha before the Ca conquest. Tamil inscriptions attest to the settlement at Maht.ttha in the eleventh century. Inscriptions of the twelfth century reveal the existence of Tamil settlements in the region bordering on the Puttalam district. But there is not sufficient evidence to warrant the conclusion that the whole of the western coast from Nanr to Chilaw was settled by Tamils or Ker4as by about the twelfth century. In the thirteenth century, the soldiers of lgha and Jayabhu are stated to have set up fortifications in nmatta, ?4annra,and }ahtittha According to the Maakkaj.appu-

1. F. de ueyroz, 2. Thid., p. li6. 3. See supra, ppL 1k. See supra,

. cit., p. k7.

if2l

2p1-

5. See supra, p. 3c1.

37
xniyam, the Xaliga ruler (!gha) gave the principality of Naiu (I4qr) to 'those who bore the flag of the Tiru-kula vamnicam' (Kurukula vaziaa) As in the other parts of northern Ceylon, }gha and his associates probably created chieftaincies in the Mar district, too. The Kurukulas who are the leading caste in this region were probably among the Tainils who settled there in the thirteenth century. As mentioned earlier, some Sinhalese works of later times and traditions of the Puttalani district refer to the establishment of Nukkuva settlements on the western littoral. According to the Sinhalese wors, the area between the Kalya and the Mya was occupied by the Kk Mukkar (Mukkuvar) in the time of Bhtika Tissa As we have pointed out, it is unlikely that the
I4ukkuva

settlement on this coast

originated at such an early date. The Sinhalese works seem to refer to a later event. A tradition recorded by Casie Chetty seems to preserve a germ of truth regarding a !4ukkuva settlement in the Puttalam district. It runs as follows:-

1. ! P . 75. 2. See upra, p.

?j11.

330
When the coast of Na].abar was overrun by the Muhainmadana from Arabia, the natives were persecuted, with the view of causing them to embrace the doctrines of the Koran; in order to avoid which the Nukwas transported themselves to Ceylon, and established their residence in the Malabar Provinces amil provinceJ. It appears that the place where the Mukwas first landed was Kudramalai, whence they emigrated to other parts of the Island, and in course of time formed several settlements. Some time after the arrival of the Mukwas in the District [PuttaiamJ, their chieftain named Vedi Arasan, had to contend wilh a rival called Manikk Taleivan Mikka Talaiva4, who them yresided over the people, denominated Karaiyr CKurukul , and possessed a settlement on the south side of the District. Nanikka Taleivan despatched some of his officers to Vedi Arasan for the purpose of soliciting his daughter in marriage, but, meeting with a refusal, he collected a considerable body of armed men and declared war against the Mukwas, threatening their total destruction. As the Mukwas were at that time a weak and defenceless people, they concerted with a crew of an Arab vessel which was then at anchor at Kudramalai, and with their assistance slew the rival chieftain and put all his troops to flight. In turn for the service rendered them by the Arabs, the whole of the Mukwas embraced the I4uhammadan religion which many of their descendants renounced in favour of Christianity, through the influence of the Portuguese. After the defeat of the Karaiyars, the Nukwas determined to send an embassy to the court of the emperor in order to ingratiate themselves into his favour. They accordingly made choice of certain individuals for the pnrpose and despatched them to Sitawaka with many costly presents. 'When these delegates reached the capital and presented themselves to the emperor, he received them with uncommon kindness, and granted them several copper sannasas or receipts, whereby the land in the whole District of Puttalamaand Kalpentyn were allotted to them for their maintenance as paraveni Besides the assignment of land, the emperor constituted a royal tribunal at Puttalam called Mutrakudam, and appointed ei hteen of the Mukwas to be members of the same, under the authority of a Dissawa or Pro-Consul, who was to be annually sent from the court ; and also conferred on the said members the title of Wanniya, ......l

L. S.Casie Chetty, Ceylon Gazetteer, p. 278.

381
This tradition seems to refer to a ukkuva settlement of the ItT period (].k15-1505). It ap ears that the Mukkuvar had to contend with the Kurukulas who were settled along the Puttalam-Chilaw coast in that period. The Vanni chieftaincies of the Mukkuvar in the western coast probably originated in the Sinhalese rulers of period. The

seem to have been acknowledged by them

as their overlords, although sometimes the Jaffna rulers, too, seem to have claimed suzerainty over them. The

Kf

rulers

appear to have assigned lands as parave1 to the Mukkuvar in the Puttalam district, We know of at least one instance when a ruler, Bhuvaneabhu VII, granted the region of Pomparippu to the Mukkuva chieftain Navaratna Vanniya as his paravei in 15

7.

The reference to the Muslim invasion of Ker4a and the subsequent migration of the Mukkuvas from there may preserve the memory of an earlier wave of Mukkuva mi ration in the fourteenth century when the Muslims invaded South India. It is not known whether the Nukkuva migration to the westrn coast of Ceylon had begun earlier than the fourteenth century. In the thirteenth century the invasion of }gha seems to have led to the occupation of this part of Ceylon, too, by South Indian mercenaries. It is stated in the C1avaisa th t Dami. warriors dwelt as they

1. See supra, p. 377.

382
pleased in sin le villages in Iyratha and that Vijayabhu III drove them a ay from there The Puttalazn-Chilaw region, which formed part of Nyraha, was under the direct rule of the Sinhalese rulers in DaThadeiya and it is doubtful that there ercTamil chieftaincies in that region during the thirteenth century. The place-name evidence in this coastal region claarly indicates that at one time a large part of the present Puttalam district was occupied by Tamils. The high percentage of Tamil names along the coast may mean that there was a concentration of Tamils there. One of the revenue divisions of this district still continues to be called Demaa Hatpattu (Seven Tamil Divisions) although a large section of this division is now occupied by Sinhalese speakers. Traditions in this area preserve the memory of Tamil chiefs having ruled in the 2 Ravanni and Kunravanni Pattus of the Dema.a Hatpattu. This was probably after the thirteenth century. The vara-znmiyam contains a detailed account

of the Tamil settlements established by K4akk t a in the region of this account, varam, in the Chilaw district. According to akka after having completed the renovation

of the temple of Kvaram went to Nu.varam in the Kali

].. V.,

l:J.k.

2. S.Casie Chetty, eylon azetteer, p. 86.

383
year 512 (2590 .c.) and undertook the renovation of the Mivaram temple. After the completion of this work, the Brhaaa It1akaa ivcriyr, hi8 wife Vislaki AmriAj. and several learn d Br.haas were invited from the Ca country to con uct the Kunibhb1i.eka festival. In order to ensure the continuance of the various services in the temple, Kujakka decided to invite settlers from South India. He, therefore, went to places like Nadurai, To4aimyalam, Eraikkl, Tiruccirppa.i, K'alUr NarufikIr, selected people from among the Pirmyar (BrThmaas), Caivar (aivas), Ceis, Ve.3ar (cultivators), VTra-niuti Cakmr (a class of VTra aivas), Ttar (Vaiava mendicants)of the Sdra caste), Kollar (blacksmiths), Kar (braziers), Tab (goldsmiths), Cipar (sculptors), Taccar (carpenters), Ypi (minstrels), Eai-viyar (oil mongers), Akampaiyr (Agampai mercenaries or servants of the inner apartments), Nul1ai-Naapp4iyr, Caruku-Maappa4iyr, CakuMaappa U iyr1 ,

Kaikkar (weavers, also temple officials and

C

Loldiers ), Cniyar

a class of weavers), Ilai-viyar (sellers

of betel-leaves), Viaku-vei (wood-cutters), Ttar (me sengers) vitar (barbers), Va4r (washermen), Timilar (boatmen), Valaifiar (caste of fishers), Varua Kulattr (those of the Va.ra kula),

1. See supra, p. 2' 2. See supra, p. pt. 2, p. 116. ; Travancore Archaeological Series, VI,

381
Kuyavar (potters), I4aavar (Maava tribesmen), Paflar (a low caste), Kattikkrar (swordsmen) and Paaiyar (drummers). These immigrants were assigned different services in the temple and were granted lands in the vicinity of the temple for their settlement.and maintenance. A nobleman from Madurai, named was appointed as their chieftain. The lands belonging to the temple were divided thnto sixty-four villages. Twelve officials from among the Cutta Ve

.ar (Pure Ve3as) and the AkRmpai

Ve3ar were appointed at Nuvaram and in ten of the villages, namely Pampa3ai (Pambala), Pakala-pirmaa-t4uai (PahalabrThmaa-daluva), Kokkvil (Kokvi.la), Tampakal (Tabagala), Curuvela (Suruvela), Pajamai (Pajama), Takampavai (Debambava), Naalai (Ma4alna), Miikkt4ani (?4inikk4ama) and tikampitti (Uhampiiya). The two who were assigned duties at Muivaram were conferred the titles of Cantira-okara-mutali Pattakai (Sinh, Paabdi) and Pak4ti Kr.ai (Sinh. Paabdi Kra) while the others were conferred the title of AttukkZr. i (Sinh. Atukrja). The assignment of villages for the settlement of the different castes was as follows:- a) The Akampati Ve3ar were settled at Elivei (Elivia),

Kkkppa3i

(Kk pa3iya),

Maavari (ianavari), Karavei (Karavita), Ka Thala-pirmaia-t 4uvai (Ihala-brhmaa-daluva),

ikj

(Gaanke),

} TTMtil vet tuva

385
(Mugwiuvaavana), Villattavai (Vilattava),
M

alai (Maalna),

V!ra-koripan-t4uvai (Vra-komban-daluva), Pirap aiik4i (Pirappak4iya), Ol].jtt4uvai (Olidaluva) ii Marutafikujam (Narudankulania) an Tittakkaai (Tittakae); b) the Vrami Params were settled at Pariyni1lai in Nivaram; c) the Ciperis were settled

in the southern part of Nuvararn; d) the blacksmiths were

settled in Karaveti (Karavia) and Va.kattaai (Bafigadeiya); e) the Taccar were settled in Va..kattaai (Bagadeiya); f) the Kar were settled in F'lakk4am (Plakuama) and Ca].pain (Chilaw); g) the 1 akrar (drummers) and the Tvatcis (temple dancers) were settled in Nuvaram and Chilaw; h) the Kuyavar were settled in Vatakal ].ai (V 4akahagala ?) and Nu!svaram; i) the Cua-viyar (lime sellers) were settled in Narave.i (Maravila); j) the Kaikkar were settled
in

Chilaw and Nu6varam;

in

k) the K5trikkrar (wood-cutters) were settled

Pakalakamam

(Pahalagama) and Po.ki (Ponnankanniya); 1) the Timilar were settled in Tmm.i].ai (Timilla); in) the Cr were settled in Maakk4am; n) the Karuppakkaikkrar (jag ery makers) were

Iikotav4i (Inigoavela); o) the Ca.ktis (conch blowers) were ettle in Kkkp ai (Kkpafl.iya), Iluppatai (Iluppadetiya), Ciy nip lakasv4i (Siy balagasvel ), Karukkuji (Karukki4iy )
settled
in

an T,*nive.i (Di amy 1 ); p) the Nlai-kati (the arland makers) were settled in uvaram and Timilai ( imilla); q) the Var were settle in Mukant4uvai (Mnngandaluia) and Cempnkketti

38
(Sembukattiya); r)the Paaiyar were settled in rapiya (V!rap4iyana) The above account of the var#iTiyam is in

many respects untrustworthy. In the first place, it has fla rantly incorporated traditions from the Trincomalee chronicles and claims them as those of the uuvaram temple. The renovation of the temple by K4ak the invitation of different castes

from South India, the assi nment of different duties to them and their settlement in several villages surroundin the temple, the invitation of Tai- A-ptT from Nadurai and his

appointment as their chieftain are elements from the Trincomalee chronicles. Secondly, some other elements h ye been borrowed from the Jaffna chronicles. For instance, the invitation of the BrThmaa lakaa Sivcriyr, the BrThinaa la y Vislaki Anun and other learned BrThmaas from the Ca country is found in the Ypp -vaipava-mlai in connection with the story of Vij ya The list of castes and the places from which they came is very similar to that found in the aiy and the Vaiyp1 in connection with the Dravidian settlements in other arts of the Vanni It appears that such pera nalities as Kuaka and the Brhmaia a ivcriyr have be en associated

1.
2. Yv.,

vara-mmiyam, pp. pp. k,6.

8-li.

3. See upra, p. 31i.

381
with Nu&varam in order to give it a special sanctity and anticjuity which it lacked in comparison with the temple pf Kivaram. The appointment of Atukras is something based on later knowledge. It is in the Sinhalese inscriptions of the fifteenth century that we first hear of the officials called Atukras Similarly the account of the settlement of the different castes in the villages adjoining Mivaram seems to be based on later knowled e. It is difficult to accept this as authentic. It is possible that South Indian families were invited to perform services in the temple and were settled in the lands belonging to the temple. But the Dravidian settlement at Munvarain, Chilaw and other nearby villages may not be the result bt a policy of settlement followed by the temple authorities. These settlements have to be treated as part of the Dravidian settlements along the whole coast from Iar to Chilaw. Thus, we see that neither the Linhalese traditions nor the Taxnil traditions help us to know anything definite about the Dravidian settlem nts alon the western coast. The evidence of place names, though indicating extensive settlements of Tamils along this coast, does not help us to fix the chronology. All that we can day now is that the settlements of Nuk.kuvar 1. U.C. . , I, pt. 2, p.739.

383
Kurukulas and other South Indians was a process that seems to have one on for a long period. New bands of inuni rants probably settled down on the western coast in the thirteenth century as well. As some of the traditions suggest, the invasion of South India by the 1usli*s probably further migration of }lukkuvar and Kurukulas in the fourteenth century. During the Kt period there were Taniil Vanni chieftaincies in this region, some of which were contr011ed by Nukkuvar. The Tamil chieftaincies of the Puttalam-Chilaw region seem to have owed allegiance to the rulers, although according to de Queyroz the lands

as far south as Chilaw belonged to the kings of Jaffna. This coastal region appears to have been a bone o contention between the Sinhalese and Tamil rulers owing to its in ortance in the control of the island's pearl fishery. The ________ vaipava-mlai refers to the struggle between Ceyav1raciMqii riya, one of the kings of Jaffna, and Bhuvanekabhu, probably the fourth of that na e, over the control of the pearl banks Another invasioa of ?yrattha, in which was included the Chilaw region, by the Tamils from the northern part of the island in the time of Parkramabhu IV (1302-1326) is alluded to in the }Iranbdda inscription The Tamil chieftains of the Chilaw-Puttalam

1. Yv., p. k23 S.Paranavitana, 'The 4rya Kingdom of North Ceylon', p.22

2. U.C. .C., I, pt. E, p. 63k, fu. 7k.

389
region may have been forced to change their alliances from time to time. In the thirteenth century probably there were no Tami]. chieftaincies in the Puttalam district. But probably there were such chieftaincies in the Nar district, here }gha and JayabThu had established fortifications. Outside the Northern and Eastern Provinces and the Puttalam district, there were Vanni chieftaincies in the Northcentral Province in the thirteenth century. This province, which formed the central part of Rjaratha, was the scene of much ravage and plunder during the occupation of Ngha. As we have seen earlier, it was this region which was most affected by the foreign invasions and the consequent drift of the Sinhalese population to the south-west. The claim of the Clavaisa that ?1gha's soldiers seized tillages, fields and houses in Rjaraha may not be wholly untrue. We have seen that there is some archaeological evidence in Polonnaruva and Minnriya which attest to the destruction of Bu dhist establishments and the building of Saiva shrines by the invaders. We h ye also pointed out that a good percentage of the place names in this province is Tamil in origin. Certain traditions of the Nuvarakalviya district of thi province indicate the aettl nt of T mu Vnniy s in that re 1Ofl The

Tmi1 an

Ker4a soldiers of N ha as well

1. See upra,

p.33

390
as the mercenaries of the later invaders like Chan rabhnu probably established settlements in parts of the North-central Province. The Clavawa has references to Sinhalese Vanni rinces ruling in Rj raha in the thirteenth century. Probably there were also Vanni chieft ins of the Taniils in this area at that time. It is difficult to determine with t e evidence at our disposal how extensive or strong the new Dravidian settle ents were in the thirteenth century. The North-central Province was lar ely abandoned after the thirteenth century by both Tamils and Sinhalese and only a few pockets of Tamil and Sinhalese settl ments seem to have been left. In the thirteenth century the Dravidians were probably scattered all over the province in small group. These settlers have not left behind Tamil inscriptions or Dravidian-style temples. The only evidence of their settlement is found in the local nomenclature. The disa earance of the earlier Sinhalese names s eaks both for a su den and a violent occupation of the area by the forei ners as well as for settle ent by Tamils after the total abandonment of the area by the earli r settlers. Th retention of the Taniil na es by the resent-day opulation which is largely Sinbalese- eakin , shows that the Ta is, howev r small in number, bad continu d to live in this province till the ti e of the Sinhale e resettlement-

1. .C. . , I, pt. , pp. 713-71k.

391
The wi e rea occurrence of Tamil p1 ce n s sugge te that the province w s at one time extensively settled by Taxnils. This was certainly not before the thirteenth century when the Northcentral Province was the heart of the Sinhalese kingdom. Tamil settlements on such a wide seal would not have taken lace long after the thirteenth century wh n this region was ab n oned to a great extent and was covered with jungle. The Tamii.l settle ents that were res onsible for many of the new lace names have to be dated to the thirteenth century. In conrse of ti e several of the settlers ap ear to have drifted to the Taniil kiagdom in the north, where there was soon a concentration of Tamil population. But small groups seem to have lingered on in the jungle viflages In the Sirthalese kizgdom of the southi there is little evidence of any Dravidian settlem nt

in the

thirteenth

century. The only evidence relating to the presence of Tamils is that concerning the mercenaries. Tamil ercenary forces continued to be in the service of Sinhalese rulers in the thirteenth century as in earlier times. ih aba e'i-as a refers to twelve thousand amil soldiers, doubtless an exagg rate number, who drew their pay from the p u]. chest in the tim of Parkramabhu 4ong theni were the Agamp i forces and

the Svlakkr s. 1.B.Ariyapa1a attempts to relate theto the

1.

vidence of this found in the adzninistr tion reports of the

Dutch an British erio s. 2. Da adei-asna. p.

392
Cava.aikkrar, a caste of weavers in the Tinnevelly district He also oints out the similarity betwe n the name Zvlakkra and the Tamil Cav4akkrar, a class of fishermen or ferrymen The Svlakkrar are not mentioned in the South Indian sources among the mercenary bodies. Perhaps they were a minor body of mercenarie In the fourteenth century, oo, there were Tamils in the army of the Sinhalese ings, as is evidenced by the Gaaldei inscription of Bbuvaneka Thu IV (l3kl-l35l) No Tarnil inscription of the thirteenth century has been found in the area covered by the Sinhalese kingdom. But Tamil inscriptions of the fourteenth, fifteenth and sixteenth centuries have been found in some places there. There is, however, an inscription written in the Sinhalese and Tamil scripts and datable to the thirteenth century, found at ?iysnkndura in the Kamdehna Tea Estate at Naniunukula, Badulla district The Tamil script has been

1. M.B.Ariyapapal, ociety in Medieval Ceylon, p. 161. 2. Ibid.

3. S.Paranavitana, 'Gaaldei ock Inscription of .bhuvanekabThu

IV',

r.z., Iv, p. 106.

k. A.S.C.A.R.

for

1952, p.

1f2.

393
used in the inscription for 'words specialising magical power' This is hardly evidence for any Tamil settlement in that area. In the unsettled conditions of the thirteenth century few Taznil settlers would have found their way into the Sinhalese kingdom which was at war with the foreign invaders. The only Tamil settlers there were probably the mercenaries. In the fourteenth and fifteenth centuries when better conditions prevailed in the south and new ports were epened up for foreign trade the South Indians who were affected by the Muslim invasions as well as Tamil traders seem to have settled in Ceylon, not only in the Tamil kingdom but in the ports and towns of the Sinhalese kingdom as well. This is attested to by the Tamil inscriptions, Sinhalese literary sources and Saiva arcliaeolo ical .2 remains.

1. A. .C.A.R. for

1952, p.k2. p.335;

4 . Tamil inscriptions: a) S.Paranavitana 'The Tamil Inscription on the Galle Trilingual Slab', (1kb), .Z., III,

b) In cription from Nyimmra, A.D. lk22, see S.Pa.ranavitana, Upulvan emple, Memoirs of the Archaeological urvey of Ceylon, VI,

pp. 71-75; inscriptions

pp. 189, 191, 212.

from Cob bo, Kalutara and Kurunigala

districts, S.Paranavitanal3 'Epigra hical Summary', C.J.Sc. (G), II,

iva te les: a) H.C.P.Be].l, Kigalla eport, pp. 63-65;

b) S.Paranavitana, 'Epigra hical Summary', .J.Sc. (G), II, p. 191;

394
c) S.Paranavitana, Upulvan Tem1e, . cit., p.75.

Sinhalese literary references to iva temples: a) Kkila Sanda, vv.00, .C.H.C., I, pt.2, pp. 768-769; Paravi anda, v.68, U.C. .C., I, pt. 2, p. 768; Slalihini Sanda, v. 22.

395
From the foregoing analysis it becomes clear that the settlements of the thirteenth century mark the most important stage in the course of the early Dravidian settlements in Ceylon. From the beginning of this century for nearly dtoadee decades a quick succession of foreign invasions, which brought into the island fresh mercenary forceB, led to the establishment of new Dravidian settlements. The nature of the invasions and of the settlements that followed was in many ways different from that of earlier invasions and settlements. While the earlier invasions, including even the Ca invasion of 1017, could be treated as episodes in the history of the island, the invasions of ?gha, Chandrabhnu and the P4yas in the thirteenth century cannot be dismissed as mere episodes. The settlements of th. earlier periods, though not quite unimpressive, did not result in the visible dislodgement of the Sinhalese population from any area. As far as we can see, those were not the result of forcible occupation of the lands of the Sinhalese. Those early settlers may have become assimilated to the Sinhalese population in due course. But it was the events of the thirteenth century that prevented such an assimilation in the greater part of northern and eastern Ceylon. The invasion of gha with the help of Ker4a and Tami]. mercenaries was far more violent than the earlier invasions. Its chief importance lies in the fact that it led to the permanent dislodgement of Sinhalese power from northern

393
Ceylon, the confiscation by Tamils and Ker4aa of lands and properties belonging to the Sinhalese and the consequent migration of the official class and many of the common peopleto the southwestern regions. These factors, more than any other, helped the transformation of northern Ceylon into a region occupied b Tami]. speakers and directly led to the foundation of a Tamil kingdom and several 2mi.]. Vanni cbieZti1nies tbere. It may be concluded that in the major part of the thirteenth century it was the Tamil and Ker4a mercenaries who founded. the new settlements. These were spread over a good part of Ijaraha. Once the kingdom o Uaffna and the Vanni chieftaincies were founded, it appears that Tamil rulers invited fimt11es from South India for settlement. Towards the end of the thirteenth and the beginning of the fourteenth century, therefore, a peaceful migration of settlers from the Tamil and Kera3a countries seem to have taken place. The prominent mention of Ker4a mercenaries in the Sinhalese and Tamil sources and the occurrence of Xer4a places among the original omes of the new settlers, as listed in the Tamil chronicles, indicate that there was a strong Ker4a element in the new population of the northern districts. This is also revealed by the similarity pr that exists between the social organizations of the Tamila of Ceylon and the Nalayis of Kera3a and the affinity between the

397
Marumakkattyarn laws of Kera.a and the Mukkuva and Th!sav4amii laws. The settlement of peaceful migrants seem to have been confined mainly to the nortbernmaet regions of the island. The difference in character between the settlements in the Jaffna district and those of the Vavuniy, Mullaitvu and

Nuvarakalviya districts is not only borne out by the evidence of the literary sources but is also demonstrated by the placename evidence. Whereas in the Jaffna district we come across a large percentage of place names with Sinhalese elements, in the local nomenclature of the other districts the Tamil element is predominant. The former indicates slow and peaceful penetration of the Tamils and the latter a violent and sudden occupation. The survival o Sinhalese place names, especially of Sinhalese territorial names, in Jaffna tells against a wholesale extermination or displacement of the Sinhalese living there. At the same time, Tamil names of estates denoting family settlement which are found scattered acroos the peninsula remarkably confirm the evidence of Taxnil chronicles regarding the settlement of prominent families from South India by the early kings of Jaffna. The settlements of the orth-central Province and of the major part of the North-western Province did not last long and soon there was a concentration of Tami].s in the Northern Province. The Trincomalee district of the Eastern Province and the Puttalam district of the North-.estern

398
were also areas bwere
la8tiflg

Tamil settlements were established.

Dravidian settlements in the Batticaloa district had also begun in the thirteenth century although the bulk of the settlers may have migrated to this district in later times.

399
CHAP!ER VI

THE BEGINNINGS OF THE XINGDOX OF JAFPNL - I )gha and Candrabhnu

After the drift of Sinhalese power to the southwestern parts of Ceylon, the only political authorities in the northern regions about rhom we hear in the Pli chronicle are the Vanni chieftaina If we had to depend solely on the Clavasa for information, we should conclude that after about the middle of the thirteenth century the whole of the former Rjaraha was split into minor chieftaincies under Vanni rulers who owed allegiance to the Sinhalese ruler at the capital in the south-western region. But the Rehla of Ibn Batuta the Sinhalese chronicle NiIya-sagrahaya3 and the Sinhalese poems Slalihinisanda Gir-sanda and Kkila-sand!a 6 as well as later chronicles like the Rjvaliya and the Tamil and Portuguese

1. Cv., 88:87, 89:51.

2. The Rebla of Ibn Batuta, Tr. Nandi Hussain, Baroda 1953, p. 217.

3.

p. 27.

k. Slalihini-sanda, ed. LC.Fernando, Moratuva 1956, v. 29.

5. Gir-sanda, ed. T.Sugathapala, Alutgama 192k, vv. 138_111O. 6. Kkila-sand g a, ed. P.S.Perera, Colonibo 1906, vv. 263-26k.

400
works testify to the existence of an independent 1r{ndom in the northernmost part of the island. The rulers of this kl-4gdom are referred to in these sources as Ayri Shakravarti Iriya Sakviti Ariyaxaca Varati3and riyar 1 . The first three are variants of

the Sanakrit ryacakravartin while the last

is

the Tainil form

of the Sanskrit rya. The capita]. of these rulers, according to the Sinhalese sources, was at flp-pauna (modern JaZfna) Ibn Batuta visited this northern kingdom in A.D. 13k 1. and his notice is the earliest clear reference to the kingdom. Of the Sinhalese references, the earliest is that of the Nikya-sagrahay-a, written in the last decade of the fourteenth century We have, therefore, definite evidence in our literary sources regarding the existence of an independent kingdom in the northernmost

l..The Rehia of Ibn Batuta, p. 217 2. Kkila-sand!a, v. 3. 1. de Queyroz,

263. k9.

. cit., p.

4..,pp.

25ff. ;Km.,p.6.

5. Kkila-sanda, vV. 263-26k.

6. U.C.H.C., I, ptl, p. 57.

401
part of the ia].and in the fourteenth century 1. In the latest issue of the J.P.A.S. (C.B.), Paranavitana has claimed the discovery of a Sanakrit inscription dated aka 1211 (A.D. 1289). According to him, it is 'inscribed in faint letters on the earlier writing on a stone slab in the Abhayagiri-vihra at Anurdhapura (Ep. Zey., Vol.8 , No. 20)' and 'refers to the king who ruled at Subhapatana (Jaffna) on that date with the full style of ripati Sri Sr ya-rryae Sri Candrabhnu Nahrja' (J.R.AIS. (C.R.), N.S., VIII, pt. 2, p.

26k). This

would, therefore, be the earliest known reference to the kingdom of Jaffna. But unfortunately the inscription has not been published yet and considering the nature of the inscription it is somewhat difficult to use Paranavitana's note with confidence. The editor of the original inscription, over which this inscription is claimed to be incised, has made no mention of any later writing on the slab nor is any such writing visible on the photograph of the estampage appearing in the Epigraphia Zeylanica (Vol. I, No. 20 and not Vol. 8, No.20). In a aper read at the University of Ceylon (see infra, pp45c Paranavitana has stated that these later writings are of such a nature that they may be totally overlooked when one's attention is focussed on the original inscription. In these circumstances, it preferable to wait till the inscription is published by Paranavitana before its evidence is used in a work of this nature. ),

is

-
When did this independent kingdom originate and who were its founders 2 These are questions which have led to some amount of controversy among Ceylon historians in recent years. The origin of the independent kingdom in northern Ceylon or, to be more precise, in the Jaffna district has been traced back to pre-Christian times, as far back as the fifteenth century B.C.,by some while some others have traced it to the eighth century A.D. The generally accepted theory is that the

402

k(ngdom was founded some time in the thirteenth century. The basis for the first claim is a reference in the Mahbhrata as well as some references in the PV.i chronicles of Ceylon and the Tamil epic M$imkalai to a 1ga kingdom in northern Ceylon, then known as NgadTpa, in the time of the Buddha The basis for the second claim is the evidence of the Tamil chronicles of Jaffna, especially the pa-vaipava-ir1!lai. In a learned

article contributed to the Journal of the Royal Asiatic Society (Ceylon Branch) recently, Paranavitana has analysed at length the spurious arguments brought forward by some writers in support of these two claims In our opinion, he has convincingly set

1. C. Rasanayagam,

. cit., p.

7 If.

2. S.Paranavitana, 'The Xrya I1ngdom of North Ceylon', J.R.A.S. (C.B.), N.S., VII, pt.2,

1961, pp.

l7k-22k.

403
aside these arguments and shown that there is no reliable evidence for the existence of an independent kingdom, ruled by Tamile or others, in the Jaffna district during the period preceding the fail of Po].onnaruva. It is not our intention, therefore, to discuss these arguments here. But since these theories have a direct bearing on our aubject, we shall briefly outline them along with the counter-arguments before we proceed on our inquiry. C.Rasanayagam is the chief protagonist of the theory of an independent ktndom having existed

in the pre-

Christian centuries His argment that an independent kliigdom of the Ngaa existed in Jaffna from the fifteenth century B.C. is based on a reference

in the Mhbhrata that Arjuna married

a princess of Maipura, a place beyond KalifLga. Rasanayagam has identified Naipura with Jaffna, for the latter is sometimes referred to in the Tamil works of the fourteenth century A.D. 2 as Maavai. He has argued that 'Maavai seems to be a contracted poetical form of Maavr or Naipuram' and that, therefore, Arjun.a married a ga princess from Jaffna Such an identification 1. C.Rasanayagaxn, . cit., p . 33 if. 2. Ccm., p. 80.
3 C. Rae anayagam,

. cit. ,

ft1 3 f 'f

401
of place names without regard to chronology or relevant historical facts is hardly acceptable. As ffrtber evidence of his theory that there existed an independent kingdom in norbhern Ceylon in the pre-Christian centuries, Rasanayagani adduces the reference in the Mahvasa and the Mainikalai to a 1ga kingdom in the Jaffna district In the Mahvasa there is a legend about two Nga kings of NgadTpa who fought over a gem-set throne and were reconciled through the efforts of the Buddha The same legend appears

in

the Taniil Buddhist epic Naim!ka1ai in which the

scene of the event is given as )1aipallavam, which is identified by Raeanaygam as NgadTpa or the Jaffna district. We agree with Paranavitana that these legends are not based on any historical event and that 'in the Mahvaipsa and the

Main1kalai, as

indeed

in the ancient Sanskrit and Pli literature in general, the Ngas are never represented as human beings, but as a class of superhuman beings, who inhabited a subterranean world' ' We have also pointed out earlier that Rasauayagazn's attempts to prove that the gas of Ceylon were Tamil in language and culture and that their independent kingdom is referred to in Tamil literature

1. C,Raaanayagam, p. 7 if. 2. Mv.,

l:kk-70.

3. Maimkalai, XKvII.
1 S.Paranavitana, 'The Arya Kingdom of North Ceylon', p. 181.

405
as well are baaed on the erroneous identification of some place names in the Saitgam texts Neither the visit of the Buddha to Ceylon nor the existence of a Nga kingdom there can be taken seriously on the strength of these legends. Rasanayagain has argued the continued existence of the northern kingdom in the seentb century A.D. on the basis of a statement by Cosnias Indicopleuates, who visited Ceylon in that century, that there were two kings in the island This cannot be an argument for a kingdom in northern Ceylon for, as Paranavtana baa pointed out, the account of Cosmas itself indicates that by the two kingdoms he meant those of Anurdhapura and Rohaa The Tamil cbronicles refer to a person called Ukkiracii1ka as a ruler of some part of Ceylon in the eighth century A.D. It is stated that be was a descendant of a brother of Vijaya Rasan&yagam has argued that he was a ruler of northern Ceylon and that he was a Eliga. He has also contended that the Kaliga rulers with whom the Zinhalese kings from l4ahinda IV bad alliances were actually the Kliga rulers of flaffna

1. See supra, :p.3J 2. C.Rasanayagam, . cit., pp. 120-121.

3. S.Paranavitana, 'The irya Kingdom of North Ceylon', p. 18k.

k. Yvm., pp. 13-23. 5 C. Rasanayagam,

. cit., p 272 U.

dOG
There is, however, not the slightest evidence for a line of 1iga rulers in Jaffna in this period Besides, as we shall see presently, the legend of UkkiracifLka annot ha relied upon or the history of the northern kingdom. While rejecting the chronlogical baeis of the account of 1Jkkiracifka,, Paranavitana has attempted to identify this personage with Cifikakuzzra of the car-alveu

According to him, 'if we call have faith in the legend given in the !car-kalvetu, the lion-faced king, Ukkirac1-ikp or

Ci1'1k9kun1rafl, may be taken to have flourished about the same time as Igha, whether be was identical with the latter or not' By making such a statement be does not seem to have much doubt about the historicity of UkkiraciAka. But the authenticity of he whole account o Ukkiracfi1,, as it appears in the Tamil chronicles, is questionable. We have already briefly pointed out that the story of Ukktrac{ik and. his queen ?rutappiravalli is based on the Vifaya legend and has also certain elements borrowed from folk-etymology. It has also been indicated that there is a confusion between the legend of Ukkirac1ik#p and

1. S.Paranavitana, 'The Lrya ingdorn of North Ceylon', p. 2. Ibid., p. 191. 3. Ibid.

186.

407
the account of Ku!aa in some of the Tamil chronicles Here we shall digress a little to show why the story of Ukkracizka, has to be treated as just another of the popular versions of the Vijaya legend. The story appears in different versions in six of the Taniil chronicles, namely the

ppa-vaipava-mlai,
!c ar-kalve u,

at akk4appu-nmiyam , Tri-1c.c ala-pitham,

Vaiyptal and the Kailyan1lai.

It has hardly any historical

foundation and is clearly based on the Vigjaya legend. It appears that the Vijaya legend found its way into the historical tradition of the Tamils in this garbled version through the residents of the northern regions who in course of time became assimilated to the Taini]. population. The striking resemblances that one finds between the UkiraciMr story and the story of SithabThu as given in the Pli and Sinhalese sources cannot be explained as mere coincidences. The different versions of the story reveal the stages by which the story of Si4tiabhu became transformed into that of Uracika.
The a-v&ipava-n1lai

has the following version

of the story Ukkiracifika,, a prince of the line of Vijaya'a

1. See supra, 2.

pp.337f.

pp. 13-22.

403
brother having a leonine face and a human trunk, invaded Ceylon from North India in the Saka year 717 (A.D. 795) and conquered half of the island. In the eighth year of his rule in Ceylon, a Ca princess named rutappirava11i, who had an equine face and was suffering frm a strange disease, came to Ceylon on a pi1griine. On the advice of a sage, she bathed in the springs of !rimalai, in Jaffna. As a result, she was cured of her illness and her equine face became human. She tarried there for some time and effected repairs to the temple at that site. One night, when she was asleep in her camp outside the temple, she was abducted by tTkkirac1t k who was enchanted by her beauty. Ukkiraci)rs married her and ruled from Cekataka-nakari (Senk4agala, i.e. Kandy). In time )1rutappiraval1i bore twin children, a som named Nara-c( 1cs-rca( Skt. Nara-ai4ha-rja), who had the tail of a lion, and a daughter named Capakvati (Skt. Campakavat!). Naraci1rca married his own sister and reigned from Ce.kajaka-nRknri after his father's death. It was during his reign that a blind lutist or ______

came from the Ca country to Cek4 alre l.akiri, sang the praises of the king and obtained the peninsula of Jaffna as his prize. Thus be became the founder of the kingdom of Jaffna. The story
of Ukkiracifi1ca end8 abruptly with this event.

The character Ukkiracizka, (Skt. Ugra Si$ha) of this legend, who is described here as a descendant of Si4.habhu,

409
resembles in many ways the lion who is the father of Si*babThu in the Vijayallegend. In the different versions of this story one can distinguish the stages by which the lion of the Vijaya legend becomes transformed into an ordinary hnm ii being. In the

Kailyamlai, on

which the author of the

a-vaipava-mlai

seems to have depended heavily for his story, we find this character as a half-leonine and half-human being whose abode was in a big cave in the hills (kla malai n muficu), to which he carried away )rutappiravalli when he abducted her The p!a-vaipava-nlai states that he as well as his son

bad leonine features, but it does not mention that his abode was a cave In the Vaiypal we find that be is a normal hum'ui being although his son is said to have possessed certain leonine features In the

Tir*-cala-puram
K4a

even that element

disappears altogether In the K!car-ka1veu and the Ma t akkajappu and flahsena respectively -nmiyahescofudwith

With the exception of the last two sources, al]. the others give the name of this personage as Ukkiraci?tka (Fierce Lion) or use variants of this name, such as Uira-cac iA1c Sena Ziha and U kira-n-ca-cl , (Zkt. Ugra

, (St. Ugra )1ah Sena Siiha)!

1. tm., p. 2.

2. lviii., pp. 13, 21. 3. !;a. v.15. e, supra, p.'37

5. y., v.17 ;

., Tampain&c r-patalam, v. 31.

410
This also indicates the m i nner in which the lion of the original story gradually became a person called Fierce Lion and later Sena the Great Fierce Lion. The transformation of the Vaz1.ga princess SuppdevT of the Vijaya legend into IZrutappiravalli, or akacavuntari as she is known in some versions, can also be seen to an extent in the Tamil versions. In the car-kalveu and. the Maakk4appu-

1 - miyam the Vaga princess becomes a lifiga princess. In the Kail;anflai, a-vaipava-nilai and the Tiri-kc ala-

puram she becomes a Ca princess That she was 'very fair' and 'very amorous', like the Vafiga princess is clearly stated in the a-aipava-n1].ai and is also borne out by the

name .akacavuntari (Skt. Faka Sundar) in some of the versions. Just as it is said of uppdevT in the Mahvaisa that 'alone she went forth from the house, desiring the joy of independent life', it is stated that rutappirvaUi went out with her maidens and led an independent life, but, of course, with a different mission Like Suppdevt, she was abducted and carried away to a cave, or palace in some sources, where she bore twin

1. Kk., p. 32 ;

p.30. Tapainskar_paalam, v. 3.

2. Tym., P . 15 Km., p.2 ;

3. Mv., 6:3. k. Mv. 6:li.; Yvm., pp. 15-16.

411
bhildren, a son and a daughter. In the Tiri-k;cala-puram, car-kalveu and the Vaiypa1 the birth of only a son is mentioned. In a similar mxr1ner, it appears that the children of Ukkiracika and )rutappiravaUi are no other than Sihabhu and Si*hasTvalT, the children of the lion and SuppdevT in the Vijaya legend. In the !4ahvasa it is stated that SiI?habThu's 1 'hands and feet were formed like a lion'. In our sources, the son of Ukkiraciks, is stated to have had the face and the tail of a lion2 or only the tail of a lion His name is given in the Iailyaxn1ai and the ppa-vaipava-n1lii as Nara-c1Mc,-

rca (Skt. Narasiha Rja = Man-Lion King) in the Vaiypa]. as Ci!ka (Skt. Si$ha = Lion) in the )1aakk4appu-nmiy and !car-kalveu as Cika-kuira (Skt. &ihakumra)= Lion

Prince) while in the Tiri-cala-puram his name appears

with the fulsome epithets CeyatuIca Vira Pka Vararca Cik (Skt. Jayatuiiga VTra Bhoga Vara Rja Si4ha)? In a].]. these names

1. !. 6:10. 2.
3.

., v. 18 ;
Km., P .

3 ; Yvm., p. 23.
p.

1,

3; i!.!, 18. p.

23.

5. !' v.
6. Kk.,

35; Mm., p. 35.

7.

Tampainakar-patalam, v. 32.

412
the element ei*ha (lion) is preserved. Like S1ihabThn, this prince married his own sister and attained kingship. Although this prince is the counterpart of S4habhu in our legend, certain elements of the story associated with Vijaya have also been included in the story of this prince in the Vaiypal and the Tirik cala-pur4am. The Vaiypal states that this prince, Naracifikarca, sent emissaries to Nadurai and sought the hand of the Pya princess. The princess arrived in Ceylon with a large retinue of people belonging to the various castes as well as sixty maidens This reminds us of the wooing of the Pya princess of Madhur by Vijaya and the arrival of 'craftsmen and a thousand families of the eighteen guilds' as well as seven hundred maidens The Tiri-cala-puram, though not containing all these details, states that the son of Ukkiracii'iktuj married a Pa princess This element in the story further strengthens our contention that the legend of tIkkiraci1ra is clearly based on that of Vijaya. It is interesting to note the position occupied by this legend in the traditional history of the Tamils of Ceylon as it is recorded in the Vaiypal, Xai1yalai,

1. !E.
2. Mv.,

TV. 21-22.

7:55-57.
Tanipainakar-paa1am , v 33.

3. 2

413
and the avvaiapva-ifl!lai. In theBe sources it is associated with the beginning of the independent Tamil kingdom in northern Ceylon in much the same way as the Vijaya legend marks the beginning of the Sinhalese kingdom in the Sinha].ese sources. The manner in which the Sinhalese legend came to assume this position in the traditional history of the Tamils may not be difficult to explain. The Zinhalese of the Jaffna district, as we have already seen were at no time completely dislodged by the Tamils. Many of them probably became assimilated to the Tanzil population in due course. The story of Vijaya would have been current among these people at the time of the Tamil settlements0 When the Sinhalese became assimilated to the Tamil population, a garbled versio@ of the Vi&ya legend would have still lingered in their memory. At a time when their origins were forgotten, these people may have used this legend to expktin the origin of the Tamil kingdom instead of that of the Sinhalese kingdom. The legend may also have been current among the other inhabitants of the Jaffna district. Gradually it appears to have undergone changes that would have made it more suitable to explain the origins of the Tamil kingdom. Hence the representatiom of Mrutappiravalli as a Ca princess and Ukkiraci'iii as a

1. See supra,

414
Ca prince, in some of the versions. In the a-vaipavaiIi and the Kai].yarnlai it ends abruptly and is used to introduce the story of the blind lutist who is claimed to have founded the kingdom og Jaffna. This is an attempt to combine the Ukkiracika, legend with the story of the lutist based on tolk-etymology In the Vaiypal it is used to explain the origin of the Ta.mil settlements in

northern

Ceylon and of the Vanni

chieftaincies. Here the activities of some of the early rulers of the Jaffna kingdom as well as those of the associates of gha seem to have been attributed to Ukkirac1ik and his son NaraciA1cnrca. This is why some are &nc]..ixied to think that UkkiraciMt of the Tamil chroniles is gha Gaanapragasar has attempted to identify him with JayabThu, the associate of )gba Among the other elements in the Ukkiraci?'ktin story are those derived from folk-etymology. One of these is the account of Xavi VTra Rkava (Poet Vira 4ghava), a blind or (lutist). It is said that this lutist

visited the court of Naraci1rirca,, sang a panegyric on him

1. See infra, p.l-'-O 2. S.Paranavitana, 'The rya Kingdom of North Ceylon', p. 191. 3. S.Gnanapragasar, pa-vaipava-vimarcan, p. 6k.

415
and obtained the arid peninsula of Jaffna as his prize. He then invited settlers from South India, gave them lands in that peninsula and ruled ever them, thus becoming the founder of the Tamil kingdom of Jaffna. Since it was founded by a the kingdom was named Thpam This is the traditional account of the foundation of the kingdom of Jaffna. It has been rejected by all serious scholars as lacking any historical basis We are inclined to agree with Gnanapragasar that this story is based partly on the popjlar etymology of the name T1pp 4am (Jaffna), a Tamilised form of the Sinhalese name Tp-pauna, and o4 the story of the blind South Indian poet Kavi V!ra Rkava, who
lived in the sixteenth or seventeenth century Unfortunately

the earliest forms of the name flppam are not known. This makes it difficult to trace the process of Tamilisation. The early forms of the name, as recorded in the South Indian inscriptions are : a) Iy1ppam (lk35) b) flppam (1532 1) flp 4am (i6Ok) c) ppa-paaaa (1685)? d) Iyalpa-tcam (l7l5)

pp. 23-2k.

2. S.Paranavitana, 'The Arya Kingdom of North Ceylon', pp. 176, 201-202;

S. Gnanapragasar, 3. S.Gnanapragaaar, 14 S.I.I., VU, No.778. a-vaipava-vimaraa,c, pp. 15-19. a-vaipava-vimarcan, p. 18.

5.

LE.It. for 1916, No. 61k of 1915.

6.J.Burgess, Tamil and Sanskrit Inscriptions, Setupati Grant No.1, p.62.

7. Ibid., Setupati Grant No.9, p. 81. 8. Ibid., S tupati Grant N .L0,

416
a) Ypia-tcam (173k) 1and g) !pam ( 1 73 k ) The earliest occurrence of this name in the Sinhalese sources is around ].k50, in the sanda poems of the time of Parkramabhu VI The form in these poems is I-pauna. In all the Sinhalese works tifl the British period this form has been retained without change. In the Tamil chronicles the name occurs in the present form of Yppam, except in one instance in the Vaiypal when the form rAfram occurs ' This exception is evidently a copyist's erroz, for there are several orthographic mistakes in the Vaiypal. In the Portuguese works the form Jafanapata is common According to de ueyroz, 'its name without corruption is said to be Jafana-en-Putalam, which means the 'Town of the Lord Jafaxia', and is the name of him who first peopled it' Jafana-en-Putalam has been restored as (the Town of Tppa)? De ueyroz also mentions another

1. J.Burgess, Tarnil and Sanskrit Inscriptions, Setipati Grant No. p. 90. 2. Ibid., Setupati Grant No.1k, p. 93.

13,

3. Gir-sanda, v. 138; Kki1a-sanda, v. 263; Slalihini-sanda, v. 29. k. 5. F. ., v.

as Queyroz,

. cit.,

p.k7.

6. Ibid., pp. k7.-k8. 7. id., p. 11.7, fri. 1.

417
derivation of the name and that is 'Jafanapatanature, which means long harbour' It is not slear how Jafanapatanoture could be interpreted to mean long harbour. S.G.Perera has suggested that it may be Tva-paaattuLai meaning deep harbour This is not very convincing. From the first interpretation given by de Q3teyroz it appears that in the period of Portuguese rule the legend of the or lutist was already current

among the people. In the Dutch records there are several forms of the name, among which Jaffnapatam is common By the time of the British rule the final element was dropped and the place came to be known only as Jaffna. The present Sinhalese name of Jaffna, namely
A

seems to be a recent form derived from

flppam. It does not occur in any of the Sinhalese works before the nineteenth century. We do not agree with Gnanapragasar that fllppam is a Tamilised form of Tpan4.nd that the latter form is a variant of np-pauna Gnanapragasar 's opinion that !p5-pauna is only a Sinhalese translation of the name Nalilir, which is now applied to aprt of Jaffna town where the last

1. F.de Queyroz,

. cit., p. k8.

2. Ibid., p. k8, fu. 1. 3. l4emoirs of Rijckloff van Goens, 16 65, Tr, S.Pieters, Colonibo, 1910, p. 105. Ii. S. Gnanapragasar, Y pa-vaipava-vimarc an, p 18.

418
rulers of the Tanill kingdom had their court, is also imacceptable It seems improbable that the $inhaleee translation of the name of a city founded by the Tamul rulere of Jaffna came to be so popularly accepted by the Tamula who applied it not only to the city but also to the whole peninsula and to the entire district. From the early forms that we have shown above, it appears that the ixihalese Tp-pauna first became flppa-paaam, then Yppam and then Yppam. By about the first bale of the fifteenth century, when we get the earliest recorded form of this name in Taniil, the form Ippam seems to have been current. The earlier form Yppa-paa4am, however, continued to be in use as late as the seventeenth century. Paranavitana's attempt to derive in the name flp-pauna, from Jv is not
may

convincing The element y

be either the Sinhalese word

meaning good, as in flpahuva (Pli Subhapabbata), or the title used by Sinhalese princes in medieval times, the variant of In a medieval inscriptional document entitled Mgha-vttnta, the discovery and contents of which were announced by Paranavita.na

1. S.Gnanapragasar,

p. 18.

2. S.Paranavitana, 'The Lya Kingdom of North Ceylon', p. 202.

3. U.C.E.C., I, pt. 2, p. 732.

419
recently, there oceurs the name Subha-patana which is identified as Jaffna The same name occurs, according to Paranavitana, in a pa].impsest inscription from Anurdhapura dated aka 121]. (A.D. l289) Subha-paana is a direct translation of the Sinhalese name Yp-pauna an& seems to provide a clue to the meaning of this place-name. If subha is not just a wishful rendering into Sanskrit of the Sinhalese element yp in these inscriptions, then we may well accept the meaning of the placename as the 'good port'. But Paranavitana's account of the nature of these records and the ma nn er of their discovery throws much doubt on their authenticity and even their exiatence Until these inscriptions are peroperly edited and. published, it is impossible to base any argument on their evidence.

1. S.Paranavitana, 'Newly Discovered Historical Documents

Relating to Ce1on, India and South-east Asia', Paper read on

k.l1.196k at

the University of Ceylon, Peradeiya,

Ceylon (unpublished). See infra,

p4..c
#.oI

2. S.Paranavitana, 'An inscription from Padaviya', J.R.A.S. (C.B.), N.S., VIII, pt. 3. See infra,

2, p. 26i,

fn. 13; see supra, p.

p.4$i

42
Although we are inclined to believe that the Tami]. name Yppam is only a rendering of the Sinhaleae Tp-pauna, it must be admitted that this derivation is not certain. The earliest forms of this name in both Sinhalese and Tamil occur around the same time, the Tanii]. form occurring nearly fifteen years earlier than the first recorded Sinhalese form. It may well be that the Sinhaleae name is just a Sinhalese rendering of a Tamil name which may or may not be associated with the term But this seems unlikely. It is more likely

that Y ppani is one of the large number of Tamilised forms of Sinhalese toponyms in the Jaffna peninsula. The other element that has gone into the creation of the legend is the story of the blind poet Vra

Rkavar, wh lived in the sixteenth -or seventeenth century. He is said to have visited the court of one of the laat kings of Jaffna who bore the consecration name of Pararcackara, 1 and received an elephant and a plot of land as gifts. While the earlier work Kailyamlai does not mention the name of the lutist or that he was blind, the a-vaipava-mlai

names him as VTra Rkavar and states that be was blind

1. S.Gnenapragasar,

a-vaipava-viniarcan, pp. 16-18 ;

S.Kiim' racuvarni, 'Vaa }1katt4a Cila Iappeyarkaji Vara1u', pa-vaiyava-kaumuti, op. cit., p. 127. 2. !vm., p. 23.

421
This statement is evidently due to the confusion of the story of the lutist created by popular etymologists and the later story of the blind poet. The whole story of the blind otist has to be treated as a mere legend devoid of any historical basis
O1tii.'r tt c.vtr4 o+L

\Pv.

Another element of folk-etymology in the UkkiraciMc legend is the story of the equine-faced )TArutappiravalli. While this charqcter is based on Suppdev of the Vijaya legend, certain features in the story are based on the popular etymology of the place-name Nvifapuram, in Jaffna. In the pa-vaipava-nilai

an ingenious derivation has been given to this place-name It is stated that it was so named by Mrutappira'valli on account of the miraculous change her equine face underwent near that place (Tamil n= horse, vita = left, puram = city - The City where the (face of the ) Horse Left). As Gnanapragaaar has explained, }viapuram seems to be a Tnnrilised form of the Sinhalese name M-vata-'rra The change of the Sinhalese element v!ra into puram in Tamil is demonstrated in several names in the Jaffna peninsula Besides these main elements, several others draw from a number of folk tales can be found in the different versions

1. Tvm., p. 19. 2 S . Gnanapragasar, a-vaipava-vimarc an, p 13.

3. Ibid.

422
of the Ukkiracika story. One of them, for instance, is the story of the legendary Pya princess with three breasts named Taftakai. In the versions where ?rutappirava1li's name is given as *akacavuntari, she is said to have had the epithet Mummulai (Three Breasted) This physical abnormality, the Amazonian natire of the two princesses as well as the similarity of the namTatakai and Aaka suggest some affinity between the two legends As we have already pointed out, some other elements in the story of iakaavuntari are based on the account of VihradevT as found in the hvasa In this mrner it could be shown that the story of Ukkiraci.ka, in the Tamil chronicles baa no historical basis and is only another garbled version of the Vijaya legend with elements from popular etymology and several other folk tales. In our opinion it has to be rejected outright. Any argument for the existence of a Tamil kingdom in Jaffna before the thirteenth century based on this legend is unacceptable. Rasanayagam has further argued the continued existence of the kingdom of Jaffna in the twelfth century on

1.

., p. 32.

2 S . Gnanapragasar,

a-vaipava-vimarc aam, p 10.

3. See supra,

423
the basis of certain references in the literary sources. The Ariyadesa referred to in the Clllavaqisa as the place from where a king named Vradeva invaded Ceylon in the time of VikramabThu I (1111-1132) is identified by him with Jaffna This reference in the Pli chronicle, as Paranavitana has indicated, is to a country outside Cey1on Probably it was a kingdom in India Rsanayagam' a argtments for the rule of Tamil kins in Jaffna in the twelfth century, based on the late Tamil-nvalarcaritai and the a-matala-catakam, are also unacceptable. As Paranavitana baa stated, neither of these works can be considered as having been written in the twelfth century.ifThe reference to Koumpu (Colombo) in the verses attributed to Pukanti 1. C. Rasanayagam, P. 286. 2. S.Paramavitana, 'The Arya Kingdom of North Ceylon', p. 187. 3. The C1avaisa states that V!radeva was a sovereign of Palandpa, (Cv., 61:37). The only Paladlpa that we hear of in our sources in this period is the Maldives, which are re referred to as PaantIvu in the Ca inscriptions (s.I.I., II, p. 91) But it is very unlikely that the two are one and the same place.
if. C.Rasanayagam,

. cit., pp. 2811_286;

S.Paranavitana, 'The irya Kingdom of North Ceylon', pp. 187-188.

424
in the Tamil-n!valar-caritai and the reference to Kai (Kandy) in the other work showethat the evidence of these works is unreliable for the events of the twelfth century, for both places were known by that name only after the ixteentb century Rasanayagan's identification of Zbg with Jaffna in order to argue for the existence of an independent kingdom in Jaffna in earlier than the thirteenth century is also erroneous Zbg of the Arab writers, as is well known to students of the history of South-east Asia, corresponds to Jvaka (Sumtra or Nalay Peninsula) and. not to a place in Ceylon. Certain Ca inscriptions of the eleventh century refer to the defeat of death of six Ceylonese kings at the hands of the Ca rulers. These kings are Yikramabhu (Ilaikaiyarkn -King of the Ceylonese) Vikrama-piya. (I1a.kcura -

Lord of Ceylon) VTra-cai1nka (Cik4attaraica, - King of SiAh$a C!rvailava-mataarija, (attaraica - King of Ceylon)

1. C.Rasanayagam,

. cit., p. 287 ;

2. Ibid., pp. 81, 192. 3. S.I.I., III, p. 5k. k. Ibid. 5. Ibid. 6. Ibid.

d2
Vra-ca1nika, (Ka1ifikar-ma - King of the Ka1igas) 1 and

1parata (I1atkaiyarkkiaiva - Lord of the Ceyloneee)

It has been possible to identify the first three kings as three of the rulers of Robaa mentioned in the Cttlavaisa The identification of the others presents some difficulty. Rasanayagam and Gnanapragaear have attempted to overcome this Jfffl difficulty by arguing that they were rulers of But there is no justification for such an identification. Nowlire in the Ceylonese o South Indian sources do we get any reference to the existence of a kingdom in Jaffna which reisted the Ca occupation. On the other hand, after A.D.

1017, the

northern

half of the island was securely in the hands of the Cas and it was in the south that they encountered opposition. The rulers mentioned in the Ca records were probably in control of parts oS southern Ceylon. As we know from the Ctllavarjisa, there were several petty rulers in the south during the period of Ca rule, offering resistance to the foreigners The fact that

1. S.I.I., III, 2. Ibid.

p.

61.

3. U.C.H.C., I, pt.2, pp. kl8-k20. k. C.Rasanayagam, S.Gnanapragasar, 5. Cv., 56 & 57.

.cit., pp. 278-28k ;

a-vaipava-vimarcan, pp. 52-53.

426
some

of them are not mentioned in the Clavasa

is

no argument

to treat them as rulers of a kingdom in the north, the existence of which is nowhere mentioned. The arguments of Rasanayagam and Gnanapragasar for the existence of an independent kingdom in Jaffna before the thirteenth century are, therefore, untenable. Rasanaysgam's methods of historical analysis are wholly unsatisfactory and, therefore, it is unnecessary to examin e all his other arguments besides those pointed out above. Many of these are based on erroneous identification of place-names on the strength of superficial similarities without regard to chronology or relevant historical facts, as shown by some of the examples above. A number of unidentified or unnamed places in some of the literary sources and inscriptions are used as evidence for his theories by a process of argument that they have been' or 'ought to

have been' Jaffna. ror instance, it is stated in the VTdileri plates of Clukya KTrtivarman II, dated A.D. 757, that
Vinayditya Satyaraya levied tribute from the 'rulers of Kvera,

Praika, Sih4a and other islands ' According to Rasanayagam, 'by the words 'other islands' were meant Jaffna and its dependent islands' and it can be inferred that a separate king was ruling over th5Ifl' This is a conclusion unwarranted by the statement

in
p.28.

1. L.Rice, 'The Chalulyaa and Pallavas', l.A., VIII, Jan. 1879,

2. C. Raeanayagam, . E

p. 239.

427
the inscription. Similarly, the conquest of the 'old islands of the sea numbering twelve thousand' by Rjarja I, lie says, 'Lust indubitably refer to the Jaffna is1ande' In. his opinion, the Maldives and the Laccadives were the dependent islands of JafZna Such methods of historical analysis, it is needless to say, need not be taken seriously. The paucity of references to the Jaffna region in our sources may suggest that this area was not of much significance in the island. The few references that we come across in the Pli chronicle seem to suggest that it was part of the Sinhalese kingdom till the twelfth century. The authority exercised by DevThampiya Tiasa over this region in the third century B.C. is indicated by the account of his reign in the Mahvaisa The port of Jainbukola in the Jaffna peninsula was under the control of the Sinha].eae monarch and it is stated that he built the Jambukola vihra there. There is no mention of that region having been ruled by any other independent monarch at that time. Ls Paranavitana has suggested, it appears that this northern region, whicli was then known as NgadTpa,was administered by 1. C.Rasanayagam,
2. Ibid., p. 262.

. ' p. 239.

3. flv., 11:23, 38; 18:8; 19:23 ff., 60; 20:25.

428
a provincial governor who seems to have held the title of Tparja The evidence for this is the reference in the - 2 Sammohavinodan! to a prince ca].ledDTparaja ruling over N5gadpa. There is also a reference in one of the pre-Christian BrhmI inscriptions at Mihintal to a certain DTpara who was the son of a king of Anurdhapura DTparja was probably the title of the governor of Ngadpa. That Ngadpa was undoubtedly a province of the Anurdhapura kingdom in the second century A.D. is clearly established by the gold plate inscription found at Vallipuram in Jaffna This inscription informs us that in the time of Vasabha (67-111), Nakadiva (1 acI!pa) was administered by a minister (amete) of that king. In the P.li chronicle there are only a few references to NgadTpa. Sometimes for long periods there is no mention of this palce at all. Such silence is explained by Rasanayagam in an incredible m1iner. To him, 'the presuiirption therefore, is that in those years the northern principality was quite independeiit and quiet' It is a well-known fact that the 1. S.Paranavitana, 'The Arya Kingdom of North Ceylon', p. 189. 2. SammohavinodanT, P.T.S. ed. , p. k43. 3. U.C.H.C., I, pt.]., p.229. In No.231 of 1930/31 (M.E.L
for

1930/31)

a subordinate of I?Ajendra I is called DTpattaraiya (Skt. dv!parja). It is not known whether he as associated with Jaffna in any way. li. See supra, p.'1O-

5. C. Raaaiiayagam, 2.

p. 66.

429
P11 chronicle deals mainly with the rulers at Anurdhapura and the later capitals and does not inform us of what happened in the other parts of the island, except when the affairs of these parts affected the course of events at the centre. The silence of the CbroniJ,e on matters regarding gad!pa, Rohaa and Ka].yIi is no argument to say that these regions of the island were independent. From time to time, when the rulers of Anurdhapura built a vihra or sent troops either to suppress rebellions or to repel invaders, the northern district finds mention in the Chronicle. In the seventh century, for instance, when Aggabodhi II (60k-6ik) presented the Ualomagbra temple to the Rjyatanadhtu vihra and an umbrella to the 4mlacetiya, both in NgadIpa, the event was considered to b& important and has been recorded in the Chronicle In the time of Siltmeghavaa

(619-628), when

Siringa attempted to take possession adtpa formed a part, the king

of Uttaradesa, of which

promptly advanced to that district and regained control of it In the time of Nahinda II

(777-797), when

the district chiefs

of Uttaradesa revolted, they were immediately crushed by the king At this time Uttaradesa appears to have been under a

]..Cv.,

42:62. 44:70-75. 48:83-85.

2. Ibid., 3. Ibid.,

430
prince of Anuridhapura, who had the title of Idipda In the time of Nahinda IV (956-972), when a Vallabha ruler invaded Ngadpa, the king sent an army from &nurdhapura to fight him. Nahinda' a troops were able to free that part of the kingdom from the South Indian invader In the eleventh century, Vijayabhu I restored the ancient Jambukola vihira in N gadpa These references in the Pli chronicle show that the northernmost part of the island was considered to be an integral section of the Sinhalese kingdom and treated as such by the rulers of Anurdbapura and Polonnaruva. The evidence of the later inscriptions also shows that in the twelfth century, too, Jaffna was under the control of the Polonnaruva rulers. In that century, Parkramabhu I had strong naval bases at ttuai (rtoa now layts), MaivLL

(Naffuvil)

and VallikAmam (Val( kmam), all situated in the

western part of the Jaffna peninsula, According to the Tiruvlagu inscription of jdhirja Ca II (1178), Parkramabhu built
ships and assembled troopsin these places in order to launch

an attack on the Ca kimgdom A Tamil inscription from NaitTvu,

L. Cv., 448:155.
2. Ibid., 511:12-15; W.N.LWijetunga, 'Who was Vallabha, the invader of North Ceylon', TJ.C.R., XX, pt.2, Oct. 1962, pp. 287-291.

3. Cv., 60:60.
44. V.Venkataaubba Liyyar, 'Tiruvalangadu Inscription of

Rjdiiirja II',

E.I., xxxi, pp. 86-92.

431
an island off the peninsula of Jaffna, contains an edict promulgated by Parkramabhu These inscriptions attest to the autbority wielded by Parkramabhu over the northernmost parts of the island. Thus we see that whatever little evidence is available to us regarding the Jaffna district indicates that it was part of the kingdom ruled by the kings of Anurdhapnra and Polonnaruva. There is no evidence to suggest that it was independent at any time during the historical period before the thirteenth century. We are informed by the Sinhalese sources that in the thirteenth century gha had fortifications at rtoa (Pli Sllkaratittha), besides several others in northern eylon This would mean that in the early part of that century Jafmna was under the rule of the monarch at Po].onnaruva. ?gha was the last ruler of Polonnaruva who wielded authority over the whole of RIjaraba. Re is known to have been ruling at Polonnaruva at least till l236 We may, therefore, reasonably conclude that at least up to 1236 there was no independent kingdom in the Jaftna district. The earliest definite mention of the kingdom of Jaffna is in the travelogue of Ibn Batuta, who visited the

1. E.Indrapaa, . cit. p. 70. 2. See supra, 3. See infra, -p.4.

432
kindoin in 13kk It must have been, therefore, founded before that date Hence we have to place the foundation of the kingdom of Jaffna between 1236 and 13 L 4. An emination of the events of this period should help us to narrow down these limits and to understand the circumstances that led to the rise of the new kingdom. We have already pointed out the significance of the invasion of I4gha and of the events of his regis in the bistorl of the island !gha began his rule at Poloanaruva in 1215. It is agreed that he was still ruling there in 1236 when Par]cramabThu II ascended the throne at Dabadeiya But there has been some amount of difficulty in determining the date of his defeat and the duration of his reign. The CUavaisa gives the length of his rule as twenty-one years In the Pjva1iya, too, it is generally given as twenty-one years, but in one version it appears as nineteen (ekunvisi) years It has

1. See supra, p. L-0.

2. See

infra, p.Ji

; if the palinrpsest inscription of

Ca.ndrabhnu at Anurdhapura, discovered by Paranavitana, is authentic, the lower limit for the foundation of the kingdom of Jaffna can be advanced to 1289. 3. See supra,

lf

tT.C.H.C.,

I, pt.2, pp. 616, 8119.

5. Cv., 80:79.
6. Pv.,

p.109; U.C.K.C., I, pt.2, p. 8119.

433
been suggested that this may be an error for twenty-one (ekvisi) This explanation is palusible for the Pt!jvaliya, in another place, seems to imply that }gha was in occupation of Ijaraha for twenty-one years when Parkrarnabhu II ascehded the throne2(1236) If we are to accept the evidence of the chronicles, we have to conclude that Ngha's reign ended in 1236, some time after the accession of Parkramabhu II. It is not stated in our sources how gha met his death or how his reign came to an end. In the account of Parkramabhu's campaign against the Dami.a a2d Ker4a forces, it is claimed that the mercenaries were completely routed but no mention is made of the fate of ?gha What then happened to Igha 2 Did he die in 1236 or later 7 Was he defeated by the Sinhalese army or did he meet his death before the final debc1e2 These are questions to which our sources do not provide any definite answer. It has been suggested that ?gha's rule did not end in 1236, although the C1lavazsa and the P1!jva1iya have allowed him only a reign of twenty-one years According to this view, ParkramabThu II, unlike his father,Vijayabhu III, proclaimed himself as the sole monarch of Ceylon

in 1236.

1. U.C.E.C., I, pt.2, p. 8149. 2. Py ., p.116 ; aae_.iu,.-. 3. Cv., 83.20 ff. ; Pv., p. 117. U.C..C., I, pt.2, pp. 620, 621, 8k9 ; A.Liayauagamag!, 'Decline of Polonnaruva and the Rise of Da1tbadeiya', thesis submitted to the University of London, 1963.

434
It is, therefore, explained that the 'historians who wrote under Parkraniabhu II would thus have reckoned }1gha'e reign as having terminated with the accession of their sovereign, even though IZgha continued to maintain his position in the Rjaraha for several years after This seems to be a reasonable e

explanation, for the silence of our sources regarding the fate of Igha casts some doubt as to whether his reign really ended in the year when Parkramahhu's accession took place. 1f ?4gha had died or left Ceylon in that year, the chronicles would not have failed to mention this. It is, therefore, very likely that gha continued to rule even after

1236, although

the Sinhalese historians did not recognise him as a legal ruler of Ceylon after that date. This brings us to the question of when }Zgha's rule really ended. Paranavitana and Liyanagamage are inclined to think that )gha continued to rule in Polonnaruva until 3255 and that he was finally defeated by the Sinhalese armies in that year Paranavitana bases his arguments on two considerations. Firstly, he says: 'the Pjvaliya definitely states that igha bad been ruling at Polonnaruva for forty years before he was

1. U.C.H.C., I, pt. 2, p. 8119. 2. Ibid., p. 621 ; A.Liyanagarnage, .

435
obliged to abandon it ' This would mean that 'Zgha abandoned Polonnaruva in or about l255' But there is room for doubting that the figure 'forty' in the Pvaliya statement is orrect. In the first place, the campaign against the Dem4a and N$ala forces, in connection with which the reference to the forty-year occupation oceurs, is dealt with in the PJvaliya before the the campaign against Candrabhlnu, which took place in l2i7. The Ctt].avasa, too, deals with the two campaigns in this order. If we assume that the campaign against gha's forces took place in 1255 after the invasion of Candrabhnu, it is difficult to explain why the two chronicles chose to deal with them in the reverse order. It is not possible to argue that at this point 'the chronicles do not follow a chronological order in dealing with the various campaigns' for the P!jIvaliya specifically states that the invasion of Candrabhnu took place at a time when the island of LAk was freed of the foreign enemies and
rehabilitation was being

undertaken in the country (mes Lakdiva

parasaturan sdh raa samruddha karave mi n jti kalhi) This shows that the chroniclers was consciously placing the defeat

1. TJ.C.H.C., I, pt. 2, pp. 620-621. 2. Ibid., P. 621. 3. Ibid. Pv., p. 117.

43i
of Ngha's forces before the invasion of Candrabhnu. Secondly, the C!fllavaisa judiciously avoids any mention of the forty-year occupation, although this work is later than the P'jvaliya and its accouut of the campaign against the forces of )gha is remarkably similar to that of the P11jvaliya. This, too, casts some doubt on the validity of the Pjvaliya statement. It is, therefore, possible that the figure 'forty' in the statement of the P11jva1iya is a mistake. It may also be possible that MAgha was ousted from Polonnaruva before l2k7 but his forces were not completely routed till 1255. The forty-year occupation may refer to the presence of the enemy in the northern parts of Ceylon and not to MAgha's rule in Polonnaruva. Parana'uitana's second argument in favour of 1255 as the date of MAgha's defeat is that If we take that Polonnaru and the Rjara ftha were recovered by Parkramabhu before l2k7, in which year the invasion of Candrabhnu was repulsed, he had no enemies to contend with until the second invasion of Cand.rabbnu, which, as will be shown later, occurred in or about 1260, and the P4ya invasions, the first of which was in or about 125k. He was, therefore4 free on this supposition, to realise his ambition of being crowxied at Polonnaru, and of restoring the Tooth Relic to its ancient shrine at that city. But, for about ten or fifteen years after the first defeat of Candrabbnu, Parkramabhu paid no attention toolonnaru. All his religious and other activities during these years were in the MAyraf a, or in the south-western or central districts of the Island. The reason for this must have been that Polonnarurwas still in the hands of the enemy, who abandoned it after an occupation of forty years, as stated by the PUjva1!, in 1255. 1

1. U.C.H.C., I, pt.2, p. 621.

43?
This is not quite convincing. During the whole period of his r reign, Parkramabihu had to contend with a number of enemies. First ?1gha, then Candrabhnu and possibly the Pyas as we shall see later, gave constant trouble to the Sinhalese ruler. If one asks why Parkramabhu failed to hold his coronation at Polonnaruva between 12k7 and 125k when there were no foreign invasions, one may also ask why he failed to achieve that between 1255 and 1260 when, according to Paranavitana, Parkramabhu bad freed Polonnaruva of the enemy. The date of Candrabhtnu's second invasion is not definitely known to be 1260. Some place it in 1262 and some others later on If we accept 1262, the time lapse between 12k7 and 125k is almost the same as that between 1255 and 1262. If during the seven-year period between 12k7 and 125k Parkramabhu was pnevented from realising his ambition by !gha'a presence in Polonnaruva, what prevented him from achieving his aim during the seven-year period from 1255 and
b.CAtler

1262, whenIgha o any other enemy was at Polonnaruva I It is 2, not possible , therefore, to argue that )gha was ruling at Poloxinaruva till 1255 on the basis that Parkramabhu failed to hold hid coronation in that city before 1255. As we shall aee later on, whether )Zgha was defeated in 1255 or before 12117, what is important is that the Sinhalese rulers could not get

1. A.Liyanagmge, 2

438
rid of the foreign enemy from northern Ceylon after the fall of Polonnaruva. The enemies were there even after 1262. Northern Ceylon had permanently slipped out of the hands of the Sinhalese rulers in 1215. As we shall see in the sequel, what made it possible for Parkramabhu to enter Polonnaruva in 1262 was apparently the temporary subjugation of the enemy

in northern

Ceylon by the Pyas. This, more than any other factor, prevented ParkramabThu not only from holding his coronation but also from restoring Polonnaruva to its pristine position. We are, therefore, inclined to think that the defeat of ?gha and his forces took place before the first invasion of Candrabhnu in

12k7. )gha

was ousted from Polonnaruva

possibly not long before that event. Hi forces may have continued to resist in their fortifications in Rjaraha even after that date. The inha1ese sources do not inform us that ?gha was killed in battle by the Sinhalese. It appears that )1gha was only dislodged from Polonnaruva but not ousted completely from Rjaraffha. However, we have no evidence at all as to what happened to gha or about the events in Rjaraha after his defeat. In a].]. probability, }gha and his associates established their authority somewhere in the northernmost part of the island. In

12k?, some

time after the defeat of }1gha,

Candrabhnu invaded Ceylon. He was defeated by ParkramabThu

439
and driven away from the Sinhalese kingdom We do not know what Candrabhnu was doing between this event and hi second abortive attempt to capture power in the Sinhalese kingdom, which probably took place in l262 Between the defeat of gha and the second invasion of Candrabhnu certain important events seem to have taken place in northern Ceylon. The Sinhalese chronicles mention nothing of these events. But the information contained in some o the contemporary P4ya inscriptions, inadequate though it may be, helps us to extent in conjecturing the course of events. The inscriptions of Javarma, Sundara Ipya I, from the year 1258, refer to a

P4ya invasion

of Ceylon which

presumably took place before 125S No details of this invasion

1. The invasions of Candrabhnu have been discussed in great detail in the unpublished thesis of A.Liyanagamage mentioned above. 2, Paranavitana dates this event in 1260 (U.C.H.C., I, pt.2, p.621), Krom in 126k (Ned. Kon. Akad. van Wetenschappen, Afd. Letterkunde, Deel 62, Serie B. Nd.5, pp.

8-9) and

Coed&s in 1256 (B.K.I. ,

83, pp. 1f59-472). Liyanagamage (. cit.) has discussed

these views and dated the event in 1262. We are inclined to accept this. 3. There is no evidence to date this event to 125k.

440
are given except for the fact that Sundara

P4ya exacted

tribute

frok the Ceylonese ruler The Sinhalese chronicles make no mention of this invasion. It has been surmised by Paranavitana that there was an alliance between the Sinhaleae and the Pyas and that Sundara Pya invaded Ceylon to give assistance to Parkramabhu against gha According to him, 'the silence of the monks who chronicled these events with regard to the assistance which their patron received atrom a great Tamil power for subduing a band of Thmil and Nalayi mercenaries is easily understandableZ Ee also contends that Ngha was in league with the Cas ' But there is hardly any evidence for such alliances between Ceylonese and South Indian rulers at this time. In support of his conclusion Paranavitana adduces the statement in the chronicles that the monarchs of foreign lands bad come under the influence of Parkramabbu But it must be remembered thatParanavitana himself has stressed that the account of Parkramabhu's campaign against 1gha in the chronicles, 'given by monks who wished to glorify their patron,

1. Sen Tainil, IV, pp. 51+-5l6; K.A.Nilakanta Sastri, The Kingdom, p. 162 ; LE.R. for 189k, No. 166 of 189k. 2. TJ.C.H.C.,I, pt. 2, p. 621 3. Ibid. 4 Ibid., . 622.

5. Ibid., p. 621.

441
is meant for those who were accustomed to believing without iuestion anything stated by the+nd does not carry conviction to a critical historian of toda.y' In the light of this

much credence cannot be attached to the statement that foreign monarchs came under the influence of Parkramabhu. Further Paranavitana tries to identify the Ceylonese prince Parkramabhu, who died fighting for the Cas in South India in 1230, with
Parkramabhu Niafika MaU.a of the Pa44uvasuuvara Tamil

inscription The latter is termed

i1akaikk (King of

'South Ceylon') in this inscription. On this basis Paranavitana argues that 'if the ruler of South Ceylon took the side of the enemies of the C2a empire, it is reasonable to infer that gha ranged himself on the side of the Coas' The unlikelihood of this contention has been fully demonstrated by Liyanagamaget It has also been pointed out in an article on the Pai4uvasnuvara inscription that Parkramabhu Niik' Nalla of that record is no other than the K1iiga ruler NlMai3.ka Nalla who reigned in the twelfth century When the Pauvasnuvara inscription

1. U.C..C., I, pt.2, p. 620.

2. Ibid., p. 622. 3. Ibid.

k. A.Liyanagamage,

5. X.Kanapathi Pillai, 'A Tamil Inscription From Pa4uvasnuvara', U.C.R., XVIII, No.1,3 & 11, July-Oct. 1960, p. 157 ft.

442
expressly mentions that the Parkramabhu in whose reign it was set up had the surname Niki Malla it is not clear why Paranavitana prefers to identify him with another ParkramabThu who is known to us only from a South Indian inscription. 'om his Siniialese inscriptions we know that Ni Aka Malla had the name ParAkrainabThu as wel1 Since it has been discussed elsewhere in detail, we do not propose to enter into a discussion of this point. But we would like to point out here that it it erroneous to take Teilakai to mean the southern part of Ceylon and, on that basis, to build up a theory of the ruler of South Ceylon ranging himself on the side of the enemies of the Cas and the ruler of North Ceylon tk1-rg the side of the Cas. Tei1afkai has been used
in

Tamil literature as well as in inscriptions

1. S.Paranavitana, 'A Slab Inscription of NitaM )ialla at Polonnaru, wrongly attributed to Viayabhu II', E.Z., V, pt.2, p.202. 2. a) Tiru-flacampantar Tvra Tiruppatikaikaj, Ka 1aka ed., No.243, p. 520 (7th century); b) Cuntarar Tiruppatikam, No.147 (c.8th century c) ikkavcakar Tiruv'cakam, ed. C.Pillai (Mad. 1949), p. 326j Cc. 9th century); 4) M.E.R. for 1923, No.5 05 of 1922 (12th century);

e) M.E.R. for 1915, No.406 of 1914 (13th century); f) mid., No.407 ., Ciappuppyirani & v.k, p.k (c.l5th century); h) Km., P . 6 (c.l6th century); i) ., vv .90 , 91 (c.l6th century); j) ., pp. 18, 170 (c.l8th century); k) Mni., Tta.,ka1veu, p. 18 Cc. 18th century).
of 1914 (13th century); g)

443
almost always in poetry, for a long time since the seventh century. In all these instances this name stands for the whole s of Ceylon and not for any part of it. The name I l Aki was also used for some places in south India, besides Ceylon The prefix tea,, meaning south, was used to distinguish Ceylon from the other Ila.kais. In the early centuries of the Christian era, one of the South Indian Ilakais had the prefix n 2 (-ilki).

In the Ca period, we find that it had the prefix uttara, meaning north (Uttara Laz.k) The Lek in the south was, therefore, known as Te-ilakai (Lak in the South). Perhaps the best example that one can show to prove that Te-ilkii does not stand for the southern part of Ceylon only is its occurrence in the ailyanilai. In this Tamil chronicle of Ceylon, one of the kings of Jaffna with the throne or consecration name of Cekarcacka.ra is styled CekarcaTei1akai mapav (Cekarca, the king of Teilakai) No one would contend that the king of the northern part of the island is here referred to as the king of southern Ceylon. What is really meant is that the ln.g of Jaffna is the monarch of the whole of Ceylon. Teilafkei, in

1. See supra, p. 32. 2. Thid.; there was also a }yilafigai in }'sore, E.G., III, pp.1k7-].kB. 3. M.E.R. for 1913, No.77 of 1913; K.A.Nilakanta Sastri, The Cas, p. kk3, n83. k. Km., p. 6.

441
this instance, stands undoubtedly for the whole of Ceylon. In the Pauvasnuvara inscription, too, as in all the other instances, it stands for the whole island. It is, therefore, difficult to accept Paranavitana's theory that Parkramabhu of the South Indian inscription, mentioned above, was a king of southern Ceylon who aided the enemies of the Cas. Consequently there is no basis for the theory that the P4ya invasion that took place some time before 3258 was aimed at assisting Parkramabhu Iiagainst Ngha, the aUeged Ca ally. The silence of the Sinhalese chronicles on the ff'a invasion and the true significance of this event ma be understood to some extent if we analyse the evidence of some P4ya inscriptions of 1263 and 126k, which refer in detail to another P4ya invasion of Ceylon under Jat varma V!ra P 4ya I (acc. 1253). In one of his inscriptions of 1263, Vra

Pya is credited with the feat of having taken 'Iam and the CAvaka's crown together with his crowned bead' In an inscription

of 126k, a detailed account of the campaign in Ceylon is givex&

1. N.E.R. for 1911, No. 588 of 1916+ LA.Ni].akanta Sastri, 'The Ceylon Expedition of Javarman Vra Pya' , Proceedings and Transactions of the Eighth AllThdia Oriental Conference, Bangalore 1 937, p. O9. 2. Ibid., pp. 523-525.

445
The text of this inscription from Kuumiyma1ai is corrupt at certain places and the details are, therefore, not quite intelligible. In the words of Nilakanta Sastri, we can see that there was corns dispute in Ceylon, that one of the ministers had invoked Pyan intercession, and that the king's aim was to uphold in proper form the ancient practive of royalty (araiyal va,akkam ippaa n Then we learn that among the kings of Ceylon one was killed in the battlefield and all his troops, treasures and parapherhalia confiscated (araiu keAu dyam aaia van), after which the double carp (the Pya emblem) was put upon the fine flags waving on the Kamalai and the Tnikagini, another kixig(of Ceylon) was compelled to sur;ender his elephants as tribute. Finally, the son of the Evaka, who bad formerly disregarded commands and evnced. hostility, came and prostrated (before Vra Pya) and was duly rewarded. The tex is difficult here and so far as I can make it out, the Svaka's son was presented with the anklet of heroes (vrakka), was taken round n procession on an elephant and was permitte& to proceed at once to Lnunipuri because it was thought (by Vtra Pya)that it was only proper that the son should rule the vast land of Ceylon formerly ruled by his father. 1

uipp4 ).

The two inscriptions are generally taken to refer to the same

expedition and rightly so. The expedition is not mentioned in Vra Pya's inscription of the early part of l262 It appears to have taken place either in the latter part of 1262 or in 1263. As we shall see later, it took place probably in 1262. These Pya inscriptions inform us that around 1262 there were two kings in Ceylon and that one of them was a Cvski (Jvaka).

1. LA.Nilakanta Sastri, 'The Ceylon Expedition of Javarman V!ra Pya', . cit., pp. 511-512.

2. l'.E.R. for 1929-30, iq o.k8o of 1930.

44'i
The campaign of Sundara Ptya, some time before

1258, was

probably directed against this Cvaka and this may be the reason why the Sinhalese chroniclers took no notice of it. The reference in the Ku4uzniymalai inscription to 'the son of the vaka, who bad formerly disregarded commnnds and evinced hostility' seems to support this conclusion. For, the person who was recalcitrant was the Jvaka and not his son, as some are inclined to take. The phrase 'who had formerly disregarded comrnnds and. evinced
hostility' qualifies Svaka and not the 'son of the Zvaka'.
S S

The Jvaka was killed in battle in or about 1262. Then the son was placed on the throne by the Pya ruler. Before this
event, the P4ya's coni m ds would have been directed to the

father and not to the son. It is not possible that the son is accused in our record of having formerly disregarded commands'. It is the father who would have disregarded commnnds earlier and paid for it with his life. Once the father was punished, the son was given full royal honours and placed on the throne, 'because it was thought that it was only proper that the son should rule am, surrounded by the vast sea, which was ruled
by his father' It folOows, therefore, that it was the father and not the son who was recalcitrant earlier. For the elder

1. Nili1ranta

Saetri's rendering of this phrase is not quite

accurate; 'which was formerly ruled by his father' should read 'which was ruled by his father'.

447
1vaka to be accused of having been recalcitrant, have submitted to Pya authority on an earlier occasion and rater failed to be submissive. When did this happen 7 It seems to have occurred some time before 1258, when Sundara P'ya claims to have obtained tribute from a ruler of Ceylon. The silence of the Sinhalese chronicles regarding any Pya invasion of the Sinhalese kingdom suggests that it was the Jvaka and not ParIkramabhu II who was subdued by the PI4yas. On the evidence of the Pya inscriptions we may, therefore, say that some time before 1258 Sundara Pya invaded Ceylon and exacted tribute from a Jvaka king who was ruling part of Ceylon, that this ruler soon became recalcitrant and was killed by Vira P4ya
in 1262 and that the son of that Jvaka was then placed on the

throne by VTra Pya"because it was thought that it was only proper that the son should rule am, surrounded by the vast sea, which was ruled by his father'. Now, we have to identify the Jvaka of the Pya inscriptions and. the kingdom ruled by hint. The Sinhalese
chronicles refer to the activities of only one Jvaka in Ceylon

at this time, He is Candrabhnu, wh is recorded to have invaded the Sinhalese kingdom on two occasions. The defeat and death of this invader on the second occasion enabled the Sinhalese princes to enter Polonnaruva and start restoration work there Since

1. A.Liyanagamage, .

443
the Pya inscriptions refer to the defeat and death of a Jvaka in Ceylon around the same time, it has been rightly surmised that these inscriptions and the Sinhalese chronicles refer to the same incident Paranavitana, who supported this view earlier, has lately attempted to identify the Jvaka of the P4ya inscriptions with )gha This change of opinion has been due to his new theory that gba hailed from 'l4alaysia' and not from Kalifiga, in In&ia, as recorded by the Sinhalese and Tainil aources We find this theory unconvincing and agree with Nilakanta Sastri that it is based mainly on 'vague surmises k and plays with phonetic similarities' Recently Paranavitana has claimed to have discovered epigraphic materials which conclusively prove his theory until these materials have been published, we will not be in a position to offer comments on this theory. For the present, we are inclined to accept the statement of the chronicles that lgha came from Kalifiga in India. Consequently, we believe that the only Jvaka known to

1. A.Liyanaganiage,

. cit. ; LA.NilAknta Sastri, 'The Ceylon i'a P4ya', . cit., p.520,

Expedition of Javarma.n

2. U.C.H.C., I, pt.2, p. 627 ; S.Paranavitana, 'The irya Kingdom of 1orth Ceylon', p. 19k. 3. S.Paranavitana, 'Ceylon and Malaysia in Medieval Times', J.R.A.S. (C.B.), N.S., VII, pt.l, pp. 11I2. k. K.&,N.Sastri, 'Ceylon and Sri Vijaya', JR.A.S. (C.B), N.S.,VIII,p.12 5. See supra, p. JoI

449
have been active in Ceylon in the third quarter of the thirteenth century is Candrabhnu. It may not be wrong to conclude, therefore, that the JAvka of the Pya inscriptions was Candrabhuu, that be was ruler of a kingdom in Ceylon for some time and that he met his defeat and death in 1262, the year in which the Sinlialese princes entered the old capital, Polonnaruva It has been mentioned earlier that the campaign of V!ra Pya in Ceylon can be dated either in the latter part of 1262 or early in 1263. Since the death of Candrabhnu, on the basis of the Sinhalese chronicles, appears to have taken place in 1262, V!rapPya's campaign against the J'vaka ruler has to be dated in 1262 and not in 1263 As Paranavitana and Liyanagamage believe, the Pyaa were probably the allies of Parkramabhu II The Pya inscriptions are not clear on this point. The Kuuniiyma1ai inscription refers to the request made by a minister, presumably to intervene in the war in Ceylon between ParkramabThu and CandrabhAnu. It is not stated, as has been claimed sometimes, that the minister came from Cey1on

1 A. Liyanagamage, 2. Ibid.

. cit.

3. Ibid. ; U.C.H.C., I, pt.2, p. 627.

k. U.

.H.C., I, pt.2, pp. 627, 685.

45k)
Re could very well have been mfnister of V!ra P4ya himself. But probably he was a ninister of Parkramabhu who appealed to the Pya ruler to intervene in the Ceylonese war. Such an appeal would have been made to the Pya ruler because, as we have indicatedk earlier, Candrabhnu was probably a tributary of VTra Pya, at least in name if not in fact. The Pya inscriptions claim that after the Jvaka was killed, V!ra Pya ].evied tribute from the otber king' If this other king was Parkramabhu, it would mean that the Siniialese ruler was treated only as a subordinate ally. Raving successfully intervened in the war and punished his recalcitrant tributary with death, V!ra Pya raised the Jvaka's son to the throne of his father. It is clear that the Pya monarch did not intervene in the Ceylonese war as an ally of Parkramabhu with the intention of conquering the vaka'a kingdom for the Sinhalese ruler. He appears to have been settling a dispute between his own tributary, who had become refractory, and. another subordinate ally, who too was probably one of his tributaries. This is how we could possibly interpret the evidence of the Sinhalese chronicles and the Pya

1. !ai vntanaj (ojher king) is the phrase used in the Kuumiyn1ai inscription, LA.Nil&nta Sastri,'The Ceylon Expedition of Javarman VTra Pya', . cit., p. 52k.

451
inscriptions regarding the events of the period, between 12k 7 and 1262. If the little evidence we have favours the identification of the Jvaka of the PIQya inscriptions with Candrabbnu

the next question we are faced with is the location of his

king

dom. The Linhalese sources do not inform us of the existence of an independent kingdom in southern Ceylon, other than that ruled by ParkramabThu II, in the middle of the thirteenth century. ut we do not know whether there was any independent kingdom in northern or eastern Ceylon, which at this time was not under the control of the Sinhalese monarch. Candrabhnu'a activities prior to his second invasion of the Dabadei;a kingdom were confined to the northern part of the island. He is stated to have landed at Mahtittha with Tami]. mercenaries from the P14ya and Ca countries and 'brought over to his side the STh4aa dwelling in PadT, Eurund and other districts'

1. Paranavitana has claimed that a certain Candrabhnu Nahrja is mentioned as the ruler of Subhapajana (Jaffna) in 1289 A.D.
in an inscription found at Anurdhapura. If this record is

authentic, this ruler may be the son of Candrabhnu, the invader. This may confirm the identification of the vaka of the Pya inscriptions with Candrabhnu. See infra,

2. Cv., 88:63-6k.

452
Pad! and Kurund, as we have noted earlier, are Padaviya and Kuruntanttr in northern Ceylon. Possibly not
azL

the Taniil

mercenaries of Candrabhnu were from South India. Re may have recruited some from the northern parts of Ceylon, oo. If Candrabhnu had a kingdom in Ceylon, this must have been in the northern region of the island rather than in any other part. It was probably the forerunner of the Tamil kingdom of Jaffna which was ruled

in the

fourteenth century by a line

of lnga called the Aryacakravartins. The evidence of certain place names in the Jaffna district, revealing Jvaka association, also points to the conclusion that it was in northern Ceylon that the vakas had some sort of authority at any time in the history of the island. There are at least two village nawes in the Jaffna district with the element Jvaka. ..,,Jivy) Cvaka-cri (Jvaka-cri = Jvaka settlementand CvkVai (Cvaka k t ai - Jvaka kai = Jvaka fort). These two names, the

stifl in use, find mention in the

Kkila-sanda and in some of the Linhalese Kaaimpotas (Boundary Books) It appears that the Jaffna pen4nula and some parts of the Jaffna and Hullaitvu districts had the territorial name Jvagama. This name OCCU8 in one of the Sinhalese Kaaimpotas

1. Yvm., p. 60; Rkila-sanda, y.Jo ; Tn Siha Kaaim saha Vitti, ed. A.J.W.I.rambe, 1926, p. 21. 5.

1acv"I,

)-o4..

2. Tn Sii.ha Kaaim saha Vitti, op. cit., p.21.

453
Paranavitana takes JAvagama to be derived from Jvaka, through Taniil, 'just as Sinhalese nA4agama is derived from Sanskrit nAaka through Tamil' This is plausible, although it need not necessarily be so. It it is derived from Jvaka, it indicates JAvaka rule in the northern regions more than the other two place names. Paranavitana has adduced further evidence in support of the conclusion that the JAvakas were the predecessors of the Aryacakravartins in the kingdom of Jaffna In a fourteenth century Sinhalese inscription found at davala a personage named }rttA 4anL Peruniun, who entered into a treaty with VikramabAhu III (1357-137k ) is mentioned. He has been identified by Paranavitana with IrttAa Cifikai Ariyag, one of the Aryacakravartins of Jaffna mentioned in the ppAia-vaipava-n1lai.

In this inscription, he is given the epithet Sav4u-pati. Paranavitana is inclined to equate the word sava4u with Jvaka. He argues that 'JAva is pronounced in Tamil as CAva or SAva, to which A., 'person', has been added on the analogy of Ma].ayA1i from Malaya + On this basis, he says, SAVA]. or

1. S.Paranavitana, 'The Arya Kingdom of North Ceylon', p. 195. 2. Ibid., pp. 197-200. 3. Ibid., p. 199.

451
Svii would denote a person of Thvaka race. 'The final vowel ii (in sav4u) suggests the influence of Telugu which is known to have been the language of the rulers of the vaka kingdom in the Malay peninsula' Having thus derived Sav4u front Jva, Paranavitana explains the m yrner in which An Iryacakravartin came to bear the title of Sav4upati. If, as we have demonstrated, )gha came front Malaysia with a following of Malay warriors, and if he founded a kingdom in the North, the ruling class of that kingdom And further, if an would have been Jvakae or Iriyan fro Rmvaram became master of this kingdom as a result of a matrimonial alliance, the Jvakaa or or Sav4u people would have referred to this Lriyan and his descendants as their lord. 2 This seems to be a far-fetched theory. In the first place, the identification of ?rttam of the ?davala inscription with an ryacakravartin of Jaffna is not certain. Even if this is granted, the derivation of Sav4u-pati from Jiva is rather ingenious. It is true that Jiva is pronounced in Tamil as Cva or SAva. But the analogy on which this is made the first element of SAvAji is certainly wrong. MalayAi is not derived from the two words Malaya and , but from MalaAam, the Tamil name for Ker4a, meaning 'valley', in the same way as Vki (BegAi) is derived from Vk.am (BagAa - Bengal).

1. S.Paranavitana, 'The irya Kingdom of North Ceylon', p. 199. 2. Ibid.

455
No one would say that the latter is derived from Vaitic (VaAga)

and

, although it would appear quite logical. The derivation

of Sav4u, occurring in the Sinhalese works, is disputed by scholars. Various other interpretations have been given to it It does not occur in any Tamil work and it is doubtful whether any ryacakravartin bore this title. Further, the inscription

in which it is claimed to occur as the epithet borne by a certain

Mrtt4am is badly damaged. Paranavitana admits that 'the record is badly weathered, and from its sixth line, only a few letters are legible here and there' and. that even 'some letters in the first five lines are also indistinct' The term Sav4u-pati occurs in the fourth line and two letters of this word, namely va and ti, are not clear. Under these circumstances, one cannot be sure that the epithet is Sav4upati and not some other word. Thia evidence adduced by Paranavitana in support of the rule of in northern Ceylon is, therefore,

unacceptable, although it does not go against our conclusion arrived at from other evidence.

1. D.B.Jayati].aka, Shitya Lipi, 1956, p. Wi ; E.LCodrington, 'The Gampola Period of Ceylon Bistory', J.R.A.S. (C.B.),

XXXII, No. 86, p. 301.

2. S.Paranavitana, 'The 4rya Kingdom of North Ceylon', p. 197.

In the light of the meagre evidence that is available to us w may not be wrong

in locating

the kingdom of Candrabbnu

somewhere in the northern part of Ceylon. In all probability, it was the kingdom of Jaffna which the 4ryacakravartins inherited

1. The Saziskrit inscription from Anurdbapura mentioned earlier (see supra,

p.43 )

refers,to a Candrabhnu as the ruler of

Subhapatana (Jaffna). If this is an authentic record and if Candrabhnu is identified as the son of the vaka who invaded the Sinhalese kingdom, the identification of the Jvaka's kingdom with Jaffna will be beyond dispute. See infra, p. 4c9 In the Vaiypal

(v.36) occurs

the following reference:-

'Tanikkal eum varaiyatail Cakarr eurn Kar!r enruin nta kula v'ar pa t ai_ut a!_ki tara ilCkXum_njil'

(1ihen the Cka, with the army of low caste hunters called Cakarr and Karr, were ruling this world from TaikkAl). The only manuscript of the Vaiypal now extant, from which this chronicle has been edited, and published, is full of copyist's errors. The words Cka vum can be emended as Ckavarum by adding one letter ra which would then mean 'being ruled by the Ckavar' (a varia nt of Cvakar). If this is admissible, the reference here may be taken to preserve some memory of Jvaka rule in Taikkal, in the Northern Province. But we cannot be certain that only this emendation is possible. Cka vum is obviously an error.

457
If, as we have suggested earlier, the invasion of Javarman Sundara Pya I, some time before 1258, was directed against this new kingdom, its foundation has to be placed before that date. We have already laid down that the upper limit for the establishment of an independent kingdom in northern Ceylon

is 1236. The

lower limit may now be reckoned as 1258. Our

sources do not mention anything about the events in northern Ceylon between these two dates, except for the defeat of gha, which, we believe, occurred some time before 12k7. In the present state of our knowledge we can only resort to conjecture in reconstructing the course of events that led to the foundation of the new kingdom in northern Ceylon. The only basis for our conjecture is the vague evidence of the Tamil chronicles. Although gha was ousted from Polonnaruva some time after 1236, be seems to have continued to exercise authority further north. The Linhalese monarch was in no position to recover the whole of jaraha or even to secure his position at Polonnaruva for a long time. The reason for ParkramabIhu's failure appears to have been the presence of the enemy in the northern part of the island. The failure of the Sinhalese to oust the foreigners from the island was an important factor that led to the rise of the new kingdom in

453
the north. 1gha presumably 5et up a new capital somewhere in northern Ceylon, probably in Jaffna, and exercised authority in that region. We do not know what fate eventually overtook him and it is useless to surmise on this point. He probably died a natural death and was succeeded by someone else. In all probability this new kingdom of northern Ceylon is the same as that ruled by the Tvaka prince around 1262. It is not known how a Jvaka came to be on the throne of a kingdom in northern Ceylon. As we are inclined to identify the Jvaka of the Pya inscriptions with Candrabhnu, it is possible to conjecture that this I'a1ay ruler, after his defeat at the bands of the Sinhalese, fled to the northern kingdom. In course of time, by some means he was in a position to succeed to the throne.there. Probably he won the favour of Mgha, if he was still living at that time, and succeeded him. Or, it is possible that be was able to wrest power from the ruler of the northern kingdom. If such was the course of events, it would appear that it was as ruler of the northern kingdom that Candrabhnu launched his second attack on the Sinhalese kingdom, which turned out to be fatal to him. After his death, his son ascended the throne as a feudatory of the PIyas with the blessings of Jatvarman YTra Pya. This reconstruction of the events in northern Ceylon seems to fit the meagre and vague information that we are able to extract from our epigraphic and literary sources, chiefly

1 the T mi1 chronicles.

459
627),

1. In his contribution to the TJ.C.H.C. (I, pt.2, P .

Paranavitana held that 'the independent kingdom in North ceylon appears to have originated with Candrabhlnu'. Later, be changed this view and attributed the foundation of the kingdom to gba and agreed that it was 'possibly also ruled for sometime by Candrabhnu' (J.R.A.S. (C.B.), N.S., VII, pt.2, p.19k). But in a paper read at the University of Ceylon recently, he claims to have discovered an epigraphic document entitled Mgha-vttnta_(which he sometimes calls Ngharja-vttnta)

which deals inter aiim with the 'foundation o the kingdom of

Subhapatana (Jaffna) with Ga4agopla, }Zgha' a son as its first ruler, under the protection of the Pyaa, the career of CandrabbThu, the son of Gaagop.la, ....., the relations between Candrab1inu of Subhapaana and Pa4ita ParkramabThu of Kurungala, the supplanting of gha's descendants at Subhapaana by Rjaputra hakusa from RAmevara (rya Cakravarti)...
(S.Paraxiavitana, 'Newly Discovered Historical Documents

Relating to Ceylon, India and South-east Asia', Although the light thrown by such a document would alter our picture in many ways, it would confirm the connections of gha and Candrabhnu with the northern kingdom. However, until the new document is published no comments can be offered on this matter.

4Bi
In the Tamil chronicles , as we have already Been, the foundation of the kingdom of Jaffna is attributed to a blind minstrel. It is said that this minstrel died without an issue and that Jaffna was without a king for some time. Eventually one of the nobles went to Nadurai and invited a prince called

Ciiki Liya

(Irya of Si4ha(nagara) ) alias silcai Cakkaravartti to be the

Kafkai 4riya or Vicaya

king of Jaffna. The invitation was accepted and Vicaya

KU]i3kn i

Cakkaravartti became the first prince to rule Jaffna.

This is the account found

in the

Kailyanflai and the

a-vaipava-nlai As we have discussed earlier, it is stated in these chronicles that several noble families as well as members of the different castes were invited from South India in the time of the king Vicaya
nCi Cekkaravartti

and given lands in the Jaffna peninsula and the Vanni districts

to settle down There are some reasons to believe that this first princely ruler of Jaffn4wae no other than Mgha. In the first place, we are inclined to agree with Gnanapragasar that the name Vicaya KlMc!ni may be a corruption of Vicaya K]i?ilcq (Vijaya Kliiga) gha, as we know from the

1. !vm., p. 25; 2. See supra,

., p. 6.

3 S. Gnanapragaaar,

ppa-vaipava-nlaiga, p. 65.

Nia-sagrahaya and the 4tbavagalaviraaa, had al8o the name Kliiga Vijayabhu This name could be easily rendered into Tamil as Vicaya Klia. Here the second element of the name Vijayabhu is dropped and the name Kliga is used at the end, in the same ner as Ca or Pya in Tami1

,.&+ra k.'ro

461

Gnauapragasar has explained that in tha manuscripts K].iikp or its variant Ilixkai may have been mistaken for Since ____

aikai makes no sense, it may have been altered to

KU4ksi (crippled hand) in course of time. Hence the explanation of the author of the that the king got

this name because one of his hands was disabled. But the posibion

1. 2.

ks., )t7 ;

&Yw

Rjarja Ca, V!ra Pya, etc. Vijaya is. written as Vicaya in Tamil and Klifia as Klika or IVlifdcai.

3. In the cursive style of the ola manuscripts, where an angular style is avoided to prevent the ola from splitting, 111P1kR would have been written thus:

9nJL'

This could have

been easilyymistaken for P,4YV'Ua (I't!.ai.ka). Li.. p. 30.

462
of IC 1a.kai after Vicaya suggests that it may not be a nickname. If it were, it would have been used in fromt of Vicaya, the surname (akai Vicaya Cakkaravartti), as in the case of normal nicknames in Taniil Gnanapragaaar explanation sems to be plausible but it is rather unlikely that sich a well-known name like K1ifga was misread as IVt!aka. The corruption of Kliga into Ktax.ka may have occurred in some other way. That of the ?4a ki is a mistake for ICliga is confirmed by the evidence akk4appu-xnnm y a. This chronicle of Batticaloa

deals mostly with the history of the Eastern Province of the island, but in one place it gives an account of the manner in which the riyas came to establish their rule in Nkatvn (Pii Ngad!pa - Jaffna district) While the Ypp;a-vaipavanlai and the Kailyamlai introduce the story of the blind lutist in between the story of Nara-cik-rca and that of the first rya ruler of Jaffna, the M akk4apu-nfmiyarn has no reference to the Ypi legend Instead, the story of the first riya ruler follows that of Cii'ik R-kuIxra (Nara-cifika-rca of the pia-vaipava-mlai). According to this account,

1. ., Antaka (Blind) Kavi VTra Rkavar, Kaikkl (Short-legged) Irunrpoai, Taikka (Ele hant-eyed) Cy.
2. Mm., pp. 36-37.

3. See supra, p.4-O

463
a Ca prince named lit'iki riya, went from iriya-nu to Nkatvia (Jaffna district), invited several families from riyantu to settle down in that region and became their king. The name of the first Iriya, ruler of Jaffna is thus preserved in the Maakk4appu-rn iyam as ljkai Iriya which corresponds to KaiUcai iriya of the a-vaipava-lai This strengthens

our supposition that KtT]akai is a mistake for Kl4i3ki. Vicaya 1ki appears to be Vijaya Klif&ga or ?gha, who, as we shall seepreaently, has been confused with the first Lyacakravartin in the Tamil traditions. Secondly, it appears that Gikai Iriya was not the name of Vicaya K] 1c i but a later addition of the

chroniclers. Cikai Iriya (4rya of Cifkai) was the dynastic name of the Lyacakravartins who ruled Jaffna in the fourteenth and fifteenth centuries' In the list given in the vaipava-ni1ai, there are ten rulers after Vicaya K1afi.kai who bear this dynastic name and they all have their personal names prefixed to it But in the case of the first ruler, Vicaya Kaikai, there is no personal name prefixed to the

1. Yvm., p. 30. 2. See infra, ; i?ikii is an abridged form of Ci?dci-n&tr (Siha-nagara) which was the capital of the Aryacakravartins. , ulackara Cikai Ariyaa, Ku)ZttuiTh Ci.kai Ariya.

3.

461
dynastic name. He is just called C4i1ca1 Iriya in the Xailyanflai. and the
reign the

At the end of the account of his

a-vaipava-nilai gives Vic&ya K'tI]akai CaJ dca-

vartti and Kafikai Ariya as the other borne by Cii3.kai Iriya. This confusion is not difficult to explain. In the historical. traditions of Jaffna the Aryacakravartins have overshadowed all earlier rulers of the Jaffna kingdom. At a time when these early rulers were being forgotten, the name of Vicaya K%1'ikni may have been still, lingering in the memory of the story-tellers, who, ignorant of the identity of this person, may have identified him with the first Ariya of Jaffna. We agree with Paranavitana that the chronicles of Jaffna were 'written when the CiMti riya,s had ceased to exist, at a time when, after the dynastic name had been attached to the rulers of Jaffna for about three centuries, the belief had gained ground that all rulers of that kingdom bore that name' As a result of tn1drg Vicaya akM

identical with the first Aryacakravartin, the accounts of their two reigns bave also been mixed. It appears that somehow the personality of 14gha, under the name of Vicaya survived in the traditions of Jaffna. ati, had

I.. S.Paranavitaaa, 'The Arya Kingdom of North Ceylon', p. 203. The Vaiypal also refers to Kkai as an riya,
v. 57.

463
The conclusion that Vicaym

1ci of

the Jaffna

chronicles is gha is further strengthened by other considerations. In the Sinhalese chronicles it is stated that Ngha settled several Tamils and er4as in northern Ceylon In the Tamil chronicles, Vicaya K . Mrai is said to have been responsible for the settlement in northern Ceylon of people from India But more important than this is the consideration that the adoption by the Jaffna rulers of the couchant bull (nandi) as their royal insignia, Gaga as their vaisa or kula name and Cifikai-nakar (Siha-nagara) as the name of their seat of government indicates the Kliga origins of the northern kingdom The emblem of the couchant bull (nandi or vabha lfichana), with the crescent moon above it, was used by the Eastern Gagaa of Kaliga, as is evidenced by the seals of their copper plates as well as by their inscriptions The kinEs of Jaffna used the same emblem on their flags and coins On these coins, not only do we find
1.

1. See supra, 2. See supra, ; Tvnt., p. 27 tX.; Mm., p. 37.

3. S.Gnanapragasar, !,pa-vaipava-vimarcan, pp. 60-61. 14 E.I., III,

5. ., IV,

p. 130. p. 192.
., p. 7; Ki4jai-vi fu-t'
tu , v.152;

6.

!' p. 3k; Km., p. 5;

S.Gnanapragasar, 'The Forgotten Coinage of the Kings of Jaffna',

C.A.L.L.V, pt.4, pp. 172-179.

IA.
the couchant bull but also the crescent moon above it. In the Tamil literary works produced in the Jaffna kingdom, the rulers of that kingdom are sometimes referred to as those of the Gazga country (K__kai-n) 1 or simply as 4ryas of Gazga (KdkaiAriyar) In the Kai].yanilai some of the noblemen who served under the first ruler of Jaffna (Vicaya

ii1i)

are stated

to have belonged to the Gaga kula (Kazk-kula) These references seem to reveal the Gaga connections of the founders of the Jaffna kingdom. We agree with Paranavitana that the Kaligas who touiided the norbhern kingdom must have regarded their Gaxga connections with pride and that the 4riyas who inherited the kingdom might very well have continued these traditions. Gaga in these epithets seem to

1.

Ciappuppyiram, v. 11; Ciappuppyiram.

2. Irakuvammicani, I, v.223, XIII, v.107.

3. .,

p. 12. This term may refer to the Vefla caste

called Xkai-kulam .

467
denote some connection with the Eastern
GaAgas. Paranavitana

originally expressed the opinion that the 'claim of the riya rulers to be of the GaAga lineage can be upheld if they are taken as successors of the Jvaka kings of the aliAga-vaiaaZ and that when the KaliAgas founded a kingdom in northern Ceylon 'they must have regarded their Gaiga connections with sentiment and pride' But in a subsequent article, while attempting to refute the view of Ni1aknta Sastrit s, he has argued against his own conclusion above. Here he baa stated that KaAkai is the form that the Sanskrit Gag, and not Gafiga, assumes in Tamil and that CaiThsi may be ta.ken to be the name of a community. He then quoted from the Madras Tamil Lexicon, in which Kafikaikulam is explained as a Ve4a tribe who claim to have migrated from the Gangetic region. 'As the Jaffna tradition refers to *rya Cakravartis who had Ve3a consorts', he has argued, 'it is very likely that Kafikai in both these epithets is used with that meaning' It is true that normally GaAga takes the form K.fl1ca in Tamil. But just as Kalifiga sometimes assumes the form Kalikai and Sifiha becomes Ciiki, Gaga can take the form Kaflkai. The name GaAga-.p!4i, for instance, often
1. S.Paranavitana, 'The Arya Kingdom of North Ceylon', p. 201.

2. S.Paranavitana, 'Ceylon and Malaysia', J.R.A.S. (C.B.), NZ., VIII, pt.2, pp. 370-371.

463
occurs in the Tantil inscriptions as KMc-pi, but sometimes it was also written as Kaikaippi, as in the Pya inscription found at VmbatT!, in Trichinopoly district The fact that one of the epithets, mentioned above, refers to i-nu

(the Gafiga country) shows that Kaii.kai, in this instance at least, does not refer to a Veja community. For the kings of Jaffna to have used the epithets Ka.kai riyar and. Kax.kai
Nar, it is more likely that Kei ks j was the name of a dynasty

rather than that of a caste into which they married. ut it must be admitted that we cannot be certain that Ksik4- in these epithets used by the Jaffna kings refers to the Eastern GaAgas only. The capital- of the early rulers of the northern kingdom was known as CiAkai-nakar (Siia-nagara) or _____ Consequently the ryacakravartins who ruled from there were known as Ciftkai Iriyar or C1ki-nakar iEiyar The capital seems to have been named after the aliAga city Sifthapura. Zifihapura was the 6eat of one of the dynasties of Kaliga in the fifth century A.D't After about the sixth century LD.

1. M.E.R.

for 1923, No. 366 of 1922.

2. See infra,

p.2i.

3. See infra, p.c/" k. The Classical Age, ed. R.C.Yajumdar, (Bhratya Vidya Bhavan), pp. 212-213; E.I., IV, p. 143.

463
nothing is known about this city from the Indian sources. In the Sinhalese inscriptions of the twelfth century, the liga ruler Niafi1ca Mall4laims to have gone to Ceylon from Shapura, the capital of his father Jayagopa, in KaliAga om this it appears that as 'ate as the twelfth

century Si*hapura was the capita]. of a dynasty in Kaliga, probably a minor branch, which had escaped notice in the Indian sources. gha and some of his associates probably hailed from Sihapra, like Ni1c Ma].la. The new capital founded by the invaders in northern Ceylon was probably named after their city in Kaliga. liere, too, we agree with Paranavitana that )gha 'would have nanied the capital of his new kingdom after the city which was the home of the Kaligaa! although we do not support his contention that the home of the KaliAgaa was in 'Malaysia' These considerations lead us to think that the kingdom established in northern Ceylon in the thirteenth century had its origins in a dynasty which was connected with the Eastern Gafigas of Kalii3.ga.

1. D.M.de Z.Wickremeinghe, 'The Slab Inscription of Krti NiAka Mafla at Ruvanv.li Dgaba, Anurdhapura', E.Z., II,

pp. 80, 85.

2. S.Paranavitana, 'The Arya kingdom of North Ceylont, p. 201.

470
These ICZiI l-ga origins of the northern kingdom must date from a period before 1262. There is no evidence of a Ka1izga invasion of Ceylon after that of gha. Prom at least 1262 the northern Jr(ngdom appears to have been ruled by JvRk who seem to have enjoyed the protedtion of the
powerful Pyas At the turn of the century, the Iryacakravartine

of South India inherited this kingdom It is unlikel7 that

there was aXaliAga invasion of Jafmna between 1262 and the

date of the cession of the first 4ryacakravartin, which is not definitely known. This was the time when Pya influence seems to have been at its height in northern, and even in southern, Ceylon. Kaliga rule in the independent kingdom of northern Ceylon should, therefore, be dated to a time

1. Some place names in the Jaffna peninsula 4 seem to preserve these Kaliga origins. There is a place called Kal4rsrIya-cTma and another

called

The persona]. names Xa].iAk-rya, (Ka1ia Rja) and a-kkp-rya, (Coa-gafiga-rja), which form the first elements in these names ma denote their association with people from the Kaliga country. But we cannot be too sure of this, for both these personal names were used as titles in the Tamil country. 2. See sura, p. 3. See infra, t.-vjJ

471
before 1262. As we know, between 3236 and 1262, the X1iigas who wielded power in northern Ceylon were Mgha and his associates, among whom there seems to have been at least one probable Gaiga prince, Coaga In the light of these considerations, the foundation of a kingdom in northern Ceylon by Igha and his followers after their defeat by the Sinhalese seems to be a strong possibility. The beginnings of this new kingdom are shrouded in obscurity. The earliest rulers have not left behind any datable coins or inscriptions. The chronicles of this kingdom are very late and do not seem to preserve many genuine traditions about its beginnings. With the meagre evidence that we have, it is ot possible to.assert anything with any degree of certainty. Basing our assumptions on the course of events

in the

middle of the thirteenth century,

as reflected by the Sinhalese chronicles and the Pya inscriptions, it appears that )gha and his associates and the Jvaka invaders were closely connected with the beginnings of the northern kingdom and that Igha and his associates rather than the vakas were responsible for its fpundat ion.

1. See supra, 4-

472
We may now sumniiriae the main conclusionsoof the foregoing discussion. In the first place, the date of the foundation of the Jaffna kingdom cannot be traced exactly. That this took place between 1236 and 1262 appears to be more or less certain. It was probably between 136 and 1258, possibly between 1236 and 12k7 that it was founded. Zgha and his followers who seem to have been defeated some time after 126, in all probability, shifted their seat of government further north to the Jaffna peninsula, and founded a new kingdom. The Jvaka invader Candrabhnu appears to have found his way to the throne of this kingdom some time after 12k7. Re was probably subdued by Sundara

P4ya

around

1258. and

killed by V!ra Pya in 1262. It was probably his

son who was allowed to be crowned in full regal style as the ruler of the northern kingdom in 1262, under the protection of Vtra Pya. Whatever the uncertainties regarding the begin.nings of the northern kingdom may be, the circtunstances that led to its foundation are 14 .d.ifficult to understand. In the first place, the foreign invasions of the thirteenth century played a significant part in paving the way for the rise of an independent kingdom in northern Ceylon. The invasion of gha led to the control of northern Ceylon by Ier4a and

Tmil elements

and possibly some from Kalizga like Coagaiga.

473
These invaders could not be ousted from the inland, although they were defeated and driven away from Polonnaruva. Once they lost Polonnaruva, the pext natural step would have been to set up another capital and continue their hold on northern Ceylon. If 1gba and his associates were led, by their loss of Polonnaruva, to found a new kingdom, the vaka invaders who came next seem to have found refuge in this kingdom and helped to consolidate its position. The Pyas, who invaded the island after the Jvakas, seem to have given the new kingdom flU recognition and protection, thus rendering it difficult for the Sinhalese rulers to wipe it out. In this m p nner, the invasions of the thirteenth century, while being

fatal to

the Polonnaruva kingdom

and. limiting the power of the Dabadeiya kingdom, helped the rise and consolidation of a new kingdom in northern Ceylon. The foreign invaders of former times were able to wield power in the island only for short periods. On all the earlier occasions Sinhalese princes were in a position to drive the enemies out of the island after some t&.ine. But in the thirteenth century, when enemy after enemy sacked the country, the Sinhalese kiTiga were in no position to oust them completely from the islad. gba and his Ker4a-Tamil forces left Polonnaruva but not the inland, The vakas were defeated but were apparently allowed to find refuge in the north. Thepansion of the Pyae could not be contained by the Sixihalese. They were allowed to reduce

474
the new kingdom in the north to the position of a tributary and thus give it recognition, dashing the hopes of the Sinhalese to wipe it out. If Parkramabhu II,invited the help of the P 4yas , which he probably did, against Candrabhnu and if be

had any hopes of annexing the new kingdom and re-unifying Ceylon, the results showed that the Pyas were only: willing to keep the peace in the island but not to allow the annexation of the northern kingdom which was tributary to thei. The fall of Polonnaruva and the drift of Sinhalese power to the south-west were also notable factors that helped the rise and survival of a kingdom in northern Ceylon. With the shift of the Sinhalese capital to the south-western parts of the island effective control of northern Ceylon was lost. This made it easier for an independent kingdom to emerge in that region. The foreign elements, who were driven to that part of the island, exploited the circumstances to found a new kingdom. With the abandonment of the region around Polonnaruva, the northernmost regions were virtuafly cut off front the south, The chieftaincies in the southern part of the formex RAjaraha acted as a buffer between the northern and southern kingdoms. The re-unification of the island became difficult even on occasions when either of the kingdoms was subjugate the other, as they did in the fourteenth and fifteenth centuries. In the thirteenth century perhaps the abandoned regions of Rjaraha may not have provided such a barrier to the subjugation of the north. But it would have

473
given protection to the new kingdom from the Sinhalese capital which now lay farther south. The political conditions of the first half of the thirteenth century were, therefore, favourable in many ways for the rise of an independent kingdom in northern Ceylon. The conditions in the second half helped to consoliiate its position. The power of the Dravidian elements in northern Ceylon, with whose support the invaders held that area under their controi, is a facbbr that cannot be overlooked in this respect. In the thirteenth century, the Dravidian elements were more powerful than ever before. They bad grown from strength to strength with almost every invasion that rocked the island since the onslaught of the Cas. There was a steady increase of the Dravidian element in the island, especially in the northern parts, from the time of the Ca invasion. The invasion of 1gha let loose in the island a further band of Ker4as and Taiwtls whose leaders established their authority in.imany parts of northern and eastern Ceylon. These were the elements that provided the greatest strength to the anti-Buddhist regime of )1gha. More South Indians seem to have been invited for settlement at this time. The inevitable culmination of the forces that were at work there was the establishment of an independent authority that enjoyed the support of the Dravidians. Those South Indian

41(0
leaders who established petty chieftaincies in the Vanni districts seem to have acknowledged the authoriti of the new kingdom. However, the northern kingdom was not a Tamil kingdom at the beginning, ,though Tamils and Ker4as probably formed a substantial section of the population. It was with the advent of the Aryacakravartins that it became a kingdom ruled by a dynasty from the Tainil country and. gradually evolved into a kingdom of the Tamil-speaking people.

Ipl,.

CHAPTER VII

477

THE BEGINNflGS OP THE XIDOI4 OP JAPPNL - II The Dynasty of 4ryacakravartlns

The Tamil chronicles do not mention any Klizga or J5vaka ruler as having ruled over the northern kingdom. It is the ryacakravartins who are described as the first princely rulers of this kingdom and. are given the credit for its establishment on a firm footing. In his paper entitled 'The Lrya Kingdom North

Ceylon', Paranavitana has dealt with the origins of the dynasty of ryacakravartins and their rule in northern Ceylon This paper forms the latest contribution to this subject and revises in many ways the views held earlier by Rasanayagam and other writers on the history of the Jaffna kingdom. We are inclined to agree in the main with the conclusions of Paranavitaua, aLthougb we find some of his arguments unacceptable. His contention that the 4ryacaJavartins of Jaffna came from Rnfvaram in South India is convincing and is supported by evidence not adduced by him. The earliest source in which an ryacakravartin is mentioned is an inscription from ChaturvdixnafLga.am, in the Tiruppattr

1. J.RA.S. (C.B.), N.S., VII, pt.2,, 1961, p. 17k-22k.

473
tluk of the Rmnd district The astronomical data in this record has been worked out by Swe m1 kki nnu Pillai as being probably equivalent to September
9,

l27l But since it was inscribed in

the fifth regnal year of IAavarma Ku1akhara I, who ascended the throne in 1268, this inscription may be dated to 1272. 'An agent of the chief Ariyachakravartig4' is referred to in this record Another inscription, from Sivapuri in the same tluk and dated in the same regna]. year, mentions a certain Dvar 4riyaccakkaravarttika Re was probably the same as the chief mentioned in the first inscription. The astronomical data in this record, according to SwmUdcnnu Pillai, correspond to September 5, l27k But this seems to be a mistake. The third inscription, which Paranavitana erroneously quotes as the first record mentioning the Xryacakravartins, is found in rTrai.gam in the Trichinopoly district and is dated in the tenth regnal year of ravarxna Kulakhara, which is 1277

1. M.E.R. for 1927/28, No. 290 of 1927/28. 2. Ibid., p.

li6.

3. Ibid., p. 57. As the record is unpublished, no details are available. If. I .E.R. for 1928/29, No. 21 of 1928/29. Unpublished. 5. Ibid., p. 53. 6. I'LE.R. for 1936/37, No. 7 of
1936/37,

pp. 8, 75.

479
The astronomical details in the record, which correspond to December ,

1277, confirm

this date It reoords the grant of a

plot of land by one Matitu lc a (Natitufiga) who bore the title of Iryacakravartin as well as the epithet Tai niu vea perun1 ('the chief who stood alone and wthn)). He bailed from Cakkaravartti
Nallr, in Cevvirukkai-nu. Cevvirukkai-nu has been identified

2 as a territorial division in modern Ramnad distr&ct. A fourth inscription from Truvaraigu.am, in Pudukktai, 'registers a political compact between Iba, Iryacakaravartin and Srya 2 on the one sde and AiyrkunaUnd Kuppai on the other' The record is dated in a regna]. year of V!ra P3yadva. But since it is damaged, neither the year nor the throne name of the monarch is clear. There were two Vra Pyas in the period of the second Pya empire. Both had the throne name Javarma. One ruled between 1253 and

1268 and

the other between

1296 and

l3k0

There was also a prince called V!ra Pya whose rule was

confined to the South Arcot district, as is evidenced by his inscriptions He may not be the same as the V!ra Pya of our

1. M.E. T .

for 1936/37, p. 48. fn.127. 276 of 191k. 174, 201, 240. 147.

2. S.Paranavitana, 'The rya Kingdom in North Ceylon', p. 2 07, 3. M.E.L for 1915, No.

4. K.A.Ni1sknta Sastri, The Pyan Kingdom, pp. 5. Ibid., pp.

233, 245 ; T.V.C.Pantarattar,

Piyar Vara1z2, p.

48
inscription, which is from PudukkVai. The latter may, therefore, be either Javarma Vira Pya I or II. Lastly, there are two inscriptions from Tiruppul1i, in the Bamnad. district, mentioning the jryacakravartins One of them is dated in the thirty-eighth regnal year of Z"avarma, Kulakhara I, which is either 1305 or 1306 It registers a grant by var 4riyacakkaravarttik4. He is identifiable with var riyaccakkaravarttik4 of the second inscription mentioned above, which is also from the same district and belongs to the same reign. The other inscription from Tiruppulli, which is badly damaged, gives the names of two personages, Teyvaccilaiy Aaka, alias Ariyaccakaravartti and Irma alias Va. kkai *riyacc akkaravart ti, the amm of

Parkkirama Piya (Parkrama P4ya) Unfortunately the date of the record is not known. But the fact that one of the persons mentioned here is called. an ainn (uncle or father-in-law) of

Parkrama P4ya may help us in the dating. We know of at least five Parkrama Pyas who ruled in the southern territories of Panmad and udukkfai in the fourteenth century, after the fall of the second empire The first of them, Jatvarma Parkrama

1. N. .R. for 190k, No. 110 of 1903; S.I.I., VIII, Nos. 396 & 398. 2. .1.1., VIII, no. 39G.

3. Ibid., No. 398. 4. LA.NilskRnta Sastri, The Pyan Kingdom, pp. 215 - 246.

481
Pya I, began his rule in 115 and the last of them ruled till at least 1k15 There are certain considerations which lead us to think that this inscription belongs to the earty part of the fourteenth century. This record is from TiruppuUi, in the Rannad district. The other inscription from the same site mentioning an Aryacakravartin is dated 1305/06. Further, all the other inscriptions from the Pimnad district referring to the Aryacakravartins are datable to either the latter part of the thirteenth century or the early part of the fourteenth century.There was no Pya ruler called Parkrama in the thirteenth century. It is 1 therefore, likely that our inscription belongs to the time of Javarma Parkrama_P4ya I, who lived in the early part of the fourteenth century.

Irm

alias Va..kkai

Ariyacc&ckravartti was probably the anim of this Pya ruler. The evidence of these six inscriptions is practically all th*t we learn about the iryacakravartins from the IncIia.n side. That the 4ryacakravartins were niiior chieftains is clear frog the information we get from the rrafgam inscription

1. K.A.Nilakanta Sastri, The Pyan Kingdom, p. 2k5. The five P74 rulers are: a) Javarma Parkrama Pya, 1315-13k?; b) avarzna Parkrama Pya, 1335-1352; c) Javarma Parkrama Pya, 1357- c.1380; 4) Parkrama P1ya, 1367till after 1387; and e)Paiikrama Ptya, c.].38 k- till after 1k15.

482
about one of them. As Paranavitana has remarked, 'the title 'Devar' applied to this Xrya-cakravarti, the fact that his order is called a tirumukam (verbal order) and that there was an officer whose duty it was to write down his orders, indicate that he was a ruler; but his record being dated in the regnal years of the Pya emperor establishes that he was a feudatory' Al]. the other inscriptions, except the last one,are also dated in t'egnal years of the Pya kings. The last inscription, from Tiruppulli, shows that some of them were related to the Pyas, presumably through marriage. From the Sinhalese sources we learn that in or about 128k there was a Pya invasion of Ceylon under the command of an Iryacakravartin This, too, supports the conclusion that the iryacakravartins were subordinates under the Pyaa. It is, therefore, clear that they were feudatory chieftains of the Pya country. The fact that two persona with the name ryacakaravartin are mentioned in one record, namely that from Tirnppu11i, map indicate that this name was used as a family title or name, unless the two persons mentioned here are of different generations. This is not clear from the record.

1. S.Paranavitana, 'The irya Kingdom in North Ceylon', p. 207. 2. Cv.,

90:k3-k7; Dalad-sirita, ed.

Sorata Nayaa Thera,

1950, p.1+5.

483
Of these six inscriptions, four are from the Pamnad
district. Of the other two, that from Trichinopoly gives the

home of the iryacakravartin mentioned in it as Cevvirnkkai-nu, which was an bid terr&torial division in the Raiinad district. The sixth,from Puduktai, does not indicate that the Aryacakravartina held authority there. This inscription records a political compact that an Iryacaiq,avartin and another person entered into with two others Such a compact was probably made at the end of a battle or a political mission in which an ryacakravartin was representing his PIya overlord, as an Iryacakravartin did in Ceylon, or which be conducted himself as a petty chieftain. The details are not known owing to the damaged nature of the record. We do not get information regarding any iryacakravartin in any other part of South India. We are, therefore, inclined to agree with Paranavitana that it is reasonable to conclude that the home of the Lyacakravartins was in the Ramnad district,, as is indeed claimed by one of them in the inscription from Trichinopoly They probably bade their chieftaincy in that district. Of the six inscriptions under discussion, the four datable ones belong to the period between 1272 and 1306, which

tLct

1. See supra, p.47q

2. See supra,

p.4)1k.

484
is in the reign of Iavarma Ku1akhara I (1268-1310). Of the other two, that from TiruvaraJa4am may very well belong to this period if the VTra PIp4ya nientined. in it is Javarzna, V!ra P44ya II If be is the first of that name, whose accession took place in

1253, this

inscription woizld be a few, at the

moBt nimmeen, years earlier than the four datable ones. On the other hand, if the inscription belongs to the latter part of the reign of V!ra

P4ya II (1296-l3kO), it

would be

later than the others by a few years. The sixth inscription, as we have already seen may belong to the early part of the fourteenth century. We see, therefore, that most of the South Indian inscriptions mentioning the Aryacakravartins, or probably all of them, belong to a period of about four or five decades in the latter part of the thirteenth and the early part of the fourteenth century. It is in tha same period that the Siithalese sources refer to an ryacakravartin who led invasions on behalf of ?avarzna, Eulakhara I. It is interesting to nte that all the datable inscriptions in which the 4ryacakravartins find mention belong to the reign of ?avarma Kulakhara I. It appears that the ryacakravartins belonged to a short-lived dynasty of fendatory chiefs who held away in the Rmriad district in the reign of Zavarma, Ku1akhara and possibly a little longrr.

1. See supra, p. 1.'V 2. See supra, p. i1.3(

485
Their inscriptions are not found in the Ramnad district after this period. As mentioned earlier, Paranavitana is right in tracing the origin of the ryacaicravartins of Jaffna to this family from Ramnad. Apart from the fact that the latter are the only other ryacakaravartins th t we know of, there are other considerations that lead ud to this conclusion. The chronology of the iryacakravartins is in favour of such a conclusion. The last reference to the iryacakravartins in South India occurs in the inscriptions of the early part of the fourteenth century. The first definite reference to an Iryacakravartin ruler of Jaffna is made in l3kk not long after the South Indian references. But more important than this, as Paranavitana has indicated, is the fact that the Aryacakravartins of Ramnad and Jaffna used the word ctu in their records in the manner of a benediction, as is evidenced by the Tfruppulli inscription of an ryacakravartin of Ranxnad and the Koagama inscription of an Aryaeakravartin ruler of Jaffna The word ctu was also inscribed on the coins of the Aryacakravart ins of Jaffna They probably

1. See supra, p. 400

2. S.I.I., VIII, No. 396 ;

H.C.P.Bel]., Report on the Kgalla

District, p. 8. 3. S.Gnanapragasar, 'The Forgotten Coinage of the Kings of Jaffna',

A C)

had this word inscribed on their shields as well, for de Queyroz, while describing the defeat by the Portuguese of one of the last kings of Jaffna, mentions that be 'had a black inscription on his white ahield' The Aryacakravartins of Jaffna had also the title ctukvala ('protector of the ctu') C!tu (= darn or causewa) is the name applied to Ir.mar Lai ('Rnia's Bund' or Adam's Bridge), the narrow coral reef that ionnecte the island of Nanr with that of Rmvaram on the Thdian side. Ctu is also applied to the temple of Rnivara4zid it is in this sense that it is better known, The Aryauakravartns of Jaffna, till the rise of the 6tupatis of Raninad, considered themselves as the protectors or custodians of the temple of nvaram The occurrence of ctu in their records shows that the Aryacakravartins of Ramnad also had a special association with the temple of Rni!varam which, being situated in Ranuiad, probably came under their protection. Further, there were traditions in Ceylon which connected the ryacakravartina of Jaffna with Rxx1varam. In the Cekarca-ckara-nlai, one of the earliest Tami]. works

1. F.de Queyroz,
2. v. 86;

. cit., p. 366.
., Ciappupp7iram.

3 S. Gnanapragasar,

pa-vaip ava-vimarc an, p 62.

48?
produced in Jaffna, the ancestors of the ryacakravartius of Jaffna are said to have been Brhmaa rulers of antamtaam It is stated that Rma, after his campaign in Ceylon, want to Kantanitaam, established a shrine there after his own name (inivaram) in honour o Siva. and invited five hundred and twelve Pupata BrThmaas to perform service in the temple. Two of them were made kings and given the title of '4riya-vntu'

(Lya king)

together with the insignia of umbrella, Brahmanic

thread and the bull standard. Kantamtaam is the name given in the literary works of South India to a hill in the vicinity of R iiivaram 1.de Queyroz, too, records a tradition that the ancestors of the kings of Jaffna lived in nvaram Although the late ,a-vaipava-mlai claims a Ca descent for the

4ryacakravart ins, the earlier work KaiIyanlai states that the first ryacakravartin came from the P4ya country and that be was a Pya prince We have mentioned earlier that some of the iryacakravartins of Ramnad were related to the P4yas, presumably through marriage. But there is no evidence to suggest

,. Tvai U1, vv. 179, 220 ; Kampa Rmyaaxn, VI, Yutta-ktam, Paikkci Pa t alam , v.15, 1'fftci Paa1am, v. 168; Kantapuram, Makntira Kam, I, v.18. t. Ccm., Citappuppyiram. . F.de Queyroz, . cit., pp. k8-49.

k.

Km., p.

6; Yvm., p. 25. The Yvm. states, however, that he came

from Nadurai, the P14ya capital.

483
that they were scions of the PIya family. The Kai1yalai statement may be an exaggeration. That the ancestors of the Iryacakravartins of Jaffna cxnae from the Pya country is further etidenced by certain traditions relating to them which are recorded in the Cekarca-ckara-nlai In the jappupiram (introductory section) of this work, an eulogistic description of the achievements of the ancestors of Cekarca-ckara, the Jaffna ruler in whose time this work was composed, is given. In this account, there is an allusion to 'the king who fought and defeated the KrunI Lkar (1Car1as) at Antaravalli' (ce

Karunakarai Antaravalliyil porutu ceyitta vntu) 1 and to 'the king who, having dismembered the trunk of the rutting elephant that dashed angrily towards him, defeated the Pc4a, (Boysaja)' (i varum mata vam karam tuittu Pc4ai_katinta_vntu) These two statements appear to be allusions to the achievements of one person. It is not known that a ruler of Jaffna ever got involved in a battle against the Hoys4as, the last of whom ended his rule around l3k3 It is not possible that a ruler of i1Jaffna carried out a succesful expedition against p powerful Koys 4as. The reference is evidently to an 4ryacakravartin of

1. Ccm., v.6. 2. Ibid. 3. K.A.Nilakanta Sastri, A History of South India, 1958, p. 231.

489
Ramnad who, as the ally of the P4yae, probably won certain victories over the Eoys4as. This was probably before 1310, for by that date there were internal dissensions in the

Pra

kingdom

and the Piya princes were in no position to attack their neighbours The epigraphic sources inform us of victorious

P4ya

campaigns against the Hoys4aa only after the accession

of 11avarma, Sundara

P4ya

in 1251. In an inscription of

his seventh regna]. year, Sundara P4ya claims ba have inflicted a severe defeat on the aoys4as Following this success, the Pfras seem to have been in occupation of Kaar, the Hoys4a capital, for quite some t1me The second time we hear of a campaign against the ops4as is in the time of }avarma Ku1akhara I (1268-1310). In an inscription from Tinnevelly, Ku].akhara claims the subjugation of the Uoys4as After this we do not hear of any successful Pya campaigns against the

Hoys4a neighbours.

The Iryacakravartin, whose achievement in

a battle against the Hoys4as is alluded to in the Cekarcackara-n1ai, may have been fighting on the side of the Pyas in one of these campaigns. Probably he was involved in the

1. K.ANilakanta Sastri, The Pyan Kingdom, p. 201 ff 2. Ibid., p. 161 If. ; .E.R. for 1 3. LA.Ni]akanta Sastri, he P.yan 11. . . for

9k, No. 166 of 189k. Kingdom, p. 16k ; El.,

III,pp.11f

1926/27, No. 29 of 1927.

49k)
battle of the time of Kulaflhara for, as we know from the Cflava4sa , it was in the latter's reign that an Iryacakravartin
led some of the PIya campaigna The allusion in the Cekarica-

c!kara-nl!lai seems to be a genuine tradition preserved in the courts of the iryaca1cravartins of Jaffna. In the light of afl. these considerations, the conclusion that the Iryacakravartins of Jaffua were descended from those of Panmad becomes irresistible. In his paper on the kingdom of northesin Ceylon, Paranavitana has made some interesting comments on the origin of the Iryacakravartins of Pinmad. Arguing on the basis that the word 'Ariya (Irya) has a distinct connotation in T2nil
literature', namely that 'it denotes the language, literature

and people of North India, as distinct from those of the Dravidian lands', he baa euggest.d a North Indian origin for the Iryacakravartina Be argues his case as follows: Even if, as stated in the Cekarca-cTkara-nilai and

reported by de Queyroz, the rulers of Jafina were called 1riyaa due to descent from a Bralimin of .ni!varam, it is not the fact of this ancestor being a Brahain, but of his belonging to a particular class of Brnhmlnn, namely Iriya Brahins, that would have conferred on them the titi. of 4riya. For there are still at

1. Cv., 9O:k3-k7 2. S.Paranavitana, 'The 4rya Ifngdom in North Ceylon', p. 2O+.

491
Izvaram a sect of Brhm{ns called 4riyappirlm4ar, who hay, special rights in the temple and who claim to be immigrants from North India. The North Indian origin of the Brahmins to whom the 1riyaa of Jaffna traced their origin is admitted also by the CeIcrca-ckaranfElai, for it says that they came with Rma. A stray verse, ascribed to Puka2nti, included in the anthology called Tami-n1valar-caritai, seems to support the Northern origin of the rya kings. This stanza, which expresses the poet's grief at the death of an lrya king, refers to him by a phrase which, in the printed text, is given as vaal ri r-kThifAp,. The compound vaalIriyar can only be analysed as vaal+L'iyar. But the first of these two words, according to the Tamil Lexicon means banya, and gives no sense in this context. As v can be confused with ]. in Tamil manuscripts, the avriyar-knf, the correct reading appears to be king of the Northern Kryas. De Queyroz definitely mentions that these Brahmons of RInvaram came from Gufar1t which, together with the adjoining regions, is called &riaka (Iryaka) by classical geograpber8. De ueyroz further states that these Irya Br.hmirn claimed royal descent. This is rather puzzling, for the rigidity of the Hindu caste system of those days would not have permitted a qatriya being accepted as a Brhmin. It is possible that originallythere were at RLfl!varam and its neighbourhood Br2h mine as well as KSatrlyaa who called themselves lryae, and that, at a later date, when the Brahmina alone_succeeded in preserving their separate identity, all 4riyas of Rn1varam known to tradition were held to be Br.hm(n i. The fact that the rya rulers of Jaffna wore the sacred thread need not necessarily prove that they were of Brahmin origin; the qatriyas, too, were entitled to wear it. 1 Having thus argued

in favour

of a North Indian, and possible

EatrLya, origin for the Iryaca.kravartins, Paranavitana proeeda to 'ascertain who the Arias were' For this purpose he takes 1. S.Paranavitana, 'The Zrya Kingdom in North Ceylon', p. 20i4 205. 2. Ibid., p. 205.

492
two independent references to Iryas in the sources relating to the thirteenth century and, on their basis, attempts to identify the iryae as RIpute. The first reference is that in the C!!1ava to 4riyk-Ichattiya (Irya-katriya) warriors

who were mercenaries in the service of Vijayablhu IV (l27O-1272) These iryakatriya soldiers have been identified by Co&rington as RIjputa The second reference is that in the inscriptions of Ja1varma, Sundara PIya I (l25l-c.l270) In this record, certain ryas are mentioned as the 11feef MudugUr. Irishna Sastri identified these Iryas

as the

Cas, as it was believed

at that time that the ryas of the CUlavasa references were CVas Disagreeing with Saetri, Paranavitana argues But,_aa it has now been established. beyond doubt that the Aryas of the CUlavqtaa were Bjputs, the Aryas who fought iwith the Telugue must also be similarly identified. In later writings, the Irya families of Jaffna are associated with a place named 11aap4i, the name of which was borne as a title by the descendants of the last king of Jaffna. A place of this name is said. to have existed in the do mi nion of,the XTkatTyae, which is called the kingdom o l4otupal].i by )arco Polo. It is possible that the Iryas referred to in the inscriptfon of Javarman Sundara Pya as the allies of the Telugus lived in this place, and later, after the conquest of

1. .' 90:1 ff. 2. H.W.Codrington, 'Notes on the Dabade.i Dynasty', C.A.L.R., X,

p.88.

3. M.E.L for 191k, lbs. ii. Thid., pp . 91-92.

332, 3k0

and

36].

of

1913.

493
the 7katTya kingdom by the Muslims, they migrated southwards and joined forces with the Irya-cakoavartia of RInvaram, to be mentioned in the sequel, and ultimately fount their way to Ceylon. It was at this time, or somewhat earlier, that the Rljpit kingdoms in North India collapsed under repeated attacks by Muslim invaders, and bands of warriors who survivd the disasters, but were not prepared to lead a dishonourabl. existence under the yoke of the foreigners, might very well have come southwards seeking new homes, and taken service under rulers of Indian faiths and culture who welcomed them and were ready to take advantage of, and pay for, their military prowess. I! these Rljput exiles came as far as Ceylon, they might as well have sought their fortunes under theirulers of South India. And. there is epigraphical evidence for the presence cZ in the country near RIzfl!varam of chieftains named Irya-cakravartis about the close of the thirteenth century. 1 Whatever the possibility of the Aryacakravartins having been Rljput in origin, Paranavitana's arguments in favour of it is not quite convincing and the evidence he adduces is not always correct. In the first place, ho puts forward five arguments to establish the North Indian origin of th. Iryacakravartine. The first argument that the word ir ya has a special connotation in Ti1, in that it denotes the language, literature and the peopl. of North India is generally correct. But there seem to have been aertain exceptions to this ma'i- An inscription from KuttUa.a, in the Tinnevelly district, dated in the fifth year of )avaraa Vikrama Pa (1288), refers to two Tmi1

1. S.Paranavitana, 'The 4rya ingdom in North Ceylon', pp. 207-208.

491
riyar, Lakkum4a

DvanIya)cn, and Lekkum4a

Iapijii They

were assigned a certain extent of laud by the local village assembly and required to cultivate it and pay taxes to the temple at Tiru-kuttilam. Ws do not know who the Tem{. iriyar were and how they got that name. But the fact that they were called
T&flLt

I 4riyar shows that some people from among the Tmila.

were also known as 4riyar in the thirteenth century. Therefore, one cannot be too sure that the element Irya in the name *ryacakravartin denotes North Indian origin beyond any doubt. The second srgument of Paranavitana that it is the fact of their ancestor belonging to the class of irya BrThmaas that would have conferred on the 'Zryacakravartine, if they were Brhn4as, the title of riya, is certainly a possibility, although it is only an assumption. Thirdly, his coutentiom that the North Indian origin of these rulers is 'admitted also by the Cekar1ca-ckara-nilai, for it says that they came with Pma', is wrong. This Tamil work clearly states that Rma built the temple of RImvaram on his way back from Ceylon, 'invited five hundred and twelve PIupatas'(Pcupatark4

1. ?4.E.R. for 1918, No. 1126 of 1917. The astronomical data in this record corresponds to October 29, 125k, but there seems to be a mistake somewhere for )avarma Vikrama PIyas accession was in 1283 (Ibid., p. 112).

495 aii2rir.ii pairuvarai

varavalaittu) and r.queated theni to perform service at the temple (ptcaai ce y ini, nr ea karuai purint*) It is not stated from where the Brli m.4as were invited. In any case, on. cannot attach much importance to this legend. lourthly, he takes a stray Terse in the Tamil-nvalar-caritai, makes his own emendatiozi to a phrase in it and uses it in support of hi. argument. Th. following ar. the first two lines of the verse in which the verse, given by hi m as vaailriyar kThn and emended as avriyar occurs:-

A I vitiy ,al Ariyar k3n E valar;l iranta 2

(1hZ Is this fate I (Cure.) th. day the valiant king of the Iriyar died at the hands of the messengers (of Death) ). The phrase in question is actually al Iri yar k5n and not a). iriar In the above verse, the word aa]. is preceded

by vitiy. When the initial vowel a ( ' 9') of aa1 combines with the final vowel! () of yitiy, th. consonant v (ii) is introduced for euphonic reasons, in accordance with Tamil gr. mm*tical rules. The whole line would then read as

1. 2

. TV. 1-4.

2. T n " i 2-nIvalar-caritai, ed. T.XannRuntarampillai, Mad. 1921, p. 52.

496
A I When the words are separated, they would rea& as in the Vera. quoted above. That the word which qualifies Iiyar- is

aal and not vaaL is further demonstrated. by the fact that the initial letter of this word, namely a ()4), a1U.terat.s with the first letter of the whole line, in keeping with the rules of the pj metre Aal, me.nlng strong, 'valiant', 'tough' h or

or 'ability to kill', is a very common epithet for

heroes, elephants, lions and armies, and has been in use from

1. Note the elliteration in this verse: A I vitiy aal riyai, kn E Ivalar*l iranta nI - O Tarukk4ilum kulirnta ta4i tan'cu ia Tirjum_cuunf_ti?

The alliteration: I ---------j------0 a (vowel-vowel)

E1

497
very early tim.e We see, therefore, that there is no reason to

1. 1)

gi_nikI, 11,

, v. 16 (ad. Cuv1' l-Ita Paitar,

Mad. 1909); 2) Pikala-n ik a , X, stra 15, (Rippon Press, Madras 1 9 17); 3) aa1 kari (valiant elephant) in

}dirv'carksr

Tiruvcakam , Ilittal Vi4 appam , v.32, p. 182 (ad. Cuppir.aViya Pi D ai , Mad. 19119); 11) aal an (valiant lion) in N!lak!ci, Tarumavurai Car ude m, V.
55 p. 2k (ad.

L.Chabavarti, Mad, 1936);

5) aa1 !y matann earam (the fatal dart of Cupid) in Kanparinflya, PIla-kaiu, Ka(m4a Paalain, v.11, P. 83k (ed. V.WIGopIla ri4anicIriyar, Mad. 1953); 6) aal arakid (th. strong Rkasi), ibid., Ir.4iya-krLam, CUrppaakai 'UI Paa1a', v.]Q, p. 54 (1953); 7) a] Irva4 (the valiant RIv.4a) , ibid., Cuntara-kIam, Poil-irutta Paa lam , v.20,

p. 53k (1955); 8) aealvaliarakkan (the Rkasa with strength

and the ability to kill), ibid., Yutta-kaa, Irv4axj, Yatai Paa1an, v.18, p. '+8 (pt.2) (19k8) ; 9) aa]. kein tit vran (the hero with strong broad shoulders), Kant a-puniam, - Mak!ntira-k4ani, VTravku Kantantaan Ccl Paalaa, v.22, p.111, (ad. M.T.flukavi, Mad. 1907); aa1 katir vl valiant kiig with the shining spear), Puk4Tnti, r,litoar..k7am, v.37, p. 386, (ad. K.RghavcIri and T.C.PirttacLati, Mad. 1938). (the

498
emend the epithet of 4rirar k5a in the abovs vere. and to clai North Indian origin for the riyacaavartina on that basis. Besides, the authenticity of the veres under diectesion is open to question. It would not be inappropriate to quote here Paranay ita.na's own comments, in an .arlier section of the same article, on this and another verse attributed to Puka3nti in the Tami].-nva1ar-carit ai: it may be stated that the anthology in question is a recent compilation in which stray verses attributed t b y tradition to various poets, together with anecdotes about the poets, have been collected togeth.r. It is a work of the same tpe as the Sanakrit Bhojaprabandha, and in the attribution made in suck works have to be critically .rm(ned before they are accepted as correct. Th. verses in qi.stion do not occur in any of the works which are attested to be of Pu]anti ............. Nilknta Sastriialso admits that works with little or no claim to literary merit have been fathered on Puk flnti. It thus follows that the verses attributed to Pn1rnti are not beyond question from the hand of that poet, and that his date too is a matter of controversy, literary crtticG being inclined to place him in the lat. thirteenth century. 1 In addition to the doubt that has to be cast on the authenticity of the verse, there is nothing in it to indicate that the

Iriyar knf referred to then, was none other than an lryacakravartin. Finally, Paranavitana adduces th. evidence of de Queyroz who mentions that the Br1a ancestors of the ryacakravartina came from Gujanit. In this instance, despit. th. many obvious '

1. S.Paranavitan&, 'The irya Xiugdoa in North Ceylon', p . l7.

499
"4-

found in his workAaeema to b recording a genuine tradition for we learn from inscriptions in Gujar&t that there was a community of peopl, known as Iryas in th. time of the Pratihiras. A Sansbit inscription of Kakkuka of the PratThlra dynasty, dated in the Samvat year 918 (A.D.861) and found at GhaiyV.I, a few miles north-west of Jodhpur, mentions 'that lciilrp

obtained great renown in the countries of Trava!, lTalla, Ma, amongst (the people known as ) rya, in Gurjjattarl, and in Parvata in the La country' Another inscription has a reference to the AjJa, which D.R.iandarkar has equated with the L'ya ot the first inscription It is possible that some of these Irya people migrated to RImvaram after the Muslim invasions and the Iryacakravartins may have been among the descebdants of these people. Of the arguments put forward by Paranavitana in favour of a North Indian origin of the Lyacakravartine, that based on the evidence of do neyroz is the only convincing one. The case presented by Paranavitana for the identification of the ancestors of the ryacskravartins with Rljputs rests entirely on th. identification of the Liy.kirhittiyas of the C5lav4ia with Rjput mercenaries. The latter identification

1..., IX, Inscription No.38, pp . 277-281. 2. Ibid., p . 278.

500
need not be questioned. Codrington's identification that the Liya mercenaries vho were in the service of VijaabThu IV were Ijputa ma well be accepted. As Paranavitana has stat.d, it is quite poeible that there were jput soldiers in South India, too, at this time. But Rljput$ were not the only Iryas known in South India in the thirteenth century. The South Indian inscriptions refer to different grotipe of people who were known as

Iryas,

all of whom cannot be identified as

Rljpnts. As mentioned earlier, an ii$ecription from Iuttlam refers to two Ta mil riyar, who appear to have been cultivators Another from Taujore, dated in th. fifth year of Eu]Zttu.dga Ca, probably the third, refers to the Mah5vara Iryas who were attached to a. templ. Besides the inscriptions of Ja r aYarma Sundara Piya mentioning th. ryae of Xudugr, another record of th. same monasch, found at Cidambarani, alludes to the defeat of the 'fierce army of riyar' (Tern paai riyar) Another Piya record x.fers to order authorising certain

Iriyar to guard the gold treasury (kval) of the temple of

1. See eupra, P. 2. M.E.R. for 1918, No. 23 of 3. N.E.R. for 1893. No. 172 of

1918. 1092.

501
rTraAgam All these ryae of the South Indian inscriptions, belonging to the thirteenth and fourteenth centuries, cannot be identified as Rljpute on the ground that the Iryas mentioned in the C!lavasa were Rljpute. It is clear from the evidence of the South Indian inscriptions that there were several groups of people who were known as Iryas. The Mhvara &ryas, for instance, were a. class of Brhms$aa. In the CV2a inscriptions we get several references to these 1Qivaraa in connection with the administration of temples and Brh m a s.ttlenients We do not

know the identity of the army of Iriyar mentioned in the inscription from Cidambaram and of the riyar who were temple guards at rTragam. Judging from their occupation, the last two groups may very well have been Rjput mercenaries. The foregoing evidence, shows that it is not easy to trace the origin of the element Ir y a

in the

name Iryacakravartin to Rljputa.

An analysis of the little evidence that we have shows that there are several ways

in which

the origin of the

ryacakravartina can be exL4ined. The first possibility is the Iryacakravartins belonged to a commirnity of people called the lriyar who lived

in the

southern parts of the PIya country.

1. kLR. for 1938/39, No. 84 of 1938/39. 2. M.E.R. for

1923, No537 of 1922; M.E.R.

for 1926, No.9k of 1926;

M.E.R. for 1927, No.279 of 1 92 7; X.A.Nilnta Sastri, The Cas, pp. 11.27, k91.

502
We have seen .arli.r that the home of almost all the 4ryacaa.. vartins mentioned in the South Indian inscriptions could b. traced to the P mnId district Although the titi. Of lryacakzavartin seems to hay , gone out of us. in this region after about the early part of the thirteenth century, we come across some other titles, with the element ir y a, which were used ir ths Remn aci and inneveUy districts in the thirteenth century and later. One such title was Ayya, (irya), as in Ayya MaavarIya

Cc. thirteenth century) 2 and Ayya Nayi

(A.D.l582) Another

was riyapperun occurring in an inscription from rT-villiputtt!r in the Pamnad diatrict It is dated aka 1k82 (A.D.1560). About the same time, a territorial division called Iriya-nIu) (rya country) is mentioned in the inscriptions of Pnmad district and of the adjoining district of Tinnev.11y. For instance, one inscription from SrT-vilhiputt!r, in the Rrnad district, records the gift of a pio of land

in a

village in 1riya-nIu But a

large number of inscriptions mentioning riya-nu are found in

1. See supra, p.4t3. 2. LE.R. for 1918, No. 428 of 1917; E.I., W, p. 72. 3. LLR. for 1918, No. 600 of 1917. 4. M.ER. for 1926/27, No. 531 of 1926. 5. Ibid., No. 524 of 1926.

503
Te 1c!iR i tluk, in the Tinn.v.11y diatrict Most of them belong to the sixteenth c.ntury In many of them the variant ri-nIu is used instead of riya-nu, together with the prefix or vaa (north) That Iri-nu is a variant of 4riya-nIu is established by the mention of certain villages as being situed in Iri-nIu in some inscriptions and Ariya-nau in some otbers The mention of Puliyili TeRi Tiru-kuLLLa$ ?!lIgaram' and other villages all situated in modern Tek.i tiluk of th. Tinnevelly district, as being situated in Zrya-nIn helps to locate this old territorial division in the modern Tinneveily district. The reference to Ariya-nu in the inscription8 of Rp.pd district indicates that this territorial division covered parts of the modern Pnmed district as well. In this connection it should be noted that the inscription referring to Iriyar (south)

states that they were from Tiru-kuIlam, which was in the

1.

LE.L for 1918, No.. 397, 1+01, 1+03, 40 7, 409,

1+10, 1+12,

1+16, 417,

1+18,

529, 639, 532, 582 and 603 of 1917.

2. There is at least one inscription dated aka 1202 (A.D. 1280) belonging to the reign of a certain Parkrama PIya, who may be identified with Parlkrama Pya 1be1i Vlnldirlya. M. E. B. for 1918, No. 401 of p.

1917 ; K.A.Nilake n ta

Saatri, The PIran Kindorn,

187. 529 and 603 of 1917. k0lIO3 f 407 Q9,.4l0,

1+12 and

3. M.E.R. for 1918, 4. Ibid., No.. 397,

1+16 of 1917.

504
riya-nu division The occurrence of at least three titles, one territorial name and the name of a community or a body of persons with the element r y a in the Tinnevelly-Pamnad region suggests that there must have been some association with this region of some people calle d Iryas. These Iryas may hays been a fmi1y of chieftains from North India, possibly from Gujart, as de Queyroz informs us, who established their authority in that region, or may have been Brhm 4aa , possibly lrya Br h as, whose descendants are to be found still in RIn!&varam,

who wielded political authority there. The Iryacakravartine may have been their leaders or may have sprung from these fmi1ies. In this connection, there is an interesting reference in the Maak4appn-xnmiyam, which may throw some light on our problem. This Battica].oa chronicle refers to the first Iriya, who established his authority in Jaffna as having hailed from riya-nIu If this place is the same as the Zriya-nIu in South India, its name may be connected with the 4riyar who became rulers of Jaffna. On the other hand, the riya-nu of the Battica]oa chronicle may refer to North India. But this is unlikely, for the chronicle states that the first iriya ruler of Jafmna settled in that region

1. See eupra, 2. ., p.

p.1j4..

36.

505
people from ya-nZu These settlers, according to the chronicles of Jaffna, came from th. Tamil country. A second possibility is that Iryacakravartin was just a title conferred on certain chieftains of the Raiiia& region and Jaotes only a fictitious connection with ryaa. In some of the South Indian inscriptions belonging to tb. t thirteenth and fourteenth centuries we hear of such cakravartin titles being borne by officials and chi.ft4a. For instance,

in an

inscription dated

in the

fourth year of a. V!ra PIyadva,

we come across the title ) uva-c ldr1r avartti, which was borne by an official.

in the

PIya count ry In a similar

the

title PIya-cakravarti was borne by some persona in the K,nn4a country who had no authority over the PIya 1ringdoni Another Piya inscription refers to a person with the titl. of G4ita_ci idrsravartti' It is, ther.for., possibi. that Iryacakravartin was also a sirilar title conferred on certain cbieftLi in Rnd by the PIya rulers. But this seems unlikely. The

TamI1 chronicles of Jaffna repeatedly refer to the Zryacakravartina

1. Mm., p. 37. 2. M.E.R. for 19128/29, Ne. 413 of

1928/29.

3. Ibid., Nos. No.

360

li 75

and 488 of

1928/29; LLR. for 1930/31,

of

1930/31.

4. M.E.R. for 193]J32, No.

104

of

193]/32.

506
as Iriyaa and 1 ngs of the riyar (iriiar tarn suggest that their name was more than a title. A third aM strong possibility i. that the 4ryacaIavartins belonged to a BrThmir4a community. We know from th. Ca inscriptions that Brhma4as served in the army as comnenders
As Nilaknta Sastri baa commented., 'it is remarkable that many

which may

of the leaders (senatis) in the army were of Brahmin extraction' It may be that one such eenji distinguished h1ielf in battle, earned the title of lryacakravartin and was granted a chieftaincy

in the

1?p "mad district by one of the PIya kings of the thirteenth

century. It may also be that one of the BrIimea chieftains of the Ramna4 region earned the title in the service of the PIya rulers. We have already seen that at least one 4ryaca,kravartin was .n the service of avarma, Xulatkhara I as a eenpati?

There is a strong possibility that this eenpati was the first Lyacakravartin. l.a discussed earlier, the earliest datable source mentioning an iryacakravartin is an inscription of the fifth year of )avarma Kulakhara I (l268-l3lO) Around 1284,

1. LA.Jile1 aiita Saatri, The Caa, p. k6. 2. Ibid. 3. See supra, p. (1t2 4, See supra, p.411

507
nearly ten years after the date of this inscription, an Iryacakravartin who was a senati in the service of )avarma, Kn1akhara led an invasion of C.ylon ince the date of the first mention of an Aryacaavartin is very close to that of the invasion of Ceylon by an 7ryacakravartim, the identification of the latter person with the former seems possible. That the ryacakravartins who ruled in Jaffna were of Brih m a extraction is stated by

de Queyroz as well as the author of the CekarIca-ckara-nIlai, a work produced in the time of an Iryacakravartin. According to de Qneyroz, the rulers of Jaffna were 'Brsnes, natives of Guzarate, called Arus, who claimed Royal deecent' The Cekaricac!kara-mflai refers to the ryacakravartin in whose time it was composed as a Brhm4a from i, who belonged to the Xyapa

of the Iltylyana stra and who knew the truths of the four Vedae The evidence of this Tamil work cannot be set aside easily.

1. See supra, p. 2. P.de Queyroz, . 3. Ccm, v.20k.

p.48.

Vaiya "ann KItya,a clitrattu

nniya Rc 1pm kVtzat4aruvu k4mai

Ce yy a catur maai v5ymai ci vanta CekarIca-ckaram Tuyya puka2 p'!!cura mz[avar Iti

508
If the Iryacakravartins were Eatriyae, the author of the Tam(1 work could not have aiade a mistake in caU 1hia patron a Brhmia

ruler. If the ryacakravartiu in whose time the Cekarca-ckara.lai was composed were a Brh mia, his ancestors must also have been Brlhma4aa. They were probably Irya Brahma4as and hence the name riyar. We are, therefore, inclined to believe that the Iryacakravartine were Brhmaas, probably L7a Brhm4as, who may have hailed from Gujart, as is cla(m.d by de ueyroz. We have seen that in Gujarit there were at this time a people called The ancestors of the Iryacakravartins were probably from this co mmunity. As 4rya chieftains they may have used the title of iryacakravartin. How and when the Iryacakravartins of South India became masters of the kingdom of northern Ceylon is not clearly known, The Ti mI 1 chronicles state that the first 4ryacakravartin was invited by one of the nobles of Jaffna, called PIi Maavaa, to rule over the northern kingdom, which was without a ruler for some time This is rather doubtful. It is unlikely that the kizmen of the previous rulers would haie kept quiet without

1. See eupra, p.

2. XI! p. 25 ; !!. P . 7.

509
attempting to seize power. Do Qu.yroz has recorded another tradition about the in which the Iryacakravartina became kings of the northern kingdorn According to him, the Arue (ryas) of uacor (PInvaram) began to have trade and friendship with the binge of Jafanapata, and one of them married a daughter of that King; and finilly his descendants became heirs to that kingdom. Of these th. first that tried to free himself from the subjection t the ring of Cota (Xotte, was iriaxaca Varati (ryacakravartin) who being naturally proud and not brooking the haughtiness of the officers of that King, took the life of one who governed there....2 Paranavitana accepts this tradition and concludes that an Zriya from RIiI!varam espoused a princess who was a descendant of the Jvaka prince who-was- ruling the northern kingdom in the thirteenth centur7 In support of this conclusion be cites an inscription front Idavala. In this record, a personage named }rtt1am-perunun-vahanse, with th. title Sav$u-pati, is mentioned. Paranavitana identifies him as )rttIa Iryacaicrai'artin of Jaffna and puts forward 1ge far-fetched hypothesis on the basis of the title av4u-pati, which he interprets to mean lord of the JIT_R According to his argument 1. 1.d. Queyroz, 2. Ibid. 3. S.Paranavitana, 'The 4rya Xingdom in North, Ceylon', p. 197. Lf. Ibid., p. 199. . s.i.. p. L.9,

510
if an Iriya from Rn&varain became master of this kingdom as the result of an matrimonial a1].iance, the J1v,1r gr 1vIi or Sav4u people would have referred to this Ariya and hi. descendants as their lord. The epithet 'Sav4u-pati' applied to an Ariya-caicravarti in the )davala inscription can thus be satisfactorily exptiined on the hypothesis that the ryal femlIy- into which the Iriya married was that of the 1 As we have pointe& out earlier, Paranavitana's interpretation of the titi. of Sav4upati cannot be accepted Consequently there i. no evidence in the }d.avala inscription regarding a matrimonial alliasce between the ryacakravartins and the f41y of their predecessors in Jaffna; and de Queyroz's information stands uncorroborated. Under these circumstances ws can take this matrimonial alliance to be only a possibility. Another possibl manner in which the Iryacakravartins of Ramnad gained comtrol of the northern )rtngdom in metrm is by conquest. The successful invasion of an ryacakravartin around 128k resulted in the eubjuation of the Sinhalese kingdom in southern Ceylon by the PIyas It is not known whether the general ryacakravartin invaded northern Ceylon, too. His troops presumably landed in northern Ceylon. The ruler of northern Ceylon possibly continued to be a f.udatory of the PIyas and, therefore, there was no need. to wage a war against him. If that

1. S.Paranavitana, 'The Irya Kingdom in North Ceylon', p. 199. 2. See sura, 3. See enpra, p.

511
was the case, the invasion of Iryacakravartin may not have resulted in any control of the northern ldiigdom by him. On the other band, if the northern ruler had been recalcitrant, the Piya genera]. may have subjugated him, oo, and begun to extend his authority there. Thee. are matters of speculation. The suggestion that this ryacakravartin. who invaded the Sinhalese kingdom around 128k was already a ruler of Jaffna is not acc.ptable The C!lavasa specifically states that this general was 'sent with an army by the five brothers, the kings who held sway in the

P.4u

realm'

and, having seized the Toots relic as well as other co.tly treasures, he 'returned *th them to the Pa4 Ir4iigdom' The PI].i chronicle clearly mentions that he was a 'Damia generai.' If the L'yacakravartins did not capture power in northern Ceylon around 128k, this event has to be dated to the time of the Muslim invasion of the Pya country, when the forces of }ik KafUr rode triumphkntly down to 1 !varani This was in 1310. As Paranavitana has stggeeted, in the fateful years between the first Muslim invasion of South India and. the establishment of the Nadura Sultanate in 13k'i, the political confusion brought 1. C.Rasanayagam, 2. Cv., 90:k3. 3. Ibid., 90:44, 4. K.A.Nilakta Sastri, The PIyan Kingdom, p. 206 if. . cit., p. 3kk.

512
about by these event4ay have led the ryacakravartine to seek their fortunes in Ceylon. The eall and probably weak kingdom in northern Ceylon would have been a tempting target for their designs. This is, however, another possibility. In the present state of our knowledge we cannot be certain about the "sjer in which the ryacakravartins came to occupy the throne of Jaffna. Prom the account of Ibu Batuta, we find that the 4ryacaavartina were firmly established on the throne of northern Ceylon by l3kk and were in contmnd of the sea around, which was infested with their piratica]. boats Their rule in the island must, therefore, have begun some time before l3kk. In the absence of any evidence regarding the date of the accession of the first Iryacakravartin ruler, we can place this only within rough Itinite. It certainly took place before l3kk, probably in the first quarter of the fourteenth century. possibly in the last quarter of the thirteenth. The independent ki-gdom of northern Ceylon that emerged in the thirteenth century continued to be in existence till 1620, when the laat of the Tamil rulers was beheaded by the Portuguese and the k4ngdom became part of the Portuguese possessions in the island This medieval kingdom has been commonly known

1. The Rehia of Ibn Batuta, . cit., pp. 217-22k. 2. T.de Queyroz, . cit., pp. 628 ft, 691.

513
to historians as the kingdom of Jaffna. Commenting on this, Gnanapragaear has the following to say: It is an anachronism to call the North Ceylon of the Temil period by the name of Jaffna. Nor is it correct to say that any Ruler of the North of the Island was king of Jaffna.- The name Jaffna, now designating the entire peninsula, was first given to the new town Nallur in the 17th cntiry. The kings, whose brief history is to be rec1Ied in the following pages, reigned first at 51i1r1 Nagar, a town situated probably on the sea-shore near Point Pedro,aud then at Nallur, till their downfall. Their k4gdom was known in their own days as that of T1m, a ne given also to the whole s1an& of Ceylon. As this old name is no more in use, and as "Jaffn.a" has come to indicate the northern lringdom, we conform to modern usage in calling our ancient rulers the kings of Jaffna. 1

in

This seems to be a fair criticism, although it is wrong to say that the name Ja.ffna was first used only in the seventeenth century The Tamil form of Jaffna, namely flppi4am, does not seem to have been used for the northern. k{ngdom.in the early period of its existence. Like the Sinhalese rulers

in

the d

south, the rulers of the north considered themselves to be the

kings of

the whole island. The South Indian inscriptions refer

to the northern ruler

as one

of the kings ot am or Tlt4b2i

(Cey1on) The Tam4l works of Jaffna, written..be1or th fall of the kingdom, also refer to him as the king of Ceylon

1. S.Gnanapragasar, 'Sources for the Study of the Kiatory of Jaffna', Tamil Culture, II, Nos. 3&k, Sept. 2. See supra, p. 3. See infra p. .i.
Ii.. 1953,

P.

303.

S.. infra, p. cit

514
The titles of the Iryacakravartina gen.rally refer to their overlordship of the city of C4ki, as for instance, ClMralriyaA (Iriya, of Ch1ri) C1t'ki-takum-Iriy.z-kV (King of the Iryas who resides at Ci] c ai) 2 and Cl r i-em-k5n Sovereign of C(Mr iil) But there are a few other titles occurring in the earliest Tamil works of Jaffna which may give a clu. to the names by which the northern 1 ngdm. w.- npwn. In the medical work entitled CekarIca-ckaram,ritten 1n the time of an ryacakravartin who had the consecration nam. of Cekarica-ckara, (Skt. Jagat-rja-ekhara) and atable to about the fourteenth century, refers to this ruler as 'Ceyam (Our

u Ci ai-nI, Cekarca-cftara,' (th. victorious CekarIca-ckara

of Ci i-u) This is the only occurrence of tha name

Cliflr.i-nIu (Country of C4iri). It is obviously a reference to the kingdom ruled by Cekarca-cUara,, the capital of which was Cikai. Thia practi.. of extending the rame of the capita]. to the whole kiigdom is common in Tamil literatmre and tradition, as it is in several other countries, too. The naeKa4i.niu

1.

., CiappuppIyiram, p. 7.

2.., v. 36. 3. IbId., v. 76.

1 Cekarca-c!karam, quoted in
ws v - t.

p. xiii. Ltir 'v.4.

f-i

c.y.. 'vi -

51
and !I.ppam for the kingdoms of Kandy and Jaffna, for instance, are both derived from the names of their respective capitals. CiMii-ziu was, therefore, one of the names applied to the northern kingdom after the capital c4ty, but it is not known whether it was commonly used. Perhaps it was used only in literary works. Another title of the Aryacakravartina that provides a clue to the name of the northern kingdom is the one based on the name Ma4vai. In the Cekarca-ckara-mlai, the Xxyacalcravartin in whose time the work was written is caUed Maavai lriya Va.rtayag (rya Vardaya of M.4avai) 1a4avaiyar-kVn Cekaricac!kara-ma (King Cekarea-ckara, Monarch of the people of Ma4avai)2 and Ma4avai-tanta-nlI]. (the Lord produced by Naavai) Scholars differ regarding the identification of the place )4a4avai. Some take it to be a place near Ini!varam and. conclude that the Cekarca-ckara who bore the above titles was born in Ma4avai , in South Iudia1 But the title Ma4avaiyar-k, meaning 'Monarch of the people of Maavai' may suggest that Ma4avai was not dust a small viflage but a larger territory over which

1..

v. 158.

2. Ibid,, v. 269. 3. Ibid., v. 173. If. U.C.H.C., I, pt.2, p. 698.

516
CekarIca-ckara wielded authority. In Paranavitana's opinion, Z4a4avai may have been an alternate name of Ci 1 i i If we turn to the Tamil chronicles of Ceylon, we find that the ancient Tm{1 name of the Jaffna peninsula was J(aai, ?aial or N aijar. All three ar. variants of the same name. Paranavitana and some other writers on the history of the Jaffna kingdom have stated that M4alr and NaavUr also occur as variants of these names But no references are given to the sources where these occur. We have not been able to trace these two variants in any of the Tamil chronicles. The pa-vaipava-

mlai gives two of these forms, namely Ka4ar.a1 and I4aaiar The akka4p-iya gives the two forms Naai and Maaiar

In addition, it gives two other names of the peninsula, namely Naipuram and the older name IkatTpam (Pii 1gadTpa) The form N4ari also occurs in a poem called ti-nlai-pfu This form is evidently an abridgement of 1a4aiar or Naarial.

1. S.Paranavitana, 'The rya Xingdom in North Ceylon', p. 202. 2. Thid. ; C.S.Nayaratnam, Tamils and Ceylon, p. 75. 3. !z. p. 9, 2k. k. Mm., pp. iii, 14, 56. 5. Ibid., pp. J7, 51. 6. CItivnlaipiftu, quoted in A.Muttuttampi Pillai, carittiram, Jaffna, 1912, p. 1.

!4a-

51?
Ma4arial is a compound of two words, ma4al and tia1 (sand + mound). Tiar is another form of

tifal. How

tbs name came to be applied

to Jaffna is not known. It is possible that it is a translation of the Sinhalese name VIli-gama, which was applied to a part of Jaffna as early as the twelfth century, when it first occurs in a C5a record and is now applied to the western half og the peninsula. The Tsinhl translation of Vili-gama would be Naal-Ur, which is also considered to be one of the names applied to the peninsula. N aial may be a variant of this Tamil form. Possibly the name Viligama was applied to the major part or the whole of the peninsula at one time, although it is now restricted to only one half of the region, in much the same way as Nigadipa, which was once applied to the whole of-_the moern. JaLfna district, is now restricted to an island off the pe rIn Rula. Whatever th. origin of the name I aial, it appears that Naavai -may have been a variant or an abridged form of this name, just as Taficai is a shortened form of Tafic1v11r Naavai and Maaia]. may have been applied to the entire kingdom in the thirteenth and fourteenth
centl.ries.

If these names represented only a

VIli-gama,

Tami l rendering of the Sinhalese name

their disappearance

1. See snpra, 2. Cf., Intftain> }nt ai, C1t1ren1rjis yamputtr> vai, AurItapuram > &urai,

)C4?r1.

518
is not difficu].t to explain. The Tamil rendering of the Sinhalese name would have been used only in literary works while the old Sinha].ese name would have continued to be used by the orthary people in its Tamilieed form of Vl4i mai, which is still

current. Such a practice is in keeping with Ind{n literary tradition. In the Ceylonese chronicles, we find that very often Sinhalese names are rendered in Phi, as for instance Mahltittha for Etoa. However, neither C4ra1nIu nor N aial and its variants appear to have been commonly used as the name of the northern kingdom. Th*r disappearance

in the

later literary

works supports this conclusion. We are inclined to agr.e with Gnanapragasar that this northern kingdom was lmown as am or Ili Akai, without any special epithets to distinguish it from the southern kingdom. The Taki4a-kailIca-1,uriam refers to a CekarIca-ckara,, one of the rulers of Jaffna, as the king of Ilp Mr2 i The Kai1hyanlai calls the first $rya ruler as Teilei-m2mava (ling of IlMi in the South) The Kuumiymz1ai inscription mentioned earlier refers to the ruler of the northern kingdom as one of the kings of Ceylon We have

1. Tkp., Ciappupphyirain. 2. ., p.

6.

3. See supra,

519
also seen that the aa of the inscriptions of )avarma Sundara PIya I may refer only to the northern {ndoa The inscriptions of the Vijayanagara period, too, refer to the nortber{'dom of Ceylon as am From about the beginn ing

of

tb

lnof

Earihar* II (acc. 1377), JaZfna was subdued by_ the Vlj&yanagara rulers. But the subjugation of Jaffna is regard& in_the earlier inscriptions and literary works, anch as the Nria4T-vilIsam, as the conquest or re-conquest of Cey1on But in the fifteenth century, the name flppIam came to be applied to the northern kingdom and am was reserved for the Linhalese kingdom in the south. Thus, we see for the first time a Vi4ayngara.i4scription from Iakri, dated aka 1357 (A.D. 1k35), referring, to the northern and southern kingdoms as IIppIam and. am.respectively The name TIppe m suBt have gained currency in the liLteenth century. In all the grants of the Setnpatis of Bmnad the northern bingdom is referred to as !1ppIam or flppa-tcam (the land of !ppiam) In the Portuguese works, the ugdoa is

1. See supra, p. 2. LE.R. for 1918, Woe. 128 and 13k; M.E.R. for 1923, No. 92 of 1923.

3 S.Krishnaswamy A.iyangar, Sources of Vifayanagara History, p. 153; U.C.H.C., I, pt. 2, p. 687.

k.

LE.L for 1901, No. 128 of 1901; S.1.1., VII, No. 778.
P"i-iC

5. See enpra,

520
often referred to as Jafanapata The Sinhalese literary works of the fifteenth century refer to tb. 4ryacakravartins as the rulers of flp1pauna (the Sinhalese form of JafZna) flpIpauna certainly designated the capital of the northern kingdom. This is clear from the context in which it occurs in the Sizihalese owrks But whether it was applied to the whole of the nothern kingdom as well is not clear. Probably it did. The evidence of the Sinhalese Nampota suggests that the whole of the Tamil kingdom, including parts of the modern Trincomalee district, was also known to the Sinhalese as Dem4a-paanama. In this work, a number of villages which are now situated in the Jaffna, I4uflaitTvu and Trincomalee districts, namely Ngak'vila (NkarkZvil), Kadurugoa (Kantarfai), Telipola (Tellippali), Malligama (Mal].kam), Ninivagamu (VTm1cnuin), Taini-divsyina (Kayts), Mni-divayina (&alai-tTvu), NIga-divayina (NIkatTvu or NayiI-tvu), Puvagi-divayina (Pwi$u-tTvu), ra-divayina (Krai-tTvu), )4olliylvala (N43iyav4ai), TrikiTm1 aja (Trinoozna].ee), Vilgamvehera (NItar-kvil or Vilgam-vihra), Tiss m ahavebera and Ilandago4a, are mentioned as places in Dem4a-paanama1

1. F.de Queyroz,

. cit., p. k8.

2. See supra, p. o.o 3. See eupra, p.

lj.00

k. Nampota, pp. 5-6.

521
Dem4a-paaji.ma (Tamil Port) was probably first applied to the seat of the Tamil rulers in the northern region and then extended to mean the whole of the Tamil kingdom. There is also another Sinhalese name for the Tamil kingdom in an old K4aimpota (Book of Boundaries). In this work, the whole of the northern kingdom is referred to as J1vaama, as mentioned earlier This is the only work mentioning this Tl eme. Perhaps it was not widely used by the Sinhalese. The capital of the florthern kingdom, which we may now call the Jaffna kingdom for purposes of convenience, is often given in the early Tamil works of Ceylon as

Ciiea1 or

C4kai

nakar. These works, the Cekarica-ckara-mlai, Cekarca-ckaram and the a-kai1ca-nlai, mgkp it clear that Cl

?ks( was

the place where the Xryacakravartins r.eided The only Tamil inscription

in the

island mention4 ng an Xryacakravartin also

refers to C41ki-nakar as his seat Cikii is also mentioned, along with Aurai (Anurldhapnra) in the inscriptions of Arikari Parkrama PI 4ya (].k22-lk6l) 1 It is clear from these references that ClMi or Ciki-n,kar was the capital of the

1. See aupra, p. lF&3 2. Ccm.,

v.36; Cekarlca-c!karam,

quoted in the Cern., p.xiv,b. ;

., p.78, v. 109. 3. See supra, p . 4wk. M.E.R. for 1912, No. 1i of 1912.

522
Jaffna kingdom in the early period of its existence. The later chronicles, however, do not mention Ci-Mai as the capital of the Taniil.kingdom, although these refer to the Iryacakravartins as C lcpi Ariyas The ppa-vaipava-n1ai and the Iailyanlai

give Na111r as the capital of the first iryacakravartins Na1].r is not mentioned in any of the earlier Tamil works or in inscriptions. It has, therefore, been suggested that C1?ik2i-nkr was the first capital and. NallUr the second capital, established in the fifteenth century after the conquest of the 11ngdom by Sapumal EunrayI The Linhalese works of the fifteenth century refer to the seat of the Jaffna rulers as TIppauna The fourteenth-century traveller Ibn Batuta states that he visited an ryacakravartin at Battala, which some have attempted to identify with Putta].am, on the western coast of the island But the topographical details furnished by the Muslim traveller show$ that this town was somewhere to the north of Manr. Probably Ibu Batuta was referring to !IE1pauna. The element una

in this

name has the variants

anama and ffa1ama_in Sinhalese In fact1 de Queyroz gives

1. Tym., p. 27. 2. Ibid., p. 26; !! p. 7. 3. S.Gnanapragaaar, pp. 106-107.

k. See supra, p. 1frD.

5. S.Gnanapragasar, pa-vaipava-vimarcana, p.

88.

6. S.Paranavitana, 'The Irya ingdom in North Ceylon', p. 211, fn.136.

523
Jafana-en-putalain as a variant of Jafaua-pata5 Battala of Ibm Batuta, like the putalam in the name given by de Queyroz, may be related to alama. However, it is not impossible that Ibm Batuta met the ruler of Jaffna at a place which was not the capital. The Portuguese sources inform us that at the time of the arrival of the Portuguese in the island, that i. in the first decade of the sixteenth century, the seat of the rulers of Jaffna was Na11ttr Thus, we have Ci %Jri-n.kir, npip una or !Ilppaa and Na11x' being mentioned in the literary works and inscriptions as the capital of the Jaffna klngdom. Whether theme three names refer to the same city or not is a matter of some controversy. Since Sinhalese literature contemporaneous with the 4ryacaavartins refers to their capital by the name of Tpipauna, Paranavitana is- inclined to presume that 'the name Ci?'ical- or C1#i{-naker was restricted in its use to court circles and literary men, as GagIsiripura for Gaip4a, for instance, and. that 'Yipipauna' was the popular name' Natesan has suggested that it is possible that C1?ir,i-nakar was re-named as NaUtr in later centuri.st Rasansyagam and Gnaiiapragaear hav 1. Y.de Queyroz, 2. Ibid., p. 47. 3. S.Paranavitana, 'Tb. 4rya Kingdom in North Ceylon', p. 201. 4. U.C.E.C., I, pt. 2, p. 695.

' p. 47.

524
maintained that NallUr was founded in the fifteenth century after the invasion of Sapu mal Kumray1 of the KT kingom and that C1ki-nkar was the earlier capital, to be located in the region of Vallipuru, near Point Pedro According to these writers, the ancient potsherda and brickbate near the sea at Vallipuram indicate that this place was the anciant capital of the 1d-ngs of_Jaffna. This is not an acceptable argument. The occurrence of ancient artefacts in a place does not necesmarily prove that that place was the seat of kings. VaUipurwn, where a gold plate of the time of Vaeabha (67-111) was discovered, is the site of an ancient vibra, as evidenced by the go]4-plate inscription But there is no evidence to indicate that it was the capital of the kings of Jaffna. Lf we group the different references chronologically, we find, that the Tamil works datable to the fourteenth century mention C(1Thski-nakar as the capital. The Loagama inscription, datable to the fourteenth century, also refers to the

4ryaa of

Jaffna as the *ryaa of Ci1 1r i-ne l ar. The Sinhaleae works of the fifteenth century give flpIpauna as the capital of the T.i1 rulers. In the sixteenth century, when the Portuguese weiit to

1. C.Rasanayagam, . cit., pp. 117-118; S. Gnanapragaear, a-vaipava-vimarca, p. 67.

2. S.Parsnavitana, 'Vallipuram Gold-plate Inscription of the Reign of Vaaabha', E.Z., IV, p. 237.

525
the island, it was Na11Lr that was the capital of the Jaffna kingdom. Is it possible that the kings of Jaffna changed their capitals Be often ? This is unlikely. Un l11re th. south-western region of the illand, where the Sinhalese kings at this time changed their capitals frequently, a change of capitals in a sm1l flat area like the Jaffna peninsula would not have afforded any real strategic advantages. Further, there is no archaeological evidence in support of such a change of capitals. So far the only place where evidence of secular buildings datable to the tine of the kings of Jaffna is found is Nall'ar. When the Portuguese went there in the sixteenth century Na11r seems to have been the only city worthy of mention in the peninsula. In the words of de ueyroz, 'they never had any other city save Ne].ur (Naillir) which is not half a league distant from the town and praa of the Portuguese' The latter is the port of Jaffna which later grew into the modern Jaffna town. This port is identifiable with the flplp4una (Port of Tipi or good Port) of the Linhalese sources and the fllpp4a-paffa4am of the Setupati grants Na11r, where the royal palace was situated , was within two miles of this town in the sixteenth century (today Nalfltr comes within the limits of the Jaffna Municipality). It is possible that original-ly

1. 7. de Queyroz, 2. See supra, p.

., p. 50.

rn

the royal palace was also in the port of Jaffna and that it was later shifted towards the interior to Nallr. The foudation of NallUr is sometimes attributed to Sapii m. l KurayI who conquered and ruled the Jaffna kl-ngdoa in the middl, of the fifteenth century This view is based on *evidence that associates him with th. building of the kknda templ. at NallUr. A formula called the kaiyam which is recited by the priests in the temple refer to a person named SrT Saghabodhi BhuvanekabThu Ue is identified as Sapw nal Kunray who, after his rule in Jaffna, became the ruler of (VI, with the name of Bhuvanekablhu

lk70-lk78) The

ppa-vaipava-iflhlai also credits a

person called Puvanftavku (Bhuvanekablhu) with the building of the Skanda temple at NallUrti Be is described in this chronicle as a BrThmaa minister of the first Iriya ruler and is also credited with the building of the outer city walls ef Nal1Ur A stray verse, published along with the

Kai].Uym1ai, attributes

the building of the Skanda temple at NallUr and of TIppa-ni1reri (the City of lIppan) to a certain Puvaakavku The verse

1. S. Gnanapragasar,

a-vaipava-vimarcag, pp. 106-107.

2. S.Paranavitan.a, 'The 4rya Kingdom in North Ceyloa', pp.

192-193;

U.C.H.C., I,
3. Ibid. I. lvii., pp. 5. Ibid.
6. ., p.

pt.2, p.

695; C.Raaanayagam,

22 cit.,

p. 332.

31-32.

23; lvii.,

p. 32, fn.

527
does not state whether he was a minister ox' a king. Both this verse and the a-vaipava-nlai dat. these activities of Bhuvanekabhu to Saka 870 (A.D. 9148). The Puvan!kavku of the stray vers, and of the pa-vaipava-mlai is evidently the iyam

same as rT Sa4ighabodhi BhuvanekabThu mentioned in the of the Skanda temple. A recitation preserved in the form of

a formula and recited regularly it the temple for centuries is likely to be more authentic than the late Taniil chronicle and the stray verse. The title rI Saghabodi used in the kaiyam f or Thuvanekabahu indicates that he was a ruler of the Sixihalese kingdom, for this title, as far as we know, was used only by the Sinhalese rulers as their consecration name. The oni; Bhuvanekabhu who bad any association with the Jaffna kingdom was Bhuvanekablhu V or Sapumal Kunfirayl, who ruled there for some time before be ascended the throne of reasonable to identify Bhuvanekabhu of the KunrayI. The It is, therefore,

kaffiyaiu as Sapiinil

TIlppa-vaipava-n1lai has an 4ryacakravartin. is

evidently confused him

with a minister of

The date given for the

building of the temple the

also unreliable. If, as i.e claimed in

a-vaipava-nlai and the stray T2mil verse, BhuvanekabThu

had anything to do with the building of the city of Na1].r or f 1ppia- n i r '.ri, l it

is

possible to conjecture that he may have

been responsible for the shift of the court from Jaffna (town) to NaUUr. It

is

not possible to hold that he shifted the capital

52d
from some other place to Jaffna, as it.intained by Gnanapragaaar,

for the latter place was weU known as the seat of the 4ryacaavartina before the time of Zapnml EunrayI's occupation. The Linhales. eand!a poems composed at the time of this occupation refer to Sapnml KuiflrayI's sack of flplp4una, which is described as th. seat of the Iryacakravartins A Vijayanagara inscription dated Saka 1357 (A.D.1k35), nearly fifteen years before the occupation of Sap nmal KuuIrayI, mentions th. campaign undertaken by I ldr

46a D.4anyaka

to destroy IyI]ppIam (Jaffna) It is,

therefore, clear that in the fifteenth century, certainly in the first halt of the century, TIppauna or IIppIam (Jaffna) was the capital of the northern kingdom. When the Portuguese went to the island, NaUr was the place where the ruler of Jaffna resided. As we have suggested, it was probably in the time of &apllmRl Kumlrayl that the court was shifted to It may not be necessary to treat this as a change of capitals, considering the close prori-4ty of NallI!r to Tppia-paa4am. As indicated above, C(1ri-nakar is mentioned as the capital of the ryacakravartina in the Tamil works of th. fourteenth century. This is confirmed by epigraphic evidence as well.

1. S.. supra, p

t4OO

2. S.I.I., VII, No.

778.

529
But till very recently there was no contemporary evidence regarding the capital of the northern kingdom in the thirteenth century, that is, during the early decades of its existence. But the Sanskrit inscription from Anurdhapura, recently deciphered by Paranavitana as we have already noted, refers to Subha-pkana, identifiable with Jaffna. Subha-pafana is the anskrit rendering of flpIpauna and, if we are to accept the decipherment of Paranavitana, we have to conclude that in the early years of its existence, too, the capital of the rulers of the northern kingdom was flpIpauna or flppam. This would mean. that in the thirteenth and fifteenth centuries flIp4una was the capital. In the fourteenth century it was Cif&kai or Ci-fikLi--nakar. Was C1-k1, therefore, another name for flpIpauna I The answer seems tobbe in the affirmative. Inscriptions of.4.rikari Parkrama PI.ya (lk22-].1+6l), belonging to the period between 11149 and
11+51+,

refer to the victories won at c{Mei and. Aurai

(Anurdhapura) It would appear, therefore, that in the middle of the fiftbenth century the capital of the Jaffna kingdom continued. to be C(MC1. Since the Sinhalese works of the fifteenth century and a Vijayanagara inscription of
11+35

refer to the

capital as flppaina or IyIppIam, it is reasonable to conclude

1. See supra, P.tp) 2. Travancore Archaeological Series, VI, pp. 89-91; l.A., XLIV, p. M.LR. for LE.R. for
1891+, 1922,
251+;

No, 17 of 1891+; M.E.R. for 1907, No. No.

561+

395

of 1906;

of

191.

53t
that C1rai was another name for TIpIpauna. But there is one difficulty. The mention of Aurai, which in these inscriptions, as in the Koaga.ma inscription, evidently refers to the Sinha.lese capital of the time of the epigrapha, shows that C1jai could. very well have been the name of an earlier capita]. of the Jaffna. kingdom applied by South Indians, by mistake, to flplp4una as well. This is possible but unlikely. The evidence of de queyroz appears to go against such a possibility. According to him, when the Portuguese landed at Koumputtuai, about two miles from Jaffna town, in 1590 and marched towards Nallr, they had 1 to get past a stockade by the name of Chunguinaynar. We are inclined. to agree with Gnanapragasar that this name is a corruption of CiMr i nakar Thia may indicate that Ci1i-nakar was in the vicinity of Nallr. In the sixteenth century, the name C(1c.1nakar was probably zised only for a fortified section of the capital. Before the court was shifted to Nallar, the Jaffna rulers may have held. their court in this place. Ci ?k

i -

ii

and IIpIpauna may,

therefore, be considered as one and the same place. The silence of the later chronicles regarding tbeexistence of a capital called C4-'ai-nakar also points in the same direction. It i. not likely

1. P.ae Queyrom,
2. S.Gnanapraga.ear,

., p.k52; pa-vaipava-vimarcan,

p.68.

531
that CiA1t2ink,r and flpipafuna were two different plac.e and that the capital of the Jaffna kidom was shifted. from the latter to the former place in the fourteenth century and back to the latter place in the fifteenth century. It seems more likely that Cik,i-n*kar, like NallUr, was another fortified place in the vicinity of the port of Jaffna. C4ai-nakar, NallUr and flppauna have to be treated as sections of the same city rather than as different places. Probably the court of the Jaffna rulers was held in all three places at different periods. The capital of the northern kingdom has, therefore, to be located in the region of the present-day town of Jafmna. The emblem of the kings of Jaffna, as we know from their coins, was the couchant bull or nandi. That this emblem was used on their flags, too, i.e evident from the references in the Tamil literary works These works are the CekarIca-ckaramlai, Kailic a-Dure4
and

Takia-kailc a-pur 4am, Kaillyainllai

ai-viu-tUtu. But in the account of Ceyavra Cifzkaiin the ga-vaipava-iai, it is said that this

Iryacakravartin brought the whole is1rid under his mitup,a I21i

(Mitu,a lute Zlag) This raises the question whether the kings

1. Ccm., v. 76 ; KailIca-purlam, CiappuppIyiram ;

pyiram ; ., p. 5 ; Ki#ai-vit u- ttn , v. 152.

., Ciappup-

2. Nitua (Gemini) lute is one of several kinds of lutes need in India.

532
of Jaffna used the lute emblem on their flags at any time. Gnaxiapragasar has drawn our attention to the reference in the Kal11I1rpttn-para4i to the v!;ai ( kt. v1 - lute or lyre) flag which was among the many flags that were lowered by the Cae when their tiger flag was raised everywhere Guanapragasar has posed the question whether this would mean that the lute emblem was used oi the flags of the Jaffna rulers in the eleventh century, since we do not know of any other Indian dynasty having used that flag As there is no evidence that an independent kingdom existed in Jaffna

in

the eleventh century, it is not

possible to suggest that the lute flag was that of the Jaffna rulers. But there is iinm1takable epigraphic evidence that points

1. ICalikattu-para4i, v. 18, p.

25 ;

al ffii kalai yIli vai cilai Keai e!Jaiya pal koi a M!ruvil uyartta Cempiyar i pu].i koi ta1a{1dv. The single tiger flag of the Cempiyar (Cas) that is planted. on the )Zru rises high,

as the

boar, plough, deer, lion, lute,

bow and fish flags and others are lowered. (Boar - Cllukyas, plough - fldavaa, deer - Pilas, lion - Zinhalese, bow - Clrae nd fish - PIyas). 2. S.Gnanapragasar, ppa-vaipava-vimarcap, p. 50.

533
to the use of the vT4 flag in Ceylon in the twelfth century. An inscription of the second regnal year of Kul5ttuga C5]a III (1.180) mentions the vTai-koi_Cik4ar (the Sinhales whose flag is the v) There is no reference to the vT flag in any of the Sinhalese sources. Since the i1ant bad been unified by 1180 under Parlkramablhu I, it is not possible to surmise that the reference in the South Indian inscription is to one of the minor rulers of the island. It would, therefore, appear that the Sinhalese ruler. 5the

v41 flag in the

twelfth century, although it was not probably considered to be the main banner. The reference

in the

Kali.kattu-pare4i is

probably to the Sinhalese. Paranavitana is inc1ind to believe that the vt flag of the Sinhalese must have been used by the Kaliga rulers of JaZfna. He argues that the 'Xaliga kings of Polonnarnva claimed in their inscriptions to be the legitimate successors of ParamabThu I' and that they, after founding the northern kingdom, 'must have continued to use the royal insignia of the Polonnaru kings, which included the lyre-flag' This, however, remains only a possibility. Since the evidence of the earlier Tamil works and of the coins of the 4ryacakravartina

1. S.I.I., V, p. 269. 2. A.Paranavitana, 'The rya Kingdom in North Ceylon', p. 222.

534
clearly inform us that the couchant bull (nandi) was the emblem of the Jaffna rulersand since it is reasonable to presume that )gha and his associates, who seem to have founded the kingdom, introduced the emblem of their home-land, we may conclude that the main emblem of the Jaffna. kingdom was the couchant bu1l The lute flag may also have been used as one of the minor banners of the kingdom. The exact li'n4t* of the Jaffna kingdom are somewhat difficult to determine with the evidence at our disposal. It is only in the time of the Portuguese occupation that we get proper information of the boundaries of the various kingdoms. A valuable description of the territories of the Jaffna kingdom is given by dc Queyroz:The modest kingdom is not confined to the little district of Jafanapata, because to it are also added the neighbouring lands, and those of the Vani tVanni] which is said. to be the name of the Lordship which they held before we obtained possession of them, separated from the preceeding (ale) by a salty river, and connected only in the extremiby or isthmus of Pachalapali accilai-p4aJ , within which were the lands of Balig&mo Tai Ucmam1, Temerache e-marcci) , Bedamrache aa-marccj) , and Pachalapalt }accilai-p4a for ing that pe4nsula, and outside it there stretch the lands of the Vani crosswise, from the side of Manar to that of Triquilem.]. rincomaleeJ , being sepaated also from the country of Nantta nt5ffam or Nahtittha] in the jurisdiction of the Captain of Nanr MaIr by the river Paragali;

1. See supra,

535
which (lands) end in the River of the Cross in the midst of the lands of the Vanl and of others which stretch as far as Triqnilemal, which according to the map appears to be a large tract of country. These lands are divided into Patue and the first near the River of the Cross is Tanamavaraddi eM mi'a-viiJ , a very fine country, but almost uninhabited because of war, and because it was thenroute of our arrayale, the husbandmen who escaped from the war bet1c1-ng themselves to the woods, leaving ver few for cultivation. Prom thence to the side of I4anar is the Province of Muliauali ufliyav4ai] , which consists of three p8tus, Varcama, Valadadi and )!lpatu. This Province is th. principal one of all the lands of the Vant, and is fruitful, though badly peopled on account of war and because it is unhealthy. Next comes Carnaptu tjCarunvai-pattij1 and the province called krmaiirJ, the ne of the Vani who resided Panag.mo there. It consists of the P&tu of Urugare and of Valavi which border on the lands of NantSta, and along the coast of the sea or gulf of Ceylon there are the villages of Parangali, Uerauil Punari, and others of lesser importance. 1 This evidence of the Portuguese writer is generally corroborated by the Sinhaleae and Tamil works. Prom the references in the a-vaipava-mlai and the VaiyIpIal it is clear that the kings of Jaffna directly ruled over the peninsula of Jaffna and the adjoining islands. The villages that were assigned, according to the ppj$a_vaipava-ntlai, to the nobles from South India

by EMti Ckkaravartti, the first princely ruler of the kingdom, are all in the Jaffna penlnula and the adjacent islanda But it

is often

entioned that the seven Verni chieftaincies outside

the peninsula were subordinate territories of the Jaffna kingdoin

1. 7.d. Queyroz, 2. See era,

. cit., p. 51.

3. Tm., pp. 38, kO.

53
Whenever these revolted ag{Tist the authority of the Jaffna ruler and attempted to support the Sinhalese rulers, the former took necessary steps to subjugate them The Sinhalese Kaainrpotas (Boundary Books) of about the fifteenth or sixteenth centurl refer to the territorial divisions of the northern kingdom and and mention that stone pillars with Ta il writing on them were set up as boundary stones in those divisions But these works give neither the boundaries of the Jaffna kingdom nor a full list of all the territorial divisions .f therCa%eaterritorial divisions of the Jaffna district are usually given along with the other divisions of the whole of Pihiiraa (former RjaraHha). But one of these Kaainrpotas, the Tn Si1ha Kaaim saha Vitti, has a reference that appears to be useful to our inquiry. This work mentions that in Jvagama there were five ham districts, namely Javariparaa, )rac c i-naa, Balat a4i-.raa, I4udundu-ma

lliyI..

raa and K41kkii-r4a Some manuscripts give the form Cvagacri or Civagatcini for Javaripa This has enabled the identification of Java.riparaa with the Clvakaccri region of the Jaffna peninsula.

1.

pp. 38, 'io.

2. Tn Si$h4! Kaaim saha Vitti, ed. A.J.W.Narambe, (1926), P. 21; Sini-lak Kaaimpota, ed. Sri Charles de Silva, (1961), pp. 22-23 3. Tn Siha! Kaaim saha Vitti, . cit., p. k. S.Paranavitana, 'The Irys. Kingdom in North Ceylon', p. 195.

537
Mracci-r4a has been identified as the area now comprising the two revenue divisions of North and South Marlcci in the eastern half of the Jaffna peninsula Balata4i-raa is to be located in the vicinity of M4iyav4ai, in the Mullaittivu district. This location rests on the reference in de ueyroze Conguista to Valadadi, identifiable with Ba].atai, as one of the three Ptus (pattus - divisions) of the district of }fuliauali O44iyav4ai) Madundu-m.11iy-r4a appears to be the present MuUiyav4ai division. There are certain considerations for this identification. In the first place, Mudundu-mallij-raa occurs with lat4i-raa and Ie4u1dc1i_raa which can be located in the areas adjoining Muiya-v4ai. The element mafliy seems to be a variant of Further, in the VaiyipTal, Muiyav4ai is called )i-nf-nakar as well as Mi_ma anutai_n' r Mudundu may be

a corruption of )-maanuai. In view of these considerations, the Mudundu-me 11 iy-raa of the Kaa1nipota may be identified with the modern Muiyav4ai division, in the Mullaittvu district. Ka u {i-raa is, of course, the same as modern Kaukk!i,

referred to in the VaiyIpIfal ma Kauk inskar1 in the same

1, S.Paranavitana, 'The 4rya Kingdom in North Ceylon', p.195. 2. F.de Qneyroz, . cit., p. 51 ;

P.E.Peirie, Celon: The Portnguee Era, II, p. 152.

3.!'
I. Ibid.

YT. 35, k2.

538
district. Thus, the five main districts of Jvagama, mentioned in. the Kaaimpota, can be located in. the Jaffna and )4ullaitTvu districts of the Northern Province. This raises the question whether Jvagama was another name for the Jaffna 1 ngdom. Paranavitana is inclined to think that Jvagama is a Tamil form of Jvaka, 'no doubt due to the reason that the region was under the rule of Jvaka princes' This is possible, though by no means certain. Whatever the origins of the name, it probably stood for the northern kingdom. In that case, it would appear that the kingdom was mainly confined to the Jaffna peninsula and some parts of the MullaitTvu district that adjoins it. An earlier work, the PL]lvaliyp, refers to the domains of the Tamils as lying beyond Salgal-kaMura It has not been possible to identify this place. It is generally considered to be somewhere north of Polonnaruva. It is, however, difficult to define the exact l4m{t of the Jaffna kingdom and it is unlikely that it ever had any well-defined limits. It is certain that the Jaffna peninsula and the adjoining islands formed the main. section of the 1ngdom. The V yini chieftaincies of the rest of the modern. Northern_Province usually owed allegiance to the kingdom and. were considered to

1. S.Paranavitana, 'The Arya Iingdoa in North Ceylon', pp. 1914-195. 2. Pv., p. 111.

539
be subordinate territories of it. The Mar region, judging from the evidence of Ibn Batuta and of the !ppla-vaipava-fflE.ai, which refer to the control of pearl bnkR by the ryacakravartins, may have also been directly under the rule of the Jaffna kings But the region of the pearl bsnkR appears to have remained a bone of contention between the Sinhalese and the Tamil rulers It is not clear whether the Trinconialea area was also considered to be part of the Jaffna kingdom. In the a-kailIca-pur ,

one of the Iryaca.kravartins is closely associated with the temple of Kvaram at Trincomalee This may mean that the Jaffna rulers had some sort of jurisdiction over the affairs of that region. The Nampota, as we have seen , include Triku$m1aya (Trinconialee) and Vilgam-vehera (NtaIr-k'5vil) in ihe list of places in Dem4a-paanama which were sacre& to the Bud.dhists1 As we have pointed out, Dem4a_pat .numa appear&to have, been another name for the kingdom of Jaffna. Possibly Trincomalee was considered to be part of the northern kingdom. We have seen that in the Tamil works of the fourteenth century and later the kings of Jaffna are given the epithet

1. The Rehla of Ibn Batuta op. cit., p. 217 ff ; Tvm., pp.k2-43. 2. 3. , pp. k2-k3. ., Tiruakara Carn1dcm, v.107 U.

k. Nampota, p. 6.

540
Ctn-Ivalar, meaning 'Protectors of the CTtu' (RImvaram) In th. Ta4ia-.kaiIIca.-pn4am, an 4ryaca1avartin is referred tp as Tvaiyinma ( King of Tvai, i.e. RImvaram) and C!tu u yar karai kva1 vnta ( the Guardian King of the Nigh Coast of C!tu) Tradition has it that the kings of Jaffna used to send flowers and milk for the daily services at the temple of RIn!varam from some of the illands of! the coast of Jaffzia The epithets Ctu-kivala and T!vaiyi-ma, say not imply any control of the Rimvaram area by the rulers of Jaffna. We are inclined to agree with Paranavitana. that these were inherited from the earlier Iryacakravartins ofT!!Iad. Tb. rulers of Jaffna probably continied to be closely associated with the temple at RZntvaraa and considered themselves to be its lawful protectors. We may, therefore, conclude that the Jaffna kingdom was mainly confined to the peninsula of Jaffna and the islands adjoln lr g it. It had some sort of suzerainty over the Vanni chieftaincies of northern Ceylon, probably those that th were in the modern Northern Province. Thus, with the advent of the iryacakravartine, a stable Tamil ringdom came into existence in the northernmost part of Ceylon. As we have mentioned, earlier, the foundation. of this

1. See aupra, 2. ., Tiruna1rara Car nkk s m , iv. 107, 116. 4a-carittiram, (1912), p. 52.

3. A.}uttuttampi Pillai,

541
kingdom marks the cul mination of the forces that began to work in the northern part of the island with the C5]a occupation in the eleventh century. Gradually the Dravidian elements, chiefly the Tamils and the Ker4as, increased their numbers and influence. The invasion of )1gha and the downfall of Polonnaruva soon led to the emergence of a kingdom in the northern part of the island, which, though at first controfled by )gha and then by the vkas, remained their sphere of authority for over three centuries. From there they attempted to control the rest of the island and at times met with partial success. The Sinhales* who remained behind in the northern areas were in all probability assimilated to the Tamfl population and by the time of the Portuguese occupation the present Jaffna district had been transformed into a predo m i nantly Tamil-speH ng area.

542
CORLUS ION

We have seen in the foregoing chapters the stages by which the Dravidian-speakera from the neighbouring subcontinemt settled in Ceylon and the circumstances under which a new kingdom, dominated by theBe South Indians, emerged into existence in northern Ceylon. The Dravidian settlements that began at the turn of the tenth century gradually covered several parts of northern Ceylon and culminated in the rise of the Tamil kingdom. We have, therefore, considered the foundation of thiB kingdom a convenient point at which to stop. As we have seei, two main stages can be distinguished in the course of these settlement8. The first begins about the tenth century and extends till the end of the twelfth century. The process of settlement during this stage may be said to have reached a fairly notable scale in the eleventh century. The Ca conquest of the island was certainly responsible for this. Although there is evidence of several settlements in northern Ceylon, it cannot be said that there was a mass scale migration of peaceful settlers in the wake of the C2a conquest. The mercenary and the mercantile bodies appear to have been the predominant elements among the Dravidians present

543
in the island in this period. The main areas of settlement lay outside the Jaffna district which in later centuries bad the highest concentration of Tsimils. In fact several of the places which yield, evidence of Tamil settlement in this period are no more occupied by Tamils. Four main areas of settlement can be seen in this period. One is in the north-eastern littoral, another in the western region or what is now known as the North-western Province and the other two are in the region of Anurdhapura and Polonnaruva. There is also evidence of Tamil settlement in the main ports, eapecial].y Mahtittha and rbba. The most important feature of these settlements is the presence of a number of trading communities, such as the Aiffffuvar, Ceis, and mercenary bodies like the Aganrpa4is. In the first stage of the Tamil settlements, therefore, the main areas of settlement were still outside the Jaffna and Batticaloa districts. Of the present-day Tamil areas, only the upper half of the Eastern Province and parts of the western coast had Dravidian settlers in the eleventh and twelfth centuries. The main stage in the process of Tamil settlement which eventually led to the transformation of the present Northern Province into a predominantly Thmil-speaking area had not yet Nakarattr and 11drar and the

544
been reached in the twelfth century. This stage was reached with the conquest of ?gha and it is doubtful that the Tamil settlements of the period before the thirteenth century would have resulted in. a permanent division of the country into two linguistic regions. cept during the period of Ca occupation, migration from South India seems to have been slow though steady. The bulk of the settlers in. this first stage appear to have gone to the island within the half-century of Ca rule. The predominant elements among the settlers appear to have been mercenaries and traders. The absence in the Tamil chronicles of traditions relating to the C]a period may also suggest that the settlements

were founded in the period after the twelfth century. The second and most important stage of the Dravidian settlements is covered by almost the whole of the thirteenth century. tn this second stage, two different phases can be distinguished. The first phase covers roughly the first half of the thirteenth century and the second almost the whole of the latter half. As in the first stage, the arrival of fresh mercenary forces and a quick succession of invasions from the mainland led to the establishments of new settlements in the first phase. But the nature of the invasions and of the settlements that followed wamein many was different from the nature of the earlier invasions and settlements. While the earlier invasions could be treated as episodes in. the history of the island,

545
the invasions of )!gba and the Pya rulers in the thirteenth century cannot be dismissed as mere episodes. The settlements of the earlier period did not result in the vidible dislodgement of the Sinhalese population. As far as we can see, those were not the result of the forcible occupation of lands belonging to the Sinhalese. Those early settlers may have become assimilated to the Sinhalese population in due course. But it was the events of the thirteenth century that prevented such an assimilation in the greater part of the northern and eastern districts. The invasion of )Tgha with the help of ier4a and Tamil mercenaries was far more violent than the earlier invasions. Its chief importance lies in the fac that it resulted in the permanent dislodgement of Sinhalese power from northern Ceylon, the confiscation of lands and properties belonging to the Sinhalese and the consequent migration of the official class and several of the common people to the south-western regions. These factors more than any other helped the transformation of northern Ceylon into a Tamil region and directly led to the foundation ofkingdom there, which soon became a kingdom of the Tamlia. In the second phase, with the foundation of an independent kingdom and several chieftaincies, a deliberate policy of settling South Indians in the Jaffna district and the Vanni regions seems to have been followed by the northern rulers

546
and chieftains. This led to a migration of peaceful settlers from South India. It was this peaceful mogration that was largely responsible for the Tamil settlement of the Jafina district, The settlement of Dravidians in the northernmost regions may, therefore, be said to have been radically different in cbaracter from the process of mercenary or military settlement in parts of the present North-central and North-western Provinces.and Vavuniya district. This is demonstrated by the evidence of literary works and place-name materials. Whereas in the Jaffna peninsula we come across a large perceatage of place-names with with Sinhalese elements, the Tamil element is predominant in the local nomenclature of the North-central Province and the Vanni regions. The former indicate a slow and peaceful penetration of Tamils in Jaffna and the latter a violent and sudden occupation of the other areas. The survival of Sinhalese place-names, especially of Sinhalese territorial names, in Jaffna tells strongly against a wholesale extermination or displacement of the Sinhalese living there. At the same time, Tamil names of estates denoting family settlement which are found scattered across the peninsula, confirm the evidence of the Tami2. chronicles regarding the settlement of prominent Xml1ies from South India by the early kings of Jaffna. The settlements of the thirteenth century, therefore, mark the most important stage

in the

course of the early TRWIil

and other Dravidian settlements in Ceylon. The political conditions

541
of the thirteenth century favoured the rise of an independent kingdom in northern Ceylon. There is little doubt that the Dravidian elements were the force behind )4gha and his associates when they founded the new kingdom. With the establishment of the Jaffna kingdom and the Vanni chieftaincies of the ortbern and Eastern Provinces, the first steps towards the division of the island into two linguistic regions were taken. The process of settlement and assimilation did. not end with the fotuidation of the northerh kingdom. Tamil settlers continued to n.grate to Ceylon and presumably most of them went to the Tamil kingdom. But there were also several Tamils, traders and others, who settled in the Sinhalese areas as is indicated by the Tamil inscriptions and Sinhalese literature and traditions. These settlers were evidently assimilated to the Sinhalese population in due course. In the Jaffna kingdom a similar assimilation seems to have gone on in the centuries after the foundation cZ the kingdom. Here many o the earlier Sinhaleae settlers seem to have been assimilated to the Tamil population. The emergence of an independent kingdom and some minor chieftaincies

in the

northern and eastern regions of the

island and the lack of intimate intercourse between the Tamil and Sinhalese kingdoms as a result of the abandonment ef the North-central Province were factors that led to the division of Ceylon into two linguistic regions. The Sinhalese

kings, as

a rule, did not show any interest in subduing the northern

43
kingdom and unifying the whole country. Even when a prince from the Ee kingdom succeeded in conquering the Jaffna kingdom in the middle of the fifteenth century, he ruled it as an independent kingdom. The Ke rulers were aati8fied with receiving nominal all.giance Ifrom the rulers of Jaffna. It was after the arrival of the Europeans in the island that the Sinhalese and Tamils were brought together, though the country was not unified politically and administratively till the nineteenth century.

549
A SELT BIBLIOGRAPUT

I. Primary Sources (Printed) A, Tami]. 1 C d.ka-ilakkiyam (Complete afigam works), ed. S.Vaiyapuri Pillai, )Ldras 19k0. 2, Cekarca-ckara-m1aj, ed. I,C.Irakunata Aiyar, Kokkuvi]. 19k2. 3. Ci1appatikrani, ed. V.P.P.M.Kaci Vicuvanatan, Madras 1950. If. Cint.ma4i-nikau, ed. Cuvaminata Pantitar, Madras 1909. 5. Kailyam].ai, ed. C.V.Jampu].ikani Pillai, Madras 1939. 6 Kalinkattu-paraj, ed. Puliyur Kecikan, Madras 1958. 7. Kamparmya, ed. V.M.Gopa].a Krian.amacaryar, Madras 19k88. Kantapuram, M.T.Panukavi, Madras 1907. car-ka1veu, 9. ed. A.Canmukarattina Liyar, JaZfna 1909 (Appended to the Takiia-kai].ca-purpam ) 10. )ikkavcakar Tirnvcakam, ed. Cuppiramania Pillai, Madras 19k9. 11. Maakk4appn_inThmiam, ed. F.X.C.Nataraca, Colombo 1962. 12. Musvara-nfnmiyam, r! Vavmpik Sata Muantasvmi Tvastam Kiyarccaai Malar, Coloznbo 1961. 13. N4a-vep, , ed. LRkavacarj, Madras 1938. 1k. lT1akci, ed. A.Cakravartti, Madras9l936. 15. Patuppn, ed. U.V.Caminata Aiyar, Madras 1918. 16. Piflkala-n{ktu, - ed. Rippon Press, Madras 1917. 17. Takja-kailc a-puram, ed. P.P.Vaittiyalinka Tecikir, Pt. Pedro 1916, 18. Tami].-nva1ar-caritai, e d T Kanakacunt arampillai, Madras ,3.921. 19. Tiri-kc a1a-puram, ed. &.Canmukarattina Aiyar, Jaffna 1909.

550
20 Tiru-fia-c amp ant ar Tv!ra Tirppat ikafik4, Tami 1Ka 1akazn, Madras 21. To1kppiyam,I, ed. B.Vaiyapuri Pillai, )Iadras 193k. 22, Vaiy!, ed. S.Gnanapragasar, Jaffna 1921. 2 3. Vaiypa1, ed. J.W.Arutpirakacam, Jaffna 1921. pa-vaipava-in1ai, 2k. ed. K.Capanatan, Colombo 1953.

B. Sinhalese and P1i: 1. Cilava, ed. W.Geiger, 2 vols., P.T.S., London 1925,1927 tr. W.Geiger, tr. from the German into English by C.Xabel Rickmers, 2 pte. Colombo 1953. 2 .Dabadei-asna, ed. D.D.Ranaainha, Colombo 1917. 3. Da1ad-sirita, ed. Sorata Nayaka Thera, Colombo 1950.
k. Dthvaisa,

ed. and tr. B.C.Law, Lahore 1925. Eu-attanaga1uva1sa, 5. ed. P.Aryaratna, Colombo 1932. 6. Gir-sandaya, ed. T.Sugathapala, Alutgama 192k. 7, Hat thavanaa11a-vihra-vawa, ed. C.E.Godakumbure, P.T.S., London 1956. 8. Katikvat-sagar, ed. D.B.Jayatilaka E1ci1a-sand&aya, 9. ed. P.S.Perera, Colombo 1906. 10, Nahvaipsa, ed. and tr. W.Geiger, assisted LH.Bode, Colombo 1912, Reprinted 1950. 11. )1ayra-s and aya, ed. W.LLDharmavardhana, Colombo 1960. 12. Nampota, M.D.Gunasena & Co., Colombo 1955. 13. Nikya-sathprahaya, ad. D.P.R.Samaranayaka, Colombo 2.960. tr. C.Llernando, revised ad. D.Y.Gunawardhana, Co].ombo 1908. 1k. Paravi-sandaya, ed. Ziri Sunandasabha Mahathera, Natara 1925.

551
15. PLTva1iya, 16. RLva1iya, 17. 15. 19. 20. 21. 22. ed. B.Gunaseka c i, Colombo 1953 (2nd,ed.) Saddharm1akraya, ed. Bentara Saddha Tissa Thera, Colontbo 1951. Saddbaramaratnkaraya, ed. I.Nanavlmla Thera, Coloinbo 19k8. S1a1ihini-sand!aya, ed. M.C.Fernando, Moratuva 1956, Sammohavinodan!, ed. A.P.Buddhadatta, P.T. S. London 1923. Sih4a-smitya-lipi, ed. D.B.Jayatilaka, 14a.h aragama 1956. Siri-1ak-kaa1m-pota, Sri Charles de Lilva, Colonibo 1961. aha Vitti, _________ ____ ed. A.J.W.Marambr, Randy 1926. Vaisatthapaksinj, ed. G.P.Malalasekera, P.1.5., London 1935,1936. ed. A.V.Suravira, Colombo 1961.

2k.

C. Others: 1. Batuta, Ibn, 2. 3. The Rehla of Ibn Batuta, tr. Nandi Eussain, Baroda 1953.Cosmas Indicopleustee, Christian Topography, ed F.O.Winstedt. 6e ueyroz, Ferna, The Temporal, and. Spiritual Conquest of Ceylon,11 tr. &.G.Perera, Colombo 1930. Krishnaswaniy Aiyangar (compiler), Sources of Vijayanagara Bistory, Madras 1919. Pring].e, A.T. The Diary and Conu1tation Book of the Agent, Governor and Council of Fort St.George, let Series, III, Madras 1b95. Rijckloff van Goens, Memoirs of Rijckloff van Goens 1665, tr. S.Pietere, Coloinbo 1910. Varthema, L. The Travels of Vartheina, tr. T.W.Jones, 1863.

k.
5.

6. 7.

552
II. Manuscriptss (in the National Museum Library, Colonibo) 1. Vanni Rjvaliya 2. Vanni Upata

III. Epigraphic and Archaeological Sources 1. Ancient India, Bulletin of the Archaeological Surrey of India. 2. Archaeological Survey of Ceylon Annual Reports,

1890-196k.

3. Ceylon Journal of Science, Section G, I & II, 1928-1933. 4. Epigraphia Carnatica , I 5. Epigraphia Indio, I 6. Epigraphia Zeylanica, I-V. 7. Inscriptions (Texts) of the Pudukkoai State,

1929.

8. Eharot1iI Inscriptions, (Corpus Inscriptionum Indicarum, II) S.Konow, Calcutta

1929.

9. Madras Epigraphical Reports, (Annual Reports on South Indian Epigraphy),

1887-1945. 1914.

10. Memoirs of the Archaeological Survey of Ceylon, I.yVI. 11. Memoirs of the Colonibo Museum, Seiies A, No.1, Colonibo

12. Nellore Inscriptions , see Butterworth, A. under modern works.

( ' 3. Tamil and Sanskrit Inscriptions, ed. J.Burgesa, tr. 1

I(Archaeological Survey of South India, IV, Madras 14. Travancore Archaeological Series, I - VI. 15. South Inj Thriptions,v.la. I-Till.

1886).

553
IV. Modern Works (Articles on epigraphic and archaelogica]. materials ae included in this section). 1. Appadorai, A Economic Conditions in Southern 'India, 17& II, Madras 1936. 2. Ariyapala, M.B. Society in Medieval Ceylon, Colombo 1956. 3. Balendra, W, 'Trincomalee Bronzes,' T.C., II, No.2, Ap. 1953, PP . 176-198. k. Bashan, A.L. The Wonder that was India, London 195k. 5. Bell, H.C.P. Report on the Kga1la District, Colonibo 1892. 6. Brito, C. Mukkuva Law, Colombo 1872. 7. Burgess, James1 Tamil and Sanskrit Inscriptions, (Archaeological Survey of South India IV tr. S.}1.Nateea Sastri, Madras 1886. 8. Butterworth,A. and Venugopal Chetty, V. A Collection of Inscriptions on Copper Plates and Stones in the Nellore District, 3 pta. Madras

9. Casie Chety, Simon

Ceylon Gazetteer, Colombo 183k. 'On the History of Jaffna, from the Earliest Period to the Dutch Conquest', J.R.A.S. (C.B.),
I, No.3 , 18k7-k8,

1905.

10. Chanmugam, P.1. 'Anthropometry of Sinhalese and Ceylon Tamils', C.J.Sc. (G), IV, pp. 1-18. 11. Civanatan (no initials) , Pt.1, Kuala Lunxpur, 1933. PQa-kuti y 12. Codrington, H.W. 'The Inscription at Fort Frederick, Trincorna1ee,

pp. 69-79.

J.R.A.S. (C.B.), No.80.

'Notes on the Dabadei Dynasty', C.A.L.R., X,

pp. 37-53, 88-89. 'The Gampola Period of Ceylon History', J.R.A.S.(C.B. mTI,No,86, PP . 258-309.

554
13. Collingwood, R.G. and Myree, J.N.L. Roman Britain and the English Settlements, (Oxford History of En1and, I) , Oxford 1936. 1k. Coomaraawamy Anauda K. Bronzes from Ce yl on (Memoirs of the Colombo Museum, Series A, No.1), Colombo 191k. 15. Cox, H.F. A Manual of North Arcot, I, reviAed by H.AStuart, Madras 1895. 16. de Laval, Pyrard, Discours du Voyage des Francai aux Indes Orientales I, 1887. 17. Derret, J.D.M. The Hoysalas, Madras 1957. 18. Ekwail, E. 'The Celt&c Element', Introduction to the Survey of English Place-names, ed. A.Mawer and F.N.Stenton, Cambridge (England) 192k. 19. Fernando, P.E.E. 'Palaeographical Development of the Brhni! Script in Ceylon', U.C.R., YII No. 11, Oct. 19k9. 20. Fleet, J.P. 'Sanskrit and Canarese Inscriptions', l.A., VI, May 1877, pp. 136-1k2. 'Inscription at Nanag.i', E.I., V, 1898-99, pp.9-31 21. Godakumbure, C.E. 'Bronzes from Polonnaruva', J.R.A.S. (C.B.), N.S., VII, pt.2, P p . 239-253. 22. Geiger, Wilhelm 'Die Vannis', Sitzungsberichte der Bayerischwn Akademie der Wissensohaften, II, Heft k, Juni 19k1, 131nohen, pp.3-il. Culture of Ceylon in Medieval Times, Wiesbaden,1960. 23. A Grammar of the Sinhalese Language, Coloabo 1938. 23. Gnanapragasar, S. Ya-vaipava-vimarcan, Aocuvli 1928. 'The Forgotten Coinage of the Kings of Jaffna', C.A.L.R., V, pt.k, pp. 172-179. 'Ceylon Originally a Land of Tamils', T.C., I, No.1. , Feb. 1952. I, No.2, June 'The Tamils Turn Siiiha1Lese',

1952.

'Begilninge of Tamil Thile in Ceylon', T.C., I,No.3, Sept. 1952. 'Sources for the Study of the History of Jaffna', T.C., II, Nos.3&k, Sept. 1953.

555
2k. Indrapa].a, Karthigesu 'The Nainativu Tamil In8cription of Parkramabhu ' U.C.R., XII, No.1, April 1963. 25. John, S. .a-carittiram, Jaffna (American Ceylon Missi.

1a79.
26. Kanakasabhai, V. Tamils Eighteen Eunred Years A go, Madras 1956) first edition 190k. 27. Kanapathi Pillai, K. I1afikai-v1 Tamilar Varalr, Peradeniya 1956. 'A Pillar Inscription fro Moragahawela', U.C.R., XVIII, Jan-Ap. 1960. 'A Tmil Inscription from Pa4uvasnuvara', C.R., XVIII, Noa. 3&1+, July-Oct. 1960. 'Maik;ai Inscription of GajabThu II', U.C.R.., XX, No.1, Apr. 1962. 28. Kumaracuvami, S. 'Vaa mkattujja Cila Iappeyark4i Vara1u', ppa-vaipava-kaumuti by K.VeluppiAlai, Vas y i.A, Jafina 1918. 29. Lewis, J.P. Manual of 'che Vanni Districts, Colombo 1895, ' .ohaeo1ogy of the Vanni', J.R.A.S. (C.B.), No.11-5, 1891. 30. Mahalingam, T.V. South Indian Polity, Madras 1955. 31. ?4ajumdar, R.C. Corporate Life in Aticient India, Calcutta 1922. ed. Classical Age,(History and Culture of the Indian People)11), Bombay 195k. 32. Matiyaparanam, K. Y,pa-prvika-vaipavam, Jaffna 1927. 33. Mawer, A. aad Stenton, g.M. An Introduction to the Survey of English PlaceNames, Pt.1, Cambridge En gland 1929. 3k. Menon, C.A. Cochin state Manual, 1911. 35. }iinakshi, C. Administration and Social Life under the Pallavas, Madras 1938. 36. Nodder, F. A Manual of the Puttalam District, Colombo Gazetteer of the Puttalam District, Colombo 1908. 37. Moore, L. Nalabar Law and Gust oml,1 Madras 1900. 38. Murphey,Rhoads, 'The Ruin of Ancient Ceylon', J.A.S., XVI, pp.181 ff

556
39. Muttuttampi-piilai, A,

pa-carittiram, Jaffna 1912,

140. Nagam Aiya, V.

Travancore State Manual , I, 1906. h i. Narayana Rao, C. 'The Brhn Inscriptions of South India', N.IIA., I, Sept. 1938 142.Natesan, S. 'The Northern Kingdom', U.C.H.C., I, pt.2, pp.691 fZ 143.Navaratnam, C.S. Tamils and Ceylon, Jaffna 1958. Vanni and the Vanniyars, Jaffna 1960. 14k. Nicholas, C.W. 'Historical Topography of Ancient and Medieval Ceylon', J.R.A.S. (C.B.), N.S., Vt, 195g. k5. Nicholas and Paranavitana, S. Concise History of Ceylon, Colombo 1961. 46. Nilakanta Sastri, K.A. The Cas, Madras 1955 (Revised ed.). The Pan Kingdom, London 1929. 'A Tamil Merchant Guild in Sumatra', Tijdschrift Voor Indieche Taal-, Land-. en Volken Kunde, LXXII, 1932, Batavia. 'The Ceylon Expedition of Jara'varman VT.ra Pya', Eighth All Indjan0riental Conference Proceedings and Transactions, 1937, pp. 50b-526. 'Vijayabhu I, the Liberator of Ceylon', J.LA.S. (C.B.), LS., IT, 19. 'Parkramabhu I and. South India', C .H. 3., IV,

pp. 32-51. 'Ceylon and rI Vijaya', JR.A.S. (C.B.), N.S., VIII,

pt. 1, pp. 125-l+0.

A History of South India, 1958. 147. Pantarattar, C.V.Cataciva Piyar VaralHri, Madras 1956 (3rd ed.) 148. Paranavitana, Senarat Art and Architecture of Ce y lon - Polonnaruva Period, Arts Council of Ceylon, o1ombo 195k. Sigiri Graffiti, I, London 1956. The Shrine of Upulvan at Devundara, Memoirs of of the Archaeological Survey of Ceylon, VI,
Colombo 1953.

The Stpa in Ceylon, Memoirs of the Archaeological

Survey of Ceylon, V, Colombo 1914-6.

557
k8. Paranavitana,
Senarat (contd.) 'The Polonnaruva Inscription of Vijayabhu 1', E.I., XVIII, 1926. 'A Tamil Slab Inscription from P4amai', E.Z., IV,pp. 3.91-196. 'Ga1apta Vibra Rock Inscription', E.Z., IV. 'Vallipuram Gold-Plate Inscription of the Reign of Va8abha', E.Z., IV. 'Gaa1dei Rock Inscription', E.Z., IV. 'The Tamil Inscription on the Galle Trilingual Slab', E.Z., III. 'Ktip Slab Inscription', LZ., V, pp. 157-158. 'A Fragmentary Sanskrit Inscription from Trincomalee', E.Z., V. 'Pa4kauva Copper-plate Charter 6f Vijayab.hu I', L.& . V, pt.1. 'Colombo Museum Pillar Inscription of Kassapa IV', III, 'Polonnaruva Council Chamber Inscription of Abhaya Salamevan', E.Z., IV. Giritale Pillar Inscription of Udaya III' III. 'Two Tami]. Inscriptions from Budunnittva', E.Z., III. 'An Inscription from Padaviya', J.E.A.S. (C.B.), N.S., VIII, pt.2. 'TamilUouseholders' Terrace ', A.B.I.A., XIII, .13: Epigraphica]. Summary', C.J.Sc.(GI, II, pp 17-29;

99-128; 175-228.

'Archaeological Summary', p.J.Sc. (G), II, pp. 160 ff. 'A Bas-relief_at Isurumuni', Artibus Asiae, XIX. 'anxples of Andhra art recently found in Ceylon', 'Evidence of Earliest Sinhalese Art', Ceylon Observer 11 .2.1950, p. 6. 'The Irya Kingdom in North Ceylon', J.P.A.S. (C.B.), N.S., VII, t.2, pp. 1711-22k. 'Ceylon and Malaysia in Medieval Times', J.R.A.S. (C.B), N.S., VII, pt. 1, pp. l-k2. 'Ceylon and Malaysia', J.R.A.S.(C.B.), N.S,, VIII, 'pt. 2. 'Newly Discovered Bistorical Documents Relating to Ceylon, India and South-east Asia', Lecture delivered at the University of Ceylon, Peradeniya, on k.11.l96k (unpublished).

A.B.I.A.,XI, pp. 15-18.

558
If

9. Parker, Henry

'Irrigation in the Northern Province', Papers Laid Before the Legislative Council of Ceylon (Sessional Papers), No.XI, 1886, pp. 105-116. 50. Peiris, Paul E. Ceylon: The Portuguese Era, II , Colombo 191k. 'gad!pa and Buddhist Remains in Jaffna',

5].. Ra.ghavan, M.D.

J.R.A.S. (C.B.), XXVI, No.70,

1917, pp. I1O67.

The Karva of Ceylon, Colombo 1961.

52. Rae anayagam, C.

Ancient Jaffna, Madras 1926.
53. Ray, Hem Chandra (ed.)

University of Ceylon History of Ceylon, I, 2 pts. Colombo 1959 and 1960.

k. Rea, A. 55. Rice, L.

Catalaogue of Prehistoric Antiquities from dichchana11Ur and Perumbair, Madras 1915.

'The ChaJ.ukyas and Pallaas', l.A., VIII, Jan. 1879, pp. 23-29. 56. Saletore, B.A. Social and P3itical Life in the Vijayanagara Empire, 2 vole., Madras 193k. 57. Satasiva Pillai, VV. ppa-vaipavam, Madras, 188k. 8. Sircar, D.C. 'Karfas Outside Kara', J.N.Banerjea Volume, Calcutta 1960. 51. Srinivasan, K.R. 'The Megalithic Burials and Urn-fields of South India in the light of Tamil Literature and Tradition', Ancient India, Bulletin of the Archaeological Survey of India, No.2, July 19k6,

pp.9 fZ.

52. Srinivaan, LR. and Banerjee, N.R.

'Survey of South Indian Megaliths', Ancient India,

No.9, 1 953 special number, pp. 113 ff.

53. Sriskantaraca, A. 'TiruImlai Vara1u Tirnick!car Ilaa Kumppika Malar, Trincomalee 1963. 5k. Stanley, E.J. s of Vasco da Gama and His Viceroa1

55. Sturrock, J.

das da India,of Gaspar Correa, 1tS69. Manual of South Canara, 189k.

56. Subramanya

Ayyar, K.V.

'The Tirumuk.ka]. Inscription of Vrarjndra',

XXI,

559
57. Taylor, William,
Catalowue Raisonn of Oriental ernment Library, 4L, iaras .00u. tion and Analysis of the Mackenzl Manuscripts, Nadra8 1b.

58. Tenneht, Emerson Ceylon: An Account of the Island Physical,

Historical and Topographical, London 1860. 59. Thambiah, LW. Laws and Customs of the amils of Jaffna, Colozubo 60. Thiagarajan, N. A Manual of the Pudukottai State, Puthkottai 1921. 61. Vaithianathan, Kanthiah Tiruketheesvaram Papers, Colombo ,1957.
62.Veluppillai, . Ya-vaipava-kaumuti, Vasavilaii Jaffna 1918. 63. Veluppillai, T.K.

Travancore State Manual, II, Travancore 19140. 64. Venakatasubba Ayyar, V. 'TiruvLafig4u Inscription of Rjdhirja II', E.I., XXII, PP . 86-92. 65. Vogel, J.Ph. 'Ngrjuakoa Inscriptions', E.I., XX, pp. 1-37.
66. von FUhrer Haimendorf, Christof

'New Aspects of the Dravidian Problem,' T.C., II, No.2, pp. 131 ff. 67. Wickrsmainghe, Don Nartino de Zilva 'Po1onnar*va Xnau1und.va Slab Inscription', E.Z., II, T baganiuva Inscription of Vijayabhu I', E.Z., II. 'The Slab Inscription of 1rti Ni&afika Mafla at Ruvnv1i Dgba, Anurdhapura', E. Z., II. . ,II, 'Bpii Slab Inscription of Xa1yavtT', pp. 190-192. 'Anurdhapura Slab Inscription of Mahinda IV', E.Z., I, pp. 29-38. D 'The Slab Inscription Marked of Queen LflvatT', E.Z., I, pp. 176-182. 'Polonnaruva Slab Inscriptions of the V.{kkras', E.Z., II, pp. 242-255. 'Kant4ai Ga1-sana Inscription of Eitti Ni?ika Malla', LZ., II, pp. 282-290. 68. Wijesekera, N.D. People of Ceylon, Colombo 1951.
69. Wijetune, W.M.K.

70. Wilson, KJ. - - Mackenzie Collection, I, Calcutta 1828.

U.C.R., XX, No.2, Oct. 1962, pp. 287-291.

'Who was Vaflabha, the invader of North Ceylon

56
V. Unpublished MoflOr'i1s 1. Kanapathi Pi].la.i, K. A Study of the
JJ. L&

Thesis submitted to the University of London for the Ph.D. degree, 1936. 2. Liyanagma,ge, Amaradasa The Decline of Polonnaruva and the Rise of Dabadeiya Thesis submitted to the University of London for the Ph.D. degree, 1963. 3. Marr, J.R. tholoies with svecial refere Thesis submitted to the University of London for the Ph.D. degree, 1958. Wickramasinghe, Sirima The Age of Parkramab.hu I Thesis submitted to the University of London for the Ph.D. degree, 1958. Wijetunge, W.LK. The Rise and Decline of CN.a Power in Ceylon

if.

5.

VI. Reference Works 1. Cintniai-nika4u Cd. Cuvi 1Iilnata Pantitar, Madras 1909. 2. Dictionary of Pli Proper Names, 2 vole. G.P.Malalasekera, London 1937, 1938. Dravidian Etymological Dictionary, 3. Burrow, T. and Emeneau, LB. Oxford 1961 if. Encyclopaedia of Buddhism - Volume of Specimen Articles e -in-chief G P. Nalalasekera, Colombo 5. Gazetteer No. 119; Ceylon Official Standard Names Approved by the U.S. Board on Geographic Rames, Office of Geography, Department of the Interior, Washington D.C. 1960. 6. Hobson-Jobson Yule, K. and Burnell, A.C. London 1903. 7. Imperial Gazetteer of India, Vol. XVIII, 8. Madras Tti"1 Lexicon, S.Vaiyapuri Piflai, Madras 192k9. Sanskrit-English Dictionary 14. l4onier-williama