Early  Journal  Content  on  JSTOR,  Free  to  Anyone  in  the  World  
This  article  is  one  of  nearly  500,000  scholarly  works  digitized  and  made  freely  available  to  everyone  in  
the  world  by  JSTOR.    

Known  as  the  Early  Journal  Content,  this  set  of  works  include  research  articles,  news,  letters,  and  other  
writings  published  in  more  than  200  of  the  oldest  leading  academic  journals.  The  works  date  from  the  
mid-­‐seventeenth  to  the  early  twentieth  centuries.    

 We  encourage  people  to  read  and  share  the  Early  Journal  Content  openly  and  to  tell  others  that  this  
resource  exists.    People  may  post  this  content  online  or  redistribute  in  any  way  for  non-­‐commercial  
purposes.  

Read  more  about  Early  Journal  Content  at  http://about.jstor.org/participate-­‐jstor/individuals/early-­‐

journal-­‐content.    

JSTOR  is  a  digital  library  of  academic  journals,  books,  and  primary  source  objects.  JSTOR  helps  people  
discover,  use,  and  build  upon  a  wide  range  of  content  through  a  powerful  research  and  teaching  
platform,  and  preserves  this  content  for  future  generations.  JSTOR  is  part  of  ITHAKA,  a  not-­‐for-­‐profit  
organization  that  also  includes  Ithaka  S+R  and  Portico.  For  more  information  about  JSTOR,  please  
contact  support@jstor.org.  
 COMPANION CELLS IN BAST OF GNETUM AND
ANGIOSPERMS
W. P. THOMPSON

(WITH SEVEN FIGURES)

The presence of companion cells in the bast of angiosperms

is one of the constant anatomical featureswhich distinguishthat
great group of plants from the gymnosperms. On account of
the technical difficultiesin the study of bast tissue in general,
little emphasis has been attached to this distinction,although
it is really quite as valuable as the familiarone based on the pres-
ence of vessels in angiospermicwood. The elements in question
are designated companion cells, because almost invariably one
of them is associated with each sieve tube. They are small,
vertically elongated, parenchymatous cells which have special
characteristicsthat will be described later, and which have no
counterpartin the bast of gymnosperms.
In a study of the anatomy of the Gnetales (3, 5), it became
clear that the same elements or ones remarkablysimilar are to
be foundin the bast of Gnetum. Anotherclear-cut characteristic
is thereforeadded to the long list of features in which Gnetum
departs fromgymnospermicstructureand resemblesangiosperms.
It has been shown (5), however,that one of the most strikingof
these resemblances,the possession of vessels, is not the result
of a geneticconnectionbetween Gnetumand angiosperms,because
the vessels have been evolved in entirelydifferent ways in the two
groups. It becomes necessary,therefore,to examine and compare
the angiospermic and Gnetalean companion cells both as to
structureand as to development.
Companioncells of angiosperms
Angiospermicbast as foundin Aristolochiamacrophylla(Lam.)
is representedin fig. i. Two chiefkinds of elementsare visible:
the clear, irregularlyshaped sieve tubes, and the richly proto-
plasmic parenchymatouscells (in most sections of bast thereare
451] [Botanical Gazette,vol. 68
452 BOTANICAL GAZETTE [DECEMBER

in addition thick-walledfibers constituting the so-called hard

bast). The parenchymatouselementsare in turn plainly of two
kinds: large cells containingstarch grains, which are represented
in black as if stained with iodine, and much smaller, starchless
elements. The latter are the companion cells, whereas the former
are ordinarystorage parenchyma. A furtherdifferencebetween
the two kinds appears in longitudinal sections; the companion
cells are highly elongated vertically,while the ordinary storage
parenchyma cells are rectangular. It should be noted that a

4.~~~~

re~~~~~~~~~~~~~~~~~~~g
~

~~~~~~A1
~ ~ ~ ~ ~ ~ "

FIG. i.-Mature bast of Aristolochia

macrophylla(Lam.); X iooo

companioncell is associated with each sieve tube and is frequently

fittedsnuglyinto one cornerof the tube.
The bast of Aristolochiawas chosen for illustrationbecause
its elementsexhibit a more orderlyarrangementthan is usual in
angiosperms. In many species the rows of phloem cells formed
at the cambium very quickly become so distortedthat the orderly
sequence is lost, and the cells of the mature bast then appear to
be haphazardly placed. In Aristolochia,however, the rows of
cells are to a certain extentpreservedin the mature bast, as may
be seen in fig. i. When this is the case, there is a significant
I919] THOMPSON-COMPANION CELLS 453

indicationof relationshipbetween sieve tube and companion cell.

The relative positionsof these two elementsis such as to indicate
that they are formedfromtwo successive cells in a row, and that
they are productsof the same mothercambial cell. This relation-
ship was mentioned
longago by DEBARY (I): "Both fromtheir
arrangementin cross-sectionand on tracing them in the longi-
tudinal direction,it oftenhas the appearance as if the cambiform
cells arose with the elementsof the sieve tubes fromone mother

ie.~~~~~~~~.

41 ~ ~ JZ 4&

FIG. 2. Cambium and young bast of Aristolochiamacrophylla(Lam.) showing

each sieve tube and its companioncell formedfromcontiguouscells in same row;
X I000.

cell, the latter dividinglongitudinallyinto a daughter cell which

becomes the sieve tube element, and another which becomes a
cambiformcell withoutfurtherdivisionor is divided by cross walls
into several of them." STRASBURGER (2) later made the definite
statementthat companioncells and sieve tubes are sistercells.
This statement is of course to be confirmedor disproved by
an examinationof the method in which the bast develops in the
cambial region. The process as it occurs in Aristolochiais illus-
trated in fig. 2. As one traces the rows of cambial cells in the
454 BOTANICAL GAZETTE [DECEMBER

lowerpart of the figureinto the bast of the upper part, it becomes

perfectlyclear that each sieve tube and its companion cell are
derived fromcontiguouscells in the same row, and that as the
bast matures the companion cell tends to become shiftedto one
cornerof the sieve tube. I have not been able to show that sieve
tube and companion cell are products of one division. In fact,
judging fromwhat we know of cambial activity, it seems more
probable that they are products of two successive divisions, in

FIG. 3.-Mature bast of Gnetumlatifoliumshowingsieve tubes and companion

cells; X IcOO.

each of which one daughter (the innermost) remains cambial.

In other words, they are probably not sistersbut aunt and niece.
However that may be, for the purposes of the present discussion
the importantconclusion is that they are successive members of
a singlerow of cambial products.
Companioncells of Gnetum
The mature bast of several species of Gnetum(fig. 3) consists
of only two kinds of elements,large clear sieve tubes and small
parenchymatous cells. The latter evidently correspond to the
1919] THOMPSON-COMPANION CELLS 455

companion cells of angiosperms,as is shown by their small size,

their association with sieve tubes, their location in the corners
of the sieve tubes, theirlack of starch,and theirgreat elongation
in the vertical direction. Storage parenchyma of the ordinary
type (not rays) as found in angiospermsis absent. Sieve tubes
and companion cells are arrangedwithgreat regularity,the former
in very uniformrows and the latter in the angles between the
tubes. Usually there is an interruptedrow of companion cells
for every row of sieve tubes, but occasionally the former are
lacking,as may be seen towardthe rightOffig.3.1
The developmentof the bast at the cambial regionis illustrated
in fig. 4. It will be seen that the companion cells are formedin
radial rows which are continuousthroughthe cambium, and that
in these rows sieve tubes are not formed. There is a tangential
alternationof cambial rows which formsieve tubes with cambial
rows which formcompanioncells. Both are never formedin the
same row. As the bast matures and the sieve tubes expand, the
rows of companion cells become interruptedand the individual
cells pushed to the cornersof the sieve tubes.
This process is evidentlyquite different fromthat whichoccurs
in angiosperms,because in the latter sieve tube and companion
cell are invariably successive cells in the same row. There is no
separate cambial mother cell for the companion cells. Thus the
similarmature conditionis broughtabout in quite different ways
in the two groups.
A study of the wood adjacent to the cambium, also illustrated
in fig.4, revealsan unexpectedrelationshipbetweenthe companion
cells and the parenchymaof the wood. The same cambial -rows
which form companion cells outwardly also form rows of wood
parenchyma cells inwardly. The latter with their protoplasm
are easily distinguishedin the figurefromthe empty,thick-walled
wood fiberswhich are formedin rows by the same cambial cells
whichformsieve tubes. The rows of wood parenchymaformedin
this way are readilymistakenforuniseriaterays, but longitudinal
I It may be remarkedthat the bast of Gnetum is extremelyfavorablematerial,
both fororiginalstudy and forclass use. The sieve areas are remarkablyabundant
and of great size, and the individualpores are verylarge and clear. Moreover,the
wholetissueand the connectedcambiumare not easily crushedin sectioning.
456 BOTANICAL GAZETTE [DECEMBER

sections show that they are typical, vertically elongated wood

parenchyma. The rays of Gnetumare multiseriateor broad.
The appearance of this parenchyma in the mature wood is
shown in fig. 5. At the end of the season's growth,where the
fibersare small and vessels lacking, the wood parenchymaforms
continuousrows (lower part of figure),but with the increase in
size of the fibersand the introductionof vessels at the beginning
of the next season's growth, the parenchymatous cells become

FIG. 4 FIG. 5

FIGS. 4-5.-Fig. 4, wood, cambium,and youngbast of Gnetumlatifoliumshow-

ing sieve tubes and companioncells formedfromdifferentrows of cambial cells;
also wood parenchymaformedinwardlyfromthose cambial cells which formcom-
panion cells;. X iooo; fig.5, wood of Gnetumlatifoliumshowingradial rows of wood
parenchyma; X ioop.

separated just as do the companioncells of the bast. Oftenwhere

the vessels are large the.rows of elements of all sorts become so
distortedthat they can no longer be traced. The elements then
appear to be haphazardly arranged.
In certainregionsof some species of Gnetumparticularlyinter-
esting and primitiveconditionsof the companion cells are found.
In the root of G. scandens,for example, the companion cells do
not become separated and relegated to the corners of the sieve
tubes, but formcontinuousrows, even in old bast. This condi-
igig] THOMPSON-COMPANION CELLS 457

tion of affairsis illustratedin fig.6. At the top of the figure,

althoughthebast is old and beginning to collapse,the companion
cells are still in continuousrows. This conditionis frequently
foundin roots,youngstems,and reproductive axes. It evidently
represents a retentionof a primitivecondition.

FIG. 6 FIG. 7

FIGS. 6-7.-Fig. 6, cambium and bast of root of Gnetumscandens showing

uninterrupted rows of companion cells; XIOOO; fig.7, wood, cambium,and bast of
young stem of Gnetummoluccenseshowingabsence of companion cells and wood
parenchyma; X iooo.

Again,in certainregionsof some speciescompanioncells are

entirelyabsent,the bast thenconsistingentirelyof sieve tubes
in fig.7, fromtheyoungstemof G.
and rays. This is illustrated
moluccense. When this is the case, wood parenchymais also lack-
ing,as mightbe expectedfromthe relationship whichhas been
shownto existbetweencompanioncells and wood parenchyma.
The absenceof companioncells has been notedin seedlingsof
axes, as well as in rootsand
severalspeciesand in reproductive
youngstems.
458 BOTANICAL GAZETTE [DECEMBER

Discussion
The presencein the bast of Gnetum of companioncells which
are in most respectsof the angiospermic type is at firstsight
anotherstrikingindicationof real relationship betweenGnetum
and angiosperms.They are to be comparedwith the vessels
of thewood,broadrays,generalhabit,style(4), absenceof arche-
gonia,free-nucleate embryosac, endospermformation(4), and
otherreproductive characters. In regardto all of these points
Gnetum is angiospermic.
The studyofthedevelopment ofthecompanioncells,however,
showsthat the resemblance does not necessarily indicategenetic
relationship.Whereasthe companioncells of angiospermsare
formedfromthesamerowof cambialcellsas are the sievetubes,
and each one is contiguousto a sievetubein sucha row,thoseof
Gnetum are formedin rows quite separatefromthe sieve tubes
and are the productsof different cambial cells. Of course,it is
possiblethatthe companioncellsof Gnetum are reallygenetically
relatedto those of angiosperms, and that, afteroriginating in
Gnetum in the method described, theirformation has later been
takenoverin angiosperms bythesamecambialcellsthatformsieve
tubes. On theotherhand,it seemsmorelogicalto concludethat
we are dealingwitha case of parallelevolution,just as I have
shownto be truewithregardto thevesselsofthewood (5). The
vesselof Gnetumwitha singlelargeperforation in the end wall
is almostidenticalwiththatofmanyangiosperms.Nevertheless,
it has been evolvedin an entirely way. The perforation
different
of the angiospermic porousvesselhas resultedfromthe breaking
down and disappearanceof scalariform bars, whereasthat of
Gnetumhas resultedfromthe enlargement of typical,circular,
haphazardlyarranged,borderedpits accompaniedby the dis-
appearanceof the middlelatnellae,the enlargement proceeding
until the intervening portionsof the vessel wall have vanished.
In thecase of thecompanioncellwe seemto have a closeparallel
to thatofthevessel. Whilethe completedstructure is similarto
thatofangiosperms, thecourseofitsdevelopment is quitedifferent.
If two such strikingpointsof resemblanceas vesselsin the
wood and companioncellsin the bast are reallyresultsof inde-
,9i9] THOMPSON-COMPANION CELLS 459

pendentevolutionin Gnetumand angiosperms,

the inferenceis
naturalthatotherresemblances
maybe in thesamecategory.

Summary
i. Companion cellsresemblingthoseof angiospermsin size,in
theirassociationwithsieve tubes,in theirusual locationin the
angles of the sieve tubes,and in theirverticalelongation,are
presentin thebast ofsomespeciesofGnetum.
2. The development of these companioncells, however,is
quite differentfromthatfoundin angiosperms.Whereas,in the
latter,each sievetubeand its companioncellare derivedfromtwo
successivecells in a singlerow of cambialproducts,in Gnetum
sieve tubesand companioncellsare producedfromdifferent rows
ofcambialcells.
3. Althoughthe completedformsof companioncell in the
two groupsare similar,theyhave probablybeen independently
evolved.
4. Primitiveconditionsin whichcompanioncells are lacking,
or in whichcontinuousrowsof companioncells are present,are
foundin certainregionsofsomespecies.
5. The parenchymaof the wood is formedby thosecambial
cells whichformcompanioncells; the distribution of the wood
parenchymais consequentlyin radial bands, which frequently
becomeinterrupted by theexpansionofvesselsand fibers.
UNIVERSITY OF SASKATCHEWAN
SASKATOON,SASK.

LITERATURE CITED
i. DEBARY, A., Comparativeanatomyof Phanerogamsand Ferns. Oxford
UniversityPress. i884.
2. STRASBURGER,NOLL, SCHENK, and KARSTEN, Textbook of Botany.
Macmillan. I9I2.
3. THOMPSON, W. P., The anatomyand relationshipsof the Gnetales. I.
The genusEphedra. Ann.Botany 27:I077-II02. I912.
4. , The morphologyand affinitiesof Gnetum. Amer. Jour. Bot.
4:I35-i84. i9i6.
5. - . Independentevolutionof vessels in Gnetales and Angiosperms.
BOT. GAZ. 65:83-90. I9I8.