Vol. 47, No.

2, August 1951

Journa l of Research of the Na ti onal Bureau of Sta nda rds

Research Paper 2229

Critical Temperatures, Pressures, and Volumes of

Hydrogen, Deuterium, and Hydrogen Deuteride
Harold J. Hoge 1 and James W. Lassiter 2
The cri t ical temperat ures, press ures, and vo lumes of H 2 , HD, a nd Dz have bee n determi ned by m easurin g shor t sections of P- F isot herm s in t he ne ighborhood of t he cri t ical
poin ts. T he H 2 a nd Dz sa mpl es wer e ca taly zed to or tho-para equilibrium at 20.4 K and
a re des ignated e-H z a nd e-D.. Fo r e- H z t he values found a re T , = 3 2.9 94 o K, P , = 12. 770
atm a nd 1', = 65. 5 cm 3 mole-I. Fo r HD the corresponding valu es a re 35.908 K , 14.6 45 at m,
and' 62.8 cm 3 mole- I; for e-Dz t he values a re 38. 26Zo K , 16. 282 atm, an cl 60. 3 cm3 m ole- I.
A m et hod of correlat in g isotherm data fo r different s ubstances, based on t he use of g rap hs of
log P " ers us log Y, is desc ri bed.

de cimal p oin t. Th e directl y observed (unadjus ted)

values of temperat ure and press ure ar e given in
column s 2 and 3 of each ta ble. T emperature
va ri a tion among the point s of a given isotherm was
small ; non eth eless. correct ions wer e mad e to bring
each group to a comm on LempC'ralure. Thi s was
accom plish ed b.v using app ropriate valu es of
( l /P )(clP l d T ) obtained whenever possible from
tables 5, 6, and 7 of [1], which give log P as a function
of ] IT. Above th e range of th ese tables, es tim ated
valu es of (J IP ) (dPl d T ) wer e used. Th e correction
is exaC't onl y wher e Lwo ph ases are presen t in the
eq u iIi bri um chamber , but fortuna tely this is th e
place wh ere th e co rrec tions are m os t importan t .
Adju sted values of T and P a re given in columns 4
a nd 5 oJ th e tables. The ad justm en ts seldom
changed th e press ure by more th an 1 mm H g .
The molar volum e cOlTesponding to each obser vation i s the e ffective volume of th e equilibri urn chamber d ividecl by the num bel' of m oles of hydrogen presen t in it when th e obsC'l'vation was m ade. This
q uan t ity, V , is given in column 6 of th e tables. The
effective volume of th e equilibrium ch amber and the
volumes of t.he various parts of th e press ure-transmittin g lin e werr determined in a sep arate series of
exp erimen ts. The number of moles of sample
confined in th e equilibrium ch amber and in th e pressure-tr ansmitting line, ineluding t h e U-tube, was
determined separa tely for each run. F or each observation of a given run, th e number of mol es no t
in th e equilibrium ch amber was compu ted by means
of th e gas laws, and sub t racted from th e total number to give t he qu an tity actualJ .ITinside th e ch amber .
All th e necessary knowledge of volumes may b e
ob tained aft e), the apparatus is assembled, provided
th e inside diameter of the U-tube is known.
If n moles of gas are confmed in a volume VI at
temp erature 1'1, t.hey satisfy the equation PVI =
nR TI 7J, wh ere th e compressibility factor 7, accounts for the devi ation of the gas from ideality.
If th e volume containing th e gas is not all a t uniform
temperature it must be divided in to subvolumes.
The number of mol es of gas in the i th subvolume is

1. Introduction
The presen t paper and th e accomp anyin g one on
vapor pressures an d dew-poin t pressllres r]] 3 comprise a full rC'por t of a progr am of de tel'min a tion of
proper ties of th e h ~-drog('lls und C' I' Laken at tlte
N ational Bureau oJ S ta nd a l'd s earl y in 1950. Th e
situ ation wi th regard 10 C'ri tieal da ta at th at time
was very simil ar to the situation with regard to
vapor pressures ftboye 1 a im ; th e field h aving b een
neglected from th e tim e of th e las t L eiden measuremen ts [2] in 19]7, un lil very recen tly , wh en IiYhi tC' ,
Frirdman , ftnd Johnston [3] p ubli sh ed critical data
for normal hydrogen, and a pl'eliminar.I' report [4]
was m ade of th e presen t work .

2. Experimental Methods
Th e measuremen ls wer e m ade with an appara tus
used previo u sl~T in an in vestigation of t li e vapor
pressure of oxygen [5]. Min or modifi ca ti ons are
men tion ed in [1]. Pressure exer trd by the h.ldrogen
in the equilibrium ch amber was transmi t ted ou t of
th e cryostat thro ugh a sm all metal Lub e th at ('onnected to one end of a U-tub e of stainless steel nf-in.
ou tsid e diam eter) . Th e lower-part of th e U ~tub e
contained m er cury, and the arm not fill ed wi th
hy drogen con tained water through whi ch th e pressure was transmi t ted to oil and thence to the piston
gage with which pressures wer e m easured. By
means of a pump, wa ter could be for ced in to or
removed from the line. This caused the m ercury
to rise in one side of the U- tube and to fall in th e
o th er , thu s varying th e space available for hydrogen
in the U-tub e and causing m aterial to flow into or
out of th e equilibrium chamb er. This made it
possible to measure th e sample in the C' quilibrium
ch amber at differ en t molar volumes.
The experimen tal data ar e given in tables 1,
2, and 3. All observation s ar e numbered in chronological order , wi th the r un numb er preceding the
1
2

P resent add ress: Leeds & N ort hr up Compa ny , P hiia., 1:)a.

P rcsc nLadd res" U. S. Coast Guard , Cape )I,[ay, N. J .
F igures in brackets ind icate the lite rat ure references at the end of this pa per .

75

-------------------

TARLE

1.

I sotherms of

TAR L E

e -H 2

Obser-

vation

T
(unadjusted)

P
(unadjusted)

T
(adj u sted )

O b ser-

vaUDn

(adj u sted)

nmnbcr

18.6a
18.6b
18.6c
19. 1g
18. 6d
19. 4h
19. 4i
18. Be
19. 4j
18.61
19.4k
19.41
18. 5a
18. 5b
19. 3h
18.5c
19. 3i
18.5d
18. 5g
18.5e
J9.3j
19.3k
18. 5f
19. 31
19.3m

33. 0455
33.0463
33.0453
33.0455
33.0453
33.0458
33.0452
33.0457
33.0453
33. 0453
33. 0460
33.0456
33. 0456

17. 5e
16. 12a
19.2q
18.41
17. 5f
16. 12b
18.4m
19.2r
16. 12e
19. 28
18. 4n
19.2t
16. 12d
18.40
17.5g
18.4 p
19.2u
17. 5h
19.2,
17.5i
17.5j
17.5k

33.2935
33.2933
33.2928
33.2963
33.2933
33.2928
33.2931
33. 2928
33.2944
33.2925
33 . 2935
33. 2925
33.2939
33.2925
33.2942
33. 2931
33.2925
33.2935
33. 2922
33.2967
33. 2938
33. 2938

a. b

mmHg

96.14.4
9579.8
9566.8
9564. 4
9566.1
9556. 0
956 1. 1
9565.9
9560.8
9566.0
9558.2
9554. 6
9925. 1
9837. 0
9815.2
9799. 7
9784.2
9786.7
9782.5
9778.8
9774. 5
9768.7
9769.3
9756. 3
973 1. 6
10733
10755
1041i
10390
10334
10324
10244
10238
10193
l Oi n
IOJ75
IOJ36
10136
10133
10105
10091
10076
10055
10008
9963
9798
9543

OJ{

1nmHg
9653.7
9579.4
9566.4
9563.6
9565.8
9560.6
9559.7
9565. 5
9559. 7
9565.6
9557.4
9553. 8

32.893
99694)

( 1'I 1' , ~O.

33. 045
(0155 )

33.293
00906)

cm3 mole- l
52. 07
55.24
58.61
60.69
62.50
62.86
66.09
66.95
70. 23
7t. 84
76.81
80.55

T
(uH adj ustcd)

10732
10755
10417
10385
10334
10324
10244
10238
10 191
10 178
JOl74
10 137
10 135
10 134
10 103
10091
10077
1005<1
10009
9958
9797
9542

52.20
55.35
56.58
58.87
61. 95
62.79
65. 12
67.44
67.49
71. 47
72.70
76.09
81.17

9924.4
9835.1
9814.8
979U.0
9783.8
9785.6
9782. 2
9777. 8
9774. 1
9768. 3
9767.9
9755. 4
9730.7

(TI 1' ,~ l.

( 1'I T ,~ 1.

48. 76
048.77
52. 21
52.26
53. 49
53. 50
56.04
56.52
58.99
60.24
60.4 7
64.97
a

P
(u nad
justed)

T
(adj usted)

P
(adj us ted)

65.74
70.68
72. 40
74. 47
77..\6

82.38
86. 28
96. 60
108.67

b
b
b

No measurements of m ass of material were made correspond ing to the two

given b y n ;R= PV d T;'7- t , and summing over all

subv olumes
nR = R~ni = P~ ( Vd Ti Zi).

(1)

With all t.h e volumes Vi at th e same known t emperature (room temp erat ure), and with enough gas in th e
apparatus t o ob t ain convenien t press ures, th e position of th e m ercury m enisc us in th e U-tube was
varied and th e resulting changes in P observ ed .
In this way th e t otal volume occupied by the sample
was found t o b e
~Vi(cm3 )=,=84 . 66 - 0.096 8hl'

J<:

mm H g

3.16b
3.16e
3. 16a
3.16d
3.16e

35.033
:15.033
35.033
35. 034
35.034

9844
984 1
984 1
9843
984 1

5.2h
5.2g
5.21
5.2d
5.2e
5. 2e
5. 2b
5.2a

35. 7499
35. 7494
35.7494
35.7<186
35.7496
35.7487
35.7486
35.7491

10889
10886
10889
10387
10885
10886
10880
10860

5.3a
5.3b
5.3c
5.3d
5.3e
5.3f
5.3g

35.8798
35.8796
:35.8798
35.8796
35. 8795
35.8795
35.8793

!J095
11091
11089
11 091
11078
11056
11021

3.17d
3. 17e
3. 17c
3. 17b
3.17a

36.022
36.02:3
36.023
36.023
36.023

11 3 19
11 308
11298
11257
11132

4. lO b
4.lOc
4. lOa
4. JOd
4. We
4. 101
4. 109
4.1011

36.250
36.25 1
36.25 1
36.250
36.250
36.249
36.249
36.249

11 715
11684
!J 648
11 626
11603
115(15
11508
11427

0 ](

35. 033
97563 )

(TI T ,~ O.

:l5.749
99557)

(1'I T ,~ O.

r:.r'
984 1
984 1
9842
9840

10888
10885
10888
10888
10884
10886
10881
10860

cm:J 7nole- 1
59. 48
63.60
68. 10
73.2:3
8 1. 75
/

57.76
60.91
64. J2
67.68
67.84
72.1 1

57.83
61. 04
64.68
69.36

11092
Uoo'
11089
11092
11079
!J 057
11022

78.75
82.68

11 32 1
1 11:l08
:36.023
11298
00320)
11257
II J32

59.86
63.22
66.89
75.71
86.76

117 15
11682
11 646
]]626
11 603
11567
11 510
11429

57.88
60.84
65. 15
67.87
70.6 1
74.39
79.05
84.35

35.880
( TI T ,~ O. 99922)

( T!7' , ~ I.

3(;.250
00952)

(7'I T ,~ I.

(2)

where hi was th e scale reading in millimeters of the

m erc ury m eniscus in the arm of the U-tube cont aining hy drogen . Previously, when th e apparatus
h ad b een used to m easure th e v apor pressure of
oxygen [5], similar m easurem en ts had given 84.45
for th e constant term . The agr eement is considered
satisfactor y. Having obtained eq 2, th e effective

74. 15

volume of th e cry osta t was determined by no ting th e

pressure, first with all volumes a t room temperat ure,
and again af ter cooling th e cryosta t with liquid air ,
and applying eq 1 to th e r esul ts . The volume was
found to b e 5.88 cm 3
Once th e volum es were known, th e number of
moles of gas confin ed could b e determined from any
set of pressure and temperature read ings, provided
none of th e sample was condensed . Norm ally, su ch
a determina tion was m ade at the completion of each
run , after t h e cry ostat h ad w armed to liquid-air
temp erature and th e pressure h ad risen t o 25 t o 35
a tm . The m olar v olume V corresponding to each
individual observation of th e run could t h en b e
computed by summing eq lover all volumes excep t
that of th e cryostat itself, and subtracting th e result
from th e t otal number of m oles in th e system. The
temp erat ure of t h e U-tube was indicated b y a m ercury-in-glass th ermometer . The table of Z-values
given b y Woolley, Scott, and Brickwedde [6] was
used for all three gases. The table is fo r Hz, bu t was
considered adequate for HD and D2 also, in view
of the fact th a t oth er quant ities involved in th e
calculation wer e not known with ver y high accuracy.
In two instances we fail ed to make an y m easuremen t of the t otal mass of sample in th e system .
The two groups of obser vation s involved are for t h e
high es t isoth erm of e-H 2 and are indicated in table l.
Cal culations of th e molar v olume were made for
1h ese obser v ations in th e usual way, excep t that
the unknown to tal number of moles appeared as an
arbitrary constant. Graphs of press ure vers us

76

II
.,

75.72
82.59

65. 63

grou ps 01 observations marked respectively by a and b. Masses were determined

as descr ibed in tbe text, from t he su pe rposition of gra pbs.

I sotherms of H D

---

-----OJ{
32.8935
32.8933
32.8933
32.8936
32. 8932
32.8897
32.8943
32. 8933
32.8938
32.8933
32.8936
32.8936

2.

Observat io ns of the de pendence of pressure o n vo lum e in the cri tical reg ion

Observations 01 the dependence 01 press ure on vol u me in t he cri tical region

numbp,r

------

I
TA BLl c

3.

/ sothenns oj

C-D 2

ObSNv a tion s of t he depe nd ence of pressure on volume in the critical region

Obse rvaLian
number

( un ad
jlls ted)

(un adjus t ed )

0 I(

6. 21 c
6. 21a.
6.21 b

3i. 1983
37. 1988
3i. I99 1

mm H g
J0708
J0709
10708

7. 4a
7. 4b
8. 6a
8.6b
6.22c
6. 22a
6.22b

38. 2029
38. 2036
38.2038
38. 2031
38. 2017
38. 2022
:3 8.201 8

13969
12292
12276
12272
12274
12272
12163

8. 7a
8.7 b
8.7e
8. i el
8. 7e
8. 7f

38.3978
38.3979
38.3978

126.30
12610
12599
12588
12570
12536

7. 5a

38.5556
:38.5502
38.5.?54
:38.5558
38.5554
38. 5., G:3
:18.5556
38.5554
38.5566
38. 55(in
:l~. 5565
:18. 5563

i.5b
7.5e
7.5d
i .5('
8. 4a
7.5f
7.5g
8. 4b
8.4e
8.4d
8. 4e
7.6a
7.6b
7. 6e
7. 6el
7.6e
8.5a
7.6f
7. Gg
8.5b
8.5e
8.5d
8.5e

38.39ii
38.3977

38. 39ii

3S.8794
3S.8766
38.8795
38.8794

:38. 8i94.

150 10
14n6
1365 1
13295
l:llOG
1:100:]
12991
12952

cm 3 mole- l

9i222)

10i09
10708

63.74
70.60
80. i3

38. 203
99846)

13969
12291
12275
12272
12276
12273
12165

40. 19
52. 48
56. 44
63. 04
64 . 20
72. :35
84. 13

12630
12610
12599
12588

54. 12
57. 15
60. 49
64.25
68.57

( T/ '1', ~ O.

( T/ T ,~ O .

38.398

( T/ T ,~ 1.00155 )

125iO

38.55f>
12'/ T ,~ 1.

00i(8)

1.50i5
14421

1401 i

i.7b
7.70
6. 23c
6. 23a
6. 23b

39.1990
39. 1985
39. J988
:19. 19SJ
39. 1995
39. 1980

169:li
.1 5:878
1411 3
J3839
13627
13211

6. 24a
6.24b

40. J960
40.1959

14876

6.2501
6.25b

41. 2076
41. 2085

7. 7"

mm H g
10709

o J(

37. J99

J.\98 1

3S.879 1
38.8807
3S.8801
38. 880 J
38.8806

3S.8i94

P
(adju s t ed )

1290i
1286G
I282i
12742

l :li7.\
13666
13608
13515
13492
J:3:399
133 12
13 155

38.8SOI

I_-.:~usted
T
)

38.880
(T/T, ~ 1. 0 1 6 J 5)

12536

n.34

150 11
14237
13655
13295
l:lI07
1:\002
12992
1295:3
12906
I28(i5
12826
12742

40.30
42. ~O
'14. :t]
46.58
'18.9 1
.50.74
51. 85
53. 46
.15.57
(i0. 28

15982
15082
14422
140 18
13776
13666
J3609
135<17
1:)49 1
1:\:399
1:1012
13 154

40.37
42. 40

(-)5. 72

7:l. 99

'14.6i
47.05
49.5:3
50.86
52. G6
M.50
5.\.9 1
GO. 87
66.51
75.23

1. 02,149 )

]69:37
15879
14 11 3
J3841
13626
13213

Si. OO

141 5 1

40. 19G
(7'/ T,~ 1. 0.5055)

H Si6
1<IJ 51

74.9S
89.63

HiD66
15020

41. 208
(T/ T ,~ J.

16067
150 19

92.33

39. 199
( T/ 7' ,~

OnOO )

40.4.\
42.58
55. 40
64.74
7:3.5<1

i6.42

nurnbel' of moles of h.Hlrogen in the eq uilibriurn

chamber were made on transparent paper for these
data (with the zero of abscissas undeLermined) ;
and a similar graph was made of the rest of Lhe data
of the isotherm, for which the zero of abscissas was
known. The graphs of the " unknown" data were
s uperposed in turn on the graph of the " knO'w n"
data, and moved along the axis of abscissas until
known and unknown graph s merged into a single
continuous curve. By comparing the two abscissa
scales the total number of moles of sample employed
in each unknown run was deLermined .

modification. It was difficult Lo deLe rmine th e

pOSItIOns of the domes with t he desired accuracy,
however , so we sou gh t some m eLhod of corr elaLing
th e data for the three kind s of hydrogen in order
to obtain more accurate critical constants for all .
If the cri tical tcmperature and press ul'C of each
variety were known, P IP e could be plot ted verSllS
V I V c with all three modifications on the same
gr aph; and , if the law of corresponding states wer e
obeyed all th e isotherms would form a single family.
Since onl y approximate valu es of P c and V c were
available, this method of co rrelation co uld be used
only by a process of sll ccessive approximations.
However, it was l'ecognized that thi s diffi.culLy
could be avoided by ta kin g logarithms. If log
(P IP e) were plotted versus log (V I V e) , the isotherms
of the three modifieations would, of course, form a
single family just as the~r would whcn Lhe variables
were PiPe and V j V e. Bu t a graph of log (P /P e)
v crSLlS log (V ( Vc) diA'crs from a graph of log P vel'SUS
log 11 o nl.\~ bv a Lranslation of - log P e along lhc
ax is of ordinates and of - log V e along Lhe a'\i s 0 f
abscissas. I-Icn ce, if separate graphs of log P
ve rSllS log V for each mod iJica tion a re prepared on
transparcnt s Lo ck, it should be possible Lo s upcrpose
Lhem in s uch a way as to form a sin gle family of
isothcrms, lo draw a single liquid-vapor dome for
t he family, anclLhu s to obLain cri tical consta nLs
for t he three moclifica t ions with improved absolute
accuracy and. much g rcaLer relative accLlracy Lhan
could oLherwise be obtained .
The result of applying Lhis Lype of co rrclaL ion is
shown in fig ure 1. The th rec sets of daLa were
ini tially plo tted on threc separaLe sh ecLs or LrfLcing
cloth and supel'Dosed for viewing b? Lrans mi tted
light. The graphs wcre shifLed relaLi Vl' Lo each
oLhc r along Lhe axes or coo rdinates until lhc? most
ne arhr blended into a single famiJ.') . The relaLion
beLwee n Lh e various coordinate scales was t hen noted
and all the data plotted on a single sh eet. On this
gmph. the liquid-vapor dome ,vas drawn and the
cri tical point selected. The coordinates of this
point wiLh respect to the three ori ginal g raphs give
log P e and log V e for the three modifLcations of
hydrogen . . Th e co rresponding cri tical tcmoe ratLll'es
were found by s ubs titu t ing each val ue of P c into
th e approp riate vapor-pressure table (tablc 5, 6,
or 7) of [1]. The critical constants are given in
table 4. Valu es of t he compressibility factor
Z = PVjnR T at the cri tical point are included.
AfLcr Lhe critical cons tan ts had been determined,
TI Te was computed for each isotherm of figure 1,
and coordinate scales in terms of log (PIP e) and log
(V I V e) were s ubstit uted for t hose originally used, so
that the same scales could be used for all three
modifications.

3. Analysis of the Data

TAHLl!;

4.
T,

The preliminary anal.vsis of the isotherms was

carried out in t he usual way by plotting P versus V
for each modifieation of hydrogen, drawing in the
liquid-vapor domes , and taking the highest point of
each dom e as the critical point of that particular

e-U, ___
UD ___
e-D , ._

77

1{
32. 99~

:35.903
33. 262

Critical constant s oj e-H 2, HD , and e-D2

p,

v,

mm U g

atm

psia.

crn 3mole- t

9, 705
11, 130
12, 3 7~

12.770
14. 6~5

187. 67
215.22
239.27

(i5.5
62.8
60.3

W.282

p , \ ',/ RT,

0.309
.3 12
.3 12

V/V c -

1.00

.75

1.25

1.50

1.20

1.15

1.1 0
<Dc?
999

ri:
~

0 2~--~--------~~~------------~-----------------+-----------------~~

1. 0 5

Ed
_H __ -

HD

.~ -- -- --

0. 99846

. -2--

.--. -0 TITe ' 0.9955 7

- .10

.10

.20

- 0.9 0

log 10 (VIVc ) --

FIGC"R " l.

Isotherm s of e-H2, HD , and e-D, i n the neighborhood oj the critical point .

log P ver sus 0.81 log r, th e resul ting graph could

be superposed satisfactoril y on th e hydrogen graph .
The differ en ce in dome-shape of h ydrogen and CO2
is undoub tedly real. I t would be inter es ting to
compar e the cri tical regions of other triatomic
molecules, such as N 20 and 80 2 wi th [that of CO 2 ,
to see th e exten t to whieh triatomic molecules follow
on e law of corresponding states and diatomic molecules ano t her.
The vapor-pressure tables of [IJ are believed t o b e
accurate near the cri tical point to ::i~ 8 mm H g in
the case of e-H2 and e-D 2 , and to 15 mm H g in
th e case of RD . The un cer tainties of th e criti cal
cons tan ts are grea ter than these values, how ever ,
because they include th e additional uncer tainty in
loca ting the proper spo t on figme 1 to choose as t h e

T o help in properly drawing th e liquid-vapor

dom e, sever al oth er se ts of isoth erm data were
plotted on log P versus log T- graphs and compar ed
wi th our own r esults by superposition . The data
plotted included those of Mi ch els, Bl a isse, and
Mich els for CO 2 [7], and those of Onn es, Crommelin ,
and Ca th for hydrogen [2]. The isotherm data in th e
latter paper arc given graphically ra th er than numerically , so th e coordinates of the plotted poin ts and of
various points on th e dome had to be read from the
pu blish ed graph. The dome drawn in figure 1 is
in agr eem en t with th at of the L eiden observers
wi thin t he accuracy of th e obser va tions. Th e
CO 2 data could no t be satisfactorily superposed on
our hydrogen da ta because th e dome 'was too wide.
It was estimated that if for CO 2 on e should plot

78

criti cal poin t. I t is estimated that the criti cal point

selected is un cer tain by O.OOI in log (P IPe). This
is eq uivalent to 22 mm Hg for e-H 2, 26 mm H g
fo], RD , and 28 mm Hg for e-D 2 To these may be
added the estimated uncertainti es in the vaporpressure tables, giving (after rounding off) ~t 30 mm
R g for e-H z, 40 mm Hg for HD, and 35 mm Hg
for e-D2' Using values of dP jdT from the accepted
vapor-pressure tables of [1], these uncertain ties are
approximately equivalent to 20 mdeg for e-Hz and
e-D 2 , and to 25 mdeg for HD. An ad di tional allowance of 20 mdeg must be made for uneertainty
in the temperature scale.
The position of the eritical point in figur e 1 is
un certain by perhaps 0.005 in log ( rj 1 ~c), whi ch is
equ ival ent to slightly more than 1 percent in F.
Errors in determining the volumes of the equilibrium
chamber and the various parts of the pl'eSSUl'etransmi tting line arc no t included in thi s fi gure.
These would affect t be determinations of th e total
mass of maLerial employed in each run , a nd would
introduce a systemaLic errol' in to log ( rj1 ~c) . An
un certain ty of 2 per ce nt is Lhou gh t to be adequaLe
to cover errors clue to all causes. The relaLive
elTors of the criti cal volume of Lh e three varieties are
mu ch smaller th a n th e Lotal errors. This Lype of
error m ay be estimated from the accurfl CY wi Lh
which the three original log P-Iog V graphs co uld be
correctly superposed . I t is estimated that Lh e ra tio
of any two cri tical volumes is correct Lo 5 pel' mille.
The relative accuracy of any two criLi cfll pressures is
probably 1 per mill e, and that of any two criLi cal
temperatures about 3 parts in 10,000.
The pu blished cri tical data for hydrogen were
summariz ed by Pick ering [8] in 1926 . H e selected
as best values: t,=- 239.9 C, P c= 12.8 fLtm , and
p,= 0.0310 g cm - 3 (equivalent to r c = 65. 0 cm 3
mole-I). These ar c essenLially th e valu es of Onnes,
Crommelin, fLnd Cath [2] for the cri tical temperature and pressure and t hat of Math ias, Crommclin ,
and Onnes [9], for the cri tical density. In [2] the
criti cal temperature is given bo th in ce ntigrade and
in Kelvin uni ts (-23 9.91 C = 33.18 K). Since tIle
L eiden auLh ors were unaware of the n ecessity of
controlling o rtho-para composiLi ons, we may assume
that th ey work ed with normal (nonnal = 0.25 parahydrogen, 0.75 orthohydrogen; sec [1]) hydrogen ,
possibly with an appreciable conversion in the
direction of equilibrium hydrogen. In the recent
determination of critical constants at Ohio State

U niversity [3], normal hydrogen Wfl S also employed,

so there arc no publish ed data exac tly comparable
wi th oms. Th e Ohio State values fLr e T ,= 33.24 4 o K ,
and P c = 12.797 atm.
The above cri tical points are shown in fi gure 3
of [1], which is a graph of dev iaLions of ob erved
vapor pressures from fLn accepted vapor-press ure
table. Note that the n-Hz data in this graph arc no t
to be expected to give zero devi ations because the
table from whi ch deviations arc calculated r epresents not n-Hz but e-H z. To our knowledge th er e
are no previoll s measurements of the cri tical constants of HD and e-D2 with which our resulLs may be
compared.
The valu able ass ista,nce of Dino Zei and Raymond
A. Nelson in making the measurements is acknowledged .

4. References
II ) lI a rold J . H oge an d R obe rt. D . Ar no ld , Vapor press ures
of h ydroge n, deu te rium, and h yd rogen d eut e rid e and
d ew-point p ress ures of t.heir mixtu res, J. R esearch
N BR 47 , 63 (1951 ) RP2228.
[21 II. I(ame rlin gh Onn es. C . A. C romme lin, a nd P . G. Cath,
A p reli min a ry determ inatio n of t he cri tica l point of
hvd rogen, Co mmun. Ph.vs. Lab . U niv. Leide n 151 c
( 19 17).
l3) Dav id . White, Ab "aham Solomo n Friedman , and H e rri ck
L . J ohnsto n, Th e direct de te rmin at ion of th e criti cal
t e mperat ure a nd cri t-ical pressu re of normal 11ydrogen,
J . Am. Chem. Soc. 72, 3565 (H) 5Q).
[11 Hobert D. Arnold and H aro ld J . H oge, A test of id eal
solu t io n laws for 1-I 2, ED , and D 2 Vapor pressu re and
critical co nstan ts of the individual compo ne nts, J.
Che m. Phys. 1R, 1295 (1950).
[5) H a rold J . H oge, Vapor press ure and fixed po int s of ox~' gen
and heat ca paci ty in t he cri tical region , J. R esea rch
NBS 44, 32 L ( 1950) RP2081.
[6) Harold 'N . Woolley, Ru ssell B . ScotL, and F. G. Bri ckwedd e, Co mpila tion of thermal p rope rties of h yd rogen
in its va ri o,u s isotopic a nd or t ho-para mod ifi cat ions, J .
R esea rch 1'\ BS 41, 379 (\ 948) RP] 932.
(7) A. Michels, B. Blaisse, and C. Micll els, Th e isotlwr ms of
CO 2 in t he neighbo rh ood of the cri tical point and roun d
the coexistence line, Proc. Roy . Soc. London A160, 358
(1937).
[8) S. F. Pi cke rin g, A rev ie w of the lite rat ure relating to th e
cri tical co nsta nts of va rious gases, BS Sci. Pap. 21,
597 (l926) S54 J.
[9) E. Mathias , C. A. Cro mmelin , a nd H. Kam e rl ingh Onn es,
The rectilinear diameter of h ydro ge n. Co mmun. Ph ys.
La b. U ni v. Leiden I5 4b (1921).

\iVASI-IING'l'O N, April 27 , 1951.

79