T H E R A M A N S P E C T R A OF SOME I N O R G A N I C C O M P O U N D S .

BY Q~. S. Vt~NKATESWARAN. (From the Department o] Physics, Indian Institute o… s Bangalore.)

Received February 5, 1938. (Communicated by Sir C. V. Raman, Kt., F.R.S., ,,r.S.)

7. Introduction.

THE Raman frequencies of simple molecules have been of great help in the past in verifying the dynamical theory of molecular vibrations and Placzek's polarisability theory of the Raman lines. The complete spectra of these substances often gire valuable clue to the ultimate structure of their mole- cules. 111 the course of the investigations of the author on the Raman spectra in the field of inorganic chemistry, which has appeared in these Pro- ceedings during the past four years, he has obtained some new unpublished results in the spectra of telluric acid, and of the chromate, molybdate, tung- state, periodate, bisulphate, chlorate, bromate and azide ions as both solids and aqueous solutions. The present paper gives a discussion of these results in relation to the structural properties of these compounds.

2. Experimental.

The substances included in the s tudy were Kahlbaum's extra-pure chemicals and were purified and obtained in good crystaUine Ÿ by slow erystallisation from the solution of these salts in distilled water. A Fuess spectrograph with a dispersion of about 20 ~_ a n d a I-Iilger two-prism spectro- graph with a dispersion of 28 ~- in A 4358 region were used in the investiga- tion. The exciting lines ate ;I 4046 and ~ 4358 of the mercury arc except for the chromates for which the green l i n e ~ 5461 was used. For sodium azide and tellurie acid crystals, Ananthakrishnan's (1937) technique of com- plementary filters was employed. The plates were measured by means of a I-Iilger cross-slide microscope in comparison with an iron arc spectrum.

3. Results.

The results of the investigation ate given in Tables I -VI along with the visual estimates of intensities of the lines. The letters in the tables have the following significance: sh = sharp, d = diffuse, b = broad, vb = very broad, st = strong, m = medium intensity, w = weak, P = polar- ised (p < 6/7) and D _--_ depolarised (p = 6/7). The concentration of the aqueous solutions is expressed in gratas of salt per hundred gratas of water.

144

The Raman Speclra of Some Znorganic Compouncts

TABL~ I . Telluric Acid (Raman Frequencies in cm.-~).

145

S t a t e of sub- s t ance

H2TeO 4. 2H~O c r y s t a l s

[-I2TeO ~ 2 6 % uq. so lu-

t ion

670 (10 .~h)

647

V2

657 (o)

624 (lb) D

365 (2)

357 (3b) D

O - - H

3121 (0 b)

W a t e r b a n d

ki X10 "5 dyaes /cm.

3 .97

3 . 5 4

ki -[- kiii X 10-: kii X 10 -5 q dynes/cm, dynes/cm.

0 .054

0 .091

0 .283

0- 251

' E rABI,,, I I . A X 4 Ions (Raman Frequencies in cm.-~). I

No. S u b s t a n c e l v, ,! v., v 3 v,

10

11

12

N % C r O 4 (crys . ) . .

Na,2CrO a 4 6 % so lu t ion

KzCrO 4 (crys.) . .

K~CrO 4 6 0 % so lu t ion . .

(NH4).oCrO4 20 o�91 solu- t i on . .

(NHa),_,MoOa-2H~O (crys.)

(~NI].4) z MoO 4 s a t u r a t e d so lu t ion

Na2 M o O , s ~ t u r a t e d so lu t ion

Na,%VO 4 40 % so lu t i on .

K I O 4 e rys t a l s

K H S O 4 a n h y d r o u s sol id

H~SO4 (50%)

859 (7)

850 (10)

859 (6)

853 (10)

838 (5)

936 (10)

944 (10)

897 (8)

�9 934 (10)

..[ 7o5 (lO)

944 (4) 1026 (6)

. . 082 (2) 1038 (5)

491 (1)

481 (2)

486 (1)

486 (3 d)

483 (0)

218 (2)

218 (3)

241 (0)

325 (4 b) 50 cm. -1 b r o a d

277 (�89 295 (~)

455 (3 b)

426 (3)

863 (1)

874 (0 d

877 (0)

879 (1 b)

915 (1) 893 (3) 874 (0)

896 (4 b)

841 (1 b)

84o (2)

828 (0) 842 (3) 851 (1)

1187 (1) 1240 (~) 1360 (0)

1172 1234 1341

504 (0)

503 (1)

513 (0)

513 (1)

360 (1)

367 (2 b)

317 (4 b)

45 . ~ (o)

337 349 736

737 853

588 897

(2) (1) (od) ?

(0 d) (1)

(4) (3)

146 C. S . V e n k a t e s w a r a n

TABLE I I I . (Force constants of A X 4 iom

fxl0-S f 'X 10-6 PXl0-S N'o. Ion. vi v2 va v4 dynes/cm, dynes/cm, dynes/cm.

C r O . l y

MoO4"

W 0 4 "

IO 4'

S04"

859

936

934

795

835

481

218

325

286

342

877

896

840

842

875

503

360

452

343

415

4 .82

6 .1

4 .5

4 .6

4-22

O. 53

O. 61

O- 92

1.14

O-59

1 .64

0 .09

0-07

0 .74

0 .51

TABLE IV. A X 3 Ions Raman Frequenci› in cm.-').

_No. [ S u b s t a n c c v i li V2 11 V a _L v~t ..l.

1

8

9

10

11

]2aC10 a (crys.) . .

KC10 a (erys.) . .

KC1Oz 7 o/ /O solution

~aC10 3 (crys.)

NaC1Oa 1 0 0 % solut ion

Mg (BrO3)2 crystuls

K B r O a crys ta ls

K B r O a 6 % solut ion

N a B r O a crys ta ls

N a B r O 3 45 % solut ion

LiIO3 crystuls

910 (0) 930 (10)

915 (1) 929 (10)

930 (10)

905 (2) 930 (5)

927 (10)

789 (0) 799 (10)

778 (3) 796 (10)

796 (5)

797 (10)

806 (lO)

765 (10) 781 (3)

4~o2 (o)

478 (3) 493 (0)

47s (4)

495 (1)

473 (5 b)

334 (0) 368 (1 b)

333 (.1) 357 (2)

357 (2) 378 (2)

346 (3)

332 (1) 309 (6)

977 (4)

977 (1)

971 (2 vb)

828 (o)

s13 (1) 830 (2)

s22 (o) 842 (3)

813 (1 799 (6)

612 (0)

615 (1)

615 (0)

611 (3b)

452 (o)

452 (1)

461 (0)

459 (0)

T/~e Raman Speclra o f 5"ome /norganic Compounds

TABLF, V. (Force constants of AXa ions.)

147

No. Iol!

i

I CIO a'

i i

BrO a'

I O a'

Author

Shen ~nd others

Author

Shen and others

Author

Shen and others

Y l

930

930

799

806

744

779

P2 Ya

615 977

610 982

452 842

421 836

I i

423 796 i

390 826

724

478

479

357

356

320

330

f X 10 -6 f ' x 10 .5 dynes/cm, dynes/cm.

7.26

5.55

6.73

5.25

6.15

5.35

3.79

6.42

1.07

3.52

0 . 2 7

3.19

t~

53~ '

51 ~

55o_30 '

53~ '

55 ~

53o_18 '

TABLE VI. Azide Ion (Raman Frequencies in cm.-1).

No. Name of State of sub- v 2 vi v a 2 v 2 ~uthor st~nce

1 Author 636 (0) 1356 (10) 2077 (0) 1267 (2)

2 Author

Langseth & others

Kohlrausch & Engler

~ a N s (solid)

~ u N a satu- ruted solution

NaNa s~tu- ra ted solution

H N 3 liquid

1346 ( 1 0 ) P

1348 (st)

1300 (1)

2066 (�89

2389 (�89

1260 (2)

:1258 (vw)

4. Discussion of Results.

1. Telluric acid. H,,TeO 4 �9 2H~0. - - -Nis i (1929) has reported one

R a m a n line of medium in tens i ty at 8481 for a 15 per cent. aqueous solution of

telluric acid, and has a t t r ibu ted it to the ' b rea th ing oscillation ' of the te t ra-

hedral ion TeO4". In the present inves t igat ion bo th the solid and aqueous

x The numbers in these pages signify wave-numbers in cm. -x

148 C . S . Venka teswaran

solutions of this acid has yielded four lines, ~ (Tab!e I), out of which the broad band at 3121 obviously belongs to the O-H group. The frequency of this linkage is too low to be a t t r ibuted to the free water of crystaUisation a n d a s has been pointed out in ah earlier paper (Venkateswaran, 1938), it is characteristic of the O-H group involved in a hydrogen bond. The ab- sence of the well-known bands of the water of crystallisafion (Anantha- krishnan, 1937), in telluric acid indicates that the two molecules of water in the crystals of H2TeO,.2 I-I20 do not remain as such in the compound, but goes to forro the molecule Te (OH)e. If this molecule has octahedraI symmetry, it would have three normal modes of vibration, whieh are active in the Raman effect. Out of these three, one corresponds to the symmetri- cal oscillations of the molecule and is single and would appear a s a very intense and polarised Raman line. Of the other two, one is doubly degene- rate and the other is triply degenerate and they would-appear as weak, diffuse and depolarised Raman lines. The spectrum obtained for telluric acid, both in the solid and in aqueous solution, clearly indicates that it is octa- hedral in structure. The study of polarisation of these lines has shown that the intense l i n e a t 647 in the solution is more or less completely polarised and therefore, belongs to the symmetrical oscillation v 1. The other two frequencies 624 and 357 are depolarised and have to be assigned to the doubly degenerate frequency, v 2, and to the triply degenerate frequency va, respect- ively. These three lines of the solution are shifted to higher frequencies in the solid state.

Nagendra Nath (1934) has worked out the theory of the oscillations of octahedral molecules and given the following relations for the frequencies Pi, v2 arld i23.

1 4 1 (k + 4 k ~ i Ÿ - 2 k ' + 4 k vi

1 ~ / 1 kii k v k~i v~ --2~c ~ ( k + + 2 -- 2 (II)

1 ~ I ( k'-~#u' ) v3 - - 2~rc ~ 4 13 + 2 k" (III)

where mB is the mass of the peripheraI atoms. Neglecting the constants k V and k "i which are usually smaU, the primary valence constant k, the repulsion constant k ii, and the sum of the directed valence constant and the

2 A reference to the spectrum of telturic acid and its structure was made by the author in Curr. Sci., 1937, 6, 5. Subsequently, Gupta (Nature, 1937, 140, 685) has also rep0rted three lines for the aqueous solution.

T]le Raman Spectra o[ Some 7norr Compounds 149

k i + k iii intravalenee constant, 12 , may be calculated from a knowledge of the

three Raman-active frequencies. These constants for Te (Oi-1)6 are given in Table I and are of the right order of magnitude. Though the frequencies in the solid and the aqueous solutions vary slight!y, these three constants are approxhnately the same in borla cases, thus indieating that the molecule preserves its ident i ty throughout. I t may be mentioned that the spectrmn of tellurie acid is strikingly different from those of its homologues; viz., sul- phurir acid and selenie acid (Venkateswaran, 19�91

AX4 Ions.

Moleeules of the AX 4 type have nine normal modes of vibration and have been treated theoretically by Nagendra Nath (1934) in an exhaustive manner. Ir they possess tetrahedral symmetry, they reduce to four; the oscillation corresponding to the totally symmetric mode v~ is si~gle and polarised, v2 is doubly degenerate and v 3 and vi are triply degenerate. All the four modes of vibrations are active in the Raman and only v3 and v, are active in the infra-red speetra.

1. Chromate ion.--The Raman spectrum of 13 per cent. aqueous solution of sodium chromate has been studied by Nisi (1929) and has yeilded one diffuse line of medium intensity at 855. In the present investigation, the spectra of the sodium, potassium and ammonium salts have been obtained both in the solids and in the aqueous solutions. In general, four frequeneies have been observed, of which the l i n e a t 859 is sharp and very intense. I t is accompanied by a comparatively faint l i n e a t about 877. Nisi has probably observed these two together a s a single diffuse line. Of the other two lines, 486 is more intense than 503. The spectra indicate that the chromate ion is tetrahedral and the probable assignment of the lines are given in Table II. The force constants of the ion ate calculated on the basis of Dennison's (1926) formul~e and entered in Table III . The infra-red frequency, 870, observed by Taylor (1929) in K~CrO 4 is to be a t t r ibuted to va and agrees well with the eorresponding Raman frequency 877.

Molybdates and Tungstates.--The Raman spectra of molybdates and tungstates have reeeived fair attention in the past. The reported frequeneies ate grouped together in Table VII along with those of the author. The results of the author agrees well with those of the others, exeept tha t in the erystals of ammonium molybdate, the triply degenerate frequeney v 3 at 881 is observed as three lines 915, 893 and 874 on the author's plate. The free- queneies reported by Ghosh and Das (1932) for sodium tungstate solution do not appear to be correct.

150 C.S. Venkateswaran

The R a m a n spec t ra of m o l y b d a t e s and t ungs t a t e s h a v e been the sub- ject of a recent no te in Nature b y G u p t a (1937) and bis resul ts ate g iven in Table VI I . Besides the compounds c i ted above he has r epo r t ed more or less s imilar f requencies for K~MoO4 and K2WO , solutions. H e has, however , been able to record only th ree f requencies for the aqueous solut ions of molyb- dates and t u n g s t a t e s and the re f rom he has concluded t h a t these ions which

TABLE VII.

SubstŸ Author r ~ v 2 v a r 4

d

v

o .~

v

P . K .

G . J .

C. S. V.

G. ~nd D.

G . J .

C. S. V.

G . J .

C. S. V.

G. Ÿ D.

G . J .

C. S. V.

932

927

936

943

938

895

944

898

891

897

931

929

934

216 880

223 881

218 915 893 874

�9 �9

256

�9 �9

218

�9 �9

241

371

320

325

890

885

820

896

845

823

841

1017

834

840

:N. 909 332 852 835 795

356

362

360

�9 ~

356

326

367

830

319

317

(583)

452

400

P. K.--Krishnamurti, P., 1930; G. J.--Gupta, J., 1937; C. S. V.--Veakateswaran, C. S., this paper ; N.--Nisi, H., 1929 ; G. and D.--Ghosh and Das, 1932.

T~e Raman Spec/ra of Some ]norgauic Compounds 151

are di-hydrates in solution are hexa-co-ordinated octahedral units, hi support of bis conclusion, he has deduced an expression vi ~ = v �91 + { v �91 for octahedral symmetry and indicates tha t this relation is 'strikingly obeyed, the deviation (2 per cent. for molybdate and 0- 1 per cent. for tungstate) being very much less than what has been observed in some accepted octahedral molecules like the hexafluorides of S, Se and Te (10-17 per cent.). ' From Nagendra Nath's formul~e for octahedral molecules given in the earlier part of this paper, we get, neglecting k" and k v~,

4 ~'2C~tt4/~ v l 2 ~-~ k -1- 4 h ii ( I V ) 6 (k i + k iii)

(v) 4 Ÿ (v2 z + ~ v � 9 1 = k + 4 kii-] - q

k i + k ~ii has a valne which is nearly equal to (in some cases greater t}lan)

k ii and therefore, i t is not iustifiable to conclude that the equations (IV) and (V) are equal. Hence the relation connecting v I, v~ and v3 used by Gupta to examine octahedral symmetry in molybdates and tungstates is not valid and the reported agreement goes rather to disprove the possibility of any such structure for these ions. Secondly, assuming that the water of crystal- lisation is co-ordinated to the metallic ion, this co-ordination will exist in the hydrated crystals as well. In the case of the dihydrate of amnloniu'n molybdate crystals, Gupta has confirmed the results of Krishnamurt i (1930) and reported four lines which indicate a tetrahedral symmetry for the ion. Thirdly, ii1 the present investigation, strong solutions of ammonium and sodium molybdates have yielded four lines as in the crysta! state and shows tha t Gupta's results ate incomplete. The spectrum of sodium tungstate resembies the spectrum of molybdate; but the frequency v4.452 being generaliy very weak, is partly obscured by the continuous spectrum aecom- panying the picture which could not be avoided even after carefully purify- ing the substance by repeated crystallisations and hence could not be measured accurately. In scheelite (Ca\u Nisi (1935) has obtained this frequency and his results ate given in Table VII. These results ate strongly in favour of a tetrahedral structure for these two ions. The force constants calculated on this basis ate given in Table III , and it may be seen that they

are of the proper magnitude. Potassium periodate.--A weak aqueous solution of 1)otassium periodate

has been studied by Nisi (1929) and be has reported one Raman line of medium intensity at 702. In KIO~ erystals Krishnamurti (1930) has observed two lines, 794 (st.) and 8 t l (f). In the present investigation, the crystalline KIO 4 has yielded nine lines and their frequency shffts are given in Table II. On the assumption tha t 10~' ion is very nearly a tetrahedron, the freqnencies

152 C . S . Venkateswaran

ate classified in four gIoups. From the table it is clear tha t the ' breathing frequency ' is given by v 1 = 795 (10), the doubly degenerate frequency yo is split up into two, 277 and 295, one of the triply degenerate frequencies, v3, is split up into three lines, 828, 842 and 857 and the other one v4 has yield- ed three lines, 337,349 and 736. The exaet origin of 736 is, however, doubt- ful. These results indicate that the ion is approximately tetrahedral, but slightly distorted in the crystal lattice, resulting in the removal of the de- generacies. The average of these frequencies for vl, v2, va and v4 is made use of to calculate the force constants for the ion given in Table III . The frequeney reported by Nisi (1929) for the solution does not agree with any of the lines in the crystal.

Potassium bisulphate.--This substance has been studied by Nisi (1929) a s a 28 per eent. aqueou,~ solution and he has obtained the following Raman frequeneies : 427 (f), 593 (f), 978 (m) and 1048 (st). The spectrum of the anhydrous solid has yielded eight lines and differs from that of the solution, but closely resembles that of 50 per cent. snlphurie acid (Woodward and Horner, 1934) as may be seen in Table II. This clearly shows that 50 per cel~.t, sulphuric acid is predominantly HSO4' ions.

AX3 Ions.

The fundamental vibrations of the molecules of the AX~ type have been treated by Dennison (1926) ; for pyramidal molecuŸ there are four normal modes of vibration, two of the normal frequeneies vi and v2, being single and parallel to the symmet ry axis and the other two v 3 and v4 being perpendi- cular to the axis and doubly degenerate. All the four vibrations are active in the Raman effect and in infŸ absorption. In a previous communi- cation (Venkateswaran, 1935), the author has shown that the Raman spectra of alkaline iodates indicate that they ate pyramidal in structure.

Chlorates and Bromates.--The Raman frequencies of these ions have been determined by Krishnamurti (1930) from the spectra of the crystals of sodium and potassium salts. Sodium and cadmium bromates have been examined by Sehaefer, 1Katossi and Aderhold (1930) and aqueous solutions of sodium chlorate by Dickinson and Dillon (1929). Table IV gives the results obtained by the author for these ions both as erystals and solutions and eontains a number of new lines not reported before.

Recently, Shen, Yao and Ta-You Wu (1937) have studied the polarisa- tion of the Raman lines in 10 N .NaCiO3, 3N.NaBrOs and 6.1 N .I-IIO3 solu- tions. Though they were no t ab l e to record the four eharacteristie Raman lines well separated from eaeh other in these eompounds, they have deduced the four frequencies from the variation in the degree of depolarisation of

T/te Raman Spectra o[ Some Inor#anic Compouuds 153

different parts of the iudividual lines. Parodi (1037) has supported the eonclusions of the above authors by his observations in the infra-red absorption of chlorate, bromate and iodate ions. In the present in- vestigation, these four frequencies have been recorded f o r a few chlorates and bromates in the state of solid and of aqueous solutions. In the solufion of the bromate, however, the frequency v3 is not" resolved from the in- tense line due to v~. But in the crystalline state the lines are sharp and hence al1 the frequencies could be identified. The force constants and the angle of the pyramid ate calculated according to Dennison's formulm and given in Table V. The classification of the frequencies by the Chinese workers is also given for comparison. The angle of the pyramid has a fairly constant value of 35 ~ for all the three halogenates. The valence force, f , is the greatest in C1Oa' and the least in 103' as could be expected from the relative stabilities of these ions. In the case of iodic acid, the lat ter authors have a t t r ibuted the line 826 to v3; but as has been shown by the author (Venkateswaran, 1934) this tine is present only in iodic ae~d and not in alkaline iodates and is more appropriately assigned to the polymerised group Io O('. The lat ter conclusion is confirmed by the fact tha t this line is observed to be polarised (Venkateswaran 1936), while va should be completely depolar- ised :

The main parallel vibration, v 1, is split up into four sharp lines in the crystals of Na, K and Li iodates (Venkateswaran, 1934), and has been tenta- t ively explained as due part ly to positional degeneracy of the pyramid and part ly to accidental degeneracy v 1 = v2 + v4. Ir ~s more probable that the first two lines in this group, represent vi and the last two v a. On the basis of this the average frequencies for the IOa pyramid are v 1 = 744, v, =423 , va = 7 9 6 and v4 =320 . From Table IV, it may be seen that v 1, v, and va in all the three halogenates are split up into two lines in the crystals. The splitting of the doubly degenerate frequency, va, may be explained as due to a slight distortion of the ionic group in the lattice. But the splitting of the single and parallel vibrations, v 1 and v2, cannot be due to this cause and is probably due to the positional degeneracy pointed out in connection with the iodates. The splitting of single, symmetrical Raman lines in crystals is significant and deserves further investigation.

The Azide Ion. The structure of the azide ion has been the subject of considerable

discussion in recent years. Penny and Sutherland (1936) have summarised the existing evidence furnished by the infra-red and the Raman spectra and by the X-ray studies of azides. The two most probable structures for this

154 C.S. Venkateswaran

ion about which the decision should lŸ ate (1) the linear and symmetzical

forro, �9 = !~ �9 and (2) the linear ana unsymmetrical form, N - N -- . The X-ray studies of cyamuric triazide (Knaggs, 1935) is in support of the second formula. In both these forros, there ate three normal mo(tes of vibra- t ion; v 1 and v a are valence oscillations and yo is a deformation oscillation. For the linear and symmetricaI model, v 1 is active in the Raman effect, but inactive in in•ra-red, while v= and va are forbidden in Raman and allowed in infra-red speetra. Ir ir were linear and unsymmetrieal, all the three frequencies would be active both in the Raman and in the infra-red spectra. Garner and Gomm (1931) have obtained two ground frequencies, 1352 and 2040, for fl lead azide in the infra-red absorption. The Raman spectrum of sodium azide solution has been studied by Langseth, Nielsen and Sorensen (1934) and they have obtained two lines 1348 and 1258 (1348 being a hundred times stronger than 1258). They have at t r ibnted 1348 to the valence oscillation v I and 1258 to the overtone of v2. From the non-appear- ance of v2 and v a in their spectra, they have concluded in favour of the linear symmeLrical form. This eonclusion, however, cannot explain the appear- ance of v, in fl lead azide in its infra-red absorption. In a recent paper Engler and Kohlrausch (1936) have recorded two lines, 1300 (1) and 2389 (�89 for hydrazoic acid (HNa) and eoncluded that the N3 group is linear, but unsymmetrical. In the present investigation, four lines are obtained for NaNa in the solid state and three lines in the aqueous solution. The strongest liue in the two spectra has almost the same frequency as reported by Langseth and others. The line 636 in the solid and 2066 in the solid an(t solution are extremely weak, but are positively present in the elear and intense plates. In comparison with the infra-red, we have to attr ibnte 63~ to the deformation vibration v2 and 2066 to the valence oscillation va. These results ate in support of the linear and unsymmetrical form (2). On this model the force constants [12, f23, and f13 as calculated come out to be 8.8 • 105 , 15.7 • lO s and 1.3 • 105 dynes/cm, respectively. The magnitude of these constants strongly support the view given above. The v 1 freqtteucy in the solid is distinctly higher than that in the solution and shows the influence of the crystal forces on the vibrations of the ion.

In conclusion the author wishes to express his humble thanks to Prof. Sir C. V. Raman for his kind interest in the work.

Summary. The Raman spectra of tellurie acid, and the chromate, molybdate,

tungstate, periodate, bisulphate, chlorate, bromate, and azide ions ate obtained. From these spectra, Te (OH)e is shown to be octahedral, CrO4", MoO4", WO�91

The Raman Spectra o[ Some lnorganic Compounds 155

ancl IO4' ions as tetrahedral, C1On' and BrO 3' ions as pyramidal and N~ as linear and unsymm~trieal, in structure. The splitting of the frequencies due to the removal of degeneracy by distortion of ions in the ery~tals, are observed in a few eases.

REFERENCES.

Ananthakrishnan, R., Proc. Ind. Ar Sci., (A), 1937, 5, 76. , ibid., 447.

Dennison, Phil. Mag., 1926, 1,195. Dickinson and Dillon, Proc. Nat. Aca.4. Sol., 1929, 15,334. Garner and Gomm, Jour. Chem. Soc., 1931, 2133. Ghosh ancl Das, Jour. Phy~. Chem., 1932, 36, 586. Gupta, J., Nature, 1937, 140, 685. Knaggs, Proc. Roy. Soc., (A), 1935, lS0, 576. Kohlrausch, K. W. F., au:l Eagler, W., Zeit. f. Phys. Chem., 1936, 34, 214 Krishnamurti, P., lnd. Jour. Phys., 1930, 5,633. Langseth, A., Nielsen, J. R., and Sr J. U.,Zeit . f. Phys. Chem., (B)., 1934, 27, 100. Nagendra Nath, N. S., Proc. [nd. Acad. Sci., (A), 1934, I, 250.

, Ind. Jour. Phys., 1934, 8, 581. Nisi, H., Jap. ]our. Phys., 1929, 5, 119.

, Zeeman, Verhe~rtelbzgeJ~, 1935, p. 261. Parodi, M., Maurice, Compt. Rend., 1937, 205, 607. Penny, W. G.,and Sutherland, G. B. B. M., Proc. Roy. Soc., (A), 1936, 156, 654 and 678. Schaefer, Matossi and Aderhold, Zeit. … Phys., 1930, 65,289. Shen, S. T., Yao. Y. T., and Ta-You Wu, Phys. Rey., 1937, 51, 235. Taylor, Trans. F~~rad. Soc., 1929, 25, 856. Venkateswaran, C. S., Proc. [nd. AcaL Sci., (A), 1935, 2, 119.

, ibid., 1936, 3,307. , ibid., 1936, 4, 174. , ibid., 1938, 7, 13.

Woodward and Horner, Proc. Roy. Soc., (A), 1934, 144, 129.

~6