British Chemical Abstracts. A. Pure Chemistry, October

BRITISH CHEMICAL ABSTRACTS

A.—PURE CHEMISTRY

OCTOBER, 1929.

G eneral, P h ysical, and In organ ic C hem istry.

H y d ro g e n s p e c tr a e x c ite d b y e le c tro n ic shock.

S. Ve n c o v (Compt. rend., 1929, 189, 279-^-280;

cf. this vol., 96 8 ).--Continuous and secondary spectra of hydrogen were obtained together in the neighbourhood of the ionisation potential of the hydrogen molecule (16-/5 volts). Variations in gas pressure or in th e accelerating field produced the strong Fulcher bands in the la tte r spectrum , whilst the former appeared as a homogeneous b a n d . F urther variations in conditions of excitation produced independent effects on the intensities of the two spectra. The appearance below 20 volts of the first Balrner lines indicates th a t ionisation of the hydrogen molecule a t 16-5 volts is accom panied by

dissociation. , J . Gr a n t.

S e p a r a tio n of h y d ro g e n lin e s in p a r a lle l a n d c ro s se d e le c tric a l a n d m a g n e tic field s. N. R.

Se n (Z. Physik, 1929, 56, 673—683).—M athematical.

The theory of Dirac and Darwin is used to calculate the separation of the hydrogen lines in parallel and crossed electric and m agnetic fields, small relativity effects being neglected. F or parallel fields the same conclusion is reached as w ith the older quantum mechanics. F or crossed fields the arrangem ent of the separated H a lines and the possible switches are

given. A. J . Me e.

D e n sitie s of h y d ro g e n s p e c tr a l lin e s a s a fu n c tio n of th e e le c tro n ic v e lo c ity of e x c ita tio n . C. J . Braseeield (Physical R ev., 1929, [ii], 34, 431—437).—The densities of the principal lines of the singlet and trip let system s of molecular hydrogen, as well as H a, Hg, and H y, were measured and plotted against electronic velocities for a range of 19—220 equivalent volts, a t constant electron emission and gas pressure. The trip let lines approach a maximum density below 19 v o lts ; the singlet lines show a maximum between 30 and 35 volts. The densities of H„, Hg, and H y decrease rapidly below 30 volts, indicating th a t the num ber of dissociating collisions producing an excited atom m ust be very small a t

19 volts. N. M. Bligh.

T e r m s of th e h y d ro g e n m o le c u le . G. H.

Di e k e (Z. Physik, 1929, 55, 447—450).—Suggestions towards a complete theory of the origin of the various hydrogen term s are made. R . W . L u n t.

S tr u c tu r e of th e b a n d s p e c tr a of h y d ro g e n a n d h e liu m m o le c u le s. G. H. Di e k e (Nature, 1929, 123, 979).—The au th o r’s interpretation (cf. preceding abstract) of regularities . in the spectrum of the hydrogen molecule was based on an incomplete

4d 1115

analogy w ith the helium band spectrum . The missing helium bands have now been observed1, a n d ; the a u th o r’s views are stren gth ened. The new bands consist of three groups, all belonging to the trip let system, one in the red, one near 535 mu., and one ric&r'495 .m p. A . A . E l d r t o g e . ’

E le c tr o n t e r m s in th e s in g le t s y s te m s of th e fin e lin e s p e c tr u m of h y d ro g e n . W . We iz e i. (Z. Physik, 1929, 55,483—501).—A complete analysis of th e electron term s of the hydrogen molecule is described in which these are a ttrib u te d to a pho to ­ electron, w hilst a second electron always remains in th e lowest s t a t e ; th e ultra-violet bands, for example, are th en symbolised th u s : A <— B, l.,a21S—

l Ja2pa1S ; A -<■- ■;■; C,\ l.,a21S— l,o2i,w1IT. In this way i t has been possible to evaluate all th e elec­

tro n term s. The ionisation potential is found to be 15-75 in agreem ent with experim ent, 15-9 volts.

The calculated value for th e h eat of dissociation H a+— ^ H + H 4, for the first vibrational quantum , and for the nuclear spacing of H 2+ are, respectively' 46-5 kg.-cal./mol., ab ou t 2100 cm .-1, and about

1-OOxlO-8 cm. R . W. Lu n t.

D o p p le r effect w ith h o m o g e n e o u s h y d ro g e n c a n a l ra y s . W. Rie z l e r (Ann. Physik, 1929, [v], 2, 429—444).—The positive-ray stream contains H*, H 2+, and H 3+, b u t when a variable m agnetic field is applied it is possible to determ ine th e Doppler effect for homogeneous beams. The velocity calcu­

lated from the deflexion by m eans of an electric or a m agnetic field agrees w ith th a t calculated from the Doppler effect, according to which th e velocities observed are in th e ratio 1 : i : agreeing with H +, H 2+, and H 3+. The Doppler distribution for inhomogeneous positive rays observed by Krefft is in fair agreem ent w ith the superposed effects of all the homogeneous beams obtained by m agnetic resolution. The intensity distribution of the Doppler effect for the complete stream is conditioned by processes occurring in th e beam itself.

R . A . Mo r t o n. N e w c o n n e x io n b e tw e e n th e a b s o r p tio n s p e c tr u m of h y d ro g e n a n d th e m a n y -lin e d s p e c tr u m . 0 . W. Ri c h a r d s o n (Nature, 1929,124, 408).—A num ber of band system s ending on Dieke and Hopfield’s C level have been found; th ey all have initial states which are identical with one or other of the initial states of the bands ending on the 21(S level, and they h a v e . the character of band sequences similar to those of the a, ¡3, and other systems ending on 23S . They all have P ' , Q, and B '

1116 B R IT ISH CHEMICAL ABSTRACTS.-— A.

branches. The correctness of H ori’s analysis of th e W erner (C— >-A) bands is confirmed. There are four, or probably five, 2-levels in the H2 spectrum.

A classification of a num ber of the m ore im po rtant levels is suggested. A. A. El d r i d g e.

B a n d s p e c tr a of lig h t m o le c u le s . I. S p e c tr a of h e liu m a n d h y d ro g e n . W . We iz e l (Z.. P h y s ik , 1 9 2 9 , 5 6 , 7 2 7 — 7 3 9 ).— T h e c h a r a c te r is tic fe a tu r e s sh o w n b y th e b a n d s p e c tr a o f lig h t m o le c u le s are : (a) th eir approxim ate conform ity to one or other of the two expressions deduced by H und (A., 1926,.

657), according to th e rotation al energy invo lv ed;

(b) the complete decoupling of th e spin, th e /-de­

coupling attaining a very high degree; ..(c) th e non- appearance of spin-m ultiplet structure, and (d) th e rare appearance of intercom bination between term system s of different m ultiplets. From this point of view the molecular spectra of hydrogen and helium are discussed, new term s being identified and previous interpretations corrected. The decoupling is dealt w ith in detail. An explanation is offered for the relative intensities of the Io n ia n and W erner bands of hydrogen and th e interp retatio n of th e former as singlet-singlet combinations is discussed.

J . W . Sm it h. N e a r in f r a - r e d s p e c tr a of h e liu m a n d m e rc u ry . T . Ta k a m in e and T. Stjga (Sci. Papers In st. Phys.

Chem. Res. Tokyo, 1929, 11, 131— 137).—Using a m ethod of heating a plate sensitised with neocyanine, and also by phosphoro-photography, the near infra­

red spectra of helium and m ercury were studied.

W ith a strong-current Geissler tube as a source of light th e helium line 2ZS —23P could be photographed w ith 5 min. exposure. The m ercury arc spectrum showed a group of lines in th e region 0-7— 1 ¡x. The rem arkable in ten sity of the two oxygen lines 0-846 and 0-777 ¡x appearing as im purity is noted and its possible bearing in astrophysics is mentioned.

N. M. Bl ig h. H e liu m b a n d s p e c tr u m . I I I . S. Im ani.shi (Sci. Papers In st. Phys. Chem. Res. Tokyo, 1929, 1 1 , 139— 149; cf. this vol., 616).;—New m easurem ents on the higher m embers of the helium 2 3aS—wipnll series bands in the ultra-violet region were m ade. Two new bands, 2s1— S^ 1 and 2s1—Dp1, were found and analysed. Constants are calculated for these bands, and a R ydberg formula is suggested which expresses th e electronic frequencies accurately, especially of the higher members. The convergence value of co0, the vibrational constant for the ^-states, is found to be

1627-2+0-2 cm.- 1 N. M. B lig h .

P r o p e r tie s of a c la s s of m o le c u la r te r m s , e sp e c ia lly t e r m s of th e h e liu m m o le c u le . G. H.

Di e k e (Z. Physik, 1929, 57, 71— 105).—M athe­

m atical. The rotational stru ctu re of th e term s of a molecule is investigated. A calculation due to Hill and Van Vleck is used to follow th e transition from stronger to weaker ro tatio n or from weaker to stronger coupling. The results obtained are tested on th e term s of th e helium band spectrum . A. J . Me e.

E le c tr o n d is tr ib u tio n in n o r m a l h e liu m . H.

Be t h e (Z. Physik, 1929, 55, 431—436).—An ex­

ceptionally good agreem ent is to be found between the values of th e energy term s and the eigenfunk-

tion of th e norm al helium atom calculated by the self-consistent field m ethod of H artree and by the analysis of Hylleras. R . W . Lu n t.

M e a s u r e m e n t of e x c ita tio n [fu n ctio n s] of the h e liu m s p e c tr u m . W. Ha n l e (Z. Physik, 1929, 5 6 , 94— 113).—The principle of th e design of an apparatus for examining th e spectrum of helium m ay be examined as a function of th e velocity of the exciting electro ns; this is discussed a t length together w ith th e errors involved in th e procedure finally adopted. The gas pressure was adjusted to a value below th a t a t which pressure effects become appre­

ciable, and spectrogram s were taken of the lumin­

escence in th e range 6678—3819 A. produced in a field- free space by electrons of velocity up to 450 volts. The curves for th e intensity as a function of th e electron velocity show th e following characteristics : a well- defined maxim um lying a t or below about 100 volts;

lines of the same spectral series have curves of the same type, characterised by the sharpness and position of th e maximum ; and the; triplet lines exhibit a very m uch sharper m axim um th a n do the singlet lines, th e 2 p —m s serie s having th e sharpest

maxim um. R . W . L u s t.

S p e c tro s c o p ic o b s e r v a tio n s of th e low -voltage n itr o g e n a r c . H . Ha m a d a (Sci. R ep. Tohoku, 1929, 1 8 , 155—164).—The distribution of spectra over th e region 3S00'—7000 A; in th e low-voltage nitrogen arc produced in a two-electrode hot-cathode tu b e has been studied. .From the intensity distri­

bution in the negative band system y it would appear th a t the greater is th e voltage applied th e greater is th e probability of tran sition from the initial states a t which the molecule has th e greater vibrational quantum num ber. R . Cu t iiil l.

E x c ita tio n of th e a r c s p e c tr u m of n itro g en . 0 . S. Du f f e n d a c k and R . A. Wo l f e (Physical Rev., 1929, [ii], 34, 409—420).—The complete arc spectrum of nitrogen was excited and m easured in th e region 8800— 3400 A. w ith a low-voltage arc in a mixture of helium and nitrogen. The first step of th e excitation process is probably th e dissociation of nitrogen mole­

cules into n eutral atom s by im pacts of th e second kind w ith m etastable helium atom s, and th e second step th e excitation of neutral nitrogen atom s by second im pacts with m etastable helium atom s. Wave­

lengths, classifications, and a num ber of new lines are

tabulated. N.vM. Bl ig h.

C la ssific a tio n of th e a r c s p e c tr a of n itro g e n a n d c a rb o n . S. B. I n g r a m (Physical R ev., 1929, [ii], 3 4 , 421—430).—The wave-lengths and approxi­

m ate intensities of about 125 lines of th e arc spectrum of carbon between 9000 and 2 0 ,0 0 0 A. were measured.

A bout 90 of these lines are classified as combinations between term s in C l and N I; new term s are fixed in both of these spectra and m any identifications m ade by Fowler and Selwyn in C i (cf. A., 1928, 450) and by Compton and Boyce (cf. this vol., 365) are confirmed. Using the d a ta of Duffendack an d Wolfe (cf. preceding abstract) a num ber of lines is cla ssified in th e visible spectrum of N I. The designations of the 2P and 2D term s found by Kiess (cf. A., 1925, ii, 911) a t 5100 and 6800 A., respectively, are corrected by

GENERAL, PHYSICAL, A N D INORGANIC CHEM ISTRY. 1117

assigning them to th e 2s22_p2 . 3p configuration, basing them on the 1D term of :N n . A complete term table of N i is given. N. Mi Bl i g h.

A d d itio n a l s e r ie s lin e s in th e s p e c tr a of C II and N II. I. S. B o w e n (P hysicalR ev., 1929, [ii], 34, 534—53G).—Several of the strong lines of C n were identified as com binations of the 4P term of the sp- configuration w ith the quadruplet term s observed by Fowler an d ■ Sehvyn (cf. A., 1928, 1165). This fixes the term values 4P 1=20G810-7, 4P S=206789‘2, 4P 3=206760;6. In N n the term value lS term of the slp i and the lP term of the sp3 configurations were fixed a t 206,159 and 72,084, respectively, relative to Fowler and F reem an’s term values (cf. A., 1927, 489).

The lS term combines w ith the lD term of the same configuration to give the nebular line a t 5754-8 A.

The relative term values are recorded for a num ber of terms, including those designated “ a ” ' by Fowler and; Freem an, and these were shown to belong to the quintuplet system. , N. M. B l i g i i .

N itro g e n a fte rg lo w . E. J . B. W i l l e y (Nature, 1929> 124, 443—444).—The decay of the afterglow is only p a rtly a homogeneous reaction in clean glass vessels, the am ount of divergence from homogeneity being determ ined by th e p u rity of the nitrogen. The influence of th e necessary traces of electronegative gas is confined to the surfaces. U nder conditions of minimised wall d ecay no appreciable change in the order of th e reaction takes place with tim e,;and no apparent alteration occurs in the afterglow spectrum so far as 4000 A. No simple relationship exists between the intensity of the glow and the concen­

tration of chemically; active nitrogen, except possibly under certain conditions determ ined b y :th e state of the walls. A ddition of small am ounts of oxygen or nitric oxide to a stream of glowing nitrogen causes a large developm ent of radiation in the blue and regions of shorter wave-lengths. The intensity of the green flame w ith nitric oxide (5%) is related to th e con­

centration of chemically active nitrogen as is th a t of the lum inosity of th e nitrogen alone. The decay of a stream of luminous nitrogen can be varied by addition of photogens w ithout affecting the am ount of the chemically active species present. I t therefore appears th a t atom s are concerned in both phenomena, and it is suggested th a t tho chemical activity is probably7 due to 2-3-volt m etastable atoms.

A. A. El d r i d g e. R o ta tio n a l s t r u c t u r e of th e r e d b a n d s of so d iu m . W. R. Fr e d r i c k s o n (Physical Rev., 1929, [ii], 34, 207—212).—Three bands of the red band system of sodium photographed a t high dis­

persion were m easured. Two strong series of lines run through th e bands and these series are shown, to be P and R branches, and the com bination constants are evaluated. The final state values and those of the blue-green system agree closely and th e constant for the upper sta te gives I 0' — 255-3 X10-40 g.-cm.z and V = 3 - 5 2 x 10-8 cm. The electronic transition is con­

cluded to be of th e 1S —1S type. N. M. Bl ig h. U ltra -v io le t s p e c tr u m of m a g n e s iu m h y d rid e . II. M a n y -lin e d y -s y s te m . R. W . B. Pe a r s e

(Proc. R oy.'Sop., 1929, A, 125, 157— 179; cf. this vol., 376).—The bands of the y-system of magnesium

hydride have been measured in th e region 2560—

3240 A. The distribution of band origins ab ou t th e origin of th e system is represented by th e equation v0= 3 5 9 0 4 -5 + 1 1 3 8 -4 (« .'+ !)-9 -5 (w '+ i)2-

1702-2(?i"~|-£)+34-2(tt"+!)2. The rotation al stru c­

tu re has been analysed and th e bands have been found to consist of a single It and P branch, w ith one line missing a t th e origin. The stru ctu re is characteristic of th e *S— >lS typ e of transition, and th e values for the vibrational constants show definitely th a t th e two new' 3 levels are quite d istinct from the 2S level of th e a- and (3-systems. Using a rotational energy term of the form F ( jk) = B nj t (jk+ \ ) ~ A i[ji(j< t+1)]2, the following values are calculated : B n’ = 4-3020—0-0492(?i + i ) —0-0050(«+ A)2, B u" = 6-37S2—0 -1 8 5 4 (« + |), £»„'=2-16 x 10-4, £ „ " = 3 -4 5 x 10-4, /„ '= 6 -4 3 9 x 1 0 -« g.-cm.2, /0" = 4 -3 4 3 x 10-4» g.- cm .2, r0'= 2 - 0 2 x lO-8 cm., 1-65‘x 10- 8 cm. The observed values for th e isotope effect, m easured for th e (0,2) and (0,1) bands, are in good agreem ent w ith those calculated on th e assum ption th a t th e bands are em itted by the diatom ic molecule, MgH, magnesium having isotopes of Weights 24, 25, and 26.

L . L . Bir c u m s h a w. O p tic a l e x c ita tio n of p h o s p h o r u s v a p o u r. A.

J a k o v l e v and A. T e r e n i n (N ature, 1929, 124, 337).—Fluorescence em itted in th e region 3500—

1900 A. was observed in phosphorus vapour (0-1 mm.) a t 600—700° illum inated by spark radiation Cd 2195 , 2144 A., Zn, 2100, 2 0 6 2 1 ., or A1 1990, 1935 A. The first vibration q uantum and the dis­

sociation energy of the norm al P2 molecule are estim ated, respectively, to be 775 cm.“1 and 6 volts.

A. A. El d r i d g e. S p a r k s p e c tr a of s u lp h u r. L. Bl o c h and E.

Bl o c h (Ann. Physique, 1929, [x], 12, 5—22).—A more detailed account of work already published (this vol., 225).

A rc s p e c tr u m of c h lo rin e a n d i t s s tr u c tu r e . C. C. K i e s s and T. L. d e B r u i n (Bur. S tand. J . R es:., 1929, 2, 1117— 1136).—More th a n 200 lines between 4000 and 9900 A. have been m easured in th e arc spectrum of chlorine excited b y an uncondensed discharge in a Geissler tu b e a t low pressure. Of these lines 62%, together w ith those observed by Turner in th e Schum ann region, have been classified as resulting from com binations between term s of the doublet and quadruplet systems. They arise m ainly from the basic term 3P of th e ion, th e term s coming from *S and 1D no t being definitely established, owing to th e faintness of th e lines. From these, th e distance separating 2P2 from 3P2 is 104,991 cm .-1, giving 12-96 volts as the ionisation potential of neutral chlorine. C. J . S m i t h e l l s .

A rc s p e c tr u m of c h lo rin e . K . Ma j u m d a r

(Proc. R oy. Soc., 1929, A, 125, 60—68).—From a consideration of the fact th a t in the arc spectra of a group of successive elements (e.g., alum inium , silicon . . . potassium ) th e wave-numbers of th e strongest lines of th e elements due to th e transition N — N 2 increase linearly w ith the atom ic num ber (cf. Saha and Mazumdar, A., 1928, 1296), it is found th a t the group of chlorine lines due to th e tran sition 4 M 2 (iVj-i— N 2) should lie in th e region 9300—

m s

7700 A. Therefore the group a t 4700—4200 A.„

identified by L aporte (A., 1928, 805) as being due to the transition 4j¥ 2 [Nx<— N t ), m ust be actually ascribed to 4 M 2 (N < — 0 2). W ith a view to discover these predicted infra-red lines th e spectrum of chlorine has been photographed in the region 6400—

8700 A. The means taken to surm ount the experi­

m ental difficulties are discussed. The new lines observed in the infra-red are identified as belonging to th e transition — iV2 and are tabulated in th e form of m ultiplets. The ionisation p otential is calculated to be 13-1 volts. L. L. Bi r c p m s h a w.

S p e c tr u m of tre b ly -io n is e d a r g o n . D. S. Jog

(Nature, 1929, 124, 303).—The lines of treblv- ionised argon have been classified ; all the q u ad ­ ruplet m ultiplets due to the transition 2M % (N 2<— N have been obtained. A. A. El d r i d g e.

N u m b e r of e x c ite d a to m s a n d th e a b s o r p tio n s p e c tr u m of n ic k e l v a p o u r. A. T. Wil l ia m s; (Nature, 1929, 124, 373).—The equation N '/ N = e -EIRT, where N \ is the num ber of excited atoms, N the to tal num ber of atoms, and E the energy, is modified to N ’jN ~ e ~ lE+AEilRT, since it is necessary to consider also the energy consumed in passing from, one norm al configuration to another : (3d)8(4s)2— >- (3d)9(4s)= A E . " A. A. El d r i d g e.

N ew z in c h y d rid e b a n d s in th e u ltra -v io le t. E . Be n g t s s o n and B . Gr u n d s t r o m (Z. Physik, 1929, 5 7/ 1—1 0).—The new band system investigated extends from X2500 to 2 0 0 0 A., and five bands a t 2092, 2152, 2240, 2332, and 2351 A. are analysed.

These bands have only simple P and R branches, the.

P -b ran ch being always somewhat more intense. The term-differences of the initial and end states are obtained. They represent a XS' — el ect ron switch of an ionised Zn+H dipole. From these five bands i t is possible to place a large num ber of bands in a level scheme. The dissociation energies of both electron states are calculated from th e convergence lim it to be D '= 4 -4 and D " = 2 -5 volts. A. J . Me e.

R e g u la ritie s in th e a r c s p e c tr u m of a rs e n ic . K . R. Ra o (Proc. Roy. Soc., 1929, A, 125, 238—246).

—The spectrum of arsenic has been photographed between 8800 an d 1370 A., using as source the arc between m etallic arsenic contained in poles of carbon or aluminium. No lines have been observed in th e infra-red region. The arc in nitrogen was used for investigating the Schum ann region down to 1650 A.

and below this region the arc in a vacuum between carbon poles containing arsenic and the spark between metallic arsenic in hydrogen. Several new lines have been measured which have led to the identification of combinations due to the electron transitions from the deepest i p state to the higher 5s, 4p ’, and 4d states.

The resonance trip let corresponds with the com bin­

ation 4p‘iS — 5siP . L. L. BirCumsiiaw.

M a g n e tic s e p a r a tio n in th e s p e c tr u m of io n is e d k ry p to n . C. J . Ba k k e e and P. Ze e m a n (Proc. K.

Akad. W etensch. Am sterdam , 1929, 32, 565—577).—

The m agnetic separation of the K r II spectrum has been investigated. The j-values of th e term s arising from the coupling of the 5s and 5p electron have been fijced and it is shown th a t ^-values with strong

*! anom aly ” appear. The g-sum rule has been confirmed. A comparison of the «¡'-values of analogous term s in the spectra of Ne n , A n , and K r i i is given.

R . A. Mo r t o n. F i r s t s p e c tr u m of k ry p to n . W. F. Meg g er s, T. L. d e Br u i n, and C. J . Hu m p h r e y s (Bur. Stand.

J . Res., 1929, 3, 129— 162).—A bout 200 fines (3302-54 to 9751-77 A.) characterising the spectrum of neutral krypton atom s have been photographed, and a list of estim ated intensities and measured wave-lengths has been obtained. The m ain spectral terms, analogous to those in th e related neon and argon spectra, have been identified and nearly all the KrT lines classified in series of various types. From the com binations and series lim its absolute term s are derived and th e ionisation potential of 13-940 volts is deduced. The K r i spectrum is closely analogous to Ne I and A i and good agreem ent with theory is obtained. The proposed substitution of the krypton line 5649-56 A. for th e cadmium line 6438-4696 A:

as a prim ary stan d ard of wave-length is open to objection because the k ry p to n line has relatively low in ten sity and involves a m etastable level. The line 5870-92 A. is in these respects preferable, but it cannot be recom m ended as a standard until it has been re-examined for hyperfine structure.

R. A. Mo r t o n.

E x te n s io n of th e C d i-lik e isoelectronic se q u e n c e to S b i v a n d T e v . R. C. Gi b b s and (Miss) A. M. Vie w e g (Physical Rev., 1929, [ii], 34,400—405).

—The spectra of tin, antim ony, an d tellurium were photographed w ith a vacuum spectrograph. New S n m lines, additional to those classified by Green and Loring (cf. A., 1928, 2), were identified, particularly second members of series. Sb iv lines were identified arising from transitions 5s5^?—5s2, and from 5p\

5sod, and 5s6s to 5 s5 p; some second m em bers were also found. The Cd i-like isoelectronic sequence was extended through T e v by th e classification of lines resulting from transitions as in Sb iv.

N. M . Bl ig h. F i r s t s p e c tr u m of x e n o n . W. F. Meggers, T . L. d e Br u i n, and C. J . Hu m p h r e y s (Science, 1929, 69, 406; Chem. A bstr., 1929, 2655).—A pre­

lim inary notice of a new list of estim ated intensities and measured wave-lengths for about 300 lines of the X e i spectrum between 3442-7 and 9923 10 A.

Spectral term s accounting for m ost of the lines have been identified; th e largest term , 1<S10, represents the norm al state of the neutral atom and has a value 97,835, from which an ionising potential of 12-078 volts is derived. L. S. Th e o b a l d.

A rc s p e c tr u m of p la tin u m . J . J . Lt o n g o o d

(Physical Rev., 1929, [ii], 34, 185— 198).— Eight new levels and 44 new com binations were found in P ti, and 56 new lines were measured in the ultra-violet.

M any levels were interpreted from an exam ination of the Zeeman effect and from the ^-values. Low structures are d9s, # s 2, and d 10; middle term s arise from drjp and dss p ; high configurations are d9 . s, d8s . s, with an indication of dss . d. Q uintuplet terms appear in the middle and high sets. Evidence shows th a t th e R ussell-Saunders coupling has almost completely broken down. Sim ilarity with the

G ENE RAL, PHYSICAL, AND INORGANIC CHEMISTRY. lilt)

theoretically analogous spectrum of N i l is only approximate. The ionisation potentials calculated approximately are 8-9 for d?s to. d9 and 9-7 volts for

dds to d8s. N. M. B lio h .

M ean life fo r th e m e r c u r y s p a r k s p e c tr u m . L. R. Ma x w e l l (Physical R ev., 1929, [iij, 3 4 , 199—

206; cf. this Vol., 112).—Mean life values found for the lines 3114 and 2572 of H g iV were 9 x l 0“7 and 8x KH sec., and for the lines 3090, 3312, and 4797 Â.

of H g m were 6 X 10~7, 4 X 10~7, and 4 x l 0~7 sec., respectively. The prom inent lines of H g n have an estimated m ean life of the order 10 8 sec. These results indicate th a t the greater the charge, of the ion producing the line the longer is the m ean life. Spark lines are produced by single electron collisions.

N . M . Bl ig h. L ine a b s o r p tio n of m e r c u r y v a p o u r fo r th e line 2 5 3 7 À. H. Ko f f e r m a n n and W. Tie t z e

(Z. Physik, 1929,5 6 , 604— 616).—The to tal absorption of the 2537 Â. m ercury resonance line by a layer of mercury vapour was m easured by a photo-electric method a t five different tem peratures between —1 1° and 20°. T h e/-v alues for th e line were calculated by the use of a formula connecting it with the product of th e m axim al absorption coefficient and th e layer length. F or th e comparison of th is/-v alu e with th a t derived by other methods, e.g., m easurement of the absorption of thepressure-w idened lines and anomalous dispersion, it m ust he remem bered th a t the 2537 Â.

line has a complex structure, being made up of five lines a t a distance of about 0-01 Â. from each other.

T h e /-v a lu e is interm ediate between those obtained for the extrem e lines and is practically equal to one fifth of th e to tal /-values. The normal life-period of the 3P 1 state of m ercury calculated from five times the /-value is in good agreem ent w ith th e life-value obtained by other methods. A. J . Me e.

E fficiency of e x c ita tio n b y e le c tro n im p a c t a n d a n o m a lo u s s c a tte r in g in m e r c u r y v a p o u r. W. H.

Br a t t a in (Physical Rev., 1929, [ii], 3 4 , 474—4 8 5 ; cf. Messenger, A., 1927, 85; Maxwell, A., 1926, 989;

Jones, A., 1928, 1168).—The efficiency of excitation by electron im pact of th e 6-67-volt resonance level in the m ercury atom was studied as a function of the energy of th e incident electrons. The electrons which have lost energy are separated out by a small retarding field and measured. The num ber of collisions is calculated from th e experim ental value of th e mean free path. F or th e range 6-67—7-07 volts th e m axi­

mum efficiency is 0-06 a t 6-77 volts, falling to 0-04 a t 7-0 volts. The num ber of electrons scattered elastically a t large angles by mercury vapour as a function of th eir energy was measured for an energy range of 2— 10 volts. Singularities were found corre­

sponding with an increase in large angle scattering, the most prom inent being a t 4-9, 5-7, and 6-3 volts, and the less prom inent a t 9-6, 10-3, and 11-1 volts.

N. M. Bl ig h. O rig in of lo n g in f r a - r e d r a d ia tio n of m e rc u ry . W. Kr o e b e l (Z. Physik, 1929, 56, 114— 130)— In order to te st th e speculations of F ranck and Grotrian as to the origin of th e radiation from m ercury vapour in the region 400—200 ¡j. th e conditions under which this radiation is em itted and absorbed have been

investigated. The m ethod consisted in measuring the indioations of a radio-microm eter on which w as allowed to fall th e to tal long %vave radiation from a specially designed m ercury arc lamp, th e whole being contained in an atm osphere of dry air to avoid absorption by w ater vapour. R adiation of wave-length below 200 (*

was removed by a series of filters, th e final one being blackened cardboard, for which R ubens gives a trans- m issibility of 9% for ^X ^ 200 ^¡j.. I t was found th a t the radiation ) , £ 200 (J. was quenched by th e addition of hydrogen to th e arc source, b u t slightly increased by th e addition of nitrogen ; th e radiation is absorbed by excited mercury vapour which is produced in th e arc, and by m ercury vapour excited by 4-9-volt electrons. I t is therefore concluded, in agreem ent w ith th e theory of Franck and Grotrian, th a t th e radiation in question arises from a m etastable excited

Hg2 molecule. R . W. L u n t .

R e la tio n b e tw e e n th e in te n s itie s of m u ltip le ts of m e r c u r y a n d of n e o n a n d th e e n e rg y of th e e x c itin g e le c tro n s . W. E n d e (Z. Physik, 1929, 5 6 , 503-—515).—The relative intensities of m ultiplets of m ercury and of neon have been determ ined from spectrograms of th e luminescence excitcd in .these gases in an approxim ately field-free space by electrons of energy from 12 to 40 volts. The relative intensities of the components of th e m ercury trip let 23P —23S V 5461, 4358, and 4047 A. were 60 : 1 00 : 4 9 ± 5 % ; the ratios were independent of th e electron energy and alm ost independent of th e strength of th e electron beam. F or th e trip let 23P 2~ 3 3D, 3663, 3655, 3650 A . th e ratios were 53 : 51 : 100, again independent of th e electron energy. The m easurem ents in neon were carried out a t a constant pressure of 0-8 mm. and relate to th e red and yellow groups of lines. The relative intensities differ widely from those observed by Dorgelo in th e positive column. R . W. L u n t .

S p a r k s p e c tr u m of th a lliu m , T i m . J . C.

McLe n n a n, A. B . McLa y, and M . F. Cr a w f o r d

(Proc. Roy. Soc., 1929, A, 1 2 5 , 50—53; cf. Carroll, A., 1926, 214).—The term structure of th e second spark spectrum of thallium has been investigated, and th e real F and G first members and members of each of th e S, P, and D series higher th a n those already located by Carroll (loc. cit.) have been id en ti­

fied. The term values are based on th e value zero for the state 1S (od10) reached a t ionisation. The ionis­

ation potential, com puted from th e value 240,600 cm.-1 for 6s2*S'j term , is 29-7 volts. The wrave-lengths assigned to T1 m , together w ith th eir wave-nuinbers, intensities, and series designations, are tabulated.

L . L . Bir c u m s h a w. S e c o n d s p a r k s p e c tr u m of le a d , P b i n . S.

S m it h (Physical Rev., 1929, [ii], 3 4 , 393— 399; cf.

this vol., i l3 , 227).—The m ultiplets 6aP0,i,2—83<S'j,

^ o . U —7*9i&3, 73A), 1.2—'73^> 1.2.3, and 63P 0.i,2—PPzP i arising from combinations between trip let term s of Pb i n wrere found. Twenty-one lines arising from combinations betwreen singlet term s and in te r­

combinations between singlet and trip let term s were identified. The 'pp^-D, term is found in com bination w ith 63P J>2, 6lP j, 6 3P2,3> and <SlF z. Seven new lines of T1 i i are given, corresponding w ith some of th e Pb i n combinations. N. M . B l i g h .

1120 £ R iT iS H CHEMICAL ABSTRACTS.— A.

B a n d s p e c tr a . II. Ru e d y (J. Pliys. R adium , 1929, [vi], 10, 129— 160).—A theoretical review based on H u n d ’s theory slightly modified in regard to rotato ry “ doubling.” T he phenom enon is a general property of molecular electronic levels ¿ „ > 0 and is no t characteristic merely of sym m etrical molecules.

The relationships between th e spectra of copper, silver, and gold atom s and th e spectra of compounds of these m etals disclose evidence of transitions between m etastable states. The theory of th e intensities of th e bands of molecular oxygen is in agreem ent w ith experience concerning th e vibration state. The selec­

tion rules for transitions between rotation levels (P , Q, and R branches) are no t as strict as for atom s and depend on th e speed of rotation and th e n atu re (electric moment) of th e molecule. R . A. Mo r t o n.

S p e c tr u m of th e a u r o r a b o re a lis . J . K a p l a n (Science, 1929, 69, 296—297).—Two weak lines of wave-length 5176 and 5149 A. in th e auroral spectrum can be accounted for b y assuming th a t collisions of th e second kind tak e place between nitrogen atom s in th e zD-atate and m etastable nitrogen molecules in th e A 0 and levels. L. S. T h e o b a l d .

S p e c tr u m of s u n lit a u r o r a r a y s a s c o m p a r e d w ith th e s p e c tr u m of lo w e r a u r o r a in th e e a r t h ’s sh a d o w . C. St o r m e r (Nature, 1929,124, 263—264).

F o r m a n d s t r u c t u r e of s p a r k s . VI. T.

Te r a d a, U. Na k a y a, and R. Yam am oto (Sci. Papers In st. Phys. Chem. Res., 1929, 10, 271—290).—The effects of various volatile organic compounds on th e form of long sparks in air has been investigated. The vapours of m ethyl iodide, ethyl iodide, chloroform, carbon tetrachloride, pp-dichloropropane, and ethylene dichloride smooth ou t a zigzag sp a rk ; ethyl brom ide lias no effect. Alcohols, ether, and acetone cause branching of th e spark and a gap in lum inosity near the negative .electrode. Benzene vapour is supposed to favour the form ation of positive brush discharges.

A transition stage in th e type of spark is produced by benzoyl chloride and benzotrichloride—a smooth p a rt of th e spark near each electrode only. Halogen com­

pounds in general, particularly those of chlorine, have th e effect of smoothing zigzag sparks if their concen­

tra tio n exceeds a critical value depending on th e ratio of th e mass of th e halogen atom to th e mol. w t. of th e compound. This effect is held to be due to enhanced ionisation near the electrodes favouring th e form ation of brush discharges. C. W . Gi b b y.

T e m p e r a tu r e a n d th e C o m p to n effect.

G. E. M. J a u n o e y and H . B a u e r (Physical Rev., 1929, [ii], 34, 387—392).—On D ebye’s theory th e intensity of the unmodified X -rays scattered by crystals should increase w ith rise of tem perature, and th e ratio of modified to unmodified rays should decrease. This was investigated, using De F oe’s m ethod (ibid., 1926, 27, 675). An alum inium absorb­

ing sheet was transferred from th e prim ary to th e scattered beam and th e ratio of th e two ionisation currents was found a t —140°, 25°, and 565°. X -R ays for th e jvave-length range 0-32—0-62 A. were scattered by carbon a t angles of 60°, 75°, and 90°, and by aluminium and copper a t 90° and 130°. No effect of tem perature on th e ratio of modified to unmodified

rays was detected w ithin th e lim its of experimental

error. N. M. Bl iq h.

C o m p to n s c a tte r in g a n d th e n e w s ta tis tic s . S.

Ch a n d r a s e k h a r (Proc. Roy. Soc., 1929, A, 125, 231—237).—M athem atical. The Compton scattering by an electron gas is considered on th e Fermi-Dirac statistics. E quations are derived which indicate that th e distribution of intensity of th e radiation scattered by a degenerate electron gas follows a parabolic and n o t an exponential law. The theo ry predicts the peak of maxim um intensity a t a place where the Compton theory for a free-stationary electron predicts a line. Evidence is obtained th a t th e Compton scattering of an electron gas should n o t be influenced by tem perature or by th e presence of a magnetic

field. L. L. Bir c u m s h a w.

R e fle x io n a n d a b s o r p tio n of X -ra y s of large w a v e -le n g th . M. A. Va l o u c h (Compt. rend., 1929, 189, 283—285).—The reflecting powers of m irrors of different m aterials (lead and flint glasses, aluminium) have been m easured as a function of the angle of incidence for th e monochrom atic K«. carbon rays (44-9 Ä.) produced a t 320 volts. P rin s’ formula (A., 1928, 451) was applied to th e experimental results, and th e refractive indices and absorption coefficients were obtained. Gr a n t.

C o n tin u o u s X -ra y s p e c tr u m . C. Eckart

(Physical Rev., 1929, [ii], 34, 167— 175).—B y the m ethod of wave mechanics th e quantum theory is shown to be capable of accounting for th e various characteristics of th e continuous X -ray spectrum.

The theoretical principles are deduced from a schem­

atic model, assuming th a t th e ta rg e t of an X -ray tube is a plane.m irror which completely reflects th e electron

waves. “ N. M. Bl ig h.

R e s o lu tio n of th e lin e Lß2 in to i t s d ia g ra m c o m p o n e n ts a n d th e re la tiv e w id th s of some X -ra y s p e c tr u m lin e s . S. K . A l l i s o n (Physical Rev., 1929, [ii], 34, 176— 180).—The line Xß2 in irid­

ium, thallium , an d uranium was investigated using th e double spectrom eter. Theoretically ' th e lino should be double, its weaker com ponent being designated £ ß 15.- In iridium no separation was obtained owing to the w idth of Xß2 ; in thallium an incomplete resolution was obtained. In uranium practically complete separation, averaging 1 -S6 against a theoretical 1-94 Ä., was found. Some experiments were m ade on th e intrinsic w idths of certain lines in the L series of thallium and lead. N. M. B l i g i i .

I i-X -ra y a b s o r p tio n s p e c tr a of s o m e ch lo rin e c o m p o u n d s in a q u e o u s so lu tio n . O . St e l l in g

(Naturwiss., 1929, 17, 689).—The wave-length values for the X -absorption spectrum of chlorine are slightly different when solid sodium chloride, potassium chloride an d chlorate are used. The results for nearly satu rated solutions of chlorides differ, no t only among themselves, b u t also from those of th e solids.

The values for potassium chlorate in solution and in th e solid sta te are alm ost th e same. R . A . Mo r t o n.

S c a tte r in g of X -ra y s in m e r c u r y v a p o u r. P.

Sc h e r r e r and A. St a g e r (Helv. phys. Acta, 1928,1,

GENERAL, PHYSICAL, AN D INORGANIC CHEMISTRY. 1121

518—533; Chem. Zontr., 1929, i, 1081).—The .F-curve corresponds w ith th a t obtained for crystals.

A . A . El d r i d g e. M otion of a n e le c tric a r c in a m a g n e tic field u n d er lo w g a s p r e s s u r e . R . Ta n b e r g (Nature, 1929,124, 371—372).

U se of s e r ie s in d u c ta n c e in v a c u u m s p a r k sp ec tra. R . C. G ib b s , (M is s ) A. M . V i e w e g , and C. W. G a r t l e i n (Physical Rev., 1929, [ii], 34, 406—

408).—An inductance was used in series with the spark gaps and condenser, as an aid to the identific­

ation of lines in the spark spectra of antim ony and tin in the region below 2600 A. The general effect is to strengthen lines arising from lower states of ionis­

ation and to weaken those from higher states. Line identifications for th e two elements are corrected.

N. M. Bl ig h. E le c tric a r c in m ix e d g a s e s . F . H . N ew m an (Phil. Mag., 1929, [vii], 7, 1085— 1091).—The au th o r’s third electrodem ethod has been found to be inadequate for the startin g of arcs in helium and neon atmospheres a t pressures of 10-3 to 1 mm. This difficulty was eliminated b y th e fusing to the cathode of a small bead of borax, th e effect of,which is assumed to be a reduction in th e cathode potential fall, which is then sufficient to ionise the molecules. W ith th e arcs in the pure gases th e characteristic radiations were absent, the arc core being white in colour with a surrounding column of yellow radiation showing only the Z>-lines. The Balm er lines were present with the CN bands. The effect of the addition of w ater vapour to either neon or helium was to enhance the Balmer lines and other lines of im purities while reducing considerably the spectra of the m ain gases.

A . E . Mit c h e l l. N ew e le c tro sc o p e . B. F . J . Sc h o n l a n d (Proc.

Camb. Phil. Soc., 1929, 25, 340—343).—A new type of ionisation electroscope is described which combines a high voltage-sensitivity with a small capacity, and has a sensitivity for q u an tity of electricity of the same order as a com bination of an ionisation chamber arid a Compton electrom eter. J . W. Sm it h.

P h o to - e le c tr ic th r e s h o ld of a d o u b ly -e v a p o r­

a te d film . R . B . J o n e s (Physical Rev., 1929, [ii], 34,,227—232).—An investigation was m ade to deter­

mine whg,t energy fraction is spent in separating the electron from the original atom , and w hat work is done to tak e th e electron through th e outer surface when light is incident on a m etallic surface. The preparation of a platinum , nickel, and platin u m - nickel film by evaporation, sim ultaneously and under the same conditions, is described. Threshold wave­

lengths found from saturation photo-electric currents obtained with monochromatic light of different wave­

lengths were N i 3333, P t 2804, N i-P t 3318 A. Similar investigations with platinum and tungsten gave W 233S, P t 2831, P t-W 2804 A. I t is concluded th a t the threshold of the m ixtures of two m etals is the same as th a t of th e constituent having the threshold of lower frequency. N. M. B l i g h .

P h o to -io n is a tio n of th e v a p o u r s of csesiu m a n d ru b id iu m . E . 0 . La w r e n c e and N. E . Ed l e f s e n

(Physical R ev., 1929, [ii], 34, 233—242).—Light intensities and photo-ionisation were measured simul-

taneously by an im proved form of Foote and M ohler’s space-charge m ethod (A., 1925, ii, 919), and an equation is given connecting th e ionisation per unit light intensity and the frequency in th e case of cæsium and rubidium over th e range 2200—3130 A.

The equation is in agreem ent w ith th e relation th a t the effective collision capture cross-section of th e ions for electrons varies inversely as th e square of the energy of th e electrons relative to th e ions, or the Thomson recom bination law, confirmed by Oppen- heim er by wave mechanics (A., 1928, 456). P h o to ­ ionisation by absorption of principal series lines in cæsium was confirmed and was observed also in

rubidium . N. M. Bl i g h.

L o n g itu d in a l d i s t r ib u tio n of p h o to -e le c tro n s . A. Ca r r e l l i (Z. Physik, 1929, 5 6 , 694—701).—

M athem atical. The longitudinal distribution of photo- electrons is calculated by applying th e new wave- mechanics w ith the theoretical treatm en t of Sommer - feld. The Sommerfeld expression is only a first approxim ation for th e case when th e wave-length is large com pared w ith atom ic dimensions, b u t the form ula here derived applies w ithout restriction of wave-length. The form ula shows a dependence h ith erto unknown on the atom ic num ber of the

substance. A. J . Me e.

E ffe c t of h y d ro g e n o n th e th e r m io n ic e m is s io n fr o m p o ta s s iu m . H . R . La i r d (Physical Rev., 1929, [ii], 3 4 , 463— 473).—The therm ionic currents a t a given tem perature were observed for a potassium surface successively cleaned by distillation and con­

tam inated b y hydrogen, and it was concluded th a t the large currents frequently observed from potassium are due to hydrogen contam ination. A uniform field appeared to be an essential condition for the sa tu r­

ation of the therm ionic currents a t 150— 185°, and probably to below 100°. The R ichardson-type equa­

tion was obeyed over the range 150—210°. A value of approxim ately 1-3 volts for th e therm ionic work function <f> was characteristic of the potassium surface after hydrogen contam ination of a certain typ e (cf.

Richardson and Young, A., 1925, ii, 343). This value was concluded to be due to a layer of K H on the potassium surface, and th e emission being due to its decomposition was chemical rath er th a n therm ionic in origin. Values of <f> of 0-26 volt were found a t 100— 110°. The conductivity of a thin layer of potassium on a pyrex insulating tube was greatly decreased by exposure to hydrogen. N. M. Bl i g h.

D is tr ib u tio n of e le c tro n s b e tw e e n th e p la te a n d g r i d of a th re e -e le c tro d e tu b e a s d e te r m in e d b y p o s itiv e c æ s iu m io n s. J . M. Hy a t t (Physical Rev., 1929, [ii], 3 4 , 486—492).— Previous work was con­

tinued (cf. this vol., 228) using a plane-anode type of tu be containing cæsium vapour, th e source of ions being a tungsten filament. The plate current and positive ion current em itted from the filam ent were observed for several negative grid potentials as th e plate potential was varied; th e num ber of electrons om itted from th e cæsium-covered grid per positive ion was calculated and found to increase uniform ly from zero, a t 95 volts to 0-24 a t 600 volts. An investigation of the ratio of the positive ion current

1 1 2 2 BR IT ISH 'CHEMICAL ABSTRACTS.— A .

to the plate to th e positive ion emission from th e filament is described. N. M. Bl ig ii.

E ffe c t of l ig h t o n th e s e c o n d a ry e le c tro n e m is s io n of a lu m in iu m . E . Fr e y (Helv. phys.

Acta, i .1928, 1, 385—416; Chem. Zentr., 1929, i, 1084).—Illum ination of nonToutgassed aluminium plates with a quartz m ercury lamp causes a dim in­

ution of secondary electron emission in a high

vacuum. A. A. El d r i d g e.

C ritic a l p o te n tia ls of m e th a n e . E. Pie t s c h

and G. M. Sc h w a b (Z. Phvsik, 1929, 55, 231—233).—

The decomposition of m ethane by 15-4-volt electrons- is discussed in relation to the recent work of Hogness and Kvalnes (this vol., 242), in which it was con­

cluded th a t CH3t ions were produced by electrons of this energy. I t is shown th a t the auth o rs’ earlier results could be a ttrib u te d either to direct decom­

position of the m ethane molecule and the sim ultaneous b u t independent form ation of CH3+ ions, or to a rapid secondary reaction between CH3+ ions and monatomic hydrogen adsorbed on the electron source.

R. W. Lu s t. E m is s io n of e le c tro n s f r o m m e ta ls on i r r a d i ­ a tio n w ith A '-ray s. W. Es p e (Ann. Physik, 1929, [v], 2 , 381—426).—T he action of X -rays in liberating electrons from m etals has been studied quantitatively in relation to the different variable factors. On the basis of E instein’s law, of equivalence formula) have been derived for the num ber of electrons set free when the X -rays strike the m etal a t 90° and a t <f>°, and i t is found th a t the emission should increase as 1/cos <j>. For the rapid electrons this proportionality is strictly in accord with experim ent, b u t for the slower secondary electrons th e theoretical value is too low, the.error am ounting to 10% when </> is 50°. The error is less for roughened surfaces th an for highly polished m etal. Comparison of the strength of the electron emission for smooth and roughened surfaces shows th a t the emission of slow secondary electrons from a polished plate is a t a minimum and increases w ith increasing roughness. The emission of rapid electrons decreases a little when a sm ooth plate is roughened. The electron emission increases w ith the atom ic num ber of the m etal irradiated, up to a point when the frequency of the radiation in use is exceeded by th e excitation frequency of the m etal. The rela­

tion between electron omission and the wave-length of the incident radiation has also been studied. The results disclose deviations from the Einstein law of equivalence which are accounted for by variable absorption of electrons by the m etal. Since emission is proportional to the intensity of the exciting ra d i­

ation, it is possible indirectly to determ ine th e relation between intensity of X -rays and voltage applied to th e tube. E lectron emission E (for anticathodes of iron, nickel, and copper) varies as V3 (F = v o lta g e ) minus a constant. Palladium charged w ith hydrogen shows no measurable increase' in emission of slow or rapid electrons as compared with uncharged palladium .

R . A. Mo r t o n. R a d ii a n d co llisio n p r o b a b ilitie s of m e ta s ta b le n e o n a n d m e r c u r y a to m s . M. W. Ze m a n s k y

(Physical Rev., 1929, [ii], 34, 213—226).—Theoretical.

Assuming th a t m etastable atom s either diffuse to the

walls of the container and g iv e u p th eir energy there, or are raised to a higher or lowered to th e normal state by im pact w ith other atom s, it is deduced th at for large values of th e tim e elapsing after the cut-off of the excitation th e average num ber of metastable atom s per c.c. decays exponentially with th e time.

N . M . Bl ig h. A n a ly s is of e le c tro n ic v e lo c itie s b y e le c tro ­ s ta tic m e a n s . A. L. Hu g h e s and V. Ro j a n s k y

(Physical Rev., 1929, [ii], 34, 284— 290).—Theoretical.

From the equation of the electronic orbits it is shown th a t an analysis of electronic velocities should bo possible by a radial electrostatic field as well as by th e usual m agnetic field. The plane of the receiving slit should be a t an angle of 127° 17' to th e plane of th e entrance , slit. Expressions are found for the resolution a t this angle, between two electrons with slightly different velocities, and for the departure from perfect re-focussing of two electrons w'ith the same

velocity. N. M. Bl ig h.

D is tr ib u tio n of m o b ilitie s of io n s in m o is t air.

J . Ze l e n y (Physical Rev., 1929, [ii], 34, 310—334).—

A m ethod which was used to measure th e mobilities and distribution of th e ions in air for various degrees of hum idity is described, and a critical stud y of the effects of diffusion on th e distribution of ions moving in an electric field was m ade; the results obtained were in fair agreem ent with theory. The mobilities of ions varied from 45% greater, th an th e slowest for positive ions to 30% for negative ions. Groups of ions having different mobilities were n o t detected.

Theoretical explanations of the distribution are dis­

cussed. The average m obility found for th e negative ions was 2 -0 0 and for positive ions 1-22 cm./sec. per volt/cm . for w ater contents of 3-2 and 2-7 mg. per litre of air, respectively, N . M . Bl ig h.

L o n g itu d in a l m a g n e tic effec t o n b e a m s of slow e le c tro n s . P e rio d ic c o n c e n tra tio n a n d d ila t­

a tio n . J . Th i b a u d (J. Phys. R adium , 1929, [vi], 10, 161— 176).—See this vol., 231.

Q u a n tu m th e o r y of e le c tro n ic s c a tt e r i n g by h e liu m . N. F. Mott (Proc. Camb. Phil. Soc., 1929, 2 5 , 304—309).—The approxim ations introduced by Born (Z. Physik, 1926, 38, 803) in th e calculation of th e variation of electron scattering w ith angle in the case of hydrogen are discussed. Using th e same approxim ations, similar calculations are m ade for the case of helium, the results being in good agreement w ith th e m easurem ents of Dymond and W atson (this

vol., 368). J . W. Sm it h.

D is p e r s io n e le c tro n s in th e o n e-electro n p ro b le m . J . Ha r g r e a v e s (Proc. Camb. Phil. Soc., 1929, 2 5 , 323—330).—M athem atical. D irac’s rela­

tiv ity quantum mechanics is used to derive the K ram ers-H eisenberg dispersion formula for an atom w ith one electron and the dipole m om ent to which is due th e incoherent scattering. The formulas obtained are sim ilar to those obtained by Klein (Z. Physik, 1927, 41, 407) for th e case of a central field, b u t are not limited to this case. This yields an expression for / (the num ber of dispersion electrons for any fine of the optical spectrum ) in term s of th e solutions of the four wave equations of Dirac’s theory. F or any