April 14, 192 1] 



NATURE 



203 



correspondence between isotopic numbers which differ j 

 by 4, or by the formula pxCt., which may be assumed | 

 to represent an a-particle plus two cementing or 

 y3-electrons. The relations pf the light atoms are 

 thus very similar to those of the radio-active atoms. 



It may be of interest to note that during an a-change 

 there is no change in the isotopic number ; in a 

 i8-disintegration the isotopic number decreases by 2. 

 Of these two units one is due to the decrease of the 

 number of negative electrons in the nucleus by one, 

 and the other to the resultant increase of the atomic 

 number (M) by one. The addition of a proton to a 

 nucleus would increase the atomic number and de- 

 crease the isotopic number by one each. Thus the 

 addition of a positive electron to the nucleus of 

 Mg 122^' would give i3i^% which is ordinary alu- 

 minium. 



The negative electrons in atom nuclei seem to be 

 usually associated in pairs. Thus in the ^-disintegra- 

 tions of the radio-active elements two electrons escape 

 in succession. This pairing mav explain the fact that 

 while most atoms have the formula yp^eSn-y with M an 

 even number, extremelv feAv have 'he same formula 

 when M is odd. Thus if p^e should prove to be the 

 Drimary group in atom-building, nevertheless the 

 most abundant group in existing nuclei would be 

 expected to have the formula ip^e) or that of an 

 o-particle. William D. H.arkins. 



University of Chicago, February 4. 



Light and Electrons. 



With reference to Sir Oliver Lodge's letter in 

 Nature of April 7, some few weeks ago I fitted a 

 flat speculum mirror to a centrifuge capable of being 

 run at 150 revolutions per second. The other arrange- 

 ments — not yet completed — are as follows : — The 

 image of a brightly illuminated slit is focussed on the 

 mirror ; a second slit is placed at a distance of about 

 10 metres. 



(i) The eye is placed behind the second slit and the 

 centrifuge increased in speed until the flash is no 

 longer seen. If the slits are i mm. wide a duration 

 of flash of 10- ' sec. is attainable. If necessary, the 

 radius of the rotating beam may be increased, 



(2) The eye is replaced by a photographic plate. 

 This is a test for electrons released from the sensitiser. 

 Below a certain duration of flash there should be no 

 latent image formed, however often the flash is 

 repeated. 



(3) A light-sensitive photo-electric cell is also tested. 

 In this manner Mr. J. H. Poole and I have planned 



to test the very point raised by Sir Oliver Lodge, and 

 also to seek for evidence respecting the quantum 

 theory of vision. 



At present there is only this much to go on. It is 

 stated (Halliburton's "Physiology") that a flash of 

 1-25x10-' sec. duration is still visible. This (if it is 

 the limit) affords a length of 4 x 10* cm. for the length 

 of the train of waves activating an electron in the 

 retina. If it is allowable to go further we find the 

 energv of a single wave (of green light) to be about 

 6x 10- " erg. J. JOLY. 



Trinity College, Dublin, April S. 



Molecular Structure and Energy. 



The question which Prof. Partington raises in his 

 letter under the above title in Nature of April 7, 

 p. 172, is an important one which' I would prefer 

 should be answered by others more qualified to do so 

 than myself. I intervene principally to correct the im- 



NO. 2685, VOL. 107] 



pression given in the letter that the structures of the 

 various molecules with which I have dealt in recent 

 communications have been proposed by me. This is 

 not so. All that I have done is to show that the 

 structures of certain polyatomic molecules, including 

 some halogen gases, carbon dioxide, and nitrous oxide, 

 postulated by Lewis and Langmuir are consistent 

 with viscosity data and X-ray crystal data taken 

 together. The procedure deals with the external 

 shapes of the molecules only, and not with the internal 

 energy of their nuclei ; and it appears to be justified 

 by the calculations of Prof. S. Chapman (Phil. Trans., 

 vol. ccxvi., p. 347), who says: "... the internal 

 energy which prevents the application of our formulae 

 to the conductivity of polyatomic gases hardly affects 

 viscosity." 



Prof. Partington's views appear to be open to 

 criticism even if we leave out of account entirely the 

 necessitv for revising earlier ideas of energy partition 

 on the basis of the quantum theory. For example, 

 Langmuir 's proposed structure for the nitrogen mole- 

 cule is not spherically symmetrical in the same sense 

 as are the atoms of the inert gases. There are two 

 separate massive nuclei instead of one, and this in- 

 volves the possibility of rotational internal ene.rgy of 

 the same type as in the oxygen molecule, so that the 

 ratio of the principal specific heats could not be ex- 

 pected to be so high as 1-667. Also, is it not possible, 

 indeed probable, that the nuclei of all polyatomic 

 molecules are capable of vibration to and fro? Such 

 motions are, I believe, known to exist in the gaseous 

 hydrogen halides, as well as the rotations to which 

 attention has recently been directed by Prof. W. L. 

 Bragg and Mr. H. Bell (Nature, March 24, p. 107). 



A. O. Rankine. 



Imperial College of Science and Technology, 

 April 7. 



The Normal Orbit of the Electron in the Atom of 

 Mercury. 



Researches on ionisation and resonance potentials 

 of mercury vapour and on its ultra-violet absorption 

 in a non-luminous state, together with considerations 

 from the serial type of the mercury spectrum, lead 

 to the definite conclusion that in the absence of excit- 

 ing agencies the spectral electron remains on the orbit 

 iS, the normal orbit of the atom of mercury. On 

 the other hand, R. Dearie has shown the presence 

 of a strong infra-red absorption band at A=ioi40, 

 and this fact has suggested the possibility of a second 

 normal orbit in the mercury atom, namely, the orbit 

 2P. The corresponding ionisation and resonance 

 potentials have, however, never been observed. This 

 problem induced us to make an absorption experiment 

 with non-luminous mercury vapour in the infra-red 

 region, using a photographic method which enabled us 

 easily to reach A =11300 A. All the photographs 

 showed complete absence of a marked absorption 

 at A =10140, although the pressure of mercurv vapour 

 reached i atm. The efficiency of the method having 

 been established, the absence of a strong an^ charac- 

 teristic absorption of A = 10140 by mercury vapour has 

 been shown and the necessity for a second normal orbit 

 is avoided. " A. Terenin. 



Optical Institute, Petrograd, December, 1920. 



Doublets in Spectral Series. 



The physicists of Petrograd have recently become 

 acquainted with a paper by Wood and Mohler {Phil. 

 Mag., April, 1919) on resonance in sodium vapour. 



