December 2, 1920] 



NATURE 



457 



potential diflference very close to zero. With retard- 

 ing fields the currents fall off rapidly as the applied 

 potential difference increases. Like similar thermionic 

 electron currents, they approach the voltage axis 

 gradually and not sharply, as in the photo-electric 

 case. Ihe true zero on the voltage scale is difficult 

 to determine on account of fluctuations in the contact 

 potential difference. In the case of COCl, it has 

 been possible to locate the zero to within o-io volt 

 by a photo-electric method. The proportion of the 

 chemically emitted electrons the kinetic energy of 

 which lies between u and a+du is very closely repre- 

 sented by 



1^ 



where fe is Boltzmann's constant and T is a certain 

 temperature. For the case of COCl,, T is near 

 3300° K, and for the case of Cl„ T is about 4900° K. 

 The formula above represents a Maxwell distribution 

 for the temperature T. Thus the distribution of 

 kinetic energy among the chemically emitted electrons 

 IS the same as that among the molecules of a gas at 

 the uniform temperature T.— Dr. A. E. Oxiey : Mag- 

 netism and atomic structure. This communication 

 is an extension of former papers on "The Influence 

 of Molecular Constitution and Temperature on Mag- 

 netic Susceptibility" (Phil. Trans. Roy. Soc. 

 vol. ccxiv., .A, 1914; vol. ccxv., A, 1915; and 

 vol. ccxx.. A, 1920). From Tyndall's work and recent 

 experiments of the author on the characteristic deport- 

 ment of diamagnetic and paramagnetic crystals in the 

 magnetic field, it appears that in non-ionised crvstal 

 structures the fundamental unit of the space lattice is 

 the molecule. It is shown that the electron orbits in 

 atoms must be distributed in space round the nucleus, 

 each electron describing a small orbit, or alternatively 

 the electron itself may be a complex unit endowed 

 with magnetic properties. In either case the distribu- 

 tion must be such that the aggregate projected area 

 of the electron orbits on a plane perpendicular to the 

 principal cleavace is a maximum in both diamagnetic 

 and paramagnetic crystals. This result is consistent 

 with a closer packiner of the molecules in a direction 

 parallel to the principal cleavage. In crvstals of the 

 simple cubic form X-ray analysis has indicated that 

 the structure is an ionised-atomic one, and the 

 cleavages are all of equal value. .Such crystals show 

 no appreciable structural deportment in the magnetic 

 fieW. The above views relating to electron distribu- 

 tion arc consistent with the cubical atom theorv of 

 Lewis and I-ingmuir. but not with Bohr's theorv. 

 The coupling forces between atoms and molecules in 

 non-ionised crvstals are due to the mutual magnetic 

 induction between pairs of electron orbits. .A model 

 of the hvdropen molecule is piven. in which the 

 arrangement of the coupling units determines a dia- 

 mapnefic molecule as required bv experiment. It is 

 considered that the above views' and those of Bohr 

 may evenfuallv be broujjht into line by a fuller 

 rerocnition of the possible differences between 

 r.idinting and non-radiating matter. -Prof. \. O. 

 Raaklnc : The proximity of atoms in paseous mole, 

 rules. In this invcstir'ation a closn examlnition is 

 made of the relations b/>fween th»> e«»imate« of atomic 

 diameters obtaine<l bv Prof. W. I,. Bragg from X-rnv 

 crvstal measurements and those deduced from the 

 kinetic theorv of fases. The examination is carried 

 out from the point of view of the I.ewis-I.anumiiir 

 moleoihr theorv. It is shown that if, for example, 

 a hvpothetir.nl molecule be ronsfrurfed of two arpop 

 atoms with th'-ir centres separated bv the di-t'anre de- 

 mand-vt bv Prof. W. I.. B^aff's figures, the beh.nviour 

 f such mol<TiiIes in thermal .Tpilation would be 

 'lost identir-.! with the ncfiiil behaviour of chlorine 

 NO. 2666. VOT,. Io61 



molecules. Similar relations are shown to exist for 

 the pairs ol gases neon-oxygen, krypton-bromine, 

 and xenon-iodme. The toliu>vmg cuuv-iusions un- 

 regarded as justified : (a) There is substantial qua»\.- 

 titalive agreement between the esiimaies ol aiumic 

 dimensions deduced from X-ray crystal measurements 

 and from the kinetic theory o! gases, (b) In size and 

 shape the atoms of the monatomic inert elements are 

 nearly indistinguishable from the atoms respectively 

 of the neighbouring diatomic eleirenls in the periodic 

 table. (c) The Lewis-Langmuir molecular theory 

 accounts satisfactorily for the kinetic behaviour of 

 the molecules of oxygen, chlorine, bromine, and 

 iodine in relation to the behaviour of the correspond- 

 ing inert atoms neon, argon, krypton, and xenon. — 

 Prof. A. O. Rankine : The similarity between carbon 

 dioxide and nitrous oxide. The two gases in ques- 

 tion have been shown by Langmuir to have almost 

 identical physical properties. In particular, they have 

 the same viscosity, and the application of modern 

 kinetic theory indicates that their molecules have the 

 same size and shape. In the present paper it is 

 shown, by the extension of methods already described 

 by the author, that the kinetic behaviour of the mole- 

 cules of both CO2 and N,0 is consistent with their 

 being identical in size and shape with three neon 

 atoms in line and contiguous, i.e. with outer electron 

 shells touching. This is in accordance with Lang- 

 muir's view of the constitution of these molecules. — 

 Dr. .\. M. Wllllanw : Forces in surface films. 

 I. : Theoretical considerations. II. : Experimental 

 observations and calculations. III. : The charge on 

 colloids. I. and II. : .Attention is directed to the 

 effects of (i) accessibility of surface and (ii) adsorption 

 on the apparent specific volume of finely divided solids. 

 A simple theory of these effects is developed, with 

 which observations are in agreement. The true 

 specific volume of a specimen of charcoal, which 

 appeared to be 051 in water and 046 in chloroform, 

 was evaluated as 067 c.c. per gm. The attractive 

 pressure on the surface film on the charcoal was 

 calculated and found to be of the order of 10.000 atmo- 

 spheres, while the internal pressure of the charcoal 

 itself was evaluated as of the order of 50.000 atmo- 

 spheres. III. : It is shown that compressive 

 forces of the order previously determined mav give 

 rise in the adsorption l;iyer to a diffusion potential 

 difference of the magnitude observed in the case of 

 susoensoids. The influence of the diffusion of 

 hvdrogen- and hvdroxyl-ions on the potrnHal differ- 

 ence is emphasised, and the neutralis.ition of the 

 charge on susoensoids and thrMr consequent precipita- 

 tion explained in terms of diffusion potential. 



Phyilcal Society, November I2.--Sir W. H. Bragg, 

 president, in the chair. — Dr. F. H. Ooocher ; lonisa- 

 tion and excitation of radiation by electron impact in 

 helium. Measurements have been made of the critical 

 potentials for helium by the method used in the ex- 

 periments of Davis and Ooucher, these biding com- 

 pared with the ionising potential of mercury vapour 

 taken as a standard. Assuming the ionising potential 

 of mercury to be 10-4 volts, two critical potentials 

 occur in helium, one at about 20 volts and the other 

 at about 26 volts. These critical values agree well 

 with those obtained by Norton and Davies. The 

 effect of radiation alone on the metal parts of the 

 apparatus was studied under conditions which woiiM 

 yieW evidence of use in the . interpretation of the 

 results obtained when the production of both ionlsa- 

 tion and r.Tdiation was takinp place simultanemisly. 

 The conclusion was <lrawn tlvf the lower critirnl 

 fwtential was a radiation potential, thonph some 

 ionisation was produced also at this potential. This, 

 however, was attributed to the presence of impurity, 



