November 26, 19 14] 



NATURE 



351 



elements is another difficulty; the variation seems to 

 be determined by some reciprocal relationship between 

 the interacting elements, valency being a dependent 

 variable and not an absolute property. Fresh signifi- 

 cance has lately been given to the problems of valency 

 by the conception, due to Barlow and Pope, that it is 

 10 be regarded as a function of the volume occupied 

 by the atom. They have in this waj- succeeded to a 

 remarkable extent in correlating crjstalline form with 

 molecular structure. 



Prof. Hicks dealt with the subject from the spectro- 

 scopic side, i.e. from the consideration of an atom as 

 a configuration capable of emitting definite sets of free 

 vibrations. He reviewed the evidence as to the posi- 

 tion of helium in the order of the elements. It is 

 assumed by many physicists, including Rutherford, 

 Bohr, and Moseley, that H and He are consecutive 

 elements, having nuclear charges of one and two 

 units respectively. There are several reasons, how- 

 ever, for supposing that there are places for at least 

 two intermediate elements between them ; in particu- 

 lar, the theory of Rydberg requires that the atomic 

 number of helium should be 4. In addition to the 

 electrons and positively-charged nuclei, we have to 

 recognise the presence in certain substances of mag- 

 netic doublets the strengths of which are multiples of 

 a definite unit, called by Weiss the magneton ; whether 

 the magneton has an independent existence or is a 

 consequence of electronic motion is an open question. 

 The formation of spectral series is a verv difficult 

 question; neither Thomson's nor Rutherford's model 

 of the atom has shown any aptitude for explaining 

 spectra, and it would seem that the actual structure 

 must be something much more complicated. There 

 is now a general consensus of evidence, theoretical 

 and experimental, that in the case ol_geries the cause 

 of spectral emission is a change in a few atoms (at 

 any one moment) from one configuration to another, 

 with sudden emission of energy. The contrarv view, 

 that it is due to small internal vibrations in all the 

 atoms, presents too great difficulties. It does not 

 follow, however, that the w'hole of the spectral lines 

 are of the character of the series lines, and Nicholson's 

 theory of the coronium and nebulium spectra points to 

 the existence, under the special conditions in the 

 corona and nebula?, of emission of a different type. 

 Prof. Hicks also reviewed Bohr's theory of spectra, 

 and, whilst admiring its suggestiveness, pointed out 

 the arbitrary character of some of the assumptions, 

 and its limitation to series of the Balmer t>pe. 



Mr. H. G.-J. Moseley explained the results of his 

 classification of the elements by their X-ray spectra. 

 The principal frequency v is given approximatelv by 

 the formula yv = constant x (N — B) where B is a con- 

 stant and N an integer increasing by a unit as we pass 

 from element to element up the periodic table. The 

 order of the elements determined by N is nearlv that 

 of increasing atomic weight ; there are one or two 

 exceptions, and in such cases the order given by N 

 is evidently the correct order corresponding to chemical 

 properties. For example, the atomic weight gives 

 the order CI. K, Ar, whereas the X-rav frequencv 

 gives the order CI, Ar, K. The latter is the order 

 required by the periodic table. 



Prof. Nicholson defended the main principles of 

 Bohr's theory. Its most striking success is that it 

 gives with great accuracy two fundamental constants 

 — the universal constant of spectra and (as Prof. 

 Fowler has shown) the mass of the electron in terms 

 of the hydrogen atom. But analysis shows that we 

 cannot obtain other spectra from it, e.g. the 

 ordinary helium spectrum, without abandoning at 

 least one of Bohr's premises, which is vital to the 

 deduction of the hydrogen formula. Referring to the 



NO. 2352, VOL. 94] 



elements postulated by him in the corona and nebulae, 

 Prof. Nicholson was prepared to admit that they 

 might not be chemical elements in the ordinary sense ; 

 they might rather be the bases of elements. The 

 atomic weights of these elements have been measured 

 by Buisson and Fabr)' by an ingenious method and 

 agree with those calculated theoreticalh- from the 

 spectra. Nuclear structure (as distinct from the 

 resultant charge) appears to play a considerable part 

 in series spectra, and, in opposition to Rutherford's 

 view, it seems that the nucleus of a hydrogen atom 

 must be something more complex than the positive 

 electron. 



Prof. Bassett discussed the bearing of the periodic 

 law on the number of the elements. The periodicity' 

 is not of the simple character at first supposed. There 

 are two short f>eriods of 8 elements followed by two 

 long periods of (2x8)4-2 = 18 elements. These are 

 followed by much longer periods, possibly of 

 (2x18) — 2 = 38 elements. It is tempting to suggest 

 that hydrogen represents a period of 3 elements, 

 because then the short periods would be represented 

 b\' (2x3) + 2 = 8, following the same rule. 



Prof. Kerr Grant summarised the difficult}- as to the 

 stability of a system consisting of one nucleus and 

 one electron. It was difficult, too, to account for the 

 non-magnetic character of the hydrogen atom with 

 this structure. 



The discussion occupied the whole morning, and was 

 followed with keen interest. So many different points 

 of attack on the problem were, represented that it 

 could scarcely be expected that any one aspect would 

 be argued out to a conclusion by the different speakers, 

 but, regarded as a symposium, the discussion provided 

 a most useful survey of the present state of our 

 knowledge. A full report is to be printed in the 

 annual volume of the British Association. 



The section met again in the afternoon, when Prof. 

 E. Goldstein read a paper on salts coloured by the 

 kathode rays. When the rays fall on certain salts 

 coloration is produced immediately. These colours 

 disappear on prolonged exposure to daylight or when 

 the salt is heated; the rate of disappearance varies 

 greatlv for different substances. The characteristic 

 colour and the behaviour on heating afford a means 

 of identifying the substance, and impurities present 

 in ver}' minute quantities can often be discovered in 

 this way, for a very small admixture of the salt 

 suffices. It was at first supposed that the phenomenon 

 was due mainly to a chemical reduction ; Giesel and 

 Kreutz obtained analogous colours by exposure of the 

 salts to the vapour of sodium and potassium. But 

 there are important differences between the Giesel 

 colours and those here considered. For example, the 

 former are much more stable when exposed to day- 

 light. Prof. Goldstein showed that stable coloration 

 identical with that obtained by Giesel can indeed be 

 produced by the kathode rays, but as the result of 

 prolonged bombardment in which the salt becomes 

 heated strongly. He calls these "after-colours of the 

 second class " to distinguish them from those of the 

 " first class," which appear immediately on exposure. 

 The colours of the first class depend not only on the 

 metal, but on the acid constituent ; indeed, they can 

 be produced in some organic salts which contain no 

 metallic base. There can thus be no question of 

 chemical decomposition ; but apparentiv the matter 

 passes into some changed — possibly pohinerised — state 

 which deser\-es a careful study. Amongst other things 

 the new state is characterised bv greatlv increased 

 absorption of light. Connected with the loss of 

 coloration by exposure to daylight there is generally 

 a phosphorescence, which follows the exposure. It is 

 possible to produce the same colours by ultra-violet 



