VOA/C CONTEMPOK.Ih'y ADr.lXCf.S IX rUVSlCS-ll- 491 



imli'ss wo nvt tlu- c|iianlil\ .V r(iual, not to tlic onliT-mimlur nf tin- 

 tU'iiKMit ii) tlio procession, l>ut to tin- order-nunilRT minus one. 



Dot's this mean that the micKMr charge of the Hth elemciii in tlu- 

 |HTi(Hlir tal)le is («— Df Inr all the values of « exceeding 11 (.the 

 \alues for which Mosi'le> 's law holds)? I fear this could not i^e 

 contradicted from the direct exjx-ri mental evidence, for ("hadwick's 

 values of the nuclear charges for the elements h — 78, 47, 21) fall just 

 short of iR'ing exact enougii to prove that they are 78<', 47e, 29e instead 

 of 77f, -Hie, 28f, respectively. However, we should do too much 

 violence to the Ix^auty of the principle if we admitted that there arc 

 only eight values lietween 2e and llf to be distributed among the 

 nuclear charges of the nine elements l)etween helium and magnesium, 

 atui happily it is not necessary, for the apparent discordance can 

 plausibK- 1h> blamed upon too simpe a view of the internal economy 

 of the atom which we took in deriving equation (1). Instead of 

 assimiing that the deepest-lying electron of the atom revolves in an 

 otherwise vacant space surrounding the nucleus, witlc enough to 

 contain the first two of its permissible orbits, we should do bicttcr 

 to assume that there are several deep-lying electrons similarly placed 

 and interacting with one another, or at least that there is no single 

 deepest-lying electron too far inw-ard to be affected b\' the others. 

 The effect of thus changing the assumption is to change the calculated 

 value of the /i a- frequency, for an atom of nuclear change Ne, from 

 the value (1) to a value ^R (\ — k)-; in which k depends on the par- 

 ticular configuration assumed for the internal electrons. It is clear, 

 therefore, that we are in no wise compelled by Moseley's law to 

 conclude that the nuclear change of the atom of the «th element is 

 {n—l)e when «>11, and may continue to accept the much more 

 satisfying principle that the nuclear change of the nth element is we." 



Before stating the conclusion let me restate the evidence in a 

 briefer form and an altered order. Originally the elements were 

 arranged in the order of their combining weights. It was seen that 

 when they are arranged in this way, there is a periodic variation of 

 the ensemble of chemical and physical properties frf)m clement to 

 element. But to make the perio<lic variation ([uite smooth and 

 imbroken, it was found necessary to violate the order of the com- 

 i)ining weights at several places in the series; three pairs of con- 



" The agreement with cxperinuTit inilced Ijccomcs very gfKxl, .it least over a 

 rcrtain range of elements, if we assume that there are tiormally .? electrons in the 

 innermost or one-quantum ring and nine in the second or two-(|uantum ring orbit 

 <}. KrfX)i. But nofMxIy wants to accept this particular repartition of electrons, 

 and it is customary to assume that the inner orbits are mostly elliptical. But it 

 would lie gratifying to attain a (|uantitatively successful theory. 



