TEANSACTIONS OF SECTION B. 571 



•with this not unimportant difterence — the citron band is as prominent as any 

 other line. Ya consists, therefore, of samaria with the greenish bkie of yttria and 

 some of the other yttria bands added to it. 



I may aptly call the Ya spectrum my Rosetta stone. It threw a flood of light 

 on all the obscurities and contradictions I had found so plentiful, and showed me 

 that a much wider law than the one I had been working upon was the true law 

 governing the occurrence of these obscure phenomena. For what does the spectrum 

 of Ya show ? It proves that what I had hitherto thought was one of the chief 

 bands in the yttria spectrum — -the citron baud — could be entirely removed, whilst 

 another characteristic group — the double green of yttria — could also be separated 

 from the citron. 



It would exceed legitimate limits were I to enter into details respecting the 

 chemical and physical reasons which led me to these definite conclusions. To 

 settle one single point more than 2,000 fractionations have been performed. 



The meaning of the strongly marked symbolic lines had first to be ascertained. 

 For a long time I had to be content with roughly translating one gi'oup of coloured 

 symbols as ' yttrium ' and another group as ' samarium,' disregarding the fainter 

 lines, shadows, and wings frequently common to both. Constant practice 

 in the decipherment has now given me fuller insight into what I may call the 

 grammar of these hieroglyphic inscriptions. Every line and shadow of a line, each 

 faint wing attached to a strong band, and every variation in intensity of the shadows 

 and wings among themselves, now has a definite meaning which can be translated 

 into the common symbolism of chemistry. 



In a mineral containing the rarer earths those most widely separated in chemi- 

 cal properties are most easily obtained in a state of comparative purity by simple 

 chemical means. For instance, in separating didymium from lanthanum, or sama- 

 rium fi-om yttrium, a few simple chemical reactions and a little waste will give 

 these bodies in a state of purity ; but when it comes to splitting up yttrium into 

 its components ordinary- chemical separation is useless, and fractionation must be 

 pushed to the utmost limit, many thousand operations and enormous waste of 

 material being necessarj' to effect even a partial separation. 



Returning, therefore, after this explanatory digression, to the idea of heavy and 

 light atoms, we see how well this hypothesis accords with the new facts here 

 brought to light. From every chemical point of view the stable molecular group,, 

 yttrium, behaves as an element. Excessive and systematic fractionation has acted 

 the part of a chemical 'sorting Demon,' distributing the atoms ofyttriuminto several 

 groups, with certainly different phosphorescent spectra, and presumably diflerent 

 atomic weights, though all these groups behave alike from the usual chemical 

 point of view. Here, then, is one of the elements the spectrum of which does not 

 emanate equally from all its atoms, but some atoms furnish some, other atoms 

 others, of the lines and bands of the compound spectrum of the element. And as 

 this is the case with one element, it is probably so in a greater or less degree with 

 all. Hence the atoms of this element difler probably in weight, certainly in the 

 internal motions they undergo. 



Another important inference which may be drawn from the facts is that the^ 

 atoms of which yttrium consists, though differing, do not differ continuously^ 

 but per saltum. We have evidence of this in the fact that the spectroscopic bands 

 characteristic of each group are distinct from those of the other groups, and do not 

 pass gradually into them. "We must accordingly expect, in the present state of 

 science, that this is probably the case with the other elements. And the atoms of 

 a chemical element being known to difler in one respect may differ in other respects, 

 and presumably do somewhat differ in mass. 



Restricted by limited time and means, even a partial separation of these atomic 

 groupings is possible to me only with enormous ditticulty. Have we any evidence 

 that Nature has eflected such a separation ? The following facts I think supply 

 this evidence. 



The earth yttria occurs in several minerals, all extremely rare. These minerals 

 are of very diverse chemical composition, and occur in localities widely separated, 

 geographically. Does the pure yttria impure in respect to every other known 



