THE EVOLUTION OF THE CHEMICAL ELEMENTS. 1 87 



the less soluble, it appears that the apparently 

 elemental yttrium may be split up into several 

 closely related substances, which, though in some 

 cases their chemical properties may be indistinguish- 

 able, yet show marked differences in their spectra. 

 And so, instead of a single metal, yttrium, with five 

 bright lines in its spectrum, we get five substances 

 with one line each in their spectrum. Similar 

 results have been obtained with didymium and 

 other metals, and quite lately (1889), even such 

 common and apparently well-known metals as 

 cobalt and nickel have been found to be constantly 

 alloyed with a third substance, and the multiplica- 

 tion of such results seems simply a question of time. 

 § II. Now, says Mr. Crookes, what are we to 

 make of these facts ? Are we to give up our tests 

 as worthless, or are we to dub all these me7nbra 

 disjecta of an element elements ? To do this we 

 should require some graduation of the conception 

 of elementicity, which would dispense us from 

 putting the constituents of yttrium and didymium 

 on a par with oxygen and carbon with respect to 

 their elementicity. But Mr. Crookes propounds 

 another interpretation, which may startle old- 

 fashioned chemists, but has the merit of being both 

 sensible and philosophic. It is a mere prejudice, 

 he says, to regard a thing as an element, because it 

 has resisted all our reagents and all our tests : for 

 each test can only cleave it in two, can only divide 

 a compound into two portions, which are elements 

 as fa7^ as that test is concerned. But if a new test is 

 applied, the supposed element splits up with perfect 

 ease. All that can be inferred from our "elements" 

 is that the tests which would subdivide them further 

 have not yet been discovered. And these experi- 



