46 Mr. William Croolces [Feb. 18, 



supposed. Between tlie molecules we are accustomed to deal with in 

 chemical reactions and ultimate atoms as first created, come smaller 

 molecules or aggregates of physical atoms ; these sub-molecules difier 

 one from the other, according to the position tbey occupied in the 

 yttrium edifice. 



Perhaps this hypothesis can be simplified if we imagine yttrium 

 to be represented by a five-shilling piece. By chemical fractionation 

 I have divided it into five separate shillings, and find that these 

 shillings are not counterparts, but like the carbon atoms in the benzol 

 ring, have the impress of their position, 1, 2, 3, 4, 5, stamped on them. 

 These are the analogues of my Ga, G/3, &c. If I now bring in a 

 much more powerful and searching agent — if I throw my shillings 

 into the melting-pot or dissolve them chemically — the mint stamp 

 disappears and they all turn out to be silver. I submit my yttrium, 

 or my Ga, G(3, &c., to the intense heat of the electric spark, the little 

 differences of molecular arrangement vanish, and the atoms of which 

 the molecules of yttrium, Ga and G(3, are alike composed, reveal 

 their presence in identical sj^ectra. 



An alternative theory commends itself to chemists, to the effect 

 that the nine bodies shown in the above table (Table I.), are new 

 chemical elements differing from yttrium and samarium in basic 

 powers and several other chemical and physical properties, but not 

 sufficiently to enable us to effect any but a slight separation. One of 

 these bodies, G8, gives the phosphorescent citron line, and also the 

 brilliant electric spectrum I have just exhibited. The other eight do 

 not give electric spectra which can be recognised in the presence of a 

 small quantity of G8, whilst the electric spectrum of G8 is so sensi- 

 tive that it shines out in undiminished brilliancy even when the 

 quantity present is extremely minute. In the process of fractionation, 

 Ga, G/3, GB, &c., are spread out and more or less separated from one 

 another, yet the separation is imperfect at the best, and at any part 

 there is enough G8 to reveal its presence by the sensitive electric 

 spark test. The arguments in favour of each theory are strong and 

 pretty evenly balanced. The compound molecule explanation is a 

 good working hypothesis, which I think may account for the facts, 

 while it does not j^ostulate the rather heroic alternative of calling 

 into existence eight or nine new elements to explain the phenomena. 

 However, I submit it only as an hypothesis. If further research 

 shows the new element theory is more reasonable, I shall be the first 

 person to accept it.* 



* Neither of these theories agrees with that of my distinguished friend M. 

 Lecocq de Boisbaudran, who also has worked on Ihese earths for some time. He 

 considers that what I have called old yttrium is a true element, characterised by 

 the spark spectrum already exhibited, but not giving a phosphorescent spectrum 

 in vacuo. The bodies giving the phosphorescent spectra he considers to be 

 impurities in yttrium. These he says are two in number, and he has pro- 

 visionally named them Za and Z/3. By a method of his own, differing from 

 mine, M. de Boisbaudran obtains fluorescent spectra of these bodies; but their 



