so-called Meta-Elements. 113 



very often quite different from the spectra of the elements 

 composing them. It may be thought that two earths of such 

 similar nature as yttria and samaria are scarcely likely to 

 form a compound together ; but yet how else can it be ex- 

 plained ? I think if we examine the facts of the case more 

 closely, we shall see that it is not improbable that yttria and 

 samaria should form a compound together. It is well known 

 with what ease the earths of the aluminium group form com- 

 plex compounds with silica and lime, and may not part at any 

 rate of the minerals, as samarskite and gadolinite, from which 

 these rare earths are obtained, be considered to be in a similar 

 manner complex compounds of the earths ? 



Crookes has therefore shown that it is possible, by combining 

 two earths together, to obtain a spectrum with a bright line in 

 it which exists in neither of the spectra of the two earths in 

 question : this line has in fact been artificially produced. 

 This being the case, why should it not be possible to manu- 

 facture, if we may use the term, other lines and other spectra 

 by the combination of various earths together in varying pro- 

 portions ? Is it not possible that the various spectra of the 

 earths G«, G$, G n , &c. are nothing more than would be pro- 

 duced if we combined together in various proportions the two, 

 or at most three, earths of which these bodies might be sup- 

 posed to be really constituted ? Do not the experiments of 

 Crookes support this supposition ? Thus, about the only pro- 

 perty in which these new earths, G a , G«j, &c, were found to 

 differ, except as regards their radiant-matter spectra, was that 

 when they were ignited with sulphuric acid and varying pro- 

 portions of other earths, as lime and barium oxide, they gave 

 phosphorescent spectra in this case also, differing according 

 to the nature and quantity of the base used. 



Does not this serve to prove still further the unreliability 

 of the radiant-matter test? If spectra are formed by the 

 union of one of these new earths with lime, say, which differ 

 materially from the spectra of either the earths or the lime, 

 does not this show that no reliance at all can be placed in such 

 spectra, especially with regard to proving that these supposed 

 new earths are really the oxides of new elements. For just 

 as a rare earth mixed with lime gives a spectrum different 

 from that either of the earth or of the lime, so it is just as 

 probable that the earth samaria combined with the earth 

 yttria should give a spectrum different from that either of 

 samaria or of yttria, and also that the spectrum should vary 

 according to the proportions of these two earths present ; and 

 it must be remembered that the earth yttria, which has been 

 fractionated into these new earths, always contains a small 



