646 SCIENTIFIC RECORD FOR 1885. 



greater part of the chemical work undertaken in the hunt for the spec- • 

 trum-forming element was performed upon material which later knowl- . 

 edge showed did not contain sufficient to respond to any known chem- 

 ical test. Except in few instances, as water analysis and the detection 

 of poisons, chemistry takes little account of traces, and when an anal- 

 ysis adds up to 99*99 the odd O'Ol per cent, is conveniently put down 

 to "impurities," "loss," or "errors of analysis." When however the 

 99*99 per cent, constitutes the impurity and this exiguous error 0*01 is 

 the precious material to be extracted, and when, moreover, its chemistry 

 is absolutely unknown, the difficulties of the problem become enor- 

 mously enhanced. The author was therefore obliged to build up a new 

 chemistry, and after six months' work he obtained the earth didymia in 

 a state which most chemists would call absolutely pure, for it i^robably 

 contained not more than one part of impurity in 500,000 of didymia. 

 But this one part in half a million, profoundly affected the character of 

 didymia from a radiant-matter-spectroscopic point of view, and the per- 

 sistence of this very minute quantity of interfering impurity entailed 

 another six months' extra labor to eliminate these llnal traces, and to 

 ascertain the real reaction of didymia pure and simple. The earth form- 

 erly called didymia proved to be a mixture of didymia and samaria. 

 During the long process of purification the 1,000 grams dwindled away 

 bit by bit until less than one-half a gram remained of the pure material. 

 Didymia thus purified shows no trace of the orange double band, which 

 is characteristic of samarium, an element discovered in 1879 by Lecoq 

 de Boisbaudran. 



The earths ceria, lanthania, didymia, and samaria possess in their 

 purest state the following characters: Ceric oxide is almost pure white; 

 in strong solution it has no absorption spectrum. The atomic weight 

 of the metal was taken and found to be 141*1. Ceria gives no orange 

 band spectrum in the radiant-matter tube. Lanthana is snow white, 

 and the metal was found to have an atomic weight of 138*3. It gives 

 no orange band spectrum when absolutely free from samaria. Samaria 

 is white, with the faintest possible tinge of yellow; its absorption spec- 

 trum is much more feeble than the spectrum of didymium. Pure 

 samaric sulphate by itself gives a very feeble phosphorescent spectrum. 

 When however the samaria is mixed with lime before examination in 

 the radiant-matter tube, the spectrum is very beautiful, consisting es- 

 sentially of three bright bands — red, orange, and green — these being 

 nearly equidistant, and the orange being the brightest. With a nar- 

 rower slit the orange and green bands are seen to be double, with 

 faint wings. 



The spectrum of samaria becomes highly modified by mixing with 

 the earth other metallic oxides. Mr. Crookes divides these modified 

 spectra into three groups; for particulars of each, with diagrams of 

 each type, we refer to the original paper. In a mixture of samaria and 

 yttria the former possesses a remarkable power of obliterating the spec- 



