456 Royal Society : — Mr. H. C. Sorby on some Spectra 



only one part of the crystal, probably depended upon the presence of 

 some substance which either reduced the uranic oxide or prevented 

 the oxidization of the uranous. 



These facts thus clearly show that the various spectra which seemed 

 to indicate the presence of a new element existing in three different 

 physical conditions, are in reality only characteristic of the two oxides 

 of uranium combined with zirconia, or not in combination. Perhaps 

 some may think that my having been thus led astray shows that little 

 or no reliance can be placed on the method of investigation employed; 

 but I contend that the mistake was due to its being such an unex- 

 pectedly delicate test for uranium ; and, as explained above, the error 

 was ultimately corrected by a further development of the same 

 method. As far as the interests of science are concerned, there is 

 no need to regret the general result. We have lost what appeared 

 to be good evidence of a new earth, but have gained an almost en- 

 tirely new system of blowpipe testing, which enables us to detect 

 such a minute quantity of some substances as could not be recog- 

 nized by the ordinary means. I shall not now attempt to give any 

 thing like a full account of this subject, since it would be much better 

 to let it form part of a paper on various improvements in blowpipe 

 chemistry, but will merely mention a few facts which have a special 

 bearing on the question before us. 



In the first place, I would say that zirconia and the oxides of 

 uranium are most useful reagents in detecting the presence of certain 

 substances with which they unite to form compounds having very 

 special characters. The most striking of these are : — the compounds 

 already described, which are distinguished by the spectra, and not 

 by any well-marked colour ; the compound of eerie oxide with uranic 

 oxide, which is of a splendid deep blue colour, but shows no absorp- 

 tion-bands ; and that of yttria with uranic oxide, which is charac- 

 terized by a deep orange-colour and extreme fusibility. Thorina and 

 oxide of lanthanum form with uranous oxide compounds which 

 give spectra with absorption-bands in special positions, but of the 

 usual type, and not of such a marked character as to be useful in 

 detecting minute quantities of those substances in mixtures. 



In order to see the spectra of the zirconium-uranium compounds, 

 it is requisite that both elements should be combined in a crystalline 

 condition. When both constituents are melted in borax and are 

 held in solution, or if when crystals are deposited any other sub- 

 stance replaces either the zirconia or the oxides of uranium, the 

 characteristic spectra cannot be seen. The most simple application 

 of this test for uranium is in the case of various zircons. As much of 

 the powdered mineral as will dissolve should be melted with borax in 

 a circular loop of platinum wire about ^ inch in diameter, so as to 

 give a bead of moderate thickness. A little boric acid should then 

 be added, which not only tends to keep the uranium in the state of 

 protoxide, but also facilitates the crystallization of tbe borate of zir- 

 conia, which is far less soluble when there is excess of boric acid. 

 The bead should then be kept at a bright red heat, just within the 

 deoxidizing flame, until so much borax has been volatilized that small 



