204 Mr. H. C. Sorby on some Spectra of Compounds [Feb. 10, 



deposited, when it must be allowed to cool rapidly. It should then be 

 transparent with scattered crystals, and the uranium all in the state of 

 protoxide. On gently reheating it, the bead ought to suddenly turn white 

 and almost opaque ; and care must be taken not to heat it any more than is 

 just requisite to cause the borate to crystallize out, or else the uranium will 

 rapidly pass into the state of peroxide. Such beads must be examined by 

 strong direct light from the sun, or from a lamp of very great brilliancy, 

 condensed on them by means of an almost hemispherical lens of about h 

 inch focal length ; and in addition to the means described in my former 

 paper, I have since found it very convenient to place them over a hole in a 

 black card, so as to entirely prevent the passage of any light which has not 

 penetrated through them, even when so arranged in the focus of the micro- 

 scope that the spectrum of their thin edges may be examined, if the centre 

 be too thick and opaque. If thus properly prepared, the presence of more 

 or less uranium will be shown by the greater or less intensity of the absorp- 

 tion-bands of the spectrum described and shown in fig. 1 of my former 

 paper. This test is so delicate that there is no difficulty in seeing the 

 darker band in the green in the case of zircons which contain no more 

 than -fa per cent, of uranic oxide ; and I find that very few localities yield 

 this mineral so free from it that it cannot be easily detected. Those from 

 Miask, Siberia, are the only specimens in which I have not been able to 

 recognize it. The jargons from Ceylon contain an amount varying up to 

 about 1 per cent., although in no published analysis that I have seen is 

 there any allusion to the presence of even a trace. It has also been over- 

 looked in several other cases ; and it now becomes important, because it gives 

 rise to various well-defined spectra, which are so characteristic of the 

 different minerals, that they can be very conveniently identified, even when 

 cut and mounted as jewels, by means of the number and position of the 

 absorption-bands, as I intend to explain in a paper on the spectra of 

 minerals. 



On flaming the bead at a moderate red heat, the protoxide passes into 

 the peroxide, and the spectrum No. 6, given above, may be seen, if suffi- 

 cient oxide be present, but considerably more is required than in the case 

 of the protoxide. I may here say that the examination of better prepara- 

 tions has enabled me to detect another distinct band in the extreme red, 

 not shown in fig. 2 of my former paper, and also an additional faint band 

 in the blue, not shown in fig. 1. 



In applying this test to detect minute quantities of uranium in other 

 minerals, it is requisite to bear in mind that zirconia may play the part of 

 both an acid and a base, and that various oxides and acids so combine with 

 the zirconia or with the oxides of uranium as to prevent the formation of 

 the compounds which give rise to the characteristic spectra. The zir- 

 conia appears to combine with some rather than with the uranous oxide, 

 and with others rather than with the uranic, so that, if one spectrum 

 cannot be obtained, the other may ; and there are few, if any, cases when 



