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



to be every reason to believe that spectra which thus differed so much from 

 those of any previously known substance were characteristic of this new 

 earth. Judging from the facts then known, it was more probable that spectra 

 of such a new type were due to a new element, than that they were due 

 merely to a combination of two such elements as zirconium and uranium. 

 Some of these chemical and physical facts can now be explained by the 

 presence of uranium ; but besides this and several of the more common 

 earths and oxides, I have detected in some zircons erbium, didymium, 

 yttria, and another substance which exists in such small quantity that I have 

 not yet been able to ascertain whether or no it is the suspected new earth. 

 These accidental constituents do not indeed occur in sufficient quantity to 

 be of importance, except as modifying the physical and optical properties, 

 the didymium giving its usual characteristic absorption-bands (zircons from 

 Sveneroe, Norway), and the manganese the same spectrum as that of 

 garnets (zircons from an unknown locality in Siberia*). This method, 

 however, fails to give evidence of a new earth ; for since the publication of 

 my former paper, I have proved that the very abnormal spectra, which 

 seemed sufficient to establish its existence, are really due to compounds of 

 zirconia with the oxides of uranium, which have such a powerful action on 

 light, that an almost inappreciable amount is sufficient to produce the 

 spectra to great perfection — in fact so small an amount, that the total 

 quantity which misled me was only a few thousandths of a grain ; and its 

 presence might easily have remained unsuspected, if I had not made a 

 number of experiments, which at first did not seem to have much con- 

 nexion with the subject. 



In studying the spectra of crystalline blowpipe-beads, it seemed desirable 

 to examine those made with carbonate of soda, with or without a little 

 borax. This when melted dissolves certain oxides ; and though it crystal- 

 lizes on cooling, so as to be only partially translucent, yet with strong 

 direct sunlight well-marked spectra may be seen. For example, in the 

 oxidizing flame uranic oxide is easily dissolved by carbonate of soda alone ; 

 and when quickly cooled, an orange-coloured bead is obtained, probably 

 containing uranate of soda in a vitreous condition, which gives a single 

 well-marked absorption-band in the green (see spectrum No. 4) with so 

 small a quantity of the oxide ; that in a bead ^ inch in diameter shows the 

 spectrum to the best advantage, and even rojfo L , grain can be easily de- 

 tected. We need not be surprised that this spectrum differs so much 

 from the usual type of uranic salts (Nos. 1, 2, and 3), since in this case 

 the oxide plays the part of an acid. It may be only an accidental coin- 

 cidence, but this difference is analogous to the change which commonly 

 occurs on adding an alkali to neutral solutions of vegetable coloursf. When 

 gently reheated it seems as if the uranate passed into a crystalline state, 

 for the spectrum then shows four absorption-bands, and is more like the 



* For both of these I am indebted to my kind friend Mr. David Forbes, 

 t See my paper in Proc. Roy. Soc. 1867, vol. xv. p. 433. 



