Mr. D. Forbes on Evansite, a new Mineral Species. 343 



an innumerable series of minute cracks; did not fuse in the 

 strongest heat ; appeared to colour the outer flame bluish green, 

 but so feebly as to be all but indistinct. On moistening the 

 mineral with sulphuric acid, this reaction was rendered rather 

 more apparent. On charcoal it proved infusible and unaltered, 

 in both oxidizing and deoxidizing flames ; but when heated, after 

 moistening it with a solution of nitrate of cobalt, an intense blue 

 colour was communicated to the assay. It dissolved readily 

 both in borax-glass and phosphateof soda* in the oxidating flame, 

 forming colourless glasses, which remain colourless on cooling : 

 some of the faint-yellow-coloured specimens give a very light- 

 coloured yellow glass when hot, but, on cooling, become colour- 

 less, a reaction due to the presence of iron. In the reducing- 

 flame both these fluxes give the same reactions. In a few cases 

 the glass formed by phosphate of soda shows a trace apparently 

 of silica floating in the clear glass bead. 



A qualitative chemical examination showed the mineral to be 

 completely soluble in sulphuric, nitric, and hydrochloric acids. 

 The solution, when treated by a stream of sulphuretted hydrogen 

 gas passing through it, gave no precipitate whatever. The acid 

 solution gave a yellow precipitate, indicative of phosphoric acid, 

 when treated with molybdate of ammonia ; and further, alumina 

 and a trace of oxide of iron were found, but no lime, glucina or 

 zirconia, which were specially tested for. 



Fluorine was examined for by treating 12*10 grains in a pla- 

 tinum crucible with sulphuric acid at a gentle heat, the crucible 

 being at the same time covered with a glass plate waxed on the 

 under side and kept cold on the upper side; some characters 

 were traced through the wax with a fine point; no visible etch- 

 ing was remarked after the operation. 



The mineral, therefore, consisted only of water, alumina, and 

 phosphoric acid with an accidental trace of oxide of iron and 

 silica. Its quantitative analysis was conducted as follows : — 



Determination of the Water. 



22*22 grs. of the transparent colourless mineral left, after 

 heating to redness, 13*49 grs. residue ; also evolved 8*73 grs. grs. 

 water, equivalent to 39285 per cent, water in the mineral. 



15*38 grs., same quality, left under same treatment 8*93 

 residue; also 6*45 grs. water, equivalent to 41*18 per cent. 



13*365 grs., translucent but of a faint yellow colour, left 8*105 

 grs. residue ; also 5*26 grs. water, which would make 39*37 

 per cent. 



* Instead of, as commonly, using miciocosinic salt (phosphate of soda 

 and ammonia), I prefer employing the dried phosphate of soda prepared 

 by heating strongly the above until all ammonia is driven off. It will be 

 found much more convenient in practice, as it melts gently, and does not 

 froth and spit as the microcosmic salt does. 



