147 



The process I employ is to add a soluble phosphate, or arseniate, to the 

 solution of a salt of the nietal, the phosphate or arseniate of which, is 

 required, in the manner hitherto adopted, but only in such limited quantity, 

 that the mixed solution remains acid in its reaction, instead of alkaline, as occurs 

 in the usual method of procedure. If the precipitate is long in appearing, it may 

 occasionally be crystalline ; if it comes at once it will be gelatinous, as usual, 

 but in the course of a few hours, sometimes, however, a few days, it will be 

 found crystallized throughout. 



The essential features of this process are : — 



1st. The maintenance of the precipitated metallic salt in its integrity, 

 which is effected by having the surrounding solution feebly acid. 



2nd. Allowing motion to the particles of these gelatinous precipitates, 

 whereby they are amenable to the action of crystallizing force ; this is ac- 

 complished by keeping a little of the same phosphate in a soluble state in 

 contact with them. 



In this manner I have succeeded in crystallizing the following phosphates 

 and arseniates, which occur in this form in the natural state : — 



Phosphate of Zinc, — Hopeite, (Zn. 0) 3 + PO5 + HO5 



Phosphate of Cadmium. 



Arseniate of Zinc, — Kottigite. 



Arseniate of Lime, — Pharmacolite. 



The following crystallized phosphates and arseniates, produced, do not 

 occur as such in a natural state. 



Phosphate of Lime (-(CaO) 2 + HO) + P0 5 + 3 HO. This has the same 

 composition as the amorphous precipitate, produced by adding a triphosphate 

 to chloride of calcium, and then a little ammonia, (the precipitate being air- 

 dried) ; and it is isomorphous with the natural arseniate of lime above, 

 Pharmacolite. It crystallizes in the form of rhombs, and is acid to test paper. 



Phosphate of Chromium contains 24 eqs. of water, and has probably the 

 same constitution as Delvauxine, or hydrous phosphate of sesquioxide of iron, 

 the iron being replaced by chromium, its colour is the same as that of chrome- 

 alum, the substance used as the source of the chromium. 



Phosphate of Silver. Only crystallized from its solution in acetic acid. 



Phosphate of Baryta and Strontia are also easily crystallized. Those salts 

 having formulae attached, have been analyzed. 



On reviewing these salts, it will be noticed that the copper, nickel, cobalt, 

 and iron, phosphates and arseniates, are absent. Indeed, I have not been able 

 to crystallize any of them in this manner ; although I am aware that it has 

 been affirmed, that phosphate of nickel has been artificially crystallized. But 

 I find that all these metallic phosphates, etc., are capable of forming double 

 phosphates, etc., with phosphates of magnesia and ammonia. The metal may, 

 I think, be looked upon as substituting one equivalent of magnesia in the 

 common ammoniacal phosphate of magnesia, thus (Met 4- Mg + NHO) 

 — P0 5 12 HO in place of 2 (Mg O) + NH 4 O + P0 5 + 12HO. 



I also find that phosphate of zinc forms a crystallizable compound with 

 either phosphate of cobalt or nickel. It may be remarked here that the 

 crystalline mineral, Kottigite, an impure arseniate of zinc, always contains a 

 little of both these phosphates. . 



Lastly, it appears that crystalline precipitates are readily prodticed by 

 contact of soluble phosphates with solutions of the metals cobalt and nickel, if a 

 salt of ammonia is also present. These precipitates contain ammonia, in small 

 quantity, but it appears to be as an essential element in their composition, and 

 not a mere accidental impurity ; its quantity has not yet been determined. 



The inferences T would draw from these results are : — 



1st. That several of the crystalline, simple, natural phosphates and 



