Nelson. — Tricalcic Phosphate in Carbonic-acid Solutions. 577 



while consistent with one another, were not so with those of series VI. They 

 show a regular increase of solubility with pressure. Here again after a 

 few runs the presence of copper was found on evaporating the solution, 

 rather more being present at high than at low pressures. The gold appeared 

 to be quite intact, except for a ring about | in. wide round the top, above 

 the part wetted by the solution during shaking. 



Series IX (1 gram) and Series X (2 grams) also gave very regular 

 results. 



In order to show more clearly the relations between amount of phos- 

 phate taken, pressure, and solubility, all the curves are placed together 

 in fig. 3. 



Analyses were made of the following materials : — 



A. The original material from the bottle. 



B. The undissolved phosphate left in the bottom of the copper tube 



was collected, and dried at 100° C. It contained a very slight 

 trace of copper, which was not estimated, and gave a very slight 

 effervescence with nitric acid. 



C. The phosphate was obtained by evaporating on the water bath 



the solution not required for estimation. The sample which 

 was analysed for copper was from several high-pressure solutions, 

 and contained more copper than usual. This material also gave 

 an effervescence on treatment with nitric acid, showing the 

 presence of a carbonate, derived probably from calcium bicar- 

 bonate in the solution. This carbon dioxide, however, was in 

 too small a quantity to be estimated directly. 



The results of average analyses were :- 



CaO 



P 2 5 



Loss (at red heat) 

 CuO .... 



98-52 98-72 99-79 



The loss at red heat in A must be due to water only, and the analysis 

 corresponds to 3-18CaO. P 2 5 . 2-43H 2 0. If in B the loss is assumed to be 

 due wholly to water, which is nearly true, the analysis shows the material 

 to be 3-2CaO. P 2 5 . 2-3H 2 0. In C, the ratio of CaO to P 2 5 is 3-44. But 

 the maximum of CaO which can combine with P 2 5 is 3CaO. Hence 

 the remaining 0-44 CaO must be combined in some other way, which under 

 the circumstances may be assumed to be with C0 2 , asCaC0 3 . Similarly, 

 if the CuO is in the form of carbonate, the amount of C0 2 present will be 

 5-72 per cent., leaving 5-04 per cent, of loss at red heat as 1-07 molecule 

 of water. Thus the analysis gives 3CaO. P 3 O 5 .0-44CaCO3.0-02CuCO 3 . 

 l-07H 2 O. It is also possible that some of the CuO may be combined with 

 P 2 5 instead of with C0 2 . 



The soluble compound which is formed is unstable, since it is deposited 

 as a precipitate when the carbon dioxide is removed, either on standing, 

 or more completely on boiling. The presence of calcium carbonate, or in 

 solution calcium bicarbonate, indicates that an interaction has taken place, 

 the carbonic acid having attacked the tricalcic phosphate with the forma- 

 tion of some calcium bicarbonate. 



19— Trans. 



