OF ARTS AND SCIENCES. 95 



Therefore I conclude that the formulae do represent the composition, 

 both empirically and rationally, and that there is no suboxide of cop- 

 per present, as also no arsenic acid. 



Wagner, after giving Ehrmann's receipt, publishes the following from 

 M. Braconnot: '• 15 kilos of sulphate of copper (CuSO^-f-'^^^-.'O) are 

 dissolved in the smallest possible quantity of boiling water, and mixed 

 with a boiling ami concentrated solution of arsenite of soda or potash, 

 so prepared as to contain 20 kilos of arsenious acid. There is imme- 

 diately formed a dirty greenish-colored precipitate, which is converted 

 into Schweinfurt green by the addition of some 15 litres concentrated 

 wood-vinegar. This having been done, the precipitate is filtered off 

 and washed. . . . 



" It thus appears that the preparation of this pigment aims first at the 

 least expensive preparation of neutral arsenite of copper, which is then 

 converted into aceto-arsenite of cojjper, by digesting the precipitate 

 with acetic acid." 



Kow this is very probably the method, by which the samjile 

 obtained from the Massachusetts Institute of Technology was made, and 

 accounts for the basic copper sulphate found in it. The insoluble 

 matter is used to dilute the color and as adulteration. 



These latter are therefore impurities, and are to be disregarded in 

 speaking of its composition. 



Now with regard to the pure pigment: Ehrmann's analysis is with- 

 out doubt correct, and is sustained by two samples prepared by myself; 

 but this cannot be regarded as the universal composition of the salt, 

 and does not, as has be^u already said, cover the commerrial samples 

 that I found in the market. Therefore we must look farther, and we 

 find there is another salt having the formula, — 



2(CuOC,HA) 3(CuOAsA), 



in two commercial samples ; hence we infer that the composition varies 

 between these salts, that it may be CuOC^H^Og 3(CuOAs.,0.j), or 

 2(CuOC^Hy03) S^CuOAs^Og), or something between the two; and we 

 also see that the amount of arsenic trioxide may vary considerably, 

 from the difficulty with which it is removed by washing, even with hot 

 water, in process of manufacture. 



In the extract from N. Reilter, we have a series of results which 

 vary from any obtained by Ehrmann and myself (with one exception), 

 where the amounts of acetic acid are much less. And he states 

 that Schweinfurt green is copper mono-arsenite combined with varying 

 amounts of basic copper acetate. We have a variance of from 



