OF ARTS AND SCIENCES. 15 



In the first of these formulas, there is an evident mistake ; it is a 

 copy of the formula given by Watts, without takiug into account that 

 Watts, while doubling the atomic weight of oxygen, retained the old 

 weight of copper, so that, corrected, this formula reads (CuO)2 ASjOj ; 

 or, in other words, is Berzelius's formula. 



But both of the formulas, one and two, are wrong, from the fact that 

 they contain no water. No. 3 is an evident attempt to represent the 

 acid arsenite which Berzelius mentions, and No. 4 is Bloxam's 

 formula. 



As will be seen, the whole literature of the subject is founded upon 

 three sets of facts. Scheele's prescription, which all the authors, whom 

 I have quoted, have given, making only such alterations as were necessary 

 on account of chancres in weights and measures. And it is a sintjular fact, 

 that not one of these authors has taken the trouble to see if the quan- 

 tities of copper sulphate and arsenic trioxide taken would produce a 

 salt of the formula given ; or have discovered the fact tiiat nearly 

 twice as much potassium carbonate is used as is necessary to saturate 

 the sulphuric trioxide of the copper sulphate, and Scheele's foot-note 

 has been totallv ignored. 



Secondly, Berzelius's account of the salt, which has evidently been 

 misunderstood. 



Thirdly, Bloxam's analyses of salts, which he would have found 

 difficult, if not impossible, to reproduce, had he been so inclined. 



After comparing the various works cited, it became a matter of in- 

 terest to find out, in the first place, what Scheele's green really is, 

 what are its properties, and whether there is more than one copper 

 arsenite. 



The experiment was tried of making copper arsenite according to 

 the method given by Berzelius ; that is, by dissolving copper carbonate 

 in an aqueous solution of arsenic trioxide. 



Hydrocopper carbonate was prepared by precipitating copper sul- 

 phate in the cold by an excess of sodium carbonate, and washing the 

 precipitate with cold water until free from sulphates. Some of the 

 precipitate was boiled with a saturated solution of arsenic trioxide, its 

 blue color soon changed to a bright green, which it maintained, 

 although boiled for upwards of an hour. The green i^recipitate was 

 filtered off, and washed with hot water, until the wash waters were free 

 from arsenic. 



The substance remaining on the filter was of a bright sfreen color, 

 scarcely inferior to Schweinfurth green in brilliancy, although of a 

 vellowish shade. 



