IN CASES OF POISONING. 277 



nitre would interfere with the subsequent treatment of 

 the mass. It is better to make a preliminary test with 

 a small portion of the mixture, by introducing it into 

 a small red-hot crucible, and observing whether the 

 mass is perfectly white after deflagration. If it be 

 black and carbonaceous, more nitre must be added. 



The mass, which now consists essentially of carbo- 

 nate, nitrate, and nitrite of potassa, and may also contain 

 arsenate of potassa, is dissolved in the smallest possible 

 quantity of boiling water, and the solution, without 

 filtering off from the suspended phosphate of lime and 

 silica, gradually mixed, in a porcelain dish, with a 

 slight excess of sulphuric acid. The pasty saline mass 

 thus produced is carefully heated till all the nitrous 

 and nitric acids are expelled, a point to which great 

 attention must be paid. On cooling, the mass is stirred 

 up with a little cold water, and the solution poured off 

 from the large deposit of sulphate of potassa. The 

 latter is washed several times with cold water, the 

 washings mixed with the first solution, and the liquid, 

 treated as above, with sulphuretted hydrogen. The pre- 

 cipitate then only requires to be oxidized with nitric 

 acid, with the precaution that the acid must be entirely 

 removed by evaporation before the solution is intro- 

 duced into Marsh's apparatus. 



It is rarely of importance to the evidence that the 

 weight of arsenic existing in a body should be deter- 

 mined. Such an estimation can only be relative, since 

 it is impossible to extract and weigh the whole of 

 the arsenic contained in all the parts of a body. In 

 such a case, a somewhat longer reduction-tube should 

 be employed, into which is introduced a closely twisted 

 spiral of pure bright copper, about two inches in 

 length ; this spiral is accurately weighed with the tube. 

 The latter is then heated in two places, one nearer the 

 evolution-bottle, for the deposition of a mirror; the 

 other, at some distance, where the strip of copper is 

 24 



