Manchester Memoirs, Vol. xlviii. (1904), No. IT. 3 



My first test with the Committee's electrolytic appa- 

 ratus showed that it gave much better arsenic mirrors 

 than I had obtained by the first electrolytic apparatus 

 which I devised some time ago, and I felt that after all 

 Lord Kelvin was probably on the right track when he so 

 strongly advocated the electrolytic process. 



I found however that it was impossible to get mirrors 

 by the Committee's electrolytic apparatus when using 

 arsenic in the form of trioxide beyond the 3-^th of a 

 grain of arsenic trioxide per gallon when working on 

 50 c c. of liquid, equal to a total quantity in the apparatus 

 of O'0O23 milligramme of arsenic trioxide (AsiOe), equiva- 

 lent to O'OO 1 77 m.g. of the element arsenic. 



I must apologise for mixing the metric system with 

 English weights and measures, but this seems necessary, 

 as the English method which has been accepted is to 

 speak of grains or fractions of a grain per gallon or per lb., 

 whilst I have worked in the laboratory with the metric 

 system. In all my experiments I have used solutions 

 computed as containing fractions of a grain (per gallon) 

 of (As^Os) whilst employing 50 c.c. of such solutions or its 

 equivalent for the tests, as this quantity of beer or other 

 liquid is the one I have found most convenient for working. 



I devised an apparatus on the same lines as that of 

 the Committee, but greatly increased the areas of the 

 platinum anode and kathode, and, curiously enough, I 

 obtained with them exactly the same sizes and intensities 

 of mirrors as by the Committee's apparatus. This 

 arrangement also failed to detect anything beyond the 

 g^^th of a grain per gallon of AS4O6 when working on 50 c.c. 



The use of other metals than platinum as kathode. 



Lead was first tried in fine sheet, and after getting rid 

 of the arsenic which it originally contained, and which 



