Manchester Memoirs, Vol. xlviii. (1904), No. V\. 13 



The first tube (scarcely visible) is the blank with the 

 reagents alone showing no trace of mirror, the second is 

 without either amyl alcohol or glucose, and shows a short 

 dense mirror. The third tube is with 2 c.c. of amyl alcohol 

 alone, and shows a less dense mirror than the one without 

 alcohol, but the difference here is not nearly so marked as 

 it is in the duplicate experiments shown in the previous 

 plate, which proves the advantage of the cooling over the 

 non-cooling method. The following five tubes show the 

 influence of the combined effects of 2 c.c. of amyl alcohol 

 with I, 2, 5, 10, and 20 grammes of glucose respectively, 

 and indicate that glucose has a marked influence in pre- 

 venting the formation of the arsenic mirrors. 



The sixth series, Plate X., Fig. i, is a repetition of the 

 fifth series, but using double the quantity of arsenic, viz., 

 J^th of a grain per gallon, and without -^ooling the tubes 

 at the point at which the mirrors were produced.* 



The seventh series, Plate X., Fig. 2, shows the arsenic 

 obtained from eight samples of beer, marked /^ to H. 

 Those on the right are obtained by the Marsh-Berzelius 

 process when using 25 c.c. of each sample, after completely 

 destroying the organic matter with sulphuric and nitric 

 acids, and removing the latter as already described by me. 

 Each shows different sized mirrors by the Marsh-Berzelius 

 process, A containing x^o^h B i^g-th, C x^tr^'^' -^ 2^0^^, 

 E x^gth, F xl^th, G x^th, and H x^th of a grain of 

 AS4O6 per gallon ; whilst with the Committee's electrolytic 

 process, using the beer direct, no trace of mirror is obtained 

 from any of them, and the tubes are therefore all but 

 invisible, their positions being indicated by the alphabetical 

 letters. 



*The mirror in the experiment in which 5 grammes of sugar were 

 employed shows Ughter in the print than it does in the photograph or in the 

 tube itself. 



