Manchester Memoirs, Vol. I. (1906), No. \%. 3 



in a stoppered bottle a reddish form of arsenic is pro- 

 duced which gradually falls from solution. 



Improvement in the cooling method for co7uiensing the 

 arsenic mirrors in arsenic determinations. 



The results obtained by the method which I formerly 

 adopted of running a stream of cold water over a piece of 

 tissue paper covering the drawn out portion of the tube, 

 leave something to be desired, especially in the estimation 

 of exceedingly minute quantities of arsenic, for the water 

 at the edge of the paper next the flame does not give an 

 exceedingly sharp line of cooling surface, and we have 

 improved on this by taking a piece of block tin tube 

 8 inches long, having an internal bore of je^hs inch. 

 This tube is flattened so that the internal measurements 

 become \ inch x |th inch, and it is bent downwards in a 

 curve about f inch from the end. Two holes are made 

 in the edges of the flattened tube opposite each other 

 2 inches from the bent end, and a thin silver-foil plate 

 soldered over one of these holes. This plate is perforated 

 by drilling in it a hole of a diameter sufficient to allow 

 the drawn out portion of the largest sized glass tube to 

 fit it. The tube is passed through the silver plate to the 

 point at which the mirror is to be formed, and the water 

 turned on through the tin tube, which is connected with 

 the main by a thin rubber tube. A gentle flow of water is 

 kept running through this tube, and it is found that if 

 the length of the bent end of the tin tube is properly 

 adjusted the water does not escape at the silver plate or 

 the hole in the tin tube opposite, even when the tubes fit 

 comparatively loosely. By this device o-(/ooth of a grain 

 per gallon of arsenic trioxide when working on 50 c.c. of 

 the solution can be readily seen as a distinct narrow 



