562 UEPORT— 1880. 



By the above table it can be perceived that in the English sample there -was 

 only -251, and in tlie Swiss -43, and in the laboratory sample -44 percent, of chlorine 

 yet to be accounted for, and on washing the samples twice vdth distilled -water, 

 each time with 200 c.c, and again estimating the chloruie in the filtrates, they 

 \eere found to contain as follows : 



English Bleach. Swiss Bleach. Laboratory Bleach, 

 Per cent. Per cent. Per cent. 



Chlorine in 1st 200 c.c. -198 -270 -19 



2nd „ -007 -051 -12 



Though the samples were now washed -n-ith successive 100 c.c. of water up to 

 1500 c.c, only -0392 per cent, of chlorine could be estimated in the English bleach, 

 ■09 per cent, in the Swiss, and -112 per cent, in the laboratory bleach. After this 

 the filtrates were not examined ; but an accurate account was kept of the quantity 

 of water used, and when each sample had been washed with 1800 c.c. the reaction 

 with the iodine and starch paper was very uncertain ; but on adding iodine and 

 starch solution to the filtrate, together with acetic acid, an immediate reaction 

 occurred, which was the case with these particular samples imtil 3000 c.c. of water 

 had been used to wash each 10 grammes of bleaching powder ; after that quantity 

 of water had been used, chlorine could no longer be detected, either with the starch 

 paper or starch solution and acetic acid, though each sample was kept covered with 

 100 c.c. of water for many hours, by closing the bottom of the funnels. This part 

 of the investigation was repeated many times, and the quantity of water used to 

 extract the chlorine varied widely with diflerent samples, as well as the rate at 

 which it was extracted (that depending greatly on various circumstances : 1st, On 

 the percentage of chlorine in the sample ; 2nd, On the temperature of the water, 

 and degree of fineness of the sample ; 3rd, On the speed at which it is washed by 

 the pump). On various occasions the quantitj- of water required to extract the 

 chlorine from 10 grammes amounted to as much as 5000 c.c. This probably arose 

 from not having the cone of parchment paper large enough, thereby allowing a 

 small quantity of suspended matter to be carried through the Swedish filter paper ; 

 or it not improbably arose from the difficulty of preventing the first portions of the 

 strong hypochlorite solution from coating the inside of the tube of the funnels, and 

 thereby contaminating the latter filtrates. But even when more than 5000 c.c. of 

 water had been emplo3'ed to wash 10 grammes, the filtrate, which had long ceased 

 to give a reaction with starch paper, still continued to do so very decidedly with 

 the iodine and starch solution together with an acid ; but on the addition of an 

 extra cone of Swedish paper to the one already in the funnel, the reaction no 

 longer appeared. It was considered ad^^sable to estimate the remaining chlorine 

 in the residue of various samples wlien they no longer gave a reaction with the 

 starch paper, but continued to do so with the acid and starch solution. This was 

 done by washing the residue into a beaker and dissolving it by the addition of as 

 small a quantity of acetic acid as possible, and then diluting, after which the chlorine 

 was estimated by the arsenious acid process as rapidly as possible. 



Though numerous residues were examined at this stage of the washing, the 

 highest percentage of chlorine found was only -007 per cent., and the lowest -0032 

 per cent. 



The percentage of residue in the various samples was determined in two stages : 

 (1st) When the samples had been washed till they no longer gave a decided reaction 

 with the starch paper; (2nd) After complete washing. 



The percentage of residue in the partially washed samples was found to 



