182 MR. H. A. SMITH ON 



were found. On analysis this deposit gave, as an average 

 of nine analyses, 43*43 per cent. As a O r This flue had, 

 however, been in working for some years. 



Passing from the flue to the towers, the next analysis 

 made was of the coke itself. Even when examined by 

 the highest microscopic power at my disposal, there was 

 no appearance of any arsenic crystals; but arsenic was 

 very easily detected in it. For each analysis 10 lb. of 

 the coke was used. It was digested first with distilled 

 water, and then with perfectly pure HC1. 



The results of three analyses gave 2*88 as the average 

 percentage of arsenic present, the variation being from 

 2*64 per cent, to 3*18 per cent. As 2 r The towers had 

 been in use for about a year. 



Table II. — Coke from Condensing-towers. 



No. of analysis. Arsenic trioxide. 



per cent. 



1 2*64 



2 < 3' 18 



3 2-84 



8-66 

 Mean 2-88 



The air in the flue leading from the condensing-towers 

 to the chimney was next examined. 



For each analysis 50c cubic feet of air was employed, 

 and twelve analyses were made. 



The amount of air passing at this flue was 31,722 cubic 

 feet per hour. 



In the following Table the results have been calculated 



out to 



Grains As z O ? per 1000 cubic feet, 



„ per hour, 



„ per day. 



The numbers here are comparatively large, and the 



