408 Dr. Thomson on [June, 



what is usually exposed to sale. But I have likewise subjected 

 to analysis a quantity of bleaching powder, which I have reason 

 to beheve constitutes a pretty fair specimen of the average 

 strength at which the powder is usually exposed to sale. The 

 analysis was conducted precisely as the former, and upon nearly 

 as great a scale. To be quite sure of extricating the whole 

 oxygen gas from the powder, I raised it in the retort to an inci- 

 pient red heat, and kept it in that state for a considerable time. 

 The quantity of bleaching powder analyzed was 1123 grains. 

 J found its constituents as follows : 



Subbichloride of lime 410-135 



Muriate of lime 207-720 



Water 190-100 



Uncombined Hme, and impurity. . . . 315-045 



1123-000 



When these are reduced to the proportional numbers for 100 

 parts of the powder, we find it constituted as follows : 



Subbichloride of lime 36*52 



Muriate of hme 18-50 



Water 16-93 



Uncombined lime, &c 28-05 



100-00 



The great difference between this powder and the former con- 

 sists in the quantity of uncombined lime. The proportion of 

 muriate of lime is nearly the same in both. This I think renders 

 it probable that the muriate of lime is formed, at least chiefly, 

 during the process of manufacturing the bleaching powder. 

 Both of these samples were quite fresh, and were transported 

 into my laboratory in bottles well sealed up. 



Besides the variation which frequently occurs in the original 

 quality of the powder, arising sometimes from defective satura- 

 tion of the lime with chlorine, and sometimes from a conversion 

 of that gas into muriatic acid, or to these two causes combined, 

 it is also subject to deterioration by keeping, being gradually 

 converted into muriate of lime. 



Article II. 



On the ComposilioN of Cliloride of Siilphm-. By Thomas Thom- 

 son, M.E>. F.R.8. 



This substunce was discovered by me in 1803, and described 

 jn a paper published in Nicholson's Journal, for October, 1803, 



