506 



Dr. II. E. Armstrong on 



[June 16, 



manner will give rise to carbonic oxybromide, together with sulphurous an- 

 hydride and free bromine. I do not believe that a pyrosulphuric bromide 

 will be formed. 



Action of Sulphuric Anhydride on Chloroform. 



The experiment was made under exactly the same conditions as the 

 former one. The action set in immediately on addition of the chloroform, 

 aud was also accompanied by a rise in temperature and an evolution of 

 gas, which gas proved to be pure carbonic oxide, and, as was ascertained 

 by experiment, free from hydric chloride, whose formation by the reaction 

 was not impossible. 



The reaction was easily completed by aid of a gentle heat, and the liquid 

 remaining in the flask was then subjected to distillation. After repeated rec- 

 tification the greater part boiled constantly at 139°-140 3 , and was obtained 

 as a colourless, mobile liquid, differing only in B.P. from the chloride ob- 

 tained by the action of the anhydride on the tetrachloride. 



The analysis yielded the following results : — 



•1051 grm. gave -138 grm. AgCl =32*4 CI. 

 •1150 grm. gave '242 grm. BaSO, =28*8 S. 



Calculated for Calculated for 

 Found. SO.,HOCl. S 2 0 5 C1 2 . 

 S = 28-8 27-46 297 

 CI = 32-4 30-47 33'1 



It is evident from the above comparison of the numbers obtained with 

 those calculated for each of the formulae S0 2 IIOCl and S 2 0 5 C1 2 , that they 

 lie midway between the two, and there is therefore no doubt that this liquid 

 is a mixture of both. 



Pyrosulphuric chloride and sulphuric chlorhydrate are therefore derived 

 from sulphuric anhydride and chloroform, as is explained by the following 

 equations, — 



2S0 3 + CHC1 3 = CO + HC1 -f- S 2 0 6 C1 2 . 

 S0 3 + HC1 = S0 2 H0C1. 



There was a probability of formylic chloride, HCOC1, instead of the 

 products of its decomposition, CO-f-HCl, being obtained by this reaction, 

 as is evident from the equation, — 



2S0 3 + CHCI3 = HCOC1 + S 2 0 5 C1 2 . 



This is not the case, however, as is proved, 1, by the fact that carbonic 

 oxide is evolved immediately on adding chloroform to the anhydride, which 

 is also the case on reversing the experiment, and allowing sulphuric anhy- 

 dride vapour to act on an excess of chloroform ; 2, had it resulted it must 

 have been detected on distillation, either as such, or if it became decom- 

 posed, by an evolution of CO and HC1 ; but the distillation only yielded the 

 above-mentioned chloride, and was not accompanied by any further disen- 



