352 PROCEEDINGS OF THE AMERICAN ACADEMY 



phuric auliydride, or, what seems to us more probable, it is brought 

 about by the concurrent attraction of the sulphuric anhydride for the 

 benzoldisulphonic acid. It is possible, however, that one of the two 

 following reactions may express the action more correctly, — 



KHSO, + C,H,(S03H), = C,H3(S03H),SO,K + H,0. 

 KaSA + 2 C.H.CSOgH), = 2 CeH3(S03H),SCUv + H.O. 



But in these reactions the excess of sulphuric acid must be supposed 

 to act as a dehydrating agent, and thus produce the action, since we 

 have proved experimentally that no action takes place when an excess 

 of sulphuric acid is absent ; and this observation seems to us to make 

 these two reactions less probable than the series in which sulphuric 

 anhydride is formed. 



We have not multiplied experiments with different metallic sul- 

 phates, as it has not seemed to us that such work would throw any 

 additional light on the theory of the process ; we may mention, how- 

 ever, that we have also tried alumiuic sulphate, which gave rise to the 

 formation of a little trisulphonate when heated with metadisulphonic 

 acid and sulphuric acid, but, as far as we could judge, was much less 

 efficient in its action than the potassic sulphate. 



Preparation of Benzoltrisulphonic Acid. 



As the result of many experiments we have worked out the following 

 process for preparing benzoltrisulphonic acid from potassic benzolmeta- 

 disulphonate, — 15 grams of potassic benzolmetadisulphonate, which 

 need not be freed from its water of crystallization, are mixed in a 

 porcelain dish with 18 grams of common strong sulphuric acid,* and 

 heated with the free flame, care being taken to avoid heating the 

 edges of the liquid too intensely during the first stage of the process, 

 as at this time tliere is danger of charring. After the salt has dis- 

 solved, the sulphuric acid begins to come off in heavy white fumes, but 

 without the formation of bubbles ; in fact, there is no bumping or 

 tumultuous evolution of vapor in any part of the process. If the 

 operation is stopped at this point, the disu][)hr>nate will be found un- 

 altered. After heating for a few minutes, little bubbles appear in 



* Double these quantities can be used, but it is not convenient to work 

 with much larger quantities than 30 grams at a time. The benzolmetadisul- 

 j)honate may contain a small quantity of potassic carbonate or sulphate without 

 injury to the process, but should be free from colored tarry impurities. 



