246 PROCEEDINGS OP THE AMERICAN ACADEMY 



has been fulfilled, as by heating potassic benzolmonosulphonate with 

 sodium amide we have obtained aniline, and in the same way meta- 

 phenylenediamine from potassic benzolmetadisulphonate ; but unfor- 

 tunately the yield, especially in the latter case, is so small, that the 

 method could hardly be used with profit for the preparation of new 

 amido compounds, so that its usefulness must be confined to determi- 

 nations of constitution. 



The sodium amide was prepared according to the method of Beil- 

 stein and Geuther, that is, passing dry ammonia gas over sodium 

 heated in a series of small flasks, except that after the first experi- 

 ment we dispensed with the use of hydrogen for expelling the air from 

 the flasks before the eodium was heated, using the ammonia gas itself 

 for that purpose. 



Action loith Potassic Benzolmonosulphonate. — One of the flasks 

 containing the sodium amide was broken, its contents pulverized as 

 rapidly as possible, and mixed (without separating the powdered glass 

 from the flask) with anhydrous potassic benzolmonosulphonate in about 

 the proportion of one molecule of each, calculated from the amount of 

 sodium used in making the amide. The mixture was then heated 

 gently over the lamp in a test-tube until the brown fused spots, which 

 appear at first, had spread throughout the entire mass ; but the process 

 should be stopped before these spots show signs of decomposition, as, 

 if the heat is carried too high, the yield is not so large. During the 

 heating some aniline vapor escapes, but the amount of this is so in- 

 significant, that it is not worth while to carry on the process in a retort. 

 The product was then carefully treated with water, and distilled with 

 steam as long as anything passed over. The presence of aniline in 

 the turbid distillate was determined by its smell, by the strong purple 

 color which the distillate gave with a solution of bleaching-powder, 

 and by the analysis of the chloride made by adding dilute hydrochloric 

 acid to the distillate, filtering, and evaporating to dryness. 



0.4966 gr. of the salt gave 0.554 gr. of argentic chloride. 



Calculated for CcHjNHaCl. Found. 



Chlorine 27.41 27.60 



The yield varied a great deal, but was in the neighborhood of 10 per 

 cent, rising once as high as 15 per cent of the theory. The portion 

 of the distillate insoluble in dilute hydrochloric acid consisted of white 

 scales, which, after repeated recrystallization from ligroine, melted 

 between 53° and 54°, and gave an intense blue color with strong nitric 

 acid. We therefore inferred that they consisted of diphenylamine, an 



