206 PROCEEDINGS OF THE AMERICAN ACADEMY 



to remove the excess of hydrochloric acid previous to the precipitation 

 of the tin, tlie sohition is evaporated on the water-bath to dryness. 

 The temperature should not, however, go beyond this point, as we had 

 opportunity to observe that even at lOC^ small quantities of aniline 

 were formed, while at 12t>'' the para-amido-benzoic acid was decom- 

 posed completely iuto carbonic acid and aniline. It would be of 

 interest to examine the action of SnClg on other aromatic acids under 

 the conditions mentioned. The solution of para-amido-benzoic acid 

 was treated with sodium carbonate, and the acid precipitated out by 

 means of acetic acid. After a single crystallization in water, it 

 showed the melting point 18G°. 



The para-amido-benzoic acid obtained in this way was placed in a 

 flask, and a sufficient quantity of slightly fuming sulphuric acid 

 (s. g. 1.850) added to dissolve the mass. Tiie flask was then exposed 

 to a temperature of 170* — 190", for from three to four hours, in a 

 paraffine bath. After cooling, the contents of the flask were removed 

 by means of water, and barium-carbonate was added to the solution 

 until it was neutralized. The liquid was boiled for several hours, and 

 then filtered. It was necessary to heat the residue repeatedly with 

 hot water before a slight pinkish Ime could be removed. Exactly 

 enough sulphuric acid was added to the filtrate to precipitate the 

 baruun, and it was separated from the barium sulphate formed. By the 

 evaporation of this solution, groups of crystals separated out, showing 

 various shades of yellow, orange, and red. Several crystallizations 

 from water freed tliem from small quantities of a by-product consisting 

 of minute red crystals, and after boiling with animal charcoal they 

 retained a faint pink color. The substance was dried at 110°, and 

 submitted to a series of analyses. The sulphonic acid which was 

 expected as the natural result of the reaction, would have yielded the 

 following percentages: — 



\ 



COOH 



J 



C — 38.7 

 11= 3.3 

 S = 14.5 



LN= 6.4 



The analyses showed, however, that we had obtained a body of 

 an entirely diiferent nature, and led to the following formula : — 



