OP ARTS AND SCIENCES. 189 



is vigorous, the benzoic acid fusing and forming a black crust, which 

 was made to spread over the whole surface by moving the dish when 

 necessary ; the crust was then removed,* and the same action repeated 

 with fresh benzoic acid, until all the superfiuoride was exhausted. In 

 this way 180 gr. of the crusts could be obtained in a single operation. 

 If the stream of superfiuoride was not rapid, the benzoic acid was apt 

 to become covered with chromic anhydride, probably from a trace of 

 moisture in the acid, and this not infrequently gave rise to unimpor- 

 tant deflagrations ; but the amount of chromic anhydride formed even 

 by this secondary reaction is very small, almost the whole of the 

 chromic superfiuoride being converted directly into the green chromic 

 fluoride Cr^Fg. 



To obtain the difluorbeuzoic acid from the crude product it was dis- 

 solved in a strong solution of sodic carbonate, and, after the chromic 

 hydrate had been removed by filtration, fractionally precipitated with 

 hydrochloric acid, the liquid being boiled with the separated benzoic 

 acid after each precipitation. We found it wisest to divide the whole 

 into two about equal fractions at first, and then to fraction the least 

 acid portion again, when all the difluorbeuzoic acid collected in the 

 least acid fraction, which should be made very small ; this was then 

 boiled with water, and the residue purified by re crystallization from 

 boiling benzol and sublimation. The yield of difluorbeuzoic acid is 

 only 1.5 per cent of the benzoic acid used. 



Having obtained the difluorbeuzoic acid, we next tried to detect the 

 presence of monofluorbenzoic acid, and for this purpose examined the 

 acid contained in the hot aqueous filtrate from boiling the difluor acid 

 with water, and also the fraction immediately preceding the least acid 

 one. As the metafiuorbenzoic acid melts at 123°-124° according to 

 Paterno and 01iveri,t the melting-point was of no assistance to us, and 

 we have accordingly made a number of analyses of the barium salts of 

 these fractions, both before and after a repetition of the fractional satu- 

 ration, which have given results varying from 35.81 to 35.99 per cent 

 of barium ; baric benzoate contains 36.15 per cent, baric fiuorbenzoate 

 33.01 per cent.t TVe also thought it might be contained in a mixture 

 of acids, obtained by boiling the crude product with insuflicicnt water, 

 melting sometimes as low as 102' ; but this on fractional neutralization 



* "We tliink it right to mention that both of us were poisoned, one severely, 

 by the crude product, which was spilt on our clotiies in removing the crusts 

 from the platinum dislies. 



t Ber. d. ch. G., 1882, p. 1197. 



I Tile most acid fractions were also pure benzoic acid, as was to be expected. 



