30 OECHSLIN. 



phorus compounds " and then warmed on a water bath. As the chloroform distilled 

 ofT, the solid passed into solution, with the exception of a very slight residue 

 in a relatively small quantity of water. Since dibenzarsinic acid is insoluble 

 under such conditions, and is soluble in dilute acid only as its quinine ester, it 

 was completely esterified. 



Sodium carbonate produced a very fine precipitate from this solution, insoluble 

 in an excess of the reagent and differing widely from quinine in its appearance. 

 No precipitate of dibenzarsinic acid was noticed on acidifying the filtrate. The 

 white residue was washed and dried at 13Q,°. This treatment rendered it 

 slightly yellow, but it was still completely and very easily soluble in dilute 

 hydrochloric acid. The ester still contained some free quinine, as the alkaloid 

 was used in excess. To remove the latter it was dissolved in a small quantity 

 of absolute alcohol and precipitated by ether. The benzarainic acid derivative 

 precipitates, and quinine remains in solution. Di-quinine dibenzarsinic acid is 

 a fine, white powder, which is filtered and washed dry with ether. The yield 

 was 8 grams. 



I. 0.2614 gram substance gave 0.0408 gram MgjAsjO,. 



II. 0.2802 gram substance gave 0.0443 gram MgoAsjO,. 



Calculated for 



C64H5.-,N408As Found per cent. 



(CcHiCOJz: (CooHogNsOo):: AsOaH I- II. 



Per cent. 



As=7.8 7.4 7.6 



III. ACETYLPHENYLGLYCIK-AHSINIC ACID QUININE ESTER. 



Besides the two esters already described, I endeavored to prepare the 

 quinine ester of phenylglycin-arsinic acid, with the intention of reducing 

 it to the di-quinine ester of arsenophenylglycin, because the latter drug 

 has proved so far to be the most successful specific for surra. 



Phenylglycin-arsinic acid was prepared by boiling the sodium salt of pamido- 

 phenylarsinic acid" (1 gram molecule) with monochloraeetic acid (2 gi-am 

 molecules) in aqueous solution for five hours on a reflux condenser. The addition 

 of sodium carbonate (0.5 gram molecule) did not noticeably improve the yield. 

 Hydrochloric acid (1 gram molecule) was then added to keep any unchanged 

 acid in solution. Although phenylglycinarsinic acid is only slightly soluble in 

 water or dilute hydrochloric acid, it sometimes persists obstinately in remaining 

 in solution and can then only be precipitated on cooling with ice and scratching 

 ■\vith a glass rod. Unchanged acid was reprecipitated from the filtrate and used 

 for further preparation. The residue, crystallized once more from dilute hydro- 

 chloric acid, was pure and free from pamidophenylarsinic acid : 

 I. 0.3248 gram substance gave 0.1812 gram MgoAsjO,. 



II. 0.2879 gram substance gave 0.1586 gram MgjASjO,. 



Calculated for C8HioN05As Found npr rpnt 



COoH . CH^ . HN . CoH, . ASO3H., ^ ^ tt 



Per cent. ^' ^^• 



A3=27.2 26.9 26.5 



It would be reasonable to expect that the carboxychloride could be obtained 

 by the action of phosphorus trichloride on the free acid, but, on the other hand, 

 it might also be possible that the acid chloride itself, especially during its 



" When a small quantity is treated at once with warm water, & considerable 

 quantity decomposes to its constituents. 

 "Chem. CentralU. (1909), I, 234 (pat.). 



