QUININE ESTERS OF PHENYLARSINIC ACID DERIVATIVES. 27 



rine of course is easily removed by heating the body in dilute alcoholic alkali; 

 for then the ester is saponified, the benzarsinic diehloride changing to benzarsinious 

 acid and the hydrochloride of the alkali. The presence of benzarsinious acid in 

 the solution of the chloride makes no difference; its silver salt, although insoluble 

 in water, is readily soluble in dilute nitric acid. This affords another quick 

 method for the chlorine determination. 



Analysis. — Two and a half hours heating to 320° with nitric acid (specific grav- 

 ity 1.5) was sufficient for complete decomposition; the filtrate from the chlo- 

 ride of silver was used for the determination of arsenic after the excess of silver 

 nitrate had been removed. 



I. 0.2351 gram substance gave 0.0617 gram MgjASoO, and 0.0383 gram AgCl. 

 II. 0.2349 gram substance gave 0.0621 gram Mg,AsA and 0.0368 gram AgCl. 



Calculated for C07H27N2O3ASCI2 

 Per cent. 



Found per cent. 

 I. II. 



As=13.1 



12.4 12.7 



CI = 12.2 



11.4 11.8 



When tested in the Biological Laboratory, in continuation of the re- 

 search by Strong and Teague ® on chemotherapeutic speciiies against surra, 

 this compound showed an exceedingly high toxicity for trypanosomata in 

 vitro. This is as would be expected, since in its chemical character it 

 corresponds closely to pamidophenylarsinic oxide ; but, on the other hand, 

 it was equally toxic for the cells of the host, in vivo, and so the drug could 

 not be used therapeutically in this connection. 



It was therefore oxidized on the one hand to the arsinic acid and re- 

 duced on the other to the arseno-derivative. The chemical character of 

 the arsenic in the former would correspond-to that in atoxyl, in the latter 

 to that in arsenophenylglycin. 



A number of oxidizing agents were tried but without satisfactory re- 

 sult. Iodine, which has been used before in several cases," is excluded, 

 since in evei-y solvent that was used it combined with the alkaloid to form 

 periodide before it reacted as an oxidizer. Red mercuric oxide, in several 

 experiments with small quantities, gave rather satisfactory results when 

 suspended in water and heated with the finely ground ester. However, 

 as the primary product is so exceedingly toxic and there is no way of sep- 

 arating the unoxidized from the oxidized product without loss, it is 

 important to oxidize completely. The usual analytical tests are not 

 sufficiently sensitive to detect 1 to 2 per cent of unoxidized substance, 

 so that the very convenient physiological test was employed. Ehrlich 

 found in test tube experiments that a 1 per cent solution of pamido- 

 phenylarsinic acid has no action whatever on trypanosomata, whilst 

 pamidophenylarsinious acid acts under identical conditions in dilutions 

 up to 1 to 100,000. My arsinic diehloride derivative under the same 

 conditions was found to act in a dilution of 1 to 20,000, but when it was 

 completely oxidized its action corresponded only to the toxicity of its 



* Loc. cit. 



''Ehrlich and Berthenn, Ber. d. deutschen chem. Ges. (1910), 43, 917. 



