ADMINISTRATION OF SALVARSAN AND NEOSALVARSAN 873 



with the production of amino-oxyphenyl-arsenoxid. This arsenoxid 

 might be present in the drug through insufficient reduction, although 

 with care this is not apt to take place. 



The commercial product always contains some arsenoxid, but it is 

 usually less than 1 per cent. Ehrlich states that the best preparations 

 will contain from 0.5 to 0.8 per cent, of arsenoxid. 



Ehrlich and Bertheim 1 state that the toxicity of arsenoxid is very 

 much greater than that of salvarsan. Like other arseno compounds, 

 the hydrochlorid of dioxydiamino-arsenobenzol possesses the property 

 of readily undergoing oxidation. Exposed to the air it will soon contain 

 amino-oxyphenyl-arsenoxid; indeed, the production of arsenoxid com- 

 pounds takes place if the preparation is kept in an ordinary glass 

 container. This fact is, therefore, of the greatest importance in the 

 practical use of this remedy, because the amino-oxyphenyl-arsenoxid is 

 about twenty times more poisonous than the pure hydrochlorid of the arseno 

 compound. 



Variations in the Medicament and their Relation to the Immediate 

 Reactive Symptoms. We are firmly convinced that the complex of 

 symptoms classed under group a, arid characterized by flushing, edema, 

 etc., and followed by pallor, and in rare instances by syncope, are due 

 to the drug administered and not to extraneous causes. Various descrip- 

 tive adjectives have been applied to this syndrome, such as vasomotor, 

 angioneurotic, anaphylactoid, and nitritoid. The milder grades of re- 

 action bear a resemblance to the symptoms following the use of nitrate 

 of amyl, and the severer types simulate rather closely the picture of 

 anaphylaxis. Indeed, Swift maintains that they represent true anaphy- 

 lactic phenomena. Swift 2 demonstrated that guinea-pigs which have 

 been sensitized by the injection of a mixture of guinea-pig serum and 

 salvarsan, and have been reinjected after a suitable time with the same 

 mixture, show symptoms like those seen in anaphylactic shock. Swift 

 believes that this phenomenon depends on an alteration of the native 

 serum by salvarsan so that the homologous serum acts like a foreign 

 protein. It is pointed out that the symptoms resembling those of ana- 

 phylaxis which at times follow the administration of salvarsan, occur 

 usually not after the first but only after one or more injections. 



The vasomotor symptoms above mentioned we believe to be due 

 to something in the drug which directly or indirectly (perhaps through 

 anaphylaxis) induces a paresis of the blood-vessels, characterized by dila- 

 tation and not infrequently by leakage of the serum into the tissues. We 



1 Berichte d. Deutch. Chem. gesell., vol. 45, 1, 1912, p. 764. 



2 Jour. Amer. Med. Assoc., 1912, lix, 1236. 



