474 Journal of Agricultural Research voi. xiii. no. lo 



logical blood should have indicated to Dzierzgowski that this was too 

 short a digestion period, for only three years later Pick {12, p. 386), 

 using but 10 guinea pigs in his digestion experiments, correctly found 

 that two-thirds of the antitoxin was destroyed by trypsin if given time 

 enough — 9 days. Mellanby's statement {10, p. 405) that 0.2 per cent 

 hydrochloric acid has no destructive action on diphtheria antitoxin is at 

 variance with that of numerous workers who found that it was destroyed. 



On the practical side the net results of the numerous investigators, a 

 few of whom are mentioned in Table I, indicate the impracticability of 

 administering antitoxins by mouth. On the theoretical side they throw 

 very little light on the problem of the nature of the antibodies themselves. 

 The various investigators who attempted to separate a pure antitoxin 

 from associated protein naturally sought a method by which the pure 

 antitoxin would be obtained free or nearly free from protein, regardless 

 of whether the proteins were destroyed during the purification. The 

 converse did not occur to any of the workers whose publications have 

 been studied — that is, to separate the antitoxin from the associated pro- 

 teins, leaving the proteins entirely intact, but destroying the antitoxin. 



For the purpose of proving that the antitoxin is not of protein nature, 

 it is obviously immaterial whether one obtains pure antitoxin free from 

 protein or the pure protein free from antitoxin so long as the separation 

 actually is made. No one has succeeded in making the first separation; 

 the writers have succeeded in making the second. 



In the experiments here described the difi"erent antitoxic preparations 

 were exposed to the action of trypsin-sodium-carbonate or to pepsin- 

 hydrochloric-acid solutions for comparatively long periods of time, with 

 suitable controls. (See Table III.) The extent to which digestion took 

 place was then measured. The antitoxin remaining in the mixtures was 

 estimated by inoculation tests on guinea pigs, carried out with slight 

 modifications according to the methods described by Rosenau and 

 Anderson {13). Anthrax serum was studied first, but as the results 

 obtained were inconclusive, they are only briefly mentioned after the 

 data on tetanus serum and tetanus antitoxin. 



EXPERIMENTAL MATERIALvS 



Table II contains a brief description of the antitoxic preparations and 

 proteolytic enzyms used. 



Guinea pigs. — A total of 440 guinea pigs were used. These were 

 obtained from the Bureau of Animal Industry's stock of 3,000 or more 

 continually on hand at the Bureau's Experiment Station at Bethesda, 

 Md. These animals have been bred for years from selected hardy stock 

 and are undoubtedly very superior for experimental purposes. Only the 

 obviously healthy animals, weighing between 320 and 380 gm. were 

 selected at the station and forwarded to the Bureau's animal room 2 or 



