488 Journal of Agricultural Research voi. xm, no. lo 



potency, partly to allow for deterioration with time and partly in the 

 desire to furnish promptly the product when requested. In a similar 

 way, antitoxin X and serum 268 were found to have potencies about 

 50 per cent higher than that stated. The other preparation used, 

 antitoxin 420F, stated to contain 300 units per cubic centimeter, pro- 

 tected guinea pigs for 96 hours in o.i unit doses, but not in 0.07 unit 

 doses (see Table VI). This preparation had received a final and correct 

 standardization before shipment. Therefore all the doses in Table V 

 should be multiplied by 10/7. In mixture B the only guinea pigs that 

 lived more than 96 hours were those that had received a quantity of 

 serum calculated to contain 0.80 unit if none of the potency had been 

 destroyed. The results with mixture A show that this dose contained 

 originally, at least, 0.80 times 10/7, or i . 14 units. At the time of injection 

 this dose (B) contained at least o.i unit, because the guinea pigs were 

 protected for over 96 hours. Hence, at most, the destruction was 



1.14 — 0.1^ or 91 per cent. The guinea pigs that received the next 



1. 14 

 lower dose — that is, 0.40 times 10/7, or 0.57 units — died in less than 96 

 hours; therefore this dose contained less than 0.1 unit, or, over 0.47 out 

 of 0.57 unit was destroyed (82 per cent). Consequently in this experi- 

 ment over 82 but less than 91 per cent of the antitoxin in mixture B 

 was destroyed. The higher figure was used throughout. 



In Table VI the results with mixture A show that the doses actually 

 contained the number of units stated in the table. In mixture D, for 

 example, the 0.8-unit dose contained less than 0.1 unit, or the destruction 



was more than — — — ^ , or 87 per cent. It is believed that the figures 



for percentage of antitoxin destroyed may be incorrect by a few points, 

 but this is not of importance. 



Comparison of antitoxin destroyed with protein digested. — 

 Twelve inoculation tests were made, but all are not recorded here in 

 detail, because Tables V and VI are thoroughly typical of the others. A 

 discussion of the results of experiments 5 and 2 1 will apply to the others, 

 because of their general concordance. With the exception of experi- 

 ments I and 2, which were preliminary, the remaining 10 experiments, 

 in which 360 guinea pigs were used, are summarized in Table VII. 



