After the vaccine was completed, cultures were made from the final dilutions and were watched for 48 hours. Mice and guinea pigs were injected with the first samples to make certain that the material was non-toxic. Two laboratory employees also volunteered and received full doses before the first batch of vaccine was released. The first five litres were turned over to the Red Cross on October 31, one week from the day the work was begun. In three more days the laboratory reached a capacity of 10 litres a day, and on the fifth day the order was received to discontinue preparation of the vaccine. Relatively little of our vaccine was given out, and in the rush it was not possible to determine which physicians had been given our vaccine and which had received commercial mixed products, so there is no data on its protective powers. As soon as we found that there was no call for prophylactic vaccines, we planned some animal experiments; but inasmuch as we were unable to get our cultures of Pfeiffer bacilli virulent enough to kill mice or guinea pigs, the minimum lethal dose could not be determined, and without it it was impossible to determine the protective value of the vaccine. Mr. Purwin, in our laboratory, injected a 25–gram mouse intravenously with 2 c.c. of a milk thick suspension of Pfeiffer bacilli without killing the animal. He was successful in getting a small needle into the tail vein and in slowly injecting the whole amount. The mouse was sick for about 36 hours, but entirely recovered. Guinea pigs were insusceptible to very large doses. Had we succeeded by means of a vaccine in completely immunizing a man against Pfeiffer bacilli, we still would have been uncertain that he was immune to influenza in its “epidemic” form. The absence of virulence in our laboratory strains may not mean that the cultures were non-virulent when first isolated, but it suggests the uselessness of attempting to make active vaccines from strains kept on artificial media for months or years, such as those commonly offered for sale by commercial houses. The loss of virulence in strains that have been isolated for some time is interesting in the light of Parker’s (12) work upon toxine production by Pfeiffer bacilli. She found that toxic filtrates appeared in infusion broth cultures in from 6 to 8 hours, and that 2 c.c. of a 20–hour filtrate would kill a medium-sized rabbit in from 1 to 3 hours. It was also found that the poison deteriorated so rapidly that, in order to determine its toxicity, the tests had to be made on the same day that the filtrate was obtained. Parker succeeded in making an anti-serum against the poison, which appeared to be antitoxic for it both in vitro and in vivo. This work is interesting, and may be a step toward the development of a practical prophylactic serum. _Conclusion_