1. In our hands the assay of a plague vaccine, in terms of its power to protect mice against what was hoped to be a standard challenge, was found to be impracticable.

2. Assay by mouse protection tests in terms of a Standard vaccine was found to be possible, but it proved to be a very time-consuming procedure and demanded the use of an almost prohibitively large number of mice for a sufficient number of replicate tests which are required to offset the variability in the method. Theoretically a single test might be adequate if the numbers of immunized mice were very large, but there are obvious limitations to the numbers of mice that can be included in a single test of this nature.

3. The Ouchterlony plate technique proved to be useful in indicating the antigens present in a particular vaccine, but not on a strictly quantitative basis.

4. The vaccine prepared from the virulent human strain of P. pestis was found to protect mice more efficiently than those prepared from avirulent strains and when the protective potency was expressed in terms of content of antigen and high molecular weight substances the vaccine prepared from the virulent culture was found to be even more potent than the Standard acetone-dried vaccine.

5. The use of formalin as a preservative results in some loss of potency of a vaccine but, provided the strain of P. pestis used for the preparation of the vaccine is highly virulent, this need not be serious and is offset by a number of advantages.