of the injection. Lacy saw a number of these cysts aspirated, and the contents were found to be a sterile, oily fluid, with many leucocytes present. In one instance the primary reaction disappeared within a few days after vaccination, but recurred after four months and persisted for several weeks. NOTANDA.—The army lipovaccine apparently offers a certain definite amount of protection against pneumonia, which was the most dangerous complication of influenza. The protective substances do not appear in the serum until eight days have elapsed after the vaccination, and while no definite evidence has appeared to show that there is a temporary increase in susceptibility immediately after vaccination, the best results would undoubtedly be obtained where the dose is given something more than eight days before the appearance of the epidemic. The indications are that the vaccine will not protect against influenza, but that the complication of pneumonia is less likely to occur in the vaccinated than in the unvaccinated individual. _Summary_ Records of attempts to confer immunity to influenza by the use of vaccines have been separated into related groups and studied. Those where pure Pfeiffer strains were used have been considered in one group. Those where mixed vaccines were used have been analyzed in three sub-groups or series, depending on the relation between the times of vaccination and of the advent of the epidemic, upon whether or not a week-by-week comparison of the occurrence of influenza among vaccinated and unvaccinated groups was made, and upon whether or not statistics for total comparison alone were available. The third group included the reports of the use of army pneumo-lipovaccines for the prevention of the secondary pneumonia complications of influenza. _Conclusions_ From our statistics we conclude that: 1. There is as yet no evidence that vaccines composed purely of strains of Pfeiffer bacilli will confer immunity to epidemic influenza. 2. The only data which can be used as a basis for estimating the value of mixed vaccines as a preventive for epidemic influenza must be obtained from experiments in which vaccination was either completed before the epidemic appeared, or in which week-by-week comparisons between the number of cases occurring in the vaccinated and unvaccinated groups can be made. 3. Data obtained from experiments conducted under the above qualifications is inconclusive, but presents little evidence of the value of mixed vaccines in protecting against influenza. There is, however, an indication that mixed vaccines used prior to the arrival of the epidemic will lessen the number and the severity of secondary pneumonias, and will probably lower the death rate to a small degree.