=Lipovaccines.=--The suspension of bacteria in neutral oil was first used by Le Moignac and Pinoy who gave the name “lipovaccines” (λιπος = fat) to them. It was claimed that the reaction following injection of these vaccines was less severe than with saline vaccines in many instances; also, that the bacteria were much more slowly absorbed. For these two reasons it was hoped that much larger numbers of bacteria could be injected at one dose and one injection would suffice instead of three or more as ordinarily used. The technique of preparation, standardization and killing of the organisms has not as yet been sufficiently well established to warrant the general substitution of lipovaccines for ordinary saline suspensions. Vaccines are either “_autogenous_” or “_stock_.” An “autogenous” vaccine is a vaccine that is made from bacteria derived from the individual or animal which it is desired to vaccinate and contains not only the particular organism but the particular strain of that organism which is responsible for the lesion. Stock vaccines are made up from organisms like the infective agent in a given case but derived from some other person or animal or from laboratory cultures. Commercial vaccines are “stock” vaccines and are usually “polyvalent” or even “mixed.” A “polyvalent” vaccine contains several strains of the infective agent and a “mixed” contains several different organisms. Stock vaccines have shown their value when used as preventive inoculations, notably so in typhoid fever in man, anthrax and black-leg in cattle. The author is strongly of the opinion, not only from the extended literature on the subject, but also from his own experience in animal, and especially in human cases, that stock vaccines are much inferior and much more uncertain in their action when used in the _treatment_ of an infection, than are autogenous vaccines. This applies particularly to those instances in which _pneumococci_, _streptococci_, _micrococci_, and _colon bacilli_ are the causative agents but to others as well. The following are some of the reasons for this opinion: The above organisms are notoriously extremely variable in their virulence. While there is no necessarily close connection between virulence and antigenic property, yet since virulence is so variable, it is rational to assume that antigenic property is also extremely variable. Individuals vary just as much in susceptibility and hence in reactive power, and generally speaking, an individual will react better in the production of antibodies to a stimulus to which he has been more or less subjected, _i.e._, to organisms derived from his own body.