If we inject a serum, containing a toxin, into the blood stream of a susceptible animal, certain things happen. The animal will become ill, but, provided that the amount of serum which has been injected was not too great, it will recover. If the toxin be again injected a reaction occurs, but the animal does not become so ill as on the first occasion, and after a number of injections the dose administered may be so great as to kill a susceptible animal but may yet produce no effect on the animal which is the subject of the process of immunisation: immunity has been conferred on it. Now can we compare the two operations, that of the solution and precipitation of the metal and that of the immunisation of the animal? We can to some extent, but the analogy soon fails, and indeed we should not attempt to formulate a theory of immunity on a physico-chemical basis if we did not start with the assumption that the series of operations was one in which only physico-chemical reactions were involved, that is to say, there is nothing in the phenomena of immunisation that suggests that what occurs in the animal body is similar to what we can cause to occur in inorganic materials outside the tissues of the living organism. We start with the assumption that the administration of the toxin causes the formation of an antitoxin in very much the same sort of way as the administration of hydrochloric acid to a solution of nitrate of silver causes the formation of chloride of silver. This antitoxin then neutralises the dose of toxin which may be administered after the process of immunisation has been effected, very much in the same sort of way as a certain amount of some acid can be neutralised by an equivalent amount of some base with which the acid can combine. If the reader will analyse any of the theories of immunisation current at the present day he will find that these are the physical ideas that are involved in it.[4] But physiological science has the much more formidable task of explaining the persistence of the immunity. The animal rendered immune to the toxins produced by certain species of bacteria may remain so for many years, that is, for a very long time after the antitoxins originally produced by the reaction of the tissues to the toxins first administered have disappeared. We must imagine, therefore, that the anti-substances produced originally by the reaction of the toxin are produced again and again by the tissues of the susceptible animal, for the latter may resist repeated infections, that is, repeated doses of toxin, without illness. But then the tissues of the animal body are transitory substances and they do not persist unchanged. Muscles, glands, connective tissues, even nerve-fibres and nerve-cells undergo metabolism, and the chemical substances of which they are composed break down into the excretory products, pass out into the blood stream, and are eliminated from the body; while at the same time these tissues are continually being