The facts already noted indicate that acquired and induced immunity have a different explanation from natural immunity. It is not here phagocytosis or direct cellular action, but the production of antitoxines. It is consistent with this that the serum of animals possessing natural immunity has ho antitoxic effect. Nevertheless, the antitoxines are the products of cellular activity. They are secretions from the cells. They have been determined chemically to be albuminoid substances, belonging to the same class of substances as some of the toxines (toxalbumins). Not only so, but there is reason to believe that the antitoxines do not neutralize in a chemical sense the toxines, and that when introduced into the body they act, in part at least, by stimulating the cells to further production of antitoxines. This view is supported by the fact that a mixture of antitoxine and toxine which is innocuous in one animal may be toxic in another, a fact that we should not expect were there a proper chemical neutralization of the toxine (see especially Roux's articles).

The most probable theory of both acquired and induced immunity is that the presence of the toxine induces the cells of the body to produce an antitoxine, which as an active albuminoid antagonizes the action of the toxine. This affords an explanation of the Periodicity of certain infective diseases. Such diseases as typhoid fever, typhus fever, pneumonia, cholera are known to run their course and come to a crisis in a certain number of days. The explanation of this seems to be that on the average a sufficient amount of antitoxine is produced by that time to antagonize the toxine. It may be perhaps inferred that, in the case of diseases which possess no periodicity, and in which one attack confers no immunity, antitoxines are not produced and treatment by blood serum is not available. This applies to tuberculosis, leprosy, etc.