Bacterial and other toxins have proved to be of so complex a nature and subject to such alterations that a group of terms relating to them has developed. These terms are not devoid of confusion, owing to their multiplicity and similarity. The word toxin has come to have a limited and technical application. Ehrlich has concluded that it should be applied to those poisonous products of vegetable or animal metabolism by immunization with which one is able to obtain a specific antitoxin. Diphtheria and tetanus toxin, abrin, ricin, etc., are examples of the former, and snake poison of the latter. Toxins must also have a definite structure, that is to say, a haptophorous group by which they link themselves to cells and substances derived from cells, especially antitoxin, and a toxophorous group. This definition, then, excludes such poisonous substances as the vegetable alkaloids, as strychnin and morphin, and poisonous minerals,