To be of value, antitoxins must be used as early as possible, before tissue change has occurred and before the toxins have, so to speak, got the upper hand. When the toxins are in the ascendency the patient suffers more and more acutely, and may succumb before there has been time for the formation in his own body of the antitoxins. If he can be tided over the "crisis," theoretically all will be well, because then his own antitoxin will eventually gain the upper hand. But in the meantime, before that condition of affairs, the only way is to inject antitoxins prepared in some animal's tissues whose disease began at an earlier date, and thus _add_ antitoxins to the blood of our patient, early in the disease, and the earlier the better, for, however soon this is done, it is obvious that the toxins begin their work earlier still. It should not be necessary to add that general treatment must also be continued, and indeed local germicidal treatment, _e. g._, of the throat in diphtheria and the poisoned wound in tetanus. Further, in a mixed infection, as in glandular abscesses with diphtheria, it must be borne in mind that the antitoxin is specific and may therefore probably fail in such mixed cases. After these preliminary remarks we will now consider shortly some of the methods employed for the production of antitoxins. An animal is required from whose body a considerable quantity of blood can be drawn without injurious effect. Moreover, it must be an animal that can stand an attack of such diseases as diphtheria and tetanus. Such an animal is the horse. Now, by injecting into the horse (_a_) living organisms of the specific disease, but in non-fatal doses, or (_b_) dead cultures, or (_c_) filtered cultures containing no bacteria and only the toxins, we are able to produce in the blood of the horse first the toxins and then the antitoxins of the disease in question. The non-poisonous doses of living organisms can be weakened, or, as we say, attenuated, by various means. Dead cultures have not been much used to produce immunity except by Pfeiffer. In actual practice the third method is much the most general, viz., filtering a fluid culture free from the bacteria, and then inoculating this in ever-increasing doses. The preparation of diphtheria antitoxin may be taken as an example, but what follows would be equally applicable to other diseases, such as tetanus.