In experiments carried out on animals this immunity to bee poison has been also induced by repeated application of the irritant. It was formerly generally supposed that the irritant nature of a bee's sting was due to the presence of formic acid; but inasmuch as bee poison can retain its poisonous character in spite of being submitted to heat, which would effectually volatilise the formic acid present, this assumption must be abandoned, and opinion is more inclined now to regard this irritant substance as partaking of the nature of an alkaloid. Before closing this brief review of some of the most recent discoveries which have been made in the domain of immunity, we must mention some extremely suggestive and important researches on the poison of tetanus, or lock-jaw, which have emanated from Dr. Roux's laboratory at the Institut Pasteur in Paris. It will perhaps be remembered that Pasteur, when working at hydrophobia, experienced the greatest difficulty in exciting rabies in animals with certainty, and that it was only when the fact of its being a disease which essentially affects the nervous system of the animal was taken into account that it occurred to him to cultivate the virus in the medium for which it had seemingly the greatest affinity, viz. the nervous tissue of an animal; it was only on taking this step that he succeeded in invariably provoking rabies in the animals under experiment. In the case of tetanus we have another disease affecting the nerve-centres of the body, and although many authentic cases have been cited in which the treatment with anti-tetanic serum has been entirely successful, a great many instances have occurred in which it has been of no avail at all, more especially when the disease has obtained a firm hold on its victim. Now Dr. Roux has not only been carrying out experiments to ascertain what is the result of directly attacking, as Pasteur did in the case of rabies, the nerve-centres of an animal with the tetanus toxin, but he has also taken another and very important step further, and has investigated, not only the action of the toxin, but also that of the anti-toxin on the nerve-centres of an animal suffering from tetanus. In describing the cerebral inoculations which he has conducted on animals, Dr. Roux points out that the operation, in itself, is attended with no pain or even inconvenience to the animal in question, that subsequently it eats with its usual appetite, and shows no signs of discomfort. First, as regards the infection of an animal with the tetanus virus introduced directly into the brain, it has been found that very much smaller quantities produce a fatal result than when subcutaneously inoculated. Thus, a rabbit which received two cubic centimetres of the poison under the skin took four days to succumb to tetanus, whilst one-twentieth of the quantity inoculated into the brain sufficed to kill another rabbit of the same size in less than twenty hours.