The size of the scarification is important. The vesicle is always somewhat larger than the scarification, and the larger the vesicle the greater danger that the surface may be broken, and the more opportunity there is for the introduction of extraneous infection. A spot as large as the head of a medium pin is about as small as can be easily scarified, and vesicles formed on such scarifications are least liable to have inflammatory complications. If, as certain evidence tends to show, a larger area of scar guarantees greater protection, and if a larger area is therefore desired, it is better to vaccinate in two or three small spots than in one large one. It is somewhat difficult to rub the virus from a bone point on a spot of the minute size described, and as this form of virus is usually more dilute than glycerinated virus, a larger area may safely be employed. The scarification may be made with any sharp instrument, or with the point itself. The only precaution necessary is that the instrument should be free from infection. As a scarifier the ordinary cambric needle presents the advantages that it is usually clean, is easily sterilized, and is so inexpensive that a fresh one can be used for every operation. It is not necessary that the scarification should draw blood, although blood is not objectionable unless it flows so freely as to wash away the virus, or unless the subject has hæmophilia. Although with a notably susceptible subject or with especially active virus it may be sufficient simply to smear the virus on the scarified area, it is usually necessary and always advisable to rub in the virus with a wooden slip or with the point firmly and thoroughly. Other methods of introducing vaccine virus are by puncture, by deep injection, and by the mouth. In the method by puncture either a grooved lancet or a hollow needle may be used. A shallow puncture is made and the virus is deposited in it. The resulting vesicle is usually small and nearly circular, and generally remains free from infection; but as the hole in which the virus is placed is small, it is possible that the issuing blood may wash it away completely, and the percentage of success with this method of inoculation is not quite so large, even in careful hands, as by the process of scarification with the same virus. Animal experiments with deep injection of virus through a hypodermic syringe and with administration of virus by the mouth show that there is no certainty of successful vaccination by these means, and that when success results there is no proof of it without a subsequent vaccination on the skin to test or to demonstrate the immunity.