In the practical application of any remedy the _dosage_ is most important. Unfortunately there is no accurate method of determining this with a vaccine. Wright recommended determining the number of the organisms per cc. as before mentioned, and his method or some modification of it is still in general use. From what was said with regard to variation, both in organisms and in individuals, it can be seen that the number of organisms is at least only a very rough guide. This is further illustrated by the doses of micrococcus (staphylococcus) vaccines recommended by different writers, which vary from 50,000,000 to 2,000,000,000 per cc. The author is decidedly of the opinion that _there is no way of determining the dosage of a vaccine in the treatment of any given case except by the result of the first dose_. Hence it is his practice to make vaccines of a particular organism of the same approximate strength, and to give a dose of a measured portion of a cubic centimeter, judging the amount by what the individual or animal can apparently withstand, without too violent a reaction. If there is no local or general reaction or if it is very slight and there is no effect on the lesion, the dose is too small. If there is a violent local reaction with severe constitutional symptoms clinically, and the lesion appears worse, the dose is too large. There should be some local reaction and some general, but not enough to cause more than a slight disturbance, easy to judge in human subjects, more difficult in animals. In cases suitable for vaccine treatment no _serious_ results should follow from a properly prepared vaccine, though the process of healing may be delayed temporarily. Wright claimed, and many have substantiated him, that always following a vaccination there is a period when the resistance of the animal is diminished. This is called the “negative phase,” and Wright considered this to last as long as the opsonic index remained low, and when this latter began to increase the stage of the “positive” or favorable phase was reached. As has been stated the opsonic index is pretty generally regarded as of doubtful value, though the existence of a period of lowered resistance is theoretically probable from the fact that antibodies already present in the blood will be partially used up in uniting with the vaccine introduced and that the body cells are called upon suddenly to do an extra amount of work and it takes them some time to adapt themselves. This time, the “negative phase,” is much better determined by the clinical symptoms, general and especially local. It is good practice to begin with a dose relatively small. The result of this is an indication of the proper dosage and also prepares the patient for a larger one. The second dose should follow the first not sooner than three or four days, and should be five to seven days if the first reaction is severe. These directions are not very definite, but clinical experience to date justifies them.