The dogs that are still healthy should have spacious, well-aired dwellings, open air exercise (as much as possible in the fine season), good but not too stimulating food (in part at least fresh animal food), pure water, and protection against undue fatigue, cold, icy baths, especially when exhausted, rain or snow storms and cold stone or metallic beds. In the cold season artificial heat in the kennel is desirable. _Immunization_ may be sought in various ways based on the use of the toxins and antitoxins on the one hand, and of a weakened type of virus on the other. Bryce (1882) and others inoculated with the blood and pulmonary exudate, and produced in three-months-puppies, local swellings mainly, with subsequent immunity. The mortality from the inoculation did not exceed 10 to 15 per cent. These losses imply that in certain cases the material inoculated conveyed the microbe of the disease, and the survivors acquired all that immunity which comes from a first attack. In the cases that show local lesions only, it may be presumed that few microbes or none were inserted, while the results came mainly from the toxins or antitoxins. This would be entirely in keeping with Lignieres’ observation that the blood and pulmonary lesions often failed to furnish the pathogenic microbe, as tested even by attempts at artificial cultures. The protection secured from the antitoxins alone is shortlived, terminating with the elimination of these elements, while that coming from the action of the toxins on the leucocytes, and the stimulation by these to the production of defensive products, is much more lasting and in ratio with the quantity of the stimulus introduced and the profundity and duration of its influence on the leucocytes. This may partly explain the occasional early exhaustion of the immunity and the reinfection of the animal within a year after inoculation.