1. Regarding attempts at immunization by means of immune serum. The experiments of Nocard and Motas indicated that protective substances occur in the blood of dogs which have recovered from European piroplasmosis. These authors claimed that immune serum destroys the parasites and that it exerts a protective and curative effect upon the disease. Similar experiments conducted by Robertson in Cape Colony gave contrary results to those of Nocard and Motas. In the absence of further evidence no conclusions can therefore be drawn from these conflicting results. Two of Robertson's dogs which received injections of hyperimmune serum every day after they showed fever, died but no parasites could be found in their blood at death. This result suggests that a partial immunity may have been obtained by the treatment.

2. Regarding attempts at immunization by means of inoculations with blood containing dead parasites. Of the five dogs which we attempted to immunize in the manner we have described, three died of acute piroplasmosis; one died without parasites in its blood although parasites had been previously found; one dog died on the 36th day from chronic piroplasmosis. The absence of parasites in the blood of one dog at autopsy and the occurrence of chronic piroplasmosis in another dog may or may not indicate a partial acquisition of immunity. In any case the experiments afford no evidence that practical results are likely to follow further investigation of this character.

3. The duration of immunity following recovery is undetermined since there are no experiments to prove that animals are immune after the parasites have completely disappeared from their blood. In the socalled “immune,” “recovered,” or “salted” dogs, the animals harbour parasites in small numbers for weeks, months or years, consequently such dogs are suffering from a mild chronic form of piroplasmosis. Whilst subject to this mild form of the disease dogs usually escape acute infection when reinoculated with the parasite.

4. The parasites of piroplasmosis canis may persist in the blood of apparently recovered dogs for a considerable length of time, 6 months to 2 years, and so long as they are present in the blood the latter remains fully virulent for clean dogs. Consequently there is no evidence that the African Piroplasma canis becomes modified in its virulence during the course of chronic piroplasmosis. According to Nocard and Motas the European P. canis does become modified in its virulence in dogs suffering from chronic piroplasmosis. It should be noted that the European disease is a milder affection than the African.

5. The passage of the African P. canis through two series of dogs, that is through upwards of 90 animals in the course of two and a half years, has shown that the parasite may be communicated by inoculation from dog to dog for an indefinite period. No evidence was obtained that P. canis is in any way modified in its virulence by passage through dogs. In a number of dogs which were inoculated with post-mortem blood the disease developed more slowly than usual but nevertheless led to a fatal issue.

6. The onset of fever after inoculation with virulent blood may precede or succeed the appearance of P. canis in the peripheral circulation. Occasionally the disease may run its course to a fatal termination without the appearance of febrile symptoms.