G. H. F. Nuttall and G. S. Graiiam-Smitii 223 
words the chronic form of the disease as a rule protects against the 
acute form of piroplasmosis. We have no experimental evidence that a 
true immunity follows upon complete recovery from piroplasmosis. 
(See Summary (3) on p. 227.) 
3. The persistence of parasites in apparently recovered dogs. 
That Piroplasma canis may persist for a considerable time in salted 
dogs, or those which have apparently recovered, has been amply proved. 
Thus Robertson (1902, p. 682, fully cited by Nuttall, 1904, p. 246) 
observed that a salted dog’s blood remained virulent for 4^ months after 
recovery (4 Nov. 1901 to 12 March 1902). He also reports a case (1906, 
p. 110, Dog 11) in which a dog’s blood remained virulent for other dogs 
for a period of two years (17 Nov. 1901 to 4 Dec. 1903) and adds, “all 
this time the animal had been kept in a run where it was free from any 
infective influence.” This dog’s blood was tested for virulence on 
seventeen occasions during the two years, and in every case inoculation 
practised with its blood produced fatal piroplasmosis in fresh dogs. The 
dog’s blood proved no longer infective on 11 Feb. 1904. In the case 
of another dog (No. 45) the blood remained infective for 13 months 
after apparent recovery. 
In the only two instances in which we have had an opportunity of 
observing the occurrence of spontaneous or natural recovery in dogs 
infected with the South African disease the blood of the apparently 
recovered animals remained infective for periods of four and six months 
respectively but not longer. After these periods the dogs could be 
regarded as recovered animals in the strict sense. 
In all of the above cases the blood of the dogs remained fully 
virulent as long as it remained infective, for it produced an acute and 
fatal disease when it was injected into fresh dogs. In other words there 
is no evidence that the parasite which causes South African piroplas¬ 
mosis in dogs becomes in any way modified in its virulence in animals 
suffering from the chronic type of the disease. 
Robertson’s (1906, p. 113) statement that keepers of hounds or packs 
of dogs believe “ that if an animal recovers from malignant jaundice 
he should be housed apart from his fellows for the rest of the season,” 
affords a piece of collateral evidence that animals which are recovering 
from piroplasmosis are a potential source of danger to others. 
The facts above cited are in direct contradiction to the statement of 
Nocard and Motas (1902, p. 273), but it must be remembered that they 
