538 
Proceedings of the Royal Society of Edinburgh. [Sess. 
diminution in virulence — least marked in nutrose and agar cultivations, 
most marked in milk cultivations. So gradual a process is this loss of 
virulence that the minimal lethal dose of a subcultivation from a year-old 
culture may only be twice as large as that originally noted. Rapid sub- 
cultivations upon artificial media effect a more rapid diminution in virulence, 
especially when combined with incubation at or near the upper limits of 
the temperature range of the organism. It is, I believe, to this tenacity 
with which M. melitensis retains its virulence, almost as much as to the 
absence of specific bactericidal substances in the normal blood serum, 
that the bulk of the laboratory infections in research workers must be 
attributed. 
Exaltation of Virulence. 
Durham ( Journ . Path, and Bact., 1899, vol. v. pp. 377-88) and myself 
(Pep. Med. Fev. Com., 1905, vol. ii. pp. 67-80) showed that after intra- 
cerebral passages the virulence of M. melitensis became so exalted for 
rabbits and guinea-pigs that acute and rapidly fatal infections could readily 
be produced by this method of inoculation. Continuing my experiments, 
I found that a strain of M. melitensis thus exalted in virulence for the 
guinea-pig was practically as virulent for the rabbit when introduced into 
the brain of this rodent; but although highly virulent (in proportion to the 
number of passages) for the guinea-pig intraperitoneally and subcutaneously, 
further passages through the brain of the rabbit were necessary to render 
it highly virulent for the rabbit by intraperitoneal or subcutaneous methods 
of inoculation. In like manner I found that M. melitensis thus exalted 
was highly virulent for the white rat and the mouse when injected intra- 
cerebrally ; but again passages through the tissues of the central nervous 
system of these animals respectively were needed before the coccus would 
provoke a fatal infection when introduced into the peritoneal cavity or the 
subcutaneous tissues of these small rodents. 
Carbone, too, by a series of intravenous passages (five in number) through 
young rabbits, was able to so exalt the virulence of a strain he had isolated 
as to render it highly pathogenic for the guinea-pig when injected intra- 
peritoneally ; and this observer was the first to note the suppurative inflam- 
mation of the tunica vaginalis which follows intraperitoneal injections of 
the organism when a small male guinea-pig is employed — a phenomenon 
quite comparable to that observed under similar conditions when inoculat- 
ing B. mallei. Previous to this I had attempted to infect the male of the 
common goat, with, as I considered, negative results ; subsequently, however, 
Zammit showed that the goat was a susceptible animal, and Horrocks, 
