STUDY OF DELAYED OH « LATENT » TUBERCULOUS INFECTION 611 
2. Cobayes of the same size and kept under identical condi¬ 
tions as the last, after being inoculated with a much greater 
quantity of the same tuberculous products previously heated 
as stated above, developed very slight lésions at the seat of 
inoculation in the course of 11 weeks. 
3. The viscera and the majority of the lymphatic ganglia of 
the cobayes inoculated with the untreated tuberculous pro¬ 
ducts, were the seat of extensive tuberculous lésions 6 weeks 
after inoculation. 
4. In the cobayes inoculated with the heated tuberculous 
products, the viscera and lymphatic ganglia were alîected 
nearly (o the same extent as those of the cobayes inoculated 
with the untreated matériel, but this resuit was obtained after 
11 weeks only. 
5. From these results it would appear that the heated 
bacilli were at first barely capable of overcoming the résis¬ 
tance offered to them by the phagocytes in the région first 
infected, but. that some bacilli had escaped and had been 
carried away probably by some of the migratory cells. 
The se bacilli had been able to multiply and to regain their 
pathogenic power. 
6. The facts observed seem also to indicate that the récu¬ 
pération of this pathogenic power became more marked as the 
distance from the seat of inoculation increased, for, while the 
local lésions remained insignifiant, those of distant organs 
were practically as extensive as those found in the animais 
inoculated with untreated milk. 
7. The relation between the extent of the tuberculous 
lésions and the distance from the seat of inoculation, indicates 
that the bacilli which had successfully overcome the various 
obstacles offered to their progress, had gradually become not 
only more numerous but also more virulent. This suggests 
strongly a process of reinforcement of virulence by relays , 
similar to that observed in the production of a “Virus fixe” by 
passage through a sériés of animais according to Pasteur’s 
method. 
8. In the experiments referred to in this paper, the period 
of latency did not exceed 3 or 4 weeks, but it was sufficiently 
long to reveal an aspect of infection processes which must be 
