606 
ANNALES DE L’INSTITUT PASTEUR 
It would appear that the number of bacilli bas a greater 
influence on the size of the lésions tban on the spread of infec¬ 
tion. Time lias greater influence on the spread. 
By the method previously described, it is possible to show 
that under the influence of certain agents the normal viru¬ 
lence of tuhercle bacilli may he considerably reduced. It is 
also possible to demonstrate that the virulence of bacilli 
obtained at different times from sources of a similar kind is 
capable of a considérable variation. 
Thus a certain strain of tubercle bacilli, obtained from a 
human lésion and exhibiting normal virulence so long it was 
cultivated on blood sérum at the optimum température, pro- 
duced only slight and slow spreading lésions after being kept 
absolutely dry in the dark for 2 months. After dessiccation 
for a period of over 4 months, the same bacillus was inca¬ 
pable of producing any infective lésion. Other cultures were 
affected in the same way, but some were able to resist des¬ 
siccation for longer periods. Pathological products such as 
tuberculous sputa, pus, milk are affected in a similar way by 
drying. 
The action of radiant sunlight is so rapidly léthal that it is 
not easy to regulate the exposure so as to détermine accu- 
rately the réduction in the virulence of the bacilli. 
A réduction of virulence was also observed in the case of 
tuberculous milk kept at a low température (below 6°G.) in the 
dark for a considérable finie. This milk, which hadbeen col- 
lected with aseptie précautions and direct from a tuberculous 
udder in a sterilized vessel, conlained a large number of viru¬ 
lent tubercle bacilli and a few other hacteria. When fresh it 
produced extensive uncomplicated tuberculosis in cobayes 
inoculated subcutaneously with it. The lésions were as usual 
extensive in ail the lymphatic ganglia connected with the 
seat of inoculation 3 weeks alter the inoculation. After this 
milk had been kept for 490 days (1), as stated above, it was still 
(l) The same milk was tested again after being kept 4 1/2 years, when it 
was found that the bacilli were ail dead or at any rate incapable of produ¬ 
cing tuberculosis. Cobayes inoculated with large quantities of this milk, 
r.emained well for over 1 year after inoculation and ail their organs were 
found healthy post mortem. 
