154 The Philippine Journal of Science 1915 
corded in the tables, but are not included in the averages. Ani- 
mals 6078 and 5983 are included in the averages of Tables II 
and III, respectively. 
If we compare the several groups of series 2, Table II, we 
have: 
TABLE VI.—Average survival of guinea pigs inoculated with different strains 





of bacilli. 
pe 
| | Aver- 
2 4 | Ani- age 
Bacillus strain. oily pad 
vival. 
| SS Se eS = | 
| / Days. | 
let bag ial tyes a) 11 | 335.7 
Wel aye yt) ayaa eee Se 7 | 283.1 
eo D premise eee | 4 | 291.3 
“Tb emuls b” ____-_-. 3 | iad 


If we compare the average weights of controls with those 
of the treated animals, both taken at the same time of inoculat- 
ing the virulent dose, we have, including only animals which 
subsequently died of tuberculosis: 
TABLE VII.—Average weights of treated and of control guinea pigs. 




} 
| 
| Series. ae | cs | 
weight. | 
ae = Ly | 
| | Grams.| 
Lh bn: Bees 9 | 544.4 | 
ee ee 25 | 573.2 | 
eee eee 11 | 595.0 
| contol Ustad 10 | 539.0 
In view of these averages it is evident that the greater re- 
sistance of the controls was not due to a selection of larger 
animals. 
Considering the averages of all series, it appears that the 
preliminary treatment with avirulent bacilli in agar has afforded 
no protection against a subsequent dose of virulent bacilli. If 
of any effect, it has apparently tended to diminish the resistance 
of the animals. The avirulent bacilli, without agar, has also 
failed to immunize, so far as can be judged by the comparatively 
small number of animals in this series. 
In the case of the two animals which survived the virulent 
dose over two years, however, there is some evidence of partial 
immunization. Their weights taken at the time of the inocula- 
