126 THE DEATH-POINT OF BACTERIA 
inoculation with living units, if maintained at a tempe- 
rature of about 70°F,; whilst when other more change- 
able fluids are inoculated, putrefaction ensues with 
equal certainty, though with much greater rapidity. 
What we have learned, therefore, concerning the 
invariable uniformity of simple inoculation experi- 
ments should of itself teach us how difficult it would 
be to account for cases of delayed putrefaction, or for 
cases in which a mere smouldering fermentation is set 
up, by the old though now well-nigh exploded notion 
of contamination by preexisting germs. Where 
living ferments really exist, the course of events is 
definite and almost invariable in its rapidity; but 
where fermentation takes place as a result of chemical 
changes occurring in the fluid itself (either by its own 
unaided powers, or under the stimulating influence of 
a less-heated organic ferment) there is abundant room 
for all the irregularity and variation actually en- 
countered. These cases of irregularity and variation 
have always, on other grounds, defied all legitimate 
attempts to bring them individually within the pale 
of a narrow and exclusively ‘vital’ theory of fer- 
mentation ; and now a wider experience with living 
ferments equally tends to show the impossibility of 
legitimately explaining a great mass of irregular 
phenomena by means of agents whose action is shown. 
to be constant and almost invariable. 
