81 
GENERAL SUMMARY AND CONCLUSIONS. 
It is difficult to determirie precisely at what moment or at what 
temperature a micro-organism dies. The fact tliat a micro- 
organism ^\^I1 not grow upon artificial media is not always a sure 
sign that it is dead; but ^^*ith the exception of the tubercle bacillus 
it is safe to assume that ordinarily bacteria that fail to vegetate 
upon suitable media imder favorable conditions have at least lost 
their virulence and power to infect, especially when ingested by the 
mouth. 
We know, however, that the vegetability of micro-organisms in 
vitro does not always correspond to their ability to grow in the animal 
host. This is especially true of the tubercle bacillus. On account 
of the reluctance with which this bacillus grows upon culture media, 
it is necessary to resort to animal inoculations in order to determine 
its thermal death point. It is proper to assume that the tubercle 
bacilli in milk which are so attenuated as to be unable to cause 
tuberculosis when injected into the peritoneal cavity of a young 
guinea pig would be harmless when ingested by man. 
In these experiments upon the gumea pig it is important to differ- 
entiate the lesions produced by dead tubercle bacilli, which closely 
simulate those caused by the live tubercle bacilli. In doubtfid cases 
it is necessary to inoculate the products of the lesions into another 
animal to determine the presence or absence of liffing micro- 
organisms. 
The cultural experiments upon thermal death pomts are sur- 
rounded by many sources of error and numerous pitfalls, all of which 
must be avoided. 
My experiments were designed to imitate the conditions of practical 
pasteurization. The micro-organisms were heated in open test tubes 
and the temperatiu'e and other factors accmately controlled. Scum 
formation was disregarded, as it was my intention to reach results 
that might be applied vdth confidence in practical pasteurization on 
a large scale against natural difficulties. 
My results of rune series of tests upon guinea pigs with five cidtru’es 
plainly show that m milk the tubercle bacillus loses its virulence and 
infective power when heated at 60° for twenty minutes : in other words, 
it may be considered dead. When heated to 65° C. a much shorter 
time is necessary. These results correspond in the main with those 
previously recorded by Yersin, Bonlioff, Schroeder, Th. Smith, Russell 
and Hastings, and Hesse. 
It should be remembered that the milk m these tests was very 
heavily infected vdth ^ffiulent cultirres, indicated by the prompt 
death of the control annuals. Milk practically never contains such 
an enormous amoimt of infection imder natural conditions. It is 
