8 
and bottom layers of the fluid may be found and misleading results 
obtained. Again, unless a reasonably accurate thermometer is used 
further inaccuracies may result, for ordinary thermometers not 
infrequently have errors of several degrees. 
Amons: bacteria some strains or races are more resistant to heat 
than others. These differences, which correspond to similar knovm 
variations in all animal and vegetable species, must be taken into 
account. 
The individual micro-organisms in any particular culture vary so 
far as their resistance to heat is concerned. Xot all the bacteria are 
killed at once. Thus it vdll be seen from the tables that the vast bulk 
of the bacteria succumb at a certain temperature, while a few hardy 
individuals resist several degrees higher heat or several minutes longer 
exposure. It is evident that when we reach the point where only a 
few organisms remain alive in the fluid, the results may be irregular 
or vary, depending upon the amount of fluid transplanted into the 
culture medium. 
The physical and chemical composition of the fluid in which the 
tests are made also exerts a decided influence. Some bacteria are 
sensitive to acid, others to alkali. If the tests are made in pure 
water, we have to deal vdth differences in osmotic pressure (plas- 
molysis). If the tests are made in an albuminous, viscid, or 
fatty medium, protecting envelopes may form and act as insulating 
coverings. 
Evaporation takes place so rapidh^ from exposed fluids that the 
smface layer may remain cooler than the body of the liquid. This is 
especially a matter of concern vdth milk, which sometimes forms a 
scum above 60 ° C., ovdng to rapid evaporation of the surface layer. 
This scum consists of coagulated albumins in which is enmeshed 
much fat. The bacteria entangled in this surface pellicle may escape 
the heat indicated by the thermometer in the deeper layers. 
A matter of very great importance is the suitability of the medium 
into which the bacteria are planted, as well as the temperatiire and the 
other conditions of incubation, etc. The optimum conditions ma^^ 
only be realized by a special study of each micro-organism tested. 
Controls must always show prompt and ^dgorous growth. 
The concentration of the suspension and other differences in the 
conditions of the experiment account for some of the variations in the 
results recorded by different observers. If a test tube is used, care 
must be taken that the glass be tliin and that the test fluid be well 
below the surface of the heated water. It is especially important to 
prevent soiling of the walls of the test tube above the fluid. This 
may give false results, for the reason that the pipette or instrument 
used to vdthdraw the milk or whateyer fluid is used may come into 
contact with the upper portion of the test tube and thereby pick up 
