Tur INFLUENCE OF Low TEMPERATURE ON Sort Bacteria 1049 
unlikely, however, for if such was the case the count following the thawing 
of the ground in the spring would not be so nearly the same as it was 
before the freeze. He attempted to explain this increase as due to a summer 
and winter flora but was unable to do so. There was a surprising similarity 
between the predominating types of bacteria found in different soils and 
in the same soil at different seasons. 
In an attempt to explain the presence of large numbers of bacteria 
in frozen soils, Harder (1916) studied the effect of heavy frosts. He 
found that the number of bacteria in surface soil increased markedly 
after heavy frosts and maintained this high average during the winter 
months. Contrary to the findings of Conn and of Brown and Smith, 
Harder noted that the increase was directly related to the moisture content. 
His findings verified Conn’s conclusion that the bacterial flora was more or 
less the same thruout the year, with the exception that after heavy frosts 
there was an increase in the proportion of small transparent colonies. 
Potted soils did not show this increase even when enriched with sugars. 
In fact the enriched soils showed fewer organisms in the frozen than in 
the unfrozen samples. Harder concludes that the increase in numbers 
was due to mechanical transportation by moisture coming up from below 
during heavy frosts. 
Conn (1918), in his microscopic study of bacteria and soil fungi, found 
that there was more or less clumping of the bacteria on the soil particles, 
and in certain cases the organisms occurred in large sheets one individual 
thick. In practically all cases there was an increase in the individual 
count over the group count. In one table Conn gives the results of thirty- 
four groups and individual counts of soils made at different periods 
following the addition of manure. The results show that the individual 
count gives an increase of 50 per cent over the group count. 
The findings of Brew and Dotterrer (1917) in their study of milk were 
very similar to those of Conn. Brew and Dotterrer noted a higher count 
with the microscope than with the plate method. The difference they 
attributed to the clumps of bacteria that do not break up on plating. 
MILK FLORA 
Stiles and Pennington (1909), in their investigation with ice cream, 
found that the samples showed what seemed to be an increase in the 
bacteria count for a few days, followed by a decrease on the fourteenth 
