636 BACTERIA IN WATER. 
number, according to Bolton, would be less than one-hundredth 
of a milligramme, and at least three-fourths of this must consist of 
water. Theorganic material represented by this number of bacteria 
would therefore be so minute that it might be supplied by dust par- 
ticles accidentally falling into the distilled water. 
Rosenberg has shown that while many of the species which he 
obtained in pure cultures from the water of the river Main multiplied 
in sterilized distilled water, other species quickly died out in such 
water. The growth of certain bacteria depends not only upon the 
quantity of nutritive material present, but upon its quality, the con- 
ditions in this regard being widely different for different species. 
In view of the facts heretofore stated bacteriologists are now giv- 
ing more attention to a careful study of the kinds of bacteria pre- 
sent in their examinations of water. Rosenberg, in his examinations 
of the water of the Main in the vicinity of Wtrzburg (1886), found 
that before the river reached the city the water contained more 
micrococci than bacilli, but that after receiving the sewage of the 
city the number of bacilli was greatly in excess. 
Adametz (1888) has described eighty-seven ‘species obtained by 
him from water in the vicinity of Vienna; Maschek found fifty-five 
different species in the drinking water used at Leitmeritz; and Tils 
(1890) has described fifty-nine species obtained by him from the city 
water supply at Freiburg. 
Among the pathogenic bacteria which are liable to find their 
way into water used for drinking purposes, the most important, from 
a sanitary point of view, are the bacillus of typhoid fever and the 
spirillum of Asiatic cholera. Both of these microdrganisms are pre- 
sent in great numbers in the excreta of persons suffering from the 
specific forms of disease to which they give rise, and are consequently 
liable to contaminate wells and streams which receive surface water, 
when such excreta are thrown upon the surface or into sewers, etc. 
Epidemics of these diseases have frequently been traced to the use 
of such contaminated water, and in a few instances the presence of 
these specific disease germs in water has been demonstrated by bac- 
teriological methods. Laboratory experiments indicate, however, 
that an increase of these pathogenic bacteria in drinking water is not 
likely to occur, except under special conditions, and that they die 
out after a time, being ata disadvantage in the struggle for exist- 
ence constantly going on among the numerous species which have 
their normal habitat in water. 
Bolton, Frankland, and others have shown that the anthrax ba- 
cillus, not containing spores, dies out in hydrant water within five or 
six days. In the experiments of Kraus the anthrax bacillus added 
to well water, not sterilized, ata temperature of 10.5° C., was still 
