MICHIGAN ACADEMY OF SCIENCE. 
117 
TABLE VI. 
Kind of Water. 
No. of 
Samples. 
Ammonia. 
Cl. 
Free. 
Alb. 
Well water. 
322 
.126 
.142 
25.3 
58 
.124 
.175 
7.3 
Water from great lakes. 
181 
.040 
.076 
15.3 
Waters from streams. 
51 
.091 
.153 
15.7 
Spring waters. 
28 
.072 
.088 
8.8 
Cistern waters. 
3 
1.011 
.558 
6.0 
Artesian waters. 
8 
.164 
.092 
It will be seen from (be above that the waters of the Great Lakes make 
the best showing, so far as both free and albuminoid ammonia go 
and the cistern waters the worst. Indeed the cistern waters were the 
most filthy of all examined. It should be said that these waters were 
from neglected cisterns and it was supposed by the families using these 
waters that they were to be employed only in daily ablutions, but children 
and roomers had washed their teeth with these waters and occasionally 
drank them when the drinking water pitcher was empty. In one family 
every person under thirty years of age, six in number, acquired typhoid 
in this way. Contamination from a near-by privy vault was evident. 
It will be understood that these figures do not indicate the composition 
of the waters in either the small or great lakes, but do show the com¬ 
position of such waters as locally contaminated. The artesian waters 
were from wells recently opened and much of the organic contamina¬ 
tion probably came from the surface. 
Bacteriologically 300 samples of well water showed an average of 4,261 
bacteria per c. c. and 107 (38 per cent) of these contained bacteria that 
killed animals. 
Fifty-eight small lake waters gave an average of 1,731 bacteria per 
c. c., and 25 (43 per cent) of these contained bacteria that killed ani¬ 
mals. 
One hundred and eighty-one great lake waters had an average of 
12,067 bacteria per c. c. and 39 (21 per cent) of these contained bacteria 
toxicogenic to animals. 
Fifty-one samples of water from streams gave an average of 5,979 
bacteria per c. c. and 27 (53 per cent) of these contained toxicogenic 
bacteria. 
Twenty-eight samples of spring water gave an average of 1,309 bacteria 
per c. c. and 3 (10 per cent) of these contained toxicogenic bacteria. 
Five (83 per cent) of the cistern waters contained toxicogenic bac¬ 
teria and 3 of these gave an average of 85,040 bacteria per c. c. It will, 
of course, be understood, that these figures do not indicate the composi¬ 
tion of the waters of well-kept cisterns any more than the figures for 
both the small and great lakes indicate the composition of the waters of 
the open lakes. 
The number of bacteria in the artesian waters varied too much (from 
80 to 56,400) to make an average of any value, and out of nine of these 
waters 4 (44 per cent) killed animals. 
It will be seen from the above that spring waters, taking conditions 
