102 
TYPHOID FEVEK IN DISTRICT OF COLUMBIA. 
it was found in 30 per cent of the specimens of ‘Opiate’’ ice examined. 
This corresponds to the fact that can ice is almost invariably made 
from distilled water, whereas plate ice is made from river or well 
water unsterilized. 
BACTERIOLOGICAL SUMMARY OF ICE EXAMINATIONS. 
Number of specimens examined: 
Natural ice 6 
Artificial ice 28 
Total 34 
Average number of bacteria per cubic centimeter: 
In natural ice 118 
In artificial ice 123 
Percentage of specimens of natural ice containing B. coli: Percent. 
In 1 cc 16. 6 
In 10 cc 33. 3 
Total 49.9 
Percentage of specimens of artificial ice containing B. coli: Percent. 
In 1 cc 14. 3 
In 10 cc 14. 3 
Total 28.6 
Bacteria per 
cubic centimeter. 
Average of 9 plate-ice specimens 216 
Percentage of plate-ice specimens containing B. coli: Percent. 
In 1 cc 20 
In 10 cc 10 
Total 30 
Bacteria per 
cubic centimeter. 
Average of 7 can-ice specimens 18 
Percentage of all can-ice specimens containing B. coli: Percent. 
In 1 cc 0 
In 10 cc 0 
Total 0 
Note. — In making the above average counts, ice specimen No. 29, which gave the 
unusual number of 5,000 bacteria per cubic centimeter, was not included; it was, 
however, included in the percentages of specimens containing B. coli. 
Our examinations show for the most part that manufactured ice 
contains more bacteria than the water from which it has been frozen. 
The contrary of course should be the case, for it is now well known 
that the act of freezing causes the death of a large number of bac- 
teria. This discrepancy, we believe, is readily accounted for by the 
unclean methods used in the manufacture of ice, given in detail in 
the report on the inspection of ice factories (p. 93). The added 
dirt accounts for the increased number of bacteria in the ice. 
