314 WATER HYGIENE 
in obtaining the B. coli index. If a negative confirmation test is obtained, 
B. coli is considered present in the next larger quantity without confirmation. 
If a negative result is obtained from the largest quantity planted, either by no 
gas formation in the original planting or by a negative confirmation test, B. coli 
is considered present in the next larger amount that would have been planted, 
except that a negative result in 100 c.c. is recorded at zero. The assumption 
is value only in the case of waters such as are encountered in water purification 
plants where the approximate coli content of the water is known. Such assump- 
tions are rarely necessary as the quantities of water chosen for examination can 
ordinarily be changed in sufficient time to always obtain a positive test. It is 
probable that this assumption causes less error than would be introduced by 
either omitting the figure from the averages or calling it zero. 
“The laboratory work above described with the numerical estimations of 
B. coli derived, are recorded for the month under headings of waters in the 
consecutive stages of purification. The form used is convenient and compact 
and lends itself readily to a grasp of conditions throughout the plant at any 
time. Five days from the collection of the water sample, B. coli may, following 
the above scheme, be satisfactorily and conclusively isolated and identified 
to conform to the definition. 
The following table will show the number of samples examined, according 
to this method, from January to October, 1916, and the percentage of samples 
confirmed as B. colt. 
Fir rerReD WATER. 
waver Rettled | Apphed 
ater ater. ater Effluent from Outlet Clear 
Filters. Reservoir. 
Samples tested Lous 271 257 266 258 194 
Percentage confirmed . 97 4 91.0 96 6 *94 9 *86.1 
* Includes one sample that showed B. Cloace. 
The B. coli Index. This was proposed by Phelps and was incor- 
porated in a report of the Committee of Water Purification Plants. 
The following is taken from that report. The B. cold index is the approx- 
imate number of B. colz per cubic centimeter as determined from qual- 
itative tests made on different quantities of water. For any individual 
sample, it may be taken as the reciprocal of the smallest volume of 
water used in the test which gave a positive result. Thus, if a sample 
gave a negative test with 0.1 c.c. and a positive test with 1.0 ¢.c. and 
10.0 c.c., the B. colz index would be 1—0.1=10. The B. coli index for a 
single sample is not very accurate. The index becomes more accurate 
as the square root of the number of tests increases. 
