(5) Gelatin tubes should not be pronounced negative until they have been incubated at least 15 days. The following group reactions indicate the source of the culture with a high degree of probability: Methyl red + │B. coli of fecal origin. Voges-Proskauer − │ Gelatin − │ Adonite − │ Indol, usually + │ Saccharose, usually −│ Methyl red − │B. aërogenes of fecal origin. Voges-Proskauer + │ Gelatin − │ Adonite + │ Indol, usually − │ Saccharose + │ Methyl red − │B. aërogenes, probably not of fecal │ origin. Voges-Proskauer + │ Gelatin − │ Adonite − │ Indol, usually − │ Saccharose + │ Methyl red − │B. cloacae, may or may not be of fecal │ origin. Voges-Proskauer + │ Gelatin + │ Adonite + │ Indol, usually − │ Saccharose + │ 12. ROUTINE PROCEDURE FOR EXAMINATION OF SAMPLES OF WATER. _First Day_: 1. Prepare dilutions as required. 2. Make two (2) gelatin plates from each dilution, and incubate at 20° C. 3. Make two (2) agar plates from each dilution, and incubate at 37° C. 4. Inoculate lactose broth fermentation tubes with appropriate amounts for B. coli tests, inoculating two (2) tubes with each amount. Note:—Where repeated tests are made of water from the same source, as is customary in the control of public supplies, it is not necessary to make duplicate plates or fermentation tubes in each dilution. It is sufficient, in such circumstances, to make duplicate plates only from the dilution which will most probably give from 25 to 250 colonies per plate. _Second Day_: 1. Count the agar plates made on the first day. 2. Record the number of lactose broth fermentation tubes which show 10 per cent. (10%) or more of gas. Note:—In case only the presumptive test for B. coli is required, fermentation tubes showing more than 10 per cent. (10%) of gas at this time may be discarded. _Third Day_: 1. Count gelatin plates made on first day. 2. Record the number of additional fermentation tubes which show 10 per cent. (10%) or more of gas. 3. Make a lactose-litmus-agar or Endo’s medium plate from the smallest portion of each sample showing gas. Incubate plate at 37° C. Note:—In case the smallest portion in which gas has been formed shows less than 10 per cent. (10%) of gas, it is well to make a plate also from the next larger portion, so that, in case the smallest portion gives a negative end result it may still be possible to demonstrate B. coli in the next larger dilution. _Fourth Day_: