One or more colonies typical in appearance. │ (a) If only “partially confirmed” test is required│None (b) If completed test is required, select two │ typical colonies for identification. │V 2. No typical colonies. │IV IV. Replace plate in incubator for an additional 18 to 24 │ hours; then, whether colonies appear typical or not, │ select at least two of those which most nearly resemble B.│ coli. │V V. Transfer each colony fished to: │ 1. Lactose broth fermentation tube; incubate not more │ than 48 hours at 37° C. Observe gas-formation. │None 2. Agar slant; incubate 48 hours at 37° C. │ (a) If gas formed in lactose broth tube inoculated│ with corresponding culture │VI (b) If no gas formed in corresponding lactose │ broth tube, test is completed and negative. │None VI. Make stained cover-slip or slide preparation, and │ examine microscopically. │ 1. If preparation shows non-spore-forming bacilli in │ apparently pure culture, demonstration of B. coli is │ completed. │None 2. If preparation fails to show non-spore-forming │ bacilli or shows them mixed with spore-bearing forms │ or bacteria of other morphology. │VII VII. Replate, to obtain assuredly pure culture, select │ several colonies of bacilli and repeat steps V and VI. │ ────────────────────────────────────────────────────────────┴───────── In order that tests for B. coli may have quantitative significance, the following general principles and rules should be observed: Ordinarily not less than three portions of each sample should be tested, the portions being even decimal multiples or fractions of a cubic centimeter; for example, 10 cc., 1 cc., 0.1 cc., .01 cc., etc. It is essential that the dilutions should be such that the largest amount gives a positive test (unless the water is such as to give negative tests in 10 cc.), and the smallest dilution, a negative result. To insure this result, it is often necessary to plant four or five dilutions, especially in the examination of a sample of entirely unknown quality. The quantitative value of a series of tests is lost, unless all or at least a large proportion of the smallest dilutions tested have given negative results.