BACTERIOLOGICAL INVESTIGATION OF WATER. 251 



water, the capillary end of the bulb is broken under the 

 water and the vacuum in the tube causes a suction of the 

 fluid into the bulb until same is about three-fourths full. 

 When this is accomplished, the bulb should again be sealed 

 and packed in ice. 



When making gelatin and agar plates a known quantity 

 of water should always be used to each plate, say '0.01, 

 0.05, 0.10, or 1 c.c., the quantity varying according to the 

 amount of suspected contamination of the water. The plates 

 are made as already described in the chapter on the making 

 of plates, only plate No. 1, however, being made. After 

 twenty-four or forty-eight hours the colonies on the plate are 

 counted ; and as it is assumed that each colony is grown 

 from a single bacterium, the number of colonies on the plate 

 represents the number of bacteria originally in the sample. 

 Two sets of plates should always be made, one on gelatin, which 

 is kept at the room temperature, and another on agar for the 

 incubator. When the number of colonies on the plates is 

 very large, plates should be made with still greater dilution, 

 so as to obtain plates with only a moderate number of bac- 

 teria, in order to facilitate the count. Absolutely sterile water 

 should always be used to make the dilution. In expressing 

 results, the number of bacteria in 1 c.c. of the water is men- 

 tioned. For counting the colonies on the plates the counter 

 of Wolfhuegel (Fig. 31) is used, which consists of a wooden 

 framework, at the bottom of which may be put a glass plate, 

 white or black, in order to form a background for the plate 

 containing the colonies. Over this background the plate of 

 which the colonies are to be counted is placed, and over this 

 a transparent glass plate ruled in square centimeters, and 

 held in position just above the colonies, but without touching 

 thorn. It is easy in this way to count the colonies in each 

 square centimeter, and so make out the total number of col- 

 onies on the plate. When this is found too tedious, a number 

 of squares at different portions of the plate may be counted, 

 an average established, and that average multiplied by the 

 number of squares will give approximately the number of 

 colonies on the plate. By multiplying the number of col- 



