THE COUNTING OF COLONIES 65 



b and d is placed a stout cover-glass of suitable size (c) ; d is separated 

 from the plate of by a square plate of glass (c) (a portion of an ordinary 

 glass-slide for microscopical purposes does well). Two small metal 

 tubes (/) are inserted through the rubber d. Method of use : Fix up 

 the apparatus as shown above, the screws being just tight enough to 

 keep the parts in position, and sterilise in the steam steriliser. Screw 

 up more firmly so as to make the rubber bulge slightly. Fill a 

 hypodermic syringe with some sterile glucose bouillon, push the needle 

 through the rubber d, and, tilting the point of the needle against the 

 glass c, slowly inject enough to form a drop on the under surface of c. 

 Withdraw the syringe and inoculate its point with the bacterium, again 

 introduce and inoculate the drop. Pass hydrogen througli one of the 

 tubes for fifteen minutes, close the ends of the tubes, and incubate at the 

 required temperature. The apparatus can be put on the stage of a 

 microscope and examined from time to time. 



The Counting of Colonies. An approximate estimate of the 

 number of bacteria present in a given amount of a fluid (say, 

 water) can be arrived at 

 by counting the number 

 of colonies which develop 

 when that amount is 

 added to a tube of suit- 

 able medium, and the 

 latter plated and incu- 

 bated. An ordinary plate 

 should be used in such a 

 case, and the medium 

 poured out in as rect- 

 angular a shape as pos- 

 sible. For the counting, FIG. 29. Apparatus for counting colonies, 

 an apparatus such as is 



shown in Fig. 29 is employed. This consists of a sheet of glass 

 ruled into squares as indicated, and supported by its corners on 

 wooden blocks. The table to which these blocks are attached 

 has a dark surface. The plate-culture containing the colonies 

 is laid on the top of the ruled glass. The numbers of colonies 

 in, say, twenty of the smaller squares, are then counted, and an 

 average struck. The total number of squares covered by the 

 medium is then taken, and by a simple calculation the total 

 number of colonies present can be obtained. Plate-cultures in 

 Petri's dishes are sometimes employed for purposes of count- 

 ing. The bottoms of such dishes are, however, never, flat, and 

 the thickness of the medium thus varies in different parts. 

 If these dishes are to be used, a circle of the same size as the 

 dish can be drawn with Chinese white on a black card, the 

 circumference divided into equal arcs, and radii drawn. The 



