424 G. S. WILSON 



observers, or at any rate, not clearly recognized. It is obvious 

 that the greater the number of bacilli distributed in a given space, 

 the less is the interval between each of them, and the greater the 

 chance of two being coincident. In every case in which two 

 bacilli are coincident or are placed very close together only one 

 colony will develop. Further, when two bacilli are situated at 

 such a distance from each other that each is able to develop, yet 

 at such a distance that continued development of both will result 

 in fusion, it is clear that a single colony must arise. \Miether one 

 continues to grow and the other desists or whether both develop, 

 the result must be the same — namely, the appearance of one 

 colony in place of two bacilli. On pure a priori grounds one 

 would expect this overcrowding factor to be of considerable im- 

 portance in determining the number of colonies which will develop 

 in a given space. One would expect it to play but a small part 

 so long as comparatively few bacilU were inoculated, but as the 

 number of the latter increased so should the percentage which 

 fails to develop into colonies become greater. So much for theo- 

 retical considerations. It was decided to make a quantitative 

 estimation of this overcrowding factor and the following experi- 

 ments were performed. A dilution in tap water of a 3 hours' 

 broth culture of Bad. suipesiifer was put up, of such strength that 

 there was approximately one bacillus in every drop. A series of 

 agar tubes was then inoculated with varying numbers of drops, 

 the tubes being put up in such an order as to equalize the possibly 

 deleterious action of the diluent. As a rule in each experiment, 

 16 to 20 tubes were inoculated with one drop, 6 to 10 wath 5 

 drops, 6 with 10 drops, 4 with 15 drops, and 4 with 20 drops. 

 After three daj's' incubation the tubes were counted, and the 

 actual number of colonies developing in the difTerent sets of 

 tubes compared. It is clear that if one colony developed per 

 tube when 1 drop was inoculated, then 5 should develop in the 

 5-drop tubes, 10 in the 10-drop tubes, 15 in the 15-drop tubes 

 and 20 in the 20-drop tubes. After a number of experiments had 

 been performed with emulsions containing from 1-15 bacilli per 

 drop, another series was instituted in which the emulsions con- 

 tained from 15 to 30 bacilli per drop and later on, a further series 



