FORMS OF THE COLON BACILLUS 133 



adaptation of the principle of index numbers to the problem at hand, 

 the index being the mean variation in form of all of the cultures being 

 compared. The value so computed may be called the index of vari- 

 ation of form. It is admittedly not an entirely satisfactory method, 

 being rather artificial and involving to some extent personal judgment 

 in classifying the cells, but in the absence of anything better may 

 serve to supply at least an approximate quantitative expression of 

 the degree of variation in form of the cells, which may be compared 

 from day to day and from culture to culture. 



The incidence of the various cell classes in the five cultures at 

 various stages of growth is given in Table XXVII together with the 

 indices of variation calculated from them. The mean frequencies of 

 the ten cell types of the five cultures at all stages of growth are shown 

 graphically in Figure 35. 



The fluctuations of the index of variation day by day are some- 

 what irregular and do not appear to be highly significant; this is 

 perhaps largely due to the small number of cells in each sample. 

 The alkaline culture showed very little variation; there was a slight 

 rise on the nineteenth day followed by a decline, corresponding with 

 the period of decrease in death rate. In the neutral medium there 

 occurred a well-defined increase reaching the maximum on the third 

 day, corresponding with the beginning of the death phase; this was 

 followed by a decrease and a second increase with its maximum on 

 the fourteenth day, at which point the death phase had nearly reached 

 its end, the number of viable cells remaining nearly the same during 

 the remaining period of observation. The medium containing NaCl 

 showed a very pronounced rise in the index of variation on the sec- 

 ond day, due largely to the remarkably high incidence of budding 

 and branching cells, followed by an irregular decline. The acid cul- 

 ture exhibited two maxima, one on the seventh day corresponding 

 with the beginning of the death phase, and one on the nineteenth 

 day when the death rate was decreasing. The medium containing 

 CaClg gave a slight initial rise followed by a decline, then a pro- 

 nounced increase which was sustained until the nineteenth day, there 

 being a marked reduction in the variation of the cells on the twenty- 

 fifth day. 



