FORMS OF THE COLON BACILLUS 135 



mean index of variation of the five cultures for each day we increase 

 the number of cases and obtain a correspondingly more accurate pic- 

 ture of the variations with time. There occurs a sharp rise at the 

 beginning of the death phase followed by a decrease, then a second 

 more sustained rise with another marked decrease on the last day. 

 This final drop in the degree of variation in the form of the cells, 

 occurring during the period of decreasing death rate as measured 

 by the plate counts and of increasing rate as measured by the micro- 

 scopic counts, can readly be explained by the final dissolution of 

 those cells which presented the greatest departure from the "normal" 

 form, and may be considered to support the argument that these ir- 

 regular cells are actually dead. I am at a loss,- however, to explain 

 the initial rise and fall in the index of variation. 



While the daily fluctuations in the index of variation did not 

 prove very significant, and particularly there could not be established 

 any clear relationship between the period of greatest degree of vari- 

 ation and the period of most rapid death, important results were ob- 

 tained when the different cultures were compared with each other 

 with regard to their mean indices of variation. In order of magnitude 

 of this value they are arranged as follows: Alkaline (67.71), neutral 

 (87.56), salt (93.01), acid (93.42), and calcium chloride (157.25). 

 It will be noted that the values for the salt and acid cultures are 

 nearly identical. But the figures for the salt culture are not directly 

 comparable with the others, as no observation was made on the twen- 

 ty-fifth day. Since in all of the cultures there was observed a de- 

 crease in variation on the last day, it is probable that this would also 

 have occurred in the medium with sodium chloride, and that the mean 

 index of variation would have been somewhat reduced. 



The rank order of the cultures with regard to their mean indices 

 of variation does not correspond with their rank order as regards 

 mean rate of death, maximum rate of death, or favorableness of the 

 medium for growth. In fact the cultures containing calcium chlo- 

 ride, which gave the highest^ degree of variation in cell form, had 

 nearly the lowest death rate. It is clear therefore that there cannot 

 be established a correlation between the index of variation and the 

 rate of death. 



This rank order does, however, correspond definitely with the 



