130 MORPHOLOGIC VARIATION 



increase was 141 per cent, in the alkaline culture 58 per cent, in 

 the acid 41 per cent, and in the salt agar 12 per cent. It is interest- 

 ing in this connection that the first two cultures named also had a 

 high incidence of constricted (dividing) cells. 



At first an attempt was made, as in the preceding study, to 

 follow variations in form by means of the area-length index. This, 

 however, failed to yield any significant results, the figures so ob- 

 tained showing no regular changes with increasing age of the culture, 

 nor noteworthy differences when the averages for the various cultures 

 were compared. It was quite apparent, however, by simply inspect- 

 ing the slides that the important changes in cell form were not so 

 much variations in relative slenderness or thickness as they were 

 departures from the normal symmetry, i.e., development of budding, 

 branching, bulging, and other unusual cell types. To measure the 

 degree of variation in form then some scheme had to be adopted 

 which would not only consider the relative slenderness and plump- 

 ness of the cells, but also these asymmetrical types, as well as curva- 

 ture in the cells. What was desired was a quantitative expression 

 of the degree of total variation encountered which could be com- 

 pared directly with rates of death and autolysis determined as above. 



This was accomplished in the following way. The cells were 

 (more or less arbitrarily) divided into ten classes, illustrated in 

 Figure 34. This illustration has been made by tracing from the 

 cards, cells of the various classes from many different samples, 

 simply to show the types of cells included in each class. They 

 were chosen more or less at random, although an attempt was made 

 to include extreme cases. It should be borne in mind that these 

 composite fields do not indicate the frequency with which any of 

 the cell types occurred in any particular culture; they are intended 

 merely to indicate the nature of the classification followed. They 

 are arranged and numbered in the order of their mean frequency 

 in all of the cultures studied. 



These ten classes include four classes of straight symmetrical 

 cells: viz., slender rods (Class IV), normal rods Class I), oval cells 

 (Class III), and spherical cells (Class VII). This is of course an 

 entirely arbitrary division, since there occurs every degree of gra- 

 dation from long filaments to perfectly round cells. But having 



