132 MORPHOLOGIC VARIATION 



measured the area-length index, there was provided a means of 

 separating these into classes which would not depend at all upon 

 personal judgment. Area-length indices of 20, 40, and 60 were 

 chosen as dividing points for these classes. Long, slender filamentous 

 forms were considered as having an area-length index of 20 or 

 less; normal forms from 21 to 40; oval forms from 41 to 60; spheri- 

 cal cells, those with an index higher than 60. 



Further classes were provided as follows: All curved cells (Class 

 II), regardless of their size or slenderness as long as there was 

 a definite curvature to their axis; all cells which were greater in 

 diameter at one end than the other (Class V), either provided with 

 a knob-like expansion or showing a uniform graduation in diameter 

 from the larger end to the smaller; all cells which showed a central 

 constriction (Class VI), i.e., cells probably in the act of dividing; 

 all cells which showed a lateral projection (Class VIII) whether 

 in the form of a small bud or definite side branch; all cells which 

 showed a sub-terminal constriction, i.e., which appeared to have 

 a terminal bud (Class X) ; and finally all cells which exhibited a 

 central bulging (Class IX). 



Having decided upon this classification, all of the cards bearing 

 the tracings of cells were gone through, and the proportion of the cells 

 in each class determined for each sample. The average percentage 

 of each class for all the the cultures at all stages of growth was then 

 calculated, and the reciprocal of this percentage determined. The 

 product of the average percentage incidence of any class by its recipro- 

 cal will of course equal one (or for convenience 100), and therefore 

 the mean of these products for all ten classes will also equal 100. If 

 now in any given sample the percentage frequency of each class is 

 multiplied by the reciprocal of the percentage frequency of that 

 class for the entire series of samples, and the average of these prod- 

 ucts for the ten classes is taken, there is obtained a value which in- 

 dicates whether in that sample the degree of variation is greater or 

 less than the average of all of the samples. Where there occurs more 

 than the average number of the more frequent cell types (and there- 

 fore fewer of the less frequent types) this value will be less than 100. 

 while conversely if there occurs more than the average number of the 

 rarer cell forms the value will be greater than 100. This is really an 



