S. C. BROOKS 



67 



ance less and more, respectively, than the niode, possessed by the 

 most fragile and most resistant classes. Here Xi + X2 is the total 

 range of resistance of all the cells. A curve of this type (a) and its 

 integral, or time curve, (b) are shown in Fig. 3; the time curve has a 

 form not unlike that of the curves for the course of hemolysis. 



If we suppose, with some investigators, that the resistance of the 

 cells varies around an average from which it does not deviate to an 

 extent sufficient 'to influence the course of the process, we must con- 

 sider all the cells to be in a single class with respect to resistance. 



Fig. 3. (a) variation curve whose equation is 



3' = 3'oll-l- — ) (1 ) , where Xi = 2, X2 



6, /fe = 0.5. 



If (a) is considered as the curve of a differential equation, (h) is the curve of the 

 corresponding integral. The ordinates of (b) are proportionate at any position 

 on the X axis (abscissae) to the area to the left of the ordinate at that position, and 

 under the curve (a) . 



The frequency curve will then be so narrow as to be approximately a 

 straight line normal to the axis of the abscissae at some point. Its 

 integral, which is the time curve of the process, will follow the axis of 

 abscissae to this point, and then rise perpendicularly to its ultimate 

 height. In other words, if the process were hemolysis, all the cells 

 would lake at the same instant. This conclusion may be avoided, 

 theoretically at least, by making one of the special assumptions 

 which are discussed below. 



