RATES OF LETHAL ACTION IX A CELL POPULATION 



593 



rate is no longer sigmoid bnt is of a hyperbolic or exponential form (Fig. 

 12-34, curve C). This latter type of curve actually is approached in some 

 cases and most of the published data indicate that the variation curve is 

 indeed generally skewed to the left. 



Much has been written about death rates that obey a so-called "mono- 

 molecular"' time course and the demonstration of such often has been in- 

 terpreted in terms of the nature of the reaction of the lethal agent with the 

 cell or the processes that this reaction produces. Such a monomolecular 

 death rate is usually adduced by obtaining a straight line when the loga- 

 rithm of the fraction of surviving cells is plotted against time. It should 



5 10 



TIME CMIN)- 



Fig. 12-34. Death rate curves obtained from the distribution 

 curves in Fig. 12-33 bv summation. 



be evident that such a procedure does not provide information about the 

 reaction of the inhibitor at all but only about the variation between the 

 cells or organisms. Plots of this type are given in Fig. 12-35 for tiie distri- 

 butions shown in Figure 12-33. It may be seen that the curves become more 

 linear as the distribution curve is more skewed to the left and that a truly 

 linear curve is obtained only from the extreme distribution of curve C. 

 Linearity of such curves merely implies that the distribution of cells more 

 resistant than the cells of the commonest susceptibility is exponential. 

 Perhaps the first work clearly showing that a skew variation in suscep- 

 tibility to an inhibitor can account for apparent monomolecular behavior 

 was that of Peters (1920). Cultures of Colpidium were exposed to 0.2 mM 

 Hg++. and the percentage of organisms killed after different time intervals 

 was determined. Figure 12-36 shows the curve and logarithmic plot of the 

 per cent survivors. A straight line was obtained between 10 and 45 min 

 but this monomolecular behavior was correctly interpreted as arising from 

 a variation curve actually not much different from curve C in Fig. 12-32 

 with a latent period of 10 min. The forms of these curves thus have no bear- 



