PART I 



THE PROBLEM OF THE LOGARITHMIC 

 ORDER OF DEATH IN BACTERIA 



I. THE ORDER OF DEATH 



Organisms which die as a result of a chemical reac- 

 tion die in an orderly, predictable way. Two methods of 

 graphic presentation of such orderly deaths are cus- 

 tomary. One consists in the use of mortality curves, 

 which are represented by black blocks in Figure 1, and 

 which give the number of individuals dying in each suc- 

 cessive time unit; this type of curve is commonly 

 used in medical statistics. In the other method use is 

 made of survivor curves, which show the number of indi- 

 viduals still living at any time after the beginning of ex- 

 posure. This type of curve, shown by the thin lines of 

 Fig. 1, is used by bacteriologists, entomologists and plant 

 pathologists. 



The order of death is, in principle, the same for all 

 multicellular organisms. The mortality curve begins 

 at zero, and frequently remains at zero for some time, 

 as long as all individuals can recover after short ex- 

 posures ; then the first individuals die, the frequency of 

 death gradually increases to a maximum, then decreases, 

 usually at a slower rate, until only a few very resis- 

 tant organisms remain, and at last even they succumb. 



Whether tadpoles, insects, or plant seeds are tested, 

 whether heat or rays or chemicals are used, the mor- 

 tality curve remains essentially the same. These mor- 

 tality and survivor curves agreed so well with our con- 

 ception of gradation in resistance that not much atten- 

 tion was paid to them. They were taken for granted until 

 Madsen and Nyman in 1907 and Harriet Chick in 1908 



