Report on the Bacteriology of Water. 463 



At length a temperature is reached where the curve is infinitely 

 short, i.e., no growth occurs at all. This temperature is, however, 

 above 39 C., and indicates the death-point. 



Taking temperatures below the optimum. There is a point, some- 

 where below 8 C., where the curve is indefinitely postponed, i.e., no 

 growth can occur at all. Then comes a temperature, also below 

 8 10 C., where the curve ascends slowly and never attains the 

 steepness of the curve at optimum temperature. This is the minimum 

 temperature. 



At temperatures above the minimum the curve attains more and 

 more nearly, and in shorter and shorter times, to the steepness of the 

 normal curve, the nearer the temperature in question is to the 

 optimum temperature. 



This optimum temperature is either 25 C. or some point very 

 near it. 



The above case of the normal curve is the hypothetical one where 

 all the conditions are constant, a state of affairs never realised.* 

 During the growth, between the period when the germination is 

 completed and the organism no longer obtains any supplies from the 

 spore, but is totally dependent on the food-materials given it, and 

 the period when the curve begins to ascend less rapidly, there is a 

 period of maximum growth, during which the filament doubles its 

 length in equal minimum times. This is the critical period of the 

 curve. The more closely the curve approximates to the normal 

 curve the longer this phase of equal minimum doubling periods lasts ; 

 the more external conditions affect the curve the shorter this phase 

 is, and the longer the doubling periods become. 



The factors affecting the curve may be regarded as of two kinds, 

 internal and external, though they probably never vary entirely 

 independently. 



The internal factors are such as (1) irregularities of cell-divisions : 

 if a single cell fails to divide in due order, the curve is at once affected, 

 because the regularity of the intercalary growth of the filament is 

 destroyed, and this occasionally happens. (2) The separation of the 

 segments : several observations suggest that the growth is slowed at 

 once when the new surfaces of the broken ends come in contact with 

 the food-medium. (3) Nutations and oscillatory movements, though 

 possibly these affect the measurements rather than actual growth. 

 '4) Unknown internal factors which affect the rapidity of germina- 

 tion, the ability to assimilate the food- materials, and so forth. In 

 some cases these may be due to pathological conditions, as in the case 

 given on p. 392. 



* Theoretically, with absolute uniformity of conditions, including food supply, 

 the curve would go on to infinity, and the doubling periods be equal th/oughout; 

 the fall above would then be indefinitely postponed. 



