The Cytotoxins 167 



substance, so that the bacteria all fall to the bottom in a short time. 

 This is independent of bacteriolysis. The bacterial destruction is 

 gauged by the disappearance of the bacteria which have been dissolved, 

 or by their failure to grow when transplanted to appropriate culture 

 media, showing that they have been killed. 



By making the bacterial suspension and complementary serum 

 constant quantities (taking care that not too many bacteria be present), 

 one is able to estimate the value of the immune serum. By using the 

 bacterial suspension and a heated immune serum (containing no com- 

 plement) as constants and varying the addition of complementary 

 serum, one can estimate the respective values of several complementary 

 serums. By using both serums as constant factors and varying the 

 number of bacteria, one can determine the exact bacteriolytic value of 

 the mixture. By taking out and planting drops from time to time the 

 rapidity of bacteriolysis can be determined, and by plating out the drops 

 and counting the colonies one may arrive at percentages of destruction 

 and express the bacteriolytic process in the form of a curve. 



A peculiar phenomenon is sometimes observed and has 

 been studied by Neisser and Wechsberg,* and is commonly 

 called by their names. It is the result of "deviation of the 

 complement," the condition being as follows: When an 

 animal whose blood-serum is normally possessed of a high 

 degree of germicidal power is immunized by repeated injec- 

 tions of a bacterial endotoxin, its serum when examined by 

 the usual methods fails to show the usual increase in the 

 specific bactericidal action toward that particular organism, 

 though it retains its general bacteria-destroying power. 

 If, however, the serum be greatly diluted, its action is 

 changed, so that it loses its general bacteria-destroying power 

 and develops marked increase in the specific destructive 

 action upon the particular bacteria used in the experiment. 

 Neisser and Wechsberg attribute the peculiar reaction to the 

 fact that there being more amboceptors than complements 

 in the serum, some of the former satisfy their combining 

 affinities by attaching themselves to the bacteria, some by 

 attaching themselves to the complement, instead of forming 

 combinations of all three. If under these circumstances the 

 serum containing the amboceptors is diluted until their 

 number becomes approximately equal to the number of com- 

 plements introduced, any deviation resulting from inequality 

 of the combining affinities becomes improbable. Bordet 

 and Gay,f however, have performed experiments tending 

 to show that these elements do not really unite, thus seem- 

 ing to controvert the theory of Neisser and Wechsberg, and 



*" Munch, med. Wochenschrift," XLVIII, No. 13, p. 697, April 30, 

 1901. 



t "Ann. de 1'Inst. Pasteur," xx, June 25, 1906, No. 6, pp. 267-498. 



