196 MECHANISM OF COMPLEMENT ACTION 



This agrees exactly with our hypothesis, both as to the order of 

 the reaction of thermoinactivation, and in that the destruction of B' 

 with its high temperature coefficient is predominant at temperatures 

 above 50°, while the limiting reaction at temperatures below 37° is 

 the breakdown of A which has a relatively low temperature coefficient. 

 The gradual decrease of B' then keeps pace with the gradual decrease 

 of A. Whenever, as after brief heating to 56° for example, the con- 

 centration of B' is disproportionately reduced and a large amount of 

 A is still left, the latter will act as a reservoir from which B' will be 

 restored to the concentration appropriate to the temperature and the 

 remaining amount of A . The hemolytic power being proportional to 

 B' will be low immediately after heating and will then be regenerated 

 as Gramenitzki^" and the writer* have shown to be the case. 



Turning now to the question of what occurs during photoinactiva- 

 tion we find that if the time curves of several experiments are aver- 

 aged the process follows the course of a monomolecular reaction.^^ 

 Now if only B' were hemolytic and the members of the left-hand branch 

 of our hypothetical series (i.e. A, B, and B') contained the same 

 photosensitive group, the proportion of each substance destroyed 

 would be the same; that is, one-tenth or one-half or nine-tenths of 

 each of them would be destroyed, but not one-half of A and nine-tenths 

 of B and B'. Under certain conditions which were defined in a previ- 

 ous paper-i the destruction of a substance by light will follow the 

 course of a monomolecular reaction, and if B' is so destroyed, and no 

 disproportionate amount of A and B is left (the latter case occurs 

 when complement is briefly heated), then there can be no regenera- 

 tion. The theory then accounts for the observed course of photo- 

 inactivation and for the fact that no regeneration follows photoinacti- 

 vation. 



The reader should now note carefully the curves shown in Fig. 2, 

 of the preceding paper. They are all characterized by the fact that 

 at some time during the process the hemolytic activity exceeds that to 

 be expected if the process followed exactly the course of a monomo- 

 lecular reaction; there is a "wave" of excess activity. If we suppose 

 that b' is hemolytic and that it forms and decomposes at a relatively 



^^ Gramenitzki, M., Biochem. Z., 1912, xxxviii, 501. 

 21 Brooks, S. C, J. Gen. Physiol., 1920-21, iii, 169. 



