S. C. BROOKS 195 



in serum, and undergoing transformation into 5 at a rate which is 

 not very greatly accelerated by an increase in temperature. The 

 change B -^ B' is very rapid and is probably never a limiting factor. 

 B' is the lysin and undergoes hydrolysis at a rate markedly influenced 

 by changes of temperature, and proportional to the amount of B' 

 present at the moment, as in any monomolecular reaction such as 

 most hydrolyses. 



Let us suppose serum to be collected and placed at 0°C. The 

 breakdown of B' may be retarded by the lowered temperature, and 

 B' will accumulate until there is so much that the amount broken 

 down in a unit of time will equal the amount formed from A . (We 

 may suppose the process jB -^ 5' to be so rapid as to have no percepti- 

 ble influence on variations in the amount of B'.) Since the reaction 

 ^ -^ J5 is also monomolecular, the amount of B formed in any time 

 interval will be less and less as A is used up; and to keep pace with 

 this change B\ the lysin, must also gradually decrease in amount. 

 Thus the hemolytic power, which is proportional to the amount of 

 B', will gradually decrease. This is a well known phenomenon. 



Let us suppose that serum which has reached this steady state of 

 equal formation and destruction of B' is suddenly raised to a high 

 temperature, e.g. 56°C.; the destruction of B', which we have assumed 

 to have a high temperature coefficient, will be enormously acceler- 

 ated while its formation will still be relatively slow. As a result the 

 hemolytic power of the serum will rapidly decrease, and, since its rate 

 of formation is at first relatively negligible, the decrease will follow 

 the course of a monomolecular reaction. This will continue until so 

 little B' is left that it decomposes at a rate comparable with that of 

 the change A -^ B. Madsen and Watabiki^^ present data which not 

 only show that thermoinactivation of complement follows the course 

 of a monomolecular reaction (about as closely as photoinactivation) 

 but that the temperature coefficient at 50-56°C. is very high {Qio 

 lying between 123 and 366.8), but between 3 and 37°C. is about that 

 of typical dark reaction {Qio between 1.98 and 2.94).^^ 



^^ Madsen, T., and Watabiki, T., Oversigt. kong. danske Videnskab. Selskabs 

 Forhandl., 1915, 125. 



^ ^ Madsen and Watabiki express their results in terms of the temperature coef- 

 ficient of the van't Hoff-Arrhenius formula. The values of Qio given above are 

 calculated directly from their data. 



