MODE OF ACTION OF BACTERICIDAL SERA. 127 



All these experiments show that the effect produced by a given amount 

 of complementing serum, just sufficient to reactivate a definite quantity 

 of inactive immune serum, was diminished when large amounts of 

 immune serum were employed. In like manner it was possible to in- 

 hibit the activity of a normal serum which was bactericidal by itself, 

 by the addition of large amounts of the immune serum. 



It seems to us that an explanation of these important phenomena 

 is possible only on the basis of the newer views of Ehrlich and Mor- 

 genroth. From the work of these authors on haemolysins and from 

 our own bacteriolytic experiments we know that the immune serum 

 contains a thermostable interbody (amboceptor) which while itself 

 inactive renders the complement effective by linking itself, on the 

 one hand, to the bacterium or erythrocyte to be dissolved, and on the 

 other to the complement. The complements, as is well known, are 

 thermolabile and are contained in normal sera. But the interbody 

 may also be normally present in a serum. This follows from the 

 side-chain theory, and has already been emphasized. 1 An instance 

 ef this is shown in Table IV. The normal active guinea-pig serum 

 contained complement and interbody. But besides this it contained 

 additional complement, which became manifest when more inter- 

 body, in the form of inactive immune serum, was added. In example 

 II it was impossible to demonstrate an interbody hi the normal 

 serum, for this by itself did not kill the bacteria even though inactive 

 normal serum was added. It did, however, contain complement, 

 and this became manifest when inactive immune serum was added. 

 These phenomena are exactly like those observed with haemolysins 

 which have recently been so carefully studied. This one phenomenon, 

 however, of the ineffectiveness of large doses of immune serum has 

 not thus far been encountered in haemolysins. This is apparently 

 due to differences in the affinities of the interbodies, as we shall pres- 

 ently show. 



In Fig. 1, on the next page, All represents schematically a bac- 

 terium a with a number of receptors; for there are many reasons 

 why we should assume that each bacterium possesses a number of 

 receptors of the same kind. According to the side-chain theory, 

 if we inject this bacterium into an animal an overproduction of the 

 corresponding group will occur, resulting in a serum which is rich 

 in body b. This body 6, however, is not able by itself to injure the 



1 Deutsche med. Wochenschr., 1900, No. 49. 



