228 THE ANTIOEN-ANTIBODY REACTIONS 



The Lysis of Bacteria. 



Although we talk of bacteriolysins, antibodies that lyse bacteria, and bacterici- 

 dins, antibodies that kill them, the two effects are not differentiated in practice, 

 and though we often talk of the bacteriolytic titre of a serum it is the bactericidal 

 effect that we actually measure. It is true that in many instances, as in Pfeiffer's 

 classical experiments on the lysis of the cholera vibrio, gross degenerative changes 

 in the bacteria have been observed, and there is no doubt that those bacteria that 

 are susceptible to the lethal action of complement in the presence of a specific 

 antibody undergo a change in structure that is analogous to the lysis of red blood 

 corpuscles. The change is not, however, of the same dramatic kind ; and the 

 naked-eye observation of the changes in a turbid bacterial suspension is not a 

 practical method of observing bacteriolysis. The principles involved in the reaction 

 do not differ from those concerned in haemolysis, and the method employed in 

 titration is essentially similar. 



A very light bacterial suspension is employed, and the serum under test, after its 

 natural complement has been inactivated by heat, is mixed in increasing dilution with 

 a constant amount of the bacterial suspension and a constant amount of complement. 

 The surviving bacteria in the mixtures are counted after varying intervals by some suitable 

 cultural method, and the highest dilution of serum that produces a significant killing effect 

 is noted. The method contains many possible sources of technical error. Gengou (1899), 

 for instance, drew attention to the fact that bacteria are agglutinated as weU as lysed 

 by a specific antiserum, and that this effect may greatly reduce the number of colonies 

 in a plate count, since a clump of bacteria will produce a single colony. When the original 

 mixtures are incubated for long periods before plating, this effect wiU be counterbalanced 

 by the multipHcation of those bacteria that have been agglutinated but not kiUed ; but this 

 possible source of error must be considered in estimating the significance of any reduction 

 of the viable count over a short period, and a control mixture containing the bacteriolytic 

 serum without complement must always be included in the test. 



It was noted by Neisser and Wechsberg (1901) that a marked pro-zone often 

 occurred in tests of this kind. A particular dilution of serum might exert no 

 bactericidal action, while a much higher dilution resulted in a complete killing of 

 the bacteria. This phenomenon was seized on by the Ehrlich school as an example 

 of the union of complement with free amboceptor, the hypothesis being that comple- 

 ment united indifferently with all the amboceptor present and that, when this 

 was present in excess, chance would favour the union of all the complement, or the 

 greater part of it, with the amboceptor that was not attached to bacteria. We 

 know, however, from our experience with precipitation and agglutination, that a 

 similar inhibitory effect is produced by a great excess of antibody in reactions in 

 which complement plays no part ; and it is clear that the use, in bactericidal tests, 

 of very thin bacterial suspensions ■^ill favour the frequent occurrence of zones of 

 gross antibody excess. The Neisser- Wechsberg phenomenon cannot, therefore, be 

 regarded as affording any support to the view that complement combines directly 

 with antibody. 



A point of considerable importance in relation to the bactericidal, or bacteriolytic, 

 reaction is that its occurrence depends in large part on the nature of the bacterial 

 cell. Certain bacteria, such as the cholera vibrio, the typhoid bacillus, and most 

 Gram-negative bacilli, are readily killed and lyscd when acted on by complement 

 after sensitization by a specific antibody. Other bacteria, such as the Gram- 

 positive cocci, are insusceptible to the direct killing action of antibody and com- 



