538 IMMUNITY 



ment which may, however, show slight variations in behaviour 

 towards different immune-bodies. There is at least no doubt 

 that all the complement molecules in a serum are not the same. 

 For example, Muir and Browning have shown that the treat- 

 ment of a normal serum with a small amount of emulsion of a 

 bacterium will remove the bactericidal action for another 

 bacterium, whereas the amount of complement as tested by 

 haemolysis is practically unchanged. They accordingly con- 

 sider that there is a moiety of complement, " bacteriophilic 

 complement," which is specially concerned in bactericidal action. 

 On the other hand, many of the arguments adduced by Ehrlich 

 and his co-workers in favour of a multiplicity of complements 

 are open to another interpretation ; the truth probably lies 

 between Ehrlich's and Bordet's views. Workers of the French 

 school also hold that complement does not exist in the free 

 condition in the blood, but is liberated from the leucocytes when 

 the blood is shed. This cannot be held as proved. On the 

 contrary, there are facts which are strongly in support of the 

 view that complement exists in the free condition in the circu- 

 lating blood. There is, however, evidence that the amount of 

 free complement increases after the blood is shed and some time 

 later gradually diminishes. 



The hsemolytic action of a normal serum can be shown in many cases 

 to be of the same nature as that of an immune-serum, that is, comple- 

 ment and the homologue of an immune-body can be distinguished. For 

 example, guinea-pig's serum is hsemolytic to ox's corpuscles ; if a portion 

 of serum be heated at 55 C. the complement will be destroyed ; if 

 another portion be treated with ox's corpuscles at C., the natural 

 immune-body will be removed and only complement will be left. 

 Neither portion is in itself hremolytic, but this property becomes manifest 

 again when the two portions are mixed. Hsemolytic sera are of great 

 service in the study of the question of specificity. Each is specific in the 

 sense already explained (p. 521), but the serum developed against the 

 corpuscles of an animal may have some action on those of an allied 

 species, that is, some receptors are common to the two species. This fact 

 can be readily shown by the usual absorption tests, for example, in the 

 case of an anti-ox serum tested on sheep's corpuscles. A close analogy 

 holds to what has been established in the case of agglutinins. It is 

 further of great interest to note that by the injection of red corpuscles 

 into an animal its serum not only becomes hremolytic, but in many cases 

 when heated at 55 C. possesses also agglutinating and opsonic properties 

 towards the red corpuscles used. And further, it would appear that in 

 some cases at least the immune-body, haemagglutinin, and heemopsonin 

 are distinct substances. These facts abundantly show how close an 

 analogy obtains between anti-bacterial and haemolytic sera, and how 

 important a bearing hsemolytic studies have on the questions of im- 

 munity in general. 



In addition to hsemolytic sera, anti-sera have been obtained by the 



