I 



IMMUNITY IN PNEUMOCOCCUS INFECTIONS 19 



The following experiment solves this question. A large quantity of 

 blood was drawn, and the corpuscles driven down to the bottom of the 

 capsule by centrifugalisation. Quantities of the serum were then drawn 

 off at intervals, and stored in sealed glass pipettes. The contents of 

 these pipettes were all tested simultaneously with normal corpuscles and 

 bacterial emulsions. 



Serum removed from clot after 5 mins. - - - 6-7 



- - - 6-9 



7'5 



- - - 7-8 

 8-8 

 97 



lO'I 



10*7 

 10-3 



lO'I 



- 10-5 



IO*2 

 107 



Thus, contact with the clot seems to be necessary for increase in the 

 opsonising power. This increase was observed in every instance, and it 

 seems a constant law that the opsonic power of the serum increases up 

 to a period of two hours, after which it gradually diminishes so that, by 

 the seventh or eighth day, it has lost 50 per cent, of its opsonin. 



A probable explanation of this property is that the clot acts as a 

 permeable filter, which, as it contracts, squeezes out constituents in a 

 definite order, according to the size of the molecule : 



1. Sodium chloride and salts ; 



2. SemicoUoids ; 



3. True colloids, to which there is much evidence that substances of 

 the opsonin type belong. This hypothesis is substantiated by the fact 

 that the serum squeezed out last from the contracting clot contains more 

 opsonin per volume than that squeezed out at any earlier period in the 

 contraction. 



From these figures, it is evident that, before the sera of two indivi- 

 duals can be compared accurately with regard to their opsonic indices, 

 they must be subject to the same physical conditions for similar periods. 



(383) 



