128 PROPERTIES OF BLEMOLYTIC SERA 



when the twenty-four hours' fluid was filtered first the 

 corresponding filtrate contained the fewest receptors of the 

 three, and the corresponding condition obtained when the 

 quarter of an hour fluid was filtered first. We therefore 

 found no evidence that in course of time a greater number 

 of receptors passed into the surrounding fluid when the 

 corpuscles were laked with water. 



The explanation of the results stated cannot be con- 

 sidered as quite clear. Of course, as filtration goes on the 

 pores of the filter gradually become choked with stromata 

 and a higher pressure becomes necessary ; in this way more 

 of the receptors might be forced through. (On the other 

 hand, it might a priori have been considered likely that 

 more receptors would have passed through when the pores 

 of the filter were freer.) It is, however, also possible that 

 the receptors become deposited in some mechanical way 

 on the filter, and we are at present not justified in con- 

 cluding that the size of particles alone is the determining 

 factor. Experiments which we have made on the filtration 

 of fresh serum show that the amount of complement which 

 passes through varies greatly : in some cases the comple- 

 ment falls only slightly in value, in other cases the greater 

 part is kept back. Immune-body, on the other hand, 

 appears to pass through practically unchanged. Further 

 results obtained in a subsequent research are given 

 above (p. 90). It may be noted that Graham-Smith 1 

 found on filtering serum through a Berkefeld filter that the 

 amount of precipitable substance in the filtrate first passed 

 was diminished ; but that it gradually rose to the original 

 amount after a certain quantity had been passed. In the 

 case of a Chamberland filter the amount of precipitable 

 substance diminished rapidly and fairly uniformly as the 

 filter became choked. 



1 Graham-Smith, Journ. of Hygiene, 1903, vol. iii, p. 357. 



