144 PROPERTIES OF ANTI-SERUM TO SERUM 



serum necessary to produce deviation is, however, out of all 

 proportion smaller than the amount of red corpuscles (treated 

 with immune-body) necessary to show an appreciable 

 absorption of complement. There are also differences, to 

 be mentioned below, when the combination of the molecules 

 and anti-molecules is used in varying proportions in the 

 two cases. It has been shown by Morgenroth 1 that a 

 haemolytic serum may be developed by the injection of serum, 

 the latter apparently containing receptors with the same 

 combining-group as the hsemolytic receptors of the red 

 corpuscles. The anti-serum to ox's serum used by us has 

 hsemolytic action, the minimum haemolytic dose for ox'& 

 corpuscles being 0-05 c.c. ; as already stated, it gives devia- 

 tion of complement with 0-000,001 c.c. of ox's serum. The 

 hsemolytic serum acting on ox's corpuscles has a minimum 

 haemolytic dose of 0-001, 5 c.c. : it also gives deviation along 

 with ox's serum, but not with a smaller dose than 0-001 c.c. 

 of the latter. The former anti-serum has thus only about a 

 thirtieth of the haemolytic action of the latter, but has about 

 a thousand times more deviating power when tested with the 

 homologous serum. Furthermore, if the anti-serum to ox'a 

 serum be left for a time in contact with a sufficient amount 

 of ox's corpuscles, practically all the haemolytic immune- 

 body can be removed, but the precipitating and deviating 

 properties remain in the serum. It is thus evident that 

 the molecules in the serum which, in association with the anti- 

 serum, deviate complement in such experiments, are different 

 from the hcemolytic receptors. 2 



1 Morgenroth, Munchen. med. Wochenschr., No. 25, 1902. 



* Gay has suggested that probably many errors have arisen in haemolytic 

 experiments through non-recognition of the deviation of complement by a 

 serum + its anti-serum, and has pointed out the difficulty in freeing red 

 corpuscles completely from the serum by washing. Even, however, 

 if a small quantity of serum be left, the amount of anti-serum necessary 

 to produce deviation is relatively great, and as all the hsemolytic sera 

 which I have used have been powerful (0-003 c.c. being generally the 

 hsemolytic dose), I am certain that no error of importance can have arisen 



