130 Behaviour of Precipitins in the Body 



and one of the others were poured together, when to our astonishment 

 a massive precipitation occurred. Fortunately small separate samples 

 of these sera had been preserved which enabled us to make further tests 

 to see if we could find a reason for this remarkable behaviour of the 

 antisera. On mixing the samples we found that 



Sera I. and III. gave no reaction, 

 „ II. and I. gave marked reaction, 

 ,, II. and III. gave powerful reaction. 



Consequently it was due to some peculiaritii in Serum II., and it 

 seemed to us that the only reasonable explanation to be found was 

 that some of the precipitahle substance, viz. crystallized horse albwtnin, 

 ivas actually present in the rabbits' serum. The matter was not pursued 

 farther at the time, but subsequently I directed the attention of 

 Mr Strangeways in our Laboratory to this observation, and asked 

 him to mix various antisera for human blood which he had made, 

 especially those taken from rabbits which, through prolonged treatment, 

 had shown a decrease in the amount of precipitin. On adding a 

 certain anti-human serum to another, he also obtained a precipi- 

 tation, and this was interpreted in the same manner as before. In 

 the latter case, the antiserum which contained precipitin likewise 

 contained precipitahle substance. It appears somewhat remarkable 

 that precipitation does not occur in such antisera on standing in sealed 

 tubes ; perhaps the explanation of the reaction which takes place with 

 a similar antiserum from another animal of the same species will be 

 found to depend upon an individual difference in the constitution of 

 the precipitin. It is however premature to draw further conclusions. 

 The matter certainly deserves further enquiry. Owing to stress of 

 other work, it has been impossible for me hitherto to pursue the 

 question, but I hope to do so shortly. I have not ))ublished this 

 observation before for the reason that I hoped to have had more data 

 before doing so. The reason that I do so now is that similar observa- 

 tions have been made by others, as follows. 



Obermayer and Pick (1902) find that when egg-white is injected 

 intraperitoneally into a rabbit which has precipitins for this substance 

 in its serum, both precipitin and egg-white may be present in its 

 blood, and, although this is the case, no precipitation occurs. Ascoli 

 (26, VIII. 1902) also found that precipitin and precipitahle substance 

 may coexist in the serum of immunified animals. He thinks the 

 condition is similar to that Avhich some claim to have observed in vitro, 



