BY H. G. CHAPMAN. 



539 



times. The tubes of a series have been weighed immediately 

 after each other. Owing to the hygroscopic nature of the pre- 

 cipitates the error in the weights of the tubes has been found by 

 experiment to be 0-3 mg. The mean figure of the several weigh- 

 ings has been taken in all cases. 



Experiments were performed to ascertain the weight of pre- 

 cipitate obtained when a measured quantity of antiserum was 

 allowed to interact with increasing weights of homologous pro- 

 tein. It has already been shown {supra vide) that no precipitin 

 can be detected in the superfluid at the end of an interaction, 

 provided that the amount of homologous protein exceeds a certain 

 quantity. Table vi, records some of the results obtained by 

 testing the superfluid with a further addition of protein and 

 noting the presence or absence of a further precipitate. In the 

 series to be described, the quantity of protein was sufficient to 

 neutralise or precipitate* the precipitin in the antiserum. The 

 superfluids were considered free from precipitin since they yielded 

 no precipitate on the addition of 144 rag. dried egg-white. The 

 results are recorded in Table ix. 



Table ix. 



Here tlie antiserum was formed by the injection of hen's egg- 

 white, and the fresh antiserum allowed to interact with fresh 

 effw- white. A portion of the egg-white was dried to determine 

 the solid content of the solution. The precipitates are stated in 

 terms of the amount yielded with 1 c.c. antiserum. The weight 



* Compare Welsh and Chapman, Proc. Roy. Soc. London, B.78, p.297, 1906. 



