104 The Nature of Precipitin Reactions 



that clouding occurred at the top of the tubes, and also took longer in 

 forming." 



Graham-Smith and Sanger (p. 266) moreover studied the behaviour 

 of aqueous blood dilutions to antisera, and report thereon as follows : 



" It has been noticed by many observers that solutions of fluid, or 

 dried, sera in distilled water become cloudy, and that after 24 hours a 

 precipitate occurs. In '5 c.c. of a 1 in 21 dilution of human serum in 

 distilled water this precipitate amounts to about '001 c.c. We have, 

 however, found that including this precipitate "1 c.c. of human 

 antiserum produces a smaller quantity of precipitum with blood diluted 

 with distilled water than with the same specimen diluted with normal 

 salt solution. The mean of three experiments in each case gave 

 0384 c.c. of precipitum in salt solution dilutions and '0328 c.c. in watery 

 dilutions." 



Regarding the Claim that Precipitins permit of Distinguishing 

 different Albumins of the same Animal. 



The investigations of Nolf (v. 1900) indicated that precipitins are 

 formed for serum-globulin and not for serum-albumin and solutions 

 of blood corpuscles. Leclainche and Vallee (25, I. 1901) however state 

 that antiserum for serum-albumin is very active for its homologous 

 substance, but almost indifferent towards globulins contained in the 

 albuminous urine with which they experimented. The urine from three 

 cases of interstitial nephritis gave marked reactions, whereas the urine 

 from a case of parenchymatous nephritis, which contained much globulin, 

 gave but slight precipitation. Albuminous horse and cow urine gave 

 no reaction. They add that human pleuritic exudate gave a reaction, 

 but not human blood serum when antiserum for human albuminous 

 urine was added thereto. Mertens (14, in. 1901) treated rabbits with 

 human serum and found the antiserum to precipitate albumin in urine, 

 concluding therefrom that the latter must be derived from the blood. 

 The urine reacted both to this antiserum and to one obtained by treating 

 the rabbits with human albuminous urine. Dieudonne (2, IV. 1901) 

 found that peritoneal exudate and blood of man reacted similarly to 

 anti-human serum. Zuelzer (4, iv. 1901) treated rabbits with urine con- 

 taining 1 9 /oo albumin, and obtained antisera as did Mertens, to whose 

 generalization regarding the origin of albumin in the urine from the 

 blood he objects, although he considers that his results warrant the 

 conclusion that at least one albuminous body in urine is derived from 



