150 Normal Precijntins 



NORMAL PRECIPITINS. 



Noguchi (xi. 1902) and M. Ascoli (1903), report having found preci- 

 pitins in certain normal sera. I have noted elsewhere that normal sera 

 may contain antibodies, which are usually present in small amounts only. 

 This appears also to be the case with the normal precipitins. Ascoli was 

 stimulated to search for normal precipitins by a statement of Obermayer 

 and Pick (see p. 97), that they had found normal rabbit serum to preci- 

 pitate egg-white. Ascoli found ox serum to contain precipitins for the 

 sera of man, dog, pig, goat, rabbit, guinea-pig, fowl ; dog serum to contain 

 precipitins for the serum of the fowl, and for egg-white; goat serum 

 contained precipitins for fowl and guinea-pig serum. He judged of 

 the occurrence of a reaction by observing if any clouding occurred at 

 the zone of contact between a serum and serum dilution, made in 

 0*85 7o saline. He used small tubes 4 cm. long by 3 mm. wide, and 

 observed the reactions after the tubes had stood 30 — 60 minutes 

 at 38"' C. By diluting the one or the other of the normal sera he 

 was able to note which contained normal precipitin, and in some 

 cases found both sera to contain it. As has been observed for other 

 normal antibodies, the amount of normal precipitin present varies 

 considerably in different animals of the same species. 



Ascoli also states that he has found " autoj^recipitins " in normal 

 human and animal sera, comparable to the auto-agglutinins found by 

 him and other observers. 



The occurrence of the precipitins in normal sera may account for 

 some aberrant results obtained by various authors, but they do not 

 appear to affect my results in the slightest. Although Ascoli does 

 not state with what concentrations he worked, any more than do 

 Obermayer and Pick, it amounts to a certainty that but low dilutions, 

 and, in the latter case, even concentrated sera were used. That the 

 use of concentrated sera or dilutions constitutes a grave source of error 

 when making tests with specific precipitating antisera has already 

 been pointed out, the source of error being apparently in part 

 the normal precipitins. The stronger an antiserum the greater will 

 be the dilution of homologous blood with which it will react, and it 

 appears now all the more necessary to work with the highest possible 

 dilutions when attempting to identify a blood. This being in accordance 

 with what Wassermann (10, ii. 1903) has just found in the case of the 

 agglutinins for bacteria. 



The furegoing remarks apply also to the observations of Noguchi 



