140 On the Specificity of Precipitins 



treated rabbits with the serum-globulins of man, horse, mule and sheep, 

 and found that the antisera acted on solutions of all these globulins, 

 only to a different degree. On the other hand he observed " scarcely any 

 quantitative differences" upon adding anti-horse or anti-mule sera to 

 both of the corresponding globulin solutions. Much less, and quantita- 

 tively decreasing in the order named, were the reactions obtained with 

 similar solutions from the ox, sheep, and man, this again being in accord 

 with my results. That different substances are acted upon in homologous 

 and non-homologous sera (see Strube, quoted above) is clear, for as 

 Ascoli found, if non-homologous serum is added in excess after precipi- 

 tation has occurred, no more precipitum forms, but on the addition of 

 homologous antiserum one obtains further precipitation. According to 

 Ascoli, for example, when anti-horse serum produces a large reaction with 

 horse blood and a small one with human blood, there are two possibilities 

 to be considered regarding the differences of reaction. Either the anti- 

 serum contains several (in this case two) precipitins acting on different 

 constituents of each serum, the one being common to both horse and 

 man, and the other present only in the horse ; or the precipitin is simple, 

 there being less precipitable substance in the human serum. The latter 

 supposition is not borne out by experiment, for even when we add an 

 excess of anti-horse, the human serum continues to give less precipitum. 



Uhlenhuth (11-18, ix. 1902, p. 661) dwells upon the fact that the 

 zoological relationship of animals is brought out by an antiserum ; he 

 however does not bring new facts to bear on the question. 



Further details regarding the specificity of the precipitin reactions 

 are to be found under Tests with different antisera, in Section VI 

 which follows. As will be seen, my most generalized reactions amongst 

 the mammalia were obtained by a powerful anti-pig serum. 



It is evident from the foregoing that there are many points of 

 similarity between the precipitins and the haemolysins. Ehrlich has 

 pointed out that the haemolysins of different species may possess 

 receptors which are identical but quantitatively different. 



The Selective Action of Precipitins in Blood Mixtures. 



In my paper of 1, vil. 1901, p. 384 I described some tests made 

 with a view of determining whether or no a mixture of several kinds of 

 blood in solution would prevent a reaction taking place upon the 

 addition of an antiserum which was effective when added to one of these 

 bloods when alone in solution. It was found that when two to six 



