Blood Relationship 139 



ox serum. Here again my observation was confirmed. Referring to 

 the specificity of the reaction, Linossier and Lemoine (loc. cit ) remark 

 "La ou on a cru voir une action spe"cifique, un examen attentif ne 

 perniit de voir qu'une action particulierement intense," which sums up 

 the question in few words. 



Passing over my paper of 5, IV. 1902 (see below) in which I described 

 my method for measuring the degrees of reaction, by volumetric 

 estimations of the precipitum formed, I shall mention some more 

 papers referring to the subject of these generalized reactions. Strube 

 (12, vi. 1902) confirms my observation that the stronger an antiserum 

 is the more powerfully does it act upon non-homologous bloods. He 

 thinks that the quantitative differences of reaction may be explained by 

 assuming that different species of blood do not possess an identical 

 constitution (which is a self-evident proposition) but a closely related 

 constitution, with the result that the antiserum for one reacts but to a 

 limited extent upon the other, as in the case of the agglutinins. 

 Secondly, let us assume that the serum is composed of different 

 albumins (and there appears to be evidence of this) which we shall 

 style a, b, and c. Substance a is present in other animals, but b and c 

 are not. With the homologous antiserum a, b, and c are precipitated, 

 whereas in the non-homologous bloods only is precipitated. He offers 

 this in explanation of my "mammalian reaction," and we have no other at 

 present. Strube, I might add, makes no mention of a time limit, saying 

 only that " several hours " may be necessary for reaction to take place. 

 Test sera which take several hours to act would however be scarcely 

 desirable in practice. Whereas Strube found anti-human serum to 

 produce an equal, though slight, reaction in the bloods of the ox, sheep 

 and pig, Kister and Wolff (18, xi. 1902) claimed to have noted that 

 human blood is more strongly acted upon by anti-horse and anti-ox than 

 by anti-sheep or anti-pig sera, this observation being evidently due to 

 some experimental error. They also note that differences are not well 

 observed in concentrated solutions, or when a too large proportion of 

 antiserum is added (note in this connection what I have said on 

 page 74). They note, as I have also observed, that clouds may 

 form in concentrated solutions, whereas flocculent precipitates will 

 form in more dilute solutions. Oppenheimer and Michaelis (18, VII. 

 1902), working with but a limited number of bloods, conclude that 

 precipitins are specific for a " bestimmte Thierart." They evidently 

 overlooked the generalized reactions. 



Ascoli (26, vin. 1902) has also observed generalized reactions. He 



