9 6 



ALAN ARTHUR BOYDEN. 



sera would give us a truer picture of the relationships of rodents 

 to other mammals than would rabbit antisera. 



600 



100 



if . 



ioo7. so so ? sr o.i 



C 



SO So 3.1 j) li 



j., 



Ar.ti< 



1 



FIG. 5. Comparison of two anti-beef sera, one from a rabbit, the other from a fowl. 



Table X. shows the results obtained with four antisheep sera, 

 two obtained from rabbits and tw r o from fow r ls. Fig. 6 gives a 

 comparison of the two rabbit antisheep sera. While there is a 

 good correspondence in the tw'o sera obtained from rabbits, and 

 this in spite of a great difference in their strength, there is not 

 such a good correspondence in the chicken sera. Some of the 

 values obtained with the chicken sera correspond closely with 

 the others but especially in Cu no reaction for pig. horse, and 

 human blood was obtained at all. Apparently this chicken 

 produced no antibodies for these bloods. This seems evidence 

 a multiplicity of antibodies present in an antiserum produced 

 in response to the injection of a single blood serum. 



