STUDIES OX H.EMOLYSINS. 97 



blood-cells are used, the group 7- will be left behind, for this possesses 

 affinity only for the goat blood-cells. 



The following protocol shows the results of two series of experi- 

 ments, which exhibits the effect of such reciprocal binding: 



To each 5 cc. of a o^J mixture of ox blood-cells or goat blood- 

 cells (freed from serum by centrifuge) varying amounts of the immune 

 body of a rabbit which had been immunized with ox blood are 

 added. The amount of immune body is seen in the first column 

 of the table; in the second and third columns the complete solvent 

 doses (for ox blood and for goat blood) contained in each specimen 

 are given, as they were determined by tests made at the same time. 

 The mixtures are made up to 6 cc. with physiological salt solution, 

 kept at 37 C. for 1J hours and then centrifuged. Two equal por- 

 tions of each of the decanted fluids are then taken and again mixed 

 with corresponding amounts of blood-cells. Finally guinea-pig 

 serum is added to activate the mixtures. The haemolytic action 

 which the decanted portions exerted on ox blood-cells and on goat 

 blood-cells is seen in the table. (See Table III.) 



The union of the immune body with the ox blood-cells has resulted 

 in a considerable abstraction of both portions of immune body. On 

 the other hand, the union with goat blood-cells, by which the action 

 of the fluid is considerably decreased for goat blood-cells, causes very 

 little decrease in the solvent power for ox blood. 



In contrast to this experiment we here reproduce an analogous 

 experiment which shows a directly opposite behavior of the two 

 fractions of immune body of a rabbit immunized with goat blood. 

 (See Table IV.) 



Here the goat blood-cells bind both portions of the immune body, 

 while after treatment with ox blood-cells the fraction acting on goat 

 blood is left almost intact. 



Hence by means of this crossed immunization and reciprocal elec- 

 tive absorption we succeed in demonstrating that in the case of the 

 rabbits treated respectively with goat blood and ox blood two 

 large fractions of immune bodies can be separated. Of these, one 

 fraction is common to both sera; the other is peculiar to each of 

 them. The main groups of receptors shown in the above illustra- 

 tion and designated a and ft for ox blood, and ft and 7- for goat blood, 

 are thus to be differentiated. 



We have deemed it important to supplement this analysis by 

 experiments on a second species of animal, and have therefore treated 



