340 B. NAKIC, A. KASTELAN AND N. AVDALOVIC 



observed in the strain combination used in the present series of 

 experiments indicates that the difference in reactivity might be 

 under strict genetic control. This is further shov^n by the fact 

 that preimmunization of Wistar parabionts does not result in 

 increased incidence of tolerance in Y59 partners. It seems that the 

 only advantage possessed by the preimmunized animal remains 

 the one inherent in the genetic set-up of the relative strain 

 combination. Previous immunization may only accentuate the 

 difference in reactivity between the partners either as regards the 

 number or quality of immunized cells. This may be the answer 

 to the question why cells from a preimmunized donor can induce 

 "parabiotic disease" in a strain combination where cells from a 

 non-preimmunized donor cannot. 



More precisely, a situation favouring a stronger immunity 

 reaction in one direction than in the other would obtain in 

 combinations where the number of strong antigens possessed by 

 one partner exceeds that of the other partner. In a very simplified 

 form, it could be presented in the following way : 



x:AB-DEFG 

 y:A-CD--G 



where partner "x" lacks C antigen but possesses BEF antigens 

 not present in partner "y". In this donor-host bilaterally incom- 

 patible combination one could expect that partner "y" would 

 develop a stronger immunity reaction against partner "x" than 

 would be the case the other way round. Thus, when vascular 

 anastomoses are established between parabionts around the fourth 

 postoperative day, the number of blood-borne cells exchanged 

 may be the same, but the number of immunologically active cells 

 v^ be different, the more slowly reacting partner receiving a 

 greater number. The greater the damage inflicted on the 

 lymphopoietic system of the recipient by the transferred donor 

 cells the less will be its capacity to react. This immunological 



