136 DISCUSSION 



tolerance produced by neonatal inoculation of isologous cells can 

 consistently be abolished by inoculating each of these tolerant animals 

 with immunologically competent cells (from the spleen and lymph 

 nodes) derived from a single female which has rejected a male graft. 



On the other hand, C57 females which accept male grafts after 

 having six or seven litters do not "pass" this tolerance on to normal 

 female parabionts. Furthermore, this tolerance cannot be abolished by 

 adoptive immunization with as many as three donor equivalents of 

 lymphoid cells. This may suggest that these animals are not chimeras 

 although the possibility of the persistence of Y antigen at very low 

 levels is not ruled out. 



Billingham: Prof Medawar, haven't you been a little too disparaging 

 about routes of immunization in considering pregnancy and other 

 problems ? I would agree that with non-cellular and non-living anti- 

 gens your points are well made, but it seems to me that it is only in the 

 rabbit that there is evidence of a peculiarity with regard to the sen- 

 sitization by the intravenous route; namely that homologous leuco- 

 cytes injected intravenously in rabbits do not sensitize and may cause 

 prolongation of survival of subsequent skin homografts and likewise 

 with epidermal cells. But as Brent, Mitchison and I found, if you inject 

 suspensions o{ splenic cells intravenously into rabbits then you do get 

 sensitization. In guinea pigs, hamsters and mice the intravenous route is 

 a perfectly effective sensitizing route with homologous leucocytes. I 

 am wondering, therefore, whether the intravenous route is as peculiar 

 as you have implied. 



Medawar: I think it is not only the rabbit. As I mentioned earlier, 

 repeated intravenous transfusions of whole blood fail to sensitize 

 human beings. However, what I said in my paper refers only to 

 extracted antigens in a fairly soluble form. I think that point ought to 

 be made quite clear. 



Lawrence: The interference with the development of homograft 

 sensitivity resulting from the intravenous access of tissue antigen 

 suggests many analogies to a similar situation in relation to delayed 

 hypersensitive states where bacterial cells serve as antigen. In each in- 

 stance, intact cells are the vehicles of antigen and in each instance the 

 intravenous route produces an effect quite different from that produced 

 by intradermal, subcutaneous or intraperitoneal routes of administra- 



