IMMUNITY IN TUMOR TRANSPLANTATION 427 



We have discussed those aspects of immunity against transplanted tumors, 

 in which organismal differentials function as antigens. It may be further 

 stated that the organismal differentials in tumors, are essentially the same 

 as those of the normal tissues from which the tumors are derived. There are, 

 however, indications that other substances present in tumors, besides the 

 organismal differentials, may be antigenic. A brief outline of some of the 

 principal data which point to the presence of these secondary antigens will 

 now be given. 



(7) The presence of antigens other than organismal differentials in tumor 

 cells. By serological tests the same types of antigenic constituents have been 

 found in certain cancers, which normal cells in corresponding organs of the 

 same species possess, namely, species-specific, organ-specific, blood-group 

 and heterophilic Forssman antigens ; also alcohol soluble substances corre- 

 sponding to Wassermann antigens not characteristic of either organ or species 

 may occur. Thus the cells of a carcinoma developing in individuals belonging 

 to blood group A may contain these same blood-group antigens and the partial 

 Forssman antigens which are associated with blood group A. Mouse carcinoma 

 may contain Forssman antigen, in accordance with the fact that the mouse 

 belongs to the group of those species which possess heterophilic antigens. 

 However, there are apparently some exceptions to this parallelism between 

 normal tissues and tumors. According to Kritchewski and Rubinstein, also 

 the Flexner-Jobling rat tumor contains Forssman antigens, although normal 

 rat organs do not contain them ; this would constitute a difference between 

 cancerous and the corresponding normal tissues. Moreover, while human 

 carcinoma cells of individuals belonging to blood group II, like some normal 

 cells, may possess the A differential, it has been stated that it is never found 

 in sarcoma cells; but it is not certain that normal connective tissue cells of 

 such individuals contain it. Hence we find in cancer cells a complex condi- 

 tion, which makes the search for constituents characteristic of carcinomatous 

 cells, and not present in normal cells, difficult, and this may explain at least 

 in part the contradictory nature of some of the results obtained by various 

 investigators. Some of these obstacles to the discovery of specific tumor anti- 

 gens were overcome by using for immunization carcinomatous material from 

 persons belonging to blood group I, which is free of antigens A and B, or by 

 first extracting the blood group antibodies by means of erythrocytes possessing 

 group A. Furthermore, the attempt was made to modify the material to be used 

 for immunization by destroying or eliminating the species and normal organ 

 antigens before injecting the antigen. Hirszfeld and his collaborators found 

 that by immunizing rabbits with human carcinoma of a certain organ, immune 

 sera developed in a small minority of the rabbits which reacted with different 

 types of human carcinomas, irrespective of the organs in which they origi- 

 nated. Witebsky and Lehmann-Facius, by using boiled, instead of fresh, 

 unheated carcinoma suspensions as antigens, obtained antibodies which were 

 specific for carcinoma, and not merely for the species or the organ in which 

 the cancer occurred. Witebsky, in addition, used boiled globulins of cancer 

 tissue in these tests. According to Lehmann-Facius, the complement fixation 



