512 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1961 



Biirnliam, 1941). Other systems, including the ABO system, are 

 sometimes involved. The problem of natural selection due to this 

 disease is interesting and important but a full discussion of it would 

 lead us too far from our main topic. It has however been shown 

 mathematically (Li, 1953) that it should not lead to the establishment 

 of a balanced polymorphism such as we discussed in the case of the 

 hemoglobins. Apart from hemolytic disease of the newborn, some 

 half-dozen diseases have been shown to have an association with par- 

 ticular blood groups, groups in all cases belonging to the ABO sys- 

 tem; the most marked example is the association between duodenal 

 ulcer and group O. There is, however, no evidence that, in the case of 

 any of the diseases studied, blood-group heterozygotes are relatively 

 favored as are the hemoglobin heterozygotes by malaria. Moreover, 

 none of the diseases proved to have a connection with blood groups 

 has an incidence sufficiently early in life to affect appreciably the 

 blood-group composition of the next generation. On the anthropologi- 

 cal side too, it is (fortunately for practical applications) true to say 

 that resemblances between populations known to be related but long 

 separated suggest that ABO frequencies are relatively stable for 

 periods of the order of 2,000 years. 



It would be unsafe, however, to accept the ABO blood groups, even 

 for periods of under 2,000 years, as completely stable population 

 markers. One possible cause of sudden large frequency changes is 

 epidemic disease. If the blood groups show a differential survival 

 among sufferers from any of the diseases responsible for major epi- 

 demics, frequencies may perhaps remain stable for periods of centuries 

 and then suffer sudden very large changes as a result of an outbreak 

 of one of these diseases, or of a series of outbreaks. That this is a 

 possibility is suggested by the recent work of Vogel, Pettenkofer, and 

 Helmbold (1960), who have examined the micro-organisms respon- 

 sible for plague and smallpox for the presence of antigens resembling 

 the blood-group substances. They find an antigen like that of blood- 

 group A in the smallpox virus, and in the plague bacillus an antigen 

 resembling the blood-group substance H which is most abundant in 

 group O cells. Basing their argument upon the hypothesis that an 

 individual will have difficulty in elaborating a protective antibody to 

 an organism antigenically resembling any of his own blood-group 

 substances, they suggest that group-A persons are particularly sus- 

 ceptible to smallpox and group-0 persons to plague. They compare 

 the world distribution of the ABO blood groups and of smallpox and 

 plague epidemics: these correspond sufficiently well to suggest the 

 desirability of further investigation of the hypothesis that such epi- 

 demics have played a major part in determining blood-group 

 distribution. 



