230 PROTOZOA 



Since the group A system can arise merely by loss of two features 

 of the group B system, the repeated independent evolution of the 

 A from the B system seems not unreasonable. 



2. Evolutionary relations within groups A and B. If the three 

 different features which characterize all group A varieties and the 

 alternative three which characterize all group B varieties indicate 

 a fundamental and ancient evolutionary divergence of the two 

 groups, we should expect to find evidences of this even in those 

 features which distinguish varieties of the same group. For them, 

 we should expect closer relations among varieties of a group than, 

 on the average, between varieties belonging to different groups. 

 This is indeed found with respect to the two sorts of chemical 

 specificity expressed as antigens and as mating types. 



In the account given earlier of the serotypes, it was pointed out 

 that the specificities of the immobilization antigens were different 

 for each variety, but that the differences were least for varieties 

 4 and 8 of group B. Most of the antigens that occur in one variety 

 are matched by somewhat different but related antigens in the 

 other. Resemblances are detectable between several of the anti- 

 gens in these varieties and several of those in variety 2, the other 

 investigated group B variety, but the resemblances are fewer and 

 weaker. Of the three principal antigens in variety 1, two bear 

 some resemblance to the group B antigens, but these resem- 

 blances are recognizable serologically in relatively few strains. 

 Melechen ( 1955 ) reports that correspondences to all three prin- 

 cipal antigens of variety 1 are indicated by antigens in variety 3. 

 One of the two antigens found in variety 9 corresponds to one 

 of the three main antigens in variety 1. The serologic relations are 

 certainly closer for some varieties of the same group than for 

 some varieties of different groups. Whether this relation will 

 prove to be general remains to be discovered. 



The relations of mating type specificity are indicated by the 

 capacity of mating types of different varieties to interact sexually. 

 Table I shows these interactions. Within group A, they occur 

 among varieties 1, 3, 5, and 7, most strongly between 1 and 5. 

 Within group B, they occur among varieties 4, 8, 10, and 14, most 

 strongly between 4 and 8. The close relationship of varieties 4 and 



