T. M. SONNEBORN 261 



B of P. aurelia, but with regular exchange of cytoplasm between 

 mates. The existence of 11 mating types in variety 2 would, if 

 our hypothesis is correct, indicate a flux equilibrium among 11 

 cytoplasmic differentiators. Although this might seem extraordi- 

 nary and unduly complicated, it should be remembered that equal 

 complexity in flux equilibrium is present in the serotype system 

 of the group B varieties of P. aurelia. Nanney (1956b) has 

 pointed out the detailed resemblances and differences between 

 the mating type and serotype systems. Variety 2 of T. pyriformis 

 may well prove to be the material of choice for testing the inter- 

 pretations proposed above for P. bursar ia. 



Mating type inheritance in variety 1 has been analyzed by 

 Nanney and Caughey (1953, 1955), Nanney, Caughey, and 

 Tefankjian (1955), and Nanney (1956b). In contrast with va- 

 riety 2, variety 1 shows the group A pattern of mating type 

 determination and inheritance. Leaving selfers aside for the mo- 

 ment, each caryonide seems to be independently determined for 

 mating type. Hence, among the four caryonides from a pair of 

 conjugants, there are usually present hereditary differences in 

 mating type; in fact, as expected, even the two caryonides from 

 a single exconjugant are usually different in mating type. With 

 seven possible mating types, a large number of combinations is 

 found among the two caryonides from an exconjugant or among 

 the four from a pair of conjugants. 



There are, however, genie restrictions as to which mating type 

 potentialities can be realized. One gene permits the development 

 of any type except I; another gene permits the development of 

 any but IV and VII. These genes behave as alleles. The authors 

 point out some difficulties raised by these discoveries with respect 

 to Metz's scheme of pairs of complementary mating type sub- 

 stances (see earlier discussion under Paramecium bursaria), but 

 realize that these difficulties need not be fatal to Metz's hy- 

 pothesis. The mating type potentialities may be modified by 

 genes at different loci which act in different ways. 



3. The life cycle. There appear to be some varietal differences 

 in the length of the immature period, but the literature contains 

 conflicting statements of the details. Part of this is clearly due 



