296 SEX IN MICROORGANISMS 



variety are most readily explained by assuming interaction of pairs of 

 complementary surface substances and at least in the varieties that are 

 linked by intervarietal reactions, two homologous series of substances 

 may be assumed. Finally, as Sonneborn (1947) points out, De Garis' 

 (1935) report that P. aiirelia will mate with F. caudatum suggests the 

 interesting possibility that the two homologous series of mating types 

 may extend across species lines. 



The breeding pattern in F. calk'msi does not differ from the 

 systems in F. aiirelia and P. caudatwu. Wichterman (1951) finds four 

 isolated varieties in this species. Mating reactions and conjugation 

 occur only between the two complementary types within the varie- 

 ties. Breeding systems in F. ivoodniffl have not been described in 

 detail. 



Bursaria-Type Systems. Immediately after Sonneborn's (1937) 

 discovery of mating types in F. aurelia, Jennings (1938) examined 

 P. bursaria for similar breeding behavior. This study and those that 

 extended it (Jennings and Opitz, 1944; Chen, 1946a) revealed a very 

 interesting and different system of mating specificity in this form. 

 As mentioned above, sexually isolated varieties occur in P. bursaria 

 as they do in other species. However, the number of mating types 

 within a variety is not limited to two. When more than two mating 

 types occur in a variety they interbreed freely. Thus four mating 

 types occur in varieties I, III, and VI. Any one of the mating types 

 within one of these varieties will mate with the other three members 

 of the variety. Similarly, each of the eight mating types in variety 

 II will conjugate with the remaining seven types. 



The specificity relations in P. bursaria are therefore of an unusual 

 sort. There is a high order of specificity between the varieties, but an 

 apparent lower order of specificity within four of the six varieties. 

 However, if the multiple types are examined in terms of specific inter- 

 acting substances, a formal explanation for this apparent low-order 

 specificity is evident. The key to the problem lies in the observed 

 number of interacting mating types in a variety. This number con- 

 forms to the geometric progression 2", which in turn suggests that n 

 pairs of substances are involved in mating-type specificity. According 

 to the proposed scheme two independent pairs of specific, comple- 

 mentary, interacting substances would be required for a four-type 

 variety. Each mating type would possess two mating-type substances, 

 one substance of each independent pair. Assuming a random relation 



