T. M. SONNEBORN 273 



genetic basis of mating type specificity (see below) do not at 

 all preclude the existence of more types. One might actually 

 expect a very large number of types in nature, as will appear. 



Mating Relations and Inheritance of Mating Type. The mat- 

 ing relations reported by Kimball for the six mating types he 

 found seem superficially to be like those in P. bursaria, but they 

 probably differ in several respects. Conjugation occurs when any 

 two of the six types are mixed together. However, by using 

 visibly marked animals, Kimball proved that the conjugant pairs 

 are not always crosses between the two types that were mixed 

 together. In some combinations of cultures, one finds both crosses 

 and selfing. In fact, if one adds animals of one mating type to 

 fluid in which animals of certain other types have lived, the one 

 and only type introduced into the fluid selfs. The animals clearly 

 secrete "hormones" that induce conjugation in animals of certain 

 other types. Sonneborn (1947) concluded from a detailed exam- 

 ination of the data of Kimball (1942) and Powers (1943) that 

 crosses occur only when each type secretes into the fluid a hor- 

 mone which induces the other type to conjugate. As Kimball 

 showed, in such mixtures three kinds of conjugant pairs are 

 formed: crosses between the two types and selfing of each type. 



This system has a simple genetic basis and remarkable conse- 

 quences. Altogether, among the six mating types, only three 

 hormones are produced. Each hormone is determined by a single 

 gene, and the three genes are codominant alleles. There are thus 

 three homozygotes, each producing one and a different hormone; 

 and three heterozygotes, each producing two of these same three 

 hormones in a different combination. These are the six mating 

 types. Because mating type is directly controlled by the geno- 

 type and each mating type has a different genotype, each syn- 

 clone is necessarily uniform for mating type because conjugation 

 has rendered it uniform in genotype. Without going into further 

 details, the system works out so that crosses between different 

 synclones take place (along with double selfing) only in three 

 of the five possible combinations of any one mating type with 

 the others. In the other two combinations one member selfs, the 

 other does not, and no crosses occur. 



