MA riXC; ]\P\ 1)1.1 I'RMINA 1 ION 267 



an arrcnipr w as iiiadc to give a complete, if formal, explanation for 

 these and other problems. Although this explanation appeared prom- 

 isino- and early attempts at verification were encouraging (Sonnebom, 

 1951 ), more recent studies have shown that the interpretation is cer- 

 tainl\- not correct in detail. On the other hand, strong support has 

 been received for the more general features of the model proposed. 

 Before considering any explanation in detail, it is first of all nec- 

 essary to revie\\' the available information concerning mating type 

 determination in this species. Many of the observations are not avail- 

 able in the literature, and the author is deeply indebted to Prof. T. M. 

 Sonneborn for the use of his unpublished data. It is to be understood 

 that the present treatment of Sonneborn's unpublished observations 

 is subject to his later modification or extension. After reviewing the 

 observations we will return to a consideration of detailed explana- 

 tions. 



PATTERNS OF MATING TYPE DETERMINATION AND 

 INHERITANCE 



General Background, the Varietal System and 

 Cytogenetic Considerations 



Paraviec'min aiirelia is a "taxonomic" species composed of at least 

 eight so-called varieties which are themselves effective genetic spe- 

 cies (Sonneborn and Dippell, 1946; Sonneborn, 1947, 1950a). (See 

 also Metz, 1954.) Each variety contains no more than two mating 

 types and in relatively few combinations can these conjugate with 

 the mating types of other varieties. Such intervarietal conjugations as 

 are completed usually result in the death of the Fl or F2 generations. 

 However, enough intervarietal combinations give some degree of 

 reaction to provide evidence of mating type homologies for most 

 of the varieties. The two mating types within a single variety are 

 designated by Roman numerals, an odd and an even number, and it 

 is clear that the odd mating types in most varieties are similar though 

 not identical. The even mating types in most varieties are also similar 

 but not identical. 



The eight varieties fall into two readily distinguished classes, 

 those \\ hich have been characterized as showing little cytoplasmic 

 inheritance: Group A, composed of varieties 1, 3, 5, and 7, and those 



