ENDOMIXIS 661 



obtained by Sonnebom (1939), in particular, afford evidence that the 

 preciseness of the ratios and the segregation of sex or mating types, 

 following the reorganization process, is quite as regular and exact as 

 after conjugation. Thus it would appear that autogamy and not endo- 

 mixis is involved. Indeed, Sonneborn states that in one race of P. aurelia 

 — not the Yale race nor the mating-type variety which it represents — 

 the alternative between endomixis and autogamy was tested genetically 

 by determining the genotypes following reorganization in clones of type 

 Aa (genes determining mating types). Genetic analysis of the reorgani- 

 zations showed that all the resulting lines are homozygous, half of them 

 dominant and half recessive. From any one reorganizing individual both 

 caryonides are of the same genotype. Thus, under these conditions (mass 

 cultures at 31° C. and isolation lines at 27° C), autogamy, not endo- 

 mixis, takes place. 



Accordingly the combination of genetical and cytological data at pres- 

 ent available justifies the conclusion that autogamy occurs, under cer- 

 tain circumstances at least, in some races of P. aurelia. Granting this 

 raises the question whether all the investigations reported on the 

 physiology and genetics of endomixis actually are on autogamy, because 

 Paramecium is the form that has been almost universally employed in 

 such studies. If the accumulated data are really all in regard to autogamy, 

 then the question is essentially one of name. On the other hand, if both 

 endomixis and autogamy occur in Paramecium, then, for a time, confu- 

 sion is worse confounded. 



Obviously, at present it is useless to attempt to generalize in regard 

 to reorganization in Paramecium — that must await far more extended in- 

 vestigation. However, the personal judgment of the writer, at the mo- 

 ment, is that both endomixis and autogamy do occur in Paramecium — 

 an opinion reached, it is believed impartially, from a consideration of 

 the picture of the micronuclear divisions and the macronuclear destruc- 

 tion, as he saw them in the original work on the reorganization process 

 and as demonstrated in the cytological preparations of Diller. Certainly 

 the two pictures presented are quite different; indeed, in many ways as 

 different as the stages in endomixis and conjugation appeared in the 

 original study. And Sonneborn (1939a) remarks: "Probably it will be 

 found that autogamy and endomixis take place in different races or under 

 different conditions." The occurrence of these two processes, either in 



