CYTOPLASMIC INHERITANCE 27-8 



about one -third as many colonies appeared on the C plate, and of 

 these about 90 per cent were large and 10 per cent were small. In 

 all transfers a large number of cells die; from 10 to 30 per cent of 

 the cells transferred from M to M usually produce colonies, but 

 the mortality is always greater on the first transfer from M to C 

 than on transfers from M to M. Since the genotype is stabilized, 

 selection must be for some difference, independent of the genotype. 



A suspension was made of a large colony growing on a C plate, 

 and equal amounts of the suspension plated on M and C plates. The 

 same number of colonies appeared on both plates and all were full 

 size. This proved that the survivors on the C plate were adapted 

 to growth on C agar. A suspension of a small colony gave the same 

 result. The fact that there is no detectable difference between the 

 large and small colonies on C plates indicates that the small col- 

 onies are merely slower in development. If a colony is delayed in 

 development, the staling effect of the more rapidly developing col- 

 onies on the medium will prevent it from attaining full normal size. 



Equal amounts of a suspension from one of the large colonies 

 on an M plate (descended from a colony on a C plate) were spread 

 on M and C plates. Only one -fifth as many colonies appeared on 

 the C plates as on the M plate and both large and small colonies 

 were found. Therefore, cells growing on a C plate (which have be- 

 come adapted to C agar) lose this adaptation by a single transfer to 

 M agar. This suggests that the variations have affected a large pro- 

 portion of the population and have been readily reversed. 



It was also possible to adapt the cells to C agar by holding them 

 in C broth in the cold room for two days. A large colony from an 

 M plate was suspended in C broth and held two days in a cold room. 

 When samples from this culture were spread on M and C plates the 

 same number of colonies appeared on both media, although the sam- 

 ples of the untreated culture plated directly from M plates to both 

 M and C plates showed only one -fifth as many colonies on the C as 

 on the M plate. 



HAPLOPHASE VARIABILITY VS. DIPLOPHASE STABILITY 



Nearly all haplophase yeast cultures, when plated on agar, pro- 

 duce a considerable variety of relatively stable colonial variants. 

 Hybrids are produced by mixing the cells of two such apparently 

 unstable cultures together in a small amount of broth. After copu- 

 lations occur, diploid cells are produced which subsequently sporu- 

 late. Genetical analyses of the hybrid are made by picking and dis- 

 secting 4-spored asci at random after sporulation has occurred. 

 One of the puzzling things about this process is the fact that in spite 

 of the apparently considerable mutational potential of the haplophase 

 cultures, the tetrad analyses usually 3delded surprisingly regular 



