228 PROTOZOA 



parent. Superficially, the group A system seems very different: 

 there is no correlation between mating types of the two caryo- 

 nides of a clone or between these and the sexual parent. This lack 

 of correlation gives the impression that macronuclear determina- 

 tion is not brought about by a cytoplasmic agent, an impression 

 which is strengthened by the marked effect of environmental con- 

 ditions, such as temperature, on macronuclear determination. 



However, the facts are not contradictory to the interpretation 

 of macronuclear differentiation by the cytoplasm or to the control 

 of the cytoplasmic agent by the macronuclear genotype. Further, 

 some facts point to just this sort of mechanism. Such a hypothesis 

 will therefore be developed in order to have a single unified inter- 

 pretation for both the group A and group B systems. There will 

 obviously have to be some differences between the two systems, 

 but the hypothesis will attempt to reduce these to the minimum 

 and preserve maximal similarity. If it be assumed that there is a 

 different cytoplasmic agent corresponding to each possible nu- 

 clear differentiation, i.e., two in a variety with two mating types, 

 four in a variety with four mating types, there need be only two 

 differences between the group A and B systems. ( 1 ) In group B, 

 the cytoplasmic agents or differentiators of the macronuclei are 

 subject to a steady state relationship; but in group A they are 

 not. This means that in group B one or the other of the cytoplas- 

 mic differentiators is always in relatively high effective concentra- 

 tion as compared with the other or others; but that in group A 

 a continuous series of relative concentrations is possible. (2) In 

 group B, the macronuclear genes for the expressed mating type 

 also as a rule determine that the corresponding cytoplasmic 

 differentiator will have a relatively high effective concentration; 

 but in group A, although the macronuclear genotype also controls 

 the relative concentrations of the cytoplasmic differentiators, the 

 pertinent part of the genotype is not that, or not only that, part 

 which controls the expressed mating type. The genes involved in 

 group A control of the cytoplasmic differentiators an 1 assumed to 

 be the same in both mating types. 



With tin's interpretation, group A would show the absence of 

 parent-offspring correlation that is observed. Further, the ob- 



