AlATING lYPI. Hi URMINATION 279 



iiiacronuclcus controlling one ninrint:;- type and sonic containing re- 

 generated macronuclei controlling the other mating type. Since it 

 seems reasonable to assume that the cytoplasmic elements, if these 

 arc indeed independent of nuclear control, will he equally assorted in 

 the various lines of descent, all lines of descent should at a subsequent 

 reorganization give rise to the same kinds of progeny. This result 

 was not obtained. Instead, each line at a subsequent reorganization 

 regularly produced progeny of the same mating type as the parental 

 line. Therefore, the cytoplasmic conditions must have been modified 

 to bring them into agreement with the nuclear constitution. 



Since the mating type of the parent is clearly controlled by the 

 macronucleus, and since the cytoplasmic condition necessary for the 

 inheritance of mating type through nuclear reorganization is strongly 

 correlated with the parental mating type, it follows that the cyto- 

 plasmic condition is ultimately controlled by the old macronucleus. 

 Thus, though the cytoplasm determines the nature of the new macro- 

 nucleus, the cytoplasm is in turn controlled by the old macronucleus. 

 This results in a cyclical interdetermination of the nucleus by the 

 cytoplasm and of the cytoplasm by the nucleus. This system tran- 

 scends "maternal inheritance" in that the cytoplasmic influence alters 

 the new nucleus in such a way as to perpetuate a trait indefinitely 

 through both vegetative and sexual reproduction. Regardless of the 

 explanation eventually given for those rare clones which revert to an 

 original mating type, the situation observed in the majority of 

 changed clones argues strongly for the macronuclear control of the 

 cytoplasm. 



The precise nature of this macronuclear control is problematical. 

 The nucleus could produce directly some substance or substances nec- 

 essary for mating type inheritance, but it is also possible that the 

 macronucleus merely controls the rate of production of a self-repro- 

 ducing factor such as has been demonstrated for the killer character 

 in this same organism (Sonneborn, 1947; Chao, 1953). In the absence 

 of evidence for a self-reproducing factor, it appears advisable to adopt 

 the simpler interpretation. 



NATURE OF SELFING CARYONIDES 



One of the perplexing problems concerning mating types, both 

 in P. aurelia and in other ciliates, involves the nature of the selfing 



