PLASMAGENES AND CHROMOGENES IN HETEROSIS 233 



non-flowering type as seed parent grows taller than the reciprocal com- 

 bination, and flowers later. These difi"erences are statistically significant. 



Reciprocal crosses between inbred California Rice pop, having the small- 

 est seeds known in corn, with inbred Indiana Wf9 having large embryos 

 and endospersms, show differences in early seedling growth and in tillering. 

 Inbred Wf9 produces no tillers. California Rice, also inbred, produces an 

 average of 4.1 tillers per plant. The first generation cross of Rice popXWfQ 

 averages 1.0 tillers, while the reciprocal combination under the same condi- 

 tions produced 2.2 tillers per stalk. In this case the non-tillering variety, 

 when used as the seed parent, produces more than twice as many tillers. 

 This seems to be a carry-over effect of the large seed. Tillering is largely de- 

 termined by early seedling vigor. Anything that induces rapid development 

 in the early stages of growth tends to promote tillering. 



PLASMAGENES AND CHROMOGENES 



In addition to these transitory effects there are many cases of cytoplasmic 

 inheritance. Caspari (1948) has reviewed the evidence from fungi, mosses, 

 the higher plants, and from Paramecium, insects, and mammals to show 

 that many differences do occur in reciprocal crosses and that they persist into 

 later backcrossed generations. Reciprocal differences in the amount of 

 heterosis have been demonstrated in Epilobium (Michaelis, 1939) and in 

 mice (Marshak, 1936). 



Cytoplasmic pollen sterility has been found in Oenothera, Streptocarpus, 

 Epilobium, flax, maize, onions, sugar beets, and carrots. In every case that 

 has been adequately studied, the basic sterility remains the same in repeated 

 generations of backcrossing, but the amount of pollen produced varies in 

 different genotypes. There is an interrelation between plasmagenes and 

 chromogenes determining the final result (Jones, 1950). 



In maize the amount of pollen produced ranges from to 100 per cent. 

 Only by suitable tests can these cases of full fertility be recognized as having 

 any cytoplasmic basis. Interest in this problem now centers on the effect of 

 these cytoplasmic differences on heterosis. 



A series of standard inbreds have been converted by crossing these onto 

 suitable sterilizer stocks, and backcrossing a sufficient number of generations 

 to re-establish completely the inbred, and maintaining the inbred in a sterile 

 condition by continuous backcrossing. It has been found necessary to select 

 both the cytoplasmic sterile seed parent individuals and the individual fertile 

 pollen parents for their ability to maintain complete sterility both in inbreds 

 and in crosses. In some lines it has proved to be impossible to establish com- 

 plete sterility, but the majority are easily sterilized and maintained in that 

 condition. 



A comparison of fertile and sterile progenies in inbreds, in single crosses of 

 two inbreds, and multiple crosses of three and four inbreds, shows that this 



