170 INBEEEDING AND OUTBREEDING 



gation through the cytoplasm becoming adjusted to a 

 heterozygous nucleus. This theory was proposed as an 

 interpretation of a reduction in vigor which he had found 

 in parthenogenetically reproduced rotifers. The recent 

 facts, however, are more in accord with the former view 

 because (1) the stimulus actually is lost as homozygosity 

 is attained, and (2) the evidence of vigor being reduced 

 in continued asexual reproduction is not at all conclusive. 



The reasons for holding the whole stimulation hypoth- 

 esis in abeyance at present has developed from the follow- 

 ing facts. In 1910, Keeble and Pellew^^^ offered a con- 

 crete illustration of a purely Mendelian method by which 

 increased growth could result from crossing. They united 

 two varieties of garden peas, which, as grown by them, 

 each ranged from 5 to 6 feet in height. The first genera- 

 tion grown from this cross was from 7 to 8 feet in height, 

 2 feet taller than either parent, a result comparable to 

 heterosis. The second generation showed segregation 

 into four classes, one class containing plants as tall as the 

 first generation, two classes having plants similar in 

 height to the two parents, and one class made up of 

 dwarfs shorter than either parent. The two classes of 

 medium tall plants, similar in height, were differentiated 

 in the same manner as the parental races ; one had thick 

 stems and short intemodes, the other had thin stems and 

 long intemodes with fewer of them. The number of 

 plants falling into these four classes agreed closely with 

 the expectation for a dihybrid ratio (9:3:3:1) where 

 two factors showing dominance are concerned. 



Keeble and Pellew assumed two hereditary factors to 

 be involved: one producing thick stems, the other long 

 intemodes. These factors they designated T and L. One 



