THE CYTOGENETICS OF HYBRIDS 371 



attributed to the synapsis of chromosomes from dissimilar sets in the 

 complement (allosynapsis) or to secondary pairing and atypical distribu- 

 tion of autosynaptic pairs. Studies on such alloheteroploid plants, taken 

 together with comparisons of the conditions observed in many related 

 species in nature, have strengthened the theory that amphidiploidy has 

 played an important role in the differentiation of species. 



Parental Characters in Polyploid Hybrids. — There is a marked 

 tendency on the part of polyploid plants of hybrid origin to show a greater 

 resemblance to the parent contributing the larger number of chromosome 

 sets. In crosses between Raphanus and Brassica the sterile Fi hybrid 

 with one set from each parent (RB) is intermediate in capsule characters 

 (except size) (Fig. 211). The fertile F2 Raphanobrassica with two sets of 

 each kind (RRBB) and a hexaploid with three of each kind {RRRBBB) 

 are intermediate also. A pentaploid form with RRRBB is somewhat 

 more like Raphanus, while the resemblance is still greater in a triploid 

 (RRB). Thus is the ratio of parental chromosome sets reflected in 

 external characters (Karpechenko, 19276). 



When Rubus rustica inermis, a diploid species, was crossed with 

 R. thyrsiger, a tetraploid, three Fi hybrids were obtained. Two of these 

 were triploid (TTR) and resembled R. thyrsiger, while the third was tetra- 

 ploid {TTRR) and was more like R. rustica inermis. The tetraploid 

 probably arose from a diploid egg produced by rustica after some meiotic 

 aberration (Darlington, 1929/). 



Euchlcena mexicana {n = 10) crossed with Zea Mays (n = 10) gave 

 an intermediate hybrid with one set from each parent {EZ). The tetra- 

 ploid E. perennis (gametic number = 20) when similarly crossed to Zea 

 gave a triploid hybrid (EEZ) which was more like Euchlcena. One plant 

 in the progeny of this cross was considerably more like Zea, and cyto- 

 logical examination showed it to be tetraploid. Presumably the fourth 

 set was formed from the Zea set by doubling, giving the formula EEZZ 

 and a more even balance between the characters of the two parents (R. 

 Emerson and Beadle, 1930). 



The same phenomenon is manifested by polyploid moss hybrids. In 

 hybrids between Physcomitrella patens and Funaria hygrometrica the 

 peristome of the capsule becomes less like that of the former genus 

 (P-) and more like that of the latter {F-) in passing through the series 

 P\ P^F\ PW\ P^F\ PW\ P^F'', P-F', P'F\ F- (von Wettstein, 1926). In 

 narrow crosses in mosses the degree of dominance is roughly proportional 

 to the relative number of dominant genes, whereas in wide crosses (specific 

 or generic) it may increase up to a certain ratio and then decrease. This 

 is attributed to differences in the parental cytoplasms (von Wettsiein, 

 1927). 



In hybrids between Crepis capillaris {n = 3) and C. aspera (rv = 4) 

 where a doubling of one chromosome set has occurred, it is found that 



