THE CYTOGENETICS OF AUTOHETEROPLOID PLANTS 



351 



happen to associate as synapsis begins, together with the number and 

 persistence of chiasmata, is an important factor in this connection. The 

 behavior is apparently influenced in some cases by the relative positions 

 of the chromosomes in the nucleus before synapsis begins; also, it tends to 

 differ characteristically in different species or genera.- In addition to the 

 "primary association" determined by the ordinary synapsis in the early 

 prophase, there is frequently a ''secondary association" at the end of the 

 prophase when homologous bivalents previously separate move together 

 to form loose quadrivalent groups.^ 



Fig. 199. — Synaptic configurations in heteroploid plants, a-f, trivalent chromosomes 

 in Hyacinthus. g, h, quadrivalent chromosomes in Hyacinthus. i, quadrivalent in Tulipa. 

 e, f, and i are postdiplonema stages. {After Newton and Darlington, 1929.) 



The phenotypic ratio shown by a pair of Mendelian characters in 

 breeding tetraploids obviously depends upon the degree of dominance 

 exhibited, and upon the manner in which the four chromosomes carrying 

 the four differential factors are distributed in the meiotic divisions. In 

 tetraploid Primula sinensis* the several combinations, DDDD, DDDd, 

 DDdd, Dddd, and dddd, give a series in depth of flower color, the first 

 three being white or faintly tinged, the fourth weakly colored, and the 

 fifth more deeply colored. Hence various types are to be expected 



2 There are sometimes n quadrivalents in Datura tetraploids (Belling and Blakeslee 

 1923, 1924o; Belling, 1927(i). The number in most cases is smaller, as in Prunus, 

 Hyacinthus, and Primula sinerisis (Darlington, 19276, 1929c, 1931a; Dark, 1931); also 

 in Euchlcena perejmis (Randolph, unpubl.) and Dahlia (Lawrence, 1929). In Cam- 

 panula most of the chromosomes are in pairs (Gairdner, 1926). The "double-diploid " 

 condition occurs sometimes in Hyacinthus (Belling, 19256) and rarely in Datura. In 

 Zea there are quadrivalents and bivalents, and the segregation is at random (Randolph, 

 Rhoades). 



^ Darlington (1928) on Prunus, Darlington and Moffett (1930) on Pyrus, Lawrence 

 (1931c) on Dahlia. 



* Gregory (1914), Bateson, Sverdrup, Haldane (1929). 



