MEIOSIS 267 



bination of characters associated with linked genes (p. 294). That this 

 interpretation is correct has been rendered practicaDy certain by studies 

 on pairing chromosomes in which characteristic structural features 

 were exchanged at the time recombinations were produced (p. 303). 

 That such exchanges somehow involve the chiasmata observed in the 

 tetrads from the diplonema stage onward has been strongly suspected 

 from the first, although much of the evidence has been open to question. 

 It is the prominent role played by crossing-over in present-day researches 

 in the field of cytogenetics that now makes it necessary to consider the 

 chiasma in some detail. 



OC ^ ^ 



Fig. 156. — Bivalent chromosomes (tetrads) in metaphase and early anaphase of first 

 meiotic mitosis in Vicia faba, showing various numbers of chiasmata. First row: chromo- 

 somes with subterminal spindle attachment. Second row: longer chromosomes with 

 median attachment. (After Maeda, 19306.) 



As already explained, a chiasma is a region where the four chromatids 

 appear to trade partners when the tetrad opens out in the diplonema 

 stage. The number of chiasmata developing frequently varies with 

 the length of the tetrad.^" In Hyacinthus orientalis, for example, there 

 are commonly three in the longest tetrads (which may have as many as 

 six), two in the medium-sized tetrads, and one in the shortest ones. In 

 Vicia faba (Fig. 156) the commonest numbers are three in the short 

 tetrads and eight in the long one. They appear to be located more or 

 less at random in some species, but to be more localized near the spindle- 

 attachment region in others. ^^ Of considerable significance is the further 

 fact that in many organisms the number of chiasmata tends to be reduced 

 as the tetrads develop from the diplonema to the diakinesis stage. ^^ In 

 Callisia repe7is, for example, the six tetrads show a total of about 25 



'"' Darlington (1929c) on Hyacinthus, Maeda (19306) on Vicia. Darlington 

 (1932a) presents an extended discussion of chiasmata and their supposed significance. 



21 At random in Hyacinthus, Fritillaria imperialis, and probably Tradescantia; 

 localized in Fritillaria meleagris (Newton and Darlington 1929, 1930; Darlington, 

 1929c, 1930d) and Allium fist ulosum (Levan, 1933). 



22 Newton (1927) and Newton and Darlington (1929) on Tulipa, Darlington 

 (1931a) on Primula, Sax (1930cf/) on Secale and Callisia, Belling (1928a, 19316) on 

 Lilium, Philp and Huskins (1931) on Matthiola, Erlanson (1931a) on Rosa. 



