132 AN INTRODUCTION TO MODERN GENETICS 



regional differences in chiasma frequency, are determined by position* 

 and the frequency characteristic of a given region is shown by any 

 section of chromosome occurring there (in translocations, etc.), so they 

 must be due to general mechanical conditions, not to local differences 

 in elasticity, etc.^ 



Darlington has pointed out that when the chromosomes start to 

 contract (by spiralization) while they are paired in zygotene, they begin 

 to coil round one another (p. 366). This relational coiling just balances 





Fig. 65. Relational Coiling and Crossing-Over. — The left diagram shows two 

 chromosomes before crossing-over. Each chromosome consists of two chromatids, 

 or is just going to split into two chromatids; these are coiled relationally (in a 

 right-handed spiral in this diagram) and the whole chromosomes are relationally 

 coiled (left-handedly). At the point where the chromosomes cross, one chromatid 

 of each chromosome breaks, the broken ends curl away and join up in a new 

 formation after relieving some of the torsion; the final state is shown on the right. 



(From Darlington.) 



the spiralization which has brought it into being, until the division of 

 the chromosomes in zygotene turns each chromosome into a pair of 

 weaker chromatids, one of which may break under the strain; if one 

 breaks, all the stress will fall on the other chromosome and one of its 

 chromatids will also break. This double break will allow the release of 

 the stress by rotation, after which the broken ends may join up again, 

 when it will be found that a crossing-over has resulted. Accordingly 

 it is found that the relational coiHng observable in zygotene is replaced 

 by chiasmata in diplotene. A test of the hypothesis is provided by the 

 behaviour of the unpaired central regions which are characteristic of 

 the chromosomes of some species. The homologues here He too far 

 apart for chiasmata to be formed, and the relational coiling is therefore 

 ^ Offermann, Stone, and Muller 193 1. 



