THE MECHANICS OF THE CHROMOSOMES I33 



not replaced but persists more or less unaltered. (In an unpaired region 

 at the end of a chromosome, the two homologies can spiralize indepen- 

 dently and no relational coiling develops and hence no chiasmata are 

 formed.) 



An alternative method of crossing-over, which has been put forward 

 by several authors,^ is based on the supposition that the division of the 

 chromosome into chromatids takes place in two stages. It is supposed 

 that first the chromomeres (presumably the genes) divide, and that the 

 newly formed chromomeres then form a new connecting tliread which 

 becomes the chromonema. Crossing-over would occur if the newly 

 formed thread sometimes joined up with a chromomere derived from 

 the other chromosome. This mechanism is a possible one, but has much 

 less observational basis than the last; there is in fact no good evidence 

 that the connecting thread does grow out from one chromomere till it 

 meets the next. It is perhaps worth while pointing out that the usual 

 diagrams of this process look much more convincing in two dimensions 

 than when they are interpreted as representations of three-dimensional 

 objects. 



^ Belling 1933, Lindegren and Lindegren 1937. 



