130 AN INTRODUCTION TO MODERN GENETICS 



the two B chromosomes when they attempt to pair will be pulled apart 

 by the fragment of A attempting to get into contact with its homo- 

 logue, and unless the chromosomes are perfectly flexible this conflict 

 of forces will prevent some of the normal pairing. In salivary gland 

 nuclei, where pairing can be observed in much more detail than in 

 zygotene nuclei, the chromosomes are usually flexible enough for full 

 pairing of all sections, even when translocated. But this seems not to 

 be the case in meiotic prophase, since it is found that a translocated 

 segment lowers the crossing-over frequency in its neighbourhood and 

 this can be most easily explained as the result of an inhibition of pairing 



Fig. 6A. Crossing-Over in Inversions. — The figure 

 shows a cross-over in an organism heterozygous for an 

 inversion, which has paired in the usual loop (see 

 Fig. A7). Only one chromatid of each chromosome is 

 shown, for clearness. Through the cross-over, the two 

 centromeres (cm) become attached to the same 

 chromatid, which is therefore pulled in two opposite 

 directions at anaphase, and forms a "bridge." The other 

 cross-over chromatid has no centromere and remains 

 passive. The two non-cross-over chromatids (not 

 shown) are normal. 



caused by the relative inflexibility of the chromosome threads. (Other 

 explanations, have, however, been advanced.)^ 



In heterozygous inversions, pairing is possible in two ways. If we 

 have chromosomes abcdefg and abedcfg, the inverted segments cde may 

 pair, leaving the sections ab and/^ together and unpaired at each end; 

 or there may be complete pairing by the formation of a loop. The first 

 type of pairing reduces the total paired length, since only cde is paired 

 instead of the whole abcdefg; it therefore reduces the number of chias- 

 mata formed and thus reduces crossing-over. (The non-homologous 

 ends ab and j^ may become associated by "snarl'* pairing (p. 367), but 

 this probably does not lead to chiasma formation.) The second method 

 of pairing, in loops, is probably the commoner. It is usual in salivary 

 gland chromosomes and has been observed in pachytene in maize. It 

 allows all sections of the chromosome to be paired and crossing-over 

 may occur at any point. 



The results of crossing-over in inversions depend on the relation 

 between the point of crossing-over and the centromere. Suppose we 

 have two chromosomes abc,defg and abc,dfeg, where the comma repre- 



^ Dobzhansky 193 1. 



