THE MECHANICS OF THE CHROMOSOMES I23 



lengths are reasonably larger than the maps at present known for the 

 chromosomes, which is an error on the right side, particularly in view 

 of the comparatively small number of genes mapped. In a hybrid 

 between Zea and Euchlaena, Beadle^ could measure both the chiasma 

 frequency and the frequency of crossing-over in a certain section of 

 chromosome, which was relatively translocated in the two species. He 

 found a crossing-over value of 12 per cent, which should correspond to 

 a chiasma frequency of 24 per cent; the actual number of chiasmata 

 found were 20 per cent, which is a satisfactory agreement. 



Secondly, Darlington was led to predict that there must be some- 

 thing very odd about chiasma formation in male Drosophila, in which 



Fig. 59. Proof that some Chiasmata involve Crossing-Over. — A shows the 



"classical" interpretation of a heteromorphic bivalent; the two chromatids paired 

 at a centromere must be different, which is not the case. The "Chiasmatype" 

 interpretation is shown at B. In C two chromosomes have a chiasma at X, and 

 on each side of it the two pairs of chromatids are coiled round one another; 

 each pair therefore behaves as a unit and consists of two sisters, so that a crossing- 

 over must have occurred at the chiasma. 



there is no crossing-over but in which the reduction takes place with 

 such regularity that some sort of metaphase association must be sup- 

 posed to occur. This prediction has been fulfilled: in male Drosophila 

 the chromosomes associate at metaphase by a process other than 

 chiasma formation, being held together in groups of 4 chromatids by 

 some generahzed attraction, perhaps analogous to that which causes 

 somatic pairing. The only chiasma formation is between the X and Y 

 (in the inert region) and these two chromosomes show very rare 

 crossing-over.2 



There are further evidences of the general paralleUsm between 

 crossing-over and chiasma formation, (i) It has been shown that the 

 variation of chiasma frequency with temperature in mice^ and Orthop- 

 tera^ is similar to the rather comphcated variation in crossing-over value 

 with temperature in Drosophila. This argues at least a strict propor- 

 tionahty between chiasma frequency and cross-over frequency. (2) The 

 total amount of crossing-over is the same in triploid Drosophila as 



^ Beadle 1932. ^ Darlington 1934a. ^ Bryden 1935. * White 1934. 



