CONFIGURATIONS OF BIVALENTS OF HYACINTHUS. 



485 



by the microscopical phenomena, we have for the resulting 

 pollen grains and any particular long chromosome, on the 

 hypothesis of segmental interchange: no segmental interchange 

 (124 + 54), 178; single point of interchange (124 + 108), 232; 

 and double interchange, 54; out of a total of 464 chromosomes 

 or pollen grains. 



This gives in percentages: chromosomes with no interchange 

 38 per cent.; chromosomes with single interchange, 50 per cent.; 

 and chromosomes with double interchange, 12 per cent. The 

 ratio of single to double interchange on the hypothesis is thus 

 4-3 to i. 



This is not far from the numbers of no, single and double 

 points of crossing-over given for the first chromosome of 

 Drosophila melanogaster by Morgan (1925). The second and 

 third chromosomes of Drosophila, however, seem to differ in this 

 respect. 



In Figs. 4 and 5, there are examples of the separation of 

 chromatids and homologues which tend somewhat to prove the 

 hypothesis of segmental interchange. In Fig. 4, LI is especially 

 instructive. Here on one side of the constriction (and spindle 

 fiber attachment) there was apparently a horizontal ring, and 



FIG. 4. Camera drawing of the bivalents in a cell in which the four large ones 

 had not completely separated, the drawing paper being shifted after each was 

 drawn. Some are foreshortened, especially ~Li and L4. 



