VARIATION OP CHROMOSOMES 



161 



The hybrid had 30 chromosomes (20+10). He fomid that 

 when this hybrid produces its germ-cells they show, after 

 reduction, 20 chromosomes, which he interprets as due to 

 10 of the rotundifolia conjugating with 10 of the longi- 

 folia. This leaves 10 without mates. At the following 

 maturation division Rosenberg describes the 10 paired 

 chromosomes as reducing, sending one member of each 

 dyad to one pole, the other member to the other ; but the 



Egj. 



Jperm, 



Zygote. 



amete 



n 



m 



Fig. 66. — A scheme illustrating the fertilization of the egg of one species of moth by 

 the sperm of another, with reduction in I, with no reduction in II, and with partial reduc- 

 tion in III. 



10 unpaired chromosomes are irregularly distributed at 

 this division. If the account is confirmed, the situation 

 is peculiar, for if the 20 (haploid) chromosomes of longi- 

 folia correspond to the 10 (haploid) of rotundifolia it is 

 not obvious why all 20 might not find a place alongside 

 of the 10, unless chance or some difference of length, etc., 

 makes this impossible. This assumes, however, that longi- 

 folia is not tetraploid — if it is, then a further question 

 arises as to which chromosomes of each set of three would 

 be the ones most likely to conjugate, etc. 



Crosses between three species of the moth Pygcera, 

 11 



