GENETICS OF HETEROMORPHIC CHROMOSOMES 471 



mined were: whether or not the morphological constitution of 

 the chromosomes is transmitted from parent to offspring, and, if 

 so, whether their recombinations in the progeny are according to 

 the laws of chance. 



Among these twenty-eight male offspring of five crosses there 

 are fifty-six homologues (2 X 28) for each of the three pairs of 

 chromosomes (nos. 1, 7, 8) or a total of one hundred and sixty- 

 eight homologues (3 X 56) in all, which might have varied at 

 the ratio found in the wild individuals of one atelomitic to seven 

 telomitics for pair 1; one atelomitic to nine telomitics for pair 

 8, and of one telomitic to seven atelomitics in pair 7. One 

 would expect twenty-two shifts of fiber attachment in such a 

 number of homologues if there is a reorganization during ontog- 

 eny which repeats the conditions in the species. On the con- 

 trary, not a single offspring varied in its chromosomal constitu- 

 tion beyond the limits to be expected from a combination of 

 the gametes of its parents. In other words, any given chromo- 

 some reappeared in the progeny in the same form that it pos- 

 sessed when it went into the parental gamete. This is shown 

 more clearly by comparing families which carry atelomitic 

 number I's with those where the number I's are telomitic. In 

 the former (matings 2, 13, 17, text fig. B) there are fourteen 

 telomitics to four atelomitics (ratio 7 to 2) in the latter (mat- 

 ings 5 and 14, text fig. 2) there are thirty-eight telomitics to 

 no atelomitics. 



The same thing is shown by pair number 8; in four families 

 (numbers 2, 13, 5, 17) involving sixteen offspring, the thirty- 

 two homologues are all telomitic; whereas among the twenty- 

 four homologues of this pair in family number 14, where both 

 parents are heteromorphic in this respect, ten are atelomitic 

 and fourteen telomitic. 



These data I believe are sufficient to establish the constancy 

 of point of fiber attachment from parent to offspring for this 

 group, and to lay the basis for further work on the assumption 

 that structural variations in the chromosomes are correlated 

 with somatic characters in such way that it will be possible to 

 tell what the chromosomal constitution of a wild individual is 

 in regard to a given pair from an external study of the animal. 



