THE LINEAR DIFFERENTIATION OF THE CHROMOSOMES IO5 



gene and the centromere. In Neurospora, Lindegren obtained values 

 of 16-5 for the cross-over distances between pale and the centromere 

 and 6-5 between the sex factor and the centromere. Hence, disregard- 

 ing double cross-overs, the distance between pale and the sex factor 

 should be either 16-5 — 6-5 = 10 or 16-5 + 6-5 = 23. Actually we 

 saw that direct measurement gave 22-5 in extremely good agreement 

 with the latter hypothesis; the two estimates confirm one another and 

 the general Une of argument, and we can draw a map of the sex chromo- 

 some containing P and ^b on different sides of the centromere. 



8. Crossing-Over in Polyploids 



Chromatid segregation can be discovered in higher organisms if 

 some way can be found of getting two, instead of one, of the segre- 

 gating chromatids into a single gamete. This happens in a race such as 

 the attached- A" race in Drosophila, where the two X chromosomes of 

 the female are permanently attached together and form XX eggs.^ It 

 also occurs in polyploids, where there are more than four homologous 

 chromatids to be distributed to the four gametes. In both these cases 

 the occurrence of chromatid segregation can be proved. Thus in a 

 tetraploid Datura of the constitution AAAa, gametes aa could be pro- 

 duced (giving aaaa on selfing).^ This can only occur if in the chromatid 

 stage AAAAAAaa crossing-over occurs so that the two a's become 

 associated with two different centromeres which are passing to the 

 same pole so that the factors have a chance to pass to the same pole in 

 the second division. 



The best analysed cases of chromatid segregations in polyploids are 

 in triploid Drosophila melanogaster.^ One can label all the chromosomes 

 of one kind (e.g. the X) by different recessive factors distributed along 

 the whole length of the chromosome. Crossing-over between whole 

 chromosomes (in the so-called 2-strand stage) cannot give more than 

 three sorts of chromatids (2 of each kind) to be distributed to the four 

 gametes, and each sort must be completely unlike any other. At the 

 second division the two sister members of each pair must be separated 

 from each other by the separation of the attachment constriction, so 

 that the diploid gametes can only contain two chromatids which are 

 wholly unlike each other. Crossing-over between chromatids, in the 

 4-strand stage, can on the other hand give rise to six different sorts of 



^ Anderson 1925. The same phenomenon occurs in non-disjunctional eggs, 

 cf. Bridges 1916. 2 Blakeslee, Belling and Farnham 1923. 



3 Bridges and Anderson 1925, Redfield 1930, 1932, Rhoades 1933. 



