The Behavior of the Chromosomes as Studied through Linkage. 



247 



Table VIII. 



These crosses show in general the same points as do those recorded 

 in table III. The index of interference in the two cases is 3'26 and 

 r67, respectively. This again agrees with what we saw, in table DI, 

 for chromosome I; for the distances involved in the second cross are 

 somewhat longer than those in the first, and interference is markedly less. 



Chromosome HI— Variability in Linkage. 



I have reported (StüRTEVAXT '13c) the existence of a third group 

 of genes in Drosophila — Chromosome III. I reported two pairs of genes 

 (P and p, Eb and eb ; involving the pink eyed and ebony bodied mutants) 

 in tliis chromosome, and shall only consider these two loci here. 



At that time I had been unable to breed ebony females very 

 successfully, and so had very few data bearing on the question of 

 crossing over in the male. I have now succeeded in raising 781 fUes 

 from backcross tests of males (see table XXII); none of these were 

 cross-overs, so that it seems likely that the condition found in the male 

 for chromosome II occurs also for chromosome III. 



I gave the totals from a large number of backcross tests of females, 

 but stated that much variability occurred in the amount of linkage shown 

 by these females. Several thousand more flies from such crosses have 

 now been obtained, but the cause of the variability has not yet been 

 determined, though it seems to be partly germinal, and there is some 

 evidence indicating that it may also be partly due to environmental 

 effects (food and moisture, apparently). Until something more definite 

 can be said it seems advisable simply to give a few sample counts to 

 show the range of variability, reserving the rest of the data until the 

 subject can be more thoroughly investigated. The nine bottles recorded 

 in table XXI gave the results shown in table IX. 



