CALVIN B. BRIDGES 



653 



In Table VIII is given a summary of the more significant crossover 

 values as determined from all the experiments involving deficiency, 

 with normal values for comparison. 



From these values it is apparent that there is a reduction in crossing 

 over between cut and vermilion-deficiency of about 2 units, while 

 there is a similar reduction between deficiency and sable of about 3 

 units, or a total shortening of the chromosome map of about 5 units. 

 It is known that not all the reduction to the right of deficiency is 

 due to the length of the deficient region, since crossing over occurred 

 between deficiency and tiny-bristles. Tiny-bristles and vermilion 

 are normally 3.0 units apart, while there were two crossovers in a 



TABLE VIII. 



Comparison of Normal Crossover Values With Those Obtaining for Females Hetero- 

 zygous for Vermilion-Deficiency. 



total of 86 males, which showed that about 2.3 per cent of crossing 

 over occurs between deficiency and tiny-bristles. This evidence does 

 not allow us to determine the actual length of the deficient region 

 nor what proportion of the deficiency is to the left of the vermilion 

 locus and how much is to the right. The garnet forked value is 

 normal or perhaps slightly high. 



The normal values used for comparison in Table VIII are the 

 mean values derived from several experiments'* and are not, as would 

 have been more desirable, from sister cultures. While the two sets 

 of values are therefore not strictly comparable, the differences ob- 

 served were so constant in direction that there can be little doubt 

 of the reality of the changes. 



* Morgan, T. H., and Bridges, C. B., Carnegie Institution of Washington, Pub- 

 lication No. 237, 1916; also unpublished data. 



