172 ELMER ROBERTS 



wild types has in times past been recombined, then selection 

 would be expected to separate certain types again, if, at divi- 

 sion, irregular sampling of the germ plasm takes place." 



In this case Morgan desires proof that the characters were not 

 in a heterozygous condition. On the other hand, the question 

 might be asked if there is any proof of the occurrence of irregular 

 samplmg of the germ plasm in DifHugia. 



Since an environmental factor, temperature, affects the ves- 

 tigial wing to a marked degree, a consistent germinal selection 

 may not have been in operation in these experiments. In gen- 

 erations occurring at a time favorable to the production of large 

 vestigial wings, selection was easily possible, but in generations 

 when the vestigial wing was inhibited by a low temperature, 

 selection of individuals potent for, larger wings could not be 

 made except as they were included by chance. 



2. Effect of ' crossing-in ' 



The data establish beyond question that crossing the vestigial- 

 winged flies to normal long-winged had a very marked influence 

 on increasing the size of vestigial wings in the 'crossed-in' series. 

 Castle ('15) reported on some unpublished work of Dr. D. H. 

 Wenrich that the extracted vestigials from a vestigial, long- 

 wing cross showed greater variability in length of wing than did 

 the uncrossed. To what this increase is due is not known, but 

 there are two possible explanations: 



1. The increased size and variability may have been due to the 

 introduction of modifying factors from the wild stock, these 

 factors acting more effectively when the temperature was higher 

 than the normal; or 



2. A gametic contamination may have occurred when the ves- 

 tigial wing factor came into association with its allelomorph, 

 long wing (or the association of the developer for wings with the 

 absence of this developer). 



There is some evidence in favor of the first explanation gained 

 from the following experiment performed to test this view. A 

 large vestigial-winged male from F32 generation produced in the 



