INFLUENCE OF ENVIRONMENT ON EXPECTED CLASSES. 2l() 



TABLE IV. 



before those recorded above. In these earlier experiments in 

 mass cultures no attempt was made to avoid crowding through 

 use of too many parents, although the bottles were otherwise 

 maintained in good condition. The following F 2 counts were 

 made: 



Long $ and 



Strap 9 and 



Ratio. 



Whether the two very high ratios under 10 and 12 should be 

 included may be questioned. It is possible though not probable 

 that contamination took place. Excluding these the ratios vary 

 from I : 4.9 to I : 30.6. 



CONCLUSIONS. 



The experiments show that under unfavorable conditions 

 due to crowding the mutant forms, rudimentary wings and 

 strap wings, run behind expectation. That the result is due to 

 crowding is shown when cultures of sister individuals are made 

 in pairs with abundant room and food. There is then a closer, 

 and in some cultures a complete agreement with expectation. In 

 experiments in which it is only necessary to show whether a 

 character is recessive or dominant, and whether it is, or is not 



