390 Mr. I'onlton' 8 farther experiments upon 



Other Eesults. 



The effects of a few other conditions not tested by 

 large numbers of individuals are shown in the table 

 below, the percentages from white surroundings in 1886 

 being also included for the purpose of comparison. 



It is thus seen that white muslin was not nearly so 

 powerful as the white paper and opal glass employed in 

 1886. The difference corresponds to the far smaller 

 amount of light reflected from the former, and its 

 feebler illumination under the conditions of the experi- 

 ment. 



Clear glass, '.when uncrowded, ^chiefly produced inter- 

 mediate forms, while the crowded'pupae were considerably 



