156 WILLIAM L. DOLLEY, JR. 



flash-frequencies of 60, 50, 40, and 10 per second. However an 

 examination of the averages presented in the last line in table 3 

 indicates that the stimulating efficiency of intermittent light be- 

 gins to rise at a flash-frequency of about 60 per second, reaches 

 a maximum at 30 per second, and then declines. According to 

 these figures the efficiency of light of 10 per second is less than 

 that of continuous light. If this is true the stimulating efficiency 

 of intermittent light of a flash-frequency of somewhere between 

 10 and 20 per second is equal to that of continuous light. In the 

 absence of more data it is probably wiser to assume that the con- 

 clusions drawn from table 4 are more nearly correct than those 

 drawn from the last line in table 3. 



The conclusion stated above that the stimulating efficiency of 

 intermittent light varies with the flash-frequency is further 

 supported by the results presented in table 5. This table shows 

 that the total average angles of the butterflies tested in inter- 

 mittent light produced by a sectored disk with one-fourth re- 

 moved and of the following flash-frequencies: 30, 20, 15, 10, 5, 

 and 2 per second were respectively: +15.38, 4-7.67, 4.16, 

 +0.84, -14.67, and -22.76 degrees. The angles of deflection of 

 the butterflies tested in light produced by a disk with one-half 

 removed and of the following flash-frequencies: 30, 20, 15, 10, 5, 

 and 2 per second were respectively: +10.75, +6.73, 0.71, 

 +0.92, -6.58, and -15.67 degrees. The angles of deflection of 

 the butterflies tested in light produced by a disk with three- 

 fourths removed were respectively +1.37, +0.11, +1.31, +0.07, 

 2.15, and 5.26 degrees. Thus, in all three sorts of light the 

 angles of deflection decrease in general as the flash-frequency 

 decreases from 30 to 2 per second. These angles are averaged 

 from various numbers of insects but this fact does not invalidate 

 the conclusions drawn from these average angles, for, while these 

 insects showed great variation, an examination of the angles 

 made by the different individuals leads to the same deduction. 

 Moreover, the conclusions drawn thus far in this section are sup- 

 ported by numerous results presented in the following sections. 



The statement that intermittent light of flash-frequencies of 

 20 and 30 per second has a higher stimulating efficiency in 



