COLE. — IMAGE-FORMING POWERS C? VARIOUS TYPES OF EYES. 369 



employed in the trials with the lights used separately and in 100 of 

 the 200 trials with both lights together ; in the remaining trials with 

 both lights five other larvae were employed. The results of the experi- 

 ments on the larvae of 

 Tciiebrio niolitor are class- 

 ified and summarized in 

 Table IV. 



The preponderance of 

 reactions away from the 

 light is strikingly shown 

 in the first two sets of 

 experiments, where uni- 

 lateral illumination was 

 tested. With the large 

 light there was not a sin- 

 gle positive reaction, and 

 only 4 out of 50 fell in 

 the class, so that 46, or 

 92 per cent, of the reac- 

 tions can be counted as 

 distinctly negative re- 

 sponses. The records with 

 the small light alone are 

 somewhat more scatter- 

 ing, but are nevertheless 

 preponderatingly nega- 

 tive. Only 8 records are 

 towards the light, while 

 in 41 out of the 50 trials 

 the animals turned to- 

 wards the dark, giving an 

 excess of 33, or G6 per 

 cent, in that direction. 

 The responses when the 

 two lights act simultane- 

 ously are almost equally 

 balanced as between the 

 two sides, by far the 



larger number of records on each side falling near the class, which 

 itself contains 43 of the 200 records, or nearly one fourth of the whole 

 number. This result may be much more quickly and completely 

 grasped from an inspection of Figure 11, C. On the side towards the 



VOL. XLII. — 24 



Figure 11. Frequency polygon constructed 

 from results with tlie mealworm shown in Table 

 IV. ^4, 50 trials to large light alone ; B, 50 trials 

 to small light alone ; C, 200 trials to both lights 

 used simultaneously. 



