COLE. — IMAGE-FORMING POWERS OF VARIOUS TYPES OF EYES. 375 



greater efficiency of the eyes of the latter animal in distinguishing 

 between a large and a small area — an efficiency which is indicated in 

 one case by 7 per cent, in the other by 1.5 per cent. If Oniscus were 

 naturally as sensitive in its responses to light as Tenebrio, this 

 difference would probably be greater. In other words, the image- 

 forming power of the eye of Oniscus probably exceeds that of Tenebrio 

 more than appears in this method of experimentation, owing to the 

 fact that even to unilateral illumination the responses of the former 

 are so much less definite than those of the latter. 



5. Cockroach {Periflaneta americana Linn.). 



This cockroach, which abounds in slaughter-houses and other similar 

 situations where there is much waste matter that can serve as food, 

 possesses large compound eyes with a great number of ommatidia. 

 Latter ( : 04, p. 8.s), . in speaking of the related species, F. oi'ientalis, 

 says : "To what extent the compound eyes are capable of forming a 

 distinct image of surrounding objects we cannot say, but it is evident 

 that they are keenly sensitive to differences of light and shade from 

 the speed with which a cockroach makes for dark corners and crevices 

 when disturbed." This characteristic of cockroaches is very evident 

 to any one who has ever observed them. If a number of them are in a 

 glass jar, the cover may be removed with comparatively little danger 

 that any of them will escape so long as the mouth of the jar is turned 

 toward the light ; but if its position is reversed, they will usually 

 escape from it very quickly. These and similar observations led to the 

 selection of this animal also for experimentation with the lights of 

 different area. 



Cockroaches are so active and comparatively difficult to handle that 

 it took considerable experimenting to devise a scheme for getting them 

 oriented in the proper position without the introduction of disturbing 

 factors. If one attempts to liberate them from the hand, their 

 struggles usually result in their being in an undesirable position when 

 released, instead of exactly in the normal position. Finally, after 

 employing several unsuccessful devices, the following was made use 

 of. A glass cylinder some 15 centimeters in diameter and 20 centi- 

 meters high was placed endwise on the table and so arranged that 

 from one side of the bottom a low, narrow glass runway led off for a 

 distance of about 5 centimeters. This runway was large enough to 

 allow a cockroach to pass through it easily, but not large enough to 

 permit the insect to turn about. It was covered with black opaque 

 paper, which cut off the light and made its interior dark. This 



