SENSATION AND REACTION 89 



impossible its appreciation by the retinular cells. Recently, 

 however, good reason for regarding compound eyes as 

 capable of appreciating a distinct image has been given in 

 an extensive study of the subject by S. Exner (1891). His 

 work has been well expounded by H. Eltringham (19 19) who 

 by observations and experiments of his own has carried it 

 farther. Exner describes how in the common glowworm — 

 the wingless larva-like female of the beetle Lampyris noctiluca, 

 an insect active in feeble light — the pigment cells may move 

 outwards towards the corneal surface, so that the deeper 

 regions of the elements are no longer completely isolated, 

 and some of the rays traversing a cone may affect not only 

 its own retinula but also neighbouring retinulae. The 

 result is the formation of a set of erect images partly super- 

 imposed on each other, and it follows that the glow-worm 

 may thus obtain clear and definite vision of a portion of its 

 near surroundings. These superposition images are believed 

 by Exner to be characteristic of the vision of those insects, 

 such as moths, which fly in the dusk or at night time and 

 have therefore to make the best use of feeble light. Eltring- 

 ham has studied the images formed by the eyes of day- 

 flying insects such as butterflies, dragonflies, and blue- 

 bottles. In the case of the two latter groups, in which the 

 cone-forming cells are imperfectly changed into the com- 

 pletely " crystalline " substance, he agrees with Exner that 

 the general apposition image induced may be best repre- 

 sented to our imagination as *' a mosaic of light spots." 

 But in the compound eyes of butterflies which possess fully 

 developed transparent cones, ** there is at the apex of the 

 cone a tiny erect image of that part of the field appertaining 

 to each facet unit." Hence it may be inferred that an erect 

 image of the whole field of vision may be appreciated by 

 means of such an " eucone " eye. The reinversion of the 

 images by the cone, so as to make them ultimately erect, 

 disposes of the difficulty arising from the conception of an 

 extensive general image, made up of a large number of 

 minute individual images each regarded as reversed. 



But the clarity of an insect's vision as regards form is 



