THE LOCUST OR SHORT-HORNED GRASSHOPPER 363 



Moreover, the image that is formed upon the retina of an ocellus 

 must be coarse, because of the very small number of the retinal 

 cells. In the grasshopper, the ocelli are probably no more than 

 organs of light perception with lenses that merely condense the 

 rays of light. Thus, when the compound eyes are coated with 

 black paint the animal will find its way out of a box in which there 

 is only one small opening. After the ocelli, as well as the eyes, 

 have been painted, it does not find the opening except by chance. 

 The structure of the compound eye is remarkable for its com- 

 plexity, although such an eye is composed of many units, the 

 ommatidia, which are similar in structure {cf. Fig. 183 A and B). 

 As the diameter of each of these units is shghtly less at its inner 

 end, each ommatidium is perpendicular to the curved outer surface. 

 The functioning of such an eye must be very different from that of 

 the eye of a vertebrate. After much investigation and discussion 

 of this subject, the " mosaic theory," which was proposed almost 

 a century ago, is still generally accepted. 



It is thought that an image is formed by thousands of separate 

 points of hght, each of which corresponds to a distinct field of 

 vision in the external world. Each ommatidium is adapted to 

 transmit light along its axis only, as oblique rays are lost by 

 absorption in the black pigment which surrounds the crystalhne 

 cone and the axial rhabdom. Along the rhabdom, then, hght 

 can reach and affect the terminations of the optic nerve. Each 

 ommatidium does not itself form a picture; it simply preserves 

 the intensity and color of the light from one particular portion of 

 the field of vision ; and when this is done by hundreds or thousands 

 of contiguous ommatidia, an image results. All that the painter 

 does, who copies an object, is to put together patches of light in 

 the same relations of quality and position that he finds in the 

 object itself — and this is essentially what the compound eye does, 

 so far as can be inferred from its structure.^ 



It appears that such compound eyes are especially fitted to per- 

 ceive moving objects. It is a famihar observation that insects 

 like grasshoppers are sensitive even to shght movements in their 

 vicinity ; yet they may be picked up with the fingers if the approach 

 is made slowly until the animal is within the grasp. This may be 

 explained on the assumption that the moving object stimulates 

 different ommatidia in succession without the necessity of turning 

 eyes or head as in the vertebrate, and that this successive stimula- 



2Folsom, "Entomology," pp. 111-112. 



