with Special Reference to the Vision of Insects. 327 



the external world would, if not interfered with, form an 

 inverted image one inch behind this lens, that is, at the dis- 

 tance of the glass hemisphere, which we shall call the primary 

 surface ; but it is prevented from forming more than one 

 patch of that image by the blackened walls of the funnel. 

 Accordingly, only one tiny patch of the image, one-sixteenth 

 of an inch across, is actually formed. It is formed by the 

 light which passes the whole way down the funnel and, 

 emerging at its end, falls on the surface of the primary hemi- 

 sphere. Here it produces one little fragment of the inverted 

 image, the rest of the image which the lens is competent to 

 form being extinguished by the blackened walls of the funnel. 

 In the insect's eye the small portion of the image that 

 emerges is, no doubt, a portion of a rather indistinct image, 

 owing to the very small aperture of the lens ; but neither 

 this nor its belonging to an inverted image is any detriment, 

 since all the rays that go to form the little patch are trans- 

 mitted to a single one of the pieces of apparatus in the insect, 

 corresponding to the rods and cones in our eyes. They, 

 therefore, can result in only one of the optic nervelets being 

 atfected, and in some one definite way ; in other words, the 

 whole of the light forming one patch, or rather speck, of the 

 image can produce only one elementary visual impression in 

 the insect's mind. 



It will be observed that the image that is formed resembles 

 rather a mezzotinto engraving, which consists of separate 

 specks, than a mosaic which consists of patches of colour 

 large enough to touch one another ; and that it differs from 

 the mezzotinto in that the specks are specks of light, instead 

 of being, as in the engraving, specks of shade. 



If we endeavour to make out what provision is made in 

 compound eyes for enabling the insect to accommodate its 

 vision to varying distances of the object, we find, upon a 

 scrutiny of the section of such an eye, that the arrangement 

 appears to be one which gives to an insect the very singular 

 power of adjusting different parts of its field of view to 

 different distances, and operates in a remarkably simple way 

 which may be illustrated upon our model if we add somewhat 

 to it. For this purpose let a third hemisphere be provided, 

 concentric with the other two, but smaller — suppose with a 

 diameter ten inches across. Let the funnels which have been 

 spoken of, and which lie between the outer surface and the 

 primary surface, be made of some extensible material like 

 indiarubber, their outer ends beino; fastened to the lenses 

 and their inner ends to threads of glass the thickness of 

 thin knitting-needles, and extending, as in fig. 3, from the 



Z2 



