l66 THE SENSES OF ANIMALS 



received. This mistake has come about because it is possible 

 to strip the outer layer containing the lenses from the surface 

 of the compound eyes of certain insects. If this layer is 

 examined with a microscope it can be arranged so that each 

 lens gives a separate image of an object at a suitable dis- 

 tance, but that does not justify an assumption that the com- 

 pound eye works that way in life. The assumption overlooks 

 the function of the tubular part of the ommatidium which 

 ensures that only the central ray gets to the bottom. It is 

 very difficult to prepare an insect eye so that, under a 

 microscope, you can look through it from the back, but 

 it has been done successfully with the eyes of certain kinds. 

 When, in such a preparation, you look through the eye 

 from the bottoms of the tubes you find that a single image 

 of an object is formed and that it is built up from the minute 

 dots of greater or lesser intensity from each ommatidium. 



A simple experiment can be made to show how the com- 

 pound eye works. Get a gross or so of lemonade drinking 

 straws, paint them black, and stick them together side by 

 side so that the ends are level and you have a solid chunk 

 of straw with a large number of long parallel holes running 

 through it. If you fix a piece of ground glass or thin paper 

 against one end of the tubes and point the other at a brightly 

 lit scene an image of what is in front appears on the screen 

 — and furthermore the image appears sharper if anything 

 is moving or if you move the tubes across the scene. 



That is not, however, the whole story of the compound 

 eye — which is not only compound, but complex. The tubular 

 system which uses only the direct ray coming down the 

 centre is efficient enough in bright light, but it is very 

 wasteful. It wastes much of the light because the tubes are 

 so arranged that all the oblique rays are absorbed by the 

 pigment surrounding them and cannot get into nearby tubes 

 and thus produce a blurred image. But when the light is 

 dim so much of it may be wasted in this way that the 

 animal is practically blind since so little light goes down 

 the centre. For vision in dim light therefore the efficiency 

 of the tube system is abandoned, and light is allowed to 

 spread from one tube to another. The pigment cells sur- 

 rounding the tubes can very quickly expand and contract — 



