A fly's eye. 163 



and 5). In addition to this, the whole eye is surrounded by very 

 large air-vessels {A V, A F, Figs. 2 and 4). I have examined a 

 living may-fly grub — an insect with a very transparent skin — and I 

 see that where these large air-vessels are, there is a broad stream 

 of blood running into or round the eye. Insects have no blood- 

 vessels, but the blood follows the course of the air-vessels. These 

 penetrate the whole body, and the blood runs outside them. P^or 

 any nerve to feel, it is necessary that a supply of fresh blood and 

 fresh air should constantly be applied to it. The air-vessels 

 between the rods, and the matted layer of them lying immediately 

 above the first nerve-junction, are necessary to keep these two 

 extremely sensitive parts of the retina in working order. Their 

 very presence, indeed, helps us to understand how sensitive these 

 parts are. The second nerve-junction has fine air-vessels running 

 into it. 



Bearing all this in mind, I would suggest that, in the first 

 nerve-junction, there is, as it were, a preliminary combining of the 

 images ; that the crossing of the nervelets and the second nerve- 

 junction completes the process ; and that the gathering of all the 

 much elaborated nervelets into one bundle of simple fibres after 

 the second nerve-junction (oJ>. ;/., Fig. 4) truly indicates that the 

 multitude of sensations has now been turned into one perception. 

 Yet I do not think that a fly sees with all its eyelets at once. We 

 ourselves only perceive distinctly at any one moment the very 

 small space represented by a disc half an inch in diameter held 

 about one foot from the eye. We think that we have distinct 

 vision over a larger surface than this, because we can shift the 

 direction of our eyes so instantaneously ; but it is not so. 

 ^Vhether or not a fly can see distinctly out of only one eyelet at a 

 time, I cannot tell. Probably it can only, like ourselves, see dis- 

 tmctly what it looks at and attends to, and that for this, as we 

 may say, it "looks out of" only one, or some small number of 

 eyelets at a time. So that the large number of eyelets, each one 

 capable of forming a complete picture, is to supply the place of 

 the power to move its eyes, the eyelets looking in every direction, 

 except behind the head. And although each eyelet forms a 

 complete picture, it is not a picture of the fly's whole field of 

 vision, but only of that small part of it in a line with the axis of 



