162 A fly's eye. 



The rods alone of a fly's eye, I am inclined to think, may 

 correspond to the " rods and cones " of the mammalian eye, but 

 the insect's cones most certainly have nothing to do with them, for 

 they are, as I consider, a modification of the crystalline lenses.) 

 These rods of a fly's eye are made up of a central nerve fibre, 

 surrounded by finer fibres, the whole being in a sheath, like a 

 telegraph cable. They are considered to be true " nerve-end 

 ceUs ; " ?>., the portion of the nerve that receives a sensation. 

 Fig. 4, r.;-, shows a set taken across the eye, and Fig. 5 r, gives a 

 single one more highly magnified. They stand on a thin mem- 

 brane pierced with many holes {m\ Figs. 4 and 5), a hole for each 

 rod, and a hole for each one of numerous air-vessels, of which I 

 will speak directly. Passing through the membrane, the rods 

 change into an ordinary nerve-fibre {nv.. Fig. 5 ; see also Fig. 4). 

 After a short distance, each fibre swells into a cell-like body, and 

 these bodies make altogether a kind of nerve-junction or ganglion 

 [n.j. 1, Figs. 4 and 5), which stands on a second membrane, ;/r. 

 Immediately above the ganglion is a layer of nerve cells with very 

 large nuclei, between which the fibres pass. Contracting again, 

 as they emerge from the underside of this ganglion, the nervelets 

 become very thin ; they run quite straight, and those from the 

 Jiinder side of the eye cross over to the fro7it side (Figs. 2 and 4). 

 When they have crossed, they enter a mass of small nerve cells, 

 and, each individual nervelet swelling again, form a second 

 ganglion, something like the first (;/. y'., 2). The nervelets issue 

 from this nerve-junction as fibres once more, and collect together 

 to form the optic nerve, which finally loses itself in the brain 

 (Fig. 4, op. 71.) 



To describe all this is easy ; to say what is the function of 

 each part is very difticult ; and it is indeed impossible to afifirm 

 specifically, " this part does this, and the other part performs that." 

 Yet I have no doubt that the unusual complexity of this retina is 

 for the purpose of combining the thousand and one images formed 

 by the lenses, for insects with simple eyes, spiders and caterpillars 

 to wit, have not such a complicated retina. Besides nerves, a 

 fly's eye is full of air-vessels. Some, shaped like thin French 

 beans, stand up between the rods^ springing from a matted layer 

 of them which lies just below the first membrane {av^ av, Figs. 4 



