THE SENSE ORGANS 145 



This membrane is formed of a peculiar fibrous connective 

 tissue. It is not only perforated, at regular intervals, to allow 

 of the passage of the terminations of the nerve-fibres to the 

 retinulae, but it is also pierced with regular rows of larger circular 

 holes, through which tracheae of considerable size pass into the 

 eye. Thus the fenestrate membrane resembles in appearance the 

 bottom of a colander pierced with holes of two sizes. The tracheae, 

 on entering the eye, occupy the large spaces or lumina (Im) which 

 separate the slender bases of the ommatidia. It is usually 

 stated [10, 152] that four of these tracheae surround the base of 

 each ommatidium, though Calvert[30] shews six, arranged in a 

 hexagon. I find that Calvert's statement is correct for the 

 arrangement of the tracheae close to the fenestrate membrane 

 (fig. 62, /). As they pass further into the eye, however, they 

 quickly tend to become arranged in parallel rows, with only a 

 slight indication of the original hexagonal arrangement left. 



It is generally agreed that the eyes of insects possess no power 

 of accommodation. There are no eye-muscles, and the lens is 

 absolutely fixed, both as regards position and shape. Vigier 1 , 

 however, has described certain fibrillae in the eye of the Dragonfly 

 which, he believes, enable the insect to shorten or lengthen its 

 ommatidia. As the Dragonfly uses its eyes chiefly during flight, 

 it seems obvious that nothing short of a very complete accom- 

 modation-apparatus would be of any value to it, owing to the 

 rapid change of position of the objects to be viewed. It seems 

 much more likely that the fixed focal length without accommo- 

 dation is quite satisfactory, and that this is supplemented by the 

 use of the ocelli for very near objects, which would otherwise 

 be quite out of focus. 



Variations in the form of the ommatidia. In the Zygoptera 

 the ommatidia are all alike, and all deeply pigmented. In the 

 Libellulidae, however, they are sharply marked off into two groups. 

 In the upper group, the facets are larger, the ommatidia con- 

 sequently further apart, the cones larger, and pigmentation is 

 absent (fig. 63 A). In the lower group, the facets are smaller, 

 the pigmentation very dense, and the ommatidia closely resemble 

 those of Zygoptera. The line of separation between the two 

 1 C. B. Acad. Sci., Paris, 1904, cxxxvra, pp. 775-777. 



T. D.-F. 10 



