VII] THE SENSE ORGANS 147 



appear to meet along a median eye-line. A transverse section 

 of this region (fig. 63 B) shews the true nature of the junction. 

 The eyes are actually separated by an excessively narrow space, 

 occupied by two layers of hypoderm-cells, and bridged over above 

 by a narrow arch of chitin. At the borders of the eye, the chitin 

 is turned inwards on each side of this arch, and extends downwards 

 almost to the level of the fenestrate membrane. This formation 

 is clearly the last remnant of the epicranium, which is therefore 

 seen to be still actually continuous from vertex to occiput, even 

 in the highest forms. 



Function of the Compound Eye. 



We have already stated that the compound eye of the Dragonfly 

 is used for far vision ; that is to say, far in comparison with the 

 range of the ocelli. To the ordinary observer, a Dragonfly would 

 appear very short-sighted, since it sees objects best at distances 

 ranging from a few inches to five or six feet. Compared with other 

 insects, however, the Dragonfly may be reckoned long-sighted. 

 It has been stated that no insect can see beyond a distance of 

 six feet. This may be so as regards perception of form. But in 

 the case of the larger Dragonflies, perception of movement extends 

 to a much greater distance. I have found that Petalura gigantea 

 can be frightened away by waving a net in the air at a distance 

 of twenty yards ; while, in the case of Hemicordulia tau hovering 

 motionless over a pool, a sharp movement at a distance of ten 

 yards always produces a sudden dodge sideways on the part of 

 the Dragonfly. On the other hand, by moving slowly and stealthily, 

 without any jerks, it is possible, though not easy, to stalk a large 

 Dragonfly sitting on a twig, and to capture it with the hand. 

 I have caught Anax papuensis in this manner. 



Theories of Vision of the Compound Eye. 



Space will not permit us to go fully into this fascinating 

 question. We shall simply state the two chief rival theories, and 

 leave our readers to choose for themselves. 



The Theory of Mosaic Vision was formulated by J. Miiller in 

 1826 [95]. Its principal supporters have been Huxley, Lubbock 

 and Exner. It is easily understood from Lubbock's well-known 



102 



