It is an absolute fact that dragon-flies, unless it is cold or in the evening, always manage to fly at just that distance at which the student cannot touch them; and they see perfectly well whether one is armed with a net or has nothing but his hands; one might even say that they measure the length of the handle of the net, for the possession of a long handle is no advantage. They fly just out of reach of one's instrument, whatever trouble one may give oneself by hiding it from them and suddenly lunging as they fly off. Whoever watches butterflies and flies will soon see that these insects also can measure the distance of such objects as are not far from them. The males and females of bees and ants distinguish one another on the wing. It is rare for an individual to lose sight of the swarm or to miss what it pursues flying. It has been proved that the sense of smell has nothing to do with this matter. Thus insects, though without any power of accommodation for light or distance, are able to perceive objects at different distances. It is known that many insects will blindly fly and dash against a lamp at night, until they burn themselves. It has often been wrongly thought that they are fascinated. We ought first to remember that natural lights, concentrated at one point like our artificial lights, are extremely rare in Nature. The light of day, which is the light of wild animals, is not concentrated at one point. Insects, when they are in darkness--underground, beneath bark or leaves--are accustomed to reach the open air, where the light is everywhere diffused, by directing themselves towards the luminous point. At night, when they fly towards a lamp, they are evidently deceived, and their small brains cannot comprehend the novelty of this light concentrated at one spot. Consequently, their fruitless efforts are again and again renewed against the flame, and the poor innocents end by burning themselves. Several domestic insects, which have become little by little adapted to artificial light in the course of generations, no longer allow themselves to be deceived thereby. This is the case with house-flies. Bees distinguish all colours, and seldom confound any but blue and green; while wasps scarcely react to differences of colour, but note better the shape of an object, and note, for instance, where the place of honey is; so that a change of colour on the disc whereon the honey is placed hardly upsets them. Further, wasps have a better sense of smell than bees. The chief discovery regarding the vision of insects made in the last thirty years is that of Lubbock, who proved that ants perceive the ultra-violet rays of the spectrum, which we are unable, or almost unable, to perceive. It has lately been proved also that many insects appreciate light by the skin. They do not see as clearly as we do; but when they possess well-developed compound eyes they appreciate size, and more or less distinctly the contours of objects.