Plateau experimented in the following way: In a darkened room, with two differently shaped but nearly equal light-openings, one square and open, the other subdivided into a number of small holes, and therefore of more difficult egress, he observed the choices of opening made by insects flying from the other end of the room. Careful practical provisions were made to eliminate error; the light-intensity of the two openings was as far as possible equalized or else noted, and no trees or other external objects were in view. The room was not darkened beyond the limit at which ordinary type ceases to be readable, otherwise the insects refused to fly (it is well known that during the passage of a thick cloud insects usually cease to fly). These observations were made on insects both with or without ocelli, in addition to the compound eyes, and with the same results. From repeated experiments on flies, bees, etc., butterflies and moths, dragon-flies and beetles, Plateau concludes that insects with compound eyes do not notice differences in form of openings in a half-darkened room, but fly with equal readiness to the apparently easy and apparently difficult way of escape; that they are attracted to the more intensely lighted opening, or to one with apparently greater surface; hence he concludes that they cannot distinguish the form of objects, at least only to a very slight extent, though they readily perceive objects in motion. One result of his experiments is that insects only utilize their eyes to choose between a _white_ luminous orifice in a dark chamber, or another orifice, or group of orifices, _equally white_. They are guided neither by odorous emanations nor by differences of color. He thinks that bees have as bad sight and act almost exactly as flies. From numerous experiments on Odonata, Coleoptera, Lepidoptera, Diptera, and Hymenoptera Plateau arrives provisionally at the following conclusions: 1. Diurnal insects have need of a quick strong light, and cannot direct their movements in partial obscurity. 2. Insects with compound eyes do not notice differences of form existing between two light orifices, and are deceived by an excess of luminous intensity as well as by the apparent excess of surface. In short, they do not distinguish the form of objects, or if they do, distinguish them very badly. Lubbock, however, does not fully accept Plateau’s experiments with the windows, and thinks they discern the form of bodies better than Plateau supposes. =How far can insects see?=—It is now supposed that no insects can perceive objects at a greater distance than about six feet. On an average Lepidoptera can see the movements of rather large bodies 1.50 meters, but Hymenoptera only 58 cm., and Diptera 68 cm.; while the firefly (Lampyris) can see tolerably well the form of large objects at a distance of over two meters.