round, and moves as though there were no force acting upon it. As a consequence of this hypothesis, it follows that the velocity of light must be greater the denser the medium, while the undulatory theory leads to precisely the opposite result. When Foucault directly measured the velocity of light both in air and water, and found it less in the denser medium, the result was fatal to the corpuscular theory. Dr. Young called attention to another crucial test between the two theories. When a piece of plate-glass is pressed against a slightly convex lens, or a watch-glass, a series of coloured rings is formed by reflected light, with a black spot in the centre. This was accounted for by Newton by supposing that the light which was reflected in any ring was in a fit of easy transmission (from glass to air) when it reached the first surface of the film of air, and in a fit of easy reflection when it reached the second surface. By measuring the thickness of a film of air corresponding to the first ring of any particular colour, the length of path corresponding to the interval between two fits for that particular kind of light could be determined. When water instead of air is placed between the glasses, according to the corpuscular theory the rings should expand; but according to the undulatory theory they should contract; for the wave-length corresponds to the distance between successive fits of the same kind on the corpuscular hypothesis. On trying the experiment, the rings were seen to contract. This result seemed to favour the undulatory theory; but the objection urged by Newton that rays of light do not bend round obstacles, like waves of sound, still held its ground. This objection Young completely demolished by his principle of interference. He showed that when light passes through an aperture in a screen, whatever the shape of the aperture, provided its width is large in comparison with the length of a wave of light (one fifty-thousandth of an inch), no sensible amount of light will reach any point not directly in front of the aperture; for if any point be taken to the right or left, the disturbances reaching that point from different points of the aperture will neutralize one another by interference, and thus no light will be appreciable. When the breadth of the aperture is only a small multiple of a wave-length, then there will be some points outside the direct beam at which the disturbances from different points of the aperture will not completely destroy one another, and others at which they will destroy one another; and these points will be different for light of different wave-lengths. In this way Young not only explained the rectilinear propagation of light, but accounted for the coloured bands formed when light diverges from a point through a very narrow aperture. In a similar way he accounted for the hyperbolic bands of colour observed by Grimaldi within the shadow of a square near its corners. With a strip of card