As soon as this important truth was established, Sir Isaac saw that a lens which refracts light exactly like a prism must also refract the differently coloured rays with different degrees of force, bringing the violet rays to a focus nearer the glass than the red rays. This is shown in _fig. 2_, where LL is a convex lens, and S, L, SL rays of the sun falling upon it in parallel directions. The violet rays existing in the white light SL being more refrangible than the rest, will be more refracted or bent, and will meet at V, forming there a violet image of the sun. In like manner the yellow rays will form an image of the sun at Y, and so on, the red rays, which are the least refrangible, being brought to a focus at R, and there forming a red image of the sun. [Illustration: _Fig. 2._] Hence, if we suppose LL to be the object-glass of a telescope directed to the sun, and MM an eye-glass through which the eye at E sees magnified the image or picture of the sun formed by LL, it cannot see distinctly all the different images between R and V. If it is adjusted so as to see distinctly the _yellow_ image at Y, as it is in the figure, it will not see distinctly either the _red_ or _violet_ images, nor indeed any of them but the yellow one. There will consequently be a distinct yellow image, with indistinct images of all the other colours, producing great confusion and indistinctness of vision. As soon as Sir Isaac perceived this result of his discovery, he abandoned his attempts to improve the refracting telescope, and took into consideration the principle of reflection; and as he found that rays of all colours were reflected regularly, so that the angle of reflection was equal to the angle of incidence, he concluded that, upon this principle, _optical instruments might be brought to any degree of perfection imaginable_, provided a reflecting substance could be found which could polish as finely as glass, and reflect as much light as glass transmits, and provided a method of communicating to it a parabolic figure could be obtained. These difficulties, however, appeared to him very great, and he even thought them insuperable when he considered that, as any irregularity in a reflecting surface makes the rays deviate five or six times more from their true path than similar irregularities in a refracting surface, a much greater degree of nicety would be required in figuring reflecting specula than refracting lenses.