Newton then began to suspect that the rays, after passing through the prism, might move in curve lines, and, in proportion to the different degrees of curvature, might tend to different parts of the wall; and this suspicion was strengthened by the recollection that he had often seen a tennis-ball struck with an oblique racket describe such a curve line. In this case a circular and a progressive motion is communicated to the ball by the stroke, and in consequence of this, the direction of its motion was curvilineal, so that if the rays of light were globular bodies, they might acquire a circulating motion by their oblique passage out of one medium into another, and thus move like the tennis-ball in a curve line. Notwithstanding, however, “this plausible ground of suspicion,” he could discover no such curvature in their direction, and, what was enough for his purpose, he observed that the difference between the length MN of the image, and the diameter of the hole H, was proportional to their distance HM, which could not have happened had the rays moved in curvilineal paths. These different hypotheses, or suspicions, as Newton calls them, being thus gradually removed, he was at length led to an experiment which determined beyond a doubt the true cause of the elongation of the coloured image. Having taken a board with a small hole in it, he placed it behind the face BC of the prism, and close to it, so that he could transmit through the hole any one of the colours in MN, and keep back all the rest. When the hole, for example, was near C, no other light but the red fell upon the wall at N. He then placed behind N another board with a hole in it, and behind this board he placed another prism, so as to receive the red light at N, which passed through this hole in the second board. He then turned round the first prism ABC so as to make all the colours pass in succession through these two holes, and he marked their places on the wall. From the variation of these places, he saw that the _red_ rays at N were less refracted by the second prism than the _orange_ rays, the _orange_ less than the _yellow_, and so on, the _violet_ being more refracted than all the rest. Hence he drew the grand conclusion, _that light was not homogeneous, but consisted of rays, some of which were more refrangible than others_.