the aforesaid defects, I may send you some more of the experiments to second it (if it be so thought fit), in the ensuing Transactions.” Following the order which Newton himself adopted, we have, in the preceding chapter, given an account of the leading doctrine of the different refrangibility of light, and of the attempts to improve the reflecting telescope which that discovery suggested. We shall now, therefore, endeavour to make the reader acquainted with the other discoveries respecting colours which he at this time communicated to the Royal Society. [Illustration: _Fig. 4._] Having determined, by experiments already described, that a beam of white light, as emitted from the sun, consisted of seven different colours, which possess different degrees of refrangibility, he measured the relative extent of the coloured spaces, and found them to have the proportions shown in _fig. 4_, which represents the _prismatic_ spectrum, and which is nothing more than an elongated image of the sun produced by the rays being separated in different degrees from their original direction, the _red_ being refracted _least_, and the _violet most_ powerfully. If we consider light as consisting of minute particles of matter, we may form some notion of its decomposition by the prism from the following popular illustration. If we take steel filings of seven different degrees of fineness and mix them together, there are two ways in which we may conceive the mass to be decomposed, or, what is the same thing, all the seven different kinds of filings separated from each other. By means of seven sieves of different degrees of fineness, and so made that the finest will just transmit the finest powder and detain all the rest, while the next in fineness transmits the two finest powders and detains all the rest, and so on, it is obvious that all the powders may be completely separated from each other. If we again mix all the steel filings, and laying them upon a table, hold high above them a flat bar magnet, so that none of the filings are attracted, then if we bring the magnet nearer and nearer, we shall come to a point where the finest filings are drawn up to it. These being removed, and the magnet brought nearer still, the next finest powders will be attracted, and so on till we have thus drawn out of the mass all the powders in a separate state. We may conceive the bar magnet to be inclined to the surface of the steel filings, and so moved over the mass, that at the end nearest to them the heaviest or coarsest will be attracted, and all the remotest and the finest or lighter filings, while the rest are attracted to intermediate points, so that the seven different filings are not only separated, but are found adhering in separate patches to the surface of the flat magnet. The first of these methods, with the sieves, may represent the process of decomposing light, by which certain rays of white light are absorbed, or stifled,