It will be seen that the three slits can be closed or opened from the centre by a parallel motion. They also slide in a couple of grooves, so that they can be moved along the frame into any position. The position they occupy is indicated by a scale engraved on the front of the slide. Behind the grooves in which the slits move are another pair of grooves, into which small pieces of card CCCC can slide, and thus close the apertures between the slits. By this arrangement all rays except those coming through the slits themselves are cut off. The metal frame fits on to an outer wooden frame, which slides in the grooves used with the card in the apparatus as already described. It is convenient always to keep the scale on the back of this wooden slide in the same position as regards the shadow of the needle-point used for registering the position, and to move the slits along their grooves when a change in position is required. Using these three slits three different colours can be thrown on the same square patch on the screen. A very crucial experiment is to see if we can make white light by the admixture of three colours, for if this can be done it almost follows that any colour can be formed. We must use the colour patch apparatus, and begin with placing one slit in the violet near the line G, another between E and F, and a third between B and C of the solar spectrum, and fill up the gaps between them with cards as shown in the figure. For our present purpose it is better to make the colour patch and the white patch touch each other, not using the rod, as by this means we avoid fringes of colour. We shall find that the aperture of the slits can be so altered that we can produce a perfect match with the white reflected light. By placing the rotating sectors in front of the reflected beam we can reduce its intensity, so that the two patches are equally bright. By a tapering wedge we can measure the width of the slits, and thus get the proportions of these three different colours which must be used to give the white. This is a sample of the method that we employ when we match any other colour. Suppose, for instance, it be wished to measure the colour of a solution of bichromate of potash; it is placed in the path of the reflected light, and we have an orange strip of light which we have to match. In this case it will be found that the slit in the blue has to be closed entirely, and only the green and red slits opened. The intensities of the two lights are equalized by the rotating sectors as before. So again with a solution of permanganate of potash. In this instance no green light will be required (or if any of it but a trifle), and the colour of the permanganate will be formed by the rays coming through the blue and red slits.