Mr. De La Rue improved on this by giving the speculum a rotatory motion irrespective of that of the sliding plate, so that the side motion should not always be along the same diameter of the speculum. This was done by allowing the speculum to turn freely on a pivot on the sliding plate, and giving it a rotatory motion by means of a cord going round the plate carrying the speculum supports. As a further improvement Mr. De La Rue controls the motion of the polisher on the central pin, giving it motion by a crank carrying a system of wheels in place of the lowest crank, so that the pin gets a rotatory motion in addition to these. Mr. Grubb’s arrangement for polishing is different. The speculum is made to rotate, the polisher is made to execute curves variable at pleasure by altering the throw of the cranks which move rods attached to the centre of the polisher, giving it a motion similar to that of Mr. Lassell’s machine. The polisher moves a little off the edge, so that the edge is worn down more than the centre, thus giving the parabolic form. M. Foucault, of whom we have already spoken, proceeds in a different manner in parabolising his glass mirrors. He first obtains a spherical surface, fairly reflective, by grinding. He then alters the surface to a paraboloid form by handwork, only testing the surface from time to time to ascertain the parts requiring reduction by the polishing pad. The method of testing is as beautiful as it is simple. The approximate estimate of the curvature of the speculum is made by placing a small and well-defined object, such as the point of a pin, close to the centre of curvature and examining its image formed close by its side with a lens. As a nicer test, he places an object having parallel sides, say a flat ruler, near the centre of curvature, and views its image with the naked eye at the distance of distinct vision, then each point of the edge is seen by rays converging only from a small portion of the surface of the mirror, the remainder of the diverging cone from each point of the edge passes on beside the eye, and by moving the eye about, any point of the edge can be seen formed by rays proceeding from any particular part of the mirror, viz., that part in line with the eye and point of the edge examined; if the curvature be not uniform the edge will appear distorted, and points on it will appear in different positions, as rays from different parts of the mirror are received by the eye as it is moved, making the edge appear to move in waves. Finally, he allows light from a very small hole in a metal plate near the centre of curvature to fall on the mirror, and places the eye just on the side opposite to the point where the image is formed, so as to receive the rays as they diverge after having come to a focus. The whole of the light thus passes into the eye, and the mirror is seen illuminated in every part. A sharp edge of metal is then gradually brought into the focus, when the