It may be remarked that the curved surfaces of the lenses forming telescopic object-glasses must not be parts of spheres. If they are, the images will be rendered indistinct by _spherical aberration_, and the optician has to design his curves to get rid of this defect at the same time as chromatic aberration. A new form of telescopic objective, consisting of three lenses, which has many important advantages, has been invented by Mr. Dennis Taylor, of the well-known firm of T. Cooke & Sons, York, England. Such a lens as this illustrates the perfection which the optician’s art has now attained. Six surfaces of glass have to be so accurately figured that every ray of light falling upon the surface of the lens shall pass through the finest pin-hole at a distance of eighteen times the diameter of the lens. THE REFLECTOR.—In a reflecting telescope, the object-glass of the refractor is replaced by a concave mirror. In order that such a mirror may reflect all the rays from a star to a single point, its concave surface must be part of a paraboloid of revolution, that is, a surface produced by the revolution of a parabola on its axis. If a spherical surface be employed, all the rays will not be reflected to a single point and the images which it gives will be ill-defined. Yet it is astonishing to find that the difference between a parabolic and spherical surface, even in the case of a large mirror, is exceedingly small. Sir John Herschel states that in the case of a mirror four feet in diameter, and forming an image at a distance of forty feet, the parabolic only departs from the spherical form at the edges by less than a twenty-one thousandth part of an inch. [Illustration: Fig. 21.—The Newtonian Reflector]