As the reflector of a search-light is an exceedingly important factor in obtaining high beam-intensities, considerable attention has been given to it since the practicable electric arc appeared. The parabolic mirror has the property of rendering parallel, or nearly so, the rays from a light-source placed at its focus. If the mirror subtends a large angle at the light-source, a greater amount of light is intercepted and rendered parallel than in the case of smaller subtended angles; hence, mirrors are large and of as short focus as practicable. Search-light projectors direct from 30 to 60 per cent. of the available light into the beam, but with lens systems the effective angle is so small that a much smaller percentage is delivered in the beam. Mangin in 1874 made a reflector of glass in which both outer and inner surfaces were spherical but of different radii of curvature, so that the reflector was thicker in the middle. This device was "silvered" on the outside and the refraction in the glass, as the light passed through it to the mirror and back again, corrected the spherical aberration of the mirrored surface. These have been extensively used. Many combinations of curved surfaces have been developed for special projection purposes, but the parabolic mirror is still in favor for powerful search-lights. The tip of the positive carbon is placed at its focus and the effective angle in which light is intercepted by the mirror is generally about 125 degrees. Within this angle is included a large portion of the light emitted by the light-source in the case of direct-current arcs. If this angle is increased for a mirror of a given diameter by decreasing its focal length, the divergence of the beam is increased and the beam-intensity is diminished. This is due to the fact that the light-source now becomes apparently larger; that is, being of a given size it now subtends a larger angle at the reflector and departs more from the theoretical point.