In the arc next the land, in fixed lights, a great loss of light ensues from the escape of the rays uselessly in that direction. So far back as 1834, I suggested the placing a segment of a spherical mirror, with its centre of curvature coincident with F the focus of the system, so that the luminous pyramid MFM, of which the mirror MM forms the base, might be thrown back through the focal point and finally refracted into such a direction as to contribute to the effect of the lens QA _q_ in seaward and opposite arc. In the diagram (fig. 87), _r_ _ r_ indicate rays proceeding directly from F; _r′_ _r′_ rays reflected from MM through F, and finally refracted at QA _q_; and _r″_ _r″_ is the beam compounded of both. In the best glass-silvered mirrors, this accession of light would amount to nearly half of the light incident on them. In such an arrangement, a considerable radius is desirable to decrease the amount of aberration produced by a large flame. In the case of revolving lights of the first order, the radius would, of course, be limited to somewhat less than three feet, which is the focal distance of the lenses, between which and the focus, the reflecting segment must be placed; but in fixed lights, the lantern is the limit of radius, so that a focal length of five feet ten inches may be obtained. M. FRANÇOIS ground some beautiful mirrors of three feet radius, which were afterwards _tinned_ by his successor, M. LETOURNEAU, by a new process discovered by himself;[72] and that gentleman is at present engaged in the construction of reflecting spherical segments 1200^{mm.} square (about 16 superficial feet), to a radius of 1770^{mm.} (5 feet 10 inches), which subtend a vertical arc of about 40°. [72] See notice of a similar process practised about the year 1750 by Mr ROGERS of London, _ante_, p. 240. ~Arrangement of Dioptric Apparatus.~ The arrangements of the dioptric apparatus in the lightroom will be more fully understood by referring to the Plates.