If now a circular central part of the curved surface of a planoconvex lens is moved parallel to itself until at its edges the glass is very thin, the diminution of thickness will not affect the parallelism of the rays after deviation, and the absorption will be very much lessened. It' another part of the lens, of a convenient breadth and concentric with the first part, be moved as was the first part until its edges become very thin, the thickness of this will not much increase the absorption, and so of the whole surface of the lens; that is, it can be divided into thin concentric rings of convenient breadth and of nearly the same curvature as the lens, which will absorb but little light, and at the same time will send out the rays parallel to each other, and, if properly adjusted, parallel to the horizon. Buffon first imagined this manner of constructing a lens. Condorcet in 1773 suggested that the rings might be made in separate pieces, and Sir David Brewster made the same suggestion in 1811. Fresnel, without knowing Condor-cet's or Brewster's suggestions, conceived the idea of making the lenses in steps and in separate pieces, and, following it up, had the lenses manufactured and applied to lighthouses.

The vertical central section of Fresnel's lens, instead of being that of a plano-convex lens, is a figure Bounded on the side toward the lamp by a vertical straight line, and on the outside by a serrated line. This last line is a portion of the arc of a circle at its central part, and receding from the centre consists of portions of arcs of circles bounded by horizontal lines. The first lens apparatus made by Fresnel consisted of eight lenses like that above described, arranged in the form of an octagonal prism. It is evident that an eye situated in the horizon would perceive a bright flash whenever one of these lenses came in front of it; and supposing the octagonal prism to be revolved about its vertical axis, there will be eight flashes in one revolution. In Fresnel's first apparatus, and in all very large ones manufactured within 10 or 12 years after his invention, the rays in the portion of the sphere above the belt deviated by the lens were brought to the horizon by a combination of lenses and plane reflectors, and those below by the combination of curved glass reflectors similar to the slats of Venetian blinds, except that the reflectors are not precisely parallel, but are placed at such angles that all the light they receive shall be thrown to the horizon.

The interval between the flashes is diminished by arranging the auxiliary mirrors so that they will reflect the light a little to one side of the beam refracted by the lenses. This makes the flash longer, and correspondently diminishes the dark interval. In order to produce a fixed lens light which shall show uniformly entirely around the horizon, if the central vertical section of the lens (the section bounded by the serrated line above described) be revolved about the vertical line drawn through the principal focus of the lens, it will generate a solid of revolution, which when made of glass will fulfil the required condition for all rays 22 1/2° above and 224° below the horizon. Those above and below this zone are brought to the horizon by a combination of lenses and reflectors on the same principle as those described for a revolving lens. But Fres-nel was not satisfied with the use of reflectors for bringing to the horizon the rays above and below the central belt of 45°. On account of the acuteness of the angles at which these rays must be incident upon any deviating surface, it was not practicable to bend them to the horizon merely by refraction at two surfaces.

He therefore calculated the dimensions of a series of annular prisms, so arranged that the rays from the lamp incident upon the first surface of the prisms were refracted by it toward the horizon, were incident upon the second surface at an angle greater than that of total reflection, were reflected by it, and were so refracted by the third surface that they emerged from the prism horizontal. Thus all the rays proceeding from the lamp, except those obstructed by the glass chimney and the lamp itself, were utilized by the lens, forming the very perfection of a lighthouse apparatus. Fresnel did not live to see his idea of using the prisms instead of reflectors in the large lens apparatus carried out; but small apparatus were made on this principle for harbor lights with entire success. It is believed that the annular prisms were first used in an apparatus of the largest kind in one made at Paris under the direction of Alan Stevenson, engineer of the commission of northern lights of Scotland. The prismatic rings placed above and below the annular lens will with the lens throw all the rays to the horizon, and the combination will thus answer admirably for a fixed light. The annular prisms to fulfil their object must be arranged in conical or beehive shape above and below the annular lens.

For a revolving light, a vertical central section of the annular lens with a meridian section of the system of prisms was revolved around the horizontal line joining the centre of the annular lens and the principal focus of the combination. The revolution was continued far enough to generate a larger or smaller solid as the interval between the flashes was greater or smaller, the dimensions at the top and bottom of the lens regulating the amplitude of the revolution. Thus a polygon of 8 sides answers for an interval of one minute, supposing the time of revolution to be 8 minutes, one of 16 sides to an interval of 30 seconds, and one of 24 sides to an interval of 15 seconds, supposing the time of revolution to be 6 minutes. In order to lengthen the flashes, the upper and lower systems of prisms were moved a little to one side of the central annular lenses. The flash from the prisms was therefore produced a little after that from the lens, but so soon after as to appear a part of it. Another distinction was formed by revolving a system of cylindrical vertical lenses around a fixed apparatus and outside of it. These lenses collected the rays incident upon them, and emitted them parallel to each other and to the horizon.