between the combustible gas and the surrounding air. If this surface be too limited, relative to the generation of gaseous vapors, or if, for lack of an aircurrent, the burned gas is not rapidly removed, the unburned gaseous vapors must pass off as smoke. Hence, only small, solid wicks can be used. Argand met this difficulty by making the wick hollow, and by providing for an interior air-current, as well as an exterior one. By this means, not only is a large development of combustion surface obtained, but, by means of Argand's glass chimney, the air currents are made to flow rapidly and uniformly, giving steadiness, as well as amplitude of flame. The increased heat of burning, by reason of the enlarged flame, also adds essentially to the intensity of the light. Flame being really transparent, the light produced on the inner suface is mostly given out and utilized.

As it is indispensable for light-houses that the luminous flame should be quite limited in volume, in order that a proper direction may be given to the generated light, the Argand arrangement was an absolute prerequisite to any intense illumination for this purpose, whether from sperm, colza, or olive oil, or from any other known lighting, solid or fluid. Argand, at once, produced his lamp nearly in perfection; and the chief, almost the only, subsequent augmentation of its power was due to the memorable research in which Arago and Augustin Fresnel, acting as associates, produced the burner with two, three, four, and five concentric hollow wicks, with an air-current passing up through each of the open, ring-like tubes or spaces between the concentric wick or oil tubes. Thus, any number of concentric wicks might burn together, until the imperfect transparency of the flame should indicate a limit. The effect of this invention has been to gather the previous cluster of separate lamps into a single central lamp, of decidedly greater power; and which may be made to meet fully all proper demands of the highest existing towers. In such a case, the limit of range prescribed by the earth's sphericity can be effectively reached in ordinary states of the atmosphere, by a burner with four or five concentric wicks, producing a flame from four to four and a-half inches high, and four and a-half inches in diameter-the wicks being one-fourth

of an inch high, and consuming about six-sevenths of a gallon of oil hourly.

The mechanical part of this lightgenerating problem has reference chiefly to maintaining a continuous flow or supply of the oil or other fluid at a constant level, so as to preserve unaltered the conditions of combustion, and thus prevent unsteadiness of flame. The well-known Carcel or mechanical lamp is a common type of this arrangement. An expenditure of ingenuity has been bestowed upon this problem, which can best be appreciated by inspecting the great number of patent specifications for this object in the French Description des Machines. Innumerable specifications for burners, oil-cisterns, oil-pumps, oil-pump valves, lamp-regulators, oil-supply alarins, etc., -these give evidence to the difficulties and importance of the problem. The main result is, that there are now in use for light-houses, the clock-work movement lamp; the lamp of Henry Lepante; the Wagner lamp; the hydraulic lamp, and the pneumatic lamp. We must not enter the maze of specifications; for scarcely has the watch itself outrun the lamp in varieties of movement, model, and mechanism. The lamps used on board light-vessels are, or at least ought to be, simple Argand burners, properly mounted, and supplied for bearing the instability of a floating mass, and furnished with reflectors. same principles govern the shaping, formation, support, and adjustment of lamp-chimneys for light-houses, as in the common Argand lamps; though the enlarged wicks require them to assume proportionate dimensions, and make a free air-draft or ventilation of increased importance. The mechanical arrangements are usually so adjusted as to cause a flow, over the wicks, of about five times the amount of oil burned, which serves to prevent charring of the wicks, and obviates the need of frequent trimming and raising. The surplus is caught in a dripper, strained, and again used.

The

Practically, the chief materials used for light-house illumination, are sperm oil and colza, or rapeseed oil. The former is still used in this country; the latter, which is derived from a species of wild cabbage, is used entirely in France, chiefly in Great Britain, and is, indeed, altogether the main reliance of the European lights. Olive oil has

been pretty extensively employed, but is nearly gone out of use. Colza oil gives the intensest light; produces less charring of the wick; is less affected by cold; breaks fewer chimneys, and is, in most places, very much cheaper than sperm oil. Indeed, so decided is the superiority of colza over all other materials for light-house illumination, that very few lights in Europe are now burning anything else. Lard oil, cotton-seed oil, olive oil, various patent oils, both animal and vegetable, besides sundry organic compounds, evolved by modern chemistry, have been experimented on for this particular purpose, with results adverse to their relative availability, reliability, and economy. Carbureted hydrogen or coal gas, from coal, rosin, or oil, has been on trial in various light-houses here and elsewhere; but it is found to be quite too precarious, and the apparatus for its production and thorough purification is costly, and too complex to be worked by a common keeper. The same objections apply with increased force to the thres intensest known sources of light. The electric light from the brush between two charcoal points in a powerful circuit, is entirely at the mercy of battery derangements, and of the wasting, breaking, or maladjustment of the charcoal points. The Drummond light from a ball of lime, made incandescent in the oxyhydrogen blow-pipe flame, is peculiarly difficult to maintain, becauso the incandescent ball is constantly prone to fly to pieces under the excessive heat. In Gurney's Bude light, oil is burned in an oxygen jet, instead of the atmospheric mixture; hence an oxygen generating laboratory, with all its essential resources and skill, is required; and this, in the hands of a common keeper, cannot at all be relied on. Besides other economical and mechanical objections to this trio of intense lights, they would be too small to give the required vertical and horizontal divergence without bringing the parabolic reflectors or dioptic lenses too near for their safety. The sum of the whole matter is, therefore, that as lighthouses are all isolated, and kept by men of moderate skill, and as they yet require the utmost certainty of lighting, colza oil and sperm oil are the only illuminating substances which can now be safely and economically used for this purpose. Nothing is at present appa

rent which is likely to change this fact, though, of course, none can foretell what inventions or discoveries are impending in this inventive and inquiring age.

We would call special attention to the great benefits attainable, by establishing in our own country the colza culture, for the production of colza or rape-seed oil. It is clear that sperm oil cannot much longer maintain its place in our light-houses; but we must soon resort, as England already has, to the use of colza oil. The whale fishery is growing precarious, and the absolute supply of sperm oil is sensibly diminishing. Meanwhile machinery is devouring it at an increasing rate, and the price is rapidly advancing. In 1841-2, the light-house sperm oil cost $0 55 per gallon; in 1847-8, do. $1 07; in 1850-1, do. $1 17; in 1854, do. $1 39 to $1 58; and at the close of 1855, the cost of oil, delivered at the lights, was $2 25. It is, therefore, almost imperative that colza oil should be introduced into our service as soon as an adequate supply can be procured. To become dependent on importations for this purpose, is a very uupalatable contingency; but unless the colza culture is developed in our country, such importation must soon be begun. There is ample reason to believe that this culture might be made very profitable, if judiciously and vigorously undertaken. Throughout France, Belgium, Holland, and Germany, it occupies an important agricultural rank, and is in some parts the staple production. In England, and to some extent in this country, colza is cultivated for fertilizing aud grazing purposes, the oil from the seed being mostly neglected. The German population in Texas raise the colza, and express enough oil from its seeds to meet their domestic wants. In Mexico, its production is carried so far, that the lighting of streets and houses, in many villages and cities, is effected by domestic colza oil. It may, therefore, be regarded as proven, that colza oil can be readily produced in our country for domestic as well as public purposes, and that a large family consumption I could be relied on at remunerating rates. Even at present European rates for this oil, the colza culture would probably be quite as profitable as that of our chief agricultural staples. But for light-house uses, a much higher rate

would now be justified, and the expenses of importation would by no means interfere with the economy of introducing it, at least in our lens-lights. As soon as the growth and manufacture should become systematically established, an enormous consumption for houselighting might be anticipated. Shall we not, then, make haste to introduce so important a branch of production; one so needful for our light-houses, in case of not improbable failures in the whale fishery, and so essential, should we be involved in a commercial war? Would it not, even, be quite as wise for Congress to offer bounties for its initiation as it was thus to favor hemp and the fisheries?

We must now very concisely present the main features of the optical portion of the problem of light-house illumination. If, then, we conceive a simple, naked light, burning on the summit of a tower on an ocean headland, the rays would issue in all directions from the flame as a centre, though.only those portions of them which proceed in directions where they might reach a navigator's eye could be of any service. Not only would all the rays in the hemisphere-whose centre is the light, and which lies above a horizontal plane through the light-be without useful effect, but all the rays which proceed landwards, and which strike the ground, would be thrown away. In ordinary cases, not one-eighth of the light generated would be so emitted as, without artificial direction, to be of any use. Calling the range of a light the greatest distance at which the earth's curvature, atmospheric refraction, and its own elevation, would generally permit it to be seen, it is evident that what is wanted is, so to direct all the light generated as that it shall entirely fall on, or pass just over, the water within a sector traced around the light, with the range as a radius, and limited by the extreme radii passing over navigable areas. The question, then, is how to direct all the generated light within these limits of useful effect, and especially along the extreme water horizon limit.

There are two modes of changing the direction of a ray of light, leading to the two species or systems of light-house apparatus, called the catoptric, or reflecting system, and the dioptric, or refracting. As the usual dioptric apparatus has parts which use internal VOL. VIII.-14

reflection, the whole is also called a cata-dioptric apparatus-the two terms being currently applied to the same arrangement. Both reflection and refraction, in this connection, take place only at the limiting surfaces of solid, homogeneous masses, shaped for their special ends. The line perpendicular to a surface at any point is called the normal at that point.

The law governing reflection at surfaces is, that the incident and reflected rays always make equal angles with the normal, at the point of incidence. By means of this geometrical law, all the rays from any radiant point could be reflected in obedience to any given condition, if we were only able to make, with accuracy, mirrors of any geometrical form, and free from absorption. In fact, only spherical and paraboloidal mirrors are used. A mirror presenting the hollow surface of a spherical segment, is the one most easily made; but this is only an approximate instrument for throwing out the rays received from a light placed in front of it, over the water spread out before it. The focal radii and the diameters at each point of a paraboloid, make equal angles with the normal at that point: also, all the diameters of a paraboloid are parallel to each other. Hence, the paraboloid (this surface is shaped like a shallow washbowl) is characterized by the property that all the rays proceeding from its focus will be reflected at its concave surface in a beam of parallel rays. If a light-house lamp be placed at the focus of a paraboloid mirror whose axis is horizontal, the reflected light will constitute an approximate beam of parallel rays, in the precise direction for the best effect at the limit of visibility. Unfortunately, the reflected rays make only a limited portion of the whole radiation. But, what is worse still, if the light and mirror are stationary, the luminous beam having theoretically no divergence, and practically not over 15°, it would require twenty-four lamps set in a circle to illuminate the entire horizon by paraboloidal reflection. Besides this, all metallic reflections cause a great absolute loss of light, by absorption; even silver, with the best polish Lord Rosse is able to give it, absorbing from 7 to 10 per cent. His great six-feet reflector gives not much clearer illumination than the fifteen-inch Cambridge equatorial refractor possesses. Thus, paraboloid

mirrors are very far from being a correct theoretical solution of this problem; and the practical departures from theory, in so far as they correct the abstract faults, annul the abstract merits of the arrangement. Both in theory and practice, every simple metallic reflector is radically faulty; for it must waste much of the reflected light, and must leave the front radiation wholly uncorrected. Except for this loss of light and front radiation, a series of lamps on faces of a revolving frame, each lamp with its paraboloid mirror attached, would make up a satisfactory revolving light; but no satisfactory fixed light is thus possible.

As might be supposed, it was not till the Argand lamp had given an intense concentrated light that spherical or paraboloidal mirrors were to any extent used in light-houses. Some rude trials of plane mirrors, and paraboloids built of plane glass facets, preceded Borda's arrangement of Argand burners with paraboloid reflectors on a revolving frame, first set up in the Corduan tower, in 1784; but, practically, the great merit of this combination belongs either to Teulère or to Borda, who, aided by Lenoir's skill, really initiated the existing catoptric system. Probably no essential advance from Borda's arrangement will ever be made by using metallic reflection only. None has thus far been realized; and, from the nature of the case, all metallic catoptric arrangements must leave much of the light unutilized. Borda's plan, though still much in use, only survives by virtue of organic inertia, and it is now rapidly giving place to one vastly superior.

A strict geometric law also governs the refraction of light at the surfaces of transparent bodies. The sines of the two angles made by the incident and refracted portions of a ray with the normal at the point of incidence, bear a constant ratio to each other, for each substance, whatever be the angle of incidence. Each refracting medium, placed in vacuo, is characterized by its own special value of this ratio, called its index of refraction, which may once for all be experimentally determined for each substance. Knowing the indices .of refraction for the various media of a given combination, as a telescope, microscope, or a light-house refracting system, the entire course of any ray

a

therein can be accurately traced. Given, then, glass of a known index of refraction, how can all the rays of a central light-house lamp be strictly utilized by its use, and what shapes and positions must be given it?

To Augustin Fresnel the world owes enduring gratitude, for his elegant and almost faultless solution of this practical problem. Prior to his research, lenses had been tried for giving direction to light-house illumination; but these trials were very faulty, either optically or mechanically. In England and Ireland, simple spherical lenses were placed before lights a hundred years ago; but their great thickness, and the bad quality of the glass, made them, on the whole, injurious to their effect. Buffon, who was much engaged in forming burning-glasses by which the sun's parallel rays are focalized, proposed to cut away the central mass of glass, and to reduce the lens to a series of rings placed around the central lens in echelon order. As his idea was to make all these in a single connected piece of glass, its supreme mechanical difficulty made it virtually impracticable. Condorcet was the first to indicate the plan of a separate formation of the rings, with which large annular lenses might then be built up. Brewster, when treating of burning-glasses, in 1811, presented a clear exhibit of the composition and action of annular lenses, and advocated their use in the inverse problem of parallelizing the rays diverging from a light-house lamp flame as a focus. As he did not fully embody his ideas in practical forms, and as he, apparently inspired with less than his accustomed ardor, failed to procure responsive action by the inert light-house commissioners, no fruit resulted from his advanced conceptions. Brewster, of all men living, can best afford to spare a single optical laurel, but even this he is not bound wholly to forego. In 1819, Arago offered to undertake for the light-house commission a systematic series of researches, with the express object of improving light-house illumination, and for this he applied to have Mathieu and Fresnel assigned as co-laborers. It was through the acceptance of this proposal that Fresnel, being duly detailed as an officer of Ponts et Chaussées, was led to that brilliant train of researches and inventions so admirably detailed in his Memoir, read before the Academy July

29, 1822, in which, not knowing of Brewster's conceptions, he takes up the whole problem de novo. He was already recognized as the profoundest optical philosopher of his age, and as a perfect master of the most difficult analytical implements. Among the many illustrious opticians since Newton and Huygens, we think not one has possessed so excellent a blending of all the qualities and powers needed for fruitful and complete research as Augustin Fresnel. As with Snellius and Malus, his brilliant career of research was prematurely closed, yet each of this illustrious trio made fundamental discoveries which only Young has equaled since Newton and Huygens. To them we may apply Newton's saying when Coates died: "Had these men lived, we should have known something." Huygens originated the watch, and the undulatory theory of light; Fresnel approached his merit, by inventing his light-house apparatus, and by discovering the formulæ of interference, double refraction, and polarization. Before the trained powers of such a man, the difficulties of lighthouse optics vanished forever. Not content with vaguely indicating desirable combinations, he determined, with precision, their exact form, dimensions, and modifications. He left but few improvements to be made, and even these he had indicated the mode of effecting. It is not amiss, therefore, to call the dioptric or catadioptric light-house apparatus, now in general use, the Fresnel lens-a name than which no worthier or more enduring monument could be erected. We will now indicate briefly the prevailing forms of these lenses.

The light is produced by a single central lamp-flame, proceeding from concentric wicks, varying in number from one to five, the focus being the central point of flame. Around this are arranged, for a fixed light of the first order, horizontal hoops or rings of glass, so shaped and placed as to throw out in a horizontal direction all the light received on them. Thus while the horizontal divergence is duly preserved, the vertical divergence is counteracted, and nearly all the rays are brought into a flat, star-like horizontalism (as when a chestnut burr is pressed flat), and the illumination is equally diffused over all points in the horizon. The number of these rings varies with the order of the light, and, in all cases, the thickness of

glass to be penetrated is so small that absorption produces only a slight loss. The middle ring, at the level of the flame, is plane-convex in cross-section, with the convexity outwards, and is of considerable breadth above and below the focal level. The rings just above and below this have a four-sided, approximately-trapezoidal section, and with the precise curvature on the exterior for parallelizing and horizontalizing the emerging rays. The several rings, above and below, are similarly determined. All these rings are limited by horizontal top and bottom surfaces, and their interior surfaces together make up a vertical cylinder: thus all the curvatures are thrown into the outer surfaces. The horizontal glass surfaces in contact are cemented, and the segments of the rings are sustained by metal ribs placed radially, and connected with the main supporting-frame. This cylindrical refractor receives the rays for about 30° above and 30° below the horizontal plane, through the focus. It is surmounted by a domelike arrangement of prismatic zones, so adjusted as to receive and horizontalize the rays between about 30° and 80° above the horizontal. These zones give a spherical triangle in cross-section. The light enters at the under side, passes to the superior face, where it is intervally reflected, and, after a second refraction at the outer surface, it emerges duly horizontalized. It is fortunate that interval reflection is attended with far less loss of light than in metallic reflection, provided it take place as in these zones, within the angle called the angle of total reflection. The cylinder of the main cylindrical refractor is extended downward by several zones of spherical triangular cross-section, operating like the upper zones, and receiving the light from 30° to 529 below the horizontal. Thus all the rays, except an upper cone of 10° angle, and a lower one of 38° angle, are received on glass rings, and thrown out horizontally and uniformly in each azimuth. The upper opening is needed for the chimney, and the lower for the lamp and lamp-attendance-very little light being thrown down in that space any way, because it is eclipsed by the lamp and wick. Such, in general terms, is a fixed catadioptric lens-the cylindrical refractor simply bending the rays horizontally, while the upper and lower zones combine