Jointing of Stones in Rock Towers.—Various methods of jointing the stones in rock towers are shown in figs. 6 and 22. The great distinction between the towers built by successive engineers to the Trinity House and other rock lighthouses is that, in the former the stones of each course are dovetailed together both laterally and vertically and are not connected by metal or wooden pins and wedges and dowled as in most other cases. This dovetail method was first adopted at the Hanois Rock at the suggestion of Nicholas Douglass. On the upper face, one side and at one end of each block is a dovetailed projection. On the under face and the other side and end, corresponding dovetailed recesses are formed with just sufficient clearance for the raised bands to enter in setting (fig. 23). The cement mortar in the joint formed between the faces so locks the dovetails that the stones cannot be separated without breaking (fig. 24).
TableI.—Comparative Cost of Exposed Rock Towers.
Effect of Waves.—The wave stroke to which rock lighthouse towers are exposed is often considerable. At the Dhu Heartach, during the erection of the tower, 14 joggled stones, each of 2 tons weight, were washed away after having been set in cement at a height of 37 ft. above high water, and similar damage was done during the construction of the Bell Rock tower. The effect of waves on the Bishop Rock and Eddystone towers has been noted above.
Land Structures for Lighthouses.—The erection of lighthouse towers and other buildings on land presents no difficulties of construction, and such buildings are of ordinary architectural character. It will therefore be unnecessary to refer to them in detail. Attention is directed to the Phare d’Eckmühl at Penmarc’h (Finistère), completed in 1897. The cost of this magnificent structure, 207 ft. in height from the ground, was largely defrayed by a bequest of £12,000 left by the marquis de Blocqueville. It is constructed entirely of granite, and is octagonal in plan. The total cost of the tower and other lighthouse buildings amounted to £16,000.
The tower at Île Vierge (Finistère), completed in 1902, has an elevation of 247 ft. from the ground level to the focal plane, and is probably the highest structure of its kind in the world.
The brick tower, constructed at Spurn Point, at the entrance to the Humber and completed in 1895, replaced an earlier structure erected by Smeaton at the end of the 18th century. The existing tower is constructed on a foundation consisting of concrete cylinders sunk in the shingle beach. The focal plane of the light is elevated 120 ft. above high water.
Besides being built of stone or brick, land towers are frequentlyconstructed of cast iron plates or open steel-work with a view to economy. Fine examples of the former are to be found in many British colonies and elsewhere, that on Dassen Island (Cape of Good Hope), 105 ft. in height to the focal plane, being typical (fig. 25). Many openwork structures up to 200 ft. in height have been built. Recent examples are the towers erected at Cape San Thomé (Brazil) in 1882, 148 ft. in height (fig. 26), Mocha (Red Sea) in 1903, 180 ft. and Sanganeb Reef (Red Sea) 1906, 165 ft. in height to the focal plane.
3.Optical Apparatus.—Optical apparatus in lighthouses is required for one or other of three distinct purposes: (1) the concentration of the rays derived from the light source into a belt of light distributed evenly around the horizon, condensation in the vertical plane only being employed; (2) the concentration of the rays both vertically and horizontally into a pencil or cone of small angle directed towards the horizon and caused to revolve about the light source as a centre, thus producing a flashing light; and (3) the condensation of the light in the vertical plane and also in the horizontal plane in such a manner as to concentrate the rays over a limited azimuth only.
Apparatus falling under the first category produce a fixed light, and further distinction can be provided in this class by mechanical means of occultation, resulting in the production of an occulting or intermittent light. Apparatus included in the second class are usually employed to produce flashing lights, but sometimes the dual condensation is taken advantage of to produce a fixed pencil of rays thrown towards the horizon for the purpose of marking an isolated danger or the limits of a narrow channel. Such lights are best described by the French termfeux de direction. Catoptric apparatus, by which dual condensation is produced, are moreover sometimes used for fixed lights, the light pencils overlapping each other in azimuth. Apparatus of the third class are employed for sector lights or those throwing a beam of light over a wider azimuth than can be conveniently covered by an apparatus of the second class, and for reinforcing the beam of light emergent from a fixed apparatus in any required direction.
The above classification of apparatus depends on the resultant effect of the optical elements. Another classification divides the instruments themselves into three classes: (a) catoptric, (b) dioptric and (c) catadioptric.
Catoptricapparatus are those by which the light rays are reflected only from the faces of incidence, such as silvered mirrors of plane, spherical, parabolic or other profile.Dioptricelements are those in which the light rays pass through the optical glass, suffering refraction at the incident and emergent faces (fig. 27).Catadioptricelements are combined of the two foregoing and consist of optical prisms in which the light rays suffer refraction at the incident face, total internal reflexion at a second face and again refraction on emergence at the third face (fig. 28).
The object of these several forms of optical apparatus is not only to produce characteristics or distinctions in lights to enable them to be readily recognized by mariners, but to utilize the light rays in directions above and below the horizontal plane, and also, in the case of revolving or flashing lights, in azimuths not requiring to be illuminated for strengthening the beam in the direction of the mariner. It will be seen that the effective condensation in flashing lights is very much greater than in fixed belts, thus enabling higher intensities to be obtained by the use of flashing lights than with fixed apparatus.
Catoptric System.—Parabolic reflectors, consisting of small facets of silvered glass set in plaster of Paris, were first used about the year 1763 in some of the Mersey lights by Mr Hutchinson, then dock master at Liverpool (fig. 29). Spherical metallic reflectors were introduced in France in 1781, followed by parabolic reflectors on silvered copper in 1790 in England and France, and in Scotland in 1803. The earlier lights were of fixed type, a number of reflectors being arranged on a frame or stand in such a manner that the pencils of emergent rays overlapped and thus illuminated the whole horizon continuously. In 1783 the first revolving light was erected at Marstrand in Sweden. Similar apparatus were installed at Cordouan (1790), Flamborough Head (1806) and at the Bell Rock (1811). To produce a revolving or flashing light the reflectors were fixed on a revolving carriage having several faces. Three or more reflectors in a face were set with their axes parallel.A type of parabolic reflector now in use is shown in fig. 30. The sizes in general use vary from 21 in. to 24 in. diameter. These instruments are still largely used for light-vessel illumination, and a few important land lights are at the present time of catoptric type, including those at St Agnes (Scilly Islands), Cromer and St Anthony (Falmouth).Fig.22.—Courses of various Lighthouse Towers.Fig.23.—Perspective drawing of DovetailedStone (Wolf Rock).Fig.24.—Sectionof Dovetail.Dioptric System.—The first adaptation of dioptric lenses to lighthouses is probably due to T. Rogers, who used lenses at one of the Portland lighthouses between 1786 and 1790. Subsequently lenses by the same maker were used at Howth, Waterford and the North Foreland. Count Buffon had in 1748 proposed to grind out of a solid piece of glass a lens in steps or concentric zones in order to reduce the thickness to a minimum (fig. 31). Condorcet in 1773 and Sir D. Brewster in 1811 designed built-up lenses consisting of stepped annular rings. Neither of these proposals, however, was intended to apply to lighthouse purposes. In 1822 Augustin Fresnel constructed a built-up annular lens in which the centres of curvature of the different rings receded from the axis according to their distances from the centre, so as practically to eliminate spherical aberration; the only spherical surface being the small central part or “bull’s eye” (fig. 32). These lenses were intended for revolving lights only. Fresnel next produced his cylindric refractor or lens belt, consistingof a zone of glass generated by the revolution round a vertical axis of a medial section of the annular lens (fig. 33). The lens belt condensed and parallelized the light rays in the vertical plane only, while the annular lens does so in every plane. The first revolving light constructed from Fresnel’s designs was erected at the Cordouan lighthouse in 1823. It consisted of 8 panels of annular lenses placed round the lamp at a focal distance of 920 mm. To utilize the light, which would otherwise escape above the lenses, Fresnel introduced a series of 8 plain silvered mirrors, on which the light was thrown by a system of lenses. At a subsequent period mirrors were also placed in the lower part of the optic. The apparatus was revolved by clockwork. This optic embodied the first combination of dioptric and catoptric elements in one design (fig. 34). In the following year Fresnel designed a dioptric lens with catoptric mirrors for fixed light, which was the first of its kind installed in a lighthouse. It was erected at the Chassiron lighthouse in 1827 (fig. 35). This combination is geometrically perfect, but not so practically on account of the great loss of light entailed by metallic reflection which is at least 25% greater than the system described under. Before his death in 1827 Fresnel devised his totally reflecting or catadioptric prisms to take the place of the silvered reflectors previously used above and below the lens elements (fig. 28). The ray Fi falling on the prismoidal ring ABC is refracted in the directioni rand meeting the face AB at an angle of incidence greater than the critical, is totally reflected in the directionr eemerging after second refraction in a horizontal direction. Fresnel devised these prisms for use in fixed light apparatus, but the principle was, at a later date, also applied to flashing lights, in the first instance by T. Stevenson. Both the dioptric lens and catadioptric prism invented by Fresnel are still in general use, the mathematical calculations of the great French designer still forming the basis upon which lighthouse opticians work.Fig. 25.—Dassen IslandLighthouse (cast iron).Fig. 26.—Cape San ThoméLighthouse.Fig. 27.—Dioptric Prism.Fig. 28.—Catadioptric orReflecting Prism.Fresnel also designed a form of fixed and flashing light in which the distinction of a fixed light, varied by flashes, was produced by placing panels of straight refracting prisms in a vertical position on a revolving carriage outside the fixed light apparatus. The revolution of the upright prisms periodically increased the power of the beam, by condensation of the rays emergent from the fixed apparatus, in the horizontal plane.The lens segments in Fresnel’s early apparatus were of polygonal form instead of cylindrical, but subsequently manufacturers succeeded in grinding glass in cylindrical rings of the form now used. The first apparatus of this description was made by Messrs Cookson of Newcastle in 1836 at the suggestion of Alan Stevenson and erected at Inchkeith.In 1825 the French Commission des Phares decided upon the exclusive use of lenticular apparatus in its service. The Scottish Lighthouse Board followed with the Inchkeith revolving apparatus in 1835 and the Isle of May fixed optic in 1836. In the latter instrument Alan Stevenson introduced helical frames for holding the glass prisms in place, thus avoiding complete obstruction of the light rays in any azimuth. The first dioptric light erected by the Trinity House was that formerly at Start Point in Devonshire, constructed in 1836. Catadioptric or reflecting prisms for revolving lights were not used until 1850, when Alan Stevenson designed them for the North Ronaldshay lighthouse.Dioptric Mirror.—The next important improvement in lighthouse optical work was the invention of the dioptric spherical mirror by Mr (afterwards Sir) J. T. Chance in 1862. The zones or prisms are generated round a vertical axis and divided into segments. This form of mirror is still in general use (figs. 36 and 37).Fig. 29.—Early Reflector and Lamp (1763).Fig. 30.—ModernParabolic Reflector.Azimuthal Condensing Prisms.—Previous to 1850 all apparatus were designed to emit light of equal power in every azimuth either constantly or periodically. The only exception was where a light was situated on a stretch of coast where a mirror could be placed behind the flame to utilize the rays, which would otherwise pass landward, and reflect them back, passing through the flame and lens in a seaward direction. In order to increase the intensity of lights in certain azimuths T. Stevenson devised his azimuthal condensing prisms which, in various forms and methods of application, have been largely used for the purpose of strengthening the light rays in required directions as, for instance, where coloured sectors are provided. Applications of this system will be referred to subsequently.Optical Glass for Lighthouses.—In the early days of lens lights the only glass used for the prisms was made in France at the St Gobain and Premontré works, which have long been celebrated for the high quality of optical glass produced. The early dioptric lights erected in the United Kingdom, some 13 in all, were made by Messrs Cookson of South Shields, who were instructed by Léonor Fresnel, the brother of Augustin. At first they tried to mould the lens and then to grind it out of one thick sheet of glass. The successors of the Cookson firm abandoned the manufacture of lenses in 1845, and the firm of Letourneau & Lepaute of Paris again became the monopolists. In 1850 Messrs Chance Bros. & Co. of Birmingham began the manufacture of optical glass, assisted by M. Tabouret, a French expert who had been a colleague of Augustin Fresnel himself. The first light made by the firm was shown at the Great Exhibition of 1851, since when numerous dioptric apparatus have been constructed by Messrs Chance, who are, at this time, the only manufacturers of lighthouse glass in the United Kingdom. Most of the glass used for apparatus constructed in France is manufactured at St Gobain. Some of the glass used by German constructors is made at Rathenow in Prussia and Goslar in the Harz.The glass generally employed for lighthouse optics has for its refractive index a mean value of µ = 1.51, the corresponding critical angle being 41° 30′. Messrs Chance have used dense flint glass for the upper and lower refracting rings of high angle lenses and for dioptric mirrors in certain cases. This glass has a value of µ = l.62 with critical angle 38° 5′.Fig. 31.Buffon’s Lens.Fig. 32.Fresnel’s Annular Lens.Fig. 33.Fresnel’s Lens Belt.Fig. 34.—Fresnel’s Revolving Apparatus at Cordouan Lighthouse.Occulting Lights.—During the last 25 years of the 19th century the disadvantages of fixed lights became more and more apparent. At the present day the practice of installing such, except occasionally in the case of the smaller and less important of harbour or river lights, has practically ceased. The necessity for providing a distinctive characteristic for every light when possible has led to the conversion of many of the fixed-light apparatus of earlier years into occulting lights, and often to their supersession by more modern and powerful flashing apparatus. An occulting apparatus in general use consists of a cylindrical screen, fitting over the burner, rapidly lowered and raised by means of a cam-wheel at stated intervals. The cam-wheel is actuated by means of a weight or spring clock. Varying characteristics may be procured by means of such a contrivance—single, double, triple or other systems of occultation. The eclipses or periods of darkness bear much the same relation to the times of illumination as do the flashes to the eclipses in a revolving or flashing light. In the case of a first-order fixed light the cost of conversion to an occulting characteristic does not exceed £250 to £300. With apparatus illuminated by gas the occultations may be produced by successively raising and lowering the gas at stated intervals. Another form of occulting mechanism employed consists of a series of vertical screens mounted on a carriage and revolving round the burner. The carriage is rotated on rollers or ball bearings or carried upon a small mercury float. The usual driving mechanism employed is a spring clock. “Otter” screens are used in cases when it is desired to produce different periods of occultations in two or more positions in azimuth in order to differentiate sectors marking shoals, &c. The screens are of sheet metal blacked and arranged vertically, some what in the manner of the laths of a venetian blind, and operated by mechanical means.Leading Lights.—In the case of lights designed to act as a lead through a narrow channel or as direction lights, it is undesirable to employ a flashing apparatus. Fixed-light optics are employed to meet such cases, and are generally fitted with occulting mechanism. A typical apparatus of this description is that at Gage Roads, Fremantle, West Australia (fig. 38). The occulting bright light covers the fairway, and is flanked by sectors of occulting red and green light marking dangers and intensified by vertical condensing prisms. A good example of a holophotal direction light was exhibited at the 1900 Paris Exhibition, and afterwards erected at Suzac lighthouse (France). The light consists of an annular lens 500 mm. focal distance, of 180° horizontal angle and 157° vertical, with a mirror of 180° at the back. The lens throws a red beam of about 4½° amplitude in azimuth, and 50,000 candle-power over a narrow channel. The illuminant is an incandescent petroleum vapour burner. Holophotal direction lenses of this type can only be applied where the sector to be marked is of comparatively small angle. Silvered metallic mirrors of parabolic form are also used for the purpose. The use of single direction lights frequently renders the construction of separate towers for leading lights unnecessary.If two distinct lights are employed to indicate the line of navigation through a channel or between dangers they must be sufficiently far apart to afford a good lead, the front or seaward light being situated at a lower elevation than the rear or landward one.Coloured Lights.—Colour is used as seldom as possible as a distinction, entailing as it does a considerable reduction in the power of the light. It is necessary in some instances for differentiating sectors over dangers and for harbour lighting purposes. The use of coloured lights as alternating flashes for lighthouse lights is not to be commended, on account of the unequal absorption of the colouredand bright rays by the atmosphere. When such distinction has been employed, as in the Wolf Rock apparatus, the red and white beams can be approximately equalized in initial intensity by constructing the lens and prism panels for the red light of larger angle than those for the white beams. Owing to the absorption by the red colouring, the power of a red beam is only 40% of the intensity of the corresponding white light. The corresponding intensity of green light is 25%. When red or green sectors are employed they should invariably be reinforced by mirrors, azimuthal condensing prisms, or other means to raise the coloured beam to approximately the same intensity as the white light. With the introduction of group-flashing characteristics the necessity for using colour as a means of distinction disappeared.Fig. 35.—Fixed Apparatus at Chassiron Lighthouse (1827).Fig. 36.—Vertical Section. Prism of DioptricSpherical Mirror.High-Angle Vertical Lenses.—Messrs Chance of Birmingham have manufactured lenses having 97° of vertical amplitude, but this result was only attained by using dense flint glass of high refractive index for the upper and lower elements. It is doubtful, however, whether the use of refracting elements for a greater angle than 80° vertically is attended by any material corresponding advantage.Fig. 37.—Chance’s Dioptric Spherical Mirror.Group Flashing Lights.—One of the most useful distinctions consists in the grouping of two or more flashes separated by short intervals of darkness, the group being succeeded by a longer eclipse. Thus two, three or more flashes of, say, half second duration or less follow each other at intervals of about 2 seconds and are succeeded by an eclipse of, say, 10 seconds, the sequence being completed in a period of, say, 15 seconds. In 1874 Dr John Hopkinson introduced the very valuable improvement of dividing the lenses of a dioptric revolving light with the panels of reflecting prisms above and below them, setting them at an angle to produce the group-flashing characteristic. The first apparatus of this type constructed were those now in use at Tampico, Mexico and the Little Basses lighthouse, Ceylon (double flashing). The Casquets apparatus (triple flashing) was installed in 1877. A group-flashing catoptric light had, however, been exhibited from the “Royal Sovereign” light-vessel in 1875. A sectional plan of the quadruple-flashing first order apparatus at Pendeen in Cornwall is shown in fig. 39; and fig. 55 (Plate 1.) illustrates a double flashing first order light at Pachena Point in British Columbia. Hopkinson’s system has been very extensively used, most of the group-flashing lights shown in the accompanying tables, being designed upon the general lines he introduced. A modification of the system consists in grouping two or more lenses together separated by equal angles, and filling the remaining angle in azimuth by a reinforcing mirror or screen. A group-flashing distinction was proposed for gas lights by J. R. Wigham of Dublin, who obtained it in the case of a revolving apparatus by alternately raising and lowering the flame. The first apparatus in which this method was employed was erected at Galley Head, Co. Cork (1878). At this lighthouse 4 of Wigham’s large gas burners with four tiers of first-order revolving lenses, eight in each tier, were adopted. By successive lowering and raising of the gas flame at the focus of each tier of lenses he produced the group-flashing distinction. The light showed, instead of one prolonged flash at intervals of one minute, as would be produced by the apparatus in the absence of a gas occulter, a group of short flashes varying in number between six and seven. The uncertainty, however, in the number of flashes contained in each group is found to be an objection to the arrangement. This device was adopted at other gas-illuminated stations in Ireland at subsequent dates. The quadriform apparatus and gas installation at Galley Head were superseded in 1907 by a first order bi-form apparatus with incandescent oil vapour burner showing five flashes every 20 seconds.Fig. 38.—Gage Roads Direction Light.Fig. 39.—Pendeen Apparatus. Plan at Focal Plane.Flashing Lights indicating Numbers.—Captain F. A. Mahan, late engineer secretary to the United States Lighthouse Board, devised for that service a system of flashing lights to indicate certain numbers. The apparatus installed at Minot’s Ledge lighthouse near Boston Harbour, Massachusetts, has a flash indicating the number 143, thus: - ---- ---, the dashes indicating short flashes. Each group is separated by a longer period of darkness than that between successive members of a group. The flashes in a group indicating a figure are about 1½ seconds apart, the groups being 3 seconds apart, an interval of 16 seconds’ darkness occurring between each repetition. Thus the number is repeated every half minute. Two examples of this system were exhibited by the United States Lighthouse Board at the Chicago Exhibition in 1893, viz. the second-order apparatus just mentioned and a similar light of the first order for Cape Charles on the Virginian coast. The lenses are arranged in a somewhatsimilar manner to an ordinary group-flashing light, the groups of lenses being placed on one side of the optic, while the other is provided with a catadioptric mirror. This system of numerical flashing for lighthouses has been frequently proposed in various forms, notably by Lord Kelvin. The installation of the lights described is, however, the first practical application of the system to large and important coast lights. The great cost involved in the alteration of the lights of any country to comply with the requirements of a numerical system is one of the objections to its general adoption.Plate I.Fig. 54.—FASTNET LIGHTHOUSE—FIRST ORDER SINGLE-FLASHING BIFORM APPARATUS.Fig. 55.—PACHENA POINT LIGHTHOUSE,B.C.—FIRST ORDER DOUBLE-FLASHING APPARATUS.Plate II.Fig. 56.—OLD EDDYSTONE LIGHTHOUSE.Fig. 57.—EDDYSTONE LIGHTHOUSE.Fig. 58.—ILE VIERGE LIGHTHOUSE.Fig. 59.—MINOT’S LEDGE LIGHTHOUSE.Fig. 40.—Sule Skerry Apparatus.Hyper-radial Apparatus.—In 1885 Messrs Barbier of Paris constructed the first hyper-radial apparatus (1330 mm. focal distance) to the design of Messrs D. and C. Stevenson. This had a height of 1812 mm. It was tested during the South Foreland experiments in comparison with other lenses, and found to give excellent results with burners of large focal diameter. Apparatus of similar focal distance (1330 mm.) were subsequently established at Round Island, Bishop Rock, and Spurn Point in England, Fair Isle and Sule Skerry (fig. 40) in Scotland, Bull Rock and Tory Island in Ireland, Cape d’Antifer in France, Pei Yu-shan in China and a lighthouse in Brazil.The light erected in 1907 at Cape Race, Newfoundland, is a fine example of a four-sided hyper-radial apparatus mounted on a mercury float. The total weight of the revolving part of the light amounts to 7 tons, while the motive clock weight required to rotate this large mass at a speed of two complete revolutions a minute is only 8 cwt. and the weight of mercury required for flotation 950 ℔. A similar apparatus was placed at Manora Point, Karachi, India, in 1908 (fig. 41).The introduction of incandescent and other burners of focal compactness and high intensity has rendered the use of optics of such large dimensions as the above, intended for burners of great focal diameter, unnecessary. It is now possible to obtain with a second-order optic (or one of 700 mm. focal distance), having a powerful incandescent petroleum burner in focus, a beam of equal intensity to that which would be obtained from the apparatus having a 10-wick oil burner or 108-jet gas burner at its focus.Stephenson’s Spherical Lenses and Equiangular Prisms.—Mr C. A. Stephenson in 1888 designed a form of lens spherical in the horizontal and vertical sections. This admitted of the construction of lenses of long focal distance without the otherwise corresponding necessity of increased diameter of lantern. A lens of this type and of 1330 mm. focal distance was constructed in 1890 for Fair Isle lighthouse. The spherical form loses in efficiency if carried beyond an angle subtending 20° at the focus, and to obviate this loss Mr Stephenson designed his equiangular prisms, which have an inclination outwards. It is claimed by the designer that the use of equiangular prisms results in less loss of light and less divergence than is the case when either the spherical or Fresnel form is adopted. An example of this design is seen (fig. 40) in the Sule Skerry apparatus (1895).Fixed and Flashing Lights.—The use of these lights, which show a fixed beam varied at intervals by more powerful flashes, is not to be recommended, though a large number were constructed in the earlier years of dioptric illumination and many are still in existence. The distinction can be produced in one or other of three ways: (a) by the revolution of detached panels of straight condensing lens prisms placed vertically around a fixed light optic, (b) by utilizing revolving lens panels in the middle portion of the optic to produce the flashing light, the upper and lower sections of the apparatus being fixed zones of catadioptric or reflecting elements emitting a fixed belt of light, and (c) by interposing panels of fixed light section between the flashing light panels of a revolving apparatus. In certain conditions of the atmosphere it is possible for the fixed light of low power to be entirely obscured while the flashes are visible, thus vitiating the true characteristic of the light. Cases have frequently occurred of such lights being mistaken for, and even described in lists of light as, revolving or flashing lights.”Cute” and Screens.—Screens of coloured glass, intended to distinguish the light in particular azimuths, and of sheet iron, when it is desired to “cut off” the light sharply on any angle, should be fixed as far from the centre of the light as possible in order to reduce the escape of light rays due to divergence. These screens are usually attached to the lantern framing.Divergence.—A dioptric apparatus designed to bend all incident rays of light from the light source in a horizontal direction would, if the flame could be a point, have the effect of projecting a horizontal band or zone of light, in the case of a fixed apparatus, and a cylinder of light rays, in the case of a flashing light, towards the horizon. Thus the mariner in the near distance would receive no light, the rays, visible only at or near the horizon, passing above the level of his eye. In practice this does not occur, sufficient natural divergence being produced ordinarily owing to the magnitude of the flame. Where the electric arc is employed it is often necessary to design the prisms so as to produce artificial divergence. The measure of the natural divergence for any point of the lens is the angle whose sine is the ratio of the diameter of the flame to the distance of the point from centre of flame.In the case of vertical divergence the mean height of the flame must be substituted for the diameter. The angle thus obtained is the total divergence, that is, the sum of the angles above and below the horizontal plane or to right and left of the medial section. In fixed dioptric lights there is, of course, no divergence in the horizontal plane. In flashing lights the horizontal divergence is a matter of considerable importance, determining as it does the duration or length of time the flash is visible to the mariner.Feux-Éclairs or Quick Flashing Lights.—One of the most important developments in the character of lighthouse illuminating apparatus that has occurred in recent years has been in the direction of reducing the length of flash. The initiative in this matter was taken by the French lighthouse authorities, and in France alone forty lights of this type were established between 1892 and 1901. The use of short flash lights rapidly spread to other parts of the world. In England the lighthouse at Pendeen (1900) exhibits a quadruple flash every 15 seconds, the flashes being about ¼ second duration (fig. 39), while the bivalve apparatus erected on Lundy Island (1897) shows 2 flashes of1⁄3second duration in quick succession every 20 seconds. Since 1900 many quick flashing lights have been erected on the coasts of the United Kingdom and in other countries. The earlyfeux-éclairs, designed by the French engineers and others, had usually a flash of1⁄10th to1⁄3rd of a second duration. As a result of experiments carried out in France in 1903-1904,3⁄10second has been adopted by the French authorities as the minimum duration for white flashing lights. If shorter flashes are used it is found that the reduction in duration is attended by a corresponding, but not proportionate, diminution in effective intensity. In the case of many electric flashing lights the duration is of necessity reduced, but the greater initial intensity of the flash permits this loss without serious detriment to efficiency. Red or green requires a considerably greater duration than do white flashes. The intervals between the flashes in lights of this character are also small, 2½ seconds to 7 seconds. In group-flashing lights the intervals between the flashes are about 2 seconds or even less, with periods of 7 to 10 or 15 seconds between the groups. The flashes are arranged in single, double, triple or even quadruple groups, as in the older forms of apparatus. Thefeu-éclairtype of apparatus enables a far higher intensity of flash to be obtained than was previously possible without any corresponding increase in the luminous power of the burner or other source of light. This result depends entirely upon the greater ratio of condensation of light employed, panels of greater angular breadth than was customary in the older forms of apparatus being used with a higher rotatory velocity. It has been urged that short flashes are insufficient for taking bearings, but the utility of a light in this respect does not seem to depend so much upon the actual length of the flash as upon its frequent recurrence at short intervals. At the Paris Exhibition of 1900 was exhibited a fifth-order flashing light giving short flashes at 1 second intervals; this represents the extreme to which the movement towards the reduction of the period of flashing lights has yet been carried.Mercury Floats.—It has naturally been found impracticable to revolve the optical apparatus of a light with its mountings, sometimes weighing over 7 tons, at the high rate of speed required forfeux-éclairsby means of the old system of roller carriages, though for some small quick-revolving lights ball bearings have been successfully adopted. It has therefore become almost the universal practice to carry the rotating portions of the apparatus upon a mercury float. This beautiful application of mercury rotation was the invention of Bourdelles, and is now utilized not only for the high-speed apparatus, but also generally for the few examples of the older type still being constructed. The arrangement consists of an annular cast iron bath or trough of such dimensions that a similar but slightly smaller annular float immersed in the bath and surrounded by mercury displaces a volume of the liquid metal whose weight is equal to that of the apparatus supported. Thus a comparatively insignificant quantity of mercury, say 2 cwt., serves to ensure the flotation of a mass of over 3 tons. Certain differences exist between the type of float usually constructed in France and those generally designed by English engineers. In all cases provision is made for lowering the mercury bath or raising the float and apparatus for examination. Examples of mercury floats are shown in figs. 41, 42, 43 and Plate I., figs. 54and55.Fig. 41.—Manora Point Apparatus and Lantern.Multiform Apparatus.—In order to double the power to be obtained from a single apparatus at stations where lights of exceptionally high intensity are desired, the expedient of placing one complete lens apparatus above another has sometimes been adopted, as at the Bishop Rock (fig. 13), and at the Fastnet lighthouse in Ireland (Plate I., fig. 54). Triform and quadriform apparatus have also been erected in Ireland; particulars of the Tory Island triform apparatus will be found in table VII. The adoption of the multiform system involves the use of lanterns of increased height.Twin Apparatus.—Another method of doubling the power of a light is by mounting two complete and distinct optics side by side on the same revolving table, as I shown in fig. 43 of the Île Vierge apparatus. Several such lights have been installed by the French Lighthouse Service.Port Lights.—Small self-contained lanterns and lights are in common use for marking the entrances to harbours and in other similar positions where neither high power nor long range is requisite. Many such lights are unattended in the sense that they do not require the attention of a keeper for days and even weeks together. These are described in more detail in section 6 of this article. A typical port light consists of a copper or brass lantern containing a lens of the fourth order (250 mm. focal distance) or smaller, and a single wick or 2-wick Argand capillary burner. Duplex burners are also used. The apparatus may exhibit a fixed light or, more usually, an occulting characteristic is produced by the revolution of screens actuated by spring clockwork around the burner. The lantern may be placed at the top of a column, or suspended from the head of a mast. Coal gas and electricity are also used as illuminants for port lights when local supplies are available. The optical apparatus used in connexion with electric light is described below.”Orders” of Apparatus.—Augustin Fresnel divided the dioptric lenses, designed by him, into “orders” or sizes depending on their local distance. This division is still used, although two additional “orders,” known as “small third order” and “hyper-radial” respectively are in ordinary use. The followingtable gives the principal dimensions of the several sizes in use:—Table II.Order.FocalDistance,mm.Vertical Angles of Optics.(Ordinary Dimensions.)DioptricBelt only.Holophotal Optics.LowerPrisms.Lens.UpperPrisms.Hyper-Radial133080°21°57°48°1st order92092°, 80°, 58°21°57°48°2nd order70080°21°57°48°3rd order50080°21°57°48°Small 3rd order37580°21°57°48°4th order25080°21°57°48°5th order187.580°21°57°48°6th order15080°21°57°48°Lenses of small focal distance are also made for buoy and beacon lights.Fig. 42.—Cape Naturaliste Apparatus.Fig. 43.—Île Vierge Apparatus.Light Intensities.—The powers of lighthouse lights in the British Empire are expressed in terms of standard candles or in “lighthouse units” (one lighthouse unit = 1000 standard candles). In France the unit is the “Carcel” = .952 standard candle. The powers of burners and optical apparatus, then in use in the United Kingdom, were carefully determined by actual photometric measurement in 1892 by a committee consisting of the engineers of the three general lighthouse boards, and the values so obtained are used as the basis for calculating the intensities of all British lights. It was found that the intensities determined by photometric measurement were considerably less than the values given by the theoretical calculations formerly employed. A deduction of 20% was made from the mean experimental results obtained to compensate for loss by absorption in the lantern glass, variations in effects obtained by different men in working the burners and in the illuminating quality of oils, &c. The resulting reduced values are termed “service” intensities.As has been explained above, the effect of a dioptric apparatus is to condense the light rays, and the measure of this condensation is the ratio between the vertical divergence and the vertical angle of the optic in the case of fixed lights. In flashing lights the ratio of vertical condensation must be multiplied by the ratio between the horizontal divergence and the horizontal angle of the panel. The loss of light by absorption in passing through the glass and by refraction varies from 10% to 15%. For apparatus containing catadioptric elements a larger deduction must be made.The intensity of the flash emitted from a dioptric apparatus, showing a white light, may be found approximately by the empirical formula I = PCVH/vh, where I = intensity of resultant beam, P = service intensity of flame, V = vertical angle of optic,v= angle of mean vertical divergence, H = horizontal angle of panel,h= angleof mean horizontal divergence, and C = constant varying between .9 and .75 according to the description of apparatus. The factor H/hmust be eliminated in the case of fixed lights. Deduction must also be made in the case of coloured lights. It should, however, be pointed out that photometric measurements alone can be relied upon to give accurate values for lighthouse intensities. The values obtained by the use of Allard’s formulae, which were largely used before the necessity for actual photometric measurements came to be appreciated, are considerably in excess of the true intensities.Fig. 43a.—Île Vierge Apparatus and Lantern. Plan at focal plane.Optical Calculations.—The mathematical theory of optical apparatus for lighthouses and formulae for the calculations of profiles will be found in the works of the Stevensons, Chance, Allard, Reynaud, Ribière and others. Particulars of typical lighthouse apparatus will be found in tables VI. and VII.
Catoptric System.—Parabolic reflectors, consisting of small facets of silvered glass set in plaster of Paris, were first used about the year 1763 in some of the Mersey lights by Mr Hutchinson, then dock master at Liverpool (fig. 29). Spherical metallic reflectors were introduced in France in 1781, followed by parabolic reflectors on silvered copper in 1790 in England and France, and in Scotland in 1803. The earlier lights were of fixed type, a number of reflectors being arranged on a frame or stand in such a manner that the pencils of emergent rays overlapped and thus illuminated the whole horizon continuously. In 1783 the first revolving light was erected at Marstrand in Sweden. Similar apparatus were installed at Cordouan (1790), Flamborough Head (1806) and at the Bell Rock (1811). To produce a revolving or flashing light the reflectors were fixed on a revolving carriage having several faces. Three or more reflectors in a face were set with their axes parallel.
A type of parabolic reflector now in use is shown in fig. 30. The sizes in general use vary from 21 in. to 24 in. diameter. These instruments are still largely used for light-vessel illumination, and a few important land lights are at the present time of catoptric type, including those at St Agnes (Scilly Islands), Cromer and St Anthony (Falmouth).
Dioptric System.—The first adaptation of dioptric lenses to lighthouses is probably due to T. Rogers, who used lenses at one of the Portland lighthouses between 1786 and 1790. Subsequently lenses by the same maker were used at Howth, Waterford and the North Foreland. Count Buffon had in 1748 proposed to grind out of a solid piece of glass a lens in steps or concentric zones in order to reduce the thickness to a minimum (fig. 31). Condorcet in 1773 and Sir D. Brewster in 1811 designed built-up lenses consisting of stepped annular rings. Neither of these proposals, however, was intended to apply to lighthouse purposes. In 1822 Augustin Fresnel constructed a built-up annular lens in which the centres of curvature of the different rings receded from the axis according to their distances from the centre, so as practically to eliminate spherical aberration; the only spherical surface being the small central part or “bull’s eye” (fig. 32). These lenses were intended for revolving lights only. Fresnel next produced his cylindric refractor or lens belt, consistingof a zone of glass generated by the revolution round a vertical axis of a medial section of the annular lens (fig. 33). The lens belt condensed and parallelized the light rays in the vertical plane only, while the annular lens does so in every plane. The first revolving light constructed from Fresnel’s designs was erected at the Cordouan lighthouse in 1823. It consisted of 8 panels of annular lenses placed round the lamp at a focal distance of 920 mm. To utilize the light, which would otherwise escape above the lenses, Fresnel introduced a series of 8 plain silvered mirrors, on which the light was thrown by a system of lenses. At a subsequent period mirrors were also placed in the lower part of the optic. The apparatus was revolved by clockwork. This optic embodied the first combination of dioptric and catoptric elements in one design (fig. 34). In the following year Fresnel designed a dioptric lens with catoptric mirrors for fixed light, which was the first of its kind installed in a lighthouse. It was erected at the Chassiron lighthouse in 1827 (fig. 35). This combination is geometrically perfect, but not so practically on account of the great loss of light entailed by metallic reflection which is at least 25% greater than the system described under. Before his death in 1827 Fresnel devised his totally reflecting or catadioptric prisms to take the place of the silvered reflectors previously used above and below the lens elements (fig. 28). The ray Fi falling on the prismoidal ring ABC is refracted in the directioni rand meeting the face AB at an angle of incidence greater than the critical, is totally reflected in the directionr eemerging after second refraction in a horizontal direction. Fresnel devised these prisms for use in fixed light apparatus, but the principle was, at a later date, also applied to flashing lights, in the first instance by T. Stevenson. Both the dioptric lens and catadioptric prism invented by Fresnel are still in general use, the mathematical calculations of the great French designer still forming the basis upon which lighthouse opticians work.
Fresnel also designed a form of fixed and flashing light in which the distinction of a fixed light, varied by flashes, was produced by placing panels of straight refracting prisms in a vertical position on a revolving carriage outside the fixed light apparatus. The revolution of the upright prisms periodically increased the power of the beam, by condensation of the rays emergent from the fixed apparatus, in the horizontal plane.
The lens segments in Fresnel’s early apparatus were of polygonal form instead of cylindrical, but subsequently manufacturers succeeded in grinding glass in cylindrical rings of the form now used. The first apparatus of this description was made by Messrs Cookson of Newcastle in 1836 at the suggestion of Alan Stevenson and erected at Inchkeith.
In 1825 the French Commission des Phares decided upon the exclusive use of lenticular apparatus in its service. The Scottish Lighthouse Board followed with the Inchkeith revolving apparatus in 1835 and the Isle of May fixed optic in 1836. In the latter instrument Alan Stevenson introduced helical frames for holding the glass prisms in place, thus avoiding complete obstruction of the light rays in any azimuth. The first dioptric light erected by the Trinity House was that formerly at Start Point in Devonshire, constructed in 1836. Catadioptric or reflecting prisms for revolving lights were not used until 1850, when Alan Stevenson designed them for the North Ronaldshay lighthouse.
Dioptric Mirror.—The next important improvement in lighthouse optical work was the invention of the dioptric spherical mirror by Mr (afterwards Sir) J. T. Chance in 1862. The zones or prisms are generated round a vertical axis and divided into segments. This form of mirror is still in general use (figs. 36 and 37).
Azimuthal Condensing Prisms.—Previous to 1850 all apparatus were designed to emit light of equal power in every azimuth either constantly or periodically. The only exception was where a light was situated on a stretch of coast where a mirror could be placed behind the flame to utilize the rays, which would otherwise pass landward, and reflect them back, passing through the flame and lens in a seaward direction. In order to increase the intensity of lights in certain azimuths T. Stevenson devised his azimuthal condensing prisms which, in various forms and methods of application, have been largely used for the purpose of strengthening the light rays in required directions as, for instance, where coloured sectors are provided. Applications of this system will be referred to subsequently.
Optical Glass for Lighthouses.—In the early days of lens lights the only glass used for the prisms was made in France at the St Gobain and Premontré works, which have long been celebrated for the high quality of optical glass produced. The early dioptric lights erected in the United Kingdom, some 13 in all, were made by Messrs Cookson of South Shields, who were instructed by Léonor Fresnel, the brother of Augustin. At first they tried to mould the lens and then to grind it out of one thick sheet of glass. The successors of the Cookson firm abandoned the manufacture of lenses in 1845, and the firm of Letourneau & Lepaute of Paris again became the monopolists. In 1850 Messrs Chance Bros. & Co. of Birmingham began the manufacture of optical glass, assisted by M. Tabouret, a French expert who had been a colleague of Augustin Fresnel himself. The first light made by the firm was shown at the Great Exhibition of 1851, since when numerous dioptric apparatus have been constructed by Messrs Chance, who are, at this time, the only manufacturers of lighthouse glass in the United Kingdom. Most of the glass used for apparatus constructed in France is manufactured at St Gobain. Some of the glass used by German constructors is made at Rathenow in Prussia and Goslar in the Harz.
The glass generally employed for lighthouse optics has for its refractive index a mean value of µ = 1.51, the corresponding critical angle being 41° 30′. Messrs Chance have used dense flint glass for the upper and lower refracting rings of high angle lenses and for dioptric mirrors in certain cases. This glass has a value of µ = l.62 with critical angle 38° 5′.
Occulting Lights.—During the last 25 years of the 19th century the disadvantages of fixed lights became more and more apparent. At the present day the practice of installing such, except occasionally in the case of the smaller and less important of harbour or river lights, has practically ceased. The necessity for providing a distinctive characteristic for every light when possible has led to the conversion of many of the fixed-light apparatus of earlier years into occulting lights, and often to their supersession by more modern and powerful flashing apparatus. An occulting apparatus in general use consists of a cylindrical screen, fitting over the burner, rapidly lowered and raised by means of a cam-wheel at stated intervals. The cam-wheel is actuated by means of a weight or spring clock. Varying characteristics may be procured by means of such a contrivance—single, double, triple or other systems of occultation. The eclipses or periods of darkness bear much the same relation to the times of illumination as do the flashes to the eclipses in a revolving or flashing light. In the case of a first-order fixed light the cost of conversion to an occulting characteristic does not exceed £250 to £300. With apparatus illuminated by gas the occultations may be produced by successively raising and lowering the gas at stated intervals. Another form of occulting mechanism employed consists of a series of vertical screens mounted on a carriage and revolving round the burner. The carriage is rotated on rollers or ball bearings or carried upon a small mercury float. The usual driving mechanism employed is a spring clock. “Otter” screens are used in cases when it is desired to produce different periods of occultations in two or more positions in azimuth in order to differentiate sectors marking shoals, &c. The screens are of sheet metal blacked and arranged vertically, some what in the manner of the laths of a venetian blind, and operated by mechanical means.
Leading Lights.—In the case of lights designed to act as a lead through a narrow channel or as direction lights, it is undesirable to employ a flashing apparatus. Fixed-light optics are employed to meet such cases, and are generally fitted with occulting mechanism. A typical apparatus of this description is that at Gage Roads, Fremantle, West Australia (fig. 38). The occulting bright light covers the fairway, and is flanked by sectors of occulting red and green light marking dangers and intensified by vertical condensing prisms. A good example of a holophotal direction light was exhibited at the 1900 Paris Exhibition, and afterwards erected at Suzac lighthouse (France). The light consists of an annular lens 500 mm. focal distance, of 180° horizontal angle and 157° vertical, with a mirror of 180° at the back. The lens throws a red beam of about 4½° amplitude in azimuth, and 50,000 candle-power over a narrow channel. The illuminant is an incandescent petroleum vapour burner. Holophotal direction lenses of this type can only be applied where the sector to be marked is of comparatively small angle. Silvered metallic mirrors of parabolic form are also used for the purpose. The use of single direction lights frequently renders the construction of separate towers for leading lights unnecessary.
If two distinct lights are employed to indicate the line of navigation through a channel or between dangers they must be sufficiently far apart to afford a good lead, the front or seaward light being situated at a lower elevation than the rear or landward one.
Coloured Lights.—Colour is used as seldom as possible as a distinction, entailing as it does a considerable reduction in the power of the light. It is necessary in some instances for differentiating sectors over dangers and for harbour lighting purposes. The use of coloured lights as alternating flashes for lighthouse lights is not to be commended, on account of the unequal absorption of the colouredand bright rays by the atmosphere. When such distinction has been employed, as in the Wolf Rock apparatus, the red and white beams can be approximately equalized in initial intensity by constructing the lens and prism panels for the red light of larger angle than those for the white beams. Owing to the absorption by the red colouring, the power of a red beam is only 40% of the intensity of the corresponding white light. The corresponding intensity of green light is 25%. When red or green sectors are employed they should invariably be reinforced by mirrors, azimuthal condensing prisms, or other means to raise the coloured beam to approximately the same intensity as the white light. With the introduction of group-flashing characteristics the necessity for using colour as a means of distinction disappeared.
High-Angle Vertical Lenses.—Messrs Chance of Birmingham have manufactured lenses having 97° of vertical amplitude, but this result was only attained by using dense flint glass of high refractive index for the upper and lower elements. It is doubtful, however, whether the use of refracting elements for a greater angle than 80° vertically is attended by any material corresponding advantage.
Group Flashing Lights.—One of the most useful distinctions consists in the grouping of two or more flashes separated by short intervals of darkness, the group being succeeded by a longer eclipse. Thus two, three or more flashes of, say, half second duration or less follow each other at intervals of about 2 seconds and are succeeded by an eclipse of, say, 10 seconds, the sequence being completed in a period of, say, 15 seconds. In 1874 Dr John Hopkinson introduced the very valuable improvement of dividing the lenses of a dioptric revolving light with the panels of reflecting prisms above and below them, setting them at an angle to produce the group-flashing characteristic. The first apparatus of this type constructed were those now in use at Tampico, Mexico and the Little Basses lighthouse, Ceylon (double flashing). The Casquets apparatus (triple flashing) was installed in 1877. A group-flashing catoptric light had, however, been exhibited from the “Royal Sovereign” light-vessel in 1875. A sectional plan of the quadruple-flashing first order apparatus at Pendeen in Cornwall is shown in fig. 39; and fig. 55 (Plate 1.) illustrates a double flashing first order light at Pachena Point in British Columbia. Hopkinson’s system has been very extensively used, most of the group-flashing lights shown in the accompanying tables, being designed upon the general lines he introduced. A modification of the system consists in grouping two or more lenses together separated by equal angles, and filling the remaining angle in azimuth by a reinforcing mirror or screen. A group-flashing distinction was proposed for gas lights by J. R. Wigham of Dublin, who obtained it in the case of a revolving apparatus by alternately raising and lowering the flame. The first apparatus in which this method was employed was erected at Galley Head, Co. Cork (1878). At this lighthouse 4 of Wigham’s large gas burners with four tiers of first-order revolving lenses, eight in each tier, were adopted. By successive lowering and raising of the gas flame at the focus of each tier of lenses he produced the group-flashing distinction. The light showed, instead of one prolonged flash at intervals of one minute, as would be produced by the apparatus in the absence of a gas occulter, a group of short flashes varying in number between six and seven. The uncertainty, however, in the number of flashes contained in each group is found to be an objection to the arrangement. This device was adopted at other gas-illuminated stations in Ireland at subsequent dates. The quadriform apparatus and gas installation at Galley Head were superseded in 1907 by a first order bi-form apparatus with incandescent oil vapour burner showing five flashes every 20 seconds.
Flashing Lights indicating Numbers.—Captain F. A. Mahan, late engineer secretary to the United States Lighthouse Board, devised for that service a system of flashing lights to indicate certain numbers. The apparatus installed at Minot’s Ledge lighthouse near Boston Harbour, Massachusetts, has a flash indicating the number 143, thus: - ---- ---, the dashes indicating short flashes. Each group is separated by a longer period of darkness than that between successive members of a group. The flashes in a group indicating a figure are about 1½ seconds apart, the groups being 3 seconds apart, an interval of 16 seconds’ darkness occurring between each repetition. Thus the number is repeated every half minute. Two examples of this system were exhibited by the United States Lighthouse Board at the Chicago Exhibition in 1893, viz. the second-order apparatus just mentioned and a similar light of the first order for Cape Charles on the Virginian coast. The lenses are arranged in a somewhatsimilar manner to an ordinary group-flashing light, the groups of lenses being placed on one side of the optic, while the other is provided with a catadioptric mirror. This system of numerical flashing for lighthouses has been frequently proposed in various forms, notably by Lord Kelvin. The installation of the lights described is, however, the first practical application of the system to large and important coast lights. The great cost involved in the alteration of the lights of any country to comply with the requirements of a numerical system is one of the objections to its general adoption.
Plate I.
Plate II.
Hyper-radial Apparatus.—In 1885 Messrs Barbier of Paris constructed the first hyper-radial apparatus (1330 mm. focal distance) to the design of Messrs D. and C. Stevenson. This had a height of 1812 mm. It was tested during the South Foreland experiments in comparison with other lenses, and found to give excellent results with burners of large focal diameter. Apparatus of similar focal distance (1330 mm.) were subsequently established at Round Island, Bishop Rock, and Spurn Point in England, Fair Isle and Sule Skerry (fig. 40) in Scotland, Bull Rock and Tory Island in Ireland, Cape d’Antifer in France, Pei Yu-shan in China and a lighthouse in Brazil.
The light erected in 1907 at Cape Race, Newfoundland, is a fine example of a four-sided hyper-radial apparatus mounted on a mercury float. The total weight of the revolving part of the light amounts to 7 tons, while the motive clock weight required to rotate this large mass at a speed of two complete revolutions a minute is only 8 cwt. and the weight of mercury required for flotation 950 ℔. A similar apparatus was placed at Manora Point, Karachi, India, in 1908 (fig. 41).
The introduction of incandescent and other burners of focal compactness and high intensity has rendered the use of optics of such large dimensions as the above, intended for burners of great focal diameter, unnecessary. It is now possible to obtain with a second-order optic (or one of 700 mm. focal distance), having a powerful incandescent petroleum burner in focus, a beam of equal intensity to that which would be obtained from the apparatus having a 10-wick oil burner or 108-jet gas burner at its focus.
Stephenson’s Spherical Lenses and Equiangular Prisms.—Mr C. A. Stephenson in 1888 designed a form of lens spherical in the horizontal and vertical sections. This admitted of the construction of lenses of long focal distance without the otherwise corresponding necessity of increased diameter of lantern. A lens of this type and of 1330 mm. focal distance was constructed in 1890 for Fair Isle lighthouse. The spherical form loses in efficiency if carried beyond an angle subtending 20° at the focus, and to obviate this loss Mr Stephenson designed his equiangular prisms, which have an inclination outwards. It is claimed by the designer that the use of equiangular prisms results in less loss of light and less divergence than is the case when either the spherical or Fresnel form is adopted. An example of this design is seen (fig. 40) in the Sule Skerry apparatus (1895).
Fixed and Flashing Lights.—The use of these lights, which show a fixed beam varied at intervals by more powerful flashes, is not to be recommended, though a large number were constructed in the earlier years of dioptric illumination and many are still in existence. The distinction can be produced in one or other of three ways: (a) by the revolution of detached panels of straight condensing lens prisms placed vertically around a fixed light optic, (b) by utilizing revolving lens panels in the middle portion of the optic to produce the flashing light, the upper and lower sections of the apparatus being fixed zones of catadioptric or reflecting elements emitting a fixed belt of light, and (c) by interposing panels of fixed light section between the flashing light panels of a revolving apparatus. In certain conditions of the atmosphere it is possible for the fixed light of low power to be entirely obscured while the flashes are visible, thus vitiating the true characteristic of the light. Cases have frequently occurred of such lights being mistaken for, and even described in lists of light as, revolving or flashing lights.
”Cute” and Screens.—Screens of coloured glass, intended to distinguish the light in particular azimuths, and of sheet iron, when it is desired to “cut off” the light sharply on any angle, should be fixed as far from the centre of the light as possible in order to reduce the escape of light rays due to divergence. These screens are usually attached to the lantern framing.
Divergence.—A dioptric apparatus designed to bend all incident rays of light from the light source in a horizontal direction would, if the flame could be a point, have the effect of projecting a horizontal band or zone of light, in the case of a fixed apparatus, and a cylinder of light rays, in the case of a flashing light, towards the horizon. Thus the mariner in the near distance would receive no light, the rays, visible only at or near the horizon, passing above the level of his eye. In practice this does not occur, sufficient natural divergence being produced ordinarily owing to the magnitude of the flame. Where the electric arc is employed it is often necessary to design the prisms so as to produce artificial divergence. The measure of the natural divergence for any point of the lens is the angle whose sine is the ratio of the diameter of the flame to the distance of the point from centre of flame.
In the case of vertical divergence the mean height of the flame must be substituted for the diameter. The angle thus obtained is the total divergence, that is, the sum of the angles above and below the horizontal plane or to right and left of the medial section. In fixed dioptric lights there is, of course, no divergence in the horizontal plane. In flashing lights the horizontal divergence is a matter of considerable importance, determining as it does the duration or length of time the flash is visible to the mariner.
Feux-Éclairs or Quick Flashing Lights.—One of the most important developments in the character of lighthouse illuminating apparatus that has occurred in recent years has been in the direction of reducing the length of flash. The initiative in this matter was taken by the French lighthouse authorities, and in France alone forty lights of this type were established between 1892 and 1901. The use of short flash lights rapidly spread to other parts of the world. In England the lighthouse at Pendeen (1900) exhibits a quadruple flash every 15 seconds, the flashes being about ¼ second duration (fig. 39), while the bivalve apparatus erected on Lundy Island (1897) shows 2 flashes of1⁄3second duration in quick succession every 20 seconds. Since 1900 many quick flashing lights have been erected on the coasts of the United Kingdom and in other countries. The earlyfeux-éclairs, designed by the French engineers and others, had usually a flash of1⁄10th to1⁄3rd of a second duration. As a result of experiments carried out in France in 1903-1904,3⁄10second has been adopted by the French authorities as the minimum duration for white flashing lights. If shorter flashes are used it is found that the reduction in duration is attended by a corresponding, but not proportionate, diminution in effective intensity. In the case of many electric flashing lights the duration is of necessity reduced, but the greater initial intensity of the flash permits this loss without serious detriment to efficiency. Red or green requires a considerably greater duration than do white flashes. The intervals between the flashes in lights of this character are also small, 2½ seconds to 7 seconds. In group-flashing lights the intervals between the flashes are about 2 seconds or even less, with periods of 7 to 10 or 15 seconds between the groups. The flashes are arranged in single, double, triple or even quadruple groups, as in the older forms of apparatus. Thefeu-éclairtype of apparatus enables a far higher intensity of flash to be obtained than was previously possible without any corresponding increase in the luminous power of the burner or other source of light. This result depends entirely upon the greater ratio of condensation of light employed, panels of greater angular breadth than was customary in the older forms of apparatus being used with a higher rotatory velocity. It has been urged that short flashes are insufficient for taking bearings, but the utility of a light in this respect does not seem to depend so much upon the actual length of the flash as upon its frequent recurrence at short intervals. At the Paris Exhibition of 1900 was exhibited a fifth-order flashing light giving short flashes at 1 second intervals; this represents the extreme to which the movement towards the reduction of the period of flashing lights has yet been carried.
Mercury Floats.—It has naturally been found impracticable to revolve the optical apparatus of a light with its mountings, sometimes weighing over 7 tons, at the high rate of speed required forfeux-éclairsby means of the old system of roller carriages, though for some small quick-revolving lights ball bearings have been successfully adopted. It has therefore become almost the universal practice to carry the rotating portions of the apparatus upon a mercury float. This beautiful application of mercury rotation was the invention of Bourdelles, and is now utilized not only for the high-speed apparatus, but also generally for the few examples of the older type still being constructed. The arrangement consists of an annular cast iron bath or trough of such dimensions that a similar but slightly smaller annular float immersed in the bath and surrounded by mercury displaces a volume of the liquid metal whose weight is equal to that of the apparatus supported. Thus a comparatively insignificant quantity of mercury, say 2 cwt., serves to ensure the flotation of a mass of over 3 tons. Certain differences exist between the type of float usually constructed in France and those generally designed by English engineers. In all cases provision is made for lowering the mercury bath or raising the float and apparatus for examination. Examples of mercury floats are shown in figs. 41, 42, 43 and Plate I., figs. 54and55.
Multiform Apparatus.—In order to double the power to be obtained from a single apparatus at stations where lights of exceptionally high intensity are desired, the expedient of placing one complete lens apparatus above another has sometimes been adopted, as at the Bishop Rock (fig. 13), and at the Fastnet lighthouse in Ireland (Plate I., fig. 54). Triform and quadriform apparatus have also been erected in Ireland; particulars of the Tory Island triform apparatus will be found in table VII. The adoption of the multiform system involves the use of lanterns of increased height.
Twin Apparatus.—Another method of doubling the power of a light is by mounting two complete and distinct optics side by side on the same revolving table, as I shown in fig. 43 of the Île Vierge apparatus. Several such lights have been installed by the French Lighthouse Service.
Port Lights.—Small self-contained lanterns and lights are in common use for marking the entrances to harbours and in other similar positions where neither high power nor long range is requisite. Many such lights are unattended in the sense that they do not require the attention of a keeper for days and even weeks together. These are described in more detail in section 6 of this article. A typical port light consists of a copper or brass lantern containing a lens of the fourth order (250 mm. focal distance) or smaller, and a single wick or 2-wick Argand capillary burner. Duplex burners are also used. The apparatus may exhibit a fixed light or, more usually, an occulting characteristic is produced by the revolution of screens actuated by spring clockwork around the burner. The lantern may be placed at the top of a column, or suspended from the head of a mast. Coal gas and electricity are also used as illuminants for port lights when local supplies are available. The optical apparatus used in connexion with electric light is described below.
”Orders” of Apparatus.—Augustin Fresnel divided the dioptric lenses, designed by him, into “orders” or sizes depending on their local distance. This division is still used, although two additional “orders,” known as “small third order” and “hyper-radial” respectively are in ordinary use. The followingtable gives the principal dimensions of the several sizes in use:—
Table II.
Lenses of small focal distance are also made for buoy and beacon lights.
Light Intensities.—The powers of lighthouse lights in the British Empire are expressed in terms of standard candles or in “lighthouse units” (one lighthouse unit = 1000 standard candles). In France the unit is the “Carcel” = .952 standard candle. The powers of burners and optical apparatus, then in use in the United Kingdom, were carefully determined by actual photometric measurement in 1892 by a committee consisting of the engineers of the three general lighthouse boards, and the values so obtained are used as the basis for calculating the intensities of all British lights. It was found that the intensities determined by photometric measurement were considerably less than the values given by the theoretical calculations formerly employed. A deduction of 20% was made from the mean experimental results obtained to compensate for loss by absorption in the lantern glass, variations in effects obtained by different men in working the burners and in the illuminating quality of oils, &c. The resulting reduced values are termed “service” intensities.
As has been explained above, the effect of a dioptric apparatus is to condense the light rays, and the measure of this condensation is the ratio between the vertical divergence and the vertical angle of the optic in the case of fixed lights. In flashing lights the ratio of vertical condensation must be multiplied by the ratio between the horizontal divergence and the horizontal angle of the panel. The loss of light by absorption in passing through the glass and by refraction varies from 10% to 15%. For apparatus containing catadioptric elements a larger deduction must be made.
The intensity of the flash emitted from a dioptric apparatus, showing a white light, may be found approximately by the empirical formula I = PCVH/vh, where I = intensity of resultant beam, P = service intensity of flame, V = vertical angle of optic,v= angle of mean vertical divergence, H = horizontal angle of panel,h= angleof mean horizontal divergence, and C = constant varying between .9 and .75 according to the description of apparatus. The factor H/hmust be eliminated in the case of fixed lights. Deduction must also be made in the case of coloured lights. It should, however, be pointed out that photometric measurements alone can be relied upon to give accurate values for lighthouse intensities. The values obtained by the use of Allard’s formulae, which were largely used before the necessity for actual photometric measurements came to be appreciated, are considerably in excess of the true intensities.
Optical Calculations.—The mathematical theory of optical apparatus for lighthouses and formulae for the calculations of profiles will be found in the works of the Stevensons, Chance, Allard, Reynaud, Ribière and others. Particulars of typical lighthouse apparatus will be found in tables VI. and VII.
4.Illuminants.—The earliest form of illuminant used for lighthouses was a fire of coal or wood set in a brazier or grate erected on top of the lighthouse tower. Until the end of the 18th and even into the 19th century this primitive illuminant continued to be almost the only one in use. The coal fire at the Isle of May light continued until 1810 and that at St Bees lighthouse in Cumberland till 1823. Fires are stated to have been used on the two towers of Nidingen, in the Kattegat, until 1846. Smeaton was the first to use any form of illuminant other than coal fires; he placed within the lantern of his Eddystone lighthouse a chandelier holding 24 tallow candles each of which weighed2⁄5of a ℔ and emitted a light of 2.8 candle power. The aggregate illuminating power was 67.2 candles and the consumption at the rate of 3.4 ℔ per hour.