284 



NA TURE 



[July 22, 1897 



Tory Island light, fitted with these burners, and variable in power 

 from 17,500 to 326,500 candles, may be cited in illustration. 



(3) The maximum intensity of an illuminant must still be 

 sought in the electric arc. Gas and oil remain substantially 

 equal compared in lamps of the same size and sort, the superior 

 applicability of each being determined by local conditions. In 

 Ireland the Wigham expanding burners give a marked promi- 

 nence to the gas, which is copiously used in them. In England 

 and Scotland mineral oil is preferred, its quality being carefully 

 maintained at the highest standard. The old disability of this 

 illuminant, the risk of explosion, has been almost nullified by 

 the production of a petroleum whose flashing point is 230°, and 

 long experience has confirmed its great value. On the other hand, 

 the gas called "incandescent" has been introduced, the best 

 form of it being of the Auer-Welsbach type. The brilliancy of this 

 gas is perhaps only second to that of the electric arc, but the 

 perishable character of its accessories exacts great caution in using 

 it, and it has not as yet been employed in any sea light in this 

 country, although it is already so adopted in France. A mixture 

 of oxygen and coal gas, and one of oxygen and oil vapour, which 

 have been tried in street lamps and otherwise, have been recom- 

 mended for lighthouses, but not hitherto accepted. 



Electricity has been, as it were, on its trial since the South 

 Foreland experiments of 1885, and the evidence affecting it is 

 hardly yet complete enough to justify a final verdict. The 

 result established at the South Foreland was that the electric 

 light is the most powerful under all conditions of weather, and 

 has the greatest penetrative intensity in fog. The Committee of 

 the Royal Society, which examined in 1890 this report of the 

 Trinity House, found that the experiments did really justify the 

 results given. 



The twelve years that have elapsed since the trials at the 

 South Foreland have on the whole tended to qualify the conclu- 

 sion as to the penetrative power in fog. Three lights — two of 

 the group-flashing, and one of the single-flashing character — 

 constructed by Messrs. Chance, may be cited on this point. The 

 triple flash of the Tino light (near Spezia) has been unmistak- 

 ably discerned in rain and fog at a distance of more than twenty 

 miles. On the other hand, the Isle of May light has been in- 

 visible in a thick atmosphere not amounting to fog, at a distance 

 of twelve miles, and in a dense fog at half a mile ; and the St. 

 Catherine's light was equally invisible to the Eider before she 

 grounded on the Atherfield ledges. It has been, indeed, 

 asserted that the St, Catherine's has often been unseen at a 

 quarter of a mile distance. 



The truth appears to be that the electric light is very 

 sensitive to atmospheric conditions, which are so many and 

 so various, and that in thick weather it parts with its power in 

 a much greater ratio than does a gas light, or even an oil 

 light. There is a degree of fog which quenches the sun, while 

 the large luminous surfaces of the superposed gas lenses project 

 on the fog a reddish colouration, and the fog itself thus becomes 

 a signal to the mariner when, as it were, in the words of Persius, 

 " Pinguem nebulam vomuere lucernae." It would be most 

 dangerous, however, for the mariner in a fog to approach the 

 coast presuming on this quality of a light in whose vicinity he 

 supposed himself to be. The lead, the anchor, the horn should 

 be his trust till the veil lifts, and the electric beam shines out in 

 full splendour. 



There has occurred no important change in the burners of 

 sea lights, The Trinity House improved six-wick, or rather 

 five- wick, remains, if not the largest, the best working oil 

 burner, with a power of 800 candles. This is used generally for 

 lights of the first and second order, while the four-wick, with a 

 power of 360, is used for third order lights, and the eight-wick, 

 of 1200 candles, for hyper-iadial lights. The lamp is of the 

 " pump" or the "pressure" type, and contains from 10 to 100 

 gallons of oil. It is probable that, owing to considerations of 

 space and economy, the gravity system may again be resorted to 

 with enlarged reservoirs seated on the lantern-roof. The electric 

 arc is used with carbons of from 15 mm. to 65 mm. diameter, and 

 currents of from 50 to 400 amperes. The incandescent filament 

 is not found equally appropriate. The luminous intensities ob- 

 tained from the arc range from five to fifty thousand candles. 

 Both direct and alternating machines are employed. The in- 

 stallations at St. Catherine's and the Isle of May are of the most 

 complete character, and do honour to the distinguished engineers 

 of the Trinity House and the Northern Commissioners. 



It may be mentioned that for positions difficult of access, 

 lamps having special wicks and reservoirs for oil or gas, capable 



NO. 1447, VOL. 56] 



of burning from ten to thirty days, are now in use, though, of 

 course, these are only available for small isolated lights. 



(4) The improvements since 1887 in dioptric lights are few, 

 and, with one exception, are of no striking importance. The 

 great invention of Fresnel, perfected by the beautiful holophotal 

 arrangements of Stevenson, has remained the cardinal principle 

 of all modern lights of the lenticular type. But the inevitable 

 tendency to modify and improve has resulted in several pro- 

 posals of more or less merit. Mr. Alan Brebner had, in 1882, 

 submitted an ingenious plan for producing vertical and azi- 

 muthal condensation by single agency in cases where straight 

 prisms placed outside the main apparatus had been employed 

 to intensify one or more sea sectors. This method, however, 

 has been seldom, , if at all, adopted. Again, Mr. Brebner, in 

 1884, had recommended a plan of dipping a portion of the 

 beam to some intermediate distance between the lighthouse and 

 the horizon to meet the case of a fog which the strongest beam 

 can only very partially penetrate. This; or some analogous 

 plan, has been tried, but has not prevailed. The advantage 

 of withholding from the horizon a substantial part of the beam, 

 and deflecting it anywhere, is extremely doubtful. The depth 

 and direction of a fog are always uncertain elements. The 

 lighthouse may be surrounded by it while the ofiing is clear, 

 or vice versd. And it would be in equal measure mischievous 

 to encourage the mariner to stand in, looking for a signal 

 which he might never see, or see too late ; and to entrust the 

 lightkeeper with the power to deflect the light according to his 

 own judgment. In 1892, Mr. Charles A. Stevenson published 

 in Nature an account of his spherical and equiangular re- 

 fractors which remedy a certain loss of emergent light in the 

 Fresnel refractors. In 1894, Mr. Stevenson further developed 

 his design, claiming for it, within practicable limits, an ad- 

 vantage of at least 10 per cent, in increase of light as com- 

 pared with the Fresnel lens. The improved refractors have 

 been adopted with success in several of the Scottish light- 

 houses. In 1895 and 1896, Mr. John A. Purves contributed 

 an able matheimatical analysis of equiangular prisms and a new 

 form of spherical central lens, called the Inverse Refractor (the 

 facets of the lens turned inwards), to be used in connection 

 with Mr. C. Stevenson's equiangular prisms. Messrs. Chance 

 have raised the Fresnel lens from 57° vertical to 80°, using glass 

 of the same refractive index, and this angle has been adopted 

 by the Trinity House with great advantage. 



(5) The one exception referred to in regard to the minor 

 improvements of optical apparatus since 1887 is the " lightning 

 light," an adaptation of serious importance v\hich has attracted 

 the attention of lighthouse engineers in nearly all maritime 

 countries. The shortening of the interval between the flashes 

 of a revolving light, so that the mariner, especially when in a 

 fast liner, may have more speedy cognisance of the signal that 

 guides him, has become a plain necessity, and led to the gradual 

 reduction of the period from 60, 45 and 30 seconds to 20, 15, 

 10, and even 5 seconds, reducing proportionately the duration 

 of the flash. So far the principle of the feu eclair has been 

 approached. But much more than tHis is demanded. No one 

 has advocated more strenuously than Lord Kelvin, himself a 

 sailor and a benefactor of sailors, the acceleration of revolving 

 lights. No one has laid it down more clearly that one-ninth or 

 one-tenth of a second is sufficient for the eye to receive the full 

 lustre of a lens passing it. Any longer duration is a loss of 

 time, and therefore of intensity. German and French physi- 

 ologists have confirmed this, and it is the principle on which 

 they^« ilclair system is based. In its application, then, if for 

 any "order" of light the maximum of power be desired, the 

 largest available lens in vertical and horizontal angle, combined 

 with a totally reflecting mirror and a central flame of adequate 

 dimensions, must be adopted. If two or more lenses be used, 

 singly or grouped, the intensity of each flash is proportionately 

 less, the interval of time, say five seconds, between the single 

 flashes or the groups of flashes, being the same as where there 

 is only one lens. In this manner the greatest intensity and the 

 shortest interval are secured, and the characteristic as pre- 

 sented to the mariner seems theoretically perfect. The first 

 general introduction of the " lightning light " is due to M. 

 Bourdelles, the chief of the French Lighthouse Administration, 

 under whose auspices several lighthouses on the coasts of 

 France have already been endowed with it. In England 

 Messrs. Chance have constructed, in 1895, fo"^ Cape Leeuwin, 

 Western Australia, a " bivalve " apparatus of the first order, 

 with an experimental duration of beam of one-fifth of a second, 



