A feature of the past few years has been the rapid development in lamp manufacture in the United States. In 1920 the production was estimated to reach 230 million lamps, of which only 7 % were of the carbon filament type (Gen. Elec. Review, U.S.A., Jan. 1921). Considerable progress in the manufacture of miniature lamps for automobiles, flashlights, miners' lamps, etc., is recorded, an output of 125 million being attained in 1920. Progress in lamp manufacture has 'been aided by success in standardizing supply voltages, nearly 79% of the lamps sold in 1920 being for the standard pressures of no, 115 and 120 volts. In Japan a uniform pressure of 100 volts throughout the country has been established.
Physical Data Underlying the Efficiency of Light Production. Researches in the physics of light production have yielded interesting conclusions, revealing the comparative inefficiency of most artificial illuminants. Thus it is estimated that the ordinary tungsten fila- ment radiates as visible light not more than 5 % of the energy im- parted to it. Increasing temperature shifts the maximum of radia- tion nearer the visible region of the spectrum and is thus favourable to high luminous efficiency. It has been computed that a source operating at solar temperature might attain a luminous efficiency of 50%. P. G. Nutting (Bull. Bureau of Standards, May 1911) estimated that a source which produced only visible white light should yield 26 candles per watt, whereas the most efficient illuminants available do not give more than about five candles per watt. Nutting also calculated that a source producing only light of the most efficient wave-length for creating brightness, namely 0-54, would yield 65 candles per watt.
Our ideal should be to control emission of radiation so as to pro- duce only light of the particular colour desired. This has a bearing on attempts made to imitate the colour of daylight. By the in- troduction of a suitable tinted glass in the path of light from a gas- filled lamp, or by reflecting the light from a matt surface having a suitable coloured pattern, a close resemblance to normal daylight may be obtained (Ilium. Eng. Feb. 1920). Such " artificial daylight " units are of great value in industries where accurate colour matching is needed. But present processes involve the sacrifice of much light by absorption, and the overall efficiency of accurate units probably does not exceed about 15-20% of the light yielded by the lamp.
Progress in Shades, Reflectors and Lighting Appliances. Advances in the efficiency of illuminants have been accompanied by consider- able progress in methods of distributing light. Reflectors are now designed to screen the source from the eyes of persons using them, soften shadows and modify the natural distribution of light in any desired manner. Spacing rules for standard reflectors oT " Exten- sive," " Intensive " and " Focussing " types are furnished and adherence to these should ensure the provision of uniform illumina- tion of a specified value in foot-candles. Prismatic glass devices, for use with arc lamps and gas-filled lamps, have been designed to give a distribution of light favourable to uniform illumination between street lamps. An example is the Holophane street lighting lantern, which utilizes two prismatic glass surfaces, superimposed one on the other, with a smooth exterior and interior such that the lantern can be easily cleaned. Improved and simplified illumination photometers have enabled much information to be obtained regard- ing the illumination necessary for various purposes. It is now con- sidered preferable to state the illumination in foot-candles at the actual place where light is needed rather than to prescribe so many lamps of a specified consumption per square feet. This illumination can be related to the consumption of electricity per sq. ft. of area lighted. Thus with direct lighting by vacuum tungsten lamps in modern reflectors about 0-2-0-3 watts per lumen (i.e. per foot-candle per sq. ft.) is usual; with gas-filled lamps about 0-1-0-15. With indirect lighting about twice of the above values are required.
The introduction of the more efficient gas-filled lamps, which re- quire screening on account of the great brilliancy of the filament, has encouraged the use of indirect and semi-indirect methods of lighting. Small gas-filled lamps with opal glass bulbs have also been introduced. Lamps are now commonly mounted high up near the ceiling so as to be out of the direct range of vision and leave a clear space for the supervision of work. The high candle-powers available allow of greater mounting heights than those formerly used. Thus in factories lamps mounted 30 or even 40 ft. above the working plane are not unusual (see The Gas-filled Lamp and its Effect on Illu- minating Engineering by F. W. Willcox, Ilium. Eng. June 1919). Certain fine industrial processes, however, require local lighting with well shaded lamps. Reflectors have been developed for lighting large vertical surfaces, notably for picture lighting. A feature in the United States has been the development of " flood-lighting," i.e. concealed lighting by compact filament gas-filled lamps in parabolic reflectors giving a concentrated beam of light with a dis- persion of IO -I5. Thus a soo-watt lamp in a suitable mirror will yield a maximum beam-candle-power of 330,000. Such lighting units nave been used for spectacular lighting (e.g. illuminating historic monuments and buildings, large advertisement-placards, etc.), and during the war served as a measure of protection, to prevent un- authorized persons approaching arsenals or other works unseen.
For further information the following works may be consulted : The Development of the Incandescent Lamp, by G. B. Barham (1912) ; Elektrische Lichteffekte, by W. Biscan (1909); Lichtlechnik, edited by L. Bloch, issued by the German Illuminating Engineering Society (1921); Grundziige der Beleuchtungstechnik by L. Bloch (1907), translated by W. C. Clinton; The Application of Arc Lamps to Practical Purposes, by J. Eck (1910); Le Nuove Lampade Elettriche ad Incandenza, by G. Mantica (1908) ; Elektrische Beleuchtung, by B. Monasch (1907) ; The Electric Lamp Industry, by G. A. Percival (1920); Electric Lamps, by M. Solomon (1908); Electric Arc Lamps, by O. Zeidler and J. Lustgarten (1908).
Frequent articles on electric lighting appear in The Illuminating Engineer (London) ; The Transactions of the Illuminating Eng. Society U.S.A. (New York); and Licht und Lampe (Berlin). See also ILLUMINATING ENGINEERING. (J. S. D.)
LIGHT RAILWAYS, MILITARY. To transportation engi- neers, both civil and military, one of the most interesting features of railway work during the World War was the development of the network of 6o-cm. lines ("soixante") in rear of all the main fronts in France during the days of position warfare. These little narrow-gauge lines were, of course, no novelty, since similar lines had been employed in industrial plants, on large plantations and on large construction works. Lt.-Col. W. H. Cole, in Light Railways at Home and Abroad (1899), describes a 6o-cm. line built under a concession granted in 1890 to the Decauville Co. between Caen and Dives. Another 6o-cm. line of lighter rail ( 19- 1 Ib. to the yd., the same as the French military track) was built about the same time between Pithiviers and Toury, France. It was worked under lease by the Decauville Company. Nor was the idea new to the student of military affairs, as at the French siege manoeuvres before Paris in 1894 some 30 m. of such line were laid under simulated field conditions upon which complete trains were drawn by double-ended engines. German military tests in 1895 had shown that it was practicable to lay 6o-cm. (or 2-ft.) tracks, with rails weighing 10 to 19 Ib. to the yd., at a rate of from 600 to 700 yd. per hour for a distance of 31 m. of continuous laying. In the campaign in Manchuria (1904-5) both belligerents had made use of light narrow-gauge lines for distribution purposes. At Port Arthur, for example, despite the fact that the main railway supply line of the Japanese army was itself of 3 ft. 6 in. gauge, equipped with medium-weight engines and light cars of small capacity (6 tons), it was found desirable to lay 6o-cm. (2-ft.) track from a transfer station or siege junction to the adjacent artillery and engineer parks and thence to the siege batteries and smaller dumps near the troops. These lines of the Japanese were not suited to the use of locomotives. The sections were merely hooked together as laid. One hook was formed by a bend in an extension of the lower flange of a rail, and this engaged with a companion hook on the end of a fish plate bolted to the web of the rail of the adjoining section. At Fort Arthur the sections came to the front already assembled so that no field bolts were necessary. Where transportation on moun- tain roads was involved, as it was in the main campaign, the rails and ties came up separately and were assembled shortly before use. At Port Arthur the light flat cars were oushed about by soldiers, at the Sha-ho by horses.
No discussion of these special railways would be complete without some reference to the " battle of the gauges," which raged around them during and subsequent to the World War (1914-8). Many officers with railway and military experience contend that the gauge of field railways should invariably cor- respond to that of the main supply railway special light loco- motives, and, if necessary, light cars being used on rapidly laid, light rail lines. Others believe that where the main line is of 4 ft. 85 in. gauge a lesser gauge is desirable for distribution within the combat area, but think that 6o-cm. is too narrow; some have suggested 3O-in., others 36-in., and so on increasing to standard, most of them agreeing, however, that when the main line is narrow, 3 ft. 6 in. or less, no lesser gauge is necessary. Others contend for the 6o-cm., and these include many who have had long experience with these diminutive systems.
It is of value to examine into the purpose for which this character of equipment was intended at the outbreak of the World War. Gen. von Bernhardi, writing in 1912, had already indicated the use the Germans intended to make of them in the following words:
" Where, during the further advance of the army, the railways are commanded by hostile fortresses or 'forts d'arrgt,' the capture
