were crude forms of incandescent lamps; but up to that time the work that had been done consisted of separate and incomplete experiments. What was lacking was a complete set or system of apparatus which could be used to produce electric lighting in at practical way, or rather commercial way. This putting together of the necessary elements, even though they may already exist. separately, is often a more important and difficult step in the creation of a new art than the invention of the individual parts, however essential each may be. In 1878 and 1879, the times being in that peculiar state when they are ripe for very rapid advance, which condition usually precedes all great inventions or industrial enterprises, there occurred almost simultaneously the bringing forth of several more or less complete systems of electric lighting. At that time the most serious difficulty was the so-called "subdivision" of the electric light; that is, the running of several lamps from the same source of current, or on the same circuit, without interfering with each other. This bugbear was greatly exaggerated, and was much discussed by scientific and technical men at that time, some of whom maintained that the subdivision was not only practically but theoretically impossible. The overcoming of this difficulty was therefore the primary object of the electric-lighting systems first introduced. Three radically different methods were almost simultaneously brought out, and put into quite extensive practical use. These three systems were invented and developed by Jablochkoff of Paris; Brush of Cleveland, O.; and Edison of Menlo Park, N.J. In the Jablochkoff system the subdivision of the electric light was accomplished by using a form of lamp called an electric candle, which was first invented by him in 1876. It consists of two thin pencils of carbon held at a fixed distance apart by insulating material in the form of a strip of kaolin. All that was necessary to operate a number of these lights on the same circuit successfully was to connect them by wires in a simple series, so that the current flowed through them one after another. The arc formed at each lamp was necessarily of constant length, and there was no tendency for one lamp to act differently from the others, or interfere with them in any way. An alternating current dynamo was employed to supply the current, in order that the two pencils should burn at the same rate. The Jablochkoff system has the

practical difficulties of requiring a new candle to be switched on every two hours, and the cost of the candles made the light rather expensive. It was sufficiently developed to be used for lighting the Avenue de L'Opéra and other places in Paris in 1878, and it was also introduced and used in a few places in America; but the objections stated above prevented it from being a permanent success commercially.

In the Brush system, brought out in 1878, arc lamps with regulating mechanism practically identical with those employed to-day were operated in series on a single circuit. The success of this system was due to the fact that it included a complete set of apparatus; that is, a dynamo having a current regulator, and arc lamps with differential coil regulators and simple ring-clutch mechanism, which lamps could be operated satisfactorily in series. Brush also invented the "double-carbon" lamp; that is, a lamp in which a second pair of carbons are automatically thrown into action when the first pair are used up. Before the introduction of inclosed arc lamps in 1893 this feature was practically essential for all-night arc lighting, as in the case of most street-lamps. In short, Brush gave to the public a system in which the various elements were reasonably good in themselves, and co-operated to produce a fairly economical and generally satisfactory method of lighting. Good business management also contributed largely to the wide use and original success of the Brush apparatus.

The Edison system, which was developed experimentally during 1878-1879, and brought out commercially in 1880, made use of the incandescent instead of the arc lamp. The Edison system contained the necessary elements for a successful use of the incandescent lamp, which elements had not been known or used. previously, although the system is apparently very simple. These essential elements are: First, a dynamo having an armature with a very low internal resistance; and the armature introduced by Edison did not have more than one-fifth to one-tenth of the resistance of similar machines used by others prior to that time. Second, a constant potential or electrical pressure was maintained throughout the system of conductors, to which the lamps were connected in parallel, that is, in branch circuits, so that the turning on or off or breaking of a lamp did not affect the others. Third, the lamps consisted of high-resistance carbon filaments

hermetically sealed in glass globes in an almost perfect vacuum. High-resistance filaments are practically necessary to enable the use of reasonably high voltage, which greatly reduces the weight of copper required for the conductors, and a vacuum is required to prevent the destruction of the filament and the loss of heat by convection. At the same time that the Edison system was brought out, or soon after, other inventors were working on systems similar to the above. Important contributions to the progress of incandescent lighting were made in lamps and other devices by Sawyer and Man, Maxim and Weston, in America; also by Swan and Lane Fox in England. In the field of arc lighting, Thomson and Houston brought out a complete and very successful system, which had the radical advantage over the Brush system that the regulator for controlling the current and keeping it constant was superior to that employed by Brush. Indeed, the great success of the Thomson-Houston system was largely due to the very ingenious and effective regulator which they applied to their dynamo. Another arc-lighting system was brought out by Weston; but this also was defective in not having a satisfactory current regulator, although the dynamo and lamp were exceedingly well designed and constructed, considering the time at which they were made. In Europe, arc-lighting

systems have been developed by Siemens, Krizig & Piette (the "Pilsen lamp"), Crompton, Gulcher, and other inventors and manufacturers; but arc lighting in Europe is not as popular or as extensively used as in this country.

A system of electric lighting by means of alternating current transformers was invented by Gaulard and Gibbs in 1882. This system was based on the early experimental work with induction coils by Faraday in 1831, Henry in 1832, Page in 1835, Sturgeon in 1837, Ruhmkorff in 1851, and others. Gaulard and Gibbs made the fatal error of running the transformers in series, which is impracticable. In 1885 Zipernowsky, Deri, and Blathy brought out a system in which this mistake was corrected, the transformers being worked in parallel. The alternating current transformer system was extensively and successfully introduced in the United States in 1887 by the efforts of Westinghouse, Stanley, and others. The great saving in the amount of copper required for the distributing conductors in this system caused it to be

rapidly and widely adopted. In the meantime, the dynamo machine was being gradually but steadily perfected by the various inventors and manufacturers, for use in their electric-lighting systems. Edison, Brush, Thomson, Houston, and Weston, all contributed to this progress. The multipolar dynamo was developed by Elphinstone and Vincent in 1879 and 1880, and by Schuckert, Gramme, Gulcher, Mordey, and others. The theoretical study of the dynamo was taken up by Clausius, Sir William Thomson, and Frölich. The last-named brought out in 1880 a working theory which is still of practical value. A great advance in the theory and practical design of dynamos was made by J. and E. Hopkinson in 1886.* Their paper laid down the correct theory, and embodied a method of designing the magnetic circuit of dynamo machines, which up to that time had been very imperfectly understood; and most machines at that time were very bad in this respect. Kapp brought out in 1887 a similar method of designing the field magnets of dynamos; but it was largely empirical, and not so complete and scientific as the Hopkinson method, which is now generally used by the best electrical engineers. Hopkinson and Edison † independently invented the three-wire system of distribution, which makes a considerable saving in the amount of copper required for low-tension circuits. Incandescent lamps have gradually been improved in cheapness and efficiency, and the mechanism of arc lamps has been perfected from time to time. Great improvements have been made in the last few years in the construction of large direct-coupled steamengines and multipolar dynamos for central stations. Enormous progress has also been made in the general perfection of the various details of electric-lighting plants. The insulation of elec'trical conductors has been very greatly improved. Storage bat

teries have been extensively applied to electric-light stations and isolated plants to secure greater flexibility, economy, and reliability. The introduction of the inclosed arc lamp by L. B. Marks in 1893 has extended the life of a pair of carbons from 8 to 100 hours, thereby greatly reducing the cost and trouble of renewing them. Alternating-current arc lamps, both open and inclosed, have been developed to the point of rivalry with the direct-current *Philosoph. Transact. of the Royal Society.

† U. S. Patent No. 274290, March, 20, 1888.

types. The Nernst lamp,* brought out in 1899, has a filament composed of certain metallic oxides which must first be heated by a flame or auxiliary electrical device in order to become a conductor. This type has a higher efficiency than the ordinary carbon filament-lamps. The Hewitt lamp † first exhibited in 1901 has a still higher efficiency. It consists of a tube containing mercury vapor. The various electric lamps are considered in Volume II.

For further study of the general history of lighting, the reader is referred to the following:

Histoire du Luminaire depuis l'époque romaine jusqu'au XIX. Siècle, par Henri-René D'Allemagne, Paris, 1891.

This is a voluminous treatise on the history of all methods of artificial illumination from ancient to modern times. The artistic side is most prominent, but technical matters of construction and operation are also considered.

For the history of electric lighting reference may be made to The Electric Light, by Alglave and Boulard; translated by T. O'Conor Sloane; edited by C. M. Lungren.

Electric Illumination, by James W. Dredge, two vols., London, 1883-1885, contains the most complete account of the history. of electric lighting prior to the dates of publication, each form of dynamo and lamp being described in detail.

Arc and Glow Lamps, by Julius Maier, London, 1886, contains illustrations and descriptions of many forms of lamps. The Evolution of the Electric Incandescent Lamp, by F. L. Pope, Elizabeth, N. J., 1889, gives a detailed account of the early work of Sawyer and Man, and Edison. Marsden J. Perry made an address on the history of electric lighting before the National Electric Association, February, 1891 (Elec. World, Feb. 28, 1891); and Charles F. Brush gave very interesting personal reminiscences of the early history of arc lighting before the same body in February, 1895 (Elec. World, March 2, 1895).

A very complete history of the dynamo is contained in Thompson's Dynamo-Electric Machinery (Fifth Edition, pages 5-21).

* U. S. Patents Nos. 685, 724-32, Oct. 29, 1901. Transactions Amer. Inst. Elec. Eng., Aug. 1901.

† Electrical Review, New York, April 27, 1901.