Open Book Publishers logo Open Access logo
  • button
  • button
  • button
GO TO...
book cover

2. The Artificial Lighting Available to European and American Museums, 1800–1915

David E. Nye

© 2019 David E. Nye, CC BY 4.0

As the rest of the contributions to this volume attest, the study of the role of artificial lighting in the display and understanding of art is a fascinating and until now largely unexamined subject. During the nineteenth century, gas and electric lighting became available to museums and art galleries, and it is worth exploring when, why, and where these artificial forms of illumination were installed. In order to lay the groundwork for further investigation, this chapter will briefly explain the different forms of lighting and provide a chronology of their emergence, together with comments on the relative slowness of their diffusion into interior spaces. When first introduced, both gas and electric lighting were primarily used outdoors, to light streets, parks, and squares, or to illuminate large interiors such as railway stations, opera houses, or emporiums. Museums long had little or no artificial light and simply closed at dusk. This was apparently the case with the Smithsonian Institution, for example, until some time after 1880.1

It also appears that gaslight was little used at world expositions from 1851 until 1881. The Paris Exposition of 1855 contained 4979 works by 2176 artists, but it closed at dusk. The London 1862 Exposition had an equally large art display, which closed 30 minutes before sunset. Electricity was more quickly adopted in such venues beginning in 1881. Yet many exposition buildings were still closed in the evening, and the focus at every fair from 1881 until 1915 was the spectacular lighting within the grounds. Visitors to the New Orleans Exposition in 1884 could see its art displays under electric lights at night, but this was unusual. As late as 1900 the Paris Exposition’s art exhibition relied on skylights and windows and was not open in the evenings. One lighting engineer considered that Exposition to be ‘a distinct step backward,’ without a ‘uniform scheme of illumination […] The lighting was a mixture of large and small incandescents, searchlights, projectors, display lighting of the spectacular order, acetylene, Nernst lamps, Welsbach burners, gas, and other illuminants — a conglomeration which was entitled to more credit as an exhibition of all known modern forms of lighting than as a comprehensive scheme of exposition illumination.’2 Expositions sought to be technologically advanced, yet as this description suggests, a great many different forms of artificial lighting were simultaneously on display, often in a jumble. It was only after 1900 that incandescent bulb lighting became the preferred form. As the example of world’s fairs suggests, the adoption of artificial lighting in art museums was a gradual process, and in the case of both gas and electricity occurred only a quarter of a century (or longer) after each these new forms of illumination appeared. This gradual change accords with the history of energy transitions, which typically have required forty to fifty years.3

Gas lighting was first developed commercially in Cornwall in 1798, quickly improved in Birmingham and Manchester,4 and displayed in London on Pall Mall in 1807. By 1820 there were 300 miles of gas lines in London. While used primarily for street lighting, many wealthy people also installed gas in their homes. There can be no question that many works of art were being viewed in private London homes under gaslight by the 1820s. In 1814, for example, Robert Ackerman’s London home, ‘The Repository of Arts,’ 101 on the Strand, was fully lighted with gas. Many of the leading artists and collectors saw Ackerman’s paintings illuminated there.5 This does not mean, however, that museums rushed to adopt the new technology of gas lighting. In 1861 the trustees of the British Museum were ‘unanimously of opinion, that they would not be justified in allowing the collections […] to be open at any hour which would require gaslight.’6
Elsewhere, however, gas lighting was installed in many venues in order to enable the working classes to view the collections during their leisure hours. This was the case at the Edinburgh Museum of Science (1854), the Oxford University Museum (1860), the Birmingham City Art Gallery (1885) and the Victoria and Albert Museum (1857) which installed 196 gas jets in its Sheepshanks Gallery of British paintings. Michael Faraday had assured them that burning gas, provided there was proper ventilation, would not endanger the works of art.7

Faraday was optimistic in this assessment. During the nineteenth century, gas for lighting was almost always produced from coal. Natural gas only came into widespread use in the twentieth century, and it has been used mostly for heating and cooking. Victorian lighting gas was not at all the same. It produced less light per cubic meter burned, and it released sulphur, ammonia, and carbonic acid into the air. Burning coal gas tarnished metals, blackened ceilings, increased the humidity, produced unwanted heat in summer, and weakened fabrics. Its soot darkened the surface of paintings, and its acidic vapours damaged canvas and interacted with the chemicals in paint.8 Gas is also an obvious fire hazard. Electric lights were less damaging, but they also presented problems that will be discussed below. In short, nineteenth-century curators had good grounds to remain cautious about artificial lighting.

This survey examines six basic forms of illumination, depicted below, but it cannot consider the various gas mantles, the many competing arc lights, or differences between incandescent light filaments.

  1. Burning coal gas produced perhaps ten-candle power, and it was in general use between 1810 and 1920.
  2. The direct current (DC) arc light was the most common form of electric lighting between 1875 and 1900. It produced 2000-candle power by jumping a strong current between two carbon rods, which needed frequent replacement. Its brightness could not be adjusted.
  3. The early incandescent light (designed by Edison, Swan, and others) had an enclosed carbon filament, and came in a variety sizes and strengths. Invented in 1879, it was gradually replaced after 1900.
  4. The Welsbach mantle was heated to incandescence by gas and produced six times more light than the older gas lights. It was widely adopted after 1885 and was common until the 1920s.
  5. The alternating current (AC) arc light spread quickly after 1893 and was greatly improved after 1900. It was rapidly replaced after ca.1910 by incandescent lighting.
  6. The tungsten incandescent light became commercially available in ca.1910. It lasted longer than the Edison carbon filament lamp, had an improved light spectrum and more efficient use of electricity. It could be made powerful enough to replace both arc and Welsbach lamps.

Fig. 2.1 Gas lamp, New Orleans. Photograph taken by daveiam, 12 June 2008. CC BY 2.0,

Fig. 2.2 Charles Brush, DC Arc Light, patent granted 1878. Glass shade not shown in order to reveal the interior mechanism. US Patent Office, Patent 203,411. Public domain.

Fig. 2.3 Thomas Edison, incandescent light bulb with carbon filament and improved glass shape, patent granted 1881. US Patent Office, Patent D 12,631. Public domain.

Fig. 2.4 Gas mantle burning at its highest setting. Photograph taken by Fourpointsix, August 2008. CC BY-SA 3.0,

Fig. 2.5 Westinghouse Alternating Current Arc Light, 1890. A frosted glass shade was placed over this mechanism when in use. US Patent Office, Patent 428, 435. Public domain.

Fig. 2.6 Incandescent light with tungsten filament. Electrical World, August, 1911.

Each of these three technologies (gas [Figs. 2.1, 2.4], arc [Figs. 2.2, 2.5] and incandescent filament [Figs. 2.3, 2.6]) continually evolved, and consideration of their incremental development is beyond the scope of this survey, as are lighting forms that failed to reach a large market.9 I will focus on the central characteristics of these three forms of lighting and explain when they became available. Nor does this chapter much concern itself with transatlantic comparisons and contrasts. The same corporations were active on both sides of the Atlantic, and innovations diffused rapidly from one continent to the other. However, there were differences in what forms of light a city or a nation preferred, and these in turn could affect the lighting a museum or art gallery might have access to. For example, in ca.1850 Charles Dickens found that Paris was far more brightly and extensively lighted with gas than London.10 Or again, in 1900, Boston and New York City had more than four times as much electrical lighting per capita as London, which relied more on gas.11 One cannot assume that museums everywhere had ready access to gas in 1830 or easy access to electrical services in 1890. Even if they did, they may not have adopted them.

The date of a discovery must not be confused with the date when a technology was widely available. Sir Humphrey Davy demonstrated an arc light at the Royal Institution in 1808, but this did not mean London’s streets were illuminated electrically shortly afterwards. Davy used 2000 voltaic cells (batteries) to produce enough current for his demonstration. Batteries were not an economical way to generate and distribute power for a far-flung system, however. It took seven decades of research and development before efficient dynamos lowered the cost of electricity sufficiently so that it became feasible to adopt arc lights instead of gas. One of the first demonstrations of arc lighting was in St. Petersburg 1875, where Alexander Lodyguin, a Russian engineer, showed that a thin carbon rod sealed inside a glass bulb from which all oxygen had been removed could cast a brilliant light.12 Paris had arc lights in the Gare du Nord, also in 1875.13 Several Wallace Farmer arc lights were erected at Philadelphia’s Centennial Exposition of 1876.14 It was in precisely this decade that many American art museums were founded, including New York’s Metropolitan Museum of Art (1870), The Boston Museum of Fine Arts (July 4, 1876), and Philadelphia’s Museum of Art (1876). However, electric lighting was not used extensively to view the art. Four decades later, the lighting of American museums was ‘almost universally from the top, very generally by skylights.’15 When the Boston Museum of Fine Arts began constructing a new building in 1902, its design ensured ample natural light in every room, with the main galleries on the upper floors, in rooms that had skylights.16

What difference did it make whether one saw a painting under natural light, gaslight or electric light? This question was answered in scientific tests made in 1892.17

Colour spectrum for sunlight, electric incandescent, and coal gas:








Electric incandescent




Coal gas




The test examined British coal gas and the then common Edison incandescent light with a bamboo filament. In the unlikely case that a painting was mostly in shades of red, gas and electric light might not seem to deviate much from daylight. But if an artwork contained green, then gaslight was a poor choice compared to the incandescent electric light. (Paraffin was slightly better than gas; candles were about the same.) Compared with natural light, coal gaslight had roughly three times as much red, only 2 percent as much green, and 60 percent less blue. The colours seen under burning coal gas were strongly skewed toward reds and ruddy yellows, and lush green landscapes or blue marine views appeared washed out and dull. In 1892 incandescent electric lighting was better but by no means equivalent to daylight: an Edison bulb had 40 percent more red, 40 percent less green, and 60 percent more blue. These test results were only suggestive for a museum director in 1892 who was trying to decide what kind of artificial lighting to install, however, for there were constant innovations in both gas mantles and incandescent filaments, as well as improvements in arc lighting.

Electricity was not adopted everywhere at the same rate, and one cannot assume that a public building or a museum had electric lighting during the last decades of the nineteenth century. In London, the telegraph office was one of the first to install incandescent electric lighting, in 1883. Arc lights were more common, and they were quickly adopted by department stores such as John Wanamaker’s in Philadelphia. These worked by passing a strong electric current across a narrow gap between two carbon rods that gradually burned down. The illumination was emitted ‘due to the intense heat of the tips of the carbon rods, and also to a smaller degree to the arc itself.’18 When arc lights were common (1875–1905), they were often powered by direct current (DC), and the upper carbon rod produced most of the light. This rod, which was positively charged, gradually acquired a hollow centre called the ‘crater of the arc.’ At the same time the lower, negatively charged rod became more and more pointed. These details might seem unimportant, but the crater of the upper arc produced much of the light and directed that light downward to where it was wanted. A DC arc light required considerable maintenance, because the carbon rods burned down rapidly and had to be replaced once a week, a task performed in the daytime. The positive rod burned down about twice as fast as the negatively charged one, and an important innovation was a mechanism to automatically advance the rods at the same rate as they burned down, in order to keep the distance between them constant. Electric arc lights required less maintenance than gas lamps, however, which for decades had to be serviced by a lamp lighter every day. Direct current electric arc lights were particularly common during the 1880s and 1890s in the United States, somewhat less so in France or Britain.19

Due to the competition from arc lights, gas lighting was radically improved in the 1880s by the innovation of the Welsbach gas mantle. It was based on the discovery that, while burning gas itself gave off more heat than light, the high temperatures could make other substances incandescent. Invented in Europe in 1885, the mantle contained oxides of thorium and cerium that when heated cast a brilliant white light. So equipped, a gas lamp produced six times more light with the same amount of gas. Its adoption prolonged the gas systems into the twentieth century. In the US, by 1900 more than 10 million Welsbach burners were in use.20 Some US cities, notably St. Louis and Milwaukee, relied primarily on Welsbach gas lighting in the first decade of the twentieth century. London’s streets also had far more gas than electric lighting in 1900. Other cities preferred arc lights, notably Boston and Pittsburg. Yet others had both gas and electric streetlights, including Chicago, Philadelphia, and Paris.

Enclosed incandescent lights were demonstrated in 1879 and became available in a few places in 1881, gradually diffusing into many interior locations. Several inventors developed them at almost the same time, most famously Edison in the United States and Swan in Great Britain. Their corporate interests merged early on. Edison’s assistant William Hammer demonstrated the system at the Paris Exposition of 1881 and on the Holborn Viaduct in London. There he installed 230 lamps in January 1882, as well as putting on a successful exhibition at the Crystal Palace.21 However, just because a new lighting system was displayed does not mean many people understood the technology or installed it. The 1881 Exposition Internationale d’Électricité in Paris22 had 277 arc lamps, 44 arc incandescent lamps and 1500 incandescent lamps. In subsequent months, the number was increased to 2500 lamps.23 It was impressive and even overwhelming to visitors, because the electric light was so new. Yet Hammer concluded that ‘no manager of any future exhibition is likely to repeat that terrific mélange of lights that flooded the interior of the Palais de l’Industrie with great brilliancy, but with an impracticable and impossible means of comparing and judging the relative merits of different systems.’24

However, a visitor could certainly see that unlike the arc light, the incandescent light did not flicker, created far less heat, and came in different sizes. The arc light shed 2000 candlepower and could not be turned up or down. The Edison/Swan light, and its many imitators, was less brilliant and therefore far more suited to illuminating domestic interiors or individual works of art. The team of inventors working with Edison created not only the light itself but the still familiar system of wall switches, sockets, fuses, and wiring systems that would make it possible to place a light in the most advantageous position to illuminate a painting rather than an entire room. It therefore would become the preferred system, eventually, when compared to either gas or arc lights. In the early 1880s the incandescent electric light was not widely adopted in public spaces such as railroad stations or department stores. However, as with gas lighting, it was quickly adopted by wealthy families, some of whom installed their own generating plants. Edison’s list of early customers reads like the social register, including J. P. Morgan and the Vanderbilts. In the 1880s a single light bulb cost one dollar, or half a day’s common wages, and using a single kilowatt hour might cost as much as 20 cents.25 With this in mind, consider how stunning was the New Orleans Exposition of 1884–85. Twenty thousand 16-candlepower incandescent lights lighted its 33 acres of interior space. These were particularly appreciated in the exposition’s art galleries.26

Beginning in the 1890s a new kind of arc light became available, which used alternating current. This meant that its two carbon rods alternated between being positive and negative. Both of its rods became pointed, and the light was shed equally in all directions. AC arc lights therefore required a good reflector much more than DC arc lights did. Both kinds of arc light were extremely hot, burning at 5500 to 6000 degrees Fahrenheit, and they produced a light too powerful to look at directly. Therefore, arc lights were usually hung higher than gas lighting, and if used indoors they were only suitable for large spaces like opera houses, department stores, and railroad stations.

In Britain, many preferred gas to electric light. Robert Louis Stevenson, for example, disliked the glare of arc lighting, which he declared

shines out nightly, horrible, unearthly, obnoxious to the human eye; a lamp for a nightmare! Such a light as this should shine only on murders and public crime, or along the corridors of lunatic asylums, a horror to heighten horror. To look at it only once is to fall in love with gas, which gives a warm domestic radiance fit to eat by.27

By the 1870s, gas had become traditional, and was associated with the familiar routine of lamp lighting, the ruddy glow of a fire, and domestic comfort.

Americans were generally enthusiastic about the white light cast by electric arc lights, but many Europeans agreed with Stevenson’s critique of them in favour of gas. As Chris Otter notes of Victorian Britain, ‘The flight from yellowness’ of gas ‘was not universally lauded. Most people were accustomed to seeing yellow. This is how normal night appeared: ochreous, cosy, peppery. The whiteness of electric illumination was often an unpleasant shock, registered chromatically as bluish.’28 Preece, who in 1883 had famously complained of the lack of electrical light in London, noted that compared to gas it indeed appeared blue at first. But ‘Americans did not call it blue at all,’ after they became accustomed to it, and ‘the imaginary blueness rapidly disappeared.’29 However, as the tests made in 1892 would show, the blueness was not imaginary. The electric light then available was 400 percent bluer than gaslight and 60 percent bluer than daylight. It was not a simple problem of adjustment. Stage actors complained that electric light changed the appearance of their traditional makeup and costumes, which had been developed with gaslight in mind. On the other hand, textiles under electric light appeared more as they did during the day, which is one reason why drapers in London and department stores in New York enthusiastically adopted Edison’s system. In short, as late as 1900 one could make a case for either gas or electric lighting, particularly due to the Welsbach mantle, which produced six times as much light with the same amount of gas, which meant that for a given level of illumination the amount of acid, humidity, and smoke were palpably reduced. Gas lighting persisted not only in the streets but also indoors well into the twentieth century, particularly in London.30

After 1900, however, a series of improved filaments in incandescent lamps produced more efficient lighting that also more closely matched the spectrum of light found in daylight. Some of these, such as the Nernst light, drew praise at world’s fairs, but it proved less successful commercially, due to its high cost. With the development of tungsten filament lamps, however, acceptable quality was united with reasonable consumer prices. Even so, the tungsten filament was weaker in the blue and green wavelengths than natural lighting. It was the best available at an economical price, but it did not fully replicate daylight, a goal only reached in the decades after 1920.

Nevertheless, by 1915 incandescent electric lighting had decisively proved itself the best form of illumination. Compared to gas, it was safer, cleaner, brighter, and closer to daylight’s palette. However, the distribution of electricity in public places was by no means the same in each country. In general, the French, the Germans, and especially the Americans adopted arc lights more quickly than the British, many of whom preferred gaslight. The incandescent light was adopted more slowly outdoors than arc lighting and made the most rapid progress inside buildings in the United States. Nevertheless, only about 20% of American homes had electrical lighting in 1920, though the figure was considerably higher for public buildings. It seems likely that art galleries and museums adopted electric lighting more readily than gas because of its obvious advantages for both showing and protecting art works, but only a detailed survey can establish to what extent this was the case in each country. Not all areas in Europe had a robust electrical infrastructure between 1880 and 1900, when incandescent service may not have been as available to museums as the less flexible, glaring arc lights. Moreover, the quality of both gas and electric light was not then comparable to daylight, and art works still could best be appreciated under natural light. The high cost of remaining open for longer hours, coupled with the additional cost of electrical lighting also may have retarded adoption in some venues. For all of these reasons, incandescent light bulbs apparently came slowly into use in art museums, though adoption seems to have been earlier in the United States. From 1884 some American expositions displayed art after dark, while the Paris Exposition of 1900 still did not. Future research might focus on the role collectors and patrons played in the adoption of gas and electricity by museums, as they were accustomed to seeing art under artificial illumination in their homes well before it was adopted in museums. Once the chronology of adoption is clearer, it will be possible to chart how painters and curators responded to these innovations, which fundamentally transformed how art could be displayed.

1 ‘Inauguration Ball. The Largest and Most Brilliant Ever Had’, Washington Evening Star, 57.8, 5 March 1881,; C. Melvin Sharpe, ‘Brief Outline of the History of Electric Illumination in the District of Columbia’, Records of the Columbia Historical Society, Washington, D.C. 48/49 (1946–47), 191–207 (pp. 201–02).

2 Luther Stieringer, ‘The Evolution of Exposition Lighting’, Western Electrician 29.12 (21 September 1901), 187–92 (p. 189),

3 David E. Nye, Consuming Power: A Social History of American Energies (Cambridge, MA: MIT Press, 1998). See also David E. Nye, American Illuminations (Cambridge, MA: MIT Press, 2018), chapter 2.

4 Leslie Tomoroy, Progressive Enlightenment: The Origins of the Gaslight Industry, 1780–1820 (Cambridge, MA: MIT Press, 2012), p. 239.

5 James Hamilton, London Lights: The Minds that Moved the City that Shook the World, 1805–51 (London: John Murray, 2007), p. 98.

6 Geoffrey N. Swinney, ‘Gas Lighting in British Museums and Galleries’, Museum Management and Curatorship 18.2 (1999), 113–43.

7 Nicholas Smith, ‘Let there be light! Illuminating the V&A in the nineteenth century’, V&A Blog, 10 September 2013,; Panos Andrikopoulos, ‘Democratising Museums’, Heritage Science Research Network, 6 June 2016,

8 Harold Passer, The Electrical Manufacturers, 1875–1900: A Study in Competition, Entrepreneurship, Technical Change, and Economic Growth (Cambridge, MA: Harvard University Press, 1953), pp. 195–204.

9 The most significant omission is the Nernst lamp, which heated a ceramic rod to incandescence and produced light close to the daylight spectrum. Demonstrated at the 1900 Paris Exposition, it was costly, but it had a niche market and was manufactured by AEG in Germany and Westinghouse in the US. See Paul Keating, Lamps for a Brighter America: A History of the General Electric Lamp Business (New York: McGraw Hill, 1954), p. 60.

10 Charles Dickens, The Uncommercial Traveller (New York: Charles Scribner’s Sons, 1905), p. 202,

11 ‘Electric Lighting in Boston’, Electrical World and Engineer, September 19, 1903. Reprinted in General Electric Company Review, 1 (November 1903), p. 12.

12 John A. Church, ‘Scientific Miscellany: A New Electric Light’, The Galaxy. A Magazine of Entertaining Reading 30.3 (1875), 415–16 (p. 415),;view=1up;seq=425

13 Stephen Inwood, City of Cities: The Birth of Modern London (London: Macmillan, 2005), p. 281.

14 Alfred Wallace, The Progress of the Country (New York: Harper Brothers, 1901), p. 276.

15 Cecil Brewer, ‘American Museum Buildings’, Journal of the Royal Institute of British Architects, 3rd series, 20 (April 1913), p. 372.

16 Helen Searing, New American Art Museums (New York: Whitney Museum of Art, 1982), pp. 40–42.

17 Gustavus Hartridge ‘The Electric Light and its Effects Upon the Eyes’, The British Medical Journal 1.1625 (1892), 382–83.

18 Franklin D. Jones, Engineering Encyclopedia: A Condensed Encyclopedia and Mechanical Dictionary for Engineers, Mechanics, Technical Schools, Industrial Plants, and Public Libraries, Giving the Most Essential Facts about 4500 Important Engineering Subjects, 3rd ed. (New York: The Industrial Press, 1941), pp. 56–57 (1st ed. available at

19 These comparisons are developed at length in chapters two and three of Nye, American Illuminations.

20 Passer, The Electrical Manufacturers, pp. 197–98.

21 William Hammer, ‘Notes on Building, Starting and Early Operating of the First Central Station in the World, Holborn Viaduct, London, England.’ Private Memorandum book, Hammer Papers, Box 20, folder 1, Smithsonian Institution Archives.

22 Alain Beltran and Patrice A. Carré, La fée et la servant: La societé française face a l’électricité (Paris: Éditions Belin, 1991), pp. 64–72.

23 Moncel, Comte Th. Du, and William Preece, Incandescent Lights, with Particular Reference to the Edison Lamps at the Paris Exhibition (New York: Van Nostrand, 1882),, p. 49.

24 Ibid., pp. 47–48.

25 David E. Nye, Electrifying America: Social Meanings of a New Technology, 1880–1940 (Cambridge, MA: MIT Press, 1990), p. 242.

26 Luther Stieringer, ‘The Evolution of Exposition Lighting’, p. 187.

27 Robert Louis Stevenson, ‘A Plea for Gas Lamps’, in his ‘Virginibus Puerisque’ and Other Papers (Boston: Small, Maynard & Co., 1907), pp. 249–56 (pp. 254–55),

28 Chris Otter, The Victorian Eye: A Political History of Light and Vision in Britain, 1800–1910 (Chicago: Chicago University Press, 2008), p. 185.

29 Cited in ibid.

30 If any museums burned natural gas, it was a far cleaner and more concentrated energy form than coal gas. Indeed, natural gas produces double the light with the same amount of fuel. It might have been used in museums close to sources of natural gas, such as Indiana in the 1880s, but until the twentieth century there was no inexpensive means to ship natural gas to large cities such as London, Boston, or New York. Therefore, in most cases natural gas can be ruled out as a likely source of light in galleries and museums, though not in wealthy households located within gas producing regions.