Link to Home: Lighting A Revolution  


Preconditions to 20th Century Lamps

"I remember this circumstance very well because of the excitement and surprise and incredulity which he manifested at the time. He asked me over and over again what it was."
(William D. Coolidge, General Electric scientist, 1909)

Coolidge was recounting Fritz Blau's reaction to a lamp made with bendable (or "ductile") tungsten wire. Blau, an Austrian, had helped invent a "non-ductile" tungsten lamp only a few years earlier and knew well the difficulty of working with this metal. Coolidge's lamp was not the first improvement in Edison's design, nor the last. It built on previous work (such as Blau's) and fueled new work (such as Irving Langmuir's).

Inventors in the late 20th century had access to technical information unknown in Edison's time. Some knowledge came from outside the industry–like phosphor work done for television. But lighting scientists and engineers made many discoveries in the first half of the century, especially in the new industrial laboratories inspired by Edison's Menlo Park and West Orange labs. Research into the physics of electrical discharges, the metallurgy of tungsten, and chemical properties of glass all played a role in creating lamps that became available in the 1930s.

As the technology matured however, the pace of major improvements slowed. Below are some of the major developments of the 1900-1950 era important to lamps in use today.


Incandescent Lamps: Exit Carbon–Enter Tungsten

Just and Hanaman tungsten lamp, 1906
Non-ductile tungsten lamp
S.I. image #69,208

By 1900 the carbon filament lamp was a mature product in mass production. Electrical efficiency (or "efficacy") remained very low at about 3.5 lumens per watt (lpw). Aside from wasting electricity, these carbon lamps simply did not provide strong light. Inventors, especially in Europe with its high energy costs, searched intently for new filament materials.

Though carbon has the highest melting point of any element, the operating temperature of carbon filament lamps had to be kept relatively low. Very high temperatures caused carbon to evaporate quickly from the filament and coat the inside of the bulb, dimming an already low light. Experiments with various metals were aimed at finding a material that could operate at a higher temperature without so much evaporation. Higher operating temperatures meant brighter, more energy efficient lamps.

Carl Auer van Welsbach of Austria (inventor of the gas mantle) developed the first commercially practical metal filament lamp in 1898 by making filaments with element # 76, osmium. The very brittle filaments gave 5.5 lpw, a significant improvement, but osmium lamps proved difficult and expensive to make. They were replaced in 1902 by lamps invented by Germans Werner von Bolton and Otto Feuerlien, who used element # 73, tantalum. Tantalum lamps produced 5 lpw, a slight drop from osmium that was more than offset by tantalum's greater strength.

Tantalum was in turn superceded by lamps made with element #74, tungsten. Another difficult metal to work with, tungsten lamps like the one seen above gave 8 lpw, and in 1904 three different tungsten lamps appeared on the European market almost simultaneously. American manufacturers licensed and sold both tantalum and first generation tungsten lamps in the U.S.

Many of Edison's carbon lamp patents were expiring around this time and competition was heating up. In 1904, Willis Whitney used the new electrical resistance furnace at GE's Schenectady lab to bake carbon filaments at very high temperatures. The resulting filaments exhibited metal-like properties and gave 4 lpw. Sold as the "General Electric Metallized" or "GEM" lamp, this lamp still achieved only half the efficacy of the new tungsten lamps from Europe.

William Coolidge, also at GE's research lab, began exploring the metallurgy of tungsten. The European lamps were almost as fragile as earlier osmium lamps because tungsten was too brittle to bend ("non-ductile"). Coolidge developed a process to make bendable ("ductile") tungsten wire, and in 1910 GE began selling lamps made with this filament. The lamps gave 10 lpw, and also gave GE strong new patents.

Coolidge's colleague, the future Nobel laureate Irving Langmuir, discovered that by coiling the tungsten filament and placing an inert gas like nitrogen inside the bulb he could obtain 12 lpw or better. Langmuir's lamp joined Coolidge's on the market in 1913, both selling under the "Mazda" trade-name.

Various improvements in both the tungsten lamps themselves and in production machinery occurred during the following forty years. These cut costs drastically but improved lamp efficacy only slightly. By 1950, tungsten lamp technology seemed at a dead-end, especially given the growth of discharge lamps like fluorescent tubes. Some older engineers began advising younger colleagues to avoid staking a career on incandescent research.


Discharge Lamps: Lightning in a Tube

Cooper Hewitt mercury lamp, about 1919
Cooper Hewitt tube
S.I. image #lar2-1b1

An interesting curiosity of the 19th century were devices called Geissler tubes. German glassblower Heinrich Geissler and physician Julius Plücker discovered that they could produce light by removing almost all of the air from a glass tube and then sending an electric current through the tube as an arc discharge. Poor seals allowed air to seep back in and extinguish the light, but the work spurred research into discharge lighting.

In the first decade of the 20th century, two commercial discharge lamps gained modest popularity. One, invented by American D. McFarlan Moore, used carbon-dioxide or nitrogen filled tubes up to 250 feet long. Moore tubes were more efficient than carbon filament lamps but difficult to install and maintain. A second lamp, invented by American Peter Cooper Hewitt, passed an electric current through mercury vapor. Cooper Hewitt lamps (above) gave off much light and could be made portable, but the light was a garish blue-green suitable for few uses. These lamps contained about a pound of mercury each.

Coolidge's and Langmiur's tungsten filament lamps of the 1910s raised the efficiency standard for all lighting devices. Moore lamps, for one, soon disappeared from the market. Research indicated that very high efficacies might be attainable with discharge lamps however, so work continued.

Building on Moore's work, Georges Claudé of France developed neon tubes in 1910 and showed that a discharge lamp could give 15 lumens per watt–if one wanted red light. Additional European work resulted in a high-intensity mercury vapor lamp (from General Electric Company of England) in 1932. This lamp used a tiny fraction of the mercury needed for Cooper Hewitt lamps, had a screw base, and gave 40 lpw, though its color was still poor.

A collaboration of GEC in England, Philips in The Netherlands, and Osram in Germany produced a low-pressure sodium lamp also in 1932. The key to this lamp lay in a special glass that could withstand the corrosive effects of sodium. The light was a stark yellow suitable only for use in applications like street lighting, but efficacy started out at 40 lpw and reached about 100 lpw by 1960.

Reports began reaching GE and Westinghouse in the late 1920s and early 1930s of French experiments with neon tubes coated with phosphors. A phosphor is a material which absorbs one type of light and radiates another. A German patent in 1927 contained most of the features of a fluorescent tube, but the lamp was not produced.

American scientist Arthur Compton, a consultant to GE, reported seeing a green French lamp giving 30 lpw in 1934. An engineer at GE later wrote that they thought Compton had misplaced a decimal, that the true figure was 3.0 rather than 30 lpw.

The figure, soon confirmed, sparked an intensive research program. In 1936, tubes using low-pressure mercury vapor and a coating of phosphors were demonstrated to the Illuminating Engineering Society and the U.S. Navy. In 1939, GE and Westinghouse introduced fluorescent lamps at both the New York World's Fair and the Golden Gate Exposition in San Francisco. Other lamp makers soon followed.

Despite resistance from some utilities fearing loss of electricity sales, the need for efficient lighting in U.S. war plants resulted in rapid adoption of fluorescent technology. By 1951 industry sources reported that more light in the U.S. was being produced by fluorescent lamps than by incandescent.


Research After Edison: "The Science of Seeing"

Thomas Edison's lamp research focused mostly on the chemistry and engineering of the light bulb itself and its interaction within an electrical system. As researchers began building on Edison's work, the topics broadened to include subjects like optics and the physics of light itself. Edison, intent on inventing, cared little for basic research, but new professional "illuminating engineers" explored the fundamental nature of light and lighting devices.

Light output curve of Curtis fixture, April 1935
Photometric curve
S.I. image #lar2-1c1


For example, as metal filament lamps began replacing carbon lamps, the problem of glare arose. Shades for the brighter tungsten lamps had to be designed to both protect eyesight and to more effectively channel light. New applications like automotive and aviation lighting required development of a host of new lamp designs with special electrical and optical characteristics.

Researching human eye response to different colors and light levels became more important as electric lighting began to change people's lifestyles. Questions about the affect of lighting on productivity in both workplace and home carried great economic significance. Development of fluorescent lamps in the late 1930s led to experiments with "windowless factories."

The 1906 establishment of the Illuminating Engineering Society marked a formal recognition that lighting had moved from the realm of lone inventors to that of a profession. Corporate and academic researchers not only presented their work in the form of patents, but also wrote papers that appeared in scholarly journals. A prominent researcher, GE's Matthew Luckiesh described the field as "The Science of Seeing."

Researchers produced light distribution curves for fixtures (above), studied how different consumer groups used light, and developed deeper understandings of the fundamental nature of light. Expensive research equipment, needed to pursue these issues, made it difficult for smaller companies to compete. Lighting design emerged as a special field, distinct from architecture, just as lighting engineers diverged from electrical engineers.

Lighting and radio were the two electrical products that sold well throughout the Great Depression, justifying continued investment in research. The onset of World War II provided stimulated research for military uses of lighting, especially into materials like quartz and ceramics, while blackouts and materials rationing held back civilian purchases. Finally, the postwar economic boom released tremendous demand for lighting. The result proved to be a burst of lighting innovation.

To Inventing Modern Lamps To 19th Century Preconditions

To 19th
Century Hall
To 20th
Century Hall
Guest Lounge