It occurred to me that perhaps a filament of carbon could be made to stand in sealed glass vessels, or bulbs, which we were using, exhausted to a high vacuum. Separate lamps were made in this way independent of the air-pump, and, in October, 1879, we made lamps of paper carbon, and with carbons of common sewing thread, placed in a receiver or bulb made entirely of glass, with the leading-in wires sealed in by fusion. The whole thing was exhausted by the Sprengel pump to nearly one-millionth of an atmosphere. The filaments of carbon, although naturally quite fragile owing to their length and small mass, had a smaller radiating surface and higher resistance than we had dared hope. We had virtually reached the position and condition where the carbons were stable. In other words, the incandescent lamp as we still know it to-day [1904], in essentially all its particulars unchanged, had been born.
After Edison's later success with bamboo, Swan invented a process of squirting filaments of nitrocellulose into a coagulating liquid, after which they are carbonized. Very fine uniform filaments can be made by this process and although improvements have been made from time to time, this method has been employed ever since its invention. In these later years cotton is dissolved in a suitable solvent such as a solution of zinc chloride and this material is forced through a small diamond die. This thread when hardened appears similar to cat-gut. It is cut into proper lengths and bent upon a form. It is then immersed in plumbago and heated to a high temperature in order to destroy the organic matter. A carbon filament is the result. From this point to the finished lamp many operations are performed, but a discussion of these would lead far afield. The production of a high vacuum is one of the most important processes and manufacturers of incandescent lamps have mastered the art perhaps more thoroughly than any other manufacturers. At least, their experience in this field made it possible for them to produce quickly and on a large scale such devices as X-ray tubes during the recent war.
During the early years of incandescent lamps, improvements were made from time to time which increased the life and the luminous efficiency of the carbon filaments, but it was not until 1906 that any radical improvement was achieved. In that year in this country a process was devised whereby the carbon filament was made more compact. In fact, from its appearance it received the name "metallized filament." These carbon filaments are prepared in the same manner as the earlier ones but are finally "treated" by heating in an atmosphere of hydrocarbons such as coal-gas. The filament is heated by electric current and the heat breaks down the hydrocarbons, with the result that carbon is deposited upon the filament. This "treated" filament has a coating of hard carbon and its electrical resistance is greater than that of the untreated filament.
The luminous efficiency of a carbon filament is a function of its temperature and it increases very rapidly with increasing temperature. For this reason it is a constant aim to reach high filament temperatures. Of all the materials used in filaments up to the present time, carbon possesses the highest melting-point (perhaps as high as 7000°F.), but the carbon filament as operated in practice has a lower efficiency than any other filament. This is because the highest temperature at which it can be operated and still have a reasonable life is much lower than that of metallic filaments. The incandescent carbon in the evacuated bulb sublimes or volatilizes and deposits upon the bulb. This decreases the size of the filament eventually to the breaking-point and the blackening of the bulb decreases the output of light. The treated filament was found to be a harder form of carbon that did not volatilize as rapidly as the untreated filament. It immediately became possible to operate it at a higher temperature with a resulting increase of luminous efficiency. This "graphitized" carbon filament lamp became known as the gem lamp in this country and many persons have wondered over the word "gem." The first two letters stand for "General Electric" and the last for "metallized." This lamp was welcomed with enthusiasm in its day, but the day for carbon filaments has passed. The advent of incandescent lamps of higher efficiency has made it uneconomical to use carbon lamps for general lighting purposes. Although the treated carbon filament was a great improvement, its reign was cut short by the appearance of metal filaments.
In 1803 a new element was discovered and named tantalum. It is a dark, lustrous, hard metal. Pure tantalum is harder than steel; it may be drawn into fine wire; and its melting-point is very high (about 5100°F.). It is seen to possess properties desirable for filaments, but for some reason it did not attract attention for a long time. A century elapsed after its discovery before von Bolton produced the first tantalum filament lamp. Owing to the low electrical resistance of tantalum, a filament in order to operate satisfactorily on a standard voltage must be long and thin. This necessitates storing away a considerable length of wire in the bulb without permitting the loops to come into contact with each other. After the filaments have been in operation for a few hundred hours they become brittle and faults develop. When examined under a microscope, parts of the filament operated on alternating current appear to be offset. The explanation of this defect goes deeply into crystalline structure. The tantalum filament was quickly followed by osmium and by tungsten in this country.
The osmium filament appeared in 1905 and its invention is due to Welsbach, who had produced the marvelous gas-mantle. Owing to its extreme brittleness, osmium was finely divided and made into a paste of organic material. The filaments were squirted through dies and, after being formed and dried, they were heated to a high temperature. The organic matter disappeared and the fine metallic particles were sintered. This made a very brittle lamp, but its high efficiency served to introduce it.
In 1870 when Scheele discovered a new element, known in this country as tungsten, no one realized that it was to revolutionize artificial lighting and to alter the course of some of the byways of civilization. This metal—which is known as "wolfram" in Germany, and to some extent in English-speaking countries—is one of the heaviest of elements, having a specific gravity of 19.1. It is 50 per cent. heavier than mercury and nearly twice as heavy as lead. It was early used in German silver to the extent of 1 or 2 per cent. to make platinoid, an alloy possessing a high resistance which varies only slightly as the temperature changes. This made an excellent material for electrical resistors. The melting-point of tungsten is about 5350°F., which makes it desirable for filaments, but it was very brittle as prepared in the early experiments. It unites very readily with oxygen and with carbon at high temperatures.
The first tungsten lamps appeared on the market in 1906, but these contained fragile filaments made by the squirting process. When the squirted filament of tungsten powder and organic matter was heated in an atmosphere of steam and hydrogen to remove the binding material, a brittle filament of tungsten was obtained. The first lamps were costly and fragile. After years of organized research tungsten is now drawn into the finest wires, possessing a tensile strength perhaps greater than any other material. Filaments are now made into many shapes and the greatest strides in artificial lighting have been due to scientific research on a huge scale.
The achievements which combined to perfect the tungsten lamp to the point where it has become the mainstay of electric lighting are not attached to names in the Hall of Fame. Organization of scientific research in the industrial laboratories is such that often many persons contribute to the development of an improvement. Furthermore, time is usually required for a full perspective of applications of scientific knowledge. In the early days organized research was not practised and the great developments of those days were the works of individuals. To-day, even in pure science, some of the greatest contributions are made by industrial laboratories; but sometimes these do not become known to the public for many years. The whole scheme of scientific development has changed materially. For example, the story of the development of ductile tungsten, which has revolutionized lighting, is complex and more or less shrouded in secrecy at the present time. Many men have contributed toward this accomplishment and the public at the present time knows little more than the fact that tungsten filaments, which were brittle yesterday, are now made of ductile tungsten wire drawn into the finest filaments.
The earlier tungsten filaments were made by three rival processes. By the first, a deposit of tungsten was "flashed" on a fine carbon filament, the latter being eliminated finally by heating in an atmosphere of hydrogen and water-vapor. By the second, colloidal tungsten was produced by operating an arc between tungsten electrodes under water. The finely divided tungsten was gathered, partially dried, and squirted through dies to form filaments. These were then sintered. The third was the "paste" process already described. These methods produced fragile filaments, but their luminous efficiency was higher than that of previous ones. However, in this country ductile tungsten was soon on its way. An ingot of tungsten is subjected to vigorous swaging until it takes the form of a rod. This is finally drawn into wire.