All the advance made in illuminating engineering has been made by utilizing higher and higher temperatures. Attempts at producing light by the electric current were made by heating wire conductors to a high temperature, and the first attempt at an incandescent lamp was with a platinum wire; but platinum melts at about I76o ยฐ C., and the efficiency of light production by this means was very low. Then Edison discovered that a wire of carbon--a carbon filament--could be used in the place of platinum. Carbon does not melt or boil until a temperature of about 400o ยฐ C. is reached. Consequently it should withstand a much higher temperature in an incandescent lamp than platinum. The old carbow filament lamps were run at about 18oo ยฐ C., somewhat above the melting-point of platinum, ,and thus with higher efficiency and no danger of melting. About 45 watts of electrical energy were required per candle~power.
Higher temperature and correspondingly better efficiency were limited, not by the mel,ting or the boiling of the filament, but by another limitation--evaporation. The carbon filament, much below its melting-and boiling-point, slowly evaporated, the carbon vapor being deposited on the lamp bulb. With continuous evaporation the filament became thinner; thereby tl~e temperature lowered and t.he light became less. T:he deposiied carbon vapor also obstructed the light. The operating temperature of the old carbon filament lamp is therefore limited to a value where the rate of evaporation does not unduly reduce the light production within a reasona,ble ~ime of, say, 500 hours or so.
The material obtained by carbonizing the bamboo fibre employed in incandescent lamps evaporates rather rapidly, but carbon deposited from g, asoline at high temperature does noยข e-caporate so easily and can be used at higher temperatures. A higher efficiency was therefore obtained by depositing a shell of carbon from gasoline vapor upon the carbon filament produced from fibre.
The carbon filamenl of the old incandescent lamp evaporated, although it was operated at a temperature much below the boilingand melting-poin,t. Tungsten, on the other hand, can be run nearly at the temperature of the melting-point with very little evaporation. Osmium proved very efficient as an illuminant in a lamp, but is too rare to be generally used. Tantalum gave good resul,ts, having a higher resistance than carbon, but it has been superseded by tungsten,