APPROXIMATE VALUES

Lumens per
Date    Filament                      Temperature            watt
1880    Carbon                          3300 deg.F.              3.0
1906    Carbon (graphitized)            3400                 4.5
1905    Tantalum                        3550                 6.5
1905    Osmium                          3600                 7.5
1906    Tungsten (vacuum)               3700                 8.0
1914    Tungsten (gas-filled)     up to 5300 deg.F.           10 to 25

Throughout the development of incandescent filament lamps many ingenious experiments were made which resulted usually in light-sources of scientific interest but not of practical value.  One of the latest is the tungsten arc in an inert gas.  By means of a heating coil, a small arc is started between two electrodes consisting of tungsten, but this as yet has not been shown to be practicable.

Another type of filament lamp was developed by Nernst in 1897.  It was an ingenious application of the peculiar properties of rare-earth oxides.  His first lamp consisted essentially of a slender rod of magnesia.  This substance does not conduct electricity at ordinary temperatures, but when heated to incandescence it becomes conducting.  Upon sufficient heating of this filament by external means while a proper voltage is impressed upon it, the electric current passes through it and thereafter this current will maintain its temperature.  Thus such a filament becomes a conductor and will continue to glow brilliantly by virtue of the electrical energy which it converts into heat.  Later lamps consisted of “glowers” about one inch long made from a mixture of zirconia and yttria, and finally a mixture of ceria, thoria, and zirconia was used.  The glower is heated initially by a coil of platinum wire located near it but not in contact with it.  Owing to the fact that this glower decreases rapidly in resistance as its temperature is increased, it is necessary to place in series with it a substance which increases in resistance with increasing current.  This is called a “ballasting resistance” and is usually an iron wire in a glass bulb containing hydrogen.  The heater is cut out by an electromagnet when the glower goes into operation.  This lamp is a marvel of ingenuity and when at its zenith it was installed to a considerable extent.  Its light is considerably whiter than that of the carbon filament lamps.  However, its doom was sounded when metallic filament lamps appeared.

An interesting filament was developed by Parker and Clark by using as a core a small filament of carbon.  This flashed in an atmosphere containing a vapor of a compound of silicon, became coated with silicon.  This filament was of high specific resistance and appeared to have promise.  It has not been introduced commercially and doubtless it cannot compete with the latest tungsten lamps.