For many years experiments have been made upon the use of the invisible rays which accompany visible rays. The practicability of signaling with invisible rays depends upon producing them efficiently in sufficient quantity and upon separating them from the visible rays which accompany them. Some successful results were obtained with a 6-volt electric lamp possessing a coiled filament at the focus of a lens three inches in diameter and twelve inches in focal length. This gave a very narrow beam visible only in the neighborhood of the observation post to which the signals were directed. The beam was directed by telescopic sights. During the day a deep red filter was placed over the lamp and the light was invisible to an observer unless he was equipped with a similar red screen to eliminate the daylight. It is said that signals were distinguished at a distance of six miles. By night a screen was used which transmitted only the ultraviolet rays, and the observer's telescope was provided with a fluorescent screen in its focal plane. The ultraviolet rays falling upon this screen were transformed into visible rays by the phenomenon of fluorescence. The range of this device was about six miles. For naval convoys lamps are required to radiate toward all points of the compass. For this purpose a quartz mercury-arc which is rich in ultraviolet rays was surrounded with a chimney which transmitted the ultraviolet rays efficiently and absorbed all visible rays excepting violet light. The lamp appeared a deep violet color at close range, but the faintly visible light which it transmitted was not seen at a distance. A distant observer picks up the invisible ultraviolet "light" by means of a special optical device having a fluorescent screen of barium-platino-cyanide. This device had a range of about four miles.
Light-signals are essential for the operation of railways at night and they have been in use for many years. In this field the significance of light-signals is based almost universally on color. The setting of a switch is indicated by the color of the light that it shows. With the introduction of the semaphore system, in which during the day the position of the arm is significant, colored glasses were placed on the opposite end of the arm in such a manner that a certain colored glass would appear before the light-source for a certain position of the arm. A kerosene flame behind a glass lens was the lamp used, and, for example, red meant "Stop," green counseled "Caution," and clear or white indicated "All clear." For many years the kerosene lamp has been used, but recently the electric filament lamp is being installed to some extent for this purpose. In fact, on one railroad at least, tungsten lamps are used for light-signals by day as well as by night. Three signals—red, green, and white—are placed in a vertical line and behind each lens are two lamps, one operating at high efficiency and one at low efficiency to insure against the failure of the signal. The normal daylight range is about three thousand feet and under the worst conditions when opposed to direct sunlight, the range is not less than two thousand feet. It is said that these lights are seen more easily than semaphore arms under all circumstances and that they show two or three times as far as the latter during a snow-storm.
The standard colors for light-signals as adopted by the Railway Signal Association are red, yellow, green, blue, purple, and lunar white. These are specified as to the amount of the various spectral colors which they transmit when the light-source is the kerosene flame. Obviously, the colors generally appear different when another illuminant is used. The blue and purple are short-range signals, but the effective range of the best railway signal employing a kerosene flame is only about four miles.
It has been shown that the visibility of point sources of white light in clear atmosphere, for distances up to a mile at least, is proportional to their candle-power and inversely proportional to the square of the distance. Apparently the luminous intensities of signal-lamps required in clear weather in order that they may be visible must be 0.43 candles for one nautical mile, 1.75 candles for two nautical miles, and 11 candles for five nautical miles. From the data available it appears that a red or a white signal-light will be easily visible at a distance in nautical miles equal to the square root of its candle-power in that direction. The range in nautical miles of a green light apparently is proportional to the cube root of the candle-power. Whether or not these relations between the range in miles and the luminous intensity in candles hold for greater distances than those ordinarily encountered has not been determined, but it is interesting to note that the square root of the luminous intensity of the Navesink Light at the entrance to New York Harbor is about 7000. Could this light be seen at a distance of seven thousand miles through ordinary atmosphere?
The most distinctive colored lights are red, yellow, green, and blue. To these white (clear) and purple have been added for signaling-purposes. Yellow is intense, but it may be confused with "white" or clear. Blue and purple as obtained from the present practicable light-sources are of low intensity. This leaves red, green, and clear as the most generally satisfactory signal-lights.
There are numerous other applications, especially indoors. Some of these have been devised for special needs, but there are many others which are general, such as for elevators, telephones, various call systems, and traffic signals. Light has the advantages of being silent and controllable as to position and direction, and of being a visible signal at night. Thus, in another field artificial light has responded to the demands of civilization.
XVI
THE COST OF LIGHT
Artificial light is so superior to natural light in many respects that mankind has acquired the habit of retiring many hours after darkness has fallen, a result of which has brought forth the issue known as "daylight saving." Doubtless, daylight should be used whenever possible, but there are two sides to the question. In the first place, it costs something to bring daylight indoors. The architectural construction of windows and skylights increases the cost of daylight. Light-courts, by sacrificing valuable floor-area, add to the expense. The maintenance of windows and sky lights is an appreciable item. Considering these and other factors, it can be seen that daylight indoors is expensive; and as it is also undependable, a supplementary system of artificial lighting is generally necessary. In fact, it is easy to show in some cases that artificial lighting is cheaper than natural lighting.