A phantom aspect is the appearance of a light signal when external light enters the signal, and is then reflected and refracted so that it leaves like the light of a normal aspect, usually colored. The external light is usually that of the sun, and phantom aspects appear when the sun is low, as when moving west early in the morning and east in the late afternoon. Phantom aspects are usually easily seen in traffic lights under these conditions, when all three roundels show bright color, but they disappear when one is close to the light and it is seen at an angle to the sun's rays.
The same phenomenon is seen in cat's eyes, where the tapetum at the back of the eyeball reflects the light focused by the cornea and lens. The eyes will glow brightly when the beam of a flashlight is thrown on them. Human eyes show a red reflection, often seen in flash photographs, since the light is reflected from the blood-rich choroid that absorbs less intense light. If you have a large clear marble, the effect can be studied easily. Note that light from a distant source is focused by the marble at a point near its rear surface (the marble will make a good magnifying glass). If the marble is backed by a white surface, it will be seen to light up like a cat's eye when a flashlight plays on it. This is the principle of reflectorized surfaces, where a layer of small spheres gives a similar directive reflection.
It is a law of optics that a light ray is followed in either direction, or that rays are reversible, which makes the elimination of phantom aspects difficult. An obvious design for a signal lamp would be a colored converging lens with an incandescent filament at its focus, and a concave reflecting mirror behind to intensify the light. Light from the filament, either direct or reflected, would pass through the lens and be refracted into a parallel beam of colored light. Unfortunately, sunlight from the same direction would enter the lens, be reflected at the reflector, and refracted again as a colored beam just as if it were produced by the lamp filament. The signal would show a color whether or not the lamp were lighted.
Semaphore signals had no problem with phantom aspects, since the semaphore arm was most easily visible under just those conditions that would produce a phantom aspect. Any light passing through the colored spectacle would only produce the proper color anyway, and a reflector was normally used to get the most out of the oil lamp. The first light signals had no problem with phantom aspects, either, since they were located where the sun does not shine--in subway tunnels.
The inventor of the position-light signal, A. H. Rudd, did have a problem with phantom aspects. His first signals were made the obvious way, and in the morning out on the Main Line all 10 lamps of the first signals lighted up brightly in the sunlight. He said it was like the full moon. Careful attention to detail, including eliminating the reflector for the direct beam (it was kept for the close-in beam, where it would do no harm), eliminated the problem completely. No color was involved here; it was just a matter of visibility.
The searchlight signal, with one optical beam and a moving spectacle carrying colored glass, should have had no problem, since any phantom aspect would be of the correct color. Nevertheless, pains were taken to eliminate any possible unwanted reflections, including the use of an inclined cover glass, so that reflected light would not be sent back along with the desired aspect.
Color-light signals are also designed to minimize phantom aspects, though here the position of the aspect (green on top, red on the bottom) could help to distinguish it.
For all the care taken to eliminate phantom aspects, it is curious that they have never been the cause of an accident reported to the ICC. There is only one case when anything similar caused an accident, and that case is indeed curious. It happened at Odin, IL, where the B&O St. Louis line crosses the IC main line. The interlocking there was operated and maintained by the IC. On 11 September 1941, IC No. 5 was on the crossing when it was run into by eastbound B&O Second 96, a manifest freight. One employee was killed. It happened at 5.45 pm with a low western sun behind the B&O train, whose engine crew said the signal showed Approach. Tests demonstrated that it could not possibly have done so.
It was found that the B&O had given the IC signal maintainer some black paint to paint the background of the color-position-light signal, and this was gloss paint, not the proper matte black. Sunlight, reflected from this gloss paint, combined with the red from the signal lights to give the impression of yellow, and this might have misled the B&O crew. This neglects the fact that two yellow lights would be at a 45° angle to show Approach. In spite of the "position" in their name, B&O signals gave primarily a color impression, as this instance shows. The signal was probably just hard to see in the reflected glare.
Composed by J. B. Calvert
Created 17 March 2002