But few of the thousands who daily travel over the railroads of the United States in the easy riding Pullmans of our limited trains have any idea upon what their safety depends in maintaining such a rate of speed.

The railroad wrecks of this year and the large number of killed and injured, as reported by the Interstate Commerce Committee, will, if compared with the reports from other countries (England and Germany especially), lead the most fair-minded to the conclusion that something is wrong with our railroad systems, either in their management, equipment, maintenance, employes or discipline, and that a higher standard and absolute uniformity in every department of railroading is necessary to secure greater protection to life and property.

It is impossible for one person to analyze the many causes for such a number of killed and injured and the loss of property, but when it is remembered that to protect space and control the trains of a railroad system various methods of signaling are used, the visual requirements of employes actively engaged in operating trains or giving and taking signals is of no small importance.

Furthermore, this system of controlling trains requires, in the daytime, the determination of the position of a semaphore arm, about

* Read before the Western Railway Club, Chicago, November 15, 1904.

4 feet long by 6 inches wide, or distinguishing the color of a disc about 17 inches in diameter displayed on a pole about 26 feet high. At night the color of a 6-inch lens, lighted by a kerosene lamp, must be observed. This must be determined by the enginemen at. sufficient distance in which to control their trains. Mr. R. Quayle, superintendent of motive power of the Chicago and Northwestern Railway, kindly furnishes the following data, which are the result of some recent experiments in controlling fast trains:

"With the ordinary brake, 70 pounds trainline pressure, a 6-car train running 70 miles an hour can be stopped inside of 2,000 feet, approximately 1,900 feet, and when the highspeed brake, 110 pounds train-line pressure, is used on the same train and running at the same speed, the train can be stopped in 1,450 feet."

It will be seen that a signal must be read at not much less than one-half mile (2,640 feet), as a few seconds are necessary for the engineman to shut off his steam and apply his air, and in each one of these seconds a train traveling 70 miles an hour is covering 88 feet.

There is no question that, next to the necessary practical knowledge of his engine, the vision and color perception of enginemen is the most important.

The writer rode considerably over 5,000 miles in the cabs of engines on various roads, at all times of day and night, and in all kinds of weather, to obtain data as to visual requirements of enginemen and firemen, and arrived. at the conclusion that the best known standard of binocular single vision and color perception is none too good, and it must be quick vision.

There are so many conditions that severely test the best vision and color sense that a person with these faculties diminished to even a small extent is a great source of danger.

Conditions existing about an engine, such as the escape of steam when an injector is used, when the whistle is blown, from leaking valves or connections, from the poorly packed piston rods of air pumps, cylinders and steam chests, will often completely envelop the engine and cab. Steam and soot from the smokestack is often blown back against the cab windows, covering them with steam and dirt and making it next to impossible to see through them, to say nothing of obtaining a view of anything through the condensed steam and smoke. This

drifts shuts off all vision for the time. Some considerable complaint is made of the drivers throwing mud and dirt on the front windows in moist weather.

The glare from the furnace door when the engine is stoked makes the reading of signals much more difficult. There is an iron shield above the furnace door on the engineman's side which protects him somewhat. Many enginemen have their seat curtained off to relieve them of this glare. After looking into this glow from the position of a fireman during the time required to shovel in five or six shovels of coal, it is an utter impossibility for a novice to read a signal. There was a black scotoma before my

Fig. 1. Conditions about engine interfering
with vision of signals.

is especially true in passing under the viaducts or bridges entering the railroad yards of large cities, where there are many moving engines and trains, and signals must be closely watched for. In freezing weather the escaping steam is especially bad, as the windows are coated with ice, and vision through them is out of the question. The escape of steam inside the cab in cold weather often coats the side windows with frost and ice, but rarely the front ones, as the rush of cold air against them while in motion keeps them clear. Double glass is also used, with an air space between, which prevents their freezing. The dust raised by passing trains often coats the windows, especially if they are damp from escaping steam; the engineman's position, being on the right side, escapes most of this, however, on roads which run their trains right-handed on double tracks. The cloud raised from plowing through snow

Fig. 2. Medium and high switch target.
Bad background.

eyes the size of the furnace door for fully ten minutes, and I could distinguish nothing except eccentrically. Firemen have told me they could not even see their steam gauge for several minutes after a stoking, and when one takes into consideration that from three to ten tons of coal are handled in a two to five hours run there is not much letup from looking into the firebox, and when this is done daily for five or six years, or even longer, before a fireman can expect to become an engineman, it must be a good pair of eyes that can stand it and pass the required standard examination for promotion without a rest or change.

The divergent rays from the headlight make it more difficult to read signals if there is fog, snow or mist in the air; with the electric headlights, while they are of the greatest advantage in all other ways, this is much more aggravated; an immense disc of reflecting and

refracting particles is directly before one, and it is almost impossible to distinguish signals through it. If these headlights were built on the same principle as the searchlights of the navy, so that they would throw parallel rays and could be deflected to strike the track at say 100 yards' distance in front of the engine, they would be greatly improved. This effect was obtained for Dr. Young and myself on the Burlington road by Mr. J. A. Carney, the master mechanic of the Burlington division, by making an extension to cut off the divergent rays and stopping the light down with a diaphragm. An 18-inch extension with the light stopped down to a pencil about 9 inches in diameter made a perfect light.

Certain atmospheric conditions are not only a source of great annoyance in reading signals, but often completely obscure them at a distance sufficient to control a train. Fog, snow, mist and rain take precedence in the order given, and when it is necessary for better vision to have the head out of the cab window the impingement of the fine particles of snow, mist or rain against the eyes blinds one almost instantly. The force of the wind traveling at a high rate of speed causes the tears to flow and blurs the vision after a very short expo

Fig. 3. Interlocking R. R. crossing signal
with bad background.

It is impossible to estimate distances accurately behind an electric headlight or to tell how far off an oncoming train lighted with one is.

The effect on the eyes of looking into an approaching electric headlight on double tracks is very much the same as looking into a searchlight, but it is unnecessary for enginemen to look directly into the source of light.

The constant jarring, swaying and rolling of an engine traveling at a high rate of speed is another factor in making signal reading difficult.

The constant supervision of an engine takes no small part of an engineman's time and attention, and his duties are far more than sitting on the seat and watching for signals. This is especially true when there is any trouble with the various mechanisms under his care.

Fig. 4. Switch targets wi h bad background.

Night signals are seen at a greater distance in these atmospheric conditions than day. The foreground and background of day signals make a great difference with their being easily seen. A sky background is the ideal; the kind. of day does not seem to make much difference. Signals displayed on roads running through mountainous country, and especially where there are many curves in valleys, are very hard to distinguish from the elevated position on either side. Woods stripped of their leaves or in full foliage, the proximity to buildings and bridges, all tend to make the position signals less distinct, while the crossbars of telegraph poles are very confusing. Undoubtedly night signals are much easier read, as the contrast is

so much greater, but as to their being seen any further, or so far, atmospheric conditions being the same, is questionable.

The reflection from snow, from a body of water, from the solid green of foliage, running toward the sun when near the horizon, the shimmer in the atmosphere on hot summer days, all are features which add much to the difficulty of seeing signals.

Thus it will be seen from the factors enumerated above requiring good vision and those tending to interfere with it, the best known standard of binocular single vision and color perception is none too good, and not only must

them and seeing the various mechanism about. the cab distinctly is out of the question; the color signals are made very indistinct. Green lenses are very good, but they blot out the red lights entirely, making one blind to red. Red lenses make the green appear dirty white, producing green blindness. Blue lenses do not cut out enough of the bright rays, and one is troubled with scotoma: A light amber lens, such as referred to in the German abstracts in the Annals of Ophthalmology and by Dr. H. H. Seabrook as having been tried in the German army for target practice, give a perfect result. I could look into the firebox of the large type engines, see all parts distinctly, relieved

Fig. 5. Block signal with smoke background.

a man have this, but it must be "quick vision," for he may for an instant be able to see through some break in the interfering media, and must be able to read his signal in that instant.

There seems to be no means of overcoming the interference with vision by escaping steam. An engine may come from the roundhouse in perfect condition, but when partly through its run something occurs which allows a leak, and the enginemen have this to contend with.

Lenses of various colors were experimented with in looking into the firebox from the fireman's position, and then trying to read day and night signals. The night signals were harder to see because of the glare being more intense in the surrounding darkness, the daylight relieving part of its intensity. Smoky lenses of the deeper shades allowed one to look into the firebox without producing a scotoma before the eves, but distinguishing signals with

Fig. 6. Bracket pole semaphore block.
Ideal background.

of the blinding glare, see about the cab perfectly, climb on the seat with no spot before my eyes, and the signal lights maintain their clear definition of color, white being tinged to slight yellowish. The green light can not be seen at so great a distance as with the naked eye, but red and white can be seen an equal distance and with equal distinctness. The effect produced by watching the lights with green lenses before the eyes was rather weird, as red lights were practically blotted out, and on removing the lenses flash into view in a startling manner; the white and green lights were not affected. With red lenses the green lights were observed as white coming through dirty glass; the white lights were tinged with a faint red.

As to the effect on the eyes of looking directly into the white heat of a furnace, various authorities disagree.