on the brakes has been eliminated, because it was thought that this might in course of time weaken the feeling of responsibility and the attention of the drivers; moreover, it would make the apparatus more complicated and consequently more unreliable. The recording instrument has only been tried in the Danzig trials, as it is not of great importance if the slippers only act at the distant signal, which it is permissible to over-run even when it is standing at "caution." We may, however, at once draw attention to the fact that in the case of experiments made with this apparatus it is primarily of minor importance, how the movement which the treadles produce on the locomotive is utilized. The chief difficulty consists in making the arrangements for the transmission reliable and durable. The experiments, particularly those carried out with great care by the Halle directorate, have given valuable information for determining the value of this apparatus. The turning on of the steam whistle is reliably effected at speeds of from about 20 kilometers (12.5 miles) per hour upwards, if the slippers and the treadles are in order. Unintentional turning on has occurred. It is true that there is not much risk of this, because as the van Braam apparatus is arranged, it is necessary for both slippers to be raised before the whistle or brake is affected. In several cases the slippers were broken, although special care was taken on the experimental sections that there should be the proper clearance. The cause of this is that a depression of 5 millimetres (3/16 inch) of the slippers suffices to make them strike objects, which are outside the loading gauge proper, but very near to its boundary, such as guard baulks on bridges, check rails at cross-overs, stiffener brackets on bridges, etc. Experiments made, however, show that oscillations of up to 25 millimetres (31/32 inch) actually occur. The guard baulk of a bridge, which was only 35 millimetres (13/8 inches) above rail-level, and hence still 15 millimetres (19/52 inch) from the boundary of the loading gauge, was found to bear slipper marks. Similar observations were made at other places. It follows that as at present constructed the slippers are liable to strike objects and become broken.

During the trials morever, difficulties arose in the connection between the treadles and the distant signals. During the snowfalls of 16, 17 and 24 November of last year the treadles could not be moved into position. The consequence was that the indications of the signals became uncertain and dangerous. In order to make the operation of the signals possible, it became necessary to dis

connect the treadles. It seems at the very least very doubtful whether it is possible to improve the treadles in such a way that they will always act reliably, even under unfavorable conditions in winter. For this reason the experiments are to be continued in a modified form; instead of using movable treadles, having a position dependent on that of the distant signal, fixed treadles or bars are to be tried. The locomotive driver is then informed not how the distant signal is set, but only that he is approaching such a signal, just as in the case of the electric apparatus described above. Experiments with this arrangement have not yet been made.

The proposal to use wireless telegraphy for transmitting signals to the locomotive has also been taken into consideration. The Stettin directorate has made trials of this kind on the Angermünde-Neukünkendorf section. The installation was provided by the firm C. Lorenz of Berlin; electromagnetic waves which were continually produced at a sending station were transmitted by induction to the telegraph lines along the track, and from them to a receiver placed on the locomotive. At those places where an approach to a signal was to be indicated, the waves going to the receiver on the locomotives were intercepted by a screen. This brought a wavedetector back to the zero position and a colored disk was shown. The experiments have shown that it is possible, within certain limits, to transmit signals to the locomotive; the strength of the waves, however, became materially reduced as the distance between the sending station and the locomotive increased. Then also the effect which the screen at the distant signal was intended to produce, was also produced in another way, for instance by telegraph lines crossing the railway, by iron bridges, by iron signal poles, and even when the locomotive ran through a deeper cutting. It is as yet hardly likely that this apparatus will give useful results.

On the other hand, good results were obtained with another device tried, the so-called recording alarm. This apparatus is intended to record the over-running of signals. A rail contact is placed next the home or block signal. If this is run over while the signal is at "danger," a recording alarm is set ringing at the place from which the signal is operated. In stopping the alarm a counter is advanced one, and thus records the overrunning of the signal. This makes it possible forthwith to investigate every case of overrunning. If a signal is overrun at all frequently, the cause can be investigated and any existing unfavorable conditions can be remedied. This apparatus has been tried by five different director

ates, and in each case were good results obtained. A larger number of these recording alarms has accordingly already been ordered.

Thus the number of experiments made on the Prussian State Railway in order to determine, in actual practice, the utility of the proposed devices, is by no means inconsiderable. Proposals of the most varying kinds have been considered. Probably no other administration has made more systematic and comprehensive experiments in order to solve the problem of the prevention of overrunning; and if, without prejudice, we consider the results of the experiments, we obtain an answer to the question why the administration of the Prussian State Railway has not yet decided on the general introduction of one of the proposed devices. An "absolutely reliable" action can not be demanded of such appliances, and will probably never be expected. But it may with justice be demanded that failures shall be the exception, when that care is exercised in the maintenance of the track appliances and the rolling stock which is to be expected in the proper operation of a railway system of the magnitude of the Prussian State Railway. But this standard has not yet been attained in the case of any of the appliances in question; nor, as may be specially emphasized, has it been attained in the case of the van Braam apparatus. Certainly the method adopted by the Prussian State Railway (practical trial of a large number of different appliances) is the best for finding out which of the appliances in question will satisfy all reasonable requirements. In adopting this procedure, the administration need have no fear that it will lag behind foreign administrations in the utilization of valuable inventions; this is shown by the second report, just published, of the American block signals and train control board, which was instituted in 1907 in order to test block signals, automatic stop signals and signals shown on the locomotive. The work of this board extended over two years, and during this time several hundred inventions were submitted to and examined by them; the result was that at the end of the second year of the period the report applies to, one sole appliance for stopping the trains from the track had been constructed and was ready for being tested in practice. Thus this report shows that the large majority of the railway-signal engineers of the United States considers the question of the automatic rail contacts and signals on the locomotive as not yet sufficiently advanced to consider these appliances as ready for practical service. The same is indeed at present the case in Germany.

PREVENTION OF RAILROAD ACCIDENTS

BY A. M. SCHOYER,

Gen. Supt. Pennsylvania Lines West of Pittsburg.

SPEECH BEFORE ANNUAL CONVENTION OF RAILROAD MEN,
INDIANAPOLIS, INDIANA,
NOVEMBER 9, 1910.

In discussing this question, the speaker does not pretend to be an expert, nor does he expect to exploit the railroad with which he is connected.

The subject, as a whole, is one of the utmost importance to the public, to the Government and to the railroads. Its importance is more apparent as the traffic of the country increases, and it is to the credit of the Railroad Commission of Indiana that they have set aside this Annual Conference for its consideration.

The scientist tells us today that the battle with disease is threequarters won when its direct cause is discovered. After the germ is located, its destruction is inevitable. The same thing must be true with a matter of this kind. To prevent railroad accidents, they must be analyzed until the main cause has been located, and the efforts of the doctors must be applied to the destruction of that cause.

The consideration of railroad accidents naturally divides the subject into four parts:

1.

Accidents due to travelers on the highway.

2. Accidents to trespassers.

3. Accidents to employes in shops and along the railroad in the handling of material and tools, or in the conduct of the average every-day employment.

4. Accidents to passengers and employes growing out of collisions, derailments, and the performance of essentially railroad services required of the average train and yard men, including engine crews.

In studying the Accident Bulletin of the Railroad Commission of Indiana for the calendar year 1909, you will note that there were 147 persons killed or injured on highway crossings in the State of Indiana, resulting in 45 deaths and two serious personal injuries. Of these accidents, 17 resulted from horses becoming frightened.

It is somewhat difficult to locate the germ for accidents to travelers on the highway, or so-called crossing accidents. It is not the speed of trains, for many of the accidents happen to slowgoing trains rather than to fast ones. It is not because of the fact that there are no watchmen on the crossings or safety gates, for a considerable percentage of the accidents occur at crossings equipped with watchmen or safety gates. It cannot be said to be the failure of the railroad man to give warning of the approach of trains, for in most cases it is found that the enginemen have given the signal in the way prescribed by law. It is true that in isolated cases these accidents are partly due to the failure of warning signals or appliances, yet in many cases the warning has been properly given. Undoubtedly, therefore, this class of accident must be largely due to carelessness on the part of the public, assisted, occasionally, by similar carelessness on the part of crossing men or other railroad employes.

Our attention is often called to the fact that in Europe there are but few accidents of this kind, and the fact is proclaimed that there are no grade crossings on the railroads there. It was the privilege of the speaker to be in Great Britain in 1906, and again in 1909, at which time conditions surrounding the railroads were looked into rather carefully. As a matter of fact, there are a large number of grade crossings in Great Britain, both of single track and of double track railroads, and the same is said to be true on the Continent.

In England the traveler on the highway, in his vehicle, approaches a crossing at grade. He finds the way barricaded by substantial gates. There may be no train in sight-and usually there is not-but the vehicle stops and waits until the gate-keeper, usually advanced in years and in a greater or less state of decrepitude, comes out of the house, carefully looks up and down the line, unlocks the gate and lets the patient traveler by. It will make no difference to the gate-keeper that another vehicle is in sight approaching the crossing. The gates are again closed and locked, the gate-keeper again enters his house, and when the next traveler approaches, the same procedure is gone through with. If a passenger train be due, whether in sight or not, the gates usually remain down until the train passes, which occasionally is discomforting when the train is late and the crossing watchman or watchwoman, as the case may be, has not been notified.