It will readily be perceived that the above-mentioned rule announced by the court of appeals is in conflict with that construction of the statute, and that it would, if finally affirmed, permit the use of many different types of couplers, each type working automatically with those of the same description, but not with any other, and authorize the hauling of cars in the same train which could not be coupled or uncoupled without the aid of men going between the cars. The case was probably presented to the court with main reference to facts relating to the plaintiff's claim for damage, and with but little attention to the great interest of railway employees and the public generally in an interpretation of the law involving the hauling of cars in the same train not equipped with couplers coupling automatically with each other.

That a remedial statute which has proved of such benefit to a large and important number of citizens should be rendered nugatory by a decision in a case brought by an individual to recover personal damages without the Government's representative being heard upon a proper construction of the statute is unfortunate, to say the least, and an effort will be made to have it properly presented.

The validity of the statute of the State of Iowa which makes a railroad corporation liable to all persons, including employees, for personal injuries due to negligence of its servants, has been affirmed by a decision of the United States circuit court for the northern district of Iowa (O'Brien v. C. & N. W. R. Co., 116, Fed. Rep. 502). An express messenger was killed in a derailment; and in the defense made by the railroad company, in a suit for damages, it was averred that no liability existed, because the messenger, O'Brien, had in writing released the express company and the railroad company from liability for injury or death resulting from the gross or other negligence of the railroad or its servants. The court holds the contract to be invalid under the statute, thus making the railroad company liable for the death of an express messenger when such death is due to the negligence of an employee of the railroad.

ACCIDENT REPORTS.

The act of March 3, 1901, requiring common carriers engaged in interstate commerce by railroad to report accidents monthly to the Commission, has been in operation since July 1, 1901, and the Commission has issued four quarterly bulletins (for the four quarters of the fiscal year ending June 30, 1902) giving summary tables of the statistical information derived from the railroad companies' reports, and particular information concerning the causes of some of the more serious accidents."

a See Appendix for tables showing totals for the year.

So far as casualties to persons are concerned, this law does not require any statistics which were not before required in the railroad companies' annual reports; the new requirement is that the causes shall be reported in detail. The two classes of accidents to employees which have for years received the most attention, because of their importance, are "coupling and uncoupling" and "falling from cars or engines." These have been made the subject of Tables Nos. 3 and 4, in each quarterly bulletin, classifying the causes in as much detail as was practicable. These tables serve to put on record useful information which, until their appearance, was to a considerable extent matter of conjecture, or concerning which different observers held conflicting views.

In Table No. 3 the causes of coupling accidents are divided into 21 classes; and this table, published quarterly, makes it possible to see what proportion of the accidents reported occur by reason of the use of couplers which are not automatic, as in coupling an engine to a train (the law not explicitly requiring automatic couplers on engines and tenders); when coupling with link and pin, as has occurred on certain narrow-gauge railroads; when using the hands in connection with an automatic coupler because it is not in working condition; when going between the cars to uncouple because the uncoupling rod is out of order; and other common causes. As in many other classes of accidents, a considerable proportion of these cases have to be classified as "Not clearly explained." In the table for the twelve months of the last fiscal year 29 out of 143 killed and 427 out of 2,130 injured are thus classified and unexplained.

Many men are injured while opening by hand the knuckle of a coupler on a moving car, when such car is dangerously near to another car. A considerable share of these cases is due to the negligence of the man himself, or to his lack of caution. There is a small but constant percentage of injuries occurring when a man undertakes to adjust a coupler laterally by pushing it or kicking it with the foot. This is often done when the car is in motion, and is manifestly a risky proceeding. 'Going between cars on the inside of a curve is an explanation which frequently occurs. This, in nearly every case, must be taken to indicate contributory negligence, for the added danger of going between cars on the inside of sharp curves as compared with the situation on a straight track is obvious.

66

وو

In Table No. 4 the reports of accidents due to falling from cars are divided into eighteen classes. Here again it is necessary to group a large percentage under the head of "Not clearly explained;" but in these cases such a statement is often necessary. Sometimes in fatal accidents there are no witnesses, and in those which are nonfatal the victim himself may be unable to explain the circumstances.

Mention might be made in this connection of the danger attending the movement of the later designs of large-capacity cars in respect to their increased width. This reduces the clearance between tracks to such an extent that it is often difficult for a man to pass between the cars. Some of our inspectors say that they hear many complaints regarding this. While it does not, perhaps, properly come under the head of safety appliances, its relation to accidents to yardmen warrants a reference to it.

In the matter of collisions and derailments, the statistics which have been gathered under the law of 1901 constitute the first authentic record of the kind which has ever been published relating to the railroads of the whole country. The tables that were published for many years by the Railroad Gazette were incomplete and unofficial. The figures which have been compiled and published by the Commission for the twelve months ending June 30, 1902, show approximately two and one-half collisions and one and eight-tenths derailments per 100 miles of railroad per year, and the losses by these accidents, not including damage to freight or sums paid to persons for bodily injuries or on account of death, average, roughly, $3,800 per 100 miles of road annually. There is nothing to indicate that in the number of accidents or in their cost the showing per mile is materially worse than it would have been in many former years, but the publication of official data has drawn renewed attention to the subject. In three of the four accident bulletins that have been issued the magnitude of the losses and of the number of casualties due to collisions, and the urgent need of some thorough-going action looking to the prevention of this class of accidents has been emphasized by the publication of notes concerning the causes of specific collisions.

In Bulletin No. 2 there was given a list of the 27 most serious butting collisions occurring in the quarter ending with December, 1901. In these collisions 70 persons were killed and 234 were injured, and the money loss was $306,511, this sum being more than five-eighths of the damage caused by all of the butting collisions of that quarter. The principal causes occurring in this list are: Forgetfulness on the part of conductors and enginemen, who run their trains past stations at which, according to written orders in their pockets, they should stop; to overlooking one out of a number of orders which they hold; to mistakes in reading hours or names in written orders; to misreading watches or miscalculating time; to misreading time-tables and train registers; to carelessness in identifying trains at meeting stations, and other errors. Besides these errors of the men on the trains, there are mistakes by train dispatchers in issuing telegraphic meeting orders and by telegraph operators at stations in receiving, copying, sending, and delivering telegraphic orders which are sent by the dispatcher to the men in charge of trains and are repeated back to the dispatcher,

Collisions also occur by reason of complications following deliberate neglect to carry out certain safeguards, the negligent employees, like a bank clerk who "borrows" from his employer expecting to repay the loan, being unable to foresee the results of his omission of a seemingly unimportant duty.

In Bulletin No. 3 a similar table was given of 41 rear collisions, this number including in that class all which caused more than $2,000 damage each to property. In this class the causes are mostly "neglect in flagging" and approaching stations or water tanks at unduly high speed. Of the 41 cases, more than one-half—that is, 26—were due to one of these causes or to a combination of the two. Other causes were: Misplaced switches, neglect of the five-minute interval rule, engineman falling asleep, runaways on steep grades, and other negligence of the men in charge of the trains. Besides these there were 5 cases on lines where the block system is in use, 3 of them due to misconduct or neglect of signalmen, 1 to failure of an automatic signal, and 1 to misconduct of the engineman. As the remedy most generally approved for the great majority of these collisions is the adoption of the block system (which is discussed below), it is proper to remark here, in passing, that these 5 cases, which occurred not withstanding the fact that the block system was in use, do not materially affect the argument in its favor. In the first place, the block system, even when imperfectly administered or applied, is universally agreed to be superior to the timeinterval system; and, secondly, the many investigations that have been made into collisions occurring on lines worked by the block system have shown in nearly every case that the error or errors leading to the collision have been due to palpable deficiencies in regulations or in discipline, rather than to any fault in the system.

In Bulletin No. 4 a list was given of the 26 most costly collisions in that quarter (to June 30, 1902), including all classes-rear, butting, and miscellaneous. This number (26) represents cases involving damage of $5,000 or over, each, those below $5,000 being excluded to bring the list within readable compass.

The totals of this list are: Persons killed, 30; injured, 187; damage to cars, engines, and roadway, $228,597. This sum is a little more than one-fourth of the aggregate damage thus reported for all of the 1,094 collisions occurring in that quarter.

The enforcement of the regulations, the neglect or violation of which has caused these collisions, has long been the subject of extended discussion among railroad officers, but it can not be said that these discussions have resulted in marked improvement, except in so far as they have led to the adoption of the block system. Without the block system the movement of trains in safety depends, except where the speed of trains is very low, on the invariable execution of rigid rules regarding rights of superior trains over inferior trains, H. Doc. 181-5

on the correct calculation of time (on single track lines) at meeting stations, and on the readjustment of the relative rights of trains when one or a few trains (out of a large number) are delayed. But these rigid rules are disobeyed, or are inefficiently administered, as appears from the statement of causes just given, and the result is a large number of collisions every month.

No attempt has been made to compare the totals of one month or year with another, or of one railroad or section of country with another railroad or section, for the reason that no comparison would be useful except for long periods. Indeed, an extensive railroad, or the railroads of a whole State, may be free from disastrous or notable collisions for a whole year, or even longer; so that from a superficial view the conclusion might appear to be warranted that defects in methods or practice which formerly led to collisions had been done away with. Or, it might be claimed that the infrequency of serious collisions, as compared with the total number of possible chances of a collision, afforded ground for the conclusion that the rules now in force, which are very nearly uniform throughout the country, provide for all reasonable precautions, and that those collisions which do occur, from the negligence or incompetence of the men who are charged with the safety of trains, should be classed as due to those indefinable infirmities of the human mind which elude rules and discipline. It is generally conceded, however, that under the block system collisions are very infrequent as compared with the number occurring under the time-interval system. The block system is now universal on the passenger lines of the railroads of England, Scotland, Ireland, and Wales, and the high degree of safety enjoyed by the passengers on those railroads is well known. In the fifteen months ending March 31, 1902, as appears from the accident records published by the board of trade, not a single passenger was killed in those countries by a train accident.

The block system is in use on about 25,000 miles of railroad in the United States, as appears from statistics published by the American Railway Association in April last. This is only about one-eighth of the whole railway mileage of the country, but it embraces many of the most important lines. The system has been introduced gradually during the past fifteen or twenty years. In many cases its introduction on a particular line or division of railroad has followed soon after a disastrous collision on that line, a fact which is quite significant. The term "block system" means simply a method whereby, by the use of the telegraph, telephone, or electric bells, or by automatic apparatus, each train is prevented from leaving a certain point until the last preceding train has passed beyond a certain point farther on. It is introduced primarily for the purpose of preventing rear collisions, though where it is desired to run trains one after another very frequently the