such annual reports and the figures below therefore differ materially from those shown by the monthly reports.

The chief service of statistics of railway accidents is to point out to those in charge of railway operations and to the Government that class of accidents which may be lessened by greater care on the part of railway employees or greater uniformity in railway equipment and conditions of management. To render this service it is necessary that casualties should be classified according to the class of persons by whom they are sustained and according to the kind of accidents from which they result. The summaries of railway accidents in the report present in detail statements of accidents which are classified under the general heads of "Accidents resulting from the movement of trains, locomotives, or cars," and "Accidents arising from causes other than those resulting from the movement of trains, locomotives, or cars."

The total number of casualties to persons on account of railway accidents, as shown for the year ending June 30, 1902, was 73,250, the number of persons killed having been 8,588, and the number injured 64,662. Of railway employees, 2,969 were killed and 50,524 were injured. These figures show a very considerable increase in the number of employees injured, a result due in part to the unusual increase in traffic and the consequent use of all kinds of equipment and the employment of untried men, and in part to the fact that since July 1, 1901, the carriers have been obliged by law to render monthly reports, under oath, to the Commission, detailing the causes and circumstances surrounding all accidents to employees, the reports being carefully scrutinized and frequently corrected, which results in the return of numerous accidents that, if they had occurred prior to that date, would not have been reported.

These casualties were distributed among three general classes of employees, as follows: Trainmen, 1,674 killed, 21,503 injured; switch tenders, crossing tenders, and watchmen, 200 killed, 1,443 injured; other employees, 1,095 killed, 27,578 injured. The casualties to employees resulting from coupling and uncoupling cars were, employees killed, 167; injured, 2,864. The corresponding figures for the year 1901 were, killed, 198; injured, 2,768. The casualties connected with coupling and uncoupling cars are assigned as follows: Trainmen killed, 141; injured, 2,475; switch tenders, crossing tenders, and watchmen killed, 17; injured, 285; other employees killed, 9; injured, 104. The casualties due to falling from trains, locomotives, or cars in motion were, trainmen killed, 371; injured, 3,821; switch tenders, crossing tenders, and watchmen killed, 40; injured, 276; other employees killed, 80; injured, 570. The casualties due to jumping on or off trains, locomotives, or cars in motion were, trainmen killed, 78; injured 2,681; switch tenders, crossing tenders, and watchmen killed, 12; injured, 203; other employees killed, 50; injured, 452.

The casualties to the same three classes of employees from collisions and derailments were, trainmen killed, 534; injured, 3,350; switch tenders, crossing tenders, and watchmen killed, 12; injured, 70; other employees killed, 87; injured, 640.

The number of passengers killed during the year was 345 and the number injured 6,683. The corresponding figures for the previous year were 282 killed and 4,988 injured. As a result of collisions and derailments 170 passengers were killed and 3,429 injured. The total number of persons other than employees and passengers killed was 5,274; injured, 7,455. These figures include casualties to persons classed as trespassers, of whom 4,403 were killed and 4,854 were injured. The total number of casualties to persons other than employees from being struck by trains, locomotives, or cars were 4,021 killed and 3,973 injured. Casualties of this class occurred as follows: At highway crossings-passengers killed, 3; injured, 9; other persons killed, 824; injured, 1,326; at stations-passengers killed, 29; injured, 382; other persons killed, 343; injured, 482; and at other points along track-passengers killed, 7; injured, 19; other persons killed, 2,815; injured, 1,755.

The summaries giving the ratio of casualties show that 1 out of every 401 employees was killed, and 1 out of every 24 employees was injured. With reference to trainmen—including in this term enginemen, firemen, conductors, and other trainmen-it is shown that 1 was killed for every 135 employed, and 1 was injured for every 10 employed. One passenger was killed for every 1,883,706 carried, and 1 injured for every 97,244 carried. Ratios based upon the number of miles traveled, however, show that 57,072,283 passenger miles were accomplished for each passenger killed, and 2,946,272 passenger miles accomplished for each passenger injured. The corresponding figures in these latter ratios for the year ending June 30, 1901, were 61,537,548 and 3,479,067 passenger miles for each passenger killed and each passenger injured, respectively.

COLLISIONS-THE BLOCK SYSTEM.

The most prominent fact in this year's record of train accidents is the appalling loss of life and property in collisions. The showing can not be called materially worse nor materially better than that made a year ago, and the total number of passengers killed and injured does not differ from the record of the preceding year to such extent as to call for comment; but the mere continuance of the record, though it be no worse than in former years, is so painful and alarming as to excite the indignation of the public.

That 130, or 118, or any large number of passengers are killed in the United States in the course of a year in a single class of accidents indicates a condition which should not pass without serious attention. These fatalities are due to causes which have never been adequately

considered by any department of the Government, either of the United States or of any of the States, and there is urgent need for such consideration. Railroad accidents, their causes and their results, have been considered in judicial decisions and in the deliberations and verdicts of coroners and coroners' juries, and to a very limited extent by State railroad commissions; but none of these has dealt comprehensively with the subject, and apparently no improvements in railroad service or reformatory measures of any kind have been accomplished by these means.

DESIRABILITY OF THE BLOCK SYSTEM.

That collisions inflict the most pitiable sufferings on the passengers and trainmen who are their victims, and the most heartrending experiences on those who escape, is too trite to be enlarged upon; that the distress is in many cases no worse than that produced by bridge failures, by shipwrecks, and fires in large buildings is obviously true; that railroad companies make financial reparation to passengers who are injured and to the heirs of those who are killed is admitted; that railroad officers sincerely deplore the existing situation is well known; that many railroads have equipped some of their lines with block signals, a measure which greatly reduces the danger of collisions, is much to be commended. But the painful fact remains that disastrous collisions occur frequently. Some important railroads have not adopted the block system. Most companies use it on parts of their lines, but not on other parts. Some use it a part of the time or for a part of their trains. Some adopt the principle, but have insufficient regulations. There occurred in the United States in the year under review the enormous number of collisions shown in the Appendix, viz, 5,219, and included in these were 10 collisions in each of which 7 or more persons were killed, a total of 104 persons in the 10 cases, all on lines not using the block system.

These ten collisions, with one occurring where automatic block signals were in use (Item 6), are listed below," the reference by number being to the table of causes in the Quarterly Accident Bulletins. The circumstances of the sixth case were explained in Bulletin No. 7.

Nevertheless, as we believe, the problem of reducing the annual collision record by a very large percentage is comparatively simple. It may be dealt with successfully within a reasonably short time, without inconvenience to the public and without causing serious financial burdens. So far as we can see, the numerous railroad companies which have introduced the block system have adopted the proper remedy, and their action should be imitated by other companies. England uses the block system universally, and the immunity from collisions on English railroads is so nearly perfect, and the casualty records there so low, as to be a powerful argument in favor of its adoption. The Commission therefore recommends the consideration of a law like that in force in Great Britain and Ireland, requiring the adoption and use of the block system in the United States; and a tentative draft of a measure for the purpose of aiding discussion is given in the appendix.

This proposed bill has been drawn on the theory that the expense necessary for the construction of new signals, or for electrical wires, or apparatus necessitated by the use of the block system, as well as the. increased expense for wages of signalmen, should be distributed over a term of years; and it is suggested, therefore, that each railroad company be required to adopt the block system on one-fourth of its passenger lines by January 1, 1906; on another fourth by January 1, 1907; another fourth one year later, and on the whole by January 1, 1909. Under this plan many of the principal lines would be required to make no important additions to their expense accounts for the first two years, and some would feel no burden for the first three years. For the purpose of dealing with separate parts of an extensive system of railroads, a section has been included enabling the Commission to deal with one part of a company's lines independently of the other parts. It is a well-known fact that some companies use the block system on certain of their lines, while on certain other lines over which a large amount of passenger traffic is carried, the older and less safe methods are still retained. Sections 4 and 5 of the proposed bill provide for the introduction of the block system on the more important lines of the country in two years less than the term just mentioned. This requirement is deemed reasonable, because of the fact that many of these railroads already use, on a good portion of their lines, block signals which substantially comply with the requirements of the suggested statute. Section 5 provides for roads on which the passenger traffic is very light, but on which the number of freight trains is comparatively large. This is necessary because passenger trains suffer not only from collisions with other passenger trains, but from collisions with freight trains.

DEFICIENCY OF INTERLOCKING.

A difficulty in the use of the block system on many thousands of miles of railroad in America is found in the deficient equipment of

outdoor fixed signals—a deficiency in what is termed "interlocking "on these railroads. The railroads of America, unlike those of England, have introduced the block system (for maintaining a station to station space between trains following one another) while not adequately equipping their stations, junctions, and crossings with suitable appliances for regulating the movements of trains at and near stations. As a large portion of the train-accident losses sustained by railroads is due to collisions and minor mishaps occurring at and around switches, stations, and yards, this is a serious deficiency.

The importance of the matter may be appreciated by consideration of the British law, which requires practically the same degree of progress in interlocking and concentration of point (switch) and signal levers as in the block system. In that country the block system is not deemed complete without the other improvement as a corollary. But to require such a radical improvement in the equipment of American railroads, even if the cost were to be spread over so long a period as five years, would necessitate very unusual expenditures on the part of the companies, and the Commission is not prepared at this time to recommend the requirement of such an outlay. The block system, however, is an efficient means of safety, even under the imperfect conditions necessitated by the absence of interlocking and concentration of switch or signal levers. This is proved by the experience of the past ten years on railroads in all parts of the United States.

The absence of suitable fixed signals at stations and yards necessitates more complicated and less satisfactory regulations for the guidance of the trainmen in the management of trains under the block system, and this additional complication is objectionable; but though an evil it yet is a lesser evil-one decidedly less than the evils which appear to be inherent in the present system of train management, based on time intervals. Section 9 of the proposed act therefore provides for recognizing this incompleteness of appliances by prescribing a regulation for limiting the speed of trains at places where the signaling is not adequate for high speeds.

Any statute requiring the use of the block system should be so drawn as not to apply to railroads, or parts of railroads, over which the traffic is wholly freight, and the draft which is proposed is thus drawn. At the same time it is to be remembered that a considerable number of stock drovers, and other persons riding on freight trains to care for freight, are injured in accidents every year. These men usually ride in the caboose, the car occupied by the trainmen, and a large part of their journey is usually in the night, and the men sleep as they ride. This is a branch of railroad traffic which is not unworthy of attention. The proposed statute also exempts roads or sections on which there is usually only one locomotive at work, and roads on which passenger trains run only every other day.