mining districts of a single State. The descriptive accounts of accidents, when reexamined and retabulated, often yield results which differ more or less from the tabular presentations in the annual reports of the mine inspectors of the several States. The terms used in mining are not of precisely the same meaning in different coal fields, and often, no doubt, the reports are made by mine officials who fill out the required forms in a perfunctory manner. This criticism, however, applies more to the minor causes than to the leading causes, which are of such a nature that the liability to serious error is small.

FALLS OF COAL OR ROOF.

In the present analysis out of 18,346 fatal accidents, 2,722 were due to fall of coal and 5,828 were due to fall of roof, rock, or slate, etc. For general purposes it is rather immaterial whether the accident is due to one or the other of these causes, or to both combined, but when the facts are stated in their relation to the probable degree of safety in working different kinds of coal fields, seams of varying degrees of thickness, etc., accuracy in the descriptive account of the fatal accidents is of considerable practical importance. Combining the two classes of casualties, as being more or less equivalent terms, it appears that of all the fatalities 46.6 per cent were the result of conditions inherent in all coal-mining operations. This average is for the coal fields of North America as a whole, and wide divergencies from the average will presently be pointed out in the case of the several coal fields and the separate States. The average is the result of accumulated experience both in time and area, and while for some of the States the returns are for shorter periods than a 10-year period, the available facts for each State are fully indicated in Table XXIII of the appendix.

Accidents due to fall of coal or roof are, therefore, by far the most important single and well-defined group of fatal accidents in coal mining, and this is true not only for the United States, but for most of the other coal fields throughout the world. Exceptional disasters, due to gas or dust explosion, causing a great loss of life in a single year, must necessarily disturb the percentage distribution of the several causes responsible for coal-mining fatalities, but normally the percentage of deaths from fall of coal or roof will not vary much from year to year. The occurrence of a very disastrous accident in West Virginia in 1907, for illustration, resulted in a marked decrease. in the percentage of deaths from falls of coal, roof, etc., although the number and proportion of deaths from this cause remained about the same as in the previous year. According to the reports of the United

States Geological Survey for 1907 and 1908, fatal accidents due to falls of coal or roof were distributed as follows: (a)

FATAL ACCIDENTS IN COAL MINING IN THE UNITED STATES DUE TO FALLS OF COAL OR ROOF, 1907 AND 1908.

This table illustrates and emphasizes the very serious statistical error which may invalidate conclusions based upon percentages alone. The table shows an increase in the proportion of deaths from falls of coal or roof to have taken place in 1908 compared with 1907; but, in fact, in proportion to the average number of men exposed to risk of death the fatality rate for deaths from fall of coal or roof decreased from 1.65 in 1907 to 1.56 in 1908.

As has previously been said, both methods of statistical analysis have their use, provided they are employed with due caution and a full knowledge of all the facts which have a bearing upon the question under consideration. The liability to error is diminished in proportion as the subject is considered from the broadest possible standpoint and rather as an approximation to the truth than as a statement entitled to the claim of scientific accuracy. No such claim can rightfully be made for any of the statistical information relating to coal mining in North America at the present day, but, on the whole, it may safely be assumed that the available data for this country conform favorably in accuracy and detail to the corresponding information for the other coal-producing countries of the world. The table of principal causes may therefore be relied upon as a trustworthy presentation of the true facts for the coal fields of North America, as far as these facts are known at the present time, and the conclusion is fully warranted that the most important and determining cause in coal-mining fatalities is fall of coal or of roof, rock, and slate, as the case may be, singly or in combination with each other. This cause, or group of causes, then, is by far the most important element in coal mining as regards the safety of mine laborers. The deaths do not occur in the mass, but they take place from day to day, singly, or, at most, a few at a time, (') but in the aggregate they mount up to from one-third to one-half of all the coal-mining fatalities during the year. Occasionally a fearful calamity will cause a

• Mineral Resources of the United States, 1908, Pt. II, p. 56. U. S. Geologica! Survey, Washington, 1909.

This, of course, does not apply to "cave-in" accidents, which may cause a large loss of life at one time. See Report of the Department of Mines of Pennsylvania, 1896, p. 79.

great loss of life by a single accident, but when the casualties are considered in the mass and with a due regard to length of time in mine experience it will invariably be found that no other single cause or combination of causes is responsible for as great a loss of life in coal mining as falls of coal and roof or slate and rock, as the case may be.

MINE CARS.

In the preceding tabular summary of causes (p. 454) one of the most important causes of coal-mine fatalities is mine cars. In the aggregate of casualties at present under consideration there were 2,204 deaths from this cause, or 12 per cent of the total. The proportion must naturally vary according to the motive power employed, and the degrees of variation in this respect will be fully brought out in subsequent tables in which the facts are given in detail for the different coal areas and the separate coal-mining States. Combining the deaths from mine cars and the deaths from falls of coal or roof, it is shown that out of the total number of fatal accidents from all causes, given as 18,346, as many as 10,754, or 58.6 per cent, were due to these two particular groups of causes. In some of the coal fields and in some of the States this proportion will be found to be much greater, so that it may be stated as a broad conclusion, sustained by a whole decade of American coal-mining experience, that primarily and chiefly the causes responsible for fatal accidents in coal mining are falls of coal, rock, and roof and fatal injuries caused by mine cars.

EXPLOSIONS.

Explosions due to gas or dust, or both, caused 2,571 deaths, or 14 per cent of the whole. There is the possibility of error in a return of this kind in that the closely related mortality from explosions of powder, dynamite, and blasts, or explosions "other" and "not specified" may include deaths which should properly be charged to explosions due to gas or dust or both. When these are considered together, it appears that out of the total of 18.346, in addition to the 2,571 deaths resulting from gas and dust explosion, there were 968 deaths from explosions of powder and dynamite, 793 from explosions of blasts (which, of course, is practically the same thing), and 292 from other explosions not specified, a total of 4,624, or 25.2 per cent. Comparing this total with the aggregate of deaths due to falls of coal, roof, and rock, it appears that the result is as follows:

COMPARATIVE MORTALITY FROM PRINCIPAL CAUSES IN COAL MINING IN NORTH AMERICA.

When, therefore, all proper allowance is made for a possible erroneous classification, it appears that fatalities due to falls of coal, roof, and rock far outnumber the corresponding fatalities caused by explosions due to gas or dust, or the use, handling, transportation, and storage of explosives of all kinds.

MISCELLANEOUS MINE ACCIDENTS.

The foregoing three principal causes, or groups of causes, account in the aggregate mortality experience for 83.8 per cent of the fatal accidents from all causes. Among the most important of the minor causes come, first, the accidents due to falling into shafts, slopes, manways, etc. Accidents of this kind are due to a large variety of causes, the exact nature of which can be fully understood only after a careful study of a large number of individual cases. Badly guarded shafts and openings are probably responsible for the majority, but many of the deaths are due to falling out of the hoisting cages, or to falls while climbing in or out of the shaft when the hoisting apparatus is not working or while climbing in and out of mine shafts not provided with hoisting apparatus, etc. Falls into shafts separately account for 2 per cent of the total number of acidents, and falls into slopes, manways, etc., account for 0.7 per cent additional.

Outside car accidents caused 470 deaths, or 2.6 per cent of the total. This is the principal cause of accidents outside of mines, or overground, as separate and distinct from accidents underground. Many, if not most, of the other overground accidents are included under "miscellaneous" causes, which comprehend a total mortality of 1,105, or 6 per cent of the total. For some of the States returns in more detail are available and these will be considered later. Manifestly specific details are of particular importance in an inquiry of this kind, for large aggregates tend rather to obscure the true underlying conditions responsible for the occurrence of fatal accidents in coal-mining operations. However, when the statistical analysis is carried too far the actual numbers often become so small as not to warrant safe conclusions. "Mining machinery" is a rather indefinite term, but largely inclusive of coal-cutting machines causing fatal accidents due to mechanical causes. There is otherwise no considerable amount of working machinery, generally so called, underground. Boiler explosions occur occasionally, but in the present analysis deaths resulting from these are not separately accounted for. Perhaps the deaths from "motors" should have been included in the total of deaths from machinery, since the motors may be such as operate the coal punchers, or coal-cutting machines, or electric motors employed in underground haulage. Deaths from this group of causes numbered 30, or 0.2 per cent of the total. While electric haulage is gradually replacing mules used for haulage pur

poses underground, a large number of mules are still employed, but the deaths caused by the kicks of these animals, or otherwise in connection with the handling, driving, etc., of the same, number only 73, or 0.4 per cent. The two remaining specific causes are asphyxiation and electrocution. The former caused 271 deaths, or 1.5 per cent of the total, and the latter 193, or 1 per cent. Both of these are important causes, which for their full understanding require to be considered in detail, as disclosed by a careful consideration of a sufficient number of individual cases.

PRINCIPAL CAUSES OF FATAL ACCIDENTS BY COAL FIELDS.

The relative proportion of principal causes naturally varies widely, according to the geological character of the coal fields, the method of mining, the presence of dust or dangerous gases, the use of coal-cutting machinery, compressed air and electricity, etc. In the table which follows a comparison is made of the distribution of principal causes in the different coal fields, amplified in the appendix by tables for the several coal-mining States. It would carry the present analysis too far to discuss the variations in causes, as determined by the percentage basis in detail, since they will be further considered on the basis of the exposed to risk of death one year for the several States. (See Table XXIII of the appendix.)

PER CENT OF FATAL ACCIDENTS IN THE COAL MINES OF NORTH AMERICA DUE TO EACH CAUSE DURING A TEN-YEAR PERIOD, BY GEOGRAPHICAL SECTIONS.