February 18, 1916.	THE COLLIERY GUARDIAN.	325

MINE ACCIDENTS AND UNIFORM
RECORDS.*

By Albert H. Fay, U.S. Bureau of Mines.

(Continued from page 267.J

Uniform Time Basis for Computing Fatality Rates.

The customary basis for computing fatality rates is on
the number of men employed in the industry. This
basis is faulty to the extent that, under existing condi-
tions, it is not possible to obtain the exact number of
' men actually at work, for the reason that not all the
men work full time. There are always a number who
work a few days at one mine, and either quit, or arc
discharged, and go to other mines. The actual number
of men on the pay roll is, therefore, much higher than
the number of men at work in the mines; and unfortu-
nately it is the former figure that the operators, too
frequently report. It is also imperfect in. that no dis-
tinction is made concerning the number of working hours
per. day. In many of the mines the eight-hour day pre-
vails, while in others nine or 10 hours still constitute a
legal day.

The comparison of fatality rates on a percentage basis
is far from being correct. For example, there may be
1,000 fatalities due to various causes in a certain group
of mines, 50 per cent, of which accidents- may be attri-
buted to some one cause, as falls of rock, and 25 per
cent, to a serious disaster, in which 250 men are killed.
The defect with the percentage basis is that the number
killed in a large disaster enters into the total number of
men killed for that particular group of mines. When
the total is thus abnormally increased, it decreases the
percentage in all the other causes, to balance the excess
which is due to the one large disaster. In the case
above cited, in which 25 per cent, of the fatalities were
due to a mine disaster, the elimination of this particular
accident from the total would leave only 750 fatalities,
and although there were the same number of fatalities
due to falls of rock, which hi the first case was 500, or
50 per cent., the percentage of falls of rock in the latter
case would be increased to 66J per cent. The actual
number killed by falls of rock remains the same, and
roof conditions have not changed.

In the coal mines of Utah and New Mexico, the per-
centage of fatalities due to falls of roof-is, respectively,
26-76 and 28-17, indicating exceedingly safe roof condi-
tions in both States, being about one-half the average
for the United States. These low percentages for falls
of roof are each due to one large disaster in each State,
wherein more men were killed at one time than by all
other causes during 21 years. The fatality rate due to
falls of roof, for this period, based on the number of
men employed, in New Mexico is 3-22, and in Utah 2-48
per 1,000, both of which are much larger than the
average for the United States, which is 1-55 per 1,000
men employed. Here is a concrete example in which
the fallacy of making comparisons by causes on the
percentage basis is shown.

A comparison on a tonnage basis is also open to
criticism in that mining conditions are not the same in
all mines, States, or countries. One mine may be
operated in a thick coal bed or an ore body that is large
and massive, and another, in a thin bed or vein. It is
evident that the man working in the large ore body or
thick coal can produce many more tons in a year than
can the man working in the narrower vein. But the
amount of ore or coal produced also varies with the
mining method in vogue, which is largely governed by
the type of deposit. The fact that the man in the
large ore body, or thick coal bed, can produce more tons
per year than his fellow worker in the narrow ore body,
or thin bed, shows a lower fatality rate on the tonnage
basis, although the two men were exposed to the mining
risk the same number of days per year. As a matter of
fact, the narrower ore body and thin coal bed are the
safer of the two, while on a tonnage basis, they would
be rated as more hazardous.

For certain purposes, each method has its champion,
but for true hazard rates, that one serves best which
takes into consideration the number of men and the
time they were engaged in a hazardous occupation.

In the comparison of mine accidents, the time element
should be taken into consideration. The length of time
during which miners are exposed to the various hazards
of the industry varies in different States, due both to
commercial requirements and labour organisations. The
number of hours per day is largely regulated by labour
organisations and State laws, and the number of days
worked per year is regulated by labour and market
conditions. When the demand for mineral products is
slight, the output is curtailed by working part time or
by closing the mines entirely for several weeks, awaiting
better markets.

The ideal method would be for the operators to report
the actual number of hours’ labour for which wages
wore paid during the year. This divided by 3,000 gives
the equivalent number of men who were exposed to the
mining hazard in 300 10-hour shifts. The standard
year of 3,000 hours can easily be converted, for com-
parative purposes, to a year of 2,000 hours, which more
nearly conforms to existing conditions in the coal mines
of the United States. Complete data along the lines
above outlined are not as yef available for the metal
mining industry of the United States, but it is hoped
that as compensation laws are enacted,'and State mine
inspection laws arc revised, that these matters will be
taken into consideration.

Data of this character are available for a 10-ycar
period in the coal mines of the United States, in which
data the varying hours, number of employees and the
working days have been given proper weight, with the
result that the total number of working hours in each
State , has been computed for the 10-year period, and

* From a paper read before the Second Pan-American
Scientific Congress, held in Washington, on Dec. 31, 1915.

their sum gives the total number of hours worked in
all the States. This divided by the number of men
reported working gives an average of 1,909 hours for
the United States as a unit.

The number of working hours per year, over a period
of 10 years, in the coal mines of each State varies from
1,314 in Arkansas to 2,447 in Virginia. The Pennsyl-
vania anthracite field is- 1,985 hours, and the
Pennsylvania bituminous mines 2,034 hours, per year.
In view of the fact that this range varies from a
few hours below to a few hours above 2,000, and that
the average for the United States is 1,909 hours, a 2,000-
hour year has been adopted for these calculations. It
is more nearly in keeping with the actual condition in
■the coal mines than it would be to base the rates on a
3,000-hour year, as is frequently done in factories,
metallurgical works, and other industrial plants. The
working hours in each State divided by 2,000 gives the
equivalent of 2,000-hour workers, upon which fatality
rates may bo based. The number of calculated 2,000-
hour workers is given in parallel with a column showing
the number of employees as reported. Anyone desiring
a 3,000-hour basis may obtain it by adding 50 per cent,
to all of'the figures in the table.

The actual average number of days that the mines are
operated each year varies from year to year in the same
State. For example, in 1906, Colorado averaged 208
days, while in 1911 the mines were operated 207 days.
Arkansas, in 1903, operated 223 days, while in 1910, only
128 days. Ohio varied from 161 days in 1908 to 206
days in 1913. With such wide differences in the number
of days that the mines arc operated, it is evident that
the hazards in each State arc not the same.

It is certainly not fair to place the fatality rate of an
eight-hour State on an equal basis with a 10-hour State,
when the latter men are exposed to the mining risk 25
per cent, longer. Every hour adds 12^ per cent, to the
hazard, as compared with an eight-hour day. As an
example of unfair comparison may be cited Ohio and
Montana. The fatality rate for Ohio coal mines is 2-94
per 1,000, based on the actual number of men employed,
regardless of the time element; and the rate for Montana
is 4-08 per 1,000. When the rates are reduced to a
common basis of 2,000 working hours per year, the Ohio
rate becomes 3-94, and the Montana rate 4-23, so that
the difference on the 2,000-hour basis is 0-29, as com-
pared with 1-14 per 1,000 on the customary basis. The
same comparison may be made for other States. In
Virginia, many of the coal mines are worked 10 hours
per day, and 280 days per year, thus increasing the time
of exposure, and giving, on the customary basis, a
fatality rate of 6-47, which, if reduced to the 2,000-hour
year, becomes 5-26, comparing favourably with the coal
mines of 'West Virginia, at 5-18 per 1,000. While the
above figures are based on coal mine experience, the
principles which they bring out are applicable to any and
all metal mines, metallurgical plants, and quarries in the
Pan-American countries.

Prevention of Mine Accidents.

In a brief paper it is impossible to give details of the
various means of accident prevention. As a general
statement, it may be said that there should be more
stringent laws concerning the operation of mines; rigid
rules and regulations on 'the part of the mining com-
panies ; a strict enforcement of the laws and regulations
by State inspectors, operators, and employees, with
penalties for all who violate them, whether he be
operator or employee. The education of the miner to
realise the dangers under which he works will have
much to do with accident prevention. A common
language, that is, a language understood by both fore-
man and miner, is of prime importance. Improved
safety appliances will reduce the dangers incident to the
industry. Among these may be mentioned the general
use of safety lamps; the marking and guarding of all
dangerous places, as shafts, winzes, raises, and other
openings in the mines; a systematic inspection of all
working places for gas, ventilation, and roof conditions;
safeguarding of machinery and electric wires; the
employment of shot-firers, or, better yet, electrical firing
apparatus, and shootings only when all men are out of
the mine; the use of permissible explosives; the co-oper-
ation of State and Federal organisations with operators’
associations, labour organisations, operators, and
employees; and, last, but not least, eternal vigilance on
the part of all.

(To be continued.)

Dutch East Indies Coal.—The coal in the Dutch Indies
is all “ young,” and of brown-coal character. It is roughly
divided into three types :—(1) The oldest is a hard, bright
bituminous coal, with a heating power exceeding 6,000 h.u.
(2) A younger type, softer, dull, and dark brownish-black in
colour, contains more water and much bitumen ; burns with
a very long flame, and is therefore unsuitable for steamship
or railway use; the ash content varies considerably, from
below 1 per cent, to over 20 per cent.,- and the caloric
value between 3,000 and 6,000 h.u. (3) Bituminous shale,
in thin layers and forming whole stretches of hilly ground;
hitherto considered quite worthless. Only type (1) has been
exploited to any appreciable extent. The Ombilien and the
Poeloelaoet coals are of this class. Apart from the possi-
bility of increasing the output of these mines (that of the
Ombilien has been about 400,000 tons annually of recent
years, and that of the Poeloelaoet mine was 110,000 tons
in 1914), the question is vdiether the method of working
can be improved. Similar coal occurs in Java, but mostly
in thin beds alternating with thicker beds of shale coal.
The thickness is irregular, generally less than 40 in., and
the workings are small. Bituminous coal is also found in
Bembang, New Guinea, etc. Type (2) is much more
com-mon, particularly in Sumatra and Borneo, wdiere it is
chiefly found in close relation to the petroleum formations.
In boring for oil many coal seams are pierced, equivalent to
a total deposit of about 4,000,000 tons per square kilometre
in East Borneo. In Java, brown coal seams over 40 in.
thick are found in numerous places, and though unsuitable
for steam raising, is of value to the local brick and tile
works.

THE CONNECTION BETWEEN THE
N.W. EUROPEAN COAL FIELDS.

As an appendix to his paper (Colliery Guardian,
February 11, 1916, p. 263) read before the Manches;er
Geological and Mining Society, Prof. X. Stainier made
the following remarks.

The numerous dales- w-hich intersect the outcrops of
the northern and eastern margin of the Lancashire coal
field afford an unequalled opportunity of studying the
strata between the Middle coal measures and. the
carboniferous limestone. The classification of these
deposits, made as the result of the observations of tw’o
generations of geologists, has become classical, and was
adopted by Hull in his classification of the carboniferous
deposits of the British Isles. Among the divisions
drawn in these carboniferous rocks is that of the Lowmr
coal measures. Taking, first, the most persistent
particulars exhibited by these Lower coal measures of
Lancashire, these present, in ascending order, the
following sequence :—(1) Barren measures, with grits
and occasional coal, some marine beds; (2) productive
measures, with one to three thin but persistent coals.
The Middle seam has sometimes a fresh woater or
brackish fauna (Carbonicola) in the roof. Twm other
seams display marine fauna in their black shale roof,
especially the uppermost coal (bullion coal) where roof
■of shale with limestone, nodides (bullions) is crowded
with marine shells, Pterineopecten and Goniatites being
especially conspicuous.

(3) Barren measures, with imp ops is tent thin and
unworkable coals, and near the top a persistent series of
flagstones or grits; (4) Middle coal measures with Arley
coal at the bottom. This regular coal seam is note-'
worthy for it-s roof, wdiich contains abundant fish
remains. E. Hull w-as of opinion that the Low’cr coal
measures are the upper limit of marine Carboniferous
fauna. The inaccuracy of this statement is known, and
the utility and limits of the Lowor coal measures division
have been questioned. There seem to be good reasons
for retaining the limits of this division as drawn by the
Lancashire geologists. The same sequence and parti-
culars hold good in the coeval beds on the eastern wing
of the Pennine chain in Yorkshire and Derbyshire. From
the description of the Cheadle coal field by G. Barrow,
it w-ould seem that, owing to favourable outcrops, one.
may observe there quite the same particulars in the
Lowber coal measures, and the same fauna. Long ago
the author studied, with the late Mark Stirrup, some
typical sections of the Lancashire Lowrnr coal measures,
and wras able to identify the minute particulars he had
observed in Belgium. Taking into account the distance
betw’een Lancashire and Belgium, and the fact that the
coal fields of both countries do not belong even to the
safne syncline, the identity of their lithological, faunal,
and stratigraphical characters is really vnnderful, and
one of the strongest evidences of the former extensive
connection between the British and Continental coal
measure areas. The author has been able to follow
the Lower coal measures all around the Belgian coal
field over hundreds of miles, and they ahvays display
the same characters. They had been found by W. C.
Klein in the Dutch Limburg and in the Wurm coal
fields, and they are known to exist in Westphalia. A
coal seam, presenting all the characters of the Arley
seam of Lancashire, is found at the bottom of the Middle
coal measures in the Belgian, Dutch, and Wurm coal
fields. In the North of France, C. Barrois had already,
notwithstanding the lack of outcrops, detected the car-
boniferous level of the famous bullion coal. In the
British Isles the Lower coal measure rocks and fossils
are still unknowm in some coal fields, but in some cases,
only on account of lack of observations. A fair example
of this is afforded by the Bristol coal field, w-hich was
formerly noteworthy for its scarcity of faunal horizons
(especially of marine horizons). No sooner was H.
Bolton able to undertake the study of this coal field than
he detected many fossiliferous levels of every kind.
From the two papers he has already published on the
question, it is easy to grasp that typical Lowor coal
measures have been found in the British coal field rest-
ing on millstone grit, and displaying the same sequence
and fauna as in Lancashire. From the descriptions of
the South Wales coal field published since the re-survey
by J. Salter, the author has little doubt that the Lower
coal measures are also present in this coal field. In
conclusion, it may be said that a division winch displays
such a persistency of characters over such an extensive
area is woll worth keeping. As to the utility of these
rocks being separated from the others, it may be said
that the Low’er coal measures, by their bareness,- are
the true Farewrell rocks. If the Low7er coal measures
are not, as E. Hull thought, the upper kind of marine
fauna in the carboniferous, nevertheless, marine beds
are the rule in this division, and fresh w-ater beds the
exception; wdiereas in the Middle .coal measures the
contrary obtains. A noteworthy difference between both
divisions is the scarcity of plant remains in the Lower
coal measures—a fact w7hich is correlative of the abund-
ance of marine kinds. As stated already, the Middle
coal measures are really the period of coal formation;
and by their faunal and lithological characters the
Lowber coal measures are more allied to the millstone
grit than to the Middle coal measures. The author
acknowledges the hearty assistance he has enjoyed in
the difficult conditions under which he wwote his paper.
His gratitude is especially due to Dr. A. Strahan,
Director of the Geological Survey, through whom he
got the necessary books from the London libraries; to
Mr. Lewis Abbot, of St. Leonards, wdio put his rich
library at the author’s disposal; to Mr. Buttemvorth,.
wdio allowed him the use of books from the reference
library at Hastings; and to Messrs. Gibson, N. Arber,
and H. Bolton, w7ho- supplied him with maps and books
of their own. He is also very grateful to the hon.
secretary of the Manchester Geological and Mining
Society (Mr. N. T. Williams).