September 17, 1915.

THE COLLIERY GUARDIAN.

569

Coal entangled with sandstone, and coal associated with
clod, behave mechanically as if the coal were two
different materials; and the writer suspects that under
shearing stress its mechanical properties are intermediate
between the two classes of rock—much more yielding
than sandstone, but harder and more brittle than clod
or fireclay. Associated with sandstone coal is the usual
material which fills the cracks and fissures already
described, and among massive sandstones, as displayed
in section, ribbons of coal tailing out in ragged streamers
can be followed for many yards, usually along, occa-
sionally across, the bedding, ending in the most hair-
like of narrow cracks. By increase of the relative pro-
portion of coal to sandstone, blocks of rock appear as
if enveloped in a matrix of coal. In this case also
tapering films of coaly substance penetrate, to the thin-
ness of thin paper, any fissures that there may be in the
rock. With further increase in the proportion of coal
to sandstone, any appearances which might suggest a
solid sandstone invaded by plastic or liquid coal are
lost, and one-time plates of rock, subjected to end
pressure, but supported at the sides by solid though
yielding coal, have decreased in length and have become
puckered in the process. Among weak rocks, which in
most cases are dark in colour, the outlines of coal
inclusions are less easy to observe. Generally, any
large mass that has preserved its individuality is a com-
pound fragment, in which clod or holing dirt has failed
to separate from the coal, and such pieces as a rule
have bi-convex outlines and tail off as a spindle at each
end. Smaller fragments occur as angular splinters with
concave faces ending in cusps, and often, where on one
side these adjoin pieces of stronger material, tapering
wedges or films of coal persist bounded on that side by
the stronger material and link up with other sharp-
angled fragments which are associated by the side of
similar surfaces. Thin plates of coal not infrequently
occur strung out along the surfaces where the lenticles
of bind adjoin, but the writer has not discovered any
signs of slip along the boundaries of the coal. On the
contrary, wherever plates of coal show signs of end
pressure, they have become puckered and buckled after
the manner of the sandstone plates previously described.

14.	As already mentioned, the majority of the rock
chunks as they lie in the rock fault jumble can be
recognised as pieces which have been detached from
an already consolidated parent rock. As- fragments,
most of them show traces of the original planes of
deposition, which in the sandstones (as often as not)
are of cross- or current-bedded type. The usual altera-
tions of coarser and finer sediment also show up plainly,
and as an extreme case of this, bands of sandstone with
true pebbles, almost always of rounded ironstone, are
not infrequent. Fossils within the pieces are not
unusual, and are interesting as affording a true index to
the bedding planes, and a trustworthy indication of the
amount of mechanical distortion which has been suffered
by the pieces of rock in which they occur.

From a careful study of the lie of the bedding planes
in some thousands of rock chunks as they show in the
rock-fault exploration headings at South Kirkby,
Normanton, and elsewhere, the writer has arrived at
the conclusion that in these rock-fault aggregates the
rock fragments are hardly ever upside down. Often,
especially with large pieces, the bedding planes are
reared up as if a wedge had been pushed under the rock
mass, or as if the rock mass itself had been driven
forward against or over a withstanding obstacle. In
these cases, however, the bedding planes seldom or
never reach vertically, and the only examples of real
inversion known to the writer are in binds of medium
strength which underlie the upper member of a dupli-
cated coal seam in the marginal thickened strip of a
rock fault at Altofts Colliery, and have evidently been
bent back on themselves as that upper seam was driven
forward. (This case is illustrated in fig. 2, section B-C.)

15.	In normal coal measure ground, quite one of the
most striking characteristics is the regularity of the
jointing of the rocks, and the persistence with which the
joints maintain a direction over considerable tracts of
country. In rock-fault fillings (with the very important
exception noted below) regularity either in frequency
or in direction of jointing is unknown, and from the
absence of joint-block outlines to the rock chunks of the
jumble, it may be argued that at the time of the filling
of the rock fault the system of jointing now general over
the district had not yet affected the rocks. This con-
clusion is rendered the more acceptable by a considera-
tion of the condition of the coal inside and outside the
rock-fault areas. As mentioned above, no rock collected
from a “ jumble ” has been proved to be in any way
chemically different from the rock of similar texture in
the measures outside. This constancy the writer inter-
prets as meaning that all the displacement of rock pieces
(coal as well as stone) had been completed before the
“ sline ” was imposed, and that “ sline,” therefore, like
the jointing of the rocks, w’as due to earth movements
at a period subsequent to the formation of the “ rock
faults.”

Alternative Hypotheses.

Having assembled the available data, the writer will
now consider how the various items of evidence are
fitted by the alternative hypotheses, which he may
re-state as follow :—

I.	—Rock faults may be due to contemporaneous
erosion of coal seams by flowing streams during coal
measure time.

II.	—Rock faults may be due to lateral adjustments
between beds by earth movements, at an early period
subsequent to the consolidation of the rocks.

That rock faults affect individual coal seams [1]* at
particular places [2], and that they cut off the seams
obliquely [4], is only to be expected on either hypo-
thesis. That they are only frequent in seams in which
a normally weak roof is closely overlain by a strong

* The numbers in brackets refer to the numbered para-
graphs.

sandstone or conglomeratic rock [3] is the chief buttress
on which upholders of hypothesis (I.) rely. These
advocates of contemporaneous erosion will point out
that the change from the quiet conditions required for
deposition of fine sediment to conditions which would
allow of the laying down of coarse sand and gravel over
the same area is indicative of important alterations in
the local topography, and that during the adjustment
of these conditions continuity of the process of deposi-
tion must have been interrupted, at least for a time.
It was during that interruption that the local erosion,
which they allege, must have taken place. Advocates
of the earth movement explanation point out that, by
reason of the change in the character of the sediment,
there has resulted a corresponding variation in the
mechanical properties of the rocks into which the sedi-
ments have consolidated, and that at bedding planes
where there were abrupt changes in the character of the
sediment there are now the most marked contrasts of
rigidity between the materials on either side of the
bounding surface.

As mentioned already, the contrast in behaviour of
hard sandstone and soft clod under the mechanical con-
ditions which prevailed during the production of a rock
fault is very striking; and the writer sees no reason to
believe that where thick masses of strong sandstone,
interstratified between weaker beds of bind, etc., and
extending over square miles of country, would, when
the region is subjected to side and end pressure, do
otherwise than allow the weak rocks to slide upon them.
Such slipping is a process quite analogous to the con-
traction, under vertical pressure, of muddy sediments
past the hard concretions of ironstone, etc., in the pro-

Figs. 7 to 9.

Diagrams showing the fingering-out of a coal seam or
fragment along “ stream-lines ” which are determined by
the arrangement of the other and harder rocks of the rock-
fault jumble. Often also in the termination of the seam
the beds and partings are wedged open by bind or stony
material, and the coal comes to an end in a sheaf of
divergent wedges, the transverse sections of which recall
fluttering pennons or the fin-scales of a swimming fish.

duction of pot holes, as already described. This
admitted, the origin of rock faults follows easily. For
with the driving forward of hard rigid masses of sand-
stone separated by weak bind from coal, which, if not
strong as compared with the sandstone, was certainly
more rigid than the separating bind, the shearing would
be localised in the bind, which in shearing became some-
what schistose, and was converted into clod. It is
unlikely that mud resting upon a coal seam and covered
by sand would remain uniformly thick, and the base of
the sandstone might be correspondingly wavy. Hence, in
the shearing, portions of the bind might be gripped
beneath the sandstone and be carried forward with it. A
continuation of the movement, especially if the pressure
increased until the rock began to bend, would “ roll
up ” the weaker rock into flattened cylinders, each with
a lenticular-like section and tapering ends. These
piling one upon another at certain places form the
patches of “ bad roof ” that are frequent in many pits,
which, though themselves free from rock faults, are
working in seams which are liable to rock faults else-
where.

Between the patches of “ bad roof ” appear the “ rock
rolls,” under which for a space the clod is completely
pinched out, and the rock comes to rest directly upon
the coal. It was (in 1913) from observations made on
the structure of the “ jumbled ” bind in the roof,
between rock rolls in the Park Gate seam workings at
Tinsley Park Colliery, that the writer’s attention was
drawn to the whole subject, and he noted then, as he
affirms now, that the “jumbled” structure of the bad
roof between rock rolls differs in no way from that of a
fault breccia. From rock rolls to rock faults is but a
further stage of the same process, for, granted that a
rock roll is in contact with the coal, and a further for-
ward movement takes place, the rock roll must “ seize ”
with the upper layers of the coal, which, by reason of

their greater rigidity, will offer special local resistance.
This being overcome by the accumulation of the shear-
ing stress, a tearing both of coal and rock results, and
the rock fault is formed and filled in one and the same
process.

The plan of the rock fault area [5] and [6] is also
critical evidence. Stream courses or hollows exca-
vated by tidal scour have shapes which, although
governed by complex laws, are quite determinate. It
is unnecessary to discuss here the details either of the
shapes or of the laws which govern them, but the writer
may mention that the tributaries of streams enter the
main stream at intervals which are commensurate with
the volume of water carried, and that in present-day
geography it is difficult to match the occurrence of
several streams, each 50 or more yards wide, flowing in
sub-parallel courses and eroding channels across the
same half-mile of country. Streams also, though they
may meander, can hardly have excavated hollows
arranged en echelon, and the matter may be dismissed
with the remark that in comparing detailed plans of rock
fault areas with river causes they cannot be made to fit.

Masses of rock undergoing distortion under earth
pressure are subject to the harmonic folding which gives
rise to the wave forms of anticlines and synclines. More
rigid beds between others of less rigidity, when they
fold, push the weaker rocks aside, and, as has already
been shown, the troughs of the sandstone waves come
down and touch the coal. Rock rolls are the waves,
rock faults the breakers, and the plan of the only York-
shire rock fault area which has yet been even moderately
completely explored in mining (the workings of the
Haigh Moor seam between Normanton and Kippax)
recalls nothing so much as the irregular oval splashes of
light thrown upon a ceiling by the reflection of a sun-
beam from the surface of ripples on agitated water.

The prevalence of slickensides at the under boundary,
and the frequency with which glossy surfaces appear,
both in the partings among the unmoved coal and
between the lenticular units of the rock-fault filling, are
not explained by the erosion hypothesis. By the advo-
cates of that view slickensides are looked upon as
adventitious, and as produced by earth movement at a
period subsequent to the formation of the rock fault
itself. The burnish or slickensides is due to surface
powdering between masses which are shearing past each
other under considerable stress, and the parallel grooved
furrows and fine striae on the glossy surfaces give a
trustworthy indication of the direction along which the
shear took place. Slickensided surfaces on coal or bind
decrepitate rapidly on exposure to the air, and when
found preserved in abundance, as in rock fault material,
it may be assumed with certainty that from the time of
their formation until their present exposure the pressure
which has held them tightly has never been completely
released.

The thickening of the coal [8], whether by bulk com-
pression or by overthrusting, is the best evidence of a
large scale shearing by lateral pressure that could have
been preserved. The arrangement of the thrust planes
is not dissimilar to the arrangement of the rock fault
patches. The lines of the faults are generally parallel
with the curving boundaries of the rock fault. Each
overthrust bends round somewhat at each end as the
throw of the fault dies away, and in bending round it
often happens that two or more of them run together.
In plan their relative arrangement may be likened to
the arrangement of the wrinkles in crinkly paper, and,
if the coal seam be regarded as having been crinkled
by the overdrag of the roof, which had seized and, con-
tinuing to move, was tearing away the coal below, the
simile will bear a close analysis.

Upholders of the stream erosion theory seem to have
ignored the evidence of overthrust coal seams, and
explain “ proud ” coal as due to a sort of sweeping out
of the peaty material from the stream course and a piling
up of the debris among the marsh growth on the banks.
This process is not unknown near the mouths of sluggish
rivers, like the Lower Mississippi, which by deposition
of alluvium in times of flood has built up its banks some
scores of feet above the level of the sides of its own
valley. Such explanation, however, seems far-fetched,
and becomes the more inadequate when the associated
occurrence of “ foreign bodies ” [9] and the wedges of
rock [10] which break into the proud (and the lean) coal
both from above and from below, are taken into
consideration.

By those who accept the contemporaneous-erosion
hypothesis the “ blocks of isolated sandstone ” and other
foreign bodies are explained as “ doubtless being the
result of cavities formed in the coal about the time of
denudation and afterwards filled with sand, etc.” * On
close examination, however, each foreign body proves
■to be a fragment or complex of fragments, bent, broken,
rolled over, twisted, or sheared, but always preserving
some traces of the bedding planes which were developed
during the original deposition as part of a continuous
sheet of sediment. Tongues of rock can also be seen
following surfaces of overthrusting, and it is suggested
that both they and the “ foreign bodies ” have really
been dragged into their present position along displace-
ments which are a minor manifestation of the same
shearing as that which has produced the larger thrust
planes and rock faults (fig. 3).

Surpassing all other difficulties in the way of the
acceptance of the origin of rock faults by erosion, is
the character of the rock fault “filling.” Sand or
gravel drifting into a submerged stream course must
undergo a sorting of the grains by size, and in their
deposition the fragments could not develop any arrange-
ment less orderly than current bedding. Materials
brought for deposition might come from anywhere within
•the watershed of the stream which brought them, or, if
a tidal drift were the carrier, might be derived from
anywhere along the coast. The size of the fragment
would, however, be a measure of the strength of the

* Hendy, loc. cit., p. 435.