erosive forces, so as to cause them to excavate the valley along that line rather than any other.

When, moreover, we examine faults below ground, we find no trace of any widegaping fissures; the walls of the fault, on the contrary, are jammed tightly against each other, and show frequent evidence of immense grinding force, proving the friction of the sides to have been enormous. In hard massive rocks there doubtless occur open spaces here and there between the walls, "pockets" or "bellies" between their projecting protuberances, or where they have been partly kept asunder by fragments detached from the sides. These are often full of crystalline minerals, and form "mineral veins" below, but seldom, if ever, form valleys or ravines at the surface.

If these ideas as to the relative action of the internal and external forces at work upon the crust of the globe be well founded, it follows that none of the present features of the surface of the globe have been produced by the direct action of the internal forces, except volcanic orifices and cones, and that all others have been produced by the process of external erosion, except such as have been formed by external deposition, like hills of blown sand or alluvial flats and deltas.

The surfaces of our present lands are as much carved and sculptured surfaces as the medallion carved from the slab, or the statue sculptured from the block. They have been gradually reached by the removal of the rock that once covered them, and are themselves but of transient duration, always slowly wasting from decay. Even, then, if the internal forces could produce such external features, it can always be shown that the surface which existed when they operated has long since disappeared, together with, in many cases, vast thicknesses of rock that intervened between it and the present one.

It remains to say a few words on the nature of the erosive agencies which form these surfaces.

The ocean is the grandest of these. The ceaseless breaking of its waves against the margin of the land constantly gnaws into and undermines it, and the tides and currents carry off the eroded materials and deposit them on some part or other of the ocean-bed. This action is that of a great horizontal planing-machine, always tending to the production of level surfaces, the cutting power being confined to the sea-level, while the matter carried off tends to fill up the hollows of the inequalities that lie below it. The denuding action of the sea, therefore, produces "plains of denudation on the parts it has passed over, and long lines of cliffs or steep banks along the margin where its influence ceased. It is essential for the energetic action of the sea that it should be the open sea, where a heavy swell can roll in upon the land, and where gales of wind can hurl furious waves against it. In sheltered bays and narrow inlets and fiords its erosive agency becomes comparatively small, and in very protected places sinks to nothing.

While, then, we look to marine denudation as the cause of wide plains, of long escarpments, of bold headlands and isolated hills, and of the general outline of mountain-chains, and as the remover of the great groups of rock that were continuous over the area of the mountains before their elevation was commenced, I believe we err when we attribute to that cause the lesser features by which these greater ones are themselves modified. The river valleys that traverse the great plains, the gullies that run down the sides of the hills, the valleys, glens, ravines, and gorges that furrow the flanks of the mountain-chains, have, I believe, all been caused by atmospheric agency on the land, while standing above the level of the sea.

The only case in which the sea tends to produce anything like a valley is that in which it forms open sounds or straits between islands, where the set of the tides and currents imparts to it a river-like action. Those depressions in the crest of a mountain-chain which are called ". "passes or gaps "have doubtless been often caused by this action, but it is obvious that this ceases as soon as the summit of the pass once rises above the sea-level and prevents the currents from sweeping through it.

While the ordinary erosive action of the sea is a horizontal one, tending to the production of plains bordered by cliffs, that of the atmospheric agencies is a vertical one, always tending to the production of furrows, or more or less steep-sided channels, on all the land exposed to their influence.

Rain falls vertically, and tends to sink vertically into rocks, producing decomposition in them, both by mechanical and chemical action. A superficial coating of greater or less thickness is always thus kept in a state of decay.

In almost all granite districts, the rock beneath the hollows and flatter parts of the ground will often be found to be decomposed in situ to a mere sand, so that it could be dug out with a spade to a depth of several feet. Roundish lumps are found here and there in this sand, which were the centres of the original blocks; these, as well as the solid rock below, showing every gradation of firmness, from hard crystalline rock to a mere incoherent sand. I have observed this in granite districts in all parts of the world, and was much struck with it during the past summer in the southern part of Brittany, where the deep narrow lanes often showed both granite and gneiss thus rotten and soft, to a depth sometimes of fifteen or twenty feet. On the steeper slopes the exposed rock was much less decomposed, obviously because the particles had been washed down and carried off as fast as they became completely disintegrated.

Hard limestones, again, exhibit the effects of the action of the rain in the numerous open fissures and caverns that are always found in them, the water here having dissolved the rock and carried it off in solution, as if it were so much salt or sugar. The fantastic forms and honeycombed surfaces of all limestone crags attest the same action. In baring the surface of a limestone quarry where the beds are inclined at any considerable angle, they are often found to be furrowed by rain-channels one or two feet in depth and several inches in width, the hollows being filled with the finest earth. A deep covering of mould and turf is no protection against this action, and perhaps even aids it by contributing an additional dose of acid to the rain-water.

Even where hard siliceous rocks exhibit a weathered coat of a very slight depth, a mere skin perhaps of a quarter of an inch thick, as is the case in some Felstones, still it merely proves that the atmospheric influences cannot affect a great thickness at any one time, and does not render it impossible that many such weathered coats may have been formed outside the present surface, and successively removed altogether by the completion of the process.

The joints of rocks when first formed are doubtless mere planes of separation, without any interstice that would allow the insertion of even the thinnest edge of a knife; they would be quite insensible to the sight, and would perhaps scarcely of themselves be sufficient to cause the separation of the rock into distinct blocks. In working deep mines it is sometimes said that the rocks cease to show any joints at all. The joints, however, doubtless exist, although they are invisible, while the open joints, such as we see in all rocks near the surface, have been opened by the "weather" acting along these concealed planes of separation.

The action of the atmosphere, then (i. e. the chemical action of air and water and the various gases mingled with them, and their mechanical action, owing both to their movements of gravity and their expansion and contraction from changes of temperature), is operative in the gradual destruction of rock, to a much greater vertical depth beneath the surface than is commonly recognized.. Its superficial action is still greater, and has also, as I believe, generally failed, as yet, in receiving due appreciation. The rain that falls upon the surface and does not sink beneath it runs, of course, down the shortest and steepest slopes it can find, and is collected first into rills, then into brooks and rivulets, and finally passes by rivers to the sea. This superficial drainage of a country is often augmented and kept up by springs, which are caused by that part of the water that had sunk beneath the surface finding its way back to it.

The natural tendency of running water is to cut its channel deeper, and that at a rate compounded of the rapidity of the current and the nature of the rock below. Let any one take the basin of drainage of any great river, and trace it up to its source, following all its tributaries to their sources, and he will not fail to perceive that all the varied features of the different channels of this system of running waters are the result of these two circumstances only. In the mountain glens he will see those that traverse granite commonly with rounded open forms; those that cut through hard slates, or thick horizontal sandstones, are commonly narrow and precipitous, with jagged cliffs and overhanging ledges, perhaps, jutting from

the sides of the ravines. He will see the marks of the old cataracts that once fell over these ledges, but which now are removed to other places, or converted into mere rapids, or perhaps altogether obliterated by the cutting down and cutting back of the streams. Torrs and pinnacles will be left here and there, perhaps, rising up from the bed of the stream, showing the former islets and rocks which resisted the erosive action better than the parts on each side of them. Where a softer and more yielding mass of rock occurred, there the glen widens into an open valley; the narrowest and most jagged and steep-sided glens are just where the rocks are most hard and intractable, and best calculated for resisting the chemical and mechanical action of running water.

The scale upon which these operations have been carried out does not affect the nature of the argument. The action has been the same in the miniature glens of our own mountains and in the grander and more awful abysses that gash the sides of the Alps, the Andes, and the Himalayas.

In all cases when the river comes down now, or has formerly come down, in the form of a glacier, before springing into running water, the ice-mass has of course scooped out and deepened and widened the valley in its own peculiar fashion.

When we leave the mountains and come down into the lower lands, where the rivers wind with a more gentle stream from side to side of broad open valleys, through wide alluvial flats, still it is to the river that the present form and depth of the valley are due. Whatever may have been the undulation of the original surface of marine denudation which determined the course of the primary stream, the river has long since cut down beneath that surface, and is still occupied in cutting deeper, so long as it retains any sensible current at all. It effects this by undermining the bank now on one side and now on the other side of the valley, shaving off a little corner here and another there, so that a river not a hundred yards broad, perhaps, may eventually form a valley of several miles in width, The obstructions it accumulates from time to time in its own bed constantly deflect its channel, so that ultimately it visits every part of the valley.

In many cases the mere deepening of the valley may necessarily widen it also, since the rocks may be of such a composition, or may lie in such a way, as not to be able to form a bank of any steepness; and the materials, therefore, always slip down towards the bottom of the valley as fast as their bases are cut into.

It is true that all these processes are infinitesimally slow; but if carried on through a period of time indefinitely great, it is obvious that it is impossible to assign a limit to the amount of their results.

I have for several years been studying the origin of the river-valleys of the South of Ireland, and have, since the last meeting of this Association, been compelled to arrive at the conclusion that the great limestone plain of the centre of Ireland has lost a thickness of 300 or 400 feet at least, by the mere action of the rain that has fallen upon it. As a corollary of this conclusion, I have also been led to perceive that the longitudinal and lateral valleys of the Irish mountains-and if of them, then those of all other mountain-chains of the world-are the result of the action of the water or the ice that has been thrown down on them from the atmosphere.

If we take any mountain-chain and its adjacent lowlands, and suppose no rain to fall upon them for a time, and that all the valleys of whatever description were filled up, and the sides of the mountains smoothed over from their peaks to their bases, I believe the surface thus produced would be one representing the limits of marine denudation; then let rain begin to fall on such a country, and all the elaborate structure of valleys, gorges, glens, and ravines would be produced by it.

I believe that the lateral valleys are those which were first formed by the drainage running directly from the crests of the chains, the longitudinal ones being subsequently elaborated along the strike of the softer or more erodable beds exposed on the flanks of those chains. I do not, of course, intend to say that any country ever existed without valleys, since valleys of some kind must commence as soon as the first peaks of the mountains show themselves above the sea, and must be continued and extended in proportion to the extent of the land which gradually rises into the atmosphere. Atmospheric denudation and marine denudation have always been at work simultaneously upon the different parts of every land in the globe, and their

action may be very complex, so that it is often difficult or impossible to separate the results of one from those of the other at any particular place. Still I believe we may generally regard the external form of a mountain-chain as due to marine, and the valleys within it as the result of atmospheric erosion.

Most of you will be aware that the views I have thus endeavoured to place before you are not altogether original; other persons have before now proposed the same method of explanation of the form of ground. M. Charpentier long ago referred the origin of the valleys of the Pyrenees to the action of the rivers which traverse them. Mr. Dana had pointed to the same action as the cause of the wonderful system of ravines that furrows the sides of the Blue Mountain range in New South Wales, and of the deep ravines separated by knife-edged ridges which radiate from the centres of the high islands of the Pacific. I confess, however, that I had, up to the present year, hesitated to accept this explanation without reserve; and therefore, since I am now convinced of its truth, I am anxious to take the earliest opportunity of recording that conviction*.

Mr. Prestwich, in his recent papers read before the Royal Society, has adopted the hypothesis of the subaerial deepening of the valleys of the Somme and the Seine, and other river-valleys both in France and England, to account for the formation of the freshwater gravels which he finds on the flanks of those valleys, so high above the present levels of the rivers or of any possible floods.

Professor Ramsay has in like manner attributed the formation of the hollows in which the lakes of Switzerland lie, to the ploughing action exercised on the subjacent rocks by the action of the glaciers, when far more extensive than now. The formation of lakes lying in "rock-basins," and not formed by the mere stoppage or damming up of a river, had always been a complete puzzle to me until I read Professor Ramsay's paper in the last Number of the Geological Journal (May 1862). I believe his explanation of their origin to be the true one.

That he and Mr. Prestwich and myself should all, within the space of the same twelvemonth, have been compelled to appeal to external atmospheric action as the only method of explaining the origin of the different surface-phenomena we were studying, is of itself, I think, good evidence that we are all three pursuing the right track in our search after truth.

At the instant of penning this sentence, I see by a newspaper paragraph that Dr. Tyndall follows us in his speculations as to the origin of the valleys of the Alpst. *Had I not become previously convinced of the extent and power of atmospheric and river action in consequence of my own observations, all scepticism must have yielded to the proof of it detailed in the admirable Report by Dr. Newberry on the Geology of the Colorado River of the West, published by the United States Government at Washington in 1861. It was only in February 1863 that I saw this work through the kindness of Dr. Newberry, who himself transmitted to me a copy of it. The beautiful maps and plates and the numerous woodcuts illustrate the text in a way that puts to shame the miserable niggardliness of our own Government in such matters; for here they are either committed to the red-tape ignorance of mere clerks whose duty it is simply to curtail expenditure, or to the equally uninstructed indifference of higher officials in dread of the well-meant but blundering questioning of some man of figures in the House of Commons, or still oftener left to the enterprise of some publisher, who has of course his profit to make out of the work. An advertisement at the beginning of the American Report shows that the Senate of the United States ordered ten thousand extra copies of it to be printed, five hundred of which were given to the officer commanding the expedition.

Dr. Newberry shows in his Report that the wonderful cañons which traverse much of the country of California, and some of which are from 5000 to 6000 feet deep, and only wide enough for the waters of the rivers to flow through them, have been cut down by those rivers through horizontal and quite undisturbed beds belonging to the Carboniferous, Devonian, and Silurian periods into the Granite below, and moreover that wide valleys in other parts have also been excavated by the gradual action of atmospheric erosion, leaving numerous perpendicular torrs, crags, or pinnacles of rock here and there, all showing the same horizontal beds.

+ A subsequent reading of Dr. Tyndall's paper, and of a notice of it afterwards by Professor Ramsay, showed me that Dr. Tyndall was inclined, at the time of writing it, to attribute the Alpine valleys too exclusively to the action of glaciers. The valleys must have been commenced and many of them almost completed before the glaciers, although the

As a concluding observation, allow me to remark how curiously the threefold physical agencies that are in simultaneous operation on the crust of the globe were typified in the old heathen mythology. The atmosphere which envelopes the land and rests upon the sea, the ocean which fills up the deeper hollows of the earth's surface, and the nether-seated source of heat and force that lies beneath the crust of the earth are each personified in it as a great divinity. If one of the old Greek poets were to revisit the earth, and clothe these ideas in his own imagery, he would tell us in sonorous verse of Zeus (or Jupiter), the God of the Air, ruling all things upon the land with his own absolute and pre-eminent power; of Poseidon (or Neptune) governing the depths of the ocean, but shaking the shores which encircle it; and of Hades (or Pluto), confined to his own dark regions below, tyrannizing with all the sternness of a force irresistible by anything which can there oppose it, but rarely manifesting itself by any open action within the realms of the other divinities.

On an Early Stage in the Development of Comatula, and its Palaeontological Relations. By Professor ALLMAN, M.D., F.R.S.

The subject of this communication was a small Echinodermatous animal, a single specimen of which was obtained by the author on the south coast of Devon, where it was found attached to one of the larger Sertularidæ, dredged from about four fathoms' depth. The author regarded it as one of the early stages in the development of Comatula, and though quite distinct from the well-known Pentacrinus stage of this crinoid, believed that it had been witnessed both by Thompson and Dujardin, but not correctly described or figured by either of them. It consisted of a body borne upon the summit of a long jointed stem. The body had the form of two pyramids placed base to base. The upper pyramid is formed of five triangular valve-like plates, moveably articulated upon the upper side of the lower pyramid, and capable of being separated from one another at the will of the animal, so as to present the appearance of an expanding flower-bud, and again approximated till their edges are in contact and the original pyramidal form restored. From between the edges of these plates, long flexile tentaculoid appendages, which must not be confounded with the permanent arms of Comatula, are protruded in the expanded state of the animal, and within these is a circle of shorter, more rigid, rod-like appendages which seem to be moveably articulated to the upper side of the calyx, immediately round the centre, where it is almost certain that the mouth is placed. The lower pyramid or proper calyx is mainly formed of five large hexagonal plates, separated from the summit of the stem by a zone, whose composition out of distinct plates could not be demonstrated, and having five small tetragonal plates intercalated between their upper angles. In assigning their proper value to the several plates thus entering into the body, the author regarded the lower zone, which rests immediately on the stem, as simply a metamorphosed joint of the stem itself, while the verticil of plates, situated immediately above this, is the true basilar portion of the calyx. The five small intercalated plates are the equivalents of the radialia, and destined to carry afterwards the true arms of the crinoid; while the five triangular plates which constitute the sides of the upper pyramid are interradialia. Professor Allman considered the little animal described in this communication as of special interest, in the light which it seemed capable of throwing on the real nature of certain aberrant groups of Crinoidea, such as Haplocrinus, Coccocrinus, &c., in which the calyx supports a more or less elevated pyramidal roof, composed entirely or in great part of five triangular plates, which find their homologues in the five sides of the pyramidal roof of the little crinoid which formed the subject of his paper.

On Bituminous Schists and their Relation to Coal. By Professor ANSTED, F.R.S. The occurrence of rocks of all geological periods, and in most parts of the world, containing a sufficient quantity of the mineral hydrocarbon to be worth distilling

present depth, width, and regularity of many of them are doubtless ascribable to glacier

action.

1862.

5