idea of its structure and extent from personal observations, and to consider the most probable way in which it has been formed.

Immediately to the north-east of the village of Cheddar, which stands on ground only a little elevated above high-water mark, there is a combe-shaped valley. Its sides gradually contract, and become more rocky and precipitous, until it forms a narrow winding gorge, with walls of limestone. About a quarter of a mile from the village the gorge becomes very narrow, and the cliffs, especially on the right hand side, very steep and even overhanging. For some distance beyond, the road runs along what may be called the Strait of Cheddar, after passing which the ravine gradually opens, and its sides become more sloping, until it loses its rugged grandeur of outline. In the most contracted part of the Strait the observer is so completely hemmed in by bare rocks as to require little to make him fancy himself in a mountain solitude remote from the habitations of mankind. When all is still in the neighbouring plain, the wind here often blows violently, and is deflected from cliff to cliff with a sound which, to the mind of a contemplative geologist, might suggest the idea of audible spectres of stormy billows which once may have followed the same course, as they rebounded from side to side of a narrow inlet of the sea. During the writer's first visit to this spot, numbers of rooks were soaring from precipice to precipice, and often appeared like black dots against a narrow strip of sky; while the resemblance to white ants presented by sheep browsing on ledges near the top of the cliffs furnished a much more impressive idea of their height than any process of measurement.

Some of the old women, who importunately press their services on the tourist as guides, will tell him that the Wind Cliff is 480 feet high. This cliff (to which no drawing can do justice) is certainly a most remarkable specimen of a literally mural precipice of considerable breadth, and at least 300 feet in height. It is quite perpendicular from top to bottom, excepting where it overhangs. It is a much finer and larger face of rock than the cliff at the entrance to Goredale in Craven, Yorkshire, and nearly twice the height of the rocky part of the High Tor, near Matlock, in Derbyshire. As regards con

tinuous perpendicularity, I believe it is not equalled by any limestone cliff in the kingdom. Next to the Wind Cliff, the so-called Cathedral Rocks are the most impressive. They consist of several buttresses projecting forward from the main line of cliffs. Their sides are perpendicular, and their fronts overhanging. The height of the summit of one of these rocks above the level of the road at the bottom of the ravine has lately been ascertained to be about 420 feet. The way in

which it was measured is deserving of notice. A worthy scientific gentleman of the neighbourhood ventured to crawl on to the summit from the grass-covered down behind, until he found himself on the brink of three precipices-those on the right and left perpendicular, the one in front overhanging. He dropped a line from the most extreme part of the brink, which went down without touching rock until the plummit struck a slightly-projecting terrace near the bottom of the ravine, and then fell on the road. There is probably no other part of England where a conformation of cliff-architecture would admit of a similar feat being accomplished. But our wonder at the cool intrepidity of the performer will not be so great when we consider that he was a member of the Society of Friends.

It has already been hinted that the Cheddar ravine is very tortuous. It consists of an alternating series of recesses and projections, or small bays and headlands. In some places there is a certain degree of correspondence between the hollows on one side and the protuberances on the other, which might at first lead one to fancy that the ravine is solely or mainly the result of a violent severance of the rocks; but a little observation will be sufficient to show that the two sides were never in contact. On the right, looking from Cheddar, the cliffs are very precipitous; on the left, they generally slope down into the ravine at a small angle.

The cliffs mainly consist of large faces of moss-covered rock, but the clefts and narrow terraces furnish a habitat for various plants, which add beauty to the sublimity of the scenery. Ivy and yew grow out of the fissures, and in various places may be found liverwort, polypody, meadow-rue, crimson mountain-pink, etc. The "screes "* at the bases of the cliffs have in many places acquired a covering of grass, and do not now appear to be in course of accumulation. The positions they occupy in some places would seem to indicate that they must have been thrown up against a wall of rock, or into a recess, rather than hurled down from above. The effects of the action of frost and rain, however, may be seen in favourable situations. The frost detaches angular fragments from incoherent parts of the cliffs, and from the under sides of rocky projections. The rain carries previously-detached fragments and chips down the "rakes," or vertical passages which indent the face of the cliffs. Vegetable mould and red loam are likewise washed down by rain from the top of the cliffs, and subjacent fissures. Nearly the whole surface of the Mendip Hills is covered with red loam, which fills up the fissures, and, to a certain extent, the caverns. Its derivation, and period or

* A convenient name used in the Lake district for the accumulated wrecks of cliffs and declivities.

periods of deposition are involved in mystery; but it seems to be generally admitted that it must have been a kind of seaooze left by retiring waters during one or more submergences of the land. In later times, it has been re-arranged by subaerial and subterranean fresh-water streams.

The Cheddar caves are regarded by the natives as the greatest source of attraction. One, in particular, has become very celebrated for its stalactites. But as it is probably surpassed in this respect by caves in Derbyshire and elsewhere, the main attraction of the Cheddar ravine must ever lie in the almost unparalleled grandeur of its cliff scenery. In this ravine there are many caves, little known and seldom visited, which present phenomena more interesting to the geologist than stalactitic concretions, however much the latter may resemble any earthly or unearthly objects the guide or the visitor may fancy. On the left side, walking from Cheddar, before reaching the Strait, there is a cave with a very conspicuous entrance at some height above the road. It has apparently been scooped out, or at least enlarged, by an inwardly-directed agent, such as sea waves, and not by an out-flowing freshwater stream. On the right side, at various altitudes, there are many caves. In nearly all of them the roof is more or less rounded, and the sides here and there smoothly hollowed out into pot-shaped cavities. In short, the interior of these caves display obvious signs of the action of water, charged with a sufficient amount of solid matter to enable it to round and smooth limestone rock; and the position in which the rounded and smoothed surfaces often occur, would seem to point to the action of powerful waves as the only adequate explanation. It is true that fresh-water percolates through crevices in limestone districts, and the Cheddar brook has its visible source in several streams which flow out of subterranean cavities near the south-west end of the ravine; but it is not very difficult for one who is familiar with the peculiar forms resulting from the inward and upward gyratory action of the waves of the sea to distinguish these forms from marks left by fresh-water streams. The latter tend to wear their channels downwards, and can never produce smooth vaulted roofs, hanging or inverted potholes, arched entrances, and other characteristics of sea-worn

caverns.

Mr. W. Boyd Dawkins (to whom the scientific world is so much indebted for the exploration of Wookey Hole Cavern, a few miles from Cheddar*), believes that the Cheddar ravine is an immense unroofed cave, the abstraction of the rocks once filling the now vacant space having been effected by atmo*See first paper on Wookey Hole, in the "Quarterly Journal of the Geological Society."

VOL. XII.-NO. I.

D

spheric agency. It is true that rain-water, especially when assisted by humus derived from the vegetable soil above, is capable of enlarging crevices. But the erosion resulting from its chemical action is limited to spaces where stalactitic and stalagmitic deposition is not going on. This deposition evidently rather tends to preserve caverns than facilitate their destruction. The detachment of chips from rocks by frost in favourable situations is a process likewise limited to spaces where the chips have not accumulated to too great an extent. All atmospheric agencies which can only remove matter a short distance must tend to choke up or glut their sphere of action. Without the assistance of a powerful transporting agent they can never, in such situations as the Cheddar ravine and its caves, make permanent progress in the great work of denudation. But supposing atmospheric action during millions and millions of years to be capable of producing a vacuity as large as the Cheddar ravine, its form would still remain to be explained. Sufficient time allowed, a colony of ants might be considered capable of rearing a mountain mass equal to the Alps, but an examination of the form of the Alps would at once forbid the idea of insects having been the architects. Atmospheric agents, whether operating above or under ground, are now producing nothing similar in form to the main features presented by the Cheddar ravine. That the cliffs have been modified in exposed situations by frost is evident, but the modification has been in the direction of destroying and not developing the characteristic forms of the cliffs. The same remark applies more or less to the action of fresh-water in caves.

The common notion that the Cheddar ravine is a crack or rent may, in a very limited sense, be correct. It is possible, if not probable, that at first there may have been a narrow winding fracture similar to that behind the High Tor at Matlock. It is, however, certain that no fracture ever occurred sufficient to disturb the angle at which the strata dip in a south-easterly direction, which on both sides of the ravine exactly corresponds. A very little observation will be sufficient to convince any one that the ravine has been mainly, if not entirely produced by the abstraction of an immense mass of limestone rock. The following diagram will show the stratigraphical structure of the locality, the dotted lines representing the strata which have been removed.

From the foregoing observations I think it must appear obvious that the clean removal of a stupendous quantity of rock must form the burden of any satisfactory explanation of the origin of the Cheddar Cliffs. It is true that in some places the rocks are now crumbling, but there is no agent (with the

exception of man) to carry the detritus away. In most places the rocks have preserved their original smoothness and regularity of outline. Here and there concave undercuts run hori

Section of Cheddar Ravine.-A, rocks removed; BB, remaining rocks; cc, terraces; dd, undercuts; e, road.

zontally along the the faces of cliffs, while at the base of others there are cavernous recesses with water-worn roofs. But the most instructive forms of rock surface are the planes caused by jointing and bedding. They show that the mode in which the adjacent blocks or masses were carried away was not a process of granular dissolution, nor even fragmentary dilapidation, but a bodily displacement. The cause must have been equivalent to the translation of large blocks of limestone. To borrow an illustration from the well-known cheese of the neighbouring plain of Cheddar, if a farmer were to find one morning that a part of a cheese was missing, and that the surface left was smooth and regular, he would conclude that some person had cut a slice with a knife, not that a mouse had been nibbling at the cheese during the night; or, suppose the farmer were to find that a whole cheese had been removed from his storeroom, he would at once conclude that a power capable of carrying it away in a lump had been concerned in the theft. A puny agency, which carries on its work grain by grain, or bit by bit, cannot leave a smooth, plain, and regular surface of any extent; but a violent and powerful agency, while it is not incapable of leaving a rough surface (circumstances being favourable), mainly tends to produce breadth and uniformity of

contour.

The most philosophical way of trying to explain natural phenomena is to seek for similar phenomena now in course of being produced; and many modern sea-coasts exhibit fac-similes of the Cheddar cliffs. The forms of these cliffs are precisely those which would result from waves driven by storms into a narrow inlet of the sea. At a greater elevation, near the summit of the Mendip hills, smoothed (sometimes polished), rounded, hollowed, perforated, and grooved surfaces of rocks may here and there be traced. The way in which they have been shaped