greenish-grey streaks and patches, and containing hard nodules and ferruginous concretions. Such beds are well shown near Williton, Halse, Ninehead, Wellington, Kentisbere, and Sidmouth. Near Williton the beds are very calcareous, and are burnt for lime.

In the sandstone series of Otterton Point Mr. Whitaker found the jaw of the Lacertilian reptile Hyperodapedon; and more recently in the higher portion of the same series of beds, nearer to Sidmouth, Mr. H. J. J. Lavis has discovered several Labyrinthodon remains, which render it probable that others who devote sufficient time to minute examination of these strata may be similarly rewarded.

Beneath the Upper Sandstones come beds of Conglomerate with large pebbles of Grit and Limestone, as at Tipnoller and other quarries east of Wiveliscombe, near Stogumber, Milverton, and Wellington. Sandstone bands frequently occur between the beds of conglomerate. Between the sandstone and conglomerate there is indeed no hard line of demarcation. South of Milverton, and between Halse and Preston Bower, there is a development of pebbly sandstone. (See fig. 13.)

In some localities the limestone pebbles and boulders from the Conglomerate are burnt for lime; and the other pebbles of grit and sandstone are used for road-mending.

The Pebble-beds of Budleigh Salterton are made up largely of flattened quartzite pebbles, containing Palæozoic fossils: these have been traced by Mr. Ussher to the quartzose pebble-beds of Burlescombe.

The term 'Popple-rock' is sometimes applied to the conglomerates, and the Budleigh pebbles are locally known as Popples.' There is also the village Newton Poppleford.

The Lower Marls are used for brickmaking near Wiveliscombe and Cullompton.

In Devonshire and West Somerset there are certain beds, included in the New Red Series, which are termed Breccias.

They are made up for the most part of slaty fragments imbedded in a red sandy matrix, and are accompanied, sometimes abundantly, by pebbles of Devonian Limestone, by pebbles of Carboniferous and Devonian grits, and by boulders of Granite and various Igneous rocks.

The Lower Sandstones and Breccias, containing clayey and marly beds, occur at Stogumber, Heavitree, Dawlish, Teignmouth, Kingskerswell, &c. They contain many limestone pebbles, some of which at Teignmouth enclose Goniatites, Clymenia (C. linearis), and many Madrepores. These pebbles are burnt for lime in places.1

Some of the brecciated clay beds in the neighbourhood of Exeter and near Torre are worked for brick and tile-making.

The lower beds of Breccia between Teignmouth and Exeter have been long known to contain a variety of felspar called Murchisonite, evidently derived as pebbles from the granitic or igneous rocks (quartz-porphyry, &c.) of the district, yet differing a little in colour and texture from the mineral when found in situ. Mr. Ormerod has lately endeavoured to trace out certain horizons in this series by the occurrence of Murchisonite, whence the name Murchisonite beds,' which has been applied to these strata.

Although horizontal sections such as that of the coast, and others made up from the numerous quarries and pits, would indicate such a succession as that so well described by my colleague Mr. Ussher, and although all isolated sections agree in indicating a sequence of rocks of tolerably welldefined lithological characters, nevertheless we have as yet no positive proof that the separate divisions that are visible in the coast-section between Teignmouth and Axmouth are continuous one beneath the other, so that at the latter

1 These pebbles are much sought for on the sea shore, and are derived chiefly from the Breccia, but some perhaps directly from the Devonian Limestone.

locality all would be represented in vertical section were a boring carried through the series.

The indications of a succession of rocks furnished by the coast-section would in most instances be taken as sufficient evidence to warrant the conclusion that in travelling from west to east, newer and newer beds (excepting of course the repetitions occasioned by numerous faults) were exposed. But when the question of the method of accumulation of the strata comes to be solved some difficulties arise.

Thus we have in the country between West Somerset, and Torquay and Axmouth on the South Devon coast, bordering the Devonian and Carboniferous rocks, (1) a margin of breccia and sandstone, (2) one of marl, (3) one of pebble-beds and conglomerate, (4) one of sandstone, (5) one of marl.

It may be observed that nowhere are the conglomerates (3) far removed from the old margin, but the pebble-beds of Budleigh Salterton are seemingly the most distant from their parent rocks. The former are all of local derivation. The derivation of some of the latter is not exactly known; the fossils seem to indicate species of Devonian and Silurian (? Cambrian) age; and although at one time the pebbles were considered to have travelled from the French area, it is now considered that they may have been derived from rocks at no very great distance from the spots where the pebbles are now found; perhaps, as Dr. Hicks has suggested, from rocks destroyed in the formation of the English Channel.1

It is at least a significant fact when looked at in a large way that the coarser materials are nearer the old margin, as is the case with the Dolomitic Conglomerate of the Mendip Hills, Gloucestershire and Glamorganshire.

Nor have all the beds been traced persistently across the

1 The Rev. P. B. Brodie has recorded his discovery, in the Drift of Warwickshire, of pebbles of similar character and with similar fossils to those found at Budleigh Salterton.

country, for one or other of the members is not unfrequently absent, a feature which may be due either to faults or to attenuation. Now it would be impossible for anyone who casually observed the district to offer any opinion on this subject, and I have, I may say almost reluctantly, come to the conclusion, which my colleague Mr. Ussher had before surmised, that the local absence of members is generally due to faults.

The character of the Breccia itself seems to me to be entirely due to the nature of the bordering rocks which formed the old margin of the lacustrine area, and which consisted most largely of slates. Therefore no conclusions as to its age can be drawn from its lithological characters. Sir Roderick Murchison considered it to be Permian, agreeing with Conybeare and Buckland, who identified the Heavitree Breccia with the Roth-liegende of the Germans. It may indeed be difficult to disprove this if we regard the Permian and Trias as one system; and therefore, as the red rocks of Devonshire form a connected series, the use of the term Poikilitic may be preferable to that of simply Trias.

It is indeed a question of great interest as to whether the Red beds of Devonshire be altogether of Triassic age. The lower beds have sometimes been classed as Bunter, while the upper are undoubtedly Keuper, passing gradually upwards into the Rhætic beds east of Axmouth. There is no unconformability in the series, and the thickness would seem to warrant the idea that perhaps the Bunter and also the Muschelkalk may be represented, although without any possibility of their equivalent sediments being identified.

RHETIC OR PENARTH BEDS.

The Rhætic beds, so named on the Continent from the Rhætian Alps of Lombardy, where they are well developed, have been identified in England mainly through the researches of Dr. Wright and Mr. Charles Moore. The English and Welsh representatives of the series-the Penarth Beds, so named by Mr. Bristow in 1864-are comparatively feeble in development, but they are not on this account the less interesting and important. They form the connecting link between the Poikilitic and Jurassic strata, a fact of great significance, considering the different physical conditions under which these two great groups of strata were deposited. The Rhætic beds bind them together as one great conformable sequence of deposit. Much discussion has arisen as to whether these passage-beds are more closely allied to the Trias or to the Lias; there is no doubt, however, that they belong as much to the one formation as to the other. The Palæontological evidence, afforded by the fish-remains which occur in the middle shaly division, would, as was pointed out by Sir Philip Egerton many years ago, incline us to connect the Bonebed with the Trias; whereas the fossils of the upper strata, and the mollusca generally, would link the Rhætic beds with the Lower Lias, from which, nevertheless, they are distinguished in this country by the entire absence of Cephalopoda. As a matter of convenience, and to avoid giving too great a prominence to the strata, it seems best to include them with the Trias, as the uppermost formation in the Poikilitic system.

The Rhætic Beds consist of a series of marls, black shales and white limestones, having an average thickness of about 50 feet. The series admits of the three following subdivisions: