way in which rock-salt is likely to have been deposited in quantity except in salt-lakes, it seems to me that in the occurrence of this salt we have evidence of the upheaval of our Bunter beds into land, which remained above the sea for an indefinitely long period, during which the Muschelkalk and St. Cassian beds were deposited elsewhere, and in certain inland hollows of this land the salt was precipitated from the saturated waters. If so, it is just possible that the saltbeds may be partly the equivalents of the missing strata. But, however this may be, if the Bunter beds formed land surrounding saturated salt-lakes, we have in that circumstance a good reason why the truly marine St. Cassian beds and Muschelkalk should be absent in Britain.

Westbury or Bone-bed series, or Kössen and Rhætic beds.—I now come to those members of the British secondary rocks which are essentially fossiliferous; and before entering on the subject I wish to express my great obligations to my colleague, Mr. Etheridge, without whose wonderful knowledge of secondary species and their range in time it would have been impossible for me with accuracy to have constructed the tables on which much of my reasoning is founded*.

The Bone-beds were considered to form the base of the British Lias until, in the year 1841, they were declared by Sir Philip Egerton to form part of the Keuper series, in consequence of the generic characters of certain Fishes which they enclose. But, notwithstanding this correct determination, they were by most men persistently classed with the Lias, until, in 1860, by an elaborate analysis of the Shells of these strata and of the overlying Lower Lias, Dr. Wright again demonstrated their intimate relation with certain Triassic strata of the Continent.

In consequence of the want of frequent good sections at the base of the Lias, it is uncertain how far these Westbury or Kössen strata are continuous between Lyme Regis on the south and the coast of Yorkshire. This uncertainty is due partly to the soft and clayey nature of the strata, but especially because over more than half of England this base is hidden completely by the northern and other drifts. The chief points where the Westbury beds have been recognized are-Lyme Regis, in Dorsetshire; Up Hill, Watchett, Kelmersden, and Beer Crowcombe, in Somersetshire; Garden or Westbury Cliff, Wainlode Cliff, Combe Hill, Aust Passage, Witworth, and near Tewkesbury, in Gloucestershire; Penarth Cliff, in Glamorganshire; Wilmcote and Harbury, in Warwickshire; and in Needwood Forest, in North Staffordshire. As a general rule, however, from the south part of Leicestershire to Yorkshire these beds have not yet been observed, although when looked for they may perhaps be discovered. Important as this subformation is in a paleontological point of view, it nowhere attains a thickness of more than from 20 to 50 feet; and it is therefore the more remarkable that throughout all the

* I have for years used such tables in my public lectures; but those now employed for my old tables are improved; and the secondary ones are entirely reconstructed from tables, many yards in length, by Mr. Etheridge, in which he shows the range in time of every individual secondary species.

middle and south of England it seems to be perhaps everywhere constant between the top of the New Red marl and the base of the undoubted Lias.

The number of fossils from these beds in England is considerable, comprising 32 genera and 51 species of Shells, 6 genera and 7 species of Fishes, 2 Reptiles, and one Mammal; in all, 41 genera and 61 species. Of these, 13 genera of Mollusca are common to the St. Cassian beds, and 18 to the Muschelkalk, and one species only, Monotis (Avicula) decussata, is common to all three formations, and it is also said to be found in the Lower Lias. Between the New Red marl and the Westbury or Bone-beds no appearance of unconformity has been noticed, though there is a perfectly sudden break lithologically from red marl to black shales; but as we have no fossil Shells in England in the marl, and no Shells at all by which to compare the two formations, there is nothing to be said about the sudden appearance of the new fauna, except that, unless some strata be absent intermediate in date between our Red Marl and the Westbury (Kössen) beds, the lithological change and the difference in colour between the marl and the shale indicate a change in conditions accompanied by a fauna that migrated hither from some other area. Though specifically distinct, yet nearly half of the genera of that fauna being common to the Muschelkalk, the triassic affinities of the group are sufficiently strong.

Lias. It is remarkable, however, that when we come to the Lower Lias, only one species, mentioned above, is common to the two formations, even though there is no marked lithological break, and no unconformity has been observed between them. The exposed sections, however, are so scarce (the junction only having been observed in the southern half of England), that it is not impossible that overlaps may exist in places.

The Lower Lias averages from 600 to 900 feet in thickness, and has been always mapped as one indivisible series of clays, shales, and limestones. Nevertheless paleontologists, led by Quenstedt and Oppel (whose views have in England been worked out by Strickland, Wright, and others), have attempted, not unsuccessfully, to divide it into six zones, each of which is partly marked by its own group of fossils, especially by certain Ammonites.

The Middle Lias, or Marlstone series, in like manner has been divided (not mapped) into three zones; and

The Upper Lias into two zones, the lower being the Upper Lias clay, and the upper the sand, the latter of which, until the appearance of Dr. Wright's paper, was considered a subdivision of the Inferior Oolite.

From Devon and Dorsetshire to Yorkshire all of these divisions are said to be constant, except where, from accidental circumstances, any one of them may chance to lie on some protruding boss of palæozoic rock; and, from bottom to top of the whole series, we cannot assert that anywhere is there actual unconformity between any two of the subdivisions, whether they be analyzed minutely or taken as three grand divisions of Lower, Middle, and Upper Lias.

Zones.

TABLE II. Showing the upward Passage of Species through the Zones of the Lias.

Neither is there any marked sudden change in the lithology of the rocks, except between the brown limestone of the Marlstone proper and the Upper Lias clay.

Now in all the Lias formations of England there are about 243 genera and 467 known species; and if we construct a table of zones distinguished by special Ammonites, after the manner practised by Oppel and Dr. Wright, we find the results shown in the following Table,-premising that none of the species of Ammonites have yet been observed in any zone lower than that in which it is mentioned as commencing; and also that for the species of every kind, those which are common to a low and a high zone, but have not been found in intermediate subdivisions, are assumed to have existed during the deposition of the intermediate members here or elsewhere. For instance, a species passes from the zone of A. Planorbis into that of A. spinatus, the upper zone of the Middle Lias; and though this species may not have been seen or catalogued in any of the intermediate zones, it is clear that it must have existed during those times, because no one believes that any species, having once passed away, has been re-created or redeveloped. Furthermore, in the

Table the same species is never repeated in any two columns throughout the whole series.

On the principles stated above, it appears that 17 species, or rather more than 40 per cent., pass upwards from the zone of Ammonites Planorbis, all of which must have lived through the period of time of the zone of A. Bucklandi. From the zone of A. Bucklandi 20 species, or 18 per cent., pass upwards; from that of A. Turneri 6 species, or more than 50 per cent., pass on; from that of A. obtusus 2 species, or more than 25 per cent., pass; from the zone of A. oxynotus (which has only yielded 6 species) none have been observed higher in the Lias, while from that of A. raricostatus, which forms the top of the Lower Lias, 6 species; or about 33 per cent., pass upwards into the Middle Lias.

In the Middle Lias, from the zone of A. Davoi, or A. Henleyi, 31 species pass upwards, or very nearly 38 per cent.; and from the zone of A. margaritatus 8 species out of 38, or 21 per cent., go into higher horizons. From the zone of A. spinatus (the top of the Middle Lias) 2 species only, or about 5 per cent., pass into the Upper Lias; and from the zone of A. communis (the base of the Upper Lias) 15 species, or more than 27 per cent., pass upwards into the zone of A. Jurensis or the Upper Lias sands. From the Upper Lias 13 species, or more than 21 per cent., pass into the Inferior Oolite.

The zone of A. oxynotus, in which only 6 species are recorded in England, is one but little opened for economic purposes, and this may perhaps account for the scarcity of its fossils. Under these circumstances it is scarcely fair to argue upon it; but, with this exception, I am on the whole struck with two main facts, the first of which has been well stated for England by Dr. Wright, namely, that as regards certain species of Ammonites the lines are apt to be trenchant, though other species in a limited degree pass up and down; and, in the second place, a large proportion of the common

species graduate upwards from one division to another, in numbers varying from about 20 to 50 per cent. The only sound exception to this is where a sudden lithological break takes place between the Marlstone and the Upper Lias, in which case only 5 per cent. pass onwards into the newer formation. Further, it must be remembered that, when the Liassic fossils were named, the tendency among paleontologists to extreme subdivision of species was so strong, that there can be little doubt that trifling varieties have often been exalted into individual species, and the numbers common to more than one subdivision are doubtless larger than I have stated.

Considering that no actual unconformity is known from the bottom of the Lower to the top of the Upper Lias, some geologists may find it difficult to perceive why there should be even such partial breaks in the succession of species, except on the hypothesis that a large proportion of the old species were in each case destroyed at the close of one formation and replaced by the creation of new forms when the succeeding formation began. But I cannot accept this wellworn hypothesis; and in this matter I incline to the idea that, considering the frequent large percentages of passage (ranging as high as 50 per cent.), we are justified in supposing that migration of what were old species here into new areas elsewhere, and of certain older species from other areas into ours, may account for the very incomplete breaks in the succession of Liassic life in England, more especially if, as may have often been the case, there were occasional pauses in the deposition of the strata. An analysis, such as I now attempt, in other regions of Europe will alone help to clear up the question.

Passing upwards in the secondary series we come to the Oolitic rocks, the minor and greater divisions of which are singularly well established.

Oolitic strata. First comes the Inferior Oolite, which, both on palæontological and physical data, has been divided at Cheltenham into the following minor zones in descending order :

Ragstone, zone of A. Parkinsoni.

Upper Freestones, zone of A. Humphresianus.

Zone of A. Murchisonic

Upper Trigonia-grit.
Gryphite grit.

Lower Trigonia-grit.

Oolite marl.
Lower freestones.
Pea-grit.

It is interesting to find, in the first place, that there is no complete break in succession between the Upper Lias and the Inferior Oolite, 13 species, or about 21 per cent., passing from the lower to the higher formation. It would also repay the trouble were any one to analyze the paleontological differences between the subdivisions of the Inferior Oolite, but these differences are so analogous to those of the Lower Lias, that I forbear to state them at present. The percentages of Mollusca common to the various subdivisions are large; and, therefore, though it would be instructive to show the