DXRW~ ON CORAL REEFS. 389

face level of the sea, in no one instance emerge above it. To escape this latter most improbable admission, which implies the existence of submarine chains of mountains of almost the same height extending over areas of many thousand square miles, there is but one alternative, namely, the prolonged subsidence of the foundations on which the atolls were primarily based, together with the upward growth of the reef-constructing corals. On this view every difficulty vanishes ; fringing reefs are thus converted in barrier reefs, and barrier reefs, when encircling islands, are thus converted into atolls the instant the last pin- nacle of land sinks beneath the surface of the ocean. " Finally, when the two great types of structure--the barrier reefs and atolls on the one hand, and fringing-reefs on tile other--are lald down in colours on the map, a magr, ificent and harmonious picture of the movements which the crust of the earth has within a late period undergone, is presented to us. We there see vast areas rising, with volcanic matter every now and then bursting forth through the vents or fissures with which they are traversed. We see other wide spaces slowly sinking without any volcanic outbursts, and we may feel sure that this sinking must have been immense in amount as well as in area, thus to have buried, over the broad face of the ocean, every one of those mom~tains above which atolls Pow stand, like monuments, marking the place of their former existence. Reflecting how powerful all agent with respect to de- nudation, and consequently to the nature and thickness of the deposits in accumulation, the sea must ever be when acting for prolonged periods on the land during either its slow emergence or subsidence: reflecting also on the final effects of these movements in the interchange of land and ocean-water on the climate of the earth, and on the distribution of organic beings, I may be permitted to hope, that the conclusions derived from the st.udy of coral forma- tions, originally attempted merely to explain their peculiar forms, may be thought worthy of the attention of geologists." Pages 146-- 1't8. D. T. A.

II. TRAVELS in NORTH AMERmA, with Geological Observations ou the UNITED STATES, CANADA, and NOVA SCOTIA. By CHARLES LYELL, Esq. F. R.S. 2 Vols. 12mo. pp. 588. Map and Plates. So far as regards geology, Mr. Lyell's work is eminently valu- able, since it presents a connected view of the results of a large number of careful surveys of different parts of the Continent of North America, by means of a coloured geological map, in which the whole amount of information at present known on the subject of North American geology is incorporated. The importance of this as a means of simplifying and general]sing the notions of English geologists with regard to the succession of strata on the other side of the Atlantic, it would be difficult to estimate too highly. Mr. Lyell has also added much to the knowledge hitherto possessed on the subject of American geology by his own investi- gations in the field, and in the present notice we propose to point out in order the various geological matters touched on in the work before us, commencing with the older rocks, and so ap- proaching last of all to those of newer date. The latter indeed, although perhaps the most important, we shall here scarcely allude to, because, having formed the subject of communications to the Geological Society, they either have been already or will be hereafter described in greater detail than even in the book itself cc3 890 NOTICES OF NEW BOOKS. in the Proceedings of the Society, as published in the various numbers of this Journal. The older Palaeozoic or Silurian rocks of America appear to be, as in England, divisible into two series--the upper and lower, and Mr. Lyell has admitted in his coloured map seven sub-divisions, which are thus named:-- {ii Hamilton group. Upper Silurian Helderberg series. Onondaga salt group. Niagara and Clinton group.

Lower Silurian Trenton, &c. limestone group. Potsdam sandstone, &c. The Taconic system, named by Dr. Emmons from a chain of mountains which form a continuation of the green mountains of Vermont, and supposed to represent a group of formations more ancient than the Silurian, are not considered by Mr. Lyell as deserving an independent place among the rocks of the palaeozoic series (vol. i. p. 246.). (1) The Potsdam sandstones are chiefly developed in a narrow band on tile south-eastern range of the great chain, extending south-west from the Vermont range and along the whole of that line ; and they immediately succeed and overlie the granitic and gneissese rocks of Virginia, North Carolina, and Georgia. They are also found occupying a broader space on the banks of the great river St. Lawrence, between the north-eastern extremity of Lake Ontario and the city of Montreal; and here these most ancient of the fossiliferous rocks are loaded with the remains of .Lingula, and a small placunoid shell nearly allied to a fossil also occurring in company with the Lingula in the lowest English silurian beds at Builth in Brecknockshire (ii. 157.) Beds of the same Lower Silurian date have also been traced on the banks of the Wisconsin river, a tributary of the Mississipi, and Captain Bayfield is inclined to consider a band of sandstone on the southern coast of Lake Superior as the equivalent of these oldest fossiliferous sandstones of Potsdam. (2) The Trenton limestone is much more widely distributed in North America than the inferior beds of sandstone; but it appears difficult to separate it in some cases from the beds of the over- lying Hudson river group (3)* (ii. 49). The "blue limestone," as it is called, forming the hills and table lands around Cincinnati and elsewhere in Ohio, belongs to the upper part of the lower Silurian group, and abounds in organic remains, consisting chiefly of trilobites, brachiopodous shells, crinoidea, and many corals, the latter differing considerably in specific character from those in the

The Hudson river group of Mr. LyelI ineludesa number of sandy and argillaceous slates containing Lower Silurian fossils, and separating in some eases the Trenton from the Niagara (Upper Silurian)limestones. LYELL'S TRAVELS IN NORTH AMERICA. 89i

Lower Silurian strata of England (ii. 50.) With reference to the subject of Silurian fossils, Mr. LyeU takes occasion (i. 20., ii. 51.) to combat the notion that species were more cosmopolitan at this early period than they are at present. (4) The Niagara and Clinton groups form the base of the North American Upper Silurian series, and are not represented on the eastern flanks of the Alleghanies, the uppermost Upper Silurian bed being there in immediate contact with the Hudson river series. A vast tract of country on the west of this chain is, how- ever, occupied with a group consisting of these Niagara beds associated with (5) the Onondaga, and (6) the Helderberg series, but the sub-division is not here made out, although the different members are well seen in detail on the southern shore of Lake Ontario. The tIamilton group (7), the uppermost of the Upper Silurian series, is everywhere seen coming out from beneath the old red sandstone, which it seems to enclose as in a basin. The Niagara limestone and shale correspond in their fossils with the Wenlock and Dudley limestone of England, and overlie the Clinton group, which might almost be looked upon as lower Silurian, but which it is thought better in the present state of palmozoic geology to class with the upper members of the series. The Onondaga salt group (5)is a remarkable formation of red and green argillaceous shale, marl, and shaly limestone, sometimes of great thickness, but partially developed;and the Helderberg series (6), although consisting in the State of New York of a number of distinct beds, passes so insensibly into the lower group towards the west, that, as we have already observed, no well-defined line of distinction can be drawn. The Hamilton group (7) includes some shaly and slaty beds con- taining Ludlow fossils, and is widely distributed. It concludes the great Silurian series of North America, concerning which it may be remarked on the whole that with regard to the Alleghanies, the inferior or older beds range chiefly along the eastern or south- eastern flank, and are distinctly marked, while the newer groups of the same series, togegher with the Devonian and carboniferous formations, make their appearance as we proceed further west- ward. (ii. 9, 10.) The rocks of the Devonian period--the old red sandstone series of North Britain and Herefordshire -- are exhibited in a prominent and characteristic form, surrounding each of the great coal fields of the United States, and perhaps no where more strikingly than in the State of New York not far from Niagara. These beds here consist of olive-coloured slate and grey sandstone, containing occasionally impure coal, and in some sandstones (seen near Tioga) fragments of more than one species of Holoptychius have been obtained associated with a Chelonichthys of large dimensions. In Ohio, at the distance of 400 or 500 miles to the south-west, the author was struck by the extraordinary decrease in volume of the whole group, the absence of some formations, and the complete identity of those sets of strata that remained. In cc4 392 NOTICES OF NEW BOOKS. the sandstones of this period many ripple marks are found, and the surface of the slabs is frequently covered with fucoidal impres- sions; but fossils appear to be rare. It is chiqfly with regard to strata of the carboniferous period that the investigations of Mr. Lyell in North American palaeozoic geology must be considered important. The coal fields of the United States and the British possessions in Canada are beyond all comparison the most extensive and the most valuable of any at present known in the world ; and the geological position of these deposits of vegetable matter, as well as the conditions under which they occur, are matters of very considerable importance to the future interests of the continent of America. The great coal fields of the United States are the Appalachian, the Illinois, and the Michigan: those of the Canadas are on the eastern extremity of the colony, and occupy a great space in New Brunswick, Prince Edward's Island, Cape Breton, &c. A large proportion of the coal is anthracitic. It is not easy to do justice to these formations by the hasty sketch to which our limits neces- sarily confine us in this place ; but some idea may be formed of their extent, when it is stated that the Appalachian coal fields ex- tend for a distance of 720 miles f?om north-east to south-west with a width in some places amounting to 180 miles * ; that the Illinois basin is not much inferior in dimensions to the whole of England; that the Michigan coal fields and the coal fields of Canada-are also of very considerable dimensions; and that the thickness of the seams of fossil vegetable matter in some instances exceeds even that of the Staffordshire coal, amounting at the Lehigh summit mine (where the usually intervening shales and grit have thinned out) to one mass of fifty feet without any greater interpolated matter than two thin layers of clay. In some places this vast bed is quarried in the open air ; but in others, where the coal is accessible to a degree scarcely to be imagined by strangers to tlle conditions of the country, the time has not yet arrived when the value of it as fuel is appreciated. (See vol. ii. p. 26, 27.) One of the most remarkable facts, geologically, with reference to the carboniferous deposits of North America, is the great abundance and excellence of the anthracitic coal, met with more especially in the Alleghanies; but as Mr. Lyell's views and con- clusions on this subject were communicated to the Geological Society shortly after his return to England, and have since been published in this Journttl (see ant~, p. 199.), it is hot necessary here to dwell upon or recur to them ; we rather prefer directing the attention of the reader to the coal fields of Nova Scotia as being a district in which our author has made most important additions to our knowledge of palmozoic geology, and as the subject offering the greatest amount of new matter for consideration. The carboniferous group of formations, as developed in Nova Scotia, may be conveniently divided into three series, namely : --

) Its ~uperficial area is ea]culated at 63,000 miles. LYELL'S TRAVELS IN NORTH A~IERICA. 393

The upper, composed of sandstone and shale, with fossil plants. The middle, comprising the productive coal measures. The lower, chiefly made up of red sandstone and red marl with subordinate beds of gypsum and limestone. Concerning the lower series, which extends to the island of Newfoundland (where, however, the productive coal measures hardly reach), its position was not understood at the time of Mr. Lyell's departure for America; and from the peculiar character of its rocks and the abundance of gypsum it contains, it was supposed to overlie rather than underlie the true carboniferous rocks. Imme- diately after his return to England, our author communicated to the Geological Society his opinion that the whole group was a true lower member of the carboniferous series, and this opinion is fully borne out by the subsequent careful observations of Mr. Dawson already recorded in the pages of this volume (ant~, p'. 26.), and admitted by Mr. Brown (ibid. p. 23), and Mr. Logan, al- though, according to Dr. Gesner, there is still a want of such evidence as is satisfactory to him. The determination of this point is of very great importance in working out the geology of :North-eastern America; and the similarity now known to exist between some of the fossils from beds above the carboniferous series and those at the very bottom of that series, together with some obscurity in the sections in a very difficult country, render it advisable to quote in this place the heads of the evidence on which Mr. Lyell's conclusions are founded. (vol. ii. p. 204.) The first argument offered in the work before us is, that the rocks of what we may safely call the, gypsiferous formation, in Nova Scotia and Prince Edward's :Island, always make their ap- pearance nearer to the region occupied by the older rocks, whether Silurian or Metamorphic, and also that they are, on the whole, more disturbed than the coal measures. In attestation of this point, sections are referred to on the East River, and the Minudie, and also near Windsor, the rocks in the latter case being fos- siliferous, and their evidence therefore of greater value. In the cliffs near this town, Windsor and in the estuary of the Shubenacadie, these strata are very greatly disturbed and thrown into folds, part of which are tilted at considerable angles, while the rocks are fissured in the direction of their strike and shifted verti- cally. It would appear that there are no indications ot' the true coal measures having partaken of these disturbances, so that on both these grounds, and also from the probable continuation of the dip observed near the Minudie already referred to, the evidence all points to the same conclusion with regard to the age of the rocks in question. The Albion mines near Pictou are remarkable for the great thickness of the coal worked there (estimated at thirteen yards). To the south of this district, Mr. Logan and other geologists have, it

* These uppermost beds, as they occur in Nova Scotia, are described by Dawson in the present number of tile Journal. (Se a~td, p. 322.) 394 NOTICES OF NEW BOOKS. would seem, distinctly made out an underlying group of red sand- stones and marine limestones, some of the latter beds having an oolitic structure, and the former being associated with abundance of redmarl. Mr. Lyell also states, as a second argument in favour of his view of the age of the gypsiferous beds, that he found most of the fossils of the limestones of l~ictou to agree with shells and corals obtained by him in the Umestones near Windsor, and in those of the Shubenacadie accompanying the principal masses of gypsum ; while in these same limestones he found an intimate association between strata containing carboniferous fossils, masses of gypsum, and coal grits with Sigillaria and Lepidodendron, but no seams of pure coal. It appears also, in conclusion, that although overlying the coal measures there is a formation of red sandstone, generally without fossils, and often unconformable with the lower beds, there are no great masses of gypsum or beds of marine limestone associated with it, and nothing in fact that can identify these with the gypsiferous beds already described. There seems, there- fore, little doubt that Mr. Lyelrs views, coinciding as they do with those of Mr. Dawson, Mr. ~rown, Mr. Logan, and other careful observers, are correct, and that the peculiar appearance and the minerals characteristic of the newest palaeozoic or oldest secondary groups in the old continents are in :North America found in beds of a much more ancient date, and correspond with the lowest carboniferous slates of England and Europe, the beds of marine limestone in America being the imperfect representations of our carboniferous limestone. It is the middle of the three principal groups of carboniferous rocks which in :North America contains the productive coal measures. Confining ourselves chiefly to their appearance in Nova Scotia, we find them there developed to a vast extent, and the whole series is admirably exhibited on the banks of the South Joggins River, near Minudie, a locality otherwise remarkable in a geological sense, on account of the presence of numerous fossil trees embedded in an upright posture, and at several different levels. Mr. Lyelrs attention had been drawn to the circumstances of these fossil trees by Dr. Gesner, and he was naturally desirous of satist~ng himself with regard to so very singular a phenomenon. We extract portions of Mr. Lyell's account of his visit to Minudie having reference to the subject of these fossil trees. From Minudie a range of perpendicular cliffs extends in a south-westerly direction along the southern shore of what is called the Chignecto Channel. The general dip of the bed is southerly, and the lowest strata near Minudie consist of beds of red sandstone, with some limestone and gypsum. The sec- tion is then obscure for about three miles, and then still further towards the south there comes in a vast seriesof newer and conformablebeds (the productive coal measures), amongst which the upright trees occur; the same series con* tinues many miles further to the south. If we assign a thickness of four or five miles to thid regular succession of carboniferous strata which must have been originally quite horizontal, our estimate will probably be rather under than over the mark. Where the section in the cliff is first well exhibited, we find about 44 feet of blue grit affording an LYELL'S TRAVELS IN NORTH AMERICA. ~9~ excellent grindstone, and about a mile further to the south, at a distance of about six miles from Minudie, the first of the upright trees appears. Then follows a series of coal-b,.aring strata, containing about nineteen seams of coat, oceupylng a range of coast about two miles long. At low tide a horizontal see- tion of these beds is seen on the beach, and their edges in the vertical pre- cipices. The beds through which erect trees or rather the trunks of trees placed at right augles to the planes of stratification are traceable, have a thickness of about 2500 feet, and no deception can arise from the repetition of the same beds owin~ to shifts or faults, the section being unbroken, and the rocks with the exception of their dip quite undisturbed. In the first ofthese uplight trees which 1 saw, -no part of the original plant is preserved, except the bark, which forms a tube of pure bituminous coal, filled with sand day and other deposits, now forming a solid internal cylinder without traces of organic structure. The bark is a quarter of an inch thlek, marked externally with irregular longi- tudinal ridges and furrows, without leaf-sears, and therefore not resembling the regular flutings of Sigillari~, but agreeing exactly with the description of those vertical trees which are found at Dixontbld, on the Bolton railway, of whleb Messrs. Hawkshaw and Bowman have ~ven an excellent account in the Pro- ceedings of the Geological Society of London.* On comparing Mr. Hawkshaw's drawings of the British fossils, in the library of the Geological Society, as well as a speimen of one of the Dixonfold trees presented by him to their museum, with portions of the bark brought by me from Nova Scotia, I have no hesitatiorJ in declaring them to be identical. The diameter of the tree was 14 inches at the top and 16 inches at the bottom, its height 5 feet 8 inches, and tile strata in the interior consisted of a series en- tirely different from those on the outside. Mr. Bowman has explained in the ~Ianehester Transactions the causes of the frequent want of correspondence in the strata enclosing a buried tree, and the layers of mud and sand accumulated in the interior, which vary according to the more or less turbid state of the water at the periods when the trunk de- cayed and became hollow, and according to the fieight to which it was pro- longed upwards in the air or water after it began to be imbedded externally in sediment, and various other accidents. It is not uncommon to observe in Nova Scotia, as in England, that the layers of matter in the inside are fewer than those without. Thus, a "pipe" or cylinder of pure white sandstone, re- presenting the interior of a fossil tree, will sometimes intersect numerous alter- nations of shale and sandstone. In some of the layers in the inside of the trunk, I saw leaves of ferns and fragments of plants which bad fallen in together with the sediment. Continuing my survey, I found the second of the erect trees separated from the first by a considerable mass of shale and sandstone. This second trunk wa~ about 9 feet in length, traversing various strata, aud cut off" at the top by a layer of day 2 feet thick, on which rested a seam of coal 1 foot thick. This coal formed a foundation on which stood two large trees, about 5 yards apart, each about 2½ feet in diameter and 14 feet long, both enlarging downwards~ and one of them bulging considerably at the base. The beds through which they pass consist of shale and sandstone. The eliWwas too precipitous to allow' me to discover any commencement of roots, but the bottom of the trunks seemed to touch the subjacent coal. Above these trees were beds of bituminous shale and clays with Stigmaria, l 0 feet thick, on which rested another bed of coal 1 foot thick, and this coal supported two trees, each 11 feet high, and 60 yard~ apart. They appeared to have grown on the coal. One of these, about 2 feet in diameter, preserved nearly the same size from top to bottom, while the other, which was about 14 inches in diameter at the top, enlarged visibly at the base. The irregular furrows of the bark were an inch and half one from the other. The tops of these trees were cut off" by a bed of clay, on which rested the main seam of the South Joggins coal, 4 feet thick, above which is another succession

* Voi. iii, I~P. 139. 270. 396 NOTICES OF NEW BOOKS. of strata, very similar to those already described, with occasional thin seams otv coal, and with vertical trees at five or six different levels. I observed in all at least seventeen of these upright trunks, but in no instance did I see any one of them intersectir/g a layer of coal, however thin, nor did I find any one of them terminating downwards in sandstone, but always in coal or shale. Their usual height was from 6 to 8 feet, but one which was more than 100 feet above the beach, and which I could not approach to measure, seemed to be 25 feet high, and 4 feet in diameter, with a considerable bulge at the base. They all appear to be of one species, the rugosities on the surface producing the effect of a rudely-fluted column, and they were placed very ac- curately at right angles to the planes of stratification. I found numerous flattened trunks of large Sigillari~e with their flutings and leaf-scars in the shales; but none of them resembled the erect trees with their irregularly furrowed exterior. Stigmatize are abundant in the argillaceous sandstones of these coal-measures, often with their leaves attached, and spreading regularly in all directions from the stem. It commonly happens here, as in Europe, that, when this plant occurs in sandstone, none of its leaf-like processes (or rootlets ?) are attached, but I saw one remarkable exception in strata of micaeeous sandstone, in which the stem was about 4 inches thick, and traversed obliquely several layers of fine white micaceous sandstone 2 feet in vertical thickness. I have stated that I counted seventeen upright treesin the strata of the South Joggins, and I was assured hy Dr. Gesner, and by residents at Minudie, that other and different individuals were exposed a few years ago ; the action of the tides of the Bay of Fundi being so destructive as continually to undermine and sweep away the whole face of the cliffs, so that a new crop of fossils is laid open to view every three or four years. I saw the erect trees at more than ten distinct levels, one above the other ; they extend over ;l space from two or three miles from north to south, and more than twice that distance from east to west, as I am informed by Dr. Gesner, who has explored the banks of streams inter- secting this coal-field. Many curious conclusions may be deduced from the facts above enumerated. 1st. The erect position of the trees, and their perpendicularity to the planes of stratification, imply that a thickness of several thousand feet of strata, now uniformly inclined at an angle of Y4°, were deposited originally in a horizontal position. But for the existence of the upright trees it might have been con- jectured, that the beds of sand and mud had been thrown down at first on a sloping bank, as sometimes happens in the case of gravel and coarse sand. But, if we are compelled to assume the original horizontality of beds 2500 feet thick, through which the erect trees are dispersed, we can hardly avoid ex- tending the same inference to the greater part of the strata above and below them. It by no means follows that a sea four or five miles deep was filled up with sand and sediment. On the contrary, repeated subsidcnces, such as are required to explain the successive submergence of so many forests which grew one above the other, may have enabled this enormous accumulation of strata to have taken place in a sea of moderate depth. Secondly The evidence of the growth of more than ten forests of fossil trees superimposed one upon the other prepares us to admit more willingly the opinion, that the Stigmaria with its root-like processes was really the root of a terrestrial plant fossilised in situ. Yet, if we embrace this opinion, it follows that all the innumerable underclays with Stigmaria in North America and Europe, are indications of an equal number of soils, whether of dry land or freshwater marshes, which supported a growth of timber, and were then sub- merged. Ifthls be true, and the Conclusion seems inevitable, the phenomenon of the upright trees in Nova Scotia, marvellous as it may be, shrinks into insig- nificance by comparison. At the same time, it is quite intelligible, that we should find hundreds of caseswhere the soil has remained with the roots fixed in their original matrix for one instanee where the trunk has eontinued to stand erect after submergence. Many favourablc circumstances must concur, to allow of such an exception to LYELL'$ TRAVELS IN NORTH AMERICA.. 397

the general rule. There must, for example, be an absence of waves and currents of.sufficlent strength to loosen and overturn the trees, and the water must be charged with sediment ready to envelope the plants before they have had time totally to decay. I have shown that on the coast of S. Carolina and Georgia the land sunk in modern times, and that buried trees are occasionally found in strata containing shells of recent species. The formation of low islands of saml off the shore, breaking the force of the Atlantic, has probably allowed many of these trees near the mouths of estuaries to" continue erect under water, until they were silted up and preserved, Similar low islands and sandbanks skirt nearly the whole of the eastern coast of the United States, and may assist the geo]oglst in explaining some of the phenomena of the Carboniferous period, especially the manner in which superficial beds of vegetable matter, as well as upright t~ees, escaped the denuding forces. Thirdly. It has been objected to the theory which refers the origin of seams of pure coal to plants which grew on the exact spaces where we now find coal, that the surfaces of ancient continents and islands ought to undulate like those we now inhabit. Where, they ask, are the signs of hills and valleys, and those river-channels which cut through deltas? These apparent difficulties will, I think, be removed, if we reflect that the fossilisatlon of successive forests pre- supposes both the subsidence of the ground and the deposition of sediment going on simultaneously. If so, the accumulation of mud and sand furnishes us with the levellil)g power required, and, had there been extensive denudation capable of produciug valleys, it could readily have swept away all the coal. In regard to ancient rlver-courses, tile late Mr. Bud(tie often assured me, that he had in many places met with them in the coal-fields of the North of England, and he has given a detailed account of one which intersected a seam of coal in the Forest of Dean. Ev'en in these cases, however, the general evenness of the surface is immediately restored by a new sinking of the delta, and the depo- sition of fresh sedime,~t, so that the succeeding seam of coal has grown on as perfectly flat a surface as if there had been no partial destruction of the beds below. If it he objected that, according to the analogy of recent subterranean move- ments, some areas ought to have sur,k down at a more rapid rate than others, producing irregularities in the ancient level of the dry land, we reply, that there are abundant proofs in the arrangement of the carboniferous strata, that the amount of local subsidence was actually not uniform. Mr. Bowman has clearly pointed out, that the wedge-shaped or lenticular masses of sandstone and shale, which sometimes intervene between the upper and lower portions of a seam of coal, are the natural result of such inequalities in the downward move- ment. In those areas which sink so fast as to be submerged, the growth of terrestrial plants is suddenly arrested, and the depressed region becomes the receptacle of sediment, until its level is again raised. Then the growth of the former vegetation is resumed, and the result is, the intercalation of strata for a certain space between two beds of coal, which unite and become one, if they are followed to a certain distance in every direction. Vol. ii. pp. 179--193.

The continent of America is well known to exhibit but very few indications of rocks of the secondary period, or at any rate of the triassic and oolitic series. Still the lower beds are by no means wanting, and the red sandstone of Connecticut, although it has been referred by some geologists to the Permian system, and is, perhaps, to a certain extent, a passage bed, appears to be, on the whole, more properly referred to the trias, and must, at least for tile present, be looked upon as of that age. This rock has long been celebrated for the impressions of footsteps of birds with which its slabs are in many places covered, and Mr. Lyell mentions that since these impressions were first announced by Professor Hitcheock in 398 NOTICES OF NEW BOOKS.

1836, and referred by him to birds, he (the Professor) has observed above 2000 footprints, probably made by nearly thirty distinct species. (Vol. i. p. 254.) The beds of this age in Connecticut include a fine-grained slaty sandstone, black and bituminous, and about six feet thick, which alternates with a coarse conglomerate. Small fragments of fossil wood and a ripple-marked surface were ob- served in some of the strata near the fossil fish. The only other rocks considered to belong to the older secondary period in North America are the coal measures of Eastern Virginia referred by Professor W. B. Rogers to the Oolitic period, but concernin~ these Mr. Lyell does not offer an opinion. Our author's views of the cretaceous rocks of America will be found fully expressed in the earlier pages of this volume (see ant~, p. 55.), and the tertiary strata will be described in similar detail in the next number. In addition to those accounts of descriptive geology which come properly under consideration in speaking of the succession of strata, Mr. Lyell's work also contains some interesting facts and observa- tions on the subject of changes effected, in comparatively modern geological epochs, by means of causes now in action. The geo- graphical and geological features of the great Falls of 1Niagara occupied, as might have been expected, a considerable portion of his attention. With regard to the origin of these Falls, and the rate of their recession, (for of the fact of their recession there is distinct proof), we also have some speculations which cannot fail to be viewed with considerable interest. The superficial covering of gravel, the presence of erratic blocks, and the existence of horn- blendic and syenitic rocks, polished, furrowed, and striated, as seen in ttle north of Europe and in Switzerland, are also described and referred to by the author, in the course of his various journeys and traverses in various parts of North America. Numerous localities, in addition to the well-known salt-lick of Kentucky, are mentioned, in which the remains of that singular animal the Mastodon have been met with, frequently indeed associated with materials which prove how very recent must have been file date of its existence, as a regular inhabitant of that part of the world. It appears that in some places near Niagara the bones are found associated with shells, all of existing species now com- mon in the district, but buried in shell marl below the peat, and therefore agreeing in position with the large fbssil elks of Ireland. The extensive swamps of North America afford matter of con- ~ideration to the geologist, and Mr. Lyell gives (vol. i. p. 143.) an account in some detail of one of the largest of them, known as the Great Dismal Swamp. This singular expanse of soft and muddy quagmire extends forty miles from north to south, and is m some places twenty-five miles broad. It is at a higher level than the surrounding dry land, and, even in spite of its semi-fluid character, is about twelve feet higher in the middle than towards the margin. The conditions of existence of this swamp and its BRODIE ON FOSSIL INSECTS. 399 spongy vegetable soil suggest to :Mr. Lyell the possibility that we may in this way explain the manner in which vegetable matter has been accumulated to form coal. The instances of drift and erratic blocks in :North America, quoted by. Mr. Lyell, are too numerous to recount. The most remarkable instance perhaps of their occurrence in great abundance is in Long Island, where excavations lately made have exposed the boulder clay to the depth of thirty feet, and in this case the blocks are, some of them, of very large size; one which is mentioned (although not the largest), measuring 13 feet long, 9 feet broad, and 5 feet thick. Beneath the ordinary boulder clay, there appears to exist a red drift, the detritus of the new red sandstone formation of :New Jersey, and this mass is also sometimes of con- siderable thickness. The boulder formation of Long Island contains blocks of very different mineral character in different districts, but the source may in most cases be traced, at least with very considerable probability. In the work beibre us, the geological facts and observations are narrated in the order in which they were noticed and made. We have endeavoured to place before the reader an account of the materials thus brought together, leaving out of view the thread of personal narrative on which the geology is strung, and hoping in this way to add to the usefulness of what we consider to be the great characteristic of the work, the geological map of North America which has been prepared for its illustration. D. T. A.

III. A History of the FossIL INSECTS in the Secondary rocks of FZ~QLAND, accompanied by a particular account of the Strata in which they occur, and of the circumstances connected with their preservation. By the Rev. PETER BELLINGER BaODIE, M.A., F. G.S. 8vo. pp. 130. 11 Plates. T~E author of this work has been remarkably successful in the discovery of minute fossils, and is well known to geologists as having worked out points of detail, which by most people would be passed by unregarded. Dwelling for a time in the vale of Wardour, he there discovered in the Wealden beds of the Purbeck series, a curious genus of Isopodous Crustaceans (.4rchveoniscus), first described from these specimens in the _&nnales des Sciences by M. Milne Edwards, and also a bed actually loaded with frag- ments of fishes, and the small crustacean of the weald, called Cypris, but containing also the remains of insects. Afterwards, in the vale of Aylesbury, he observed a bed likewise containing re- mains of insects, and occupylng the same geological horizon. Re- moving then to Gloucestershire, this indefatigable searcher after the infinitesimal in Palmontology very soon succeeded in obtaining from