The Preservation of the Neogcne Brassington Formation of the Southern Pcnnines and Its Bearing on the Evolution of Upland Britain

The preservation of the Neogcne Brassington Formation of the southern Pcnnines and its bearing on the evolution of Upland Britain

PETER THOMAS WALSH, MICHAEL CHARLES BOULTER, MOHAMMED IJTABA & DAVID MARTIN URBANI

CONTENTS Introduction • • • 520 2 The stratigraphy of the Brassington Formation 522 3 The subsidence mechanisms 524 (A) General . 524. (B) The nature of the surface on which the Brassington Formation was formed 525

(c) Structural influences of the host rocks . • . 526 (D) The structure of the Bees Nest Pit subsidence complex 527 (E) Subsidence simulation tests 527 (F) Conclusions 528 4 The sub-Ncogcnc surface in the southern Pcnnines 53o (A) Theory . . 53o (B) Critical sections 532 (c) Palaeontological evidence . 533 (D) Mineralogical evidence . 535 (E) The elevation of the surface on which the Brassington Formation was

formed . • . . 535 5 The denudation chronology of the southern Pennines .... 537 6 The relationships with other British Neogene deposits .... 54o 7 The implications for the evolution of Upland Britain 54 x 8 References 548

SUMMARY The Brassington Formation of the southern sands of the Brassington Formation is uncon- Pennines comprises a succession of sands, gra- formable and, characteristically, the latter are vels and clays, the youngest bccls of which seen to rest on zones of lilac-stained weathered contain Lower Pliocenc fossil plants. These shale, representing a pre-subsidence land sur- strata arc generally considered to have found- face. crcd into the Carboniferous Limestone through Evidence is presented which indicates that solution subsidence mechanisms. The sub- the Brassington Formation strata arc conform- sidcncc cavities ~Iso contain detached marginal able and that the succession represents a single blocks of shale, which locally contain Namurian cycle of sedimentation. Because of this con- goniatites. | formablc relationship it is inferred that this Examination of the shale blocks from a num- land surface is also of Lower Pliocenc (or ber of widely separated localities has revealed Upper Miocene) age. that all contain Namurian spore assemblages, A tachcomctric survey of a well-exposed datable in some ~:ases as Ez (Arnsbergian) age. solution subsidence complex at the Bccs Nest The contact between the shale blocks and oldest Pit, Brassington, has enabled structure contour

J. geol. Soc. Lond. vol. x28, I972 , pp. 519-559, 7 figs., pls. I- 4. PHnted in Northern Ireland.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 520 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

maps to bc made of key horizons in the sites of the present subsidence masses. Subsid- succession. Laboratory tests have been con- ence can be shown to be of the order of 15o- ductcd which have attempted to duplicate 25o m, which implies that the Lower Pliocene this structure and so indicate the probable land surface of the southern Pennines lay at subsidence mcchanisms. Results suggest that about 45 ° m O.D. subsidence was gradual, with block-by-block Based on the evidence of the contemporary removal of the host limestone, and took place sea level, the presence of the Lower Pleistocene in deep, steep-sided pipe-like forms. deposits of s~ England, and the sub-Neogene The depth of subsidence, and hence the surface of the southern Pennines, a graph has clcvation of the land surface prior to the depo- been constructed which suggests broadly that sition of the Brassington Formation, has bccn Upland Britain has undergone a regular uplift determined by estimating the thickness of since the main Alpine episode at a rate of about limestone and Namurian shale which, as t metre in x5,25o years. In this light, the pos- suggested by previous research work in the sible age of other well known planation surfaces southern Pcnnines, was once present over the is discussed.

I. Introduction THE SO-CALLED 'Pocket Deposits' in the Carboniferous Limestone of Derby- shire and Staffordshire, masses of clay, sand and gravel which are generally supposed to have been preserved in solution cavities, have given rise to much controversy among geologists and geographers. The deposits, exploited com- mercially since at least as early as the late eighteenth century (Pilkington I789, p. I62) are still worked for their silica-sand content at one or two places, though, because older sections have deteriorated through disuse or infilling, and because few new sections are becoming available, opportunity for a scientific assessment of these controversial deposits is diminishing and will probably continue to do so. The Pocket Deposits masses are known from an area which measures 35 ° sq.km. (Fig. I). The origin of the term 'Pocket Deposits' is obscure and its earliest use known to the authors is that of Howe (x897). The term has been retained by all recent workers (e.g. Yorke I96i , Kent I957, Ford & King I968 and I969) but, since the original manuscript of this paper was written, it has become necessary to distinguish the large-scale stratified subsidence masses from the shales and solution residues which are also Pocket Deposits materials in the broad sense of that term. The former have now been termed the Brassington Formation (Boulter et al. I97i ). It is fortunate that, of the few clear sections which still exist, most show clear evidence of stratification of the Brassington Formation which is sufficiently well- developed to infer that deposition was complete before subsidence took place. Several previous writers, notably Boswell (I918) and Kent (I957) have also emphasised that a stratigraphical order can be made out from many sections (despite distortions attributable to subsidence) and that this appears to be maintained for some distance across the southern Pennines. The present paper seeks to summarise the evidence for the stratigraphy of the Pocket Deposits; to discuss the relationship of the subsidence masses to the host limestone and the 'Millstone Grit' and to establish the altitude of the high-level erosion surface on which at least the major portion of the Brassington Formation

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 The preservation of the Neogene Brassington Formation of the southern Pennines 521

NATGRID REFS SIc, 10 \ + Hindlow

÷ ÷ ÷

+''-~ I ., ~.~__~, ~ (''~, '.11tli l+b.KenslowTop

*+ . Frik~nl I I I I F~,T~I~ 60- ÷ , ,lllllllrdl i~ ,÷

x, ÷ ÷\

E= ~. + ~. Minn~glow pits

Low Moor ' I I I I L,. HilltopForm 'l~+lll,,,llllr,.gg~ l.,X," C

I,1 urossjngton~%? "tC6)' Plit ~.....¢.. iWirksworth

C~ ..g f;" . ~ "/ -\_ ~]). I • Large.:o,esu~s,*°=emo~.(o...~,, Rlbden [ . ,n~./'-~" I _ Yokeond Ford & King) )~/ I ~ Junction of Carboniferous Limestone with .'~..~ ~lleymOOr ya~~.V. I "--~a'~;rio" sho,es

.r' / J I I Parsons,Kent, Hughes & Ford & King) "El'! // I J ~,"" Probable feather edcl¢ of volcanic layers 10 [ 2'0 I I (after Shirley and Aughes)

FIG. t. Map to show (I) the location of the research area (2) the localities mentioned in the text and (3) Ex & E2, the age of the basal Namurian sediments at various localities (the latter partly after Ford (I968) and Morris (i967 and I969) and I.G.S. sources. (4) D1 & Dg, the age of the youngest preserved Vis~an marginal to the limestone outcrop.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 522 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

was originally deposited. The importance of this estimate in relation to the geomorphological evolution of the Pennines in particular and the British Isles in general is also discussed.

2. The stratigraphy of the Brassington Formation Over sixty solution-subsidence outliers of Brassington Formation are known (Yorke I96I, Ford & King I959, Fig. I). The majority of these masses are so poorly exposed at the present time that the different materials cannot be placed into a reliable stratigraphical order. However, careful field observation of a few of the larger exposed subsidence masses has established that numerous individual strata can be recognised locally and placed with confidence into a stratigraphic sequence. Especially valuable sections in this respect are those at the Bees Nest Pit (SK241546), Kenslow Top Pit (SKI82616) and Kirkham's Pit (SK21754o). Per- haps the least distorted and best exposed section is that at the Bees Nest Pit (Boulter & Chaloner i97o , Fig. 2). What is preserved of the original pre-subsidence sheet is a reflection of the opposing factors of depth of subsidence on one hand, and sub- sequent erosion on the other. The greatest known contrast, as revealed by field measurements, has occurred at the Kenslow Top Pit where, by the authors' own estimate, no less than 7 ° m of pre-Glacial sediment is discernible (cf Kent's estimate of 38 m for this locality in I946 ). This implies that the thickness of the pre-subsidence Brassington Formation sheet was almost certainly in excess of this since it would be most unlikely that the uppermost part of the sheet has entirely escaped glacial scour. This view is further supported by Kent's (i957, p. 5) suggestion that some observed wedging of the individual strata is a subsidence effect which reflects attenuation of the original thickness. Detailed analyses of what remains of the Brassington Formation sheet at the Bees Nest, Kirkham's and Kenslow Top Pits shows that it is essentially divisible into three members: 3. Grey clays, containing plant remains. 2. Coloured clays, now mottled red and green, unfossiliferous as far as is known. I. Sands with gravel beds, probably originally lilac-red though now much bleached, and occasionally cemented into a soft sandstone. Unfossilifer- ous as far as is known. The terms Kenslow, Bees Nest and Kirkham Member respectively have been applied by Boulter et al. (i97i). The fossil flora of the Kenslow Member is of Upper Miocene-Lower Pliocene age (Boulter & Chaloner I97o; Boulter I97I), and, since the tripartite stratigraphy was proposed by Boulter et al. (op. cit.) excavations at the Kirkham's Pit have revealed a further clear section of the Kenslow Member Clays and their conformable junction with the Bees Nest Member (Ford, in press). Kent (1957) suggests that the Kenslow clays are paludal or lacustrine in origin and were formed locally during a comparatively late-stage deepening of the 'sink-holes' in Tertiary times; Ford & King (I969) , using the palaeobotanical

identifications of Boulter, also suggest a localised lacustrine origin, the 'ponds' being surrounded by heathland. Though Nyssa and Trapa (modern plants of these genera inhabit wet or swampy environments) have been recorded subsequently by Boulter (I97I), the authors have found no sedimentological evidence either for or against this hypothesis of localised deposition, and would only argue that an estimated preservation of not less than 6 m of these sediments at Kenslow Top suggests a relatively widespread deposit. A further mass of probable Kenslow Member clay has in any case recently been discovered at Hindlow (SKo8469I) (Dr. N. Aitkenhead, personal communication). On the basis of its conformity with the Kenslow Member, the Bees Nest Member is thus also of Upper Miocene-Lower Pliocene age. There is also little doubt from the exposure in the three pits mentioned that the Bees Nest clays are quite conformable with the underlying Kirkham sands: sampling certainly reveals an essential petrographical continuity in the clay sediments between the two latter members. Much has been made of the supposed 'Triassic' nature of the heavy minerals in the Kirkham sands and numerous previous authors have concluded that the sands themselves are consequently of Triassic age, with, in some cases, the corollary that there must be an unconformity with the Tertiary-fossil-bearing clays. Linton (i956), Clayton (i953) and Ford & King (~968, I969) deduced from general field evidence that the Kirkham sands have been derived mainly from a Triassic parent nearby, and consequently the grains of the Kirkham sands have been little modified during the late-Tertiary erosion cycle. Further detailed petrographical work by one of the authors (M.I.) has indicated that there are, in fact, significant differences between the Kirkham sands of the Brassington Pits and the Bunter Sandstone of the Hulland Gravel Pits, 9 km to the south, the nearest exposure of undoubted Triassic to the Brassington area. Whereas garnet is a common constituent of the Hulland Bunter Sandstone, that mineral is virtually absent from the Brassington Pits sediments. The Hulland Bunter Sandstone has a log normal or nearly log normal particle size distribution, while probability graphs for the Kirkham sands of the Brassington Pits show pronounced deviations from log normality. Moreover, in the case of the Bees Nest clays, often thought to be the local equivalents of the Keuper Marl, none of the characteristic clay minerals of the Keuper Marl of the Midlands area (chlorite, swelling chlorite, sepiolite) (Dumbleton & West 1966) is present in the sediment of the Brassington Pockets. Unfortunately, there is no firm evidence yet concerning the provenance of the Brassington Formation sediments, and they may have been derived from the north rather than the south. Hughes, for example, recorded (I952, p. 79) that current bedding in the basal sands in pockets near Brassington showed evidence of a derivation from the north and east. The authors are thus of the opinion that there can be little doubt that the Brassington Formation is wholly of Neogene age; that it accumulated in a terres- trial environment on what is now, through uplift, a high-level planation surface above the Pennines, and that there is an essential continuity of sedimentation which marks a single cycle of sedimentation. The cycle opens with piedmont or alluvial sands and gravels, develops into a lacustrine phase, and terminates with semi-paludal sediments containing small quantities of lignite. Possibly other

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 524 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

cycles were developed which have not been preserved in the subsidence structures. A remote possibility exists, of course, that the Kirkham sands are neither of Neogene nor Triassic age--for instance Palaeogene. Until fossil dating is forthcoming, all that is really certain about the older Brassington Formation sediments is that they are younger than the Namurian and older than the Lower Pliocene, though some earlier attempts to fit them into the Carboniferous (e.g. Scott 192o), Permian (York 196 I), or Rhaefic (e.g. Boswell 1918) have little scientific basis and are of historical interest only.

3. The subsidence mechanisms A) GENERAL Attempts to account for the present locations of the Pocket Deposits masses have possibly given rise to more controversy than have attempts to ascertain their age. While the possibility that the Pocket Deposits have been preserved in multiple structural forms should be borne in mind, they have mainly been described as cave collapses linked to dolomitisation processes (Ford & King 1968 , 1969) , infilled limestone gorges (Yorke I961), and sink holes (Kent x957). There seems littledoubt that some of the smaller Pocket Deposits masses have been preserved as cave infillings, the sediment of Brassington Formation type having been washed in from the surface through solutionmwidened channels, and have accumulated there in quasi-horizontal layers. The authors have not observed this feature though it has been recorded by Ford & King (I969, p. 6I) who presumably refer to the sand-filled caverns in the Golconda Mine (SKe49552) . . . apparently at the base of pocket-cum-cavern systems. It is also possible that some of the masses are pre-existing erosive forms which have received a surface fillrather than underground cavern systems. It is clear, however, that the larger-scale Brassington Formation masses are neither cave infillings (otherwise the sediments would be quasi-horizontal) nor are they surface channels which have become infilled (for the same reason), nor is it easy to see how any of the masses can be interpreted as a 'sink-hole deposit'. The terminology of such phenomena is rather confusedmthe recent glossary of Karst Terminology (U.S. Geological Survey Water Supply paper 1899-K I97O ) merely defines 'sinkhole' as a general term for a closed depression, whether basin-, funnel- or cylindrically-shaped; in the authors' opinion this latter term should be restricted to a landform typically associated with karst topography in which enlargement of the feature by erosion is in excess of the opposite effect of sedimentation in it, i.e. a 'sinkhole' should not be termed a sinkhole if it is receiving and retaining sediment. True, the site of old sinkholes can, with a rise in the water table come to receive and retain sediment~in which case the layering would be horizontal as stated before. Thus, because of the evidence of the crudely synclinal form of many of the Brassington Formation masses, (conversely, of the absence of quasi-horizontal bedding in all of the larger scale masses), it follows that most if not all of the subsidence sediment is preserved in cavities which are post-depositional in origin (it is theore~ally possible, though much less likely, that the

cavities were pre-depositional in origin, but were not occupied until after the Brassington Formation sheet had formed in all its essential details). Having decided this, it remains to be determined whether the sediments are preserved due to subsidence into a large number of small holes or a small number of large holes; whether the collapses were fast or slow; from what levels the subsidences took place; when they occurred; why they occur ili what appear to be the characteristic localised masses; and what rocks were involved apart from those of the Brassington Formation and the host limestones. Relevant data are the size, shape and internal form of the subsidence masses; the relationship with the host rocks; the structural condition of the host rocks; the hydrogeological history of the area; glacial and other erosive modifications and the distribution pattern of the masses themselves. It is not intended to give here a comprehensive account of previous opinion, but to try to modify or extend earlier views that, in the authors' opinion, have not received due attention.

(B) THE NATURE OF THE SURFACE ON ~VHICH THE BRASSINGTON FORMATION WAS FORMED There seems from the previous literature to be some uncertainty as to whether the basal sands were formed on a surface cut across the Carboniferous Limestone, or the Millstone Grit, or the Triassic (the most likely possibilities) or across a composite surface formed by more than one of these, or others. If in the 60 or so known occurrences of Pocket Deposits only Brassington Formation sediments were found, this would be strong, if indirect evidence that the sheet was formed on a surface cut solely across the Carboniferous Limestone. But this is not so. At several locali, ties, masses of Namurian shale can be seen marginal to the pockets (i.e. close to the limestone or dolomite wall-rocks) which are also in clear-cut contact with the basal sands. The critical localities are described later. In the authors' opinion these shale masses are not to be confused with other shale masses which have been thought to occur in the local tills, and they are in a sense an essential pre-Ncogene constituent of the Pocket Deposits (sensu lato). At the Bees Nest, Green Clay, Low Moor and Minninglow Pits the basal bed of the Kirkharn sands is seen to rest directly on zones of weathered shale, which, for perhaps a metre below the contact have a characteristic lilac hue. At Kenslow Top, the actual contact is obscured, but immediately adjacent to the likely position occurs a basal conglomerate, whose cementation is doubtless due to the precipitation of iron compounds above an impervious shale base prior to subsidence. Thus over an appreciable part of the total area where pockets arc found, it can be shown that the Brassington Formation was not deposited directly on the Carboniferous Limestone which now acts as their host, but on an impervious shale cover which was certainly widespread enough to suggest that possibly no part of the Brassington Formation was formed on a bared limestone surface. No such interpretation of the association of the Neogene sands and the Millstone Grit shale cover is believed to have been previously given. Kent, for example, (1957, p. 5) thought that the Kenslow Top shale mass was due to 'late stage collapse' into what he conceived as a sinkhole, whereas Ford & King (1968, p. 324 and 1969 Fig. 4) clearly imply that the sands

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 526 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

were directly deposited on a limestone surface and state that the Namurian shale blocks are always part of the Pleistocene drifts. It must also be noted that Howe (I897), Barke, Hind& Scott (I92O) and Kent (x957) have recorded that 'unaltered masses of cemented Bunter sandstone' (Kent's term) are present in the Staffordshire pockets. This was taken by Kent to imply that this was evidence that the sand content of the pockets there was wholly of Triassic age, but in view of the evidence recorded here, it seems a more likely interpretation that in Staffordshire, the sub-Neogene surface was locally cut across the Bunter Sandstone rather than the Millstone Grit, again with the implication that the Carboniferous Limestone was roofed over at the time of the Brassington Formation deposition. Similar pockets in Flintshire were figured by Maw (i 867) and Strahan (1890) as having stratified sands and lignitic clays overlying a former cover of Millstone Grit. (See Walsh & Brown x97I). Whether or not any of the Neogene sediments were formed directly on the Carboniferous Limestone is a problem that cannot now be solved. Evidence can be cited both for and against this hypothesis--for instance, the dolomitisation, which Ford & King have stressed is such an important factor in cavernisation of the host rocks, and which is traditionally thought of as being due to downwards migrating magnesium-rich solutions, would have great difficulty in penetrating a total Namurian shale cover. On the other hand Pitty (1968), from a knowledge of the rate at which the present surface of Derbyshire is being lowered, has argued that the limestone may have been unroofed as recently as three million years ago. Whether or not the thickness of the Namurian shale as preserved in any of the pockets described here is representative of the true thickness of the presumably near-horizontal shales between the Kirkham sands and the top of the Carbonif- erous Limestone is not known and presumably never will be. No estimates more accurate than 6 m can be given for the maximum preserved thickness of shale (at Kenslow Top) ; only 4 m appears to be present at the Bees Nest Pit. While it is possible that this is a true reflection of the pre-subsidence thickness, it should be remembered that these shales have obviously been subjected to intense shearing marginal to the subsidence zones, and it seems to the authors more likely that only a small fraction of the original layer remains.

C) STRUCTURAL INFLUENCES OF THE HOST ROCKS In order to test the hypothesis that the Carboniferous Limestone was completely roofed over at the time of deposition of the Brassington Formation sheet, it might be profitable to enquire whether or not the larger pocket masses where shale blocks are present are located along the axes of Sudetic or post-Westphalian synclines. This might also indicate whether or not synclinal structure has concentrated the sub-surface flow and so selectively produced cavernisation there. From the authors' general field observations, it does not seem that any of the shale blocks described in this paper is associated with a clear-cut synclinal structure in the Carboniferous Limestone--indeed, the structure at Kirkham's Pit seems markedly anticlinal.

(D) THE STRUCTURE OF THE BEES NEST PIT SUBSIDENCE COMPLEX The topographic detail of the Bees Nest Pit has been surveyed using tacheometry. Fig. 2A represents a plot of this data, on which is also shown the main geological boundaries. Fig. 2B is an enlargement of the area of the Bees Nest Member outcrop, whose base is shown contoured at 5 feet intervals using borehole data, dip records and spot heights. The pattern of contours suggests the presence of at least three distinct synclinal structures, two of which are incompletely exposed towards the east. The largest, the western-most, appears to have formed from two relatively shallow sags which have coalesced at a late stage in its development. This composite structure has a faulted axis which trends east-west. A throw of at least 12 m is implied by field data in the central part of the fault. Both observations in the field and extrapolation of structure contours reveal that all three of the sags are tight in their axial portions, dips suggesting that the amplitudes of the sags are greater than their wave lengths.

E) SUBSIDENCE SIMULATION TESTS In order to try to duplicate the field model in the laboratory it was decided to construct a two-dimensional laboratory model which could be made to simulate the production of an individual void in the limestone host rock having a cross- section as small as 56 m s. The apparatus (Plate i) was basically a narrow wooden box having a perspex front, and a base formed by the ends of forty-five vertical boards (each 17 mm thick) which could be lowered independently to simulate the removal by solution of blocks of limestone. The overlying Pocket Deposit sediments were represented by layers of Leighton Buzzard sand, chalk dust, cement pigment and composite mixtures of these materials, sometimes having a soluble-oil base. In the box, 92 cm wide and IOO cm high the 'strata' could be allowed to 'founder' to a depth of 5 ° cm (representing a field subsidence of 19 ° m, the horizontal and vertical scales of the model being the same). Control of the 'subsidence' was achieved by the use of pegs fitted to a series of notches cut at 2 cm intervals. Originally a three-dimensional simulation model was considered, but because plotting of points in the central l parts of the subsidence could not have been achieved during the test, it was decided to concentrate simply on a two dimensional section simulating the central part of the deepest subsidence at the Bees Nest Pit. In this way continuous records of the developing structure could be made using wax-pencil tracings on the perspex front. After first having established that the materials used gave what might be considered a fair representation of the actual Pocket Deposits, a series of eight tests was conducted. Various opposing patterns of limestone removal were simulated: (I) random as against regular (2) small as against large collapse (3) gradual as against wholesale collapse (4) broad as against narrow collapse. A brief outline of the results is given in Table I.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 528 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

F) CONCLUSIONS As far as the authors are aware, no previous attempts have been made to simulate karstic subsidences by any method similar to that described in this paper. Without a precedent, it is therefore particularly difficult to assess whether or not the size and shape of the model, the nature of the test materials and their behaviour, the rapid rate of 'subsidence' and the two-dimensional nature of the experiment offer a fair enough simulation of the field example for this work to be relevant to their general thesis. Further, the authors freely admit that a large element of" precon- ception about the nature of the field subsidence was incorporated into the initial form of the laboratory model and the cycles of lowering.

.<, ~l Ii --

fi Z [q llllll HzLTI i IJ

o ~o "0

~-0 o

ol "2" "2

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 ~I

!~.^ \ I / \ \

. ~, °

I IE I ~.-

/ / .,~ // //

ffl

8.-a - ~~ ~=~ I ',- "a ,¢J f.

g II

o~ ,§~ ~ 8

I /

/~ [ N "a .o

! I

// ~ '!!I ~.~ I -:j, / • I~. ii / / I I .I

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 53 ° P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

However, despite this subjective element, the results of the eight tests allow some genera1 conclusions to be made which the authors consider might serve as a work- ing hypothesis until more extensive experimentation has been conducted. With regard to the Bees Nest Pit in particular and the Pocket Deposits in general, the results suggest strongly that (I) the subsidence complexes are the results of gradual block-by-block removal of the limestone rather than of wholesale collapse of the roofs of large caves (2) for each outlier, the upper surface of the limestone during collapse was even rather than uneven (3) subsidence has been intermittent over several millions of years and may still be operative, though presumably it has diminished since the time when the overlying fluvioglacial or periglacial deposits were formed and (4) the subsidence complexes have the form of deep, steep-sided, U-shaped pipes, rather than of shallow basins. Despite the use of a variety of sand, chalk-dust and soluble-oil mixtures for the 'Namurian shale' basal layer, the layer fragmented and separated in all cases in a manner strikingly similar to that observed in the field (as illustrated in block diagrams 3 and 4 of Fig. 5 and Plate 2). Several equally striking examples of convoluted and axially-faulted 'synclines' were produced in a number of tests (Plate 3). In fact, the authors would comment that despite the material and geo- metrical changes in the test conditions, very few results seemed totally unrelated to what can be seen on a larger scale in the various Pocket Deposits exposures. At present, the authors are not able to offer any firm opinion on the problem of why the Pocket Deposits masses are limited to their present geographical distribution and why they occur in the form of isolated pockets. Ford & King (I969) emphasised the association of Pocket Deposits masses with dolomitised zones of the Carboniferous Limestone. But even on their own distribution map (ibid Fig. I), there are just as many large scale masses outside dolomitised zones as within them. Similarly, Kent's (i957) concept of a cavernisation process linked to sink holes development along a receding feather edge of a suprajacent Namurian shale cover receives little support from the present work. Fig. I of the present paper suggests that there may be more than a casual connection between the probable western limit of the toadstones of the Matlock area (aquicludes of volcanic origins) and the eastern limit of the area in which the larger scale masses of Pocket Deposits are found (see also Fig. 5)

4. The sub-Neogene surface in the southern Pennines

(A) THEORY Identification of a fossiliferous shale foundation immediately beneath the Brassing- ton Formation offers some prospects for determining the amount of subsidence that has taken place, provided that the same material can be identified in outcrop reasonably close to the subsided masses. The principle is a simple one. Either regional mapping can be employed to determine the position of the projected level of the known horizon involved in the subsidence; or, the regional stratigraphy having been already determined, fossil evidence is used to obtain the stratigraphic anomaly of host rock and subsidence mass. The depths of several other British

~. .~ ~ ~ .~ • ° ~-~

• ~ ~ ~ "~

~-oo~ ~ .=~ _~ ~o .~~o~ ~ ~o~ ~ ~ ~ , "d ~.~ •~ ~ ] ~o

~ ~ o ~_~o ~ o ~~ ~ ~~ ~ ~

o~ ~ ~ ~ ~ .~ ~ ~ .~ •~ ~0 ~o

~J ~

4,J

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 532 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

subsidences have been established in this way. Examples are (I) the solution- subsidence outliers of Millstone Grit on the north crop of the South Wales coalfield (Thomas x964) , where goniatite evidence and regional mapping have combined to give an estimate of I8o m for the subsidence there; (2) Millstone Grit and supposed Neogene sediments in the Carboniferous Limestone of Flintshire (Walsh & Brown I97I), where a subsidence of I I 7 m can be demonstrated; (3) the gash breccias of the south Pembrokeshire coast, where collapse 'through several zones of the Carboniferous Limestone' can be demonstrated (Dixon I92I); (4) the Tertiary pipes of the Dorset coast near Lulworth, which are found in zones of Chalk hundreds of feet below the local base of the Tertiary (Arkell 1948). Thus in the present situation, if the blocks of Namurian shale can be accurately referred to the stratigraphy of the nearby main outcrop of the Namurian and, further, if a reasonable assessment of the stratigraphic anomaly with the host limestone can be effected, there is some prospect that the depth of subsidence can be established. Unfortunately, because of the presence of a well-marked unconformity between the Vis6an and the Namurian, demonstrable over a considerable length of the outcrop bordering the limestone area of the Derbyshire massif, the results cannot be accepted as anything more than 'educated guesses'. The effects of the Sudetic earthmovements are widespread and some remarkable examples of the preservation of a deeply serrated pre-Namurian erosional topography have been demonstrated in several areas. Existing data are well summarised by Ford (I968) who stresses the severity of the Sudetic erosion cycle and even suggests that... 'it is probable that much of the present limestone upland is an exhumed pre-Namurian surface.' Thus deductions must fully take four factors into account: (i) whether or not a Sudetic unconformity was present locally, and if so, its temporal magnitude. (ii) the thickness of limestone present below the unconformity to the horizon equivalent to that of the host rocks. (iii) the age of the shale blocks immediately below the unconformity with the Brassington Formation. (iv) whether or not this information can be projected, accurately enough, from the margins of the massif to the interior, where the Brassington Formation is preserved. (This depends upon a rather subjective appre- ciation of the pattern of palaeogeographical change.)

(B) CRITICAL SECTIONS (i) The Bees Nest Pit Namurian shale is present on the southern wall of the pit. Relationships with the Brassington Formation and solution residue are shown diagrammatically in Fig. 2A. The shale mass appears to be related to a minor pocket, marginal to the main sub- sidenee, possibly of earlier date. The stratigraphy is: 7. Fluvioglaeial or periglaeial deposits. Up to 6 m. 6. Brasslngton Formation. N43 m. 5. Lilac-stained weathered zone of shale mass. ,~I m.

4. Unweathered shales. "~3 m. 3. Solution residue (clay-with-cherts). Up to about 5 m. 2. Detached masses of dolomite wall-rock. I. Sound dolomite wail-rock. (ii) GreenClay Pit (SK.e4t548) A stratigraphy similar to that established for the adjacent Bees Nest Pit is also present. A small part of the solution residue of clay-with-chert is mineralised by iron compounds (Walsh x965). The eastern and northern walls of the pit reveal several masses of fresh black Namurian shale. (iii) Kirkham's Pit The brownish clay marginal to much of the central subsidence of this complex is believed to represent weathered Namurian shale in which the original bedding has been largely obscured. No extensive development of solution residue appears to be present, nor is the lilac staining prominent. (iv) Low Moor Pit (SKz87566) The southern wall of this pit exposes a mass of Namurian shale, weathered with the characteristic lilac hue close to the contact with the basal sands, and in fairly clean-cut contact with the limestone host rocks. No solution residue is present. The shale mass is about 5 m thick and contains a prominent band of sandstone close below the contact with the basal bed of the Kirkham member. The bedding trace of the shales is nearly vertical. Slumped shale masses also flank the outlier on the eastern wall of the pit. The field relationships of the Low Moor Pit are shown diagrammatically in Fig. 3-

[1~4~ ~7s7~ m ] Parwic~ Pcrwich"Httl HiHtoo Form Lowmoor Form I ft 600 Sampt¢ [2000 nferred sub-.Nep_.~en¢surface I 4so ~..-7-~ ~----=------~,~,-7.'~bsi-~,~-~,,~moo

180~" , shot ..... ~ 600

Fzo. 3. Section to show the geology between Parwich and Low Moor Pit. Natural scale. Length of section 2.5 kin.

(v) Minninglow (South) Pit (SK2o2574) A mass of Namurian shale (reported to contain coal seams by Hughes (i952)) is present in the south-western walls of this pit. The shale is weathered lilac close to its contact with the basal sands. (vi) Kenslow Top Pit The only conspicuous shale block marginal to the subsidence mass lies in the north-west corner of the existing pit. This is at least 6 m thick and is associated with cherty solution residues. The contact with the basal Neogene deposits is obscured by vegetation, but its probable position is close to the iron-cemented conglomerate blocks present in the middle levels of the pit wall.

Howe (I897), Jackson & Charlesworth (I92O), Scott (z92o) and Kent (z957) have all reported goniatities in the shale masses of the Derbyshire and Staffordshire pockets. Yorke (x 96z) records the identifications of Posidoniella laevis, Posidonia 2

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 534 P. T. Walsh, M. C. Boulter, M. Ijtaba & D, M. Urbani

corrugata, Nuculoceros nuculum, Gravenoceratoides sp, and Gravenoceratoides cf nitidus from Low Moor Pit, while Eumorphoceras is recorded by Kent (I 957) from the shale masses in the Kenslow Top Pit. In his analysis of the plant miospores from the Namurian and Lower West- phalian of the southern Pennines, Neves (1958 , i96i ) has identified more than 5 ° forms, of which 34 are new species. 41 of these have a limited stratigraphic range within the Namurian of the southern Pennine basin, so enabling them to be used, albeit cautiously, as a source of reference to the standard sequence of goniatite stages. A number of these miospores with characteristically restricted ranges have been identified from the shale blocks in three of the most widely separated Pocket Deposits, at Kenslow Top Pit, Minninglow Pit and Bees Nest Pit, as shown in Table 2. The stratigraphic ranges of these taxa, as determined by Neves, show that in each of the three Pocket Deposit localities that have been studied the shales are of E2 (Arnsbergian) age. It should be pointed out, however, that these ranges are based on a very limited knowledge of the occurrence of each miospore taxon. Further work on the palynology of Namurian sediments may well extend the ranges sufficiently to make correlation with the concise goniatite stages un- reliable or even impossible. The assemblage is made up largely of species belonging to the genera Lycospora, Potoniesporites and Florinites, as well as other forms which have not yet been identified. In addition to the miospores from these three genera and those listed in Table 2, single specimens of Raistrickia microhorrida Hoist, Endosporites ornatus Wilson & Coe and Acanthotriletes sp. have been recognised. These taxa have not been considered stratigraphically by Neves (I 96I) though they have been identified from the Gastrioceras subcrenatum horizon in north Staffordshire (Neves 1958), as well as from Namurian A strata in Europe. Their identification in the Pocket Deposits shale could well represent an extension to their range in the British Namurian, though a more detailed study is necessary to fully appreciate the significance of their occurrence. Other samples of Namurian Shale have been examined for microfossils from Low Moor Pit, Green Clay Pit and Kirkham's Pit as well as from the most immediately adjacent localities in the main outcrop of basal Namurian Shale surrounding the limestone, at Raper Mine (SK216654) and Parwich (SKI88546).

TABLE 2" List of Namurian plant miospores, identified from shale masses at three of the Pocket Deposits, whose restricted range has been determined by Neves ( x96x) Stratigraphic Kemlow Bees Range Top Minninglow Nest Puru:tatisporitesgiganteus Neves E2 + + Ibrahimisporites brevispinosusNev~ E2-G2 + + Mooreisporites trigaUerusNeves E2-RI + + Convolutispora obliqua Neves E2 + Densosporites spinosus Dybova & Jachowiez E2-Rt + + Proprisporites laevigatus Neves E I-R2

The miospores in these samples were found to be poorly preserved and of infre- quent occurrence so making specific identification (so important in comparisons for correlation) very tentative. However, examination of the macerated samples from these localities did not reveal any Namurian miospore taxa which have not been identified from the other three Pocket Deposits.

(D) MINERALOGICAL EVIDENCE Because palynological analysis of clay samples from exposures of basal Namurian shale has not revealed definitive miospore taxa, an attempt has been made to correlate these shales with those in the subsidence masses at Kenslow Top and Low Moor by means of X-ray analysis. Except for the Parwich sample, the clay fraction of each was obtained by dispersion and sedimentation. The shale from Parwich was compacted to such an extent that it was almost impossible to dis- aggregate it into its fundamental grain size. From this sample only the fraction which passed through a 300 mesh (B.S.) sieve was used. For each sample a powder diffraction photograph was then made, analyses of which revealed the following mineralogy: Illite. A series of basal spacings indicating illite was observed in all four samples. The presence of the (060) line at 1.5 ° A suggests that much of the illite is dioctahedral. Quartz. As indicated by lines at 4.26 A, 3-34 .A and 1.81 A, quartz was present in all four samples also. Kaolinite. A line at 7.15 A was observed in the samples from Kenslow Top and Low Moor Pits only. This was not recorded when both samples were heated at 6oo°C for two hours, confirming the presence of kaolinite. Too little is known about the bed-by-bed mineralogical variations of the Edale shale formation (the basal Namurian shales) for this analysis to be really meaning- ful. The samples from Low Moor and Kenslow Top (the latter already shown to be of E~ age on its spore content), are hereby shown to have a similar mineralogy, though, obviously, this does not necessarily indicate that exactly the same bed has been sampled at both localities. The hoped-for mineralogical correlation of the Kenslow and Raper Mine samples, and those of Low Moor and Parwich has unfortunately proved somewhat negative, though the same two minerals (illite and quartz) are present in the Parwich and Raper Mine samples (the latter having been established as of E, age on the basis of goniatite remains (Ford 1968, p. 88)).

(E) THE ELEVATION OF THE SURFACE ON WHICH THE BRASSINGTON FORMATION WAS FORMED From the evidence of the foregoing, an attempt is now made to determine the configuration of the sub-Neogene surface. The difficulties in making these estimates of subsidence have already been discussed, and it is intended to allow for the variables as far as possible. The authors have not undertaken any detailed field mapping of the Carboniferous Limestone outcrop, except in the vicinity of the subsidence outliers themselves, and, as will be seen, they rely heavily on the

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 536 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

0"~00 O~ ..---. 0 r--o 0 ~ ,.~ ,,~ ~ o o o ~ c~o o ~= ~ v°° v~ v~ ,S~ ,.~ v'.,%v~

~c~ ~ 0 ~ c~ u'~ 0 0 ~ ,-, 0 0

;> 0 rj ,.Q

0 0 LO ~0 ,~0 ,.~

,.Q

o,..~ °~ o,..~

.,,.i v ~ "~ ~,.~. I~., v °~.~ • ~, ~ ~ ~ ~ g~

~.~ °,-.4

¢0 0 ~ ~ 0 '~0 0 v~ ~ v v v v v v

-~ "~ o

m.~ 0 0 0 i.~ v v vv

el'.~ CO ~I "~" C/)

0 3 ,-~. ~ 0

v vv vv v v v v v v 2;

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 J. geol. Soc. Lond. vol. x28, I972 WALSH et al

PLATE I. The subsidence simulation apparatus, showing the results of Test 8.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 J. geol. Soc. Lond. vol. x28, 1972 WALSH et al

PLATE 2. Test 2. Final stage. Note the large void in the "basal sand" due to arching of the test materials.

PLATE 3' Test I. Final stage. The light layers are chalk dust; dark layers, sand.

~o, ; •

,. ,,l~

PLATE 4" Initial s~ages of Test i, showing formation of temporary "cavities" by lowering of the boards.

published maps of Shirley (i959) and the Geological Survey (I96Z, SK25 NE) and on the unpublished map of Hughes (z952). The following is a summary of the critical details of each of the outliers described herein. In each case, an allowance of 2z m (7 ° It) is made for the thickness of Namurian (Edale) shale which formerly lay between the Carboniferous Limestone and the basal Neogene sands; this estimate is, of course, largely subjective and it tries to account for the order of thickness of shale preserved at say Kenslow Top, and a probable additional loss due to the shearing out of the weak shale during the subsidences; (see discussion in section 2B). Figure 4 presents this data graphically. The arithmetic mean of these figures places the sub-Neogene surface at about 45 ° m (I5OO ft)--i.e, within the tolerance limits of all individual estimates and it is thus suggested that this mean figure is accepted for inter-regional purposes.

] metres 360 390 420 450 480 510 5~ feet 1200 1300 1400 1500 1600 1700 180 Kenslow Top Minninglow south Low Moor I Bees Nest I Green Cloy Kirkhoms -- ]

FIG. 4. Diagram to graphically illustrate Table 3.

5- The denudation chronology of the southern Pennines The present work establishes the existence across the southern Pennines of a sub-Upper Miocene or Lower Pliocene land surface at an altitude of about 45 ° m, i.e. something higher than the so-called z 25 ° ft surface, the highest extant 'surface' in the southern half of the limestone outcrop of the Derbyshire Massif. Higher surfaces in the south Pennines region may be inferred to be of mid-Miocene or older age, while lower ones have developed during the last 7 million years or so. Before considering the applicability of this dating to other areas of Upland Britain, however, it is pertinent to see how the sub-Brassington Formation surface accords with supposed high-level surfaces in areas near to that in which the Brassington Formation is preserved. Clearly, it is only to the north and west that such a comparison can be sought, as in the south and east nothing remains in the physiography which reaches the required altitude. Within the area of the Derbyshire limestone massif itself, very few summits reach the supposed level of the sub-Brassington Formation surface, especially if a gradient falling to the south west is added so as to allow for a possible southwards- inclined Upper Miocene fluvial regime. Possibly the higher limestone hills such as Taddington Moor (I438 ft O.D. 431 m) and Eldon Hill (z543 ft O.D. 463 m) are remnants of the surface, but, in view of Pitty's (z 968) remarks concerning the preservation of 'erosion surfaces' in a limestone area, it appears more probable that if relics of the surface still exist, they will be more easily recognised on the

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 ~o... •

rio. 5

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 FIo. 5 Block diagrams summarising the postulated sequence of events lead- ing to the present day form of the Pocket Deposits outliers. x. The sequence of models is entirely synthetic in its composition, but all have in mind the general area between Matlock and Winster and the southern margins of the Limestone massif. The southern aspect is dominant in each case and the initial height of the model is ca. 18o0 ft (54 °m). The length of each side is several kin. Model I purports to illustrate the general geological relationships in ?early Pliocene times (6- 7 m.y.B.P.) following completion of Brassington Formation sedimentation and with initiation of a strong uplift of the region. A sluggish drainage is initiated on the upper surface of the Pocket Deposits sheet. Anastomosing streams are shown to follow a general southerly direction, for which there is no particular field evidence. Warwick (i 964) however, has suggested a superimposition of the existing Derbyshire and Staffordshire mainstreams, and the early ancestors of the Manifold, Dove and Wye-Derwent may date from this event. The near-horlzontal sheet of Neogene sediment is shown to rest mainly on Namurian shales and grits (black) and also, symbolically, on the Bunter Sandstone (dots and circles) in the south west. A window in the Carboniferous Limestone (possibly related to an exhumation of a Permo- Triassic erosion surface) is shown in the north west. The disposition of the two volcanic layers (v-v) symbolises the essential features of the pre-Neogene fold structure as mapped along the Derwent Valley. Two irregular zones of dolomitised limestone are shown in the southern aspect; these are possibly related to other Permo- Triassic windows, cut into the roof of the nearby Carboniferous Limestone. 2. ?mid-Pliocene (5-6 m.y.B.P.) The general height of the model is maintained. The effect of the rejuvenation of perhaps 15om or so has been the deep incision of the drainage below the Brassington Formation surface. The dominance of downcutting over valley widening lead to incision into the upper part of the Carboniferous Lime- stone. Swallow holes developed as they have in recent times in the northern part of the Limestone massif, as near Castleton. Surface streams were thus intermittent and much solutioning took place immediately below the basal shales of the Namurian. As Ford (x968 p 3~9) comments, the locations of the exit springs from such a system are not yet known. In general, subsurface flow at this stage sought out the more permeable zones in the Limestone--dolomitised areas, fold axes, faults, etc., and special concentrations of groundwater flow doubtless took place marginal to the volcanic layers (Toadstones) since these are aquicludes and would have supported perched water tables. Thus, where a zone of dolomitisation lay adjacent to the feather-edge of a toadstone, cavernisation would have been exceptionally well promoted. Arrows denote the direction of this subsurface flow; the water table is marked by dashed lines. At times of critical instability, progressive cavernisation would cease, and subsidence of the roof take place. No large scale subsidences would have taken place in the areas of the toadstones as downwards migration of cavernisation would have been prevented by the presence of the non-soluble aquicludes. 3. ?mid-late Pliocene (3-4 m.y.B.P.) Concommitant with progressive cavernisation and subsidence of the Brassington Formation sheet into the solution plexi, a relatively stable base level relative to the existing I25o ft contour resulted in extensive valley widening. Removal of the unfoundered relics of the Brassington Formation sheet accomplished, together with most of the Namurian roof-rocks. Only where these materials had subsided into solution cavities were they preserved. The landscape assumed a senescent appearance and may have been veneered at this stage by alluvial quartzite gravels derived from wastage of the Brassington Formation sheet (or nearby Trias). Solution-lowered remnants of these may survive in the Matlock district represented by the gravel soils mapped by the I.G.S. (sheet SK 25NE I96I). Such a Davisian view of the development of the erosion surfaces of the southern Pennines has recently been challenged by Pitty (i968) as being rather meaningless when applied to a limestone terrain which has a regionally low water table. 4. Renewed uplift and incision of the dales are known to date from either later Pliocene or early Pleistocene times. All of the Namurian has now been removed from the area except for the drawn-out detached masses present in the sides of the subsidence masses. Synclinal cores of Bees Nest clays (dashed lines) are all that now remains of the upper part of the original layer. The water table is again low, while the pockets support perched water tables. The scenery is dominated now by relics of the so-called iooo ft surface; glacial events have been ignored for the sake of simplicity.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 54° P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

Gritstone areas of the Cheshire Hills to the west and the High Peak to the north. Little detailed work appears to have been published on the geomorphology of the Cheshire Pennines, but to the north-west of the area of deposition of the Brassington Formation Sissons (i954) has described a series of erosion surfaces in south-west Yorkshire at heights of between 2oo and i8oo ft (6o and 54 ° m). Of the possible correlatives with the sub-Brassington Formation surface, only the Holme Moss Surface (48o to 54 ° m) seems to be a likely equivalent. Sissons comments (ibid. p. 312) that the 54° m surface marks a very effective planation of south-west Yorkshire during which all the land above this height was removed except for Kinderscout, Bleaklow Hill and Black Hill. A marine, rather than a subaerial origin is postulated, however, these higher grounds being referred to as 'islands'. The slightly higher surface in south-west Yorkshire accords with a suggested direction of derivation of the Brassington Formation sediment (Hughes I952 ) and may also be an expression of the relatively higher degree of uplift of the grosser structure of the Pennines since the Lower Pliocene. Sissons suggests a Pliocene age for a lower surface at 375-435 m by way of comparison with a similar datable surface in France, and thus reasons that the Holme Moss surface is somewhat older.

6. The relationships with other British Neogene deposits

Wooldridge (x95o , p. I63) in his presidential address to the Geography section of the British Association wrote of the British Neogene... 'since no deposits inter- vene, it has become widely assumed that nothing can safely be said of the "vast lost interval" (between the topographical expression of ancient structures and the story of glaciation) or that it is a field at best for aimless and profitless speculation'. Regarding the initial statement, however, the authors would take the opposite view--that there are very few regions of the British Isles where small relics of Neogene sediment and of the weathering front of late Tertiary erosion are absent. Along with morphometric analysis and drainage pattern studies, these materials contribute in an important way to our knowledge of the manner in which the British landscape as we know it today came into existence. On the annotated outline map of the British Isles (Fig. 6) are depicted (I) areas at present covered by Neogene sediments, (2) areas thought to have been formerly covered by Neogene sediment, (3) unfossiliferous sedimentary masses of possible Neogene age (but which could theoretically be of Pleistocene or Palaeo- gene age) and (4) superficial weathered zones ('Rotten Rock') whose thickness or field relationships suggests a probable Neogene age (but which could theoretically be of Pleistocene or pre-Neogene age). While acknowledging that not all of the examples figured here will be generally acceptable as material relics of the Neo- gene, the authors are nevertheless convinced that those depicted represent only a small fraction of the likely occurrences of such material and that others will undoubtedly be recognised as exposure permits.

7. The implications for the evolution of Upland Britain

Reference to Fig. 6 shows that, within the area of the British Isles and on the adjacent Continental Shelf, the composite surface on which Neogene deposits can be shown to have formed is as deep as -300 m (possibly --600 m) in the sedimentary basins and at least as high as 450 m in areas of Upland Britain. The differences of altitude are an obvious reflection of the interplay of four factors: i. differential erosion terminating at the time when sedimentation began in the different parts 2. commencement of sedimentation on this composite surface at different times in different places 3. crustal movements during accumulation of the sediments, and 4- crustal movements after sedimentation. Too little is known at present on which to make any possible attempt to assess the relative effect on altitude of differences attributable to the first two factors, but even ignoring these, the altimetric range indicates that late Neogene and Pleistocene crustal movements must have had a considerable influence on the grosset morphology of the area of the British Isles. Yet despite these late Alpine movements (which patently still exist, as shown by movements on such as the Dent Fault) there is considerable evidence that over wide areas (cf Fig. 6) only a few tens of metres of hard rocks have been removed from off the post-Armorican massifs and that such large scale erosion as has taken place has been restricted to the post-Armorican sedimentary basins with their readily erodible fills of Permian, Mesozoic and Tertiary sediment. In other words, the authors feel that whereas the traditional geomorphological opinion has held that movements since the mid-Miocene Alpine episode have been small and erosion considerable, much more emphasis should be placed on the converse of this teachingni.e, that crustal movements have produced structures which have much larger amplitudes than have been suspected, and that the volume of sediment produced in this period is nothing like as much as is commonly believed. The notching of the Armorican massifs by erosion surfaces is thereby relegated to a position of relatively minor importance in the attainment of the present physiographic form. Whether or not the planation surfaces of Upland Britain (i.e. roughly those parts above the Iooo feet (3oo m) contour (Stamp I947) ) can be equated from massif to massif across the British Isles is not the concern of the present paper. Two separate problems are involved here: (I) the definition of the surfaces themselves (2) the establishment that each of the massifs has had a similar erosion history. George (i 967) concluded that... 'The pattern of landscape in Upland Britain shows no major change in kind from one region to another. On an economy of hypothesis, it is not to be ascribed to radically different agents or periods of formation in the different regions'. The authors have no wish to dispute the issue. For those who might speculate further into the possibilities, a simple graph which adaptsthe present results and those of Wooldridge & Linton (1955) to those of

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 542 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

such as Hollingworth (I938), Fleet (I938), Zeuner (I947), Balchin (I952), Proudfoot (i954) , Sissons (i954 and I967) , Brown (I96I), Brunsden et al. (I964), Godard (i965) and George (i966 , i967) could be used to give a somewhat firmer dating to other stages in the system of Upland surfaces (Fig. 7)- The meagre evidence on which the graph is constructed seems to support the concept that the Upland regions of the British Isles have undergone a regular uplift since the mid- Tertiary movements which has the gradient of about one foot in five thousand years (I m in I5, 250 years). Some suggestions as to the possible age of other

S ft m • :~x:x~ 914

- .

Holingworth1938 i5~+ .2ooo 609 Browns High Plateau 1960 Sis .... ~ "~ "~.~.~

......

+ , +.o+, , + , ,,, +. _ ++ iooo 304 ++ 22

16 15 14 13 12 II I0 9 8 7 6 .5 4 3 2 I "0 MILUONS OF YEARS B.P F xo. 7. Graph to show the possible rate of uplift of the Upland areas of the British Isles during the Neogene period. The age of the basal Pleistocene and basal Pliocene sediments after Funnell (x964).

Pennines and Welsh erosion surfaces are shown and on this admittedly rather flimsy basis, the highest of the unwarped British Upland surfaces (summit level coincidence in the Lake District, the Cheviots and the Southern Uplands of Scotland, as suggested by Hollingworth (I938)) could be established as about 13 millions of years B.P.

ACKNOWLEDO~.~m~crs. The authors gratefully acknowledge the help given by Dr. J. G. Capewell and Professor E. H. Brown, who have critically read the manuscript of this paper. Professor T. Murphy, Mr. H. M. Monfford, Dr. T. D. Ford, Dr. P. E. Kent, Dr. D. J. Blundell, Dr. A. F. Pitty and Professor B. M. Funnell have given specialist advice on certain aspects of the work. The subsidence simulation model was constructed by Mr. N. Blake. Mr. A. Standley has conducted most of the subsidence tests. Mrs. S. E. Ware typed the manuscript. X-ray analysis of the clay samples was carried out in the Soil Physics Laboratory of The City University under the supervision of Mr. J. D. Coleman and Mr. C. Begg. Special thanks are due to the owners of the various pits mentioned in this text for their unfailing support for geological research. The Steetley Refractory Firebrick Company has kindly made available all relevant borehole records. Dr. Kent and Mr. P. J. Walmsley of the British Petroleum Co. have kindly furnished weU data on the Pliocene of the southern North Sea.

Explanatory notes to accompanyFig. 6 ENGLAND Cornwall (i) St. Erth. Fossiliferous sands and clays, at an elevation of about Ioo ft O.D. Variously regarded as Pllocene or Pleistocene in age. (Mitchell I965; Edmonds et al. x97o ). (ii) St. Keverne, (iii) Crousa Down, (iv) Camborne. Unfossiliferous sands and gravels, usually regarded as of Pllocene age. (v) St. Agnes. Sands and clays, usually regarded as of marine origin. Usually regarded as of Pliocene age (Edmonds et al. I97o; McFadyen x97o), but verbally reported by G. F. Mitchell to have yielded pollen of Oligocene age. Area of probable former extent of these deposits dotted. Milner (i922) had noted mineralogical affinities among the SW England oufllers, and they have frequently been associated with a 43 °ft erosion platform. Devonshire (vi) Orleigh Court. Clays and gravels containing derived Cretaceous fossils, otherwise unfossiliferous. Origins uncertain, variously associated with the 43o ft erosion platform and others (the map of Rogers and Simpson (I937) shows the deposit to extend to as low as the IOO ft contour) and to the Lustleigh-Stlcklepath fault zone. Usually accorded a Pliocene age. (vii) and (viii)Petrockstow and Bovey Basins, containing considerable thicknesses of clays, sands and lignites (40oo ft or so at Bovey (M. Fasham personal communication); 2,xoo ft at Petroekstow (Fenning and Freshney I968)). The sediments of these basins are usually accorded a Palaeogene age, but in view of the thicknesses of sediment involved, it seems possible to the authors that the upper parts of the sequence could well be of Neogene age (cf remarks on the Mochras borehole). (ix) Hatherleigh. A small outlier of gravels of possible pre-Glacial age has recently been mapped by the I.G.S. (Edmonds et al. x97o). Somerset (x) Brendon Hills. Haematisafion close to the surface of mineral veins was reported by Sherlock in the discussion of Hollingworth (x938). This could be interpreted as a relic of the Neogene weathering front, though it may be of more recent origins. SE England (xi) Outcrops, contours and extent of the area probably covered by Neogene sediment taken from Wooldridge & Linton (I955). The deposits preserved on the North Downs and Chilterns, which are thought to be remnants of a former continuous sheet of sands and gravels, are reported to be of Pliocene age in east Kent, but younger, Lower Pleistocene, in Surrey. A westwards transgression is implied, and except on morphological grounds the inclusion of the whole of this data is not strictly correct. Several small outliers have been omitted. Essex, East Anglia (xii) Outcrops and contours on the base of the Plio-Pleistocene succession taken from FunneU (I96I) and West (I968). Leicestershire (xiii) Deep weathering of possible pre-Glacial age has been described by Ford (I968) from the Charnwood Forest area.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 o ~ i~ z~

o oU z ~- ~o tu~

~ i,,. ~ o ~o> ~

~o~~o~ o~zi~

~:Xx n..

• '=~'~::~

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 ..--;._..--.--~---.--...... -; ...... ~. ~,, ~ o ~ , .i .__.- .___...... "-..o;~-A.~.~ / o i i / .~ ~-" ..---.~ .___...... ~.-~.'rL._.~ o ..~ ,. I , t-- i 1.1- ...~.. ~.~~ ~.~ ~.'T" o (!~ / 8 .f. 8 i o ~ . <' '" <,. I x ,'~'~':~ ~,~'." It I i t" ,-" "/.~ " "-'l +~ "~,.~. ,~.~-~/ C~,".,'i.":-~

i,-i

• ~.

\. .\ .... .

• ..-> ~ ~~" '.",,:~ \ • " "i ...../ ~...... <~ .'~ • ..~ • ~

"L'.r .i" ." o

x '~.~ ._._~.~. • 0

a~ r#i i~

0~.~ l) 0

..~:~oo~. 0

L .I, i ~.I~'~"~'~.'~Z._."~'~ .....~..';-~..'~.~.~.~.. ~ ~.~~-.--.L.---.~..~,,., .£ x _.~o= 0 x ^ ,,n % ,.el ir--.xx o \ ",~ £

x , .\ i

> x .-~ ~x _~

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 546 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

DerbysMre[Staffordshire (xiv) Representative subsidences from the system described in this paper are shown. Yorkshire (xv) Sands and clays preserved in solution subsidences in the Chalk of the Yorkshire Wolds were considered by Versey (x937) to be ofpre-Glaeial age. Durham (xvi) The Castle Eden fissure deposits have often been referred to the Pliocene (Trechmann x915) though this dating has recently been questioned by West (1968) who suggests that a Lower Pleistocene age is more acceptable.

WALES Flintshire[Denbighshire (xvii) Over 20 solution subsidence outliers of probable pre-Glacial age have been preserved in the Carboniferous Limestone of these counties. Clay, sand and ?lignites have been reported, while figures of Maw (I865, x867) and Strahan ¢t al. (189o) strongly suggest that these subsidence masses are well- stratified. The occurrences have been redescribed recently by Walsh & Brown (I97I). Caernarvonshire (xviii) Werndow, Conway. A large mass of rotten rock with relic banding. (Walsh & Brown I97x). (xix) Y Llymllwyd. Gibbsitic clays of possible pre-Glacial age have been described by Ball (1964). Merionethshire (xx) Trawsfyrmydd. Gingerbread rock reported by Feamsides in the discussion of Greenly (x938). (xxi) Mochras borehole. Sediments of Tertiary aspect considered by Wood & Woodland (I968) to be of Palaeogene age, but at least the upper part, and possibly the whole, has subsequently been suggested by Miss M. Herbert- Smith to be of Neogene age (personal communication). Montgomeryshire (xxii) Rotten rock occurrence on Lower Palaeozoic slates (H. M. Montford, personal communication). Pembrokeshire (xxiii) Rotten rock occurrence on basic igneous rocks (H. M. Montford, personal communication). (xxiv) Rotten rock occurrence on Lower Limestone Shales. (Higgin- bottom & Fookes, I97o ). (xxv) Rotten rock occurrence in ?Namurian shales (Thomas 197o ). (xxvi) The Flimston outlier, often considered to be of Palaeogene age on lithological comparisons with the Bovey and Petroekstow deposits. However, the authors consider that a case for a Neogene age for this mass is no less strong, while a Pleistocene age cannot be ruled out. (Dixon i92 x). Garmarthensh ire]Breconshire (xxvii) Solution-subsidence deposits, considered here as relics of the Neogene weathering front (Thomas x964, see especially his remarks on p 56).

8C0 TLAND Aberdeenshire, Banff, Kincardineshire, Angus, Perth, Fife, Midlothian and Sutherland (xxviii) Rotten rock data from Fitzpatrick (I963) and Higginbottom & Fookes (i97o) (the latter regard certain masses of rotten rock as of Pleistocene age). (xxix) Oufliers of sands, clays and gravels after Flett & Read (x92I). Usually considered to be of Pliocene age. Sutherland, Ross and Cromarty, Inverness and Argyll (xxx) Rotten rock localities (personal communication, H. M. Montford). IRELAND Donegal (xxxi) Rotten Rock localities (personal communication, I--I. M. Montford). (xxxii) Rotten Rock localities (personal communication, T. Murphy). Sligo (xxxiii) Solution subsidences of unknown character, possibly of recent origin. (Personal communication, T. Murphy). Mayo (xxxiv) Extensive rotten rock occurrences, mostly described by Max (t97o). (xxxv) Rotten rock occurrence, similar to the Cloyne Clay deposits (personal communication, T. Murphy). Cavan (xxxvi) Rotten rock occurrence (personal communication, T. Murphy). Louth (xxxvii) Fissure deposits exposed in the Drogheda cement quarries, interpreted here as being subsidence products of the Neogene weathering front. (Murphy x966 ). Westmeath (xxxviii) Rotten Rock occurrence near Mullingar (Field observation, PTW). Meath (xxxix) Rotten Rock occurrence at Dtmshaughlin (personal communication, T. Murphy). Dublin (xx) Solution subsidence mass, similar to those at Drogheda (personal communication, T. Murphy). (xli) Solution subsidence masses, similar to those at Drogheda (personal communication, T. Murphy). Galway (xlii) Tynagh The weathered gossan of the Tynagh orebody is reported to preserve fossil wood, possibly of Tertiary type (Morrissey & Whitehead 1969). Offaly (xliii) Birr Pipeclay and silica sand deposits similar to the Ballymacadam (Co. Tipperary) solution subsidence deposits (Bishopp and McCluskey 1948).

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 548 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

Tipperary (xliv) Rotten rock occurrences, fairly widely distributed in association with the Ballymacadam Palaeogene deposits (personal communication, T. Murphy). Kerry (xlv) Rotten rock occurrences (field observations PTW). Cork (xlvi) The Cloyne Clay deposits of possible Neogene age (Bishopp & McCluskey I948; Murphy I966 ). NORTH SEA FLOOR (xlvii) spot heights on the base of the Neogene kindly provided by P. E. Kent and P. J. Walmsley of the British Petroleum Company. Generalised contours inserted from these data. Wells indicated include B. P./Hamilton 44[II-X ( --802 It) and B. P. 49/1-I (--564 It). IRISH/GEL TIC SEA FLOOR (xlviii) contours from Blundell et al. (t97I). Based on geophysical data, interpreted as Neogene in age. 8. References ARKELL, W. J. I948. The Geology of the Country around Weymouth, Swanage, Corfe and Lulworth. Man. geol. Suro. U.K. BALCHm, W. G. V. I95o. The Erosion Surfaces of Exmoor. Geogrl J., xxS, 453-476. BALL, D. F. I964. Gibbsite in Altered Granitic Rock in North Wales. Nature Land., 204, 673-674. BARK~, F., HIND, W. & SCOTT, A. x92o. Q uartzose Conglomerate at Cauldon Low, Staffordshire. Geol. Mag., 57, 76-82. BISHOPP, D. W. & MCCLUSKEY,J. A. G. I948. Sources of Industrial Silica in Ireland. The Geo- logical Survey of Ireland. Emergency Period--Pamphlet No. 3- The Stationery Office, Dublin. BLONDELL, D. J, DAVEY, F. J. & GRAVES, L. J. (x97x). Geophysical Surveys over the south Irish Sea and Nymphe Bank. J. geol. Sot., x27, 339-375. BOSWELL, P. G. H. x9x8. A Memoir of the British Resources of Refractory Sands. London (Taylor & Frances). BOULTER, i. C. x97I. A palynological study of two of the Neogene Plant Beds in Derbyshire. Bull. Br. Mus. nat. Hist. (Geol)., x9, 36o-4I I. & CHALONER, W. G. i97o. Neogene Fossil Plants from Derbyshire, England. Reg. Palaeo- botan. PalynM., to, 61-78. , FORD, T. D., IyrhBA, M., & WAIaH, P. T. I971 Brassington Formation: a Newly Recognised Tertiary Formation in the Southern Pennines. Nature Physical Science, 23t, 134-I36. BROWN, E. H. I96I. The Rdiefand Drainage of Wales, Cardiff (University of Wales Press). BRImSDEN, D., KIDSON, C., ORME, A. R. & WAanSl~S,R. S. I964. Denudation Chronology of Parts of South Western England. Field Studies, 2, x I5-x32. CLAYTON, K. M. 1953. The Denudation-chronology of the Middle Trent Basin. Trans. Inst. Br. geogr., 25-36. DixoN, E. E. L. x92x. The Country around Pembroke and Tenby. Mere. geol. Sum. U.K. DImBL~TON, M.J. & W~ST, G. I966. Studies of the Keuper Marl; Mineralogy. Road Research Laboratory Research Note No. 4 ° . Ministry of Transport. EDMONDS, E. A., McI~ow~, M. C., & WILLIAMS, M. I969. British Regional Geolog~v---South-West England. H.M. Stationery Office, London. F~AnmlDVS, W. G. I932. The Geology of the Eastern part of the Peak District. Proc. Geol. Ass., 43, I52-I9I. FEm~mo, P.J. & FRESmCEY, E. C. x968. Preliminary Results ofa Borehole at Petrockstow, North Devon. Geol. gag., xos, x88-x89.

FITZPATRICK, E. A, 1963. Deeply weathered rock in Scotland, its occurrence, age & contribution to the soils, d. Soil Sci., i4, 33-43. FLEET, W. 1938. Erosion surfaces in the Grampian Highlands of Scotland. Rapp. Comm. Cartog. des Surf. d'Appl, tert., Union geogr. Internat. 91-94 . FLETT, J. S. & READ, H. H. I92I. Tertiary Gravels of the Buchan District of Aberdeenshire. Geol, Mag,, 58, 215-225. FORD, T. n. 1968. The Millstone Grit. In Geology of the East Midlands. Leicester University Press (Eds. P. C. Sylvester-Bradley & T. D. Ford), pp. 83-94. & KING, R. J. 1968. Outliers of Possible Tertiary Age. In Geology of the East Midlands, Leicester University Press. (Eds. P. C. Sylvester-Bradley & T. D. Ford), pp. 324-33I. • 1969 . The Origin of the Silica Sand Pockets in the Derbyshire Limestone. Mercian Geologist, 3, 51--69 • in press. The Peak District Neogene Deposits and related features: Director's Report. Proc. Geol. Ass. FtmNELL, B. M. 196I. The Palaeogene and Early Pleistocene of Norfolk. Trans. Norfolk Norwich Nat. Soe., 19, 34o-364 . 1964. The Tertiary Period. In The Phanerozoic Time Scale. Geological Society of London, Special Publication No. I (Eds. Harland, W. B., Smith, A. G., and Wilcock, B.). GEORGE, T. N. 1966. Geomorphic Evolution in Hebridean Scotland. Scott. J. Geol., 2, 1-34. 1967. Landform and Structure in Ulster. Scott. J. Geol., 3, 413-448. GODARD, A. 1965 . Recherches de Gdomorphologie en t~,cosse du Nord-oest. Paris. GREENLY, E. I938. The Age of the Mountains of Snowdonia. Q. Jl geol. Soc. Lond., 94, 117-124 . HIGGINBOTTOM, I. E. & FOOKES, P. G. 197 O. Engineering Aspects of Periglacial Features in Britain. Q. Jl Engng Geol. Lond., 3, 85-1 I7. HOLLINGWORTH, S. E. 1938. The Recognition and Correlation of High Level Erosion Surfaces in Britain: a Statistical Study. Q. Jlgeol. Soc. Lond., 94, 55-84 • HowE, J. A. 1897. Notes on the Pockets of Sand and Clay in the Limestone of Derbyshire and Staffordshire. Trans. a. Rep. N. Staffs. Fld. Club, 3 x, I43-149. HUGI~tEs, E. M. 1952. The Geology of an area north of Brassington. Unpublished Thesis, Nottingham University. JACKSON, J. W. &; CHARLESWORTH,J. K. 192o. The Quartzose Conglomerate of Cauldon Low, Staffordshire. Geol. Mag., 57, 487-492. KErr, P. E. 1957. Triassic Relics and the I,OOO foot Surface in the Southern Pennines. E. Midl Geogr., I, 3-1o. LIIWrON, D. L. I956. Geomorphology. In Sheffield and Its Region. Brit. Assn. Adv. Sci. Sheffield (Ed. D. L. Linton), pp. 24-43. MAW, G. 1867. On the Distribution beyond the Tertiary Districts on White Clays and Sands Subjacent to the Boulder Clay Drifts. Geol. Mag., 4, 24°-25 I. MAX, M. D. 197 o. Some mineralisation in part of north-west County Mayo, Ireland, Scient. Proc. R. Dubl. Soc., 3, 264-273. McFADY~N, W. A. 197o. Geological Highlights of the West Country. London (Butterworth). MILNER, H. B. 1922. The Nature and Origin of the Pliocene Deposits of the County of Cornwall and their bearing on the Pliocene geography of the South-West of England. Q. Jl geol. Soc. Lond., 78, 348-377 . MITCHELL, G. F. x965. The St. Erth Beds--An Alternative Explanation. Proc. geol. Ass., 76, 345-366. MORRIS, P. G. I967. The Namurian (E 1 -- Ez) Strata between Waterhouses and Cauldon, near Leek, North Staffordshire. Proc. geol. Ass., 78, 335-418. 1969. Dinantian and Namurian Stratigraphy east and south-east of Leek, North Stafford- shire. Proc. geol. Ass., 80, 145-176. Mom~ISSEY, C. J. & WHITEHEAD, D. 1969. Origin of the Tynagh Residual Orebody, Ireland. Paper 24, 9th Commonwealth Metallurgical Congress, London. MURPHY, T. 1966. Deep Alteration of Carboniferous Strata in the Midleton, Co. Cork District as Detected by Gravity Surveying. Proc. R. Ir. Acad., 64B, 323-334.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 550 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

N~.vEs, R. t 958. Upper Carboniferous Plant Spore Assemblages from the Gastrioceras subrrcaatum horizon, North Staffordshire. Geol. Mag., 95, x-I9- I96I. Namurian Plant Spores from the Southern Pennines, England. Palaeontology, 4, 247-- 279. PmmNOTOr% J. x789. A view of th~ Present State of Derbyshire. (2 vols). Derby. PrrrY, A. F. x968. The Scale and Significance of Solutional Loss from the Limestone Tract of the Southern Pennines. Proc. geol. Ass. 79, x53-I 77- PH~MISTER, T. C. & S~PsoN, S. t949. Pleistocene Deep Weathering in N.E. Scotland. Nature, Lond., x64, 318-319 . PROIYDFOOT, V. B. I954. Erosion Surfaces in the Mourne Mountains. It. Geogr., 3, 26-35. Roo~Rs, I. & SIMPSON, B. I937. The Flint Gravel Deposit at Orleigh Court N. Devon. Geol. Mag., 74, 3o9-315- SCOTT, A. I927. The Origin ofthe High Peak Sand and Clay Deposits. Trans. Br. C~'am. Sot., 255-260. SHmL~Y, J. I959. The Carboniferous Limestone of the Monyash-Wirksworth area, Derbyshire. Q. Jl geol. Soc. Lond., xx4, 4I 1-429. SISSONS, J. B. 1954. The Erosion Surfaces and Drainage System of South-West Yorkshire. Proc. Yorks. geol. Sot., 29, 3o5-342. I967. The Evolution of Scotland's Scenery. Edinburgh (Oliver & Boyd). SMrrH, E. G., RHYS, G. H. & 2DEN, R. A. I967. Geology of the Country around Chesterfield, Matlock and Mansfield. Mere. geol. Suro. O.K. STAMP, L. D. x947. Britain Structure and Scenery. (lst 2dn) London (Collins). ST~,HAN, A. & D~ RANCh, C. E. x89 o. Geology of Flint, Mold and Ruthin. Mere. Geol. Surv. U.K. THOMAS, T. M. x963. Solution Subsidence in south--east Carmarthenshire and south-west Breconshire. Trans. Inst. Br. Geogr., 33, 45-6o. --- I97o. Field Meeting of the South Wales group on the Stack Rocks to Bullslaughter Bay Section of the South Pembrokeshire Coast; Report by the Director. Proc. geol. Ass., 8x, 241- 249. Tm~c~, C. T. x9 x5. The Scandanavian Drift of the Durham Coast and the General Glaci- ology of south-east Durham. Q. Jl geol. Soc. Lond., 7 x, 53-82. VERSEY, H. C. 1935- The Tertiary History of East Yorkshire. Proc. Yorks. geol. Sot., 23, 3o2-314, WALSH, P. T. I965. Possible Tertiary Oufliers from the Gweestin Valley Co. Kerry. It. Nat. J., 15, IOO-IO 4. & BRowN, E. H. 197 t. Solution Subsidence Oufliers Containing Probable Tertiary Sediment in N.2. Wales. J. Geol., 7, 299-32o. WARWICK, G. T. 1964. Dry Valleys of the Southern Pennines. Erdkunde Arch. wiss. Geogr., 18, I I6--I22. WEST R. G. x968. Pldstocene Geology and Biology. London (Longmam). WILSON, M. J. 1969. A Gibbsite Soil Derived from an Ultrabasic Rock on the Isle of Rhum. Scott. J. Geol., 5, 82. WOOLDRtDOE, S. W. x95o. The Upland Plains ofBritain: their origin and geographical significance. Adv. Sd. Lond., 7, I62-175- & Lm'TON, D. L. x955. Structure, Surface and Drainage in South East England. London (Philip). WOOD, A. & WOODLAm~,A. W. 1968. Borehole at Mochras, West of Llandbedr. Merionethshire. Nature, Lond., 219, 1352-1354. YORK-, C. I961. The Pocket Deposits of D~rbyshire. Birkenhead (Private Publication). Z~ul~a, F. E. 1952. Dating the Past. London (Methuen).

Submitted March 1971 ; revised manuscript received t February 1972; read 8 December 1971.

P. T. Walsh, nsc, PHD, VOS, Geology Laboratory, Department of Civil Engineering, The City University, St. John Street, London. EC xV4PB. M. C. Boulter, Bsc, P~m, FLS, Department of Biological Science, North East London Polytechnic, Romford Road, London. EI5 4LZ.

M. Ijtaba, Bsc, Msc, t,os, Geology Laboratory, Department of Civil Engi- neering, The City University, St. John Street, London, EC xV4PB. also: Department of Geology, Chelsea College, Manresa Road, London. S.W.3. D, M. Urbani, BSC(~.NO), Chief Engineers Division, The London Borough of Camden, 2 t 3 Haverstoek Hill, London. N.W.3.

DISCUSSION Mr P. EVANS, congratulated the authors on a very interesting and useful piece of work. There had been so much difference of opinion about the age of high-level erosion surfaces that it was important to have this definite evidence of a Pliocene surface. The low relief of the 45 ° m Brassington surface suggested that it had been eroded to a base-level only a little below this. Sissons's Holme Moss surface was about i oo m higher, and Mr. Evans agreed that it would have supplied sediment: base-level might well have been the Bradfield level of about 4o0 m which is trace- able at intervals as far as the south of France. Old high sea-levels back to about 200 m at something over a million years ago could be plotted to give the rate of fall of sea-level for that spell of time, but to continue back to earlier periods was guesswork. His independent guess put the Brassington Beds at 4 million years, and the authors had guessed 7. In the present state of ignorance about Pliocene dates, the difference between guesses of 4 and 7 million years was not worth arguing about.

THE AtrrHom thanked Mr. Percy Evans for his kind remarks, but would natur- ally prefer to have their 7 million year dating for the sub-Brassington Formation surface accepted as a carefully-reasoned deduction, after many years of palaeobotanical research rather than a 'guess'. Having ascertained that the Kenslow flora was approximately of Mio-Pliocene boundary age, the 7 million year figure was derived from Prof. Funnell's paper in the Society's 'Phanerozoic time-scale' memoir. The suggested correlation with Sissons' Holme Moss Surface has in mind the fluvial character of the Brassington Formation sediments and the knowledge that Hughes (i952 , p. 79) has recorded that current bedding of the (?) Kirkham Member sands of the Brassington Pits indicated a derivation from the north east. A southwards-sloping fluvial regime, having a 45 ° m elevation in the Brassington-Friden area might thus be reasoned to have an elevation roughly at the level of the Holme Moss Surface in the Aire-Don sector of the eastern slopes of the Pennines. Since the original manuscript of this paper was submitted, however, advice proffered by more than one geomorphologist who has active interests in the southern Pennines indicates that the concensus of contemporary opinion holds that it is most unlikely that any erosional remnant of a 7-million year surface could possibly have survived, even in the Gritstone areas. The authors believed that, at the present time, no one really knew whether or not high-level erosion surfaces could survive as concordant summits or spur-levels for this length of time, and, if not, to what extent they could survive as uniformly- lowered planar elements of landscape. They agreed that their Figure 7 implied a

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 552 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

degree of preservation of pre-Pleistocene surfaces comparable with those of the: Pleistocene (which perhaps was unwarranted). Accordingly, they were now more inclined to the view that if any physiographic homologue of the sub-Brassington surface existed in the Gritstone areas, it would: only be recognisable as a uniformly-lowered, fiat surface, almost certainly at no greater altitude than that of the reasoned position of that surface in the Brassington Friden area. A correlation with the Bradfield level therefore caused no particular distress to the authors.

Mr. W. B. EvANs remarked that, whilst he welcomed and agreed with that part of the paper dealing with the sequence of the deposits and their disposition, he was less happy regarding the argument relating to the original height of the Tertiary surface. In particular he was concerned that the probability of a significant sub- Namurian unconformity had been discounted. The regional stratigraphy around the nearby margins of the Derbyshire Dome suggested that, in the area described, Namurian sediments probably spread over an already eroded surface of limestone. The break was at its greatest east of Axe Edge, where Ric sediments spread over the Dovedale reef-complex onto the shelf-limestones of D 1 and D9 ages, and close to the east near Hurdlow where Dr. N. Aitkenhead has recently obtained RIo goniatites from the Namurian infilling of one of the solution pits. The likely presence of this regional unconformity throughout the area made it probable that the D9 limestones had been partly, and in places wholly, removed before the Namur- ian cover was laid down, rather than at a much later date. If this was the case the preservation of thick chert breccia lying between the wall-rock of the pits and the Namurian infill demanded an explanation. Admittedly the chert had formed as a solution residue from D2 limestones or from a bedded chert now removed, but its absence along the plane of the unconformity around the Dome suggests that it was preserved only where solution hollows existed before the Namurian transgression, and where these were partly filled by chert- scree from the surrounding limestone surface exposed before this transgression. Clearly the hollows had been reactivated in late-Tertiary times, and it was even possible that there had been other periods of reactivation. If subsidence had taken place in stages, an estimate of the height of the Tertiary surface depending upon an assumption of a single Tertiary phase of solution and emplacement might give misleading results.

The AUTHORS were grateful to Mr. Wyndham Evans for raising the problem of the origin of the chert-clay residues as they wanted to say more about this in the paper but could not do so for lack of space. The senior author had discussed the problem at length with Mr. Evans in the feld, and there was no dispute concerning. the field data--merely the interpretation to be placed on these. Mr. Evans was~ correct in assuming that the authors had discounted the possibility of a significant pre-Namurian unconformity over the areas where subsidence outliers of Brassing- ton Formation were found. They also believed that, because they consider that there is a total lack of positive evidence otherwise, the subsidences were no more ~

than a single phase effect, which took place beneath a Namurian shale cover and which post-dated the Brassington Formation deposition. The difference of opinion stemed from the second sentence of the second para- graph of Mr. Evans' contribution; in a sense, this is an intellectual debate, not a geological one, The authors disagree that ... 'absence [of the cherts] along the plane of unconformity around the Dome' ... necessarily ... 'suggests that it is preserved only where solution hollows existed before the Namurian transgression and where these were partly filled by chert scree from the surrounding limestone surface exposed before this transgression', They submit that the absence of chert residues along the wave-trimmed Sudetic unconformity around the Dome suggests, instead, with no less validity, that chert residues were equally absent from the wave-trimmed interior of the limestone massif. There is a vast outcrop of cherty Carboniferous Limestone in Western Europe, unconformably overlain by Nam- urian or younger sediments. But, except in the case of the Pocket Deposits, where solutioning was not less than partly post-Namurian, the writers have never found any reference of any sort to localised pockets of cherty-residue (regardless of the subsidence issue) being preserved beneath an unconformable cover. The authors believed that the exposed present-day contacts of the Namurian shale blocks and the cherty residues were not a reflection of Namurian shales coming to rest on solution residues, but were entirely due to the physical effects of solution and subsidence in late-Tertiary times. The facts that I. the solution residues are almost without exception uncemented and unmineralised 2. shale blocks are sometimes present apparently without the association with cherty residues (Low Moor, Hindlow etc) 3. cherty solution residues exist in many of the Flintshire and Denbighshire pockets figured by Maw (I 867) and Strahan (i 890 ) (see Walsh & Brown 197 I), represent positive evidence against Mr. Evans' concept of a localised pre-Namurian origin for the cherty- residues and negative evidence in favour of the authors' own view. However, the authors would agree that there is much more to be learnt about the cherty residues, and the solution to the problem might be approached if a more detailed examination of the fossil content of the chert fragments were made.

Mr. H. M. MONTFORD said there were remarkable similarities between the pocket deposits of Derbyshire (and similar deposits elsewhere) and recent modern tropical soil profiles. The common sequence of white sands overlaid by coloured sands and clays corresponds with the pallid and mottled zones of tropical soils, and the large block of sandstone at Low Moor pit, which is embedded in dark red sand containing numerous ferruginous nodules, could well be a fragment of a surface duricrust. This interpretation was not necessarily inconsistent with such features as sedimentary stratification and current bedding, since many tropical soils show evidence of transportation, having undergone pedogenetic development after deposition. It was suggested that this may have been the history of the Derby- shire deposits, although, in the absence of detailed mineralogical and pedological analysis, it was not possible to be certain at present about the details of their development.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 554 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

The Derbyshire deposits could be matched closely with many similar occurrences on the mainland of Europe, providing evidence of widespread similarity of conditions over a wide area, probably for a long period of time. Among such deposits was the remani~ of the Siderolithique of France; to the south and west of the Massif Central some outcrops could be dated from their relationship to fossiliferous deposits as pre-Lutetian (Upper Eocene). The Miocene volcanic rocks of the Massif Central are said to show no signs of tropical weathering; in Derbyshire, so much further to the north, tropical conditions must have come to an end even earlier. It was suggested that the Derbyshire deposits are remnants of a sheet of tropical soil which covered large parts of Europe in the early Tertiary, and that they had been emplaced in their present position and had undergone pedogenetic development before the end of the Palaeogene. On this basis, and accepting the dating of the plant fossils, there would have to have been a considerable time gap somewhere in the succession below the base of the Kenslow clays. This would accord with the suggestions of Kent and of Ford and King that the fossiliferous clays may have accumulated in small local lakes over the pockets; or it might be that the surfaces of the pockets were the only places providing the right conditions for fossilisation. The hypothesis of a late local origin for these clays was not necessarily negatived by the field appearance of conformity within the Bees Nest Member and between it and the Kenslow Mem- ber, since many apparently conformable tropical soil profiles have been shown by detailed analysis to be made up of units of different ages and provenances. However, there were wider grounds for doubt about the age of the fossils. Land plant communities tend to maintain a corporate identity in close relationship with climatic and ecological environments often over long periods of time. The correlation of the Kenslow clays was essentially with the Brown Coals of Central Europe; given the wide geographic separation and the relatively more oceanic environment of Derbyshire, there could be no certainty that the similar vegetations of the two areas (if they are similar) were not right out of phase, as for example, are the Taxodium swamps of the German Palaeogene with the floristi- cally similar swamps of modern Florida. The caution which palynologists currently observe about distant correlations of the comparatively well known Pleistocene floras is surely even more desirable when dealing with older and stranger floras. In view of these doubts about the age of the deposits and of their contained fossils, the geomorphological speculations in the paper appear to rest on very insecure foundations.

The AUTHORS replied that Mr. Montford had advanced a fundamentally different opinion concerning the interpretation of the field data, and they felt it necessary to reply in detail. The age of the Kenslow Member flora had been determined by comparison with a large number of plant assemblages throughout Europe, only a small number of which demonstrated the browncoal facies. These European assemblages show a very consistent floristic composition with little variation, other than for facies, from east to west. The complex problem of Neogene oceanic elements (such

the Tertiary equivalents of Hedera, Ilex and Empetrum) was much confused by their occurrence not only in the Kenslow flora but right across Europe to Poland and Russia. Nevertheless, there was no reason to believe that plants which gave rise to pollen types which make up the Pliocene and Miocene Components (from which the age of the Derbyshire flora has been chiefly determined) were effected by greater oceanicity. There were many differences between the Flora of the European Tertiary browncoal deposits and modern American Taxodium swamps. But there was a striking similarity between the Kenslow Member assemblage and those from Neogene Europe. The Upper Miocene--Lower Pliocene age of the Derbyshire flora has been confirmed by a number of European and American palaeobotanists, The authors had approached the problem of whether or not there is an unconformity below the Kenslow clays with a completely open mind. They had given a great deal of attention to the critical sections at the Bees Nest, Kenslow and Kirkham's Pits, over the past five years. The sections had been carefully dug, both manually and using mechanical excavators, and they were aware that others had done likewise. None of the authors could find any criterion whatever which indicated that there was other than a gradational contact between the Kenslow and Bees Nest Members. From the initial results of Mr. Ijtaba's sedimentological analyses, they were surprised to find a notable content of gibbsite in the Kenslow clays (up to about IO % by weight), whereas that mineral is absent from the lower parts of the succession. Careful checking of the mineralogy, however, had shown that gibbsite does not suddenly enter into the composition of the clays, at the same time as the fossil plants, (as the authors had originally supposed), but that the content reduces downwards across the Kenslow/Bees Nest boundary, and that the mineral still appears to be present not less than 15 ° cm below the level at which the oldest fossil plant material is preserved. To say that the mineralogy of the Brassington Formation sediments had not been studied in detail is incorrect by any yardstick. Mr. Ijtaba had amassed a considerable amount of data on clay minerals, heavy minerals and particle shape and size analysis for the sediments preserved at the Bees Nest and Kirkham's Pits. None of these data suggests that the sedimentation of the Brassington Formation sheet was other than a single, comparatively rapidly deposited, 'normal' fluviatile sequence. There is a continuity of mineralogical and sedimentological features which rendered the suggestion that hidden non-sequences are present to be a totally unnecessary complication to the interpretation of both field and laboratory data. Accordingly, the whole sequence had to be regarded as of Neogene age, with the corollary that, if anything, the sedimentation (as represented by what is now preserved in the solution cavities) probably spanned a relatively short period of Neogene time. The authors could claim no particular expertise in pedology. They admitted that their suggestion that the lilac-stained, weathered Namurian shales lying immediately below the basal bed of the Kirkham Member represent a lateritic or quasi-lateritic soil profile, was little more than an educated guess, and rather than present an inadequate account of its substance they would prefer to leave this particular problem to those who have a greater familiarity with tropical

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 556 P. T. Walsh, M. C. Boulter, M. Ijtaba & D. M. Urbani

soil profiles. They insisted, however, that while there were certain pedological aspects about the Brassington Formation as a whole--for instance the reddening of the Bees Nest clays--the main physical features were those of a fluvial sedimentary succession and not, grossly, those of a soil profile. To reason that the Brassington Formation is largely a soil profile would be analogous to saying that the British Old Red Sandstone or the Bunter Sandstone is a soil profile. In our present state of knowledge, this was playing ducks and drakes with terminology. Mr. Montford draws attention to the Low Moor sandstone block; the authors consider that this is a Namurian sandstone, contained within the mass of a larger block of Namurian shales, reported by Ramsbottom (in Yorke, 1961) to contain E2c goniatites, and not a duricrusted mass of Brassington Formation sand (see Ford in press). The point should easily be resolved by anyone who was familiar with thin-section petrography of local Namurian sandstones. In any case, the block lies stratigraphically below the main mass of the Kirkham sands here, which, as revealed from field notes made by the senior author in 1964 (before the section became concealed by dumped industrial waste) line the collapse structure, younging inwards from the walls in the usual development.

Mr. B. C. WORSSAM said he had listened to the reading of this paper with much interest, but was concerned to note that the authors had used Wooldridge and Linton's Pliocene Platform as a datum plane for correlation. He would like to know if the authors had devoted as much care to assessing the evidence for this platform, as they had given to determining the level of the Pennine land surface prior to formation of the Pocket Deposits. The concept of the Pliocene Platform depends largely on evidence provided by the Lenham Beds. The only well-preserved parts of this formation are piped into a Chalk surface which is about 630 ft (I 92 m) above O.D. at Lenham. An attempt to determine the pre-subsidence level of the Lenham Beds would be of great interest. Clement Reid suggested on faunal grounds that the Lenham Beds accumulated at a depth of 4 ° fathoms. His estimate had not been contradicted in print, so far as the speaker knew. With no allowance for subsidence it implied a sea level at 870 ft (265 m) above O.D. The implications of a sea level at this height were taken into account by Jukes-Browne in his 'Building of the British Isles' in 1892, but they appear not to have been given due consideration by the geomorphologists responsible for currently-accepted theories about platforms.

In reply to Mr. Worssam, the AUTHORS said they had not examined the Lenham Beds, and know of them only from the literature. They agreed that it would be interesting to determine the amount of subsidence to which they have been subjected. Possibly, as with their Derbyshire subsidence analogues, this could partly be diagnosed by reference to the Chalk fossils preserved in the marginal residues of non-calcareous matter. The authors believed (perhaps wrongly) that, although it was very reasonable to link the Lenham Beds and the 63o-ft East Kent Surface, there was no substantiated evidence for the presence of unsubsided Lenham Beds masses and the association was therefore assumed and not proven. As a

suggestion, could the East Kent Surface have been cut across the Chalk and Len- :ham Beds pipes alike, as had happened in Derbyshire in the case of the so-called I ooo-ft surface (whatever the origins of the latter) ? The use of the 6oo-ft sea-level of S.E. England in Fig. 7 refers to the Netley Heath deposits of supposed Red Crag (basal Pleistocene) age, which, being younger than the Lenham Beds were possibly much less affected by solution subsidence from their original position and which might, in any case, represent rather more of a shallow water deposit than do the Lenham Beds. Further small outliers bearing Red Crag fossils had, of course, also been identified at Rothampstead in the Chilterns (for recent commentary, see Hey, Krinsley & Hyde 197 I). Being further away from the fringe of the North Sea Basin, these basal Pleistocene relics seemed to offer a more reliable datum for the inter-regional correlation of planation surfaces than do the Lenham Beds. The authors would again stress the point made in the paper and elaborated during its reading. They believe that their Fig. 7 cannot possibly have any more worth than a crude, rule-of-thumb guide as to the significance that these high level deposits of known age and elevation have in the estimation of the age of other proposed planation surfaces. If more high level fossiliferous late Tertiary deposits were found in the British Isles, it might then become possible to analyse their physiography in absolute temporal terms.

Professor F. W. SHOTTON remarked that it must have been around 1960 that he and Dr. M. Kelly first examined similar deposits in Friden Great Pit and then, due to the pressure of other work, had to relinquish claims to research on them. They had already, however, come to certain conclusions which the authors had amplified at Brassington. There seemed no doubt that all the deposits in the pocket were continuous, and presumably of Cainozoic age since pollen from the topmost part of the succession yielded Tsuga and Car_ya which could not be younger than very early Pleistocene. The inclination of the strata, approaching verticality at the pit edge, conclusively proved that the solution and subsidence had been subterran- ean. Professor Shotton had also found the 'clay with cherts' margining the pocket and like the authors had interpreted this as a solution relic of cherty Carboniferous Limestone. He had examined the quartzite and sandstone pebbles in gravels near the base of the sequence and concluded that these were quite unlike any Bunter pebbles that he knew. So he had discounted the explanation of a Triassic age for the pocket deposits. He wished to ask the authors two questions--what environment of deposition did they envisage, and had they any idea of the original areal extent of the Neo- gene deposits ?

The AUTHORS thanked Professor Shotton for his support for many of their conclusions. Until Mr. Ijtaba has completed his sedimentological studies, they were not able to say anything more definite concerning the environment of deposition beyond what had been conveyed in the paper. Of the environment of deposition, they could certainly say that no marine influences had been detected either from the sedimentology or the palaeobotany. Essentially, they followed previous opinion

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021 558 P. T. Walsh, M. G. Boulter, M. Ijtaba & D. M. Urbani

that the succession was dominantly 'fluviatile', and it was hoped shortly to under- take studies of the provenance and distribution pattern of the sediment by careful analysis of the current-bedding structures preserved in the Kirkham sands (al- lowing, of course, for rotation into the solution cavities). The authors were of the opinion that the sheet or sheets of Kenslow clay were very much more widespread in their pre-subsidence condition than had been thought previously (cf. Kent 1957; Ford & King 1969) and, from existing data, they felt certain that the clays were formed very close to sea level. It would be profitable to speculate whether, had another to m thickness of Brassington Formation sediment been preserved, a marine development would have been forthcoming; certainly, in the unlikely event of a future excavation of a pocket even larger than that at Kenslow Top (the Friden Great Pit), this possibility should be borne in mind.

Professor E. H. BROWN told the meeting that the lithology, altitude and geomorphological significance of the deposits described by the authors bore a close resemblance to similar deposits in northeast Wales described by one of the authors, and himself (Walsh, P. T. and E. H. Brown, x97 I, Solution Subsidence Outliers containing probable Tertiary sediment in North-East Wales, Geol. J. 7, 299-32o). The estimated amount of subsidence undergone by the deposits in both areas suggested that they were originally laid down at a level well above the plateau surface at around 300 metres, into which they had now subsided. In north-east Wales the original surface upon which the deposition occurred was probably the Middle Peneplain circa 500 metres above present sea level rather than the High Plateau, circa 600 metres. By analogy the Brassington Formation was likely to have been deposited at a lower level than that of the Holme Moss Planation surface. Did the authors consider that the nature of the Brassington Formation is such as to suggest that it may originally have been part of a continuous sheet extending from the southern Pennines into north-east Wales ?

The AUTHORS replied that both Professor Shotton and Professor Brown had en- quired about the original extent of the Neogene deposits. There was no real evidence of a limit to the basin of deposition anywhere, except between Flagg (SK x367) and Hindlow (SKo868). As reported during the reading of the paper, Kenslow- type clays bearing fossil plants had been found at the latter, in direct contact with Namurian shales and without the intervening presence of Kirkham or Bees Nest Members. This might imply a local western limit to the two lower members until late in the history of the basin, with the hint that the trend of this boundary had a NE-SW component. Elsewhere, there was no evidence that the outermost preserved masses of Brassington Formation sediment lay close to the margins of the area of deposition; thicknesses of not less than 7 o m appeared to be present at Kenslow Top, >43 m at the Bees Nest Pit and an unknown, though considerable thickness at Ribden implied an extension of the sheet to the NE, SE and SW, well beyond that area within which subsidences had taken place. Recent research into the Flintshire, Denbighshire, Staffordshire and Derbyshire deposits had shown that the post-Armorican Massifs of Wales and the Pennines had been upwarped to a significant degree in late Tertiary times (the age of the

Welsh material is not proven, but is assumed to be roughly the same as that from Derbyshire). A fascinating feature was that these massifs could be shown not only to be suppliers of sediment (which might have been expected), but also, sur- prisingly, that the massifs had been recipients of sediment, at least on their fringes. The authors would have had no need to invoke the presence of this sediment had it not emerged from the solution cavities; they knew of it only because karstic subsidence had proceeded faster than the general lowering of the nearby landscape. From general physiographical considerations, it seemed difficult to envisage that, if the fringes of the massifs were receiving sediment, the post-Armorican Basins of Cheshire and Liverpool Bay were not. While it was not impossible that the lateral equivalents of the Brassington Formation might yet be detected by bore- holes through the Drift of the Cheshire Basin and on the floor of the Irish Sea, it seemed likely that fluvial and glacial erosion had largely cleaned out any soft Neogene fill that was originally present in these areas and the authors felt that the key to the ultimate solution to this problem lies in a constant observation of new sections in the Carboniferous Limestone. This has a wide outcrop on the fringes of the massifs and basins alike; it had already proved its worth as a tool for geomorphological research and it was up to future generations of geologists and geographers to ensure that temporary sections of karstic subsidence material did not go unrecorded.

Mr. H. E. P. SPENCER said he found this paper of particular interest. Recently he had made a new examination of the mammalian fauna discovered early this century at Dove Holes, Derbyshire, and originally described by Boyd Dawkins. The remains included the Upper Pliocene Mastodon, M. arvernensis, Archidiskodon meridionalis, 'Machairodus crenatidens' now Homotherium sainzelli, Hyaena, 'Cerous etuerarium" now Dama nesti nesti and Eucladoceros spp., also Equus bressanus. These are members of the Villafranchian mammalian Series which range from the Upper Pliocene to the Lower Pleistocene and are represented in Britain in the East Anglian Crag Series from the Red Crag to the Cromer Forest Bed Series. Their Age is regarded as probably belonging to the continental interval between the Pliocene and Early Pleistocene marine stages. The fact that these fossils occurred in a cave deposit is an indication that considerble topographical changes must have taken place in the region, a view which seems to have been regarded with disfavour by some geologists. The results of research on the Brassington Formation is highly commendable and has provided positive evidence of Mio-Pliocene deposits in Britain.

The AUTHORS thanked Mr. Spencer for his kind remarks.

Downloaded from http://pubs.geoscienceworld.org/jgs/article-pdf/128/6/519/4884604/gsjgs.128.6.0519.pdf by guest on 24 September 2021