Archaeology: The Lower Palaeolithic Age

ARCHAEOLOGY

THE LOWER PALAEOLITHIC AGE

Most of the Lower Palaeolithic archaeological material in Middlesex has come from either the river gravels or the brickearths which in many areas overlie them. So far no evidence of human occupation is known from the period when the Thames was flowing through the vale of St. Albans and, later, through the Finchley Gap, and the earliest Middlesex material can be dated to a period no earlier than the Boyn Hill stage in the valley of the modern river.

Although Middlesex is a comparatively small county and that part bordering the river of no great length, a considerable amount of material has been found over the last hundred years. The discoveries are restricted, as elsewhere in Britain, to the stone tools or waste chippings made by the hunters of the Palaeolithic period, or the bones and teeth of contemporary fauna. Rarely are these remains found in situ, where they were actually made, used, or dropped, for the great majority have been swept by flood-waters off land surfaces which no longer exist, and have become incorporated in the deposits of gravel and sand which represent ancient courses of the Thames or its tributaries. Sometimes, however, silt and clay have been deposited in slack water and have buried an ancient land surface intact, with its 'floor' of flint tools and flakes, as at Creffield Road in Acton and at Stoke Newington.

The flint artifacts found in the ancient river gravels are frequently rolled, battered, and stained brown by iron oxides. Conditions have been unfavourable for the preservation of anything but flint or bone, although Worthington Smith claimed to have found pointed birch stakes in the 'floor' he examined at Stoke Newington. Objects of wood, leather, and bone, fashioned or utilized by the Palaeolithic hunters, have perished without trace, and any reconstruction of their life and environment must be largely conjectural.

Archaeologists have distinguished various groups of stone tool types, referred to as 'industries'. The main industries identified in the Thames valley are Clactonian (after the type site at Clacton-on-Sea, Essex), Acheulian (type site, St. Acheul, a suburb of Amiens (Somme)), and Levalloisian (type site, Levallois, a suburb of Paris), and in principle chronologically related to each other in that order. They show a progress in stone-working ability that probably reflects human development during the latter part of the Pleistocene period. Sub-divisions within the industries can sometimes be distinguished.

Faunal remains in the river gravels of cold-loving animals such as the mammoth and woolly rhinoceros, boulder clay, and solifluction deposits testify to glacial periods, and the main problem of the archaeologist is to determine into which warm or cold phase of the complicated Ice Age sequence an industry belongs. Much of the evidence comes from Thames gravel which dates from the Great or Hoxnian Interglacial, which began, according to the interpretation of the Milankovitch astronomical time-scale, about 400,000 years ago. The earliest stone industry is the Clactonian. No human remains have ever been found with it, although it seems very likely that Pithecanthropus was associated with it. The Swanscombe Skull, associated with Middle Acheulian handaxes, is devoid of its frontal bone, but is most closely paralleled by the Steinheim Skull, generally regarded as intermediate in type between Pithecanthropus and Homo sapiens.

The typical industry of Neanderthal Man, the Mousterian, is not found in south-east England, but the Levalloisian may have been his work. These primitive hunters span a period of at least 300,000 years in Britain. About 35,000 years ago, according to estimates based on radiocarbon datings, Homo sapiens was to be found in Europe, and from this time until the final retreat of the glaciers in northern Europe is the period of the Upper Palaeolithic with its improved hunting techniques and cave painting. This period is not represented in Middlesex.

The greater part of the evidence for the Lower Palaeolithic in Middlesex was collected at a time when the acquisition of specimens was of more interest than the geological context from which they came. The material so collected is now of very little scientific value, since in many cases the stratigraphical position of the objects is not known, nor sometimes the exact locality. Much of the collecting was done, not by controlled excavation, but by purchase from the gravel diggers. Payment of comparatively large sums to the workmen for specimens led not only to forgery, sometimes of a very high standard, but also to the salting of pits with material from more prolific localities. (fn. 1)

While little value can be placed on this earlier collecting, its history is not without interest, as it reflects, to a very large extent, the pattern of archaeological thought throughout Europe. The digging of these Thames gravels over the centuries must have brought to light a considerable quantity of archaeological material, but this does not appear to have aroused any curiosity. Some speculation as to the possible age of stone implements began as far back as the early 16th century, but it was not until the late 17th century that a clear association of artifacts and extinct fauna was demonstrated. At the end of the century a Mr. Conyers found, opposite 'Black Mary's' near Gray's Inn Lane, a handaxe associated with the tooth of an elephant, (fn. 2) probably that of either a mammoth or the straight-tusked elephant (E. antiquus). John Bagford in a letter dated 1715 suggested that this elephant was probably introduced by Claudius, but at the same time accepted the implement as being of human workmanship. This discovery made no impression on the antiquarian world, nor did the finds of John Frere at Hoxne (Suff.) in the later 18th century fare any better, in spite of his prophetic remark as to the possible age of the implements which he had found. (fn. 3) Further finds of Palaeolithic implements were made during the 19th century in Gray's Inn, Drury Lane, and off Oxford Street and Piccadilly, (fn. 4) but only with the work of Lane Fox, Allen Brown, and Worthington Smith can systematic work in Middlesex be said to have begun.

In 1869 Lane Fox's attention was directed to the river gravels at Acton. (fn. 5) In 1872 Evans published the first edition of his Stone Implements which inspired Worthington Smith to keep watch on the gravel excavations in the neighbourhood of Stoke Newington and in west Middlesex, and to undertake extensive work in Hertfordshire and Bedfordshire. In 1879 Worthington Smith published his results from west Middlesex, (fn. 6) and these were later extended, with his East London results, into the well-known account, published in 1894. (fn. 7) A short paper was also published by Greenhill in 1884 on Palaeolithic implements from the same area. (fn. 8) Allen Brown's research was mainly in the west of the county around Yiewsley, West Drayton, Dawley, (fn. 9) Southall, (fn. 10) and Ealing, (fn. 11) although he also made excursions into east Buckinghamshire as far as Iver.

Since very little work has been done in the Lower Palaeolithic in the county for the last sixty years, it is necessary before trying to put the work of these three investigators into perspective, to make brief excursions into the adjoining counties of Kent and Buckinghamshire where the archaeological and geological sequence has received considerable attention over the last forty years, and, although still not complete, is a great deal clearer than that of Middlesex.

By far the best-known locality from which Palaeolithic material has come is that of Swanscombe (Kent). (fn. 12) At Barnfield Pit near Swanscombe Halt gravels of the Boyn Hill Terrace overlie Thanet Sand, which in turn rests on chalk. These gravels are divisible into several horizons. The Lower gravel, directly on the Thanet Sand, contained a Clactonian industry associated with a rich fauna of Middle Pleistocene type. Immediately above is a loam, the Lower Loam, which, although originally a flood loam of the river, has been subjected to extensive weathering, indicating that during the period in which this weathering was taking place the river had retreated to a lower level. There is evidence of this low-level stage at Clacton-on-Sea (Essex), (fn. 13) where gravels containing slightly later Clactonian industry have been found in a buried channel. (fn. 14) Following this low sea-level, which was responsible for the corresponding drop in river-level, there was a rise which took the river to a higher point than when it laid down the Lower Gravels and Lower Loam. This high sea-level stage left gravels above the Lower Loam, and contained an archaeological industry with abundant hand-axes, which is referable to the Middle Acheulian; there was also a rich fauna, slightly later in type than that from the Lower Gravels. This Middle Gravel and the earlier Lower Gravel are both referred to the same Interglacial, that of Mindel/Riss, (fn. 15) and are the same age as the Boyn Hill gravels farther up the river in Middlesex and Buckinghamshire. The level of the Thames appears to have dropped sharply following the deposition of the Middle Gravels, probably in response to the oncoming cold conditions of the Riss Glaciation. Directly above the Middle Gravel is the Upper Loam, which may be a flood loam of the river similar to that of the Lower Loam, or possibly a hill wash. Above the Upper Loam is the Upper Gravel which is a solifluction deposit due to cold conditions, and further evidence of the onset of the Riss Glaciation. From the Upper Loam came a series of hand-axes, mostly cordiform in shape and frequently with the edges twisted.

From the evidence provided by the Barnfield Pit sequence, it is clear that during the early part of the Mindel/Riss Interglacial there was a Clactonian industry, followed by a Middle Acheulian, which in turn was followed by a later stage of Acheulian, but all three within the same Interglacial. At the end of this Interglacial there was a period of intense cold, the Riss Glaciation with marked erosion as evidenced by the solifluction deposits of the Upper Gravels. There is also evidence that during the low sea-level interval, represented by the weathering stage of the Lower Loam, Clactonian man was living on the Thames near Clacton-on-Sea. (fn. 16)

Further evidence for a drop in river-level in response to the oncoming cold conditions is provided by marked down-cutting by lateral streams through the deposits of the Boyn Hill Terrace, causing channels which are filled with later gravels; examples of this can be seen in the Wansunt Channel, between Swanscombe and Crayford (Kent). (fn. 17) Associated with these channels are hand-axes similar to those from the Upper Loam.

The stages in the fall of the river during the Riss Glaciation are not yet clear, but the glaciation is known to have had two peaks of cold with slightly milder conditions between, (fn. 18) during which the river would presumably have risen slightly. The next well-marked stage is that of the Taplow Terrace, which is generally dated to the rise of the sea in the Last Interglacial, Riss/Würm. It is possible, however, that there are more terraces between the Boyn Hill and the Taplow: there are, for example, gravels at intermediate heights at Grays Thurrock (fn. 19) and Ilford (Essex), (fn. 20) which may represent short halts in the rise and fall of the river; but their exact position on the cycles of change is still uncertain. From the former came a Clactonian industry, possibly slightly later typologically than that of the Lower Gravels of Swanscombe and the buried Channel at Clacton-on-Sea, and from the latter fauna which suggests a stage within the same (Mindel/Riss) Interglacial as the Swanscombe Gravels.

The Taplow stage is represented in the lower Thames by thick deposits of brickearth, which are assumed to have been the result of either estuarine conditions or a very slowrunning stream. These deposits are clearly seen at Crayford (fn. 21) and Ebbsfleet (Kent). (fn. 22) The industry from these brickearths is Levalloisian of a rather late type. In Baker's Hole, (fn. 23) just east of Swanscombe, Levalloisian has also been found, associated with hand-axes. (fn. 24) At present there is some uncertainty as to the age of this material. These brickearths have generally been considered as being of Taplow age, that is the Last Interglacial or Riss/Würm, and the material from Baker's Hole even earlier, the solifluction deposits from which they came being attributed to a stage of the Riss Glaciation. (fn. 25) The dating of the early Levalloisian to Riss has been advocated for many years by some French prehistorians. (fn. 26) But more recent work has cast doubt on such a high antiquity for the Levalloisian at Baker's Hole, (fn. 27) and the evidence that the Levalloisian is earlier than the Last Interglacial is open to question. In northern France the Levalloisian continues into the first and possibly into the second stage of the Last Würm Glaciation.

In Buckinghamshire the terrace complexes are more clearly defined than in the lower Thames, and the archaeological content of many of them is well known. The complete succession of terraces is outlined in the previous section, and only those terraces which contain archaeological material will be discussed here. In the Middle Thames area no archaeological material has been found in a terrace earlier than the Boyn Hill. Following work at Burnham Beeches, (fn. 28) Furze Platt near Maidenhead (Berks.), and Lent Rise near Taplow, (fn. 29) the Boyn Hill Terrace is generally thought to represent at least two stages. The upper or original Boyn Hill is at approximately 180-5 ft. O.D. or 103-8 ft. above the river. There is, however, about 20 ft. below, a further terrace which is visible at Furze Platt, Lent Rise, and at Baker's Farm south of Farnham Royal. Originally these two gravel spreads were mapped without distinction as Boyn Hill, but the archaeological content of the two series differs considerably. The upper series at East Burnham contained derived hand-axes which typologically resemble those from the Upper Loam at Swanscombe. In the lower stage at Furze Platt, Lent Rise, and Baker's Farm the material is unrolled and is similar to the Middle Acheulian from the Middle Gravels at Swanscombe, and is clearly earlier than Burnham. It has recently been argued that the material from the gravels of the higher stage represents a sweeping of the area, possibly the result of considerable erosion due to cold conditions, and that, archaeologically, the equivalent of the Swanscombe Middle Gravels are the Furze Platt and Lent Rise terraces. (fn. 30) It is already known that following the deposition of the Upper Loam at Swanscombe there was a period of intense erosion which gave rise to the solifluction deposits which are referred to as the Upper Gravels, and it is tempting to equate the erosion on the higher of the Boyn Hill terraces to the same cold phase.

At Iver (fn. 31) several pits lying at a lower level than those of Furze Platt and Lent Rise were investigated. Here was a poorly-sorted gravel with derived hand-axes of the same type as the Upper Loam and the derived specimens from East Burnham. The composition of these Iver gravels suggests that they were the result of intense erosion which brought material into the river, which it was unable to sort in the ordinary way. It thus seems clear that following the deposition of the Furze Platt and Lent Rise gravels there was here, as in the lower Thames, a period of marked erosion, which is probably due to one of the stages of the Riss Glaciation.

Immediately above the river gravel at Iver, shown particularly clearly at Mansion Lane Pit, (fn. 32) is a solifluction deposit, followed by two clearly separated brickearths, containing Late Levalloisian. (fn. 33)

It has been suggested that the Lower Boyn Hill Terrace and the deposits at Iver should be grouped as one terrace, the Lynch Hill, which should be considered as extending across the Colne into west Middlesex. (fn. 34) The marked difference in the archaeological content of the Lower Boyn Hill and Iver, the former with unrolled Middle Acheulian hand-axes of Swanscombe Middle Gravel type, and the latter with hand-axes of Swanscombe Upper Loam type, rolled, with earlier material obviously derived from the Lower Boyn Hill Terrace, suggests that this grouping is not justified.

At the type locality at Taplow the so-called 50-ft. terrace is clearly defined, and it is also well marked between Maidenhead and Cookham (Berks.). Very little archaeological material has come out of the gravels of this terrace. Finds, particularly from the Taplow district, consist of hand-axes and flakes which have been derived from the older gravels where these lie above the terrace. The Taplow Terrace deposits lie on chalk and consist of well-sorted sands and gravels at the base, with river gravels above. Clearly distinguishable from them is a mass of ill-sorted and ill-washed material, which appears to have resulted from a hill wash probably due to extensive erosion. There is evidence of frozen soil action at the junction between the solifluction deposit and the underlying river gravel. Above the solifluction is a brickearth, part of which has been subject to weathering. At Langley Levallois material has been found in the brickearth. (fn. 35)

As far as the evidence from Buckinghamshire goes there appears to have been a welldefined terrace of the same age as the 100-ft. terrace at Swanscombe with the same archaeological material, Middle Acheulian. Following the formation of this terrace, there appears to have been a drop in the river-level to an unknown extent accompanied, as far as can be seen, by a period of considerable erosion. This erosion appears to have taken place immediately after the formation of the Upper Loam at Swanscombe, since material of this age occurs in the gravels at Burnham and Iver. The next well-marked stage in the rise of the river is represented by the gravels of the Taplow Terrace, which, except for derived material, has no archaeological content, although faunal remains have been found at the base of the river gravels. This Taplow stage was in turn followed by a period of solifluction with signs of frozen soil action, the whole capped by a brickearth, which at Langley yielded Levalloisian of the same type as that from Crayford. A late Levalloisian was also found in the two stages of brickearth at Iver, so there is a possibility that these two series of brickearths, Taplow and Iver, are of approximately the same age. Following the deposition of these brickearths, presumably the lowest of the Flood Plain terraces of the Middle Thames was laid down, but this has not produced any archaeological material in the Middle Thames.

Into this archaeological sequence the material from Middlesex can, to some extent, be fitted. The amount of material which can be attributed to the High Terrace is small, and only stray pieces have been found. Allen Brown collected implements from the Town Pit, Hillingdon, at an altitude of 177 ft., and although this is higher than the general level of this terrace in the area, there is no reason to suppose that they did not come from this terrace. Brown also found a hand-axe on the lower slopes of Castlebar Hill, Ealing, which also belongs to the same terrace.

The Boyn Hill Terrace through London is much denuded and no implements have been found in quantity, as at Swanscombe and in Buckinghamshire. The locality of many of the hand-axes which have been found in the past is not clearly marked, so it is not possible to attribute them with any accuracy to the Boyn Hill Terrace. There are several specimens in the Sturge Collection from the higher ground of North London. One specimen from Pentonville is a small ovate with a markedly twisted edge resembling those from Burnham, Iver, and the Upper Loam at Swanscombe. (fn. 36) It appears from this that the erosion phase which followed the deposition of the Boyn Hill Terrace in the Middle Thames also occurred in the London area. The extensive finds made by Worthington Smith came from a lower level than that of the Boyn Hill Terrace, and what he refers to as the 'Palaeolithic floor' ranges from 50-70 ft. O.D. This floor, according to Smith, covers a wide area, Stoke Newington Common, Abney Park Cemetery, and Clapton, and he further claimed that it extended well into Hertfordshire, and eastward across the River Lea. (fn. 37) Although there is a wide variation in height over this area, the general pattern of the sections is very uniform. At the base of the deposits is a coarse gravel with a surface about 65-70 ft. O.D. A feature of these gravels is the large blocks of sandstone, some weighing as much as 5 cwt., and material derived from Hertfordshire and the Lower Greensand of Kent. These large blocks and the foreign stones suggest the breakdown of a glacial drift, probably that underlying the Boyn Hill gravels at Hornchurch (Essex). (fn. 38) The Stoke Newington gravels have produced rolled hand-axes at the base and less-rolled specimens on the surface of the gravels. Separating these gravels from Worthington Smith's 'floor' are bands of sands and clays with land and fresh-water shells. The floor, only a few inches thick, contains hand-axes and flakes as well as hammer-stones, and is undoubtedly a working floor, covering, if Worthington Smith is correct, a large area. Above the floor is evidence of very cold conditions. In many places the floor is disturbed by frost action resulting in marked festooning and frost cracks. These festoons and cracks are overlaid by what Worthington Smith describes as 'contorted drift', which is clearly a solifluction deposit.

The material from these deposits falls into two groups, at least at Stoke Newington. The rolled series is similar to that from the Middle Gravels at Swanscombe, that from the 'floor' is also Acheulian, but characterized by very small hand-axes made of pebbles, sometimes only two inches long. The flora from the floor is temperate and there is little doubt that this deposit is in the same Interglacial as Swanscombe, although possibly a little later than the Middle Gravels. Worthington Smith assumed that the 'floor' was one geological unit covering a wide area in East London, but the morphology of the ground is against this, as there is a very marked fall between Stoke Newington and Clapton, clearly indicating at least two terraces. This is also borne out by the archaeology since Lower Clapton has produced typical Levalloisian.

From western Middlesex, particularly from West Drayton and Yiewsley, a considerable amount of material is available of which the largest series is the Galloway Rice Collection in the London Museum. West Drayton, Yiewsley, and the adjoining hamlet of Dawley lie on a well-defined terrace at about 100 ft. O.D. Both the gravel and the overlying brickearth have been dug commercially for many years, and several of the old pits are still being worked.

This terrace is generally referred to as the Upper Taplow Terrace or Taplow Terrace I to distinguish it from the main Taplow Terrace below. This terminology is somewhat confusing as there is no connexion between the two terraces, but in the early geological mapping they were both attributed to the Taplow stage as defined in the Middle Thames. In spite of the vast amount of material which has come from this area there has never been a definitive publication on the deposits. (fn. 39) These consist of three main units: a river gravel at the base covered by a solifluction deposit containing much chalk and in turn covered by brickearth. In some pits, for example Pipkin's at West Drayton, the junction between the solifluction and the brickearth is very irregular, and clearly the solifluction deposit was subjected to considerable erosion before the deposition of the brickearth. At Eastwood's Pit, also at West Drayton, this irregularity is not so marked.

The artifacts in the Galloway Rice Collection are marked with the pits from which they came, and can be divided broadly into four groups. The first consists of slightlyrolled flakes which are similar to the Clactonian series from the Lower Gravels at Swanscombe, and the second is a group of slightly-rolled hand-axes which belong to the same Middle Acheulian as the Middle Gravels at Swanscombe, and the Lower Boyn Hill series from Furze Platt and Lent Rise. These Middle Acheulian hand-axes are derived from the breakdown of a section of the Boyn Hill Terrace, which must have passed to the north at a higher level than West Drayton; as the rolling is generally slight they have not travelled very far. The remaining two groups are both un-rolled. One is a series of hand-axes in mint condition which are typologically later than the Middle Acheulian but do not resemble the later material from the Upper Loam. On the evidence from the Somme in northern France, where a very complete series of Acheulian has been found, thesehand-axes appear to be later than the ovates from the Upper Loam, and they resemble forms which, in France, occur towards the end of the Last Interglacial. The fourth series is typical Levalloisian, very like the material from Crayford and Ebbsfleet. There are, however, a few pieces which are slightly rolled and patinated, suggesting the possibility of there being two groups of Levalloisian.

The position of these four groups in the deposit has been clearly stated by Allen Brown. (fn. 40) All of them come either from the gravel itself or from its surface; nothing was found in the unstratified solifluction deposit nor in the brickearth above. The absence of material from the brickearth is further confirmed by one of the gravel diggers, who supplied Galloway Rice with much of his material. Marsden, however, found a series of Levalloisian flakes, including one in situ 7-9 ft. from the surface, (fn. 41) which, on the evidence of both sections illustrated by Allen Brown-Pipkin's and Eastward's pits- suggest that this material is from the solifluction gravel above the river deposits; or, in the case of Pipkin's Pit, possibly even from the base of the brickearth. He does not say which of the deposits was being dug at the time when he found the flakes.

Allen Brown found part of the skeleton of a mammoth at Southall, about 13 ft. below the surface at about 88 ft. O.D. (fn. 42) The bones lay in what he describes as 'sandy loam between coarse stratified sandy gravel', the whole covered by about a foot of brickearth. It is not very clear from this description what the deposit was; in altitude it would belong to the same spread as that at West Drayton, in which case the mammoth, judging from the depth, would have been just above the river gravels. As the bones, however, were completely unabraded it is unlikely that they came from the river deposit. In addition to some hand-axes, there were, in association with the mammoth, flakes, which from Brown's description, are clearly Levalloisian.

Allen Brown's best-known sites are those at Creffield Road to the east of Ealing Common. (fn. 43) These pits were cut into a spread of gravel overlaid by brickearth. Above the brickearth and just below the present surface was a 'trail', an unsorted deposit suggesting a minor solifluction. At the base of the brickearth Allen Brown found what was clearly a Levalloisian working floor, with over four-hundred flakes in a very small area. The Levalloisian resembles that from Crayford, Ebbsfleet, Yiewsley, and Langley. From the gravels below came a series of small hand-axes, generally pear-shaped and often much rolled. Similar hand-axes have been found in other pits at Ealing and also in the pits at Hanwell, which are at approximately the same level. The date of these pearshaped hand-axes is difficult to establish since knowledge of the Acheulian of the Thames Valley is still very incomplete. Similar examples occur in the Boyn Hill gravels but they are by no means typical, and the large forms, such as occur at Furze Platt and Lent Rise, are rare at Ealing and Hanwell. The few hand-axes which have been recorded from the Taplow Terrace are also of this type and, like the specimens from Ealing, heavily rolled. The Ealing and Hanwell gravels have not produced hand-axes of the Upper Loam type nor the markedly twisted ovates from Dartford, the Globe Pit, Greenhithe, or Rickson's Pit. (fn. 44) This twisted form is not found in any of the Yiewsley pits. At both Ealing and Hanwell there are rolled Levalloisian flakes, which in the case of the former must be earlier than the undisturbed floor at Creffield Road. On the basis of these rolled Levalloisian flakes, several workers have claimed that there are two distinct terraces in this area, (fn. 45) the higher and older being that of Yiewsley, Dawley, and Hayes, where there is no rolled Levalloisian, and the later terrace with rolled Levalloisian in the gravel, this terrace being represented by Ealing and Hanwell. There is also a series of rolled Levalloisian flakes from Brentford in the Cooke Collection in the London Museum, but their exact locality is not known. From an archaeological viewpoint the separation of these terraces into two distinct types cannot be proved conclusively, but the general evidence suggests that Yiewsley and Ealing/Hanwell are not the same, and until the date of the unrolled hand-axes from Yiewsley and the pear-shaped hand-axes is established this problem cannot be settled.

In 1872 Lane-Fox published the results of his work at Churchfield, Acton, and Brown's Orchard, Turnham Green. (fn. 46) The former area consisted of several pits with a surface height of 75-83 ft. O.D., on the average about 20 ft. below the deposits at Ealing. The deposits consist of sands and gravels resting on London Clay; only in the section at Chaucer Road was there any suggestion of a covering brickearth, and here it contains seams of white sand. The hand-axes from these gravels are rolled, and resemble the pear-shaped specimens from Ealing and Hanwell. At Chaucer Road, however, unrolled flakes were found underneath the gravel in a thin seam of white, sandy clay which rests on the London Clay. Lane-Fox's material from Acton, now in the PittRivers Museum, Oxford, is not marked with the pits from which the individual pieces came, but the only unrolled flakes, clearly marked as coming from below the gravel, consist of typical Levalloisian flakes, similar to those from Creffield Road. If these are the flakes to which he refers in his report, (fn. 47) then clearly these gravels are later than those from Ealing, and this suggests that we are dealing with two distinct terraces, as Burchell has suggested.

In the Brown's Orchard pits, at a surface level of 24-29 ft. there were sands and gravels overlaid by brickearth, the whole resting on the London Clay. From the base of these gravels came remains of mammoth, hippopotamus, ox, horse, and deer.

It is clear that the archaeological pattern of Middlesex follows that of the adjoining areas very closely, although it is not always easy to make direct comparisons with deposits above and below stream. Even over a comparatively short distance, for example from Maidenhead to Gravesend, the terrace system is extremely complicated, largely because the terraces can be built up in two completely different ways. In the lower part of the river they are the result of eustatic change, an adjustment of the river profile to changes in sea-level. A high sea-level during warm interglacial conditions causes aggradation and terrace buildings; during cold or glacial periods the sea-level falls and the river degrades its bed, cutting down to a low level. The lowest point of this down-cutting is the bench on which the subsequent aggradation was built. In theory the benches and gravels should give a precise chronology, since cold and warm conditions are clearly indicated, but even in the lower Thames there are more benches than known cold phases.

Farther up-stream the problem is complicated by the fact that some of the terraces are the result of climatic as well as eustatic action. In the case of climatic terraces severe erosion fills the river with material which it is unable to sort, with the result that the river bed is aggraded, owing in this case to cold climatic conditions. In the middle and upper reaches of a river both climatic and eustatic terraces are formed, causing some confusion in the chronology. There is a further complication caused by lateral streams such as the Colne, Brent, and Lea, which are either cutting through the older Thames deposits or building terraces of their own, with altitudes related to their own profile and not to that of the Thames.

While it is generally conceded that it is the geological evidence which dates the archaeological material, in areas as confused as the middle reaches of the Thames the archaeological material is often the only means by which river deposits can be dated, although this is only possible if the archaeological sequence and its relative dating is known from more stable areas.

The major datum for the Thames from Maidenhead to Gravesend is the Lower Boyn Hill Terrace, which is well represented throughout the whole stretch of the river, although rather eroded in the Middlesex section. The archaeological material from Furze Platt and Lent Rise is the same as that from the Middle Gravels at Swanscombe, which is dated on the fauna and its altitude to the Great Interglacial, Mindel/Riss. Similar material also occurs in the 30-m. terrace of the Somme, with the same fauna. The Lower Gravels and the Lower Loam of Swanscombe, belonging to the earlier part of the same Interglacial, have no demonstrable counterpart, either in Middlesex or Buckinghamshire, and the Clactonian of the Lower Grave or the slightly later material from Clacton-on-Sea have not been found as a separate horizon, although Clactonian flakes are frequently found in deposits in Buckinghamshire and Middlesex, as, for example, at Yiewsley and Stoke Newington.

At Swanscombe the Upper Loam produced hand-axes of a later type than those from the Middle Gravels, and in some respects these resemble those from Hoxne, which, on the evidence of the flora, belong to the Great Interglacial. The evidence from the Globe Pit, Dartford, and Craylands Lane suggests, however, that these hand-axe forms continued into the initial phases of the erosion of the Boyn Hill deposits, since these three sites are in channels cut into the older gravels. That the erosion phase continued is shown by the presence of this type of hand-axe, derived and rolled in the deposits at Burnham, Iver, and the filling of the Ancient Channel between Reading (Berks.) and Henley (Oxon.). (fn. 48)

Following this major erosion phase, the archaeological and geological pattern is very confused, and our knowledge of the archaeological sequence is anything but precise. The erosional phase which appears to follow immediately after the final aggradation of the Great Interglacial is very probably due to the oncoming cold conditions of the Riss Glaciation, which, although in at least two parts, has left no tangible evidence by which its phases can be separated in the Thames basin.

In France the Last Interglacial was a period during which the later stages of the Acheulian, the Upper Acheulian, Micoquian, and Levalloisian industries developed. Up to the Second World War it was thought that the late Acheulian and the Levalloisian were two distinct industries, which if they ever came together did not do so until the later stages of the Last Interglacial. More recent work, however, has shown that such a separation is perhaps not so definite as was originally supposed, and that both the Upper Acheulian and the following Micoquian were both familiar with the Levalloisian technique, and that far from there being two clearly separated streams, there is in fact a very confused interchange of techniques and tool types. There are, in the Seine area, both an Upper Acheulian and Micoquian with lanceolate hand-axes, a flake element made with a Levalloisian technique, and a Levalloisian with cordiform hand-axes. (fn. 49)

In the Seine area the Micoquian and the Levalloisian with cordiform hand-axes occur at the very end of the Last Interglacial and are immediately succeeded by the deposits of the first stage of the Last Glaciation, Würm I, represented in northern France by a loess deposit. The idea put forward by Breuil and Koslowski in 1931-4 that the Levallois began in the Riss Glaciation does not seem to be supported by the evidence.

If the recent view expressed by French prehistorians, particularly Bordes, is accepted, (fn. 50) then the archaeological pattern in England following the Great Interglacial becomes very much clearer, and the apparent absence of an Upper Acheulian in Britain, with its place taken by the early Levalloisian, ceases to be a reality, and much of the rather ambiguous Acheulian material takes its place in the Last Interglacial. The Great Interglacial finished with the Acheulian of the Upper Loam, with a possible slight continuation into the oncoming glaciation, and with the end of the Riss erosional phase a new cycle begins, which, in the French terminology, would be classed as Upper Acheulian and Micoquian.

In France both the Upper Acheulian and the later Micoquian made use of the Levalloisian or prepared core technique, as it is better called, and traces of this technique are discernible as far back as the Middle Acheulian at Baker's Farm, and also at Rickson's Pit, both of which have produced rough prepared cores, and there is a suggestion of a similar technique, although not so clear, in the Middle Gravels at Swanscombe and possibly Dorney Wood (Bucks.). (fn. 51)

In the Seine deposits this continuation of Acheulian hand-axes and flakes from prepared cores is well established and continues throughout the Last Interglacial, where at the end it is succeeded by the typical Levalloisian industries either with or without small cordiform hand-axes.

In Britain the geological position of this Levalloisian with cordiform hand-axes is still unsettled. These hand-axes are an integral part of the Levalloisian of Baker's Hole, and similar hand-axes occur at Frindsbury, near Rochester (Kent). These forms are also present in the unrolled series from Yiewsley, but at this last site it is not easy to separate them from the lanceolate types, which on the French evidence, belong to the Micoquian or last phase of the Acheulian. Small oval hand-axes also occur in the lower brickearths at Crayford. A slightly later form has been found in the Gipping valley (Suff.), with the Levalloisian at Ebbsfleet, and also at Iver.

There is only one hand-axe in the Levalloisian from Creffield Road. (fn. 52) In the upper brickearth at Iver there is none, and the same applies to the material from the base of the gravels at Churchfield, Acton. This Levalloisian is, in France, later than the Levalloisian with hand-axes, and there is considerable evidence that this stage continued into the Last Glaciation, possibly even as late as the second cold phase, Würm II. If this dating is accepted, then it appears that the terrace at Yiewsley with the Late Acheulian on the surface belongs to the Last Interglacial, and that the solifluction deposits and the brickearth above belong to a stage of the Last Glaciation. It is also possible that the Coombe Rock which overlies the Levalloisian with hand-axes at Baker's Hole also belongs to the Last Glaciation.

The Middlesex material so far discussed has come from the terrace deposits along the southern part of the county. This treatment has been necessary on chronological grounds, since it is only because of its connexion with these terraces that the material can be dated. Several other localities have produced implements, but these are generally stray finds, many of which are listed by Vulliamy (fn. 53) and Worthington Smith.

The implements from Middlesex are well represented in local collections in the county. In addition to its own collection the British Museum has that of Allen Sturge; the Galloway Rice Collection is in the London Museum as are also the Cooke and Layton collections. The Sadler Collection is in the Gunnersbury Museum.