5

Palaeolithic Series and Pleistocene Sequences, 1900–1918

The Modern Archaeologist

At the Paris School of Anthropology, Professeur Capitan’s final lesson of the old century had been a review of the methods of prehistoric archaeology (Capitan 1899). For Capitan, competent modern prehistorians did not rely on a single method to answer their questions about the deep past, but employed a range of analytical, comparative, experimental and deductive methods of investigation, underpinned by careful and reasoned observation. Nor did they rely on any one strand of evidence or a standard mode of enquiry. Stratigraphy, palaeontology, botany and meteorology were essential prerequisites, providing the geological and environmental contexts for the study of material objects, which might also involve technology, typology, function, materials, aesthetics, mineralogy, experimentation, ethnography and sociology. Specialists were routinely consulted, their knowledge adding fine detail to enrich the general synthesis. The young science had certainly come a long way in 40 years, but Capitan warned that the existing frameworks were oversimplified and would not withstand the force of an increasingly complex data set. He urged that future frameworks should be treated as nothing more than works in progress, capable of responding to new facts, and should be constructed only on the most secure and most detailed of observations. The reports and artefact assemblages that had amassed since the 1860s could not always be taken at face value and needed careful individual assessment, because not every observer had been equally competent or impartial.

The need for an adequate and accurate framework was not the only moral of Capitan’s lesson. From the different strands of evidence Capitan believed it was possible to construct logical chains of reason that helped guide interpretation from the stone tool to the sociological. The study of the handaxe, for example, led Capitan (1899) to thoughts of hunting, in turn prompting the examination of the animals killed, their migration patterns and the various habitats they occupied. This led him to consider feeding customs and clothing, and from there climate and flora. By grouping artefacts by inferred function, Capitan thought he might understand their social function and identify the customs of different ethnic groups, synchronically as well as chronologically.

Capitan (1901) continued his campaign with a review of variation in handaxe form. The dominant (i.e. Mortillet’s) view was that different types of handaxe were all variations on the same tool, with various modifications to suit different purposes. The origin and persistence of this idea stemmed, Capitan thought, from the intellectual bias that primitive people could only have the simplest idea of a rudimentary tool. For Capitan, however, these modifications were just too extreme: the coup de poing class subsumed everything from a tiny discoid to a giant lanceolate. During the Chellean, handaxes had been generally coarse with sinuous edges and a thick butt, but even here four distinct types could be identified (Table 5.1). By the time of the Acheulean, handaxes had been perfected and further elaborated into eight types and two sub-types (Figure 5.1), La Micoque being seen as especially noteworthy for its series of small yet finely made lanceolate forms. This did not consider the many other varieties or ‘passage forms’, but for Capitan infinite subdivision was not the point. It was (as d’Acy had argued) that handaxes fell into a number of discrete and deliberate sub-categories, forming a group of implements, not just one. Furthermore, different sites were dominated by one type or another, although he noted that there was significant overlap and that all types could be encountered in a single assemblage. Libert (in Capitan 1901) wondered whether some of Capitan’s distinct types might not be regional variations created by isolated but contemporaneous populations. Central European archaeologists were of a different mind altogether: Professor Moriz Hoërnes of Vienna rejected the validity of the Chellean, Acheulean and Mousterian in Austria-Hungary, seeing them as one evolving line that was poorly represented in his study area (Hoërnes 1903).

Figure 5.1 Dig the new breed. Three of Capitan’s handaxe types (after Capitan 1901). Left to right: Acheulean ovale, lanceolate (from La Micoque) and disc. To illustrate his types, Capitan used several of the same images employed by d’Ault du Mesnil. Scale = 5cm.

Table 5.1 Capitan’s (1901) Chellean and Acheulean handaxe types.
Chellean	Acheulean
Well worked only at the tip, which may be elongated. Smaller forms may be drills	Lanceolates (large, small and tiny)
Pointed with a straight cutting edge along the tip and margins	Bevelled (tranchet) tipped
Forms with a lateral working edge	Lateral edge
Oval, approaching the Acheulean type	Twisted edge
	Notched edge
	Pointed ovate
	Regular ovate
	Discoid (some tiny)

Capitan’s intended readership was not the old guard whose ideas would probably never change, but the younger generation, those who had no personal knowledge of a time before 1859, and for whom deep history was another established fact. They sought answers to different questions and pushed interpretation in new and exciting directions that we have been dealing with ever since. It is to this generation and their ideas that the present chapter is devoted.

The British Palaeolithic (Finally) in European Context

Many British researchers would no doubt have recognised in themselves the qualities of Capitan’s competent modern prehistorian, and most would have shared his misgivings about Mortillet’s classifications. Franks’s successor as keeper at the British Museum, Charles Hercules Read (1857–1929), opined in the official Guide to the Antiquties of the Stone Age that the range of Palaeolithic implements was too poorly defined to allow them to be securely dated, and that form was in any case a poor guide to age (Read 1902). He also found it difficult to correlate Mortillet’s framework with the sequence the British Geological Survey had mapped for East Anglia (Skertchly 1877; Table 5.2). This showed three warm periods during which humans had occupied Britain, separated by periods of cold inhospitable conditions. But handaxes of St Acheul and Chelles types were common in all three warm periods, while High Lodge was among the oldest but contained scrapers of Le Moustier type. Read could detect “no chronological order such as the French scheme implied” (Read 1902, 9).

Table 5.2 Pleistocene sequence in East Anglia, as mapped by the BGS (after Read 1902, 9).
Period or Deposit	Climate	Archaeology
Neolithic Period	Postglacial
Hessle (Brown) Boulder Clay	Glacial
Late Palaeolithic	Interglacial	Handaxes in the Little Ouse (Santon Downham, Broomhill, Redhill)
Purple Boulder Clay	Glacial
Intermediate Palaeolithic	Interglacial	Handaxes in the old channel system at Brandon Field and Lakenheath. Hoxne
Chalky Boulder Clay	Glacial
Early Palaeolithic	Interglacial	Handaxes in High Level gravels at Botany Bay, Brandon. Handaxes and Scrapers at High Lodge
Lower Chalky Boulder Clay	Glacial

Most local British archaeologists likewise continued to ignore Mortillet. Discussing his discoveries at Savernake (Wiltshire), Edgar Willett (1901) noted that each assemblage had its own unique character, and that a practised-eye could identify a handaxe to its place of origin. The most distinctive feature of the Savernake ovate-assemblage, however, was its high-gloss surface, which Willett thought might have been a coating of silica, but which Henry Stopes bizarrely believed to have been caused by friction from passing worms (ibid.). Llewellyn Treacher’s (1859–1943) review of the rich Palaeolithic find-spots around Reading and Maid-enhead in the Middle Thames Valley similarly avoided Mortillet’s terms. Treacher (1896,

1904) noted the different dominant forms at sites such as Caversham, Groveland’s Pit (Reading) and Furze Platt (Maidenhead), but doubted that there was any evidence of progress. The handaxes from the highest and presumably oldest terrace at Caversham were considerably better made than those from Furze Platt, on a terrace some 40ft lower down the valley. That said, Treacher wondered whether the ability to produce the same implement in a more efficient manner, using fewer blows and using less time and less flint, might be seen as progress.

Progress of a different kind was made at Foxhall Road, Ipswich, where in 1902 Nina Layard (1853–1935) discovered handaxes in freshwater deposits overlying boulder clay, much resembling the situation at Hoxne (Evans et al. 1896) and Hitchin (Reid 1897). Miss Layard’s informal advisory team, which included Sir John Evans, botanist Sir Clement Reid (1853–1916), geologist Horace B. Woodward (1832–1921) and William Boyd Dawkins, was hardly one to promote correlation with the continental framework. Her finds were thus referred only to the Palaeolithic, by which she meant they belonged to the older drift subdivision, not the cave or reindeer period. They did encourage her, though, to undertake her own excavations, the first by a woman and which she conducted with a care unknown in other open-air Palaeolithic sites of the time (White and Plunkett 2004).

Miss Layard chose an unquarried area on the edge of Foxhall Road pit and spent two field seasons (1903, 1904) supervising labourers specially hired to dig for artefacts. The workmen were instructed to remove the sediments stratigraphically while Layard recorded the context, vertical depth and horizontal position of most finds, preserving a three-dimensional record of the site. Most of the finds were made in beds of silty-clay and red gravel, which sloped between 8ft and 12ft beneath the surface and represented the former banks of a river or pool. Ovate handaxes, many with twisted edges, came from the clay, whereas the gravels contained a wider diversity of forms including some classic points. She at first (probably correctly) believed them to be two separate horizons, but later concluded that the sharp rimmed ovates had cut their way down into the clay from the original gravel location. She warned against assuming that finds from different levels were of different ages, rather than just parts of an undulating surface. Among the finds were perfect miniature handaxes, children’s toys according to Miss Layard, plus a trio of beautifully made twisted cordate handaxes, found lying close together and so exactly similar in form and technique that she concluded they must be the work of the same hand (Layard 1904). She also emphasised the abundance of small tools, knives and simple scrapers (Layard 1906), but made no comment on whether the site should be classed as Acheulean or Mousterian. Layard was more interested in the types of activities to which each type may have been put, again using her own hands to infer types of hold and motion. Some forms had a waist (probably natural damage) which she suggested might have been intended for hafting (Layard 1903).

For a younger generation of British workers, however, the continued use of Lubbock’s 40-year-old division into drift and cave periods was “hardly conducive to clearness of thought” (Warren 1902, 97). The deposits within caves and rock shelters might contain evidence for later time periods than those found in the drift, but otherwise the archaeo-logical succession was the same in both settings (Warren 1902; Read 1902). Samuel Hazzledine Warren (1872–1958), who retired from the family business in 1903 to devote his time to geological and archaeological research, saw greater utility in Mortillet’s framework, with certain modifications (Warren 1902, 97). Warren thought the epochs were generally ill-defined, their ‘characteristic’ types not confined to one epoch but frequently found in younger and older periods. This did not mean that the epochs were contemporaneous but demonstrated the perils of definitively assigning any individual handaxe to a particular period. For Warren, the relative age of artefact types or series of types should be based on stratigraphy and the physical condition of the artefacts, arguing that rolled derived arte-facts at a site would always be older than fresh contemporary ones. He therefore proposed adding a stage before the Chellean-Acheulean represented by the rolled palaeoliths from the high-level plateau of Kent (not the eoliths, which he rejected, but the associated genuine handaxes, see Chapter 4). He also queried the validity of the Chellean, although conceded that there appeared to be a gap between the earliest Palaeolithic and the Acheulean, which the Chellean comfortably filled. He was equally uncertain whether all so-called Chellean assemblages were truly contemporary with the industry at the type-site at Chelles.

Figure 5.2 Miss Layard excavates. Photograph of the 1903–1904 excavations at Foxhall Road, showing (from left) Nina Layard, Arthur Barker, Clement Reid and F. Fox (after White and Plunkett 2004, courtesy Suffolk Records Office). Despite the poor working conditions, Layard recorded the position of most artefacts according to stratigraphy and their 3D-coordinates. The main find horizons were on top of the light grey clay at the base of the section and the overlying red gravel.

Warren’s solution to these problems was to introduce a system of sequence dates, based on the one devised by Flinders Petrie (1899). To do this properly required a secure sequence with actual super-positioning, plus samples large enough to allow the relative proportions of different types and patterns of development/degradation over space and time to be meaningfully established (ibid.). Despite lacking in all these requirements, Warren defined a series of six Palaeolithic (P) sequence dates, ranging between P. 30 and P. 90, each point on the scale aligned with the Mortillean framework (Table 5.3). The top and bottom of the range were left blank to accommodate later discoveries, and the scale had the potential for finer subdivision.

Table 5.3 Warren’s (1902) sequence dates for the Palaeolithic correlated with Mortillet’s epochs. The values shown are the basal values for each epoch. Depending on the proportion and form of different types, sites might be more or less like those above or below and would be assigned a sequence date to reflect this.
Period	Epoch	Type Station	Sequence Date
Later Palaeolithic	Magdalenien	Rock shelter of La Madeleine, Dordogne	P. 90
	Solutrean	Le Clos du Chanier, Solutré, Saône-et-Loire	P. 80
Middle Palaeolithic	Mousterian	Cave of Moustier, Dordogne	P. 70
	Acheulean	Contemporary implements in the drift of St Acheul, Somme	P. 60
Earlier Palaeolithic	?Chellean	The earlier drift of Chelles (Seine-et-Marne)	P. 50
	Earliest known Palaeolithic	Drifted implements of the Plateau Drift of Kent	P. 30 to P. 40

To demonstrate the scheme, Warren reclassified Smith’s sequence at Stoke Newington, assigning the material from the main floor to P. 65–P. 75, the 12ft floor to P. 55–P. 60 and the 30ft horizon to P. 30–40, with the main floor straddling the Acheulean-Mousterian, thus the relatively high frequencies of handaxes and scrapers (High Lodge was assigned to the same sequence date, P. 65–P. 75). In the high terrace at St Acheul, Warren found a derived series (P. 30–P. 40) and three primary context series in the gravel (P. 40–P. 45), middle beds (P. 60) and the brickearth (P. 65–P. 75). The lower terrace at Montières contained a P. 60–P. 80 in the lower drifts and P. 75–P. 90 in the brickearth. Mixed assemblages could readily be divided by their condition and the different components assigned their own sequence date.

Warren clearly felt the need for a standard frame of reference and recognised the value of the familiar French terms. His sequence dates were intended to help fine-tune an existing framework, communicating subtle differences and similarities in the character of each industry and its place in the overall scheme. But he lacked the data to do this on anything but a coarse scale and his deliberately vague sequence dates simply substituted a formal name with a numerical code. It had the air of mathematical precision, but the numbers were as arbitrary as the names and added little clarity or detail to the meaning. Unsurprisingly, the system was never widely adopted, but it further exposed the urgent need for a new synthesis and Warren’s certain belief that a local archaeologist was qualified to provide it.

Palaeontologist Martin Hinton (1883–1961) and malacologist Alfred Santer Kennard (1870–1948) similarly found the British Palaeolithic to be in “a less than satisfactory state” (Hinton and Kennard 1905, 76). Collectors shouldered much of the blame, for failing to gather critical contextual information, but British archaeologists had done much to add confusion by the continued use of vague obsolete terminology and a general disinclination to use the classifications of Mortillet and Rutot fashionable on the continent. They demanded a more stratigraphical approach, believing that once the proper relative order of the deposits had been settled, the succession of Palaeolithic industries would all fall into place. Only Prestwich (1889) and Warren (1902) were praised for their bold attempts at dividing the British Palaeolithic.

Hinton and Kennard focussed their study on the Lower Thames Valley, starting from the datum provided by the Lenham Beds, which represented the final marine inundation of the Tertiary. The geomorphological model behind their work was fairly robust. The Lower Thames Valley had been affected by uplift of the Weald, which saw intermittent periods of crustal elevation forming new base levels for the river, which then down-cut and began to fill its new valley. From this they identified a series of eight ‘base-levels’ to which the river had cut, which they then correlated with the deposits of the Upper Thames Valley (Monkton 1904) and with the continental archaeological sequences (Table 5.4). They considered it a complete scale of progress with little missing and no breaks. The eoliths graded into the transitionals, which passed into older and younger handaxes. The Mousterian implements were ‘obviously’ a development from handaxes, and in turn gave rise to the Solutrean.

Table 5.4 Hinton and Kennard s 1905 correlation of the Thames stratigraphy, artefact type and European cultural phase.
Lower Thames Geological Unit	Upper Thames Geological Units	Artefacts	European Correlation
Plateau Gravel	Upper Hale gravel	Eoliths	Rutot’s Ruetellienne and Reutelo-Mesvinianne
Hill Gravel	Goring Heath	Transitional Forms	Transitional from Rutot’s Mesvinian to Chellean
1st Terrace (Dartford Heath)	100ft Terrace	Older Palaeoliths – ovoids	Acheulean
2nd Terrace (Swanscombe)	100ft Terrace	Palaeoliths from the 100ft terrace – points predominate	Acheulean
3rd Terrace (Crayford)	Plain gravel Crayford brickearth	Palaeoliths of the brickearth – pointed and ovate handaxes, long thin flakes (Levallois)	Mousterian
4th Terrace	Not recognised	Palaeoliths of the fourth terrace, rare	?Solutrean
Buried Channel	Not recognised	Rock Shelter implements at Oldbury?	?Solutrean
Holocene Alluvium	Modern alluvium	Uxbridge Type, Cissbury Type. Neoliths	Magdalenien Neolithic

As members of Harrison’s Ightham Circle (Preece 1990) Hinton and Kennard were pre-disposed to believe in the much-debated eoliths (as did Rutot and Mortillet), but their invention of a new high terrace at Dartford Heath was equally controversial, as several other workers had already correlated these gravels with the 100ft terrace at the relatively new site at Swanscombe (Chandler and Leach 1907, 1911, 1912). Furthermore, as Hinton and Kennard (1905, 93) acknowledged, the sparse artefacts from the Dartford Heath Gravel were similar to those at Swanscombe and the two were perhaps best grouped together. At Swanscombe, artefacts fell into two Acheulean series – an abundant unpatinated lower series and a sparse white-patinated upper series – but they could detect no indisputable Chellean presence in the Thames, as all handaxes were associated with scrapers. Their positioning of the Solutrean and Magdalenian, furthermore, were based only on what should occur next in the sequence, rather than on the character of the collections, which were practically non-existent. Ultimately the best that can be said of the Hinton and Kennard’s framework is that was a good idea which made a genuine attempt to drag the British Palaeolithic into line with France and the rest of Europe, but it was premature, poorly executed and failed to make good use of either authors’ specialism.

The Alpine Glacial Sequence 1901–1909

James Geikie’s idea of multiple Quaternary glacials and interglacials (Chapter 4) had never gone away. In the much-revised third edition of The Great Ice Age, Geikie (1894) had summarised the expansive evidence for glacial phenomena amassed by workers across the globe, from which he proposed that Europe had experienced six glacial periods and five interglacial periods during the Quaternary. In his reading of the faunal, geological and archaeological evidence humans had been absent from Europe during the oldest interglacial, only appearing, in the form of the Chellean, during the second interglacial, along with the southern elephant fauna. The Acheulean belonged to a later part of this period. The Palaeolithic industries most frequently found in caves probably belonged to more than one period, parts of the third glacial, the third interglacial and during the formation of younger loëss, while the Neolithic was placed towards the end of the fourth interglacial and not during the current Holocene era. In northern Europe, however, there was no evidence that humans returned after the end of the third glacial. The problem for Geikie was that British archaeologists found it more convenient to ignore him, or at best pay his ideas lip-service before retreating behind the less controversial monoglacialist viewpoint.

By contrast, the evidence for multiple glacials and interglacials identified by geologists Albrecht Penck (1858–1945) and Eduard Brückner (1862–1927) in the Alps was well received by German and French archaeologists. The pair had spent years studying the glaciofluvial gravels and glacial moraines of the range, Penck tackling the northern and northeastern Alps as far as Vienna (e.g. Penck 1882), and Brückner those of the western Alps in Switzerland, their results jointly published in three volumes between 1901 and 1909 (Penck and Brückner 1901–1909). They identified four distinct periods of glacial expansion, which they named from oldest to youngest the Günz, Mindel, Riss and Würm after tributaries of the Danube, each separated by a prolonged period of warm interglacial conditions (named after the glacials either side of it, Riss-Würm being the last: Table 5.5). They also identified three periods of wind-blown loëss formation. In terms of vegetation, each cold-warm cycle had seen arctic tundra give way to interglacial forests, then steppe, before returning to tundra.

Table 5.5 Alpine correlations: table showing correlations of the Alpine Glacial Sequence with Mortillet’s cultural framework and Pleistocene fauna.
Period	Alpine System	Penck1903; Obermaier 1905a	Obermaier 1905b	Obermaier 1909; Boule 1908b
Current		Neolithic	Neolithic	Azylian. Neolithic
Postglacial		Magdalenian	Magdalenian	b) Magdaleniana) Solutrean
4th Glacial	Würm		Magdalenian	Mousterian
3rd Interglacial	Riss-Würm	b) Solutrean, steppic faunaa) Mousterian, warm fauna	b) Solutrean, warm fauna then cold steppic faunaa) Mousterian, warm fauna	b) Acheulean/early Mousterian, steppic faunaChellean, warm fauna
3rd Glacial	Riss	Mousterian, cold fauna	Mousterian, cold fauna
2nd Interglacial	Mindel-Riss	Chellean, warm fauna	c) Phase of La Micoqueb) Acheulean, cool faunaa) Chellean, warm fauna
2nd Glacial	Mindel
1st Interglacial	Günz-Mindel
1st Glacial	Günz

Rueing the fact that archaeology was moving away from geology, Penck (1903) attempted to correlate Mortillet’s system with the expanded glacial chronology (Table 5.5). Penck’s study area contained a number of Magdalenian sites within the terminal moraine of the (last) Würm glaciation, and he safely assigned this epoch to the postglacial. As Mousterian implements were known only beyond the limits of the second, Rissian glaciation, Penck concluded that the Mousterian and associated cold-adapted fauna was synchronous with that event, although he also recognised a Mousterian with warm fauna at Villafranche, showing that the Mousterian survived into the beginning of the succeeding Riss-Würm interglacial. The steppic fauna and Solutrean industries belonged to the last part of the same Riss-Würm period. The Chellean was absent from Penck’s study area, but as it had to come before the Mousterian, he correlated it (or its warm fauna) with the second, Mindel-Riss interglacial.

In 1905, Baverian archaeologist Hugo Obermaier (1877–1946) enthusiastically reported Penck’s results to a French-speaking audience, declaring the geology and archaeology to be in accord (Obermaier 1905a). Obermaier had been a student and field-assistant of Penck at Vienna, where he had only recently completed his doctoral thesis on the Palaeolithic of Central Europe. He was now spending a year in France working with Capitan, Emile Cartailhac and rising French star Henri Breuil. Breuil was a priest and scientist whose interest in the Palaeolithic had first been inspired by a meeting with Geoffroy d’Ault de Mesnil during a teenage holiday on the Somme in 1894–1895, and subsequently encouraged by his tutor, Abbé Guibert, at the Jesuit seminary at Issy-les-Moulineaux (Brodrick 1963). After working on Edouard Piette’s (1827–1906) excavations at the cave site at Brassempouy in the late 1890s and having studied the old man’s personal collection of Magdelenian portable-art, Breuil decided on the path he wished his career to take. Gaining a dispensation from his order to pursue his scientific research rather than his religious orders, he rapidly established himself in the field of Upper Palaeolithic archaeology (ibid.), conducting pioneering work on cave art and winning victory in the ‘Battle for the Aurignacian’, which resulted in the reintroduction of that long defunct industry (Breuil 1906, 1907).

Obermaier’s time in Paris quickly gave him a different perspective on the Palaeolithic period, leading him to slightly modify Penck’s original scheme to include the Acheulean and the Phase of La Micoque (Obermaier 1905b, see Table 5.5), the latter based on the primitive non-handaxe industry recently discovered beneath the final Acheulean level at the eponymous site (Déchelette 1908b). Just a year later, after studying sequences at Villafranche and in the Pyrenees, where the four terraces of the Garrone provided a bridge between the Alpine glacial extensions and the non-glaciated regions of northern and southern France (Obermaier 1906b), Obermaier concluded that Penck’s original correlations had made everything too old. By the time his first paper had been translated into English (Obermaier 1906a) he had developed a shorter chronology, with humans only present during the Riss-Würm interglacial, Würm glacial and postglacial periods (Obermaier 1906b). Boule (1908b), who only accepted the evidence for three glaciations (rejecting Günz), independently reached similar conclusions.

When Penck (1908) saw no reason to alter his earlier conclusions Obermaier (1908b, 1909) provided detailed rebuttals. As there was no Chellean or Solutrean in the Alps, Penck’s correlations were little more than theoretical postulation, determined solely by the placement of the Magdalenian and Mousterian. Obermaier’s observations led him to place the Mousterian exclusively in the Würm Glacial. There was no warm Mousterian in the Riss-Würm interglacial, the apparently warm fauna from Villafranche being, in Obermaier’s opinion, mixed. Obermaier also refuted his own earlier correlation (1905b) of the Acheulean with the later (cooling) part of the Mindel-Riss Interglacial, which Penck had since adopted. Obermaier (1908b, 1909) placed much emphasis on a single Upper Acheulean handaxe from Challes de Bohan (Ain), which had been found in situ in red loëss by Adrien Arcelin (1901). This deposit lay above a glacial gravel of the Rhône glacier that Penck had firmly attributed to the Riss Glacial, meaning that the Acheulean must have post-dated the Riss, probably belonging to a later part of the Riss-Würm interglacial with the missing Chellean covering the earlier part of the same period. The sequence Obermaier constructed had many parallels with d’ Ault du Mesnil’s (1889) earlier correlations for the Somme. He agreed entirely with the glacial chronology and recognised its value as a framework for understanding the human past, but he thought Penck’s correlations with archaeological periods were ‘shipwrecked’ (1909, 517).

So began the never-ending task of accurately correlating the Palaeolithic record with an independently constructed, continental-scale Quaternary framework, one presumed to give both chronological and environmental security. It was a far more complex Pleistocene framework onto which archaeologists now sought to fit a Palaeolithic cultural sequence that was itself larger and more diverse. It will probably never be right.

Victor Commont in the Somme Valley 1904–1917

Victor Commont (1866–1918) was a science teacher at the École normale in Amiens, a post he had held since 1895. His workplace and his home in Avenue d’Edimbourg were a few minutes’ walk from the famous St Acheul quarries and brickyards, allowing Commont to monitor them on an almost daily basis. He also maintained a network of correspondents to pass on information from quarries in other parts of the Somme Valley and would, less frequently, visit other locations such as Abbeville (Commont 1910b; Tuffreau 2009). Between 1908 and 1914 he followed the construction of the canal du Nord along the tributary rivers Verse and Mève, recording the stratigraphy and identifying a large number of Mousterian stations, with refitting Levallois knapping episodes and characteristic Mousterian handaxes, mostly small cordates and triangulates (Commont 1916).

Commont complained that much confusion surrounded the sequence at St Acheul, a situation that had been created by a lack of rigour from the beginning (Commont 1906). To understand the true sequence, he argued, one had to look at the details with great precision. Through such rigorous geological observation, Commont (1908, 1909a) initially identified three terraces in the Somme, based on their height above the lowest river deposits, eventually adding a fourth, higher terrace that was visible in only a few isolated patches (1910a, 1911a, 1912a, 1912b, 1912d, 1913):

1. The low or 10m terrace

2. The 2nd or 30m terrace

3. The 3rd or 40m terrace

4. The 4th or 55m terrace

By precisely measuring the levels of the different fluvial deposits between Daours in the east and the estuary at St Valery, Commont was able to map the Somme terraces in long profile, showing that they dipped downwards towards La Manche (Commont 1910a, 1910b). This meant that the Low Terrace disappeared beneath the floodplain around Eaucourt, leaving only three terraces exposed at Abbeville (1910a, 1910b; Figure 5.3). From his geological and palaeontological data, Commont (1914) created a palaeo-geographical picture of northern France, at times joined to Britain when the Somme, Seine, Thames and Rhine flowed into an enormous English Channel river (cf. Hull 1912), at others flooded by marine incursions. The Channel finally flooded at the end of the Palaeolithic. When occupied by Palaeolithic humans, the sites had been rich riperian oases, well-stocked with herds of prey animals: the living had been easy. He also built on Ladrière’s (1891) work on the brickearths (ergeron) that spread “like a mantle” over the valley sides, identifying them as a series of loëss deposits (which he suspected was a mixture of slope-washed and wind-blown sediments) and recording the associated faunal and artefact assemblages (1912a, 1912b). Within this upper sequence he identified the limon fendrille (a soil of the last interglacial) as a major stratigraphical break.

Figure 5.3 The Somme terraces 1910. Victor Commont’s mapping of the Somme terraces in long profile from Daours to the mouth at St Valery (after Commont 1910c).

These detailed observations led Commont to radically modify Mortillet’s archaeological framework. In December 1905, Commont noticed that work in Tellier’s Quarry had started to penetrate a rich seam of flint artefacts, located at the base of some red-brown sands, 8.2m below the surface of the Second Terrace (Layer H in Figure 5.4). He visited the site every day from 2 December until 30 December, when the deposits were exhausted, collecting over 5,000 artefacts from an in situ flint knapping floor, characterised by the manufacture of Chellean handaxes, scrapers, cores and flakes (Commont 1906–7). This was the famous Atelier Commont, or ‘Commont’s Workshop’.

Figure 5.4 Atelier Commont. Photograph of Tellier’s Pit, St Acheul (after Commont 1909a). A) brickearth; B) coarse ergeron (loëss); C) small gravel with industry; B1) fine ergeron (loëss), very sandy; C2) gravel or bed of white sand with industry; D) clean red silt; E) yellow-grey silt with calcareous nodules; F) loose sand, yellow or white, with zones of manganese at the top; H) brown sand, very compact. The workshop is in this layer, at the foot of the ladder; L) lower gravel, well developed to left; to the right is a chalk boss.

Through his analysis of the material on the workshop floor (Figure 5.5), Commont distinguished two types of flakes, one resulting from handaxe manufacture, the other from core reduction. Chellean (and Acheulean) cores lacked a definite method and the nuclei were irregular in form. After removing each flake, the knapper would have had to examine the core and choose the next most suitable face. The resultant flakes were thus irregular in form but were often short and thick. Handaxe manufacturing flakes were of all sizes yet tended to be largest at the beginning of the reduction process when the flint was being prepared, and smallest at the end, when the knapper was concentrating on making the piece symmetrical and sharp. Many handaxe manufacturing blows were delivered obliquely, often producing thin, delicate flakes. Above the workshop, at the top of the red sands, was another layer of flints characterised by better made handaxes. This was the true Acheulean.

Figure 5.5 Chellean artefacts from Commont’s Workshop, Tellier’s Quarry, St Acheul, including a ficron handaxe (in the French sense), two amygdaloid handaxes and flake tools (after Commont 1906–1907). Scale = 5cm.

In the adjacent quarries operated by messieurs Bultel and Fréville, however, Commont (1906, 1906–7) identified two separate Acheulean levels: an older (ancien) or lower (inférieur) Acheulean from the middle gravels (F) and coarse sand (E) characterised by red-stained limande (dab-fish shaped) handaxes, and an upper (supérieur) Acheulean layer from the red silts (D) containing fine elongated pointed handaxes (lanceolates) with a lustrous white patina. The limandes of the older Acheulean were well-made, thin in section and worked all round, the tip often beveled. The industry included tools that were transitional between the Chellean coup de poing and the Acheulean handaxe. The elegant points in the Upper Acheulean, on the other hand, marked the apogee of the form, the delicate working and slender points singling them out as true art. Scrapers, grinders, points, blades and flakes were common in both, just as they were in the Chellean. Thus, for Commont, it was not the presence of general tool types such as scrapers and flake tools that distinguished the different periods, but the specific types of handaxe they contained, which improved in form and technique over time. Handaxe shape variation was therefore a valuable chronological tool (Figure 5.6). Scrapers and flake tools were similarly increasingly refined over the course of the handaxe-making epochs.

Figure 5.6 Subdividing the Acheulean. Handaxes from Fréville’s Quarry illustrating different phases of the Acheulean. Bottom: transitional form (left), Chellean coup de poing (middle), lower Acheulean limandes (right). Top: upper Acheulean handaxes described by Commont as lanceolate (after Commont 1906, 1909a). Scale = 5cm.

In the upper gravels at St Acheul, Commont (1906–1907) found only a Mousterian industry, in which Levallois types became most frequent (Figure 5.7). He initially believed it to be a modification of the Acheulean. As attractive as some handaxes were, Commont thought that aesthetics was probably not the primary reason why handaxes looked as they did. It was more likely that the knapper had a clear purpose in mind, needing something that was sharp, easy to hold and good for cutting up carcasses and skins. When they realised that a large blade-like flake would do the same job, handaxes were slowly abandoned. Economising with flint resources might have been an additional factor, because handaxe manufacture was costly and error prone, the flakes often oblique and too delicate to use. Mousterian knap-ping was more efficient, and highly suitable at a time when flint deposits were becoming concealed under layers of brickearth. Consequently, Mousterian handaxes were not as well-crafted as Late Acheulean ones.

Figure 5.7 The Mousterian of Northern France. Top panel: Levallois cores and flakes, with examples of facetted butts. Refitting examples such as that shown top right was positive proof that the ‘disc-cores’ (as Commont continued to call them) were the parent blocks of Levallois flakes. Lower panel: scrapers, points, notches and knives/cutters (after Commont 1909a). Scale = 5cm.

In February 1906 Commont made another enviable discovery at 54 Rue de Cagny, St Acheul, where groundwork was underway for a new house (Commont 1908). This was close to the old Tattegrain Quarry where Rigollot, Prestwich and Gaudry had collected material from the gravels of the Second Terrace. During the next three months Commont observed the workings, recorded the geology and collected every artefact discovered from an area covering some 1,752m² (Commont 1908). In contrast to the earlier reporters, Commont found that handaxes were not restricted to the lower gravel but also occurred in the red sand higher up the section. He assumed that these beds had not been sieved for commercial purposes when the earlier reports had been written, and that any finds had therefore gone unnoticed, although somebody who had known the area in the 1860s thought that the workmen attributed them to the lower units to get a better price. From a total sample of 540 handaxes and 450 other objects, Commont identified four series (Table 5.6 and Figure 5.8).

Figure 5.8 Commont’s three series from Rue de Cagny. Top: pre-Chellean handaxes from the basal lower gravel; Middle: Chellean handaxes and flake tools (Commont emphasised their presence) from the top of the lower gravel; Bottom: lower Acheulean handaxes, one with a twisted edge, from the red-brown silts (after Commont 1908). Scale = 5cm.

Table 5.6 Commont’s first (1906, 1909a, 1909b, 1909c) Palaeolithic framework for the Somme.
Industry	Stratigraphy	Characteristic Tools
Mousterian	Loëss, horizons C and C1Leval	Levallois, scrapers, points. Handaxes rare and usually heavy
Pre-Mousterian	2nd Terrace, base of loëss	Small La Micoque-type handaxes
Upper Acheulean	2nd Terrace: red silts	Finely made lanceolates
Older (ancien) Acheulean	2nd Terrace: middle gravels and base of red sands	Well-made ovates with an edge all-round, sometimes twisted
Chcllcan	2nd Terrace: upper part of lower gravel	Ficron and thick ovate handaxes
Pre-Chellean	3rd Terrace: lower gravel at Leclercq’s Quarry	Crude thick handaxes
	2nd Terrace: base of lower gravels

By combining the evidence from different localities around St Acheul, Commont devised a comprehensive archaeological sequence for the Somme Valley that subdivided Mortillet’s epochs based not on the presence/absence of flake tools, but on the form of the handaxes each contained (Table 5.7). The Mousterian was further distinguished by the frequent use of the Levallois method of core working which had not been employed during the Acheulean. In fact, for Commont, Levallois was the defining feature of the Mousterian, and the products of Mousterian and Acheulean core working were so distinctive that a single example could determine the age of the layer from which it came (Commont 1909b).

Table 5.7 Four series of handaxes from Rue de Cagny (after Commont 1908).
First Series	From the base of Lower Gravels (L). This series was primitive and characterised by roughly worked handaxes (about 30, many shaped at only one end), choppers and edge-worn flakes. Commont assigned it to a Pre-Chellean phase.
Second Series	From the upper part of the lower sands and gravels (L). This series was unpatinated, sharp, and characterised by pointed (French) ficron handaxes with thick cortical butts. Numerous flake tools and scrapers, including some bifacial scrapers made on flakes, but these were often outnumbered by unretouched flakes. Commont assigned it to the Chellean.
Third Series	From the red sand (sable roux/brun) (D). This series included 300 handaxes, plus a smaller number offtake tools. Ficrons had disappeared and ovate forms dominated, 116 of them showing deliberately twisted edges. These may have taken the role elsewhere served by small tools. Commont assigned it to the Acheulean.
Fourth Series	From the cailloutis (gravel parting) in the red silt, at the base of the brickearth (C). Handaxes disappeared, elongated (Levallois) flakes appear. This was the Mousterian.

Commont’s definition of Levallois included the flat pyramidal forms that had previously been considered discs, and those designed to produce more laminar products from both ends. Levallois could produce numerous predetermined flakes, each serving the purpose of the handaxe (which therefore gradually fell out of use), and Commont thought that he could detect an evolution from more massive to more delicate Levallois forms over time (Com-mont 1909b, 1913). The character of the bulbs and platforms were identified as useful discriminatory attributes for identifying Levallois methods of production (1909b, Figure 5.7).

Commont found essentially the same sequence of sediments, fauna and artefacts at Abbeville and Montières (Commont 1906–7, 1909a, 1910b, 1910c, Tables 5.8 to 5.10), although the latter provided evidence of still further stages. The Montières pits were situated at altitudes between 73m to 22m O.D., corresponding to the First and Second Terraces at St Acheul (1909a, 56; Figure 5.9). At the Buhant, Muchembled and Tattegrain pits, at ~25m O.D. (~10m above the Somme), he identified a Chellean industry in the lower gravel, with two Acheulean industries from gravel partings above and beneath the white silts (Pipe earth, G in table 5.8). Handaxes were rare in the Acheulean layers, but Commont compared them to the upper and lower Acheulean respectively. The Chellean handaxes from the bottom of the lower gravel were characteristically coarse, but those found higher up this bed were more refined, triangular and pointed. These had no parallel at St Acheul, and Commont doubted they could be contemporary with the limande-dominated lower Acheulean found at the top of the lower gravel at the type-site. As the two sites were only 6km apart he saw no reason to suspect that two tribes could have been isolated over such a short distance (Commont 1910c). They might belong to a different period, have served a special function around low-lying ponds or found their way into the river from settlements higher up the riverbanks (dropped from boats, perhaps), but how could one tell?, Commont asked (ibid.).

Figure 5.9 St Acheul and Montières. Commont’s composite section through the middle and low terrace deposits at St Acheul and Montières (after Commont 1909a). For description of sediments and their contents, see Tables 5.8 and 5.9.

Table 5.8 Commont’s sequence for the Montières Low Terrace (after Common: 1909a, 1910c).
a	Black Marsh Earth	Polished Axes of Grey Flint	Robenhausen
a 1	Grey marsh earth with freshwater shells	Small cow, small horse, deer, boar Tranchet axes, blades and end-scrapers on yellow flint	Campignienne
A	Red loamy silts, used as brickearth	Blades and end-scrapers on blades, patinated blue	Upper Palaeolithic
B	Ergeron (loëss)	Mammoth, horse, large bovid, horse	No industry
C1 and C2	Gravel lenses separated by grey-black silt	Mammoth, horse, large bovid, horsea) white patinated flakes, sharp, finely worked bifacialb) long, thick blades on black or brown flint, anticipates Upper Palaeolithic	a) Mousterianb) Lower Mousterian blade industry (originally thought to be early UP)
C3	Gravels over white silts	Well-made triangular handaxes, sharp with blue patina	Upper Acheulean
G	White silts (pipe earth)	Horse, large bovid, mammoth, giant deer Little archaeology, a few triangular handaxes
I	Red Gravel	Triangular handaxes, sharp, made on large thin slabs	Older (ancieri) Acheulean
L	White Gravel	Triangular and pointed handaxes, crude at the base, becoming increasingly refined through time	Advanced Chellean Chellean

Table 5.9 Commont’s (1909a) combined sequence for the Second Terrace deposits in the contiguous quarries of Messieurs Bultel and Tellier at St Acheul.
Bed	Description	Archaeology	Industry
A	Brickearth
B	Coarse ergeron (= loëss)
C	Small gravel	Levallois cores and retouched tools	Mousterian
B1	Fine sandy ergeron
C2	Gravel lens or white sand	Triangular handaxes and flake tools	Micoque-like (Tellier’s)
D	Red silts	Lanceolate handaxes	Upper Acheulean (Bultel’s)
E	Grey-yellow silts with calcareous nodules
F	Marbled yellow-white sands, with zones of manganese	Limande handaxes	ancien Acheulean
H	Red-brown silts	Commont’s Workshop	Chellean (Tellier’s)
K	White sands (in hollows)
L	Basal gravels	Sparse crude handaxes and flake tools	Chellean

Table 5.10 Commont’s correlations of Quaternary deposits and archaeology at St Acheul, Montieres and Abbeville. The Acheulean is somewhat disingenuously absent from the 30m Terrace (after Commont 1910b).
Lowest Fluvial Beds (asl)
	Saint-Acheul	Montierès	Abbeville
	13m	10m	-15m
a) Quaternary deposits: formed after the last downcutting. Lower gravels, clay, recent loëss (= Upper Quaternary silts of Ladrière = brickearth) Cold fauna: Elephas (Mammutkus) primigenius, woolly rhinoceros, small horse and reindeer Industries: different facies of the Mousterian, Aurignacian, Solutrean and Magdalenian b) Filling deposits: peat, tuffs and recent silts (limon de lavage) Temperate fauna: red deer, Bos longifrons, dog of the peat bogs, beaver Industries: Neolithic, Bronze Age, Iron Age
10m Terrace (Above Lowest Fluvial Bed)
	Saint-Acheul Station Quartier de la Gare La Vallée	Montières Boulevard de Montières St Roch	Abbeville Under the peat
	22–28m asl	20–25m asl	-5 to -1m asl
Deposits: gravels, sands, clays, upper part of the Middle Quaternary silts: red silts and soft silt with black spots = older loess Fauna: Elephas antiquus, Elephas primigenius, hippopotamus, red deer, lion Industries: Advanced Chellean, Early Acheulean, Upper Acheulean
30m Terrace
	Saint-Acheul Ancient Extractions and Tellier Quarry	Montières Route de Savuese, Muchembled Quarry	Abbeville Menchecourt, Mautort L’Heure
	40–45m asl	36–42m asl	10–17m asl
Deposits: gravels, sands, clays, lower part of the middle silts (limon panache), plateau silts (ancient loëss) Fauna: Elephas antiquus, ancestor of Elephas primigenius, hippopotamus, large horse. Level with Corbicula fluminalis and B elgrandia marginata. Industries: typical Chellean
40m Terrace
	Saint Acheul Route de Cagny (Cemetery)	Montières Route de Saveuse Quarries at 55m	Abbeville Moulin-Quignon, Champs du Mars, Saint-Gilles, Mareuil
	50–58m asl	55m asl	27–34m asl
Deposits: a) marl, gravels; b) fiuviatile sands, silts of the high plateau Fauna: Elaphas (Mammuthus) trogontherii, Elephas antiquus, hippopotamus, Merck’s rhinoceros, Etruscan rhinoceros, horse, Machairodus (scimitar-toothed cat) Industries: a) nothing b) Pre-Chellean
55m Terrace
	Saint Acheul Route de Cottenchy	Montières Ferme de Grâce	Abbeville Caubert
	60–6 8m asl	65m asl	40m asl
Deposits: gravels, possibly PlioceneFauna: noneIndustries: none

As Commont could not physically access the low terrace at St Acheul, which was underneath the railway station district, he could not test these ideas at the type-site, but in the gravels at Thennes, 10 miles upstream of Amiens, Commont again saw triangular handaxes above a classic Chellean industry. Both had the same ancient warm fauna. On this basis he concluded that the Montières and Thennes material represented an Advanced Chellean (évoluée), that had evolved in situ from the typical Chellean (Commont 1910a, 1910b, 1910c, 1911b). Montières also yielded a peculiar blade industry (laminar Levallois) that Commont had assumed to be Upper Palaeolithic (pre-solutrean) until it was found to lie underneath the Mousterian (Commont 1910c). He concluded it must represent an earlier facies of the Mousterian, with the associated fauna, which included hippopotamus and straight-tusked elephant, suggesting that the climate had been warm (Commont 1909a, 1912c, 1913). He noted that other ‘warm’ Mousterian sites could be found at Taubach and Krapina.

Commont had always found it paradoxical that the type-site for the oldest epoch at Chellessur-Marne lay on the lowest terrace, because geologists would normally expect this to be the youngest (Commont 1910c). At Montières he faced a similar problem but concluded from the fauna and handaxe forms that the river must have down cut to practically its modern level prior to the deposition of the higher gravels at St Acheul. In later publications (Commont 1914), the Acheulean was removed from the lower terrace at Montières, the assemblages above and beneath the pipe earth being re-assigned to the Advanced Chellean. This was now directly overlain by the Mousterian, with Commont explaining the newly problematic absence of an Acheulean as the result of the waters of the Somme having covered the area at the time, making it uninhabitable.

Understanding the age of the Somme terraces staircase and the deposits preserved therein had never been a simple matter of counting up or down the levels, the deposits at any given height were polygenic and belonged to different climatic cycles. Still, it appears that while Commont certainly began his work by building objective sequences based on sound stratigraphical and palaeontological principles, he ultimately allowed a linear evolutionary model for stone tools to take precedence, interpreting faunal and stratigraphical sequences on the basis of their associated archaeology, not vice versa. To account for the Chellean being located so low in the valley required Commont to invoke a sequence of earth movements involving uplift, subsidence and/or immense aggradation/inundation. The distribution of the Acheulean across several terraces and the high plateau, was more easily explained, by (re-)interpreting the relevant loamy silts and sands as subaerial in origin, slopewash deposits and wind-blown dust, thereby rendering their elevation irrelevant for determining their age. Rutot (1911) thought this was mistaken, and that the silts were fluvial.

Nevertheless, Commont (1912a, 1912b, 1912d, 1913, 1914) provided a definitive, linear sequence for St Acheul and its environs that would have a major impact on archaeological theory for decades to come (Table 5.11). His work transformed the handaxe epochs, turning them from long monotonous periods marked by technological stasis to stages in an evolving lineage, wherein form and method gradually improve over time. He created several sub-periods, each characterised by a distinctive dominant handaxe-type, with hints of other, perhaps transitional forms (twisted ovates, La Micoque). He recognised variation and overlap, emphasising (but perhaps not always practising) the principle that whole series were required to classify an industry, not a single implement. He appears to have been the first to use handaxe form in this fashion.

Table 5.11 Commont’s later Palaeolithic framework for northern France (after Commont 1910a, 1910b, 1913). Commont thought correlations with Penck’s glacial scheme were premature but tentatively followed Obermaier in thinking that the Chellean and Acheulean belonged to the Riss-Würm Interglacial, the Mousterian just before and during the Wurm glaciation. The Upper Palaeolithic was all postglacial, in Penck’s meaning of that term. Likewise, while he considered correlating his scheme with adjacent regions, such as Belgium and east of the Rhine, he felt the details were lacking and decided to wait. In England, he correlated the boulder clays of East Anglia with the Riss glaciation (Commont 1914).
Stratigraphical Designation	Description	Industry	Type Fossils
	Beneath modern deposits	Magdalenian
A	Upper silts	Solutrean	Isolated finds
A	Upper part of recent loess and upper silts	Upper Aurignacian
A	Pebble parting in upper part of recent loess, Low Terrace at Selle	Typical Aurignacian
C	Pebble parting in lower part of recent loess on 2nd Terrace	Upper Mousterian	Quina type, no handaxes
Cl, C2	Gravels in lower part of recent loess on 2nd Terrace	Lower Mousterian	a) with plano-convex handaxes (Micoque)b) with triangular handaxesc) Levallois cores and crude pseudo-handaxes
C3 – D	Gravel parting of Low Terrace; red-brown sandy silt of 2nd Terrace	Upper Acheulean	Lanceolate handaxes dominate
I – F	Red Gravel of Low Terrace/ marbled sands of 2nd Terrace	Lower Acheulean	Ovate handaxes dominate
D, E	Base of red sand and from the bottom of pockets of white sand, 2nd Terrace, Rue de Cagny	Lower Acheulean	Twisted ovate handaxes dominate
L	Lower Gravel of Low Terrace at Monti eres	Advanced Chellean	Triangular handaxes made with large flakes
K, H, L	Lower sand and top of lower gravels, 2nd Terrace	Typical Chellean	Handaxes, small tools
L	younger deposits of 3rd Terrace, lower gravel of 2nd Terrace	Pre-Chellean	Crude handaxes, many small tools

Commont divided Chellean and Acheulean technology into two schemes, one based around handaxe manufacture, the other core-and-flake reduction, and noted clear differences between the flakes produced by each. He identified another core-and-flake reduction method, Levallois, as the most diagnostic feature of the Mousterian, a period equally marked by gradual technological improvements. As Levallois was absent in the handaxe periods, it made a far better marker fossil than the ubiquitous scrapers and flake tools used by Mortillet. He thought the Acheulean and not the Mousterian marked the perfection of the handaxe, which occurred in more refined, more varied and more specialised forms, situating the plano-convex handaxes from La Micoque in the early Mousterian. He also recognised the role of raw materials in determining handaxe form and refinement, citing examples from the red silt at Tellier’s Pit that had only been sharped to a point because they were made on tablets of flint that might have been easily broken during more elaborate working (Commont 1908). He latterly wondered whether the Acheulean and Mousterian were made by different races (Commont 1912a).

In 1906, two years after he began his research at St Acheul, Commont noted that the Acheulean no longer had the same meaning as it carried when Mortillet had first defined it 30 years earlier. In the intervening years it had been exiled, redesignated and reinvented. By the end of the decade, Commont had unleashed it.

Golden Slumbers

At the inaugural meeting of the Prehistoric Society of East Anglia (hereafter PSEA), held in the Norfolk and Norwich Library on 26 October 1908, Dr William Allen Sturge (1850–1919), retired Royal physician and first president of the fledgling society, lamented how Britain had “dropped behind” its neighbours in the “attention paid to prehistory”. There were remarkably few papers to be found in the Society journals, he moaned, and the glory days of Evans, Christy and Pengelly were long gone. Britain had seen few discoveries of real note since the Palaeolithic floors of London were found a quarter of a century earlier, although he was quick to add that Miss Layard’s new discovery at Foxhall Road was more than worthy of remark. Even the eoliths had gone into ‘abeyance’. Prehistory in England, thought Sturge, had simply gone to sleep, it had become the “Cinderella of archaeology”. They knew hardly any more in 1908 than they had when Boucher had first announced his discoveries. “Everything remained to be done”, warned Sturge (1908, 10), and he was not alone in this view. Boule (1915), thought that British prehistory had become insular after 1875, never fulfilling its early promise, while Breuil was heard to remark at a lunch in Cambridge that the English knew nothing of prehistory (Burkitt cited in O’Connor 2007, 239). British geologists had also all but lost faith in the chronological value of artefacts, the revised memoir for The Geology of the London District (Woodward 1909) repeating verbatim the British Museum’s (Read 1902) opinion that the French chronological system had not yet been established in this country.

Having woken from its own slumber, Sturge thought France now led the way, particular advances having been made by Breuil, Cartailhac, Peyrony and Capitan in the archaeology of the cave period, inspired by their own dazzling discoveries of Upper Palaeolithic paintings and engravings. In Belgium, Rutot had been equally busy redefining the eolithic and Early Palaeolithic periods and was to be praised for his zeal if not his beliefs. Even the Germans had finally started to gain momentum after a long period of disinterest and were now moving ahead of Britain in some respects. It was a national disgrace, declared Sturge, because apart from the cave period, Britain had a record that was as rich as anything seen on the continent, if not richer. He blamed a spirit of obscurantism that had affected those in positions of authority and looked forward to a new halcyon era to come when we could move beyond the unsatisfactory Chellean or Acheulean, as he had heard Commont was now doing in France.

To be fair to his compatriots, Sturge’s opinion of French archaeology might have been favourably coloured by his long residence in Nice between 1880 and 1907, where he had acted as physician to Queen Victoria, or simply by the cave art. While prehistoric France might have had a more healthy publication record and dedicated journals such as L’ Anthropologie (founded in 1890) and the Bulletin de la sociéte préhistorique de France (founded 1904), the Lower Palaeolithic articles were mostly reportage, the announcement of discoveries or syntheses littered with casual and ill-defined attributions to the Acheulean, Chellean or Mousterian, each based on idiosyncratic readings of Mortillet’s many schemes or the eolith-heavy classifications of Rutot (see for example, Chantre 1900; Laville 1902; Coutil 1904; Thiot 1904; Desailly 1907; Desmazières 1908; Baudet 1909). This remained true even after Com-mont’s main conclusions were in the public domain, when few workers took advantage of the finer time-slices it offered when announcing new finds (Desforges 1911; cf. Givenchy 1911; Reynier 1914; Socley 1916). Indeed, as one of Sturge’s unnamed French colleagues had quipped, if prehistory slept in England, it was dying in France (Sturge 1908, 10).

Part of the problem in France was that the old doubts still lingered. Joseph Dechélette’s magisterial Manuel d’archéologie préhistorique (1908a) was up to the minute in terms of the work of Commont, Obermaier, Breuil and Boule, but was largely of the opinion that fauna was the best way to classify the Lower and Middle Palaeolithic because tool forms were too rudimentary to command confidence. He accepted that forms developed over time, but their size and shape also varied by region. In some areas, for example Central Europe east of the Rhine, handaxes were completely absent. At Taubach, Germany (Reinach 1897), an ancient fauna of straight-tusked elephant and Merck’s rhinoceros was associated with a diminutive flake industry that approached the Mousterian. Its character, however, belied its age. It was probably, argued Dechélette, a response by Chellean peoples to indigent local raw materials that were unsuitable for handaxes. Similarly, in areas with no flint, humans had used other rock types and had often made do with cruder handaxe forms, which were thus a poor guide to age. This was certainly true in Portugal, where the quartzite handaxes had outline shapes like Chellean and Acheulean flint examples from France, but they were coarse and unrefined in their manufacture (Fortès 1911, 1912). The question of bone tools in the Chellean and Mousterian also enjoyed a brief celebrity in Paris (Thieullen 1901; Martin 1907; Pittard 1907).

Commont knew all of this, however. Like many before him he noted that the nature of flint can influence the refinement and size of handaxes, and that a coarse aspect did not always indicate great age (Commont 1914). He also observed that raw material availability could influence the character of the industries found in open-air sites. In Picardy, where flint was abundant, the tools were little used and minimally retouched. By contrast, at Roisel and Busigny, where humans had to go some distance to find suitable materials, they were used to exhaustion, with the cores completely flaked to small discs and the flake-tools heavily retouched. They looked identical to those from Spy and Le Moustier or the new Neanderthal cemetery at La Ferrasie (Peyrony and Capitan 1910).

Several other well-worn issues similarly refused to die. At the Société préhistorique in Paris, Vauvillé (1910) challenged the cogency of the named industries at Mont-Notre-Dame (Aisne), where so-called Chellean handaxes occurred with an abundance of small tools, as d’Acy had pointed out 20 years earlier and as Commont had recently reaffirmed in the Somme. What he failed to acknowledge was that Commont had used this information to redefine the Chellean, which now included a range of rudimentary scrapers and flake tools and was defined by characteristic handaxe shapes. Laville (1910) likewise re-examined the correlation of the Chellean with warm faunas, suggesting the traditional associations of certain animals with specific climates might be oversimplified. This was certainly true, but the well-preserved botanical remains from the Chellean station at La Celle (Déchelette 1908b) left little room to doubt that climate had been warm during Chel-lean times there.

Regardless of Commont and the Alpine glacial sequence, very little changed in wider French thinking on the Lower Palaeolithic until World War I brought everything to an earth-shattering halt. The conflict would rob France of Commont (evacuated to Abbeville following the shelling of Amiens and died in April 1918 of a lung disease contracted during a field trip to abandoned Somme battlefields), Déchelette (killed in action, October 1914), Bourlon (killed in action, August 1914) and others. Despite Sturge’s jaded views of archaeology at home, it was the British who would embrace the new ideas most fully, for better or for worse.

Swanscombe: England’s St Acheul 1912–1914

In Britain, Commont had a powerful advocate in Reginald Allender Smith (1873–1940). Smith had been assistant Keeper of British and Medieval Antiquities at the British Museum since 1898, working under Charles Hercules Read. Smith was serious, hard-working, fastidious in appearance and lifestyle and somewhat cliquey (Tonnochy 1953, 87; Kendrick 1971). One of his main duties at the Museum was the revision of the guidebooks that accompanied the displays in his charge, including the second edition of the Guide to the Antiquities of the Stone Age (Smith 1911a). These were relatively inexpensive, at one shilling affordable to many of the Museums approximately one million annual visitors. Smith may have had little power to radically change the arrangement of the displays, but he could alter the interpretations the Museum presented to the public, using the guidebooks as institutionally backed shapers of opinion.

Whereas Read (1902) had found little merit in Mortillet’s scheme, Smith openly embraced Commont’s work, which he felt had restored confidence in the link between artefact form and age. He endorsed Commont’s sequence as the continental standard, which he tabulated for easy reference and inserted sparingly in his description of palaeoliths in the collections, in doing so introducing terms such as St Acheul I (Commont’s Lower Acheulean), St Acheul II (Commont’s Upper Acheulean), and the French understanding of Levallois cores and flakes to a large English-speaking audience (Table 5.12). He also accepted the Alpine glacial sequence of Penck but was undecided which of the many available correlations with the Palaeolithic record was correct. The quality of the British Lower Palaeolithic record, which had been unsystematically excavated and contained a confusion of forms, prevented him from attempting a relative chronology, but Smith had hopes that Commont’s classification would soon be extended to British rivers.

Table 5.12 A summary of Commont’s findings in the Somme, as presented by Reginald Smith in the British Museum Guidebook.
Series	French Bed Description	English Description	Period	Occurrence and Character of Implements, etc.
Upper	Terre-à-briques	Upper brickearth Lower brickearth	Neolithic Madeleine	Flints of many forms Bluish blades, planing blades and graving tools
	Ergeron (loëss)			Barren except in intercalated bands of gravel
	Cailloutis	Upper gravel	Moustier	Bluish white flints, Levallois flakes, handaxes rare
Middle	Limon rouge fendillé (sable des fondeurs)	Foundry sand	Acheul II	Flat oval implements called limandes, lanceolate and small specimens, lustrous white patina
	Limon grisâtre à poupées caliare	Grey loam, calcareous concretions
	Sable roux (doux) à points noirs. Cailloutis	Red sand with manganese specks Middle gravels	Acheul I	Reddish limandes lustred by sand but not patinated; also small implements; grey or yellow flint, unrolled. Twisted ovates common
Lower	Sable gras (terre-à-pipe)	Sticky sand (yellowish white)		Upper part with fossil land shells, lower part with river-shells
	Sable aigre	Sharp white sand	Chelles	Handaxes including ficrons (long points) and implements with thick butts, unrolled and unpatinated; black or yellow lustred flint
	Gravier du fond	Lower gravels (resting on chalk)	Early Chelles	Worked flakes with primary chipping, deeply patinated (orange and olive) and retaining much of the greenish grey crust

Smith was one of a number of quiet eolith sympathisers, and in his Guide both Harrisonian eoliths and Rutot’s Belgian examples were given equanimous treatment, noting that they were “not accepted by some and differently interpreted by others” (1911a, 1). Eoliths occupied the top half of just two display cases (106 and 107) but took up almost half of the summary table that opened the description of the older Stone Age galleries (Table 5.13). Smith also provided a summary of Rutot’s scheme, which is worth rehearsing here as context for some of the work to follow. For Rutot eoliths were not deliberately shaped but natural forms selected for their inherent properties. They were therefore not divided on the basis of style, which was the same throughout, but on context. The oldest Belgian eoliths, belonging to the Reutelian, were found on the 80–100ft river terraces, the Mafflian on the next terrace down, and the Mesvinian on the alluvial surface of the middle and low terraces. This was followed by the Strépyan, an industry characterised by long rods of flint chipped at one end, possibly simple daggers, which marked the beginnings of deliberate design, the origin of weaponry and the transition between the eolithic and Chellean. It was also found under the sable aigres at St Acheul (see also Sollas 1911).

Table 5.13 The British Museum’s arrangement of prehistoric flint artefact industries (after Smith 1911a).
Period Named After Site	Characteristic Forms of Flint Implements	Classification	Geology
La Madeleine	Long blades and minute tools, borers, etc.	Caves and	Upper Quaternary
Solutré	Points with shoulder on one side; lozenge and leaf-shaped blades, finely chipped	shelters
Aurignac	Planing tools with steep sides, pointed blades with one edge battered
Le Moustier	Scrapers made of flakes worked on one side
St Acheul	Amygdaloid (almond-shaped) implements, finely flaked and often twisted	River drift, alluvium	Middle Quaternary
Chelles	Handaxes often pear shaped or fiat oval, coarsely flaked
Strépy	Nodules flaked generally at the point
Mesvin		Eolithic (Rutot)	Lower Quaternary
Maffle	Nodules, not designedly chipped
Reutel
Kent Plateau	Ochrous, the edges much worn by rolling, usually of the scraper form	Eolithic	Early glacial/ preglacial

Smith then introduced another new term to the English language. Thinking a name was still needed to describe the normal type of Chellean and Acheulean implements, Smith offered ‘hand-axe’ as a useful translation of coup de poing. Although Mortillet would certainly not have approved of the word ‘axe’, hand-axe or handaxe was quickly and widely adopted and was clearly preferable to Sollas’s (1911) suggestion of ‘Boucher’ (in honour of Boucher de Perthes, and perhaps also an indication of its function).

The Geologists’ Association excursion to the British Museum on Saturday 10 February 1912 gave Smith (1912) the opportunity to explain the new Continental sequence to 85 visiting members and the wider readership, journal readers who were not at the event being given clearly labelled illustrations of characteristic types borrowed from the new guidebook. The European sequence, Smith maintained, had been proven in every way and was accepted, but the national collections were not in great shape. Collectors again shouldered much of the blame, but geologists were also at fault for failing to record archaeological sequences in sufficient detail. To test whether the Continental sequence could be applied to Britain required new evidence built on better collaboration between archaeologists and geologists.

A week later Smith visited the Barnfield Pit, just off Milton Street in Swanscombe, Kent, where the Associated Portland Cement Manufacturers were about to start clearing a new area of gravel from the 100ft terrace of the Thames to expose the Chalk beneath (McNabb 1996b; O’Connor 2007, 241). The archaeological richness of these deposits was well known (Spurrell 1883; Hinton and Kennard 1905), and Smith was immediately concerned that the site was in danger of extinction before its archaeological sequence had been properly studied (Smith and Dewey 1913, 177). Enlisting the skills of BGS geologist Henry Dewey (1876–1965), who since 1910 had been remapping the deposits around London for a new series of large-scale (six-inch to the mile) maps, Smith wrote to Hercules Read at the British Museum and Clement Reid at the BGS proposing a joint venture, on their own vacation time and with private funds if necessary (McNabb 1996b).

By the end of March 1912, Smith and Dewey were excavating at Swanscombe on behalf of their respective institutions, spending ten days in the main Barnfield Pit, two days in the smaller Colyer’s Pit 200ft (61m) to the south, and three days at Globe Pit in Greenhithe (Smith and Dewey 1913). At the Barnfield Pit they established the now-classic tripartite sequence: Lower Gravel and Lower Loam; lower Middle Gravel and upper Middle Gravel (current-bedded sand); Upper Loam and Upper Gravel (Figure 5.10). The Middle Gravels produced a rich in situ assemblage of flattened pointed handaxes, roughly flaked and with heavy cortical butts. Smith compared them with Chellean handaxes from the Somme, although noted that there were stratigraphical difficulties. The Upper Loams at Swanscombe had reportedly produced St Acheul types, white patinated ovates with frequent twisted edges, although none were recovered during the 1912 excavation.

Figure 5.10 Swanscombe, Kent, 1913. Plan of Barnfield Pit, Swanscombe, and a section through the entire sequence, drawn from a photograph. The overhanging scale is divided by cards into spaces 1ft long (after Smith and Dewey 1913, 1914).

Further excavations in 1913 also failed to produce artefacts from the upper deposits in Barnfield Pit, or in Colyer’s Pit, where the upper part of the sequence was completely missing (Smith and Dewey 1914). Happily, they were able to confirm the old reports at a new pit on the other side of Craylands Lane, 600ft (183m) east of the Barnfield section. A request to the workmen at the New Craylands Lane Pit to carefully collect all artefacts from the upper deposits resulted in a collection of 16 white patinated handaxes from the junction of the upper Middle Gravel (current-bedded sand) and an even-bedded upper gravel, several of them twisted ovates (Smith and Dewey 1914). Supporting the workmen’s testimonies, either Smith or Dewey was present to witness the recovery of one white handaxe in situ. At Barn-field Pit, where the even-bedded upper gravel was absent, a shaley Upper Loam rested on the cross-bedded sands instead. What was clear to Smith and Dewey (1914) was that regardless of their coarseness or origin, the last cycle of deposits in both pits was younger than the Chellean Middle Gravel, and contained younger implements of a different, Acheulean, type. The highest deposit, a clayey Upper Gravel, apparently contained a blue-white patinated industry with numerous large flakes similar to Baker’s Hole (Southfleet Pit), Northfleet, another Flaxman Spurrell site (Spurrell 1883; Smith 1911b; Abbott 1911). In Commont’s scheme these would represent an early Mousterian occupation, and Smith thought the resemblance between St Acheul and Northfleet itself so striking that the two must have been connected in culture if not in blood (Smith 1911b).

Twisted ovates were again recovered during Smith and Dewey’s (1914) work at Wansunt Pit, Dartford Heath. Here they agreed with Leach and Chandler (contra Hinton and Kennard) that the deposits were not on a higher terrace but were part of the 100ft formation seen at Swanscombe. Their reconstruction of the geology and geomorphology of the site differed from Chandler and Leach’s (Chandler and Leach 1912; Leach 1913), however, as they interpreted the Wansunt loam to be the final deposit of the Thames at this height, not a later tributary. As a group, they thought the handaxes from the Wansunt Loam belonged to St Acheul II, although St Acheul I and Mousterian elements were also found at the site.

If the top part of the 100ft terrace at Swanscombe and Dartford conformed to expectations, a neat progression of Chellean overlain by Acheulean, the industry from the Lower Gravels was “a total surprise in both quantity and quality” (Smith and Dewey 1913, 182). Nearly 200 flints were recovered from this deposit during 1912, two-thirds of them from a red band immediately above the Thanet Sand. Most were thick flakes with large striking platforms, prominent bulbs and minimal retouch, the largest being nearly three-quarters of a kilogram (1lb 10oz) in weight, but disc-like cores were also present, an accident of chipping rather than design, they thought. A few chipped rods of flint were also found, which Smith and Dewey suggested might correspond to Rutot’s Strépyan, but no handaxes of any type were present. Stratigraphically it was Pre-Chellean, but the character of the industry did not match Commont’s equivalent at St Acheul (e.g. Commont 1909a), which contained crude and irregular handaxes. Commont, in fact, refused to accept the Strépyan and thought all Rutot’s eolithic stages to be much younger than proposed (O’Connor 2007, 260), meaning the Strépyan would have been post-Würmian, had it been real: it had nonetheless managed to creep into usage to describe stratigraphically Pre-Chellean industries (e.g. Sollas 1911).

A remarkably similar assemblage was recovered from the shell bed at Dierden’s Pit, Ingress Vale in 1913, which, lying directly on Chalk at a similar height, was correlated with the Lower Gravels at Barnfield Pit (Smith and Dewey 1914; Dewey and Smith 1914). Over 500 thick flakes were found in the shell bed, but not a single handaxe, contradicting earlier reports of twisted ovates from the shell bed by Newton (1901) and Stopes (1904). Smith and Dewey seized upon a report by Francis Corner, who found twisted ovates just 3ft beneath the surface, speculating that they had come from a higher and now worked-out bed near the centre of the pit. A later review of the Stopes Collection from Greenhithe revealed Acheu-lean, Mousterian and Aurignacian materials, leaving “room for some ingenious theory to reconcile the flints and the fauna” of the 100ft terrace (Smith 1918a, 187; see T. White et al. 2013; White et al. 2019).

Smith and Dewey (1913, 197) initially compared the sequence from the 100ft terrace at Swanscombe with that described by Commont at Carpentier’s Quarry, situated on the third (40m) Terrace of the Somme at Abbeville (Table 5.14). A year later (Smith and Dewey 1914, 205) their main point of comparison was the Second Terrace at St Acheul. Neither match was perfect, but the sequence of industries in the ‘complementary’ pits at Swanscombe generally contained a full Palaeolithic sequence from Pre-Chellean to Mousterian, identical to France and Belgium (Smith and Dewey 1914). The form of an implement was a reliable indicator of its age, especially when combined with surface condition (see Sturge 1909, 1912a, 1912b for an over-emphasis on patination) They had not forced these correlations upon the facts, they insisted. The facts had been allowed to speak for themselves, pushing the analogy with Commont’s work upon them (Smith and Dewey 1913).

Table 5.14 Correlation of the sediments and archaeology at Swanscombe with that at Abbeville (after Smith and Dewey 1913).
Period	Strata at Abbeville	Implements CarpentierPit	Implements BarnfieldPit	Strata at Swanscombe
Neolithic	Hill wash (black sandy soil) Splintered white flints	Neolithic scrapers	Either Neolithic or cave period flints	Hill wash (black sandy soil) Bleached clayey gravel
St Acheul II	Laminated loam	One white implement	Reputed horizon of white implements	Laminated loam
St Acheul I	Flint seam, only prolific stratum	All ovate implements, none with thick butt. Unpatinated or with yellow or reddish patina, unrolled, rarely twisted and lustred by sand	Chelles types	Middle Gravel, replaced laterally in part of the pit by current-bedded sand
Chelles	Current-bedded sand, gravel of rolled flint	Chelles types	Chelles types	Current-bedded sand on thin seam of gravel
	White marl	Barren, fauna resembling Pliocene		Greenish loam
Pre-Chelles	Coarse red gravel		Flakes, Elephas antiqitus, no handaxes	Coarse red gravel

There were, in fact, significant differences in both empirical data and their interpretation to what was seen in France. There was no equivalent of the flake industry from the Swanscombe Lower Gravels in the Somme; it was a surprise and could not be confidently attributed to any stage. It was just earlier than the Chellean. The small and often acutely pointed handaxes from Swanscombe might certainly have warranted a Chellean designation in France, but perhaps the advanced rather than the typical variant. There was no St Acheul I (Commont’s lower Acheulean), because Smith and Dewey thought the twisted ovates in the upper deposits at Swanscombe and Wansunt represented a stage of perfection, late forms from the very end of the so-called drift period. They assigned most of them to their advanced St Acheul II stage (Commont’s Upper Acheulean), noting the occurrence of some Mousterian scrapers. In the Somme, however, the Upper Acheulean was characterised by slender lanceolate handaxes, some of which continued into the Micoque stage of the Early Mousterian, whereas limandes and twisted ovates belonged to the lower Acheulean. Both included scrapers. Smith’s demonstration at the Museum had shown that he certainly understood Commont’s original definitions (Smith 1912), suggesting that these modifications were deliberate. It would soon become the standard British sequence.

A Rope of Sand

Smith and Dewey’s two great institutions had rather different aims, the unfortunate result of which saw some key figures at the Survey deride the Museum’s purpose at Swanscombe as nothing more than finding implements (O’Connor 2007, 254). They also felt that Dewey’s work with Smith was interfering with his official duties, and that he was trespassing on other Survey geologists’ allotted territory. The collaboration could not continue. Their final campaign, which saw Dewey working in an unofficial capacity only, tackled deposits further up the Thames Valley at Croxley Green and Mill End near Rickmansworth, located on the junction of the rivers Gade and Colne (Smith and Dewey 1915), and first brought to attention by the late Sir John Evans (Evans 1907). The aim here was to determine whether tributary valleys contained the same archaeological sequence, and to establish the relationship between the archaeology and the deposits mapped as glacio-fluvial gravels on top of the Chiltern Hills. Their own and earlier finds had been individually attributed to the Chellean and Acheulean, but these gravels were a tumbled mass. They could feasibly be of glacial meltwater origin, which would make them older than the chalky boulder clay, but for Smith and Dewey (1915) the presence of Acheulean artefacts suggested otherwise. Perhaps, they mooted, these objects had been discarded on ancient land-surfaces on the gravel and had lain there ever since, although Smith would later describe this as a ‘desperate’ theory (Smith 1917). The problems of palaeoliths in plateau drifts would certainly not be solved simply by continuing to reject the French classifications, they argued. These would certainly outlast common views on the age and origin of different gravels (Smith and Dewey 1915, 195).

This time though the British did anything but reject the French classification. Kendall (1909) had already applied Commont’s terms to the handaxes from Savernake, identifying Pre-Chellean, Chellean, Lower Acheulean and Upper Acheulean types. Prestwich’s successor to the Chair of Geology at Oxford, William Sollas (1849–1936), also drew heavily on Commont and Obermaier in his popular Ancient Hunters and their Modern Representatives (Sollas 1911). It was soon applied to other regions outside of the Thames Valley. Clarke (1913) identified a complete sequence from Chellean to Solutrean at Barnham in Suffolk, while at Dovercourt in Essex, Colonel Underwood (1913) found 150 handaxes in the highest terrace of the Stour, which he divided into Chellean, St Acheul 1 and St Acheul 2, although he noted these attributions were not consistent with their physical condition. Cambridge pre-historian, Miles Burkitt, who had worked with Breuil and Obermaier in France, also used the French classification on material from the Cambridgeshire gravels (Marr and Burkitt 1919). Crayford’s Mousterian status was both confirmed and refuted by new finds (cf. Chandler 1915, 1916; Higgins and Smith 1914).

At Farnham in Surrey, local geologist Henry Bury identified four main terraces of the River Blackwater-Wey (Figure 5.12), each of which contained a distinctive Palaeolithic industry that he described in French terms, and which he also correlated with the Thames sequence (Bury 1913, 1916, 1919). The oldest deposits, on Terrace A, yielded crude pointed handaxes belonging to the Chellean and Advanced Chellean, with Acheulean present in deep pockets and near the surface. Terrace B contained better-made and smaller handaxes which fell into eight types, mostly of St Acheul 1 and 2 age, including twisted ovates. The Mousterian was found on Terrace C, in the form of Levallois flaking, but this terrace also produced elongated pointed forms, some with concave edges (the British ficron), which Bury thought belonged to the Advanced Chellean.

Figure 5.12 The Farnham River terraces. Top left: Bury’s (1913) correlation of the four higher terraces of the Farnham River with the Thames; bottom: an early terrace staircase diagram for the Farnham River (after Dines and Edmunds 1929); top right: handaxes from Terrace A (upper), Terrace B (lower left) and Terrace D (lower right).

On the basis of their contained archaeology, Bury suggested that the river had laid down its terraces in a non-uniform sequence, Terrace A was first, but this was followed by C, then B and then D, involving at least one period of submergence and two periods of uplift. The Levallois on Terrace C was hypothesised to come from a subaerial deposit contemporary with Terrace D. There was nothing in the artefacts that could serve as a good zone fossil, Bury observed, but a distinct archaeological succession was evident. That did not stop others dividing their own surface finds into early and Late Acheulean entirely based on their form (e.g. Dewey 1915; Peake 1918).

Other problems soon emerged with these continental comparisons. Foremost, different workers might have been using the same names, but they were often operating under quite different definitions to either Commont or Smith. Personal differences of opinion on typo-logical classification could have major implications for interpretations of the geology and the archaeology. Bury (1916) and Smith (Wade and Smith 1934), for example, classified two near identical handaxes from Farnham Terrace C (Figure 5.13) as Advanced Chellean and Micoquian respectively. It was a serious divergence, as the former attribution required the river to yo-yo up and down during periods of subsidence or uplift, whereas the latter allowed a regular sequence of progressive downcutting to have occurred.

Figure 5.13 What’s in a name? Two handaxes from Terrace C Farnham, differing only in thickness. The one on the left was labelled Advanced Chellean by Bury, the one on the right classified as Micoquian by Wade and Smith (after Bury 1913; Wade and Smith 1934). Scale = 5cm.

Smith (1915) also soon realised the problems of reconciling the 100ft Terrace of the Thames with that at Amiens. The French thought that their river valleys had been excavated to their present depth in the Early Pleistocene, as shown by Chellean artefacts in the lowest terrace gravel. British rivers seemed to have begun downcutting only after the 100ft terrace had been formed. Without assuming repeated and considerable adjustments to the level of the river it was

difficult to understand how St. Acheul types come to be included in high-terrace gravels of the Thames and Somme if both rivers cut down below their present beds in the Chelles period, and their sunk channels were cut at a period of maximum land-elevation corresponding to Le Moustier.

(Smith 1915, 3)

Commont’s solution had been to assume all the sediments containing Acheulean assemblages were later rain-wash deposits, but this had been disputed by Rutot, and was certainly not true at Swanscombe, where the deposits were undoubtedly fluvial (see Bury 1919). There were additional problems with correlating the marine sequences either side of the Channel. Dewey’s (1913) reconstruction suggested that the raised beach at Brighton was formed in Pre-Chellean times, while Commont thought the marine deposits at Menchecourt were formed at the end of the Chellean.

To explain this, Hinton (1910) proposed that the English Channel must have acted as a tectonic pivot, Britain submerging as France was uplifting during the Chellean, the pivot then switching so that the Thames was uplifted at the end of the High Terrace (Acheulean) phase, as the Somme began to sink. More uplift and downcutting followed the 50ft Terrace phase, the river excavating its valley to below modern levels. It was during this late period of maximum uplift that glaciers formed. In Hinton’s mind then, the great ice age was much younger than most people thought. In his work on Pleistocene voles and lemmings, species that evolved relatively rapidly and were highly sensitive to climate change, he had found few differences between the Cromer Forest Bed and the 100ft Terrace of the Thames (Hinton 1910). He certainly saw no evidence that a major glaciation separated the two formations. The small mammals from the Middle (50ft) Terrace at Grays and Ilford also contained many survivors from the 100ft level. In fact, compared to the Somme, the British fauna from these terraces was decidedly archaic and insular, suggesting to Hinton that Britain was already an island. The first major faunal turnover occurred at Crayford and Erith, also on the so-called Middle Terrace, from which Hinton concluded that while they appeared to be one physical structure this terrace contained deposits of different ages. Despite the well-known presence of musk-ox at Crayford, Hinton refused to believe there had been any major climate change, rather the re-connection of a land-bridge between Britain and Europe allowing an influx of new species. In fact, Hinton could find no evidence for a major early glaciation. The entire Palaeolithic was preglacial, not postglacial and certainly not interglacial.

Hinton’s only significant ally in this belief was Kennard (1916, 259), who argued that the British molluscan fauna also showed an “unbroken succession from the Pliocene to the end of the Pleistocene”. The only evidence for a period of extremely cold conditions occurred very late in Pleistocene, where it was marked by the ‘arctic bed’ described at Ponders End by Warren (1912a) and by Reid at Hoxne (Evans et al. 1896). To maintain his position, Kennard conveniently dismissed the many reports that handaxes occurred above the boulder clay at Hoxne, arguing that the correlations were largely based on borings and were thus unreliable, and rejected as pure inference the widely accepted view that at Hornchurch the till was over-lain by deposits of the 100ft Terrace. He doubted things would have been so complicated if ancient humans had not been involved and saw no merit in trying to fit the British data onto the Germanic glacial scheme, displaying more than a little anti-German sentiment (and not just because it was wartime). Real advances would only come once the British had their own secure Pleistocene framework, they both argued.

Kennard and Hinton were largely alone in their views (Savage 1963). Warren (1912a) had been clear that the Ponders End arctic bed was not the same age as that at Hoxne, but much younger, possibly corresponding to the ‘trail’ at the top of many Palaeolithic sequences. Warren also accepted the multiple phases of the Alpine glacial chronology, although he found it hard to establish a firm correlation. Nonetheless, whether one accepted the long or the short correlation, the Ponders End stage was definitely post-Würm (Table 5.15).

Table 5.15 Warren’s (1912) suggested correlation for the Arctic beds at Ponder’s End (Note the almost incomprehensible sequence dates).
Period	Seq. Dates	Ponders End .	Admirality Buildings Westminster	Hoxne	Stoke Newington	Norfolk	Boule	Penck
Prehistoric	10.00–9.3
Hiatus	8.xx	Arctic	Arctic	Trail	Trail		Postglacial	Postglacial
Magdalenian	7.9							Glacial (Wurm)
Solutrean	7.8							Interglacial
Aurignacian	7.73–7.77
Mousterian	7.7						Glacial (Wurm)	Glacial (Riss)
Acheulean	7.6			Temperate	Temperate	Palaeoithic. river	Interglacial	Interglacial
Chellean	7.5					gravel
Early Palaeolithic	7.4			Arctic				Glacial (Mindel)
	7.3			Temperate				Interglacial
Chalky Boulder Clay				Glacial		Glacial	Glacial (Riss)
Contorted Drift etc.
Cromer Till etc.						Arctic freshwater bed		Glacial (Gunz)
Forest Bed						Temperate	Interglacial Glacial (Mindel)	Preglacial

Explaining the origin of the brickearths and gravels of the high plateaux, and the mixture of industries they contained, posed another geological conundrum (Evans 1908; W.G. Smith 1916; R. Smith 1915, 1917, 1918b). Reginald Smith (1917, 393) complained that “some violence has been done to the Palaeolithic classification” by prioritising the geological opinion that the plateau deposits must be very old, which made the highly refined implements they sometimes contained older than the Chellean. The point had been reached, he suggested, when it was easier to move the gravel than the industry. Worthington Smith’s (1916) recently announced high-level finds in the Chiltern Hills at Round Green and Gaddesden Row, where Late Acheulean handaxes occurred with Mousterian racloirs, were a good example. They appeared to be related to some kind of sinking of the land, creating large swampy areas. The contorted drift above the floors had flowed over the areas during a later glacial period but contained handaxes derived from older land surfaces. The notion that the plateau deposits were necessarily ancient had long been abandoned in France (Reinach 1916), because they contained everything from Acheulean to Aurignacian and must have formed at different points throughout this long period. Smith was certain it was the same in Britain, and that patina was not an accurate guide to age (contra Warren 1902; Smith 1915; Sturge 1908). Echoing Commont, he wondered whether the terrace deposits were really all fluvial in origin, rather than mantles formed by wind, rain and gravity.

Warren was having difficulties of another kind at Clacton-on-Sea, where since 1908 he had been collecting bones and fossils from Pleistocene fluvial beds exposed on the fore-shore. The site had produced an interglacial fauna including straight-tusked elephant, Merck’s rhinoceros, aurochs and deer along with the tip of a wooden spear (Warren 1911) and a stone industry dominated by simple flakes, scrapers and ‘pseudo-Mousterian’ forms (Warren 1912b). There were no typical Chellean or Acheulean implements and no Levallois, the only shaped tool present being side-choppers (which Warren confusingly and unhelpfully decided to call handaxes). It was not Mousterian, and despite containing a warm elephant fauna it was not Chellean or Pre-Chellean but Post-Chellean. Warren (1912a) had many doubts about the chronological value of the elephants. Elephas antiquus was not limited to the oldest deposits and overlapped with mammoth on the 50ft terrace at Ilford, where it was associated with Mousterian lithics. There was also more than one type of mammoth within the Elephas primigenius taxon, the Ilford type (Mammuthus trogontherii) being warm adapted and smaller than the classic cold form (Mammuthus primigenius). They did not provide a straight-forward sequence. He agreed with Dawkins that the straight-tusked elephant had been a southern species, the mammoths a northern species, and that the two groups overlapped at their edges of their range during seasonal (or longer) climatic fluctuations (Warren 1912a, 220). Warren thought that artefacts were the more useful chronological tool, for wherever comparisons had been made between Britain and France the same succession of industries was found.

Warren had no immediate answer to the industry from Clacton, although a potential solution had existed for some time, contained in the post-doctoral thesis of Hugo Obermaier, published in German in 1908. Although his archaeological classifications came straight from Commont and his experiences in France (Déchelette 1908b), Obermaier’s interpretation came from the German tradition. His understanding of past populations was influenced by the school of Kulturkreise diffusionism, particularly the culture-historical notions and cultural cycles associated with Wilhelm Schmidt and Fritz Graebner at Vienna (Vila and Estévez 2010). For the Vienna School, cultures emerged and developed from individual centres, whence they spread to various parts of the world. “The complex cultural variations observed on every continent resulted from the mingling of cultures at different levels of development” (Trigger 1989, 152). To understand the patterns of prehistory one had to identify the areas where cultures had emerged, and the areas to which their ideas had been transmitted.

Obermaier (1908a) had thus made some subtle but critical amendments to the French sequence. He assigned the primitive (Pre-Chellean) industry at the base of the lower gravels at St Acheul (which consisted of flakes, retouched flakes, points and notches, but no refined handaxes) to an Early Chellean (früh-Chellean) without true handaxes. He also assigned the Mesvinian from Helin, near Spiennes (part of Rutot’s eolithic) to this period, as it contained genuine flakes with parallel negative scars on their dorsal surfaces. Summarising the work for an English audience, Sollas (1911) agreed that the Mesvinian contained genuine artefacts and should be removed from under the cloud of the eoliths but could not accept Obermaier’s nomenclature. For Sollas, the rehabilitated Mesvinian could not belong to the Early Chellean, as it contained none of the characteristic fautkeil and the Strépyan intervened between the two cultures. Sollas had an amendment of his own to make, re-assigning the Pre-Chellean at St Acheul to the Strépyan (Commont’s Pre-Chellean handaxe shown top right in Figure 5.11 was illustrated by Sollas (1911, 113) as Strépyan), giving this dubious eolithic industry a Palaeolithic pedigree, despite Commont’s own misgivings.

Figure 5.11 Chellean handaxes from the Middle Gravel at Swanscombe (top) and Acheulean forms from Wansunt (bottom) (after Dewey and Smith 1914). Scale = 5cm.

Obermaier (1908a) put additional strain on the classic French definitions by assigning the non-handaxe flake-tool industry from the lower unit at La Micoque to an early phase of the Lower Acheulean, a period otherwise defined by limande handaxes. If there could be non-handaxe regions or facies within the handaxe period, where did this leave the traditional definitions and stages? He also argued that Levallois had first appeared in the Acheulean, where it had been developed independently of handaxes from large flakes (first seen in the Chellean). It was not the reliable marker fossil for the Mousterian that Common had deemed it. Based on the new research of Maurice Bourlon (1875–1914) in the upper shelter at Le Moustier, it was apparent the handaxe had not gradually disappeared at the start of the Mousterian as once thought but had vanished suddenly and then reappeared later. The Mousterian could now be broken down into six facies (Bourlon 1906), the first dominated by thick scrapers that resembled handaxes. This was overlain by a typical Mousterian with fine scrapers and dominance of Levallois. Handaxes were again found in the Mousterian in the third level, close to the transition with the Upper Palaeolithic. They were a technological atavism.

Obermaier was not alone in holding an idiosyncratic understanding of what the archaeo-logical stages were supposed to represent. The by now ancient Dawkins continued to regard the French terms as describing contemporaneous series of types with no chronological significance, while Smith and Burkitt followed Commont in believing them to be highly distinctive industries that defined specific periods of time that occurred in a strict order (Marr and Burkitt 1919). Breuil on the other hand quickly published his support of Obermaier’s divisions of the Lower Palaeolithic, which they had discussed often and which were largely in harmony with his own (Breuil 1908).

Clearly influenced by Obermaier and Breuil (who had proofread the text of Ancient Hunters), Sollas (1911) argued that the archaeological periods marked only stages of culture, not divisions of time. “The industries overlapped the epochs” (Sollas 1911, 120), and there was no reason to suspect that the Chellean, for example, was the same age across the world. Tools from different stages could co-exist in different parts of the world, meaning handaxes could still be in use at the peripheries of human settlement long after they had been abandoned at the centres. Indeed, the Tasmanians, a race that had become extinct only in 1877, were for Sollas the perfect analogy for understanding daily life in the Lower Palaeolithic, they had been isolated living fossils who had even used a hand-held implement very similar to the handaxe (Sollas 1911). Sollas’s understanding of archaeo-logical stages thus allowed geographical and temporal variation in material culture, just like modern populations, and allowed for groups to come and go over time (cf. Lubbock 1865; Breuil 1913; contra Mortillet). Reconstructing a consistent story from the evidence was, wrote Sollas, like trying to weave “a rope of sand” (1911, 97), and as finer and finer subdivisions were proposed it would only get worse. Sollas further noted that the supposed distinction between Acheulean and Chellean handaxes was not as great as it appeared on paper, and that it was generally impossible to attribute an individual implement to one or the other, although when large collections were compared the differences became obvious (also Bury 1919).

Fossil Man in Spain (and Elsewhere)

Hugo Obermaier was in Spain when hostilities broke out between Germany and France, continuing the research that had already seen celebrated discoveries of Upper Palaeolithic cave art at El Castillo and La Pasiega in Cantabria. As a citizen of an enemy country, he could not return to his post at the Institut de paléontologie humane in Paris, where he had worked alongside Breuil under the direction of Marcelin Boule since its establishment (under the patronage of Prince Albert of Monaco) in 1910 (Breuil and Obermaier 1913, 1914a, 1914b). Obermaier had already fallen out with his old mentor Albrecht Penck, who had blocked his appointment at Vienna in 1909, and with Boule now turned against him,¹ he stayed in Madrid where he took up a post at the Museo Nacional de Ciencias Naturales. One of the key results of Obermaier’s enforced exile was El Hombre Fósil, first published in Spanish in 1916, and translated into English in 1924 under the inappropriate title of Fossil Man in Spain. The book certainly provided an essential summary of work in Spain, including a valuable gazetteer of known sites, but the book’s English title belied the true breadth of Obermaier’s vision.

Obermaier was now in full command of the Spanish, French and central European Palaeolithic records, which he was able to situate in the wider Eurasian and African context. This showed that in Europe the Chellean was restricted to Italy, Spain, England and France and was entirely absent from central and eastern Europe. He suggested that the Chellean had arrived from the south, through Turkey and the Levant, following the Mediterranean coast. Contemporary industries in central and eastern Europe completely lacked handaxes, however, those from Taubach and Krapina being prime examples (Figure 5.14). Obermaier now called these non-handaxe assemblages ‘pre-Mousterian’ and considered them to contain the germ of the true Mousterian industries that later spread west.

Figure 5.14 Pre-Mousterian flake tools from Taubach. Scale = 5cm. Note the diminutive size (after Adrien de Mortillet 1920).

The Acheulean had a similar distribution to the Chellean. It was not totally absent from eastern and central Europe, but Acheulean handaxes were remarkably rare, found only at a handful of sites in Germany (Hundisburg in Saxony-Anhalt; Kösten and Klause Caves in Bavaria) and in Poland (Mount Smardzewitz, Korytanja, Miechow). The material from Markkleeberg near Leipzig, which included scrapers, handaxes and Levallois, had also been assigned to the Early Acheulean or even Chellean by the excavator Jacob-Friesen (1911, 1913), although the latter attribution did not sit well with the cold fauna (including mammoth) from the site. Breuil, Commont and Obermaier attributed Markkleeberg to the Mousterian (Grahmann 1955).

Obermaier (1916) postulated that the Acheulean may have spread from several centres: a western Acheulean extending from North Africa, Spain, France, England and Italy; a southern Acheulean, essentially similar to the western variant, that probably extended through the Levant into the Balkans, Hungary and Poland; and an eastern Acheulean, coming from Asia and represented by a few very late, almost proto-Solutrean (i.e. leaf-shaped and thin), bifacial tools from Germany (Figure 5.15). The western Acheulean reached its apogee with the miniature forms at La Micoque. Concerning the makers of these various industries, Ober-maier equated the Pre-Chellean with Homo heidelbergensis of Mauer, the Chellean with the infamous and entirely fraudulent Piltdown skull (Dawson and Woodward 1913, 1914; Keith 1915; Weiner et al. 1953), and the Acheulean, pre-Mousterian and Mousterian with the Neanderthaloid race.

Figure 5.15 Obermaier’s Cultural Geography. Obermaier’s distribution and migration routes for Early Palaeolithic industries of Europe. Dots indicate the zone of Chellean industry with handaxes (coup de poing). Dashes indicate the zone of pre-Mousterian industry with no handaxes. Arrow 1 – western Acheulean. Arrow 2 – southern Acheulean. Arrow 3 – eastern Acheulean. Note the question mark over SE Europe (after Obermaier 1916).

Obermaier’s approach introduced the notion that different cultural zones existed in the Lower Palaeolithic (a term he was using by 1909 to include the Mousterian and everything before it), each representing contemporaneous populations of ancient humans with different sets of stone tools. There was a clear geographical patterning to the record too, some groups having their origin in the south, some in the east. Climate change during the Mousterian had had a marked influence on European populations: open air camps became scarcer and groups more scattered, seeking the shelter of caves. They became less mobile and more isolated, rarely encountering each other. This led to the development of many local industries, making the classification of the Mousterian a very difficult matter (Obermaier 1924, 79). Furthermore, as it originated in the east from the pre-Mousterian, the Mousterian reached Spain relatively late, creating an ‘odd’ mixture of fine small tool with handaxes and large ‘primitive’ flakes. Premature appearances of techniques and atavistic cultural survivals were also seen. By contrast the Chellean and Acheulean were southern industries that had appeared earlier in Spain than further north (Figure 5.16).

Figure 5.16 Obermaier’s Cultural Chronology. Chronological succession of Palaeolithic industries in southern, western and central Europe, and the contemporary faunal and geological phases of the late Pleistocene. It does not show the relative duration of these phases. Warm fauna is indicated by horizontal lines, cold fauna by vertical lines (image and caption after Obermaier 1916).

Breuil had already introduced these new concepts to the Upper Palaeolithic. At the 14th International Congress of Prehistoric Archaeology and Anthropology, held in Geneva in 1912, Breuil (1913) had argued that the time when archaeologists could look to simple linear models for cultural evolution was over. As the evidence grew, it became more obvious that regional differences existed and that the archaeological record did not show the same synchronic development everywhere. Rather, it showed the existence of numerous regional populations who had interacted with each other and influenced each other through trade, assimilation or invasion. None of the Upper Palaeolithic industries were indigenous to Europe, rather all had come from the contiguous continents of Asia and Africa, where their origins should be sought. Their arrival in Europe also marked the appearance of modern humans, and the disappearance of the earlier Lower Palaeolithic/Neanderthal lineage. Breuil thought that these latter groups had shown a similar development across the Atlantic and Mediterranean zones, although reversals and reoccurrences complicated the picture, the Mousterian with ancient fauna at Grimaldi and the quartzite handaxes from the upper Mousterian at El Castillo being just two examples.

The two friends had clearly learnt a lot from each other during their time working together at the IPH, cross-fertilising ideas derived from both the French and German traditions, and from later periods of prehistory (e.g. Kossinna 1911). The war did not break their friendship or their collaboration. Like Obermaier, Breuil spent much of it in Spain, sometimes doing archaeological research, sometimes working as an auxiliary to the naval attache in Madrid, and sometimes wandering the country disguised as a sailor (Brodrick 1963). He would not, though, turn his attention fully to the Lower Palaeolithic for another decade.

In North America, such ideas found an early champion in Henry Fairfield Osborn (1857–1935), president of the American Museum of Natural History and research professor in zoology at Columbia University, New York. Osborn’s (1915) Men of the Old Stone Age was based on a series of lectures given at the University of California in 1914, which had themselves been precipitated by a three-week tour of France guided by Obermaier, Breuil and Cartailhac. Osborn aimed to provide a popular account that gave the subject a more historical treatment, a synthesis of “geology, palaeontology, anthropology and archaeology, a correlation of environments and of human events in the European ice age” (1915, viii). He drew widely on Commont and his two IPH guides, as had Sollas (1911), but also adopted materials from Schmidt, Bourlon and others. What Osborn failed to take from his European colleagues was the sense of developing geographical variation, his Lower Palaeolithic sequence is a unilinear pan-European one, the major disjuncture being the arrival of modern humans and the Upper Palaeolithic. He followed Obermaier in assigning the invention of Levallois to the Late Acheulean, seeing it as a ‘masterpiece’ emerging from handaxes (as did Mortillet) or large Chellean flakes. Yet he followed Schmidt in assigning Taubach and other German sites to the Acheulean, missing the regional variations detected by Obermaier (1916). This remained true in all subsequent editions of his book (Osborn 1916, 1918).

In Oxford, Sollas’s second edition (1915, preface) declared that cultural views considered heretical four years earlier had now become orthodox, noting that at the 1912 Prehistoric Congress in Geneva, there had been much talk of migrations (e.g. Breuil 1913), but very little about indigenous developments. Bury (1916, 177) and Moir (1918) followed Commont, wondering whether the Acheulean and Mousterian were made by different races who might have overlapped, the newcomers initially adopting the handaxe before finding it not to their liking (Moir 1918, 508). Moir (1916) further traced the origins of the Mousterian and Acheulean to two different types of eolith. By the third edition of Ancient Hunters, Sollas (1924) had fully integrated Obermaier’s and Breuil’s conclusions concerning the ebb and flow of people and industries during the Lower and Upper Palaeolithic. Burkitt (1921) was teaching the same approach at Cambridge, and for students of Professor Macalister at University College Dublin ideas of migrations, technological convergence, the extermination of the Mousterians (Neanderthals) by Aurignacian (modern human) migrants, and the need to treat each region on its own merits, had by been part of standard undergraduate teaching for a number of years (Macalister 1921), although he disapproved of adjectival terms such as Chellean, thinking them arbitrary and meaningless. He suggested replacing them with the equally unhelpful Lower Palaeolithic I and Lower Palaeolithic II, following the practice in the Cretan tombs. Lewis Abbott’s (1911) main contribution to the understanding of Levallois technology went similarly unremarked. Abbott proposed that a Levallois core should be called a ‘Prestwich’ and the Levallois flake an ‘Evans’, and suggested that they had served as tokens used to seal a social transaction or bargain; perhaps by husbands and wives who would each retain a Prestwich or its Evans and who would refit them after a prolonged period of separation to recognise their spouse(s).

The End of the Line for Linear Models

Trigger (1989, 145–147) made the argument that inherent theoretical weaknesses in unilinear evolutionary models, particularly of the sort that dominated Palaeolithic archaeology for the first 60 years of its development, almost inevitably brought about their demise. Advocates of these Eurocentric models had no direct way of dealing with geographical or environmental variation in stone tool industries. Time was the only variable, placing the artefacts in the correct chronological order the only mission, regardless of any geological or biological implausibility this might entail. In the absence of documentary evidence, the diversity of the ethnographical record had been taken as “the matrix for archaeological data”, but this had primarily been used to construct a series of proxy evolutionary stages which, when arranged from the simplest to most complex, “recapitulated the sequence through which European societies had evolved” (Trigger 1989, 146).

The main role of Palaeolithic archaeology had thus become to demonstrate that cultural evolution had occurred, to chart its progress and to assign the various industries to their relevant stage of development, sometimes from eolithic beginnings (ibid.). Artefacts merely illustrated the past, in an antiquarian sense, rather than provided the tools for an understanding of human societies in evolutionary context. The ethnographic record supplied all one needed to know about how people at each stage of civilisation might have lived, but there were no systematic attempts to correlate the ethnographical and archaeological tool-kits, and comparisons were impressionistic or alarmingly direct, even from those adopting the nascent culture-historical models (ibid., 146). Sollas (1911, 1915), for example, directly equated the Tasmanians and Australians with the Lower and Middle Palaeolithic periods. Obermaier (1916) described Chellean and Acheulean groups as nomadic tribes, who lived by the chase and camped in the open air close to rivers. Gravels supplied raw materials for a range of simple weapons and tools of stone, some of which were used to produce wooden implements. Fires were lit for protection at night. The animals were fierce, but humans were canny. Macalister (1921, 282) distilled the mode of life and psychology of Lower Palaeolithic humans to three propositions:

1. They were fully developed humans, capable of progress in such arts as had evolved.

2. They lived on the product of the chase.

3. They dwelt in the open air, on the plateaux and not far from rivers.

And three possibilities:

1. They were acquainted with the use of fire, and with means for its artificial productions.

2. They built huts of boughs and hides.

3. They trapped larger animals, as the weapons at their disposal were inadequate for direct ‘warfare’.

Anything beyond this was for the novelist, chided Macalister (1921, 283), who derided the claim made by Rutot at the 1910 Prehistoric Congress that there had been a handaxe cult in the Lower Palaeolithic, based on the presence of giant and miniature forms that served no obvious function and which must, therefore, have been for worship, as having “nothing to recommend it” (Macalister 1921, 282; Dubus 1911). The Palaeolithic record may have been arranged to the satisfaction of older French, British, Belgian, Spanish and maybe German savants, but there was “a growing sense of impasse and sterility” (Trigger 1989, 147). Sollas’s rope of sand would have to be spun.

Note

· 1 Breuil and Obermaier had never got on particularly well with Boule, the senior scholar possibly feeling somewhat threatened by the dynamic and unmanageable pairing (Brodrick 1963). When Breuil returned from his trip to South Africa with 54 cases of ‘stones’, Boule refused to house them at the IPH. Paul Rivet, professor of anthropology at the Museum d’Histoire Naturelle found Breuil a room in one of his laboratories, until they were eventually moved into the new Musee de l’Homme at the Trocadero, Paris in 1937.