4

The Fall and Rise of the Acheulean, 1873–1900

Materiaux and Method

In the decade after the events at Abbeville and Amiens three different approaches to understanding the Palaeolithic rose to prominence: Evans’s technological method, Lubbock’s ethnographical path and Mortillet’s cultural chronology. They were highly complementary, and all would continue to influence archaeological method and theory for the remainder of the century, but it was the question of chronology, fundamental to any understanding of the past, that would dominate debate. In France this would revolve around archaeological sequences, whereas in Britain the most heated arguments concerned glacial stratigraphy and its bearing on the antiquity of humanity, these national differences reflecting the nature of the evidence available in each country.

The March 1876 issue of Materiaux pour l’Histoire Primitive Naturelle de l’Homme, now in its second series and edited by prehistorian (and later pioneer of parietal cave art research) Emile Cartailhac (1845–1921), published a list of Palaeolithic sites then known from Europe (p124–135; Table 4.1). The list was incomplete and biased towards France. None of the known sites in Spain were mentioned, and the corpus of 11 English sites was confused and partial: Evans (1872) alone had recorded 50 drift sites, four cave sites and over a dozen surface sites.

Table 4.1 The number of known Palaeolithic sites, published in the 1876 edition of Materiaux pour l’Histoire Primitive Naturelle de l’Homme.
	Tertiary	Drift	Surface	Caves	Total
France	6	44	144	489	683
Belgium	0	2	0	47	49
England	0	11	0	0	11
Italy	1	0	0	8	9
Germany	0	1	0	2	3
Switzerland	0	0	0	3	3
Total	7	58	144	549	758

There are nonetheless some interesting trends. More than 530 cave and drift sites had been identified in France since 1859, an average of 30 per year. They had been found in 136 communes, representing 32 departments (Mortillet 1876). Cave sites outnumbered drift sites by 10:1, the latter noted as sometimes containing only a single axe. This bias was also apparent in Belgium, Switzerland and Italy, but not in England, where drift sites outnumbered cave sites by 12.5:1 (by Evans’s figures). Although regarded as an age of discovery, outside France the rate at which new finds were reported was remarkably slow; in England an average of only three drift sites were reported each year between 1859 and 1872 (the date of Evans’s synthesis), in Germany an average of <1 site was found every five years, if we accept the journal’s figures. This trend was mirrored in the small and, frankly, disappointing number of articles published on Palaeolithic topics in each of these years (see McNabb 2012a).

It is therefore not too difficult to understand why national research biases arose. In little over a decade, caves had gone from being repositories of dubious worth to archives of unparalleled value. In France, not only were caves fairly abundant, but they also offered greater rewards for relatively little effort. Caves were finite in size, they produced more artefacts per unit of sediment and they were dug for research rather than commercial reasons by men (and they invariably were men) specially employed for the purpose. They contained well-stratified sequences of Pleistocene animals and different prehistoric industries that showed increasing complexity and sophistication over time, preserved flint working areas, food refuse, needles for making clothes, evidence of art, mystical objects and abundant human remains (e.g. L. Lartet 1872; Massénat et al. 1872). They were prehistoric homes that had been occupied exclusively by humans (Evans 1872), and discoveries of novel household items were reported with pleasing regularity.

The drift, by contrast, was an immense and poorly understood repository. There was no lack of new discoveries but there were few surprises. The range of artefacts and mammals was limited and repetitive, and their archaeological value often depended on the observations and honesty of the quarrymen who found them (see Evans 1872; Mortillet 1883 on the problems of forgery). There weren’t even any French glacial deposits to break the monotony. The major debates of this era in France belonged to Mortillet and his critics, and were largely methodological, testing not only the validity of his arrangement, but the methods by which that arrangement was constructed. Mortillet responded to each new attack with rebuttal or revision, but his belief in the both the veracity of the approach and the basic sequence was unshakable.

The British, however, had few caves, and most of those had already been excavated by earlier workers, often rather badly. British caves were also far less rich than their European counterparts, and new discoveries such as the series of virgin caves at Creswell Crags, first excavated in 1875 and which Boyd Dawkins had hoped would rival Les Eyzies, only ended in disappointment and acrimony amid accusations of fraud (Dawkins 1876, 1877; Mello 1875, 1876, 1877; Heath 1879, 1880; White 2017). With much of their evidence coming from flu-vial contexts and falling into a single epoch, the British naturally had little use for Mortillet’s scheme, and instead concentrated their efforts on understanding the complexities of the drift deposits and the Palaeolithic implements these contained. When not geological reportage, much of the archaeological literature published in Britain during this period steered a course close to the technological approach of Evans or the interpretative framework of Lubbock, a safe observational method that courted little controversy, and a more interpretative or imaginative approach that might raise eyebrows but which did not easily lend itself to empirical testing.

The rest of this chapter examines the period from 1872 until the end of the nineteenth century. It was certainly a period of entrenchment for some of the major savants, and somewhat pedestrian in terms of the development of archaeological theory, but when examined on a multinational scale, it was far from a time of interpretative stagnation, as I hope to show. The focus, of course, remains on Britain and France, two colonialist countries that between them controlled much of Africa, the Indian subcontinent, the Near East, south-east Asia and Australasia, and which would intellectually dominate Palaeolithic research until the 1960s (Dennell 1990).

Another British Antiquity Debate

The British were lukewarm to Mortillet’s scheme and only rarely used it. For them, the validity of the epochs, as chronological zones, had never been satisfactorily verified by independent means such as geology or fauna. The archaeological and stratigraphical differences between the drift implements and later cave series was clear enough, but the distinction between some of the cave epochs was no greater than that between two tribes of modern ‘eskimo’ (Dawkins 1874). They might simply reflect local differences: tribal customs, activity facies, or the relative abundance of resources, such as antler (Dawkins 1874, 353, 1880, 201–202). Levels of sophistication or crudeness were likewise not a certain indication of age. With few caves from which to sort out the problem for themselves, and with those that they had providing only an imperfect record marred by taphonomic and recovery issues,¹ the British generally paid lip-service to the epochs before quickly locking them up in yet another suspense account (Evans 1872; Lyell 1873; Dawkins 1874).

Indeed, in the fourth edition of the Antiquity of Man, Lyell (1873, 135) managed to dispense with the French archaeological method in little over a hundred words:

M. Mortillet has grouped the various cave deposits under four heads, classifying them according to slight differences in the forms of the flint tools, which he pointed out to us in the museum at St. Germain, and by the number and character of bone implements. But the differences in the fashion of the flint is not great, and is less apparent on the ground where all the rough specimens are also present, than when a selected series is examined. The absence of the forms of the stone implements [handaxes] in what he considers the newer deposits, and the rarity of bone instruments in the older forms the chief distinction.

Either the epochs were too well known to mention by name (which seems unlikely) or Lyell thought them not worth naming. Dawkins eventually accepted the Acheulean in his Early Man in Britain (1880, 199), but used it only once, when defining it. He rejected the other epochs.

For the British, understanding the chronology of the Pleistocene depended on geology and fauna, not artefacts. Geologists more or less agreed that the Pleistocene drift deposits had been laid down by river action, but how the fluvial beds related to the history of valley formation and to each other were still matters for debate (for example, had the drift filled older pre-existing valleys from bottom to top or were the valleys gradually down cut and the drift deposited in stages from top to bottom; see Prestwich 1866; Dawkins 1867; Wood 1867, 1868; Tylor 1868a, 1868b; Flower 1869). Evans (1872) had devoted his final chapter to the issue, repeating his tour of English rivers, this time seeking to show how the nature and geomorphology of all the known implementiferous beds matched Prestwich’s (1864, 1866) hypothetical terrace formation model and preferred postglacial age. The late Hugh Falconer (d. 1865), though, had long believed the faunal succession, once properly identified and in the right stratigraphical order, held the key to understanding the chronology of the Pleistocene and humanity’s place within it. This vision was taken up with enthusiasm (and not a little belligerence) by William Boyd Dawkins, who provided nineteenth-century British archaeologists with an evolving but oft-disputed faunal framework for understanding the Pleistocene occupation of Britain (see White 2017 and O’Connor 2007 for details of the various geological debates involving Dawkins and others).

Dawkins (1874) rejected Lartet’s faunal scheme for the same reasons as had Mortillet, because its signature species were not temporally restricted marker fossils, but animals that had co-existed over long periods of time. Variation in the relative frequency of different species could relate to several issues, including human prey selection; mammoth, woolly rhinoceros and reindeer did not naturally die in caves. Dawkins took a broader view. Between 1867 and 1880 he published several revisions to a biostratigraphical scheme for the British drift and cave faunas, amassing a database of some 150 faunal assemblages and clarifying the status and nomenclature of several taxa (White 2017: Tables 4.2 and 4.3). For most of his career he stuck to the basic three-fold sequence proposed by his Oxford mentor John Phillips – preglacial, glacial and postglacial – generally dealing with complexity by lumping sites and species together (1867, 1874) or subdividing them within the basic framework (1869, 1874). His postglacial faunas showed a curious mixture of warm-adapted animals such as hippopotamus and cold-adapted species such as reindeer and musk ox. Lubbock (1865) thought the association to be illusory, and that the Arctic and African animals had occupied Britain during different periods with different climates. But if this were the case, then geologists should expect the different groups to occur in discrete strata, a situation unknown in Britain or the Continent, according to Dawkins (1874). Dawkins’s own interpretation envisaged a more strongly continental climate for ‘postglacial’ Britain, biting winters and scorching summers, with hippopotamus and reindeer making seasonal migrations from their different ‘headquarters’ in southern Europe and the Arctic (see also Lyell 1863). Evidence for humans in the Pleistocene was only ever found associated with the postglacial fauna. The first humans must thus have entered Britain after the Great Glacial. In this Dawkins agreed with Prestwich, Evans, Whitaker (e.g. 1889), Lyell and most other British workers. A postglacial antiquity for humans in Britain thus became the accepted view.

Table 4.2 Dawkins s faunal divisions for the Pleistocene. Entries in italics mean that the fauna included humans. Following geological convention, time runs from bottom to top.
1867	1869	1874	1880
	Main postglacial series	Late Pleistocene Division	Late Pleistocene Division
Postglacial fauna	Transitional postglacial fauna of the Thames Valley	Middle Pleistocene Division (upper)	Middle Pleistocene Division (upper)
Great glacial	Great glacial	Great glacial	Great glacialWith mid-glacial warm phase
Preglacial fauna of the Thames Valley	Transitional preglacial fauna of Clacton	Middle Pleistocene Division (lower)	Middle Pleistocene Division (lower)
Preglacial fauna (Pliocene)	Preglacial fauna (?Pliocene) of the Norfolk Forest Beds	Early Pleistocene Division	Early Pleistocene Division

Table 4.3 List of characteristic preglacial and postglacial fauna, according to Dawkins 1869.
Characteristic Preglacial Fauna	Characteristic Postglacial Fauna
Extinct Etruscan Bear (Ursus arvernensis)	Palaeolithic humans
Cave Bear (Ursus spelaeus)	Wolverine/Glutton
Shrew	Cave Bear? (Ursus spelaeus)
Desman	Grizzly Bear?
Water Vole	Cave Lion (Panthera spelaeus)
Mole	Cave Hyaena (Crocuta crocuta spelaeus)
Roe Deer	Panther?
Red Deer	Musk-Ox
Extinct Deer (Cervus sedgwickii)	Woolly Rhinoceros (Coelodonta antiquus)
Extinct Deer (Cervus ardeus)	Woolly Mammoth (Mammuthus primigenius)
Aurochs (Bos primigenius)	Lemming
Hippopotamus major	Cave Pika
Horse	Pouched Marmot
Extinct Rhinoceros (Rhinoceros megarhinus)	Ground Squirrel/Marmot
Extinct Etruscan Rhinoceros (Rhinoceros etruscus)	Reindeer
Straight-Tusked Elephant (Palaeoloxodon antiquus)	Bison
Southern Mammoth (Mammuthus meridionalis)	Hippopotamus major
Beaver
Extinct Giant Beaver (Trogontherium cuvieri)

Dawkins vociferously defended the postglacial antiquity of humans in Britain (e.g. 1878, 1880), but in truth humans were rather epiphenomenal to Dawkins’s reconstructions of the deep past, his tableau basically being wildlife scenes, and he never strayed far from the Evans/Lubbock approach to the archaeological record. His first essay on the topic (Dawkins 1866) is a close homage to Pre-historic Times, dividing the Palaeolithic record into flint folk (older) and reindeer folk (younger) and speculating on the conditions of Palaeolithic life (tools, hunting, clothing, architecture, art and religion) as informed by archaeology, geology and ethnography. His few original interpretations, such as the notion that north, including the Somme, Seine and Thames, formed part of a common hunting ground for ‘drift-folk’ (Acheulean hominins), whereas the caves sites of the west and south-west were the hunting grounds of a different tribe who used different toolkits, came from biogeographical and palaeogeographical reconstructions first and foremost, supported by ethnographical analogies derived from indigenous North American tribes (Dawkins 1874). In fact, the many controversies he entered were more to preserve the sanctity of his faunal divisions and their interpretations than any real objection to preglacial humans. He nonetheless added to the growing number of self-imposed obstacles to progress faced by archaeologists and geologists of the period.

This was well exemplified by the controversy surrounding the findings at Victoria Cave near Settle, North Yorkshire, between 1870 and 1878 (Figure 4.1). Dawkins had originally directed the Victoria Cave excavations but in 1874, shortly after George Busk had identified a human bone from the cave (Busk 1874), he transferred this responsibility to Richard Hill Tiddeman (1842–1917), a BGS geologist whose research area included Yorkshire. Whereas Dawkins found only one postglacial faunal suite in the cave, Tiddeman identified two distinctive faunas – a lower, warm series that included hippopotamus and hyaena and an upper, cold-adapted group containing reindeer and horse (Tiddeman 1873, 1875, 1878). The two cave earths containing these different mammalian groups were separated by a layer of boulder clay and 4m of laminated clay, making the lower fauna preglacial and the upper fauna postglacial. The human bone was associated with the lower.

Figure 4.1 Ursus sapiens? Top: Victoria (C) and Albert (B) Caves, Settle, Yorkshire. Bottom: the fibula from Victoria Cave (Figures 1 & 2), identified by Busk as human (Figure 3 as comparison), and by Dawkins as a bear (after Dawkins 1874).

There was much for Dawkins to dispute here, including species identification (he thought the ‘human’ fibula was that of a bear), the supposed glacial origin of the clays, and the stratigraphic integrity of the two strata and their contained faunas. Dawkins believed them to be one and the same layer exposed at different heights inside and outside of the cave (Figure 4.2). Worse still in Dawkins’s view, the separation of the warm and cold elements of the postglacial fauna into different groups of different ages added unwelcome credence to the controversial claims of Scottish BGS geologist James Geikie (1839–1915). Geikie believed that the geological record of Europe showed evidence of not just one period of intense cold – a Great Glacial – but several glacial periods separated by warm ‘interglacials’ (Geikie 1874, 1877a, 1881; see also Croll 1864, whose mathematical models had predicted the existence of multiple glacials and inter-glacials). Geikie therefore thought Dawkins’s use of the term postglacial not only misnominal but believed the fauna it contained came from different warm and cold phases of a complex and cyclical Ice Age. Only the modern wild fauna, that which settled after the last glaciation, was truly postglacial. Geikie was also highly critical of Dawkins’s seasonal migration theory, thinking it ethologically ridiculous (hippopotamuses simply do not migrate thousands of miles each year) and based on numerous false climatic and meteorological premises (Geikie 1872a, 1872b, 1873, 1874, 1877a, 1881).

Figure 4.2 Glacial wrangling: Dawkins’s (top) and Tiddeman’s (bottom) sections through the deposits at Victoria Cave. The two geologists had different interpretations of the relationship of the deposits and the age of the contained fauna (after Dawkins 1874; Tiddeman 1878).

Dawkins had clashed with Geikie on many occasions (see White 2017). To quash the further spread of the Scotsman’s interglacial/preglacial heresy, Evans, Dawkins and Thomas McKenny Hughes² (1832–1917), all mono-glacialists and all voices of authority, staged a conference on the ‘current state of the Antiquity of Man question’, which took place at the Anthropological Institute in May 1877. It focussed on the question of Victoria Cave, but only Tiddeman had been invited to speak in favour of a preglacial age, and by the time he finally took the stage (last) his two opponents and the chair had undermined any case for preglacial humans or animals he might have hoped to make (Evans 1878; Dawkins 1878; Hughes 1878; Tiddeman 1878). Dawkins also took the opportunity to point out a number of fallacies in Geikie’s most recent arguments (Geikie 1877a), while Hughes acknowledged the evidence from Thetford and Brandon in East Anglia, where Survey geologist Sidney Skertchley (1850–1929) had found stone tools in brickearth both above and beneath boulder clay, but placed it in his own suspense account pending verification. When Busk later changed his mind on the putative human fibula, admitting that it probably was a bear, the case (and the BAAS funding for further excavations) collapsed (Tiddeman 1879). Dawkins and Evans were vindicated: humans were officially postglacial arrivals in Britain, in the accepted sense of the term.

Geikie (1877b) wrote to the journal Nature complaining that nothing new had come from the ‘conference’, where the same people had defended the same hackneyed theories to the detriment of more progressive views, all on the dubious grounds of ‘due caution’. And while Evans was famously cautious, Dawkins was anything but. He had, in fact, toyed with the idea of preglacial humans on several occasions, his evolving understanding of the Thames Valley sites (which by 1880 had produced abundant artefacts) suggesting that the same suite of Middle Pleistocene animals, including humans, might have come and gone during preglacial to postglacial times in response to climate. He would later (Dawkins 1880) concede that humans might even have made fleeting incursions into Britain during a brief warm interval within the Great Glacial (an ‘inter-glacial’ although not in the sense of Geikie), largely to accommodate the new evidence from High Lodge (Skertchly 1877; Miller and Skertchly 1878) where it was realised that both the scraper and handaxe assemblages were sandwiched between two glacial boulder clays. All the time, however, he was really defending his faunal divisions.

Dawkins, Evans and Hughes had all made major contributions to the new science of Palaeolithic archaeology, but they had become entrenched behind a new orthodoxy of their own making, one every bit as stubborn and long-lasting as that it had replaced. Instead of embracing complexity as the evidence grew, this cohort of elite geologists perpetuated a compressed Pleistocene framework that correlated the entire Palaeolithic with just two ecologically confused fauna suites and left little room for detecting patterns in the archaeological record. It is hardly surprising that none were found. Prestwich (1887), for his part, was more receptive to the possibility of pre-glacial humans, as evidenced in the high-level gravel of the Somme and at Reculver (Kent), Victoria Cave and High Lodge, but not in the sense of their being present long anterior to the glacial. He thought humans might have arrived immediately before the glacial, when ice was advancing but had not yet reached its maximum extent, suggesting that this might more correctly be termed a glacial or mid-glacial occupation. He was still, however, reluctant to accept age estimates that placed the Glacial period between 240,000 and 80,000 years ago, preferring a much shorter chronology. If humans were glacial, then they appeared, at most, 30,000 years ago, if postglacial, no more than 15,000 years ago (ibid., 408). Unsurprisingly, Prestwich failed to convert Evans, Dawkins or Geikie to this particular cause, while in France Marcellin Boule (1890) also confessed to being unconvinced. The stagnation in British thinking began at the top, and there was little room in the social structure for it to change from the bottom, or the peripheries.

Geikie (1881) would soon write his own Prehistoric Europe, in which humans first arrived at the close of the Pliocene, when conditions were warm, and were subsequently present during at least three separate interglacials. Geikie provided a wealth of geological, floral, faunal and climatic information spanning the Palaeolithic to the Iron Age. He rejected Dawkins’s homogenising approach, preferring Lartet’s subtler faunal divisions and arguing that communities of animals were controlled by prevailing climate, not human hunting preference. But he lacked the data to summarise conditions in different interglacials, just managing a generic one, and for the same reasons made no attempt to map the archaeology onto different interglacials (he even failed to mention Mortillet’s epochs by name, merely noting their existence). Geikie’s theories would not be fully vindicated until the early twentieth century, when Albrecht Penck (1858–1945) and Eduard Bruckner’s (1862–1927) sequence of four glacial-interglacial cycles in the Alps would gain unstoppable momentum (see Chapter 5). Until then the formally received view of authorities such as Evans and Dawkins would remain dominant in Britain, despite murmurings of dissent from the ranks (e.g. Smith 1884; Shrubsole 1885), and while Dawkins would eventually be forced to accept that humans were present in preglacial and glacial Britain (e.g. Dawkins 1894), he remained a monoglacialist (Dawkins 1910), thinking the evidence of multiple glaciations in the Alps (and Scotland) to be of local relevance only (see Chapter 5).

Tertiary Distractions

The French were far less doctrinaire on the issue. Mortillet (1873) reasoned that the Acheu-lean appeared before the Great Glacial in France, because it was associated with hippopotamus and straight-tusked elephant (Palaeoloxodon antiquus, then known as Elephas antiquus), two ‘southern’ warm-adapted species that belonged to the terminal Pliocene or earlier part of the Pleistocene (i.e. Dawkins 1869, 1874). This, however, marked only the beginning of the Palaeolithic. Acheulean handaxes were fairly sophisticated objects, and it was deemed unlikely that they represented the earliest stone tool technologies (e.g. Evans 1872; Ameghino 1880). The human fossil record, derived exclusively from caves, also contained essentially modern skeletons, plus a few archaic looking yet clearly human specimens, as found at Engis in Belgium, the Neander Valley of Germany and Gibraltar (Huxley 1863; King 1864; Busk 1865; Quatrefages and Hamy 1874). The ‘flint folk’ of the drift were still unknown as fossils (bar the Moulin Quignon fake), although Dawkins suspected they were a different race entirely, “as extinct as the mammoth and woolly rhinoceros with whom they lived” (Dawkins 1866, 338). Theoretically, both Darwinian natural selection and a progressive view of culture expected to find less human-like fossils and cruder pre-Palaeolithic tools deeper in time. While Africa or Asia were both fancied as the ultimate cradle of humanity (Wallace 1870; Darwin 1871), the possibility that humans would be found in the Tertiary period of Europe remained very real.

The list of sites published in the 1876 issue of Materiaux contained seven locations attributed to the Tertiary period, six in France and one in Italy. As early as 1863, Desnoyers had claimed to have evidence for Pliocene human occupation in France, in the form of cut-marked bone from St Prest, Eure-et-Loir (Desnoyers 1863). Human antiquity was pushed still further back in time by Abbé Louis Bourgeois’s (1819–1878) flakes and burnt flint ‘splinters’ from Thenay, Loir-et Cher, which had been found in Miocene deposits (Bourgeois 1868, Figure 4.3). At the 1872 Congress of Anthropology and Prehistoric Archaeology in Brussels, the same meeting at which Mortillet had announced his Acheulean epoch, Bourgeois (1873) presented the evidence from Thenay before a special commission that included Carthailhac, Quatrefages and Franks. Nine of the 15 panel members accepted the integrity of the deposit and a human origin for at least some of the finds.³

Figure 4.3 Eoliths. Putative Miocene and Pliocene stone tools and burnt flints from Thenay, France (top, bottom), and apparent flakes of the same age from the Tagus Valley, Portugal (centre). Scale = 5cm (after de Mortillet and de Mortillet 1881, drawings by Adrien de Mortillet).

Claims for Miocene humans from Otta in the Tagus Valley, Portugal (Ribeiro 1873, 1880), were investigated by yet another committee, convened at the 1880 Congress in Lisbon. The largely new membership,⁴ which included Evans and Mortillet, was similarly split on the anthropogenic origin of the objects and the supposed age of the deposits.

Neither could a consensus be reached on whether flints found in situ with an Upper Miocene fauna at Puy Courny (Auvergne), France (Ramès 1884), were natural or man-made. Boule, Evans and American archaeologist Henry Haynes (1831–1912) were unconvinced, but, crucially, Mortillet accepted this and all other claims as genuine, and in 1881 added a new period, the Eolithic (Dawn of the Stone Age) to the base of his influential sequence. It initially consisted of the Thenaisienne and Ottaienne (Mortillet and Mortillet 1881), but the latter was dropped (Mortillet 1883) and later replaced by the Puycournienne (Mortillet 1897b). He also proposed the names Anthropopithecus bourgeoisii and Anthropopithecus ramesii for the hypothetical missing links who made these eolithic industries. Gaudry wondered if they might have been made by the Miocene anthropoid ape Dryopithecus.

Despite Mortillet’s approval, there was strong opposition to eoliths in both France and Britain (e.g. Evans 1872; Magitot 1876; Dawkins 1880; Stirrup 1885; d’Acy 1885; Newton 1897). Dawkins (1880) dismissed the claims on evolutionary grounds. As there was no evidence for placental mammals in the Eocene, and no members of the animal community to which humans belonged in the Miocene, it was absurd to think that the highest form of animal could have existed alone, when all its conspecifics were absent. He also found the claims for Miocene artefacts from France unconvincing: if they were artificial they were probably the work of an anthropomorphic ape that had gone extinct or evolved into something else, a view he shared with Gaudry. He put up no such theoretical barriers to the presence of humans in the Pliocene yet, like Evans (1872), rejected existing claims because their geological contexts were simply not secure enough.

None of this prevented eolith-mania from spreading into Britain. During the later 1880s, Benjamin Harrison, a grocer from Ightham in Kent, began to collect both genuine drift implements and deeply stained and heavily worn eoliths from the High-Level Plateau gravel above the valley of the River Shode (Harrison 1928; McNabb 2012a; Figure 4.4). Harrison, who was well aware of the claims coming from Europe, believed the latter group to be very ancient artefacts. They were crude, exactly what might be expected at the dawn of human culture, and they came from gravels above 500ft O.D., deposits that could not be related to the postglacial drainage of the region. They must therefore belong to older Tertiary river courses.

Figure 4.4 Harrisonian eoliths (aka chocolate brownies, the ‘curved sides and pointed head’ type (after Harrison 1904). Scale = 5cm.

Over a period of several years Harrison also managed to convince Joseph Prestwich that eoliths were genuine artefacts, although the latter would only go as far as declaring them ‘possibly’ preglacial (Prestwich 1889, 1891, 1892). The eolith ‘cult’ (Dawkins 1910) quickly gained popularity among a new breed of enthusiast, many of them friends of Harrison (e.g. Abbott 1894), while Dordogne veteran Rupert Jones (1894) joined Prestwich among Harrison’s more influential supporters. Theirs was the minority view, however, and other influencers including Evans, Dawkins and survey geologist William Whitaker (1836–1925) were far from convinced (see meeting discussions recorded in Prestwich 1889, 1891, 1892; Whitaker 1895). They thought the eoliths were naturally chipped stones and, given that the whole collection was made up of surface finds, believed that the genuine Palaeolithic arte-facts simply belonged to the drift period, although their elevation and condition was worthy of remark (e.g. Evans 1897; Newton 1897; Cunnington 1898). With opinion so strongly divided, Harrison’s eoliths became locked in another suspense account. When the British Museum eventually devoted display case 106 to Harrisonian eoliths, largely for completeness, the accompanying guidebook gave no guarantees concerning their authenticity (Read 1902; Smith 1911a; cf. Thompson 1892).

Eoliths and other ‘Pre-Paleolithic’ objects enjoyed a second flush of celebrity during the early twentieth century, when they were energetically promoted by Aimé Rutot (1847–1933), curator of the Museum of Natural History in Belgium, and by Sir Ray Lankester and James Reid Moir, respectively the ex-director of the Natural History Museum in London and a reluctant tailor from Ipswich (e.g. Rutot 1900a, 1900b, 1902, 1904, 1909, 1911; Abbott 1905; Moir 1910, 1911, 1919, 1927; Underwood et al. 1911). They were again met with vehement opposition, increasingly based on experimental evidence that eoliths were formed naturally and with the increasingly influential Marcellin Boule (1861–1942) and Henri Breuil (1877–1961) at their vanguard (Boule 1905; Obermaier 1905c; Warren 1905a, 1905b, 1923b, 1923c; Smith 1907, 1908; Commont 1909d; Breuil 1910; Haward 1912). At times the debate became personal and unpleasant, words being twisted to make the oppositions claims look as ridiculous as possible (e.g. Warren 1923b, 82). There was little room for compromise, “the arguments of each side [being] constructed upon such different premises that there was little engagement between their respective views” (O’Connor 2007, 146). What for some were just naturally shaped stones were for others stones selected by humans because they possessed a useful natural shape. The skeptic’s natural damage on a useless steep edge was the believer’s deliberate modification to a robust tool of unknown function. The two sides were just not listening to each other. They were not even having the same conversation.

The eolith debate is an interesting history lesson into nineteenth- and early twentieth-century constructions of humanity’s place in the history of life on Earth, and where the different phases of the Palaeolithic sat within that history. It was taken incredibly seriously by both sides, consuming much emotional and mental energy and for periods dominating the pages of learned journals and the outputs of individuals. Opinions turned not only on the strength of evidence but on the power of personality, reputation, authority, religion and relationships. Ultimately the whole episode was little more than a sideshow, a distraction from a genuine understanding of the distant past.

Advocates for the eoliths liked to cast themselves as the new Boucher, prophets ignored in their own countries, but unlike the late esteemed customs officer their claims were bogus. The debates did provoke a greater appreciation and understanding of flint fracture mechanics under different conditions, but they contributed little to contemporary understandings of the core themes of this book beyond the belief that as handaxes could not have sprung from nowhere there must have been earlier, more primitive expressions of human craftsmanship whose evolution could be traced though time. Further accounts of the various convulsions of this debate can be found in Grayson (1986), de Bont (2003), O’Connor (2007) and McNabb (2012a). Newton (1897) provides a comprehensive contemporary review of the supposed Tertiary lithics, human fossils and cut-mark bones, including the ‘cut-marked’ Miocene whale bone reported by Cappellini from Poggiarone, near Monte Aperto, Italy, quite reasonably suspected by many, including Evans, to have been gnawed by a carnivorous fish.

Ernest d’Acy and the Acheulean Under Siege

To remain current, any chronological framework had to be able to accommodate both new discoveries and refinements to known sites and sequences. For Mortillet, this meant that his cabinet might need tidying-up occasionally and might sometimes require a whole new layer to be added to the stack, but he seemed perfectly happy to accept this. What he could not easily tolerate, however, was mixing the contents of a drawer, so that it contained objects from more than one epoch, because that would undermine the ordered progression and unique character of each stage. If this were allowed to happen, then the value of the cabinet and its contents would be seriously reduced. For others, however, Mortillet’s chronology was less of a precious family heirloom (e.g. Mortillet and Mortillet 1881) and more of an experimental design that demanded independent testing (e.g. Sirodot 1875). With some seven years of experience of collecting at the St Acheul type-site, Jacques Louis Ernest Cadeau d’Acy (1827–1905) had the perfect testing ground while Materiaux gave him the perfect mouthpiece (d’Acy 1875).

Mortillet (1875a, 174) was very clear that handaxes were found only in the lower part of the sequence at St Acheul, where they occurred in abundance and to the exclusion of all other tools. Scrapers, Mousterian points and blades were only ever found in the higher strata. The same stratigraphical relationship was seen at lower elevations, in younger terrace deposits, although Mousterian tools were here relatively more abundant. This was true in the Somme at Menchecourt, and in the Seine at Levallois-Perret⁵ and Clichy; it was the same everywhere, in fact, Acheulean beneath and Mousterian above. The two epochs also had characteristic faunas, reflecting fundamental differences in climate (Mortillet 1877). There was a certain degree of overlap, mammoth, for example, being present in both epochs, but the Mousterian was associated with cold-adapted species such as reindeer, musk-ox and groundhog, which could not have lived under the same conditions as the warm-loving animals such as lion, straight-tusked elephant and hippopotamus found with the Acheulean (contra Dawkins and most of the British).

D’Acy (1875) had seen thousands of Palaeolithic flints from St Acheul, enough for him to know that all types of implement were found at all levels, from the top of the Chalk to the bottom of the coarse silt (where polished axes appeared). Handaxes were undoubtedly the most abundant type in his collection (n=385), but Moustier types formed almost a third (n=230) and were found in similar proportions in both the upper and lower layers. This was easily demonstrated because the flints from each layer bore a characteristic patina (d’Acy 1875), something Mortillet knew perfectly well. Even so, these data were biased in favour of handaxes, which commanded higher prices from collectors and were thus preferentially sought and retained for sale by quarrymen (d’Acy 1875, 282). With regret, d’Acy was forced to disagree with his erudite associate.

Although he felt obliged to use the terms Acheulean and Mousterian, which he suggested had already become axioms, d’Acy was a long way from recognising the primacy of the handaxe at St Acheul. In his experience, Acheulean and Mousterian types were contemporaneous and were found together at St Acheul, in the lower terraces of the Somme, and in the Seine, although he doubted whether the high- and low-level terraces of the two river systems were directly comparable. He was equally skeptical of faunal chronologies, having witnessed the ‘beating’ of Lartet’s scheme in France and the continued confusion over the age of the fauna at the supposed Pliocene site at St Prest. Mortillet’s fauna distinctions suffered from the same overlap and stratigraphical problems. Taking the example of the straight-tusked elephant, found associated with an Acheulean industry in the high- and low-level terrace of the Somme and with the Mousterian station of La Pecq in the Seine, he wondered how animals that were supposed to have become extinct in the higher terraces, could then reappear in the lower terraces? Like Evans and Dawkins in Britain, d’Acy thought it more likely the two assemblage types were the products of neighbouring contemporaneous tribes, constantly in communication, continually quarrelling.

Mortillet’s (1875b) short reply in the same journal is worth considering in detail because it reveals much about his working practices, some facets of which are rarely acknowledged. Having been warned by Cartailhac to stick to the facts and not play games, Mortillet summarised d’Acy’s attack in two propositions. The first, that there was no ‘Acheulean type’, had in Mortillet’s opinion been ably contradicted by d’Acy himself, who throughout his paper referred to two types, Acheulean and Mousterian. That the characteristic Acheulean handaxes varied greatly there could be no doubt, but they all shared a set of common characteristics, “a family resemblance” (Mortillet 1875b, 343). “When we talk about knives, who misunderstands?” demanded Mortillet (ibid.).

Knives vary infinitely, much more than the Acheulean type: sharp pointed knives, round-ended knives, pocket knives, table knives, kitchen knives, butcher’s knives, hunting knives, sacrificial knives, knives of wood, steel, silver, bronze, flint, etc. Does this prevent the knife from forming a real family, a natural distinct group, an easily recognisable type?

The answer to this rhetorical question was emphatically ‘non’, so why, he asked, should one apply different standards to Acheulean knives, that is, to handaxes?

The second and related proposition targeted another foundation stone of Mortillet’s model: that the Acheulean and Mousterian could not be distinguished solely on the basis of supposedly characteristic artefact types. Here, Mortillet focussed on facts and figures. The Saint-Germain Museum had 241 accessioned handaxes from St Acheul but only 29 ‘blades and flakes’; d’Acy’s own contribution of St Acheul material to the national museum consisted of eight flakes and 42 handaxes.⁶ Similarly, at Thennes, where d’Acy had trained the quarrymen to keep everything they found, he had been rewarded with 28 handaxes and not a single flake. In Mousterian stations the proportions were reversed. La Pecq produced nine handaxes and 44 blades, Montguillian five handaxes and 82 blades – but even here it is likely that handaxes were over-represented because of the biases to which d’Acy had already referred. Summing up, Mortillet noted that in Acheulean sites handaxes made up at least 80% of the tools, whereas in Mousterian ones they rarely constituted 20%, “a difference sufficient to distinguish the two eras, and which, I hope, will satisfy our dear Director of Materiaux” he sarcastically concluded (Mortillet 1875b, 344). So, taking a step back from exclusive types, Mortillet now looked to overall assemblage composition to define an industry. This is a highly significant development in Mortillet’s thinking, one that is usually associated with the much later work of Francois Bordes (see Chapter 6), and not entirely dissimilar to Hamy’s (1870) notion of a type.

Whether it satisfied Cartailhac is unrecorded but d’Acy was certainly not convinced. Extending his research to the plateau 15 km southwest of Amiens (D’Acy 1878), he found further evidence that Mortillet’s scheme was flawed. D’Acy concentrated on the silts that covered the high plateau and extended into the valleys, mantling both the high-level and low-level terraces. This blanketing deposit was certainly younger than the terrace gravels it covered, although it was not necessarily all of the same age or origin. Sites within the silt on the plateau at Hangard, Aubércourt and Demuin C yielded assemblages of handaxes, alongside flakes and scrapers (Figure 4.5). Demuin B produced only cores and flakes, a dozen of which he conjoined onto their parent block, in a very early example of archaeo-logical refitting (Figure 4.6). Demuin A produced just five handaxes, a core and a flake. D’Acy (1878, 15) concluded that

the relative frequency of one or the other of the two forms, called by M. de Mortillet the Saint-Acheul type and the Moustier type, has no correlation with the level of the diluvium in which it is observed… and can in no way be used to determine the relative age of the deposits.

Figure 4.6 Demuin B. A core and flakes from the silts at Demuin B, a site where d’Acy achieved some of the earliest archaeological refits (after D’Acy 1878). Scale = 5cm.

Figure 4.5 Scrapers in the Acheulean. Scrapers and handaxes together at Hangard (after D’Acy 1878). Scale = 5cm.

Mortillet’s response was short and sanguine (Mortillet 1878a). He refused to ‘surrender’ to d’Acy’s reasoning, repeating the fact that the age of a deposit could not be reliably inferred from a small sample, only a large one, deferring to d’Acy’s own collections on display at the Universal Exposition in the Trocadero as proof. He did not think the scrapers at Demuin C could be regarded as intentional forms, like the Mousterian racloirs, but saw them as flakes with retouch to a convenient edge. The presence of flint points might suggest the assemblage was transitional, but more information was required.

Chelles-Sur-Marne and the Acheulean in Exile

The Pleistocene deposits above the village of Chelles in the Marne Valley had been quarried since the building of France’s eastern railway line in the 1850s. By the late 1870s, seven quarries existed on the right-bank of the river, from which some 650,000m³ of aggregate had been removed (Chouquet 1878, Figure 4.7). Mammalian bones and palaeoliths from Chelles became widely known in 1878 through the work of local railway engineer M. Leroy, French prehistorian E. Chouquet and the Argentine palaeontologist Florentino Ameghino (1853–1911), who provided consistent reports of the site and its contents (Chouquet 1878, 1884; Mortillet 1878b; Ameghino 1880, 1881).

Figure 4.7 Chelles-Sur Marne 1879. Ameghino’s original location plan for the quarries just outside the village of Chelles, and his published geological section. The sequence shows (4) basal grey sand and gravel, 0.8–1m thick and often cemented by lime; (3) 3.5–4m pebbly gravel and brown sand; (2) 0.4–3m grey sand; and (1) topsoil, locally lying on a reddish sand layer (after Ameghino 1880).

The deposits at Chelles lay on a low terrace between 7 and 20m above the river and consisted of 5–7m of laterally variable Pleistocene sediments (Figure 4.7). The agglomerated basal sands of Bed 4 contained only Acheulean handaxes, flakes and blades. These were exclusively associated with an ancient warm-loving fauna that included straight-tusked elephant, Merck’s (or forest) rhinoceros (Stephanorhinus kirchbergensis), hippopotamus and the giant beaver (Trogontherium cuvieri) but the cold indicators such as mammoth and woolly rhinoceros, often found alongside handaxes at St Acheul, were absent. Bed 3 was largely sterile, having produced only a few handaxes from near its base, but the sand of Bed 2 yielded a Mousterian assemblage consisting of scrapers, points and regular blades but no handaxes. Fauna was less well preserved in this layer, but included mammoth, bovid and horse with no trace of the older thermophilous animals.

It was all perfectly clear. The first two epochs could be divided on the bases of stratigraphy, palaeontology, climate and artefacts. The makers of the St Acheul or ‘Chelles-type’ handaxes (a term used by Ameghino as early as 1880) had lived during an ancient period of warm, wet climate, surrounded by frost-sensitive plants and forest dwelling animals. The Mousterian, on the other hand, belonged to a more recent era of cold conditions, characterised by mammoth, woolly rhinoceros and other open-adapted species. Chouquet further correlated the basal deposits with the tufa site of Lacelle (La Celle), situated in the Seine Valley at the same altitude at Chelles, where the presence of fig, laurel, Judas tree, sycamore and willow added texture to the picture of humid conditions, with mild winters. In those quarries where one of the archaeological layers was missing, so was its characteristic artefact type. The only instance where overlap might be found was in the areas where Bed 3 had been cut out or ravined, and these were stratigraphically obvious (Chouquet 1878, 1884; Ameghino 1880).

Chouquet (1878) drew special attention to the fact that the evidence from Chelles conformed in every way to Mortillet’s (1877) most recent interpretation of the Quaternary period, a conclusion with which the famous curator, who had certainly visited the site by the Autumn of 1879 (at the latest, see Ameghino 1880), unequivocally agreed. Chelles was the perfect site and the perfect foil to d’Acy’s criticisms (Mortillet 1883, 138). Mortillet accordingly decided that Saint-Acheul was not pure, but contained a misleading mixture of material of the first two epochs. He thus renounced it as a type-site, installing Chelles in its place and adopting the name ‘Chellean’ for the first phase of human cultural evolution (Mortillet and Mortillet 1881, Figure 4.8). At the same time, he proposed replacing the commonly used term hâches (axes) with coup de poing (ibid., 21), because the Chellean-type was not an axe, but a tool that lent itself to many tasks.

Figure 4.8 The industry formally known as Acheulean. Plate of handaxes used to illustrate the Chellean Period in Musée Préhistorique (after Mortillet and Mortillet 1881). No. 25, Chelles; Nos 26, 28, 29, St Acheul; Nos 27, 30, Abbeville. Scale: Object 29, bottom right, is 19cm in length.

Mortillet announced these changes in Musée Prehistorique, ‘the portable museum’ he published with his son Adrien, where he no doubt hoped it would reach a wide audience. That the term was already in use among Mortillet’s supporters (e.g. Ameghino 1880; Maufras 1880) might suggest that the problem, and its solution, had been widely discussed. Mortillet was reluctant to abandon such a well-known term as the Acheulean, and saw the move, forced upon him by d’Acy’s “tireless” attacks, as a major inconvenience (Mortillet 1883, 133). To ease the passage, he retained the term Acheulean in his 1881 summary table (Table 4.4), placing it in lower case letters beneath the capitalised Chellean, indicating that it was no longer of epochal status. By 1883, Mortillet was using the term Acheulean to indicate a transitional period between the Chellean and Mousterian, during which primitive Mousterian scrapers appeared, handaxes gradually become more refined and elegantly made, and a faunal turnover occurred (Mortillet 1883, 254; Table 4.4).

Table 4.4 Revised classifications for the Palaeolithic after Mortillet and Mortillet 1881 (a) and Mortillet 1883 (b).
1881
Era		Age	Period	Epoch
Current	Prehistoric	Stone Age	Neolithic, polished stone	ROBENHAUSIENNE
Quaternary			Palaeolithic, chipped stone	MAGDALÉNIENNE: mostly caves, reindeer in almost all
				SOLUTRÉENNE: reindeer and mammoth
				MOUSTÉRIENNE: mostly mammoth
				CHELLÉENNE: Acheuléenne, large bears, Elephas antiquus
Tertiary			Eolithic, broken stone	OTTAIENNE
				THENAISIENNE
1883
Era		Age	Period	Epoch
Current	Prehistoric	Stone Age	Neolithic, polished stone	ROBENHAUSIENNE
Quaternary			Palaeolithic, chipped stone	MAGDALÉNIENNE: mostly caves, reindeer in almost all
				SOLUTRÉENNE: reindeer and mammoth
				MOUSTÉRIENNE: Large cave bears
				CHELLÉENE: Acheuléenne, mammoth in part, Elephas antiquus
Tertiary			Eolithic, burnt stone	THENAISIENNE

By relocating the type-site Mortillet was able to reinforce the fundamental tenets of his method, and to revoke any concessions he had been forced to offer d’Acy. Transitional periods like the Acheulean were, after all, only to be expected within a framework of gradual progress, and it was not the theory that was at fault, but a faulty type-site. To deny the fact of periodisation in the archaeological record because there were transitional phases was like denying the existence of night and day because of the passage of dusk and dawn (Mortillet 1883, 22). Furthermore, while these quotidian rhythms might not be synchronous everywhere, the division between night and day was sharp, and one always followed the other (ibid.).

D’Acy was not so easily fobbed off and detected at Chelles the same confused faunas and overlapping artefact types that he had found at St Acheul (d’Acy 1884). He identified scrapers from the lower ‘Chellean’ levels, which others had ignored or dismissed, either because they did not regard them as true Mousterian types or because they believed them to have been later introductions through imaginary cuts or ravines. He also cited Gaudry’s identification of mammoth teeth among the fauna from the lower units at Chelles that had been similarly argued away. It was all too simplified, selective and subjective for d’Acy, requiring special geological pleading to explain unwelcome associations. D’Acy simply refused to be silenced, even after Chouquet’s (1884) robust public rebuttal, but he did have the self-awareness to realise he was beginning to look like a troublemaker, as theory bound as those he sought to challenge. Nobody at the time seems to have made much of the fact that the type-site of Mortillet’s oldest epoch was on a terrace only a few feet above the modern river and according to normal river behaviour should be the youngest. The ‘ancient’ warm fauna demanded an earlier date, regardless of Prestwich’s terrace model.

Form Versus Function

Mortillet and d’Acy extended their disagreement to the basic function of handaxes. The latter (d’Acy 1887) followed Boucher de Perthes (1860, 1864, 171), thinking that handaxes fell into two functional categories. Pointed forms with rounded butts were designed to be hand-held, serving as hunting knives or daggers for resolving violent interpersonal conflicts, whereas oval forms with a sharp edge all-round were meant to be hafted, like the axes used by Australian aborigines (Figure 4.9). Oval handaxes were simply too dangerous to be used in the hand and many of them did not sit comfortably; any chopping action on resistant materials such as wood was just as likely to cut the hand. Chopper-cores and various (Mousterian-type) flake tools, all of which were found with handaxes throughout the so-called Chellean epoch, were much better suited to this and a range of other tasks. In d’Acy’s view this showed that Chellean humans had at least reached the same level of sophistication as indigenous Australians, and had not existed in some half-bestial state with only a single implement in their toolkit.

Figure 4.9 Emmanchement probable des hâches. Boucher de Perthes’s (1864) plate showing possible ways in which handaxes and other flint tools were hafted.

This was anathema to Mortillet (1883, 147), who thought the handaxe was the only tool of the Chellean and that it was a tool that did everything a Palaeolithic person could possibly have needed. According to its size and design, it could be used as an axe, a knife, a saw, a chisel, a chopper and so on. On some pieces, different parts served different purposes: the points to pierce, the sides to cut and the butt to chop. Ovate and pointed handaxes, and all the shapes in between, were not different tools but the same hand-held implement, they were simply designed to be held in different ways (see Figure 4.10). Ovates were always thicker and wider at one end, intended to be carefully held between the thumb and forefinger. Many retained unworked portions on their edges to provide a safer grip, although some may have required the use of some protective barrier (hide, cord) to shield the hand. If hafted, they would have soon cut through their bindings, Mortillet argued, not unreasonably, although this was apparently not a problem for the Australians (see discussion in d’Acy 1887). Twisted handaxes, furthermore, nestled comfortably in the heel of the hand and were poorly suited to hafting, as were the many pointed forms with thick unworked butts, which had a natural handle anyway. The handaxe was thus central to Chellean life, a universal multitool that provided a range of functional extensions to the hand. He chose the name coup de poing because it reflected both the origin and function, although there has never been a satisfactory English translation. Brown’s (1887b) ‘fist striker’ is literal but fails to capture the nuance of the original, handaxe probably does come closest, although axe is unfortunate, given that Mortillet abandoned using the term hâche because handaxes were not, in any real sense, axes (Mortillet 1883, 148).

Figure 4.10 Coup de Poing. Mortillet’s (1883) image indicating how a handaxe with an all round edge would have been safely used in the hand.

There was a third option, that handaxes had been throwing weapons. This was an early notion that had been quickly abandoned by Evans (1861, 1872) and summarily dismissed by Boucher (1864), who wondered why such elaborate objects would be used for this purpose, when any suitably sharp rock would have presumably done the same job. Re-inventing the idea in the popular weekly journal the Revue Scientifique, Jules Meunier (1886) argued that the handaxe had evolved from simpler offensive and defensive solutions, providing a hard, heavy and sharp object that was thrown at the heads of prey and enemies and which would have torn open a wound at the point of contact. All handaxes were based on a common idea, a template shared by all Palaeolithic humans, with variation in shape, size and finish emerging from raw materials, the choice of hammerstone, knapping technique and individual skill, he opined. The basic form was the disc with an edge all round with specific areas left for gripping and throwing. They were just one stage in an evolutionary arms race, later replaced, in turn, by Mousterian points, Solutrean leafpoints and Magdalenian bone harpoons.

Despite some resonance with the ideas of Mortillet and Evans, nobody who had studied large collections of handaxes from France or Britain could readily accept Meunier’s main thesis. Discs were not a particularly common form, and most handaxes were neither aerodynamic nor easily thrown (d’Acy 1887). Giant and miniature handaxes, of which there were by now many examples, were just too heavy to launch or too light to do much damage on impact. Although open to many possible uses for the handaxe, Mortillet (in d’Acy 1887) rejected the idea that they had functioned as weapons, except expediently in times of peril, when everything to hand served as an ad hoc weapon. He wondered whether the giants had served any practical purpose at all, or if they had been luxury or prestige items. For Mortillet, the main Chellean weapon was a pointed wooden stick, which may have been sharpened using a handaxe.

In reaching their conclusions these men relied upon personal experiential evidence, combined with analogy, experiment and ‘common sense’. They assumed that because their own hands and Palaeolithic hands were the same, they could also share the sensory and emotional experience of the original holders, instinctively and logically knowing how each handaxe was manipulated, with the fine workmanship seen on some examples effectively eliminating the possibility that there could have been major differences in the morphology and fine-motor control of ancient and modern hands. Neither Mortillet or d’Acy doubted that many pointed forms evidently had natural handles, but d’Acy (1887) believed that the supposed finger holds and palm-rests on ovates were happy accidents or parts left unworked because they did not interfere with function and/or because their removal would have spoilt the overall symmetry of the piece; they were certainly not the intentional ergonomic features imagined by Mortillet or Meunier. Ultimately, though, this was not really a debate about function nor about the intelligence or design capabilities of Chellean hominins, but another expression of the deep-rooted opinions held by the main protagonists on the nature and number of types found in the Palaeolithic record. Lines had been drawn.

Flint Workshops and Palaeolithic Living Floors

D’Acy’s (1878) main purpose in presenting the refitting core and flakes from Demuin B had been to demonstrate the integrity of the assemblages found in the silt on the plateaux above the Somme. He used it to show that the Chellean and Mousterian artefacts were contemporaneous tools lying almost exactly where they had been dropped by Palaeolithic hands, not the result of mixing by natural processes.⁷ It was just another part of the tireless feud with Mortillet. Others saw much greater potential.

In March 1880, Flaxman C.J. Spurrell (1842–1915), an independently wealthy man of leisure who devoted his time to archaeology and geology (Scott and Shaw 2009), discovered a thin seam of artefacts in the brickearths at the well-known site of Crayford in Kent (Spurrell 1880a, 1880b). Spurrell, along with his father Flaxman Spurrell Senior and William Boyd Dawkins, then a fellow-resident of the nearby Belvedere Estate, had been active at Crayford since the 1860s, during which time the site had produced a rich Pleistocene fauna and the occasional unassuming flint artefact (Fisher 1872). The new find, lying at 36ft (~11m) beneath the surface in fluviatile silts banked up against a small chalk cliff, represented an in situ knap-ping horizon extending at least 10ft (~3m) north-south and 15ft (~4.5m) east-west (Spurrell 1880a, 1880b). Everything was preserved, from the smallest chips to broken hâches and hammerstones. Heaps of refitting flakes were revealed, some piled several inches high, lying where they had once fallen between the knapper’s legs. They had been buried so incredibly gently that there were still cavities between them. Spurrell was able to put together several ‘restorations’ including an almost complete manufacturing sequence of a crude and broken hâche (in fact a Levallois core, Figure 4.11), the first time that such a feat was accomplished.

Figure 4.11 Spurrell’s restorations from the knapping floor at Crayford. (a) A major restoration with the resulting ‘hâche’ (a Levallois core); (b) a minor restoration showing a long parallel sequence of blade removals. Scale = 5cm.

Equally significant was the direct association of artefacts and the classic Crayford mammalian fauna, which included the bones of woolly rhinoceros, musk-ox and horse, showing that Palaeolithic humans had sat and worked right here, among these cold-adapted animals, on the banks of the Pleistocene Thames. Spurrell suggested that the spot has served as a knapping ground over a considerable period of time, detecting another six floors above the main one, with other possible floors below (1880b).

It was clear to Spurrell (1883) that these locations could not have been directly in the river, but represented adjacent land surfaces, sandbanks and quiet areas that had only occasionally been inundated by rising flood waters. He was not surprised archaeologists had had no success in finding complete stone-tool manufacturing sequences in river gravels, but in these marginal locations were preserved undisturbed workshops or knapping floors which could provide direct evidence of methods of manufacture and help understand the intentions and decisions of the knappers while they were working (Spurrell 1884). He identified several methods of manufacture and noted that the flint raw materials used had been of indifferent quality, suggesting that humans had relied on cliff falls and lacked the ability to quarry flint directly from the chalk. Several blocks had one or two flakes removed from the cortex to expose the flint within, presumably to test its quality. Some more extensively worked cores showed little purpose other than the production of coarse flakes, while some had been used to produce a regular series of long, parallel flakes. The most interesting cores had been carefully worked to produce knife-like flakes for retouch, a method that required the “continual rectification of the superior plane of percussion from which the large flakes were struck” (Spurrell 1884, 110), a description today recognisably a form of recurrent Levallois reduction (Boëda 1986). Several sequences were incomplete, the missing flakes those the makers had deemed worthy of retention for later use, probably as skinning implements or scrapers for dressing skins (Spurrell 1880b, 1884).

Evans (in Spurrell 1880a) was quick to point out the similarities between these ‘knives’ and those from the Lower-Level Gravels of Menchecourt, but nobody seems to have yet made the link between these familiar Levallois-type flakes and their parent cores. When Blanquet (1888) presented a small (10cm by 5cm) but classic Levallois ‘tortoise’ core from the site of Mont Roty, near Pont Audemer (Normandy), it was interpreted as a type of disc-scraper, the negative scar left by the Levallois flake seen as a deliberate thumb hold (Blanquet 1888) (Figure 4.12). Capitan (1891) regarded it a Mousterian type, noting it was widespread and that notable examples existed in the Yonne Valley and at Mesvin in Belgium. Mortillet (in Capitan 1891) was of the opinion that they probably occurred in most Mousterian assemblages but had been lumped in with discs. In Britain, it was more likely that they would have been regarded as rude handaxes, exactly as Spurrell had done (Spurrell 1880a; also Brown 1887a; Smith 1894), although at his other knapping floor at nearby Northfleet, Spurrell had recognised and described tortoise-cores and the ‘turtle-backed’ flakes removed from them, remarking how much preparation had gone into shaping the core, and how success was measured by the single final blow (Spurrell 1884, 113, Plate III; Spurrell 1885; Figure 4.12).

Figure 4.12 Tortoise cores, turtle-backed flakes and disc-scrapers. Top: turtle-backed (Levallois) flake from Northfleet; middle: disc-scraper (Levallois cores) from Roty; bottom: tortoise core from Northfleet (Northfleet after Spurrell 1884; Roty after Blanquet 1888). Scale = 5cm.

Similar jeopardy was faced by the makers of the handaxes at Northfleet, on which Spur-rell noted the removal of a final, longitudinal flake from the tip (a tranchet flake). Through experimentation and replication of these pristine artefacts, Spurrell realised that much of what had been interpreted as use-related damage on the edges of flakes and implements may have been caused by deliberate striking-platform preparation during manufacture. Refitting lithic assemblages had the potential to unlock the minutiae of Palaeolithic technologies in a manner that was previously impossible.

Another discrete horizon full of mint condition artefacts, some conjoinable, was discovered by archaeologist Worthington George Smith (1837–1917) in the sands and gravels of the Lea and Thames in north-east London (Smith 1879, 1882a, 1882b, 1884; see also Greenhill 1882, 1884, 1885), close to his home at Highbury. The most important finds were from the floor at Stoke Newington, at an elevation of ~85ft (~26m) OD, which yielded small but finely made handaxes, scrapers, cores, choppers, flakes, hammerstones, refitting groups, implements broken in manufacture and even sharpened wooden sticks. Not being a rich man, Smith collected not just the finer forms, but everything offered to him, providing a more representative sample of the types of artefacts originally present, from which, probably influenced by the presence of scrapers and the diminutive size of the handaxes, he concluded that the finds partly agreed with the period of Le Moustier (Smith 1882b). Ernest d’Acy would no doubt have seen it differently.

Unlike Spurrell, who observed his workshop in three-dimensions but over a fairly limited area (about 13m²), Smith’s floor was mostly based on correlating a patchwork of finds made over several years from sections in various pits and foundations excavated for housing, as London expanded to absorb the erstwhile village of Stoke Newington, although when the opportunity arose he had the overlying sediments carefully removed to expose the floor in plan (although he seems not to have drawn or photographed it as such). Even though the elevation and depth of the floor beneath the surface varied, it was always in the same stratigraphical position, underneath a bed disturbed by warp and trail formed during the most recent period of intense cold. Beneath the main floor were two similar horizons (at 12ft and 20–30ft beneath the surface), each with a characteristic set of artefact types, with the physical condition and quality of the workmanship improving through time (Figure 4.13).

Figure 4.13 As classified by Mr. Worthington G. Smith. Top left: implement of the oldest class; top right: implement of medium age; bottom: implements of “more recent Palaeolithic age from the ‘floor’ at Stoke Newington. More neatly wrought and unabraded” (after Brown 1889, Plate II). Scale = 10cm.

Smith confidently traced the Palaeolithic floor over a much wider area that crossed river valleys, straddling the Brent at Hanwell, the Lea at Clapton and the Thames at Acton and Ealing. On moving his permanent residence to Dunstable, Bedfordshire in 1888, Smith transferred his attention to the small and often fleeting commercial brickpits that peppered the Chalk of the Chiltern Hills, where he found further evidence of his Palaeolithic floor, the same type of artefacts again lying beneath beds that had been contorted by later (cold climate) geological processes (Smith 1889, 1894). At the Cottages Pit in Caddington, Bedfordshire, he was able to refit over 500 artefacts (Smith 1894, 77), including failed handaxe manufacturing attempts (Figure 4.14). He now traced the floor over 40 miles from the hilltops of Bedfordshire to Blackwall.

Figure 4.14 From the hilltops of Bedfordshire to Blackwall. Top: refitting handaxe manufacturing attempt from Caddington; bottom: shell of a handaxe manufacturing episode, the handaxe (cast by WGS on far right) has been removed from the spot. By contrast the several failed episodes reconstructed by WGS usually preserved both flakes and the broken handaxe/abandoned roughouts (after Smith 1894). Scale: the roughout top-centre is ~16cm long.

Further information on the Palaeolithic floor at Ealing and Acton (Lane-Fox 1870) was provided by another local archaeologist John Allen Brown (1833–1903). Brown’s main site was located in a series of small pits around Creffield Road, Acton (NW London), where he identified three distinct horizons of artefacts at 4–6ft, 7–8ft and 10–14ft beneath the surface, each marked by a seam of black matter he believed to be decayed vegetation (Brown 1886, 1887a, 1887b). The stones in these layers were bleached (patinated) white on their upper surfaces, probably as the result of sub-aerial weathering. Where the black staining was absent, the line of whitened stones could be used to trace lateral continuation of the floor.

The material from the upper floor was in situ, like Smith’s at Stoke Newington, being perfectly sharp and including refitting ‘nests’ of flakes. The prevailing types were pointed (Levallois) flakes 3–6in. (7.5–15cm) long, made from cores that had been specially prepared to produce a double-ridged flake suitable for hafting on the tips of javelins or spears (Brown 1886, 1887a). Brown (1887b) compared these with Mortillet’s Mousterian points, and while the parent (Levallois point) cores were present in small numbers, he failed to recognise them, describing two illustrated examples as rough celts (Figure 4.15), variations of which (i.e. crude axes, choppers) he thought to be ubiquitous in the Palaeolithic. Finely made handaxes were absent from the main floor, although examples had come from elsewhere on the same terrace, which Brown compared to the St Acheul or Acheulian type (Brown 1887a, 1887b, 1888, 1889). In a classic display of British suspicion of Mortillet’s epoch, however, Brown based the age of the find not on the artefacts but on fauna, specifically Lane-Fox’s discovery of the Middle Pleistocene species Rhinoceros hemitoechus in the next terrace down, making Creffield Road much older than the woolly mammoth and woolly rhinoceros fossil assemblages usually associated with the Mousterian (Brown 1887b). He thought it was best correlated with the Chellean period.

Figure 4.15 Early Man in north-west Middlesex. John Allen Brown’s plate of stone tools from the floor at Creffield Road, Acton, which included numerous Levallois points and flakes. Two Levallois cores are also illustrated (Nos 78 and 87), although Brown failed to recognise them as cores, describing them as rude celts (after Brown 1887a). Scale: Object 118 is ~10cm long.

Brown was equally unafraid to extrapolate his floor across hills and valleys, although he was less certain than Smith that the assemblages from different depths were of vastly different ages, detecting a gradual series of blending implement types. Judging the cortex remaining on the Creffield Road artefacts to be fresher than the river-drift crust found in the local gravels, Brown (1886, 1887b) suggested that the material had been brought in from Chalk exposures, or more likely the high-level plateau gravels of Hertfordshire, where such crusts were common. Whether he was correct in this assessment or not (see Scott 2011), he was nonetheless implying that humans had physically moved these raw materials over distances of ~20 miles. This was an important insight into how Chellean people might have organised themselves in the landscape, suggesting extraction and production zones separated by long distances, requiring planning and the conceptual division of places and resources into distinct activity zones. Within the prevailing theoretical context, however, which assumed the Chellean peoples to have been ‘nomad hunters’, wandering tribes whose spread over three continents must have taken vast ages and who were stuck in the same stage of culture for a very long time (Dawkins 1880), little was made of the observation. The Chellean people may have been more primitive than the most primitive of modern primitives (Lubbock 1865), but they were still expected to behave, more-or-less, like the Tierra del Fuegans, Australians or Esquimaux, “possessed of the faculties which distinguish man from the beasts” (Brown 1887a, 75).

Local knowledge and perseverance were thus critical to the development of this multi-scalar understanding of the archaeological record. Through refitting sequences at workshop sites, Spurrell, Smith and Brown were able to provide biographies for individual artefacts, validating and enriching the experimental approach long advocated by Evans and Mortillet, and providing future insights into the Palaeolithic mind. By stepping out from the sites to the wider landscape, using the concept of a contemporaneous or penecontemporaneous floor that contained all the artefacts discarded in and beyond a river valley over a relatively brief period, they also developed a clearer understanding of the range of technological and typological variation one might expect to find on a dynamic and undulating marginal fluvial environment. In each case the perfect preservation suggested that humans had abandoned the floor in a hurry, perhaps in the face of floods, pestilence or other danger (Spurrell 1884; Smith 1894). It was precisely the type of stratigraphical care and objective observation of which d’Acy would have approved and which did not sit comfortably with the dominant French model. The grand narratives written about the deep past were becoming more complex.

Man the Primeval Shadow

In reviewing the evidence for the Chellean epoch in his grand synthesis, La Préhistorique, Mortillet (1883) noted that bona fide handaxes had been confirmed from three main geological contexts (Pleistocene river deposits, the silts of the higher plateaux, and on the surface) across Europe, in the Levant, in India and in both northern and southern Africa. They were particularly abundant in England and France, Mortillet (ibid., 158) estimating that some 25,000 had been found in the Somme Valley alone, but growing numbers were known from Spain, Portugal and Italy. Central Europe was more of a blank, with nothing known from Switzerland, Germany or Austria, but where such lacunae existed, it was probably due to the distribution of collectors, Mortillet suggested, although he also noted how handaxes became less common away from the Cretaceous Chalk regions. In regions where flint was entirely lacking, Brittany for example, other materials such as quartzite had been used instead. Mortillet also accepted the authenticity of handaxes from North America, citing the example from Trenton, New Jersey (Abbott 1876), and that found by the Italian archaeologist Giovanni Cappellini at Burlington on the Mississippi in support of a global Chellean epoch. Four years later Thomas Wilson (1832–1902), curator of prehistory at the Smithsonian Institute, listed nearly 7,000 palaeolithic finds from 23 states, plus Canada (Wilson 1890), but it was quickly realised that these were unfinished implements left by more modern people, rejects that had been thrown away at an early stage of manufacture and never used (Holmes 1892a, 1892b, 1893). As the century closed, very few North American archaeologists still believed in a Chellean presence on that continent (Brinton 1897).

The people behind the handaxes remained little more than shadows (Smith 1894, 2). Mortillet’s book was more concerned with describing objects, sites and sequences, his few forays into the mode of life during the Chellean focussing on tool function, climate and habitat. He associated Chellean handaxes with Neanderthal fossils, drawing on the evidence of physical trauma to suggest that the possessors of these relics had been belligerent and violent towards each other, the shape of their jaws suggesting they had no proper language, communicating via shouting and singing. They lived by the rivers and used caves only rarely, and then only in hot locations (Furninha Cave, Portugal, and Oussidan Cave, Algeria) where they offered respite from the sun; he doubted that they had made clothes, the multifunctional handaxe not being particularly well-suited to tailoring. In this, Mortillet showed the same reluctance, or disinterest, to engage with the more anthropological side of the Palaeolithic as his English counterpart, John Evans.

It was left to the less eminent, those possessed of knowledge but unhindered by the weight of international authority, to lend greater definition to the mode of life of the handaxe makers (Spurrell 1884; Brown 1887a; Smith 1894). Smith (1884, 370) thought that there must have been a great number of people living on the Palaeolithic floor at north-east London, as shown by the thousands of stone tools found in the area he had examined, which he estimated to be about 1/10,000th of its total extent. By collecting and organising everything, he was able to better appreciate the range of objects that genuinely belonged together and the variety of technologies that lay behind them.

The lithic assemblages found by Spurrell and Brown, while contributing to the development of new approaches, were dominated by elaborate (Levallois) core-working and flake tools (retouched knives and javelin-tips), with handaxes absent. They both had much in common with the Mousterian,⁸ and are still today regarded as Early Middle Palaeolithic. Smith’s floor(s) at Caddington and Stoke Newington also contained scrapers and flake tools, like High Lodge and Le Moustier, but technologically the material was very different to Crayford or Ealing, lacking Levallois and being dominated by handaxes and the simpler forms of core working described by Spurrell. Smith estimated that his floor was between 300 and 100,000 years old, older than the cave series but younger than the most ancient river drift implements, but he did not attribute it to any of Mortillet’s epochs. In fact, he seems never to employ the terms Chellean or Acheulean at all, a disregard that probably stemmed from a combination of Evans-inspired skepticism, personal experience with large and more representative assemblages, and a lack of engagement with the continental literature.

Smith emphasised the high frequencies of rough handaxes at Stoke Newington, tools that had been made quickly using a few well-directed blows (Smith 1879, 279). There was much to learn from these about the daily life of Palaeolithic people, technological solutions designed on the spot to fulfil an immediate need, insights that were not found in the more ornate and elaborate forms, which instead spoke of culture, well-honed skills and the luxury of time. Smith reasoned that the rougher types should always dominate an unbiased assemblage, being more speedily made and more readily disposable.

He was able to develop these themes at the Cottages Site floor at Caddington, where by 1890 he had painstakingly refitted over 500 artefacts (from a total of 2,259), including near-complete handaxe manufacturing sequences (Smith 1894). The finished handaxes were both ovate and pointed in shape and usually small, but the refitting groups contained mostly failures and pieces broken in manufacture (Figure 4.14), or handaxes that had been removed from the place, leaving just the waste flakes. By reconstructing the outer flake shell from one such group, Smith was able to make a cast of the missing handaxe. It was unfinished. Smith supposed the maker had taken it away for finishing elsewhere, perhaps just a few feet outside the excavations (Smith 1894, 152, Figure 4.14). Further evidence for the movement of objects came from orphaned cores, made on a material type for which he found no flake equivalents, or similar groups of flakes that could not be matched in material or colour to any core he recovered (Smith was also careful to note, however, that despite his diligence the tenacity of the clay meant that much had inevitably been missed). The floor also yielded collections of tested blocks, hammerstones, cylindrical ‘punches’ with damage at one end, cores, piles of flakes selected for later use and a number of more unusual forms, make-shift implements with ad hoc cutting edges in any convenient place. One handaxe from Kempston showed evidence of recycling: an old iron-stained implement that had been picked up by the occupants of the floor and refreshed at the tip.

Smith (1894, 132–155) illustrated 16 conjoined groups of varying completeness and complexity, but the accompanying text does them little justice, emphasising the refitting process rather than the technological decisions and actions that created each artefact. The reader was left to mentally remove the flakes for themselves. He did, though, summarise some basic technological schemes, noting that ‘quartering’, the breaking up of a large nodule into smaller pieces suitable for making handaxes and other tools, was common at both Caddington and Stoke Newington, heavy hammers of quartzite or other hard rock being used for the purpose. Handaxes were made either on large flakes produced by quartering or directly from suitable nodules. They were roughly chipped into the desired shape using a stone hammer before being trimmed and finished using more delicate percussors. The refitting evidence thus largely confirmed what Evans (1872) had been able to work out from replicating flint handaxes (see Chapter 3), in terms of the sequence of actions and knapping stages involved in making a handaxe, but Smith had a very different idea of how those blows had been delivered. Evans had an essentially modern understanding, that handaxes had been made by a single person using direct percussion: the target handaxe was held in one hand and received direct blows from a hammer held in the other. The size and material of the hammer might vary at different stages of completion.

Hints that this was true came from Spurrell’s (1883, footnote) observations on the flaking patterns used to create twisted handaxes, which showed that these implements had been “chipped on one side [at one end], then turned over with the same end toward the workman, and worked on that”. This produced a handaxe where margins across the diagonal plane had been worked in the same direction but adjacent margins had been worked in opposite directions, creating a sinuous, twisted edge (White 1998b). The direction of the twist, whether an S or a reversed-S, depended on whether the hammer was held on the right or left hand, the higher frequency of reversed-S shaped edges suggesting that Palaeolithic humans had been predominantly right-handed. Worthington Smith, on the other hand, believed that while the initial stages may have been accomplished using a free-hand approach, finishing was accomplished by two people using indirect percussion, one holding the handaxe against a supporting anvil the other using a punch and hammer to deliver a precise blow (1884, 376, 1894), illustrating the method in several reconstructions. It was here, in his visual and textual reconstructions of Palaeolithic life, that Worthington Smith went much further than his contemporaries. His Palaeolithic people were recognisably modern in their appearance (the moustaches on the women and pointed ears notwithstanding) and were depicted as a nuclear family (adult male, adult female, juvenile female and a child), sitting on a log beside a river, a pile of handaxes by their feet and a club at the ready (Figure 4.16). They gaze into the foreground, as if concentrating fixedly on the artist.

Figure 4.16 Top: drawing showing a man and woman work together to finish a handaxe using a punch and hammer; bottom: frontispiece to Man the Primeval Savage showing a somewhat modern Stone Age family (after Smith 1894).

By the time these primeval savages arrived in Britain, after the last Great Glacial period and during a period of genial climate, they were already skilful designers and makers of “geometrically correct” stone tools (Smith 1894, 3), able to use the correct ornamental forms and possessing an uncommonly skilled touch. Presumed burnt vegetable matter from the floor at Stoke Newington showed that they knew how to produce fire, using grass as tinder and flints to generate sparks. To keep the home fires burning humans would have had to send parties of women, children and the elderly to collect untied bundles of firewood, picking up what seasoned wood they could find and tearing green branches and twigs off trees. Smith believed that stone toolmaking would have been a similarly organised ‘industry’, the same section of society being sent out to collect flint from the Chalk or other sources and bring it back to the human ‘haunt’. At night, while huddled together around fires built for comfort and protection, “the more skilled and light-handed human creatures would… fabricate pointed stone weapons and keen-edged oval choppers and knives” (ibid., 55). His reconstructions of toolmaking show a man and a woman working together, naked except a belt to hold the essentials, which was, of course, a handaxe. Smith thus not only gave functions for the two main handaxe designs, but also provided the relaxed, co-operative social context within which the better made examples were produced, in contrast to the rough pieces made in action in the spur of the moment.

The presence of so many small implements showed that people commonly engaged in fine and neat work, and that life was not a constant round of heavy toil (Smith 1884, 379). People would often just laze around under trees, bushes and other natural shelters. Few of the implements could have served as weapons, they were all obviously tools, and there was nothing in the armoury that would be suitable for hunting large wild animals. Humans had probably depended on nuts, roots, berries, eggs, fish and insects for survival, supplemented by (putrid?) meat scavenged from carcasses (found by yet more scouting parties) and the occasional small animal from hunting (1884, 379). Bones were smashed for marrow and brains. The numerous scrapers led Smith to infer that animal skins may have been used as wraps for additional warmth, while the fossil (Cosinopora globularis) beads he had found at Bedford (Smith 1884) indicated that they were adorned with personal jewellery.

Smith made good use of the environmental and landscape evidence in his reconstructions. At Stoke Newington, he saw that people had roamed the floodplains and valley of the river Thames, while at Caddington they had gathered around large ponds situated high on the Chalk downs. People probably felt safer in these open landscapes, away from the dark forests, and anyway the lack of secure containers meant that humans were forced to stay close to water. He noted that artefacts occurred in dense clusters, and would then suddenly cease, representing the places where primeval people were in the habit of gathering. Some large branches may have been used to make hovels or dens.

Despite a declared intention not to delve “deeply into any uncertain branches of the subject” (1894, ix), Smith then strayed far from the evidence, weighing, in his terms, probability versus improbability. Although certainly fanciful it is quite unlike any other scholarly work of its time and simply for that reason it is worth discussing in some detail. For Smith, Palaeolithic people had not been domesticated (1894, 53), but equally “they were not solitary savages, few and far between, but… a peaceful, timid, playful, inoffensive, and intelligent community, living in friendly intercourse with each other” (ibid., 379). On the other hand, they could be clannish and exclusive, shunning strangers. There was no marriage, but there was parental-pairing on a birth-by-birth basis, with women entering the breeding pool at an early age. During pairing season, males would compete for females, the strongest, handsomest and hairiest having most success. Females, being much smaller and weaker, were more often killed and thus in prime demand. Males would lead raiding parties into neighbouring territories, where they would capture females by force, armed with clubs and weapons of flint. Children would frolic and dance around, climb trees, paddle and play-fight. Young boys would chase the young girls.

These Palaeolithic humans lacked proper language, so ideas were passed through signs and gestures or a monosyllabic “jabber, shriek or roar, accompanied by a strange laughter” (1894, 49). It was rudimentary but sufficient for their means. They had a basic level of social care, but the severely injured would be abandoned or killed and the dead were not buried, primeval people having very little understanding of life versus death. Infanticide was common, as was suicide, particularly among the infirm, starving, weak-headed and tired of life, who would have thrown themselves off cliffs or cut their throats with handaxes. Decency and cleanliness were unknown. Human skulls were used as cups at mealtimes that were frenzied, salivating affairs involving fighting and quarrels (1894, 59). Early humans were hardly more understanding of natural phenomena than a horse (ibid., 52); everything that moved was alive, even their own reflections, and revenge would be sought on inanimate objects. People may have sat on their haunches for hours, seemingly deep in thought, but in reality they were just staring vacantly (Smith 1894, 49): “Men would dream but their dreams would be esteemed as realities” and would often lead to killing.

Smith combined plausibility, a third-hand knowledge of ethnography and his own interpretation of the archaeological record to bring the primeval savage to life. His Palaeolithic societies were structured around the fundamental relationships forged between men, women and children, extended families who worked, rested and played together in fairly large but exclusive communities. They had technology and intellect sufficient for their needs and suited to their wandering lifestyle, taking what they needed from Nature as they found it and depending on instinct, emotion and native wit for survival. But they had no morals, no decency, no reasoned understanding of the world around them and lived at the basest level of human existence. In this, Smith offered his late Victorian readership a primal avatar of themselves, an atavistic shadow, a prehistoric Mr Hyde. It was a compelling image, one carried to the wider masses via the writings of Anglican populariser of evolutionary science, Rev. Henry Neville Hutchinson (1858–1927), whose Prehistoric Man and Beast (1896) drew liberally and often verbatim from Smith’s “fanciful picture” in its own recreations of Palaeolithic life (Figure 4.17). Equally stimulating for Hutchinson (1896, x), however, had been the series of satirical Prehistoric Peeps cartoons (Reed 1894) published in Punch Magazine from 1893, in which artist Edward Tennyson Reed (1860–1933) had entertained an enormous readership with his prehistoric reflections on modern life (Figure 4.18).

Figure 4.17 Mammoth hunters in the south of France (after Hutchinson 1896, drawn by Cecil Aldin).

Figure 4.18 “The Unsociable Mammoth: Owing to his notorious eccentricity their relations with the local Mammoth were somewhat strained”. Cartoon from Punch’s Prehistoric Peeps series, popular in the 1890s (after Reed 1894). Mammoth hunting was (and remains) a popular theme for artists.

D’ault Du Mesnil: The Acheulean Reloaded

Beneath the main floors at Stoke Newington and Ealing were two deeper horizons of flint artefacts, separated from each other by varying depths of archaeologically sterile sand and gravel. As neither horizon could be the source of the floor material, nor of each other, Smith (1884, 1894) and Brown (1887a) concluded that three distinct phases of occupation were represented. Smith thought his floors were of vastly different ages, separated by periods when humans had abandoned the place, and captured different points in the development of handaxe form and manufacture. The lowest artefact horizon contained only rude and abraded handaxes, with no scrapers or knives. The material from the middle horizon contained large, well-made pointed handaxes, choppers and medium-large scrapers, more sophisticated than anything found below. The youngest group, from the main floor, was even more advanced, containing diminutive, thin handaxes of beautiful workmanship and small, dextrously made scrapers. That said, the same two basic handaxe types occurred throughout, Smith suggesting that after a period of ‘exaltation’, during which pointed and ovate handaxes had been developed, humans entered a stationary period when few new ideas emerged. He was also emphatic on the fact that crudeness alone was no measure of age, makeshift forms being found everywhere.

Brown held a different conviction, that humans had occupied the Thames Valley continuously for an enormous period of time, their implements forming a series from “rudely worked nodules wrought simply to a point” to the better made and specialised tools and weapons from sites such as Ealing (Brown 1889, 63). He later traced the development of axes from the Palaeolithic to the Neolithic, via a transitional or ‘mesolithic’ period (Brown 1893). It provided a very long lineage for the English but brought Brown into conflict with Dawkins and Evans (Dawkins 1894), both of whom saw a sudden rupture between the two main periods of prehistory.

These differences aside, the theoretical and empirical implications were obvious. The handaxes of the drift period had not been the same from start to finish but had developed over time. So, the differences observed in handaxes from different heights in, for example, the Little Ouse downstream of Thetford (Flower 1867, 1869), or the Thames-Kennet near Reading (Stevens 1881;

Shrubsole 1885, 1890), might represent different periods of time, as some original collectors had hinted, rather than local or tribal variations. It was an idea that the next generation would exploit to the full and which Evans would never accept (Evans in Gaudry 1889, 77; Evans 1897).

The same concept had also become popular in France, Adrien de Mortillet noting the importance of progress in handaxe form in the long discussions on function at the Société d’anthropologie in Paris (in d’Acy 1887). The idea could also be inferred from the elder Mortillet’s transitional Acheulean phase between the Chellean and Mousterian (e.g. 1883), although this distinction was predominantly based on the presence and character of associated flake tools. Unfortunately, while the British carried on with the task of describing and ordering their Palaeolithic record without worrying too much about what to call things, the French seemed determined to turn every question and every new discovery into a discussion about Mortillet’s chronological classification, conducted with varying levels of hostility (e.g. Reinach 1889; Gaudry 1889; Salmon 1891; Vauvillé 1891; d’Acy 1894a, 1894b, 1895; Laville 1898a, 1898b). It is today quite difficult to see what was achieved, other than the swelling of personal enmity.

Progress finally came from prolonged and meticulous local observations, just as it had in Britain. Geoffroy d’Ault du Mesnil (1843–1921) was already a respected mineralogist and geologist by the time he moved to Abbeville in 1874, and in 1877 was elected to the town’s Sociétè d’Emulation, replacing the recently deceased Charles-Joseph Buteux (d. 1876), one of Rigollot’s two geologists (López-Ramero 2015). From 1875 he studied the geology and archaeology of the Somme, working in collaboration with Albert Gaudry to establish secure faunal and climatic sequences onto which he could map the archaeology (see also Ladrière 1891, whose contemporary geological interpretations largely agreed with d’Ault du Mesnil’s). His work became well known through word of mouth, but he published little, a failing for which he was criticised in his own time (López-Ramero 2015).

D’Ault du Mesnil’s work at the Carriere Leon on the Champs du Mars at Abbeville, close to Boucher’s Moulin Quignon site and at the same or slightly higher level, described a sequence of ten fluviatile beds showing a clear and fine-grained succession of mammalian fauna and archaeological industries (d’ Ault du Mesnil 1896, 1899; see Table 4.5 and Figure 4.19). The beds at the base of the sequence were older than any previously seen at Abbeville and contained the ancient ‘southern’ mammoth Elephas meridionalis and straight-tusked elephant (E. antiquus) in association with large, coarsely made lanceolate (long, pointed) and amygdaloid (almond-shaped) handaxes with wavy edges. Roughly worked almond-shaped handaxes also occurred in Bed 3, but a “more advanced style of chipping” and straighter edges characterised those from the base of Bed 4 (d’ Ault du Mesnil 1899, 144). These beds saw the appearance of rare unifacial (flake) tools and a primitive form of mammoth, alongside surviving populations of straight-tusked elephant.

Figure 4.19 Sand and gravel pit at Abbeville, 1890s. This section shows strata 1, 2, 4, 5 and 7, but omits 3, 6, 8, 9 and 10. See Table 4.5 for explanation of deposits (after d’ Ault du Mesnil 1899).

Table 4.5 D’Ault du Mesnil’s description of the geology at Abbeville (see Figure 4.19).
Phase	Stratum	Description
IV	10	Yellow calcareous loam (limon), remains of Elephas primigenius, rare, without fauna or human industry
	9	Blocks of broken flint with white patina
III	8	Red-clayey sandy loam with Elephas primigenius , objects of human industry at the base
	7	Blocks of broken flint with white patina, with Tertiary pebbles
II	6	Yellow sandy loam with beds of clay; objects of human industry at the base
	5	Thin bed of gravel
	4	Gray or yellow gravels. Elephas primigenius and objects of human industry at the base.
	3	Rolled gravels and sand in layers and distorted stratification mixed (type of fluvial alluvium) with Elephas primigenius and sometimes Elephas antiquus. Numerous blocks of sandstone, chipped flints. Erosion of deposit next below.
I	2	Gray sandy marl, horizontal; stratification with Elephas primigenius, Elephas antiquus, Elephas meridionalis, Rhinoceros merckii; human industry at the base
	1	Large gravels, slightly rolled, stratification horizontal with remains of Elephas antiquus, Elephas meridionalis, and Rhinoceros merckii

Further technological refinements were seen in the delicately flaked-handaxes from Bed 6, d’Ault du Mesnil noting that their white patina made them highly desirable to collectors, while a third step in the evolution of chipped stone implements was identified in Bed 8. This was marked by more abundant scrapers and unifacial flake tools with handaxes comparable to those found at St Acheul or in the brickearths of Normandy, and was associated with a younger fauna that contained no elephants except mammoth⁹ (d’ Ault du Mesnil 1899). Each bed of gravel was overlain by a body of silt, within which d’Ault du Mesnil detected soils and plant matter suggesting that fluvial sedimentation had ceased and that dry land surfaces had temporarily formed. The human industries were all concentrated at the top of a gravel bed or at the base of the overlying silt. When found in the brickearth they were invariably associated with lines of angular, broken flints. They were, in couched terms, living floors.

There was overlap between the forms found in the different beds, older forms persisting throughout, but each industry had its own flavour, with the introduction of new forms and changes in the dominant or most characteristic types (d’ Ault du Mesnil 1899, Figure 4.20). The general types could be argued to be the same everywhere, but there was a gradual evolution in both form and technique over time. The same types and associations were also seen beyond Abbeville by d’Ault du Mesnil and/or his good friend Dr Louis Capitan¹⁰ (1854–1929): at Rouen, Mesnil-Esnard and Oissel (Ault du Mesnil and Capitan 1893), at Tilloux (Boule 1895; Capitan 1895), at La Micoque (Capitan 1896) and at St Acheul (d’Ault du Mesnil & Capitan in d’Acy 1894b). Indeed, for both men there was a very strong case for recognising the Acheulean as a distinctive epoch in its own right, marked by finely worked handaxes, the constant presence of flake tools and a characteristic Middle Quaternary fauna in which mammoth was the only remaining elephant species. It was a transitional industry that gradually evolved over time (as did mammoth teeth), emerging from the Chellean and ultimately morphing into the Mousterian, but it was distinctive and well-defined in terms of artefacts, fauna and stratigraphy (see Table 4.6).

Figure 4.20 Handaxe evolutionary sequence identified by d’Ault du Mesnil at Abbeville. Left: earliest and rudest implement, associated with E. meridionalis and E. antiquus in stratum 2; middle: second step in evolution of chipped stone implements, associated with E. primigenius in strata 3–6; right: third step in evolution of chipped stone implements, associated with E. primigenius in Stratum 8 (after d’ Ault du Mesnil 1899). Scale = 5cm.

Table 4.6 D’Ault du Mesnu’s 1889 geological, environmental, palaeontological and archaeological framework for the Palaeolithic period. Note the newly reinstated Acheuleene and author attribution.
Stages		Palaeontological Divisions			Geological Character	Industrial Character	Climate	Palethnological Divisions
Upper Quaternary		Predominance: Reindeer. Mammoth continues to exist			Red silt with reindeerSandy clay loam with broken flintRed diluvium Weathered. No stratification.	Predominance of stone worked into narrow blades	Cold and dry. Return of cold conditions	MAGDALÉNffiNNE (G. de Mortillet)
Middle Quaternary	Upper	Mammoth and woolly rhinocero	Transition: mammoth and horsePredominance: a mammoth (teeth with thin, tight plates) and horse. Woolly rhinoceros disappears		Silt and Sand with mammoth and horse. Horizontally stratified. Fluvio-marine in origin (contains marine shells). All traces of glacial action have gone.	Predominance of stone worked into wide bladesA transitional epoch	Ameliorated and dry	MENCHECOURIENNE (G. d’Ault du Mesnil)
	Middle		Period of great glacial extension	Predominance: a mammoth (teeth with wide, spread plates), woolly rhinoceros, Hippopotamus has left	Gravel, Marl, Clay and Sand with mammoth and woolly rhinoceros. Stratification is contorted. Traces of glacial action.	Predominance of stone worked on only one face	Cold and wet	MOUSTÉRIENNE (G. de Mortillet)
	Lower		Transition: mammoth and Elephas antiquusHabitual association of mammoth and Elephas antiquus		Gravel, Marl, Clay and Sand with mammoth and Elephas antiquus. Stratification is contorted and sloping. This bed cuts into the deposit beneath.	Predominance of implements made from small flakes and worked on both faces.A transitional epoch	Cooling and wet	ACHEULEENE (G. d’Ault du Mesnil)
Lower Quaternary		Predominance: Elephas antiquus and Merck’s rhinoceros. Hippopotamus			Marl, Gravel and Sand with Elephas antiquus and Merck’s rhinoceros. Stratification generally horizontal	Predominance of implements roughly worked on both faces	Hot and wet	CHELLEENE (G. de Mortillet)

D’Ault du Mesnil and Capitan employed multidisciplinary methods similar to those used by Mortillet (despite plentiful modern assertions that his scheme was entirely artefact-based) and they saw the value in his classification, even if just as a heuristic device, a convenience to be maintained until something better could be found (Capitan, in d’Acy 1894b). They disagreed on some details, such as the precise sequence of Pleistocene elephants, but they were most definitely on the same side (cf. d’Acy 1894b). D’Ault du Mesnil’s transitional Acheulean epoch was essentially an elevation in status of Mortillet’s own 1883 transitional Acheulean period, another small refinement, and Mortillet had few qualms in formally accepting it as the second of five Palaeolithic epochs or in giving d’Ault du Mesnil credit for helping to reinstate it (Mortillet 1891, 1894, Mortillet, in d’Acy 1894b Table 4.7). It was small tribute to a key ally in the fight against d’Acy (e.g. d’Acy 1894a, 1894b, 1895), designed no doubt to ensure the continued support of both d’Ault du Mesnil and Capitan (Mortillet, in d’Acy 1894b). In the same year, Salmon (1891) credited both d’Ault du Mesnil and Mortillet for working out the basic framework.

Table 4.7 Mortillet’s 1891 classification, formally reinstating the Acheulean as a (transitional) epoch
Era		Age	Period	Epoch
Current		Stone Age	Neolithic	Robenhausienne Dolmens, first lake dwellings
				Campignienne Kitchen middens
Quaternary	Upper		Palaeolithic	Magdalénienne Mostly caves, reindeer ubiquitous
				Solutréenne Reindeer and mammoth Menchecourienne
	Middle			Moustérienne Cave bear
				Acheuléenne Mammoth, Elephas antiquus ends
	Lower			Chelléenne Elephas antiquus
Tertiary			Eolithic	Puycounienne Upper Miocene
				Thenaisienne Lower Miocene, Oligocene

In his final statements on the matter, Mortillet (e.g. 1897a, 1897b) emphasised the differences between the crude and massive Chellean handaxes and the elegantly shaped Acheulean forms, which were more finished and delicately worked with great care (Figure 4.21, Table 4.8). Palaeolithic people had not been inventive, they had used handaxes for many millennia, but during the Acheulean these had gradually improved until people developed the habit of using the large flakes produced in their manufacture. Retouching these flakes created new tools, which ultimately took over, during the Mousterian. By 1894 at the latest (see Mortillet, in d’Acy 1894b), the true nature of the so-called discs had also been realised: they were the missing Levallois cores, designed to provide broad or pointed flakes proper for the manufacture of Mousterian points and scrapers, probably first seen in the lower Mousterian (moustérien inférieur, aka Acheulean). None of this satisfied d’Acy, who saw in the new scheme all the problems inherent in previous versions (d’Acy 1894a, 1894b), and it was practically ignored by the British. Evans (1897, 435, 528) grudgingly adopted the 1881 version of Mortillet’s scheme, for the expedient reason that it had become almost universally accepted, but he bemoaned the significant differences with his own ideas and seemed unaware that the Chellean and Acheulean had since become separate epochs. He would no doubt have disapproved of the division.

Figure 4.21 Chellean (left) and Acheulean (right) handaxes, as illustrated by Mortillet 1897b. Scale: the handaxe on the left is ~18cm in length.

Table 4.8 Gabriel de Mortillet’s 1897 (and final) classification.
Era		Age	Period	Epoch
Modern Quaternary	Prehistoric	Stone Age	Neolithic	Robenhausienne (Robenhausen, Zurich)
				Campignienne (Campigny, Seine-Inférieure)
				Tardennoisienne (Fère-en-Tardenois, Aisne)
Ancient Quaternary			Palaeolithic	Tourassienne (La Tourasse, Haute-Garrone) Ancient Hiatus
				Magdalénienne (La Madeleine, Dordogne)
				Solutréenne (Solutré, Saone-et-Loire)
				Moustérienne (Le Moustier, Dordogne)
				Acheuléenne (St Acheul, Somme)
				Chelléene (Chelles, Seine-et-Marne)
Tertiary			Eolithic	Puycounienne (Pay-Courny, Cantal)
				Thenaysienne (Thenay, Loir-et-Cher)

Fin De Siècle

In his presidential address to the Geologists’ Association in February 1898, BGS palaeontologist E.T. Newton (1840–1930) reviewed the evidence for human fossils of possible Palaeolithic age, providing a timely reminder that as the century drew to a close, archaeologists still had little idea of what Chellean or Acheulean humans had actually looked like. Mortillet’s (1897a) celebration of France’s deep rich past provided a similar but more detailed list. Most of the claimed Palaeolithic-age human fossils had been found in caves, if their origin was known at all, where they fell into two types: the Canstadt or Neanderthal type and an essentially modern human type.

There were no uncontroversial human fossils from any Pleistocene drift deposits, the existing claims being either fraudulent or intrusive, and many of the cave examples had been thrown open to doubt (Table 4.9). Dawkins (1874, 1880) had established a principle that Palaeolithic people had not buried their dead, and thus rejected the claimed burials from Cro-Magnon, Bruniquel and Aurignac in France, Paviland in Wales, Engis and Trou du Frontal in Belgium, and Baoussé-Roussés (Menton, Grimaldi) in Italy as intrusive Neolithic interments. Mortillet (e.g. 1883, 1897a) nursed similar doubts about the last, because the presence of bone tools did not fit with his Solutrean age for the lithics, although he thought the modern human skeletons from Cro-Magnon and Laugerie-Basse were genuinely Palaeolithic.

Table 4.9 Newton’s catalogue of human fossils of claimed Palaeolithic age (after Newton 1898). The most spurious claims had already been expurgated.
Site (Find Date)	Specimen	Type	Context	Newton’s Suggested Age
Canstadt, Germany (1700)	Skull	Neanderthal	Unknown	Probably Palaeolithic
Gailenreuth, Germany (1774)	Skull	Modern, brachycephalic	Cave, Pleistocene fauna and pottery	Neolithic, intrusive
Lahr, Germany (1823)	Human bones	Unknown	High-level gravels of the Rhine	Uncertain
Paviland Cave, Wales (1823)	Partial skeleton	Modern	Cave, Pleistocene and later fauna	Neolithic, intrusive
Engis, Belgium (1833)	Two skulls	Modern, dolichocephalic	Cave, Pleistocene fauna and pottery	Neolithic, intrusive
Staegenaes, Sweden (1843)	Two skeletons	Modern or Neanderthal	Fluvial gravels 100ft above modern river	Uncertain
Denise, France (1844)	Human bones, skull fragments	Modern, primitive features	Volcanic tuff overlying ancient gravel	Younger than Pleistocene fauna
Aurignac, France (1852)	17 human skeletons	Modern	Cave, burial?, Pleistocene fauna and pottery	Neolithic above Pleistocene layers
Bruniquel, France (1863)	Human bones, skulls	Modern, dolichocephalic	Cave, Pleistocene fauna	Neolithic, intrusive
Neanderthal, Germany (1857)	Partial skeleton	Neanderthal	Unknown	Probably Palaeolithic
Arcy-sur-Cure, Yonne (1860)	Mandible	Low type with receding chin	Cave, with Pleistocene fauna	Palaeolithic
Moulin Quignon, France (1863)	Mandible	Modern	High level gravels of the Somme	Fake, modern
Larzac, France (1864)	Human bones,Skull	Neanderthal	Cave	Palaeolithic
Eguisheim, Germany (1865)	Skull	Neanderthal-like	River deposits, Plesitocene fauna	Palaeolithic
Naulette (Furfooz), Belgium (1865)	Mandible	Robust, primitive, not modern.	Cave, Pleistocene fauna under five layers of stalagmite	Palaeolithic
Cro-Magnon (1868)	5 partial skeletons	Modern, dolichocephalic	Rock shelter, Pleistocene fauna, Magdalenian industry	Neolithic
Clichy, France (1868)	Near-complete skeleton	Neanderthal?	Gravel pit, with Pleistocene fauna	Recent or Palaeolithic
Brux, Bohemia (1872)	Human bones, skull	Neanderthal	Quaternary fluvial deposits	Palaeolithic
Trou de Frontal, Belgium (1871)	16 skeletons	Modern, brachycephaplic	Rock-shelter with modern fauna	Neolithic
Mentone, Italy (Baoussé-Roussés, Grimaldi, 1872)	‘Several’ skeletons	Modern, dolichocephalic	Cave, with Pleistocene fauna	Neolithic
Lombrive, France (1872)	Two skulls	Modern, brachycephalic	Cave, with reindeer and Bos longifrons	Neolithic
Colombi Cave, Italy (1873)	Human bone, cutmarked, burnt	Modern	Cave, with flints. Pottery and modern animals	Neolithic
Schipka, Moravia (1881)	Mandible	Neanderthal	Cave, with Pleistocene fauna	Palaeolithic
Tilbury, England (1883)	Partial skeleton	Modern	Thames gravel 34ft below floodplain	Neolithic
Spy, Belgium (1886)	2 partial skeletons	Neanderthal	Cave, with Pleistocene fauna	Palaeolithic
Galley Hill, Northfleet, England (1895)	Partial skeleton	Modern, dolichocephal	Gravel 100ft above Thames (same height at Swanscombe	Not certain
Java, Indonesia (1894)	Skull, teeth, femur	Pithecanthropus erectus	Fluvial deposits with Pliocene fauna	Tertiary, the missing link?
Hunstanton, Norfolk, England (1897)	Skeleton	Modern	Gravel, not alluvium	Uncertain

For Mortillet the most likely authors of the Chellean, Acheulean and Mousterian were the Neanderthal race (Mortillet 1873, 1897a, 279), for which Schaaffhausen (1888) had provided a useful facial reconstruction (Figure 4.22). The Pithecanthropus erectus fossils from the island of Java, discovered in 1894 by Dutch military surgeon Eugène Dubois (1858–1940) in association with a Pliocene-type fauna (Dubois 1894, 1896a), were a good candidate for the makers of the eolithic (Mortillet 1897a). Mortillet followed Dubois (1894, 1896a, 1896b), thinking that the new Asian species was a missing link between humans and apes, with a distinctively human femur and primitive, relatively small-brained (~1,000cc) cranium. The claim was highly controversial and had little support outside France (Dubois 1896a, 1896b). Most British scholars saw both the femur and skull as essentially modern, comparing them with the Neanderthal specimens, whereas German scientists thought the femur to be human but suggested the skull belonged to a large ape similar to the gibbon (Dubois 1896a, 1896b; Mortillet 1897a, 222). In 1900 Dubois stopped all research on the finds and hid them away for almost a quarter of a century (Shipman 2002).

Figure 4.22 The first reconstruction of the Neander Valley skull (after Schaaffhausen 1888).

The lack of human fossils may have meant that it was impossible, for the moment, to put a face on the handaxe makers, but this was not an impediment to the archaeologists’ efforts to understand their customs and habits from their material culture. These obstacles were of their own making.

After 1872 Evans’s thinking on the Palaeolithic hardly changed (cf. Stopes to Layard 1 September 1904); the aims on page one of the second edition of Ancient Stone Implements (1897) were unchanged from the first, despite a lapse of 25 years. In it, Evans accepted that several new types of handaxe had been identified but still saw no reason to alter his earlier divisions. He continued to avoid questions of chronology, and where he noted that Kent’s Cavern now contained a complete cultural sequence through the Acheulean, Mousterian, Solutrean and Magdalenian, the crude handaxes discovered in the basal breccia at the site in 1872 completing the series (Pengelly 1884), he remained deeply skeptical about the Mortillet scheme as a whole and was still not convinced that the Acheulean and Mousterian were not just two facets of the same culture. Where new interpretations were offered, they were characteristically mundane. Evans suggested, for example, that the apparent lacunae in Palaeolithic settlement north of a line drawn between the Severn and the Wash was not because ice had lingered in northern England and Scotland preventing human ingress (e.g. Dawkins 1880), but was collection biased caused by the difficulties involved in spotting non-flint artefacts in non-flint gravels and the smaller number of quarries in these gravels providing fewer opportunities for collectors. In other words, the absence of evidence was not evidence of absence. Elsewhere, the extraordinary refits of Spurrell and Smith were mentioned, but they were not integrated into his earlier experimental analysis of stone tool manufacture. The catalogue may have been updated, but interpretatively the second edition was the same book as the first.

The same is true of Sir John Lubbock (who was ‘raised to the peerage’ as Lord Avebury in January 1900). The fifth (1890) and sixth editions (1900) of Pre-historic Times varied only insubstantially from the first four (the hyphen was only dropped with the seventh (1913) edition), and he was still debating the case for deep prehistory decades after the battle had been won. In terms of the drift period, the sixth edition is notable mostly for the inclusion of Spurrell’s floor at Crayford (20 years after it was first announced), the recognition of several new handaxe types and possible functions (rough-butted axes intended to be hand-held, worked-butted axes intended to be hafted, oval implements with an edge all round and oval implements with a deliberately twisted edge) and the use of photographic plates instead of line drawings. In a display of staggering indifference or disrespect, Lubbock (1900) mentioned Mortillet only once, in relation to the relative hardness of flint and bronze, as if the past 30 years of debate in France had never happened. Dawkins’s opinions remained similarly rooted in the 1870s (White 2017).

Mortillet had fared little better at home during this period, expending much mental energy on an interminable debate on classification that led to a few key refinements to his unilinear model but no finer understanding of the true complexity of the Palaeolithic period. In the quarter-century up to his death in 1898 his main theoretical achievement seems to have been the recognition that transitions were important and the division of his earliest epoch into two parts, one transitional.

Instead, advances in method and theory came from archaeologists occupying a lower tier in the archaeological hierarchy, who did most of the primary work. Across both France and Britain, knowledgeable, local enthusiasts (generally men but some women of advancing years) used their time and proximity to amass large and unbiased assemblages from well-understood and detailed contexts, information carefully gathered over many years and sometimes involving controlled, horizontal excavation. Some understanding of regional and global variation was also emerging. Almost without exception, these studies showed a more complex, nuanced Palaeolithic record that did not fit smoothly into the bald epochs, classes and categories created decades earlier. However, researchers at this level still depended on the writings and personal opinions of Evans, Mortillet, Lubbock, Dawkins and other international synthesisers to place their work in wider context and to give their published conclusions legitimacy. Those who were taken seriously in the societies and journals rarely sought to bring down the entire edifice (Érnest d’Acy being an obvious exception). To make their contribution, they largely concentrated on the significance of their own regional discoveries and how these might confirm, enhance or amend the details of existing frameworks. It was certainly not unusual for them to defer outright to one or other higher authority. In this way, the greatest archaeological achievements of the century became its strongest theoretical shackles.

Notes

· 1 In the quest to understand the Palaeolithic archaeology of their caves, the Victorians were not helped by the fragmentary and uneven nature of their evidence. The British record is, we now know, overwhelmingly dominated by Magdalenian, with very limited evidence for Mousterian or Aurignacian, and no trace of the Solutrean.

· 2 BGS geologist and, from 1873, Woodwardian Professor of Geology at Cambridge.

· 3 Brussels congress panel members and how they voted:

 1) Japetus Steenstrup (1813–1897, Denmark): rejected; 2) Rudolf Virchow (1821–1902, Germany): rejected; 3) G. Neirynck (Belgium): rejected; 4) Jean-Bastiste-Julien d’Omalius (1783–1875, Belgium): accepted some; 5) Armand de Quatrefages (1810–1892, France): accepted; 6) Emile Cartailhac (1845–1921, France): accepted; 7) Giovanni Capellini (1833–1922, Italy): accepted; 8) Oscar Fraas (1824–1897, Germany): rejected; 9) Jens Jacob Worsaae (1821–1885, Denmark): accepted; 10) Pierre-Joseph van Beneden (1809–1894, Belgium): abstained; 11) Pierre Desor (1811–1882, Germany/Switzerland): rejected; 12) Conrad Engelhardt (1825–1881, Denmark): accepted many; 13) Carl Friedrich Schmidt (1811–1890, Russia/Germany): accepted a number; 14) Paul Hurault, Marquis de Vibraye (1809–1878, France): accepted some; 15) Augustus Wollaston Franks (Britain): accepted one.

· 4 Lisbon congress panel members and how they voted:

 Cartailhac (accepted), Quatrefages (doubted), Virchow (doubted) and Capellini (accepted) served on both panels. New members were Evans (rejected), Mortillet (accepted), Léon Paul Choffat (1849–1919, Switzerland: accepted), Paul Cazalis de Fondouce (1832–1922, France: doubtful), Juan Vilanova y Piera (1821–1893, Spain: doubted Miocene age), Giuseppe Bellucci (1844–1921, Italy: accepted), Gustave Cotteau (1818–1894, France: rejected Miocene age) and Nery Delgado (1835–1908, Portugal; accepted). Ribeiro was unwell and absent from the debate.

· 5 This was based on Mortillet’s interpretation of the site at Levallois, not Jules Reboux’s (see Chapter 3).

· 6 I assume that by ‘ lames et éclats’, Mortillet is referring to tools made on such. Otherwise he would appear to have somewhat moved the goalposts on d’Acy.

· 7 d’Acy’s (1878) long paper on the plateaux above the Somme is frequently cited as a reference to early work by d’Acy at Chelles-sur-Marne. The site of Chelles is not, as far as I can see, mentioned in this paper.

· 8 Both sites are now regarded as belonging to the Early Middle Palaeolithic, dating to after the MIS 8 glaciation.

· 9 The Middle Pleistocene mammoth being described by the early workers was probably in many cases a form of Mammuthus trogontherii, also known in dwarf form in some interglacials, rather than the Late Pleistocene elephant today known as Mammuthus primigenius and associated with the classic Mousterian industry.

· 10 Joseph Louis Capitan was at this point in his life a practicing medical doctor. In 1898 he became Professor of Prehistoric Archaeology at the Paris School of Anthropology (previously occupied by Gabriel de Mortillet) and is perhaps best associated with his later work on Palaeolithic cave art, notably at Combarelles and Font-de-Gaume in the Dordogne.