The Neolithic Age


Victoria County History



J.S. Cockburn, H.P.F. King, K.G.T. McDonnell (Editors)

Year published


Supporting documents




Comment on this article
Double click anywhere on the text to add an annotation in-line

Citation Show another format:

'Archaeology: The Neolithic Age', A History of the County of Middlesex: Volume 1: Physique, Archaeology, Domesday, Ecclesiastical Organization, The Jews, Religious Houses, Education of Working Classes to 1870, Private Education from Sixteenth Century (1969), pp. 29-36. URL: Date accessed: 18 April 2014. Add to my bookshelf


(Min 3 characters)


The first Neolithic settlers reached Britain about the 4th millennium B.C., (fn. 78) but for the thousand or so years which elapsed before the appearance of bronze there is little evidence to supply an absolute, let alone a relative, chronology for the different cultural features of the Neolithic in Middlesex. A distinction can sometimes be made in the British Neolithic between settlers and native (Mesolithic) peoples who learned the new ways of life, presumably from the migrants. (fn. 79) It is not possible, however, to divide finds in Middlesex definitely into the two divisions, for items peculiar to one group of people can appear legitimately in the equipment of another.

Only two Neolithic sites of importance have been discovered in Middlesex. (fn. 80) One comprises two small hollows or pits at Heathrow, about one mile from the Colne and over five from the Thames. These contained pieces of at least nineteen pots showing considerable variety of decoration such as might be found on 'Peterborough' or 'Mortlake' ware. (fn. 81) The other site is a causewayed camp at Yeoveney, near Staines, excavated in 1961, which yielded a large quantity of pottery similar to that from Windmill Hill (Wilts.). (fn. 82)

The presence of pottery of the Windmill Hill type suggests that its makers were the first farmers to reach the county, settling initially on the better-drained river gravels, from which they may have extended their occupations. Settled farmers possibly existed contemporaneously with comparatively nomadic groups. Even at this time there was considerable movement, at least of individuals, for axes or axe stone from as far away as the Lake District, Cornwall, and even Ireland have been found in the London and Middlesex area. (fn. 83) Some of the imported axes were manufactured from specially graded close-grained stone, and these were probably used for felling trees, although the more adze-like ones were possibly used for working timber, and even, as on the Continent, in building dwellings.

There is, however, little direct evidence of agriculture or cattle-keeping in Middlesex during the period. Nevertheless, the discovery of pottery similar to that of the Windmill Hill farmers suggests that grain was grown in Neolithic Middlesex, and an elegant flint sickle from the Thames at Chelsea is evidence of harvesting. (fn. 84) Hunting and food-gathering, however, almost certainly continued. Flint arrowheads have been found at a few places in Middlesex as well as in the Thames, although they are relatively few when compared with the hundreds from the right bank of the river at Ham, opposite Twickenham and Teddington. (fn. 85)

No evidence of salt-winning in Neolithic Middlesex has been found, nor is there any trace of the earthen long barrows found in other parts of south-east England. Conclusions about daily life can be based only on such isolated finds as a piece of worked deer antler from the Thames at Hammersmith. This 'antler comb', an implement often found with pottery of the Windmill Hill type, (fn. 86) has one end cut in a fringe with parallel teeth some two inches long, and may have been used for removing hair from deer skins (fn. 87)

The Neolithic pottery of Middlesex falls into three main groups, named after sites outside the county where similar wares have been found: Windmill Hill, Peterborough (Northants), and grooved-ware. (fn. 88)



Windmill Hill pottery (fn. 89) is good quality, hard ware, made in baggy or shouldered, usually round-based, shapes (possibly derived from leather containers), plain or sparingly decorated round the upper part with incised lines and impressed dots (fn. 90) and very occasionally with impressions of short lengths of twisted cord. (fn. 91)

In southern England the Windmill Hill people settled mostly on the chalk uplands, (fn. 92) but the discovery at Yeoveney of one of their causewayed camps shows that there was some penetration into Middlesex. The considerable quantity of Windmill Hill pottery from Yeoveney has not yet been fully studied. It includes some pieces resembling rare types found at Windmill Hill which seem close to continental forms, and others similar to varieties found in East Anglia and at Abingdon (Berks.) causewayed camp, which are considered to represent a late stage in the British evolution of the pottery. (fn. 93) The only other Windmill Hill pottery so far recognized in Middlesex is one complete pot and a few stray sherds from the Thames at Kew Bridge, Syon Reach, and Mortlake.

Peterborough ware (fn. 94) is the other major group of Neolithic pottery found in Middlesex. Its makers seem usually to have lived in low-lying areas like the Thames valley, where one of the largest concentrations of the pottery in Britain occurs. (fn. 95) In Middlesex the largest finds have come from the Heathrow site and, in lesser quantities, from the Thames at Mortlake and Hammersmith, while at least one sherd was recovered from Yeoveney. (fn. 96)

Within the Peterborough group a sequence of types, all represented in Middlesex, has been worked out. (fn. 97) The first is Ebbsfleet ware, named from pottery found in the Ebbsfleet valley in north-east Kent. (fn. 98) The fabric of Ebbsfleet pottery tends to be inferior to that of Windmill Hill ware, but the vessels are normally quite well made, in rather globular shapes with simple rims, distinct necks, rounded or occasionally sharp shoulders, and a moderate amount of decoration round the upper part of the pot, sometimes extending inside the rim. Decorative motifs include incised patterns, finger-nail and, more rarely, vertical cord impressions, and pits in the neck usually made with the finger-tips. (fn. 99) Ebbsfleet pottery developed into Mortlake ware, typified by a bowl (now in the British Museum) and other pottery from the Thames at Mortlake. Mortlake pottery is generally cruder and clumsier than Ebbsfleet ware and made of thicker, softer fabric. (fn. 1) The vessels usually have heavy projecting rims, deeply hollowed necks, sharp shoulders, and round bases, although flat bases are not unknown. (fn. 2) Close-set impressed decoration, often arranged herring-bone fashion, covers most of the exterior and sometimes the upper part of the interior. Some of the devices, like pits in the neck, resemble those on Ebbsfleet pottery, but cord impressions are much more common and new techniques, such as rustication (pinching the clay between two finger-nails), appear. (fn. 3) The pottery from Heathrow is Mortlake ware and illustrates many of the techniques used. (fn. 4) The final stage in the Peterborough sequence is the Fengate style, (fn. 5) in which the bases and contours tend to become flattened and the rims to overhang the walls of the pot in a way that suggests the overhanging-rim urns of the Bronze Age. (fn. 6) The bowl from the Thames near Hammersmith Bridge is a good example of Fengate pottery.

The relationship of Peterborough to Windmill Hill pottery is not clear, except that the Mortlake phase seems later than the earliest Windmill Hill. (fn. 7) The evidence of Heathrow, supported by the modest accumulation of Peterborough pottery from the Thames at Hammersmith and Mortlake and the habitation site across the Thames at Thorpe (Surr.), (fn. 8) shows that Peterborough people lived in the Middlesex area. Indeed, the close sequence of Peterborough pottery types from the Middlesex stretch of the Thames, which includes forms apparently transitional between Ebbsfleet and Mortlake, (fn. 9) suggests that some of the development may have taken place in the district. (fn. 10)

The third group of Neolithic pottery from Middlesex, Rinyo-Clacton or grooved ware, is characterized by flat-based pots with grooved, incised, dotted, and applied decoration. It is represented in the county only by a few sherds from the Thames, which are of some importance because they come from the middle of an area in which grooved-ware was previously unknown. The affiliations of grooved-ware and its makers are uncertain, but chronologically it seems to belong to the latter part of the Neolithic era, contemporary with later Peterborough ware and beakers, and possibly overlapping the beginning of the Bronze Age. (fn. 11)


Yeoveney (TQ. 024726): causewayed camp. Large quantity of Windmill Hill pottery; some Peterborough ware. (fn. 12) B.M.
Harmondsworth, Heathrow, 'Caesar's Camp' (TQ. 084766): settlement site consisting of two pits. Mortlake pottery; some worked flints (fn. 13) L.M. 49.87
Yiewsley: two sherds, apparently one Ebbsfleet and one Mortlake in box labelled. 'pottery found at Yiewsley with the flints'. L.M. unnum.
Mortlake, perhaps S. side of river: (fn. 14) plain Windmill Hill sherd with slightly out-turned rim. L.M. A.10560
Four Ebbsfleet sherds. L.M. A.10215, A.10573, A.13671, (fn. 15) C.953 (fn. 16)
Three Peterborough sherds, apparently transitional between Ebbsfleet and Mort-lake styles L.M. A.13693, (fn. 17)
C.954, C.955
Mortlake bowl (alleged to have been found on same site and beneath same calcareous seam as beakers B.M. 1909.5.18.13-14); six sherds. B.M. unreg., 1909.6.25.1 (fn. 18) L.M. A.10213, (fn. 19) A.13666, (fn. 20) A.13668, A.13670
Fengate sherd. (fn. 21) Dr. Isobel Smith has recognized two other sherds from same vessel. L.M. A.13667
B.M. unreg.
Sherd, probably grooved-ware, with flat rim. Exterior of wall decorated with horizontal herring-bone pattern of grooves and top of interior with four horizontal grooves. L.M. A.9964
'Nr. Kew Bridge, off Strand Island, from ballast' (c. TQ. 194776): Windmill Hill bowl; hard black ware with some grit; suspicion of out-turned rim; plain. L.M. Lay.P.13
Strand-on-the-Green: grooved-ware or beaker sherd. Flat base and wall with rough grooved decoration of horizontal lines and herring-bone pattern. L.M. A.27166
Hammersmith, perhaps S. side of river: (fn. 22) three Ebbsfleet sherds. L.M. C.940, C.941, (fn. 23) C.944
Peterborough sherd; simple rim; no neck; decorated with two rows of finger-nail impressions on outside and one on inside rim. L.M. unnum.
Two grooved-ware or beaker sherds. One with slight horizontal grooves and finger-nail impressions. Other with small circular depressions and grooves. L.M. C.946, (fn. 24) C.948
'Nr. Hammersmith Bridge' (c. TQ. 229781): Fengate bowl (half); flat base and incipient overhanging rim; decorated with twisted-cord impressions, pits made with round implement, and irregular semicircles of finger-nail impressions set end to end. (fn. 25) Partial analogies from Peterborough and Ford (Northumb.). (Also said to have come from Thames at Wandsworth). (fn. 26) L.M. Lay.P.21
Syon Reach: Windmill Hill sherd; small irregular out-turned rim; plain. L.M. A.23378
Mortlake sherd. (fn. 27) Weybridge Mus.
Windmill Hill sherd. Decorated on rim with oblique incised lines and below rim with roughly horizontal lines of hyphen-like impressions possibly made with string of small seeds or toothed implement. (fn. 28) Ebbsfleet sherd. L.M. A.3624
L.M. A.3624

Some hundreds of axes (fn. 29) which can be called Neolithic have been found technically in Middlesex, although usually off its banks in the Thames. Most of them, whether chipped or polished, are of flint. A smaller proportion is of ground or polished stone, which was sometimes imported from other parts of Britain. Few of these axes are broken, although some seem to have been reconditioned by chipping. The quantity of axes is large, but it should be remembered that they represent the losses of durable objects over a period of nearly two thousand years. The evidence on which a chronology for the various types of stone axes could be constructed does not exist; nor is the relationship of the Neolithic axe to the Mesolithic 'pick' clear.

Axe finds are concentrated at several points. (fn. 30) Most of the land axes have been found south of a line from Yiewsley to Finsbury. They are, therefore, from the surface of the Pleistocene gravels, the brickearths, and the more recent alluvial areas of Middlesex. There is hardly any on the London Clay. Two landward concentrations are notable: one in the Lea valley between Hackney and Walthamstow, and another, consisting of some twenty-five axes, covering the City of London and parts of Holborn. The London area lies on the Taplow gravels, which elsewhere in Middlesex have not produced so many axe finds.



The Thames concentrations are of some interest. Staines, near the causewayed camp, seems to be more prolific, in contrast to the lesser scatters of Sunbury and Hampton. The densest find areas are in the Teddington-Twickenham and Isleworth-Brentford stretches. Downstream the finds are thinner, except at Hammersmith and Mortlake. Below Chelsea Bridge axes are not plentiful except where the river passes the older parts of London and the City. Down-river from the Tower finds are few.

Only in the area around Yiewsley, West Drayton, and Dawley is there any rough correlation on land between axes, pottery, and leaf-shaped arrowheads. A discoidal knife was found at Yiewsley, and at Harefield, an outlier on glacial gravels, three axes have been found. The best correlation between river finds of axes and pottery appears in the stretches of the Thames at Mortlake and Hammersmith. (fn. 31) There is a possible correlation with the relatively plentiful finds of leaf-shaped arrowheads between Twickenham and Ham and, to a lesser degree, off the Fulham bank.

In addition to the axes, and the numerous indeterminate or unstudied flint types, there are from Middlesex a small number of flint objects of distinctive forms. Those with a claim to be Neolithic are included here; but, since all are stray finds, they can be dated only by reference to similar objects found outside the county. They come, generally speaking, from areas such as Yiewsley and the Thames to the west of London which have produced pottery and axes, and so confirm the overall distribution pattern. They fall into four groups.

The first form is the petit-tranchet derivative arrowhead, (fn. 32) which has a straight chisel edge instead of a point. These arrowheads are derived from a type predominant in later Mesolithic times in Britain, Scandinavia, and other areas. (fn. 33) Outside Middlesex they have been found at various Neolithic sites, (fn. 34) particularly with grooved-ware pottery, (fn. 35) but they were evidently still in use in the earlier part of the Bronze Age. (fn. 36)

The second form is the leaf-shaped arrowhead. (fn. 37) This was one of the innovations of the Windmill Hill people and one of their most distinctive pieces of equipment, (fn. 38) but such arrowheads have been found in various parts of England and Wales with Peterborough (fn. 39) and grooved-ware pottery. (fn. 40) They may still have been used in the Bronze Age, (fn. 41) and have even been reported in an Iron-Age context. (fn. 42) The known Middlesex examples are all of the true leaf-shaped form, not the lozenge-shaped variant.

The third form is the discoidal knife, (fn. 43) a roughly disc-shaped or triangular implement with polished edges, found chiefly in eastern Britain, and perhaps used in the preparation of hides. (fn. 44) This is not a common type, but the Thames valley in and near Middlesex has produced a comparatively large concentration. (fn. 45) Discoidal knives seem to have been exclusively British tools, possibly of Mesolithic antecedents. (fn. 46) They have seldom been found with associations, and are usually assumed to be Neolithic, (fn. 47) although there is some evidence that they were used in Beaker times. (fn. 48)

The fourth form is the sickle. (fn. 49) The two so far recognized in Middlesex are of the long, curved, single-piece variety presumably designed for hafting at right angles to a straight handle, (fn. 50) and probably used for cutting corn and reeds. (fn. 51) The type has analogies in Holland, (fn. 52) and its distribution in Britain, where it is not very common, is confined to eastern districts, one group occurring along the Thames valley and estuary. (fn. 53) Available dating evidence suggests that the sickles were used in Neolithic times although not introduced by the Windmill Hill farmers. (fn. 54) How long they continued in use is uncertain: metal sickles are not a feature of the earliest part of the British Bronze Age. (fn. 55)

A skull, trepanned and partly healed, was found in the Thames at Hammersmith. Trepanning was particularly common in the Beaker period but pollen samples from the skull analysed by Professor G. W. Dimbleby suggest that it belongs to the Neolithic period.


77 The authors of this and the following sections are indebted to Prof. W. F. Grimes for his help and guidance.
78 The authors are indebted to Dr. Isobel Smith for details of British radio-carbon datings which put the first Neolithic settlements at about 3700 B.C.
79 For a full account of the period see N.C.B.I.
80 It is not possible in the limited space available to list every Neolithic find or site. Full details for Mdx. are available in the L.M. card indexes. Fresh material and reinterpretations (from Oct. 1961) of the material discussed below is submitted from time to time to the editors of T.L.M.A.S.
81 E.D.S. 186-97.
82 The excavation is to be published by Mr. R. Robertson-Mackay, who has kindly furnished preliminary details. Aerial photographs taken in 1966 confirmed the existence of a second causewayed camp at East Bedfont (TQ. 080737): Antiquity, xl (158), 145.
83 Dr. F. S. Wallis carried out petrological analysis of selected Mdx. stone axes in time for inclusion of preliminary details in this section.
84 Said to be 'Secondary Neolithic': L.P.A. pl. 111 (8).
85 Many of the arrowheads and small flint scrapers from Ham are in the J. Orrell Knowles Collection in L.M. (60. 176). Since BA arrowheads were found in that area there was possibly some continuity in its use for hunting.
86 The authors are indebted to Dr. Isobel Smith for pointing this out.
87 This interpretation is based on N.C.B.I. 43-84.
88 The term Windmill Hill is here used for pottery of a generally Primary or Western Neolithic character as defined in N.C.B.I. 66-75, and not merely the particular variety found at Windmill Hill. Since this was written, views on Neolithic pottery have been revised. It is now generally believed that Ebbsfleet pottery is akin to Windmill Hill types, and that Mortlake and Fengate pottery were developed out of Ebbsfleet styles in the later Neolithic period. The term 'Peterborough' should probably be dropped. See L'Europe a la fin de l'age de la pierre, ed. J. Böhm, 568; Antiquity, xl. 177-81.
89 A. Keiller, Excavations at Windmill Hill, CompteRendu 1st. Internat. Congress Prehist. Sciences (Lond. 1934), 135-8; N.C.B.I. chs. ii, iii.
90 Arch. Jnl. lxxxviii. 67-158; N.C.B.I. 66-75.
91 Antiq. Jnl. vii. 455.
92 Arch. Jnl. lxxxviii. 85-88; N.C.B.I. ch. ii, fig. 1.
93 Information on Yeoveney supplied by Mr. R. Robertson-Mackay. For Abingdon and E. Anglian ware see N.C.B.I. 72-74. See below, n. 12.
94 Archaeologia, lxii. 333-52; Antiq. Jnl. ii. 220-37.
95 Arch. Jnl. lxxxviii. 120; N.C.B.I. 303, fig. 48.
96 Ex inf. Mr. R. Robertson-Mackay.
97 Antiq. Jnl. xix. 405-20; E.D.S. 194. This school of thought is not followed by all archaeologists.
98 Antiq. Jnl. xix. 405-20.
99 Ibid. 409-20; N.C.B.I. 308, fig. 49.
1 Some Mortlake-type pots from Peterborough show that they were built up from rings of clay: Arch. Jnl. lxxxviii. 112; N.C.B.I. 308.
2 Antiq. Jnl. ii. 223, 227, 237; Arch. Jnl. lxxxviii. 113-14; N.C.B.I. 308-10.
3 Arch. Jnl. lxxxviii. 72, 110-58; N.C.B.I. 308; E.D.S. 194-7.
4 Ibid. 186-97. See illustration facing p. 52.
5 Named from a site at Peterborough.
6 Dawn, 332.
7 N.C.B.I. 310, 312-16.
8 E.D.S. 181-5.
9 Sherds from Thorpe show similar features: ibid. 185.
10 Ibid.
11 P.P.S. ii. 178-210; N.C.B.I. 321-46, fig. 48; Dawn, 334; Antiquity, xl. 180-1.
12 R. Robertson-Mackay, 'The primary Neolithic settlement in southern England: some new aspects', Proc. International Cong. for Prehistoric and Protohistoric Sciences (Rome, 1965), ii. 319-23.
13 E.D.S. 186-97.
14 Arch. Jnl. lxxxvi. 82-84.
15 Ibid. fig. 1 (2).
16 Cf. Thorpe, no. 1: E.D.S. 185, fig. 71 (1).
17 Arch. Jnl. lxxxvi. fig. 1 (1): internal decoration on rim, similar to external, not shown.
18 Ibid. lxii. 340, pl. xxxvii (3); lxxxviii. 122, fig. 18 (2).
19 Ibid. lxxvi. 82, fig. 1 (4).
20 P.P.S. xx. 228. n. 3.
21 Arch. Jnl. lxxxvi. 82, fig. 1 (3). Decoration inside rim not shown.
22 Ibid. 86.
23 Cf. Thorpe, no. 1: E.D.S. 185, fig. 72 (6).
24 Cf. Lyonesse (Essex) pottery: P.P.S. ii. 190 sqq.; 'Woodhenge' (Wilts.) pottery: N.C.B.I. fig. 58 (2); P.P.S. ii. 29, fig. 14 (1).
25 Archaeologia, lxix. 11-12, figs. 9, 10.
26 Antiq. Jnl. ii. 220-37; Arch. Jnl. lxxxviii. 67-158; N.C.B.I. 314.
27 Antiq. Jnl. v. 431-2, fig. on p. 431; Arch. Jnl. lxxxviii. 152.
28 Cf. N.C.B.I. 86; Abingdon: Antiq. Jnl. vii. 436-64, pl. lii, fig. 2; viii. 461-77, pl. lxxii, fig. 2, pl. lxxiv, fig. 2; Avebury (Wilts.): ex inf. Dr. Isobel Smith.
29 These include the 'celts' of former writings. Objects which might better be described as adzes are included under the heading of axes. The economic background of stone axes is summarized in Ec. H.R. 2nd ser. xviii (1), 1-28. Jadeite axes are a special problem and it has been suggested that they came over from the lower Rhineland with Beaker groups: P.P.S. xxix. 133-72: Vauxhall Br., B.M. 1907, 6-19, I and L.M. 31.48.
30 Lists and indexes of the axes are in the L.M. Many axes are still in private hands or in museums which have not made them available for study.
31 The pottery in these reaches seems to be of Ebbsfleet, Mortlake, and Fengate types. Two discoidal knives with polished edges have been found in the same parts of the river.
32 Examples are known from Harefield (one, L.M.), Yiewsley (c. 3, L.M.), Thames, Wandsworth Park (one, Nat. Mus. of Wales).
33 Clark, Mesolithic Settlement in N. Europe, 142, 144, fig. 51; Clark, Prehistoric Europe: the Economic Basis, 36; N.C.B.I. 285-6, 369; L.P.A. fig. 1 (lower, 2).
34 N.C.B.I 78 et passim.
35 Ibid. 343.
36 Arch. Jnl. xci. 32-58.
37 Examples are known from Acton (one, Gunn.), Thames, Barn Elms (one, L.M.), Brentford (two, L.M.), Mortlake (one, L.M.), Richmond (one, Colchester Mus.: Arch. Jnl. lxxxvi. 77), Twickenham (Nat. Mus. of Wales and L.M.), Wandsworth (one, L.M.). Smith, Man, 311, mentions one from Shacklewell.
38 Proc. Soc. Antiq. iii (2), 168-72; Archaeologia, lxxvi. 81-106; Arch. Jnl. lxxxviii. 42-43, 78, 368, et passim.
39 N.C.B.I. 311.
40 Ibid. 343.
41 Prehist. Communities, 154.
42 Archaeologia, lxxvi. 106.
43 Most of the Mdx. examples are listed in P.P.S.E. Anglia, vi. 40-54. Two additions are now in L.M.: one 'found during filling of Ham gravel-pit with filling from Chelsea Reach', the other from the Thames at Richmond lock and weir.
44 P.P.S.E. Anglia, vi. 44.
45 Ibid. 40-54.
46 Ibid. 46, 51; N.C.B.I. 286.
47 Ibid. passim.
48 P.P.S.E. Anglia, vi. 40-54.
49 Examples are known from the Thames, Chelsea (B.M.; presumably J. Evans, Ancient Stone Implements (2nd edn. 1897), 357; P.P.S.E. Anglia, vii. 67-70, 72, no. 10, pl. viii); Hammersmith (L.M.: P.P.S.E. Anglia, vii. 72, no. 11, fig. 1).
50 P.P.S.E. Anglia, vii. 67-68; cf. Antiquity, xii. 152, pl. 11.
51 N.C.B.I. 365.
52 P.P.S. ii. 208.
53 P.P.S.E. Anglia, vii. 67-81, fig. 6.
54 Ibid. 76-81; P.P.S. ii. 181; iv. 32; N.C.B.I. 311.
55 P.P.S. iv. 33-35; cf. Clark, Prehist. Europe, 111.