^{page 97 note 1} The writer started without experience of excavation under such conditions. In a frost, for example, a floor cannot be studied for stake-holes or other features: the conditions destroy colour-contrasts. In a thaw the floor (with its stake-holes) comes away on one's boots. Vertical sections disintegrate and collapse. The only way to deal with the difficulty that he could discover was to examine and record all floors, faces, or deposits immediately on exposure. But the limitations which this imposes on the study of difficult problems are obvious.

^{page 98 note 1} Sheeplays 279’ shows a slight variation on this procedure.

^{page 99 note 1} Archaeologia, lxxxvii, ‘Two Bronze Age Cairns in South Wales’, by Cyril Fox, pp. 129–80, esp. p. 145.

^{page 99 note 2} In an Appendix to a forthcoming paper in Arch. vol. lxxxix, by the present author.

^{page 100 note 1} A ditch surrounding the barrow was searched for by extending the north-south trench in both directions, but no satisfactory evidence of such was obtained.

^{page 100 note 2} For range of diameters, see Appendix I, p. 123.

^{page 100 note 3} Stake-holes of exactly this character ‘3—4 inches in diameter outlined by a ring of oxide-impregnated soil’ and ‘8 inches to a foot deep’ were found by H. Noel Jerman in a barrow on Caebetin Hill, Kerry, Mont., in 1931 (Montgomeryshire Collections, 1932, pp. 176–81). The circle was 17 ft. 4 in. to 19 ft. 6 in. in diameter; there were 36 holes, from 15 to 32 in. apart. From the fragmentary remains of B.A. pottery the present writer judged the barrow to ‘be contemporary with the first phase of the overhanging-rim urn’.

^{page 102 note 1} The holes in the photograph (pl. xxi,a) are 2½ × 2½ in. and the stained circles, probably the size of the stakes, 3½ × 3½; the depth of the holes is 10 in. or more.

^{page 103 note 1} The range is probably greater than these figures suggest; only a small portion of the circle was proven.

^{page 103 note 2} By Mrs. P. Murray-Thriepland; they were indeed the first holes to be located in the barrow.

^{page 103 note 1} Compare the 28-ft. circle, where a stained ring was 3½ × 3 in. and the central hole 2½ × 2½ in.

^{page 106 note 1} The only place in the area covered by the turf-stack where hard-pan was not present.

^{page 110 note 1} A dark smear on the ancient surface, freshly exposed with the polished blade of a spade, is the primary indication (p. 104).

^{page 111 note 1} This area has another distinctive feature; in it the hard-pan rises here and there in hummocks (over decayed heather or turf?).

^{page 111 note 2} Mortimer (Forty Years’ Researches, pp. 154 ff. and figs. 397 and 400) plans a Yorkshire barrow, his no. 23, which showed a double circle of stake-holes one within the other, which he (I think correctly) regarded as the framework of a wattled circular hut. The diameter of Mortimer's outer circle is 28 ft. He describes also a series of stake-holes in his barrow no. 41 (p. 182), but the account is inadequate. Mortimer regarded these as ‘abandoned’, not symbolic dwellings. They were of Early Bronze Age date.

^{page 111 note 3} The holes may seem small to accommodate posts capable of supporting such a large roof. In this connexion a remark made in a letter from my friend T. C. Lethbridge, F.S.A., is relevant. ‘In the Viking houses’ (in Iceland), he says, ‘there are rows of posts marking aisles. I saw three cleared ones, and was interested to see how small the post-holes were. They could hardly have had anything in them much bigger than a prop for a clothes-line, and yet one of the houses was huge, 100 feet long inside at least.’

^{page 112 note 1} I cannot explain the absence of vertical marks of decayed wattling in the two faces of the north trench (section AA’) where the 28-ft. circle—the wall of the hut —crosses this trench. It is the only serious weakness, I think, in the argument. But the destruction of the wattled wall on this side, in order to make a solid and continuous turf-stack, does not appear improbable.

^{page 112 note 2} The branch-covered dome was no part of this accidental irregularity: no hardpan formed on it; it had not been trodden.

^{page 112 note 3} Archaeologia, lxxxvii, pp. 147–8, 157.

^{page 112 note 4} As well, no doubt, as serving the practical purpose of providing a means of dumping turf into the hut from the top, after the filling from the bottom had reached the limits of practicability.

^{page 113 note 1} The only hint provided by the stake-holes of such an axial lay-out is in the N.W. quadrant, where a radial alignment of additional holes points, somewhat uncertainly, to the primary burial. A duplication of stake-holes at two other points in this quadrant should also be noted (see plan).

^{page 114 note 1} Mortimer included, in the observations already referred to, evidence that the ‘huts’ in two of his barrows had been wrecked before envelopment, op. cit., pp. 156 and 183. See also an important paper by Mr. Stuart Piggott—Timber Circles, a Re-examination, Arch. Journ. 1939, esp. p. 218. Here he urges that the majority of the Dutch palisade barrows began with a burial made on the floor of a hut, and cites evidence for the collapse of a hut in one barrow.

^{page 114 note 2} Archaeologia lxxxvii, p. 157.

^{page 118 note 1} Dr. F. J. North tells me that the hollow may well be natural with such an underlying rock as lias, and due to collapse.

^{page 118 note 2} For Breach Farm see W. F. Grimes in Proc. Prehist. Soc. 1938, pp. 107–21. For Crick see H. N. Savory in Arch. Camb. 1940, pp. 169–91; Sutton 268’ will be published in Archaeologia lxxxix. The subject of stone rings in South Wales is discussed in Archaeologia lxxxvii, pp. 160–3.

^{page 120 note 1} Of considerable thickness. The site is close to an ancient house, and may have at one time been a dump for soil.

^{page 121 note 1} The closest parallel to Six Wells 267’ I have come across is Combe Beacon, Somerset, a barrow excavated by Mr. H. St. G. Gray, F.S.A. Mr. W. J. Varley also records what may be a similar structure—Robin Hood's Tump, Cheshire (refs. Proc. Som. Arch, and N.H. Soc, 1935, lxxxi, pp. 83–107, and Prehistoric Cheshire, W. J. Varley and J. W. Jackson, 1940, fig. 6). A barrow, CCLXX, in the parish of Fylingdales, Yorks., opened by Greenwell and recorded in Archaeohgia, lii, p. 41, contained a structure which is not like mine, but which conforms to my definition of a ‘ritual’ barrow. Barrows with graves or cists lacking burials are not uncommon; I do not regard these as ritual barrows, but rather as a form of cenotaph. For examples, see Greenwell, loc. cit., pp. 23–4, 35–6, 60. ‘Ritual Pits’ ancillary to burials are of course common enough. See e.g. Archaeologia, lxxxvii, p. 141. Mr. L.V. Grinsell has been very helpful in searching for these parallels.

^{page 121 note 2} The destruction of a broad belt across the barrow by agricultural operations introduces an element of doubt; but the plan does not suggest this zone as a likely area for a primary burial deposited by a people of this culture, having regard to the centralization of the primary deposits in Sheeplays 293’ and 279’.

^{page 122 note 1} ‘Timber circles: a Re-examination’; Arch. Journ. 1940, xcvi, pp. 193–222Google Scholar; esp. p. 219 and figs. 12, 13.

^{page 123 note 1} To the list of barrows in Archaeologia, lxxxvii, p. 162, add Crick, Arch. Camb 1940, pp. 169 ff., and Sutton 268’, Llandow, Glam. (publication pending).

^{page 123 note 2} Clark, Grahame, ‘The Timber Monument at Arminghall and its Affinities’, Proc. Prehist. Soc. ii, 1936, pp. 1—51CrossRefGoogle Scholar: Mrs. M. E. Cunnington, Woodhenge, Devizes, 1929, and The Sanctuary, Hill, Overton, Wilts. Arch. Mag. xlv, 1931, pp. 300 ffGoogle Scholar.: Leaf, C. S., Bronze Age Barrows at Chippenham, Cambs., Proc. Camb. Antiq. Soc. xxxix, 1939, p. 36Google Scholar: J. R. Mortimer, op.cit.: Piggott, loc. cit.:W.J. Varley, ‘The Bleasdale Circle’, Antiq. Journ. 1938, pp. 154–71: Cf. Giffen, A. E. Van, Proc. Prehist. Soc. iv, 1938, pp. 266–71Google Scholar.

^{page 123 note 3} Arch, lxxxvii, loc. cit., pp. 156 ff.