Site of Late Celtic Settlement at All Cannings Cross Farm.

Scale 6in. = 1 mile.

VOL. XXXVII.-NO. CXVIII.

+ Spots dug over.

2 T

useless to collect them further. They are all either of flint or sarsen; the former are by far the most common, being apparently in proportion of about six to one of sarsen.1 The smallest measures 6in., the largest 13in. in circumference, but these are exceptional, the majority being from 8in. to 10in. They vary somewhat in shape, but this may be chiefly due to the amount of usage to which they have been put, and it is possible to trace stages in the development of the hammerstone, from an angular youth to a rounded old age. In some cases it appears that the flint chosen was first roughly chipped into a form resembling that of a rude Palæolithic implement, but sometimes a naturally-broken flint was used. It seems that the object of the initial shaping was, at any rate partly, to obtain sharp edges, for it is invariably these sharp edges that are bruised and worn down by the pounding or hammering action to which they have been put. In the same way when a naturally broken flint was used, it seems to have been chosen for the sake of some point, or sharp edge, desirable for the purpose in view, Merely as a result of thus hammering down all the angles, stones that have been much used tend to become rounded in form, and in the last stage when all the sharp angles and points have been knocked off they become almost perfectly round and ball-like. In these balllike specimens the surface is bruised all over, and this suggests that when the sharp edges were worn away the stones were then put to some further use that caused the bruising of the whole surface.

The sarsen "hammerstones," on the other hand, have been used for grinding or rubbing on some hard substance, and scarcely, if at all, for hammering or pounding; their surfaces are not bruised

66

1 With regard to the relative frequency of flint and sarsen as a material for "mullers or hammerstones" in the county, it seems that about Marlborough and Clyffe Pypard in the north, sarsen examples are found more often than flint, whereas in the south of the county and over most of Salisbury Plain the hammerstones are almost invariably of flint, so much so that the only examples of sarsen hammerstones in the Blackmore Museum come from the north of the county. This may well be cited as a further piece of evidence that sarsens never existed in any numbers on the Plain or anywhere in the south of the county.

like those of the flints, but are smoothed, and sometimes even slightly polished. The rounded flint hammerstones have also in some cases evidently been utilised for the rubbing or grinding process as is shown by their extra smooth surfaces, and by the characteristic ridges produced on their rounded outlines.

Several suggestions have been made as to the possible industry in which these hammerstones may have been used. The probability of their having been used in the manufacture of other flint implements is negatived by the absence of flint debris such as rough flakes and broken flint in any quantity, and of broken and imperfect implements. It was suggested that they might have been used in the preparation of clay, and in crushing flint, etc., to mix with the clay, for making pottery, but no evidence of pottery having been made on the spot was found.

There are, however, a great many small fragments of sarsen stone scattered over the site, and many sarsen boulders on the adjacent downs, and it is not unlikely that at one time there were many sarsen boulders in the immediate vicinity. A quern stone, mealing stones, and rubbers of sarsen, found on the site, show that sarsen was used for other purposes than hammerstones, and the presence of fragments of broken sarsen suggests that boulders have been broken up on the spot. It seems, therefore, probable that some sort of industry of which sarsen was the raw material was carried on here. To be worn as they are, the hammerstones must have been used on some hard material, and, with the exception of flint, sarsen is the only hard stone to be found on the site.

That similar hammerstones of flint and sarsen were actually used for dressing sarsen appears to be proved by the tools found in the excavations at Stonehenge in 1901. The implements found at Stonehenge seem to have been chiefly used for hammering or pounding, and rarely for rubbing or grinding. The surface thus dressed by pounding would no doubt do well enough for the stones of Stonehenge, but it might well have been found necessary to supplement this with a further process of grinding if querns, nealing stones, and such-like smaller objects were being prepared. It is suggested, therefore, that it was in the manufacture of sarsen

querns and mealing stones that the hammerstones on this site were mainly used.1

THE POTTERY.

A large quantity of pottery was found, but only in fragments, the half of a small bowl (Plate II.) being the most complete piece.

The majority of the pottery appears to be hand-made, but some of it has been turned on a wheel of some sort. Generally speaking it may be described as well made and well baked, some of the pieces being burnt to a uniform red right through, and as a rule the ware is thin for the size of the vessel. Most of the pottery is a reddish brown in colour, the shades varying according to the sharpness of the baking, but some of it is black right through, the result, probably of a fuming or dipping process.

Similar hammerstones are actually used in making querns in Africa at the present time. For the purpose of hollowing out the quern, the method employed there is the constant dropping, or throwing, of a stone held in the hand, on to the surface of the quern in the course of preparation. Mr. A. D. Passmore, of Swindon, who has during the past year been engaged in excavating in the Soudan, and who has seen querns in the course of preparation by the women there, has kindly given the following account of the process:- I observed near Wad Medani, south of Khartoum, a woman (to what tribe she belonged I cannot say) making a plain hollowed quern or mealing stone of granite for use with a rubber. She continued the work for nearly an hour whilst I watched her. At times she threw the hammerstone hard on the quern, which made it re-bound a considerable height, but becoming tired she gradually reduced the height between her hand and the quern till the stone re-bounded, itself, into her hand, from about four inches. No effort was required to do this, and becoming refreshed she gradually brought force into play and the harder she worked the greater the re-bound. When I first saw her the hammerstone was fairly round, and just before leaving I examined it carefully and it was exactly like the so-called hammerstones of our Downs. What the stone was like before, I cannot say, but it is obvious that only more or less round stones would be selected. The hour's work observed by me produced a hollow in the quern stone, more than might have been expected. I questioned the woman as to the finish of the quern stones and I understood from her that they are finished by rubbing." This process explains the bruised sufaces and ball-like roundness of so many of the hammerstones, and the use to which they were put; but Mr. Passmore's account does not explain the use of those stones on which the sharp edges and points only are bruised; these latter were probably used in the shaping and rough dressing of the outer parts of the querns.

It does not seem likely that it all came from the same kilns, or that it was made in the immediate neighbourhood, for some of it is mixed with pounded flint, some of it is very full of sand, while much of it is mixed with broken fossil shells from the oolitic beds.

It is difficult to determine the exact shape of many of the vessels from the broken fragments, but it is clear that the greater number of them were characterised by a decided shoulder (Plate III., 2, 3; Plate IV., 1). Most of the shouldered vessels seem to have had a row of ornamental indentations round the shoulder, and many had a similar row round the rim as well (Plate III., 2), so that comparatively few vessels were quite devoid of ornament. The indentations vary in shape from that of the print of a finger tip to that of a slight thumb-nail impression, but they do not appear to have been actually made with the fingers, but by tools doubtless designed for the purpose. Some of the vessels were provided with eyelet handles, and several of these were found, all of the same type as that shown below.

Late Celtic pottery handles from All Cannings Cross Farm.

Fragments of the type of bowl shown on Plate II., 1, 2, were fairly numerous. These bowls are better made and of a finer ware than the bulk of the pottery, and appear to have been turned on a wheel; they are small vessels, generally measuring some 5in. or 6in. in diameter at the rim, and some 4in. or 5in. in height.