In 2012, Cotswold Archaeology recovered a largely complete upper stone of a rotary quern on a late prehistoric and Romano-British settlement site located c. 800 m from the north coast of Somerset, England, west of Wick Moor at Hinkley Point (ST 205 453). Full details of the site will be presented in the site reportFootnote 252 along with illustrations of the quern but, due to the object's significance, a fuller discussion of it is offered here than was within the remit of the site report. The quern has a maximum surviving diameter of 34 cm, is 13 cm thick and weighs 25 kg. It was placed in an inverted position towards the top of a pit along with some other, unworked, stones.Footnote 253 It was one of two querns from the site found in this type of placed deposit. At some point before deposition, the edges had been trimmed resulting in a crude sub-oval shape, but it was probably circular or slightly oval during use.
When examined by eye, the quern is a hard, pale greyish brown sandstone. Damage to one edge of the quern allowed the removal of a small sample for thin-section analysis. This revealed the stone to be a fine to medium-grained, quartz-cemented sandstone. The rock has undergone pressure solution and contains infrequent feldspar, mica and clay minerals; it is typical of the Devonian Upper Old Red Sandstone, probably from the Portishead area, where the sequence has not been subdivided and is all known as the Portishead Beds.Footnote 254 This fine-grained rock was used for querns much less often than its coarser and more pebbly counterpart found both in the same beds, and more often in the Quartz Conglomerate of the Forest of Dean and Wye Valley.Footnote 255 Full details of the petrographical analysis can be found in the site report and the reader is referred there for further information.Footnote 256
The quern has a rounded bun-shaped profile. In the centre is a wide cylindrical feed pipe into which the grain dropped from the shallow basin-shaped hopper, little more than token in its size (figs 36d, 37 and 39f); for fuller illustrations the reader should refer to the site report.Footnote 257 Around part of the circumference is a shallow wide groove and projecting from one side of the circumference is a lug containing a socket for a vertical handle. The inside of this socket is worn very smooth suggesting that the handle was loose within it, rather than being secured with lead as is sometimes found. The grinding surface is steeply concave and curved and the whole quern was originally pecked, although the innermost and outermost sections of the grinding surface have been worn smooth through use.
FIG. 36. Form of querns from south-western England. a and b. Meare West, Somerset (after Watts Reference Watts and Shaffrey2017, fig. 8.5); c. Wanborough, Wilts. (after Anderson et al. Reference Anderson, Wacher and Fitzpatrick2001); d. Hinkley Point (after Mudd et al. Reference Muddin prep.).
FIG. 37. The Hinkley Point rotary quern.
The most noticeable characteristic of this quern is the socket for a vertical handle that protrudes from the circumference of the quern. This late Iron Age or early Roman rotary quern is the first of such a design to be found in southern England. Although early Roman Mayen (German) querns were operated via the use of a vertical handle secured in an iron fitting,Footnote 258 none of the surveys of querns in southern England have found native-made querns with any type of socket for a vertical handle.Footnote 259 This indicates the absolute preference for horizontal handles in the region. In southern England, true beehive querns such as those seen in northern and eastern England were never in use and a flatter but still rounded bun-shaped form was preferred. The thickness of these querns is within the lower range of thicknesses seen on the northern beehive querns.Footnote 260 They typically have handle sockets cut horizontally or close to horizontally into the thick sides (fig. 36a–b). Occasionally they are laid laterally across the top (fig. 36c) and rarely both appear on the same quern.Footnote 261 When querns developed into the disc type from the first century a.d., the lateral handle socket remained in use.
This difference between a horizonal handle socket and a vertical one is significant because of the implications for the way the quern was operated. Ethnographic observations indicate that querns with a vertical socket can be operated either by a short handle with the user (or users) kneeling next to or leaning over the quern, or with a much longer pole, perhaps a ceiling pole, and the user standing above the quern. In contrast, the horizontal sockets, whether laid laterally across the top of thinner disc querns or cut into the side of thicker bun-shaped and beehive querns, were probably oscillated rather than fully rotated. Operating a rotary quern with a horizontal socket in this way seems counter intuitive because it would be easier to keep the momentum of a heavy quern turning in one direction. However, it is now widely accepted by quern specialists that a push-pull motion was employed.Footnote 262 This theory is based on the very uneven wear found on a large number of beehive and bun-shaped querns in England and on wear to the inside of the feed pipe that can only have been caused by the ‘upper’ side of the upper stone leaning on the spindle.Footnote 263 The pressure exerted by the user on the handle side of the quern where they are pushing and pulling causes greater wear, and on the thicker beehive querns is often later righted by the addition of a second handle on the other side.
This bun-shaped quern with a projecting socket for a vertical handle is most closely related to a form of quern that occurs in north-eastern Ireland and south-western Scotland. In Ireland, Caulfield categorised this as his type B2 to broadly include all querns with vertical handle sockets. He identified 48 examples of this form, many with the handle in an extension from the main stone.Footnote 264 In Scotland, the form is known as the Fintry type after the location where the first example was found; it occurs beyond the Roman Forth/Clyde frontier, in the far north of Argyllshire, Stirlingshire and on some of the Hebridean islands.Footnote 265 A flatter form of the same design also occurs in Scotland but because of the difference in overall form is not considered to be the same type of quern.
In addition to the finds in Scotland and Ireland, the Fintry quern form has also been found on the Isle of Man at Ballacagen Lough A (fig. 39c),Footnote 266 where a single example is known. Occasional querns with vertical handle sockets have also been found in Wales. Two with flatter profiles are known from Lampeter and Carmarthen Museums but these are more in keeping with the flatter Scottish disc types with sockets for vertical handle mentioned above.Footnote 267 However, bun-shaped Fintry types are known from Tomen-y-Mur in north Wales and from Carmarthen Museum (fig. 39d).Footnote 268
FIG. 38. The locations of ‘Fintry-type’ querns mentioned in the text. 1. Hinkley Point; 2. Tomen-y-Mur; 3. Ballacagen Lough A; 4. Fintry; 5. Appin; 6. Balfron; 7. Lismore; 8. Strontian; 9. East Coldoch; 10. Derry; 11. Picona.
FIG. 39. Examples of ‘Fintry-type’ querns. a. Balfron, Stirlings., Scotland (after MacKie Reference Mackie1995 with suggestion of original depth); b. Derry City, Northern Ireland (after Caulfield Reference Caulfield1977, 111); c. Ballacagen Lough A, Isle of Man (after Bersu Reference Bersu1977); d. Tomen-y-Mur, Wales (after Grover Reference Grover1871); e. Picona, Spain (after Maya and Angel de Blas Reference Maya and Angel de Blas1973); f. Hinkley Point, England (after Mudd et al. Reference Muddin prep.).
Perhaps most striking is the parallel that can be drawn between this quern and an example from Castro de la Picona in northern Spain (fig. 39e).Footnote 269 This unstratified Spanish example was first recorded by archaeologists in 1965.Footnote 270 Stylistically it is very different to the Hinkley Point quern because it is elaborately decorated, but its overall form is very similar. It has a concave grinding surface, wide feed pipe and shallow hopper with a vertical handle socket set just outside the main circumference of the quern.
A greater range of profiles and handle fittings was in operation in the Spanish Iron Age and Roman period than is recorded from southern Britain. This includes querns with horizontal or vertical handle sockets, as well as more complex fittings secured in double slots and sockets in the top of the stone;Footnote 271 it also includes forms of Iron Age date with double projections for vertical handle fittings (with concave grinding surface, wide feed pipe and funnel-shaped hopper). Roman querns with a projection or lug for a single vertical handle socket are also found, but were not commonly recorded in a survey of the querns of southern Spain.Footnote 272 These querns typically have a less concave grinding surface with wide feed pipe and shallow hopper typically defined by a kerb around the circumference.Footnote 273 They are not absolutely comparable with the Hinkley Point quern but are as similar to it as either the Scottish or Irish examples.
DISCUSSION
The recovery of this quern in Somerset is extremely significant. It is the first quern of a Fintry variant found in southern England and it is noticeably different to its Scottish and Irish counterparts. The Irish, Scottish and Manx varieties always have flat grinding surfaces and the feed pipe is always narrow and paired with a funnel-shaped hopper that is quite distinct from the feed pipe (fig. 39a–c).Footnote 274 In contrast, the form of the Hinkley Point quern is typical of querns from South-West England. Such querns have a similarly rounded profile but paired with a steeply concave grinding surface, wide cylindrical eye and only a shallow hopper in the top of the quern (fig. 36a–b).
Given the rarity of querns with any sort of socket for a vertical handle in southern Britain, it can reasonably be presumed that the Hinkley Point quern came about because of direct contact with someone who had seen such querns in action. The quern was not imported, however, but was made locally either by someone, or under the instruction of someone, who had seen the projecting handled Fintry querns elsewhere. It is not the object itself that travelled, but the idea of the object. This movement of design in place of an actual object is also true of the examples on the Isle of Man, which, although following the Fintry design more closely, were made of the local Foxdale or Dhoon granite.Footnote 275
So, how did the idea of the Hinkley Point quern arrive in Somerset? One possibility is that the quern was copied from continental examples, which provide the closest parallels in form. However, it is also possible that the idea of a projecting socket for a vertical handle was taken and applied to a local form, thus explaining the very different shape of this quern to the classic Fintry querns from the rest of the British Isles. In fact, the Hinkley Point quern is closer in form to that from Tomen-y-Mur, with its wider feed pipe and concave grinding surface (fig. 39d). The quern might not represent direct contact with Ireland or Scotland, but a cascade of ideas down the Irish Sea, culminating (so far) in the quern at Hinkley Point. Contact across the Irish Sea, a feature of much of prehistory,Footnote 276 might explain the presence of Fintry querns on the Isle of Man and such contact is also indicated by the recent recovery of an unusually decorated quern from Somerset, for which the only parallel is in northern Ireland.Footnote 277
The issue of dating the bun-shaped Fintry quern type has been problematic. The Irish examples tend to be from unstratified contexts because they were usually deposited away from settlement sites, often in boggy locations.Footnote 278 In Scotland, similarly few bun-shaped Fintry querns appear to be from closely dated contexts. Those from Fintry, Lismore, Balfron, Strontian and East Caldoch, for example, were all unstratified. MacKie has dated the form, based on its placement within the typology of other quern forms in the Scottish sequence, to the first to second century or even the third century a.d.Footnote 279
The stratified example from the Isle of Man is from Phase 3 of Ballacagen Lough A. The site broadly dates to the second century b.c. to the first century a.d.,Footnote 280 with Phase 3 towards the end of the occupation period, suggesting a late first century b.c. to early first century a.d. date for the quern deposition; a date towards the beginning of MacKie's time-frame or even slightly preceding it. The example from Picona in Spain is not directly dated but its general shape and decoration were taken to imply an origin slightly predating the Roman period,Footnote 281 which would make it approximately comparable with the other dated examples.
The Hinkley Point quern was recovered from a pit that was one of a curved arrangement of ten clay-lined pits.Footnote 282 The pit has been phased to the late Iron Age to early Roman period by association with other pits in the group that contained pottery of this period. One of the associated pits also contained material which produced two radiocarbon dates: 361–184 cal. b.c. (SUERC-79728) and 54 b.c.–a.d. 65 (SUERC-79732; both 95.4 per cent probability), the latter of which is in keeping with the pottery.Footnote 283 Overall the evidence is taken to indicate a late first century b.c. to mid-first century a.d. date for the whole curved alignment of pits. This would make the Hinkley Point quern broadly contemporary with the quern from Ballacagen Lough A. In addition, the Hinkley Point example is an Iron Age quern form. In southern England, quern production moved away from bun-shaped querns to flatter (although often still rounded) disc querns at the time of the conquest. It is therefore typologically likely that the Hinkley Point quern was produced before the Roman conquest and that ideas about quern forms were being exchanged at that time, perhaps between the Continent and the British Isles or along the Irish Sea and into the Bristol Channel.
The study of rotary querns is continually evolving, as with other areas of archaeological study. In recent years, we have come to identify the importing of querns to mainland Britain from France and Alderney, as well as the long identified German and French lava querns.Footnote 284 Movement of querns across the English Channel is also evidenced by querns of stone from Brittany and Bordeaux found on Jersey and Guernsey respectively.Footnote 285 However, within Iron Age and Roman Britain there were clearly defined regional identities, and quern form was an integral part of these identities, evolving quite differently between regions. Scotland demonstrates a different evolution to northern England, for example, and both of these are very different to southern and western England. Whilst there is a blurring in quern form between the regions, and differences within those regions, there has been very little evidence of the ‘intrusion’ of one design into another region. The Hinkley Point quern is striking because it does not fit within this very regional specificity of Iron Age and early Roman quern design in the British Isles. Querns were essential household tools, to which individuals and families were probably very attached because of their long-lived nature. Such an attachment is occasionally demonstrated by the unusual placed deposits in which we find them (including this example). The absence of variation suggests an unwillingness to move away from the designs that people were comfortable with and the methods by which the querns were operated.
The Hinkley Point quern demonstrates that during the late Iron Age to early Roman period, travellers occasionally absorbed ideas and applied them to objects at home or carried with them the memory of their belongings at home and sought to have things made in their image, perhaps liking the comfort of using tools that reminded them of home. In this case, the different operational method indicated by the handle may have been a specific requirement of a user who was unable to adapt to oscillation and preferred full rotation.
ACKNOWLEDGEMENTS
The author gratefully acknowledges the contributions of Susan Watts, John Cruse and the late Fiona Roe in the hunt for English and Welsh parallels for this quern. Fiona initially recorded the quern and left notes to the effect that she had never seen a parallel for it in her working area (the Midlands and South of England). Susan Watts provided information on querns in Wales, particularly Carmarthen Museum and confirmed the absence of this form in South-West England. John Cruse provided information about the quern from Picona in Spain and confirmed the absence of this form in the North and North-East of England as well as providing helpful comments on the article itself. The article undoubtedly benefited from their contributions but any errors or omissions remain entirely the author's responsibility. Clarity of some points and figures has been aided by comments of the referees and the author is grateful for these.
The author also thanks Cotswold Archaeology for the opportunity to study this quern and for their permission to publish this note, and especially Andrew Mudd, who provided useful comments on the text.
