The existence of a significant Roman period site at Adel, to the north of Leeds ( fig. 18), has been known for some considerable time having entered the archaeological literature through the work of the Leeds antiquary Ralph Thoresby. Recent work has shown the site to incorporate a fort, centred on SE 277 411,Footnote ¹²⁹ and a substantial civil settlement.Footnote ¹³⁰

FIG. 18.

Map showing places mentioned in the text. (Prepared by Kelly Hunter and Sophie Coy of MAP Archaeological Practice)

Thoresby was a great collector and acquired much material from Adel for his private collection.Footnote ¹³¹ In his diary entry for 14 August 1702 Thoresby discusses possible Roman period names for Adel:

Danum, or Dano (Doncaster) appears in Iter V of the Antonine Itinerary and Rivet and SmithFootnote ¹³⁴ associate the preceding name Segeloci with Littleborough located between Doncaster and Lincoln; thus Thoresby was in line with modern scholarship in rejecting it as a possible name for Adel. His speculation regarding possible transposition and/or the possibility of there being more than one ‘Adelocums’, appears to be exactly that — speculation. Some seven years later, in a letter to Dr Hans Sloane, the Secretary of the Royal Society, Thoresby concluded, having had ‘an opportunity of perusing that Venerable Record Domesday Book in the Exchequer, I found besides Adel and Echope …’, that ‘… Burghedurum or Burgedunum, was the Ancient Roman Name of this Station: … within a Mile of it there two scattering[s of] House, that do to this day retain the Name of Burden- (for Burgdun-) Head’.Footnote ¹³⁵ Rivet and Smith, in an Appendix to their Place-Names of Roman Britain entitled ‘Doubtful names, ghost-names and inventions’, reject Burgodunum:

Faull states that the Domesday name ‘Burgedunum’ related to a vill ‘the name only surviving today in that of Burden Head in Weardley township, 2.4 km to the north-east of Adel Roman site’.Footnote ¹³⁷ Adelocum or Adellocum does not even get considered by Rivet and Smith as a falsa. The modern name Adel derives from the Old English adela meaning ‘filth’ or ‘filthy place’.Footnote ¹³⁸

Thoresby returned to the subject of the Roman period name for Adel in Ducatus Leodiensis:

He then goes on to discuss Adelocum/Segelocim (see above).

Richmond and Crawford interpret Pampocalia as a conflation and distortion of Cambodunum and Calcaria,Footnote ¹⁴⁰ a view endorsed by Rivet.Footnote ¹⁴¹ Rivet and Smith seek to explain it as the conflation of ‘adjacent names presumably misread from a map’.Footnote ¹⁴² Iter II of the Antonine Itinerary must pass close to Adel, but has serious problems with regard to the mileages given in the section between Calcaria (Tadcaster) and Manchester — a distance of 55 Roman miles as the crow flies with only 38 miles being recorded in the Iter. Following Richmond and Crawford,Footnote ¹⁴³ Rivet and Smith propose that a complete stage has been lost and suggest, that ‘the missing station must be Camulodunum, which is placed in these parts by Ptolemy and whose name could easily be taken by a scribe as a duplication of Cambodunum …’, concluding, after some adjustments to the mileages to account for possible scribal error, that Camulodunum equals the known Roman fort at Slack. Following further discussion of the mileage they conclude that ‘Cambodunum could, then be found in the Leeds area’ and that its etymology, which they suggest as ‘fort at the bend’ suits a location on a notable bend on the river Aire,Footnote ¹⁴⁴ and propose an unlocated fort at Leeds, at the confluence of the Sheepscar Beck with the river Aire.Footnote ¹⁴⁵ The name Cambodunum is known at various sites in continental Europe,Footnote ¹⁴⁶ including Kempten in Germany where it has been translated as ‘crooked fortress’ i.e. ‘fortress on the bend of the river’.Footnote ¹⁴⁷ Camboduno, the form attested in the Antonine Itinerary, also appears as the variant Campoduno which suggests assimilation of the first element of the name from the British by Latin speakers as campus ‘field’.Footnote ¹⁴⁸ The etymology is interesting, with the second element apparently having a Celtic origin; as a (reconstructed) neuter dounon it was regularly represented in Latin as dūnum. In sense it seems to have developed from ‘hill’ to ‘fort’, as at Branodunum (Brancaster, Norfolk) where there is no hill.Footnote ¹⁴⁹ While the first element ‘Cambo-’ is basically ‘curved, crooked’ in Celtic as indicated above; however forms with –mp- may indicate an assimilation of the first element with the Latin campus ‘field’.Footnote ¹⁵⁰ Such possible assimilations are known elsewhere, including at Catterick (Cataractonium), where the British ‘catu-’ meaning ‘battle’ may have been assimilated with the Latin ‘cataracta’ meaning ‘waterfall’ or ‘rapids’.Footnote ¹⁵¹

It was the assimilated version that was conflated with Calcaria to eventually appear in the Ravenna Cosmography as the falsa ‘Pampocalia’ around a.d. 700. The assimilated version survived to be used by Bede as ‘in Campoduno’ in his description of the events in a.d. 627–33 following the conversion of King Edwin of Northumbria and prior to his defeat and death at the hands of Penda of Mercia:

Faull discusses the location of Cambodunum at length and with great care, starting with the statement that it is ‘the most difficult problem of Roman West Yorkshire’.Footnote ¹⁵³ She goes on to state:

She accepts both Camulodunum as Slack and the conflation of Calcaria and Campodunum and by implication the conflation as the explanation for the loss of a stage from Iter II. In a thorough discussion of the linguistic evidence and what little archaeological material is known from Leeds in the Roman and post-Roman periods in general, Faull concludes that ‘Leeds [is] the best candidate for Edwin's villa regia of Campodunum’Footnote ¹⁵⁴ and that the new palace that replaced the destroyed villa regia was elsewhere. Two possible sites for Roman forts have been proposed in Leeds — Walflat/Wall Flat (Quarry Hill)Footnote ¹⁵⁵ and Woodhouse Moor.Footnote ¹⁵⁶ Faull favours an Iron Age date for the latter, but regards a Roman fort on Quarry Hill as ‘a distinct possibility’ — which lying only 400 m north-east of the confluence of the Sheepscar Beck and the river Aire, could certainly justify the name Cambodunum, ‘fort in the bend of a river’.

In a careful discussion of the evidence of the Antonine Itinerary for the location of Cambodunum, Faull notes that the distance of 20 Roman miles (30.3 km) from Tadcaster would give a location ‘in the vicinity of Cockersdale about 1.6 km east of the intersection of road 721 with 712 — an area that has produced virtually no Roman material at all’ which she rejects. She also dismisses Villy's suggestion of CleckheatonFootnote ¹⁵⁷ (34.8 km from Tadcaster) as being too far.Footnote ¹⁵⁸ Accepting that Leeds at 21.2 km is too close to Tadcaster, Faull agrees with the suggestion made by Rivet and Smith that Slack should be inserted as Camulodunum and the missing mileage is that from Cambodunum to Tadcaster.Footnote ¹⁵⁹ She goes on to suggest that the 28 km from Slack to Wall Flat/Quarry Hill is not too far removed from the required 30.3 km to equate Leeds with Cambodunum. She supports the suggestion with a discussion of Roman finds from the Leeds area and the possible nature of a settlement at the Aire crossing.Footnote ¹⁶⁰ While accepting that such an identification is possible, it is appropriate to observe that despite the possibility of a Roman fort at Wall Flat/Quarry Hill and the certainty of Roman period occupation in the Leeds area, there is another possible location for Cambodunum — Adel. Adel lies approximately 30.3 km from Slack and is the location of a substantial Roman period settlement and fort. The location of a significant settlement at the junction of roads M712 and M72b seems entirely reasonable, as does the strategic imperative for the establishment of a fort controlling the road junction. Margary speculates on the line of M712 to the north-east of Raistrick, suggesting a route through Brighouse and Cleckheaton and suggests that ‘the present road to Leeds [the A58] … almost certainly represents it … It must have continued north-east to join near Thorner roads 72 or 729 leading to Tadcaster, but its course has not yet been traced’ (my emphasis).Footnote ¹⁶¹ Faull and MoorhouseFootnote ¹⁶² do not trace M712 with any certainty north-east of Grimescar and with decreasing certainty from Cleckheaton and abandon any attempt to suggest a course north-east of Cockersdale, although they do add a number of possible east–west routes to those suggested by Margary, including the suggestion of two short lengths of road at Wellington Hill (c. SE 353 384) and to the north of Roundhay Golf Course (c. SE 332 389) that might represent extensions of ‘road E’ as proposed by Ramm around TadcasterFootnote ¹⁶³ and suggest the existence of a road roughly paralleling M712/M729 some 2 km to the south. Antiquarian sources repeatedly refer to a road ‘… of which the remains leading to Adel are sufficiently conspicuous on the moor near Whitkirk [Temple Newsam], and on Hawscaster Rigg [Chapel Allerton]’,Footnote ¹⁶⁴ which could have crossed the river Aire at Sheepscar; however neither Margary nor Faull and MoorhouseFootnote ¹⁶⁵ mark the road and it must be regarded as ‘unproven’. Such a road, or any of the other putative east–west trending roads, if they were to exist, might indicate that M712 did not extend to Adel, but turned east or north-east as Margary suggests, however, again this is unproven.

In a recent contribution to the topic Mike HakenFootnote ¹⁶⁶ offers a series of suggestions with respect to possible routes for Roman roads and the location of Roman period place-names in West Yorkshire. His discussion derives from a scepticism of the view that Camulodunum can be equated with SlackFootnote ¹⁶⁷ with any certainty. He suggests that:

He then clearly articulates the limited physical evidence for the road and outlines a number of potential alternative routes for Iter II in the area, with suggestions that sites at Birstall might represent Cambodunum and either Cloise Farm, Littleborough or Worlow Camulodunum. Evidence of Roman period occupation exists or is claimed at all these locations: there are antiquarian references to a tessellated pavement at Birstall,Footnote ¹⁶⁹ Cloise Farm has been proposed as a fortletFootnote ¹⁷⁰ and Worlow as ‘a small Roman station’.Footnote ¹⁷¹ None of them is, on the evidence currently available, of a comparable size and importance to Slack and this suggests that Richmond and Crawford and Rivet and Smith are correct in claiming that a stage must be missing from Iter II.Footnote ¹⁷² Therefore, given the correlation in terms of distance from Slack, it would seem reasonable to conclude that Iter II, in the form of road M712, ran via Adel which can therefore reasonably be identified with Cambodunum.

That having been said the identification of Cambodunum with Adel leaves one outstanding question — how does the toponym ‘Cambodunum’ relate to the site at Adel, which cannot be described as a ‘fort in the bend of a river’? Looking to Rivet and Smith's suggestion of ‘fort at the bend’ we are forced to examine the topography, and while accepting that a bend in the river Aire would be a more obvious landscape feature, it is noticeable that the area south-west of Adel fort is in fact a low promontory defined by Adel Beck and a stream running south-west. It is possible that it is that promontory, or rather the streams that surround it on three sides, to which the toponym refers. Alternatively if the name reflects the variant Campoduno, and the assimilation of the first element to Latin campus ‘field’,Footnote ¹⁷³ the difficulties of the name disappear — it may simply mean the ‘fort in the field’.