Sunday, October 1, 2017


Aerial View of the Site of Castlesteads/Camboglanna Roman Fort

I've never really had cause to search a map for the Castlesteads fort.  Why?  Because there is no fort at the site anymore.  It was completely destroyed during the modern period.  The other day, however, I was curious about the situation of the fort in relation to both the River Irthing and the Cambeck.  So, I went to to take a look.  

Kellwood by Camboglanna

What I found was both exciting and disturbing.  Assuming, of course, I had actually found anything at all!

I'm talking here about the presence very near the fort of a place called Kellwood.  Why should this place matter to anyone?

Well, it occurred to me immediately that as there were other Cumbric names in the area, Kell- might preserve a word like celli (GPC: grove, corpse, woodland).  This brought to mind Arthur's court at Kelliwic, supposedly in Cornwall.  -wood could easily have spawned the (g)wig (GPC: wood, forest, grove) of Kelliwic.  Or, although it is a stretch, -wood could be an English substitute for an earlier (g)wig.  According to O. J. Padel (CMCS 8, p. 19), Kelli Wic is 'the forest grove'.

To investigate this possibility, I had to first gather the early forms of the Kellwood place-name.  The following information comes courtesy Stephen White of the Carlisle Library:

The Placenames of Cumberland, Volume 1, CUP, 1950, simply says
KELLWOOD is Kelwood, Parish Registers Lanercost, 1731. It is also noted as Kilewood in 1794
The placename elements in volume three of this work contains the Old Norse Kelda for spring

Thomas Denton’s Perambulation of Cumberland 1687-88 notes under Walton
Henry [Dacre] is now lord thereof. He hath a demesn within this mannor called the  Kellwood, worth 50 li a year, the countrey being hereabouts covered with wood on each side of irthing

I next had to go to Brythonic place-name expert Alan James.  He was kind enough to pass along his ideas for this place-name:

"A Cumbric origin is quite likely here, but the very late documentation leaves it uncertain: Kelwood 1731, Kilewood 1794. It could be celli, or MIr/Gaelic coille, or indeed Cumbric *cīl (Mn Welsh cil) 'a nook'.

However, Kelwood is a place near - now a suburb in - Dumfries, with better, earlier documentation: Kelwode 1323, Kellwod 1342, Keldewod 1370, Keldwod 1433, so probably ON kelda 'a spring', and of course the Cumbrian Kellwood could have had the same origin. 

Still I think celli might be behind Kellwood..."

Tautologies are fairly common in British place-names.  What this means is that two elements are present in a place-name, both from different languages, but both indicating the same kind of feature.  You end up with a formula that reads something like 'Hill-hill.' The Welsh kelliwic or gelliwig is an unusual form in and of itself, a sort of reduplication, and as Alan James reminded me in a separate communication, a -wic ending would probably precipitate an English ending such as wick (cf. Latin vicus).  Unless the meaning of Cumbric -wic/-wig were known to the non-Cumbric speakers of the area.

The notion that Kellwood could be the original Kelliwic is intriguing, even if unprovable.  But the Arthur name (from Artorius) was definitely associated with the Celtic word for 'bear', and the Irthing Valley of Ceidio's/Arthur's father Arthwys ('man of the Arth', perhaps an eponym) was a 'Little Bear' River.  I've in the past tried to make a case for Uxellodunum/Petriana at Stanwix being his headquarters, but that argument is really based only on the fairly late tradition that Etterby hard by Stanwix was known as "Arthur's burg" and by the proximity of other Arthurian sites. We might surmise that Birdoswald/Banna with its Dark Age timber hall was Arthwys's fort, while Camboglanna belonged to Arthur.  

I will say in closing that I feel Kellwood's proximity to Castlesteads seems too coincidental.  Yet a coincidence is all it may be.

POSTSCRIPT:  This would appear to be a false alarm.  I accessed earlier maps, like the following from  They clearly show a small stream originating from Kellwood.  And this means that a derivation from Old Norse kelda is preferable.  Alternately, as the farm sits in a bend of the river, Cumbric cil is also possible here, as this last word has the sense of corner as well as nook.  I would say, geographically speaking, that celli is last on the list of possibilities for the place-name Kellwood.

Early maps also shows a woodland called 'Kellwood Alders' on both sides of the Irthing River.  See  According to Alan James, Kellwood Alders "looks like a plantation, probably no earlier than 18th ct and not necessarily a guide to where the eponymous wood was."

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.