Chapelfield Birdoswald

Description

Possible bastle or stonehouse site.
Haverfield showed a plan of the foundations of a thick-walled building excavated by T.H. Hodgson and Mr. Ashby which they considered 'not specially Roman... probably.. an old cottage...or, as has been conjectured, of a small Peel tower. There sre some faint trations of such a tower east of Birdoswald.'
This does not coincide with any of the tenements shown on the 1603 Survey or Map. It may be the same as the 'kind of old ruin' where an alter was found, some 100 yards east of Birdoswald.
SMR record reads '...only foundations courses survive', so presumably nothing to see above ground. (Perriam and Robinson)

The other building was found in the western part of Chapel Field and at its highest point. Only foundation courses survive, placed on the subsoil of yellow clay. It measures 34 by 21 feet and has an annexe about 10 feet square : a trench taken obliquely across its interior and some trenches outside shewed neither internal partitions nor adjacent buildings. No smaller objects were found to fix its date, nor do the rough foundations give clear clue. But the masonry is not specially Roman in character and the whole is probably the relic either of an old cottage, as the ground plan suggests, or, as has also been conjectured, of a small Peel tower. There are some faint traditions of such a tower east of Birdoswald, but the cottage provides perhaps the safer if the less attractive hypothesis. (Haverfield 1899)

This building was interpreted as a cottage measuring 10.36 x 6.4m with a 3.04m square annexe. It was built directly on the clay subsoil, and no dating evidence was recovered from it. The building could either have been a Roman vicus structure or a post-Roman building. It certainly does not appear in the Gtlsland Survey of 1603, unlike the farm in the north-west corner of the fort and the cottage contained within mile castle 49. If post-Roman, therefore, it must either pre-date or ante-date 1603. The field name is of unknown antiquity, but suggests the presence of a ruined stone building. It is possible that this building was Roman and that the field was named after its remains, though any conclusion on the date of the structure must remain speculative. (Wilmott 1999)

Comments

The plan in Haverfield is not inconsistent with a bastle although the walls appear to less than 1m thick. As with many other places some confusion may have been made with the term pele which can refer to three storey small tower houses or, in the use pele-house, to bastles. The plan is not that of a pele tower. Without dating evidence the site is enigmatic but it may be a late C17 bastle derivative house, some of which were used as chapels by non-conformists or it maybe, as the field name suggests, an early modern non-conformist chapel site. Although this site is within the scheduled area for a section of Hadrian's wall the medieval bastle is not specified in the online scheduling report and does not appear to be recorded, as a separate identifiable, site.