(Village centred NT 78940852)
(Chapel at NT 78940852). (1)

Among the Roman camps at Chew Green are mediaeval garths and the site of a wayside chapel excavated by Mr C C Hodges in 1883. The name of the place in the earlier Middle Ages is not known although Kelso Abbey owned pastures and tithes in Redesdale (grant of pasturage dated 1227, but tithes much older). The village had grown up and decayed by 1550 when the place was known as 'Kemylpethe'. The medieval garths overlie and obscure the Roman work. Excavations in 1936 revealed medieval sherds, most from the garths and chapel. They fall roughly within the period AD 1250-1450. (For reproduction of plan see Northum 31 SW 3). (1)

About 40 years ago (c.1883) excavations at Chew Green revealed the foundations of a small stone building in the middle of the East Fort (Permanent fortlet V). Mr C Hodges came to the conclusion that it had been a mediaeval chapel. Some years later (1889) a shepherd, Thomas Glendinning, found a small cross with pointed arms to the north east of this spot and near the Roman road. Mr Hodges thought the cross may have been on a gable of the chapel but Mr Dixon thinks it was perhaps one of the boundary crosses of the liberties of the monks of Kelso.
Traces of later (than Roman) settlement can be seen on the slopes descending to the Coquet. They were possibly enclosed grazing grounds. The foundations of four houses can be seen between the north camp (North Marching Camp III) and the Roman Road. (2)

Visible on aerial photographs. (3)

A small sandstone cross found about 20 inches below the surface at a point about 20 yards from the Chew Green and 40-50 yards from the excavations (of c.1883 referred to in authority 2) on the north east side of the camp. The cross, which may have been a boundary cross has pointed arms. Within the arms a smaller cross is inscribed. It was presented to the Museum of the Society of Antiquaries, Newcastle, by Captain J R Carr-Ellison. (4)

A survey of 1550 refers to Kemylpethe and houses built in times past called Kemylspethe Walls. (5)

A Kemblepath is referred to in 1456. (6)

Centred NT 78940852. The remains are typical of deserted medieval villages in Northumberland, with earth banks forming enclosures and garths. In some cases the banks follow the line of the Roman works or cross and mutilate them. Four building steadings could be certainly identified with banks of earth and stone representing the foundations. Other vaguer enclosures and platforms may also represent the remains of dwellings. (7)

NT 78940852 The remains of the chapel referred to by authorities (1) and (2) consists of a rectangular foundation measuring 18m x 9m and oriented south west-north east. It is bounded by a bank of earth and large undressed stones 2m to 3m wide with a maximum height of 1m. There is no trace of any dressed masonry. The interior of the building is badly disturbed and no features could be identified. The present location of the pottery referred to by authority 1 was not ascertained.
Correctly published on 6 inch. (8)

(Subsequently published) Village of Kemylpethe (site of) and Chapel (remains of). (9)

These remains are as described in the report of 9/5/57. Published survey (1:10,000) correct. (10)

Scheduled. (11)

This site has been re-assessed in connection with RCHME's survey and publication of Roman Camps in England. The following descriptive account is taken from the published text.
Remains rejected as those of DMV of Kemylpethe, though that name may well have been applied to this area, the only habitable one in the immediate vicinity. As a convenient halt on the route over the hills, Chew Green had a long history of later use. As early as 1249 it was formally established as a trysting-place for the settlement of cross-Border criminal cases (Ridpath 1810, 138 (12b)). This activity probably centred on the small building, apparently a Norman chapel, the site of which is marked by confused, roughly rectangular earthworks in the centre of the fortlet (V). The chapel was excavated in 1883 (Proc Soc Antiq Newc 3 ser 4 (1909-10), 162 (12c); Bosanquet 1925, 68 (see auth 2)). Small enclosures, probably for stock, bounded by banks up to 0.5 m in height, were built beside Dere Street. In places they overlie the Roman earthworks but are otherwise undated. They are concentrated to the NW of the fortlet (V) but there are others in the N quarter of the fort (IV) and also to the SE of the fortlet. One bank follows the SE counterscarp of the fort and forms part of an enclosure beside the river. There are traces of platforms, which might have been occupied by timber buildings, immediately inside and outside the S angle of the fortlet (V). Full information is included in the NMR Archive. (12)