When informed of the historical importance of the site the farmer readily consented to an excavation. This was carried out by members of the Carmarthenshire Antiquarian Society and friends, under the direction of W. H. Morris during the spring and summer seasons from 1966 to 1969. Notes on the progress of the work are contained in his day book.16

THE CHAPEL

The Chapel was a simple rectangular building measuring 26 feet by 15 feet internally. The land fell significantly towards the west through the building, which was orientated east-west. The walls were about 2½ feet thick, with moderate sized undressed facing stones inside and out, in a matrix of rubble, plaster and smaller stones. There were no dressed stone fragments on the site. A foundation course had been laid external to the south wall. No structural division existed between nave and chancel walls.

These walls had survived to a height of only two or three courses above the footings. The best-preserved section was the north wall, and the northeast corner in particular. Much of the east and south walls had been robbed apart from one large internal facing stone near the southeast corner. The floor which lay one foot below the exterior ground level was formed of irregular flat, mediumsized stone laid on clay and was level with the wall footing inside the northeast corner. No trace of windows or door-way survived but presumably the entrance been through the west wall. Numerous broken slates and fragments of ridge tiles on the site suggested that the roof had been of that construction.

A crude step of white stones was sited about 8 feet from the east wall, and extended for 3 feet from the north wall. The footing of the altar on its north side was consistent with an altar width of 7 feet, assuming symmetry. A layer of plaster extended west from the altar appeared to have formed the chancel floor.

Outside the east wall and northeast corner there was a layer of loose cockleshell mortar several inches thick which extended irregularly for 4 or 5 feet eastwards. Its purpose presumably to prevent water seepage into the building from the higher ground. This area been cut across by a modern water-pipe. Beneath the mortar layer and set adjacent to the wall footing at the northeast corner were larger flat slabs of slate, set irregularly and up to three deep. This layer extended for 2 or 3 feet from the wall with smaller slates lying beyond fading at about 6 feet from the east wall. There was no mortar associated with this slate layer. The outside the north wall was not investigated. The areas external to the south and west walls described below.

The rough walls of undressed stone in a building with an internal width of only 15 feet suggested a twelfth or early thirteenth century date. The ridge tile fragments were probably thirteenth to fifteenth century, in keeping with later improvements supported by the presence of a 7 foot wide altar and chancel step.17

EVIDENCE FOR AN EARLIER BUILDING

Beneath the Chapel floor of irregular flat slabs, which were laid on 6 inches of yellow clay, was a firm surface of small, wellset cobblestones. This surface extended beyond the line of the robbed out west wall, where there were traces of what might have been a raised altar17 and an apse. This lower floor also extended along the north and west walls of the Chapel.

Remains of an earlier wall were found partly overlain by the later north wall. On the south side the earlier wall was about 2 foot thick and ran immediately inside the course of the later south wall. No mortar was associated with this stonework.

Outside the south walls there was an ill-defined surface of small irregular stones set in clay, fading towards the edge of the Cwm. To the west of the Chapel on a level some 6 inches below the wall footing was a well defined surface of irregular grey stones firmly set on clay, which extended westwards under the bed of the stream (dry during summertime). It was not clear whether this was a natural or man-made surface.

There was a suggestion that the field hedge to the north and east of the Chapel may originally have been part of a bank which encircled the site.17

THE BURIALS

The skeleton of a male aged about 15 years was found beyond the mortar layer, to the east of the Chapel in a shallow grave. The burial was orientated facing east with the arms folded on the chest. The bones were soft and poorly preserved, the spine, pelvis and legs being no longer discernable. Overlaying the chest was a single copper button of eighteenth century date presumably belonging to a cloak, as no other buttons were found. Immediately surrounding the skeleton was an indeterminate blackish layer, perhaps the remains of textile.

The skeletons of three or more infants were also uncovered near the southeast corner of the Chapel, buried superficially and in close proximity, not orientated. There were no gravestones on the site and no burials were discovered inside the Chapel.

It was not customary for burials to take place other than at the parish church until the nineteenth century. The isolated burials at Capel Teilo imply a continued veneration being accorded to the site. The young man may have died by drowning or suicide, while the infants may have been stillbirths or neonatal deaths.

THE FINDS

Numerous finds of North Devon 'gravel-tempered' ware, 'sgraffito' plate, slip ware posset pot and other miscellaneous redware sherds of seventeenth to eighteenth century date were found both inside and outside the building. Similar types were found at Kidwelly Castle. These North Devon wares from the kilns of the Barnstaple-Bideford area were traded extensively along the south Wales coast and far beyond.18

The ridge tile sherds, some with shallow, lopsided crests were of thirteenth-fifteenth century date.18 A Charles II farthing (1672-9 type) was found on the later floor inside the south wall. A variety of Victorian and later china fragments were found superficially within the Chapel. There were cockleshells in the eastern sections of the excavation but very few animal bones were found, and there was no evidence of fires.

PISTYLL TEILO

This was noted in references to nearby 'Chappell Tylo' in 1593 and 1622.19 Though the two were clearly associated they were not co-located. At the end of the nineteenth century steps led down from the field to Pistyll Teilo, though no trace of these now remains. The water was said to be especially good for rheumatism and sprains.19 Legend refers to a ghost which formerly haunted this spot and cried in pitiful tones:

Mae'n hir ac yn o'r i ares
I orwyr Wil Wattar20

The site of the spring was not investigated during the excavations as it was completely overgrown and inaccessible at that time.

POSTSCRIPT

In recent years the Chapel site has been filled in and a new entrance made into the field opposite the farm lane leading to CaeGwyllt. The trees and bushes on the north rim of Cwm Teilo have been cut down exposing the site, and is again under grass. The large stone which forms the northeast corner of the Chapel building is still visible on the surface however, which will allow orientation for astute visitors to the site.

ACKNOWLEDGEMENTS

My thanks are due to Terry and Heather James who provided me with the opportunity to write this article, and who waited patiently for the draft version and have helped prepare the illustrations. Also those colleagues who toiled away with me on that quiet hillside in an earnest attempt to discover the secrets of Capel Teilo. Finally, to the late Bill Morris himself, who gave us all so much insight and understanding of local history.