The Upper Holt

The Upper Holt ridge is part of a long, narrow outcrop of greensand, which runs from the north of Wilton until near Hindon. Previous archaeological investigation in the area includes work by amateur archaeologists in the early and mid-twentieth century. Pottery and earthworks were first noted by G C Engleheart in 1913, and in 1920 a trench was dug through a probable burial mound by the local land agent. In 1959 a local amateur archaeologists dug into the west side of the mound and discovered Roman activity, including a samian ware cup and other Roman material. The site has previously been interpreted as a shrine or temple, and as a military look-out post, although this seems less likely given the structural sequence of the site, and lack of military finds. 

The Teffont Archaeology Project has undertaken both non-invasive and invasive fieldwork in the Upper Holt Wood, including earthwork survey, geophysical survey, vegetation survey, auger survey, test-pitting and small scale excavation. 

The project’s work has revealed that almost the entirety of the Upper Holt Wood is a large Roman enclosure, bounded to the north by a wall, to the south by an existing lynchet, and to the east possibly by the woodland bank. The northern and southern boundaries narrow at the western end, leaving only a narrow track to the Glebe. In essence, the modern wood covers the former shrine enclosure.

Considerable resources were invested into the construction of the northern boundary wall and stone gate at its north-east corner. Excavations at the bank just inside the northern edge of the wood discovered the collapse of a substantial wall. This was a revetting wall between the interior of the enclosure and the natural slope beyond, and was built on an earlier bank, possibly dating to the bronze age. A hollow in the articulated wall collapse contained the remains of a newborn baby, and two fragments of an adult tibia which showed signs of infection. This suggests a ritualised element to the establishment of the wall. A trench situated at the opening into the enclosure at the north-east corner revealed substantial stone footings and post holes of a gate into the enclosure. Along with the wall at the northern boundary, this stone gate indicates that entry to the northern side of the enclosure was perhaps symbolically controlled.


Base of probable gateway and associated post-hole at north-east corner of Upper Holt Wood

The second bank at the south of the enclosure runs from the eastern edge of the wood and continues to the north-east corner of Glebe field. This is accompanied by a ditch and an additional bank, as well as a flat terrace at the southern edge of the ridge. Excavations on this terrace revealed that Roman structures were present, while pebble floors and post settings suggest that there was possibly an open walkway close to the shrine enclosure’s core, with the further away elements of the terrace a simpler trackway reinforced with stone. The bank to the immediate north was found to be built up using high status Roman building material. This stone was from local sources, which contrasts with the widespread stone sources utilised for the structures in the Glebe. Therefore, although both sites are late Roman, they were constructed under different economic circumstances or with different architectural aims. An additional stone structure was situated to the north of this bank.

Auger survey and test pitting have enabled a density analysis of the archaeological material within the enclosure at this site, thereby indicating of the extent of the area which contains archaeological features. It is therefore apparent that archaeological activity is present in the area west of the mound described above, and also in an area covering 30x60m to the east. The central area of the enclosure is empty of archaeological features.

At the eastern occupation zone, excavations revealed a building constructed from wood and stone, with a slaughtered lamb interred beneath it as a foundation deposit. Test pitting also revealed a stone flue characteristic of a hypocaust or grain dryer, succeeded by a stone lined tank or cist constructed from roof tiles. Other features in this area include a possible roundhouse, a terrace, a midden and a cobbled surface.


End of excavation tour showing the final excavated building including limestone wall, postholes and scooped features below the first phase floor, one of which contained a slaughtered lamb foundation deposit.

The central area containing no archaeology was probably occupied by natural features, as even wooden structures would have left some trace. This natural feature could be either an area of open grass or woodland. A religious focus to the site is likely due to its hilltop setting, the monumental nature of the site and the ritual deposits of human bone and sacrificial lamb, therefore an area of woodland in the form of a sacred grove is most plausible. The complex at the Upper Holt is significant as it is much larger and more elaborate than a typical Romano-Celtic shrine or temple in the region, although is still significantly less elaborate than the shrine complex of Nettleton Shrub in north Wiltshire.