CPAT logo
Cymraeg / English
Back Home
Projects index

Parks and Warrens

This Scheduling Enhancement Project (SEP) involved a largely desk-based study of medieval and post-medieval parks and warrens. The work comprised an initial desk-top study of relevant sources, followed by field visits to selected sites during March 2014.

Cwmhir Abbey Great Park and Little Park (Radnorshire). Parts of the northern boundary of Little Park and Great Park can be traced within the modern forestry plantation, surviving in places as a low bank. The eastern and western sides of the parkland appear to have been defined by streams. The southern boundary can no longer be clearly established.


In recent years there has been a dramatic growth in interest in parks, both deer parks and landscaped parks. It would not be an exaggeration to say that the identification of deer parks in east Wales has been something of a voyage of discovery. In the absence of any substantive work on the topic, a wider range of approaches has been necessary in the collation of material than has been usual in the medieval and early post-medieval SEPs that have been completed to date.

What has emerged strongly from assessment of the records in the HER is how very little field observation has been undertaken on deer parks, other than perhaps during the compilation of the Parks and Gardens Register in the later 1990s. This omission is of course not ubiquitous. We can mention André Berry’s earlier fieldwork on the parks of the Dyffryn Clwyd lordship, an interim statement on which was published in the local historical society transactions in 1994. CPAT itself undertook the examination of the Cwmhir Abbey deer park, producing an interim statement in the late 1990s, as well as recognising part of the Vaynor Park pale whilst doing fieldwork there in 2003.

Collating the data from the various sources has produced a very different picture from what was known previously. Molewick Park in Denbighshire is the best example, a genuine medieval deer park attached to the lordship of Denbigh which despite our efforts, we have so for been unable to inpoint,even though it was known to commentators at the end of the 17th century. Further research should locate it, but other parks are at present simply names where there is less optimism about uncovering its location. Other medieval parks must be suspected but as yet no evidence for their existence has been encountered.

The distribution too is significant with a remarkably even spread across the southern counties that now make up Powys, but many more in the north-east where longstanding lordships and gentry estates and based on greater wealth will have facilitated the creation of these symbols of status. The known corpus of parks was increased to around 50 as a result of this study.


The idea of the rabbit warren – a defined area in which rabbits were bred and protected from predators in times before the rabbit became a widespread feature of the British countryside – is one which can be readily grasped, but on occasion, all that we have to denote the presence of a warren is a place-name. This may include Welsh or English language elements, such as ‘cwningen’ or ‘coney’, but it is generally not possible to define the extent of the site by these means. As always, some caution needs to be displayed when considering place-name evidence, in case the name has been derived in another way, but it is generally taken that the evidence is likely to relate to a local memory of land-use which has found its way into the names of localities or dwellings. An anomalous situation can sometimes be seen with place-names such as ‘Giant’s Grave’ and ‘Soldiers Graves’, where pillow mounds have been attributed a mythical function based on their perceived appearance, clearly demonstrating a significant time gap between the termination of rabbit farming and the attribution of the name.

Rabbit farming was certainly established in Britain by the medieval period, although the use of the produce (meat, skins etc) was initially the prerogative of the wealthy members of society. Williamson and Loveday (1989, 298) suggest that early large warrens would have generally been sited in favourable environments of coastal or inland sand, something which holds true for the warren at Borras, near Wrexham and some of the place-name locations on the Flintshire coast. There seems to have been a general move away from rabbit production on better quality land during the post-medieval period, in part owing to the practice being labelled as ‘wasteful and archaic’ by the agricultural improvers of the 18th century (Rackham 1986,48). The activity continued, nevertheless, and perhaps migrated to rather more marginal agricultural land in the process.

Pillow mound forming Esgair y Ty Warren at the head of the Elan Valley.

In the post-medieval period and continuing into the 19th century, rabbit farming became a more organised activity, and utilised a range of common features, such as raised mounds and sunken hollows. Other features which might be present include a dwelling for the warrener, whose task was to look after the rabbits, a surrounding enclosure designed to protect the rabbits from ground-based predators, and smaller enclosures which may have been used to grow crops for fodder.

In summary, the study has provided some new insights into the distribution and preservation of warrens in the study area. Warrens fit into the wider pattern of landuse in the medieval and post-medieval periods, when they provided a useful adjunct to other methods of pastoralism.

Privacy and cookies