EVENTS AND PROCESSES THAT HAVE HELPED SHAPE THE HISTORIC LANDSCAPE OF UPLAND CEREDIGION

Below are described several of the themes which have been identified in the course of this study as having shaped, influenced or otherwise moulded parts of the historic landscape of upland Ceredigion. The subject list is not exhaustive, but major themes are described.

Palaeoenvironmental
Pollen and other plant remains locked and stratified in peat deposits provide a means of analysing and describing past climate and vegetational change, and by inference man’s impact and influence on the landscape. The process by which peat forms and the methodology of extracting the required palaeoenvironmental data from it are well known and documented, and it is not intended to repeat them here. Upland Ceredigion contains some of the premier sites in Britain for the study of palaeoenvironments, including the well-known site of Cors Caron/Tregaron Bog (for a list of pollen analytical sites see Caseldine 1990, Figure 11 and p. 127). Caseldine (1990) describes the results from all palaeoenvironmental work from Wales on a period-by-period basis. It is sufficient to state here that almost every historic landscape character area in the study contains natural or semi-natural deposits the study of which would assist in our understanding of the human impact on the environment over the past 10,000 years.

Bronze Age ritual, funerary and settlement sites
Evidence for the pre Iron-Age settlement of the study area is limited to knowledge obtained through the examination of palaeoenvironmental remains (see above) and from Bronze Age ritual and funerary monuments. These monuments - which mostly consist of round barrows/burial cairns and standing stones - are common elements in the landscape, but have a marked concentration on high ground and on the fringes of high ground. This distribution is at least in part the result of post Bronze Age, and in particular modern, agricultural practice; almost certainly there was once a more even distribution of these monuments across the study area. The presence of large numbers of these monuments, often in what is now considered to be quite remote areas, indicates a settled population in the uplands. However, evidence for contemporaneous settlements is scarce, with the best evidence provided by burnt mounds or hearths. Clearance cairns may also indicate settlement and agriculture of Bronze Age date. In some of the upland areas historic landscape components are almost entirely composed of Bronze Age ritual and funerary monuments. For a full discussion of the Bronze Age in Ceredigion see Briggs (1994).

Iron Age/Romano-British settlement 800BC - AD400
The distinctive monument of the Iron Age is the hillfort, and these are more sparsely distributed in the inland and upland areas of the study than further west and south. There are none of the rectangular enclosures, now surviving as cropmarks that are densely distributed in southern Ceredigion. Castell Rhyfel and Dinas are located in open, unenclosed upland, others tend towards the high ridges of north Ceredigion on land enclosed in the early modern or post-Medieval period. However, it is worth noting that another, previously unknown hillfort, was discovered during aerial photography undertaken for this study. No obvious patterns of coincidence between hillforts and later territorial units of cwmydau/commotes can be discerned

The Roman impact on the county was primarily military, and included the creation of a north-south aligned route corridor from the gold mines at Dolaucothi, via the fort of Llanio on the Teifi, to Trawscoed and thence north to a crossing of the Dyfi south of Pennal. The study area was thus by-passed. Whilst civilian settlements grew up outside the gates of forts (i.e the vicus at Trawscoed,) and the demands of the military garrisons for supplies and tribute may have quickened economic exchange, they did not survive the withdrawal of garrisons in the third century. Impact on Iron Age patterns of life was therefore minimal, but a viable north-south route corridor was created which survived into the post-Roman period, as the traditional name of Sarn Helen attests (albeit generally to the west of the study area). This picture could change if more evidence is forthcoming for the Roman exploitation of Ceredigion lead mines - the lead in two objects from excavations of the Roman fort at Trawscoed appears to have been of mid Wales origin (samples subjected to gamma-ray spectrometry). For a full discussion of the Iron Age and Romano British periods in Ceredigion see Davies (1994) and Davies and Hogg (1994).

Early Medieval AD 400 - 1100
The long period of time between the end of Roman rule in Britain and the beginnings of the Norman Conquest of Wales is important in terms of cultural heritage, and in laying the foundations for later medieval settlement and land-use, but very little evidence for this period can be seen within the present landscape. Archaeology has to date produced evidence of early Christian burial sites and monuments, but little of settlements at any social level. Research into the period is urgently needed, as our knowledge is only likely to be extended through detailed landscape analysis. Consequently the historic landscape characterisation process has thrown up as many questions as it has answered.

In terms of the historic landscape characterisation of this area of inland and upland Ceredigion, three major themes need to be appreciated in order to have an in-depth understanding of how the present landscape evolved during this period. These are:
1. the introduction and spread of Christianity, its influence on settlement through cult centres and burial grounds.
2. the emergence of Welsh and the likelihood that many of the place-names that define and distinguish the landscape for its inhabitants and visitors are of early Medieval origin.
3. the creation of territorial and administrative units that were to remain powerful influences on subsequent social organisation.

The distribution of place-names with the personal name elements of the early saints indicates the spread - or contraction - of their cults over several centuries, not the wanderings of 5th - 6th century ‘missionaries’. Some centres developed into ‘mother churches’ and medieval parish churches, but others did not. At Trisant three 9th century cross inscribed stones from the churchyard suggest its origin as an early Christian burial ground, but it became simply a hamlet within the larger parish of Llanfihangel y Creuddyn where the archangel Michael place-name element indicates Anglo-Norman reorganisation. Significantly, though Trisant is a ‘llan’ or enclosure of three saints, their names have been lost. At the other end of the scale, the important nucleated settlement of Tregaron appears to have been a pre-Norman focus of power, as it has a Class I ECM (early Christian monument) and the place-name, combining tref (farm, township) and a saint’s name Caron, which can probably be equated with Carannog - the ancestral deity or saint of the 5th to 6th century Irish settlers of Ceredigion.

An important component and influence on the medieval landscape of this study area is the Cistercian monastery of Strata Florida, founded as a daughter house of Whitland in 1164, but generously endowed and supported by Lord Rhys, prince of Deheubarth. It is likely that the large territorial endowments made by Lord Rhys to the Abbey, that combine both upland and lowland, were existing pre-Norman estates and land units.

Finally, in terms of territories and boundaries, and peoples’ perceptions of where they belonged, it is worth quoting Professor Ieuan Gwynedd Jones (1998, 480) in his study of the patterns of nonconformist belief in 18th and 19th century Ceredigion, its conservative nature and his discussion of the strengths of Calvinistic Methodism in the inland, upland areas:

There existed therefore, deep-set feelings of belonging to ecclesiastical and civil structures which were of great antiquity, that the world people inhabited was the world that their ancestors had known, and one that was not likely to change

Medieval political - conquest, co-existence and change
The early territorial divisions of ancient Ceredigion are succinctly summarised by Dodgshon (1994). Ceredigion was composed of a gwlad, which was subdivided into two cantrefi (literal meaning, a hundred settlements or trefi). Cantrefi were further subdivided into two cymydau or commotes, each cwmwd divided into four maenorau. Of the four maenor in each cwmwd, three were in the hands of freemen, the fourth in the hands of bondmen. Free maenorau contained 13 townships, bond maenorau seven. To the 46 townships in the maenorau should be added two held in demesne by king, one in the hands of the reeve and one in the hands of the chancellor, to bring the total in a cwmwd to 50, or half a cantref.

It is not proposed to enlarge on this discussion of earlier territorial divisions here, as their impact on the historic landscape is not currently known apart from one possible exception. This exception comprises the large tracts of land granted to Strata Florida Abbey by Rhys ap Gruffudd. The details of this grant, which was to form the foundation for the abbey’s granges, has not been researched, but its boundaries must have borne some relation to pre-existing territorial divisions. Arrangement and organisation of the granges by the abbey, and in the post-Medieval period by estates, has affected the historic landscape in a manner that is still detectable, as can been seen in the descriptions of the historic landscape character areas.

By the end of the 13th century these old territorial divisions had broken down, and Ceredigion was organised on county lines - the foreruner of the modern county. This transition is described by Jones Pierce (1959). It is, however, unclear how the former territorial divisions were translated into the forerunners of the administrative units that now exist, if indeed they were.

Population 1750 - 2000
Prior to 1801 there are no reliable sources for assessing the population of Ceredigion. Howells (1974/75) estimates that by the mid 16th century numbers may have recovered from the late Medieval low to stand as high as 17,000, and may have reached 27,000 by the early years of the 18th century. Population changes in Ceredigion for 1750-2000 have been the subject of a recent study by Aitchison and Carter (1998); this section summarises parts of their work. Ceredigion experienced a rapid increase in population from at least the mid 18th century when it has been estimated that 32,000 people lived in the county, until a maximum 73,441 in 1871. The reasons for this were the rising demand in agricultural produce and rise in farm prices following the Napoleonic War, and an increase in output from industries, in particular lead mining. This increase in population has left its mark on the historic landscape: new settlements, especially on the fringes of marginal land were created; moorland and common land was enclosed and brought into cultivation, new lead mines were opened and production from older mines increased. A fall in lead production, and agricultural depression coupled with greater mechanisation on farms, led to a steady decline in population from the 1871 high point. Deserted farmsteads and cottages, and abandoned lead mines are the most tangible evidence of this decline. The greater reliance on farm mechanisation did not, however, lead to a large-scale withdrawal of agriculture from the highlands, though fields and paddocks associated with deserted cottages were abandoned. Since 1961 there has been a slow increase in the population. This overall pattern of increase followed by decline from the late 19th century was exaggerated in upland parishes of the study area, where most of the reserves of lead are to be found and where in the 18th century acreages of unenclosed land far out-numbered those of enclosed farmland. In one parish, Llanfihangel-y-Creuddyn, population rose from 800 in 1801 to 1800 in 1861, only to fall to below 500 by 1961.

Medieval and modern settlement
Deserted settlements in the Ceredigion uplands
The deserted rural settlement sites of upland Ceredigion are both numerous and of great significance to the archaeology and history of the county. They represent the homes of farmers, agricultural labourers, shepherds, peat cutters and lead miners. It is evident that there are very few corners of the county that have not been settled at some time; even on the highest reaches of Pumlumon the ruined huts of 18th and 19th century shepherds can still be found.

Although there has been a tendency to discuss deserted rural settlements in a Medieval context, most deserted settlements in the Ceredigion landscape are demonstrably post-Medieval in date (i.e. 16th century or later). Some specific sites are identifiable in early documentary sources, which enable us to confirm that they were occupied in the 16th or 17th century. Many more sites appear on estate maps, tithe maps or Ordnance Survey maps from the mid 18th century onwards. Through these sources, we can chart their latter stages of occupation and, sometimes, their period of abandonment - we know little of their origins.

Deserted settlement sites survive in the landscape mostly as much ruined and eroded dry-stone or earthwork structures. They are characteristically rectilinear in shape, many being single compartment structures, but often they are sub-divided into two or three compartments and may have any number or form of extensions added to the original structure. It is usual for the dwelling to be associated with ancillary buildings and garden enclosures or field systems of varying size and complexity. Some deserted settlements now consist of empty building platforms cut into the hill-slope, probably the foundations for timber structures.

A common type of deserted settlement of the Ceredigion uplands are the lluestau, once occupied by shepherds who stayed on the commons throughout the year with their flocks. The lluestau are mostly situated on level, natural terraces, close to the streams of the sheltered valleys that run off the mountains. A typical lluest consists of the dry-stone or earthwork wall-bases of the dwelling, to which is appended a small enclosure, defined by an earthwork bank and often showing signs of cultivation. This small-scale cultivation made the lluestau into oases of green in the otherwise drab upland landscape. Ancillary structures can often be found around the dwelling, probably representing animal or fowl pens, or storage structures.

Farmsteads are in many respects comparable with the lluestau in terms of physical remains, but on a larger scale with their field systems making a more substantial impact on the landscape. Like the lluestau, upland farmsteads can often be found in remote and isolated locations, with enclosed pasture and arable lands being carved out of the extensive mountain wastes. Claerddu Farm, north of Llynnoedd Teifi is an example of such a holding. With a true upland farmstead, the amount of enclosed land might appear relatively small, but the true extent of the farm would be measured in its sheep-walk, which would often be between 500 and 1,000 acres of unenclosed moorland on which the farm enjoyed sole grazing rights. Although originally unenclosed, the 19th and 20th centuries have seen these extensive grazing areas defined and sub-divided by wire fences.

It is evident that many of the platform sites of the district are located on higher, often more exposed, ground. A group of three such platforms on the southern slopes of Disgwylfa Fach to the west of Nant-y-moch reservoir, are found on an exposed south-western facing slope. The significance of this observation is unclear, but it is possible that there are both chronological and functional differences between the settlements of the sheltered valleys and the platforms found on the higher ground. It is tempting to think of the platforms as representing the sites of medieval hafotai or dairying-stations, only occupied during the summer months and therefore not so dependent on shelter from the prevailing winds.

Notes on the place-names Hafod, Lluest and Magwyr
There are a number of place-name elements which can act as pointers to the location and possible function of deserted settlement sites in the Ceredigion uplands. Occasionally, terms such as tai hirion (long houses) or hen dy (old house) will be encountered. But the most commonly encountered terms are Hafod, Lluest and Magwyr, place-name elements that are important to our understanding of the landscape and its development and worthy of further explanation.
Hafod (plural hafodydd) is an ancient term, used in medieval law texts, which refers to upland summer pastures on the commons and wastes. It literally means ‘summer-place’ and was traditionally associated with the summer grazing of dairy cattle, but its application has broadened during recent centuries with the decline of dairying in the upland areas of Wales. The dwelling was known as the hafoty (plural hafotai). The term cadw hafod (keeping a hafod) has been used up to present times and refers to any form of transhumant activity that involved human settlement in an upland environment. Hafod is a relatively uncommon place-name element in the Ceredigion uplands, although examples where the name applies to potentially early deserted settlements are found at Hafod Frith and Hafod Eidos, east of Strata Florida. Both sites may have been associated with the abbey of Strata Florida during Medieval times. Another use of the term in relation to the properties of the abbey comes from the former monastic grange of Cwmystwyth, which was divided into four hafodydd in Medieval times.
Lluest (plural lluestau) in also an ancient term used in medieval Welsh lawbooks. It originally referred to a temporary camp or settlement, not necessarily associated with agriculture or animal husbandry. Lluest is a very common place-name element in Ceredigion, although by the end of the 19th century the term had often been dropped in speech in many place-names, only being recalled in documentary and cartographic sources. In the Ceredigion dialect, lluest means much more than a ‘camp’. It can be applied either to a dwelling or to the land associated with the dwelling - particularly with reference to the old shepherding cottages of the uplands and their associated sheep-walks. Documentary research has shown that before the 19th century, the lluest was not necessarily only associated with shepherding. There are examples of Ceredigion lluestau being described by the Latin term domus lactaerius (translated as ‘dairy’) in 16th and 17th century sources at the National Library of Wales, although whether it was sheep or cows that were milked at such places is not known. It seems that the lluest in Ceredigion was either a regional variation of the hafod system or perhaps a development of it. It is evident that the lluestau on the more remote hills were being abandoned in large numbers during the latter half of the 18th century (Vaughan, 1966). By the 19th century, the lluest was most certainly characterised by the cottages where shepherding families eked out a rather precarious existence on the mountain pastures.

Magwr, Magwrn (plural Magwyr) is a frequently encountered place-name in documentary and cartographic sources and a term very much alive in the Ceredigion dialect. Its root is the verb magu (to raise, nurture), probably indicative of efforts to improve and nurture a piece of land, but the modern meaning of the word in an upland context applies to ruinous stone buildings or walls, usually an abandoned dwelling. The occurrence of the element magwyr in upland place-names is a fairly reliable indicator of human activity and settlement.

The role of Post-Medieval estates in shaping the landscape
The large tracts of land granted to Strata Florida and divided into granges or estates had initially been worked by lay brothers and administered directly by the abbey, but by as early as perhaps the 13th century financial and economic pressure compelled the abbey to lease out of the land in return for cash rents. Each grange was divided into a number of separate farms that were leased out on a commercial and individual basis. It is likely that some of the farms pre-date the granting of land to the abbey, but others would have been established during monastic tenure. At the Dissolution, the embryonic land-owning classes covetously viewed monastic land. Strata Florida’s land was granted to the Earl of Essex as receiver-general. In 1547, his son obtained a lease on most of the abbey’s former lands. However, other families staked a claim to the bounty. For instance, within Cwmystwyth Grange the Herbert family had been busily acquiring leases of farms let out by the abbey for 99 years (Morgan 1991), providing them with a strong claim to the land, and at Strata Florida itself John Stedman acquired the abbey buildings and demesne land. Herbert acquisitions formed the basis of what was later to become the Hafod estate. John Stedman built a mansion adjacent to the abbey; this was to form the nucleus of a small estate until it was absorbed into the Nanteos estate on the death of Richard Stedman in 1746. Crosswood became the largest estate in the study area mainly due to the purchase of the Earl of Essex’s holdings of Strata Florida Granges in 1630 (Cadw 1992, 21).

In conjunction with the opportunities afforded for the large-scale acquisition of land by the Dissolution of the monasteries, the acceptance of the concept of privately held land from the late Medieval period profoundly affected tenurial systems, with concomitant knock-on effects on economic structures and farming practice. It is now impossible to judge what elements of entrepreneurial flair, business and political acumen or sheer good fortune allowed just a few of the many small estates which were carved out of the Medieval tenurial systems to expand and develop into successful units. Once they had achieved a certain size, estates were able to expand by purchase of other holdings, exchanges of land, and, most commonly, judicious marriage. Thus it was that in the study area by the 19th century almost all land lay within the estates of Crosswood, Nanteos, Gogerddan and Hafod. The families of these estates dominated political, cultural and, most importantly for this study, economic life. For a good description of the methods by which the gentry accumulated land see Moore-Colyer (1998, 54-56). Metal mining, with its obvious effect on the historic landscape (see below), was controlled by the estates. The pace and extent of enclosure was often also governed by estates (below). Apart from the vast visionary schemes of Thomas Johnes at Hafod - the creation of a picturesque landscape, upland afforestation, and the creation of new farms - other estate influences on the historic landscape were slight. Agricultural improvement was slow, and no examples of large-scale estate housing are known.

With agricultural depression, and the cessation of metal mining, the 20th century has witnessed the collapse of the estate system. The core areas of Crosswood, Gogerddan and Hafod are all now part of government agricultural or forestry institutions.

Encroachment and enclosure
Enclosure of the commons, wastes and moors has had a most dramatic effect on the landscape of upland Ceredigion over the past two to two-and-a-half centuries. A detailed examination of enclosure leads one to conclude that it was not the result of a concerted action by a group of landowners or one process, such as Parliamentary Acts, but the result of many actions and processes undertaken by many individuals and organisations fuelled by rising populations in the late 18th century and the first three quarters of the 19th century. The changing pattern of enclosure can, in part, be gleaned from an analysis of the landscape itself. Extra information is available through the study of documentary and cartographic sources. There are no good published accounts of changing enclosure patterns in Ceredigion, and therefore primary sources were consulted. The most useful sources are late 18th century and early 19th century estate maps and tithe maps of the 1840s. These maps capture an image of the landscape as it was undergoing dramatic transformation: from a mostly unenclosed landscape to a largely enclosed landscape. Prior to the late 18th century, apart from landscape analysis, manuscript documentary sources provide the only evidence for the changing pattern and pace of enclosure, but the examination of these is beyond the scope of the present study.

This drive to enclose wastes common and moor should be set against the general economic and agricultural background of the late 18th century and early 19th century. During the Napoleonic Wars Parliament recognised the need for food security. Moore-Coyler (1998, 20) states that public concern became so great that the conquest of waste became almost synonymous with the conquest of France. Coupled with this desire for increased food production was a steady rise in population in most parishes in Ceredigion throughout the 19th century. These two factors were instrumental in encouraging the enclosure of land, and resulted in the creation of a vastly different landscape at the end of the 19th century from that that had existed at the beginning.

Below are listed some of these processes and actions that influenced the pattern and pace of enclosure in upland Ceredigion.

Parliamentary Enclosure
Enclosure by Act of Parliament did not have a profound or lasting effect upon the landscape of upland Ceredigion. By these acts landowners could gain possession of common land, enclose it, and bring it within the control of their estates. The only major areas to be enclosed were blocks of high common and waste in Llanfihangel-y-Creuddyn parish to the north and south of the upper Ystwyth valley (Chapman, 1992, 53). This Enclosure Act was awarded in 1866; its area is shown on a map in the National Library of Wales (Card CC Deposit 6). Almost all of the land included in this award has been assigned unenclosed upland historic landscape character areas, though occasional wire fences now subdivide it - or a post World War Two forestry plantation. An exception to this is a small area to the north of Cwmystwyth. Here dwellings within and surrounded by small fields were defined and shown on the map. These were illegal settlements and encroachments on the common. Those settlers that could demonstrate that their dwellings had been established over 20 years were granted ownership. Others had the opportunity to purchase their houses, or abandon them.

A second upland area that seems to have been earmarked for Parliamentary Enclosure was Mynydd Ffair Rhos. A map was surveyed in 1815 (NLW Crosswood 347) in preparation for the Act or Award, but nothing further appears to have been carried out. As with the Llanfihangel-y-Creuddyn Award, areas of illegal encroachment were noted; these were concentrated around the village of Ffair Rhos. Although not enclosed by Act of Parliament, lower-lying portions of this area as depicted on tithe maps of the 1840s and later Ordnance Survey maps, show that they were formed into large fields in the 19th century by the tenants and landowners of adjoining farms - higher ground to the east was not, and has not been, enclosed.

Enclosure by estates
Within upland Ceredigion large estates such as Crosswood, Hafod, Nanteos and Gogerddan, were the main driving force behind much of the reclaiming and enclosing of common and wasteland. Most unenclosed land was considered the property of the Crown - in the early 19th century two-thirds of land in Ceredigion nominally belonged to the Crown - a right that many landowners refused or were unwilling to recognise. As formal boundaries on unenclosed land were non-existent, and Crown agents few and far between, enclosure and land claims were made by the simple expedient of pushing back boundary fences onto moor or common. Concern was expressed in a report of 1809 that the Crown was being deprived, but expressed caution at legislation for fear of ‘exciting great dissatisfaction’ among landowners (Moore-Colyer 1998, 56). Moore-Colyer states that the Crown pursued one of two solutions: either it sold its rights by auction or accepted the encroachments on payment of a peppercorn rent.

This large scale appropriation and enclosure of Crown lands is evidenced right across the study area and is elegantly attested by the clear differences between unenclosed land shown on late 18th century estate maps, and unenclosed land depicted on tithe maps of 1840. Clearly areas most susceptible to enclosure were those on the fringes of moor and waste.

It was unusual for new farms to be carved out of the waste as part of this process of enclosure, but not unknown. For instance in the early 19th century, Thomas Johnes of Hafod, a notorious and rapacious appropriator of Crown land, established an experimental farm known as New Farm, now Gelmast, on Crown land north of the Ystwyth. The farm is now surrounded by forestry. The farm was intended to exploit the vast sheep-walks surrounding it, but it was also designed for dairying, and land was drained and enclosed with this latter aim in mind. Woodland was also planted. After Johnes’ death the Crown surveyor estimated the encroachments and plantations amounted to 467 acres (Suggett 1998-99). A second example is Daren, within the Gogerddan estate. A 1788 estate map (NLW R.M. 108) is annotated by a later hand showing a formerly unenclosed ridge divided into fields with two new farms founded. Lewis (1955, 66-67) describes the landscape development of Daren from 1788 to 1943.

Enclosure by individual farmers
In tandem with estates rapaciously grabbing and enclosing common, moor and waste, small farmers were active in both extending the bounds of their land and improving what they had by enclosure and drainage. Unless documented, it is impossible, and, for the purpose of this study, unnecessary, to differentiate between common, waste and moor enclosed under the aegis of a large estate, and that carried out by a farmer.

Owing to the availability of estate maps it is often possible to identify where small-scale improvements and enclosure have taken place within the bounds of farms. This was a common practice in the late 18th and 19th century and there were few farms that bordered, or were close to, unenclosed land where it did not occur. Most farms that bordered unenclosed common or moor had sheep-walks attached to them. The bounds of these sheep-walks were usually shown on estate maps with a high degree of accuracy, though they rarely seem to have had a formal built boundary such as a hedge, bank or fence. Farmers would chase away neighbouring sheep that strayed onto their sheep-walks. From this it is clearly only one step away from ring-fencing the sheep-walks to separate them from neighbouring properties and common. A 1764/65 estate map of Bwlchcrwys Farm (NLW Nanteos 312) shows a property at this stage of the enclosure process. A house is shown, the boundary of the farm seems to be delineated by a bank, hedge or other physical boundary, but farmland is marked ‘Clear Open Ground’ and ‘Turbary’. By the tithe survey the ‘Clear Open Ground’ has been subdivided into fields. It would be tedious to list more examples, but this pattern, usually not as dramatic as the case above, can be paralleled on many other farms for which estate maps exist, and reflects the methods by which farmers gradually improved their properties.

Tai-unos - squatter settlements
By far the most well-known, written about and discussed form of enclosure are small-scale encroachments and settlements on unenclosed land - so-called tai-unos or squatter settlements - even though the amount of land concerned is minute compared with Crown land illegally appropriated by large estates. The reason for this notoriety was the animosity they generated from farmers who complained that their rights to common were threatened, from Crown agents who functioned to protect and guard the Crown’s rights over unenclosed land, and from large estates and their agents who perceived squatters as general ne’er-do-wells who occupied land outside the estate tenurial system.

Against a rising population and an inflexible estate system which was unwilling or unable to create new farms - the new farms created by the Hafod estate and the Gogerddan estate cited above are matched by farms abandoned in the same period - it is hardly surprising that the disenfranchised poor sought to settle on marginal land, either as a temporary measure until a larger farm became available, or on a permanent basis practising a subsistence agriculture supplemented by farm labour and craft employment (Knowles 1998, 81).

Squatting has been described by several writers (Lewis 1955, 65; Knowles 1995, 79-82; Moore-Coyler 1995, 21-23), but it is a description of the 1790s by a London lawyer quoted by Morgan (1997, 210) that best summarises the activity:

The Party Incroaching assembles his relations & Friends to his Assistance & they run up a Structure & inclose a small Quantity of Ground between sunset & sunrising the next morning. In this Structure without any Alteration they must reside a year & if in that time he has met with no interruption he claims such ground & Structure as his freehold property, pulls down the first rude Edifice & constructs another of more permanence & encroaches gradually on his Enclosures.

A squatter settlement at Cnwch Coch aroused the indignation of the Vaughans of Trawscoed; one of the complaints was ‘being miserable looking Huts, to be seen from the Windows in front of the house’. Morgan (1997, 209-213) describes the unsuccessful attempts to evict these squatters. Cnwch Coch consisted of a small nucleation of cottages and houses located in a landscape of enclosed pasture fields and scattered woods on the side of a moderately low hill was in many ways an atypical squatter settlement. The typical squatter settlement comprised cottages and houses dispersed at irregular intervals of 50m to 200m across poor quality, often rush-covered or peaty ground of high altitude at or beyond the margins of cultivation. Rhos y Gell, Brynafan, part of Ystbyty Ystwyth, Ffair Rhos, parts of Ystumtuen and Tai Unnos) are all examples of such settlements. Houses in all except the last example are all still occupied. In the late 20th century these areas have become desirable places to live either as permanent residences or as second homes, offering a degree of remoteness in, to modern sensibilities, attractive settings, without the burden of being attached to large tracts of farmland. Many of these former squatter dwellings have been recently modernised or rebuilt or are currently undergoing such treatment, a trend reflected in the recent population increase seen within parishes of upland Ceredigion.

While these loosely dispersed squatter settlements are readily identifiable in the field and from maps, the recognition of single or one-or-two tai unos is more difficult, if not impossible without considerable cartographic and documentary research, which is beyond the scope of the present study. On occasions it has been possible to tentatively identify a single squatters cottage from maps, but it is likely that many, if not all, of this type of settlement were established and abandoned before they could be surveyed and mapped. In these instances the archaeological record is the only evidence.

Enclosure of sub-divided fields
There has been no comprehensive study of the character, form, date, extent and distribution of sub-divided field systems in Ceredigion. That they existed in the lowlands, particularly along the coastal strip is not in doubt, and is attested to by the still partially extant system at Llan-non (Davies 1973, 526-27, Jones 1985, 165-67). Outside the lowlands, sub-divided, intermixed and dispersed holdings have been described by Davies (1973, 522-24) at Llandewibrefi, but how typical the field system of this collegiate centre of the Bishop of St David’s was across the uplands is uncertain, though other examples have been identified during the course of this study.

An estate map of 1781 (NLW Crosswood Deeds 5, Ser IV, Vol 1) titled ‘Map of Sputty Intermixed Lands’ shows a small sub-divided field. This field which lies immediately to the south of the village of Ysbyty Ystwyth at the relatively high altitude of 220-250m must surely have been an element of a once much larger system, a system that had by the late 18th century been consolidated into one or more holdings and enclosed. The landscape area to the east was probably once part of this sub-divided system, as was that to the west of the village of Ysbyty Ystwyth. By the tithe survey of 1848 all traces of the sub-divided system had been removed. The modern landscape is one of small fields of rough grazing, small forestry plantations and improved pasture; and morphology provides no clue as to whether the modern field patterning has evolved from a sub-divided field system or not.

A sub-divided field is recorded on manuscript maps of 1791 and 1819 (NLW Vol 45, 70; Vol 36, 151) on gentle northwest and southwest facing slopes between 200m and 320m to the east of Tregaron. As with the system at Ysbyty Ystwyth, this sub-divided field was once probably part of a much more extensive area of intermixed strips that was, by the advent of large-scale estate mapping, consolidated and enclosed. It may be possible to at least partially reconstruct the extent of this former sub-divided system by careful analysis of estate maps and the tithe map, but this is beyond the remit of the current study. The estate maps of 1791 and 1819, and the tithe map of 1845 record the system in decline. By 1845 only a few strips remained. The present-day landscape of regular fields provides no indication of the former strips.

In addition to the unequivocal map evidence described above, there are certain morphological features of the built landscape which imply the former presence of sub-divided field systems, but which cannot be substantiated without documentary research. Surrounding the village of Llanfihangel-y-Creuddyn is a landscape comprising enclosed pasture, a landscape that has changed little since it was first mapped on a large scale during the tithe survey of 1847. Enclosed long narrow fields may indicate that they evolved from a strip or sub-divided field system. Similar evolved landscapes can be postulated for Capel Bangor and the landscape to the west of Tregaron.

Buildings
General
Apart from modern structures, most of the buildings in the upland Ceredigion landscape date to the 19th century, with the majority built in the middle through to the final decades of that century. Several categories of building are present: farmhouses; agricultural buildings; non-agricultural rural dwellings; village housing; industrial housing; and industrial buildings. All share a common ancestry and are built of local stone. Indeed it is not easy to detect whether some dwellings were primarily agricultural or industrial. Perhaps the division is not meaningful, as most smallholders of the region probably worked seasonally in the lead mining industry, and many industrial workers supplemented their earnings with farm work.

In common with other parts of west Wales very few buildings were erected during the 20th century prior to the 1960s/1970s. There are exceptions such as the small housing estate at Ponterwyd, and small groups of social housing in other communities. Since the 1970s new housing has been constructed at an ever-increasing rate, and now in what were essentially 19th century villages, such as Llanafan, houses built in the last 10-15 years now dominate.

The settlement pattern of the area is overwhelming one of dispersed farms and cottages interspersed with 19th century industrial settlements such as Pontrhydfendigiad, Ponterwyd and Goginan. Tregaron is the only example of an earlier nucleation, but it is not a large settlement and the buildings are mostly 19th century. Of particular note are the settlements of squatters – tai unnos (one-night houses) in Welsh. The clustering of this type of house into communities of closely positioned houses on marginal land is a distinctive component of the upland Ceredigion landscape. Some of these communities are deserted, as at Gwar Castell, partially abandoned as at Rhos-y-Gell, and still functioning as at Cnwch Coch.

Materials
Local stone is virtually ubiquitous in older buildings. Generally the stone available in upland Ceredigion – Silurian sandstones and mudstones - varies little from east to west and from north to south. In most rural buildings the stone is not particularly good quality, being small and slabby and roughly coursed. However, in the few gentry farms and larger farmhouses better quality stone, roughly squared and more regularly coursed is evident. Towards the end of the 19th century better quality masonry is also often evident in the industrial housing, perhaps reflecting a more systematic, industrial extraction of building stone. Brick used for window- and door-jambs is relatively common on the later 19th century industrial housing. However, buildings constructed wholly of brick are relatively rare and are mainly confined to the later 19th century and to the far northeast of the area close to Aberystwyth where the bricks were either imported or manufactured. It is highly likely that the cement render (see below) over many buildings, but in particular on rural cottages, masks clom – earth construction. This type of construction probably ran in parallel with the stone building tradition, and was mostly employed in more modest buildings. Its use was replaced with that of stone in the mid to late 19th century. A few survivors are listed.

Masonry is left bare, whitewashed (more recently painted) or cement rendered (stucco). Whitewash/paint over masonry on houses although relatively common in upland Wales is not usual in west Wales, where bare stone or stucco are the preferred finishes. Stone-built agricultural buildings are invariably left bare, but a careful examination reveals that most of these were once whitewashed. Very few remain in this condition. Use of stucco as opposed to bare stone is not related to the age or status of houses. Thus stucco and bare stone early examples of late 19th century ‘Georgian style’ houses can be found as well as smaller houses. However, cottages, in particular single-storey examples, are invariably cement rendered, probably owing to the poor quality of their masonry.

Examples of the late 19th and early 20th century tradition of building in corrugated iron (tin) has all but disappeared, apart from the railway station at Devil’s Bridge, a few houses and cottages, and some now very decayed industrial buildings.

Commercially quarried north Wales slate is the universal roofing material. Little evidence of other types of roof is present. However, old photographs show that thatch was once common, at least on cottages and smaller houses.

Building styles
Although the vast majority of buildings date to the 19th century, there are hints of an earlier building tradition. A cruck structure within what externally seems to be a modern house at Llanfihangel y Creuddyn is one such example, and the long, low elevations of several houses in the northern part of upland Ceredigion suggest pre 19th century origins. It is also highly likely that the seemingly 19th century appearance of many houses may hide early internal elements such as chimney-end stairs.

Houses of all types – farmhouses, cottages and industrial buildings – have traits of both the polite Georgian style and the vernacular tradition. Pure Georgian style buildings with a regular floor plan, symmetrical elevations, high rooms, large high windows, and architectural details such overhanging eaves, high chimneys and door cases are rare, but can be found, such as a farmhouse at Cwmsymlog dated 1852. Running parallel to the Georgian style, but with an earlier ancestry, are buildings in the vernacular tradition. Examples of these, that are not influenced by the Georgian style, are rare and typically have an irregular floor plan (reflecting the different status between the hall and the parlour during earlier periods), with a long, low front façade, asymmetrically arranged small windows and one chimney noticeably larger than the other. In both traditions rural houses are typically quite small, two-storey, with a central front door, and one window above the door and two windows either side. Most houses borrow from the Georgian style and vernacular tradition, with mixed traits producing a subtle grading of house types termed Georgian vernacular. Thus it is common to see long low houses with one large and one small chimney (vernacular traits), but symmetrically positioned large windows (Georgian style). By the end of the 19th century traces of the vernacular have all but disappeared apart from in the most basic of homemade cottages. By this date even industrial housing, often the last type of building to shed vernacular traits, has rooms with high-ceilings and symmetrical facades of high windows. Towards the end of the 19th century the influence of the late Victorian gothic style can be detected in houses close to Aberystwyth. There are several large detached ‘villas’ here, which probably were associated with the mining industry. The distinctive tall, narrow style of gothic buildings, with high windows and chimneys also subtly influenced some late 19th century farmhouses in the region, but these are rare.

The buildings of upland Ceredigion reflect the wide social-economic society of the 19th century from the aristocracy and wealthy landowners such as Thomas Johnes’s late 18th early 19th century neo-gothic house at Hafod (now demolished), down to small single-storey home-made squatter cottages. Between these two extremes lie the vast majority of dwellings – gentry farms, estate farms, smallholder- and industrial- worker houses.

In common with dwellings, farm buildings date to the 19th century, with most originating from the later part of the century rather than the earlier. Estate farms and the farms of the gentry have large ranges of stone-built out-buildings, including barns, cow houses, stables and cart sheds arranged semi formally around a yard and sometimes set a little distance from the house. This type of arrangement is typically found in the lower-lying areas of richer farmland such as Abbey Farm at Strata Florida. However, it is noticeable that some of the larger farms on the very fringes of the high uplands have similar arrangements of outbuildings, albeit on a smaller scale. More typically of upland Ceredigion are farms with two or three modest ranges of outbuildings set informally with the house. Smallholdings are also common. These usually have just one range of out-building, typically just a small cow-house and store, often attached and in-line with the house. Most farmland historic landscape areas contain examples of all scales of farm outbuildings. With the amalgamation of farms into larger units and changing farming methods many (perhaps as many as 30%) outbuildings are no longer used. Unlike other areas of west Wales there has been little attempt to find alternative uses for these buildings, such as conversion to accommodation.

Industrial housing is found right across upland Ceredigion, usually in small communities such as Goginan and Pontrhydfendigaid. A short single build terrace is the norm, but semi detached, detached and terraces of different styles can be found. Some are in the vernacular tradition, but most have strong Georgian-style traits. Most of these worker houses are two-storey, but terraces of single storey cottages are present with particularly good examples at Capel Bangor.

Few industrial buildings, particularly from the lead mining industry, survive apart from as ruins. Even so, these ruins can be significant landscape features, such as the stone buildings at Frongoch and Cwmystwyth. An exception is a group of buildings at Pont-rhyd-y-groes, including a mine office and a school of mining, and the buildings of the mining museum at Llywernog.

Modern buildings
Modern dwellings are becoming an increasingly common feature of the upland Ceredigion landscape. The majority date to the late 20th and early 21st century. Houses and bungalows in a variety of styles and materials can be found in most areas, apart from the high uplands.

Some modern agricultural buildings are very large. These, however, tend to be located on farms that have acquired extra land and are not the norm. Generally working farms will have one-or-two medium sized modern agricultural buildings. Many farms are no longer working and therefore do not have modern buildings.

Agricultural change and development
The changing pattern and pace of enclosure has been described above. It is not intended here to describe the changing pattern of agricultural practice from the 18th to the 20th century, as this task has been undertaken by other writers (in particular see Moore-Colyer 1988). The effect of changing agricultural practice on the historic landscape has been less dramatic that that of the changing pattern of enclosure. However, there is one 20th century agricultural scheme that deserves special attention as its effects on the upland landscape of not just Ceredigion, but the whole of upland Britain have been profound. This is the Cahn Hill Improvement Scheme, devised by R G Stapledon of the Welsh Plant Breeding Station in the 1930s (Colyer 1982, 100-104).

Stapledon’s plan was to transform unproductive upland into high quality pasture. With this in view he obtained a gift of £3000 a year from Sir Julian Cahn. In 1933, the Welsh Plant Breeding Station took over 2200 acres of waste on the Hafod estate near Devil’s Bridge and a further 2700 acres of higher sheep-walk, and set about transforming it. The results were spectacular - a five-fold increase in edible dry matter. The continuing transformation of upland moor and wastes into productive grassland is a direct result of Stapledon’s work. Apart from forestry, there is no one single process that has done more to alter the historic upland landscape than improvements based on Stapledon’s research. The results can be seen right across the uplands.

Communications
Routeways in the study area are determined by topography. A main route corridor runs from north to south following the Teifi valley south and the Rheidol valley to the north. To the west of this route-way the landscape is characterised by relatively low-lying enclosed ground comprising villages and dispersed farms linked by a complex network of roads, lanes and tracks. To the east lies high, sparsely populated, unenclosed land. Crossing this mountainous area from west to east are several roads linking the coastal lowlands of Ceredigion with central and eastern Wales and England. Major settlements have developed at the junctions of the mountain roads and the north - south route-way. From south to north these are Tregaron, Pontrhydfendigiad, Ffair Rhos, Ysbyty Ystwyth/Pontrhydygroes, Devil’s Bridge and Ponterwyd. A decline in road transport with the coming of the railways in the mid 19th century led to a gradual erosion of the market and other functions of the major settlements in the study.

Pre-turnpike roads
The major north-south route-way, mirrored by the present B4343, has lost much of its former importance, but its past influence is clear, as is its antiquity. On its course lie the Medieval town and important pre-Norman ecclesiastical centre of Tregaron, Pontrhydfyndigiad village and the nearby Strata Florida Abbey, Ffair Rhos, a location for Medieval and later fairs, Ysbyty Ystwyth church and village, the associated mining settlement of Pontrhydygroes, the tourist settlement that developed at Devil’s Bridge from the late 18th-century, the medieval chapel and prehistoric site at Ysbyty Cynfyn , and the post-medieval village of Ponterwyd. To the north of Ponterwyd the natural routway is not well-defined, but the road once continued across mountainous land to Machynlleth.

The antiquity of the mountainous west - east routes is also not in doubt as most follow natural route corridors; first running in deep, steep-sided valleys before climbing the mountain ridge which lies to the east. By the Medieval Period several of these route-ways would have been employed by Strata Florida Abbey for access to the abbey’s vast upland estates, but their most celebrated use was by drovers for moving cattle and other livestock from collecting centres such as at Tregaron and Ffair Rhos to markets in England. Close proximity to a mountain road was an essential element in the settlement and exploitation of unenclosed upland for all periods. A decline in the importance of these roads in the 20th century has accompanied a recession in settlement of the high uplands - there are only a handful of high mountain farms now occupied.

Turnpike roads
Turnpike construction in the late 18th century had a profound affect on settlement and commerce. New villages were established on the routes of turnpikes; those some distance from these new roads suffered absolute or relative decline. In 1770, a turnpike was constructed from Aberystwyth to Devil’s Bridge), through to Cwmystwyth and over the mountains to Rhayader (Colyer 1984, 176-182; Lewis 1955, 42-45), with a spur from Devil’s Bridge through to Dyffryn Castell and over to the northeast. These turnpikes were superseded by a new road built in 1812 from Aberystwyth to Ponterwyd and beyond. The effects of these new roads on the landscape was dramatic; hotels constructed at Ponterwyd to serve the traveller were instrumental in the development and expansion of the village, similarly at Goginan, and all along the new roads new settlements were founded and old ones expanded.

Railways and tramways
The importance of the north - south route corridor was emphasised in 1867 when the Manchester and Milford Railway opened its line from Lampeter to Aberystwyth. The most important effect on the landscape was the development spur it provided to Tregaron, but this has to be offset against the adverse effect on the town’s cattle trading and droving industry. In 1893, the opening of a mineral line in the north of the study area permitted the greater development of several mines that were served by it, but this was a short-lived venture. A second mineral line opened in 1902 whose primary purpose was to serve mines in the Rheidol valley and had a secondary function as a tourist route in the summer months. It still runs as a tourist railway, a function that has helped Devil’s Bridge (Pontarfynach) at its eastern terminus, to develop into its present form.

Industry
Metal mining
It is probably correct to state that no aspect of Ceredigion has been more studied, investigated and recorded that the metal mining industry. There is a vast amount of published literature on the subject ranging from papers concerned with the archaeological investigation of single sites, to general histories and photographic records - for instance: Lewis (1998); Bick (1974, 1983, 1988); Hughes (1988) and Carr and Schöne (1993). It is not the purpose of this study to provide a potted history of lead mining in Ceredigion, nor is it intended as a guide to the industrial archaeological remains. The purpose of this short thematic section is to provide a brief introduction on how the metal mining industry has affected and influenced the historic landscape.

Archaeological investigations at Copa Hill, Cwmystwyth have demonstrated the Bronze Age origin of this mine (Timberlake 1995). Roman workings have also been postulated at Cwmystwyth and the mine was worked in Medieval Period. It was not until 1690 when the Mines Royal Society monopoly on mining argentiferous ore was broken that private investment flooded into the industry, creating new opportunities, new mines and new jobs (Lewis 1998, 160). Booms in the industry were just as common as slumps, mining conditions were arduous and transport always a problem in such a remote area, but despite these difficulties mines continued to operate until the 1930s.

Apart from some sites in the extreme north of the county, all north Ceredigion metal mines lie within the study area. The impact of these mines on the historic landscape is considerable: there are very few historic landscape areas that do not contain some physical evidence of mining, and many possess considerable remains. Spoil heaps are the most common, and in many instances the most obvious and dramatic, remains associated with lead mines, but other structures are also often present: engine houses, processing plants, crusher houses, wheel pits, dressing floors, shafts and levels, and inclines and tramways. Because of the industry’s extraordinary reliance on waterpower, leats, dams and reservoirs are ever-present in Ceredigion’s upland landscape. An account of mining remains is provided in each of the relevant historic landscape area descriptions.

The success or not of the metal mining industry had a direct affect upon population levels and a concomitant effect on the settlement pattern, transport infrastructure and limits of cultivation and enclosure in this upland area of Ceredigion. Aitchison and Carter (1998, 8) have noted that in three parishes in north Ceredigion, a rapid rise in population in the second half of the 19th century, followed by an equally rapid fall, can be directly attributed to an increased production of lead ore and subsequent exhaustion of veins and mine closure. A graph of 19th century lead ore production from Ceredigion mirrors that of population. Abandoned cottages, houses and a retreat of cultivation from marginal areas are some of the effects of the changing fortunes of the lead mining industry that have embedded themselves in the historic landscape.

Owing to high levels of toxicity many spoil heaps and associated remains have been, and will be, subjected to environmental improvement works. Substantial engineering can be involved in decontamination, detoxification and reclamation programmes; usually resulting in the removal or landscaping of spoil heaps. The effect is to render mining remains less visible in the landscape, though care is taken to ensure that important archaeological remains are not damaged. Goginan, Cwmsymlog, Cwmerfyn and Cwmbrwyno are mines that have experienced environmental improvement in recent years.

Peat cutting
Peat provided an important source of fuel for the inhabitants of Ceredigion until the early 20th century, when it was eventually displaced as a domestic fuel by coal. A large proportion of the county’s land surface is still covered by peat, the largest single peat bog being Cors Caron at c.1067 ha. Across much of the Ceredigion uplands and in isolated pockets away from the higher ground, substantial areas of peat remain.

The peats of the Ceredigion uplands have been accumulating since the end of the last glaciation. Analysis of these peat deposits can tell us a great deal about the environmental history of the region over the past 12,000 years. They are immensely important to the archaeologist as they preserve organic matter such as pollen, plant remains and wood, as well as animal and even human remains (a human body was discovered buried in peat, approximately 2 feet below the surface, in Gwnnws in 1811 and buried at Ystrad Meurig parish church).

The rise of peat as a domestic fuel may be closely linked with the decline of the availability of wood during medieval times. By the 17th century, peat was certainly the main fuel source in many upland districts, and 18th century estate maps show that ‘turbaries’ (areas where peat was cut for fuel) were common in Ceredigion. Indeed, the right of turbary was considered as important to the common people as the right of pasture on the extensive common lands of the Cambrian Mountains (Owen 1990).

Turbaries were not enclosed parcels of land, but rather areas of mountainous ground (usually on common land) where either by tradition or by permission of the manorial lord it was allowed to cut peat for fuel. Turbaries are readily identifiable in the upland landscape even today. The most telling sign that peat has been cut is usually a crescentic scar across the landscape, as much as 40-50m across, creating a step approximately 1m high across that portion of ground. Many cuts and depressions can be seen in the area of a turbary, often extending over a large area. The trackways along which the heavy loads of peat were dragged on sleds across the commons and down to the farms and villages were often well built and can still be followed in many upland areas. An obvious upland turbary that is shown on mid 18th century Nanteos estate maps is that of Y Dderw, near Ponterwyd. Several trackways can be followed from farms along the Rheidol Valley up onto the turbary, which apparently served the inhabitants of the valley between Ponterwyd and Devil's Bridge.

Not all turbaries were on the upland commons. Areas of peat on lower ground were also cut for fuel. Cors Caron, which fills a substantial portion of the Teifi valley bottom, is the most impressive example of this. However, most lowland bogs were relatively small and improvements in land management meant that they were increasingly drained and turned into more valuable farmland during the late 18th and 19th centuries.

Woodland and Forestry
Woodland and estate management

Many hills therabout (Strata Florida) hath bene well woddid, as evidently by old rotes apperith, but now in them is almost no woode. The causses be these; first the wood cut doun was never copisid, and this hath beene a great cause of destruction of wood thorough Wales. Secondly, after cutting doun of woodys the gottys hath so bytten the young spring that it never grew but lyke shrubbes. Thirddely men for the nonys destroied the great woddis that thei shuld not harborow theves ... a hille side Clothmoyne, wher hath bene great digging for leade, the melting whereof hath destroid the woddes that sumtime grew plentifulli therabout

So wrote John Leland in his celebrated itinerary of the 1530s (Toulmin-Smith 1964, III, 118).

Woodland that is shown on late 18th century estate maps on the steep valley sides of the Rheidol, and on the lower slopes of the Teifi valley close to Strata Florida Abbey can be attributed to estate management. Parts of the Rheidol valley lay within the Crosswood estate, documents for which record woodland management from the early 18th century (Edlin 1959, 19), while land at Strata Florida came into the hands of the Stedmans, and later became part of the Nanteos estate. Much of this woodland is still in existence today. Thomas Johnes of Hafod was the great pioneer of upland afforestation. Linnard (1970) has described the planting techniques, fencing and species of the hundreds of thousands of trees that Johnes planted in the final decades of the 18th century and the first years of the 19th century. Remnants of his planting are still evident at Hafod, but his great areas of upland afforestation have not survived. Much was neglected by later owners of the estate, though a large portion survived on high land to the north of Gelmast farm into the 20th century, but was cut down during World War One and has now been replanted by the Forestry Commission (Edlin 1959, 12).

20th century afforestation
In the 20th century, no other single process has had such a dramatic effect upon the historic landscape of upland Ceredigion than afforestation. Steep valley sides have been cloaked in woodland and vast tracts of unenclosed moorland transformed into upland forests. The ethos, methods and techniques behind this afforestation are set out in a Forestry Commission book of 1959 (Edlin). Early post-World War Two and pre-World War Two afforestation concentrated on relatively lower-lying ground and steep valley sides, and often involved replanting of old estate woodlands and filling in the gaps between old established deciduous woods. The steep valley sides of the Rheidol, Brwyno and Arian, and Ystwyth are examples of such planting. Some higher level planting was undertaken, such as that begun in 1929 at Tarenig. Later 1950s, 1960s and 1970s planting was concentrated on high unenclosed moorland, including: valley sides as well as the high ground of the Hafod estate which was acquired by the Forestry Commission in the early 1950s; further planting of Tarenig, isolated blocks in the upper Rheidol valley and the massive area known as Tywi Forest.