Thatch in Breconshire – some initial conclusions
On the face of it, the remarkable paucity of pictorial evidence for thatch in Breconshire makes it hard to believe that it was ever a common roofing material in the county. This would seem to be borne out by William Thomas’ description of the town of Brecon in 1698 as having ‘not above seven thatched houses, which I believe is less than can be found in any town in Wales’. It is clear that Thomas saw this as a matter for civic pride and implies that, already by this date, thatched roofs were considered old-fashioned, and perhaps a sign of poverty. Reducing the number of thatched buildings in a town was also a matter of practical concern as it would reduce the risk of fire where buildings were closely packed.
However, in addition to the references from Penpont documents noted before, other sources confirm that thatch was once widespread elsewhere in the county. In his History of Merthyr Tydfil, Charles Wilkins, describing the appearance of the town in 1782 wrote that ‘most of the buildings were thatched’. In their ‘Buildings of Breconshire’, Jones and Smith wrote of Nant-y-ffin cottage in Ystradgunlais that ‘like a number of other houses in the district it was roofed in living memory with thatch’ (Brycheiniog 16, p66). This was written in 1972 when ‘living memory’ might suggest some survival into the early years of the twentieth century. The authors did however add that it was ‘a material now entirely out of use’.
One reference that, for me, hints that thatch was once not just widespread but perhaps the most common roofing material, is a 1764 lease of Wern Figin farm in Trallong. The main house, formerly a gentry house but then decayed and in danger of collapse, was referred to as the ‘Tyled House’. For this description to have been useful suggests that it was unusual, and implies that most other nearby buildings were thatched. Stone tile might have been available in the area, but it had to be quarried, dressed by skilled artisans and transported. What is more, its great weight meant that strong roof structures with large and expensive timbers were needed to support it. For these reasons, stone tile may have been primarily a roofing material for relatively higher status buildings.
I don’t think it is too difficult to reconcile the evidence for the prevalence of thatch with the lack of pictorial evidence. First, we should consider the subject matter. Most images of buildings, especially those which were printed, depicted towns, ruined castles or abbeys, bridges or gentry houses. These either had no roof at all, or were likely to be roofed in slate or stone tile from an early date. The cottages, farmhouses and outbuildings which would have been thatched were usually, if they appeared at all, incidental to the main subject, and were perhaps also more likely to be depicted in a stylised manner.
Second, it is a matter of date. Depictions of ‘picturesque’ humble country cottages for their own sake were not unknown in the 1700s, but they reached their greatest popularity after the mid 1800s, for example with the works of Helen Allingham from 1870s on. One might expect the advent of photography to increase the records of such buildings, but very few photographs of the county survive from before about 1860.
I think that the story of the thatched cottage next to Honddu Mill, told in the third instalment of this blog, where we can date the replacement of the thatch to between 1820 and 1849, may be fairly typical. I have collected a number of other references to the removal of thatch. In 1779, prior to the building of the Lodge at Penpont, the ‘carrying away of all the thatch’ from the building it replaced is recorded. In 1835, Evan Powell leased a ruined cottage and field in Llanspyddid known as Pullin (unidentified); Powell was to ‘plough in the old thatch’ from the cottage into the field as a soil improver before sowing clover and oats. Another cottage in Llanspyddid on which the thatch had been repaired in 1803 was to be demolished and rebuilt in 1831. We have also seen that the thatched cottage next to Priory Bridge in Brecon had disappeared by 1834.
An outline of a chronology seems to be emerging. Most thatched roofs had already disappeared from Brecon town by 1700, but thatch but was replaced more slowly in the villages and countryside over the following century, a process which perhaps accelerated in the early 1800s. By the middle of the 1800s thatched buildings had become relatively rare, just a handful of examples surviving into the early 1900s.
What drove this change? One possibility is that improved transport links made slate a cheaper and more easily available alternative to thatch. The opening of the Brecon canal in 1800 would certainly have helped to reduce its cost, but this doesn’t mean that slate immediately became readily and cheaply available. When Penpont house was re-roofed in 1803 it was still necessary to place a bespoke order direct with the Caernarfonshire quarries, the material being shipped by coaster to the south coast ports, transhipped onto canal boats and finally onto carts. Stone tile continued to be used, for example when Bryn Fedwyn farm in Trallong was rebuilt in 1822 it was roofed in tile. It may not have been until the coming of the railways around 1860 that slate became really ubiquitous.
At least as significant were economic and agricultural factors. In theory, it might seem of benefit for landlords that their tenants bore the cost of repairing thatched roofs. In practice however, it had drawbacks. When Penry Williams III of Penpont visited Blaen Hebstea farm, Ystradfelty in 1824, he noted that the thatch was in a bad state. His main tenant John Walters promised to see that his under tenant repaired it but, despite this, on the next visit in 1835 Penry found ‘the whole of the buildings are in a bad state particularly the small barn, entirely owing in my opinion to the very neglected state of the thatch.’ In other words, the failure of the tenant to maintain the thatch adequately was causing wider damage to the buildings and so additional costs for the landlord.
A failure of maintenance on the part of tenants seems to have been a general problem for landlords, often requiring expensive repairs when a new tenant came in. In those circumstances the extra capital investment needed for a stone tile roof may have been seen as worthwhile. But this should also be seen in the context of a wider programme of rebuilding farms to a higher standard, not only to reduce maintenance costs, but also to encourage the latest agricultural methods and attract tenants prepared to pay higher rents. On the Penpont estate, a high proportion of farms were wholly or partly re-built in the first half of the nineteenth century.
However, if landlords were making significant investments in new buildings, this went hand in hand with a desire to rationalise their estates and concentrate their investments on fewer but larger farms. After his 1835 visit to Blaen Hebstea, Penry Williams III noted ‘I think these buildings very unnecessary on such a small taking’ and suggested the tenement could be amalgamated with neighbouring holdings: ‘I told the tenant John Walter that if he wished to repair the buildings I would give him the timber but the work must be done at his expense’. This process of amalgamation of holdings can be seen on a number of Penpont properties where farms which can be found on estate maps of the 1740s have disappeared by the time the tithe maps were surveyed around 1840.
As well as this reduction in the number of farmsteads, many cottages also disappeared between 1740 and 1840, and others were gone by the time of the Ordnance Survey maps of the 1880s. For cottages owned by landlords, as for farms, maintenance costs were a key consideration, and the rents payed by cottagers were not considered profitable.
Attitudes to cottage dwellers were also ambiguous at best and distinctly hostile at times. The need to house industrious labourers and deserving widows might be acknowledged, but other cottagers were considered indolent at best and potential criminals or burdens on the parish poor rate at worst. In a letter of 1817 regarding the Trallong Common cottages, Penry Williams III wrote ‘I do not wish to make any profit from the tenements, but merely retain them, in order that they may not get into improper hands’. Given such attitudes it is not surprising that many landlords simply allowed cottages to fall down rather than repair or rebuild them.
The above gives us the beginnings of a timeline for the disappearance of thatch from Breconshire, and some possible reasons for that disappearance. But it does not answer the question of why thatch seems to have disappeared earlier and more completely from this county than from its neighbours. I think that there are several strands which could be investigated as potential causes.
One relates to ownership: was the proportion of buildings owned by their occupiers, by landlords of large estates, or of small estates significant? Did the willingness of these various types of owner to invest in replacement buildings make a difference? For example, were owner occupiers or smaller landlords more likely to retain thatch than large estates which might have better access to capital?
Another might be the availability of alternative materials. Breconshire had relatively good access to stone tile from an early date. However, it was not alone in this, and those parts of Glamorgan where thatching thrived would had similar access to tilestone, and easier access to true slate from the nearby seaports.
A third is the nature of the materials available for thatching. Walter Davies in his General View of the Agriculture of South Wales of 1815, stated that the high quality of Glamorgan thatch depended on carefully hand reaped wheat straw, especially that from soils around St Donat’s and Pennarth. In Breconshire, wheat was much less grown, oats being the main crop, especially for smaller farmers. One of the attractions of thatch was that the raw material could be gathered locally at low cost, or even free from one’s own farm or from common land. If the poorer tenants of thatched buildings in Breconshire did not have easy access to wheat straw, and had to rely on less durable materials which needed frequent replacement, then the incentive to turn to alternatives might have been higher. Changing crops, an increase in pastoral farming, and the enclosure of common land are all factors which might also have affected access to thatching materials.
We might also consider whether there was what we might call a ‘tipping point’ for thatching skills. Once the number of thatched buildings, especially those of the wealthier classes, had fallen below a certain level, there might no longer have been enough work to support professional thatchers to keep those which did survive in good repair. Whatever the reasons for the disappearance of thatch from Breconshire, I remain surprised how something once so ubiquitous could have left so little trace in the records. I hope that this is something which further research might go some way to correct. There are a number approaches which might be taken. I’m sure that there must be further evidence for thatching in the records of other estates, as well as in descriptions in documents such as tour journals. It might also be possible to examine the presence of professional thatchers in the county through census and other records. Other images of thatched buildings must certainly be out there, but there may also be evidence for thatch in the physical structure of surviving buildings – for example the steep pitch typical of thatched roofs, the raised gable walls which were a feature of some thatched buildings, or even the physical presence of the remains of fixings, a base layer or underthatch.
