Hints from Hamptonne: How to build a Jersey farmhouse

When, in 1988, the National Trust for Jersey purchased a semi-derelict farmstead called Hamptonne, a detailed archaeological investigation followed as part of a seven-year restoration programme. The multi-period site, with 15th-century origins, opened to the public as the Hamptonne Country Life Museum. Building on what was learned from this project, Warwick Rodwell’s newly published research report tells us almost everything you need to know about the archaeology of vernacular houses in Jersey, as Chris Catling reports.

I said ‘almost everything you need to know’ because, as Warwick Rodwell explains in Hamptonne and the Archaeology of Vernacular Houses in Jersey (see ‘Further reading’ on p.44), there is very little ancient structural timber surviving in Jersey’s houses, churches, and castles, and what does survive is not suitable for dendrochronological dating. There are further traps for the unwary: datable architectural features were frequently reused in later buildings, and the date stones that are found on so many buildings in Jersey do not always relate to the fabric in which they are embedded: some are salvaged from earlier buildings or other sites, and some record the date of significant nuptials rather than the building’s construction.

Additionally, cultural traditions that have been superseded on the mainland have a much longer life in Jersey. Round arches decorated with rope mouldings that would be considered 12th-century in other parts of Europe continued to be used on the island to decorate 16th-century churches, such as St Brelade’s. King-post roofs of a type found in England and France in 1300 were used in 15th- and 16th-century buildings here.

Having issued his caveats about the difficulties of absolute dating in Jersey, Warwick Rodwell then gives a comprehensive and illuminating account of the ways in which traditional farm buildings, like Hamptonne, were constructed. This is based on the evidence found during the restoration of its buildings and on his further research, over four decades, on some 20 other Jersey houses of similar age and character.

These were built principally from undressed pink or speckled granite and shale from local quarries. Dressed masonry was reserved for the principal façades and quoins (or corner stones), for framing the more important doors and windows, and for fireplaces. Older buildings that survive relatively unaltered exhibit intrinsic homogeneity, whereas the use of a jumble of stone types usually indicates a building that has been rebuilt using material of different ages from different sources. Another clue to age is the degree to which the walling stones have been dressed – rubble indicates an older building and squared blocks a 19th-century date or later.

Occasionally very large stones are used in the foundations or quoins, leading to speculation that they might have been prehistoric menhirs (standing stones) or parts of a megalithic tomb, but Warwick says the evidence is slender and most are simply not large enough.

The usual building practice was to dig a shallow foundation trench, rarely more than 30cm (12in) deep, and to use undressed stones to form a footing, perhaps with a rough offset or a slight batter (inward lean). Masons raised walls in a series of ‘lifts’, representing a day’s work. At the end of each day, an attempt would be made to bring the stonework up to a common level and cap it with a thin spread of mortar. In some buildings, each day’s lift of around 30cm can be clearly detected as a continuous mortar line.

Once a wall reached a height of around 1.2m (4ft), it was necessary to provide a working platform for the masons; further platforms were needed for each additional 1.2m of walling, until the eaves were reached. Sometimes planks supported on trestles were used, but firmer support for scaffolding could be achieved by embedding the ‘putlogs’ (the ends of the horizontal scaffolding timbers) into the masonry as it was built. When the scaffolding was removed, the resulting putlog holes were often filled in with stone and mortar, but they can be detected with an analytical eye. Some remain open, though, and prominently visible, while others may well have been used to create pigeon nesting-holes.

Quoins can give a broad indication of date: medieval examples are irregular in size and shape, and only in the 16th and 17th centuries do we see an attempt at regularity. Quoins of this date often look like the distinctive long-and-short work examples associated with Saxo-Norman churches on the mainland, but with a difference. In long-and-short work, the stones are laid with the long axis alternating between vertical and horizontal. In Jersey, quoins of early modern date are laid side-alternately – that is to say, every stone is laid horizontally, but with the long dimension alternating between adjacent walls.

No attempt was made to square the quoin stones until the mid-18th century, and earlier quoin stones are tapered, with the wide end forming the wall corner and the ‘tail’ bonded into the adjacent rubble masonry. Simple chronologies like this break down, however, when (as very often happened) walls or complete buildings were reconstructed, and the quoin stones were reused. Quoin stones can thus be several centuries older than the buildings in which they are embedded. Not all buildings had right-angled quoins, either. Radiused (rounded) corners are most often used where the space between buildings was restricted, where carts needed to manoeuvre, or to prevent livestock from being damaged by sharp corners.

Between the 13th and 19th centuries, houses and farm buildings were almost entirely constructed using earth-based mortars, often based on locally occurring loess – a combination of clay, sand, and silt. Render or limewash was then applied to aid waterproofing and prevent the mortar being washed out of the joints. There are no natural deposits of limestone in the Channel Islands, and the earliest records of imported lime from France or England date from 1757. For a short period in the early 19th century, oyster shells were used for lime-production, and two lime kilns from this time have survived – one dated 1823, the other of a similar date at Jersey Zoo.

Once lime-based mortar began to be used, it was restricted to the joints around the principal architectural components, such as window and door surrounds. At Hamptonne, it was also used for some outward-facing masonry in combination with earth-based mortar. Lime mortar might be used for the outer 3cm to 5cm (1in to 2in) of the joint between two adjacent blocks of stone, with earth mortar behind that – thereby making the best use of the limited supplies of the harder and more weatherproof lime-based material.

Fireplaces and surrounds for doors and windows are the components on which stone masons lavished their greatest skills. The degree of uniformity in the dimensions and constructional details seen at Hamptonne, and elsewhere in Jersey, leads Warwick to conclude that the majority of these components were not dressed on site from roughly shaped granite blocks, but were supplied ready for use from quarry-based workshops.

The same window dimensions were repeated in numerous widely dispersed buildings: the majority of arched doorways were made up of nine stones arranged in identical combinations, while the principal components of fireplaces were also standardised, even if the decorative details vary. Blacksmiths could make iron window grilles to a standard size, too, knowing that they would fit the window surrounds being made by quarry masons.

By contrast, decorative details varied considerably and must have been created by carver masons on site in the form of chamfers, mouldings, and chamfer-stops (the decorative carving at the foot of the door or fireplace jambs). ‘There is no agreed nomenclature for describing the bewildering variety of chamfer-stops found in the Channel Islands,’ says Warwick, but they include stylised leaf shapes, fleurs-de-lys, acorns, vases, crosses, eggs, and even towers with doors, windows, and spires or pyramidal roofs. Keystones above door arches are carved with armorials, dates, and decorative motifs, and inside the house the principal features to incorporate decorative carving were the fireplace corbels and lintels – the latter providing another opportunity to display an armorial shield and a date.

The lack of oak forests with sizeable trees in the Channel Islands explains why there are so few timbers suitable for tree-ring dating, and why most of the roofs found in older Jersey buildings are of poor quality, using immature oaks or hedgerow trees, which were seldom straight, peppered with knots, and of modest dimensions. Though pine began to be imported from the Baltic region in the 17th century, its use was initially restricted to doors, panelling, shelving, and other fittings.

Jersey thus lacks the flamboyant hall roofs of the mainland, and the overwhelming majority of the roofs in the island’s historic houses and farm buildings date from the 18th or 19th centuries. The few surviving pre-19th-century roofs are usually one of three types: the king post (with Hamptonne as the most impressive example), where the eponymous central upright rises from the tie beam (the main horizontal timber) to the principal rafters; the upper king post, where the tie beam and king post are located much higher up in the truss to allow unimpeded storage space below; and, simplest of all, the A-frame truss, lacking a king post, with principal rafters crossing at the apex to support a square ridge piece.

Thatch was almost ubiquitous as the material to cover roofs in the Channel Islands until the 18th century. Where earlier slate was found at Hamptonne, it was mostly used, along with oyster shells and ceramic tile, as a packing material for levelling up the dressed stones used for doors, windows, and fireplaces. Finds of imported French slate fragments at Hamptonne show that this material was also used for limited decorative effect on the porch, the dovecote, the staircase turret, and the cistern, or dipping well.

During the course of the 18th century, inflammable thatch began to be replaced by Welsh slate or by ‘Flanders tiles’ as they were known – pantiles that had to be imported from the mainland, because Jersey lacks clays suitable for tile-making. This required structural changes to roofs in order to extend the rafters. With a thatched roof, the principal rafters are bedded into the inner face of the walls: the thatch, being some 30cm (12in) in depth, sits partly on the wall top and then extends beyond the wall to throw rainwater clear of the building. Warwick says that the evidence of the changeover is everywhere in Jersey’s older buildings, with extra timbers and packing pieces used to extend the rafters outwards to project well clear of the outer edge of the walls and provide the necessary overhang for guttering.

Prior to the 18th century, puddled loess was spread and tamped to create the terre battue (‘beaten earth’) that was commonly used for floors at ground level. Areas that received heavy use – such as doorways – could be repaired cheaply and easily. Later in the 18th century, stone paving from Swanage, in Dorset, began to be imported in quantity and used in halls and principal rooms, while smaller stones, clay pavers, and bricks were used for the service-area floors.

Timber floors in upper rooms display features that were necessary to compensate for the lack of substantial pieces of wood to form the bridging beam – the main flooring timber spanning two opposing walls on which the smaller joists would rest to support the floorboards. Some houses have a forked beam – the result of using younger oaks and retaining the primary branches as well as the trunk. Frequently the bridging beams were not straight or regular in section, so various means had to be used to ensure that the joists provided a level surface for the floorboards: lowering the joist by cutting a pocket in the top of the bridging beam or raising the joist using a wedge. The use of mortise and tenon joints to lock the joists into place along the bridging beam was not common, and Warwick cites an unusual court case that resulted from this practice, dating from 1596. Anything not nailed or pegged to the structure was, at that date, regarded as part of a deceased person’s ‘moveable estate’, and this fact was used by one legatee in an attempt to remove the recently installed joists from a house in St Helier. The court ruled against the legatee, and stated that the joists had to be left in the house.

Terre battue flooring was commonly used for upper floors until the 18th century. The use of earth flooring meant that the supporting boards could be made up from offcuts and waste timber: many such floor slats were made from old barrel ends or staves and, in farm buildings, closely spaced cabbage stalks and branchwood would also be used instead of timber joists. A bed of straw was then laid down on top of the joists as a base for the earthen floor. The increased availability of pine boards for flooring in the 18th century led to the replacement of terre battue floors, but their former presence can often be determined by the shadow left on the surrounding walls.

The underside of the beams, joists, and flooring used on upper floors is visible in the rooms below, and, where these were not designed for display, it was usual in England to hide them by means of a lath-and-plaster ceiling – either plain or embellished with ornamental mouldings and motifs. None of this was common in Jersey until plastered ceilings began to appear as a result of 19th-century refurbishments, and, even then, only in the finest houses. More commonly, from the early 19th to the mid-20th centuries, joists were concealed by nailing boards to their undersides.

Jersey fireplaces, says Warwick, deserve a detailed study in themselves, being the dominant internal feature in any early farmhouse. Typically, from the 16th to the 18th centuries, separate fireplaces were located in the gable- end walls at the opposite ends of the house: one on the ground floor of the hall, and another in a first-floor chamber – seldom one above the other in the same wall. Half the depth of the hearth would be contained within a wall recess and the other half would project into the room, so a canopy or hood was required to contain the smoke, corbelled out over the hearth and supported by a timber or granite lintel.

Ovens could also be contained within the sides of the fireplace, or at the back, in the form of a projection from the outside of the gable wall. Sometimes fireplaces incorporated what are commonly, though not necessarily correctly, described as salt boxes: a keeping-place in the back or side wall for objects that needed to be kept warm and dry.

Another feature that Warwick nominates for further full-scale study is the stair tower, or tourelle, that a great many early farmhouses in Jersey once had. These are almost invariably located at the rear of the building, opposite the main entrance, or in the angle between the wings of an L-shaped house. They can be circular, semicircular, elliptical, square, or rectangular in plan, and the stairs they contain mainly linked the ground and first floors, though some rise to the attic level, and a few ascend further still to emerge above the roofline as a freestanding tower. Fragmentary evidence has survived for timber-framed upper stages, but most of these towers were truncated in the 19th century and incorporated into the main roof slope, or raised in height to create observatories, or ‘prospect’ rooms.

Probably inspired by examples in Brittany, these attractive features were fashionable until the close of the 17th century, when they were superseded by fully internal timber staircases, of which a fair number dating from 1750 to 1800 have survived, and many more from 1800 onwards.

Great granite arches are another distinctive feature of Jersey’s domestic buildings. Hamptonne has an impressive pair forming the twin-arched entrance to the walled farmyard. Carved with the date 1637 and with the initials LH and EH (Laurens and Edouard Hamptonne, father and son), it is decorated with roll mouldings and chamfers, with a large arch for vehicular access and a smaller pedestrian gate. The farmhouse itself has no fewer than three arched entrances, each composed, like many others on the island, of nine granite blocks: three on each side forming the jambs and three forming the rounded head. Arches, in general, frequently have initialled and dated keystones, most of which are assignable to the 17th century.

Warwick detects the telltale evidence that porches were once attached to some of these doorways as part of a fashion for adding such structures in the 17th century, followed by their removal in the 19th century. None have survived intact, and it is not known how common they were, but there are two in situ pillars at Hamptonne dating from c.1640 that provide clues, and which have been incorporated into the new main entrance to the house. In this reconstruction, the slender granite pillars are used to support the wall plates for a double pitched roof covered in French slate with ceramic ridge tiles, both of which are evidenced by excavated finds. Comparable porch pillars have been found all over Jersey reused in walling, or as lintels or gateposts, and their similarities led to the suggestion that they were all the products of a single quarry workshop.

Also reconstructed at Hamptonne, based on excavated evidence and parallels, is the colombier, or dovecote, which stands apart from the main complex. It lost its original roof early in the 19th century, before being mantled in ivy by the 1920s, and converted to a two-room farm dwelling in 1961. The dovecote is significant in the history of the farmstead because the documents have survived in which permission was granted for the construction of the building in 1445. A second document, signed ‘Charles R’ by Charles II, gives royal permission for its rebuilding in 1649, the year of his father’s execution and his Scottish coronation. In the event, the colombier was not rebuilt until 1674. That date and the initials of Elizabeth Hamptonne and her husband Josué Ahier are inscribed on a granite block set above the door lintel.

Hamptonne has long been recognised for its exceptional state of preservation, but it was very nearly lost to development when planning permission was sought in 1985 to convert the buildings into holiday accommodation. Fortunately, it was saved, and Warwick Rodwell’s book contains an exemplary and comprehensive account of what happened next. He modestly warns us about the limitations of his research: in reality, he has produced a model analysis of what makes Hamptonne at the same time so special and so typical of the many hundreds of vernacular houses in Jersey that have not been as fortunate.

Further readingWarwick Rodwell, Hamptonne and the Archaeology of Vernacular Houses in Jersey (Société Jersiaise, ISBN 978-0901897930, £90).

Roger Forshaw investigates this intriguing institution and its role in ancient Egyptian society.…

Continuing his series of articles about individuals who influenced events in Egypt during their lifetimes, Wolfram Grajetzki now turns his attention to one of…

Ancient Egyptians used honey in cooking, especially to sweeten dishes in the absence of sugar, and in medicine, for its healing properties and to…

Barbara Boczar investigates the sources of the wonderful colours used by ancient Egyptian artisans.…

Lloyd Llewellyn-Jones continues his new series on the later Ptolemaic queens, focusing on the tangled family relationships of the second Cleopatra.…