FROM BRAKDAK TO BAFOKONA: A Study in the Geographical Adaptation and Cultural Transmission of the South African Flat Roofed Dwelling

Franco Frescura

INTRODUCTION

The parapet or lean-to dwelling is a flat roofed structure ubiquitous in many parts of the southern African interior. It has been associated, over the past two centuries, with a wide range of urban as well as rural environments. In the process it has also gained for itself a variety of names, having been identified, from time to time, as the preferred residential form of Karoo, Griqua and Dutch farmers as well as the newly emancipated Malay community. More recently its use has also spread to indigenous builders, to such an extent that it has also become popularly known as a highveld dwelling. Yet this nomenclature is incorrect and highly misleading. For one thing it is based upon a number of wide generalisations and stereotypical group images which have prevailed at one time or another during this country's history. For another, the implications of being a style of construction deny both the economic processes which give rise to it and the historical patterns which link this structure to a Cape and, ultimately, a wider European architectural tradition. This paper traces the origins of the flat-roofed dwelling in southern Africa and examines the social and economic processes which have facilitated its transmission and incorporation into the larger body of indigenous built forms. It also documents the changes and adaptations which take place when the dwelling forms and technologies of one culture are adopted by another.

HISTORICAL BACKGROUND

The beginnings of a flat roof tradition in Cape domestic architecture appears to have originated in Cape Town during the early years of the eighteenth century as a response to the potential fire hazard posed by thatch roofs in a densely populated urban area. This fear of fire had already manifested itself some years earlier, in August and September 1691, when Cape Town's Council of Policy, under Simon van der Stel, promulgated regulations enforcing the reduction of chimneys deemed to be too high, and the treatment of roofs with chalk, clay or other fire-resistant materials (Picard, 1968: 18). This was reinforced subsequently by Willem Adriaan van der Stel who, as Governor between 1699 and 1708, forbade the smoking of pipes in the street (Picard, 1968: 24), an activity which Kolbe, who resided at the Cape between 1705 and 1713, attributed to:

"Sailors and (Khoikhoi) (who) were continually crowding and smoaking their Pipes ... and sometimes, thro' carelessness, set 'em on Fire ... Yet now and then a Sailor or a (Khoikhoi) is seen smoaking in the Streets; and when they are so, there is no Lenity for 'em, if they are laid Hold of, but away they are hurried to the Post, and lash'd indeed very severely." (1731: 346-353)

Although at first only the buildings of the Dutch East India Company (VOC) appear to have been affected, by 1732 the dwellings of private individuals also began to follow suit (Lewcock, 1963). This move away from thatched roofs was accelerated after 1736 when five houses were set alight during a slave uprising in Cape Town. Fortunately for its citizens the south-easter, which had been blowing at the time, abated and the rest of the settlement was saved (Picard, 1968: 38-39).

ILLUSTRATION 1. View of Cape Town, c1710, as pictured by C van Stade. Three small flat-roofed structures may be seen in the immediate centre-left foreground. This picture has been edited to highlight the buildings.

One of the earliest pictorial records of a flat roofed structure was made by E van Stade, a Dutch artist, who visited the Cape in about 1710 (Gordon-Brown, 1975: 225). His view of Cape Town from the harbour shows a small village made up principally of hip and gable roofed buildings. Three units however, facing directly upon the sea front, have unmistakable mono-pitch roofs whilst a fourth may be found nestling beneath the Castle's northernmost rampart, also known as the Buuren bastion (Hattersley, 1969: 48-49) (illustration 1). A similar panorama by van Stade, drawn from the opposite direction, on the lower slopes of Table Mountain, shows the slave lodge, a long flat roofed structure located south of the Church, as well as two similar units located in the proximity of the Company's gardens (Hattersley, 1969: 47).

Half a century later, in 1762, the Danish artist Johannes Rach stopped over in the Cape on his way to Batavia (Gordon-Brown, 1975: 208). His drawings show Cape Town to be a growing community whose residences remain roofed in the older manner although a number of warehouses and larger buildings are now flat roofed. A later drawing of Greenmarket Square in 1764, also by Rach, supports this view by showing the Burgher Watch House, recently completed in 1755, surrounded by predominantly hip and gable roofed dwellings (Hattersley, 1969: 64-65).

Within the next fifteen years however, the roof line of Cape Town was to change dramatically. Andrew Sparrman, who visited the town in 1775, was able to report that:

"A great part of their houses as well as the church are covered with a sort of blackish reed ... The rest of the roofs at the Cape are like flat brick floors ... " (1975: 47-49)

The extent to which flat roof dwellings had spread throughout the settlement was illustrated by the German artist Johannes Schumaker in a panorama of Cape Town he painted in about 1777. Assuming that he has not taken too much artistic license, his drawing shows that out of approximately 349 identifiable domestic structures, only 41% still retained their pitched and, presumably thatched, roofing (Gordon-Brown, 1975: 218).

Although it is obvious that this new technology was first associated with civic buildings whose design in a neo-classical style, marked them as belonging to the VOC, such as the Burghers's Watch House, or the Town House as it became known subsequently, its use soon spread to humbler homes.

When, for example, Sparrman visited a Khoikhoi settlement in the Outeniquas in 1775, he noted that, apart from the usual mat beehive structures normally associated with this group, others "were built of straw in a square form, with shelving roofs, like the cottages of the slaves." (1975: 267)

This statement is important for it indicates that, by 1775, the use of flat roofed structures was spreading to other socio-economic groups in Cape society. Not only were burghers building these dwellings to house their slaves, but they had already done so in sufficient numbers as to create a stereotypical picture of their habitat within the minds of outside visitors. Sparrman's account is also the first reported case of this domestic form being transmitted to indigenous groups, something that was to take place with increasing frequency in subsequent years.

ILLUSTRATION 2. William Burchell pictured this portion of Strand Street in about 1810 as a series of flat roofed dwellings with two older gabled and thatched structures interspersed between them.

The replacement of pitch roofed structures in Cape Town was to continue apace over the next half century and by the time Barrow visited the Cape in 1797, just over a generation later, he was able to report that the town's houses:

"... are generally white-washed, and the doors and windows painted green;  are mostly two storeys in height, flat-roofed, with an ornament in the centre of the front, or a kind of pediment;  a raised platform before the door with a seat at each end." (1804: 340-341)

His wife, Lady Anne Barrow (nee Truter) whom he married locally, painted a panorama of Cape Town in about 1800 and shows that, by this stage, a mere handful of thatched roofs had survived (1804: 120-121), a fact supported by Burchell in 1810 (1953: 53) (illustration 2). By the 1830s few, if any, pitched roofs could be found in the town (illustrations 3 and 4).

ILLUSTRATION 3. The panoramic views of Cape Town depicted by Sir Charles D'Oyly during the early 1830s, such as the fish market above, show few signs of pitch roofed structures. This picture has been edited.
ILLUSTRATION 4. View of Cape Town and Table Bay by Sir Charles D'Oyly, May 1832. This picture has been edited.
THE ORIGINS OF A FLAT ROOF AESTHETIC

Although fire prevention must be regarded to be a strong practical reason why a flat roof aesthetic should have gained such widespread acceptance in Cape Town during the eighteenth century, the predominance of this architectural form within such relatively small geographical confines presents a number of puzzling features. Why, for example, did it not gain wider usage in other Cape settlements of that time? Or, for that matter, why was its use limited to Cape Town whose cold, wet winters made its technology somewhat impractical? The answers, I suspect, are multi-faceted and cover a wide range of concerns.

There is no doubt that although flat roofs were applied at an early stage to small scale and relatively humble structures, their aesthetic was associated from the outset mostly with governmental and hence high design buildings (Rapoport, 1969: 3-6). Later, as their use spread to domestic architecture, the burghers of Cape Town began erecting substantial double storey homes, and although many of these structures retained some obvious stylistic links with the Cape architecture of earlier times, such as the use of a gable on the front parapet, the aesthetic principles upon which they were based owed more to a neo-classical influence originating from Europe than to the nature of local architecture.

The roots of the flat roofed aesthetic have, over the years, been the subject of considerable debate. Some writers, like Clive Chipkin, have claimed that:

"... the low slope, horizontal parapet town house – is linked to maritime areas once under V.O.C. predominance; but ultimately this natural building form derives from the Mediterranean Town House and Marine Tradition." (1985: 30-34)

Barrie Biermann takes this thinking a little further. He points out that not only was the construction of nineteenth century Greek farm-houses very similar to that used in the Cape's flat roofed dwelling (Pers comm, March 1989), but that the Portuguese erected flat roofed buildings in both East Africa and the Far East. It was inevitable therefore that a maritime nation such as the Dutch should come into contact with this domestic form, as they traded with and, in some cases, conquered and occupied, former Portuguese colonies. In Sri Lanka, for example, the Dutch established a military presence in 1602 and, after a series of protracted wars, finally drove the Portuguese out in 1658. Thereafter they proceeded to implement extensive administrative reforms which included an ambitious programme of public works. Many of the resultant buildings appear to have been constructed in the flat roofed manner of the Portuguese they had replaced, and modern architectural historians often experience difficulties in differentiating between the colonial buildings of Portuguese and early Dutch settlers. It is presumed then that the Dutch of the Cape learnt of the flat roof through their colonial experience and, given the correct environmental stimuli, adopted it as their own (Biermann, 1952).

Lewcock, on the other hand, puts forward a different point of view. He states that during the mid-seventeenth century:

"... a restrained classicism was prevalent in the architecture of Holland.  Pattern books such as that of Philip Vingboom popularised this style, one which favoured a simple rectangular box-like house, with sometimes a central classic sculptured pediment crowning a forward projection in the centre of the facade." (1963: 8)

This is a description which could be applied equally to any number of double storey homes built in Cape Town during the latter part of the eighteenth century and which echoes the words of Barrow a century and a half earlier.

Considering the social and economic complexities of cross-cultural transmission, and the fact that the postulated Mediterranean influence has, to date, been scarcely documented, Lewcock's explanation appears to date to be the more logical of the two theories.

Another factor that should be taken into consideration is the availability of building materials in the Cape. Local timbers were not only found to be unsuitable, being long-grained and thus prone to splitting, but their resources were relatively small. The indigenous forests near Swellendam were quickly exhausted and the yellowwoods and stinkwoods of Plettenberg Bay did not become available until 1785 (Lewcock, 1963: 10). In 1781 le Vaillant was horrified to learn that:

"... the India directors send every year, from Amsterdam, several ships loaded with plank of all kinds, a passage of more than two thousand leagues, into a country abounding with immense forests of the finest timber in the world ..." (1790: 173-174)

Barrow, on the other hand, was more realistic in his assessment when, in 1797, he reported that:

"Timber of all kinds for building is an exceedingly scarce and expensive article at the Cape ... (trees) have been found to thrive most rapidly; but the timber they produce is generally shaken and unsound." (1801: 19)

Therefore, throughout the seventeenth and eighteenth centuries, the builders of the Cape had to rely upon expensive imports of timber from Europe and India to supplement their meagre local stock. Similarly, clay bricks and roof tiles also had to be shipped in and were thus prohibitively expensive.

This seems to indicate two things. Firstly, that a shortage of timber suitable for construction promoted the development of a Cape wall building tradition. As such its aesthetic was suited to a more classical approach which emphasised the walls at the expense of the roof structure, a factor which, outwardly at least, would seem to support Lewcock's hypothesis. However it must also be borne in mind that a Mediterranean flat roof tradition makes extensive use of domes and vaulting as part of its constructional technology. A Mediterranean masonry tradition would have used such devices, as indeed they appear to have been in the early structures of the Cape Town Castle, but these precede the main flat roof building era in Cape Town by at least a century. The fact that they never found wider application in the context of domestic architecture in the Cape would seem to indicate that although the local Dutch were no strangers to the flat roof, they never mastered the masonry techniques needed for its efficient performance. This means that the spread of flat roofs in Cape Town can be explained as an attempt, on the part of local builders, to revive a forgotten architectural form whose technology was never fully understood, in order to optimise on a limited source of expensive building materials.

Thus, whilst the burghers of the more closely populated urban centre, Cape Town, were forced to compete for a small stock of timber, and as a result changed their housing form to meet this contingency, the residents of the rural areas found themselves under no such strictures and were able to continue building their homes in their preferred manner. Presumably also, rural settlements being smaller and less densely inhabited were not as liable to the spread of fire through strong winds as was Cape Town and thus they did not feel the need to replace their thatch roofs with less combustible materials.

The question of a town's visual identity and the collective civic self-image enjoyed by many of its burghers may also have played a strong role in pre-determining the spread of such ideas. The village of Stellenbosch, for example, was largely destroyed in December 1710 by a fire attributed by Kolbe to the carelessness of a slave (Hattersley, 1969: 23-24). In spite of this, when the settlement was rebuilt soon afterwards, most of its dwellings retained their distinctive thatched roofs. Subsequent conflagrations which took place in 1762 and 1803 also do not appear to have prejudiced public opinion against this roofing technology to any great degree.

Secondly, the shortage of local timbers would have become particularly acute during the second half of the eighteenth century. This appears to coincide with a period of transition during which most of Cape Town's homes shed their gables and thatch and converted to a flat roof aesthetic. At the same time the availability of improved wall building methods and increasing population pressures would have encouraged many of Cape Town's burghers to expand their homes upwards into a second storey. The form of such houses lent itself to the principles of neo-classical design, thus indirectly also reinforcing Lewcock's arguments.

Taken as a whole these factors would seem to explain why the inhabitants of Cape Town, faced with the prospects of conflagration or, alternatively, a prohibitively expensive roofing technique, made a pragmatic choice to alter their housing to a form whose water exclusion properties were not as efficient but which, nonetheless, represented the minor of two evils. It also explains why the flat roofed house failed to gain wide acceptance in urban areas outside of Cape Town right up to the mid-nineteenth century and why, when it did spread, at first this was only in areas where arid climatic conditions made it feasible for its technology to perform satisfactorily.

TECHNOLOGY AND CONSTRUCTION

Originally the humble single storey flat roofed parapet dwelling consisted of a central doorway giving onto a living/kitchen area with a sleeping room located to one side of it. Windows were usually set equally about the door opening giving the front facade a symmetry which belied the asymmetrical disposition of the plan beyond it.

ILLUSTRATION 5. CE Boniface depicted his own house, at the upper end of Long Street, Cape Town, in about 1832.
ILLUSTRATION 6. The business premises in Cape Town of JC Poortemans, in about 1836, whose artistic skills also extended to lithography and snuff making.

An example of such a building was the residence of CE Boniface, which could be found at the upper end of Long Street in about 1832 (Gordon-Brown, 1975: 86) (illustration 5). On the other hand the house in which the artist Jacobus Poortermans located his lithographic office in 1836 had a two-roomed plan but its facade was made distinctly asymmetrical by the introduction of an additional window to the bedroom (Gordon-Brown, 1975: 93, 205-207) (illustration 6). Subsequent developments of the floor plan saw the introduction of a central voorkamer or, alternatively, of a central passage which led to further rooms to the rear. In both such cases then the symmetrical street facade became a more accurate reflection of the floor plan beyond.

However it must be emphasised that such developments in the floor plan of the single storey flat roofed house were matched, and perhaps even preceded, by similar events in the evolution of pitched or hipped and gabled domestic structures - an architecture which today is described as being Cape Dutch. Thus, in spite of its unique outward form, the flat roofed structure did not become a distinct residential type in its own right and whether dealing with the houses of affluent burghers or the newly emancipated slaves, these early dwellings remained, in their essential character, Cape Dutch houses with Cape Dutch plans. Their new aesthetic therefore did not remove them from the larger body of early immigrant Cape architecture, it merely created a variant of its original form.

One of the factors, which may have promoted the (mistaken) idea that the flat roofed house developed as a separate architectural tradition in the Cape, was the characteristic treatment often applied to its facade. Although there obviously existed a number of variants, the usual practice was to emphasize the door and window reveals with a broad plaster border and the corners were often treated as applied pilasters. A horizontal splash band could also run the length of the facade, rising as high as the window sill.

The parapet walls rose on three sides of the low-pitch roof allowing rainwater drainage to occur to the rear and away from the facade. The front parapet was often corniced although in some examples the plaster mouldings were also heavily ornamented with baroque scrolls (Hattersley, 1969: opposite 102).

Its construction was also unique. The roof often consisted of 25-35mm yellowwood or deal boarding placed on heavy beams. A crushed brick aggregate was laid next and finished with three coats of shell-lime and seashells. This method of building suffered severely from water-exclusion problems (Lewcock, 1963), most particularly in Cape Town where winters are both cold and wet.

Further afield the farmers of the Cape interior also used a roofing technology which became known locally as the brakdak. It is a method of construction common to many arid regions of the Mediterranean, from Morocco to Anatolia, and comprised a series of timber beams, supporting a framework of reeds, timber battens or slates which, in their turn carried a superstructure of earth. In lime-bearing areas waterproofing was achieved by the use of a whitewash or a quick lime coating but, in its absence, this could be substituted with brackish soil which, having a high salt content, had good water repellant qualities. This soil, known as brak, was spread in a thin pulverized layer over an earth screed laid to a low fall so to reduce any possible erosion. The process was completed after the first rains when the moisture would turn the brak into a hard continuous shell which sealed the surface (Barrie Biermann, pers comm, July 1989). The term was derived by the Dutch from the obsolete French word brack, meaning brackish water (Fowler and Fowler, 1975).

ILLUSTRATION 7. Mrs Swaving's French Academy for Young Ladies, by A Reid in 1831, shows many of the features of a fine Cape Town townhouse including applied corner pilasters, a bold parapet cornice and a stepped central gable.

The double storey Cape Town house did not differ markedly from its single storey cousin except perhaps in that it represented a larger, loftier and more polite version of the same theme. The building was placed on a raised plinth, probably of local stone, with a stoep projecting forward to the street edge providing a transitional space between the muddy street exterior and the fine residential interior. The facade was set symmetrically about a central doorway, which was often emphasised by means of applied pilasters on either side and a high skylight above it, and was often terminated at roof height by an ornate cornice and mouldings. Houses with a projecting centre bay were often given a central pediment over the cornice following a standard Vingboom detail. A dakkamer or attic room on the roof sometimes added a third storey to the building giving it the appearance of a gable. Sash windows were set flush with the exterior wall surface, the idea of recessing them being a later improvement introduced by the British. The corners of the house, and sometimes also the individual window bays, were often articulated by means of applied giant order pilasters. Walls were almost invariably plastered with lime mortar and whitewashed over (Lewcock, 1963: 8-11) (illustration 7).

This means that whilst the early dwellings of Cape Town had much in common with the farmhouse architecture being built further inland, the widespread adoption of an urban flat roof aesthetic during the 1750s began a slow and inexorable drift apart in the building traditions of town and of country. The latter retained its gabled and thatched character whilst the former opted for a more contemporary and fashionable outlook and, as will be shown below, where flat roofed dwellings were built in rural areas, these remained a direct extension of the Cape cottage tradition.

THE CAPE COTTAGE TRADITION

Like its early urban counterpart, the development of a flat roof tradition in rural architecture should be read in the wider context of the Cape Cottage, a structure which, in its humble form, bore a close resemblance to the European longhouse. Sparrman visited one of these near Eerste Rivier, in the district of Stellenbosch, in July 1775 and described it thus:

"In most places the house consisted of two rooms only ... The interior of one of these was used for a bed-chamber for the boor himself, with his wife and children. The outer one composed the kitchen, in a corner of which they spread a mat for us on the floor ... The (Khoikhoi) of either sex, who were in the boor's service, always chose to sleep in the fireplace. This mostly took up a whole gable of the house, and at the same time had no other hearth than the floor, ... " (1975: 137)

A similar residence in the Outeniquas he further described as having "but two rooms ... somewhat more than two yards in breadth, and about six long, with a peep hole at one end of it, and a small broken window at the other" (1975: 265-266)

The homes of the graziers further inland present a consistent picture. In about 1797 Barrow described them as having "seldom more than two apartments, and frequently only one, in which the parents with six or eight children and the house (Khoikhoi) all sleep" (1804: 120-121)

ILLUSTRATION 8. Veld Kornet Snyman's homestead in the Agtersneeuwberg, depicted by William Burchell in 1811.

Burchell, who visited the Achtersneeuberg north of Graaff-Reinet in 1811 (illustration 8), described the Vermeulen farmhouse thus:

"The rooms in the principal house being but three (that is, one in the middle in which the family sit and take their meals, and one bed-room at each end)" (1953: 2:82)

while the Niekerk's homestead nearby was of a more humble character:

"All the buildings were of the most miserable description ... The whole house formed but a single room; and in this a large fireplace at one end served for kitchen, where slaves, and some (Khoikhoi) maids, sat within the chimney, cooking both for the company and for themselves. At the other end a screen of mats parted off a bed-room for the female part of the family; while a few blankets spread upon a row of mats on the floor ... formed the only sleeping-place for the two young men, and for any casual visitors." (1953: 2: 85, 87-88)

Not all such reports were as impartial. Barrow, who was forced to take refuge in a shoe-maker's hovel near Cape Town in 1797, complained that:

"There were but two apartments, one of which was filled with the company; the other we occupied. This, it seemed, was made to answer a four-fold purpose of bed-chamber, work-shop, cellar, and storehouse. The heat of the weather, the closeness of the room, which had only one small aperture to admit the light, added to the mingled odours arising from stinking leather, bunches of onions, butcher's meat swarming with flies, fumes of tobacco, dregs of wine and gin and Cape brandy standing in pools on the clayed floor; in a word such "a congregation of foul and pestilential vapours", were sufficient to nauseate stomachs much less squeamish than ours. Nor was the sense of feeling less annoyed by an innumerable quantity of bugs, fleas and mosquitoes." (1804: 120-121)

More importantly however, he also reported that:

"... a true Dutch peasant, or boor as he styles himself, has not the smallest idea of what an English farmer means by the word comfort.  His house is either open to the roof, or covered with only with rough poles and turf, affording a favourable shelter for scorpions and spiders;  and the earthy floors are covered with dust and dirt, and swarm with insects ... His apartments, if he happens to have more than one, which is not always the case among the grazing farmers, are nearly destitute of furniture." (1801: 76-77)

Quite clearly then, the Dutch farmers of the Cape hinterland, probably somewhere beyond Paarl, on the road to Graaff-Reinet, were already building flat-roofed homes of their own by the end of the eighteenth century. These did not use the same technology as the more comfortable homes of Cape Town, being roofed over with simple turf blocks laid directly over the roof beams, an arrangement which could not have kept out more than a light drizzle. However, considering the nature of a grazier's lifestyle, it is unlikely that more substantial arrangements would have been needed.

These reports are all consistent with the findings of Hugh Floyd whose research has linked the early architecture of the Cape to that of the northern European region, more specifically to the longhouse folk building tradition (1983: 28-31). The longhouse is a term used in England to describe a rectangular domestic structure having opposing doorways in the long walls, with the family's residential space being located to one side of the through passage between doorways, and a cattle byre to the other side (Mercer, 1975: 34-49). Although this house form has been documented extensively in the context of an English medieval agricultural economy, its use has also been recorded in other countries adjacent to the North Sea, such as Brittany, Scotland and Denmark. Their chronology varies but examples are known to have been built in some places as late as the seventeenth century. Floyd argues that although the Cape Cottage is not a longhouse in the medieval sense of the word, its colonial use represents an adaptation of the European tradition in that the Cape dwelling was certainly accessed through the long wall and, in cases where it was divided into two cells, one was used as sleeping space for the family and the other was multi-functional. Barrow's complaints about "foul and pestilential vapours" certainly appears to bear this point out.

Translocation of the longhouse tradition to the Cape took place not because of the Dutch, whose farmers tended to build hall houses, but through the agency of the VOC whose employees were recruited from a wide range of coastal countries, from Brittany and the British Isles through to Denmark. Thus when the VOC settled its servants in Table Bay during the latter part of the seventeenth century, this group was neither homogeneous nor predominantly Dutch and chose to model its farmhouses after a pattern familiar and common to the majority of them. Floyd also goes on to demonstrate how the Cape two-roomed structure is typologically linked to a three-cell unit, the basic generator of plan form in the subsequent development of a Cape Dutch farmhouse architecture.

Barrow's accounts also make another important point: they indicate that the migrant Dutch farmers of the Cape had, and often exercised, the choice of erecting either a pitched or a flat roof over their dwellings. Presumably they were guided in this by local climatic conditions and the availability of suitable building materials. This means that, whatever the form taken by the roof over, the plan of the house remained essentially unaltered, thus retaining its links to the larger body of rural Cape cottage architecture.

The implications of this argument are two-fold. Firstly, that the roots of the Cape cottage and, by implication, Cape Dutch architecture as a whole, lie neither in the Cape nor with the Dutch. Secondly, and more significantly to this paper, that the early nineteenth century Cape flat roof tradition does not represent the development of a new and separate cultural identity in Cape society, but should rather be viewed as a series of architectural responses to changing regional and climatic conditions. These took place within the context of existing Cape architectural practices and, as such, belong to the wider patterns of local building custom.

It is true that differences in such elements as form, façade treatment and building technology may give rise to aesthetic variations, but these are not sufficient grounds in themselves for creating such radical divisions. The domestic plan on the other hand remained constant and, having been generated by the activities of its users, is the primary factor in determining the question of an architectural identity, which stands separate from, and should not be confused with, aesthetic style. In the case of the Cape, this is supported by yet more evidence.

ILLUSTRATION 9. The church square or nachtmaalplein at Piquetberg as seen by JC Poortermans in 1857. This picture has been edited.

It does not appear, for example, that, outside of Cape Town and Simon's Town, there were many practical reasons why the rural cottage should either shed its pitched thatch roof or move into a two storey development. There certainly was no rush on the part of colonial villages to follow Cape Town's practice and convert their homes to the new flat roof aesthetic, but then perhaps they were not as concerned with the spread of fire as their mother city, nor were they as bound by questions of style and fashion. Indeed, a pictorial survey of Cape towns during the first half of the nineteenth century reveals that such places as Piquetberg (illustration 9) and Stellenbosch retained their essential early Cape Dutch ridge and gable character up to comparatively recent times; that few if any flat roofs were evident in Tulbagh in 1811 (Burchell, 1953: 1: 85), Bethelsdorp in 1813 (Campbell, 1815), Grahamstown either in 1823 or 1833 (Gordon-Brown, 1975: 87, 158), Fort Beaufort in 1853 (Gordon-Brown, 1975: 165), Uitenhage also in 1853 (Baines, 1964: 304) or Port Elizabeth in about 1860 (Gordon-Brown, 1975: 172); and that even Cradock in 1844 and Graaff-Reinet in about 1850 had a large proportion of pitch roofed structures (Gordon-Brown, 1975: 39, 212).

THE MECHANICS OF DISSEMINATION

The chronology, and indeed the whole process, by which the flat roof domestic structure came into general use throughout southern Africa is a difficult one to establish. Not only do we lack detailed architectural information of this form between 1820 and 1920 but its methods of dissemination are not clear and are often contradictory. It is generally assumed, for example, that once the colony began to spread inland during the early 1800s, these settlers found the flat roofed dwelling to be ideal for the more arid conditions of the interior and within a short time this form began to be identified with the domestic architecture of Dutch farmers in the Karoo. Yet we know that once white settlement broke beyond the bounds of the western Cape, its main thrust followed the well-watered coastal path to Algoa Bay and not the more arid road into the Karoo hinterland. The exploratory journeys made by Burchell in 1811 (1953: 1: 85) and by Campbell in 1815 and 1820 (1815 and 1822), proved crucial in this regard as they opened the way to the interior via Uitenhage and Graaff-Reinet and highlighted the dangers and travails facing travellers into the interior.

ILLUSTRATION 10. Founding of settlements in the Karoo, 1786-1920.

It is also recorded that of the sixty-eight major towns and villages in the larger Karoo region, only four were founded before 1840 and just over 60% of them were begun after 1861 (illustration 10). By this time the dissident Dutch farmers had already trekked northward and founded Winburg (1836), Potchefstroom and Pietermaritzburg (1838), Bloemfontein (1846) and Pretoria (1855) and had settled as far north as Schoemansdal by 1849. Thus although Holub was led to comment in 1873 that:

"In its general aspect, Fauresmith (founded in 1850) is very like the other towns in the Free State ... consisting of not more than eighty houses, ... clean white-washed residences, flat-roofed as elsewhere" (1881: 1: 43).

A pictorial survey of other early Voortrekker towns further north reveals that only a small number of their dwellings were flat roofed and that indeed, many such structures were of a temporary nature. A closer look at Bell's painting of Colesberg, for example, done in 1844 (Gordon-Brown, 1975: 39), reveals that the more substantial buildings of the town are all pitch roofed whilst it is mainly the humbler homes and the more recent arrivals, on the outskirts of the town, which carry flat roofs. Therefore, whatever else it may have been, the flat roof dwelling was not part of the political and ideological impedimenta carried by the Voortrekkers to their new homes, and its spread into the southern African hinterland was not dependent upon the urbanisation of the region.

ILLUSTRATION 11. The mission station at Bethany, OFS, visited by Backhouse in 1839, shows how the missionaries themselves used the kapsteilhuis as a temporary shelter before erecting more permanent flat roofed houses. This picture has been edited.
ILLUSTRATION 12. Jerusalem, or Afrikaner's Kraal in Namaqualand, visited by Backhouse in 1840. This picture has been edited.

Our knowledge of individual flat roofed farmhouses is even more scattered. Early examples built by missionaries were recorded by Backhouse in 1839 at the Bethany and Beersheba mission stations in the OFS (1844: 357-359, 421) (illustration 11) and at Afrikaner's Kraal in Namaqualand in 1840 (1844: 561) (illustration 12). Baines recorded flat-roofed dwellings at Colesberg in 1848 and at Bloemfontein in 1850 (1961: 1: 99-100). There he accepted the hospitality of a Mr Colley whose store he described as being “a fine commodious building of one storey only and, like most of the others, flat roofed.” (1964: 2: 35). The experience however did not prove a happy one. The town was overtaken by torrential rains leading Baines to report that:

"As the day advanced the rain increased and poured through the mud roof, laid only upon a foundation of small bamboo reeds, almost as readily as it did outside ... every corner of the house was wet ... the floor converted into mud and the only dry places under the bed and table.

"... people were seen in every direction attempting to stop the leaks in their roofs by throwing quick lime upon them; and one person, in despair of finding a remedy, was actually obliged to bring out his goods in the never ceasing rain, load them up in wagons, and go in search of drier quarters." (1964: 2: 36)

ILLUSTRATION 13. Boer farmhouse in the OFS, recorded by Holub in about 1890. This picture has been edited.
ILLUSTRATION 14. Private house in Potchefstroom, recorded by Wangemann, in about 1872. This picture has been edited.

Flat roofed structures were also recorded in the Transvaal by Holub at Christiana in about 1890 (1881) (illustration 13), by Wangemann at Potchefstroom in about 1875 (1871-75) (illustration 14), and by Merensky at Botshabelo mission station in about 1872 (1875).

It was also during this era that the flat roofed house became identified, for a brief period of time, with the Griqua of the northern Cape and western OFS. In 1859 Mackenzie remarked how they “have all chosen the Dutch colonists as their model in social life and manners, although most of their missionaries have been Englishmen." (1971: 61)

Remarkably, once this group migrated to Griqualand East in 1862, a region endowed with a better rainfall and richer in thatching grasses, they abandoned the flat roofed dwelling and built for themselves hipped and gabled cottages (Dower, 1902), very similar in appearance to those described by Sparrman and Barrow in the Cape previously.

The Griqua represent an important case study which demonstrates how the flat roofed dwelling, whose roof in the western Cape had historically been built in a pitched or ridged and gabled form, reverted back to this once similar climatic and environmental preconditions were re-established. It is not known whether its plan conformed with the Cape two-cell longhouse derivative, but photographs taken in about 1878 indicate that its outward form, being possessed of a door and two windows on the eaves facade, and a large fireplace in the gable wall, was similar to that of the Cape cottage.

The social and economic factors which served to generate the flat roofed house during the early years of the eighteenth century are therefore becoming increasingly evident. Barrow's account of "apartments, nearly destitute of furniture", Sparrman's tale of houses of "but two rooms, somewhat more than two yards in breadth" and Baines' picture of Colesberg are all indicative of one fact: at best the rural flat roofed home was a humble structure. Some might even say that it was a poor man's residence, or at the very least, the temporary abode of graziers, occupied but a few months of the year. Even the buildings erected in new settlements such as Philippolis, laid out as a town in 1862, were not exempt from this definition. Holub, who visited it ten years later described its aspect as being:

"most melancholy ... Equally dreary were the flat roofed houses, about sixty in number, and nearly all quite unenclosed, that constituted the town; ... the majority of the houses being unoccupied, scarcely a living being was to be seen." (1881: 1: 39)

Quite clearly Philippolis, which was begun as a station of the London Missionary Society in 1821, owed its existence by 1872 to the Dutch practice of nachtmaal, a religious ceremony which brought together the farming community of an area four times a year for the purpose of breaking bread, celebrating weddings and generally reinforcing their sense of identity and bonds of common purpose. It was also important for it created opportunities for general socialising, for teenagers to go courting and for families to purchase their staple needs for the following three months. It was quite common for farmers to build themselves town residences called tuishuise to serve their housing needs during nachtmaal and these were normally covered with a flat roof. This would certainly explain why Philippolis was mostly unoccupied during Holub's visit. Other towns in the Karoo also owed their existence to the spiritual needs of the farming community. Amongst them were Jansenville and Aberdeen in the eastern Cape, founded in 1854 and 1856 respectively. In both instances the congregation was first established and served by a travelling minister working from a temporary church structure. Once enough money was collected to build a formal church and a substantial pastorie, or manse, a minister was then brought to the parish on a full time basis (Herholdt, 1986 and 1989).

ILLUSTRATION 15. “A typical Boer farm of the poorer type", Watson, in about 1902.

The humble nature of the rural flat roofed house was further emphasised by JH Watson who, reporting during the South African War of 1899-1902, illustrated such a structure, either in the Orange Free State or, more probably, in the Transvaal (illustration 15). He told that:

"This is a typical Boer farm of the poorer type; the wall is of brick, the floor of mud, and the roof which has only a slight slope to let the rain run off, is of galvanised iron, fastened on in the primitive method by putting heavy stones on the top ... Often one of the rooms has been used as a barn and contains mealies and wheat in sacks." (Undated)

This description contains recognisable elements of both the Cape cottage, recorded by Barrow a century earlier, and the parapet dwelling, built by peasant farmers of the Highveld and documented by Frescura in the 1980s (1981). More importantly however, these very same buildings were also illustrated by the Carnegie Commission in 1932 when describing the problem of poor whites in South Africa (1932). Thus it can be said that the rural flat roofed house of southern Africa has always been associated with the poorer, temporary, indigent and newly urbanised elements of rural society. This image has been compounded in more recent times by its increasing identification with urban informal settlements as well as the landless and dispossessed black farmers of the highveld region.

This then gives rise to a dichotomy for, on the one hand, the urban flat roof aesthetic was generally associated with statements of style, fashion and high design whilst in rural areas the same form belonged to a more humble tradition of folk architecture.

ILLUSTRATION 16. View of Baird Street, Uitenhage towards the DR Church, in about 1845. The unknown artist has shown a double storey building with its ridged and hipped roof masked on the street facade by a broad parapet. This picture has been edited.

This picture is complicated further by the fact that a few, and admittedly isolated cases, may be found in the Cape of large and evidently prosperous farmhouses built in the flat roofed manner (Dennis Radford, pers comm, April 1989). Also, further north, Bloemfontein's first Presidency, built in about 1860, and Kruger's farmhouse at Boekenhoutfontein near Rustenburg, built in about 1873, both followed a form reminiscent of the flat-roofed town houses of Cape Town. The design of both structures was obviously influenced by current concepts of style and fashion, something which other buildings elsewhere in the country also sought to emulate to the extent of masking their pitched roofs by means of ornamented parapets (Gordon-Brown, 1975: 41) (illustration 16). It must also be borne in mind that although the tuishuise clustered about the village church for nachtmaal may have been of a temporary nature, they were not necessarily the product of indigent farmers.

Thus, by the late nineteenth century, the relationship existing between the flat roofed town house and the rural cottage may outwardly appear to have become somewhat remote and academic. This, however, is not correct. Indeed it was the humble rural dwelling which was to provide the more sophisticated urban structure with a sense of cultural and architectural continuity which permitted it to develop and achieve urban respectability. Rapoport discussed this viewpoint in 1969 when he postulated that a vernacular or folk tradition to be the common matrix, the clay which makes high design architecture possible, and that the two combine to form a total spatial and hierarchic system (1969: 1-8). Viewed in these terms therefore, the adapted longhouse or Cape cottage may also be seen to form the basis for both the townhouse architecture in the western Cape as well as the more modest cottage dwellings of the Cape interior, regardless of their respective aesthetic and cultural status.

The chronology, and indeed the whole process, by which the flat roof domestic form was disseminated throughout the southern African sub-continent is a difficult one to establish with any degree of accuracy, and little work appears to have been done in this field by previous architectural historians. Most publications make reference to assumed British influences and Renaissance periods but fail to examine the mechanics of such movements in any great detail. Immelman, for example, states that:

"Gedurende die jare 1835-1837 het 'n groot aantal van die ervare boere van gesiene families die Kaap verlaat en noordwaarts getrek ... Van 1860 af was hierdie boere nie meer trekkers nie en het hulle permanent gevestig geraak ... Binnelandse argitektuur weerspieel die veranderde omstandighede" (Trans: Between 1835 and 1837 a large number of skilled farmers from prominent families left the Cape and migrated northwards … From 1860 these farmers had ceased their travels and built permanent homes … Inland architecture reflected these changing circumstances.) (1977/1978)

Others were content to remain silent on the subject. More recently however Radford has postulated the existence of cultural heartlands, core areas within which certain architectural traditions are strongest (1989). Using white immigrant architecture as an example, he points out that since the eighteenth century three such cores have arisen in this country, the Western Cape, the Eastern Province and Natal, and that many of our current vernacular and popular building forms, textures and technologies may be traced back to such points of diffusion. Although this theory still requires considerable fleshing out, its basic hypothesis goes a long way towards explaining the processes of geographical dissemination which may take place outside such architectural cores.

If we were to apply Radford's thesis the migration of the flat roofed house may be perceived to have taken place in a number of distinct if partly overlapping stages.

ILLUSTRATION 17. Flat roofed dwelling on Robben Island, probably dating to the 1830s, recorded by Wangemann c1872. This picture has been edited.
ILLUSTRATION 18. Flat roofed fishermen's cottages, Hermanus, depicted by JH Pierneef, 1931. The picture has been edited.

The first is well documented. It began in Cape Town during the early 1700s and had more or less run its course by the 1830s. Its spread was limited to the main urban areas of the western Cape and probably did not extend beyond other coastal towns such as Simonstown and Hermanus (illustrations 17 and 18). Its outcome was a dwelling which was essentially urban and whose form ultimately owed much to the imponderables of aesthetics, style and fashion (illustration 19a).

ILLUSTRATION 19a. Diffusion of the flat roofed house: Stage One, the western Cape.
ILLUSTRATION 19b. Diffusion of the flat roofed house: Stage Two, the western Cape and Klein Karoo.

The second stage began in the 1770s and thus overlapped chronologically with events in Stage One. It involved the dissemination of flat roofed dwellings amongst the migrant graziers of the Cape interior, and although poorly documented, their spread into the larger Karoo appears to have been somewhat limited. It is probable that the major thrust of this movement lay eastward along the rain shadow provided by the Langeberg and Outeniqua mountains, a semi-arid area which has since become more popularly known as the Klein Karoo (illustration 19b). This is supported by the fact that by the mid-1830s, the major settlements in this region were located at Calitzdorp (founded in 1821), Knysna (1817), Port Elizabeth (1820), Uitenhage (1806), Somerset East (1825), Graaff-Reinet (1786), Cradock (1816) and Colesberg (1830), all of them being located on the fringes of the Karoo, whilst only one, Beaufort West (1820), was located in the Karoo itself.

The processes of dissemination during this stage are also not easy to determine. Certainly during this time Cape Town was the major centre of economic, cultural and legislative activity and, for one reason or another, most inhabitants from the interior would have visited the town at some stage of their lives and thus come into contact with its growing number of flat roofed residences. However, considering the geographical spread of this domestic form it is probable that the flat roof was only used as an alternative to the more conventional pitched roof of the Cape cottage. The outcome was a dwelling which was essentially humble and its construction was guided primarily by pragmatic considerations of climate and an availability of materials. It is likely that this movement continued well into the nineteenth century and was consolidated by the introduction of corrugated iron sheeting into southern Africa from the mid-1850s onwards.

During stage three the focus of these developments swung to the eastern Cape where, for a number of economic reasons, Port Elizabeth began to overtake and, by the 1860s, eventually supplant Cape Town as southern Africa's premier port. Following the arrival of a large body of English immigrants from 1820 onwards, the movement of whites into the interior was given added impetus by the departure of dissident Dutch parties after 1836, the increased military garrisoning of the Border region, and the discovery of diamonds in the north-western Cape in the 1860s. As a result we find that the flat roofed tradition leapfrogged over the southern Cape where, by that time, it had had little chance of consolidating its position, and entered the eastern Cape.

Admittedly its spread there was slow and even minimal. The imported cottage architecture of the newly arrived English adapted easily to local conditions in the eastern Cape and proved singularly unwilling to convert to a local flat roofed technology. From the onset the early residents of new settlements such as Grahamstown, Port Elizabeth, Uitenhage and Fort Beaufort all showed a distinct preference for hipped roofed houses. The Dutch themselves had also not built such farmhouses in this region to any great extent, even as far north as the Sneeuwberge, immediately beyond Graaff-Reinet, where flat roofed structures are recorded to have been built only after the 1850s (Whitlock, 1989). Before this time much of this development appears to have been focused upon Graaff-Reinet where Thompson, who visited the town in 1823, described it as consisting of approximately three hundred neat and commodious brick buildings, some of which had flat roofs with simple moulded cornices (Minnaar, 1987). On the other hand it is recorded that some English immigrants who took over farms from the newly departed Dutch found flat roofed homesteads already standing there and, in many cases, restored and adopted them to their own use (Lewcock, 1963).

We need not look far for reasons why a flat roofed architecture failed to gain much ground in the eastern Cape. A notable example may be found in Grahamstown where the new jailhouse, completed with a flat roof in 1825, had to be re-roofed within three years with a more conventional pitched structure. Although this event was followed by protracted arguments about shoddy workmanship and poor performance by unskilled contractors, there is no doubt that the roof's major failure lay in its inability to exclude moisture. The yellowwood timbers had been set in clay walls and the ensuing damp had caused the wood to rot (Randles, 1987), a problem also experienced previously in Cape Town (Lewcock, 1963: 385).

Although the British administration thereafter still maintained a flat roofed aesthetic for many of its official buildings in the eastern Cape, often by the use of parapets masking a pitched roof beyond, this was probably done as a deliberate statement of political identity and only became structurally feasible after the introduction of corrugated iron into the region during the early 1860s.

ILLUSTRATION 20. Diffusion of the flat roofed house: Stage Three and Stage Four, the southern African interior.

Thus although the eastern Cape may, in Radford's terms, be seen to have acted as a cultural heartland, it did not foster a flat roofed architectural tradition within its own boundaries, such as Cape Town had done. Instead its ability to disseminate these domestic forms may be attributed to the economic and political role played by Port Elizabeth in acting as a channel for the movement of goods, resources and people into the southern African hinterland. The fact that settlers originating from this area later built flat roofed structures as their new homes was due to the environmental conditions they found there, not to any cultural preconditioning derived from their origins (illustration 20).

The fourth stage in this process began in the early 1860s and probably lasted well into the twentieth century, perhaps as late as the 1930s and 1940s. It owed its character to the introduction of corrugated iron into southern Africa. It is true that various forms of flat metal sheeting, such as lead, zinc and copper had been available in the Cape for some time, but such roofs were generally considered to be too light and labour-intensive and never gained widespread usage. Corrugated iron sheeting, on the other hand, was durable and easily portable and the first iron buildings appeared in Cape Town as early as 1847 (Lewcock, 1963: 347), whilst the missionary Livingstone is reported to have taken a corrugated iron house to the Zambesi as early as 1858 (Lord and Baines, 1976). An advertisement for a similar structure appeared in the Graaff-Reinet Herald of August 1858 to the effect that:

"PER LEONIDAS THE UNDERSIGNED ARE RECEIVING AN IRON HOUSE  of four rooms, complete and ready for immediate erection, it  is provided with flooring, and with wooden lining expressly  arranged for ventilation in hot climates; and is of the same  construction as those which have been so extensively used in Australia and the West Indies.
H. & W. BOLUS
"

ILLUSTRATION 21. Kimberley, as viewed by Holub in 1872. It will be noted that although corrugated iron roofs predominate, the majority are pitched. This picture has been edited.

The technical and economic performance of corrugated iron rapidly proved to be vastly superior to that of other waterproofing methods tried previously at the Cape, and it soon began to replace the flat roof technology of earlier times. Although at first its availability was limited to the larger urban centres, by the 1870s and 1880s, following the opening of the Kimberley diamond mines and the Transvaal gold fields (illustration 21), it rapidly spread into other areas where it found widespread usage in both the domestic and industrial sectors. Inevitably it was only a matter of time before it became synonymous with flat roofed architecture.

ILLUSTRATION 22. Settlement of the Karoo, 1786-1860.

It was during this time that most of the major towns and villages of the Karoo were established. A survey of their founding dates reveals that only four major settlements could be found there prior to 1840: Graaff-Reinet (1786), Beaufort West (1820), Calitzdorp (1821) and Colesberg (1830). Thereafter expansion into this region quickened and over the next twenty years up to 1860, some seventeen more villages were begun (illustration 22). Significantly most of this development occurred in the southern part of the Karoo where a wool farming industry was beginning to flourish.

The decade between 1861 and 1870 was a period of economic depression marked by droughts and the failure of a number of local banking institutions. This had a marked effect upon the development of the Karoo with only six settlements being founded there during this time.

ILLUSTRATION 23. Settlement of the Karoo, 1861-1920.

The discovery of diamonds in the northern Cape in 1867 marked the start of an era of strong economic development for the southern African interior. This and the spread of a railway infrastructure throughout the Cape during the 1880s and 1890s effectively opened up the northern Karoo for settlement and over the next thirty years, up to 1900, 36 villages were founded (illustration 23). This was further assisted by two separate feather farming booms, one in the Albany district in the 1880s and the other, more importantly for the Karoo, in Oudtshoorn during the 1900s. Thereafter the South African War of 1899-1902 brings the settlement of the Karoo to a virtual halt and between 1900 and 1920 the establishment of only five new villages was recorded. The one centre which benefited most from this economic growth was Port Elizabeth which, during the 1870s, became the centre of wool, feather and, for a time, diamond marketing (Wilson and Thompson, 1975).

Thus the predominant architectural character of the Karoo was not established whilst this area was under the administration of either a Dutch or even an early English colonial government; nor was it established to any great extent before the introduction of corrugated iron in the region during the 1860s and 1870s; it was established between 1870 and 1900, at a time when this region was under the predominant economic influence of the eastern Cape, a region where the English predominated. Signs of a growing English presence in the Karoo were becoming evident as early as 25 August 1852 when the Graaff-Reinet Herald published the following editorial:

"It may perhaps be a matter of surprise that an English paper should spring up in a district hitherto considered almost exclusively Dutch. It will appear perfectly natural however, when it is known that many frontier English have been compelled by the (Xhosa) wars to settle in this district, and that many educated Dutchmen are so well versed in English as to prefer it to their own language." (Smith, 1974)

Indeed, as late as 1875, 20% of the residents of the Albany District were British born (Radford, 1989).

This is supported by other evidence. Although no hard demographic data appears to exist regarding the language makeup of the Cape during this time, an analysis of the colony's civil lists reveals that, by 1860, 57% of all postal employees came from English backgrounds. By 1870 this figure had risen to 71% and by 1879 to 73% (Cape of Good Hope Blue Books, 1838-1882).

ILLUSTRATION 24. Divisional distribution of English and Afrikaans civil servants employed by the Cape Postal Administration, 1870.
ILLUSTRATION 25. Divisional distribution of English and Afrikaans civil servants employed by the Cape Postal Administration, 1879.

A division by division breakdown of these figures plotted out on a map also makes for interesting reading (illustration 24). In 1870 the suburban area of the western Cape showed a strong English bias. This was separated from the rest of the country by a belt of strong Dutch presence stretching in an arc from Calvinia in the north to Stellenbosch and Caledon in the south to Oudtshoorn in the east. Beyond that however the southern and eastern Cape were almost overwhelmingly English, the southern Karoo showed a good mixture of English and Dutch, with the English being in slight predominance, and the areas north of that showing a strong English presence. A similar map for 1879 shows essentially the same patterns with areas of English and Dutch mixes spreading into both Namaqualand, Paarl, Caledon and Bredasdorp and a growing English presence in the northern and central Karoo (illustration 25). Although it is probable that these figures reflect a bias in civil service employment, they do nonetheless reveal that the Cape middle class, which would have provided many of these bureaucrats, and who, presumably, would also have built a large proportion of the new urban homes of that time, was English to a considerable extent.

It becomes clear therefore that if the flat roofed house should, in any way be considered to be typical of the domestic architecture of the Karoo, then its construction in urban areas was as much the result of English home building as it was of the Dutch. Indeed once the brakdak was superseded by corrugated iron after the 1860s, there is every reason to believe that the flat roofed home was simply added to the range of domestic structures available to local speculative builders (Radford, 1988). Many of these, then, would not have been built as a reflection of local cultural preferences but merely because, by that stage, they represented a style of construction linked to the fashionable town house of the Cape, a structure which, it was seen, enjoyed a large degree of urban respectability.

ILLUSTRATION 26. Voortrekker house recorded by JH Pierneef at Naboomspruit in the northern Transvaal, 1931. This picture has been edited.

The spread of the flat roofed house further north and beyond the Karoo during this period is not easy to assess, largely because of a lack of reliable statistical data and eyewitness accounts. It is known, for example, that such dwellings were built in rural areas, both in the Transvaal and the Orange Free State, by missionaries and farmers alike, but their extent is difficult to establish (illustration 26). Examples were recorded between 1870 and 1883 on mission stations as far afield as Botshabelo near Middelburg, Bethanie near Rustenburg, Makchabeng near Louis Trichardt and Portjesdam near Fauresmith (Wangemann, 1871-75). At the same time however, many other missions in this same region were also built with pitched roofs. Similarly, flat roofed structures could be found in Potchefstroom in the 1870s and Pretoria in 1880 (Gordon-Brown, 1975: 112) but these were neither numerous nor representative of a wider urban norm.

On the other hand it is recorded that from about 1870 onwards the social and economic fabric of the rural hinterland was disrupted by a series of land wars and a period of prolonged drought. This was followed by a succession of plagues of locusts; the rinderpest epidemic of 1897; and the South African War of 1899-1902, during which time many Dutch homesteads and their crops were fired. As a result we find that by the 1920s the use of flat-roofed dwellings had become associated with indigent white farmers (Carnegie Commission, 1932). To this day these structures are still referred to by local whites as boer-maak-'n-plan, a tribute to the farmer's ability to improvise under conditions of economic duress.

ILLUSTRATION 27. Diffusion of the flat-roofed house: Stage Five, dissemination into Black rural and urban communities.
ILLUSTRATION 28. Informal dwelling at Tyoksville, KwaNobuhle, in Uitenhage, 1987. Drawing by Craig Billson.

The fifth and final stage involved the transmission of the flat roofed house and its technology to local indigenous builders. Although some of the preconditions for this cross-cultural borrowing were being created by missionary trade schools as early as the 1850s (Frescura, 1985: 228-231), the construction of such dwellings does not appear to have begun much before 1900 and did not take place in any significant amount before the 1940s (illustration 27). Even then it was a predominantly rural manifestation associated with the highveld region and with obvious links to the homesteads of Dutch, or, by now, Afrikaner, farmers. Since that time however southern Africa has witnessed the increasing urbanisation of its rural population, both white and black. In most cases this has not been met by an equal supply of housing, most particularly in black areas. As a result large numbers of homeless families have turned to the flat roofed structure as a quick and cost-effective means of providing themselves with a home. Today the flat roofed house has become identified with informal housing throughout the land and although in many instances its form has departed from that of the Cape cottage, the same principles of pragmatic shelter still appertain to it (illustration 28).

Although the cultural heartland, if it can still be described as such, of the flat roofed house has now technically moved to the highveld, most specifically the Witwatersrand and the northern Orange Free State, in reality this dwelling has since become forma franca throughout southern Africa and any major industrial centre in this region may now be considered to act as one of its focal points.

THE BEGINNINGS OF A "BAFOKONA" TRADITION

The word bafokona is a recent addition to the vocabulary of local builders and refers to flat roofed, square plan structure, a dwelling which in other parts of the country has also become known as an iflat or an iplata, the words having been derived from obvious English and Afrikaans roots. Its incorporation into the larger body of southern African indigenous architecture appears to have taken place in two distinct stages. The first began in the late eighteenth century and was part of a process initiated by missionaries who sought to alter the migrant pastoralist nature of Khoikhoi economy and convert and change them to a sedentary agrarian society. Evidence of this was recorded by Backhouse in 1840 who noted that Khoikhoi dwellings "were usually of rushes; they next built hartebeest houses of better quality; and many had superseded these by neat, comfortable cottages, ..." (1844: 619-620)

In view of the fact that when the Khoikhoi built such cottages, either hipped and gabled or flat roofed, they were usually associated with mission stations, their construction should therefore be considered in the context of missionary endeavours in the region.

The second stage is somewhat more complex and despite missionary claims to the contrary, was not achieved as a result of religious proselytising. On the contrary it was part of a much slower and long-term chain of events involving material, technological, social and economic transition in the southern African hinterland of which missionaries were but a small part. Unlike the Khoikhoi of the Cape, other indigenous inhabitants further north and east had a well developed sedentary agricultural economy, and were thus better equipped to meet the pressures which white immigration placed upon their lands. These events should be seen as part of a slow and evolutionary process, which did not seek the radical supplant of an existing architectural tradition; it did not directly try to undermine existing social structures; it cannot even be said to have been identified with any conscious movement or motivated by a strong ideology. It began to manifest itself from the 1900s onwards as part of a wider historical development which derived deeply from the traditional roots of local society, both architectural and social. This means that although the rural builder of today does construct square and rectangular plan dwellings, as well as many other types, he considers these to be part of his wider repertoire of domestic forms, having been derived locally and as the result of internal pressures. The transition from circular to square plan involved various stages of development which, in some areas, took many generations to achieve (Frescura, 1981). Most important however is the fact that such changes were brought about without there having been any consequent devaluation of the social and economic structures and values inherent in vernacular architecture.

The introduction of a flat roofed dwelling into the southern African vernacular tradition may be seen to have taken place through a variety of agencies:

  • Missionary trade schools, which taught the technology required before such changes could be introduced. It also created a body of skilled craftsmen who found work in urban areas where they came into further contact with new building methods. This stands in sharp contrast to their proselytising efforts where their achievements have been brought into question by more recent research (Williams, 1959).
  • White farmers who provided their workers with practical raining in the use of new roofing and wall building technologies. The presence of such words as isitene (steen or a brick), bafokona (a four cornered house) and idekspan (a thatching leggett) all stand testimony to this influence.
  • The Land Acts of 1913 and 1936 which restricted 88% of the southern African population to 13% of the land's area (Mbeki, 1964). In time this led to overpopulation, overgrazing and a concomitant reduction in such resources as thatching grass. This, in its turn has led many a rural builder to turn to more permanent, if more expensive, roofing methods. The flat roofed house has often proved the simplest and most cost-effective solution to this problem.
  • The implementation of a Hut Tax from the 1850s onwards which forced many rural farmers into the migrant labour system and brought them into daily contact with urban building technologies and materials, most particularly corrugated iron sheeting.
  • The introduction by traders of material artifacts, such as furniture, which, being based on a 90° geometry, made the furnishing of circular rooms highly inefficient and thus accelerated the move towards square plan dwellings.
  • The availability of stoves and chimney flues which replaced traditional cooking methods and facilitated the introduction of corrugated iron sheeting as a roofing material.
THE EMERGENCE OF A HIGHVELD TRADITION

The earliest pictorial record found to date of the rural use of corrugated iron on black residential structures is in the form of a picture postcard dating back to circa 1905. It shows the "(Black) location, Bethlehem, O.R.C." where there appears to have been a predominance of gable-ended cottages of a more conventional construction (Post card in Author's collection). Scattered among them however were a number of flat roofed structures which, significantly for that time, had their sheeting held down with a variety of stones or other objects. This is a practice prompted by socio-political considerations which survives to the present day and is indicative of a community with few political rights and no land tenure. Thus a family which may be forced to relocate five or six times in a man's lifetime, might not be too eager to puncture their roof sheeting and destroy the integrity of a valuable money-intensive material, knowing that they would be required to use it, and reuse it, at a number of different sites.

The adoption of corrugated iron into black rural architecture was gradual. Very little of it is evident in the photography of Duggan-Cronin in the 1920s and the 1930s (Duggan-Cronin Photographic Archives, Kimberley) and it only emerges as a force in indigenous building from about the 1940s onwards. Even at that stage its use predominated on the central highveld region, becoming identified with the housing of farm workers employed on white farmsteads.

This area is located in the northern and north-eastern Orange Free State and the southern Transvaal, the historical home of Sotho and, to a lesser degree, Tswana groups. During the Difaqane the local population was decimated by both famine and warfare and many survivors sought the refuge of the Lesotho mountains. There, under the leadership of Moshweshwe, they coalesced into the Kingdom of Lesotho, which survives to the present day. Their former home is now part of the Republic of South Africa where it is under the control of a white and predominantly Afrikaner farming community.

ILLUSTRATION 29. Flat-roofed dwellings recorded on the highveld between 1979 and 1981.
ILLUSTRATION 30. Southern Sotho homestead, about 1978. Drawing by Evan Davies after measured drawings done by architectural students, University of the Witwatersrand.

Some Sotho have since returned to these areas as migrant workers having no legal land ownership rights. They live with their families on white farms where they are given a small plot of land to plant some crops and build their homes. Their residences are almost invariably built in the form of parapet or flat roofed dwellings, having a rectangular plan and a flat, corrugated iron roof sloping to the rear (illustration 29). Kitchens are either located in a separate structure and thatched in order to allow the smoke from the cooking fire to percolate through the roof or, where they have been equipped with a cast-iron wood-burning stove and a smoke flue, they are incorporated into the main dwelling itself. Such units normally consist of a single room which is used as a general eating and socialising space with a cooking area situated to its right hand side as one stands looking out of the doorway. A crockery shelf will often be built into the gable wall alongside the stove. Additional rooms may sometimes be built onto the rear of the dwelling by extending the roof structure backwards and away from the front parapet wall. It is more usual however to extend the homestead by merely building additional units facing onto a common courtyard (illustration 30).

ILLUSTRATION 31. Analysis of prototypical flat roofed highveld dwelling.

Door and window openings are located on the front elevation which normally faces to the north or north-east. As in their Cape and Karoo predecessors the door is usually placed slightly off-centre, with a small, square window on either side implying an internal division of space into two rooms which is, however, rarely achieved (illustration 31). As such then it may be considered to be a continuation of the Cape cottage form but it also tends to follow the local indigenous tradition of single cell units within a larger homestead structure.

ILLUSTRATION 32. Flat roofed lean-to informal dwelling in the Winterveld, some 50 km north of Pretoria, c1981. Drawing by architectural students, University of the Witwatersrand.
ILLUSTRATION 33. Informal dwelling, Walmer Township in Port Elizabeth, 1987. Drawing by Craig Billson.

Since the 1940s communities of landless people throughout South Africa have adopted almost universally the lean-to principle of the flat roofed dwelling, having discovered it to provide them with a cheap and efficient method of achieving incremental housing. Current research conducted in such informal settlements as Crossroads near Cape Town, Winterveld north of Pretoria, Malukasi in Durban, and Thaba 'Nchu (Frescura, 1982) has indicated that, whatever the technology used, dwellings tended to begin as a simple, flat roofed single cell unit (illustration 32). Thereafter the structure was allowed to grow organically, dependant upon the amount of space available and the residents' levels of affordability. Thus although its form may ultimately differ quite substantially from its forerunners, the basic informal urban house owes much to its historical roots (illustration 33).

ILLUSTRATION 34. Southern Sotho wall patterns originating from the districts of Balfour, Villiers and Greylingstad respectively.

In more recent times, rural housewives on the highveld have also taken to decorating their dwelling facades, usually following the same basic pattern (illustration 31). This is subject to some simple guidelines: the parapet at the top and a low splash board at the base are expressed as long horizontal bands. The width of the splashboard however is not constant and may vary from a thin ribbon of about 100mm to a high band reaching up to window sill height. The two corners on either side are rendered as vertical elements; a broad surround is created about both door and windows and is often allowed to run into the parapet above; the parapet band is often decorated and its top profile sculpted in order to create small pediments and half-pediments over the doorway and the corners respectively. Within this basic framework then the Sotho artist can and often does exercise considerable choice as to colour, graphic pattern and texture. Most of these however are related to plant or geometrical themes (illustration 34).

The basic elements of this decorative style may be perceived to have strong links with the facade renderings of colonial, late nineteenth century, domestic architecture. In the interior too, the women often recreate, in clay, crockery display shelves which are a stylised rendering of English middle class, turn-of-the-century kitchen furniture, complete down to the presence of a cut-out paper doily trim. The frill of the doily is, at times, reflected externally in the rendering on the parapet band, thus emphasising the essentially female nature of dwelling decoration (illustration 29).

In spite of the outwardly aesthetic nature of this decoration, parallel developments among the Xhosa and Mfengu of the southern Transkei and the South Ndebele of the Transvaal indicate that the activity of women's painting is a reflection of their fertility and status within the community. When it is also considered that the polychromatic painting of rural facade walls began during the late 1930s and 1940s and coincides with the rise of white nationalism in South Africa, then this practice may be interpreted in terms of a grass root movement initiated by rural women as a protest to their group's lack of land tenure and political rights (Frescura, 1988).

CONCLUSIONS

The transmigration of the European medieval longhouse to the Cape of Good Hope during the seventeenth century and its subsequent dissemination into the southern African interior raises a number of interesting questions. What, for example, is the exact role played by climate and technical performance in predetermining a people’s choice of their own built habitat; at what stage do such choices cease to be guided by pragmatic factors and are overtaken by considerations of aesthetics, style and status; and finally at what point does architecture cease to be a matter of individual choice and become associated with the cultural mores and values of a larger group?

In the case of the flat roofed house the first of these questions may be dealt with relatively quickly. Climate and technical performance are powerful inhibitors of dwelling form in that they prescribe what may not be built rather than what may. Thus we find in seventeenth century Cape Town that a combination of fire and driving winds proved a forceful argument against the use of thatched roofs. The earth-covered flat roofed technology which replaced it was not entirely water-proof, needed constant maintenance and thus remained a viable alternative only for as long as there existed the threat of conflagration. Once the colony began to expand its settlement eastward along the southern Cape coast then the older and more conventional thatched technology was re-established. Similarly once the Griqua moved from the arid northern Cape, where many of them had built homes in the manner of Dutch graziers, to the water rich slopes of Griqualand East, they too abandoned their flat roofs in favour of thatch. Further inland however, where the Karoo's arid climate made the use of earth-covered flat roofs more feasible, the choice fell between two alternatives: an availability of thatching resources, limited at the best of times, and the relative ease of building flat roofs. Ultimately all of these debates were overtaken by the introduction of corrugated iron which at once cancelled both variables of climate and technical performance but also added a factor of cost and affordability to the housing equation.

In the second instance it was found that once a community ceased to be concerned with basic shelter and became aware of factors such as political identity and symbolism, then considerations of aesthetics, style and fashion began to overrule other, more pragmatic alternatives. This does not mean to say the people concerned became dominated by a blind desire to follow style and fashion, merely that, under certain conditions, their need to fulfill their psychological aspirations should be seen to rate as highly on a scale of human needs as the more widely accepted factors of safety, comfort and affordability. Thus when the burghers of the Transvaal and Orange Free State republics built their first flat roofed civic buildings, they were not acknowledging their own humble farmhouse architecture, they were aligning themselves with the aesthetic norms and values of urban centres further south. Because of the added costs involved in carting expensive imported building materials overland to Bloemfontein and Pretoria they were demonstrating their surpluses of material resources; by abiding by current architectural styles they were exhibiting their builder's constructional skills and their architect's aesthetic good taste. They were therefore making a political statement to the effect that their own public works were equal in every way to those they had left behind with the British of the Cape. Ironically, in doing so, they were also tacitly acknowledging Cape Town's architecture to be the country's moderator of fine buildings.

The idea that a particular building form may, in some way, become identified with the value structure of a people, a time or a place is associated, in many ways, with their perceptions of both style and stereotype (Frescura, 1985: 272-281). There is no doubt that human society has a strong propensity for creating stereotypes, whether they be environmental, material, social or, in our instance, aesthetic. They are mental tools which allow us to simplify our thinking and inform us, in general terms, as to the nature of the object, person or location under discussion. Although patently fallible when applied to many individual cases, they do nonetheless tend to describe a general norm and thus carry a degree of validity. Consider, for example, the phrase "the Mongolian Yurt". This stereotype may be taken to imply that yurts are the preferred habitat of Mongols. This is patently incorrect for whilst it is true that migrant Mongol pastoralists developed the concept of a yurt and, for a time, many Mongols lived in such structures, not all Mongols did so nor, for that matter, not all yurt-like structures were inhabited by Mongols. Similarly architectural historians may describe a church to have been built in a Romanesque style. Whilst it is true that the general feeling of the building's aesthetics may indeed be Romanesque, (whatever that might imply!), the fact that its main nave was constructed during early Christian times, that its campanile was rebuilt by a Renaissance architect and that many of its interior chapels were redecorated by Rococo artists surely makes a mockery of such generalisations.

The danger of such classifications therefore lies in two distinct areas: the assumption that the general norm finds universal application in individual cases; and that such norms are timeless and unchanging.

Thus, when architectural historians talk of a style they are in fact taking the buildings of an era, a group or a region and through them, creating stereotypes. Whether or not the significance of such stereotypes can be extended to a point where they become synonymous with culture needs to be severely questioned. The case of the South African flat roofed dwelling is a particular case in point. Not only does it owe its roots to the longhouse, originating from medieval Europe, but its links to a wider Cape cottage tradition and its subsequent evolution into a Cape townhouse architecture stands in sharp contrast to its present status as a Karoo domestic stereotype. Statements by such writers as Theron to the effect that:

"Tussen die jare 1840 tot 1860 raak die trekboere in die Karoo langsamerhand gevestigd en dorpies word oral gestig as godsdienstige en sakesentra in die uitgestrekte distrikte. ... die dorpsbeeld is een van beskeie platdakgeboutjies, meestal wit gekalk of van Karoo-leiklip gebou."  (Trans: Between 1840 and 1860 the migrant Dutch herders in the Karoo gradually gained in affluence, and small villages began to be established in the outlying districts, serving as religious and business centers … they consisted of modest flat-roofed buildings, usually whitewashed or built in Karoo slate.) (1973: 4-12)

are not only chronologically incorrect but their attempts to link the development of Karoo settlements endowed with a prototypical flat roofed architecture to a trekboer subculture is part of the larger myth which has been created about this architecture.

When we consider further that the flat roofed house has, at various times of its transmigration across the southern African interior, been identified variously with the Malay, the Griqua, the Karoo trekboer, the Transvaal and Orange Free State voortrekker, the South Sotho and, more recently, the informal settlements in our urban centres, then it becomes obvious that this form belongs to a larger southern African identity free of sectarian labels. Thus when Theron also proclaims that "Die argitektoniese samehang van die Karoo-dorpsbeeld wortel in 'n eenvormige kultuur en in die gemeenskaplikheid van oprede en gedrag" (Trans: The architectural coherence of the Karoo village is rooted in a uniform culture and in common patterns of language and behaviour) (1973: 4-12), he is writing in the context of a wider white political debate which, since 1936, has sought to gather all Afrikaans-speakers into a one homogenous group. The schism between liberal town-dweller and conservative farmer dates back to the late seventeenth century and although today it might still find a measure of social and political validity, attempts to justify it in terms of separate architectural patterns do not have a sound historical base.

Indeed efforts to differentiate between this domestic form and the larger body of Cape Dutch architecture can only be seen as an attempt to deny its origins and bring it closer to the architecture and political culture of a more conservative north. Given its evolving social identity, it would probably be more correct to describe the flat-roofed dwelling as one of the continuums linking a changing pattern of colonial settlement in the region, beginning with a transient Dutch presence up to the 1840s, changing to a more intensive pattern of English settlement until the 1920s, when it began to give way to a second wave of colonialization of the Karoo by Afrikaans-speakers, leading to its subsequent establishment in white political mythology  as an Afrikaner cultural heartland.

POSTSCRIPT

I wish to thank my colleagues Dennis Radford, Barrie Biermann, Albrecht Herholdt and Andre Appel for their constructive criticism and support during the formulation of these ideas. This paper was originally published in the South African Journal of Cultural and Art History, 1989, 3(4): 365-387, under the title “The South African Flat-Roofed Dwelling: A Study in Geographical Adaptation and Cultural Transmission”. It was reprinted in a slightly extended format as a teaching paper in the Department of Architecture, University of Port Elizabeth, under the title. “From Brakdak to Bafokona”. Colleagues wishing to use it as a reference should probably use the first publication as a source.

BIBLIOGRAPHY

BACKHOUSE, James. 1844. A Narrative of a Visit...etc. London: Hamilton, Adams.
BAINES, Thomas. 1961 and 1964. Journal of Residence in Africa, 1842-1853. Cape Town: Van Riebeeck Society.
BARROW, John. 1801 and 1804. An Account of Travels ... etc. London: Cadell and Davies.
BIERMANN, Barrie. 1952. A Contribution to the Study of the Origins of Colonial Architecture at the Cape. Unpublished PhD Thesis, University of Cape Town.
CAPE OF GOOD HOPE BLUE BOOKS. 1838-1882. Cape Town: WA Richards and Sons.
BURCHELL, William J. 1953. Travels in the Interior of Southern Africa. London: Batchworth Press.
CAMPBELL, John. 1815. Travels in South Africa. London: Black and Parry.
1822. Travels in South Africa ... Being the Narrative of a Second Journey. London: Francis Westley.
CARNEGIE COMMISSION. 1932. The Poor White Problem in South Africa. Stellenbosch: Pro Ecclesia-Drukkery.
CHIPKIN, Clive. 185. Beyond the Cape. Architecture SA, March/April 1985.
DOWER, William. 1902. The Early Annals of Kokstad and Griqualand East. Port Elizabeth: Jas Kemsley and Co.
DUGGAN-CRONIN, AM. Photographic Archives in the holding of the Alexander McGregor Memorial Museum, Kimberley.
FLOYD, Hugh. 1983. Inherited Characteristics of the Cape Vernacular. Architecture SA, Sep/Oct 1983.
FOWLER, HW, and FOWLER, FG. 1975. The Concise Oxford Dictionary. Oxford: Oxford University Press.
FRESCURA, Franco. 1981. Rural Shelter. Johannesburg: Ravan Press.
1982. A Review of South African Squatter Settlements in the Late 1970's. Proceedings, National Congress of the ISAA, Durban: Department of Architecture, University of Natal, 1982. 81-90.
1985. Major Developments in the Rural Indigenous Architecture of Southern Africa of the Post-Difaqane Period. Unpublished PhD Thesis, University of the Witwatersrand, Johannesburg, 1985. 272-281.
1988. Sign, Symbol and Resistance in African Architecture: A Southern African Rural Response to White Imperialism. Paper delivered to the African Studies Association 31st Annual Meeting on Continuity and Change in Africa, Northwestern University, Chicago, 28-31 October 1988.
GORDON-BROWN, Alfred. 1975. Pictorial Africana. Cape Town: Balkema.
GRAAFF-REINET HERALD. 1858. Advertisement, Vol 6, No 312, 7 August 1858.
HATTERSLEY, Alan F. 1969. An Illustrated Social History of South Africa. Cape Town: Balkema.
HERHOLDT, Albrecht. 1986. Die Argitektuur van Steytlerville. Port Elizabeth Eastern Cape Institute of South African Architects.
1989. Die Argitektuur van Aberdeen. Unpublished Report, Department of Architecture, University of Port Elizabeth,
HOLUB, Emil. 1881. Seven Years in South Africa. London: Sampson Low, Marston, Searle and Rivington.
IMMELMAN, Elbie. 1978. Karoo-Argitektuur. Lantern, Vol 27, No 2, December 1977 to January 1978.
KOLBE, Peter. 1731. The Present State of the Cape of Good Hope.  London: W Innys,
LE VAILLANT, Francis. 1790. Travels from the Cape of Good Hope. London: William Lane.
LEWCOCK, Ronald. 1963. Early Nineteenth Century Architecture in South Africa. Cape Town: Balkema.
LORD, WB, and BAINES, Thomas. 1976. Shifts and Expedients of Camp Life, Travel and Exploration. Second Edition. London: H Cox.
MACKENZIE, John. 1971. Ten Years North of the Orange River. London: Frank Cass and Co.
MBEKI, Govan. 1964. South Africa: The Peasant's Revolt. Harmondsworth: Penguin Books.
MERCER, Eric. 1975. English Vernacular Houses. Royal Commission on Historical Monuments, London.
MERENSKY, A. 1875. Beitrage zur Kenntnis Sud-Afrikas. Berlin.
MINNAAR, A. de V. 1987. Graaff-Reinet, 1786-1986. Pretoria: HSRC.
PICARD, Hymen W. Gentleman's Walk. Cape Town: Struik, 1968.
RADFORD, Dennis. 1988. Market Street, Cradock. Architecture SA, May-June 1988.
1989. The Origins and Diffusion of Vernacular Architecture in Southern Africa. Architecture SA, July-August 1989.
RAPOPORT, Amos. 1969. House Form and Culture. New Jersey: Prentice-Hall.
SMITH, KW. 1974. From Frontier to Midlands - A History of the Graaff-Reinet District, 1786-1910. Grahamstown: Rhodes University.
SPARRMAN, Anders. 1975. A Voyage Towards the Cape of Good Hope ...etc. Cape Town: Van Riebeeck Society.
THERON, Danie. 1973. Ons Karoo-Erfenis in Gevaar. Lantern, Vol 23, No 1, September 1973. 4-12.
WANGEMANN, Dr. 1871-75. Lebensbilder aus Sudafrika and Geschichte der Berliner Missions Gesellschast und Ihrer Arbeiten in Sudafrika. Berlin: Im Gelbstverlag des Ev Missionshauses in Berlin.
WATSON, JH. Undated. The Troopers' Sketch Book of the Boer War. Glasgow.
WHITLOCK, Peter. 1989. The Rural Architecture of the Sneeuwberg-Camdeboo Region. MArch research in progress, Department of Architecture, University of Port Elizabeth.
WILLIAMS, Donovan. 1959. The Missionaries on the Eastern Frontier of the Cape Colony: 1799-1853. Unpublished PhD Thesis, University of the Witwatersrand, Johannesburg, 1959.
WILSON, Monica and THOMPSON, Leonard. 1975. The Oxford History of South Africa: South Africa to 1870. London: Oxford University Press.

Copyright @ francofrescura.co.za