INTRODUCTION
The Mexico-United States border region, in which Latin and Germanic cultures, the third and developed worlds, touch and interact, constitutes a fascinating but poorly understood culture area. A centuries-old movement of people and ideas has crossed this zone and continues massively today. Overlap, mixing, and hybridization of Anglo and Mexican cultures are occurring in the borderland, extending well into the United States and deeper into Mexico than is perhaps generally recognized. A self-conscious border subculture, whose practitioners call themselves fronterizos, has apparently formed. Geographers and other social scientists know all too little about the way of life in the borderland, in spite of its strategic location, vigorous cultural exchange, and exciting opportunities for research.

My own timid forays into the region suggest to me that it is even more complex than simply the place of meeting and fusing of two great civilizations. While much of the special character of the borderland is indeed the result of an overlay of Anglo and Hispanic features, certain other cultural elements are confined to that area, occurring nowhere else in the United States or Mexico. Such local traits call into question the simplistic interpretation of the borderland as merely an overlap zone.

POPULATION ORIGINS
The unique attributes of border culture likely reflect a more complicated settlement history than a simply Mexican and Anglo-American one. Since at least the l830s, the seam between Anglo and Hispanic cultures has attracted a wide range of ethnic minorities, who perhaps felt more comfortable in this zone of cultural mixing, where neither of the two locally dominant groups exerted an overwhelming influence (Jordan 1986, 409). Also, their cultural influence might more readily survive there and gain wider acceptance. The Texas segment of the borderland became home to Germans, Czechs, Poles, Wends, Irish, French, Swedes, Blacks, displaced Indians, Yankees, Chinese, Vietnamese, and the curious tri-racial group known as the Seminole Negroes (Jordan 1986,411-415; Oberste 1953; Weaver 1985; Porter 1952). As early as 1850, in the Texas counties of the lower Río Grande Valley alone, the 700-plus natives of Europe and the northern United States were a tenth as numerous as Mexicans and far outnumbered the 300 free natives of the slave states (excluding Texas). Irish (102), Germans (91), and English (73) were the most numerous European groups (U.S. Census 1850). Towns such as Laredo, Monterrey, Roma, and Matamoros acquired ethnically diverse populations by the mid nineteenth century (Newton 1964, 101-102).

These exotic ethnic elements had more to do with shaping a distinctive border subculture than one might expect. Cultural geographers interested in the region ought to take a cue from music historians, who concluded that the popular conjunto style, which fills much of the radio time in the Texas-Mexico border area, grew out of contact among Mexicans, Germans, and Czechs in South Texas or northern Mexico (Peña 1985, 35-36). Similarly, the important place of beer in the "Tex-Mex" cuisine of the borderlands presumably also has Central European roots.

BORDER FOLK ARCHITECTURE

The traditional cultural landscape of the border region also offers some features unique to the area. An intriguing example is a steep-pitched, parapet-gabled roof type confined to the Texas Nuevo León-Coahuila border, along and near the Río Grande (Figure 1). These highly unusual Hispanic folk houses are distinguished by a ridgepole (viga), rafters (latias), and battens (huilotes) all positioned well below the level of the gable coping, producing a parapet, the purpose of which was to provide protection on either side for thatched roofing (Figure 2) (Robinson 1979, 147). In recent years, thatching has all but disappeared, even on the Mexican side of the border, causing the replacement roofs of shingles or metal to lie substantially below the level of the gables, producing the parapet effect. Gabled houses are numerous in eastern Mexico, but the parapet subtype is much more restricted geographically, existing almost exclusively as a borderland type (West 1974).

The parapet gable form element appears in a baffling array of constructional and architectural contexts in the borderland, making its interpretation difficult and challenging. It is built variously of sandstone, sawn caliche blocks, or fired bricks, and interior walls are occasionally of adobe (Figure 3). The exterior is normally but not invariably plastered, and mortar can be either a lime cement or mere adobe mud. Parapet gables are usually symmetrically paired on each end of the structure, but some buildings have such a gable only on one end, perhaps in anticipation of enlargement, while others display medial parapets (Figure 4). The gable someetimes has a centrally positioned chimney, but it is not unusual to find the chimney attached instead to a shed room or to be absent (Figures 5 and 6). Such exterior chimneys, which are more widespread in the borderland than the parapet gable, may be annother exotic form reflecting the ethnic diversity of the region (West 1974). A few parapet structures have cylindrical corner buttresses.

Floor plans vary also, from simple, square, one-room dwellings (Figure 2) to various elongated plans consisting of one or two main rooms (Figure 6). Shed additions frequently appear, usually on the rear, but often on a gable side. Bed or table outshots occur on a few eave walls, and windows, while rare and often of recent origin, occasionally appear on front and sides. Doors are most often of the Spanish or double-leaf type, opening in two vertically-divided halves (Figure 6), but some specimens of Dutch doors have been reported (Newton 1973, 35-36).

Setting and function also vary. Parapet-gabled buildings serve most commonly as dwellings, but some are restaurants or stores, and at least two chapels (Figures 3, 7). Many are in towns, but others stand isolated in the countryside or form part of multi-structure ranch complexes. Both in urban and rural settings, parapet-gabled structures often incongruously abut flat-roofed buildings, a type also common in the borderland (Jordan et al. 1984, 197; West 1974; Robinson 1979, 132-145).

What are we to make of such baffling contextual, constructional, and architectural variety? The parapet appears to be one of an array of interchangeable form elements, causally linked only to thatched roofing. Dating provides part of the answer. The earliest known parapet gabling dates from about 1850, and most such structures seem to have been erected in the 1870-1890 period (Newton 1964, 106, 164). Though the flat-roofed form is far older in the Valley, the gabled type could represent a later, ecologically pre-adapted form, better suited to the fairly great amounts of precipitation received in the lower Valley and to the small size of roof support beams available there, allowing it to diffuse rapidly (Newton 1964, 111-113). The builders, to judge by the names of carpenters occasionally inscribed on rafters, were Mexicans (Robinson 1979, 147-148).

PROTOTYPES
Old World prototypes are abundant, and the parapet-gabled thatched roof almost certainly represents a European introduction, though we should be aware that flat-roofed Mexican and Indian dwellings often have parapets.The flat-roof parapet, however, serves fundamentally different purposes. Thatched parapet gables essentially identical to those of the borderland occur abundantly along most of the windy, hilly Atlantic perimeter of Europe, from the Shetland Islands southward through western Scotland, northwestern Ireland, Brittany, and Andalucía (Naismith 1985, 29, 195; Meirion-Jones 1982; Gailey 1984; Soule 1924, 8; Alvar et al. 1961, vol. 1, plate III, no. 513). Throughout that area, the type is remarkably consistent in form detail. Ada L. Newton favored a northern Irish prototype and went so far as to call the Río Grande Valley parapet-gabled dwelling the "Anglo-Irish" type, an explanation unenthusiastically repeated by architectural historian Willard B. Robinson (Newton 1973; Robinson 1981, 18). Architect Eugene George prefers Andalusian origin (George 1987).

Problems exist with both of these proposals. The mid-nineteenth century Irish migration to the borderland drew mainly on Wexford and other eastern counties, where the parapet gable is unknown. Andalusian origin would virtually demand the presence ot the parapet form elsewhere in Mexico, which is apparently not the case. Moreover, the chronology leaves no doubt that parapet gables long postdate the Spanish colonial era.

I propose instead that the borderland parapet gable was derived from Celtic Brittany in western France and that one particularly influential person was responsible. Pierre Yves Keralum (181771872), a Breton priest-architect from the port town of Quimper in Finistere helped lead the Oblate order's remissionization of the lower Río Grande Vallley between 1852 and 1872. A circuit priest, he rode horseback to the numerous towns and 70-odd ranches of the area. He is known to have applied his skills as an architect, stonecutter, and mason in the Valley. A surviving Oblate chapel, the La Lomita mission in Hidalgo County, Texas, is a parapet-gabled structure. Highly revered by the local peoople, who called him EI Santo Padre Pedrito, Father Keralum had the expertise, background, and mobillity to achieve a substantial implantation of the Atlantic European roof type (George 1976; Meirion-Jones 1982). Moreover, some related features such as gable-end shed rooms and table or bed outshots are also known in Breton folk architecture (Meirion-Jones 1979; 1982, 103, 222-223).

The borderland is a far more complicated place than we have imagined, and individual persons can be much more influential in diffusion, particularly in cultural boundary zones, than any superorganic view of culture permits. Geographers need to try to make sense of the area, bearing in mind both its multiple peopling and the strange workings of cultural diffusion.

REFERENCES CITED
Alvar, M., Llorente, A., and Salvador, G. 1961. Atlas lingüístico y etnográfico de Andalucía. Vol. 1. Granada, España: Universidad de Granada.

Gailey, A. 1984. Rural houses of the North of Ireland. Edinburgh: John Donald.

George, E. 1974. MSS drawings of structures innundated by the Falcon Reservoir, Zapata County, Texas. Copies in possession of T. G. Jordan.

George, E. 1976. Pierre Yves Keralum. In The handbook of Texas, Vol. 3, ed. E. S. Branda, 4700471. Austin: Texas State Historical Association.

George, E. 1987. Personal communication, Austin, Texas.

Jordan, T. G. 1986. A century and a half of ethnic change in Texas, 1836-1986. Southwestern Historical Quarterly 89:385-422.

Jordan, T. G., Bean, J. L., and Holmes, W. M. 1984. Texas: A geography. Boulder, CO: Westtview Press.

Meirion-Jones, G. I. 1979. The bed-outshot in Britttany. Ulster Folk Life 25: 29-53.

Meirion-Jones, G. I. 1982. The vernacular archiitecture of Brittany: An essay in historical geography. Edinburgh: John Donald.

Naismith, R. J. 1985. Buildings of the Scottish countryside. London: Victor Gollancz.

Newton, Ada L. 1964. The history of architeccture along the Rio Grande as reflected in the buildings around Rio Grande City, 1749-1920. M.A. thesis, Texas College of Arts and Indusstries, Kingsville, Texas.

Newton, Ada L. 1973. The Anglo-Irish house of the Rio Grande. Pioneer America 5:33-38.

Oberste, W. H. 1953. Texas Irish empresarios and their colonies. Austin: Von Boeckmann-Jones.

Pefia, M. H. 1985. The Texas-Mexican conjunto. Austin: University of Texas Press.

Porter, K. W. 1952. The Seminole Negro-Indian scouts, 1870-1881. Southwestern Historical Quarterly 55:358-377.

Robinson, W. B. 1979. Colonial ranch architecture in the Spanish-Mexican tradition. Southwestern Historical Quarterly 83: 123-150.

Robinson, W. B. 1981. Gone from Texas: Our lost architectural heritage. College Station: Texas A&M University Press.

Soule, W. 1924. Spanish farm houses and minor public buildings. New York: Architectural Book Publishing Co., Paul Wenzel and Maurice Krakow.

Texas Historical Commission, Austin, Texas. Phootograph files for Cameron, Hidalgo, Jim Hogg, Kinney, Maverick, Starr, Val Verde, Webb, and Zapata counties.

U. S. Bureau of the Census. 1850. MSS Free population schedules for Cameron, Starr, and Webb counties, Texas. Available on microfilm at the Texas State Archives, Austin.

Weaver, B. D. 1985. Castro's colony. College Station: Texas A&M University Press.

West, R. C. 1974. The flat-roofed folk dwelling in rural Mexico. In Man and cultural heritage, ed. Bob F. Perkins, 111-132. Geoscience and Man Vol. 5. Baton Rouge, LA: Louisiana State University Department of Geography and Anthropology.