Marco Meniketti

Sugar Mills, Technology, and Environmental Change: A Case Study of Colonial Agro-Industrial Development in the Caribbean

Sugar plantation development and environmental change in the Caribbean are examined together through case studies of sugar estates and settlement distribution on a former British colony. The downscaling of production and general abandonment of the island after the mid-19th century left intact an agro-industrial landscape. Relic industrial features provide insights to pivotal economic episodes since the first founding of the colony in the 17th century. Technology, factory architecture, and artifacts were used to assess change over time and for hypothesizing phases of development congruent with socio-economic history in the region.

Introduction

And in their voyage from ye Downes landed att ye Barbadoes which they did not like, nor of Antegoa nor Mount Serratt. They came downe to Nevis ye 22th of July 1628, which Island they thought fittest for their settlement being next to Christophers, from whence they might be better supplied.

—John Hilton, Storekeeper and First Gunner of Nevis, 1675

¹Colonizing the Caribbean islands by non-Iberian European states during the first quarter of the 17th century was a bold move. Although the Caribbean was largely under Spanish dominion, Spain's preoccupation with developing mainland ore resources and, to a lesser degree, those of the Greater Antilles, left open a window of opportunity among the smaller islands for incursion by European rivals. Lacking mineral wealth, and once depopulated, these islands were easy for Spain to ignore. Newly arrived Dutch, French, and English colonizers in the first quarter of the 17th century concentrated on the production of luxury agricultural commodities. Entrepreneurial colonists also experimented with sugar—already a successful crop on plantations of Portuguese Brazil—but lacked the necessary technical expertise to ensure success. Unlike other commodities, sugar is labor intensive, may take upwards of four years to mature, and requires considerable technological skill and expertise to produce. Such skills and expertise did not reach the Lesser Antilles until after 1650. The eventual widespread development of sugar among the French and English outposts, however, changed the course of history for the region and nurtured the rise of capitalism as a socioeconomic system against the backdrop of an expanding Atlantic economy. Sugar became so vital to national economies during the 18th century that empires bitterly fought for sugar islands and willingly ceded continental land to maintain their grip on Caribbean real estate.

The fate of individual sugar colonies had ramifications reaching far beyond their shores into the centers of manufacturing and shipping in Europe, and to the slaving ports in Africa. Indeed, many of the important industries of European states were at first driven by their close association with the sugar industry. On these factors there is general agreement among economic historians. Less studied are patterns of settlement and environmental transformation wrought by agro-industrial development on the scattered islands. Moreover, the design and evolution of estate factories corresponding to economic change or technological progress has not been systematically addressed. Development of colonies prior to the 19th century in the Caribbean has often been viewed as haphazard or idiosyncratic, with limited relevance to the study of industrial capitalism. However, similarities in historic trajectories among various islands of different nationalities suggest an underlying structure having broader implications for social history. Systematic study may provide insights into the influences of industrialism, environmental transformation, and regional patterns of settlement. Further, by examining industrial and technological innovation in the sugar colonies, new insights might emerge into the much-debated origins of capitalism and globalization at the dawn of the industrial age.

The term "sugar colony" is used here to emphasize that these islands were, foremost, sites of production and resource extraction, not the loci of cultural expansion as was the case in North America. Cultural extension was a byproduct rather than the intention of settlement. As David Lowenthal has argued, these were places to make a living, not to live.² Demographic and settlement data support this view. As slave populations grew, European populations decreased in most cases. From the beginning, land was devoted entirely to commodity production so that all food had to be imported.

Since 1997 these issues have been central to research on the Eastern Caribbean island of Nevis, carried out as The Industrial Colonial Landscape Project. Nevis was once an important trading partner with the North American colonies and was the birthplace of American statesman Alexander Hamilton. The project was initiated as a collaboration between Michigan Technological University and the Nevis Historical and Conservation Society (NHCS) as a thesis project, and the work continued as dissertation research through 2004 in association with Michigan State University. In all, two seasons of archival work and five field seasons of survey and excavation culminated in the discovery and recording of ruins for more than two dozen previously unknown and undocumented sugar factories and estate complexes. Also found were associated landscape features and residential components, including laborers' villages and dwellings of plantation owners.

Following a case study approach, the industrial landscape of the formerly prosperous sugar colony is examined here to illustrate how changes in technology, factory layout, and modifications to natural environments reflect the various phases of economic history throughout the island. The focus here is primarily on comparative layouts and spatial dimensions of agro-industrial sites and on the vestiges of sugar production technologies constituting the observed colonial and postcolonial industrial landscape. The term "estate" is frequently used as a synonym in the Caribbean for semi-autonomous sugar plantations. With centralized governmental authority lacking and each plantation in direct economic competition with the others, estates operated in independent fashion through agents, brokers, and ship captains, often ignoring statutes and taxes.

Although several research questions framed the complete study, three relating to technology and industrial design are taken up here : (1) Are patterns of development, which might offer clues to broader, regional patterns of change, expressed in the form and function of sugar factory technology? (2) Can phases of development be found in the patterns of settlement and industrial landscape that correlate to economic history? In other words, what within the case study can be generalized throughout the region as a whole? (3) What has been the relationship between environmental change and agro-industrial settlement? In a related question having implications for comparative studies among industrial sugar landscapes, what is the evidence for a sugar-plantation landscape signature?

An integrated strategy was devised to answer these questions, combining field recording, use of archival sources (historic maps, probate inventories, letters), and demographic data, coupled with critical analysis of artifacts from both industrial and domestic contexts. Archaeological investigations of Caribbean settlements dated to the first quarter of the 17th century are extremely limited. To adequately address these questions, it was necessary to locate estate and factory sites spanning the period of first colonization in 1627 by English adventurers through emancipation in 1833—a time of tremendous demographic and political transformation that serves as a convenient temporal boundary. It was also deemed essential for comparative purposes to document several plantation sites and, where feasible, excavate structures to obtain relevant data concerning construction methods, site chronology, and implementation of technology. Much has been determined by merging spatial and temporal data. Individual sites were plotted according to age to generate corresponding maps for development periods. The map sequence then allowed the scale and scope of settlement to be analyzed and patterns of settlement to be followed sequentially over space and time.

Geography

Nevis is assured a place in history as one of the mother colonies of English settlement in the Caribbean along with St. Christopher (St. Kitts) and Barbados. The diminutive island is located at the northern end of the Lesser Antilles, an island chain extending from Puerto Rico southward to Venezuela like a string of pearls. Roughly circular in shape and dominated by an inactive volcanic cone rising to 3,150 feet, Nevis has been described as not unlike a "candy kiss." The first colonist's view of the landscape hearkens to an idyllic perception of a tropical Eden. Captain John Smith described the island in 1608 as "... all woody, but by the seaside southward there are sands like downs, where a thousand men may quarter themselves conveniently. But in most cases the wood groweth close to the water side ... at high-water mark, and in some places so thick of a soft sponge wood like a wild fig tree, you cannot get through it, but by making your way with hatchets." Sir Henry Holt reported in 1631 that the islands (St. Kitts and Nevis) were "... covered over in palmettos, cottontrees, lignum vitae, and divers others sorts but none like any in Christendom."³ The rich diversity of the climax rainforest was mistaken for boundless fertility. These descriptions contrast sharply with what was recorded only 50 years later.

Blessed neither with natural harbors nor year-round water, and much of its land tilted at upwards of 60 degrees, the island seems an unlikely place to plant a colony; nevertheless, industrial settlement thrived and prospered during the "sugar revolution" that began in the mid-17th century and continued unabated through economic triumphs and reversals into the late-18th century.⁴

The casual observer may be misled by the modern landscape on Nevis to imagine there has been minimal environmental impact of the sugar industry. Lush and green, especially on the western side where tourism is heavily concentrated, only stone windmill towers and a few chimneys hint at the industrial past. Yet, the picturesque coconut palms are a recent addition to the scenery. In fact, nearly all flora and much of the island's fauna are introduced or invasive species. Most native species are now extinct.⁵ The vegetation has more in common with Mexico, Polynesia, or West Africa than with what the native Carib Indians would have recognized.

Voyaging to Nevis in 1685, future founder of the British Museum, Hans Sloane, energetically collected botanical specimens and published his observations a few years later. His comments hint at the magnitude of environmental change, stating, "... the mountain on Nevis has been cleared nearly to the top ... where yet remains some wood and where runaway negroes harbor themselves ..."⁶

Today, second-growth forest cloaks former sugar works in vines and the sun-obscuring canopy. So dense can the vines become that GPS units are imperative for fieldwork in order to relocate structures that may stand invisibly only a few feet away. Often the canopy interferes with GPS reception. In a few places, modern housing developments and luxury villas have erased many of the terraced cane fields. Coconut palms, agaves, acacia, breadfruit, mango, and dozens of other shade-bearing, ornamental, or food-producing plants from distant shores now spread across the much-eroded terrain and are evidence of continuing environmental alteration.

The eastern side of the island is marked by shoreline cliffs, sculpted by Atlantic waves, and buffeted by the full force of the trade winds blowing in from the east. These same winds drove the slave ships filled with pitiful human cargo from Africa. In dramatic contrast, the wind diminishes on the leeward western side where marshy conditions prevail, interrupted by narrow stretches of sandy beach that are appealing to modern tourists. Moisture-laden clouds continuously embrace Nevis Peak, where the air is cooled, condenses, and precipitates to a greater degree on the western and northern slopes. As a result, Nevis has several distinct biomes that are influenced by precipitation, elevation, and exposure. These variables differentially influenced land-use decisions by planters regarding technological application, such as whether to construct an animal mill, windmill, or both. Such variables therefore played a significant role in industrial history by impacting agricultural productivity, settlement dynamics, rates of ecological degradation, and estate prosperity.

Weather-related events such as hurricanes could be devastating and drought, even worse. Letters from agents to absentee landlords throughout the 18th century overflow with passionate complaints about the weather and its impact on crops.⁷ Consider, for example, the human toll and distress exhibited in the following letters [original spelling retained]:

A dry time an very little suggar made in the island and a bad prospect for next year. No provisons in the island for negroes. If the weather had prov'd tolerably good he had ship'd my mother att least two hund: hoggs of suggar besides lady Russells annty and the plantation expenses. As things stand he shall not be able to shipp her above thirty.

From T. Tyrrel, Nevis 1726 June 1⁸

Informs me of an extreme drowth from beginning of December 1725 to the latter end of July 1726 ... many negroes and stock lost for want of provisions and water ... My losses are greater in suggar, not enough to supply the plantation expenses ...

From Joseph Herbert, manager in Nevis 1726 June 12⁹

Hurricanes are documented in planters' diaries on Nevis and other English and French islands and in letters as well as admiralty reports. Because the soil rests tenuously on a hard volcanic matrix, the effect of hurricane rains is to wash away the topsoil or to rush in torrents down the mountain's many drainage gullies, known as ghuts.

Another impact of hurricanes can be loss of shipping. Accounts of hurricanes between 1650 and 1790 describe several ships sunk in the harbor at Charlestown. Fluctuations in the Caribbean climate also sustained periods of prolonged drought, lasting sometimes several years with annual precipitation decreases of 50% or more. This pattern is prevalent today. Nevis has no year-round water supply, no lake or river to depend upon. Water must be stored in cisterns for all purposes, industrial or domestic, and estates devised ingenious catchment systems. Nevis experienced drought from 2001–2005, leading to water rationing. The resulting reduction in vegetation facilitated archaeological surveys considerably.

Historical Background

Arriving on St. Christopher in 1624, in the company of 80 gentlemen-adventurers and their anonymous servants, Captain Thomas Warner (1575–1649) established the first permanent English settlement in the Caribbean. A splinter group transplanted itself to neighboring Nevis within two years (see figure 1).¹⁰ Their objective was to found a colony for the production of exotic, highly prized, and lucrative commodities for European markets. Among these products were indigo, tobacco, and ginger. Early successes and the wealth derived from these products encouraged others to join in the hunt for quick riches. Settlement was accomplished by several competing joint-venture companies, each responsible for their own recruitment of labor, artisans, transport and resupply, and contracted shipping.

Figure 1. Location of Nevis in the Caribbean basin. The pullout shows the island in closer detail. Pullout map adapted by author from Nevis Ministry of Tourism map, 2004.

Nevis was soon parceled into wedge-shaped parishes, radiating from Nevis Peak like slightly bent spokes of a wheel—the first of many landscape schemas imposed by settlement. As a result of this sectioning, each parish encompassed multiple environmental zones, stretching from seastrand up the precipitous mountainside. This did not mean, however, that each parish was equally endowed with terrain suitable for agriculture or settlement and occupation. Some parishes were disproportionately comprised of mountainous slopes and deep ravines; others were host to vast tracts of fertile yet astonishingly rocky flats. Based on a few surviving estate maps, it is evident that many were of rectangular configuration, stretching upslope from the sea, providing these early estates with a variety of environments to exploit, conforming in some regard to the linear wedge shapes of the parishes themselves (see figure 2). Ideally, estates held forested land in reserve while clearing tracts for planting. In reality, the island was nearly deforested during its first half-century, and even the most unsophisticated planters recognized and described topsoil degradation.

Figure 2. An estate map from Nevis. The longitudinal layout was typical on Nevis. In this case, the estate begins at sea level and extends upslope into forested land. Map from Caribbeana, 1919 (see n. 17).

Until 1706, the well-defended colony avoided the disruptive conflicts experienced by other sugar islands during the raucous 17th century.¹¹ Conflict consumed almost the entire 18th century, and sugar plantations were commonly attacked and destroyed as strategic targets, with predictable impact on the economies of warring states. Early maps indicate fortification walls running nearly the entire length of Nevis' western shore. A series of batteries at regular intervals bristled with cannon, according to historic sources. These defensive works are occasionally visible today eroding from the shoreline as seasonal storms irregularly expose and rebury the coastal ruins.

From a macroscale perspective, Nevis was in the vanguard of regional sugar-plantation development—successful Nevisian estates became models emulated by others. Ultimately the limitations of scale besetting the colony allowed larger colonies to surpass it in production.¹² Nevertheless, for a considerable portion of the 17th and early-18th centuries, the Nevisian plantocracy achieved remarkable production quotas, and the colony acquired status and wealth out of proportion to its size.

Survey Methodology

Briefly, the chief instrument for this project was the quadrat field survey. A single survey section measured 250 meters wide by 1,000 meters in length. The principal tools were the archaeologists' two legs and eyes, aided by compass, GPS, and machete. The unit of analysis was the estate. This unit can be defined as inclusive of factory zone, domestic quarter, associated village (if apparent), and planting grounds. These areas were frequently bordered by stacked-stone enclosures running several kilometers, but occasionally boundaries were inferred from terraces or other extant features. Wide swaths of landscape were transected from which broad interpretation of landscape and settlement change was made possible. Field teams transected the quadrats at various intervals, depending on the degree of ground cover, canopy, and lateral visibility. In addition to the walking survey, available aerial photographs and extant historic maps were consulted to document landscape configuration.

The parishes of St. John (Figtree) and St. Thomas (Lowland) were the areas of earliest and most intense development prior to 1700 and continued to be significant throughout the history of the colony. It can be inferred that within these parishes would be remnants of all phases of settlement. Furthermore, these regions differ dramatically environmentally, from coastal strand to precipitous mountain slopes, thus permitting a broader sampling of possible environmental factors shaping settlement and industry. A random sample was conducted over the parishes, stratified at 500-foot elevations, from sea level to 2,000 feet, in order to test whether gradient was a factor in settlement.¹³ Our target survey was 40 quadrats, representing a 20% sample.

At the broadest scale, sites were mapped relative to island topography, using GPS to pinpoint their locations. To facilitate a more fine-grained scale of analysis, individual sites were recorded, and measured drawings of extant structures were produced. Change in environment and land use, including town development/expansion; community growth; and location on the physical landscape of plantations, dwellings, and mercantile enterprises can provide solid evidence of settlement patterning.¹⁴ Plantation communities (defined as interaction spheres encompassing multiple estates) are large spaces having numerous specialized activity areas as well as places where more subtle, often invisible, social interaction occurred. Our survey quads were designed to be sufficiently large to allow detection of these various site types yet not so large that households or important small-scale settlement features would be missed in an effort to meet time and coverage constraints.¹⁵

Episodes of Development

As a unifying framework, three temporal periods or phases of agro-industrialization are proposed here, each linked to definable technological innovation and parallel tracks of demographic and economic stages. These phase designations are fundamentally based on findings from the field and were combined with historical source data. Phase designations will be used throughout this paper to describe factory development. Landscape change or stasis inferred from field surveys and documents were integrated into phase construction.

Phase I, 1627–1655: Initial Settlement and Expansion

The first settlement was adjacent a natural spring. This phase includes the first imposition of European land schemas, boundary conceptions, and built landscape. Plantation establishment and settlement proceeded as fast as manpower was available for clearing and appears to progress along the coastline and on relatively flat land south and north. According to tradition, the first sizable settlement was Jamestown, founded just opposite St. Kitts across The Narrows.¹⁶ An important export during this phase was logs of Lignum vitae (ironwood) that dulled the axes used to cut it down. Phase I corresponds with initial plantation enterprises. Small sugar works were the norm.

Wills of estate holders prior to 1650 principally include bequeathals of tobacco with barely any mention of sugar at all. The phase ends in 1655 with major demographic changes resulting from the capture of Spanish Jamaica by English forces and resettlement by numerous Nevisian planters to the new colony. Nevis was apparently already deemed too crowded and offered few prospects in comparison with Jamaica. In terms of land use, plantations on Nevis prior to 1655 tended to be of modest scale, the majority of no more than 100 acres.¹⁷ Middling and small holders were more the norm as compared to other colonies of Antigua and Barbados. When divided, the resulting plantation acreage could scarcely qualify as estates. Although windmills impressively tower over the visual landscape, drawing considerable attention from tourists and researchers alike, much production, perhaps even most, was initially carried out with animal mills, particularly in regions where wind was unpredictable.

Phase II, 1655–1785: Ascendancy of Sugar and Plantation System Expansion

Production volumes during this phase increased in scale. Shortly after 1655, coinciding with the arrival of Sephardic Jews from Brazil (who established a community until expelled in the early 1700s), sugar became so important that it virtually eclipsed all other agricultural production. By 1680, sugar was the only plantation product of note.¹⁸ Monoculture would have a long-term adverse effect on the local environment. Indentured labor gave way to a systemic institution of slave-based capitalism, and new modes of production were sparked by an influx of experienced sugar producers. Charlestown and Newcastle grew into important shipping centers, serving as nodes of commerce. Population increases were dramatic. From fewer than 7,000, numbers of 20,000 and more are reported as the number of slaves and garrisoned soldiers rose. The number of merchants and planter families actually remained stable.¹⁹

The most important demographic shift over this entire period was the steady drop in European population and the precipitous increase in the African component, finally stabilizing at a ratio of 8:1 for nearly a century. Although a de facto English settlement, importation of Africans, failure to introduce such critical infrastructure as schools or farms, and the ad-hoc nature of governmental institutions are further indicators that the colony functioned essentially as a site of resource extraction rather than for cultural expansion. It was not unlike a mining district with the population comprised mostly of laborers and managers. This was not necessarily the case on French islands. This phase closes with an invasion by French expeditionary forces in 1785, followed by further collapse in international commodity prices.

Phase III, 1785–1833: Development to Decline

This rather complex third phase of development is highlighted by intensified consolidation of plantations, reduction in the number of estates, introduction of new production technologies, and recognition of major environmental problems such as erosion, soil depletion, and coastal land loss—all in the face of declining revenues. New products were manufactured, and some estates converted to cotton or coconut palm industries. Others relied increasingly on rum distilling (always present) for economic diversification.

Traveling to Nevis in 1825, Henry Coleridge wrote: "There are two steam engines employed in grinding canes, a thing I had not seen anywhere else except Trinidad. Surely where water and coals may be commanded, the certainty and rapidity of making sugar would in the long run be worth the additional expense."²⁰ These would have to be a reference to the two earliest engines on Nevis, which must have been installed a year or two earlier, the first indication that the heavy machinery of the industrial revolution had extended to Nevis. Coleridge's observations are that of a traveler, a tourist, yet he expresses a key rationale for the coming industrial age.

Emancipation serves as the terminus for this phase. In the wake of emancipation, many former slaves departed the island for colonies such as Jamaica or Guyana, where homesteading was more promising. Many planters sold their holdings or simply abandoned them and returned to England. For the laborers who remained, there was little land available to settle on and, owing to restrictive laws, no employment opportunities outside plantation work with former slave owners. Limited sharecropping was practiced.

Field Results: Archaeological Landscape

Successive field seasons were highly productive. Thirty-one of a planned 40 quadrats were achieved, with the majority in St. John. In addition, area-specific, nonprobabilistic surveys were carried out at three locations of interest. Every quadrat produced evidence of built or modified landscape. Sixty-five percent of quadrats contained relics of industry. In all, 22 sites were identified in the field.

Archaeological remains found during survey included mill platforms, stone-masonry boiling and curing houses, iron sugar-milling equipment (such as roller-crushers), cobble service roads between estates, masonry outbuildings, residential structures, and miscellaneous foundation traces. Steam engines were also recorded at three locations. Landscape modifications evident included terracing and stone retaining walls to level the industrial zone, road networks, boundary walls, stone clearances, and wells. Artifacts scattered near sites were collected, sorted, and assessed by type and use categories to provide texture to understanding of the varied activities transpiring at each location. These fell into two broad categories: industrial and domestic. By combining analysis of these multiple data sets with the few extant historic documents, a rich tapestry of 17th through early-19th-century landscapes emerged. Three post-emancipation villages known from an 1871 map were also recorded.²¹ Each was situated near or central to large estates. It is likely these occupation sites have pre-emancipation components, which have not yet been examined. All quadrats contained evidence of agricultural activity, usually in the form of remnant or extant terracing or furrows. Nearly all estates have palms, mango, or other introduced tree species that have commercial or ornamental value.

An image of the townscape can be conjured from a passage in the state papers:

Nevis contains 320,000 acres ... 2000 acres patented, the whole island settled, except the top of the mountain ... In Nevis, five places for trade, but two considerable; Charles Town, where are good dwellings and storehouses, built with the country timber, not exceeding 60 feet long and 20 broad, story and a half, the "Hurri-Canes" having taught the people to build low.²²

In the area most researchers believe to be Jamestown, archaeological evidence from this period corroborates the historic sources at least in terms of construction. Clearing a series of masonry foundations of similar dimensions revealed parallel structures, not unlike a warehouse district. Five structures paralleled the existing main road, which is paved over the original, dating to just after initial settlement. Walls were 24 inches thick, of dressed stone on the exterior and interior surfaces and well mortared (see figures 3, 4). Test holes captured subsurface data that suggested the buildings had been constructed directly on the shore, if not the beach. Artifacts, mainly pipestems and bowls and ceramics, suggest an 18th-century date for the feature.

Figure 3. Excavation in progress on a foundation in the Jamestown site. The structure was likely a storage warehouse. Photo by author.

Figure 4. Jamestown warehouse structure plan, showing two buildings (Features A, A-1) sharing a common wall between them. Plan by Paul White and author, 2004.

The survey discovered two mill complexes that had undergone extensive re-engineering (converted from Spanish train into Jamaica train), suggesting late-17th-century dates for first construction and strongly hinting at efforts to extend the working life of the estates by making the works more efficient in the 18th century. The critical indicator of technological innovation here concerns the position of the fireboxes of the boiling house. Sugar syrup must be boiled at specific temperatures, and impurities removed in stages. Control over this process is vital to the success of sugar production and was under the supervision of a "sugar master." Until development of the hygrometer, temperature and other elements of boiling were regulated by sight, smell, and the sugar master's elbow.²³ In earlier Spanish train technology, fires are maintained beneath separate "coppers" (boiling cauldrons) of diminishing size, whereas the Jamaica train makes use of chimney draft to pull heat under the coppers from a single fire source at the opposite end. The newer system is fuel and labor efficient while allowing suitable control of temperature over the boiling table's length. Efficiency in this case is not a measure of labor output but of product. Like stone, labor on Nevis was cheap, even if in chronic short supply. Better control ensured a higher quality and quantity yield from the juice.

The shift from actual copper to iron cauldrons likely corresponds to an earlier stage in technological refinement and cost-reduction measures implemented by savvy estate owners. But documents are mute on this point. Lamentably, coppers are mentioned in documentary sources as among the chief industrial components carried away by invading forces and with little detail. Kettle fragments on sites suggest the use of iron cauldrons (see figure 5). Actual use of copper kettles for boiling appears to have ceased during the 17th century, replaced by iron.²⁴ Boiling tables appear to have operated with only four coppers in the early years of production, advancing to five and six later. All boiling facilities surveyed that date later than 1750 had at least six, and this came to be a reliable secondary chronological indicator.

Figure 5. Iron cauldron for boiling sugar syrup at the Coxheath Estate mill. Many varieties can be found scattered across the landscape or in adaptive reuse. The smallest of these measured 3 feet across with a depth of 18 inches, while the largest had a diameter of 8 feet and a depth of nearly 5 feet. Photo by author.

Iron also was introduced at this time for other components of milling equipment. This slight technological change had profound effects on production. Copper distributes heat more evenly than iron and has other properties affecting coagulation of impurities, absent in iron cauldrons. Presumably, iron cauldrons were easier to manufacture and transport. They certainly have proven durable. Several examples of cracked and repaired cauldrons were found in the field, suggesting that replacement costs may have been dear or that the expedient of iron bars and bolts was sufficient to extend use life.

Most of the discovered factory mill complexes employed animal mills. Platforms showed no evidence of having ever used wind power. Not having towers to give away their presence, many of the sites were hidden by vegetation and most have escaped looting. Large masonry residential structures were found in association with some of these mill complexes. Preliminary dating of these complexes was based on associated artifacts (mostly ceramics) and by comparing construction techniques with structures of known dates.²⁵ Dating can be problematic as these structures were in use for a considerable length of time before abandonment. Surface artifacts or represented industrial technologies are better indicators of the final phases of operation than they are for periods of initial construction and production. Many estates operated both animal and wind-driven mills prior to the introduction of steam. Most of the estate works found during survey were of this type. Contrary to what might be supposed, steam power did not immediately replace windmills or animal-powered crushing. Nor were these innovations necessarily more efficient. Archaeological evidence suggests that even with onsite steam power, estates continued to maintain an animal mill, as illustrated in a period treatise on sugar manufacture (see figure 6).²⁶

Figure 6. This French industrial encyclopedia depicts a classic vertical three-crusher animal mill. Cane is manually fed between the crushers to extract the juice. The covered slot labeled as M in the figure represents the juice channel through which sugar syrup flows by gravity into a collection tank. Engraving from Recueil de planches, 1751 (see n. 26).

What steam mills offered was predictability. They could operate day or night and when harvest was ready. Crushing rates, however, were not significantly improved, although available crushing time was enhanced.²⁷ The number of persons replaced from the labor force and the number of persons required to man the machine was about the same as before. Bottlenecks in production remained principally in rates of harvest, for which the predictability of milling itself could not compensate, and this aspect of production required the usual complement of workers. Although the bagasse of crushed cane could be utilized as a fuel for steam mills in the same manner that it had fed the mill animals, fuel requirements were nonetheless a new consideration for plantation managers.²⁸ The advent of steam technologies in the first quarter of the 19th century occurred as the industry was in major decline, and many estates had already ceased operation. According to Richard Lobdell, many planters during the 19th century were in too much debt to acquire new loans for new technology, yet they could still manage to secure funds to maintain the status quo.²⁹ Acquisition and consolidation of middling plantations by a few elite planters during Phase III altered several landscape features, patterns of mobility, and production schedules as cane came to be transported to centralized mills.

Several known mill complexes standing in the parishes that were sampled also fell within designated quadrats. Several had operated well into the 19th century and are listed in an inventory dated 1871. Although shown on the island's 1984 ordnance survey map, these have never been documented. All have extant windmills in varied states of distress. As several were originally built prior to emancipation, these were documented by survey teams as components of the pre-emancipation landscape. Lastly, a few estates have been refurbished by descendants of the original planters' families and are operated as luxury hotels. Cordial interviews with owners and inspection of renovated buildings left untouched for "romantic and picturesque" purposes added another avenue for comparative chronologies.

The first reference for the use of iron for rollers on three-roller vertical crushers was in 1662. Rollers were cast hollow and supplied as a special order to fit each individual mill.³⁰ Prior to this, they were fashioned from ironwood. By 1721 each roller was being cast in one piece, with the gudgeon joined to the case by a spider (see figure 7). These industrial artifacts, of which several were found at various sites, provided chronological clues to periods of production. These industrial artifacts were interpreted as indicators of continual operation of facilities across phase boundaries.³¹

Figure 7. The center gudgeon from a vertical crusher in situ near the Ridge Mill complex; overall length 1.7 meters. Photo by author.

The process of sugar boiling and refining is well documented and need not be discussed here, yet no suitable chronology of cauldron manufacture and installation is available that would facilitate using these industrial components as a sensitive chronological tool. Several types were recorded in the field. Sizes vary according to their position within the boiling train. One site was located that closely resembled an illustration in Labatt's classic early-18th-century treatise on the subject and was assigned to the earliest phase. These boiling trains were open affairs. While a few were enclosed by wooden structures, others were only protected by lightly timbered and gabled awnings. Masonry was reserved for the table and for the curing house foundations. Chiseled insets in the masonry apertures supported the cauldrons by their flanges or lips. Enclosure of operations within masonry structures makes its first appearance in the late-17th century, as sugar manufacturing transitions to production at factory scale. All sites that were assigned to Phase I provided artifacts and architectural evidence pointing to continued operation transitioning into Phase II.

Artifacts were both industrial and domestic in character. Frequency analysis of ceramics, clay pipestems, and bottle glass yielded mean date ranges consistent across artifact classes (bottle glass, smoking pipes, ceramics, etc.) in correspondence with phase designations but were weighted toward the period of phase transition. No single artifact category belonged solely to any one hypothesized phase. For example, combed yellow and brown slip-decorated, tin-glazed wares were common to both Phase I and II (ranging from 1680–1750); obtaining multiple data sets therefore was critical (see table 1).

Table 1. Mean Manufacturing Dates: Ceramics Recovered during Surveys and Excavation Sampling

Note: Date ranges are shown to illustrate probable periods of use. Total sample from surveys: N = 1873. The use of these median dates, by default, places most of the investigated sites on Nevis during the present study into Phase II. Source: Meniketti, "The Historical Archaeology of Nevis, West Indies: Capitalism, Environment, and the Evolution of the Caribbean Colonial Landscape, 1625–1833" (PhD diss., Michigan State Univ., 2004).

Estate Layouts

Landscape signatures for sugar factories will be discussed here in terms of their associated development phase. With more than two dozen sites altogether, space allows for only a representative sample of mill-complex layouts and technologies implemented at each site. It should be noted that not one of the sites to be described has a windmill.

Spatial arrangements of the sugar works were reasonably predictable within specific chronological periods. With one enigmatic exception—when one feature was found, others generally were locatable as well. Mill platforms, either for animal or wind-driven crushing, were without exception positioned on the highest ground of the industrial sector, always situated close to and elevated above the boiling facility.

Where terrain was level, these platforms were raised on artificial mounds, as at the Long Point site discussed below. In other instances, mills were built on naturally elevated ground, with other structures placed downslope or in what appear to be excavated depressions below the mill. This configuration facilitated gravity feed for crushed cane juice, which sluiced into collection vats and clarifying tanks. Seventeenth-century engravings portray these conduits as little more than shallow troughs.³² Later versions were constructed of clay, were brick lined, or made use of iron pipes. Examples of each type were found over the course of this study. Adjacent to the sugar-syrup collection room would be the open boiling table or, in cases of larger operations, an enclosed boiling house. This general pattern continued even with the advent of the steam-powered mills after 1825. After analysis, three identified sites were assigned to the Phase I settlement against which comparisons could be made. In all cases, facilities were substantially smaller than the classic images of sugar factories depicted by 18th-century folios. None were found to have coppers in situ.

Most significantly, estate construction followed natural topography, rather than being aligned to cardinal directions, with terrain fundamentally influencing landscape configuration. There has been a generalized acceptance of the idea that Georgian architecture required or specified symmetry, and, by extension, industrial standardization implied rigid alignment to compass coordinates. Certainly there are splendid examples of such conformity to "Georgian rules" to be found in New England states. Symmetry, however, was not found to be central to industrial design on Nevis, although these principles of symmetry are evident in some grand residential structures of the late-18th century and post-emancipation construction on Nevis. To some degree then, lack of conformity to cardinal directions in factory layout cannot be used as a chronological indicator and may be a clue to social issues instead. However, steam mill operations of late-19th-century provenience in two cases investigated were aligned on grid plans conforming to a north-south axis, without regard to natural topography. This is significant as it dates much later than Phase III. A first conjecture is that steam power permitted the factory to be set up anywhere on the landscape without constraint. When estate consolidation was at its peak, canes could be delivered around the clock from distant fields to the centralized mill.

On the west-facing slope at an elevation over 1,100 feet, the archaeological teams encountered a factory complex adjacent to an overgrown cobble roadway (see figure 8). Operation of this complex straddled periods designated as Phase II and III as indicated by artifacts, layout, and architectural features. On an adjoining ridge within 200 meters, foundations were found for a multiroomed residence, having a small garden enclosure defined by collapsed stone walls. The following season, a detached building (warehouse?) was found 20 feet downslope. The house structure had remnants of a slate staircase at the ground floor entry. The Ridge House complex (SJ15KD5/21–1), as it was designated, was well situated for commanding outstanding vistas. Standing in the "garden," one could see to St. Kitts in the north. From the front of the building, one could make out the profile of Montserrat in the south. More significantly, the view took in the terraced fields surrounding it, allowing an estate manager to observe workers. The foundation was on a raised platform. Thirty-inch-thick walls standing more than 3 feet high gave integrity to the stone fill along this side. Tea sets, porcelain, cut crystal, and other fine domestic wares were recovered. Ceramics and bottle artifacts suggest mid-18th through 19th-century occupation of the household, prior to either abandonment or use in a nondomestic manner (Table 2). In a small adjacent rock outcrop, evidence was uncovered suggesting the promontory has been a popular location for consuming alcohol for at least 200 years.

Figure 8. The western facing wall of the Ridge Mill complex. Photo by author.

Table 2. Ridge House Ceramic Types and Vessel Forms
Ceramics	plate	cup	platter	saucer	bowls	tankard mugs	jar	serving	puncheon	salvor	storage	teapot	small vessels	lid	? Body	Totals
porcelain		20			14					1		1			128
saltglazed stoneware brown mottled						2					27
redware	13				3				2						50
terracotta
comb toothed yellow-slip decorated	18	10	2												35
whiteware/shelledged	16															16
whiteware/cream-ware	54	3		1											68	126
whiteware/transfer printed (see by color)
(blue)	10	4		4	3			2						1		24
(green)	3															3
(brown)	2															2
(red)	2															2
printed/stamped	3	3			12										10	28
delft	10	1			1										220	232
pearlware	32	1		2		2									44	81
colonoware					4										34	38
Carib earthenware															3	3
stoneware/generic							2									2
Bellarmine/Rhenish							23									23
white saltglazed		10			1	18	1						11		209	250
Rhenish/Westerwald stoneware					19										19
earthenware/slip decorated agate & marble ware	4	4			3	1										12
painted stoneware
other/unknown
roofing tile	136	26		7	24	40	26	2					11	1	588	861
	plate	cup	platter	suacer	bowls	tankard jugs	jar	serving	puncheon	salvor	storage	teapot	small vessel	lid	indistinct
Totals	303	82	2	14	65	82	52	4	2	1	27	1	22	2	1,389
Note: Combined with mean manufacturing dates, these data enabled development phases to be assigned to sites. Although overall numbers are low, ceramic types were consistent with other artifact classes, including pipestems and bowls, glass bottles, construction methods, and architectural finishes.

The mill complex associated with the house exhibits the greatest degree of episodic modification and changing industrial use of any the team surveyed (see figure 9). The enclosed boiling house is essentially intact (except the roof), but coppers had been removed. The structure was built in multiple stories with a Jamaica train as its initial design. However, later room additions along the exterior walls were constructed directly over fireboxes. Milling equipment was located in the brush, including the central shaft of the vertical crusher and several gears (see figure 10).

Figure 9. Ridge Mill complex plan. The road approaching along the east side was constructed from small cobble and appears to connect the factory to the coast road 3 miles away. Plan by author, 2004.

Figure 10. Iron center vertical crusher, gudgeon, and spider for a three-crusher vertical mill, found resting on the mill platform at Pembroke Estate (see Figure 19). Photo by author.

The mill's signature on the landscape is fundamentally the same as the others—built exhibiting the same careful planning that was cognizant of natural topography, with the mill platform on the highest ground. Of those the survey teams encountered, this platform was among the best preserved (see figure 11).

Figure 11. West-facing side of mill platform at Ridge Mill complex, looking northeast. From this side, the platform is nearly 5 feet tall. The collapsed stone wall in foreground is from a much later period and covers much of the fuel storage area. Photo by author.

Room additions were found to abut existing walls, rather than being integrated with original structures, which would have entailed dismantling the existing masonry, possibly weakening the structures at critical load-bearing points. In other cases, cheapness can be conjectured as the rationale for this construction. The mill complex has a staggering mix of masonry of variable quality. Variations in craftsmanship were apparent between additions and original structures. Whether this points to a lack of skilled masons or simply economic expediency is not clear. Courses of dressed stone were crude, over mortared, and void of continuity in spacing and style. Differential size and quality are suggestive of borrowing from other buildings and use of unskilled, imaginative labor.

The complex was not without its surprises. Excavations during the 2004 season revealed that the boiling room had been reconfigured for purposes unrelated to sugar processing. Trenching across the structure revealed the interior had been filled to create a level floor. Artifacts recovered included a cane hoe, chain links, iron hooks, and ceramics of 19th-century manufacture (chiefly whitewares). Beneath this fill, a downward sloping mortared floor was uncovered, ending at a rectangular masonry vault that had also been filled in with large stones. The recess was 50 cm deep and lined with dressed stone, bottoming at a level floor of cobble, gravel, and mortar (see figure 12, 13). Analysis is incomplete, but site alterations suggest that the facility was converted for cotton processing. A masonry vault of similar dimensions was noted at the Montpelier estate, which is known to have converted from sugar to cotton in the early 1800s.³³

Figure 12. Interior trench of the boiling room at the Ridge Mill complex. The excavation revealed a rectangular vault running longitudinally in the building inserted into the former boiling room floor. Photo by author.

Figure 13. The interior vault and modifications to the Ridge Mill complex boiling house. The shaft from the firebox was filled in, and masonry blocked it from the interior. Schematic by author, 2004.

Industrial Footprints Compared

Comparing Phase I architecture and layouts with subsequent phases was problematic. Few definitively Phase I sites were found, although one potential candidate of a mill complex that retained Phase I technology was located (see figure 14). In practical terms, Phase II and Phase III estates could not be easily distinguished within this sample from mill technology alone. Composition and frequency analysis of artifact scatters along with architectural details were used to assess sites and dates. For example, mortar finish and style of stone dressing varied over time as did quality of stonework. Uniformity of stone in terms of size, facing, and setting, along with mortaring practices, are reflected in buildings across the island, many with known provenience, particularly in churches and fortifications.³⁴ Although not uniformly sensitive to chronology, these details, when combined with additional evidence, facilitated designating phase association.

Figure 14. Early factory plan. Small in scale and with a Spanish train apparently still in place, this estate may span from Phase I to later phases. It has been tentatively identified from an 1871 map as the Beaumont estate, but the conjecture needs further testing. Plan by author, 2004.

One example of a Phase II site was located in the remote southwestern reaches of lower St. John Parish (SJ14MM5/16–2). The relatively intact four-copper boiling table, probable animal mill (almost too scant to detect), and landscape footprint were similar to a site much closer to Charlestown. It had been reconfigured as a Jamaica train and enclosed, at least from above, possibly with a shingle or even a thatched roof.³⁵ Adjacent structures were built on masonry foundations measuring 24 inches wide (see figure 15). The site was almost entirely devoid of artifacts, precluding this avenue of dating. What few were recovered from fill and a builder's trench point to a pre-19th-century operation. Conservatively, based on the technological and architectural attributes but in the absence of corroborating artifacts, a Phase II designation for this complex seems appropriate, even though its general configuration is suggestive of Phase I. Approximately 300 meters east are the remains of a possible laborers' village and the remains of a road bordering the immense Grandee Ghut (seasonal drainage ravines). As with most of the historic cobble roads, it averaged 16 feet wide with stacked-stone mortarless walls.

Figure 15. Boiling table of the Lower St. John mill complex; looking west toward inset for smallest copper of series. Photo by author.

Excavation at the complex concentrated on obtaining construction details from below the surface. The exterior north wall of the boiling table was penetrated by fireboxes that had been filled with rubble, an indication of renovation into a Jamaica train. A single 1 by 1.5 meter unit along the wall facing was excavated to expose the cleanout beneath the fireboxes. A room addition had been appended to this northern side, and a new floor level (50 cm above an earlier floor) had been created with fill and mortar. The arched aperture was masterfully crafted of cut stone. It, too, had been filled with rock and mortar (see figure 16, 17).

Figure 16. Excavated firebox at Lower St. John site, Unit 1. Photo by author.

Figure 17. Lower St. John complex plan. Boiling table held four coppers. Entire building would have supported a timber roof. Plan by author, 2004.

A few miles away, and at similar elevation, another mill site exhibited striking similarities in configuration. Because of its proximity to the new road leading to the recently completed deep-water pier, the survey team named this Long Point Road site (SJ9KR5/22–1). The complex has suffered grievously from unregulated salvage of building materials and its industrial components, just as at the Lower St. John site but with considerably more damage. The animal mill was, however, more in evidence. The base was located by means of clearing, and enough of its circular base arc remained for its original circumference to be calculated. A small, slightly elliptical cistern was sunk into the ground adjacent to the mill platform, which probably served to water the animals. The damaged boiling table once supported four coppers, but there was no indication of it having been enclosed by anything more substantial than a timber frame (see figure 18). A single base stone for supporting a timber superstructure was located. Associated artifacts span from early Phase II through Phase III, but typology frequencies, combined with details of construction, suggest Phase II as the chief operational period.

Figure 18. Long Point Mill complex, principal buildings. Note the collapsed boiling table, damaged at some time in the past when the coppers were salvaged. Plan by author, 2004.

Among Phase III sites, the Pembroke 1 Estate (SJ12MM5/16–1) incorporates critical features of Phase II sugar manufacturing along with later additions that suggest large-scale operation as a distillery. Buildings on the site from the 19th century are aligned with cardinal axes, while the structures associated with the 18th century and Phase II conform instead to natural contours and landforms. The mill platform was built on high ground and artificially supported on the north side. It appears the complex was strategically built around natural outcrops of flat bedrock utilized centrally as a natural floor (Figure 19).

Figure 19. Pembroke Estate mill complex plan. The 19th-century structures are aligned to north-south axis. Note how the complex was built to take advantage of natural topographic features. Plan by Kaitlin Deslatte and author, 2004.

Not far from the Jamestown site in St. Thomas Parish (STMM5/29-), another mill complex was discovered exhibiting the same characteristics as Pembroke, that is, structures following both organic landforms and cardinal axes with construction dating to different phases (see figure 20). The site may be the Lawrence Estate (based on the 1871 map), but extant documents do not corroborate the technological details. Inventories of estates in this area do not mention animal mills, yet the large and well-built animal mill dominates the site. No milling equipment was found in situ. Numerous bricks having maker's marks from the site provided clues to dating the fireboxes and some of the smaller structures.

Figure 20. Unknown mill complex near Jamestown. Three periods of construction were detected at the site. Plan by author, 2004.

The Richmond-Lodge Estate in St. John Parish is an excellent example of a Phase III complex. It was built during the later years of the 18th century and redeveloped in the 19th century. Situated at the intersection of a road that links it to three other mill complexes, the factory was carefully configured to conform to a grid pattern on the north-south axis. The steam engine is in situ as are boilers, clarification vats, and numerous other industrial artifacts (see figure 21). The estate's Great House stands in ruins adjacent and above the complex on a slight rise. From the front porch, the view to the sea would have been over the engine house. Two immense cisterns nearby served the factory. An even greater cistern is attached to the house. Domestic ceramics, bottles, and artifacts of industrial character persuasively indicate operations from the 18th through the 19th centuries and possibly into the early 20th. The engine dates to 1871, built by Fletcher & Co., London. Laborers likely lived in the village of Brown Hill, linked by road less than one-half mile distant. At Bush Hill Estate, less than a half-mile walk along an old road, another engine is in situ (1870s), also from Fletcher and Co., Edinburgh (see figure 22).

Figure 21. Richmond-Lodge Estate plan—(C) boiling house, (K) juice sluice, (J) clarification vat (moved from original location), (E) engine, (G) boiler. This complex operated principally in the 19th century using a steam engine. The landscape footprint is significantly different from earlier mill layouts. The Great House stands on a slight rise east, immediately adjacent to the mill in Area 3. Structure B was bulldozed in 2005. Plan by Krysta Ryzewski and author, 2004.

Figure 22. The engine at Bush Hill Estate. Bush Hill may have been in operation as soon as the early-18th century and did not cease operation until the early-20th century. Photo by author.

Settlement Landscape

Physical evidence along with documentary sources point to rapid and extensive environmental modifications. The earliest maps of Nevis dating to 1703 depict an island laced with a network of roads designed to connect estates and facilitate movement of sugar to the wharves in Charlestown. Nearly the entire road system had been completed by 1700, a network much in use today. Regular maintenance of roads in each parish was the work of slave gangs from plantations. Records suggest planters were loath to have their workers spending time on roadwork rather than estate chores, even though planters benefited most from the improvements.

The climax rainforest was clear-cut over much of the island for plantation development to at least 2,000 feet elevation. Terracing and boundary walls segmented space. Native plants and animals suffered from the introduction of non-native species or purposeful eradication as described by John Oldmixon in a 1706 account of the island's changing natural history. Settlement itself followed industrialization as planters acquired land and extended roads from the shipping centers to distant localities. On Nevis, landform change was immediate and continued unabated for a century following initial colonization. Land clearing was so complete and use of land so dedicated to production that colonists had to import food and timber. By the dawn of the 18th century, estate managers were in constant short supply of both. Natural water supplies were greatly overtaxed and supplemented by cistern construction and catchment systems. Quarrying the offshore barrier reefs for coral to use in the limekilns may have hastened coastal erosion in several sectors. Environmental degradation in the sugar colonies and its implications were not missed by that most astute gentleman of the age, Benjamin Franklin. He delivered a paper to a scientific society in Paris concerning the relationship between deforestation and local temperature increase on the French sugar colony of Guadeloupe.³⁶ French authorities responded by enacting ordinances designed to create forest buffers around streams and watercourses on the colony. The same troubles beset Nevis.

With introduction of steam mills after 1827, smokestacks appeared on the horizon, sharing the visual landscape with windmills, ushering in a new phase of development (see figure 23). Yet, the sugar industry of the Caribbean was already in steep decline during this period (see figure 24). The colony that once boasted nearly 300 estates could in 1871 account for no more than 100, and only 35 with steam power. In many instances, estates added steam-powered mills adjacent to existing windmills, quite possibly constructing their platforms on former animal mills. In other cases, such as is evident at the Richmond-Lodge Estate, steam allowed a completely new freedom of estate configuration.

Figure 23. The new landscape of Nevis after 1825. During Phase III, steam engines were introduced to the sugar industry but were adopted slowly. Chimneys came to share the visual landscape with windmill towers. Photo by author.

Figure 24. Muscavado sugar prices in London during Phase II of development on Nevis. Fluctuations represent periods of glutted markets and price crashes. The steady rise after 1733, nevertheless, spurred development. From Meniketti, "The Historical Archaeology of Nevis, West Indies: Capitalism, Environment, and the Evolution of the Caribbean Colonial Landscape, 1625–1833" (PhD diss., Michigan State Univ., 2004).

Discussion

During the early-18th century, demand for sugar grew to such an extent that it transitioned from a luxury good to a status approaching a staple. Production volumes contributed directly to price reductions as the market became glutted. Its distinction as a prerogative of the elite diminished, its use became more widespread, and production was further stimulated. Introduction of coffee and tea into the consumer markets at about the same time only drove harder the desire for sweetener. Several industries that had once stood independent of the West India trade became deeply invested. With the West Indies emerging as the hub of Atlantic commerce, shipping, manufacturing, and the slave trade grew in resonance. Sugar production is a consumer of lives as well as land. Labor shortages were chronic on Nevis. For much of its history, the sugar plantocracy relied on slavery for production. While some scholars have argued this delayed the implementation of more efficient technologies within the industry, it should be emphasized that milling of cane was only one step in a carefully choreographed process. Technological innovation was slowed by several factors, many of which were social in character. For some there was an aversion to associating with technocrats. Members of the plantocracy belonged to a social class that viewed itself as an emergent aristocracy. Many, if not most, were absentee landowners.³⁷ Being a planter was as much a lifestyle as it was an industrial enterprise, and bank loans intended for estate infrastructure were often siphoned off to support the planters' lavish lifestyles. This was not true in all cases, as plenty of middling planters, consistently in debt, could just manage to get by in the highly competitive market. Contemporary perceptions, nonetheless, beget the expression "as rich as an East India planter." Excavations at the residence of a middling estate revealed that in addition to locally produced, low-fired, coarse colonoware, fine imported ceramics and crystal were in use. The evidence suggests that status consciousness influenced consumer practice and display at all levels. One result was the evolution of a managerial class, beginning with onsite agents. Initially some were relatives, some were attorneys, and others were would-be planters. Few had expertise in all aspects of plantation operations.

An industrial signature of sugar factories on the landscape was clear and predictable. Most significantly, factory layout and construction provided evidence of a changing economic landscape as well. Nevisian settlement infrastructure neither evolved around towns nor with an agenda founded on cultural extension. It was instead geared toward systemic needs of production and agro-industrial development. At least in the first decades of the colony, focus was on land clearing, establishing estates, and production. There is little to indicate colonists were providing an infrastructure suited to long-term habitation. In fact, it is likely gentlemen-owners and laborers lived in or near factories in close association. The landscape signature of production in the earliest phases can be interpreted as one of semiautonomous settlement, with a modest allotment of land, small residences, mill platform and factory complexes built into natural topographic contours.

There appears to have been little differentiation between residences of the wealthy to middling planter during the first phase of settlement, although material culture differed to some degree. Ceramics from Phase I do not display significant variation either in form or function. Clearly rank distinctions existed, but these likely manifested through social avenues mostly unrelated to material culture. For instance, distinctions in access to foods, amount of labor required, mobility, fineness of clothing, or authority may have established rank, leaving little trace in the archaeological record. By the middle of the 17th century, however, marked distinctions arose in housing and consumer goods. In subsequent phases, estate houses and landholdings enlarged, planter wealth increased, but mill complexes grew only slightly through modification, the major improvement being structural permanence. Improved estate management and enlarged estate size, rather than technology, led the way to increased production and rising profits as prices fluctuated. Prices for sugar in London corresponding to much of Phase II are charted in Figure 24.

Terracing and land clearing expanded exponentially as latecomers to the industry arrived on the colony until only marginal land remained. Yet even these parcels were developed. Factory scale production during Phase II and III was accomplished by numerous medium-scale estates, many on marginal lands, but a handful of handsome estates and townhouses emerged as an elite class of planters came to dominate the others. Indeed, with one exception, all of the extant Great Houses date from Phase II or later. Only three towns ever grew to any significance. Anchorage at Charlestown is no better than anywhere else on Nevis, but with a great fort on a promontory overlooking the harbor and defensive walls running miles along its shoreline, Charlestown gained distinction, and Nevis came to be noted as a well-fortified colony (see figure 25)

Figure 25. Nevis, 1703. Note the extensive road network linking estates around the island. More than one-half of these roads are obscured or unusable today. The coast road has eroded into the sea over much of its length. The map was made by French cartographers with emphasis on description of coastal defenses. The text on the map points out cannon emplacements and forts. Courtesy of Horatio Nelson Museum Archives and the Nevis Historical and Conservation Society, Charlestown, Nevis.

Two economic developments may have influenced the increased differentiation of colonists. The first was the astonishing increase in demand for sugar and the infusion of capital from new planters. This appears to be a regional occurrence as French and Dutch colonies also experienced similar, concurrent expansion. Manufacturing centers in England responded to new wealth with new products. Shipping centers in England reacted to new labor demands with increased shipbuilding and investment in slaving v