The St. Kitts-Nevis Digital Archaeology Initiative conducted a two-day shovel-test-pit survey in July 2008 to confirm the location of the slave village associated with The Spring, an 18th-century sugar plantation located in St. Mary Cayon Parish on the island of St. Kitts. Eighteen shovel-test-pits were excavated in an area identified as the village on the 1828 McMahon map of St. Kitts. Domestic artifacts dating from the mid-18th century through mid-19th century, including a high concentration of African-Caribbean coarse earthenware ceramics, suggest the presence of the village. Unfortunately, heavy erosion throughout the last century has destroyed any useful spatial data as well as evidence of house platforms.

The plantation recorded on McMahon's map of 1828 as 'The Spring', was first recorded as such in 1770, when Daniel Cuningham, then of Ludlow in Shropshire but late of the island of St Christopher, in his will of that date demised to his wife Elizabeth and her trustees his plantation in the parish of St Mary Cayon called The Spring which he purchased from Clement Crooke Doctor of Physic (National Archives PROB11/1036).

The earliest detailed maps of St Kitts possibly show the Spring Plantation. Norwood's map of 1700 and De Buor's map of 1706 both show a plantation, held by Ensign Crook and Clement Crook respectively, which is possibly the Lodge; on both maps the Spring may be an un-named or un-numbered plantation shown to the south.

One Clement Crooke certainly held a plantation in the parish of St Mary Cayon by c.1706, when he made claim for losses suffered in the recent French attack on the island. The following claims by Crooke are possibly our earliest description of the Spring:

Framed house of 2 rooms 34 feet long and 15 broad of Leeward timber except the rafters, ends and sides boarded ... worth £150
A framed room 17 by 21 feet with a porch, some part boarded, £80
A boiling house 34 by 21 feet with Leeward posts and sides .... Molasses cistern of boards (National Archives, CO 243/2, Fo.86).

Daniel Cunningham esq. evidently owned the Spring by 1750 when he mortgaged it to Robert Colhoun esq. for £14,880 plus interest. By then it was described as "all that plantation in the parish of St Mary Cayon, c.168 acres 3 roods 11 perches, and all that mountain land and wood land belonging to the same bounded to the N with lands late of Samuel Crooke, to the E with lands of William Ottley esq.and Timothy Earle esq., to the S with the tops of the mountains, to the W partly with the ridge and partly with the gut which divides the same from the lands of John Burryau", together with 130 negro and other slaves (St Kitts Common Records Book H, No.4, 1833, fol.326). At this stage Clement Crooke may have retained some interest in the estate. First, in the period 1755 to 1758 the index to deeds records transactions from Clement Crooke to Daniel Cunningham as deeds 5703-4 and 5932 (St Kitts Common Records, Index Book X, no.1; unfortunately the volumes for these years were too fragile even to be microfilmed so are not accessible for this research). Secondly, Clement Crooke was shown as the owner of this estate on Baker's map of 1753.

Later title deeds show that by 1832 the two plantations known as the Spring and the Lodge formed a single estate, which had formerly belonged to one William Crooke who was now deceased. By a series of complex legal proceedings linked to the redemption of existing mortgages the two plantations then passed to David Elliott esq., formerly of Clifton, the wealthy suburb of Bristol, who in that year sold the estate to Charles Adamson, a planter of St Kitts, recorded as the owner on McMahon's map of 1828 (St Kitts Common Records Book H, No.4, 1833, fols.308-345). The deed of sale includes a schedule with the names, ages and gender of the slaves sold with the estate; this can be compared with the earlier list of 1750 and the triennial register of 1828 (St Christopher Triennial Return of Slaves Book D).

The 1832 list also provides the names and general ages of each individual owned by Adamson. That year he owned 20 men, 25 women, 7 boys, 4 girls, 14 infant boys and 7 infant girls. The slave population totaled 77. Only one name was embedded with a clue to a person's ethnicity, age, origin or parentage. In such cases, two names are provided for a single individual, one proper name and one descriptor. Here Tom Creole was of the creolized Caribbean. No men or women had occupations linked to their names and none carried names suggestive of their origins in Africa.

Four slaves had names indicating they had been born on or purchased from other plantations: Alexander Cunnyngham, Robert Cunnyngham, Mary Mathew, Glasgow Mathew and Polly Ottley were likely connected to the adjacent Cunningham, Matthew (or Brighton) and Ottley plantations.

In 2004, staff with the Nevis Heritage Project in collaboration with National Museums Liverpool, conducted a large-scale landscape survey of sugar plantations and slave villages on St. Kitts. Their work focused on two parishes on the island, St. Mary Cayon and Christ Church Nichola Town.  Using the 1828 McMahon map as their starting point, archaeologists linked modern topographic maps of St. Kitts to the McMahon Map, allowing them to locate and map eight plantations in St. Mary Cayon and three in Christ Church Nichola Town. The Spring was one of eight plantations in St. Mary Cayon surveyed by Dr. Robert Philpott and his team in 2004. The major components of the estate, including house, windmill, cisterns, and slave village, were mapped using a total station.

The Spring Estate occupies a site which slopes down strongly towards the ocean to the northeast. A modern dirt track that leads to the spring, a water source uphill from the plantation, runs along the eastern edge of the estate. The plantation works are built on a series of terraces and the remains of the plantation buildings are disposed around a rectangular yard. The upper range of buildings to the south appears to contain the main house. Though not extant, a series of rectangular cellar rooms, including one constructed with a brick vault, is visible above the surface. An adjacent detached structure with a possible fireplace and open oven was identified and mapped. This may have been a kitchen.

The center of the yard is occupied by a stone windmill tower with an entrance on the west side that is approached by a ramp. The northern side of the yard, down slope, is bounded by a rectangular building, probably the boiling house, which incorporated two cisterns in the structure, and a possible platform. The yard is cut by a track that is probably contemporary with the use of the yard. Another building survives only as a single wall and its function is uncertain .

The slave village shown on the 1828 McMahon map was located in 2004.  It lies in the base of a sloping valley. The windmill and sugar works are located east of the village, separated by a steep slope and accessed by a track. The southern edge of the village is defined by a steep earth bank with a ditch on its south side that channeled water from the plantation above. The western side is marked by a deep and steep-sided ghut while the eastern edge is defined by a steep slope up to the works. The interior of the valley gently slopes towards the north. To the north are cultivation ridges and the northern limit of the village is marked on McMahon's map of 1828.

The 2004 survey was the starting point for the St. Kitts-Nevis Digital Archaeology Initiative, which is a collaboration among archaeologists from The International Slavery Museum (Philpott), The University of Southampton and Nevis Heritage Project (Leech and Morris), and DAACS (Galle, Neiman, Hudgins, Heath). In May 2008, archaeologists with SKNDAI and students from The University of Southampton's archaeological field school conducted a one-day shovel-test-pit survey at The Spring's slave village in order to confirm its location.

The team placed 18 test pits on three N/S transects through the center of the valley that contained the village. The pits were placed on 6-meter centers using a total station. Each pit was 50 centimeters in diameter. In most cases these pits were excavated to either subsoil or an apparent sterile layer. All sediment was screened through ¼-inch mesh. Artifacts were washed on Nevis and flown to Monticello where they were cataloged to DAACS standards. The artifacts will be returned to St. Kitts in 2009.

Domestic artifacts dating from the late-18th-to-early-19th centuries were discovered. However, it appears that over the last century heavy erosion through the ghut had destroyed any spatial data and evidence of house platforms.

The 18 excavated shovel-test-pits yielded a total of 453 artifacts, including ceramics, glass vessels, tobacco pipe fragments, brick, architectural slate, nails, and a single gunflint. Forty-five percent of the entire ceramic assemblage (n = 158) was comprised of Afro-Caribbean coarse earthenware ceramics (n = 72). The remaining imported ceramics (n=86) range from North Midlands Slipware (n = 1) and Delftware (n= 5) to creamware (n= 33), pearlware (n= 23) and whiteware (n= 4).  Instrumental Neutron Activation Analysis was conducted on six of the 72 Afro-Caribbean ceramic sherds excavated from The Spring. The raw data from the INAA analysis will be made available through a query on the DAACS website in the Fall of 2010. A detailed report on the INAA results from samples taken from Jessups, New River and The Spring can be found here.

DAACS's preliminary analysis has focused on dating the site.  As noted earlier, the village lies in a narrow valley, sandwiched between the great house and mill complex on higher ground to the south and a deep ghut to the north. Shovel test pits in the valley bottom were in excess of 1 meter deep, while those closer to the ghut were 0.2 meters deep. This raises the unfortunate possibility that the site is a hydraulic jumble: both artifacts and sediment have been eroded from their original contexts and deposited along the valley center line. For this reason, our digitization team excavated only 18 STPs at this site.

If the erosion hypothesis is correct, then spatial patterning on which our digitization design relies will have been erased. Analysis of the STP data from The Spring suggests this is in fact the case: the CA results show no pattern, nor is there a significant correlation between the CA dimension-1 scores and the MCDs (Figures 1 and 2). We conclude that internal site structure has been erased. The evidence from The Spring provides an important test of our field digitization techniques at the other sites. It shows that failure is detectable. That adds further confidence in the results from New River and Jessups.

However, the site does offer a single artifact sample that can be usefully compared with the Nevis sites. These results indicate that the ceramic assemblage from The Spring dates on average to the late-18th century (MCD: 1787). The date is a puzzle, since it does not fit with the evidence of the McMahon map, which portrays a sizable village on the site 40 years later. Perhaps, as we have seen at New River and Jessups, we are capturing the eroded remains of an early village, with the location of a later village as yet unidentified.