vicinity of Rosewcll. The spring is also said to have 

 inspired the name of the house, but for this informa- 

 tion we have only the fanciful pages of Leonora and 

 the Ghost to rely upon. 



^\'hen marked on a plan, the ground hog's fi\e 

 entrances to his burrow create a circle some 12 

 feet in diameter. The majority of the artifacts were 

 found in the westerly holes — those farthest down the 

 slope towards the stream. The approach to the 

 northerly segment of the circle was impeded by a 

 large tree that the excavators did not remox'e because 

 of an agreement with the owners that the excavations 

 were not to do any permanent damage to the property. 



To ensure that the exca\ation should include the 

 entire trash deposit, an area 20 by 28 feet was hacked 

 out of the undergrowth. After initial clearance of 

 matted underbrush, many artifacts were recovered 

 from the surface and from the top 6 inches of humus 

 that had been disturbed by the removal of vines and 

 other vegetation. .Ml the artifacts from this part of 

 the excavation are described as surface finds and 

 cannot be used as dating evidence. 



PLAN 



Since the colonial deposit would in all probability 

 measure between 5 and 10 feet in diameter and 4 or 5 

 feet in depth, it was estimated that the area would 

 take less than a week to clear. In order to locate the 

 pit an exploratory trench ?) feet wide was dug in an 

 east-west direction from the top of the slope to the 

 western edge of the cleared area, a total length of 28 

 feet. 



The trench revealed that its eastern extremity was 

 excellently placed on the eastern rim of the colonial pit 

 but that its western end stopped short of extending 

 through the pit. The original estimate of the depth 

 was reasonably accurate for it was found that the 

 deposit was 5 feet 3 inches in depth at its deepest 

 point. The completed work revealed that the pit 

 was pear-shaped, with its ends lying southeast to 

 northwest, and was 34 feet long and 18 feet wide at 

 its widest point. The excavation took some 360 

 man-hours to complete. 



The first trench was di\ided into four sections (E- 

 H in fig. 3) arranged so as to leave a 2-foot balk 

 running north and south across the center of the pit 

 (fig. 4). Subsequently four area excavations were 

 laid out to cover the rest of the prepared area within 

 the right angles created by the trial trench and the 

 projected north-south balk. The.se area excava- 

 tions (.A-D in fig. 3) were dug so as to leave balks 



flanking eitlier side of the first trench, thus pre- 

 serving north-south and east-west sections through 

 the filling. R\entually the balks (lettered J-O) 

 were remo\ed. The final stage of the excavation 

 was the extending of the excavated area to the north- 

 west (P in fig. 3) in the hope of reaching a westerly 

 edge. Unfortunately, this part of the work was com- 

 pleted in some haste as the presence of chiggers and 

 swarms of mosquitoes that breed in the valley made 

 it abundantly clear that the excavating .season was 

 over. 



STRATIFICATION 



The slratitication of the pit (see fig. 4) was quite 

 simple. The top layer, of brown soil, was only some 

 2 inches thick at the top of the slope but it increased 

 to a depth of more than 2 feet at the west. This layer 

 contained very few artifacts other than those left in 

 the ground hog's burrow passages. 



The second stratum, consisting almost entirely of 

 oyster shells, bones, and artifacts, was the principal 

 source of finds. It commenced close to the top of 

 the slope l)ut thickened rapidly towards the center 

 of the pit; unlike the first stratum, it dwindled to 

 practically nothing at the west, causing the first and 

 third layers to merge together at that end. 



The third stratum, of varying thickness, was created 

 bv watei--eroded clay from the sides of the pit, into 

 which many artifacts had sunk. It rested on natural 

 clav at the west but became confused with a variety 

 of intrusions at the east. 



The fourth layer in area B was a localized deposit 

 of wood ash, burnt oyster shells, and fragments of 

 Indian pottery; it had been cut into the side of the 

 pit at the southeast. C^onsiderable reddening on the 

 pit side as a result of fire suggests that Indians had 

 built a cooking fire in the pit and probably had used 

 the side as protection from the wind. The fact that 

 the burning o\-erlay washed clay containing colonial 

 artifacts dating from the first half of the 18th century 

 clearly indicates a terminus post quem for this 

 occurrence. 



The primary filling was confined to the central area 

 of the pit, accumulating against the steep east bank 

 and spreading out thinly towards balk N-O. This 

 stratum contained fragments of early 18th-century 

 wine bottles and three thin glass flasks (fig. 29, no. 6), 

 tol)acco-pipe fragments, and the stem of a wine glass 

 (fig. 32, no. 7) of a type that could date no earlier than 

 about 1740. which provides the earliest date for use 

 of the pit as a repository for trash. 



164 



BULLETIN 225: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY 



