The most remarkable form utilizing gravel-tempered clay is found in the baking ovens which remained a North Devon specialty for over two centuries. These ovens vary somewhat in shape, and were made in graduated sizes. Most commonly they are rectangular with domed superstructures, having been molded or "draped" in sections, with their parts joined together, leaving seams with either tooled or thumb-impressed reenforcements. An oven obtained in Bideford has a flat top, without visible seams (USNM 394505; fig. 6).

An early example occurs in Barnstaple, where, in a recently restored inn, an oven was found installed at the side of a fireplace which is "late sixteenth century in character." Pipes and a pair of woman's shoes, all dating from the first half of the 18th century, were found in the fireplace after it had been exposed, thus indicating the period of its most recent use.[40] An oven discovered intact behind a wall during alteration of a Bideford house is believed to date from between 1650 and 1675.[41] That oven (figs. 7, 8) is now exhibited in the Bideford Museum.

At the other extreme, C. H. Brannam of Barnstaple in 1890 was still making ovens in the ancient North Walk pottery.[42] The following year H. W.

Strong wrote of Fishley's Fremington pottery that "shiploads of the big clay ovens in which the Cornishman bakes his bread ... meet with a ready sale in the fishing towns on the rugged coast of North Cornwall."[43]

Fremington ovens also were shipped to Wales,[44] and, according to Jewitt, those made in the Crocker pottery in Bideford "are, and for generations have been, in much repute in Devonshire and Cornwall, and in the Welsh districts, and the bread baked in them is said to have a sweeter and more wholesome flavour than when baked in ordinary ovens."[45]

[Ill.u.s.tration: FIGURE 11.--Sgraffito-ware platters from Jamestown. The platter shown above has a diameter of 15 inches; the others, 12 inches.

Colonial National Historical Park.]

Of ovens made at Barnstaple there is much the same kind of evidence. In 1851, Thomas Brannam exhibited an oven at the Crystal Palace, where it was described as "generally used in Devonshire for baking bread and meat."[46] In 1786, "Barnstaple ovens" were advertised for sale in Bristol at M. Ewers' "Staffordshire, Broseley, and Gla.s.s Warehouse."[47]

Thirty-six years earlier, in 1750, Dr. Poc.o.c.ke, who indefatigably entered every sort of observation in his journal, noted that in Devonshire and Cornwall "they make great use here of Cloume ovens,[48] which are of earthen ware of several sizes, like an oven, and being heated they stop 'em up and cover 'em over with embers to keep in the heat."[49] Poc.o.c.ke visited Calstock, "where they have a manufacture of coa.r.s.e earthenware, and particularly of earthenware ovens."[50] We have encountered only one other instance of ovens having been made at any place other than the North Devon communities around the Fremington clay beds. Calstock lies some 35 miles below Bideford in the southeast corner of Cornwall, just over the Devonshire boundary.

As for evidence concerning the manner in which these ovens were used in England, we have already seen that they were built into houses. Jewitt wrote that they "are simply enclosed in raised brickwork, leaving the mouth open to the front." They were heated until red hot by sticks or logs, which were then raked out with long iron tongs.[51] A bundle of gorse, or wood, according to Jewitt,[52] was sufficient to "thoroughly bake three pecks of dough." Poc.o.c.ke's remarks to the effect that the ovens were covered over with embers to keep in the heat suggests that they were sometimes freestanding. However, this could also have been the practice when ovens were built into fireplaces.

From an esthetic point of view, the crowning achievement of the North Devon potters was their sgraffito ware, examples of which in Brannam's window display have already been noted. Further evidence in the form of 17th-century sherds was found by Charbonnier around the site of the North Walk pottery in Barnstaple. These consisted of "plates and dishes of various size and section.... Extensive as the demand for these dishes must have been, judging from the heap of fragments, not a single piece has to my knowledge been found above ground."[53] The apparently complete disappearance of the sgraffito table wares suggests that they ceased to be made about 1700. They were apparently forced from the market by the refinement of taste that developed in the 18th century and by the delftware of Bristol and London and Liverpool that was so much more in keeping with that taste.

However, certain kinds of sgraffito ware continued to be made without apparent interruption until early in the present century. Instead of useful tableware, decorated with symbols and motifs characteristic of 17th-century English folk ornament, we find after 1700 only presentation pieces, particularly in the form of large harvest jugs. The harvest jugs were made for annual harvest celebrations, when they were pa.s.sed around by the farmers among their field hands in a folk ritual observed at the end of harvest.[54] Unlike the sgraffito tablewares, where style and taste were deciding factors in their survival, these special jugs were intended to be used only in annual ceremonies. Thus they were carefully preserved and pa.s.sed on from generation to generation, with a higher chance for survival than that which the sgraffito tablewares enjoyed.

The style of the harvest jugs is in sharp contrast to that of the tablewares, the jugs having been decorated in a pagan profusion of fertility and prosperity symbols, mixed sometimes with pictorial and inscriptive allusions to the sea, particularly on jugs ascribed to Bideford. The oldest dated examples embody characteristics of design and techniques that relate them unmistakably to the tablewares, while later specimens made throughout the 18th and 19th centuries show an increasing divergence from the 17th-century style. An especially elaborate piece was made for display at the Great Exhibition of 1851 in the Crystal Palace.[55]

Less complicated pieces, with a minimum of incising, were made for ordinary use, as were plain pieces whose surfaces were covered with slip without decoration. The trailing and splashing of slip designs on the body of the ware, practiced in Staffordshire and many of our colonial potteries, apparently was not followed in North Devon.[56]

Sites Yielding North Devon Types

Excepting the Bowne House oven and a 1698 jug (see p. 45), no example of North Devon pottery used in America is known to have survived above ground. Archeological evidence, however, provides a sufficient record of North Devon wares and the tastes and customs they reflected. Following are descriptions of the princ.i.p.al sites in which these wares were found.

JAMESTOWN, VIRGINIA: MAY-HARTWELL SITE.

The site of Jamestown, first permanent English settlement in North America, has been excavated at intervals by the National Park Service. The early excavations were under the supervision of several archeological technicians directing Civilian Conservation Corps crews. In September 1936, J. C. Harrington became supervising archeologist of the project, and until World War II he continued the work as funds permitted. Except for the privately sponsored excavation of the Jamestown gla.s.shouse site by Harrington in 1947, no extensive archeological work was thereafter undertaken until 1954, when John L. Cotter was appointed chief archeologist. Thorough exploration of Jamestown was his responsibility until 1956.[57]

One of the most interesting subsites in the Jamestown complex was the two and one-half acres of lots which belonged successively to William May, Nicholas Merriweather, William White, and Henry Hartwell. The site was first explored in 1935. On this occasion there was disclosed a meandering brick drain that had been built on top of a fill of artifactual refuse, mostly pottery sherds. The richness of this yield was unparalleled elsewhere at Jamestown; from it comes our princ.i.p.al evidence about the North Devon types sent to America.

[Ill.u.s.tration: FIGURE 12.--Sgraffito-ware cup and plate from Jamestown.

The cup is 4 inches high; the plate is 7 inches in diameter. Colonial National Historical Park.]

The May-Hartwell site was explored further and in far greater detail in 1938 and 1939 by Harrington, whose unpublished typescript report is on file with the National Park Service.[58] Harrington's excavation, in the light of historical doc.u.mentation, led to the conclusion that the brick drain had been laid during Henry Hartwell's occupancy of the site between 1689 and 1695. This was supported by the inclusion in the fill of many bottle seals bearing Hartwell's initials, "H. H." Hartwell married the widow of William White, who had purchased the property from Nicholas Merriweather in 1677. That was the year following Bacon's Rebellion, when Merriweather's house presumably was destroyed.

[Ill.u.s.tration: FIGURE 13.--Sgraffito-ware jugs, about 8 inches high, from Jamestown. Colonial National Historical Park.]

There were many hundreds of sherds in the fill under and around the brick drain, as well as in other ditches in the site. The North Devon types were found here in a.s.sociation with numerous cla.s.ses of pottery. The most readily identifiable were sherds of English delftware of many forms and styles of decoration related to the second half of the 17th century. There were occasional earlier 17th-century examples, also, as might be expected.

No 18th-century intrusions were noted in the brick drain area, and only a scattering in other portions; none was found in a.s.sociation with the North Devon sherds.

JAMESTOWN, VIRGINIA: OTHER SITES.

North Devon wares occur in the majority of sites at Jamestown, but it is not always possible to date them from contextual evidence because precise archeological records were not always kept in the early phases of the excavations. Nevertheless, narrow dating is easily possible in enough sites to suggest date horizons for the wares.

The earliest evidence occurs in material from a well (W-21)--excavated in 1956[59]--that contained an atypical sgraffito sherd described below (p.

43). The sherd lay beneath a foot-deep deposit that included Dutch majolica, Italian sgraffito ware, and tobacco pipes, all dating in form or decoration prior to 1650. This sherd is unique among all those found at Jamestown, but it is essentially characteristic of North Devon work.

Presumably it is a forerunner of the typical varieties found in the May-Hartwell site and elsewhere.

No gravel-tempered sherds occur in contexts that can positively be dated prior to 1675. A sizable deposit of gravel-tempered sherds was found between the depth of one foot and the level of the cellar floor of the mansion house site (Structure 112) located near the pitch-and-tar swamp.

This house was built before 1650, but burned, probably during Bacon's Rebellion in 1676.[60] The sherds were doubtless part of the household equipment of the time. All other ceramic fragments, with one exception, were a.s.sociated with objects dating earlier than 1660.

[Ill.u.s.tration: FIGURE 14.--Sgraffito-ware jug and cups from Jamestown.

Colonial National Historical Park.]

In sites dating from before about 1670, no North Devon wares are found, excepting the early sgraffito sherd mentioned above. Such was the case with a brick kiln (Structure 127) of early 17th-century date and two sites (Structure 110 and Kiln C) in the vicinity of the pottery kiln. In Structure 110 all the ceramics date from before 1650.[61]

The latest occurrence of gravel-tempered wares is in contexts of the early and middle 18th century. A pit near the Ambler property (Refuse Pit 2)[62] yielded a typical early 18th-century deposit with flat-rimmed gravel-tempered pans of characteristic type. a.s.sociated with these were pieces of blue delft (before 1725), Staffordshire "combed" ware (made throughout the 18th century, but mostly about 1730-1760), Nottingham stoneware (throughout the 18th century), gray-white Hohr stoneware (last quarter, 17th century), Buckley black-glazed ware (mostly 1720-1770), and Staffordshire white salt-glazed ware (1740-1770).

HAMPTON, VIRGINIA: KECOUGHTAN SITE.

In 1941, Joseph B. and Alvin W. Brittingham, amateur archeologists of Hampton, Virginia, excavated several refuse pits on the site of what they believed to be an early 17th-century trading post located at the original site of Kecoughtan, an Indian village and colonial outpost settlement which later became Elizabeth City, Virginia. Rich artifactual evidence, reflecting on a small scale what was found at Jamestown, indicates a continuous occupancy from the beginning of settlement in 1610 to about 1760.[63] The collection was given to the Smithsonian Inst.i.tution in 1950.

[Ill.u.s.tration: FIGURE 15.--This sgraffito-ware chamber pot, from Jamestown, has incised on the rim _WR 16 .._, probably in reference to the king. Height, 5-1/2 inches. Colonial National Historical Park.]

[Ill.u.s.tration: FIGURE 16.--Sgraffito-ware harvest jug made in Bideford, with the date "1795" inscribed. Borough of Bideford Public Library and Museum. (_Photo by A. C. Littlejohns._)]

JAMES CITY COUNTY, VIRGINIA: GREEN SPRING PLANTATION.

In 1642 Sir William Berkeley arrived in Virginia to be its governor. Seven years later he built Green Spring, about five miles north of Jamestown.

The house remained standing until after 1800. Its site was excavated in 1954 by the National Park Service under supervision of Louis R. Caywood, Park Service archeologist.[64] The project, supported jointly by the Jamestown-Williamsburg-Yorktown Celebration Commission and the Virginia 350th Anniversary Commission, was executed under supervision of Colonial National Historical Park at Yorktown, Virginia.

WILLIAMSBURG, VIRGINIA: EARLY 18TH-CENTURY DEPOSITS.

A small amount of North Devon gravel-tempered ware was found in sites excavated in Williamsburg by Colonial Williamsburg, Inc. These excavations have been carried out as adjuncts to the Williamsburg restoration program over a 30-year period. Few of the North Devon sherds found can be closely dated, having occurred primarily in undoc.u.mented ditches, pits, and similar deposits. However, it is unlikely that any of the material dates earlier than the beginning of the 18th century, since Williamsburg was not authorized as a town until 1699. It is significant, in the light of this, that North Devon pan sherds in the Williamsburg collection have characteristics like those of specimens from other 18th-century sites.

Also significant is the fact that no sgraffito ware occurs here. A gravel-tempered pan (fig. 23) from the c.o.ke-Garrett House site was found in a context that can be dated about 1740-1760.