(Page 84) Plain is next to the most common surface treatment of the shell-tempered ware; it is exceeded in number of sherds by less than one hundred by the fabric-impressed type. However, it has been mentioned above that some of the plain ware may be worn down sherds of other types. It may be expected that this is not an important factor in the comparative study. Figure 9, lower left, shows a plain rim sherd with tool scraping showing on the outer surface. It was recovered from the 12-inch level at H 1.

The plain surface finish is sometimes burnished. How often this may have been done intentionally is not possible to determine; too often the sherds have suffered some kind of wear. There is no slip application evident either.

Not too much may be said about the fabric impressed variety that would expand upon what has been said in this category for grit-tempered ware. All the forms of fabric described before apply here. This type is the most numerous of all the shell-tempered types. The fabric-impressed type seems to continue a long-standing tradition with very little change. The entire vessel is covered with the impressions. Figure 9, top row; middle left, and lower right are examples. Upper middle top row illustrates a sherd from C 1 which was fabric-impressed but has a row of small punctations beneath a pair of shallow incised lines that are separated by two rows of very small punctations.

(Page 85) The lower right sherd is a fabric-impressed specimen from P 6 to which parallel-line incising has been added. The center sherd is from H 1; it has been fabric-impressed, then had a series of parallel vertical lines incised beneath a row of punctations. Also the middle left (from H 4) is a rim sherd of a fabric-impressed vessel about 280 mm. in diameter. Individual cord-wrapped stick marks are applied to interior of the rim. The fabric impressing was nearly smoothed out then panels or zoned punctations were applied.

The simple-stamped technique is a radically different mode of surface finish when compared with fabric impressed. The latter has implications that it is pressed into the surface of the vessel and most of the pressure was applied on the inside of the vessel. This may be surmised from the size of some continuous fabric impressions; they seem too large for a fabric-wrapped paddle mark. In simple stamping, or stamping of any kind, the outer surface is malleated or struck repeatedly while an anvil (a rock) or the hand is held against the inner surface. All the simple stamping on shell seems to be from a carved paddle, i.e., a paddle with parallel grooves cut into it. None of the specimens suggested that a single-edged tool had been used to apply the marks. There is a uniform depth of the impressions on a given specimen, a fact that precludes the marks having been applied individually. The furrows are rather shallow, perhaps less than 1 mm. and are about 2 mm. wide and 2 mm. apart. There is variation in this but not much. All the outer surface of the vessel is treated including the lip. Figure 9, upper left, is from a jar about 300 mm. in diameter from H 2.

The occurrence of incising is quite rare--only nine specimens in more than two thousand shell-tempered sherds--but they may have a (Page 86) disproportionate importance in relating this ware with that of other areas. The incising is simple parallel lines, vertical or horizontal on the vessel, or it may be simple lines, parallel, and angular. (Figure 9, bottom row. The smaller specimen is from A 9, next right is from C 4.)

Punctations alone are even rarer than incising. Other than examples where the two occur together, three punctated sherds were found. These are simple punctates.(Figure 9, middle row right, from H 1.)

The sherd second from right in the bottom row of Figure 9 is an unique form. It appears to have a series of panels with a chevron design in each. The designs seem to have been applied by a roulette or a combedged (dentate) stamp. They are almost identical with the rows of punctations on the aboriginal pipes in Figure 10. This sherd is from C 4.

The general shape of the shell-tempered vessels was that of a globular jar not unlike the average shape of the grit-tempered ware. Where determinable, the maximum diameter of the vessel is at the shoulder, even though a pronounced break or change in curvature was never evident. The mouth of these vessels simply was a few millimeters--maybe 20--smaller than the maximum diameter which seems to average about 300 mm.

A distinct rim area, as described for the grit-tempered ware, is usually absent. Sometimes there is slight flaring. The lip is most often flat or everted, i.e., sloping downward and outward.

The base is always rounded or tending to a conical form.

Thickness of the vessels is rather uniform with 6 or 7 mm. an average, perhaps the maximum is 10 mm. and minimum about 4 mm. Basal sherds as thick as 18 mm. were found. There is no appreciable thinning of the (Page 87) lip in most examples of shell-tempered pottery.

The shell-tempered pottery could hardly be simpler in form. There are no appendages, no rim modifications, no special shapes.

It seems appropriate at this point to mention sherds of stone vessels. Fragments occurred at three sites: C 4 and P 34 each had one fragment and ten pieces were found at A 2. There was nothing unusual about the steatite sherds. The form of the vessel could be ascertained from A 2 as a simple bowl. The pieces were nearly 10 mm. thick although there was variation of nearly 3 mm. in the thickness. Bowl diameter could not be accurately determined but is about 250 mm.

A single sandstone vessel fragment came from P 35.

There were sufficient examples of aboriginal pipes found to give them some attention. A great number of fragments of white clay trade pipes were also found but they seem outside the provenience of this study. At least three major varieties of Indian manufacture were found in the surface collections.

Most common and perhaps the simplest of the pipes is that illustrated in Figure 10 bottom. lt is an elbow form of low angle, about 30^o , and having no embellishments of any kind. The pottery is a sandy paste with an occasional grit inclusion that hardly seems to qualify as a tempering material. The one complete specimen recovered (from N 3) has a well-finished appearance. The owners of P 1, Whalen Site, have from that site, an almost identical pipe. Fragments of this type have been recovered from P 1 (two specimens), P 7 (three specimens), P 12, P 13, (Page 88) P 16, P 18, P 19, P 27, P 35, P 37, A 1, H 12, R 1, and R 2. The specimen shown in Figure 10, just above the complete pipe, is a portion of a stem and most examples found are similar to it. However, found at A 1 and P 27, lower left, were nearly complete bowls.

The whole pipe from N 3 has a stem length of 60 mm. measured along the base. The stem has a maximum diameter of 16 mm. The bowl is 32 mm. as measured along its longest side and it is 22 mm. wide. The thickness of the bowl varies from about 3 mm. to 1.5 mm.

A second type may be a modification of this simple angle pipe for only a single portion of a stem was found. This specimen, found at the Vandemere Site, N 3, with the complete pipe, is illustrated in Figure 10, lower right. Its cross section is roughly triangular with the longest side 30 mm. across. It is 20 mm. from apex to base of the cross- section. A hole 6 mm. in diameter is bored through the stem. On the sides at regular 7 mm. intervals shallow notches have been made at approximately a 45^o angle to the long axis. The paste of this specimen seems identical with that of the whole pipe.

A third form of pipe is illustrated in the top row. There is only a possible remote connection in form between this type and the preceding. All specimens are, perhaps significantly, from Hatteras Island sites. The worn bowl, upper left, is from H 3; the next specimen is from H 6, and all the remainder including the long stem are from H 1.

The paste of this type has fine grit temper, smooth burnished to polished surface, and is well consolidated. Hardness is 4 to 4.5. The plain bowl specimen from H 3 was picked up from the water and it is worn but not all the surface has been eroded. It does not appear to have had any decoration. It is 20 mm. in diameter and about 40 mm. long, i.e., from base to lip. The wall is 2 mm. thick at the lip.

(Page 89) The designs on the other specimens are evident from the illustration. The regular spacing of the small punctations looks like rouletting but this can not be ascertained. The bowl from H 6 is about 40 mm. long; its shape probably approximates that of the bowl from H 3 and the stem must have been at about 30^o to 40^o to the bowl axis. The next bowl is 32 mm. long and because of its manner of breaking nothing specific may be determined about the stem. There is a suggestion that the stem was nearly in line with the bowl which would give it an appearance quite similar to the pipe found at Fort Raleigh (Porter, 1952, p. 36).

An examination of this latter pipe shows it to be of the third type, although it bears no punctate design. It is a hard, smooth, reddish brown pottery very similar in surface appearance to the bowl from H 6.

In 1938 Dr. Charles W. Porter, III, suggested that the pipe type of Roanoke Island might have inspired the form of the first English-made pipes (Porter, 1938). It would seem that the additional data supplied by these Hatteras pipes would strengthen this contention. During the construction of the Elizabethan Garden (R 2) near Fort Raleigh a fragment of pipe was found but this is of the first type (N 3) described above. Also at Northwest Point (R 1) the stems of aboriginal origin were apparently of that type. At none of the sites where pipes were found were fragments of both the first and the third type found. This implies a time differential, and on the basis of Harrington's find of the pipe in the ditch of the Fort it is judged that the third type pipe is the most recent. The other form may be a contemporary but had an earlier beginning. There is definite similarity of form between the first (Page 90) and third types.

No whole specimens of the English white clay "church warden" pipes were found, but on many probable historic sites where no other European or recent materials were evident pieces of these pipes were found.

Within a given area projectile points are usually rather varied and of little practical use in comparative studies. Within the scope of this study there would be virtually no projectile points to analyze were it not for the generosity of Mr. David Stick whose collection from the Duck Site (B 5) is illustrated in Figure 11. This group and those from P 35 shown in Figure 12 are the only considerable representation of projectile points. As a consequence, rather limited generalizations about projectile points may be made.

Two major types were found in the area: 1) a triangular form with some variety, and 2) a "lightly" corner-notched form with considerable variety.

The triangular form and its variation is shown in Figure 11. The basic form is an isosceles triangle but it may vary from a short, straight-sided type to a fairly long concave-base type. The former may be about 20-25 mm. in length (from base to point) whereas the latter are as much as 40 mm. Nonetheless, the triangular form seems to be a fairly consistent "type." The material from which the points are made is generally a brownish chert. Gray cherts occur but not as frequently as the brown. Only a few specimens were found (from P 16, P 18, and P 35) made from quartzite and another single example (from P 30) made from a granular grey quartzite locally known as Carolina slate. The triangular points are (Page 92) generally well made although not very finely chipped.

Triangular points were distributed as follows: B 5, 14; H 6, 1; P 8, 1; P 16, 1; P 30, 3; and P 35, 8.

In Figure 12 two large triangular points are shown (second and fourth from left in top row). In Figure 11 the upper right hand specimen is a large triangular form. These specimens are rather crudely chipped and do not seem to belong in the same category as the smaller triangles. At A 10 an example of the large crude triangle was found.

The second common form is well illustrated in the bottom row of Figure 12. It is basically a small triangle with either a corner only lightly or slightly removed or a very shallow side notch. The length ranges from 55 mm. to 27 mm. Their distribution is as follows: A 1, 1; H 8, 1; P 16, 2; P 30, 1; P 35, 6.

The remaining specimens all represent miscellaneous types. Another triangular variant is the right hand middle row specimen of Figure 11. This rounded base form occurs as single examples at A 10, P 30, and P 35. Perhaps the crudely chipped specimen of upper left in Figure 12 belongs to this general type; it is made of tan cert. An unusual form is the lozenge or diamond-shaped point, single specimens of which occurred at P 8, P 30, and A 12. The size of this form varies in length from 30 to 50 mm.

Only three large stemmed points were attributable to the area. In Figure 11 the large point of banded gray chert, 78 mm. by 37 mm., came from the Duck Site, B 5. In Figure 13 the two specimens of upper row are unusual. The middle one is of Carolina slate. It is of a contracting stem type that is very abundant farther inland. An additional specimen came from A 12. The other specimen is a large example of the slight corner-(Page 93) notched form. Perhaps the fragments of larger blades shown in Figure 12 middle row and Figure 13 lower right belong to one or the other of these forms.

In Figure 12 upper right is a long, narrow, thick, blade made from gray chert. Its tip is broken off. It may have been a knife rather than a projectile point.

A single small corner-notched point of the type of the large-stemmed form from B 5 was found at P 16. Its length was 33 mm.

(Page 93) Not many additional artifacts were found. Figure 13 shows most. The object in upper left, from H 4, is battered on both ends and appears to have been used as a hammer after shaping. It is of quartzite and measures 80 x 38 x 20 mm. Next to the right is a smooth-faced pebble from A 2. It too is of quartzite and was used as a polisher (of what, of course, is unknown). At the right end of the row is a rather crudely-chipped blade of Carolina slate. It measures 110 x 45 x 16 mm. and was found at P 16. In the center of the figure is a neatly-chipped scraper manufactured on a unifacial flake. It, too, is of the same quartzitic stone and comes from P 3. To its right is a chipped lump of black mottled flint; found at P 18, it is too large for a gun flint.

In the lower left is an excellant example of an arrow-shaft smoother; it was picked up from the water at Northwest Point (R 1) on Roanoke Island. It has grooves on all faces. A total of thirteen distinct grooves may be discerned. The stone is quartzite and is about 60 mm. in diameter and 45 mm. thick.

The next specimen to right, from P 34, is a blade piece of a large (Page 95) quartzite, chipped celt. Not much smoothing has been done. The last specimen is a rudely fashioned celt or perhaps whetstone.

A few other objects were found. A fine flat slab mortar was found at C 4 on Colington Island (Figure 14). It was recovered from the water at the beach edge. Hammerstones, generally quartzite cobbles, were found at several sites, as follows: A 1, A 2 (2), A 7 (4), A 8, A 10, A 12 (2), B 4, H 4 (3), N 4 (2), P 7 (5), P 16, P 18 (3), P 19, P 23, P 24, P 25 (3), P 30, and R 1. Three more specimens were found in the test excavations at A 12.

At P 25 was found a single specimen of a pebble (chert) battered on all sides that may be called an abrader.

Bone artifacts are limited to a single specimen which might be surprising in that bones were quite numerous at H 1 and A 12 where stratigraphic tests were made. However, only a small antler projectile point at A 12 was disclosed. This is a typical conical point bored to fit on a shaft. There was no additional hole on the side.