1 Webb and Dodd, 1939, Pl. 26, Psnel 3:1. Judge Harry J. Lemley, Hope, Arkansas, has a similar vessel from the Crenshaw site, obviously a trade piece from the Alto or Gahagan Focus.

2 The Gahagan site in northwestern Louisiana has yielded several perfect examples of black Holly Fine Engraved vessels, all of which may have been obtained by trade from the Davis site itself, so close is their identity. Since the Gahagan multiple burials contained very few vessels but were rich in non-ceramic offerings, it is indicated that the black engraved vessels were highly prized, if not specifically manufactured or bought for mortuary purposes. See Moore, 1912, Figs. 22, 23; Webb and Dodd, 1939, Pl. 26:3, Figs. 1, 2. Several other points of connection with Gahagan are mentioned on pages 197-9.

3 A bowl with “stepped” design was included with the Holly type in Krieger, 1946, Fig. 19. This should now be considered only a possibility. On the other hand, Moore (1912, Fig. 2) shows a peculiarly shaped jar or bottle with a similar design from L'Eau Noire Bayou, Louisiana, suggesting that “stepped” designs were of very different origin but adapted to Holly-like pottery at the Davis site.

4 These statements on the appearance of scrolls apply only to engraved pottery at this site; the principle was already present in Phase 1 on a few examples of Crockett Curvilinear Incised.

5 Typical Crockett rim profiles are shown in Krieger, 1946, Fig. 22.

6 Although a minor point, seemingly identical bisected meanders, excised on carinated-bowl rims, are known in the Miraflores phase of Guatemala (photostats provided by A. V. Kidder from manuscript in preparation, especially his Fig. 17, K). We shall have more to say of possible Miraflores connections in the interpretations (pp. 227-8), but the relatively late position of this mode of decoration in the Davis site may be noted. I am not aware of its occurrence anywhere else in the United States, and in itself it would of course mean nothing.

7 Haag, 1939, No. 4. In first figure note especially upper right and third one down on right side; in second figure, upper left.

8 Ibid., especially the first, second, fourth, fifth, sixth (large central sherd), seventh, tenth, and eleventh sherds of the first figure, and second, fifth, sixth, and eighth drawings in the second figure.

9 In Krieger (1946, Fig. 19) the names “Hardy Incised” and “Dunkin Diagonal Incised” were both used, and may indeed still be separate types. They are here placed together as Dunkin Incised. The name Hardy Incised is that used by Quimby (1942, Pl. XV: 8-12, p. 267) for a clay-tempered type “found in both the Plaquemine and Natchezan cultures.” He confines it to crudely incised horizontal, parallel lines in a band around the vessel, but it was a mistake for me to use it for Davis pottery without more investigation into Louisiana and Texas pottery as a whole, with time perspective.

10 Although a sherd count, this is theoretically equal to a vessel count, since the reduction to vessels showed the same ratio in both cases.

11 Ford, 1936, Fig. 33, a-d and p. 275; also Haag, 1939, No. 3, p. 11.

12 I do not mean necessarily that both types were contemporaneous within a Coles Creek period. These punctates or impressions might have been produced over a considerable area and not confined to these two types. They could have developed out of the very widespread practice in the Eastern States of punctating or pinching vessel bodies with fingernails, or even resulted from imitating the semilunar stamped impressions on Hopewellian pottery.

13 However, very similar neck banding occurs in the later Frankston Focus in this area in the type LaRue Neck Banded. This differs mainly in having brushed bodies rather than fingernail punctating (Krieger, 1946; cf. these types as shown in Figs. 18, 19).

14 A discussion of the technique in “Caddo” cultures of northeastern Texas and southeastern Oklahoma occurs in Krieger (1946, pp. 237-41, 268, and Pi's. 33-35). At that time I wrote that a derivation of neck banding in the Fulton Aspect from early Pueblo IV cultures in the Southwest was more likely than an earlier connection, but also pointed out a peculiar resemblance to Alma Neck Banded in the considerably earlier Mogollon culture, both in vessel form and in the confinement of neck banding (and crude corrugation of neck coils) to the vessel rims. This was before the Davis site material had been fully plotted, but it was stated that the problem was still an open one. It now appears that Duren Neck Banded could well have existed at a time comparable to Mogollon, but it must also be stressed that no other traits at the Davis site support the possibility of contact with the Southwest at this time. That neck banding is also known in Honduras and Venezuela may or may not be pertinent. It is, of course, possible that the Duren type was a purely local invention. The chronological position of Alto Focus is further discussed on pages 219-24.

15 Ekholm (1944, p. 345) describes molcajetes in the type El Frisco Black of the Tampico area (Period II). Apparently any open bowl becomes a molcajete when an area on its interior, usually in the middle, is incised in the same general fashion as a functional molcajete used for grinding. Thus, Ekholm says that in Period II, “The molcajete surface is round and small and was perhaps more decorative than practical. It consists of simple incised parallel lines or cross hatch… .” I have no idea what the distribution and time placement of imitative or decorative (rather than practical) molcajetes would be in Middle America generally. The Davis bowls are not known to have had legs, but they are very fragmentary and a few legs were found on the site (Fig. 52, G-J).

When examining the collections of Richard MacNeish from the Sierra de Tamaulipas in 1946, I was struck by numerous distinctions between highland and lowland “Huastecan” ceramics which have never been made clear; MacNeish had numerous polished black sherds from shallow bowls, in the bottoms of which were incised and engraved designs plainly of decorative value only. A brief description of these in the Pueblito culture mentions only the incised designs and does not distinguish between functional and simulated (decorative) molcajetes (MacNeish, 1947, p. 5). On the basis of figurines, Pueblito is correlated with Ekholm's Huasteca III, IV, V. Engraved decoration appears to be far more common in the Tamaulipas highlands (e.g., MacNeish's Pueblito complex) than in the lowlands material of Ekholm, a point of considerable importance in comparing “Huastecan” and “Caddoan” ceramics. Superficially, there is a resemblance between the Davis molcajele-\ike bowls with engraved decoration, the Pueblito bowls with engraved interior lines simulating functional molcajetes and both of these with Ekholm's Period II molcajetes. All are in the monochrome tradition. However, one would need to know much more about the larger distribution of such bowls and simulations in Middle America before attempting to derive the Davis vessels from any particular area or culture.

16 Baker, Griffin, Morgan, Neumann, and Taylor, 1941, Pl. 49.

17 Not all plain rim sherds were included in the type, those which are too small being omitted from the study. That is, pieces only one square inch or so in area could easily have come from between the decorative units of some types—for example, from between the triangles of Pennington Punctated-Incised. Tiny sherds from vessel lips could have been plain above the decorative zone, and those from a bend where rim and body met might have been just below the zone. Bowles Creek Plain therefore includes only rim sherds of sufficient area to be rather sure that the rims were undecorated.

18 Vessels with similar high, concave, flaring sides and convex bottoms are shown by Smith (b, Fig. 1 [Mamom] Nos. 7, 13; Fig. 3 [Chicanel], Nos. 9, 12; Fig. 5 [Tzakol], No. 10). In none of these figures do the vessels have legs, but Smith (1936i) states that in the Tzakol “legs come in abundantly” with earliest polychrome and other new traits. See also the Pennington Punctated-Incised bowls in Figures 38, L and 40, D (photographs) and Figure 39, B, D (sketches). The bowl reconstructed in Figure 39, B is like the present plain one but much smaller and the rim not so flaring. Figure 39, D does not flare at all, the mouth being slightly smaller than the greatest diameter near the bottom, but the side is also concave—it is very much like a bowl shown by Ekholm (1944, Fig. 25, j ) , type Zaquil Black incised, most common in Period IV in the Tampico area. Two more small bowls with slightly flaring concave side and convex bottom bear rows of rectangular incised panels, polished over, around the side (Fig 51, x-x’). The five Davis bowls may therefore represent three slightly deviant variations of this form, but they are nevertheless unique for the United States in general. In conjunction with other ceramic elements also pointing to Middle America, this bowl form assumes more weight as a derivation from that direction.