1 The author prefers “milling stone” as a general designation for slabs on which rotary motion was used. A previous suggestion of “seed slab” (Krieger, b, p. 51) is hereby rescinded. “Metate” is reserved for, and limited to, stones on which back-and-forth motion was used in the Puebloan and Middle American manner. The latter implement was diffused eastward into Texas about as far as Fort Worth (Krieger, 1946, pp. 79, 108-9, 126, 138) but is entirely unknown in the “Caddoan” area. If the same distinction were made elsewhere, most so-called “metates” in North America outside of Middle America and the Southwest would belong to the “milling stone” category.

2 It reminds one of the full-grooved hematite axes of northeastern Texas “Archaic” sites. One hard blow would shatter such an axe. If non-utilitarian, these fragile objects might be called “ceremonial,” or possibly they symbolized rank or property. Hematite, even in large, pure lumps, would not be very precious here, but perhaps enough so to convey a measure of prestige.

3 Identification and information from Dr. H. B. Stenzel, Bureau of Economic Geology, University of Texas.

4 Krieger, 1946, pp. 193, 202-3, 215.

5 Tables 16-17 do not reveal any specimens of Catahoula sandstone, hones or otherwise, from Phase 1 in the mound. We have mentioned the danger of drawing conclusions on absences perhaps due to chance, but collectively they point in this case to the use of Catahoula sandstone only in Phase 2 and after. I have no record of it in any “Archaic” or “Burial Mound” culture in Texas or elsewhere.

6 Moorehead, 1922, Figs. 27-8. Note both the rectangular and oblong or pointed forms, the neatly squared edges and general care in finishing all around.

7 Similar sandstone “tablet” fragments, with and without beveled edges, were found in the Harrell site in north central Texas and may be unique for the Henrietta Focus in that region (Krieger, 1946, pp. 104, 119, 142, 152). Catahoula sandstone was not carried into this region or anywhere into the Plains, so far as I know. Since the eight Davis specimens are all from shallow depths in the field (Table 17), a temporal overlap with Henrietta Focus is easily possible although the Hopewellian “tablets” show the difficulty of a definite statement. These objects do not necessarily represent the same culture trait(s) in all three places.

8 Information from Glen L. Evans, geologist and Assistant Director of the Texas Memorial Museum, Austin.

9 Identification by Glen L. Evans (see preceding note). W. S. Webb (1939, pp. 189-96) cites the smashing of “greenstone” celts as a trait of the Copena Focus in northern Alabama and mentions “the frequency with which evident intentional destruction of artifacts has been found on Hopewell sites.” This trait was not, however, noted in the seven Pickwick Basin components of the Copena Focus (Webb and De Jarnette, 1942, p. 304). Webb and Snow (1945, p. 88) mention the preference of Adena peoples for granite celts, broken specimens being “very common” on any Adena village site. Flint and hematite celts also occur. They note that many fragments are from the pole end and suggest the celts were broken when used for hammering. Only three of the 30 Davis fragments are poles, one being split lengthwise. Whether or not they were deliberately smashed, the Davis specimens are collectively of “greenstone,” including several materials, all of them imported.

10 C. H. Webb, 1944. Ten schists were recognized, talc among them, but all sufficiently alike to have come from the same outcrop. Webb cites several occurrences of stone bowls in relatively early sites in Arkansas, Louisiana, and the Southeastern States, and there is general agreement that they belong principally to the late Archaic (preceramic) horizon in the east. He says that “At each site [in the Pickwick Basin] there was evidence of a long occupation before the advent of the stone vessels, which in turn clearly antedated clay pottery usage.” See also W. S. Webb and De Jarnette (1948, pp. 17-19), who place steatite and sand stone vessels in their Archaic (3) period, the latest preceramic in the Tennessee River basin of northern Alabama; and MacNeish, 1948.

11 Information from Dr. H. B. Stenzel, geologist, Bureau of Economic Geology, University of Texas.

12 Courtesy of Catherine O'Neal, Bureau of Economic Geology, University of Texas.

13 Possible sources include the volcanic areas of New Mexico, extreme western Texas, and Mexico, to name only the nearest. Carriage from New Mexico or west Texas is possible, obsidian being found occasionally in central Texas archaeological sites. The obsidian might also have been carried from some part of Middle America, or perhaps in some round-about way connected with the Hopewell peoples who obtained obsidian from the Rocky Mountain region long ago. Obsidian, like most flint, can hardly be traced to definite source areas until its varieties are seriously collected and mapped.

14 Waring and Holder, 1945. See especially their Figure 1, Row IV, a, except that the present eye is oval rather than a diamond.

15 Burnett (1945, Pls. XI, XII) shows several concentric patterns on Spiro ear spools. Webb and Dodd (1939, Pl. 28, Panel 3) show Gahagan specimens. At both places copper sheeting on the decorated faces is fairly common, although not invariable. None of the Davis specimens shows any signs of copper or copper stains; the three bits of sheet copper found are of unknown use.

16 Krieger, 1946, pp. 185, 215, Pl. 22, c. It has also been observed, perhaps as a borrowed and minor type, in the Plains-like Henrietta Focus of north-central Texas ﹛Idem, pp. 103, 140, Fig. 7, M-o). As the Henrietta Focus crossdates with Fulton Aspect foci in the “Caddoan area,” the type seems to have survived longer among borrowing peoples than where it first appears (i.e., at the Davis site). Mr. Joe Ben Wheat reports it as the first small (arrow) point style in the pottery-bearing middens excavated by him in the Houston area (Smithsonian Institution, River Basins Survey, 1947); here the heavy (dart) points continued in use after Alba Barbed appeared.

17 Moore, 1912, Figs. 26 (note protrusions on three stems at left), 28, 29 (slim tips), and 33 (various combinations).

18 Lemley, 1936, Pl. 11.

19 I first noticed these on display in the Teotihuacan museum in September, 1943, but thought it mainly an interesting coincidence. It now appears much more likely that they could be historically related. These represent, I believe, the first appearance of “arrow” (as opposed to “dart”) points in the Valley of Mexico, dating roughly 500 A.D., and can be said to differ from Hayes Barbed only in the material: obsidian (see LinnS, 1935, Figs! 301, 302, 308, 310, and p. 148). The four examples shown by LinnS do not come as close to typical Hayes Barbed specimens from Arkansas as do several others on display at Teotihuacan. Before drawing any conclusions it would be necessary to know where this style is found, spatially and chronologically, elsewhere in Mexico—if anywhere. On this there is nothing in the literature.

20 Krieger, 1946, p. 115, Fig. 7, P-S.

21 Ford and Quimby, 1945, Fig. 8, particularly D, H-K (Tchefuncte culture, Louisiana); Ford and Willey, 1940, Fig. 45, A-F and possibly i, and Fig. 46, A-H (Marksville culture at Crooks site, Central Louisiana); C. H. Webb, 1948, Fig. 44: 1-3 and possibly 4-6 (preceramic culture at Poverty Point, Louisiana); Lewis and Kneberg, 1947, Pls. 1, 2, “tapered stem type” and “beaver tail stem type,” the latter said to be a “Late Archaic type, or Wroodland” and found “rarely at most sites, frequent only at two of the southernmost sites in area,” i.e., in the western valley of the Tennessee River, western Tennessee. The type is also quite common in non-ceramic sites in eastern Texas and Arkansas. It was probably never established in the Great Plains. Mr. Joe Ben Wheat, excavating in the Addicks reservoir basin near Houston, Texas for the River Basins Survey in 1947, found some evidence of two varieties of the type; those with more or less parallel-edged stems and straight base were relatively more common in the lower levels of the Doering site, while the more pointed stems appeared in somewhat greater number in the upper levels. This indicates a slow shift taking place about when pottery appeared, the pottery itself being very unique but resembling Woodland vessels with pointed bottoms. I am indebted to Mr. Wheat for this information, which he plans to publish.

22 Thus, in the works just cited, the following specimens seem to follow this pattern: Ford and Quimby, 1945, Fig. 8, E, M (A-C, F, G may be a third type). Ford and Willey, 1940, Fig. 45, G, H, L, and possibly K, and Fig. 46, I-L, p. Lewis and Kneberg, 1947, Pl. 1, “corner notched type” and “straight stem type,” although some of the latter look more like Gary variants than Ellis. Their “side notched type” is quite distinct and is not represented at Davis. I do not doubt that spatial, temporal, and frequency distinctions will appear with realistic and especially widespread typological studies, but without them there is little purpose in trying to define either type or its distribution accurately.

23 A feature commonly found on Folsom Fluted, Clovis Fluted, and Plainview points (Krieger, 1947c, 1947d).

24 Kelley has provided me with full data on the depth and cultural and geological associations of these points. His conclusions are tentatively stated as follows: “Following the chronology and correlations discussed by Kelley (1947), the age of these specimens would cover practically the entire time range of the Central Texas sequence. The cultural association in all instances is with the Round Rock and Clear Fork Foci, Edwards Plateau Aspect. At the Round Mountain site, two Wells specimens were found deep within the 40-foot terrace at actual depths of 230 and 124 inches, near the beginning of deposition of the terrace. This is thought to have begun at circa 4000 or 2000 B.C. and to have extended to about 1000 A.D. In my opinion, these early specimens are somewhat aberrant typologically, although they do apparently fall with the range of variation of Wells Contracting, Stem. In the Sandy Creek, Grelle I, and Grelle I I sites the type seems to have ranged from the upper part of the 40-foot terrace (circa 1000 A.D. minus X) to just prior to the Austin and Toyah Foci (ceramic cultures beginning circa 1500 A.D.).”

Typologically, the four specimens from the Grelle I and II sites are certainly closer to the long-stemmed Davis examples, the chief difference being a lack of smoothed stem edges. The two deep Round Mountain specimens have shorter stems and wider shoulders, hence may resemble Wells points only superficially.

25 Ray, 1938.

26 Kelley, 1947, esp. pp. 97, 99, 103-4, 107-8.

27 Ray, 1938; Kelley, 1947. It must be stressed that the latter does not regard either Baird or Taylor as the principal diagnostic dart point type of the Clear Fork Focus; this place is occupied by Nolan Beveled Stem, equivalent to Ray's “Clear Fork Dart 1” and “Clear Fork Dart 2” combined. The principal dart type for Round Rock Focus is Pedernales Indented Base, Baird and Taylor occupying a minor position in this focus too. No examples of the principal diagnostic types (Nolan and Pedernales) of the two central Texas foci were seen at the Davis site. On the other hand, relatives of the triangular Baird and Taylor types are found continuously from the Davis site westward and southwestward across Texas and into Tamaulipas, Mexico, where MacNeish associates them with his Abasolo complex (MacNeish, 1947). As one proceeds through southwestern Texas, toward the Rio Grande and Tamaulipas, the heavy triangular points become very common, whereas in central Texas and at Davis they are distinctly minor even though widely distributed.

28 Personal communication, July, 1948. The type was found at the Sandy Creek, Grelle I, and Grelle II sites in Travis County in circumstances enabling Kelley to determine association with the 20-foot terrace formation, including equivalent depositions on top of the older 40-foot terraces. At the Heffington site, Travis County, 50 specimens were found throughout the midden, a formation over the 40-foot terrace but deposited when the 20-foot terrace was developing along the river channel.

29 In keeping with the arguments in Krieger (1947rf), neither Folsom Fluted nor Clovis Fluted is used for such specimens when incomplete or uncertain. The present specimen is undoubtedly one of these two types, but can only be called “fluted.”

30 By “local” I mean that the shades and degree of luster are like those of most flint at the site; the actual quarries are unknown but there is no reason to think that the material was brought here from any distance, and a great range of flints could be found in the Neches River gravels nearby.

31 Krieger, 1946, pp. 166, 168, 206-7; Kelley, 1947.

32 Webb and Dejarnette (1942, pp. 301-6) called similar specimens “Copena points” because of association with the Copena Focus. Two village and five mound sites of this focus were recognized by them in the Pickwick Basin of the Tennessee River. Only two had these “points” or knives in any quantity: site HN°4, 15 specimens (Pl. 29); and site LUV65, 10 specimens (Pl. 207). Four sites had only one or two examples apiece. Three sites, including LUV65, yielded one fluted (“Folsom”) point apiece (Pls. 132, 207), which is interesting in view of the fluted point fragment from Davis (present Fig. 57, v and pp. 170-2). The Copena Focus is included in a “Southern Aspect of the Hopewellian Phase” by Webb and Dejarnette. Martin, Quimby, and Collier, (1947, pp. 351-3) conjecture that “The Copena Indians were closely related to Hopewell, Troyville, and late Adena peoples.” Lewis and Kneberg (1947, Pl. 2) illustrate similar blades from a preceramic Archaic culture in western Tennessee. There are Copena knives in the Spiro mound collections at Norman, Oklahoma. Harrington (1920, Pls. CX, CXI) shows them from the Mineral Springs site in southwest Arkansas, probably a Spiro Focus component. Copena knives were also found in quantity in multiple burials in the Gahagan mound on Red River, northwest Louisiana (Moore, 1912, Figs. 12, 18, 19, 21; Webb and Dodd, 1939, Pl. 27, Panel 1). At Gahagan these artifacts were fairly numerous and associated with black, highly polished vessels of the type Holly Fine Engraved, possibly acquired by trade from Davis or a related site. At Gahagan and Davis, at least, the type is at least as common as in the Pickwick Basin, if not more so. Hence we do not necessarily have a borrowing from the Copena Focus; the reverse could hold just as well.

33 Kelley (1947, p. 104) states that “Copena type blades or points” occur in non-pottery Round Rock Focus components in Coryell County, Texas. Several rock shelters in this vicinity (between the Brazos and Colorado Rivers west of Waco) have yielded both Copena knives and Wells dart points.

34 Ray, 1938. Kelley (1947, pp. 99-100) warns that their seeming prevalence in the Clear Fork Focus is unproved, “since no statistics have been published.” At any rate, I would not regard the present specimens as evidence of a specific influence from the central Texas non-pottery cultures.

35 “ … the flint disks found in Hopewellian mounds seem to be essentially similar in shape, material, and method of deposition” (Baker and others, 1941, p. 48). In their Plate 3 these authors show six disks, oval to roundedtriangular, selected from a cache of 137 “disks” in Mound 6 of the Havana group, Illinois. In this instance, burials were placed on flint caches, apparently the only recorded case of such a practice. The Havana mound disks were made of a blue-gray nodular flint, and “seem to be quite similar to those found in the various Illinois sites and to those from the Hopewell Mound Group in Ohio.” On page 17 these authors present a bibliography of flint caches. Naturally, a broader survey should be made before the Davis cache can be interpreted in cultural affiliation and chronology. For one thing, the present specimens have been called “axes” and “scrapers,” and may therefore not be strictly comparable to Hopewellian “disks.” From what I have seen of the latter, however, axes could certainly be present among them. Then, too, the trait of caching large flints, perhaps only reserves of material later to be converted into more refined artifacts, may be more significant than the precise shapes of the pieces cached.