Q Nfx AN INTRODUCTION NK TO HUDSON VALLEY PREHISTORY By William A. Ritchie, State Archeologist New York State Museum and Science Service NEW YORK STATE MUSEUM AND SCIENCE SERVICE BULLETIN NUMBER 367 The University of the State of New York The State Education Department Albany, N. Y. M437r-J157*2000 January 1958 AN INTRODUCTION TO HUDSON VALLEY PREHISTORY By William A. Ritchie, State Archeologist New York State Museum and Science Service NEW YORK STATE MUSEUM AND SCIENCE SERVICE BULLETIN NUMBER 367 The University of the State of New York The State Education Department Albany, N. V. January 1958 THE UNIVERSITY OF THE STATE OF NEW YORK Regents of the University With years when terms expire 1969 John F. Brosnan, A. M., LL. B., J. D., LL. D., D. C. L., D. C. S'., Pd. D., Chancellor - -- -- -- -- New York 1968 Edgar W. Couper, A. B., LL. D., Vice Chancellor - - - Binghamton 1963 Mrs. Caroline Werner Gannett, LL. D., L. H. D., D. H. Rochester 1961 Dominick F. Maurillo, A. B., M. D., LL. D. - - - - Brooklyn 1962 Jacob L. Holtzmann, LL. B., LL. D., D. C. L., D. C. S., Litt. D. - -- -- -- -- -- -- -- - New York 1964 Alexander J. Allan, Jr., LL. D., Litt. D. - - - - - Troy 1967 Thad L. Collum, C. E. - -- -- -- -- -- - Syracuse 1966 George L. Hubbell, Jr., A. B., LL. B., LL. D. - - - - Garden City 1958 T. Norman Hurd, B. S., Ph. D. - -- -- -- -- Ithaca 1960 Charles W. Millard, Jr., A. B. - -- -- -- -- Buffalo 1965 Chester H. Lang, A. B., LL. D. - -- -- -- - Schenectady 1970 Everett J. Penny, B. C. S. - -- -- -- -- - White Plains President of the University and Commissioner of Education James E. Allen, Jr., Ed. M., Ed. D., LL. D., Litt. D., Pd. D., L. IT. D. Deputy Commissioner of Education Ewald B. Nyquist, B. S., LL. D., Pd. D. Associate Commissioner for Higher and Professional Education Assistant Commissioner for State Museum and Science Service William N. Fenton, A. B., Ph. D. State Archeologist, State Science Service William A. Ritchie, M. S., Ph. D., D. Sc. 155 CONTENTS PAGE Preface . . . . . . . 7 Harris site (Scv 1-2) . . . . . . . . 8 Location . . . . . . . . . . . . . . 8 Excavation . . 11 Artifacts . . . . . . . . . . . . . 14 Conclusions . 16 Lotus Point site (Ctl 3-1) . . . . . . . . . 25 Location . . . . . . . . . . . . . . . 25 Stratigraphy and typology . . . . . . . . 26 River site (Coh 8-3) . . . . . . . . . . . 34 Location . . . . . . . . . . . . . . . 34 Stratigraphy . . . . . . . . . . . . . . 43 Geological interpretation . . 44 Lower horizon artifacts . . . . . . . . 45 Upper horizon artifacts . . . . . . . . . . 50 Conclusions . . . . . . . . . . . . . . 51 Hennessy site (Sdy 98-2) . . . . . . . . . 53 Location . . . . . . . . . 53 Features . . . . . . . . . . . 54 Artifacts . . . . . . . . . . 59 Artifact inventory . . . . . 59 Interpretation . . . . . . . . . . . . . 60 Bannerman site (Wtp 1-1) . . . . . . . . 62 Location . . . . . . . . . . 62 Stratigraphy and distribution of the artifacts . . . . . . . 62 Interpretation . . . . . . . . . . 66 Artifacts . . . . . . . . . . 68 Artifact inventory . . . . . . . . . . . . . 69 Conclusions . . . . . . . . . . . . . . . . 70 South Cruger Island site (Ctl 11-3) . . . . . 71 Location . 71 Stratigraphy . 72 Features . . . . . . . . 77 Artifacts . . . . . . . . . 79 Artifact inventory . . . . . . . . . 80 Comparison and conclusions . . . . . . . 81 Snook Kill site (Scv 19-2) . . . . . . . 91 Location . 91 Excavation . . . . . . . . . . . . . . 91 Artifacts . . . . . . . . . 92 Conclusions . . . . . . . . 93 General conclusions . . . . . . . . . 98 References . . . . . . . . . 110 [3] ILLUSTRATIONS PAGE Figure 1. Locations of Hudson Valley sites.......... . . . . . 6 Figure 2. Map of Harris site . 9 Figure 3. Map of Lotus Point site . Inside back cover Figure 4. Map of South Cruger Island site . . . ....................Inside back cover Figure 5. Cultural and chronological relationships of Hudson Valley sites, with reference tc> central and southeastern New York sequences . . . . . . . ,.. .Inside back cover Plate 1 a. Plate 1 b. Plate 2 a. Plate 2 b. Plate 3. Plate 4. Plate 5. Plate 6. Plate 7 a. Plate 7 b. Plate 8. Plate 9. Plate 10. Plate 11. Plate 12. Plate 13. Plate 14. A portion of the Harris site at the start of excavations in trench 1 . . . . . . . 17 Rock hollow bounding Harris site on the north . 17 Profile near midpoint of section C, trench 2, Harris site, showing- four artifacts in situ . . . 18 Profile in section D, trench 2, Harris site, showing two small hearths intruded from black soil layer into subsoil..... . . 18 Projectile points from the Harris site . . . 19 Artifacts from the Harris site . . . 20 Potsherds, polished and chipped stone artifacts from the Harris site The Lotus Point site, looking east from the Van Orden site...... 35 View from the Lotus Point site, looking west across Lotus Bay to the Van Orden site . Laying out the grid on the Lotus Point site. 36 36 Profile at midpoint of section A 14, Lotus Point site, showing soil stratigraphy . . . . . . . . . 37 Profile in section All, Lotus Point site, showing soil stratigraphy 38 Profile in section A 11, Lotus Point site, showing method of horizontal excavation by shallow levels of various soil layers 39 Artifacts from strata 1A, B and 2, Lotus Point site . . . 40 Artifacts from strata 3, 4, 5 at Lotus Point site . . . 42 Eroded beach at the River site, looking south, showing litter of natural cobbles and boulders, with interspersed Indian relics 55 Artifacts from or assigned to stratum 3 of the River site. 56 [4] AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 5 PAGE Plate 15. Chipped and ground stone artifacts from or attributed to stratum 5 of the River site . . . . . . . . . 57 Plate 16. Rough, chipped, pecked, ground and polished stone implements from or attributed to stratum 5 of the River site.... . . . 58 Plate 17. Artifacts from the southwestern area of the Ilennessy site.... . 63 Plate 18. Artifacts from the central and northeastern portions of the Hennessy site . . . . . . . 64 Plate 19. Section in trench 6, Bannerman site . 73 Plate 20. Chipped and ground stone artifacts from the Bannerman site...... 74 Plate 21. Rough stone, polished stone, antler and pottery artifacts from the Bannerman site . . . . . 75 Plate 22. Hearth on subsoil level at base of stratum 3, trench 6, section 6, Bannerman site . . . . . . 76 Plate 23 a. South Cruger Island, looking southeast from Cruger Island . 83 Plate 23 b. Burial of adult male in closely flexed position (No. 2), South Cruger Island . . . . . . . . ....! . 83 Plate 24 a, b, c. Simple hearth in subsoil of South Cruger Island site, show¬ ing various stages of excavation . . . . . 85 Plate 25. 7 Artifacts from stratum 1 (1-18) and stratum 3 (19-43), South Cruger Island site . . . . . . . . . 86 Plate 26. Chipped stone and pottery artifacts from stratum 2, South Cruger Island site . . . . . . . . . . . . 88 Plate 27. Rough, ground and polished stone implements from stratum 2, South Cruger Island site... . . . . 90 Plate 28. Chipped artifacts from the Snook Kill site . . . . 94 Plate 29. Rough, chipped, ground and polished stone tools from the Snook Kill site . . . . . . . . . . . . . . . 96 Plate 30. Fluted points from the Hudson Valley... . . . . 103 Plate 31. Points of Steubenville Stemmed type, chiefly from the Hudson Valley . . . . . . . . . . . .' . . 104 Plate 32. Artifacts from burial of Early Woodland period on Van Orden site . . . . 106 FIGURE 1. Locations of Hudson Valley sites AN INTRODUCTION TO HUDSON VALLEY PREHISTORY By William A. Ritchie, State Archeologist New York State Museum and Science Service PREFACE European exploration and settlement of future New York State occurred first in the coastal and Hudson Valley regions where, during the opening decades of the 17th century, the inevitable conflicts of interest arose between Europeans and Indians which were to lead, within a century and a half, to the virtual extinction or expulsion of the native tribes of this area. Apart from the fragmentary historical accounts relating to aboriginal customs and the problems of Indian- white relationships during this period, there is little to guide our cultural or historical reconstructions of protohistoric and early historic times.1 The scanty archeological data garnered from the ground relate mainly to more remote periods. The principal reasons for the lack of archeological sources, espe¬ cially for the early historical segment of the picture, are to be found in the large scale destruction of aboriginal sites by overlying Cauca¬ sian settlements, selectively established upon the same propitious situa¬ tions ; the prevailing small size and meager content of the Indian sites ; the considerable amount of irresponsible digging done thereon by relic collectors ; and the obvious rapidity of the acculturation process, through which the skills and contrivances of the stone age were largely replaced by European devices during the quarter century fol¬ lowing initial contact. Considering these facts, it is of some importance to the history, in the complete sense, of the State of New York, to salvage and make available as much data bearing upon the prehistory and early contact relations of the Hudson-coastal area as can still be obtained from the sources, now almost entirely limited to the systematic excavation of such aboriginal sites as have escaped total destruction. This task will require the cooperative efforts of professional archeologists and com- 1 Furman, 1845; O’Callaghan, 1850; Ruttenber, 1872; Murphy, 1875; Jameson, 1909. [7] 8 NEW YORK STATE MUSEUM AND SCIENCE SERVICE petent amateurs, imbued with a sense of problem. The first steps have already been made2, and the present report marks the continuing effort of the New York State Museum and Science Service and its collaborators in this direction. Of the seven sites herein described, ranging south from Saratoga to Dutchess Counties, two, the Harris and Lotus Point sites, were excavated by State Museum and Science Service parties in charge of the writer ; the remaining five sites, River, Hennessy, Bannerman, South Cruger Island and Snook Kill, were explored by amateurs, whose interest in the potential historical sig¬ nificance of their sites led them to request professional field guidance in the collection of their data and to making their findings available to the writer for study, analysis and publication. Well-deserved thanks are due, and heartily expressed, to all of them. They will be individu¬ ally identified in later pages. As the title indicates, this is a preliminary report for the Hudson Valley region. Many problems remain for future solution. The pre¬ historic outline drawn from the sites described herein is far from complete, especially for the later stages of Indian occupation, which were the most scantily represented. Numerous other sites have been excavated in the same valley, by both professional and nonprofessional personnel. In some cases the resulting data are no longer available or were collected with too little attention to provenience and associa¬ tions to render them useful for our purpose. In other instances, while the methods of collecting and recording were professionally compe¬ tent, the site samples usually proved too limited to be very informative. For still another group of sites, the collected and recorded data may be amenable to scientific treatment, and it is hoped that they will be made available either through independent publication or through inclusion in papers of this kind. HARRIS SITE (Scv 1-2) Location One of the largest known prehistoric sites in the Hudson Valley is situated on the John J. Harris farm, in the southeast corner of the township of Northumberland, Saratoga County (figure 1, site 2). It had long been surface hunted by Louis E. Follett of Schuylerville, who kindly reported it to the writer and assisted him in obtaining permission to excavate from Mr. Harris. Grateful acknowledgment is hereby expressed to both of these gentlemen. The New York State 2 Smith, 1950 ; Ritchie, 1952 ; Ritchie, Lenig and Miller, 1953 : Ritchie, n.d. a. has further references for Long Island. iw • r.-vre x?56v *,■’/ ■ / :eei [9] FIGURE 2. Map of Harris site AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 11 Museum and Science Service party of five, under the writer’s direct supervision, explored a portion of this site between August 6 and 24, 1951. The Harris site covers an area of about an acre, on a slight rocky point overlooking the Hudson River from an approximate elevation of 20 feet, lying between the mouth of Pecks Creek on the south and a broad hollow in the rocks (plate 1, a , b ; figure 2). The present easy descent into the hollow from the north edge of the site makes it probable that this feature gave convenient access to the river at a long rift, which was almost certainly the primary reason for the selec¬ tion of the site. Until dams were created in the river both below and above this point, extensive spring spawning runs of shad and other fish are known to have occurred. Indian fishing camps, ranging in time from the Archaic period to the historic Mohawk Iroquois, have left their evidence along the smaller rifts of Fish Creek, which drains Saratoga Lake into the Hudson some four miles south of the Harris site.3 Excavation The surface of the Harris site is nearly level at the point (plate 1, a) but rises westward, away from the river. The native rock, a Middle Ordovician formation known as Snake Hill shale, is nowhere far beneath the surface, and over parts of the site it is exposed as folded ridges. Under cultivation for many years, there has doubtless been some reduction by erosion and, at present, the upper 9-12 inches of the soil mantle comprise a definite tilth zone, dark brown in color and sandy in composition. In some places the plowshare has bitten into the soft bedrock, bringing fragments to the surface and, wherever it has invaded, it has thoroughly intermingled the cultural remains of the several groups who lived here at various intervals over a pre¬ sumably long time span. The artifacts of all ages are included in the surface collection of Mr. Follett and were present in the topsoil examined in our excavations. For purposes of analysis, therefore, the material from this disturbed zone was useful only typologically. Beneath the clear-cut plowline the color of the soil changed from dark brown to black, its texture became more compact, but its char¬ acter continued as homogeneous sandy loam, ranging from 3 to 21 inches thick, while for the most part varying from 8 to 13 inches 3 “It is a well known fact that in Colonial times, before the mills and dams were erected at Schuylerville by Gen. P. Schuyler in 1760, herring and shad in immense schools were in the habit of running up the Hudson in the spring into Fish Creek (hence the name), and thence through Lake Saratoga and the Kayaderosseras even to Rock City Falls.” (Stone, 1880, pp. 35-36). 12 NEW YORK STATE MUSEUM AND SCIENCE SERVICE thick over the dug area. It was thinnest where the contorted bedrock most closely approached the surface and deepest where slight natural hollows had existed. Its dark color seemed to indicate a high organic content, suggesting deposition under heavily forested conditions. Its manner of accumulation appears to have been by the slow accretion of forest duff, augmented by sandy wash from the border of rising ground to the west, and the downward migration of humates. The disposition of the few features found, chiefly small hearths, and of the artifact content, flint chippings, and fire-broken stones which oc¬ curred throughout the black deposit at no definable floor levels but with certain restricted local areas of greater concentration, would indicate a succession of small camps, probably of seasonal character (plate 2, a ) . The black soil mantle rested conformably upon a tan-colored subsoil, also sandy but with a higher clay content. This appears to have been laid down by the Hudson River at some indeterminate period in its postglacial history and in part removed by erosion prior to the Indians’ arrival. From a few inches to several feet of this material invested the bedrock under the site ; its surface presented slight natural undulations, while root channels and rodent burrows could readily be followed through it. Some of these contained obviously intrusive cultural articles ; such artifacts were, however, sparingly present within the first few inches of the apparently undisturbed sub¬ soil, especially toward the bluff edge overlooking the river. They were more numerous at or near the base of the black zone/in a narrow transitional band about an inch thick, not uniformly present over the site, which was interpreted as marking the inception of the vegetation cover (plate 2,b). Since this cultural material, apart from the wastage of flint chipping, comprised almost exclusively narrow-bladed projec¬ tile points of a general Lamoka-like character (plate 3, figures 51-57 ; cf. plate 2, b, arrow 1 ) , an Early Archaic level of occupation is postu¬ lated. A careful search was made of the exposed subsoil for post mold patterns which might furnish a clue to house structures or other fea¬ tures. Nothing of this kind was found, nor were there any hearths that originated on this level. Explorations were concentrated on, but not confined to, what appeared to be the central area of the site. Here a grid 65 feet long and 25 feet wide was staked off in 5-foot squares (plate 1, a; fig¬ ure 2). Twenty-one and a half squares were completely excavated well into the subsoil and an additional 5 similar squares (4 of which were connected in series) were dug some 40 feet southeast of the trench, close to the river bluff. In the total area of 662.5 square feet AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 13 explored, 620 artifacts, inclusive of potsherds and objects too frag¬ mentary to type, were recovered, together with several thousand flint chips, flakes and other rejectage of implement manufacture. Save for a few near-surface occurrences, refuse bone and shell and bone imple¬ ments were absent, probably due to prevailing soil acidity.4 The notable exception is a pair of deer atlases having cut perforations in the trans¬ verse processes, found together in section C7, 14 inches from the surface in the black layer, two inches below plowline and about 12 inches above subsoil. The purpose of these unique objects is unknown (plate 4, figure 20), Features found were few and unimpressive. They may be sum¬ marized as follows : Section A8. Near the top of the black soil layer, 12 inches from the surface, was a small saucer-shaped mass of much decayed fresh water clam shells, fish bones and bits of charcoal, together with three crumbling potsherds, one identifiable as of St. Lawrence Pseudo¬ scallop Shell type,5 This seemed to represent a hearth 14 inches in diameter and 3 to 4 inches deep. The thickness of the plow zone was here 10 inches, the depth of the black zone 20 inches, and the feature pertained to the latest of the probably three principal phases of the site’s occupation. Section A. A shallow hearth comprising a rough platform of fire-broken stones, about 2jA feet in diameter, lay in the upper portion of the black layer, here 11 to 13 inches thick, and had been partly dis¬ turbed by the plow. A small quantity of charcoal, a worked slate frag¬ ment, broad side-notched point and broken flint blade were found among the stones of this apparently late feature. Section D. Three small hearths, two shown in situ in plate 2, b, arrows 2, 3, occupied slight depressions dug into the subsoil, but trace¬ able to near the midpoint of the black layer, 8 inches thick in this section. They contained angular pieces of burned quartzite and crystal¬ line igneous rocks, with interspersed small charcoal granules. A small pebble hammerstone was embedded in hearth 1 (plate 2, b, arrow 4, and plate 4, figure 21). A few bits of carbonized deer bone were in hearth 2. 6 4 pH determinations made at the State Museum and Science Service, Office of Entomology, showed a concentration of 5.2 at 8-9 inches in the brown topsoil ; 6.1, 6.2, 6.4, 6.5, at various depths in the black layer from 12 to 19 inches, and 6.4 in a subsoil sample at 27 inches from the surface. 5 Ritchie and MacNeish, 1949, p. 103. 6 All suitable materials for radiocarbon analysis were carefully collected. 14 NEW YORK STATE MUSEUM AND SCIENCE SERVICE Artifacts As already mentioned, physical stratification was not present in the deposit ; depth analyses showed no clear-cut horizontal levels of oc¬ cupation, beyond the hint of a primary habitancy around the beginning of the presumed forest-made black layer, and this initial residency seems referable to an Early Archaic group using slender projectile points reminiscent of the Lamoka type (plate 3, figures 51-57). The distribution of this point form was, however, not confined to the sub¬ soil and basal black layer. A scatter of 14 occurred throughout the plow level to the surface. Of 28 examples from below the plowline, 25 or 89 percent were found between 11 inches above and 6 inches be¬ neath the junction of black soil and subsoil. Some of these points are illustrated in plate 3, figures 40-50. Except for two narrow-bladed, bifacially chipped artifacts from the subsoil, of quartzite and flint, respectively (plate 3, figures 59, 60), the first probably a knife, with general similarities to certain Lamoka forms, the second possibly an end or side scraper, although lacking signs of use, and without parallel in the Lamoka complex, there are no indubitable artifact associations with this point type at the Harris site. From the provenience of this slender stemmed or side-notched point on the site under discussion, and its vertical and horizontal rela¬ tionships to the other forms of points found there, it would appear that a relatively brief first period of occupation on an Early Archaic time level, probably somewhere within the temporal range of the Lamoka complex farther west in New York, was followed by the introduction of a more diversified assemblage of chipped and ground stone tools, identifiable with the Laurentian tradition and more specifically with the Vosburg complex thereof. Herein we recognize the characteristic broad-bladed stemmed (15 examples)7, broad side-notched (63), corner-notched (7), and heavy triangular points with excurvate edges and straight base (13), illustrated on plates 3, 4. Here also belong the simple end (6) and rudely stemmed end scrapers (3) ; retouched flake (4) and coarse ovate knives (5) ; the straight (2), stemmed (1) and expanded base (3) drills; the plano-convex adze (1) and adze blanks (2); gouge (1); rude plummet (?) (1), and notched crescentic bannerstones (3), pictured on plate 4. A native copper fragment, found just over the subsoil, 19 inches down in the black 7 The inventory includes all identifiable material, whole or fragmentary, found by excavation in the topsoil and black layer. Nothing suggesting the Laurentian came from the subsoil except the broad stemmed point shown in plate 3, figure 58, and the two artifacts already described at figures 59, 60, of the same plate. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 15 layer, likewise fits into the known Laurentian pattern, as does the single, thin, adzelike scraper (plate 4, figure 23), and the pebble hammerstones (7), shown as figures 21, 22 on the same plate. This succession seems to duplicate, though very faintly, the sequence of events worked out at Frontenac Island in central New York,8 and at the Harris site also the assimilation with partial persistence of the older complex is indicated. The dominant and best defined manifestation at the Harris site, to which the bulk of the artifacts, excavated and surface found, pertain, appears then to represent periodic habitation by small bands of Vos- burg people who camped now here, now there over the surface of the already more anciently visited site, depositing scanty remnants of their brief stay in weak concentrations and at sundry levels as the surface was slowly raised by wash, humus and human occupational debris, beginning during some part of the Middle Archaic time period. The final episode in the history of this site may be broadly compre¬ hended within a later stage of the Middle Woodland period, as judged by comparative pottery studies9 (see figure 5). At the Harris site a thin scatter of 208 small potsherds was found in 16 of the 26l/2 dug sections. In 10 squares it occurred only in the plow zone while in 5 others it was present also in the upper few inches of the black layer. Only in one section were sherds encountered at greater depths and in this instance extensive woodchuck disturbance was apparently responsible. The 208 potsherds, for the most part too small for typological analysis, and in 13 examples even for identification of the decorative techniques employed, comprise both rim and body fragments, the latter in the great majority. All are tempered with medium coarse to fine crushed stone aplastic, are of medium thickness and are tan, brown or gray in color. Coil or fillet juncture lines show on several, and interior channeling is not infrequent. While no basal sherds were found, a conoidal or semiconoidal bottom can be inferred by analogy with better preserved material. Rim sections are either straight or mod¬ erately everted and in some cases bear slight interior decoration. Well- marked scalloping occurs on three examples. Lips are wedge-shaped, or rounded, or more rarely flattened, and most of them carry corded - stick imprints. 8 Ritchie, 1945 ; Lathrap, 1956, pp. 15-16. 9 A single rim sherd of Jack’s Reef Corded Collar type (Ritchie and Mac- Neish, 1949, p. 107), shown on plate 5, figure 18, pertains to a still later ceramic stage of development in central New York, viz., Middle Woodland II. It seems to depict the most recent Indian indication at the Harris site. 16 NEW YORK STATE MUSEUM AND SCIENCE SERVICE The sherds seem to have been derived from relatively small vessels having either a plain smooth surface or an allover body decoration. Smoothed-over cord treatment appears on three, while five are comb- or brush-marked. Two rims, among the 51 plain sherds, are suggestive of the Point Peninsula Plain type.10 Another 50 body and rim sherds carry extensive ornamentation in fine, wavy-line stamping and are probably referable to the St. Lawrence Pseudo-scallop Shell type11 (plate 5, figures 20-26, 29, 30). Dentate-stamping occurs on 38 ex¬ amples (plate 5, figures 7-9, 12), rocker-stamping in fine dentate on 19 (plate 5, figures 5, 6), while cord-impressed embellishment is evi¬ dent on 25 sherds (plate 5, figures 13-17, 19, 27, 28). There is a single instance of fine crisscross incising (plate 5, figure 11) and two of broad parallel line incising, in one sherd combined with rocker-dentate stamping (plate 5, figures 9, 10). Apart from the single referred-to example of Jack’s Reef Corded Collar (plate 5, figure 18), the ceramic remains at the Harris site can be contained within the large pottery assemblage of the Point Penin¬ sula 2 complex of Middle Woodland I provenience, as known chiefly from sites in central New York and the St. Lawrence Valley. The route of infiltration of this ware into eastern New York was doubtless southward along Lake Champlain, as traced by the writer in site and collection surveys over this area. In the Hudson Valley such pottery diminishes rapidly from north to south, rarely appearing in the region south of Poughkeepsie. On the Harris site, artifacts tentatively associated with this pottery complex include a few thin, corner-notched points (plate 5, figure 4) ; a broken perforated stone pendant (plate 5, figure 1) ; the small deli¬ cately incised, shale pebble ornament shown on plate 5, figure 2, and perhaps the above mentioned worked deer atlases (plate 4, figure 20). Conclusions Despite its larger size and heavier artifact yield, the Harris site was, on the whole, less instructive for our reconstruction of eastern New York prehistory than the Lotus Point site, situated in the mid-Hudson Valley and described next in this report, primarily because it lacked good stratigraphy, presenting instead a relatively broad and thin scatter of indefinitely related objects and features. This site characteristic seems generally to prevail where the specific topography permitted the small bands of seminomadic food gatherers to disperse their flimsy 10 Ritchie and MacNeish, 1949, p. 103. n Ibid. Plate 1, a. A portion of the Harris site at the start of excavations in trench 1. Looking northeast across rifts of the Hudson River. The hollow shown below is at the upper left. Plate 1, b. Rock hollow bounding Harris site on the north. Probably provided access to the site from the river rifts here shown. Looking northeast. [17] Plate 2, a. Profile near midpoint of section C, trench 2, Harris site, showing four artifacts in situ. String segment at lower right above arrow 1 marks junc¬ ture of subsoil and black soil deposit. Artifact beneath, (arrow 1), embedded in subsoil, is quartzite knife, illustrated on plate 3, figure 59. Arrow 2 indicates flint discard in upper part of black soil zone (tilth layer has been stripped off). Arrow 3 refers to a chipped and partly ground slate object, at 3 inches above subsoil. A rudely chipped blade fragment, arrow 4, lies at the contact of the two soil zones. Plate 2, b. Profile in section D, trench 2, Harris site, showing two small hearths intruded from black soil layer into subsoil (arrows 2, 3). Arrow 1 indi¬ cates projectile point in subsoil, pictured in plate 3, figure 58; arrow 4, a pebble hammerstone (plate 4, figure 21), apparently associated with a hearth. Plate 3. Projectile points from the Harris site. 1-7, Vosburg Corner-Notched type; 8-12, 14-22, broad side-notched forms, 19-21, probably spearpoints, 22 may be a knife; 13, eared side-notched; 23-25, 27-34, narrow side-notched forms, 34 is probably a spearpoint; 26, 58, broad stemmed; 35-39, narrow stemmed dart (?) and spear (?) points; 40-57, slender side-notched or stemmed points, generally similar to Lamolca types; 59, biface knife (?); 60, uniface end or side scraper (?) Provenience: 1-50, black soil stratum; 51-60, subsoil Materials: all local Deepkill, Normanskill or Fort Ann flint, except 14, quartz; 50, crystal quartz; 59, quartzite; 36-38, 51, 58, slate; 10, 18, 42, Onondaga flint [19] [20] inch LQ pH /**N O-. o' eg co CO k. p-T O C y » • - Cy co CM CO O £ ^ *2 D J5 -a m u JS s X) O u X -M o fi OT .2 rt q3 ba ^ rt =2 X 13 O to d § o loo- O T~' CO u . o -J rd C x o H J2 t c P S m ■*=j u P co k* cu On pX 3 u, C -a 2 ^ ^ ? o CQ *S c 0,^4 •d © a> .2 -M u .2 *d a d *d o o u | bo cm >» o ■8 evf o a> X! » » 13 S JB S* VO CM O i a cm £* *d •S I 3 .2 k S a Oh s.§ - jq -oo 7 o N fc - - 43 - Is- m 0\ CM in o m ^ ^aj jo u • • *o 3 § hS U g ‘3 .3 i qj X O > « AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 61 Stemmed and Lamoka Side- Notched, respectively, constitute an Early Archaic horizon marker over, at least, a large portion of the New York and southern New England area, we may further suppose that this complex, ill-defined in the eastern section of its range, coexisted with the more fully elucidated Lamoka complex in central and western New York and northcentral Pennsylvania.38 As yet such diagnostic Lamoka traits as the beveled adze, the large bone and antler industry and the assemblage of stone devices for the preparation of wild vegetal foods, apparently chiefly acorns, have not been found on these eastern components, hence they are not grouped in the Lamoka complex, but are considered by the writer to pertain, with the Lamoka, to a common Early Archaic cultural tradition, as yet anonymous.39 The next comers to the Hennessy site, after an indeterminable lapse of time, were hunters equipped with paraphernalia more or less distinctive of the Vosburg manifestation of the Laurentian tradition. Their camping quarters were confined to the central area of the site, where their hearth-constructing activities may have been largely responsible for the intermixture of their remains with the older relics.40 Still more recently, beginning in the Late Archaic-Early Woodland transitional period, the Susquehanna Broad points were added to the record, followed by the scanty pottery of Middle Woodland times. Too little persists to tell us whether these influences from the south and north (or west), respectively, of the locus in question, were progressively assimilated to a continuing Vosburg cultural develop¬ ment, or whether they here denote minor mementos of briefer sojourn- ings by unrelated groups. (See figure 5.) It is well to emphasize that evidences of Vosburg continuity with the absorption of traits from other traditions seems to have occurred at the Bannerman and South Cruger Island stations, next to be discussed. 38 Ritchie, 1944, pp. 292-310; n.d. b. 39 The term “tradition,” as used throughout this report, refers to a socially transmitted set of ideas, values and customs, governing a way of life, or the behavior of a society, which is continuous without radical change from generation to generation. The sundry traditions referred to in the text, or indicated on the chart, figure 5, had, apparently, different historical beginnings and developments, for the most part lost in the obscurity of time and space. Subsequently, certain of them underwent convergence or blending to different degrees, like the anonymous Early Archaic and the Middle Archaic Laurentian traditions ; others, like the Laurentian, seem to have been elaborated or increased in complexity, with the addition, from time to time and place to place, of traits or complexes, derived by diffusion from direct or indirect contacts with other cultural groups. 40 A parallel instance of partially superimposed components of two cultures and periods of occupation, with mechanical admixture of the artifacts from both in a portion of the site, arising from the activities of the later people, is described in Ritchie, 1946, pp. 15-17. 62 NEW YORK STATE MUSEUM AND SCIENCE SERVICE BANNERMAN SITE (Wtp 1-1) Location Thus far the only radiocarbon dated site in the Hudson Valley series, the Bannerman site, is situated on the east bank of the river, one mile above the gateway to the Highlands section of the Hudson, which is formed by the opposing Storm King and Breakneck Moun¬ tains (figure 1, site 8). These towering masses of Precambrian Storm King granite flank either side of the river-cut notch through one of the long, northeast-southwest trending ridges of the Highlands exten¬ sion of the New England Upland physiographic province.41 The site lies opposite Pollepel Island, occupied by Bannerman Arsenal, hence the site designation, and is 3 miles south of Beacon, in Fishkill Township, Dutchess County. The location is a small, irregu¬ lar bench, measuring less than 100 x 50 feet, carved by erosion in the western foot of steep, rocky, Breakneck Ridge, which climbs to an elevation of 876 feet in Sugarloaf summit within half a mile to the northeast. This bench slopes gradually, then more abruptly, to the Hudson, flowing 20 feet below the site level and about 125 feet to the west, across the New York Central railroad main line. Discovered and excavated between 1950 and 1954 by James Shafer of Poughkeepsie, the site was brought to the writer’s attention in October 1950, and several times visited, in the company of Mr. Shafer, during 1950 and 1951. The writer’s participation in the excavations, at Shafer’s kind invitation, was concentrated in trenches 5 and 6. The stratigraphy was consistent throughout this portion of the site and, according to Shafer, the conditions of soil and artifact distribu¬ tion observed here were essentially constant within all portions of his grid of 49 5-foot squares, forming 9 parallel trenches, and including 1,225 square feet. Stratigraphy and Distribution of the Artifacts The following stratigraphic conditions were observed and recorded by the writer: A black humus layer, referred to as stratum 1, 1 inches thick, sterile of Indian remains, and comprising an A 1 soil horizon. A layer of yellowish sand with some fine clay content, having a finely laminated structure (stratum 2), 8 to 10 inches in thickness, and also without artifacts. It appeared to be the leached upper portion (A 2 soil) of the deeper deposit, designated stratum 3. The latter, a reddish sandy member, measuring from 24 to 36 inches in thickness, varied in texture from top to bottom. In the upper 4 to 6 inches small 41 Berkey and Rice, 1921, pp. 9-12, 22-26; Fenneman, 1938, pp. 368-370. Plate 17. Artifacts from the southwestern area of the Hennessy site. 1-20, 24, 26, narrow stemmed and narrow side-notched points; 21-23, rejects of point manufacture ; 25, 27, 28, 32-34, trianguloid and ovate knives ; 29-31, quarry blanks or rejects ; 35, notched netsinker ; 36-38, flake knives ; 39, pitted stone ; 40, notched flat pebble ornament (?) ; 41, battered quartz crystal; 42, whetstone; 43 chopper Materials : all Deepkill or Normanskill flints, except 35, 39, 40, 42, 43, sand¬ stone, or as noted above [63] Plate 18. Artifacts from the central and northeastern portions of the Hennessy site. 1. 2, body sherds of Vinette 1 ware; 3, 4, smooth body sherds; 5-7, dentate- stamped body sherds; 8, trianguloid point or knife; 9-11, Susquehanna Broad points ; 12, triangular point ; 13, trianguloid blade fragment ; 14, drill with ex¬ panded side-notched base; 15, strike-a-light ; 16, 17, 23, 25, 26, broad side-notched points; 18-22, Vosburg Corner-Notched points; 24, straight drill; 27-38, narrow stemmed points ; 39, narrow side-notched point ; 40-42, flake knives ; 43, wing section of bannerstone with perforated septum ; 44, 48, pebble hammerstones ; 45, ovate knife ; 46, plano-convex adze ; 47, chopper ; 49, combined hammer and anvilstone Materials : all Deepkill, Normanskill, Oriskany, Little Falls, and Fort Ann flints, except 30, crystal quartz ; 43, 46-49, sandstone [64] AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 65 water-worn pebbles and angular pieces of local granite were more numerous, giving this layer a coarser texture than found below. Randomly embedded in the deeper levels of generally finer sand and clay were angular cobbles or small boulders of the same granite (plate 19). Stratum 3, the artifact-bearing zone of the site, graded imper¬ ceptibly into a more compact deposit of gravel which apparently overlay the bedrock at depths greater than the 5- to 6-foot deep test pits sunk into it at various points. No artifacts were found in this subsoil layer although the deepest relics and numerous flint rejectage occurred in the indefinite transition zone. Here, too, a hearth was uncovered by Shafer and the writer at 45 inches from the surface, lying at the very base of stratum 3 (plate 22). This later to be described feature provided the charcoal for a radiocarbon date. Artifacts, chips and other rejectage of flint working, heat-shattered “cooking-stones” and occasional fragments of refuse animal bone42 were distributed in a generally sparse and more or less random man¬ ner throughout stratum 3. No pits or post molds were observed, nor were other hearths noted. At several places just below the pebbly upper portion of stratum 3, shallow lenses, approximately 2 feet across, of slightly darker soil intermixed with oyster shells, either whole or broken,43 seemed to denote a floor or habitation level (plate 19). On plotting the vertical distribution of the artifacts it appeared that this level comprised the top of a zone of maximum concentration, ranging in thickness from 10 to 16 inches, and occurring as follows : between 17 and 28 inches from the surface in trench 3, 14 and 24 inches in trench 4, 14 and 27 inches in trench 5, 14 and 30 inches in trench 6, and 12 and 27 inches in trench 7. Artifacts, however, occurred to much greater depths over most of the site. Depth analysis reveals that below the zone of maximum concentration two artifacts were found to a maximum depth of 46 inches in trench 1, 22 to 63 inches in trench 2, 29 to 48 inches in trench 3, 10 to 43 inches in trench 4, 2 to 34 inches in trench 5, and 5 to 40 inches in trench 6. Some of this material, and the more abundant chipping refuse with which it was associated, may have reached these greater depths through incorporation in pits dug from the major habitation zone. 42 Those identifiable related to the Virginia deer ( Odocoileus virginianus) . 43 Oyster beds are reported to have existed, before the river became polluted, at Haverstraw, some 20 miles below the Bannerman site (A Biological Survey of the Lower Hudson, Watershed. Biological Survey (1936), No. XI, p. 22/. N. Y. State Conservation Dept., Albany, 1937). 66 NEW YORK STATE MUSEUM AND SCIENCE SERVICE If so, the telltale pit outlines had suffered effacement through soil leaching or other means, for careful inspection of the scraped hori¬ zontal and vertical surfaces disclosed no trace of disturbance. Interpretation The most likely explanation, in the light of the total evidence, to account for the conditions observed at the Bannerman site, seems to be as follows : The gravel subsoil represents a surviving remnant of erosion of the gravel plateaus said to exist on the south side of Breakneck Mountain to an elevation of 80 feet above the river and believed to have been deposited by ice waters during the waning of the late Wisconsin (Cary?) ice tongue from the valley.44 Stratum 3, the thick, artifact-bearing bed of reddish sand, fine clay particles, rounded pebbles (derived from the old glacial gravel), angular fragments, cobbles and small boulders of native granite, has resulted from the slow flowage of these constituents down the sur¬ rounding steep hillsides, especially to the north and northeast, by soil- and rock-creep.45 Small areas of completely soil-denuded granite are now exposed on these neighboring hillsides, otherwise thinly mantled by soil materials identical with those composing stratum 3, a fact which was kindly verified for the writer by Dr. A. Scott Worthin, professor of geology at Vassar College, who visited the site and secured test samples from both loci for laboratory analysis. It seems probable that rain or melt water was the conveying agent, especially for the smaller soil and rock elements, while the heavier angular masses of granite, detached by frost action from the weathering hill side outcrops, were from time to time precipitated by gravity more rapidly down the slopes to become embedded in the growing deposit. The upper 4 to 6 inches of coarser soil materials in stratum 3 seem to reflect an acceleration of the soil-creep process, perhaps during a brief interval of greater precipitation. The archeologically sterile yellowish sandy layer, called stratum 2, probably the leached upper portion of stratum 3, represents a con¬ tinuation of the same process which was arrested or retarded by the development of a forest cover responsible for the formation of the equally sterile humus zone (stratum 1). During the earlier history of the site, after the accumulation of stratum 3 had begun by the assumed process just outlined, man was an infrequent sojourner upon this convenient, riverside camping spot, whose attractiveness may well have increased as a definite and larger 44 Peet, 1904, pp. 439-441. 45 Sharpe, 1938, pp. 21-33. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 67 bench was produced by aggradation. Occupational vestiges left on or near the surface by these visits would gradually become buried to progressively greater depths. On one of these earlier visits, a rough hearth of stones was con¬ structed, upon which a campfire was kindled, no doubt to heat the stones for roasting game or fish, by a small potteryless group of hunters and fishers. Shafer and the writer uncovered this feature in section 6, trench 6, at a depth of 45 inches from the surface, lying at the base of stratum 3, in contact with the gravel subsoil. While there was no indication of intrusion from above, it seems probable that a shallow excavation had been made in the then existing ground sur¬ face, now represented by the lower few inches of stratum 3, in keeping with the custom of Early Archaic people in the New York area.46 To form their rude cooking structure, the hunters had assembled 16 cobbles of graywacke, granite and quartz, all doubtless ready to hand. Two of them had become broken in situ from the heat of the fire, which had resulted in a 5-inch accumulation of intermixed char¬ coal and soil particles enveloping the stones. Fully uncovered, the spread of stones measured 19 x 24 inches, the area of charcoal con¬ centration (under the 6-inch rule shown in plate 22), 10 x 14 inches. Unhappily, no direct clue to the cultural identity of the makers was left behind. A single flint flake occurred among the stones, but in the surrounding sections of this and adjacent trenches, artifacts attributable to Laurentian provenience were found at comparable depths. The carefully gathered charcoal from this unique feature on the site, and a sample of oyster shells from the upper part of stratum 3, were submitted by the writer, through the good offices of Professor James B. Griffin, to the University of Michigan Memorial-Phoenix Project Radiocarbon Laboratory, under the direction of Professor H. R. Crane. Dr. Crane’s much-appreciated report on the charcoal sample (M-287) indicates that this campfire burned here on the shore of the Hudson River, some 4,480 ± 300 years ago, or approxi¬ mately 2524 B. C. While this date is the oldest yet obtained for an Indian site in eastern New York, it is still considerably above the 3500-3000 B. C. antiquity obtained by the C14 method on hearth charcoal from central New York Archaic sites.47 It seems, therefore, quite certain from this fact, as well as from the character of the artifact complex at the Bannerman site, that groups on the Archaic level of culture were established in the Hudson Valley well before 2500 B. C. 46 Ritchie, 1945, p. 6. 47 Libby, 1955, pp. 92, 93; Ritchie, 1951, pp. 31-32. 68 NEW YORK STATE MUSEUM AND SCIENCE SERVICE Artifacts The entire cataloged collection of artifacts from the Bannerman site was generously presented by Mr. Shafer to the New York State Museum in 1955. It comprises 261 specimens, ail of chipped, rough or ground stone, except three of bone, and 12 plain potsherds from a single vessel. A representative series is illustrated on plates 20 and 21. Analysis of the material in terms of more or less well-known cultural assemblages from other sites in and out of the general area, places nearly 90 percent within the range of traits regarded as char¬ acteristic of the Laurentian tradition. These include a variety of broad-bladed point forms, the ulo and bannerstone, as the most diag¬ nostic. Such other established traits as the gouge and plummet do not appear in the Bannerman series. Of the total of 139 projectile points, 63 percent conform to pat¬ terns we have come to regard as of Laurentian provenience (plate 20, figures 6-22), while 28 percent fall within the category of slender- bladed, stemmed or side-notched points, usually present on but con¬ stituting a minor proportion on such sites, while comprising the total or major point forms on sites of the Lamoka culture (plate 20, figures 1-5). Of the remainder, the bifurcated base form (plate 20, figures 23, 24) has not satisfactorily been attributed, either temporally or cul¬ turally;48 while the Orient Fishtail49 (figures 31, 32), Perkiomen Broad50 (figure 33), Susquehanna Broad51 (figure 34), lobate stemmed (figures 29, 30), and Meadowood Side-Notched52 points (figure 35) have associations in several more or less well-known cul¬ tures of the Early Woodland period.53 The distribution within stratum 3 of these sundry point forms, as nearly as can be determined, is not such as to indicate clearly their successive arrangement in a temporal sequence on this site. Never¬ theless, 100 percent of the Early Woodland period forms were found in the upper half of the stratum, while conversely, 58 percent of the slender Lamoka-like specimens came from the lower half. The Laurentian forms, on the other hand, were almost evenly distributed throughout. 48 Ritchie, 1940, p. 29, PL XIII, figures 17, 18. 49 Ritchie, n.d. a. 59 Witthoft, 1953, pp. 16-20. 51 Ibid., pp. 4-5. 52 Ritchie, 1944, pp. 122, 125-126; 152-160. 53 A formal description of these points will appear in a subsequent publication by the writer. All are illustrated in the accompanying plates and have been pic¬ tured and discussed in various of the writer’s reports herein cited. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 69 Artifact Inventory Chipped stone Projectile points.54 Narrow stemmed (50, plate 20, figures 1-3, 8-10), narrow side-notched (37, plate 20, figures 4, 5, 11-13), Vos- burg Corner-Notched (5, plate 20, figures 6, 7), long-stemmed (3, plate 20, figure 14), broad-stemmed (18, plate 20, figures 15-18), broad side-notched (13, plate 20, figures 19-22), bifurcated stemmed (3, plate 20, figures 23, 24), lozenge-shaped (4, plate 20, figures 25, 26), eared side-notched (3, plate 20, figure 27), eared triangular (2, plate 20, figure 28), lobate stemmed (5, plate 20, figures 29, 30), Orient Fishtail (6, plate 20, figures 31, 32), Perkiomen Broad (2, plate 20, figure 33), Susquehanna Broad (1, plate 20, figure 34), Meadowood Thin Side-Notched (2, plate 20, figure 35), equilateral triangular (2, plate 20, figure 36). Knives. Ovate (4, plate 20, figures 37, 38), trianguloid (4, plate 20, figure 39). Scrapers. End (20, plate 20, figures 40, 41), stemmed end (6, plate 20, figures 42, 43, 45), bifurcated base end (1, plate 20, figure 44), side (3, plate 20, figure 46). Drills. Stemmed (1, plate 20, figure 47), expanded base ovate, rectangular, side-notched, corner-notched ( 1 of each, plate 20, figures 48-51), indeterminate through breakage (3). Ground and polished stone Ulos. With ridged handle (2, plate 20, figures 52, 53), plain (2, plate 20, figure 54), in process by chipping and grinding slate frag¬ ment (9, plate 20, figures 55, 56). Bannerstone. Winged, with grooved septum (1, plate 21, figure 1). Plano-convex adze. Thick, with pointed poll (1, plate 21, figure 2). Rough stone Choppers. Flat form, various shapes (7, plate 21, figures 3-4), thick, boulder spalls with trimmed edge (4, plate 21, figure 5). Hammerstones, anvilstones, netsinkers. Pebble hammerstones (3, plate 21, figure 6), laterally grooved and faceted hammerstone (1, plate 21, figure 7), combined hammer and anvilstones (4, plate 21, figures 8, 9), shallow mortar and anvilstone (1, plate 21, figure 10), notched pebble sinkers (5, plate 21, figures 11, 12), double-grooved sinker (1, plate 21, figure 13), grooved pebble, flat on one side, may be netsinker or crude bannerstone (1, plate 21, figure 14). 54 Type nomenclature will be employed, insofar as developed, otherwise the older descriptive terms will be used. Quantification is given in parentheses, together with plate reference. 70 NEW YORK STATE MUSEUM AND SCIENCE SERVICE Steatite pots. Fragments of two cooking vessels (plate 21, figures 15, 16). Figure 15 from trench 6, section 7, depth 19 inches; figure 16 from trench 4, section 3, depth 20 inches from surface. Bone implements Antler tine flakers (2, plate 21, figures 17, 18), split beaver incisor knife or scraper (1, plate 21, figure 19). Pottery Seventeen, combined into 12, plain, smooth exterior- and interior¬ surfaced body sherds, of sandy paste, heavily tempered with coarse, angular particles of crushed quartzite. Ware is compact, hard (3-3.5 on Moh’s scale) and well fired. Sherds show coil or fillet lines, are tan colored on both surfaces, and light gray on the interior of the paste (plate 21, figures 20-23). They were found together in trench 4, section 2, at 15 inches from the surface. Conclusions Various explanations may be offered to account for the observed conditions at the Bannerman site, not all of equal plausibility. The first supposes an Early Archaic period of occupation by a group or groups using slender Lamoka-like points, a situation analogous to that de¬ scribed for stratum 4 at Lotus Point, the basal level (#3) at Harris, and the first component at Hennessy. The culture of this group would probably be related to the anonymous Early Archaic tradition to which belonged the Lamoka complex farther west in New York. Fortuitous physical intermixture of this older with later materials resulted from displacement by soil-creep and subsequent usage of the site by groups with a dominant Middle Archaic Laurentian cultural tradition. The second hypothesis involves intermittent, brief residency on the site, over a fairly long period of time, by groups with sundry con¬ trasting cultural assemblages. In all likelihood such groups would have been coresident in the area, and the situation envisioned in this hypothesis would essentially conform to our Type A 1 classification of culture contact situations, in which two unlike cultures coexist in the same area with retention of their respective cultural identities and little trait exchange.55 The third explanation rests upon the assumption that the small groups of wandering food gatherers, who periodically camped here over a period of many centuries, were from their first occupancy of the site part of a culturally mixed band resident within the general 55 Lathrap, 1956, p. 9. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 71 area, and that their cultural assemblage was already a composite of traits forming a Converging Tradition,56 whose essential roots lay in the aforementioned Early and Middle Archaic traditions. The complexity of this basic cultural assemblage was from time to time increased, especially through the addition of elements from Early Woodland cultures of surrounding areas. (See figure 5.) On the whole, the recorded data fit best into this latter assumption, which would seem to apply with equal force to the major portion of the South Cruger Island site, next to be described. SOUTH CRUGER ISLAND SITE (Ctl 11-3) Location Along the east shore of the Hudson River, one and a half miles below the village of Tivoli, in Red Hook Township, Dutchess County, a broad, marshy- based peninsula, with expanding, high, rocky ter¬ minus, juts into the river between North and South Bays (figure 1, site 7). Once an island in the drowned valley of the Hudson, it has become land-linked by natural and artificial accumulations to provide a roadbed from the mainland. Sometime after its purchase in 1835 by the Cruger family, it was attached by stone causeways to two small nearby islets, the southernmost of which, separated by about 500 feet of shallow water, is the locus of the South Cruger Island site, which covers the approximately one-fourth acre comprising the southern tip. An apparently similar site, situated on the north end of the main island, was partially explored by the Hudson Valley Archeological Survey of Vassar College, 1939-40, under the direction of Dr. Mary Butler. The island group consists of shale and grit of the Upper Nor¬ man skill formation (Middle Ordovician), which outcrops as ridges along both the eastern and western sides. The strata of these ridges dip about 75° to the east, with strike a little east of north. The slopes are mantled with sand and gravel deposits of postglacial age, capped with a layer of recent humus. Both these soil layers occur on the site, which occupies a relatively low, sandy, flat area between rock ridges (see plate 23, a and figure 4). Beginning in March 1947, a series of 10 parallel trenches, 6 to 9 feet in width, was excavated across the site by James Shafer of Poughkeepsie, assisted at various times by Almon Beneway and Harold Fuller, both of Poughkeepsie, and John Losee of Red Hook, all members of the Mid-Hudson Chapter, New York State Archeo¬ logical Association. In the spring of 1950, the writer, responding to 56 Thompson, 1956, p. 44. 72 NEW YORK STATE MUSEUM AND SCIENCE SERVICE Mr. Shafer’s kind invitation, participated briefly in the excavations and confirmed conditions recorded for this site. Stratigraphy (See figure 4) Stratum 1. This consisted of a black humus-filled sand with a thin duff layer (A0-A1 soil), containing in some areas, particularly near the base, large quantities of trampled unionid shells, and varied in thickness from 2 inches in trenches 8-10, where the shell constituent was lacking, to 10-12 inches toward the southern end of the site. This stratum was apparently formed like stratum 1 at Lotus Point, which it also resembles in archeological content, by the concurrent accumulation of organic refuse of human habitation and vegetation under forested conditions, with the downward migration of humic acids or humates. All the Iroquoian type sherds, the pipes, bone imple¬ ments, “spool-stones” and a proportionally large number of the thin, equilateral triangular arrowpoints came from this stratum which faintly records cultural evidences of the late prehistoric to the con¬ tact period, as attested by a single triangular point of sheet brass (plate 25, figure 10). Stratum 2. This, the major relic-producing horizon, consisted of reddish brown sand rather heavily charged with small water- rounded pebbles, hinting at a beach formation. It was rich in flint chips and fire-broken stones but yielded very little shell and no bone except fragments in a calcined state. Many flat, locally derived rocks, were present at the base of the deposit, which varied in thickness over the site, from approximately 2 feet at the lower or southern end to 18 inches in the northern section. A definite gray-brown podzolic soil profile could be traced in the upper portion of this stratum, where were found two body sherds of a Vinette 1 type pot (plate 26, figures 7, 8), a rim sherd with rounded lip, ornamented with rude, vertical, broad trailed lines (figure 6), and a small comb-marked body sherd. This comprised the only pottery produced by this layer. Although a beach formation was suggested by the nature of the deposit and by the wave-worn surfaces of certain of the flint objects, the island is not known to have been flooded in historic times and the elevation of the occupied area is at present some 9 feet above flood tide in the Hudson. Had occasional inundation and alluviation taken place during a higher stage of river level, hearths found in this stratum might have shown disturbance, but this was not the case. Probably the primary source of the soil materials composing stratum 2 was the similar mantle investing the adjacent rocky ridges and de¬ rived therefrom by the same process of gradual rock- and soil-creep [73] t Plate 20. Chipped and ground stone artifacts from the Bannerman site. 1-3, 8-10, narrow stemmed points; 4, 5, 11-13, narrow side notched points; 6, 7, Vosburg Corner-Notched points; 14, long-stemmed point; 15-18, broad stemmed points; 19-22, broad side-notched points; 23, 24, bifurcated stemmed points; 25, 26, lozenge-shaped points ; 27, eared side-notched point ; 28, eared triangular point ; 29, 30, lobate stemmed points; 31, 32, Orient Fishtail points; 33, Perkiomen Broad point ; 34, Susquehanna Broad point ; 35, Meadowood Thin Side-Notched point ; 36, equilateral triangular point ; 37, 38, ovate knives ; 39, triangular spearpoint ; 40, 41, end scrapers; 42, 43, 45, stemmed end scrapers; 44, bifurcated base end scraper ; 46, side scraper ; 47, stemmed drill ; 48, drill with expanded ovate base ; 49, drill with expanded rectangular base; 50, side-notched drill ; 51, corner-notched drill ; 52-53, ulos with ridged handle ; 54, plain ground ulo ; 55, 56, ulos in process by chipping and grinding Materials: all Deepkill and Normanskill flints, except 2, 6, 18, 33, slate; 4, quartz; 8, 42, 43, quartzite; 25, argillite; 31, 34, tan jasper; 35, 38, Onondaga flint; 37, Oriskanv sandstone flint [74] Plate 21. Rough stone, polished stone, antler and pottery artifacts from the Bannerman site. 1, grooved bannerstone ; 2, plano-convex adze ; 3, 4, flat choppers ; 5, chopper from boulder spall ; 6, pebble hammerstone ; 7, laterally grooved and faceted hammerstone ; 8, 9, combined hammer and anvilstones ; 10, shallow mortar and anvilstone; 11, 12, notched netsinkers; 13, double-grooved sinker; 14, grooved pebble, flat on one side; 15, 16, steatite pot fragments; 17, 18, antler tine flakers ; 19, split beaver incisor knife or scraper ; 20-33, plain potsherds Materials : all sandstone, except 2, syenite gneiss and 5, quartzite, or as indi¬ cated above [75] Plate 22. Hearth on subsoil level at base of stratum 3, trench 6, section 6, Bannerman site. Radiocarbon date obtained on charcoal from this feature. Knife marks base of humus zone (stratum 1). [76] AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 77 described for the Bannerman site. Its generally like character to the latter suggests the common origin of both in the local glacial deposits of sand and gravel (see page 66). It is also possible that on both the South Critger and Bannerman sites, dwelling areas were intentionally sanded for reasons of sanita¬ tion and to insure dry floors, an expedient demonstrated at a prob¬ ably contemporaneous Archaic site in central New York.57 Stratum 3. This compact, light tan colored, apparently water- laid sand, constituted the subsoil at the site. Considerable, slab-rock from the surrounding ridges occurred in the upper 12-15 inches, together with heat- shattered rocks and a thin scatter of artifacts. These did not seem to be directly associated with any of the features (pits, hearths and graves) which were intrusive into this horizon and the means of their incorporation are frankly unknown. Two deeply weathered, stemmed argillite points, the only examples from the site, were found in this stratum at depths of 42 and 30 inches from the surface, respectively (plate 25, figures 34, 40). Features (See figure 4) Three kinds of features were noted at the South Cruger Island site, viz., pits, hearths and burials, the latter unique for the sites considered in this report. Pits. Described by the excavators as fire or cooking pits, these pertained to all three horizons of the site and totaled 13. Only one (No. 1) opened from stratum 1 and penetrated deeply into stratum 2, where its outlines were clearly visible. Circular, with U-shaped bottom, it measured 23 inches in oval diameter and 34 inches in depth. No trace of fire was contained and this feature may have been a food cache pit subsequently used as a refuse repository, such as commonly occurs on Late Woodland sites.58 In this instance the black soil fill produced numerous uni on id valves, deer and bird (duck?) bones, and two triangular arrowpoints (plate 25, figures 13, 14). Seven other pits (Nos. 2, 3, 6, 8, 9, 10, 13) of smaller size had their origins at various levels within stratum 2. In this group the oval diameter ranged from 16 to 20 inches, the depth from 12 to 20 inches. The sole recorded contents consisted of burned unionid shells. Still smaller, but otherwise similar pits (Nos. 4, 5, 7, 11, 12) oc¬ curred intruded into the subsoil (stratum 3) from, apparently, the lower portion of stratum 2. The size range here was from 10 to 16 57 Ritchie, 1940, pp. 8-9. 58 Ritchie, Lenig and Miller, 1953, pp. 8-12, 32-33. 78 NEW YORK STATE MUSEUM AND SCIENCE SERVICE inches in diameter and 12 to 15 inches in depth. In three cases (Nos. 5, 7, 11) a basal deposit of gray ash with interspersed calcined bone fragments and heat-reddened surrounding sand seemed to iden¬ tify them as cooking pits with known parallels in Archaic sites of New York and southern New England. There is also a temptation to regard the whole series of pits on this site as primarily designed for food storage, with the further supposition that the progressive size increase through time marked the change from wild vegetal resources in the lower levels to the use of corn and beans in stratum 1. Hearths. Most of these were simple fireplaces consisting of shallow basin-shaped hollows, some 12 to 15 inches in diameter and 4 to 6 inches in depth, in which a fire had been kindled upon a rough platform of cobbles, of which many were heat-shattered in situ. A comparable structure has already been described from the Bannerman site (plate 22 and pages 65, 67). A few larger examples seemed to represent rather carefully paved flat or depressed circular areas. Three occurred in stratum 2 (Nos. 3, 4, 6) and five in stratum 3 (Nos. 1, 2, 5, 7, 8). An example of the first kind, excavated by Mr. Shafer and the writer, is shown in various stages of dissection in plate 24, a, b, c. From this feature, which intruded into the subsoil from the base of stratum 2, a quantity of diffused particles of sand and charcoal was recovered for possible radiocarbon dating. Burials. During 1947, while excavating trenches 3 and 4 on the west side of the site, Shafer and his friends came upon four closely grouped burials in stratum 3, each occupying a grave traceable by wall outlines for some distance into stratum 2. In every instance heavy stone slabs had been laid over the filled grave fossa. No mortuary offerings were present, but artifacts occurred in the fill of two graves. The mutual proximity of the remains and the uniformity of burial traits evince a definite cemetery and burial complex, apparently of the Laurentian inhabitants of the site, a cultural characteristic of the New York Archaic previously recorded only at Frontenac Island.59 Burial 1. Grave dimensions 45 x 22 inches, depth (from surface) 36 inches. Adult male (?) skeleton, closely flexed on left side, facing west, head to south. Very poor condition. A few inches east of the grave and 6 inches deeper in stratum 3, apparently unassociated with the burial, were found a greatly patinated purple argillite stemmed point (plate 25, figure 34) and a large triangular point, probably for a spear, of drab, lusterless quartzite (plate 25, figure 39). 59 Ritchie, 1945. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 79 Burial 2. Located about 4 feet north of No. 1. Grave dimensions 30 x 20 inches, depth (from surface) 38 inches. Adult male skeleton, closely flexed on right side, facing east, head to south. Knees and hands before face, feet against pelvis. Fair to poor condition, with numerous pressure fractures. A celt with broken poll (plate 27, figure 2), a triangular knife and a broken corner-notched point were found at scattered points in the fill. (See plate 23, b .) Burial 3. Found 3 feet east of No. 2. Grave dimensions 30 x 20 inches, depth 40 inches from surface. Adult skeleton (sex ?), closely flexed on left side, facing west, head to south, hands before face. Condition very poor. Chopper (plate 27, figure 14) from near top of grave fill just under stone slabs. Burial 4. Lay about 2 feet east of No. 3. Grave dimensions 18 x 16 inches, depth from surface 32 inches. Skeleton of infant, head to south, too badly disintegrated to determine position. Artifacts The industrial content of the principal occupation zone at South Cruger Island comprises predominantly chipped flint artifacts, for the most part projectile points. Polished stone implements are relatively and absolutely scarce ; rough stone implements somewhat more numer¬ ous. The total complex, as in the sites herein described, and in Laurentian stations in general, is evincive of a food-gathering econ¬ omy, with major emphasis on hunting, a lesser reliance on fishing and little dependence on wild vegetal foods. The direct reversal of this sequence would afford a more accurate picture, by current knowledge, of the Lamoka economy.60 Another typical feature of this complex is the absence of pottery, the few sherds derived from the top of stratum 2 pertaining either to brief contacts by the later inhabitants with Early Woodland period cultures or to very temporary use of the site by immigrant groups during this period. The occurrence only of calcined bone and the poor state of preservation of the human remains argue for the decay of the products of a bone industry, probably owing to soil acidity, but the testimony of the other sites herein discussed, where soil condi¬ tions were obviously more propitious, sustains the prevailing evi¬ dence from central New York showing the relatively and absolutely feeble bone industry of the Laurentian as contrasted with both earlier and later cultures in the central, northern and western New York 60 Ritchie, 1940, pp. 46-47, 79-85; 1932, pp. 98-110; n.d. b. 80 NEW YORK STATE MUSEUM AND SCIENCE SERVICE areas.61 In eastern and coastal New York, however, for inexplicable reasons, a weak industry in bone and antler seems to characterize sites of all cultures prior to Late Woodland times.62 Artifact Inventory Chipped stone Projectile points, stratum 1. Narrow side-notched (4, 63 plate 25, figures 7-9), equilateral triangular (5, plate 25, figures 13-15). Stratum 2. Vosburg Corner-Notched (4, plate 26, figures 46, 47, 49), narrow stemmed (20, plate 26, figures 12-19, 24), narrow side- notched (21, plate 26, figures 20-22, 25-36), broad stemmed (7, plate 26, figures 23, 45, 61), broad side-notched (32, plate 26, figures 37-44, 48, 54, 55), broad corner-notched (3), eared triangular (3, plate 26, figure 11), lobate stemmed (1), Orient Fishtail (6, plate 26, figures 50-53), equilateral triangular (12, plate 26, figures 1-3, 10), isosceles triangular (7, plate 26, figures 4, 5, 9), stemmed spearpoints (2, plate 26, figures 69, 70). Stratum 3. Vosburg Corner-Notched (3, plate 25, figures 19, 20), broad stemmed (3, plate 25, figures 34, 40), broad side-notched (4, plate 25, figures 21-23) , isosceles triangular ( 1, plate 25, figure 24) , isosceles triangular, chipped on one side (1, plate 25, figure 30), triangular spearpoint (1, plate 25, figure 39). Knives, stratum 2. Ovate (4, plate 26, figures 71-74), trianguloid (11, plate 26, figures 67, 68), flake (2, plate 26, figure 60). Stratum 3. Trianguloid (1), flake (6, plate 25, figures 27, 28, 31), prismatic flake (1, plate 25, figure 29). Scrapers, stratum 2. Simple end (2, plate 26, figure 56), stemmed end (3, plate 26, figure 57). Stratum 3. Simple end (5, plate 25, figures 35, 36), side (1), triangular end (1, plate 25, figure 32). Drills, stratum 2. Expanded base, rectanguloid, corner-notched, “Y” based, “T” based, and indeterminate through breakage (1 of each, plate 26, figures 62-66). Stratum 3. Expanded ovate base (2, plate 25, figure 25, 26). Strike-a-lights, stratum 2. (2, plate 26, figure 58). Ground and polished stone, stratum 1. Pendant (1). Stratum 2. Ulo, plain (1, plate 27, figure 5) grooved gouge ( 1, plate 27, figure 3), gouge ( 1) , grooved plano-convex adze (1, plate 27, figure 1), cylindrical pestles (2, plate 27, figure 16), celt (1, plate 27, figure 2), stemmed plummet (1, plate 27, figure 4). Ritchie, 1932, 1940, 1944. 62 Ritchie, 1944, pp. 102-106; Smith, 1950. 63 Indicates frequency in all cases. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 81 Rough stone Choppers, stratum 2. Flat form, various shapes (6, plate 27, figures 6, 10, 14), thick boulder spall with trimmed edge (1). Hammerstones, anvilstones, netsinkers, stratum 1. Notched pebble netsinkers (3), “spool-stones5’ (4, plate 25, figures 11, 12). Stratum 2. Flint pebble hammerstone (1, plate 27, figure 13), pitted hammerstones (2, plate 27, figure 9), faceted pebble hammer- stone (1, plate 27, figure 8), notched pebble netsinker (1), abrading stone (1, plate 27, figure 11), notched axe (1, plate 27, figure 15). Stratum 3. Pebble hammerstones (3, plate 25, figures 33, 37). Native copper Stratum 2. Awl (1, plate 26, figure 59). Bone implements Stratum 1. Polished splinter awls (6, plate 25, figures 16, 17), rough splinter awls (4, plate 25, figure 18). Pottery Stratum 1. Oak Hill Corded (1, plate 25, figure 2), Cayadutta Incised (6, plate 25, figures 3-5), Iroquoian plain body sherds (60). Stratum 2. Trail-decorated rimsherd with rounded lip (1, plate 26, figure 6), comb-marked body sherd (1), Vinette 1 body sherds (2, plate 26, figures 7, 8). Clay pipes Stratum 1. Bowl fragment of a plain trumpet (1, plate 25, figure 6), elbow fragment with vertical incised lines on bowl (1), obtuse angle elbow with punctated plat and line decorations (1, plate 25, figure 1). Trade goods Stratum 1. Equilateral sheet brass arrowpoint, imperforate (1, plate 25, figure 10). Comparisons and Conclusions In November 1955, the cataloged collection of South Cruger Island artifacts, inventoried above, was generously donated to the New York State Museum by Mr. Shafer. The distributional analysis made by the writer reveals no significant horizontal variation and only a slight percental frequency difference as a function of depth, within the formal projectile point series below stratum 1. The assemblage from these deeper layers of the site, as already indicated, conforms on the whole to the Vosburg complex of the Laurentian tradition, as pre¬ viously and herein described.64 (See figure 5.) 64 Ritchie, 1944, pp. 257-259. 82 NEW YORK STATE MUSEUM AND SCIENCE SERVICE As stated in the description of the upper soil layer, Iroquoian style potsherds were present in stratum 1, and these can readily be identified with early and middle prehistoric Mohawk types, viz., Oak Hill Corded (plate 25, figure 2) and Cayadutta Incised (figures 3-5). The 7 rim and 60 body sherds of this group were recovered from depths ranging between 2 and 6 inches of the surface. The bowl fragment of a plain pipe (figure 6), from 4 inches beneath the surface, is culturally congruent with this pottery as are the equilateral triangular arrowpoints with concave base (figures 13-15), which seem, par¬ ticularly at this site, to constitute the persistence of an earlier trait. The obtuse-angle elbow pipe (figure 1 of the same plate), found at a depth of 10 inches in stratum 1, is a rare item for eastern New York. It is well modeled and fired from a finely grit-tempered clay and is a rich brown in color. The zoned and platted decoration on the bowl is executed in fine punctations. Its closest parallels are to be found in the Canandaigua complex of the Owasco culture, of Late Woodland II times.65 The “spool-stones” (plate 25, figures 11, 12) and narrow side- notched points (figures 7-9) can stratigraphically be correlated with the 1 B layer at the Lotus Point site (see plate 11, figures 15, 16, 23, herein). It would therefore seem that stratum 1 at South Cruger Island closely equates both chronologically and culturally with stratum 1 A and B at Lotus Point, farther north along the Hudson, and that their physical composition reflects a humid forested environment at this time. (See figure 5.) Moreover, scanty though they be, the archeo¬ logical vestiges attest to the persistence into the recent prehistoric period of a much older pattern of life, suggested by the general evidence presented by the sites already described and discussed in this report. For the Late Woodland period, represented by stratum 1 at South Cruger Island and Lotus Point, we feel secure in postulating an occupation of this portion of the Hudson Valley by Wappinger and Mahican tribal groups, respectively,66 who still conducted, after the manner of their probable regional ancestors of earlier Woodland and Archaic times, a seasonal cycle of economic activities along the river and in the adjacent forested upland country, wherein small, probably winter hunting camp sites are known, both from rock-shelters and from the vicinity of small streams which empty into the Hudson. 65 Ritchie, 1936, p. 38. 66 Ritchie, 1956, Figures 1, 2. Plate 23, a. South Cruger Island, looking southeast from Cruger Island. “X” indicates site location. (Photograph by James Shafer) Plate 23, b. Burial of adult male in closely flexed posi¬ tion (No. 2), South Cruger Island. (Photograph by James Shafer) [83] Plate 24, a, b, c. Simple hearth in subsoil of South Cruger Island site, show¬ ing various stages of excavation. (Photographs by John G. Broughton) [85] ■. i [86] Plate 25. Artifacts from stratum 1 (1-18) and stratum 3 (19-43), South Cruger Island site. 1, decorated pottery pipe; 2, rim sherd Oak Hill Corded; 3-5, rim sherds Cayadutta Incised ; 6, plain pottery pipe bowl ; 7-9, narrow side- notched points; 10, sheet brass arrowpoint ; 11, 12, “spoolstones” ; 13-15, trian¬ gular points; 16-18, bone awls; 19, 20, Vosburg Corner-Notched points; 21-23, broad side-notched points ; 24, triangular point ; 25, 26, ovate base drills ; 27, 28, 31, flake knives; 29, knife from retouched prismatic flake; 30, triangular point chipped on one side only ; 32, triangular end scraper ; 33, 37, pebble hammer- stones ; 34, 40, broad stemmed points; 35, 36, end scrapers; 38, 41-43, quarry blanks or unfinished implements ; 39, trianguloid knife Materials: all Deepkill and Normanskill flints, except 11, 12, sandstone; 33, quartzite ; 34, 40, argillite ; 37, ? ; 38, slate, or as indicated in text [87] Stratum 2 [88] Plate 26. Chipped stone and pottery artifacts from stratum 2, South Cruger Island site. 1-5, 9-11, triangular points; 6, rim sherd of trail-decorated pottery; 7, 8, body sherds of Vinette 1 ware ; 12-19, 24, narrow stemmed points ; 20-22, 25-36, narrow side-notched points; 23, 45, 61, broad stemmed points (61 may be a knife) ; 37-44, 48, 54, 55, broad side-notched points; 46, 47, 49, Vosburg Corner-Notched points ; 50-53, Orient Fishtail points ; 56, end scraper ; 57, stemmed end scraper ; 58, strike-a-light ; 59, native copper awl ; 60, flake knife ; 62-66, drills with expanded rectanguloid, expanded, T-shaped, Y-shaped and corner-notched bases, respectively ; 67, 68, trianguloid knives ; 69, 70, stemmed spearpoints; 71-74, ovate knives Materials : all Deepkill or Normanskill flints, except 5, 11, slate ; 21, 41, quartz ; 33, 74, quartzite; 52, jasper, or as noted above [89] Plate 27. Rough, ground and polished stone implements from stratum 2, South Cruger Island site. 1, grooved plano-convex adze; 2, celt; 3, grooved gouge ; 4, plummet ; 5, ulo fragment ; 6, 10, 14, choppers ; 7, 8, 13, pebble hammer- stones (8 is faceted); 9, pitted hammerstone; 11, abrading stone; 12, steatite pot fragment with lug; 15, notched axe; 16, cylindrical pestle Materials: 1, gabbro ; 2, 9, 11, 14-16, sandstone; 3, syenite; 4, hornblende gneiss ; 5, slate ; 6, quartz ; 7, 10, quartzite ; 8, granitic gneiss ; 12, steatite ; 13, flint [90] AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 91 SNOOK KILL SITE (Scv 19-2) Location The discovery of this site and its investigation since 1953 are to be credited to Mr. and Mrs. William H. Rice of Gansevoort, N. Y., members of the Auringer-Seelye Chapter, New York State Archeo¬ logical Association. In January 1956 they kindly made the collec¬ tion and notes resulting from their work available to the writer for study and publication, a courtesy hereby gratefully acknowledged. The site is situated upon a fairly level point of land formed by a sharp northward bend of the Snook Kill, about two-fifths of a mile from its juncture with the Hudson River, in Moreau township, Sara¬ toga County (figure 1, site 1). Vestiges of occupation, in the form of flint rejectage, artifacts and heat- shattered rocks, are thinly scat¬ tered over some four acres of a sandy terrace, lying at an approximate elevation of 25 feet above the creek level. The prevailing soil of the surrounding locality varies from a clay loam to a heavy clay, sug¬ gesting that the well-drained, warm sand was one factor in the selection of this habitation area. A probable second factor was the cold, spring-fed brook emptying into the Snook Kill through a small ravine immediately to the south. The most heavily occupied portion of the site was apparently nearest the creek on the east side of the point. Here occurred the maximum amount of industrial remains and the majority of the pits so far discovered. Most of the artifacts and chipping detritus were found on or near the surface, in a tilth zone about a foot deep, and it is probable, as indicated by random testing, that such shallow midden deposits as may have existed have been thoroughly churned over by repeated plowing. Enough artifacts, however, have been recovered from pits, below plow line, to associate these features with the surface finds, an association further supported by the typological uniformity of the assemblage. This unusual fact denotes that rare phenomenon, a “closed” site, occupied by a group or groups of people with essentially a stable culture complex. Moreover, the character of the prevailing projectile point style is sufficiently distinctive to serve as a diagnostic marker for this complex. Excavation The investigations of the Rices have to date uncovered a small number of subsurface features, all save one described as pits, varying in size, shape and contents. Most of them have occurred within an area measuring about 75 x 75 feet on the east side of the site. 92 NEW YORK STATE MUSEUM AND SCIENCE SERVICE Six of these features, designated “fire pits” in the notes, are described as oval in cross section, as much as 5 or 6 feet in longest diameter and from about 24 to 52 inches in depth. Upon removal of the 12-inch tilth zone, these pits were found to contain from 6 inches to a foot or more of mingled black soil and heat-shattered rocks. Similar layers, or lenses of charcoal, also occurred at various deeper levels, leading to an inferred use as roasting devices. Calcined bone fragments, presumably representing refuse animal bone (none was submitted to the writer), were present in the burned layers, and the following artifacts were obtained : a broad stemmed point at 2 feet in pit 1 ; a similar point and hammerstone at about the same depth in pit 3 ; a celt and hammerstone in pit 4 ; a flint knife, 2 end scrapers and 2 drills in pit 5 ; and a hammerstone, drill, 2 unfinished points and 2 choppers in pit 6. Carbonized vegetable remains, thought by the finders to be wild cherry stones (endocarps), were present here and there in the black soil layers of pit 6, which also contained the usual calcined bone fragments, chipped stone wastage and burned rock. Just north of pit 6, which was located at some distance from the others, on the west side of the site, a large, compact bed of gray ash, some 25 x 30 feet in extent and up to 2 feet in thickness, was roughly delimited toward the close of the digging season and left for later exploration. Artifacts The artifacts from the Snook Kill site are neither numerous nor very diverse. None is of shell, bone or antler ; indeed, the only osseous remains on the site comprise the calcined fragments of probable food animal bone already mentioned. No scrap of pottery or steatite vessel has so far been found, nor is there any indication of the use of native copper. Nearly the entire collection of identifiable objects is illus¬ trated on plates 28, 29, the additional material consisting of 11 basal sections of the dominant broad-bladed point form, 15 point tip frag¬ ments, and a single bit section of a plano-convex adze. The material of the chipped stone category, which embraces pro¬ jectile points, knives, scrapers and drills, is apparently of local origin, including Deepkill, Normanskill and Fort Ann flint, the latter ob¬ tainable at various exposures, chief of which was probably the large quarry and workshop site near Fort Ann, Washington County. This material is light gray or blue-gray in color, with a slightly vitreous luster, and often a scoriaceous or “worm-eaten” appearance, well shown in plate 28, figure 9.67 67 For a further description of this flint, see Wray, 1948, p. 33. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 93 More than half of the projectile points pertain to a broad-bladed, sloping-shouldered, contracting-stemmed style, unlike any thus far described in this report or, for that matter, known to the writer, except as random instances, from any site in New York. Plate 28, figures 7, 8, 18-23, 25-31 portray the intact examples of this thick, apparently percussion-flaked and frequently unsymmetrical form, whose closest parallel seems to occur in the Lehigh Broad Spearpoint, best known from the area drained by the Lehigh and upper Delaware Rivers in Pennsylvania.68 Presumably associated with this form on the Snook Kill site are broad, side-notched or stemmed points (plate 28, figures 13-16) and narrow, stemmed points (figures 17, 24) ; asymmetric and ovate knives (figures 1,9, 10) ; triangular and simple end scrapers (figures 2, 3) ; broad, stemmed scrapers (figures 11, 12), and drills with slightly expanded, or expanded stemmed base (figures 4-6). Two of the broad points seem to have served as strike-a-lights, as shown by their characteristically battered tips and edges (figures 7, 8), and one (figure 23) may be a blank or unfinished point. Wood- working tools are few, rude and simple, as shown by the celt blades illustrated on plate 29, figures 1, 9; plano-convex adze blades (figures 2, 3), and shallow-lipped gouge (figure 6). On the same plate are pictured pebble and pitted hammerstones (figures 4, 5, 7) and what, for want of better identification, may be termed choppers (figures 10-12). Conclusions Comparison of this chipped, ground and rough stone complex with that from the intermediate level of the Harris site, which has been assigned to the Vosburg complex of the Laurentian tradition, of Middle Archaic age in eastern New York, leaves one with the im¬ pression that, except for the dominant Lehigh type points, the smaller Snook Kill series, despite the absence of the bannerstone and ulo, could readily enough be included in the same cultural category. This suggests a cultural relationship with the same widespread tradition but on a different time level. Witthoft states that in eastern Pennsyl¬ vania “the Lehigh Broad Spearpoints are ordinarily found on sites with soapstone sherds and apparently characterize the Transitional period [between Late Archaic and Early Woodland] of the upper Delaware.”69 He further remarks on the lack of data and especially of “any closed or pure sites of this complex,” which he attributes to 68 Witthoft, 1953, p. 21. 69 Witthoft, 1953, pp. 21-22. a; co co qj ^ 3 £ s ^ 'd c 2 04 4^ .3 D .. M j_ trt S ”0 ° Jh C OT aj 42 c » 5 1> g 3 r ? 3 ^ QJ io cn O h +j ~ xi ■ cj — i cd CD 03 & X 1-h ^ Oh O rri Cd ftM V ^ •& x) S M G « | 3 . U M 6 OO Cd _ G ^ 3 3 £ rj 04) O C/5 H ^ ^ »* a, 2 o Oh ft ■ k“H 3 ^ . G t 94 ] [95] inch [96] | M3«! j* -d •- o s g c <2 **> 8 ^ Q 43 C\f >f Oh tC ba • «K a w • ». O O tn 4-» o co Ph o f=H wT [97] 98 NEW YORK STATE MUSEUM AND SCIENCE SERVICE "a riverine, canoe-using, fishing culture much like the Susquehanna Soapstone Complex/’ 70 The writer’s present inclination is to regard the Snook Kill material as depicting a unique assemblage, possibly of the Late Archaic period in eastern New York, and to defer further speculation concerning it until more information can be gathered at this apparently closed station, or from other sites that may be recognized as significant in the light of the problem, as here defined. (See figure 5.) GENERAL CONCLUSIONS The avowed preliminary nature of this report permits but few firm conclusions to be drawn in the current stage of our compre¬ hension of aboriginal life in the Hudson Valley region. Perhaps the least equivocal relates to the obviously long tenancy of the valley, to be measured in millenia rather than in centuries, as was formerly supposed in some quarters. In its pristine state, the valley seems first to have been explored by paleo-Indian hunters of the Clovis tradi¬ tion, whose characteristic fluted javelin or spear points (plate 30) have a thin, sporadic distribution on the higher ground along or back from the river, from Orange to Essex Counties. The estimated an¬ tiquity of this opening historical drama “of time and the river,” is some 5,000 years B. C.71 By approximately 2425 B. C., as shown by the single radiocarbon date for Hudson Valley prehistory (Bannerman site), hunters of the Vosburg culture of the Middle Archaic period had established them¬ selves on the river. There are ample reasons to believe that their ancestors of the Laurentian tradition had spread downstream from the Lake Champlain-St. Lawrence area centuries before. There are almost as cogent grounds for assuming a predecessor folk, who shared with the people of the Lamoka culture, farther west in the State, in a cultural tradition, still ill-defined and of unknown provenience. The presence in eastern and southeastern New York of slender projectile points resembling those of the rather well-elucidated Lamoka com¬ plex, radiocarbon dated to about 3500 B. C. in central New York, but without other diagnostic elements of this assemblage, seems to relate the earliest definite components on stratified sites in the Hudson Valley to an Early Archaic horizon. Of the nebulous interval of perhaps only 1,500 years between the fluted point users and the makers of the Lamoka or Lamoka-like points, we are at present completely ignorant. Is it possible that the 70 Witthoft, 1953, pp. 21-22. 71 Ritchie, 1957. AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 99 gap was partly filled by another wandering hunter group who left behind the relics pictured on plate 31 ? These are broad-bladed points with thin, lenticular cross sections and sharp, convex edges ; wide, usually concave-base stems and weak shoulders. Some, indeed, are lanceolate in outline. Like the fluted points, but in marked contrast to the various Archaic period types which occur on the sites herein described, they were evidently fashioned by a pressure flaking tech¬ nique and, as a further resemblance to the fluted points, two or three present a weak, shallow, vertical channel flake part way up one face. Moreover, grinding of the lower edges and/or base, a near universal feature of the fluted points, is present, but only in a slight degree, on about one-third of the specimens. This class of points, recently recognized in New York State, has its closest typological affinities with the Steubenville Stemmed and Steubenville Lanceolate types, assigned to the Panhandle Archaic facies of West Virginia by Mayer-Oakes.72 A series of the New York specimens examined by Mayer-Oakes confirms this impression.73 Two areas of New York have so far produced most of the examples seen in private collections, the Colliersville district on the Susque¬ hanna, in Ctsego County74 and Greene County, on the Hudson below Albany. Representatives of the stemmed form have, however, been recorded from central New York (plate 31, figures 7,9) and else¬ where (figure 12). Since the Steubenville forms have not been found in direct associa¬ tion with Vosburg on other known Archaic materials, although they are present on surface sites where these occur, their cultural rela¬ tionship and temporal position in the New York sequence are still debatable. Among the most well-established deductions for Hudson Valley archeology is the dominant status of the Laurentian tradition. It would seem to have spanned a period of several millenia in this area (see figure 5) and to have constituted the basic, underlying tradition of the Vosburg complex of eastern New York, the Brewerton com¬ plex farther west, and of other regional manifestations, mostly still undefined, of Middle and Late Archaic times, in the Northeast, and elsewhere in the eastern United States.75 72 Mayer-Oakes, 1955, pp. 8-9. 73 Correspondence of November 1955. 74 Collection studied by courtesy of Mr. Rowan D. Spraker, Cooperstown, N. Y. 75 See, for example, Mayer-Oakes, 1955, p. 19; Stewart and Dragoo, 1954, p. 110. 100 NEW YORK STATE MUSEUM AND SCIENCE SERVICE In the Hudson Valley area the regionally differentiated Vosburg complex seems early to have absorbed elements of a still older Early Archaic tradition, not yet terminologically designated, suggesting a parallel with the perhaps contemporaneous assimilation in central New York of the Lamoka complex by the Brewerton complex. In the resulting Frontenac complex of this particular locality, the intruding Brewerton culture became dominant. Similarly in the east, it would seem that the Vosburg essentially maintained itself with but little change beyond the obvious addition of an older point style. As time passed, other influences appear to have been exerted from various directions on this general cultural pattern without, however, producing perceptible modifications in the way of life. At Bannerman and South Cruger Island, for example, the excavational data suggest that increments of the Susquehanna tradition (steatite pots, Orient Fishtail and probably related Susquehanna and Perkiomen Broad points) were added, seemingly, by contacts or diffusions from down river, to the already composite assemblage of the Vosburg complex. Still later, the first pottery appeared, this time, presumably, from up river sources. It is, however, very important to emphasize at this point the highly significant contribution to the archeological interpretation of our site series afforded by the small but stratified Lotus Point station. Here the cultural-physical stratigraphic correlation indicates the reality of an underlying slender point assemblage, followed by a Vosburg com¬ plex which had absorbed it, then a fishtail point and steatite horizon and, finally and successively, an older pottery horizon with certain lithic elements of preceding levels and a recent, late prehistoric ceramic and triangular point complex which persisted into historic times in the valley, as attested by the topmost level at South Cruger Island. The relative clarity of this unusual situation at Lotus Point, which provides a partial frame of reference for Hudson Valley prehistory, apparently resulted from a fortunate combination of cultural and environmental circumstances, as already explained. Somewhere within the long indicated time span at Lotus Point, another assemblage, not represented there, but present on the neigh¬ boring Van Orden site, was brought into the valley, possibly from the west, via the Mohawk River, along which its traces can be followed into central New York. The principal evidence in the Hudson Valley for this weak influx of the Middlesex culture, presumably rooted in the Adena tradition and believed to have flourished in the Northeast during the Early Woodland period, consists of lobate stemmed points. A few such points were present at Bannerman’s, a much larger number on the surface at the Van Orden site (figure 1, site 6). AN INTRODUCTION TO HUDSON VALLEY PREHISTORY 101 A multiple burial on a sandy knoll at the latter site, found by Carl S. Sundler of Albany and excavated by him and the writer in September 1955, shed further light on this complex. Very concisely, it comprised an older grave pit measuring 50 x 35 inches, and about 26 inches deep, within which lay extended on the back, a badly de¬ composed adult skeleton, with head to the east. Two rough limestone slabs leaned against the west (foot) end of the grave. Similar slabs, placed over the head end, had in part been removed by a bulldozer. Four whole lobate stemmed points, three broken points and a point blank or reject were scattered along the grave floor from head to foot (plate 32, figures 1, 2, 4-6, 8, 10, 13). Intruded into the western portion of the grave was the crushed and much rotted bundle burial of an adult, occupying a space about 2x2x1 feet. Among and around the bones were found two lobate stemmed points like those from the deeper grave (plate 32, figures 9, 12), 45 native copper beads made of thick rolled strips of metal (figures 7, 11) and a portion of a much larger copper bead (figure 3). Several of the beads were fused together in clumps of two or three (figure 7) with bits of the twisted cord string still adhering within them, preserved by the copper salts. Less than 2 feet west of the foot of the grave, the 6-inch deep bulldozer cut, which had removed the surface soil and exposed a few bones of the upper burial and some stone slabs of the lower one, also destroyed the top portions of a shallow basin-shaped depression, whose remnant measured 2 feet in diameter and 2-3 inches in depth. Within the dark soil filling this depression were found numerous flint chips and quarry blanks, a steatite potsherd, a small sherd of Vinette 1 type pottery and a copper bead, identical in character with the grave goods. This latter object served to link the feature culturally and probably temporally with the burials. The combination of traits, meager though they be, betokens an Early Woodland period of culture, only weakly represented in the Hudson Valley, chiefly by a thin scatter of Middlesex and early Point Peninsula point styles and by the infrequent sherds of Vinette 1 type pottery76 which occur in the deepest ceramic levels of open sites and rock shelters. Occasion¬ ally intermixed with this earliest ware, as seems to be the case at South Cruger Island and the Hennessy site, but more often overlying it in rock shelters or occurring without it, as in the upper level of the Harris station, are the more ornate sherds of the Vinette 2 ware group of Middle Woodland times.77 In the Hudson Valley these 76 Ritchie and MacNeish, 1949, p. 100. 77 Ritchie, 1955, p. 70. 102 NEW YORK STATE MUSEUM AND SCIENCE SERVICE comprise dentate- and dentate rocker-stamped, corded and pseudo scallop shell decorated forms78 which, like the earlier Vinette 1 type, diminish in frequency from north to south in the valley and seem to have been diffused in this direction. A ceramic trickle from the opposite direction, brought coastal and lower Hudson Valley styles upriver, at least as far as Columbia County, also during the Middle Woodland period. In this group we have tentatively identified North Beach Brushed and North Beach Net Marked types.79 Somewhat more abundant, and occurring in higher levels where stratification has been observed, are pottery types of closing Middle Woodland age, as defined in central New York. These include pri¬ marily Jack’s Reef Corded and Jack’s Reef Corded Collar.80 They are found with other types of the same age group, with large tri¬ angular projectile points, an occasional bone implement and flat-based platform pipe fragment. Like the lobate stemmed point assemblage of more ancient days, this late Point Peninsula assemblage was diffused into eastern New York via the Mohawk River.81 In central, northern and a major portion of eastern New York, the O wasco developmental sequence of the Late Woodland period succeeds the Point Peninsula developmental sequence, apparently in the main, as a cultural continuum. In the upper and middle Hudson Valley there has been found, as yet, very little evidence of the O wasco, and that little confined to its earlier stages, as indicated by ceramic types. This is somewhat strange, since a definite Owasco sequence has been in some measure worked out in the eastern Mohawk Valley.8'’ It would seem as though the little known occupants of the Hudson Valley at this time had barred the Owasco people and their culture from the former’s domain. In the coastal area and lower reaches of the Hudson, the East River tradition made its appearance at this time, apparently as an invasion of new groups from the adjacent New Jersey area, but the characteristic ceramic styles of the Bowmans Brook and Clasons Point complexes seem not to have been diffused far upstream.83 The uppermost ceramic stratum on such sites as Lotus Point and South Cruger Island, as well as others, including rock shelters, not considered in this report for want of sufficient data, produce ceramic 78 Ritchie and MacNeish, 1949, pp. 100-104. 79 Smith, 1950, p. 196. 80 Ritchie and MacNeish, 1949, pp. 106-107. 81 Ritchie, Lenig and Miller, 1953, pp. 27-48. 82 Ritchie, Lenig and Miller, 1953. 83 Smith, 1950, pp. 116-119, 120-123, 147-148, 153. [103] Plate 30. Fluted points from the Hudson Valley. 1, found atconfluence of Roundout Creek and Hudson River, Ulster County; 2, found near Bacon Hill, Saratoga County ; 3-5, from the vicinity of Coxsackie, Greene County Materials: 1, 4, brown jasper; 2, 3, 5, gray flint [ 104] inch O S £ •§« I N-t ^ O ffi - f) « ^ rt - ^ 9 s >, *§ o -y o >> n=! •g bo § & U bo rt -S bo Jn o *J“! _G <-M >> p 0 U o i.i aS CO re * -t-> G o U tol :•: c rer CNJ c\T -a O) '© to o bO u •4~> • g oC s s JU G rt G IS S S .5 * S-t Q to G s oo “C o to P o It G S o G O VO 1-0 co jC U. n ) Cfl ’■p p 1—) vH co e/3 tJ* . S *”1 ’O «'0 -T| „ cC GO to u re CO un CO w X o H << to U G o s_, to 're ft re OJ rt J2 • S -d ~ ^ o O ,Q eg rt ^ ’d & £ Vh o bJD c OT rd rt 44 u (/> o g „ eg +-> s .s o ^ * § >? o' ~ U y~i p,H eg ^ i W ^ rC o ^ •- „ ’d irT cd 0 ** g.-n s g *W 4-» 3 •rH O a. o a, as co w O KH C/2 •O 3 rfC 1/2 l-H U a> __ o o 13 3 3 .a a § O' O 0) O O u w §5| 'I ;2 TJ 3 CO 33 O co £ % S 3 « g5 ±; 33 3 ns ^ »o H £ o t-H ,1 FIGURE 5. Cultural and chronological relationships of Hudson Valley sites, with reference to central and southeastern New York sequences FIGURE 4. Map of South Cruger Island sit FIGURE 3. Map of Lotus Point site