Bernard W. Powell


THE GENERAL PROBLEM.  It has been postulated that the ceramics and other cultural traits of aborigines who dwelt in southwestern Connecticut should diverge, at some point in time, from an early, dominant Windsor culture (Aspect) and should evidence instead addition of, or replacement by, the East River culture, defined by Smith 1 and others.  This hypothesis rests on evidence obtained during the archeological investigations in areas adjacent to this region.  From such material, Rouse 2 predicted an extension of Smith's boundary which bisects culture areas of Long Island into eastern and western divisions.  Rouse's boundary extension across Long Island Sound should enter Connecticut and run northward somewhere in Fairfield County. Establishment of this boundary between the two culture areas on the mainland would be desirable.  However, lack of controlled excavation or of any methodical recording of aboriginal finds in this region, as noted by Rouse 3, Pope 4, Smith 5,and others, has kept this part of Connecticut archeologically unknown.

         In the summer of 1953, I located a site in this region which showed evidences of aboriginal occupation.  In the intervening 3 years, enough material has been recovered to justify a report on it and permit at least a partial attempt to reconstruct events here and relate them to the larger problem outlined above.  Large-scale road building is even now destroying part of the site.  A final report will be attempted at a later date after many features and deposits have been obliterated and there is little likelihood of further significant finds.


Location.  The site, hereinafter designated the IF (Indian Field) site, is on the west side of Cos Cob Harbor in the town of Greenwich, Fairfield County, Connecticut (Fig. 1).

It extends about 1670 yards north from the harbor entrance to the region of the New York, New Haven and Hartford railroad tracks, and inland about 500 yards west. The rather large area so delineated (in excess of 150 acres) contains, with one exception, all aboriginal deposits noted by me.  Another site, rather small, which I shall call the IP (Indian Point) Site, lies on an extreme outer point of land on the opposite shore of the harbor. IP is about 1830 yards south-southeast of my datum point at IF, and is included here because of apparent temporal and spatial relationships to IF.  Both localities can be found from my datum point at the approximate intersection of 41o 01' 45" latitude and 73o 36' 00" longitude on U. S. Geological Survey Map, Stamford Quadrangle, 7.5 minute series.


Geological.  The milieu of the IF site is typical of the Connecticut littoral.  A point extends southward into Long Island Sound and shelters along the eastern margin a narrow tidal cove.  A freshwater river, the Mianus, drains into the upper reaches of this cove.  Like other streams hereabouts, the Mianus heads inland in a region of rock outcrops and long parallel ridges.  This land blends imperceptibly northward into the Taconic Highlands of upper Connecticut and the Hudson River Valley. Evidences of Continental glaciation abound throughout the area.  Surface formations are mainly Pleistocene in origin, and include drumlins, kame terraces, and deposits of glacial till.  Striations and potholes mark many watercourses.6   Compared to the rough interior, the immediate area of the site is rolling and reasonably smooth. Shoreline environment is varied and ranges from a sandy beach near the outer extremity of the point, to muddy, exposed flats in the harbor, and eelgrass marshes in the more protected recesses.  Islets near the harbor opening and along the shore are typical of tidewater coves in this region.  Maximum height of the point is about 40 feet.

The aboriginal material occurs mainly along the first and second rises of land bordering the harbor.  In places along the shore, the complex, metamorphosed bedrock of the area creates steep banks.  The beach strip presents a very heterogenous collection of stones and minerals.  The water-worn, rounded cobbles include chert, quartz, quartzite, and other species.  Intrusive dikes in the bedrock contain mica, hornblende, tourmaline, garnet, feldspar and quartz.  To peoples who were "lithic-oriented," the region offers a profusion of workable material.  Near the periphery of the site, fortuitous deposition of glacial erratics created a few cramped rock shelters and overhangs.  Though much disturbed in the interim, they harbor evidences of aboriginal use.  The predominant yellow loam on the site is overlain by black, rich, organic humus.  Much of this is charged at random with marine shell fragments, pieces of mammal bone, and stone chips and spalls.  This humus layer varies in thickness from nothing to as much as 10 inches.  The porous nature of the soil and imperviousness of the bedrock are favorable to the occurrence of springs. Several perennial springs occur at the site now, and many more undoubtedly flowed in the past before farming and other practices greatly disturbed the drainage.  Dried courses support this conclusion.

Botanical.  Today the site is mainly crop and pasture land for a large shore estate.  Second growth hardwoods and stands of set pines occur everywhere throughout the area.  Criss-crossing stone fences of Colonial origin ramble over the site.  Where the land is not plowed or grazed, extensive thickets of poison ivy (Rhus toxicondendron), honeysuckle (Loniceradioica), and other obnoxious plants blanket much of the surface and make archeological survey both difficult and tiring.  It is doubtful if any of these plants grew here during aboriginal times.7 Located within the Eastern Deciduous Province of the primeval northern Sylva defined by Peattie 8, the site must originally have supported a stand of oak, tulip tree, gum, sassafras, butternut, and related types.  Some of these species are still present, but it is obvious that the land was overcut from much of its original stand; indeed, the historic Indians themselves may have been responsible for initiating this practice.9  While much of the sites altered from its appearance in former times, it seems reasonable to suppose that the lack of building and construction has preserved some of the original character of the land intact.

Zoological.  The usual complement of northeastern mammal, bird, and fish life is assumed to have been present at the site in aboriginal times.  No longer extant, but known by direct skeletal evidence from the refuse deposits, are deer (Odocoileus), beaver (Castoridae), migratory sturgeon  (Acipenser), scallop (Pecten), and possibly elk  (Cervus). Species living in the area today, and represented also in the refuse deposits, include grey fox (Urocyon), raccoon (Procyonidae), eel  (Anguilla rostrata), crab (Callinectes sapidus), clam (Mercenaria mercenaria), mussel (Mytilus edulis and Volsella demissa), oyster (Crassostrea virginica), whelk (Busycon), marine snail (Littorina), blackfish (Tautog aonitis), cunner (a Wrasse), other fishes, shark, and various birds.   Not identified as such in the refuse material, but surely known to the Indians, would have been bear (Euarctos), turkey (Meleagris), squirrel (Sciurus), rabbit (Sylvilagus), otter (Lutra), muskrat (Cricetidae), lesser voles and rodentia, and the ubiquitous dog, (Canis).  Animal life still maintaining itself is reduced to those forms than can live in the precarious balance dictated by reduced cover and the presence of man and domestic animals.


Maps, records, photography. Because of the projected highway grading and the destruction already attending cultivation of an annual corn crop, my initial decision concerning the IF Site was to survey the entire region for any deposits threatened by these agencies.  The deposits were then excavated mainly in the order in which they seemed to be threatened.  Emphasis was naturally accorded undisturbed stratigraphic deposits.  However, complete salvage of material wherever encountered was elected, and extensive surface hunting on disturbed ground was also undertaken.10   To record adequately the location of all features removed, a map (Fig. 1) was prepared using low-altitude aerial photos of the site.  This map was laid out on a very large scale: 1 inch to 100 feet.   A datum point was selected and was thought originally to be well outside the suspected area; subsequently it has proven to be well within it.  This datum point may be located as a red-painted circled "X" chiseled in bedrock at a shore outcrop.   With this as a referent, a grid was laid out on 100-footintervals.  That is, it was laid out on the map, not by stakes in the field, as this would have interfered extensively with the local agriculture. Such a large grid obviously did not serve as a guide for quadrant-by-quadrant excavation!  Its purpose was to permit entering features on the map in some approximation to their relative occurrence at the site.  Base lines on the map were measured by 100-foot steel tape.  Compilation of a vertical cross section through the site with a hand sighting level was abandoned as not worth the effort and not contributive of essential information.  General estimates of contours were obtained from a Geological Survey map.  Notes and sketches made in the field were later transposed for permanent records.   Color slides (35mm) recorded many phases of the work and appearance of some artifacts in situ in the field; slides of the collection are presently being made.

Excavation.  Test holing, limited trenching in suspected areas, and vertical and horizontal excavation of obvious features were all employed.  Specimens from different levels were collected separately; in many cases pieces of bone, pottery, and other artifacts, which matched along break lines came from tops and bottoms of pits and appear to obviate stratigraphy.  It is the author's opinion that many refuse pits contain material deposited nearly simultaneously and not over any extended period of time.  This appears to be true far more often than does definite stratigraphic difference.  When a pit was located, the ground surface was usually cleaned back for some distance beyond its margins to check
for the presence of post moulds. None have been recorded to date.  All excavations were backfilled according to an agreement with the various landowners concerned.

         In the search for subsurface features, a special tool I devised proved most helpful.  It is an earth augur about 1.5 inches in diameter, widely used by tree surgeons to feed roots.  I welded this to a cross-T pipe handle.  With it, I was able to bore to a depth of about 3 feet and bring up samples of soil from beneath the surface.  Shell layers and pits were revealed by relatively higher proportions of shell fragments in material so removed.  Originally I intended to bore at every intersection of the grid but settled, finally, for specific areas most likely to harbor pits.  Several completely hidden features were located in this manner. The main drawback with the method was the time and effort required to survey even a relatively small area.

Treatment of specimens. All material not too fragile to handle was carefully washed to remove loose soil.  Projectile points, scrapers, and small stone tools were catalogued and placed in standard Riker display mounts such as are used for entomological and other biological specimens.  Large, bulky items were filed in a steel-drawer cabinet. Pottery was washed, glued along break lines where these could be matched, and mounted in Riker mounts.  Bone artifacts, animal jaws, and significant animal bones were washed, dried and impregnated with Alvar 7/70 vinyl resin.  Such bone specimens are then easily handled without fear of breaking.  Where feasible, these have been displayed in standard Riker mounts.  Burned organic material was treated in a similar way, with the exception that some charcoal specimens were wrapped in foil and stored in airtight jars. Several professional archeologists suggested this as a precaution to permit a future C-14 test.  The collection is freely accessible to those doing study in this region.  Eventually all material will be given to the Archeological Society of Connecticut for its disposal.


Historic References

         Several writers place Indians on or near this site at the time of Contact. To quote a few pertinent passages: "Some small clans seem to have inhabited the coast from Greenwich to Fairfield, but so feeble and insignificant, that not even their names have been preserved from oblivion." 11 This same author concedes that aboriginal population in the region increased around 1643 by an influx of Indians from the Long Island (western?) and Hudson River tribes on whom the Dutch were exerting pressure (italics mine).12   Not necessarily contradictory (if we allow sufficient time lapse) is a statement that land near the confluence of Strickland Brook and the Mianus River had been thickly settled prior to Contact.13 This is the locus described by Trumbull 14 for the village of “Petuckquapaen,” and often referred to in passages relating to the early history of Greenwich.  He also states 15: "A village of the Siwanoy tribe was situated above the Westchester Path (Post Road), near what is now Cos Cob.  It was called Petuquapaen  (sic) and the chief was Mayn Mayano or Myanos." 16   The spot is less than a mile north of the IF site.  Petuquapaen had a record of harboring Indian malefactors from the Hudson River tribes who fled Dutch justice under Governor Kieft in New  Amsterdam.17    It was this flouting of white law, plus numerous quarrels with fur traders and settlers, that led to the events of February, 1644,when a combined force of Dutch and  English soldiers marched on Petuquapaen and destroyed it during the Battle of Strickland Plains.

The destruction of the village “... marked extermination of the Siwanoys.” 18   Varying accounts   place the number of native men, women, and children slain between 500 and 700 persons.  What  interests us here is a reference to the presence of  25 Wappingers in the village on the eve of the battle: " . . . they had gathered together to celebrate  one of their festivals." 19   Generally descriptive of   the unsettled times is the following account of the sequel to the massacre at Petuquapaen:  “More than fifteen hundred warriors, rallied from the confederacy of eleven clans, to constitute this avenging army...  From Manhattan to Stamford the coast was desolated, Dutch and English alike atoning to the inexorable spirit of Indian revenge   for the needless injuries that had been heaped upon the Indian race.” 20     Trumbull mentions several names relating to the immediate area of the site: “Cos’ cob: a neck   of land, in the s.e. part of Greenwich.  The Mianus river flows into Coscob harbor, on the w. side of which is Coscob village.  The name, denoting a   'high rock,' (comp.Cassacubque) was perhaps  transferred from the bluff west of Strickland' s Brook, near the Indian village."20  “Mianus river: in Greenwich and Stamford; and transferred to a village at the junction of this river with Coscob cove.  For “Mayanno’s,” - as the river and neck of land were called, from the Indian proprietor,  Mayanno or Mehanno, who was killed by Capt. Patrick, in 1643.  “Mayannoes neck”(Greenw.  Records, 1664).  “Mayane, a sachem residing...between Greenwich and Stamford.”  (N.Y. Col.  Mss. i., 186.) His name signifies “He who gathers together.” 21

         An incident pertaining directly to the site is recorded by Mead.22   I believe it to be the origin of the name "Indian Field' for this region. On February 1, 1686, Wesskum, identifying himself as a Sagamore of Wapping, attested to the validity of a sale by six Indians of all land or lands between the Mianus River and the Byram River  (later shown to be south of the Westchester Path).   This sale was accorded seven English settlers, including two of Mead's forebears.  In consideration for this, the Indians and "four papooses" related to them, received thirty acres of planting land within these same bounds.  This land lay  "...fenced in at Cos Cob Neck ye lower point."   The description fits the IF site.  On the death of the 'four papooses" all land reverted to the town for its keeping.  The four children named in the historic deed are almost surely the last representatives of their native group in point of time, and we may assume they disappear from the scene forever early in the eighteenth century.  Hurd 23 is evidently referring to this incident when he writes,   "In 1686 the Indians sold nearly their last acre of land in the town.  These lands were on the western bank of the Myanos, near its mouth.” 24

Analysis of the historic sources. First, it almost seems that De Forest might have been referring to events far antedating historic times; his tribes,   " . . . so feeble and insignificant, that not even their names have been preserved from oblivion,” could easily be taken to describe Windsor-tradition peoples absorbed or expelled from the area by the East River invaders around 700 A.D., or to the invaders themselves.25   There are good references to places, persons, and events in historic times that permit us to say the Indians of that period were Siwanoys and therefore allied with the Wappinger Confederacy.  Spiess says, 26  “It cannot be disputed that the Siwanogs (sic) were members of the Wappinger Confederacy in New York.  Their territory extended into what is now Connecticut and includes the towns of Greenwich...  tribal name seems derived from Sewan, shells, auk, ac, or og, place."   These Siwanoy, from their name alone, would seem to be coastal members of the Confederacy;27 Skinner felt they were the occupants of the historic village of Snakapins on the East River.  This is tentatively linked to the Clasons Point site in the Borough of the Bronx - within easy traveling distance of the IF site.28

From the literature emerges a picture of tribes in this southwestern corner of Connecticut being affiliated with Indians of the lower Hudson Valley and enjoying direct communication and allegiances with them. There is no documentary evidence that these Indians had much traffic with tribes to the east - in the region of that “long tidal river" which has colored so much Connecticut Indian history. These local peoples seem, therefore, outside the pale of the Windsor and Shantok cultures upstate.  As stated previously, this region is suspected to lie within the bounds of the East River domain.  The recovered artifacts end to substantiate this hypothesis.  Now I hope to have shown what scattered historic references have long implied: tribes of this culture were here in southwestern Connecticut at the time of Contact.

Today, the site is part of a secluded residential community.  Extensive grounds may yet preserve many features intact.  No prior controlled excavation is known for the actual site,29  though surface finds have undoubtedly been made by chance walkers over the years.30

(Click To Continue)