45, BURIALS IN SAND tMOUNDS OF THE DELTA REGION OF THE SACRAviENTO - SAN JOA`OUIN RIVER SYSTEMl By S. F. Cook and A. B. Elsasser Outline of Mound Complex From Tyler, Bradford, Holland and Bethel Islands, together with nearby mainland tracts, a series of sixteen sites have been recorded by the University of California Archaeological Survey. These sites, despite wide divergence in individual character, possess one feature in common: they are all located in or on small mound-like soil bodies which consist primarlrly of compacted sand which, with one or two possi- ble exceptions, may be attributed to the formation known as Piper fine sandy loam. Piper fine sandy loam, together with related soil types, has been described by Cosby (1941) and its distribution is shown in detail on his map, It occurs in isolated patches along the west side of the delta from the vicinity of Byron northward to a point southeast of Isleton. Cosby, s Survey shows two or three small areas near Werner in the Orwood Tract, a heavy concentration of exposures in the Veale, Hotchkiss and Holland Tracts, an almost Continuous expanse on southem Bethel Island, ten or more areas on Bradford Island and seven locations on the Webb Tract just to the east of Bradford Island. Separated from the latter region by approximately seven miles air-line are two c-tlying sand mounds on the southwestern end of Tyler Island. The Pipeir profile is also shown by Cole, Koehler, Eggers and Goff (1943) to occur near both banks of Old River along the southwestern edge of Union i"sland. The physical appearance of these soils is best described in the words of Cosby (1941, p. 25): "The Piper soil occupies low mounds and ridges that have a roughly aeolian configuration and appear to represent a comparatively recent emergence of dunes and similar wind-modified bodies of sands somewhat related to the nearby bodies of Oakley sand. At some time in the past these emergent minor elevations probabl.y were covered to varying depths by a mantle of peat, which gave the original terrain a more or less smooth, almiost flat, conifiguration. The level of ground water undoub-tedly was much higher than at present. Now, how- ever, Piper fine sandy loam is well drained to almost droughty, especially on the higher ridges. A thin cover of saltgrass and herbaceous annuals has replaced the former tule-reed vegetationr."t The typical profile shows a relatively thin surface layer of grayish fine sandy loam with a high organic content and considerable lime. This grades intostsi3oil layers with diminishing organic content but still heavily calcareous. Ultimately, at from 2-1/2 to 5 feet the unweathered substratum is seen to consist of a light colored, heavily indurated sand. There are many local variations in both the thickness and appearance of the individual layers and in certain localities the Piper series grades extensively into other soil types, thus presenting many transitional -26- aspects. In particular the Piper grades into the Egbert soils generating what Cosby calls the "Piper-Eghert Complex." The Egbert soils are predom- inantly derived from peat deposits (see Cosby's discussion on pages 19 to 22). Hence the Piper-Egbert Complex is characterized by conspicuous masses of compacted sand with intervening layers of almost unaltered peat. This composite horizon, as shown by Cosby, is developed almost exclusively on Bethel Island and the adjacent southeast corner of Jersey Island. Of the northem sites, which are found on Tyler Island, the most thoroughly excavated is Sac-104. It lies on the eastem ernd of a fairly large and quite conspicuous sand mound. Although it has been almost entirely destroyed by agricultural operations, sufficient has remained for archaeological investigation. In the exposures examined (Fenenga, n.d.)* it is clear that a layer from several inches to a few feet deep of Late Culture deposit covered the top, The material is black, heavily impregnated with charcoal, and filled with bone fragments, shell fragments and obsidian chips. It therefore differs in no essential respect from many other deposits of recent habitation midden. The Late habitation midden rests unconfonnably on the indurated sand base characteristic of the Piper fine sandy loam profile. It is probable therefore that the surface of the Piper fonration was bare when the depo- sition of the upper habitation material began. Nevertheless a number of burials have been found within this sandy base. According to Fenenga (op. cit.) at least one of these burials is intrusive. On the other hand, regardless of whether or not they are intrusive,, the cultural period rep- resented must be entirely different from, and antecedent to, that during which the late midden was deposited. In the files of the U.C,A.S. site Sac-45 is described as a cap of habitation deposit on a Piper fine sandy loam base. The thickness of the cap has been stated to be three feet, and burials have been removed from it. No burials, on the other hand, are known from the sandy base. Site Sac-162 is located about one-half mile to the northeast of Sac-45 in a cultivated field. It is described (U.C.A.S. site record file) as consisting of "yellow sand with dark sandy loam beneath." There is little occupation deposit on the surface yet artifacts were found and also "some fragments of charcoal; mammal bones; river mussel." Possible "inter- mittent occupation" is conceded. The mound has never been excavated and therefore no burials are known. On the whole, allowing for a much lighter occupation, Sac-162 appears to be of the same type as the sites previously mentioned. Southward from Tyler Island the next known sites are on the western shore of Bradford Island. These were recorded relatively recently and have been designated CCo-2 and 17. The latter is located 100 yards east of the levee at approximately the midpoint of the western shore line in an area comprising several acres of mixed Piper fine sandy loam and Oakley sand. The site itself is probably on Piper, rather than Oakley, formation *UCAS Ms. and Site Records on file at UCAS office, Univ. of Calif., Berkeley. -27- BRANNAN IS. WEBB TRACT SHERMAN IS. FRANK'S TRACT tEVILLE I S. :Y H K I SS TRACT MAP I Delta Area San Mound Burial Sit Legend 0- Sites -- County boundaries Scale in Miles HOLLAND TRACT id tes 138* @139 0129 128 *146 BACON IS. 0 1 2 L - I --- and consists (or formerly consisted) of a relatively deep deposit of prob- able Late occupation midden from which numerous burials were taken by the local 'inhabitants, together with stone and bone artifacts, none of which are now available for description, obsidian chips, and other characteristic debris. There is no authentic record of any skeletons having been found in the underlying sandy base at this point. The second locality (CCo-2) on Bradford Island lies close to the southwestern tip of the island, just behind the levee, where roughly an acre of Piper fine sandy loam is exposed. According to an apparently reli- able report a workman digging a ditch through the sandy base discovered a skeleton, the bones of which have long since been lost. Although it is claimed that a metate was found on the surface of this sand mound, there is no real evidence that it was subjected to occupation in the Late Horizon. In fact, examination of the soil profiles exposed by farming operations indicates no midden deposition whatever. South of Bradford are Jersey and Bethel Islands. On the former, although there are extensive exposures of Piper fine sandy loam, only one site has been recorded, CCo-5. This was found and recorded by IU.C.A.S. -n 1953. It was placed on the south bank of the San Joaquin River, three miles above Antioch, on the exact western tip of Jersey Island. The site could not be examined thoroughly and no clear description could be obtained except that, due to river erosion, burials were washing out of the sandy soil. Three burials had been noted, of which two were recovered, together with one pestle and two mortars. The soil map of Cosby shows no Piper sand at this locality; indeed the entire western end of Jersey Island is delin- eated as Staten muck. However, it is quite likely that a small body of sand existed here which escaped the notice of the soil. survey. At any rate, it is clear from the report that the skeletons were washed out of sand of some sort. On Bethel Island CCo-135 is located one-half mile north of Farrar Park, in the southeast corner of the junction of the Bethel Harbor Road and the road running east to Frank's Tract. It consists of the now almost completely destroyed remains of a small mound of Piper fine sandy loam. It was visited in 1952 by the U.C.A.S., and it was noted at the time that an exposed cut shows a lh-inch band of occupation midden lying beneath 32 inches of sterile sand. The base consisted of the typical indurated cal- careous sand. Burials had been reported from the portion of the mound which had been bulldozed. A more recent examination made in 1955 indicated that some revision of the earlier report was necessary. It was found by the present writers that a small portion of the south rim of the site is still intact. The profile exposed demonstrates three clear horizons. The uppermost is com- posed of two or three feet of loose gray sand, without stratification and apparently laid down by wind action. The second horizon is made up of eight to twelve inches of black, semi-carbonized peat mixed with sand. The structure of the organic material is unquestionably that of peat and there is no indication whatever of any occupation debris within it. The peaty layer rests unconformably on the third horizon, a deep, heavily compacted, brownish-gray calcareous sand. -28- The structure described, with minor variations of dispersion and depth in the peaty layers, and with the frequent occurrence of ashy layers, was found to be very common over the southern portion of Bethel Island wherever road or ditch construction has exposed a profile. Moreover, in all its features this formation corresponds closely to the descriptions given by Cosby for the same area. From the foregoing it must be concluded that the site designated CCo-135 was not occupied in the Late Horizon of Central California. and that the burials disinterred by bulldozing came from the underlying sand. Half a mile east of the preceding site is a large sand mound covering several acres. It is shown on the soils ma.p a.s an island of Oakley scand lying surrounded by Pan extensive area of interraingled Piper fine sandy loamr and Piper-Egbert Comiplex. The Oakley sand forms a large, dune-like mass to the south of the highway, which almost reaches the dimensions of a low hill. Into the nortl-h face of this mass a sizeable excarvation for sand has been carried, while the roa-d itself has been cut through the northernmost outlyers of the mound as a whole. The exposurcs along the highway (Frank's Tract road), along the face of the, bulldozed sand pit, and along the small dirt road on the west slope of the mound all have been examined by us. The profiles nowhere give any indication of occupation. The Oakley sand mass is poorly differentia.ted as to distinguishable layers. The Piper anna Piper-Egbert Complex soils on the periphery display numerous peaty and ashy horizons overlying the norm-al indurated sandy base. The description of site CCo-l in the U.C.A.S. records notes: '. occupation area indicated by darkened soil and burials. . . 4' It has been pointed out above that the darkened soil in this region is often due to frequent peaty layers, not to habitation midden--the dark soil at CCo-l seems also to be derived from peat. The cultur-al remains are reported to have consisted of 2 pestles only. Four burials were in "subsoiltt at a depth of h feet, and in a matrix of sterile sand. The site known as CCo-lhO is a sand mound covering perhaps 2 acres Pnd located on a side road pproximea.tely three-quarters of a mile north- northeast of CCo-l. Its surfTce, on which sta.nds a house and barn, is composed of a compact light gray sand, with no sign of cultural occupation. It has never been excavated but since it is a Piper formation its struc- ture mwy be assumed to resemble tha-t of the many similar mounds in the vicinity. An interview with the present owner disclosed only that the disinterment of skeletons in former ditchina or bulldozing operations had been reported. About a mile and a half south of Farrar Park, in the Hotchkiss Tract, is the well-known group of sites., of which CCo-138 is the largest. This Late Horizon site, which has been thoroughly excavated and has been dis- cussed briefly (Lillard, Heizer and Fenenga., 1939, pp. 70-72; Chard et al, n.d.) has a very deep deposit of habitation midden. About 350 yards directly east nf CCo-138 stands a mound which was recorded in 1939 and subsequently designated CCo-139. It was reported to have gray-brown soil and showed several house pits. It probably was -29- of Middle culture, with, perhaps, Late culture occupation on the upper surface. CCo-128, recorded in 1919 by U.C.A.S. as being one-quarter mile southwest of CCo-138, is in an area which has been long cultivated-- nevertheless bones and artifacts were found there at the time of inves- tigation. The soil is "sand" and the site was considered to be 't, . . a campsite on top of sand duncs. . . ." Since no evidence was found of occupation deposit, it is not unreasonable to regard this site as being basic3ally different from CCo-138, i.e., as a site which did not have a cap of Late midden deposit overlying the sterile sand. Another locality '. . . 150 yards south of Hotchkiss site . . also recorded in 1949 by U.C.A.S., was given the designation CCo-129. This spot was also cultivated in 19)49 but the recorder of the site states that burials and artifacts had been removed from "indurated sand dune." The burials were "mineralized ckeletal material obtained by private col- lectors .*" The Hotchkiss area is shown by Cosby (op. cit.) on his map as com- prising outcrops of Oakley sand, and Piper fine sandy loam bordered on the east by Egbert muck and on the south by Sacramento clay lo%nc. The locations of the recorded sites show that the aboriginal inhacbitants evidently took advantage of the sandy knolls which must haeve risen above the neighboring soft 2nd saturated soils. Of these sites, twqo appea.r to have been extensively used for habitation during the Late culture period (CCo-138 139), whereas the other two (CCo-128, 129) rmust probably be classed as burial mounds, without any habitation deposit. Not far to the southeast of the Hotchkiss Tract is Hcllcand Island, on which occur numerous exposures of the Piper fine sandy loam. Most of these have been examined,, and two have been recorded as archaeological sites, CCo-1L6 2nd CCo-1L7. CCo-1h6 lies in the south central part of the island about half a mile north of Rock Slough. The eastern portion of the mound was recently removed by bulldozing, and a cut face several feet in heiaht was made visible. In 195L a group of students in the archaeologi- cal methods field course at the University of California removed a number of burials from this exposure. Examriination of the mound shows a surface layer of dark gray sandy material, rather loose in structure, lying unconformiably on a light colored indurated sandy base. The general organization resembles closely the Piper mounds of Bethel Islmad, and it is clear that the upper layer cannot be occupation deposit. The buri'als were a11 taken directly from the compact sand of the base, and in this instance there was unmistakable evidence that they were not intrusive. The site of CCo-1L7, now destroyed, is situated roughly 1,000 feet north of CCo-146 and is probably of the sane type Ps CCo-1h6. It was reported in 1913 by a local citizen and very little is known about it. Definitely it was not a Late culture occupation mound and any skeletons which have been found must have been derived from the sandy base. What- ever the case, in view of the paucity of data on this site, it will not be included in our considerations here. Due south of CCo-146, and about one mile south of Rock Slough in the Veale tract, is the southernmost "1Piper" site. considered here: CCo-150. This site was recorded in 1939 and is described as lying on an S-shaped sand ridge, a ridge which is shown by Cosby on his $map as being composed of Piper fine sandy loam. The site, and the underlying ridge, lie in a cultivated field, and are now completely destroyed. However the area was seen and described by Lillard, Heizer and Fenenga (1939), who state that there was a habitation deposit on the surface, with a rmaximum depth of three feet, and attrihut--lble to Phase 1, Late Cuiture period. They found no burials from thwe unde,:rlying sandy base. Not shown on Ma-p 1 is site CCo-WI1, which is located cn the west central edge of the Pnlm Tract. Alt:hough this area is within the rrnge of occurrence of Piper fine scrndy lo-:m, Cosbyts rrtp dct-s not show cry specifically sndy type soils whatelver on the Pa,im Tr;t. Site CCc-141 nevertheless is described as located on "a long narrow sand ridge,"' with a mound mass much like tha.t of CCo-150 (see above) "except that it is more indurated" (Ibid., p. 5h). Although there was material of Late Horizon association on part of the site, def inite Middle Horizon artifacts recovered indicate that the latter was the predominating culture present (Ibid., p. 55). Field notes from the excavation (Xrieecr, n.d.) describe removal of burials from a very hard matrix., so hard that "only enough was removed to clear immediate burial earth." While it is not established whether Piper fine sandy loa-m was involved here, it is plain that some type of hardened, sterile-appearing soil, containing burials, was overlain by a softer, sandy cap of midden, representing a later period of time. The preceding descriptions demonstrate that many of the small sand mounds of the western delta region were formerly occupied by man. These mounds are numerous and all have by no means been examined intensively. However, surveys through the last two decades have revealed enough data to warrant certain tentattive conclusions with reference to the age and type of habitation. Discussion must center around two primary features of the sites under consideration. 1. In the period immediately preceding the Spanish conquest of California the region in which most of the Piper sand mounds occur was a portion of the inundated "Tulares," or Great Tule Swamp. The land surface was flat and was at, or only slightly above, sea level. Hence the area was subject to periodic flooding, either daily at high tide, or annually through the flooding of the rivers. For penmanent habitation, or even prolonged camping or fishing, protection from the open water was necessary0 For this purpose the sand mounds must have been ideally adapted. It is therefore not in the least surprising to find extensive recent habitation remains at the Tyler Island group (Sac-45, Sac-104, Sac-162), on Bradford Island (CCo-17), near Rock Slough (CCo-138, CCo-139, CCo-150),, and in the Palm Tract (CCo-141). The intriguing problem is why such occupation sites were not more numerous. The opportunities for occupation were nearly without limit. South of the San Joaquin River there are dozens of entirely suitable small eminences, raised well out of danger from the severest flood. Yet in only four localities, as far as is known at present, was use made of -31- these hpbitation facilities during Late Horizon times. Only the negative deduction is possible that in the marshy western border of the delta flood-proof terrain was not the limiting environmrental factor, and that other pressures, such as food, housing, and. the like determined the den- sity of settlement. The nature and interaction of these influences must be left to future study. 2. In ten of the sixtcen sites recorded prehistoric burials were found in the heavy, compacted, unweathered base of the Piper formation or simila