Chapter III: Introduction (1)

California has but few characteristic archaeological remains such as are found in the mounds of the Mississippi valley or the ancient pueblos and cliff-dweller ruins of the South. In the shellmounds along this section of the Pacific coast it possesses, however, valuable relics of very ancient date. These are almost the only witnesses of a primitive stage of culture which once obtained among the early inhabitants of this region.

Some years ago Professor Merriam recognized the necessity of exploring these ancient mounds and represented the facts to the University of California. Mrs. Phoebe A. Hearst generously made the undertaking possible by providing ample financial support for the exploration work.

UNIV. CAL. PUB. AM. ARCH. &. ETH. VOL. 7, PL. 1

One of the largest and best preserved shellmounds was selected as the object of the present investigation, which was entrusted to Professor Merriam and the writer. The mound selected is situated on the eastern side of the Bay of San Francisco at Shellmound Station near Emeryville, and is commonly known as the Emeryville mound. At present it forms a conspicuous feature of the recreation grounds known as Shellmound Park (pl. 1).

The water of the bay rises to within 130 feet of the base of the mound (pl. 3) during high tide. The beach is then only one foot above the water level, while the ground in the immediate vicinity of the mound is from two to three feet higher. This ground is quite level and forms a part of an extensive alluvial flat. A small creek, having its source about three miles away, in the hills back of Berkeley, passes the mound on its south side, at a distance of two hundred feet, and empties into the bay. In summer the creek runs dry, but its bed furnishes a channel for subterranean water. Another, lower mound, containing graves, lay on the site of the Emeryville race-track near by, but it has been leveled down during the construction of the track. The shellmound which was the object of the excavation has the form of a truncated cone, with a diameter of 270 feet at its base and 145 feet at the top, and rising 27 feet above the plain. On the north side its foot extends 100 feet farther over the flat, a few feet higher than the level of the ground about it.

Twenty-five or thirty years ago the shore line of the bay lay fifty feet farther out; a pile set at that time is still to be seen at that distance from the beach. It is above the water during high tide and marks the coast line on this side of which floodland was sold by the State. The top of the mound was not at that time crowned by the wooden pavilion which is there at the present time. It was still ungraded, having its natural conical form, and was covered with a wild growth of bushes and brambles. The creek, as yet unregulated, followed its own course and overflowed the land, causing it to become marshy. In the seventies and eighties of the last century, railroad tracks were laid along the eastern side of the mound, and took in a section of its eastern foot. At that time a number of graves and Indian artifacts were discovered. Few of these, however, found their way into the collections of the University, then but recently founded.

EARLY SETTLEMENTS IN THE REGION.

Fages, the first traveler who passed through the country, from south to north, traveled along the eastern shore of the Bay of San Francisco in 1774,[1] and came upon Indian settlements where he found a friendly welcome. His account of this expedition however, fails to throw any light upon the question whether or not the shellmounds were still occupied at that time. The neighboring creek bears the name of “Temescal” from a region between Berkeley and Oakland through which it passes.[2] This name appears to be a mutilation of the Nahua word “temazcalli,” hot-house, the name of sweat-houses in Mexico, and the place may have been so named by Mexicans living on the Bay, from an Indian sweat-house standing there. Hence it may be assumed that an Indian settlement was in existence on the banks of this creek at a time from which the name could pass over into the existing vocabulary.

Other evidences of early Indian settlements in this section of the eastern shore country of the Bay are the shellmounds, twelve of which may be found along the coast between Point Richmond and Alameda in a stretch of twelve miles (pl. 1). They may be seen near Point Richmond upon the eastern side, facing the peninsula, upon Brooks Island, near Ellis Landing, northeast from Stege upon a marshy ground intersected by narrow channels, near Seaver’s Ranch to the west from Stege, on Point Isabel, in West Berkeley, in Emeryville, and in the eastern section of Alameda between Mound, Central, and Lincoln avenues. There is also said to have been one in East Oakland on the canal between Oakland Harbor and Lake Merritt, but it has disappeared owing to building over that section of ground. In all probability many others may have met with a similar fate.

All these evidences of an early occupation of the country are but a few of the mounds that skirt the Bay upon all sides, continuing along Suisun Bay and the Sacramento and Feather rivers. Besides these, there are numerous mounds dotting the coast land of Northern California, those surrounding swamps and rivers along the Tulare and Kern lakes in southern California,[3] and on the shore near Santa Cruz. Others are found in the regions of San Luis Obispo,[4] of Santa Barbara,[5] and the islands opposite that place.

-----

[1] Cf. H. H. Bancroft, The Native Races, 1886, II, p. 595.

[2] Cf. also “San Francisco Quadrangle” with the topographical maps of California by the U. S. Geological Survey.

[3] Warren K. Moorehead, Prehistoric Implements, 1900, p. 258.

[4] Paul Schumacher, Smithson. Reports, 1874, p. 335 ff.

[5] Schumacher, Bulletin of the U. S. Geol. and Geogr. Survey of the Territories (F. V. Hayden), 1877, III, p. 73 ff.; F. W. Putnam, Reports upon Archaeological and Ethnological Collections from vicinity of Santa Barbara, Cal., etc.; Report upon U. S. Geogr. Surveys west of the 100th Meridian (G. M. Wheeler), 1879, VII, Archaeology. From more northern sections of the Pacific Coast may be mentioned specifically the shellmounds of Oregon (P. Schumacher, Bulletin, _l. c._), of Vancouver, and of the mainland of British Columbia opposite (H. H. Bancroft, Native Races of the Pacific States, 1886, IV, p. 739), also those upon the Aleutian Islands, explored exhaustively by W. H. Dall (in U. S. Geogr. and Geol. Survey of the Rocky Mountain Region, J. W. Powell, Contributions to the North American Ethnology, 1877, I, p. 41 ff.). Together with those of California these shellmounds are an important counterpart to those found along the Atlantic coast, found from Nova Scotia to the Gulf of Mexico, as well as in the river valleys of nearly all the southern states (Charles C. Abbott, Primitive Industry, 1881, p. 439; Short, The North Americans of Antiquity, 1892, p. 106), and almost all of which have been carefully studied in some of their aspects, although not yet conclusively.

EARLY REFERENCES TO SHELLMOUNDS OF MIDDLE CALIFORNIA.

All the publications treating of the shellmounds of central and northern California, which from the nature of their contents are different from those of the coast and the islands of southern California, may be condensed into the following bibliography:

The Smithsonian Reports of 1869 mention a collection of artifacts from the shellmounds of Alameda county presented to the Institute by Dr. Yates.[6] J. W. Foster, in 1874, speaks of a newspaper notice concerning a shellmound in the region of San Pablo.[7] James Deans follows in 1876 with a short notice (together with drawings of some artifacts) concerning a mound between Visitacion Valley and Point Bruno on the western shore of the Bay.[8] A short notice by H. H. Bancroft, accompanied by views of four objects, points to the great historical value of the shellmounds. The Marquis de Nadaillac in his well known work mentions the shellmounds in the vicinity of San Francisco.[9] Moorehead in his work gives a few remarks on excavations in shellmounds of central California.

-----

[6] Smithson. Reports, 1869, p. 36.

[7] Prehistoric races of the United States of America, 1874, p. 163.

[8] Journal of the Anthropological Inst. of Great Britain and Ireland, 1876, V, p. 489. The majority of these shellmounds have been graded down.

[9] Prehistoric America, ed. by W. H. Dall, 1885, p. 50.

THE NATURE OF THE EXCAVATIONS.

The work of exploration was commenced by Professor Merriam and the writer in February, 1902, toward the end of the rainy season, and was finished early in May. Captain Siebe, the proprietor of Shellmound Park, gave all possible assistance in the investigation. Owing to the presence of the circle of trees around the truncated top of the mound it was necessary to confine the excavations to a lateral section and a tunnel extending from it toward the center of the mound. However desirable a more extended section through the hill might have been, the results obtained in these partial excavations are as a whole similar to those which would have been obtained by a cut through the entire mound.

The western slope of the mound, facing the bay, was selected as the starting point for the operations. The entire work of excavation may in a chronological order be divided into the following four stages.

_A._ The first lateral cutting in the mound. This was made in the western foot of the mound, seven feet and a half above the level of the bay and at a distance of fifty feet from the plateau. The trench was two feet deep, eighteen feet long and six feet wide, its floor sloped towards the center of the mound.

_B._ Tunnel construction. The tunnel formed the underground continuation of the trench; it was the means of reaching the interior of the mound and down to its original base. Hence the floor of the tunnel was made to slope steeply inward. The tunnel was extended from the end of the trench A for forty-two feet into the interior of the mound, and at its terminal point it sank to two feet below the level of the bay. It was five feet wide and six and a half feet high. Several distinct strata were cut through by the tunnel section. Eleven feet of the length of the tunnel extended under the plateau of the mound. This was still sixty feet from the vertical center of the hill (pl. 4), but the observations made in this interior part of the mound were of a relatively greater value than those of the outer zone. Many difficulties were met during the construction of the tunnel, among which the porosity of the soil was one of the worst. The tunnel was therefore timbered and its sides sheathed. Another difficulty was the ground water, of which there was often a very strong flow when digging in the lower part of the tunnel. According to the advance of the season, it was encountered at different depths, and it grew less with the approach of summer. A small hand pump was used to exhaust this water, but it barely answered the purpose, and it was often with great difficulty that the inrushing water could be mastered.

_C._ The upper vertical cut of the entire mound. In order to obtain a view of all the strata contained in the mound this section was undertaken. The lowest parts of the mound having been thoroughly explored by the construction of the tunnel, it was now sufficient to make the upper sectional cut only as deep as the roof of the tunnel, while its terminal point was fixed by the circle of trees on the summit of the mound. Its greatest length from the mouth of the tunnel was twenty-six feet. The sides of the cut were sloped in order to prevent the fall of loose soil and to avoid the cost of timbering. The length of this section at its lower end, near _b_ (pl. 4), was reduced from 26 feet to 19 feet, and the width to 10 feet along the entire foot of the trench from _a_[10] to _b_. In pl. 5 there is shown the first cut into the mound, before it had been made wider by five feet throughout its length. In making this cut the earth was removed stratum by stratum. For want of other marks of division, the dividing lines of the various strata (I to VII) were chosen arbitrarily from the several visible lines of structure, and they are marked in the diagram, pl. 4, by asterisks. In order to obtain a uniform classification of the contents of the mound it was thought necessary to introduce the same lines of division in the sectional diagram of the tunnel: objects found there had been marked previously by the distance of their position from the mouth of the tunnel and their relative height. These strata in conformity with the numbering of the upper ones were marked as numbers VII to X.

_D._ A series of pits was dug from the foot of the tunnel out to the bay shore. The pits were made in order to ascertain the general outline of the base of the mound under the cuts already made, as well as under the unexcavated portion of the mound farther out toward its margin. The pits are marked as _h_ in the interior of the mound, and as _i_, _k_, _l_, _m_, toward its periphery. The two pits _n_ and _o_[11] are situated on the outside of the superficial foot of the mound, at a distance of 35 feet and 67 feet from the nearest pit, _m_. It was here seen that the terminal point of the foot of the mound lay between the pits _n_ and _o_, the pit near _n_ showing only the debris of the shellmound, while that near _o_ revealed nothing of it. These two pits were connected by a trench, which gave an exceedingly interesting section of the margin of the mound.

-----

[10] _a_ seems to have been situated at the intersection of the dotted lines separating divisions _A_, _B_, _C_, pl. 4, fig. 2.—Editor.

[11] it _o_ referred to in the text seems to be represented in pl. 4, fig. 1, by the west end of the cut extending from _n_ to _l_.—Editor.

THE BASE OF THE MOUND.

The mound consists mainly of a mass of broken or entire shells, ashes, bits of charcoal, and some artifacts. This mass extends far above the surface of the surrounding land and ends two and a half feet below the level of the ground water and two feet below the general tide level of the bay, and rests immediately upon a sharply defined yellowish alluvial clay stratum. There is no indication of a rocky elevation which might have served as an inducement for the original settlement, and would have helped to raise the mound to its present height. Some of the charcoal and small boulders brought here by man rest upon the clay soil. A slight discoloration of the upper line of the clay stratum may have been caused by a transitory plant growth during some early period, while there is no indication of a crust of good soil which would be a sign of a longer period of vegetable growth upon it.

The base of the mound is horizontal according to all indications gained between pits _h_ and _m_. A slight variation of the level of the ground near _h_ of but a few inches does not materially change this level. Between _m_ and _n_, however, the original soil lies one foot and seven inches lower for a distance of thirty-five feet, and from _n_ to _o_ the level drops a foot lower. The mound was originally founded upon a site rising two feet above the adjacent ground on its western side. A gravel stratum of 8 inches in thickness near _o_, and of 4 inches near _p_, but disappearing towards _n_, covered the clay which originally sloped to the west. This gravel stratum was examined by Professor Lawson and considered to be probably a fresh-water deposit and not a deposit formed in the bay, as the gravel is more or less angular instead of much water-worn. The mound terminates near _p_, 177 feet from its center, where it runs to a point between layers of clay, which are above and below it (pl. 4, fig. 1). It rises again toward the outside for the last 17 feet measured from the depression _n_, the difference being one and one-quarter feet, thus varying from the rest of the base which inclines to the west. A stratum of ferruginous clay, the same as that underlying the base of the mound, is here inserted between the gravel stratum and the characteristic mixture of which the mound is composed, and covers it up even with the present surface of the soil. This raises the actual height of the shellmound from 27 feet to 32 feet and the actual diameter to at least 310 feet instead of 270 feet. The volume of the mound, measured as a truncated cone, may be estimated as being 55,000 cubic yards, or about 39,000 cubic meters.[12]

From what we know of the situation it is obvious that the mound was rounded upon firm though still somewhat marshy land, near the bay shore and close to the creek. The latter was the occasion of its location[13] at this place. The ground must have been dry, since a gently rising slope was selected. The soil was alluvial and relatively new, since it has no overlying cover of good earth, yet it must have been dry long enough to allow a thin growth of vegetation to cover it, causing the slight gray discoloration of this stratum.

The situation of the base of the mound two feet below the water level cannot be explained on the assumption that refuse from a pile dwelling had been the first cause of its formation. This theory would presuppose modes of living to be followed by the Indians of this coast for which there is no parallel elsewhere, and which are not borne out by other evidence obtained in the study of the mound. If the mound has not risen from the water, then the former land surface must have sunk. The mound could not possibly have sunk below the water level from its own weight, for the original ground underneath it is still several feet higher than that to the west, for instance, near _n_, and sections of the base upon which the full weight of the mound rested, such as near _h_, are on the same level with others over which the mound rose only 14 feet. Since the sinking of the mound has not been brought about by local causes, it must have been caused by a general subsidence of this coast region. Similar subsidences of the coast, due probably to sliding motions, are frequent phenomena on alluvial coasts.[14] Evidences of this are furnished apparently by the shores of San Francisco Bay.[15] The ground under the mound having a slope of two feet, it may be assumed that the original foundation of the base was at least one foot above tide level. Accordingly the coast must have sunk three feet since the formation of this mound.[16] This sinkage was leveled up again to its former height by later alluvial deposits, in consequence of which the originally dry base of the mound is now situated two feet below the level of the bay, while the surrounding flats are three feet above it.

It is to be noted that the younger alluvial deposit, near _o_ (pl. 4) has a thickness of six feet.

Samples of soil taken from various parts of the clay stratum underlying the base of the mound were subjected by Professor W. A. Setchell to microscopical examinations, but no Diatoms were found in any of them. Hence those strata were probably formed of alluvial deposits of the creek, as Professor Lawson had at first suggested, and not of deposits of the bay. This finding is entirely in accordance with the origin of the gravel stratum as above stated.

The slope of the mound was an obstacle to the course of the creek when it became swollen. In the natural course of things it deposited a bar near the foot of the mound, which, when the edge of the latter gradually extended, grew out over this new obstacle. The creek in the same manner continued to heap up alluvial deposits against the latter. The horizontal growth of the mound and the vertical growth of the surrounding land took place simultaneously. This was the cause of the brim-like upward curve of the edge of the mound as seen in the cross section (pl. 4). While the mound increased about seventeen feet in its periphery, the vertical alluvial accumulation was about one and one-half feet. Hence the base of the mound peripherally increased one foot while the ground grew one inch, showing that the alluvial growth of the soil was much slower than the peripheral growth of the mound. About 310 cubic yards or 240 cubic meters produce a growth of one foot in a mound 9 feet high and about 300 feet in diameter at the base. If the peripheral growth of the mound had continued with the growth of the soil, the foot of the mound would have spread out so that the outer edge would rest in the highest or surface layer of the present alluvium. The wedge-like margin situated between alluvial strata is, however, proof that its peripheral growth ceased a long time before the termination of the alluvial accumulation in this region, as a result of which the alluvium has spread itself over the foot of the mound. The alluvial deposit above the wedged-in margin of the mound (at _p_) being 3 feet 8 inches in thickness, and the alluvium deposited underneath it from the beginning of the formation of the mound measuring only 1-1/2 feet, and assuming the increase to have been absolutely uniform, a period two and a half times as long has passed since the ceasing of its peripheral growth, as had been necessary for a peripheral growth of 17 feet on each side. The cessation of this peripheral growth of the mound, however, is not identical with the cessation of its growth altogether. It took place apparently when the mound began to grow more acutely conical in shape, whereby it increased to twice its former volume. Assuming that the mound was abandoned 100 years before the end of the alluvial growth of the land in the vicinity, then according to formula

100 × 2/3_f_ = 2-1/2 × 1/3_f_

it might be concluded that the mound was probably 600 years old before it was abandoned.[17] Several numerical values upon which the formula is based are unfortunately so uncertain that the result may not be considered as more than suggestive of the possible age.

The sinking of the coast and the alluvial increase of the ground since the first settlement of the mussel-eaters would in themselves give an adequate measure for an estimate of the age of the mound if the measures upon which both depend were not also unknown; according to Professor Lawson, this probably occupied centuries at least.[18] At any rate, such observations as have been made furnish good reasons for believing that the founding of a settlement and the beginning of the heaping up of the mound occurred at a remote date.

-----

[12] The shellmounds in the vicinity of the bay differ considerably in shape and size. The majority appear as extended plateaus 10 to 12 feet in height, others appear as slight undulations of the ground about five feet in height. The truncated conical form is found more rarely; the mound at Ellis Landing near Point Richmond approaches it somewhat in its proportions. Many of these mounds cover acres of ground, _e.g._, the mounds of Alameda, of Sausalito, of Sierra Point, of West Berkeley (in its older form, now much changed). In tropical regions many shellmounds are said to reach a height of 100 feet or more; this is known with certainty of some in Brazil (cf. Nadaillac, _l. c._, p. 54), and also of two near the dried-up mouth of the Ica river in Peru. Shellmounds as a rule are much smaller. On the Atlantic coast near Smyrna a shellmound is said to be thirty feet high (Short, _l. c._, p. 107), but the majority of these mounds are less than four feet high (cf. Wyman, Amer. Naturalist, 1868, I, p. 56 ff., and Abbott, _l. c._, p. 440), while many of them extend over areas of more than two or three acres. A shellmound near the mouth of the Altamaha river in Florida is estimated as having a size of over 80,000 cubic yards (Smithson. Rep., 1866, p. 358). The shellmounds of Denmark are only from 3 feet to 10 feet high, although more than a thousand feet long (Ranke, Der Mensch, II, p. 552). Southern California shellmounds generally are from 4 feet to 5 feet high (P. Schumacher, Bull., _l. c._, p. 38; and Smithson. Rep., 1874, p. 337, etc.). The same is the case with those mounds on the Aleutian Islands explored by W. J. Dall. In Oregon there are some of at least 8 feet in height (cf. Schumacher, _l. c._, p. 29).

[13] Shellmounds in the bay region are mostly in localities where there is fresh water, a creek or a spring, generally the former. W. H. Dall (Contributions, p. 34) observes that for the formation of shellmounds on the Aleutian Islands two conditions are necessary, as a rule: running water or a spring, and a site suitable for boat landing; one or the other of these conditions lacking, no shellmounds are to be found. In Oregon the shellmounds are generally to be found near a creek (cf. Schumacher, _l. c._, p. 28). The same rule probably governs the shellmounds of the East. D. G. Brinton found shellmounds in Florida generally near running water (Smithson. Rep., 1866, p. 356), but he supposes as the cause of this the greater abundance of shells near the mouths of rivers, while it is certain that the presence of drinking water was the main attraction.

[14] Parts of the eastern coast of the United States are sinking. Several shellmounds on the Jersey coast are being washed away at present (cf. Abbott, _l. c._, p. 448 ff.). The same may be observed with the shellmounds near Ellis Landing on the Bay of San Francisco.

[15] Near the mouth of the valley of San Rafael a small hill rises from the bay, the isolation of which from the mainland may be explained in this way.

[16] Between the shellmounds of Emeryville and West Berkeley the shore for a long stretch forms a steep bank up to twelve feet high, and broken down by the water of the bay. Possibly the coast at this point formed a promontory on the two sides of which these shellmounds were originally founded, as in sheltered bays, similar to other mounds of this region.

[17] In that case the sinkage would have amounted to about 6 inches, the alluvial increase to about 9 to 10 inches in a century.

[18] The rapidity of the sinkage of alluvial coasts varies greatly owing to local conditions. For the Atlantic Coast the rate of sinkage is 2 feet per century (cf. Abbott, _l. c._, p. 449). Applying this same rate to the eastern coast of the Bay, we would arrive at the absurd result that the shellmound of Emeryville had begun to form in 1750, while that date was presumably the end of its occupied state.

THE INTERNAL STRUCTURE OF THE MOUND.

The principal constituents of the mound are the shells. These have nearly all crumbled into small fragments and are slightly mixed with soil, which when damp gives the entire mass the appearance of pure soil. When this is flooded with water the washing away of the sand produces no noticeable change in its volume. This mass has mingled with it bits of charcoal, bones of animals, ashes or cinders, and stones averaging about the size of one’s fist and blackened by fire.[19] Marks of stratification may be traced through almost the entire mound. Plate 5, representing a photographic view of the excavation, shows the stratification planes in the walls quite distinctly. The strata consist of compact masses of more or less fragmentary shells, or of beds of ashes or cinders. In many cases the latter seem to extend through the entire mound. They are sometimes not thicker than a sheet of heavy paper, but show the general direction of the bedding planes, and form a clear contrast with the homogeneous, dark mass of broken shells.[20] These planes become somewhat less distinct in the deeper strata.[21] As in other shellmounds,[22] there were observed certain rounded masses of shells intersecting the lines of stratification. These are caused by holes, made by moles or other burrowing animals, being afterward refilled with shells.[23]

In some shellmounds in other regions strata of earth and sand were found between the shell layers. These give evidence of a temporary evacuation of the shellmound. No evidence of this character was obtained in the study of the Emeryville mound, where the only occurrence of a natural vegetable soil is the surface cover of 1 to 2 inches in thickness, which has formed since the mound was finally abandoned.[24] It is possible that slight differences in the state of preservation of the shell deposits which now mark the strata lines may have been caused by differences in the length of time of occupation. Other explanations might, however, be offered.

The lines of stratification mark clearly the gradual development of the strata of the mound from the base until the present truncated cone was formed. It is apparent that two different principles governed the growth of the mound. At certain periods it tended to take on a shallow plateau form. At other times a conical shape developed without the corresponding increase around the base. According to the first principle the mound grew in the form of a plateau to a height of from 9 to 10 feet. Near _C_ in pl. 4 the edge of the plateau still seems to be traceable, from which point the strata inclined downward. At that period the mound resembled in its proportions the old Indian camping places of the interior valley, some of which are still occupied; or some of the shellmounds along the Bay which have been abandoned at some earlier period. The undulating lines of the strata, such as seen near _f_ and _g_, suggest irregularities of the old plateau surface, similar to those which may be observed in the surfaces of camp locations of the interior, which have been abandoned for decades. The hollows from 20 to 40 feet in length mark the sites of former sweat-houses or council-halls; these curves, such as that from _f_ to _g_, may have a similar origin.[25]

The manner in which the mound was occupied for habitation varied in the upper strata. With the growth of the mound the diameter of the plateau decreased instead of expanding. From line _b_ upward the strata incline obliquely toward the sides. This change in the manner of forming the mound signifies a change in the character of its occupants. It would be interesting to determine, if possible, the exact line where these two types of growth have met. It might have been about 12 feet above the base, so that the mound grew in the shape of a shallow plateau as far as the middle of stratum V in pl. 4, and that it changed after this period to its conical form.

-----

[19] The descriptions of nearly all the shellmounds explored in other parts of the world tally exactly with this one; cf. Ranke, _l. c._, II, p. 532, for the Danish Kjökkenmöddinger; Schumacher, on the general similarity of shellmounds of the Pacific Coast with the mounds in Denmark, Smithson. Rep., 1874, p. 355, etc.

[20] Although no shellmound is free from stratification marks, owing to the gradual growth of the strata, Brinton maintains that this is the case with shellmounds on the Atlantic Coast (Smithson. Rep., 1866, p. 356).

[21] Compare the interesting observations of Wyman (Amer. Naturalist, I, p. 571) concerning shellmounds of New England, that there the shells of the lowest stratum were softer and more crumbled than those of the upper strata.

[22] Cf. Wyman, _l. c._, p. 365, on a shellmound in the vicinity of Portland, Me.

[23] Similar holes made by moles may be observed occasionally upon old shellmounds along the Bay, which if they had been refilled with shells might also have assumed a rounded form. In such a manner may be explained the finding of a modern steel knife, with the wooden parts still well preserved, in one of the strata of the shellmound of West Berkeley in a place to all appearances undisturbed.

[24] Cf. also Wyman, _l. c._, p. 571. The absence of true soil from the interior of the mound is proof that at no time was the mound abandoned by its occupants long enough to allow of the formation of such a stratum.

[25] Somewhat smaller but quite similar hollows are still preserved upon the surface of the shellmound of Ellis Landing, and are doubtless sites of houses of that nature.

CONSTITUENTS OF THE MOUND.

_Shells._—The shell layers of the mound are composed principally of the following species:

Oysters, _Ostrea lurida_.
Mussel shells, _Mytilus edulis_ and _Mytilus californianus_.
Clams, _Macoma edulis_ and _Macoma nasuta_.

Many other kinds of shells, including the following species, were found scattered through the mound:

_Purpura crispata_ and _canaliculata_.
_Cerithidea californica._
_Helix_, two species indet.
_Cardium corbis._
_Standella_, sp.
_Tapes staminea._

Of these last species, the cockle, _Cardium corbis_, and the clam, _Tapes staminea_, occur quite frequently.[26] All of these were used as food by the occupants of the mound. The various species of _Helix_ were probably also used, as they were in more recent times eaten by the California Indians.[27] It may be, however, that this species lived on the mound.

The state of preservation of the shells is proportional to their natural hardness. Hence the shells of the Macomas are the most conspicuous, those of the mussels, as the most perishable, are the least noticeable ones in the mound. The relative frequency of occurrence in the case of the three most important species depends on different circumstances.

The lower and the upper strata of this mound are composed of the same varieties of shells, in which point it is different from many shellmounds in other regions. It is, however, true that oyster shells predominate in the lower strata, while _Macoma_ shells are more numerous in the upper ones.[28]

Visiting the different shellmounds in the vicinity of the Bay, one finds a general similarity in the kinds of shells composing them. Rarely one or another variety of shell, the _Macoma_ or the cockle, or some other, is found to predominate. This general homogeneity of composition in the shellmounds around the Bay, and the small differences in the amount of any particular species, indicates as a whole the general similarity of the shell fauna at many points about the Bay during the period of occupation of the mounds.

The Indian camping grounds in the interior, although quite similar in form and origin to the shellmounds on the coast, when opened generally present a great difference in appearance. Traces of shells are almost unnoticed from the outside, yet large quantities supplied as food by the rivers of the interior are doubtless to be found in them. These shells have been found during excavations, or their use has been confirmed by persons who observed the mode of living of the Indians of these regions. The Indians also obtained salt-water mussels by trade, even in quite recent times. From the fact that shells are not in evidence on the surface of the camp grounds, one must conclude that their use diminished.

_Bones._—Bones of vertebrates are also found in most of the shellmounds. These together with the shells represent the debris of their kitchens. No other shellmound has been seen where so large a quantity of bones was observed as in that at Emeryville. Bones of land and sea mammals, of birds, and of fishes were found in abundance throughout the mound, and fairly evenly distributed in the strata. This fact is the more remarkable since the shellmound at West Berkeley, scarcely two miles distant, does not yield nearly such quantities of bone as this one. The occupants of the mound at Emeryville at all periods were huntsmen to a great degree, besides being fishermen; those of the mound at West Berkeley seem to have depended largely upon fishing; hence the stone sinkers were far more numerous in that mound than at Emeryville.

So far the fauna of only the lowest strata up to 3 feet above the base have been studied. The following species obtained in this horizon were determined by Dr. W. J. Sinclair.

Deer, _Cervus_ sp.
Elk, _Cervus canadensis_.
Sea-otter, _Enhydrus lutris_.
Beaver, _Castor canadensis_.[29]
Squirrel, _Spermophilus_ sp.
Rabbit, _Lepus_ sp.
Gopher, _Thomomys talpoides_.
Raccoon, _Procyon lotor_.
Wild cat, _Lynx_ sp.
Wolf, _Canis_ sp.
Bear, _Ursus_ sp.
Dog, _Canis familiaris_.[30] (_?_)
Seal, _Phoca_ sp.
Sea-lion.
Whale.
Porpoise?
Canvasback Duck, _Aythya vallisneria_.
Goose?
Cormorant, _Phlaeocorax_ sp.
Turtle.
Skates, Thornbacks, and other fish.

No traces of cannibalism have been detected. Most of the hollow bones of larger mammals, and even the smaller bones of the foot, were found to have been split to get at the marrow.[31]

_Fireplaces._—These were generally known by beds several feet in length consisting of charcoal and yellowish ashes. They occurred in many spots throughout the mound. Numberless scattered bits of charcoal[32] and pebbles, mostly about the size of one’s fist and blackened by fire, were further evidences of the continuous use of fire in the preparation of food. In no instance were there any stones set in rows for fireplaces, such as have been observed elsewhere, as in a shellmound near Sierra Point, where stones are plentiful.[33] A very peculiar feature of this mound is a yellowish layer of ashes comprising the entire depth of stratum II in pl. IV, and tapering towards the edge of the mound. Above it lies only the uppermost stratum (I), that of vegetable soil. Though calcined shells[34] occurred elsewhere in the mound, they were especially numerous in this ash stratum, and in some spots all shells were calcined. The origin of this ash stratum will be explained later. A similar bed is to be seen in a central layer of the shellmound at West Berkeley, and another one of similar thickness but shorter in a mound near Sausalito.

_Human Remains and Relics._—A large part of the Emeryville mound consists of remains which have been deposited here by man. Among these are molluscan shells with bones of fish and mammals, used as articles of food. In the narrower sense the human relics consist of the bones of man, graves, and artifacts, which are all found in greater or less abundance throughout the whole thickness of the mound. Actual human bones were not found to be common in this part of the mound except in stratum II, and in the graves of stratum VII. The artifacts obtained were only those of very resistant material, such as stone or shell. All other kinds, such as textiles of plant fibre, baskets, and implements of wood, which doubtless have also existed, had disappeared. The more resistant artifacts were distributed throughout all layers of the mound.[35]

About 200 cubic meters of earth were removed and sifted during the excavations, and yielded 600 artifacts of various kinds, averaging three specimens to one cubic meter. The volume of the whole mound we computed to be about 39,000 cubic meters, and it may be assumed that by excavating the entire mound the yield would be about 100,000 specimens, which indicates that many generations must have lived here to deposit such a large number of objects of imperishable material alone.[36]

The same computation was applied to each separate layer in the mound, and it was shown that the yield differed according to the section and the stratum explored. The open cut _A_ yielded one specimen to .75 cb. m., and the tunnel _B_ and the pits _h_ to _m_, six per cb. m. Section _C_ yielded three artifacts to one cb. m. This computation shows that sections nearer the center of the mound yielded the greater number, those toward the edge a smaller number. It also appears that the lower strata contained a larger percentage of artifacts than the upper ones. If, however, the number of flaked chert fragments were subtracted from the yield of the lower strata, their percentage would be much the same as that of the higher layers. The following are the contents of the various strata:

Stratum I had 20 artifacts per 15-1/2 cb. m. = 1.3 per cb. m.
Stratum II, 30 cb. m—133 objects = 4.4 per cb. m.
Stratum III, 20-2/3 cb. m—27 objects = 1.16 per cb. m.
Stratum IV, 11-3/4 cb. m.—41 objects = 3.5 per cb. m.
Stratum V, 9-2/3 cb. m.—34 objects = 3.5 per cb. m.
Stratum VI, 4-1/5 cb. m.—9 objects = 2.1 per cb. m.
Stratum VII, 2-4/5 cb. m.—10 objects = 3.5 per cb. m.

The specimens contained in the graves in strata VI and VII were not counted in with the rest. This comparison shows mainly that stratum II is the richest in implements. The connection of this fact with the preponderance of ashes will be pointed out later.

-----

[26] Eight-tenths of all the shells found in the Oregon shellmounds belong to the species of _Mytilus californianus_, _Tapes staminea_, _Cardium nuttalii_, and _Purpura lactuca_ (Schumacher, Smithson. Rep., 1874, p. 335).

[27] As by the Minooks and the Nishinams (Powers, _l. c._, pp. 348 and 430); and certainly the custom was a very general one.

[28] We were not so fortunate as was W. H. Dall in the shellmounds of the Aleutian Islands in being able to make “a tolerably uniform division” of the layers in the mound according to the various foods used. (These layers were: “1, Echinus layer; 2, fishbone layer; 3, hunting layer.” Contributions to North American Ethnology, I, p. 49.) The shellmound of Emeryville presents a much greater similarity in the kinds of food used during the different periods of its occupancy.

[29] Extinct in California, and in fact south of Washington; J. Wyman found the remains of elk, wild turkey, and large auk in the shellmounds of New England. The elk, though still in existence, is no longer to be found east of the Allegheny Mountains; the wild turkey is still in existence, but is not to be found in New England, while the auk lives only in the Arctic regions, or at least not farther south than the northern part of Newfoundland (Amer. Naturalist, I, p. 572).

[30] Also found in the shellmounds of New England.

[31] Cf. for shellmounds in Denmark: Ranke, _l. c._, II, p. 532, for those of the Atlantic Coast, Wyman, _l. c._, p. 575 (New England) and Abbott, _l. c._, p. 442 (New York).

[32] Analogous is the statement of Wyman, _l. c._, p. 564, about the shellmounds of New England.

[33] Cf. also Hellwald, Der vorgeschichtliche Mensch, p. 449, on the Kjökkenmöddinger of Denmark.

[34] Nadaillac, _l. c._, p. 50, states from uncertain authority that a shellmound near San Pablo was said to consist of calcined shells exclusively, which is certainly an exaggeration.

[35] It is alleged that there are shellmounds in the East which contain no implements at all, and have been used for the gathering of mussels only, and not as dwelling places (Abbott, p. 447, accord. to Charles A. Woodley). Equally uncertain seems to be the distinction made by Schumacher between shellmounds yielding few artifacts and those containing a larger number, as representing a place for temporary or permanent habitation. Similarly dubious is that classification which considers the piling up of shells in various separate heaps as proof of permanent abode and that of single mounds for the use only as temporary stopping places (Smithson. Rep., 1874, pp. 337 to 338).

[36] W. J. Dall (contrib. _l. c._, I, p. 47) states that during his excavations of the shellmounds of the Aleutian Islands he found on the average one object in one-half ton of earth. This would be 2.63 objects to one cb. m. The yield of the Emeryville shellmound is three objects to one cb. m.

BURIALS.

Shellmounds originate on the accumulated refuse deposited by people who have lived in the place when the heap has formed, and the mounds may therefore be regarded as sites for dwelling places, or abodes for the living, and not as mounds set aside as burial grounds by people living elsewhere in the vicinity. Whenever these mounds were used for burials it was not done in spite of their being dwelling places, but rather because they were such.[37]

Many tribes of a low grade of civilization follow the custom of burying their dead underneath their feet in the ground upon which they live, to protect the graves of their dead against being disturbed and also to enjoy the protection of the spirits of the departed against their enemies. Wherever graves are found in shellmounds, in all parts of the world, their presence is generally to be explained in this way.[38]

Ten graves containing skeletons were found during the excavations. They were found only in the middle layers of the mound in a zone extending from stratum VI to stratum VIII. The two lowest layers and the five upper ones contained no evidence of interment, indicating that the custom of burial underneath the dwelling places was observed in one period only. We have no evidence concerning the location of the burials previous to that period or subsequent to it. A burial site slightly elevated above the plain was unearthed some years ago under a shellmound near the principal mound in Emeryville, but as this probably dated from the same period as the graves in the shellmound no light is thrown upon the question.

In the upper strata of the mound there is, however, furnished evidence of a different manner of disposing of the dead, which was observed during the period of the deposition of strata II, III, and IV. During the period represented by strata VI to VIII the dead were buried in the ground. It has already been shown that stratum II consists mainly of ashes and calcined shells, which cannot be regarded as kitchen-midden deposit or as the remains of fireplaces, the latter forming an inconspicuous part of the stratum. Another characteristic feature of this layer is the high percentage of calcined bone implements found there. Very common among them are awls, of which stratum II contained 44 calcined specimens, or 72 per cent. of the whole number. In the other layers a much smaller number has been found, but the percentage of calcined specimens is high. The supposition that these were accidentally burned cannot be considered an adequate explanation, but the fact that a number of human bones were found at the same place in the strata gives weight to the theory that during the deposition of the upper beds the inhabitants of this region practiced cremation of their dead, a custom common among the California tribes of today. Then as now they were accustomed to burn all personal belongings with the body.[39] This accounts for the large number of calcined bone objects and shells in stratum II. Doubtless a large number of shell-fish were thrown into the fire as food for the departed on their long journey into the next world.

Doubtless the practice of cremation was not confined alone to the period of stratum II. The percentage of calcined awls in other strata than this suggests that the builders during the accumulation of stratum I, and probably also of III and IV and parts of V, practiced this custom, but to a less extent than in the period of stratum II, or mainly at other places than the mound.

Eight of the ten graves containing skeletons lay close together in the narrow space of the middle section of the excavation. Four were found in strata VI and VII of the upper cut _C_. Four of the graves were those of children, lying at different depths in the line of the tunnel. Two other graves were isolated from the others, lying in the edge of the mound. None of the burials were less than nine feet below the present surface. The lowest grave, No. 10 (pl. IV, fig. 2), was 21 feet below the surface. In accordance with the stratification lines of the mound, graves 1 and 2, as well as 6, 8, and 9, may be considered as belonging to the period of strata VII and VII_a_. The eight graves which lay close together in the middle of the main excavation were distributed over a space of about 90 square feet. The vertical distance from the uppermost to the lowest of these was nine feet. As the tunnel inclines toward the center of the mound it is not certain whether the cemetery extended through the entire mound or was only around the margin of a settlement on the summit of the mound at the time when it was in use. From the depths at which the different graves were found, varying about nine feet, it is clear that they were not constructed within a short period, but that long periods intervened, during which the mound grew very considerably through the deposition of refuse. P. Schumacher explored the graves of Oregon, which lay at a depth of from 1-1/2 to 2-1/2 feet below the surface, and probably the tribes of the Pacific Coast buried their dead in comparatively shallow graves.[40] In the case of a child’s grave (No. 9, pl. 4, fig. 2) it was seen by the stratification lines that it was not made deeper than 1-1/2 feet below the surface. Assuming this as being the general depth of the graves throughout the mound, then the graves varying between 2-1/2 and 12-1/2 feet above the base of the mound were dug at periods when the entire height of the mound was about 5 to 14 feet, hence the period of these burials would have to be placed entirely during the time of the earlier plateau-like growth of the mound. This period of burial is very closely followed by that of cremation, the two possibly overlapping.

The preparation of the graves was not elaborate. A simple pit sufficed. It was made large enough to place the body in it with the knees drawn up. The sides of the grave were left bare. If a covering existed originally it must have been of perishable materials, for none have been found in excavation. The bottom of the grave, however, was prepared somewhat like a bed. A layer of charcoal from one-half inch to one inch thick is found at the very bottom, above that another layer of like thickness of iron oxide. Upon this the body was laid on its side. It is evident that the body was buried with its clothing and personal ornaments, in exceptional cases with utensils or implements only. The body was tightly bound at the knees before burial. Owing to climatic conditions, causing excessive moisture in the strata, the greater part of the material buried with the corpse has decayed and disappeared. Five of the ten graves were entirely lacking in implements or other artifacts.