Caddo Parish, in the northwestern corner of Louisiana is bounded on the east by the present channel of Red River. In the vicinity of the Belcher Mound site (Fig. 2) this river follows the eastern escarpment of hills for about 15 miles, with all of its overflow valley lying to the west and varying in width from three miles at the upper part of this stretch to ten miles at the level of the site. The hills to the west of the valley are generally lower than the eastern escarpment and are cut through by several streams, including the Caddo Lake outlet, which follow the western valley margin to the Shreveport level, where they enter the river. The valley floor is covered by mixed red loam and sand, the characteristic overflow deposit of Red River. Along the channel sandy soils form elevated natural levees (“front lands”) so that the numerous small streams which traverse the valley floor have to flow downslope to the outlet of Caddo Lake. The western portion of the valley (“back lands”) was formerly covered by shallow lakes with a few islands of sandy loam. The soil of the old lake beds is a tight red clay containing numerous fragments of mussel shells; one of these lakes, Soda (originally Soto?) Lake, is shown in early maps to extend from the Caddo Lake outlet to the Shreveport level.

Houses 1 to 5 of this site have been described in a previous publication (C. H. Webb 1940), with some comparisons with historic house types among the Caddos, other Southeastern tribes, and Plains groups. Archaeological evidences of Caddoan house types were cited from the descriptions of Jackson (1936) and Harrington (1920). Since that time, however, other excavations in the Caddoan area have given a fuller picture of house structures, hence a repetition of descriptions of houses at the Belcher site together with a more complete comparison seems to be in order. It has been indicated that Houses 1 to 4 were immediately superimposed (in reverse order) in Mound B (Fig. 5), as were Houses 5 and 6 on the village area and 7 and 8 in the lower levels of Mound A (Fig. 6).

Four types of burials are to be distinguished at the Belcher site. The first group, Burials 1, 4, and 13, were disturbed burials, to the extent that details or relationships were indefinite. The second group, Burials 11 and 12, were in large, wide pits which originated within the primary mound of Mound B, just beneath the red clay cap. The long axes of the pits were on a northeast-southwest line, with the heads directed toward the southwest. The third type, bundle burial, is represented by Burial 21 alone. The fourth and largest group, consisting of all other burials, comprised single and multiple burials in relatively narrow pits which were placed on a northwest-southeast axis, with the heads of the burials directed toward the southeast. This group related to the last two habitation periods, which we have assigned to the Belcher focus.

The pottery from this site was described and classified in 1941 (Webb and Dodd 1941), on the basis of the sherds found in the four levels of Mound B (Houses 1 to 4), the floors of the adjoining Houses 5 and 6, and the fills between these levels, in addition to whole vessels from the associated Burials 1 through 19 and several of the house floors. This was the first attempt to set up pottery types from an excavated Caddoan site and was necessarily tentative. The classification has been subjected to continuing study and modification on the basis of reports from subsequent excavations and surface surveys in the four-state area.

Much of this reclassification has resulted from correspondence and conversations with Alex Krieger as he developed the Texas materials, from the two Caddoan Conferences held at Norman, Oklahoma, in 1946 and 1952, and from several informal conferences with Krieger, Howard, Stephenson, Cotter, Haag, Baerreis, and others interested in the area. James B. Griffin and James A. Ford gave valuable assistance in the original formulation of Belcher types in 1941 and critical suggestions since that time.

Stone artifacts are much less frequent at the Belcher site than either shell objects or pottery. Very few stone artifacts were placed with the burials. Only 155 chipped stone artifacts were recovered; 103 are projectile points, 52 of which were found on house floors, 27 with burials, and 24 on the surface (Table 4). All of the projectiles are small, thin arrow types, except one from the fill of Burial 9 which is large enough to be considered a dart point (Fig. 125 l).

Bassett points (Fig. 125 a-d). This form appears to be the resident projectile type at the Belcher site at least for the last two occupation periods, the Belcher focus. There are 56 points of this type and 20 others which have broken stems but are probably Bassett forms. All recognizable Belcher focus burial offerings of projectile points are beautifully finished points of this type. They are thin, keen projectiles, 2 to 4 cm. in length and 1.2 to 1.5 cm. in basal width, made from flakes of tan chert or argillite. The sides are straight, slightly convex or recurved to flare outward at the barbs, which are long and keen; the tips are sharp, the edges finely flaked, and the bases deeply indented to produce narrow, triangular contracting stems.

This site is one of several key sites in the Caddoan Area which Krieger used in the formulation of his classification of Caddoan cultures. It offers, in Mound B, a clear separation of habitation periods by sterile fills and thereby affords a sharp temporal segmentation of culture levels. We believe that we have been reasonably accurate in relating other habitation levels and burials to the four levels of Mound B in working out the picture of the entire site. It now seems desirable to investigate these levels or periods for (a) evidences of continuity of traits and culture patterns, (b) evidences of abrupt interruption of culture or interjection of an entirely new cultural pattern and (c) evidences of distinct culture sequences, resulting either from internal growth and differentiation or from external influences, sudden or gradual. This site may thereby contribute information as to relative depth of Caddoan culture, to the relationship between Gibson and Fulton aspects, and to temporal associations between the various Caddoan cultures and other cultures of the Southeast.

Specific elements of the ceremonial complex which has been termed the “Southern Cult” appear in Belcher III at the Belcher site and in the Foster and Friday mound components of the Belcher focus. This constitutes the only significant persistence of this cult material in the later or Fulton aspect of the entire Caddoan area.

These Southern Cult elements in the three sites accompany an equally large number of ceremonial traits to form an apparently highly organized burial complex. It seems desirable to segregate the Cult elements, because of their significance outside of this area, then to relate them to other ceremonial traits in the focus in order to complete the local picture. The elements of the Cult complex which are found in the Belcher focus are listed according to the classification of Waring and Holder (1945). The numbers in parentheses correspond to those which Waring and Holder have assigned to the various elements of the Cult. It should be noted, however, that there is not universal acceptance of the items in the list prepared by Waring and Holder. Probably some items should be added; almost certainly a distinction should be made between specific Cult items and those traits of decoration or ornamentation which happen to be widespread in Mississippi and related cultures but are nonreligious or nonceremonial. Further study of local variations in ceremonialism and shared Cult traits is needed.

The excavated portion of the Belcher Mound site seems to represent the ceremonial and possibly civil center of a small agricultural community located in the overflow valley of Red River. The village was scattered along the east bank of a stream which presently is inactive but apparently during the life of the site was the main course of Red River, building up its natural levee during the time when the mounds were being constructed. The prehistoric and protohistoric Caddoan inhabitants of the village were primarily farmers, but they were also hunters, fishermen, and gatherers of native foods. Although a riverine people, they left considerable evidence of extensive overland trade and contact.

The mounds, originally two which were later incorporated into one linear mound, were habitation substructures built up by accretion which resulted from the covering over of successively burned structures. This burning seemingly related to ceremonies attending deaths of important individuals, and burials in pits sunk through the covered-over burned structures accounted for most of the burials found in the site. A few individual burials were placed beneath house floors or at random in the mounds during construction.