The first publication dealing with materials which have since come to be included in the Tchefuncte horizon was an article by J. R. Czajkowski, published in the Louisiana Conservation Review for July 1, 1934. This was a brief summary of the first six months' exploration by a Civil Works Administration project of extensive shell deposits near Little Woods, Louisiana, along the shores of Lake Pontchartrain, east of New Orleans. All of these middens had been used as sources of shell for paving roads; consequently the center of each had been gutted previous to Czajkowski's work. Through arrangements made by Dr. Fred B. Kniffen, the materials and notes were to be turned over to Louisiana State University. Only a very small part of the collection, however, was received by the University; the remainder could not be located. The few notes obtainable were from the brief field diary of one of the foremen.

What is available of Czajkowski's collection has been used in this paper. It represents only a part of the enormous amount found, and because of the lack of notes and identifying marks it necessarily must be treated as a collection from all the middens which Czajkowski dug (see map, Fig. 2, Sites Or-1, Or-2, Or-3, Or-4 and Or-5). This is a particularly unfortunate situation, for these areas were so thoroughly disturbed that later investigations have succeeded in collecting very little information.

It appears that in the original excavations of the Little Woods site Czajkowski found skeletons in the middens at different points. The following is taken from his article in the Louisiana Conservation Review:

Most of the burials on the shell ridge were solitary and no funeral furniture or artifacts were found with them, with two exceptions: (1) On March 20th, two skeletons were found in a semi-flexed, east-west position. The following day, five more skeletons were uncovered, making a total of seven, all lying in the same position, parallel to each other about four feet apart, at a depth of about five feet …(2) On June 1st at a depth of six feet, a single skull impaled on a tree stump was found. Underneath the skull was a single fragment of decorated pottery.

Burial Ground: Seven skeletons found on March 20th, 950 feet northwest from the location. Adults, in fairly good condition; all skulls crushed except one…. Six in semi-flexed position on their backs … (the illustration to which Czajkowski here refers shows a fully flexed skeleton). All lying parallel in an east-west position with their heads west. The seventh at full length on its face. They were lying at a depth of about four feet from the surface and overlain with a shell bed from two to twelve inches in thickness. The distance between the skeletons was from three to four feet (Czajkowski, 1934, pp. 14-15).

In this section of the report we have described the artifacts representative of the Tchefuncte culture as a whole. These artifacts were made of pottery, stone, bone, and shell—pottery vessels, however, are described in the next chapter.

Clay Tubular Pipes.—Tubular pipes made of fired clay are a diagnostic trait of the Tchefuncte period (Fig. 7). Complete specimens averaged about 15 cm. in length and the stem diameter increases gradually from about 1.5 cm. to 2 cm. which is the average diameter of the tobacco chamber. Frequently these pipes have flattened mouth pieces (Fig. 7 a, d, f, p, q, s, t). A few specimens exhibit slight shoulders between the stem and the tobacco chamber (Fig. 7 l, m) and a few others have slightly flaring rims (Fig. 7 c, k).

In most of the specimens there is a slight interior constriction between the tobacco chamber and the stem. The bore tapers gradually from the chamber to the mouthpiece where it has an average diameter of 6 mm. The tobacco chambers have an average interior diameter of 16 mm. About one-half of any pipe is tobacco chamber and the other half is stem.

The pottery from the sites of the Tchefuncte period has been classified into a number of different types. As used by us, the type is a kind of average of an arbitrarily limited range of overlapping traits which have been abstracted from whole vessels and sherds that appear to be similar. Typology is checked against stratigraphy and stratigraphy against typology, to determine the utility of the type as an indicator of cultural change in time and space. Following are descriptions of the pottery types found in sites of the Tchefuncte period.

Paste.—Method of Manufacture: Coiled. Sherds break readily along coil junctions. Flattened coils average 3-4 cm. wide.

Tempering: Angular particles of clay. Small amount of fine sand. There are occasionally small particles of carbonized vegetal material and rarely bits of red ochre.

Texture: Clay of paste is very fine. Clay was very poorly wedged, and this feature, added to the large angular tempering particles, gives a laminated and contorted appearance to cross sections of sherds. Surfaces have been floated and are soft and chalky.

The only excavations of Tchefuncte sites which produced materials suitable for laboratory analyses of the temporal behavior of the pottery types characteristic of the period were those undertaken at the Tchefuncte middens and Big Oak Island. In Tchefuncte midden A, five hundred five-foot squares were dug to the bottom of the midden by six-inch levels. For purposes of analysis, adjoining squares of uniform total depth were grouped together in order to provide large units suitable for stratigraphic studies. These units were selected in the parts of the midden where the strata were most nearly horizontal. Areas on the periphery, or in which burials occurred, were not used for stratigraphic analysis. The various units, their size, and depth are as follows:

By means of these analysis units, not only are the trends of popularity of the pottery types of the Tchefuncte period made apparent, but also the relative temporal position of the Tchefuncte complex is conclusively shown.

A comparison of Tchefuncte traits suggests that there are three groupings or foci which at present comprise the Tchefuncte culture. The Copell site on Pecan Island, although a single site, represents the first group. The absence of pottery, presence of artifacts with burials, interment in a cemetery, and other traits tend to isolate this component from other Tchefuncte groupings.

The three shell midden sites, Tchefuncte, Little Woods, and Big Oak Island form another grouping. The construction and use of shell middens, presence of certain types of bone and shell artifacts, and styles of chipped flint projectile points differentiate these midden sites from the others.

The third grouping is represented by the two mound sites, Lafayette and Lake Louis. The traits of circular mounds, differences in the styles of projectile points, and absence of other Tchefuncte traits delimit this focus.

A number of traits held in common, or traits which overlap from one component or site to another, link all of these groupings together to form the Tchefuncte culture. For instance, burial traits and styles of bone implements link the Copell site with the midden sites of the Tchefuncte period. The midden sites and the mound sites are bound together by common types of pottery and pipes. For the reader who disagrees with our interpretations, or who has a better method, we have listed all of our traits in the summaries of each component (Chapter II).