SMITHSONIAN INSTITUTION—BUREAU OF ETHNOLOGY

ANCIENT POTTERY

OF THE

MISSISSIPPI VALLEY.

BY

WILLIAM H. HOLMES.

CONTENTS.

ILLUSTRATIONS.

ANCIENT POTTERY OF THE MISSISSIPPI VALLEY.

By William H. Holmes

.

INTRODUCTORY.

This paper is the third of a series of preliminary studies of aboriginal ceramic art which are intended to be absorbed into a final work of a comprehensive character.

The groups of relics selected for these studies are in all cases of limited extent, and are such as can lay claim to a considerable degree of completeness. It is true that no series of archæologic objects can ever be considered complete, but in exceptional cases the sources of supply may be so thoroughly explored that the development of new features of importance cannot reasonably be expected. If any series of American ceramic products has reached such a condition, it is that of the middle portions of the Mississippi Valley; yet, even in this case, I consider it unwise to attempt a monographic study, and prefer to single out a particular collection, making it the subject of a thorough investigation.

When the idea of preparing such a paper was first conceived, the collection presenting the greatest advantages was that of the Academy of Natural Sciences at Davenport, Iowa, which was, therefore, chosen. Other museums, especially those at Cambridge, Saint Louis, and Washington, were rich in material from this region, but none of these collections were so homogeneous and satisfactory.

The National Museum has recently received important accessions from the Mississippi Valley, through the agency of the Bureau of Ethnology, and ere the publication of this paper will probably excel all others in the number and variety of its mound relics. Some of its material has already been published by Dr. Charles Rau, Prof. C. C. Jones, Dr. Joseph Jones, and myself, and several additional examples are given in this paper.

Professor F. W. Putnam has described and illustrated many pieces belonging to the Peabody Museum, and Professor W. B. Potter and Dr. Edward Evers have issued an important work on the Saint Louis collections, in Contributions to the Archæology of Missouri.

This study is intended to pave the way to a thorough classification of the multitude of relics, and to the discovery of a method of procedure suited to a broad and exhaustive treatment of the ceramic art.

I do not expect to discuss ethnical questions, although ceramic studies will eventually be of assistance in determining the distribution and migrations of peoples, and in fixing the chronology of very remote events in the history of pottery-making races.

Some of the results of my studies of the evolutionary phase of the subject are embodied in an accompanying paper upon the "Origin and Development of Form and Ornament," and a second paper will soon follow. Before the final work is issued I hope to make close studies of all the principal collections, public and private. In such a work the importance of great numbers of examples cannot be overestimated. Facts can be learned from a few specimens, but relationships and principles can only be derived from the study of multitudes.

I shall probably have occasion to modify many of the views advanced in these preliminary papers, but it is only by pushing out such advance guards that the final goal can be reached.

Since the original issue of this paper in the Proceedings of the Davenport Academy of Sciences, a careful revision of the text has been made and much additional matter and a number of illustrations have been added.

I wish in this place to express my obligations to the officers and members of the Davenport Academy of Sciences, and especially to Mrs. M. L. D. Putnam and Prof. W. H. Pratt, whose generous aid has been of the greatest service to me.

CERAMIC GROUPS.

In studying the collections from the Mississippi Valley, I find it convenient to classify the ceramic products in three great groups, which belong to as many pretty well-defined districts; these I have named, for convenience of treatment, the Upper Mississippi, the Middle Mississippi, and the Lower Mississippi or Gulf provinces. Other pottery occurs within the limits of these areas, but the examples found in the museums are so few that very little of importance can be learned from them.

The three groups enumerated are not equally represented. The great body of our collections is from the middle province. The ware of the Lower Mississippi or Gulf district, of which we have but a small number of pieces, has many features in common with the pottery of the middle district, and at the same time is identical in most respects with that of the Gulf coast to the east. No well-defined line can be drawn between them; but the ware of the north is wholly distinct and need never be confounded with the other groups.

MIDDLE MISSISSIPPI PROVINCE.

Distribution.—It must not be inferred that there is perfect uniformity in the pottery of this, or any other, extended region; local peculiarities are always to be found. The products of contiguous districts, such, for example, as those of Mississippi County, Arkansas, and New Madrid County, Missouri, have much in common, and will at once be recognized as belonging to the same family, yet the differences are so marked that the unskilled observer could point them out with ease.

As indicated by decided family resemblances, the wares of this group extend over the greater part of the States of Missouri, Arkansas, and Tennessee, cover large portions of Mississippi, Kentucky, and Illinois, and reach somewhat into Iowa, Indiana, Alabama, Louisiana, and Texas. The types are better marked and the products more abundant about the center of this area, which may be defined roughly as including contiguous parts of Missouri, Arkansas, and Tennessee, with a pretty decided focal center, at least in the abundance of relics, at Pecan Point, Arkansas.

The borders of the district are necessarily not clearly defined. The characters of the art products blend more or less with those of neighboring sections. This is a usual phenomenon, and is probably due to a variety of causes. The mere contact of peoples leads to the exchange of ideas, and, consequently, to similarities in the products of industry. A change of habitat, with its consequent change of environment, is capable of modifying art to a great extent. Groups of relics and remains attributed by archæologists to distinct stocks of people, may, in cases, be the work of one and the same people executed under the influence of different environments and at widely separated periods of time.

Mixed conditions in the remains of a locality are often due to the presence of different peoples, synchronously or otherwise. This occurs in many places on the outskirts of this district, a good illustration being found in East Tennessee, where three or four distinct groups of ware are intermingled. As would naturally be expected, the distribution is governed somewhat by the great water-ways, and pottery of this province is found far up the Ohio, Tennessee, and Arkansas Rivers.

How Found.—All peoples have resorted, at some period of their history, to the practice of burying articles of use or value with the dead. It is to this custom that we owe the preservation of so many entire pieces of these fragile utensils. They are exhumed from burial mounds in great numbers, and to an equal extent, perhaps, from simple, unmarked graves which are constantly being brought to light by the plowshare. Fragmentary ware is found also in refuse heaps, on house and village sites, and scattered broadcast over the face of the land.

This pottery, at its best, was probably not greatly superior in hardness to our own soft earthenware, and the disintegrating agencies of the soil have often reduced it to a very fragile state. Some writer has expressed the belief that a considerable portion of the ware of this province was sun-baked merely. This view is hardly a safe one, however, as clay, unmixed with lime or other like indurating ingredient, no matter how long exposed to the rays of the sun, would, from ages of contact with the moist earth, certainly return to its original condition. I have seen but few pieces that, even after the bleaching of centuries, did not show traces of the dark mottlings that result from imperfect firing. There probably was a period of unbaked clay preceding the terra-cotta epoch, but we cannot expect to find definite traces of its existence except, perhaps, in cases where large masses, such as mounds or fortifications, were employed.

The relations of the various articles of pottery to the bodies with which they were associated seem to be quite varied. The position of each vessel was determined by its contents, by its symbolic use, or by the pleasure of the depositor. Uniformity cannot be expected in this more than in other features of burial. In other sections of the country the pieces of pottery are said to have been broken before final inhumation took place, but such was certainly not the practice in this province.

Age.—There can be no reasonable doubt that the manufacture of this ware began many centuries before the advent of the white race, but it is equally certain that the art was extensively practiced until quite recent times. The early explorers of Louisiana saw it in use, and the processes of manufacture are described by Dumont and others.

Possibly Du Pratz had in mind some of the identical vessels now upon our museum shelves when he said that "the women make pots of an extraordinary size, jars with a medium-sized opening, bowls, two-pint bottles with long necks, pots or jugs for containing bear's oil, which hold as much as forty pints, and finally plates and dishes in the French fashion."¹

Footnote 1: Du Pratz: Histoire de la Louisiane, Vol. II, p. 179.

Vessels were certainly made in great numbers by the Natchez and other tribes within our period, and it is reasonable to suppose that they belonged to the great group under discussion. If not, it will be necessary to seek the cause of their total disappearance, since, as I have already said, the pottery of this district, as shown by the relics, is practically a unit.

The introduction of metal utensils was a death-blow to the native industry, although some of the southern tribes, the Cherokees, for example, seem to have practiced the art continuously, in a very limited way, down to the present time. There is but little evidence of the influence of the art of the whites upon the ceramic products of this province, although the forms are sometimes thought to be suggestive of European models. It is certain, however, that the art had reached its highest stage without the aid of civilized hands, and in the study of its many interesting features we can feel assured that we are dealing with purely aboriginal ideas.

The pottery of this province is remarkably homogeneous in character, and we are warranted in assigning it to a single period of culture, and, in concluding, that the peoples who developed and practiced the art belonged to a group of closely-allied tribes. We can also state without fear of precipitating a controversy that the people who made this pottery were "mound-builders." At the same time, they were not necessarily of the same people as the builders of the mounds of Wisconsin, Ohio, or Georgia or contemporaneous with them.

Use.—It is difficult to determine the functions of the various forms of vessels. We are safe in stating that in very primitive times nearly all were intended for use in the domestic arts, and that as time went on uses were differentiated—form, as a consequence, undergoing many changes. Early writers on the Southern States mention a number of ordinary uses, such as cooking, the carrying and boiling of water, the manufacture of sugar and salt, and the preservation of honey, oil, and paint.

Only a small percentage of the vessels, and these generally of the pot-shaped variety, show indications of use over fire. It is well known that with most peoples particular forms were devoted to especial ceremonial uses. The construction of vases exclusively for mortuary purposes was probably not generally practiced, although a few examples, notably those illustrated in Figs. 372 and 420, point decidedly in this direction.

The simple conditions of life with these people are indicated by the absence of certain forms. Lamps, whistles, toys, bricks, tiles, and other articles in common use with many barbaric nations, are not found in this province. Pipes, so neatly shaped by other mound-building peoples, are here of a very rude character, a point indicating decided distinctions between the tribes of this province and those of neighboring sections.

Construction.—The methods of manufacture have evidently been of a primitive character. The wheel or lathe has not been used. At the advent of the whites, the natives were observed to build their vessels by a process known as "coiling," and by modeling over gourds, and over blocks of wood and masses of indurated clay shaped for the purpose.

It is probable that in many cases the support was not a mold in the ordinary sense, but was simply a rounded object of small size held in one hand while the base of the vessel was formed over it by the other. Rounded pebbles, or the mushroom-shaped objects of clay sometimes found in the mounds, would have served the purpose perfectly. Trowels, paddles, stamps, polishing-stones, and other implements were used in finishing.

Baskets were also used as molds, and pliable fabrics, such as nets and coarse cloths, were employed in some sections. The methods of baking have apparently not been described in much detail by early writers, but the ware itself bears the marks of those simple processes known to our modern tribes. It is highly probable that the work was done by the women, and that each community had its skilled potters, who built and baked the ware in the open air, going through those simple mummeries that accompany the work among most primitive peoples.

Material.—The material employed was usually a moderately fine-grained clay, tempered, in a great majority of cases, with pulverized shells. The shells used were doubtless obtained from the neighboring rivers. In many of the vessels the particles are large, measuring as much as one-fourth or even one-half of an inch in width, but in the more elegant vases the shell has been reduced to a fine powder. Powdered potsherds were also used. The clay was, apparently, often impure or loamy. It was, probably, at times, obtained from recent alluvial deposits of the bayous—the sediment of overflows—as was the potter's clay of the Nile. There is no reason for believing that the finer processes of powdering and levigation were known. A slip or wash of very finely comminuted clay was sometimes applied to the surface of the vessel. The walls of the vessels are often thick and uneven, and are always quite porous, a feature of no little importance in the storage of drinking-water, but one resulting from accident rather than from design.

Color.—The paste of this ware presents two marked varieties of color, a dark and a light hue. In a majority of cases it is dark, ranging from a rich black to all shades of brown and gray. The lighter tints are usually warm ochrey grays, rarely approaching reddish or terra-cotta hues. It is highly probable that the differences of color were, to some extent, intentionally produced, and that the material or methods of firing were regulated in a way to produce one tint or another at pleasure. This theory is confirmed by the fact that certain forms of vases are pretty generally dark, while certain other forms are as uniformly light—the latter in nearly all cases being used for the application of color, or of designs in color.

Form.—This ware exhibits a great variety of forms, many of which are extremely pleasing. In this respect it is far superior to the other prehistoric groups of the eastern United States. The shapes are as varied and elegant as those of the ancient Pueblo pottery, but are inferior to those of Mexico, Central America, and Peru. They take a higher rank than the prehistoric wares of central and northern Europe, but as a matter of course lack the symmetry and refinement of outline that characterize the wheel-made wares of Mediterranean countries.

As I classify by form farther on, and discuss the origin of form as each form-group is presented, I shall not make further reference to this topic here.

Finish.—The finish, as compared with the work of civilized nations, is rude. The surface is often simply hand or trowel smoothed. Generally, however, it was more or less carefully polished by rubbing with an implement of stone, shell, bone, or other suitable substance, the markings of these tools being distinctly visible. Nothing resembling a glaze has been found on pieces known to be ancient. The surface was sometimes washed or coated with a slip or film of fine clay which facilitated the polishing, and in very many cases a coat of thick red ocher was applied.

Ornament.—The ancient potter of the middle province has taken especial delight in the embellishment of his wares, and the devices used are varied and interesting. They include, first, fanciful modifications of form; second, relief ornament; third, intaglio figures; and, fourth, designs in color.

Modification of shape.—It can hardly be claimed that the ancient peoples of this region had a very refined appreciation of elegance of outline, yet the simple, essential forms of cups and pots were by no means satisfactory to them. There are many modifications of shape that indicate a taste for higher types of beauty, and a constant attempt to realize them. The æsthetic sentiment was considerably developed.

There is also a decided tendency toward the grotesque. To such an extreme have the dictates of fancy been followed, in this respect, that utility, the true office of the utensil, has often taken a secondary place, although it is never lost sight of entirely. Bowls have been fashioned into the shapes of birds, fishes, and reptiles, and vases and bottles into a multitude of animal and vegetable forms without apparent regard to convenience. All of these modifications of essential forms were doubtless looked upon as, in a sense, ornamental. So far as I can determine they were in no case intended to be humorous.

Relief ornament.—Decorative ideas of a purely conventional character are often worked out in both low and salient relief. This is generally accomplished by the addition of nodes and fillets of clay to the plain surfaces of the vessel. Fillets are applied in various ways over the body, forming horizontal, oblique, and vertical bands or ribs. When placed about the rim or base, these fillets are often indented with the finger or an implement in a way to imitate, rudely, a heavy twisted cord—a feature evidently borrowed from basketry. Nodes are likewise attached in various ways to the neck and body of the vessel. In some cases the entire surface of the larger vessels is varied by pinching up small bits of the clay between the nails of the fingers and thumb. An implement is sometimes used to produce a similar result.

Intaglio designs.—The æsthetic tendencies of these potters are well shown by their essays in engraving. They worked with points upon both the plastic and the sun-dried clay, as well as at times upon the fire-baked surface. Figures thus produced exhibit a wide range of artistic achievement. They illustrate all stages of progress from the most archaic type of ornament—the use of dots and straight lines—to the most elegant combinations of curves; and, finally, to the delineation of life forms and fanciful conceptions.

Generally, when a blunt implement is employed, the line is produced by a movement that I shall call trailing, in contradistinction to incision, in which a sharp point is used, and excision or excavation, which is more easily accomplished with the end of a hollow reed or bone. Impressed or stamped ornament is of rare occurrence, and anything like repoussée work is practically unknown. The practice of impressing cords and fabrics was common among many of the northern tribes, and nets have been used in the manufacture and ornamentation of vases at many points within this province. The use of stamps, especially prepared, was in vogue in most of the Gulf States, and to a limited extent in northern localities.

Designs in color.—The colors used in painting are white, red, brown, and black, and have generally consisted of thick, opaque, clayey paste, white or colored with ochers. Occasionally the colors used seem to have been mere stains. All were probably laid on with coarse brushes of hair, feathers, or vegetable fiber. The figures are in most cases simple, and are applied in broad, bold lines, indicative of a strong talent for decoration. The forms are, to a great extent, curvilinear, and embrace meanders, scrolls, circles, and combinations and groupings of curved lines in great variety. Of rectilinear forms, lozenges, guilloches, zigzags, and checkers are best known.

The decided prevalence of curved forms is worthy of remark. With all their fertility of invention, the inhabitants of this valley seem never to have achieved the rectangular linked meander, or anything more nearly approaching it than the current scroll or the angular guilloche, while other peoples, such as the Pueblos of the Southwest and the ancient nations of Mexico and Peru found in it a chief resource. The reasons for this, as well as for other peculiarities of the decorative art of the mound-builders as embodied in pottery, must be sought for in the antecedent and coëxistent arts of these tribes. These peoples were certainly not highly accomplished in the textile arts, nor had they felt the influence of advanced architecture such as that of Mexico. The influence of such arts inevitably gives rise to angular geometric figures. Taken as a whole, the remains of the mound-builders would seem to point to a hyperborean origin for both the people and their arts.

The origin of decorative ideas, the processes by which they are acquired by the various arts, and their subsequent mutations of form and significance are matters of the greatest interest, and a separate paper will be devoted to their consideration.

Classification of Forms.—Form cannot be made a satisfactory basis of classification, yet within a given group of products, defined by general characters, a classification by shape will be found to facilitate description. In making such a classification we must distinguish essential from non-essential features, that is to say, for example, that bowls must be placed with bowls, bottles with bottles, etc., disregarding the various fanciful modifications given to rims, necks, and bodies for the sake of embellishment. To recognize these adventitious features, which are almost infinite in variety, would be to greatly embarrass form classification.

There is also another difficulty in the employment of form in classification—the nomenclature is very imperfect. We cannot use Greek names, as our forms correspond in a very few instances only with the highly developed forms known to classic art. Our own plain terms, although defective, are better and far more appropriate. All necessary correlations of form can readily be made when the comparative study of the pottery of the world is undertaken.

If we take a full set of these primitive vessels and arrange them in the order of increasing complexity we have an unbroken series ranging from the simplest cup to the high-necked bottle with perforated foot or with tripod. A partial series is shown in the upper line, Fig 361. A multitude of variations from these outlines are found, a few of which are suggested in the lower line.

Fig. 361.—Scale of forms.

Compound, eccentric, and life forms are given elsewhere.

In deciding upon the order of arrangement for the various form groups, I shall be governed by what appears to be the natural order of evolution—a progress from simple to complex. First then we have basin-like vessels, such as dishes, cups, and bowls. Second, vases with wide mouths and somewhat globular bodies, the larger of which would be very generally recognized as pots. Third, vases with full bodies and narrow mouths, such as are often termed jars, but which are as properly called bottles. Fourth, vessels with high, narrow necks, universally denominated bottles. Vessels that cannot be grouped with either of these classes will have to be described in sub-groups, arranged in the order of their complexity or importance.

Origin of Form.—The derivation and subsequent mutations of form will be treated somewhat in detail as the various forms come up, and a subsequent paper will dwell upon the topic at considerable length.

BOWLS.

Basin or bowl-shaped vessels exhibit great diversity of shape and ornament. In size they range from less than one inch in diameter and depth to more than twenty inches in diameter and a foot in depth. In color and finish they are uniform with vessels of the other classes. Their uses were doubtless chiefly domestic.

Fig. 362.—Forms of bowls.

Form.—The forms are greatly varied, as will be seen in Fig. 362. Many are simply segments of spheres and vary from a shallow saucer to a hollow perforated globe. Others have elongated, compressed, or conical bodies, with round or flattened bases. Rectangular and irregular forms are sometimes found. Stands and legs are but rarely attached, and handles, excepting those of a grotesque character, are exceptional.

It will probably be safe to assume that some form of shallow vessel—a dish, cup, or bowl, was the first artificial form produced. Such a vessel would be most easily fashioned in clay and may have been suggested by accident, or by natural or artificial vessels.

Whatever the origin or whichever the method of construction, the difficulties encountered would at first prevent the manufacture of other than the simplest forms.

Ornament.—The ornamentation of bowls was accomplished in a variety of ways. These have been already described in a general way, under the head of ornament. Rim modifications constitute an important feature. The margin or lip may be square, oblique, round, or grooved, as indicated in Fig. 363 a, b, c, and d. The scallop may be employed as in e and f, and relief ornament may be added, such as fillets and nodes, and various horizontal projections, as shown in the second line, Fig. 363, to say nothing of incised lines and indentations, which are the heritage of wicker-work.

Fig. 363.—Modification of rims.

Not satisfied with these simple ideas of decoration, the fancy of the potter led him to add embellishments of most varied and often of extraordinary character. The nodes and ridges have been enlarged and prolonged, and fashioned into a thousand natural and fanciful forms. Shells, fish, birds, beasts, human and impossible creatures have been utilized in a multitude of ways. Many illustrations of these are given on subsequent pages.

The body of the bowl is somewhat less profusely ornamented than the rim. The interior, as well as the exterior, has received painted, relieved, and intaglio designs. In the painted ones the favorite idea for the interior is a series of volutes, in broad lines, radiating from the center of the basin. Groups of festooned lines, either painted or engraved, and arranged to give the effect of imbricate scales, form also a favorite motive. The exterior surface of the incurved rims of globular vessels offers a tempting surface to the artist and is often tastefully decorated in all the styles.

Illustrations.—Ordinary forms.—I have not thought it necessary to present many cuts of simple undecorated vessels, as their shapes are repeated numberless times in elaborated forms. The crude examples teach nothing as to stage of culture. They are of the same time and people as the finer specimens.

Fig. 364.—Bowl: Arkansas.—⅓.

The small bowl given in Fig. 364 is unusually well made, and is peculiar in having its interior surface decorated with a rather chaste incised design consisting of festooned lines. This was a favorite idea with the ancient potters and may be seen on both exterior and interior surfaces of a variety of vessels. The rim is beveled on the inner edge and has a beaded or indented fillet encircling the outer margin. The bottom is somewhat flattened. This specimen is from Arkansas.

Fig. 365.—Bowl: Arkansas.—⅓.

In Fig. 365 we have a good example of the dark, nicely-finished ware of Arkansas. The widely expanding rim is neatly scalloped on the margin and is finished on the inside with a pattern of incised lines. These lines appear to have been engraved in the hardened clay. The form is rendered graceful by a shallow encircling depression or groove at the base of the rim. The bottom is somewhat flattened.

Occasionally we find very deep bowls with sloping sides and flat bottoms resembling our common flower pots. One example from Arkansas is seven inches in diameter at the top and four at the base, and five inches deep. A heavy band of clay has been added to the outer margin of the rim, leaving a channel above and beneath. A number of perforations occur in this rim, as if made for the passage of thongs or filaments. A similar specimen of larger dimensions may be seen in the National Museum.

We have a number of bowls with incurved rims. This form is more characteristic of the south and is common along the Gulf coast.

A very small example is shown in Fig. 366. The lower part of the body is nearly hemispherical while the rim contracts slightly, giving a rather graceful outline. The exterior is embellished with a simple figure consisting of four linked scrolls which have been traced with a blunt point in the moist clay.

Fig. 366.—Cup: Arkansas.—⅓.

A much larger vessel resembling the above in shape is given in Fig. 367. It is of the dark brownish shell-tempered ware, characteristic of Arkansas. The lip is much incurved and the base considerably flattened, so that the form is that of a greatly compressed oblate spheroid. The outer surface has been moderately well polished, and is ornamented in a very effective manner by a series of figures, outlined by incised lines, alternate spaces being filled in with minute punctures.

Fig. 367.—Bowl: Arkansas. (?)—⅓.

A favorite form is a bowl with full deep body and incurved lip. A vessel of this class is illustrated in Fig. 368. The rim is but slightly incurved, while the body is considerably constricted below the greatest circumference. It is a unique and handsome specimen. The color of the slip is a pale, reddish-gray, a little darker than an ordinary flesh tint. The paste is seen to be yellowish where the surface has been injured. The ornament is a simple meander, consisting of three incised lines. It is said to have been found in Arkansas. Other bowls of like form and of elegant finish are found in the collection. They are generally dark in color, and have large apertures, low walls and flattened bases. The meander, mostly in its more simple forms, is the favorite decoration.

Fig. 368.—Bowl: Arkansas.—⅓.

There are many red vessels of the class under consideration, but the majority are less contracted at the aperture and thus are somewhat pot-shaped. They are rather rudely constructed and finished, and but for the color, would seem to be intended for ordinary cooking purposes. I observe in a number of cases that circular medallion-like ornaments have been set around the rim. These are from one-half to one inch in diameter, and are generally perforated or punctured in two or three places, apparently with the idea of representing a face. The effect is very much like that of the small perforated disks, riveted upon the exterior of copper or tin kettles for the purpose of attaching handles. Occasionally a tail-like appendage is added to the under side of these discoidal heads, suggesting the tadpole figures upon the sacred water vessels of the Pueblo Indians.

One large basin with slightly incurved rim has a series of triangular figures in red and brown upon both the inner and the outer surfaces. It is rudely finished and of large size, being eleven inches in diameter and seven and a half in height.

Eccentric forms.—Before proceeding with the discussion of life-forms as exhibited in bowls, I must present a few unique shapes.

Fig. 369. Fig. 370.

Cups: Arkansas (?).—⅓.

These consist of ladle-shaped vessels, and of bowls or basins with rectangular, oval, or unsymmetrical outlines. Ladles are of rare occurrence. In the Third Annual Report of the Bureau of Ethnology I have illustrated the best example that has come to my notice. The Davenport collection contains but one specimen—a rude shallow cup with a short thick handle. The form suggests the wooden and horn spoons of the modern tribes and may have originated in their archaic prototypes.

Fig. 369 illustrates a minute cup rudely made of coarse clay. The outline is oval and slightly pointed at one end, as if intended for pouring liquids.

In Fig. 370 we have another small vessel of rude finish with two pointed lips. A much larger vessel of similar shape may be seen in the Davenport collection. The projecting pointed lip is rarely found in aboriginal pottery, although I see no reason why such a feature may not readily have been suggested to the savage by the prolonged margins of his vessels of shell.

Rectangular vessels are of the rude shell-tempered ware, and, although rare, are widely distributed.

Fig. 371 illustrates a specimen from Pecan Point, Arkansas. The surface is rudely finished and without polish. The color is a dark gray, much flecked with large particles of white shell. Another example has a square rim but a rounded bottom, and is covered with a coat or slip of dark red clay.

Fig. 371.—Rectangular bowl: Pecan Point, Arkansas.—⅓.

A small vessel from the same region as the preceding has the rim pressed in on the four sides, leaving sharp, projecting corners.

One of the most notable vessels in the collection is illustrated in Fig. 372. It is a heavy casket consisting of two parts, body and lid, and is made as usual of clay and coarsely pulverized shell. It is brownish gray in color and bears some marks of the baking. It was obtained by Captain W. P. Hall from a low mound at Hale's Point, Tennessee, and is described by Mr. W. H. Pratt, in the following language: "It is of rude, irregular, quadrangular form, made in two parts. The lower, or case proper, is 12 inches long, 7 inches wide, and 5 inches deep, inside measure, the upper edge being slightly bent inward all around. The upper part or lid is of similar form and dimensions, being very slightly larger, so as to close down over the other part, about one and a half inches, and is somewhat more shallow. As the lid does not fit very perfectly, the joint around the edge had been plastered up with clay. When found, it contained the remains of a very small child reduced to dust, except that some of the bones of the skull, jaws, and limbs retained their form, crumbling rapidly, however, upon removal and exposure to the air. There were also found two or three dozen small shell beads. Excepting the remains described, the case was entirely empty. The case weighs six and a quarter, and the lid just six pounds." This is one of the very few vessels that would seem to have been constructed especially for mortuary purposes.

Fig. 372.—Burial casket: Hale's Point, Tennessee.—¼.

I wish to add to the list of eccentric forms a singular example from the collection of J. R. Thibault, of Little Rock, Arkansas. As shown in Fig. 373 it is an oblong, trough-like vessel with flat projecting wings at the ends. It is extremely well-finished, with thin walls, symmetrical form, and high polish. The color is quite dark and the material is as usual. The engraved design consists of incised lines, which form a number of rectangular compartments extending around the exterior surface of the body. The wings are perforated. The form of this vessel suggests the wooden trays of some modern tribes. A similar example, which is illustrated in the Third Annual Report of the Bureau of Ethnology, is of much inferior interest, being plain and rude.

Life forms.—A very large percentage of the bowls of this district are modified in such a way as to resemble, more or less closely, the form of some living creature—bird, beast, or reptile. Especial attention has been given to the heads. These are modeled in the round and attached to the rim or side, while other parts of the animal appear upon different portions of the vessel.

Fig. 373.—Trough-shaped vessel: Arkansas.—⅓.

[National Museum.]

It will be difficult to determine the origin of this curious practice. We shall not be able to say that it came from the elaboration of handles, simply to please fancy, for the reason that vessels of this class are rarely known to have had simple handles; nor from the modification of simple ornaments, as such were but little used. It is still less probable that animal forms were first modeled independently, and afterwards changed in such a way as to serve as vessels. There are no examples of animal forms in clay independent of vessels. It would not be consistent with primitive methods of procedure to copy nature direct, at least until some mystic significance had become attached to the form employed. It is possible, however, that the origin of this practice is not to be found within the plastic art itself, but in the shapes of antecedent and co-existent vessels of other materials in which life forms had been employed; or in the use of natural objects themselves as utensils, the original forms not having been lost sight of and having in time suggested the employment of other natural forms. Examples of the latter class may be cited.

Shells were primitive vessels. The hard cases of seeds and fruits were also much used. These were doubtless antecedent to vessels of clay. They were the natural models for the potter, the carver in wood or stone, and their employment as such served to lead up gradually to a more realistic and general use of natural shapes in works of art to which they were not essential features. The importance of the various animal forms was increased by their association with religious ideas. Nearly all the vessels of this class presented in the following illustrations come from the vicinity of Pecan Point, Arkansas.

Clay vessels imitating both marine and fresh-water shells are occasionally obtained from the mounds and graves of the Mississippi Valley. The conch shell appears to have been a favorite model, especially in its modified form, Fig. 374, a and b. The clam shell is also imitated in c and d. The more conventional forms of these vessels are exceedingly interesting, as they point out the tendencies and possibilities of modification. An instructive example illustrated in e has four groups of nodes, each, consisting of a large central node with four or five smaller ones, surrounding it, set about the rim, the conception being that of four shells joined in one vessel, with the noded apexes turned outward and the bases inward.

A still more highly conventionalized form is shown in f. The cup is unsymmetrical in outline, and has a few imperfect nodes near one corner, but its resemblance to a shell would hardly be recognized by one unacquainted with more realistic renderings of like subjects. In g we have an imitation of a shell cup placed within a plain cup.

Fig. 374.—Clay vessels imitating shells.

A very good illustration of this class of vessel is given in Fig. 375. It is evidently intended to imitate a trimmed conch shell. The apex and a few of the surrounding nodes are shown at the right, while the base or spine forms a projecting lip at the left. A coil of clay forms the apex. This is carried outward in a sinistral spiral to the noded shoulder. We have here a suggestion of the origin of a favorite decorative motive, the scroll, a clew, however, which the paucity of examples makes it difficult to follow up satisfactorily.

Fig. 375.—Bowl imitating a modified conch shell.—⅓.

Although we may not be able to arrive at any definite conclusion in regard to the origin and significance of the practice of modeling life forms in clay, we are certain of one thing, that it became an important feature in the potter's art, and that in due course of time the practice broke loose from the restraints of birth and tradition and asserted its freedom in the production of any form that superstition or fancy happened to select.

The artist probably did not follow nature with great accuracy in all the details of species and varieties, but some definite model must have been in view, in nearly all cases, and such characters as came to be regarded as essential to that creature were never lost sight of, consistency being a most notable characteristic of the art of a savage or barbaric people.

Fig. 376.—Frog-shaped bowl: Craigshead Point, Arkansas.—⅓.

Fig. 377.—Frog-shaped bowl: Pecan Point, Arkansas.—⅓.

Fig. 378.—Animal-shaped bowl: Arkansas.—⅓.

The sun-fish was a favorite model, but its form was generally employed in vessels with upright necks. A number of examples occur in the next section. Of reptilian forms the frog seems to have been the favorite.

Fig. 379.—Bird-shaped bowl: Arkansas.—⅓.

Fig. 380.—Bird-shaped bowl: Arkansas.—⅓.

Few examples occur, however, in the shallower vessels. In the bowl illustrated in Fig. 376, the various members of the body are boldly modeled, and appear about the most expanded portion of the vessel. The rim is ornamented with a series of notches, and two small loops connect the rim with the head and tail of the creature. The legs are characteristic, and the long toes extend beneath the body. The bottom of the vessel is flat. The make and finish are as usual, but the surface has been painted red. A similar vessel is shown in Fig. 377, the view being taken from the front. It is well polished and has a rounded bottom. The color is dark.

Fig. 381.—Bird-shaped bowl: Arkansas.—⅓.

Fig. 382.—Bowl with grotesque heads: Arkansas.—⅓.

Another remarkable example of this use of animal forms is seen in the vessel presented in Fig. 378. A deep globular bowl of dark, well-polished ware is made to represent the head of an animal. A long snout, with teeth and nostrils and accompanied by a pair of knobs for eyes, embellishes the right side—as seen in the cut—ears appear at the front and back, and a circular node standing, perhaps, for the severed neck, is placed at the left. The head has a decidedly porcine look, yet it may have been intended for a raccoon or an opossum.

Fig. 379 illustrates a large shallow bowl or pan of ordinary form and finish. The head of a bird resembling a turkey is attached to one side, with the bill turned inward. On the opposite side there is a small handle-like projection that represents the bird's tail.

A vessel of somewhat extraordinary form is shown in Fig. 380. The bowl is smaller and deeper than the last, and serves as the body of a bird, the head and tail of which are of unusual proportions. The neck is very long and thick and is gracefully curved, but the head is not modeled with sufficient care to make apparent the species intended.

The vessel shown in Fig. 381 is also finished in imitation of a bird. In this case the bird is placed upon its back, the neck and head being looped up to form a sort of handle on one side, while the legs answer a like purpose on the opposite side. The wings are represented by a number of lines rudely engraved upon the sides of the vessel. The resemblance of this bowl to the wooden basins made by Northwest Coast Indians is very striking.

The vessel shown in Fig. 382 is one of the most unique yet brought to light. It is a heavy, rather rudely finished bowl, to the rim of which two grotesque heads, apparently of nondescript character, have been attached. One resembles the oft-occurring plumed serpent of aboriginal American art in a number of its characters. The other has a double comb somewhat resembling that of a domestic fowl. No description can convey as clear a conception of these monstrosities as the accompanying illustration.

Fig. 384.—Grotesque heads.

A good degree of skill is shown in the modeling of varieties of birds. A fair idea of the accuracy of these potters in this direction will be conveyed by the series of heads shown in Fig. 383. Several species of ducks are apparently differentiated, one of which, resembling the summer duck closely, is given in a, while the head given in b, although possibly also intended for a duck, is much like a grouse or partridge. The pigeon or dove is seen in c, the vulture or eagle in d, and the owl in e.

Fig. 385.—Bowl with grotesque head: Pecan Point, Arkansas.—⅓.

Fig. 386.—Bowl with grotesque head: Pecan Point, Arkansas.—½.

It would be difficult to imagine more grotesque and outlandish heads than those attached to the bowls illustrated in Figs. 385 and 386. The vessels themselves are of the usual type, rudely modeled and finished and very heavy. The first is dark in color, the other red. The strange animal here represented is certainly not a close copy of anything in nature. It is characterized by upright ears, a high bulbous snout and a grinning mouth. The teeth in some cases resemble the fangs of a serpent. The eyes consist of rounded nodes; and often curved lines, incised or in relief, extend from them or the mouth down the sides of the neck. The tail at the opposite end of the vessel is turned upward and coiled. The type specimens of this form are from Pecan Point, Arkansas.

Fig. 387.—Bowl with grotesque handle: Scanlon's Landing, Arkansas.—⅓.

Fig. 388.—Animal-shaped bowl: Arkansas.—⅓.

The peculiar character of this class of heads is well shown in the series given in Fig. 384. My observations have led me to suspect that they may be the result of attempts to model in clay the mythical plumed serpent which is so graphically delineated in the engraving upon the little vase shown in Fig. 407. The fact that in one case legs have been added to the base of the body militates against this theory. Their resemblance to the gargoyle heads of mediæval architecture suggests the possibility of early European influence.

If possible, a still more novel conceit is embodied in the handle of the vessel shown in Fig. 387. It can be likened to nothing in nature more readily than to the antler of an elk. This vessel is of a dark brownish color, and is but slightly polished. A duplicate specimen of inferior size and finish has recently been added to the National Museum from a grave at Pecan Point.

Similar to the preceding in general appearance are a number of bowls or deep pans, embellished with the heads of animals. A very good example is given in Fig. 388. The head has a decided resemblance to that of a female deer or fawn. The tail appears upon the opposite side of the basin, and is pendant, as in nature. Legs have been added to the base of the bowl; these terminate beneath the body in cloven hoofs.

The small bowl, shown in Fig. 389, is nearly hemispherical in shape.

Fig. 389.—Animal-shaped bowl: Arkansas.—⅓.

A small head, representing some animal, has been attached to the rim. The exterior surface is covered with a number of groups of roughly-worked concentric ridges, which may be meant to imitate hair. These ridges have apparently been made by pinching up the clay between the nails of the fingers and thumb. Figures of similar form are generally incised. This vessel is probably from the vicinity of Pecan Point.

The creature represented by the head, shown in Fig. 390, would not be recognized from the cut, or perhaps not even with certainty from any single specimen, but with a number of examples in view, there need be no hesitation. The animal intended is a bat. In a number of features the likeness is striking. The high top head, the angular ears, and the small eyes crowded down upon the mouth are characteristic. The tail is flat, curved a little upward, and ridged along the middle in imitation of the attenuated caudal column. The general consistency of this work is demonstrated by the fact that this particular form of tail accompanies this form of head in all cases, and is not associated with any other. The face of the bat is always turned toward the vessel; in imitation of other varieties of animals, it is nearly always turned out.

Fig. 390.—Bowl with bat's head: Pecan Point, Arkansas.—⅓.

In one case, Fig. 391, we have, what appears to be, a human head attached to the side of the bowl. This head is furnished with a triangular crest, notched on the edges, and enlarged at the top. The case is a perplexing one, especially as a tail like that attached to the bird bowls occurs on the side opposite the head.

Fig. 391.—Bowl: Arkansas.—⅓.

POT-SHAPED VESSELS.

There is no hard line of demarcation between the class of vessels now to be considered and those already described. The distinction is made chiefly for convenience of treatment.

Material, etc.—As a rule, pot-shaped vessels are of coarser materials and of ruder finish than other forms, indicating, perhaps, their exclusive relegation to the culinary arts, where nice finish was not essential. In many cases they show use over fire.

In size, they have a wide range. The larger are often as much as fifteen inches in diameter, and twenty in height. There are a score or more of very large size in the Davenport museum.

Form.—The form characteristics are a full globular body—sometimes elongated, sometimes compressed vertically—a low neck, and a wide aperture. The bottom is very generally rounded. A few of the form modifications are shown in Fig. 392. The rim or neck is always short, and is upright or slightly recurved. Many vessels resembling the shapes here presented are placed with the succeeding group, as they appear to be functionally distinct from this. There are no examples with legs or stands.

Fig. 392.—Forms of pots.

Handles.—Looped handles are confined almost wholly to this class of vessels. They are generally ranged about the rim or neck. In a majority of cases there are four handles to a vessel. We rarely find less than that number, but often more. It is a usual thing to see fifteen or twenty handles set about the rim. Originally the handles may have been exclusively functional in character; they were so at least in antecedent forms. These potters have certainly, at times, employed them for purposes of embellishment. In some cases they are too fragile for use, in others they are flattened out against the neck of the vessel and united with it throughout their whole length. Again, they have degenerated into mere ridges, notched and otherwise modified to suit the fancy. In many instances their place is taken by incised lines or indentations which form effective and appropriate ornamental figures. A series of vessels showing gradations from perfect handles to their atrophied representatives is shown in Fig. 393.

Fig. 393.—Handles.

Origin of Handles.—Handles were doubtless originally attached to facilitate the suspension and handling of vessels and other articles. They probably had their typical development in basketry, and there are good reasons for supposing that certain forms of the handles upon pottery owe their existence to contact with the sister art. This idea is confirmed by their shapes, and by the fact that a large percentage of the pottery handles are useless as aids to suspension or transportation.

Ornament.—Rim margins are modified for decorative purposes, very much as they are in bowls. See Fig. 363.

The bodies of these vessels are often elaborately ornamented, mostly by incised figures, but often by punctures, nodes and ribs. The incised lines are arranged principally in groups of straight lines forming angular figures—a very archaic style—and in groups of festooned lines so placed as to resemble scales. The punctures are made with a sharp point, and form encircling lines and various carelessly executed patterns. A rude sort of ornamentation is produced by pinching up the soft clay of the surface between the nails of the fingers and thumb. Relief ornament consists chiefly of applied fillets of clay, arranged to form vertical ribs. Rows of nodes are sometimes seen, and in a few cases the whole body is covered with rude nodes.

Illustrations.—The specimens selected for illustration are intended to epitomize the forms and decorations of a very great number of vessels, and are not always the most showy examples to be found.

A vessel of rather exceptional shape is given in Fig. 394. It could as well be classed with bowls as with pots. The ware is of the rude kind generally used over the fire. The body is high and cylindrical, the rim flaring, and the bottom quite flat. The form is suggestive of our domestic crockery.

Fig. 394.—Pot: Arkansas (?).—⅓.

Another bowl-like pot is illustrated in Fig. 395. It is of the dark, rudely hand-polished variety. The body is globular, the neck is very short and is ornamented with a dentate band. Below this are two pairs of perforations, probably used for suspending the vessel. There are a number of vessels of this variety, mostly smaller than the example given.

The vessel shown in Fig. 396 is still more pot-like. The neck is higher than the preceding and is slightly constricted. It is of very rude construction and finish. The rim is furnished with two small horizontal projections, and the body is somewhat obscurely lobed. It represents a very numerous class, especially plentiful in Southeast Missouri.

Fig. 395.—Pot: Arkansas (?).—⅓. Fig. 396.—Pot: Waverly, Tennessee.—⅓.

The little pot presented in Fig. 397 has the body covered with rude nodes. The neck is surrounded by a heavy fillet, notched obliquely in imitation of a twisted cord. Four rude handles have also been attached.

Fig. 397.—Pot: Arkansas (?).—⅓. Fig. 398.—Pot: Arkansas.—⅓.

In Fig. 398 we have one of the rudest examples in the collection. The neck is furnished with four handles, which alternate with four vertical ribs. The body is misshapen and rough, and is ornamented with a series of nearly vertical ridges, a rather usual device, and one which is sometimes very neatly executed.

The body of the nicely finished pot shown in Fig. 399 is embellished with short, incised markings, arranged in vertical lines. The neck is furnished with a heavy indented band and four strong handles. The locality given is "Four-Mile Bayou, Alabama."

The specimen given in Fig. 400 illustrates the use of great numbers of handles. In this case there are sixteen. They are gracefully formed and add much to the appearance of the vessel, which is really a bowl with wide, flaring rim. In most of its characters it resembles the pots.

Fig. 399.—Pot: Alabama (?).—⅓. Fig. 400.—Pot: Arkansas (?).—⅓.

Another curious variation in the shape of handles is shown in the little cup given in Fig. 401. This can hardly be called a usual feature, although occurring in vessels of various localities. I have seen an example from the Missouri Valley in which a great number of perforated handles were set about the rim, and another in which there was a continuous, partially free, collar perforated at intervals. There is a specimen of this class in the Davenport Academy collection in which the flattened handles are so placed about the neck as to form a series of arches. These, I take it, are partially atrophied forms. The body is ornamented by a scale-like pattern of incised lines—a favorite method of decoration with the ancient potter.

Fig. 401.—Pot: Arkansas (?).—⅓. Fig. 402.—Pot: Arkansas (?).—⅓.

In Fig. 402 we have an illustration of total atrophy. The handles are represented by simple incised lines. There is no relief whatever. In many cases the form of the handles is shown in low relief, the outer surface being plain or ornamented with incised lines or punctures. The body of the vessel last mentioned is covered with rudely incised scroll designs.

Another good illustration of this class of vessels is shown in Fig. 403.

The cut is taken from my paper in the Third Annual Report of the Bureau of Ethnology. The handles are indicated by incised lines. The body was ornamented by pinching up the clay between the nails of the thumb and forefinger. Locality: Pecan Point, Arkansas.

Fig. 403.—Pot: Pecan Point, Arkansas.—⅓.

[National Museum]

Fig. 404.—Pot: Hale's Point, Tennessee.—⅓.

A good example of the larger pots is illustrated in Fig. 404. It is engraved a little less than one-fourth the dimensions of the original. The height is seventeen inches and the greatest diameter eighteen inches. It is very well made. The walls are even and only moderately thick. The dark, unpolished surface is profusely speckled with fragments of white shell. There are four wide, strong handles. The rim and neck are ornamented with encircling lines of finger-nail indentations.

Fig. 405.—Pot: Pecan Point, Arkansas.—⅓.

A masterpiece of this class of work is shown in Fig. 405. It was obtained at Pecan Point. It is not quite symmetrical in form but is carefully finished. The color is gray, with mottlings of dark spots, the result of firing. The height is eleven inches, and the aperture is ten inches in diameter. There are ten strong, well-proportioned handles, each having a knob resembling a rivet head, near the upper end. The margin of the rim has a circle of indentations. There are a few red vessels of this shape which have figures of reptiles attached to the neck.

WIDE-MOUTHED BOTTLES OR JARS.

Vessels of this class were probably not devoted to the ordinary uses of cooking and serving food. They are handsome in shape, tasteful in decoration, and generally of small dimensions. They are found, as are all other forms, buried with the dead, placed by the head or feet, or within reach of the hands. Their appearance is not suggestive of their original office, as there is no indication of wear, or of use over fire.

Form.—I include under this head a series of forms reaching from the wide-mouthed pot to the well-developed bottle. They really correspond closely to the high-necked bottles in all respects save in height of neck, and the separation is therefore for convenience of treatment only. The following illustration (Fig. 406) will give a good idea of the forms included.

Fig. 406.—Forms of jar-shaped bottles.

There are also many eccentric and many extremely interesting life forms included in this group. A number of vases, modeled after the human head, are, by their general outline, properly included.

Ornamentation.—The rims, bodies, and bases are embellished much after the fashion of the vessels already described, with the exception that handles or handle-like appendages or ornaments seldom appear. The painted designs are in one, two, or three colors, and the incised figures have been executed both in the soft and in the thoroughly dried clay.

The style of execution is often of a very high order, especially in some of the more southerly examples, a number of which are from the mounds of Mississippi and Louisiana. We note the fact that in a few of the designs there is a slight suggestion of Mexican forms.

In illustrating this group, I am compelled, for the want of space to omit many interesting examples. I present only such as seem to me especially instructive.

Fig. 407.—Bottle: Pecan Point, Arkansas.

Illustrations.—Ordinary forms.—The vessel shown in Fig. 407 may be taken as a type of a very large class. It is most readily described as a short-necked, wide-mouthed bottle. It is symmetrical in shape and very nicely finished. The lip is supplied with a narrow, horizontal rim. The body expands somewhat abruptly from the base of the upright neck to the squarish shoulder, and contracts below in an even curve, giving a hemispherical base. There are a multitude of variations from this outline, a few of which are suggested in Fig. 406. These vessels are nearly all of the dark, grayish-brown, fire-mottled ware. A few are yellowish, and such are often painted red or decorated with designs in red and white.

Fig. 408.—Bottle: Arkansas.—⅓.

Fig. 409.—Bottle: Arkansas.—⅓.

Two charming vases are shown in Figs. 408 and 409. The surface finish is in both cases very superior. The lines of the figures are carefully drawn, and seem to have been produced by the trailing, under even pressure, of a smooth rather blunt point. It is difficult to get so nicely finished and even a line by simple incision, or by excavating the clay. The design in Fig. 408 consists of eight groups of curved lines arranged in pairs, which are separated by plain vertical bands. It might be considered an interrupted or imperfectly connected form of the running scroll. This grouping of lines is frequently met with in the decorative designs of the Southern States. The design upon the other vase, Fig. 409, is still more characteristic of the South. It consists of an encircling row of round, shallow indentations, about which series of incised scrolls are linked, and of two additional rows of depressions, one above and the other below, through which parallel lines are drawn.

Many other interesting illustrations of the simpler forms could be given, but nearly all are very similar in their more important features to the examples that precede or follow.

As skilled as these peoples were in modeling life forms, and in engraving geometric devices, they seem rarely to have attempted the linear representation of life forms. We have, however, two very good examples.

Fig. 410.—Engraved bottle: Arkansas.

The first of these is shown in outline in Fig. 410. It is a large bottle embellished with four rude drawings of the human figure, executed with a sharp point in the soft clay. Height of vessel, eight inches.

The work is characteristic of a very early stage of art. The figures could be duplicated in the work of the ancient Pueblos, and in the pictographic art of many of our savage tribes. They are probably derived from symbolic art, and possibly relate to the guardians of the four points of the compass, or to some similar mythical characters.

Fig. 411.—Engraved bottle: Arkansas.—¾.

Fig. 412.
—Engraved design.—½.

The work upon the neat little bottle, presented in Fig. 411, is of the same class as the above but of a much higher grade, both in execution and conception. The engraved design is one of the most remarkable ever obtained from the mounds. It consists of two winged and crested rattlesnakes, which encircle the most expanded part of the vessel, and of two sunflower-like figures, alternating with them. These designs are very carefully engraved with a needle-like point, and are adjusted to the form of the vase in a way that suggests forethought and an appreciation of the decorative value of the figures. By dint of rubbings, photographs and sketches, I have obtained the complete drawing of the various figures which are given in Fig. 412 on a scale of one-half the original.

The serpent, especially the rattlesnake, has always taken a leading place in the mythology and the art of the more cultured American races, and crest-plumes, and wings have often been considered its proper attributes. The conventional method of representation is also characteristically aboriginal. The plumes, the figure connected with the eye, the bands upon the neck, the stepped figures of the body, and the semi-circular patches on the wings are all characters that appear again and again in the ancient art of the United States. The peculiar emblematic treatment of the heart is almost universal in temperate North America. And just here I may be permitted to suggest that the remarkable feature of the great earth-work serpent of Adams county, Ohio, which has been regarded as the "symbolic egg," and which in its latest phase has become the issue of a frog and the prey of the serpent, is possibly intended for the heart of the serpent, the so-called frog being the head. The rosette figures are not often duplicated in Indian art. There can be little doubt that the figures of this design are derived from mythology.

Eccentric forms.—A form of vessel of which civilized men make peculiar use is depicted in Fig. 413. There is a marked resemblance to a common tea-pot. A very few examples have been found, two of which are illustrated in the Third Annual Report of the Bureau of Ethnology. The specimen here given is well made and carefully finished. The neck is low and wide, and the body is a compressed sphere. The spout is placed upon one side and a low knob upon the other. The absence of a handle for grasping indicates that the vessel was probably not intended for boiling water. These characters are uniform in all the specimens that have come to my notice. Two small circular depressions occur on the sides of the vessel alternating with the spout and the knob and with these features form centers for four rosettes of involute incised lines. The origin of this form of vessel is suggested by a fine red piece from "Mississippi," now in the national collection. The knob is the head of a turtle or other full-bodied reptile, and the spout takes the place of the creature's tail. Many of the animal-shaped vases would resemble this form closely if an opening were made through the top of the body and through the tail.

Fig. 413.—Teapot-shaped vessel: Arkansas.—⅓.

Fig. 414.—Vessel of eccentric form: Arkansas.—⅓.

In connection with the teapot-like vessels it will be well to describe another novel form not wholly unlike them in appearance, an example being shown in Fig. 414. The shoulder is elongated on opposite sides into two curved, horn-like cones, which give to the body a somewhat crescent-shaped outline. It is of the ordinary plain, dark ware, and has had a low stand or base which is now broken away.

The specimen given in Fig. 415 has been considerably mutilated, but evidently belongs to the same class as the preceding. It probably also resembled the vessel which follows; it serves at least as a link between the two. The body is ornamented with carelessly drawn, deeply incised, involute designs.

Fig. 415.—Vessel of eccentric form: Pecan Point, Arkansas.—⅓.

Fig. 416.—Animal-shaped vase: Pecan Point, Arkansas.—⅓.

Life forms.—A further elaboration of the preceding forms is illustrated in Fig. 416. On one side the conical projection is greatly elongated and fashioned to resemble the head of some grotesque beast, with horns, expanded nostrils, and grinning mouth. The opposite point is elongated and looped, forming a tail, while the base of the body is furnished with four feet. On the sides of the vessel are engraved figures, consisting of clusters of involute lines, as in the specimen just given. It is of the ordinary dark pottery, and was obtained at Pecan Point.

Equally noteworthy as plastic representations are the two examples that follow. The vessel shown in Fig. 417 is modeled in imitation of a sunfish. The body is much flattened and is neatly polished. The head is well modeled, as are also the fins and tail. Many examples of this form are found, some of which are elaborately treated, the scales being minutely shown. The body of the fish is sometimes placed in the natural upright position, the neck of the vessel rising from the back, producing a lenticular shape.

Fig. 417.—Sunfish vase: Arkansas.—⅓.