The Project Gutenberg EBook of Stone Art, by Gerard Fowke

This eBook is for the use of anyone anywhere in the United States and most
other parts of the world at no cost and with almost no restrictions
whatsoever.  You may copy it, give it away or re-use it under the terms of
the Project Gutenberg License included with this eBook or online at
www.gutenberg.org.  If you are not located in the United States, you'll have
to check the laws of the country where you are located before using this ebook.



Title: Stone Art
       Thirteenth Annual Report of the Bureau of American Ethnology
              to the Secretary of the Smithsonian Institution, 1891-1892,
              Government Printing Office, Washington, 1896, pages 47-178.

Author: Gerard Fowke

Release Date: December 26, 2015 [EBook #50769]

Language: English

Character set encoding: UTF-8

*** START OF THIS PROJECT GUTENBERG EBOOK STONE ART ***




Produced by PM for Bureau of American Ethnology, The
Internet Archive (American Libraries) and the Online
Distributed Proofreading Team at http://www.pgdp.net (This
file was produced from images generously made available
by the Bibliothèque nationale de France (BnF/Gallica) at
http://gallica.bnf.fr)

47

The cover image was created by the transcriber and is placed in the public domain.

STONE ART
BY
GERARD FOWKE

48

49

CONTENTS

51

ILLUSTRATIONS

56

57

STONE ART
By Gerard Fowke

INTRODUCTION.

Basis for the Work.

The collection of the Bureau of Ethnology includes almost every type of stone implement or ornament, and as the investigations and explorations of the collaborators have extended over nearly all the eastern and central portions of the Mississippi valley, it furnishes a substantial basis for showing the geographic distribution of various forms of objects in use among the aboriginal inhabitants.

It has not been deemed advisable to utilize material contained in other collections. Should this be done there would be no reason for drawing upon one rather than another, and if it were once begun the examination would finally extend to every collection made from American localities, a study which, although perhaps desirable, would transcend the scope of the Bureau plans.

Much that has been published in regard to the distribution of relics in various portions of the country is of little value to a paper of this kind, since few of the objects are sufficiently illustrated or referred to any class in other than the most general terms; so that it is frequently impossible to determine the group in which a given article should be placed. Partly for this reason, partly because the primary purpose is description of a certain collection made in a definite way, little space is given to the descriptive work of predecessors in the field of archeology. The general results of previous work are, however, carefully weighed in the conclusions reached.

Classification of Objects and Materials.

The ordinary division into chipped and pecked or ground implements has been adopted: the former including all such as are more easily worked by flaking, and the latter including those made from stone suitable for working down by pecking into form with stone hammers or by similar means. The system of nomenclature in general use has been retained, as it is now familiar to students of North American 58 archeology, and, while not entirely satisfactory in some respects, is perhaps as good as can be devised in the present state of knowledge.

Careful study of the entire collection has failed to show the slightest difference in the form, finish, or material of implements from the same locality, whether found in mounds or graves or on the surface; hence no attempt is made to separate the two classes of objects. Allowance is to be made for the weathering of a surface specimen, but this is the only distinction.

It is not always easy to identify a stone, even with a fresh surface; in a weathered specimen it is often impossible. For this reason the material of which a specimen is made may not be correctly named; frequently the alteration due to exposure will change the appearance of a rock very much, and in such a case the best that can be done is to tell what it looks most like. The material of a majority of specimens however, or at least the classes of rock to which they belong, as granite, porphyry, etc., are correctly named; to give a more exact name would be possible only by the destruction or injury of the specimen. There are a few terms used which may be here explained.

“Compact quartzite” is a very hard, close-grained, siliceous rock, sometimes nearly a flint, and again closely approaching novaculite. “Greenstone” may be diorite or diabase, or it may be a very compact dark sandstone or quartzite so weathered that its nature can not be determined from superficial observation. “Argillite” refers to any slaty rock; it may be so soft as to be easily cut with a knife, or nearly as hard as quartzite. Usually it is greenish in color.

A comprehensive study of all available collections will no doubt modify materially the classification and system of types here presented.

The quotations from eminent anthropologists given below show the difficulties in the way of establishing a satisfactory system of types, or of assigning certain forms to particular localities. In most of these quotations the substance only of the author’s remarks is given.

According to Dr. E. B. Tylor, the flint arrows of the Dakota, the Apache, or the Comanche might easily be mistaken for the weapons dug up on the banks of the Thames;1 while cores of flint in Scandinavia and of obsidian in Mexico are exactly alike,2 and a tray filled with European arrowheads can not be distinguished from a tray of American ones.3 Prof. Otis T. Mason observes that the great variety of form in such weapons after they are finished is due partly to nature and partly to the workman’s desire to produce a certain kind of implement. All sorts of pebbles lie at the hand of the savage mechanic, none of them just what he wants. He selects the best.4 Perhaps the truth about the shape is that the savage found it thus and let it so remain.5 59

The state of things among the lower tribes which presents itself to the student is a substantial similarity in knowledge, arts, and customs, running through the whole world. Not that the whole culture of all tribes is alike—far from it; but if any art or custom belonging to a low tribe is selected at random, the likelihood is that something substantially like it may be found in at least one place thousands of miles off, though it frequently happens that there are large intervening areas where it has not been observed.6

On the whole, it seems most probable that many of the simpler weapons, implements, etc., have been invented independently by various savage tribes. Though they are remarkably similar, they are at the same time curiously different. The necessaries of life are simple and similar all over the world. The materials with which men have to deal are also very much alike; wood, bone, and to a certain extent stone, have everywhere the same properties. The obsidian flakes of the Aztecs resemble the flint flakes of our ancestors, not so much because the ancient Briton resembled the Aztec, as because the fracture of flint is like that of obsidian. So also the pointed bones used as awls are necessarily similar all over the world. Similarity exists, in fact, rather in the raw material than in the manufactured article, and some even of the simplest implements of stone are very different among different races.7

Tylor again says:

When, however, their full value has been given to the differences in the productions of the Ground Stone Age, there remains a residue of a most remarkable kind. In the first place, a very small number of classes, flakes, knives, scrapers, spear and arrow heads, celts, and hammers take in the great mass of specimens in museums; and in the second place, the prevailing character of these implements, whether modern or thousands of years old, whether found on this side of the world or on the other, is a marked uniformity. The ethnographer who has studied the stone implements of Europe, Asia, North or South America, or Polynesia, may consider the specimens from the district he has studied as types from which those of other districts differ, as a class, by the presence or absence of a few peculiar instruments, and individually in more or less important details of shape or finish, unless, as sometimes happens, they do not differ perceptibly at all. So great is this uniformity in the stone implements of different places and times, that it goes far to neutralize their value as distinctive of different races. It is clear that no great help in tracing the minute history of the growth and migration of tribes is to be got from an arrowhead which might have come from Polynesia, or Siberia, or the Isle of Man, or from a celt which might be, for all its appearance shows, Mexican, Irish, or Tahitian. If an observer, tolerably acquainted with stone implements, had an unticketed collection placed before him, the largeness of the number of specimens which he would not confidently assign, by mere inspection, to their proper countries, would serve as a fair measure of their general uniformity. Even when aided by mineralogical knowledge, often a great help, he would have to leave a large fraction of the whole in an unclassified heap, confessing that he did not know within thousands of miles or thousands of years where and when they were made.

How, then, is this remarkable uniformity to be explained? The principle that man does the same thing under the same circumstances will account for much, but 60 it is very doubtful whether it can be stretched far enough to account for even the greater proportion of the facts in question. The other side of the argument is, of course, that resemblance is due to connection, and the truth is made up of the two, though in what proportion we do not know.8

While the several authors quoted do not fully agree, and some are even slightly self-contradictory, still, if the statements are to be taken at their face value, it would seem that efforts to make such classifications are mainly a waste of time.

It may be premised that in every class of implements there are almost as many forms as specimens, if every variation in size or pattern is to be considered; and these merge into one another imperceptibly. Not only is this the case with individual types, but the classes themselves, totally unlike as their more pronounced forms may be, gradually approach one another until there is found a medium type whose place can not be definitely fixed.

THE ARTS AND THEIR DISTRIBUTION.

Districts.

As space would be needlessly occupied by attempting to name each county, the area from which specimens have been obtained is, for convenience, divided into districts. These divisions are for use in this article only, and are not intended as archeologic districts.

In the tables given under each heading, the names of counties or districts show where the types described are obtained; the columns following show the number of specimens of each material mentioned in the collection of the Bureau.

Where a limited area only has been examined in any division, the name of the county is usually given; but where specimens of any kind have been obtained from different counties near one another, they are assigned to the district including those counties. The districts are as follows:

Arkansas.

Northeastern: Between White and Mississippi rivers.

Southeastern: Between White and Washita rivers from Clarendon to Arkadelphia.

Southwestern: West of Washita river and south of Arkadelphia, including Bowie and Red River counties, Texas.

Central: From Dardanelles southward and eastward to the above limits.

Alabama.

Northeastern: Bordering Tennessee river east of Decatur.

Northwestern: Bordering Tennessee river west of Decatur. 61 Coosa: Bordering Coosa river southward to and including Dallas county.

Tuscaloosa: Bordering the Tuscaloosa and Little Tombigbee, and extending a short distance below their confluence.

Ohio.

Miami valley: The country along the two Miami rivers, including Shelby county on the north and Madison and Brown counties on the east.

Scioto valley: South of Franklin county, including Adams and Lawrence counties.

Central: Including Union, Knox, Perry, and Franklin counties, and the area within these limits.

Wisconsin.

Southwestern: The counties bordering on either side of Mississippi river from La Crosse to Dubuque (Iowa).

Eastern: The portion between Lake Michigan, Lake Winnebago, and the Illinois line.

Southern: Dane and adjoining counties.

Iowa.

Keokuk: The southeastern corner of the state and adjacent portions of Illinois and Missouri.

Tennessee.

Eastern: All the mountain district, with the extreme southwestern part of Virginia.

Western: From Mississippi river to and including the tier of counties east of the Tennessee.

Northern: The northern half of the interior portion.

Southern: The southern half of this portion.

South Carolina.

Northwestern: North and west of a line from Lancaster to Columbia. As no other portion of the state has been examined under direction of the Bureau, only the name of the state is used herein, reference being always to this section.

Georgia.

Northwestern: The portion northwest of the Chattahoochee.

Southwestern: Area contiguous to the lower Chattahoochee and Flint river.

Savannah: The vicinity of the city of Savannah, where a large collection was gathered.

62

Kentucky.

Northeastern: Between Kentucky, Big Sandy, and Ohio rivers.

Southeastern: From Estill and Cumberland counties to the Tennessee and Virginia state lines.

Central: Between Green and Ohio rivers, west of the last described districts.

Southern: From Green river southward and as far westward as Christian county.

Western: West of Green river and Christian county.

North Carolina.

Western: West of Charlotte.

Central: Between Charlotte and Raleigh.

Illinois.

Southwestern: From the mouth of the Cumberland to Washington county, and thence to the Mississippi.

Descriptive Terms.

The various forms of implements will now be considered. As stated above, the names given the various articles are those by which they are usually known; but it may be well to define some of the terms used.

In the grooved axes, edge refers to the cutting portion; blade, to the part below the groove; poll or head, to that above the groove; face, to the wider or flat portion of the surface; side, to the narrower part; front, to that side farther from the hand, and back, to the side nearer the hand when in use.

In celts, the terms are the same, so far as they are applicable; blade referring to the lower half of the implement; that is, to the portion on which the cutting edge is formed.

Ground and Pecked Articles.

Grooved Axes.

The implements known as grooved axes seem to be of general distribution throughout the United States; being, so far can be learned from various writers, much more numerous east of Mississippi river than west of it. It must be remembered, however, that thousands of diligent collectors have carefully searched for such things in the east, while in the west little attention has been paid to them; consequently, deductions are not to be made concerning their relative abundance or scarcity, until further knowledge is gained. The same remark will apply to every form of aboriginal relic. 63

In the eastern and interior states, the grooved axes are far more abundant than the celts of the same size9, because as a rule only the larger implements of this class are grooved. All the ordinary varieties of axes and hatchets are found about Lake Champlain, by far the most abundant being celts, or grooveless axes.10

According to Adair and other early observers, the southern Indians had axes of stone, around the grooved heads of which they twisted hickory withes to serve as handles; with these they deadened timber by girdling or cutting through the bark.11 According to travelers of a later generation among the western Indians, similar implements were used on the plains to chop up the vertebræ of buffaloes, which were boiled to obtain the marrow.12

These statements, which might be multiplied, show that such objects are to be found widely scattered; none, however, give information more definite than that the axes are “grooved,” no reference being made to the shape of the ax or the manner of grooving.

The various modes of mounting axes and celts in handles are illustrated in the Smithsonian Report for 1879.

Stone axes were used in Europe by the Germans at as late a period as the Thirty Years’ war, and are supposed to have been used by the Anglo-Saxons at the battle of Hastings.13

Fig. 29.—Grooved ax, showing groove projections.

Axes having two grooves occur in considerable numbers in the pueblos of southwestern United States, but they are extremely rare elsewhere and unknown in most districts; as the objects are generally small, the utility of the second groove is not evident.

The arrangement of stone axes may be based upon the manner of forming the groove. In one class are placed those which in the process of making had a ridge left encircling the weapon, in which the groove was formed. This gives the ax greater strength with the same material. Usually the groove has been worked just deep enough to reach the body of the ax; that is, to such a depth that should the projections be ground off there would remain a celt-like implement (as shown in 64 figure 29, of chlorite-schist, from Sullivan county, Tennessee). The axes of this class in the Bureau collection are shown in the following table:

In the second class the groove is formed by pecking into the body of the ax after the latter is dressed into shape; in this pattern a regular continuous line from edge to poll would touch only the margins of the groove, leaving it beneath. An apparent medium between the two is sometimes seen, in which there is a projection on the lower side of the groove only; this is due, usually, to dressing the blade down thinner after the implement was originally worked to a symmetric outline. By continuous or long use the edge of the ax becomes broken or blunted and requires sharpening, and in order to keep the proper outline to make the tool efficient, it is necessary to work the blade thinner as it becomes shorter. No such change is required in the poll, consequently a projection is formed where originally there was no trace of one.

Fig. 30.—Grooved ax, showing pointed edge.

There are different methods of finishing the ax, which may appear with either form of groove. The poll may be worked into the shape of a flattened hemisphere, may be flat on top, with the part between the groove and the top straight, convex or concave, or may be worked to a blunt point, with straight or concave lines to the groove. The blade may taper from the groove to the edge, with straight or curved sides, which may run almost parallel or may be drawn to a blunt-pointed edge. This latter form is probably due to breaking or wearing of the blade, which is reworked, as shown in figure 30, of granite, from Boone county, Missouri.

There are a very few specimens, as noted below, in which the ax gradually increases in width from the poll to the edge; but such specimens seem to be made of stones which had this form approximately at the beginning, and were worked into such shape as would give a suitable implement with the least labor.

In nearly every instance the groove of an ax with a groove projection extends entirely around with practically the same depth, and the blade 65 of the ax has an elliptical section. There are, however, a few with the back flattened; and while many of the second division may be similar in section, and in having the groove extend entirely around, yet in this class are to be placed nearly all of those only partly encircled by a groove or showing some other section than the ellipse.

Fig. 31.—Grooved ax, showing groove entirely around.

Fig. 32.—Grooved ax, slender, showing groove entirely around.

With these exceptions, the second class of grooved stone axes comprises seven groups, which may be described and tabulated as follows:

A. Grooved entirely around, elliptical section, polls dressed in any of the ways given above; three or four have the blunt-pointed edge (figure 31, of granite, from Bradley county, Tennessee).

66

B. Long, narrow, and thin, giving a much flattened elliptical section. These are classed with axes on account of the grooves, although too thin and usually of material too soft to endure violent usage. The edges are nicked, striated, or polished, as though from use as hoes or adzes (figure 32, of argillite, from Bradley county, Tennessee).

Fig. 33.—Grooved ax, showing grooved back.

Fig. 34.—Grooved ax, showing grooved back.

C. Grooved on both faces and one side; back hollowed, usually in a straight line the whole length; front drawn in from the groove to give a narrower edge (figures 33, of porphyry, from Brown county, Ohio, and 34, of granite, from Kanawha valley, West Virginia).

67

D. Same method of grooving; back is rounded, and may be in a straight or curved line the entire length, or a broken line straight in each direction from the groove. The type is illustrated by figure 35, of granite, from Keokuk, Iowa. This specimen is unusually wide and thin; generally the outlines are similar to those last described.

Fig. 35.—Grooved ax, showing rounded back.

E. Grooved like the last; same general form, except that the back is flat (figures 36, of sienite, from Brown county, Ohio, and 37, of granite, from Drew county, Arkansas).

68

F. Grooved on both faces and one side, with both sides flat. There is only one of this form in the collection; it is of argillite, from Keokuk, Iowa.

G. Grooved on faces only, with both sides flat (figure 38, of granite, from Keokuk, Iowa). There are from the same place one of porphyry, one of argillite, and three of sienite. This and the preceding form seem peculiar to that locality.

Fig. 36.—Grooved ax, showing flattened curved back.

Fig. 37.—Grooved ax, showing flattened straight back.

There are a few exceptional forms which are not placed with those just given, since they may have some features common to all except the Keokuk type, while in other respects they differ from all. Among them are some entire-grooved or grooved only on the two sides and one face; the general outline may correspond with some of the regular forms, but one face is curved from poll to edge, while the other is straight or nearly so (figure 39, of granite, from Wilkes county, North Carolina). This specimen has a depression, as if worn by the end of a handle, on the straight face at the lower edge of the groove.

None of this form are long enough for hoes, and although they may have been used for axes and hatchets, their shape seems to indicate use as adzes. Besides the one figured there are two from Savannah, Georgia; three from eastern Tennessee, one with a slight groove and very deep side notches; and three from western North Carolina, two of them entire-grooved with groove projections. 69

Fig. 38.—Grooved ax, Keokuk type.

Fig. 39.—Grooved ax, showing adze form.

Another unusual form, which may come under any of the foregoing figures, has the groove crossing the implement diagonally, in such a way as to cause the blade to incline backward (figure 40, of granite, from Carter county, Tennessee). Besides the specimen illustrated, this form is also represented by one of granite from northwestern North Carolina with projection for groove; two of argillite from southwestern Tennessee; one, widest at edge, from Savannah, Georgia; one from Ross county, Ohio; and two of granite, highly polished, grooved on faces and one side, with backs flat, from Kanawha valley, West Virginia.

Of the axes wider at the edge than at any point above (of which the specimen illustrated in figure 41, of granite, from a grave at Kingsport, Tennessee, may be taken as a type,) there are one of diorite from Kanawha valley, West Virginia, which seems to have been of ordinary pattern but broken and redressed to its present form; and from Savannah, Georgia, one of uniform taper with diagonal groove, and one widening irregularly until the blade is fully twice the width of the poll.

Fig. 40.—Grooved ax, showing diagonal groove.

Fig. 41.—Grooved ax, showing wide edge.

Many, if not a majority, of the entire-grooved axes have the groove wide enough for a very large handle, or for an ordinary withe to be twisted twice around. In those which have one side ungrooved, the intention was to admit a wedge between the stone and the curve of the handle. The handles were very firmly fastened; two axes in the collection have been broken in such a way that on one side, from the top half way down, the blade is gone, carrying away the groove on that side; yet the polish of the groove extends 70 over the fractured surface, which has never been reworked, showing that the tool was long used after this accident. As the handles could easily slip off over the top in specimens thus broken, they must have been tightly lashed; perhaps gum or glue was used.

Fig. 42.—Grooved ax, showing curved edge.

Partly finished specimens show that the groove was pecked out and the edge ground before the remaining parts of the ax were worked. Some have the edge ground sharp and the groove worn smooth or even polished by long use, while all the rest of the implement retains the original weathered surface. A stone was always chosen that could be brought to the desired form with the least labor, and very often one could be found that required but little work to make a very satisfactory weapon or implement or even ornament.

Occasionally specimens indicate by the manner of wear their application to certain kinds of work. Sometimes the edge is curved by the wearing away of one face until it has almost a gouge form; sometimes the side of the blade next the hand, again that farthest away, is more worn. This in time would give the blunt-pointed edge. A peculiar finish of the lower part of the blade, which is also seen in a few celts, is shown in figure 42, of sienite, from Carter county, Tennessee. One half of each face has been left full, and the part opposite hollowed out, giving an ogee curve to the edge. Figure 43, of granite, from Jefferson county, Tennessee, seems to have a ridge on the upper side of the groove; but closer examination shows that it once had a groove projection, and that afterwards the poll was nearly all broken away and a new groove made lower down, so that what was originally the lower projection is now above the groove, the remainder of the poll being worked down to a point.

There are a few hammers which differ from the ordinary ax only in being blunt instead of sharp. They may be nothing more than broken axes, utilized as hammers instead of being resharpened.

Fig. 43.—Grooved ax, showing single groove projection.

Under this head may be placed implements plainly used as adzes. They are much longer than axes in proportion to their other dimensions, have one face convex, the other straight or concave. They may be placed in the same class as the specimen shown in figure 39, and also those represented in figures 44 and 45, from McMinn county, Tennessee. There is also a similar adze from Saline county, Arkansas. All the specimens of this class are of argillite. 71

With the grooved axes is also placed a class of implements that may be called axes notched on the sides. Many of them were no doubt used as sinkers; but some of the same form, size, and material have the notches and sometimes portions of the face worn perfectly smooth, while frequently they are ground to a sharp edge. Again, even in those that have not the least polish, the edge shows marks that would seem to result from use as axes, adzes, or hoes.

There are three divisions of this class of implements, as follows:

A. Unworked, except notches; probably sinkers.

B. Partly ground sharp edges, mostly with polished notches, sometimes with faces polished from one notch to the other (figure 46, of argillite, from Cocke county, Tennessee). In addition there are 11 examples of argillite, besides one of mica-schist from eastern Tennessee and another of sandstone from Savannah, Georgia.

Fig. 44.—Grooved adze.

Fig. 45.—Grooved adze, showing curved blade.

C. Roughly chipped, with notches often at the middle but sometimes nearer one end. Probably most of these were sinkers; but as above stated the edges show marks of use, apparently in scraping, digging, or striking. Of these the following examples are in the Bureau collection: From several localities in eastern Tennessee, 40 of argillite; from Montgomery county, North Carolina, 24 of argillite and quartzite; 72 from Kanawha valley, West Virginia, and from Savannah, Georgia, a few specimens of the same materials.

Celts.

What is true of the uses and distribution of stone axes applies with much the same force to what are called celts—not a good descriptive term, but one which is now given to the implement in lieu of something better. It would appear difficult or impossible to do with these rude tools any work for which we commonly use an ax or hatchet; and yet, by the aid of fire, or even without it, the aborigines contrived to accomplish a great deal with them.

The Maori of New Zealand do all their wonderful work of wood carving with only a chisel or adze (of stone or shell).14 Among the Iroquois, in cutting trees, fire was applied at the root, the coals were scraped away with a chisel, and this process was repeated until the tree was felled. The trunk was divided into lengths in the same way. Similarly canoes and mortars were hollowed out.15 The Virginia Indians at an early day employed a similar process. They also cleared ground for cultivation by deadening trees with their tomahawks,16 and used adzes made of shell in cleaning out the charred wood in making canoes.17 The Nootka of the northwestern part of the continent in felling a tree use a flint or elkhorn set in a handle, this being struck with a stone mallet. In hollowing canoes a musselshell also is used as an adze, and sometimes fire is applied. The outside is shaped by similar means.18

Fig. 46.—Notched ax, showing polished edge.

Stone chisels have been found in various steatite quarries, where vessels and other utensils of this material were made, and the marks of their use is plain both on the vessels in an unfinished state and on the cores, as well as on the quarry face.19

The different ways of hafting, as shown by specimens in the Bureau collection, were as follows:

(1) A hole was cut entirely through a stick and the celt was inserted so that it would project on both sides;

(2) The hole was cut partly through, and the celt was pushed in as far as it would go; 73

(3) The top of the celt was set in a socket of deer horn, which was put into a handle as in form 2;

(4) Small celt-shaped knives or scrapers were set into the end of a piece of antler long enough to be used as a handle;

(5) A forked branch was so cut as to make two prongs of nearly equal length, and the celt was fastened to the end of one, parallel with it, the other being used to guide and steady it, a prong being held in each hand;

(6) The fork of a root or branch was trimmed so as to make a flat face at any desired angle, to which the celt was lashed, a shoulder, against which the end of the celt was set, being sometimes cut in the wood;

(7) A stick was split its entire length and a single turn taken around the celt, the ends being brought together and tied, forming a round handle;

(8) A stick was split part way, one fork cut off and the other wrapped once or twice and tied, thus forming a round handle of solid wood.

Fig. 47.—Celt, showing blade thick near edge.

Fig. 48.—Celt, showing blade thick near edge.

Forms 5 and 6 were used as adzes; forms 7 and 8 are the same methods as employed in hafting grooved axes.

A mounting similar to form 4 is seen in some Alaska specimens of celt-scrapers in which the implement is fastened to a piece of wood so as to project a short distance, and used like a plane. In all these, the celt is very firmly fastened to the handle with sinew or rawhide, which, when put on green, contracts with great force and binds like wire.

As to the forms of celts, no division is practicable based on anything but their entire appearance. The following descriptions and tabulations represent the material of this kind in the Bureau collection:

A. Round or nearly round section, pointed or flattened at the top, blade rapidly thickening from the edge; a few are polished at the top, but most of them show marks of a maul or hammer; all have been highly polished; all of this class were probably used as wedges, as their shape renders them more fit for this purpose than for any other; the 74 battered tops indicate such usage. The few not showing such marks may have been set into a bumper of wood or horn, or used with wooden mauls. They vary in length from 2½ to 7½ inches. They are represented by the specimen shown in figure 47, of argillite, from Lincoln county, Arkansas; there are also one from a mound in Sumter county, Alabama (figure 48), and one from Kanawha valley, West Virginia, both of serpentine and elliptical in section, though the form of the edge puts them in this class. The following specimens are typical representations of the class:

Fig. 49.—Celt, showing long, slender form.

B. Long, narrow, elliptical section, pointed top, curved or straight edges, sides straight or gently curved. None of these seem to have been put to any rough use, as the edges are quite sharp and the entire surface is well polished; length from 4¼ to 12½ inches. The type is illustrated by figure 49, of argillite, from a mound in Monroe county, Tennessee.

C. Thick, almost round section, round-pointed top, nearly straight to sharp-curved edge, sides gently curved, widest at edge or just above. Most of these show marks of use as cutting tools or hatchets. In 75 many the top has been roughened as if for insertion into a hole cut in a piece of wood; others have this roughening around the middle or immediately above, leaving a polish at both ends, and these were hafted probably by means of a stick or withe twisted around them. The roughening is a secondary operation, having no relation to the making of the implement; it was produced by pecking after the surface was polished. In a few cases it extends from the top well down the sides; but usually it reaches but a little way below the top, or else is in a circle around the body of the celt. Most of them have sharp edges; a few have edges either chipped or blunted and polished, showing long usage. Two from Kanawha valley (one roughened for handle) have the edges worn in on one of the faces until they almost resemble gouges; but that they were not intended as such is shown by the concavity being nearer one side and not reaching entirely across. The length ranges from 4½ to 10 inches. The type is illustrated by figures 50 and 51, both of sienite, from Lauderdale county, Tennessee.

Fig. 50.—Celt, nearly round section.

Fig. 51.—Celt, nearly round section.

This may be regarded as the typical form of celt for eastern United States, and its geographic distribution is exceptionally wide, as shown in the table.

The Bureau collection includes the following specimens of this class:

76

D. Of the form last described, except in being much thinner; some have the tops battered, showing use as wedges; length from 3 to 9 inches.

Fig. 52.—Celt, showing nearly diamond section.

E. Pointed oval, or nearly diamond section, sides straight or slightly curved; length 6 to 12½ inches. Few as these are, they vary considerably in appearance. The group is illustrated by figure 52, showing a specimen of brown flint, containing numerous small deposits of chalcedony, from Benton county, Tennessee; polished over the entire surface, the edge highly so.

In addition, there are the following examples: From Caldwell county, North Carolina, one of porphyry and one of granite, the latter roughened on sides for handle; from McMinn county, Tennessee, one of gray flint, highly polished over its surface, except the top, which is much battered; from Cocke county, Tennessee, one of argillite.

F. Elliptical section, flattened or rounded top, edge curved or nearly straight, sides straight or gently curved, tapering from edge to top or in a few cases nearly parallel. These present many variations in finish and in evidence of use. Some are well polished over the entire surface; some have only the lower part polished; while some are entirely without polish except at the extreme edge. In some the top is battered; some have the surface roughened for handle at the top, others around the middle, still others all over the upper half or even more than half. One from McMinn county, Tennessee, has a roughly pecked shallow groove at the middle. Several have the edge 77 very blunt, the faces at the edge form almost a right angle; these are thickest very near the edge and become gradually thinner toward the top. Most of this kind are from Caldwell county, North Carolina; the same form coming also from Monroe county, Tennessee, and from Savannah, Georgia. The length is from 3 to 7½ inches. Figure 53, of compact quartzite, from Monroe county, Tennessee; figure 54, of granite; and figure 55, of sienite, from Caldwell county, North Carolina.

Fig. 53.—Celt.

Fig. 54.—Celt.

Fig. 55.—Celt.

78

G. Of the same general pattern as the last, except that the sides widen just before reaching the edge, giving a “bell shape” (figure 56). The length is from 6¼ to 8 inches. In this group there are two specimens of granite, two of porphyry, and one of sienite, all from Yazoo county, Mississippi. Two have their tops roughened.

Fig. 56.—Celt, showing “bell shape” and roughening for handle.

Fig. 57.—Celt, showing rectangular section.

H. Rectangular section, occasionally with the corners sufficiently rounded to give a somewhat elliptical section; top flattened or rounded; sides straight and parallel or nearly so, sometimes very slightly curved. Most have polished surfaces; only three or four show any battering, or roughening for handle. A large one of hornblende from Lauderdale county, Tennessee, has the edge dulled and polished by use. Length is from 2 to 9 inches. Figure 57, of argillite, from a mound in Monroe county, Tennessee. The distribution of this class of celts is wide, as shown by the following table:

I. Thickest at top (wedge form), section elliptical or nearly rectangular; sides straight or curved, widest at edge or nearly parallel. A few are roughened for handling, and one or two are battered at top by hammering; most are small. The type is shown in figure 58, of granite, 79 from Carroll county, Indiana. This class of celts also is widely distributed and diverse in material.

Fig. 58.—Celt, showing wedge-shape.

Fig. 59.—Celt, showing half-elliptical section.

J. Flat on one side, convex on the other, giving a semi-elliptical section; sides nearly parallel; top flat or rounded. These were evidently intended for scrapers; none are at all chipped or battered from use, and with very few exceptions the whole surface is highly polished. The flint and jasper specimens, which have been first chipped into shape, have the facets and edge as smooth as though finished on an emery 80 wheel. Similar forms, except with flat instead of convex upper surfaces, are known to have been used as adzes, but these have no marks of such use. The length ranges from 2 to 8 inches, but most are small. The type is shown in figure 59, of brown flint, from a grave in Alexander county, Illinois.

K. Similar to last, except that the sides come to a point at the top; length, 3½ to 9 inches. Very few of either pattern are above 5 inches long, the larger ones being mostly of flint (figure 60, of sienite, from Warren county, Ohio).

L. Sides concave, top narrow. Nearly every specimen has the upper portion pecked rough; one from Bradley county, Tennessee, and another from Mississippi county, Arkansas, are entirely polished. The latter has the scraper-form edge to be described later and is of exceptionally large size; it measures 5½ inches, being the only one exceeding 5 inches in length.

M. Top flat, round, or pointed; the blade usually begins a little below the middle, and is perfectly smooth in every case; in some the blade is not over an inch in length, probably reduced by continual sharpening. They may have been scrapers, though they do not have that form; if 81 used as weapons they were probably set into the end of a piece of antler, which, in turn, was set in a club. The type is shown in figure 61, of argillite, from Monroe county, Tennessee.

Fig. 60.—Celt showing half-elliptical section.

Fig. 61.—Celt, showing concave sides.

N. Ground down thin, with a flat-elliptical or nearly rectangular section; sides straight or slightly curved, nearly parallel or tapering considerably to the top, which is either rounded or flattened. All are polished over the entire surface; none show any marks of use as wedges or hatchets, and most of them are too delicate for such use. The longer ones can be readily grasped in the hand, and are as well adapted to stripping off the hide of an animal, dividing the skeleton at the joints, or stripping the flesh from the bones, as anything made of stone can be; while the smaller ones, set in a handle to afford a grip, would answer the same purpose. There are three which are sharp at both ends, one having one symmetrical and one scraper-form edge; one having a scraper-form edge at each end on opposite sides; and one of rather soft argillite, unfinished, which has marks of pecking, chipping, and grinding, showing that any of these methods were practiced, as was most convenient. All these are from eastern Tennessee. The features are illustrated in 82 figures 62, of argillite, from a mound, Caldwell county, North Carolina; 63, of black flinty slate, very hard, from a mound, Poinsett county, Arkansas; and 64, of argillite, from a mound, Monroe county, Tennessee.

Fig. 62.—Thin polished celt.

Fig. 63.—Thin polished celt.

Fig. 64.—Thin polished celt.

Gouges.

While there are perhaps no true gouges in the collection, there are some examples of a form between a celt and a gouge, illustrated in figure 65, of serpentine, from Caldwell county, North Carolina.

Implements of this form are known to have been used to tap sugar maples, and also to hollow out wooden troughs, and are very common 83 in the north, though less abundant in the south.20 It is in those localities in which bark instead of logs was used for canoes that they are most numerous. Sometimes they were hollowed the whole length and used as spiles.21 They were also employed instead of celts in hollowing wooden mortars and the like when a more regular concavity was desired.22

Chisels and Scrapers.

The aboriginal implements known as “chisels” are round, elliptical, or rectangular in section. The flint and jasper specimens are generally widest at the edge, the reverse being usually the case with those of other material. Most of them have marks of hammers at the blunt end, though some are polished at the top and a few, from eastern Tennessee, are sharp at both ends. The top (except in the double-edged ones) is usually flat, though a few are pointed or very thin, almost with cutting edges. Jaspers and flints are chipped, with the facets polished, the edges highly so. Any form may occur in any locality. Almost invariably they have scraper-form edges. The length is from 2 to 6 inches.

Fig. 65.—Celt, showing thin, gouge-form edge.

Typical examples are shown in figure 66, of yellow jasper, from a grave in Mississippi county, Arkansas; figure 67, of novaculite, from an unknown locality in Arkansas; figure 68, of serpentine, from Bradley county, Tennessee; figure 69, of sienite, from Caldwell county, North Carolina; and figure 70, of gray jasper, from Bradley county, Tennessee. Some specimens are sharp and worn at both ends, and could have been used only with handles.

Fig. 66.—Celt, chisel-form.

Fig. 67.—Celt, chisel-form.

Fig. 68.—Celt, chisel-form.

84

The Bureau collection includes the following specimens:

The high polish sometimes found on the top of a round-pointed celt may be due to its working slightly in the socket in its handle of wood, deerhorn, or other material.

Fig. 69.—Celt, chisel-form.

Fig. 70.—Celt, chisel-form.

By celts having a scraper-form edge is meant those having the edge to one side of the median line, due to constant use of one face. This face, at the edge, is in a straight line from side to side; it may have a chisel-like flattening, or may curve toward the middle of the celt for a short distance and then have the same form to the top as the other face, which is convex or curved, as in the ordinary hatchet-celt. They form a medium between celts whose faces gradually curve from top to edge, and the celt-scrapers which are flat on one side. Among the thicker celts this form is quite rare, though several, especially one from Kanawha valley, West Virginia (represented in figure 74), are quite pronounced. In the thinner specimens, however, a majority are of this pattern, while in some types, nearly all indeed, even those up to 6 inches long, are so beveled. The type, of which an illustration is shown in figure 71, is of very hard black slate; the same form is presented in figures 66 and 70.

From Bartow county, Georgia, is a scraper made from the edge of a celt which has been broken diagonally across from one face to the other. A stem like that of a spear-head has been formed by chipping away the sides of the part broken, which gives a convenient attachment for a handle; the original edge is unchanged except in the wear which has resulted from its new use. 85

The specimen shown in figure 72 (of argillite, from McMinn county, Tennessee) is introduced on account of its undoubted use as a scraper, and because it is much smaller than some of the chipped flints thus classified, the edge being less than an inch wide; the sides are roughly incurved.

In Bradley county, Tennessee, there were found over 200 specimens of very small, thin, flat, waterworn sandstone pebbles, which were mostly in their natural condition, except that they had one side rubbed to a sharp edge. A few, more slender, were ground to a point. Some of them have a handle chipped out on the side opposite the edge, sometimes with nicks in it, made for attachment to a handle by means of a cord. Most of these specimens are less than 2 inches in length. No suggestion is offered as to their use.

Fig. 71.—Celt, showing scraper-form edge.

Fig. 72.—Scraper.

A granite implement from Union county, Illinois, with nearly rectangular section, slightly curved sides, rounded corners, and high polish over the entire surface, having nearly the same thickness (about an inch) at every part, would seem to be a polishing or rubbing stone. There are, however, one from Warren county, Ohio, and three from Kanawha valley, West Virginia, of almost exactly the same size and pattern, which have had one end ground off to a sharp edge; so the specimen may be only an unfinished celt. One of those from Kanawha valley has had the edge partly broken away, and one face has been pecked considerably in an attempt to restore it for use; but the intention was not carried out. Some celts, not of the scraper pattern, which have the edge to one side of the median line, are perhaps broken or blunted specimens redressed on one side only.

Fig. 73.—Scraper or adze, with projecting ridge.

Fig. 74.—Adze or scraper.

Figure 73 exhibits a specimen of argillite from Carter county, Tennessee, probably an adze or scraper, with a projection to keep the 86 implement from being forced into the handle. The edge is symmetrical, though much striated. The specimen shown in figure 74 (of granite, from Kanawha valley, West Virginia) represents a peculiar form. There are several like it in the collection, all but this one from islands in the Pacific.

Chipped Celts.

On account of their shape and undoubted use, a class of celts, although neither pecked nor ground, is introduced. Many of them resemble, in most respects, the so-called paleolithic implements, though sometimes of better finish. They are made with a rounded top and nearly parallel sides; rudely triangular; or with the sides curved to a point at the top. The edge may be straight or curved, and is usually chipped, though sometimes ground; a few are chisel-shaped. Usually they show no signs of wear; when they do, it is always in the form of a polish at the larger end, or on the exposed facets. One of black flint, 8 inches long, from Kanawha valley, has a scraper-form edge, smoothly polished. Many, even of those scarcely changed from their original form and natural surface, have the edges dulled and polished from use as scrapers or adzes.

Fig. 75.—chipped celt.

Fig. 76.—Chipped celt.

Fig. 77.—Chipped celt.

The collection includes the following examples: 36 of argillite, flint, porphyry, and compact quartzite, from Montgomery county, North Carolina, some with the wider edge sharp (figure 75, of flint); 12 of limestone and flint from Mason county, Kentucky; 70 of argillite, a few with the edges ground, from southeastern Tennessee (figure 76, from McMinn county); over 300 from Kanawha valley, nearly all of black flint, a few being of diorite or quartzite—some are partly polished, or have ground edges (figure 77, of black flint, from a mound).

Hematite Celts.

With the exception of two from Iowa and a few from Preston county, West Virginia, the hematite celts in the collection are from Kanawha valley, and are small, ranging in length from 1 to 2¾ inches, except one 4½ and one 5½ inches. They are illustrated in figures 78, 79, 80, and 81, the last from a mound. Nearly all have been ground directly from the 87 nodule or concretion in which this ore of iron so frequently appears. Occasionally one of homogeneous structure has been chipped into form before grinding, the facets in some cases being rubbed nearly away. Sometimes they have a rectangular outline, but usually the sides taper from the edge to the top by a gradual curve, or are parallel a part of the way and then taper either by a straight or, oftener, by a curved line. The section is rectangular or elliptical.

Fig. 78.—Hematite celt.

These implements were probably used as knives or scrapers, being set into the end of a piece of antler, which may in turn have been set into a larger handle of wood. That some were knives is shown by the edge which is dulled to a flat polished surface extending from side to side; and that many were scrapers is shown by their celt-scraper shape, a half elliptical section, or by the scraper-form edge, seen in the largest specimen. Some, however, have the edge symmetrical, as in the hatchet-celts. One has incurved sides, and is roughened on the sides and on the faces near the top.

Fig. 79.—Hematite celt.

Fig. 80.—Hematite celt.

Fig. 81.—Hematite celt.

Pestles.

The fact of the ordinary conical or bell-shaped, long-cylindrical, or somewhat pear-shaped stones having been used for pestles is so well settled that no confirmatory references are needed. A few citations may be given in regard to certain forms sometimes differently classed, especially some of the discoidal stones to be hereafter described.

According to Stevens, the corn crushers used by the Swiss lake dwellers are spherical; some are flattened on two sides, like an orange, others almost round with depressions on four sides. They are about the size of a man’s fist or rather smaller. The Africans have a piece of quartz or other hard stone as large as half a brick, one side of which is convex, to fit the hollow of a larger stone used as a mortar.23 Evans observes that disks sometimes show marks of use as hammers 88 or pestles;24 one found at Ty Mawr was thick, with a cavity on each face.25 In preparing pemmican, the American Indians are known to have pounded the dried meat to a powder between two stones.26 This gives the impression that any suitable stones may have been used; and the ancient California Indians worked out a round stone as an acorn sheller, modern tribes using any smooth stone.27

Fig. 82.—Handled pestle, with expanding base.

The pestles which have the bottom round or convex are generally found in the same localities as the hollowed stone mortars. Several forms of pestles are represented in the collection. They may be grouped as in the following description and tabulation.

A. With expanding base; bottom flat or slightly convex, often with a slight depression in the middle. Handle tapering, or of uniform diameter to the top; in a few, slightly swelling above as if to give a firmer hold. Top rounded, flat, or pointed. Bottom may be very little expanded or may have twice the diameter of the handle. Probably used for pounding grain or seeds on a flat stone, as it could not be used in a mortar even slightly hollowed. None seem to have been used as mullers or rubbers. They may have served for hammers, and would be excellent for cracking nuts, as the pit in the bottom would tend to keep them from flying out to the side. The type is shown in figure 82, of quartzite, from Sullivan county, Tennessee. The distribution is moderately wide, and the material chiefly granite and quartzite, with a few of other rock varieties, as shown in the table:

B. Almost cylindrical, from 6 to 18 inches long and about two inches in diameter. Some of the larger ones were probably rolling-pins, as 89 the ends, either from some fancy finish, or because worked to a point, are of a shape that would make their use as pestles impracticable. Even as rollers, some must have been used for crushing grain that had previously been softened or was not fully matured, as they are of a soft stone that would wear very easily. The shorter ones are blunt at the ends, and may have been used in a shallow wooden mortar; none are adapted for use in stone. The class is illustrated by figure 83, of soft clay slate, from Cherokee county, Georgia.

Fig. 83.—Pestle, long cylindrical form.

Fig. 84.—Pestle, conical.

C. Conical, or truncated cone, bottom flat, convex or curved from one side to the opposite. Some are quite smooth on the bottom as if from rubbing either back and forth or with a rotary motion; while many have the bottom pecked rough, showing use as hammers or pounders. For those with curved bottoms a rocking motion seems best adapted; with the palm resting on the longer side, good work could be done in any of these ways. Typical specimens are shown in figures 84, of quartzite, from Monroe county, Tennessee; 85, of granite, from Warren county, Ohio; and 86, of quartzite, from Saline county, Arkansas. A somewhat aberrant specimen, shown in figure 87, of granite, from Carter county, Tennessee, has an elliptical base, rounded top, and flat bottom; the longer sides grooved for handle. A similar one, of quartzite, came from Warren county, Ohio. There is considerable variety of material, quartzite largely predominating. Although 90 the geographic range is wide, the distribution is rather sparse, and several districts are not represented.

Fig. 85.—Pestle.

Fig. 86.—Pestle.

Fig. 87.—Pestle, grooved for handle.

Fig. 88.—Pestle.

D. Conical, or truncated cone, with top more or less rounded, very little worked, a stone of approximate form having been chosen and the angles and corners pecked off; bottom flat, and in some quite smooth; used as pestles or mullers. The group is represented by 17 specimens of quartzite, all from southeastern Tennessee. 91

E. Not dressed at all on the sides, but with both ends worn to a convex shape. Represented by two specimens of quartzite from southeastern Tennessee.

F. Cylindrical, flat bottom, dome-shaped top, these portions having been carefully pecked into shape. Some are smoothly polished on the bottom, but none elsewhere. Those from Miami valley, and one from Kanawha valley are much longer than the others. The type illustrated in figure 88 is of quartzite, from McMinn county, Tennessee.

Pitted Stones.

There is scarcely a locality in the country where pitted stones are not found; they are indeed of such frequent occurrence that they are seldom considered worth the trouble of gathering.

There can be no “type” among such crude implements; they are almost invariably waterworn sandstone pebbles, with a pit varying from a slight roughening of the surface to a hollow half an inch in depth pecked in each face. They probably belong with hammerstones, as they seldom show other marks of work, the edge in some being only slightly marked in one or two places, while in others it is much worn.

Various numbers of the Journal of the Anthropological Society of Great Britain and Ireland refer to pitted stones as found in every part of the world. According to Evans, slight pits aid in holding stone hammers; they also prevent the jar to a large extent. If used to pound meat or break bones, it would be hard to hold them when greasy without pits.28 Such implements may have had handles of wood with projections to fit the pits,29 though this is not probable; but if so a piece of buckskin on the handle opposite the pits would do better and be more convenient to apply.

Cupped Stones.

Conjecture and theory have had full sway in regard to the uses of cupped stones; but the question is apparently far from solution. There is a prevalent idea that they were used for cracking nuts; but why should an Indian make a large number of holes in a great many stones for such purpose? It is true there would be an advantage in having the nut stand on one end; but very few stones have depressions that will allow this.

Of the southern Indians Adair observes:

They gather a number of hiccory-nuts, which they pound with a round stone, upon a stone, thick and hollowed for the purpose. When they are beat fine enough, they 92 mix them with cold water, in a clay basin, where the shells subside. The other part is an oily, tough, thick, white substance ... with which they eat their bread.30

Lawson’s language regarding the Indians of North Carolina is even more definite. He says:

[They gather] likewise hickerie nuts, which they beat betwixt two great stones, then sift them, so thicken their venison broth therewith, the small shells precipitating to the bottom of the pot, whilst the kernel, in the form of flour, mixes it with the liquor, both these nuts [hickory and chinquapin] made into meal makes a curious soup, either with clear water, or in any meat broth.31

Neither of these statements seems to have any reference to cupped stones. The first is a good description of a mortar with a round pestle, while the second says nothing about any particular form of stone; yet they have been referred to time and again as proof of the nut-stone theory. There would be some difficulty in pounding nuts fine in small holes half an inch or more below where the pounding stone could reach.

C. C. Jones32 was satisfied that cupped stones were used for cracking nuts because great numbers of nut-bearing trees grow where they are found; while Whittlesey, noting the fact that hundreds of them are found throughout northern Ohio, considered them as sockets in which the end of a spindle rested. Dawson33 speaks of “stones having deep hollows in the sides which were mortars for grinding pigments, or sockets for fire drills.”

The cupped stones in the Bureau collection are almost invariably of reddish sandstone, of varying texture, from a few ounces to 30 pounds in weight. The holes are from one to twenty-five in number, of various sizes even in the same stone, and follow the natural contour of the surface even when that is quite irregular; the stone is never dressed or flattened to bring the cups on a level; none show any marks of work, but are the rough blocks or slabs in their natural state.

Many of the holes are roughly pecked in, but the larger ones are usually quite smooth, as if ground out, and almost complete hemispheres. They range from a pit only started or going scarcely beyond the surface to one 2 inches in diameter. The smaller ones with one cup pass into the pitted stones. Occasionally at the bottom of a large cup there is a small secondary hole as though made by a flint drill.

The polished cups may have been used for fire-drill or spindle sockets, though why there should be a number of holes when but one could be used at a time awaits explanation. The rough ones may have been for holding nuts, and so long as they were on the same plane any number could be utilized; but when they are on different parts of the stone, even on opposite sides, as many of them are, the question remains open. Slabs or thin pieces nearly always have cups on both sides, while blocks or thick slabs have them on one side only. On the 93 former a number of nuts could be cracked with one blow of a flat stone and thrown into a receptacle of some kind, either side of the stone being used at pleasure; but there would be no economy of time or work in this method, and it would be very strange that any one should not learn with so much experience that a nut should never be laid on the flat side in cracking. No theory yet advanced accounts for the greater number of such relics, namely, the irregular fragments of stone with cups at varying intervals and different levels.

No division can be made in regard either to size or material of the stone, or to form or finish of the cups. Many of the smaller ones were no doubt paint mortars. One well finished specimen of this class is shown in figure 89; it is of quartzite from 4 feet beneath the surface in Crittenden county, Arkansas.

Fig. 89.—Cupped stone or paint cup.

Cupped stones are found wherever representatives of the Bureau have worked, and numerous references might be given concerning their existence in other localities.

Mullers.

The objects known as mullers are generally flat and smooth on one side and convex on the other, sometimes with a pit in one side or both, mostly of granite, quartzite, or sandstone; rarely of other materials.

A fine specimen of white quartz from Elmore county, Alabama, has the bottom flat and highly polished, the edge perpendicular to bottom and rounding off into the slightly convex top, with a pit at center. Figure 90 represents a muller of marble or crystalline limestone from a grave in Randolph county, Illinois. It has a smooth, flat bottom, with convex top somewhat smaller than the base; around the circumference there is a depression polished by wear. A similar specimen, of diorite, from Carter county, Tennessee, seems to be the lower part of a pestle with expanding base, whose top or handle has been lost, the part remaining having a place for a handle pecked around it.

Fig. 90.—Muller, showing polished surface.

The discoidal stones with this shape were probably used as mullers; they were also used as pestles in the hollow mortars, as the edge is often chipped or pecked, which would account for the pits on the faces. Figure 91 represents a muller of granite from Savannah, Georgia. Sometimes the base has an elliptical instead of a circular outline, as seen in other specimens from Savannah.

Mullers are found wherever there are indications of occupancy for any considerable length of time.

Grinding and Polishing Stones.

Stones evidently used for grinding and polishing need only to be mentioned, as they are of widespread occurrence. Implements used for the 94 former purpose are made of any siliceous stone of convenient size and suitable texture, from a coarse quartzite to a very fine close-grained sandstone, according to the class of work to be done. The markings on them range from the narrow, sharp, incised lines due to shaping a small ornament, to the broad grooves resulting from grinding an ax or celt into form. Nearly all of those in museums are small specimens used for rubbing; but there are many large blocks in various localities, sometimes several feet square, marked and scored in every direction by grinding or sharpening the large implements on them.

Among the polishers may be included a number of small pebbles of very hard siliceous stone, generally some form of quartz, which by the high polish show long use. The larger ones may have been used for rubbing skins in tanning, as they can easily be grasped in the hand. Very few have changed from their primitive form to a greater degree than would naturally result from the wear upon them. A few very small ones, long-ovoid in shape, usually not over 2½ or 3 inches in length, were probably paint mullers, as they are well fitted for use in small paint cups. Many of the discoidal stones—which will be spoken of under the proper head—may have had these functions. The highly polished specimens are all from the southern states. There is one rubbing stone of pumice from Craighead county, Arkansas.

Fig. 91.—Muller, showing polished surface.

Hammerstones.

Hammers or hammerstones show every stage of work, from the ordinary pebble or fragment, with its surface scarcely altered, to the highly polished round or ovoid “ball.” They are usually of the hardest available material, and seem to be of more frequent occurrence in the northern districts than in the southern states, though found everywhere. Used in their earlier stages merely as tools with which to fashion other implements, they were assigned to specified purposes when brought to a better finish or form. A typical example, shown in figure 92, is of granite, from Ross county, Ohio.

The Sioux used an oval stone, with a piece of rawhide covering all 95 but the point and attaching it to a withe handle,34 while the Shoshoni and Ojibwa made use of a round stone, wrapped in leather, attached by a string of 2 inches to a handle 22 inches long covered with leather; this was called a poggamoggan.35 Rounded stones are said to have been used by the California Indians as bolas,36 though it is more probable that they were slung-shots. The ancient Californians worked out a round stone for an acorn-sheller; the present Indians use any smooth stone.37 Elaborately carved round stones, mounted in handles as clubs, are known to have been used by the Queen Charlotte Island Indians for killing fish,38 and other northwestern Indians have been observed to use a round stone inclosed in a net and attached to a line as a sinker.39

Fig. 92.—Hammerstone.

It is not necessary to quote references to the well-known fact that the Eskimo and the Patagonians made use of round stones of various sizes as bolas. There is no evidence that our Indians ever used anything of the sort.

Grooved Stones Other Than Axes.

Three subclasses of grooved stones, differing in essential features from axes, may be discriminated. They are as follows:

Fig. 93.—Grooved round stone.

A. Slightly or not at all worked, except the groove; often showing marks of violent usage. With these may be classed the large stone hammers of the Lake Superior region.

B. Round or ellipsoid stones; in the latter the groove may follow either axis. The type (figure 93) is of sandstone from Carter county, Tennessee.

C. Resembling axes in all but the edge. Of class A there are none in the collection; their form and size are such that they could have been for no other purpose than hammerstones. Of class B there are some from Savannah, which may be sinkers or club heads. According to Morgan, oval stones with grooves were secured in the heads of war clubs,40 and Carver observed that the southwestern Indians used as a slung-shot a curiously worked stone, with a string a 96 yard and a half long tied to it, the other end being tied to the arm above the elbow.41

The specimens of class C may be broken axes. Figure 94 (granite, from Butler county, Ohio) shows a form quite common throughout central and western Ohio. They are generally small, have evidently never been sharp, and were in all probability intended for hammers from the beginning.

Mortars.

The Indian mortars in the collection are nearly always of sandstone of varying degrees of fineness. As is the case with cupped stones, when made of slabs, both sides have been worked; when of rough blocks, only one.

The Senecas and Cayugas are said by Morgan to have used wooden mortars in which to pound corn after it was hulled,42 and it is possible that the long pestles of soft stone were used with wooden mortars, though some are not well adapted to this use. The Iroquois women pounded in stone mortars the stony material used in tempering the clay for their pottery.43 The California Indians made mortars by knocking a segment off a bowlder, making a flat surface, and working out with a hammer and chisel,44 while the tribes of the interior worked directly from the surface of a suitable rock. The Yokuts, according to Powers, use tolerably well made stone mortars, and sometimes place a basket-like arrangement around the top to prevent the acorns from flying out.45

Fig. 94.—Grooved hammer.

No two specimens of the mortars and metate-like stones in the Bureau collection are alike; the nearest approach that can be made to a classification is as follows:

A. Smooth and flat on one or both sides; for use with mullers; from McMinn county, Tennessee, and Allamakee county, Iowa.

B. With round cavities on one or both sides; for round or cylindrical pestles; from McMinn county, Tennessee. A cobblestone from Bradley county, Tennessee, has a shallow cavity in either side and a pit in the center of each. From Kanawha valley there is a slab weighing about 25 pounds, flat and smooth on one side, as though primarily used with a muller and the regular even cavity afterward made; on the other side a cavity and a cupped hole have been worked in from the natural surface. A slab from Warren county, Ohio, has a shallow cavity worked into one side and a cupped hole in the other. From Union county, 97 Mississippi, there is a flattened bowlder with a shallow cavity on each side; a shallow cup has been pecked on the edge of one of them. From Caldwell county, North Carolina, comes a bowlder of water-worn mica-schist, with a shallow cavity and a deeper one on one side, and on the other a cupped hole opposite each of these cavities.

C. With one side hollowed out, the other flat and smooth. Specimens of this type come from Caldwell county, North Carolina; McMinn county, Tennessee, and Bradley county, Tennessee, the last with a pit in the center and another on the edge of the flat side.

D. With a long, narrow depression on each side. A very large specimen of fine-grained sandstone from Lincoln county, Arkansas, represents this type.

There are, in addition, two pieces of fine-grained sandstone with uniform thickness of less than an inch and about 10 inches across, from Kanawha valley, West Virginia, and Hale county, Alabama, respectively. Both sides are ground perfectly smooth, and flat. The objects were probably for some culinary purpose.

Sinkers.

The sinkers in the collection may be divided into four classes, viz: A, entirely unworked; B, notched on the sides; C, encircled by a groove; and D, perforated. Conversely, stones under all these different heads may have served other and widely different purposes.

Of the functions of class A, only those who have seen them in use can speak. Stevens mentions that some tribes inclose a round stone in a sort of net and attach it to a line in fishing;46 and no other use can be imagined for some of the specimens in the Bureau collection.

Specimens of class B are found along water courses in such situations as to leave no doubt of their use as sinkers;47 they were attached to grapevines and dragged on the bottom of streams to frighten fish into nets or traps.48 Those in the collection are made of ordinary flat water-worn pebbles, with notches rudely chipped in the sides; a number are from southeastern Tennessee.

Of class C, while many were perhaps sinkers, more were club heads and slungshots or hammers. A number have been obtained from Savannah, Georgia, more or less worked, some being rounded, with grooves of varying depths and sizes. Small stones of this form are used by Greenland fishermen as sinkers;49 and according to Thatcher, a large stone is by the Indians made fast to a sinking line at each end of a net, and the net is spread in the water by sinkers at different parts of it.50

Class D will be referred to under the head “Perforated stones,” from which they can be discriminated only arbitrarily. 98

A number of roughly chipped, somewhat crescent-shaped specimens of argillite, from half a pound to 2 pounds in weight, collected in Montgomery county, North Carolina, may have been used as sinkers.

Perforated Stones.

Only the larger or rougher perforated stones used as implements are included in this class.

Several perforated pieces of steatite, some mere rough fragments, others with the edges smooth and dressed to a somewhat symmetrical outline, have been collected about Savannah, Georgia. Some of these have been drilled, others gouged through apparently with a slender flint. In the latter group the little projections left by the tool have been worn smooth. The hole may be near one end or about the center. Similar pieces have been found in Forsyth county, Georgia; one of these is worked to an irregular pentagon and smoothly finished. From Haywood county, North Carolina, there are some very rough fragments, apparently just as they were picked up, except for the perforation; and a number of pieces of perforated pottery are from Montgomery county, North Carolina.

Perforated stones were used by the southern Indians to drag along the bottoms of streams and frighten fish into their nets and traps.51 Four disks 4 to 5½ inches in diameter, with handles from 13 to 17 inches long, were found in a cave at Los Angeles, California,52 and objects of this character were, according to Schumacher, used by the Santa Barbara Indians as weights for wooden spades.53 According to Abbott many perforated stones are found close to rivers and on shores in such positions as to leave no doubt of their use as sinkers.54 Similar stones were used as sinkers by the Scandinavians in comparatively recent times; by the Bechuanas for grinding grasshoppers, spiders, etc., and also as weights for digging-sticks; by some savages in the Pacific islands as clubs; by the Icelanders for breaking up salted fish.55 They were used by the Iroquois as weights for fire drills;56 by the Eskimo as clubs, having a rawhide handle secured by a knot.57 According to Dale,58 Layard,59 Griesbach,60 and Gooch,61 they were used by natives of southern Africa as root-diggers (to remove earth from the roots), as weapons, and to give weight to digging-sticks. They were also used by the Peruvian Indians to be thrown with a stick. Disk-shaped and 99 cylindrical throwing stones, perforated for the stick, are found among the Swiss lake dwellings.62 According to Evans63 they were used mostly as hammers or clubs. They are hard and battered on the edges; sinkers would be of softer stone.

The most complete article that has yet been given concerning the forms and uses of perforated stones is that by H. W. Henshaw.64

Discoidal Stones.

There are numerous references to discoidal stones by various writers, but a majority of the objects do not fall under any explanation that has so far been given.

The Choctaw Indians used disks two fingers wide and two spans around in playing “chungke,”65 and the Indians of North Carolina were much addicted to a sport called “chenco,” played with a staff and a bowl made with stone.66 The same kind of game was, or still is, played with hoops or rings of wood or rawhide by the Iroquois,67 the Pawnee,68 the Apache,69 the Navajo,70 the Mohave,71 and the Omaha;72 also, with rings of stone, by the Arikara,73 the Mandan,74 and other tribes.

The game of chungke, however, will account for only a small part of the great number of stones of this form. The Indians of southern California, in manufacturing pottery, make the clay compact and smooth by holding a rounded and smooth stone against the inside.75 The Fijians, in making pottery, use a small, round flat stone to shape the inside,76 while the Indians of Guiana use ancient axes or smooth stones for polishing the clay in making their vessels.77 According to Evans,78 pitted disks were used as pestles, hammers, or mullers; a thick one with pitted ends was found in a mortar at Holyhead.79 Under the head of pestles and of perforated stones further references will be found that may apply as well to this form of implements.

No kind of relic is more difficult to classify. From the smooth, symmetrical, highly-polished chungke stone they gradually merge into mullers, pestles, pitted stones, polishers, hammers,80 ornaments, and 100 the ordinary sinker or club-head, so that no dividing line is possible. Theories constructed on a basis of their use may be far from correct.

They present various forms and degrees of finish; many have the natural surface on both sides with the edge worked off by grinding or pecking, the latter being produced probably by use as a hammer; the sides may be ground down while the edge remains untouched; or the sides may be pecked and the edge ground, being probably of a thick pebble originally. Some of the finer grades, as chalcedony and quartz, that have received the highest finish, appear to have had all the work done by grinding or rubbing, as even those only slightly worked bear no signs of hammering or pecking. When of the harder materials they are generally made of water-worn pebbles as nearly the desired form as can be found; in fact, some specimens which are in their natural state, entirely unworked, require a very close examination to distinguish them from others whose whole surface has been artificially produced. In the jasper conglomerates from Arkansas, however, there is a regular series from a roughly chipped disk to one of the highest polish and symmetry. The larger ones of quartz, particularly those with concavities in the sides, must have been patiently wrought for years before brought to their present state. Many of the smaller ones, especially sandstone, seem to have been designed for grinding or polishing.

Fig. 95.—Discoidal stone.

The following groups are represented in the collection:

A. Sides hollowed out, edge convex; 2 to 6 inches diameter, seven-eighths to 2¾ thick.

1. Edges of concavity sharp.

a. Cavity a regular curve from side to side. The type (figure 95) is of quartz, from Cherokee county, Georgia. There are also, from Kanawha valley, West Virginia, one of sandstone, of which one side has been worked out by a flint, the little pits being distinctly visible, while the other side has natural surface; from Loudon county, Tennessee, one of 101 quartzite, 6 inches diameter, which has been used as a mortar, the cavities being roughened, with their edges broken and scarred (the edge of the stone is battered entirely around midway between the sides as though used for a hammer); from McMinn county, Tennessee, one of quartzite, about the same size as last, with a slight pit in the center of each cavity, the edges of the concavity being considerably chipped, and the edge of the implement very smooth; from Polk county, Tennessee, one of quartzite, 3½ inches in diameter, with the edge polished except in one spot, where it shows marks of use as a hammer or pestle—it has been used also as a mortar, the edges of the concavity being much chipped and broken; one each from Craighead county, Arkansas, of novaculite; Randolph county, Illinois, of granite; Cherokee county, Georgia, of quartz; and Obion county, Tennessee, of sandstone. In the four last mentioned the entire surface is quite smooth or even highly polished.

Fig. 96.—Discoidal stone, with perforation.

b. With a small perforation at the center. The type is shown in figures 96 (of sandstone, from a grave in Union county, Illinois), and 97 (of granite, from Virginia). There is another specimen, of sandstone, from Red River county, Texas.

Fig. 97.—Discoidal stone, with perforation.

c. With a secondary depression in each cavity. Figure 98 (yellow quartz, highly polished, from Fulton county, Georgia) is typical. There is also one of quartzite, with a secondary depression in one side only, from Roane county, Tennessee, which may be supposed, from this and other imperfections, to be unfinished.

2. Edges of concavity rubbed off blunt. These are grouped simply by form, as the specimens from Kanawha valley, West Virginia, and northeastern Kentucky are nearly all roughly finished, quite different 102 from the smooth or polished ones from farther south. Some are worked out into the form of a ring, and there is every stage between that form and the flat disk whose sides show no trace of pecking. Figure 99 (quartzite, from Sevier county, Tennessee) illustrates a typical example, roughly worked but entirely perforated, and figure 97 shows the same type in another form.

Fig. 98.—Discoidal stone, with secondary depression.

Fig. 99.—Discoidal stone, in form of a ring.

103

B. Flat or slightly concave sides, edges straight and at right angles to the sides; diameter, 1⅝ to 5 inches. The type shown in figure 100 is of sandstone from Lauderdale county, Alabama.

Fig. 100.—Discoidal stone.

Fig. 101.—Discoidal stone.

C. Sides flat; edges straight, sometimes rounding off into the sides; diameter, 2¼ to 6 inches; thickness, three-quarters to 2¼ inches. A number from southeastern Tennessee, especially the smaller ones, are quite rough, being merely pecked or chipped into shape with no subsequent 104 rubbing. Figure 101 (chalcedony, from a mound in Monroe county, Tennessee) represents the type. The material is variable.

D. Like the last, except much smaller. Very few are polished over the entire surface; some are rubbed more or less on the edges or sides, but a majority have the edge rough as it was chipped or pecked out; many have either the edge or sides in the natural state. From those smoothly polished to those very rudely worked the gradation is such that no dividing line can be drawn. This is true, also, of the smaller specimens of other types. Some of the quartzite specimens are very loose in texture. From seven-eighths to 2 inches in diameter and one-fourth to three-fourths of an inch thick.

Fig. 102.—Discoidal stone, convex.

E. Convex on both sides, edges straight. One of white quartz from Caldwell county, North Carolina, has the sides much curved, making the stone very thick in proportion to its width; there is a deep pit on each side, the entire surface being highly polished. Diameter, 2 to 3½ 105 inches; thickness, three-fourths to an inch and a half. Illustrated by figure 102 (of porphyry, from a grave in Caldwell county, North Carolina).

F. Same form as the above; 1¼ to 2 inches in diameter, one-half to seven-eighths of an inch thick.

Fig. 103.—Discoidal stone.

G. Flat or slightly convex on one or both sides, edge straight, one side wider than the other. Some have the edge battered or chipped and it is always at the angle of the edge with the wider side. From 1⅝ to 3½ inches in diameter, and three-fourths to an inch and a half 106 thick. The specimen shown in figure 103 (of compact quartzite, from Bartow county, Georgia) is typical. The material is quite diverse.

There are also of this type, one of very hard black stone (not identified) from Red River county, Texas, three-fourths of an inch in diameter; one of barite from Bartow county, Georgia, one inch in diameter, three-fourths inch thick; and one of granite, from Chester county, South Carolina, an inch in diameter. There are also one of quartzite from Drew county, Arkansas, with a shallow pit on each side; one of the same material from southeastern Tennessee, with a deep pit gouged in smaller side; and from the same locality, three of quartzite, one of quartz, and one of sandstone, each with a deep pit in the larger side. All of these are small and none of them polished.

Fig. 104.—Discoidal stone.

H. Convex sides and curved edges; size as in group G. The type (figure 104) is of quartz, from Caldwell county, North Carolina.

107

I. Same form, rough and not polished; 1 to 2¾ inches in diameter, one-half to 1 inch thick.

J. Sides slightly convex, edge slightly curved; 2¼ to 3½ inches in diameter, three-quarters to an inch and a half thick.

K. Sides flat; edges convex; roughly finished, no polish; 1⅛ to 2¼ inches in diameter, three-eighths to three-fourths of an inch thick.

L. Not polished; roughly chipped edges; 2 to 3½ inches in diameter.

M. Edges V-shape; 1¾ to 2½ inches diameter, 1 to 1½ inches thick. The type (figure 105) is of granite, from Randolph county, Illinois, with insunk pecked sides and polished edge. A specimen from Kanawha valley, West Virginia, is of flint, with only the edge worked; apparently 108 a hammer. One from Craighead county, Arkansas, has flat sides and the entire surface polished; another from McMinn county, Tennessee, is also polished entire. A good specimen from Cocke county, Tennessee, is of flint, one side rubbed flat, the other a rounded cone, highly polished.

Fig. 105.—Discoidal stone, with V-shaped edges.

N. Sides hollowed out; edges straight or slightly curved; very thick; used as mortars, hammers, or pestles. This form gradually merges into disk-shaped, pitted, or entire dressed hammers, which in turn run into the ordinary hammerstones. The types are figures 106 (quartzite, from Bradley county, Tennessee) and 107 (quartzite, from Nicholas county, Kentucky). There are in this group from eastern Tennessee three of quartzite, 2¼ by 4½ inches, 4¼ by 5¾ inches, and 1¾ by 3¼ inches, and one of granite, 2¾ by 3 inches; from Caldwell county, North Carolina, one of granite; and from Montgomery county, North Carolina, three of quartzite. The last four are evidently hammers or pestles. In addition there is a specimen from Jackson county, Illinois, of ferruginous sandstone, 3 inches in diameter. On one side there is a pit and on the other a shallow, mortar-like cavity extending entirely across.

Fig. 106.—Discoidal stone, used as mortar.

Fig. 107.—Discoidal stone, probably used as hammer.

O. One side flat, the other rounded; of convenient size for grasping. In some the bottom is quite smooth. There is sometimes a pit in one or both sides, more frequently in the bottom. They were used as mullers or pestles; in the latter, either the side or the edge may have been the pounding surface. The line between these implements and the cylindrical, dome-topped pestles can not be drawn (see figure 91).

109

P. Sides flat; edge convex; same size and use as last.

Q. From southeastern Tennessee and northwestern Georgia there are many disk-shape fragments of pottery, small, thin, and coarse, with the edges roughly chipped; and from northeastern Kentucky there are similar pieces, except that they have been fashioned from fragments of limestone and sandstone. These specimens are illustrated by figure 108 (pottery, from a mound in Bartow county, Georgia).

Fig. 108.—Discoidal pottery fragment.

Spuds.

It has been a puzzle to archeologists to assign to any class the peculiar stones called “spuds.” They are usually of a comparatively soft material, carefully worked and polished, and bear no marks of rough usage. On the other hand, they seem too large for ornament. Perhaps their office may have been in some ceremony or game. Something similar in form seems to be denoted in the following extracts:

Col. James Smith81 says, speaking of the Indians of western Pennsylvania, that as soon as the elm bark will strip in spring, the squaws, after finding a tree that will do, cut it down, and with a crooked stick, broad and sharp at the end, take the bark off the tree, and of this bark make vessels. The Twana Indians, who formerly lived at the south end of Hoods canal, Washington, in barking logs use a heavy iron implement about 3 feet long, widened and sharpened at the end;82 and the tanbark workers of our day use an instrument of somewhat similar form.

The ordinary spud is too weak to endure such usage, though it is claimed by old people living in the Shenandoah valley, Virginia, that in the last century the Indians in that locality used an implement of this pattern for stripping the bark from trees. The implement may have been used in dressing hides, the hole being for attachment of a handle. 110

Fig. 109.—Spud.

A celt of argillite, highly polished, from Loudon county, Tennessee, of the pattern shown in figure 64, has a neatly drilled cylindrical hole about a third of the way from the top; but such cases are unusual. The spuds may be divided into three general classes, as follows:

A. Blade circular in outline, including 180 degrees or more, or semielliptical with either axis transverse; sides of stem straight or slightly curved, parallel or slightly tapering to top, which is either straight or slightly rounded; shoulder nearly at right angles to stem, with sharp or rounded corners or sometimes barbed; stem and blade not differing greatly in length. The type of the class, presented in figure 109, is of clay slate, from a mound in Monroe county, Tennessee. The other six specimens in the collection were distributed as shown in the table.

B. Lower part of the blade a half circle or less; top square or slightly rounded; stem rapidly widening, with increasing curve to the blade, making an angle with it; stem and blade nearly the same length. A specimen of green slate, from Mississippi county, Arkansas, is illustrated in figure 110. Another, of compact quartzite, comes from Loudon county, Tennessee.

Fig. 110.—Spud.

C. Handle or stem round; very much longer than the blade, which is semicircular or semielliptical, with square or barbed shoulders. Illustrated in figure 111 (probably of chloritic slate, from Prairie county, Arkansas).

Plummets.

The specimens known as plummets vary considerably in form, size, and degree of finish, indicating diversity of purpose, and different writers have assigned to them various uses. 111

According to Abbott, one of these relics was found at Salem, in a mortar.83 Stevens says, quoting from Schoolcraft, that the Pennacook Indians used sinkers very much like a plummet in shape.84 In Florida very rough plummets with deep grooves are found in the shell mounds, which were no doubt used as sinkers. The Indians of southern California use them as medicine stones to bring rain; the Eskimo use similar stones as sinkers, but have them perforated at the end. The larger objects of this form may have been used as pestles.85 They might be made very efficient in twisting thread, as they revolve for a considerable time when set in motion.

Fig. 111.—Spud.

The general form is ovoid, sometimes quite slender, sometimes almost round; the ends may be either blunt or pointed. They may be grooved near the middle or near either the larger or smaller end. Some have two grooves, some are only partially grooved, while others have the groove extending lengthwise. There are forms that differ somewhat from this description, but such are rare.

Fig. 112.—Plummet, grooved near one end.

Fig. 113.—Plummet, double-grooved.

Many small and otherwise unworked waterworn pebbles and pieces of steatite pots from southeastern Tennessee and from Montgomery county, North Carolina, have grooves near the middle or near one end; they were probably applied to some of the uses for which plummets were intended.

The plummets in the Bureau collection may be grouped as follows:

A. Grooved near smaller end. The types are illustrated in figure 112 (sandy limestone, from a mound in Catahoula parish, Louisiana), and figure 113 (hematite, double grooved, with notches cut in various places, from a mound in Kanawha valley, West Virginia). Other specimens are, one from Arkansas county, Arkansas, of sandstone, and one each from Brown and Randolph counties, Illinois, both of hematite.

B. Grooved near larger end. A good example, of hematite, is from Kanawha valley, West Virginia, with a second groove partially around the middle. 112

C. Grooved near the middle. The class is represented by a beautiful specimen (figure 114) of hematite, with the groove much polished and irregular, and a deep notch cut in one end, from Ross county, Ohio. Another specimen, from Kanawha valley, West Virginia, is a double conical implement of hematite, elliptical in section with both ends ground off on flatter sides only.

Fig. 114.—Plummet, grooved near middle.

Fig. 115.—Plummet, grooved lengthwise.

D. Grooved lengthwise. This class includes a plummet of quartzite, from Yellowstone park (figure 115), and another of hematite, much shorter than the Yellowstone specimen and with blunt ends, from Kanawha valley, West Virginia.

Fig. 116.—Plummet, grooveless, perforated.

Fig. 117.—Plummet, double cone in shape.

E. Grooveless. A good specimen (figure 116) is of quartz and mica, elliptical in section, pointed at ends with one end perforated, from Yellowstone park; another, from Randolph county, Illinois, of hematite, rough, perhaps unfinished.

F. Double cone, with one end ground off flat and hollowed out. The type (figure 117) is of granite, one of three from Savannah, Georgia.

G. Top flattened and hollowed out; sides incurving to the middle; lower half a hemisphere. The class is represented by figure 118 (quartzite, from Randolph county, Illinois), and figure 119 (sandstone, from Adams county, Ohio). From Kanawha valley there is one of hematite, similar in form to the last.

Fig. 118.—Plummet.

H. Ovoid, with the smaller end ground off flat.86 A good specimen of this class (figure 120) is of magnetite, from Caldwell county, North Carolina. From Savannah, Georgia, there are two of sandstone, both smaller than the type and rough; from Kanawha valley there is one of quartzite, nearly half ground away, leaving almost a hemisphere; and from eastern Tennessee there are one of magnetite and one of quartzite, the latter nearly round.

I. Cylindrical. A unique specimen, from a mound in Loudon county, Tennessee, is illustrated in figure 121. It is of sandstone; a short cylinder with incurved sides, each end terminating in a blunt cone. 113

Fig. 119.—Plummet.

Fig. 120.—Plummet, end ground flat.

Fig. 121.—Plummet.

Figure 122 represents a piece of smoothly dressed steatite from Desha county, Arkansas, with a two-thirds round section, the ends rounded, with a groove near one end, which may be classed with the plummets. There are pieces of sandstone from the same locality which connect this pattern with the simpler “boat-form” stones, except that the flat side is ground smooth instead of being hollowed. This is only one of numerous examples where the shapes of implements whose “typical forms” seem utterly dissimilar merge into one another so gradually that no line of demarkation can be drawn.

Fig. 122.—Plummet, cylindrical.

Cones.

Fig. 123.—Cone.

Fig. 124.—Cone.

The relics known as “cones” have the base flat and the side curving slightly; usually the curve extends regularly over the top, but sometimes the apex is rubbed off flat. The conic surface may form an angle with the base, or the line of junction may be rounded into a curve. They vary considerably in thickness, some being nearly flat, others having a height equal to the diameter of the base. One of steatite from Savannah, as also one of sandstone from Kanawha valley, has a slight pit or depression on the flat side. Among the best examples are 114 one (figure 123) of steatite from Bradley county, Tennessee, and another (figure 124) of hematite from Loudon county, in the same state; one (figure 125) of compact quartzite from a mound in Ogle county, Illinois, and a fourth specimen (figure 126) of granite from Kanawha valley, West Virginia. The distribution is as follows:

Fig. 125.—Cone.

Fig. 126.—Cone.

Hemispheres.

Hemispheric stones, like the cones, can receive a name only from the form and not from any known or imagined use to which they could have been applied.

All such specimens in the collection, except one, are from Kanawha valley, and of hematite; many if not most of them have been ground down from the nodule, and were probably paint stones originally; at least, the material rubbed from them was used as paint while the maker had their final form in view. One, however, has been pecked into shape and is entirely without polish. In all, the base is flat and varies in outline from almost a circle to a narrow ellipse. A section of the stone parallel to either axis of the base varies from a little more to a little less than a semicircle. Typical forms, both from Bracken county, Kentucky, are illustrated in figure 127.

Fig. 127.—Hemispheres.

115

The specimen, illustrated in figure 128 (yellow quartz, from a mound in Kanawha valley) is intermediate between cones and hemispheres. The sides are polished, while the flat bottom and rounded top are roughened. As it has faint red stains, it may have been used as a paint-muller.

Fig. 128.—Hemisphere.

Paint Stones.

The articles known as paint stones scarcely come under the head of implements. Some of the hematite pieces are incipient celts, hemispheres, or cones; but most of them were used merely to furnish paint, at any rate until rubbed down quite small. They are of every degree of firmness, some being as brittle as dry clay, others like iron. Most pieces in the collection are from Kanawha valley, but others are from southeastern Tennessee, northeastern Arkansas, and Caldwell county, North Carolina. From the last-named section, as well as from Chester county, South Carolina, and McMinn county, Tennessee, come pieces of graphite more or less rubbed; and one has been sent in from Elmore county, Alabama.

Fig. 129.—Paint stone.

The specimen illustrated in figure 129, from a mound, is a good example of the manner in which the harder hematite was ground.

Ceremonial Stones.

Functions and Purposes.

The so-called “ceremonial stones” are variously subdivided and named by different writers. They are supposed to have been devoted to religious, superstitious, medical, emblematic, or ceremonial purposes; to be badges of authority, insignia of rank, tokens of valorous deeds, or perhaps some sort of heraldic device; in short, the uses to which they might, in their different forms, be assigned, are limited only by the imagination.

According to Nilsson the ancient Scandinavians wore “victory stones” suspended around their necks,87 and the Eskimo wear charms and amulets to bring success in fishing and hunting.88 Adair (1775) says that the American Archi-magus wore a breastplate made of a white conch-shell, with two holes bored in the middle of it, through which he put the ends of an otter-skin strap and fastened a buck-horn button to the outside of each.89 An explanation of the purpose of many of the smaller perforated stones also may be found in Nilsson’s remark90 that the small ovoid or ellipsoid ones were used as buttons; a string being tied to the robe at one end, run through the hole and tied in a knot. 116

The various Indians of Guiana in their leisure hours often fashion highly ornamental weapons and implements which they never use except ceremonially, but keep proudly at home for show.91

So, too, the Yurok and Hupa Indians of California, as well as some of the tribes of Oregon, have very large spearheads or knives, which are not designed for use, but only to be produced on the occasion of a great dance. The larger weapons are wrapped in skin to protect the hand; the smaller ones are glued to a handle. Some are said to be 15 inches long.92 The Oregon Indians believed the possession of a large obsidian knife brought long life and prosperity to the tribe owning it.93

Some of the wild tribes of the interior have something which they regard as the Jews did the Ark of the Covenant. Sometimes it is known; again it is kept secret. The Cheyenne had a bundle of arrows; the Ute a little stone image, and the Osage a similar stone.94 The Kiowa had a carved wooden image, representing a human face; the Ute captured it, and the Kiowa offered very great rewards for its return; but the Ute, believing the Kiowa powerless to harm them so long as it was retained, refused to give it up.95

The North Carolina Indians, when they went to war, carried with them their idol, of which they told incredible stories and asked counsel;96 and as a token of rank or authority, the Virginia Indians suspended on their breasts, by a string of beads about their neck, a square plate of copper.97 These were worn as badges of authority. The native tribes, from our first acquaintance with them, evinced a fondness for insignia of this kind.98

Simply for convenience the ceremonial stones in the Bureau collection will here be divided into two general classes. The first, comprising those pierced through the shortest diameter, will be called gorgets, which name, like that of celt, has no particular meaning, but is in common use. The second class will comprise all others, which will have some name that may or may not be suitable to their form, but by which they are usually called. In this class are included boat-shape stones, banner stones, picks, spool-shape ornaments, and bird-shape stones, as well as engraved tablets or stones.99

Gorgets.

The relics commonly called gorgets have been found in Europe; they may be convex on one side, concave on the other, and are supposed to 117 be for bracers.100 It is said that the Miami Indians wore similar plates of stone to protect their wrists from the bowstring.101 Herndon and Gibbon remark that a gold ornament in shape like a gorget, but not pierced, is worn on the forehead by some of the Amazon Indians.102 According to Schoolcraft the so-called gorgets were sometimes used as twine-twisters;103 but Abbott holds that while some may have been twine-twisters, or may have been used for condensing sinews or evening bowstrings (that is, reducing the strings to a uniform diameter), most were simply ornaments, as they are generally found on the breast of a buried body.104 Stevens is even more conservative, holding that they were neither twine-twisters nor devices for condensing sinews or evening bowstrings, as they show no marks of wear in the holes.105

Some writers suppose the gorgets to have been shuttles; but this supposition can hardly be entertained, although it is true, according to Chase, that the Oregon Indians passed thread with a curved bone needle.106 As twine-twisters they would be about as awkward as anything that could be devised. As to evening bowstrings, it would seem that if a string were too large in places to pass through a hole it could not be pulled through; pounding and rolling the wet string with a smooth stone, or some such means, would be the remedy. The bracer theory is plausible; but no one seems ever to have seen a gorget used for this purpose.

Few of the gorgets in the Bureau collection show such marks of wear around the edges of the hole as would be made by a cord; but the majority are thus worn at the middle, where the hole is smallest. Some specimens among every lot are not perforated, or only partially so; the drilling seems to have been the last stage of the work. The hole is almost always drilled from both sides, and the few in which it goes entirely through from one side would probably have had it enlarged later from the other. A number are fragments of larger gorgets, the pieces having been redrilled.

Some of the specimens have various notches and incised lines, the latter being sometimes in tolerably regular order; but there is not the slightest indication that these marks had any meaning or were intended for any other purpose than to add to the ornamental appearance of the stone.

If they were to be worn at the belt or on any part of the dress they could easily have been fastened by a knotted string, or if the wearer desired he could have an ornamental button of some kind. If suspended around the neck, in order to make them lie flat against the breast they probably had a short cord passed through the perforation and tied 118 above the top of the object, the suspending cord being passed through the loop thus formed.

Fig. 130.—Gorget.

The principal division is into group A with one hole and group B with two holes, though in many cases this forms the only difference between two specimens.

A. General outline rectangular, or perhaps slightly elliptical, sometimes with one end somewhat narrower than the other, or with one end rounded off, or with the corners slightly rounded. Perforation commonly near one end. The form is represented by the specimen with two perforations illustrated in figure 133, which otherwise fully answers the description. The argillite specimens have the broader ends striated as though used for rubbing or scraping, but in other respects conform to those of other materials. The materials are generally the softer rocks, as shown in the accompanying table:

A related type is rectangular or with incurved sides (forming either a regular or broken curve) and rounded ends, and differs in having the perforation near the center. The same pattern sometimes has two holes. It is illustrated in figure 130 (striped slate, from a mound in Kanawha valley, West Virginia). There are also from the same place one each of slate, cannel coal, and clay slate, and from eastern Tennessee one each of slate, shale, and clay slate.

Fig. 131.—Gorget(?).

There are a number of small pebbles, thin and flat, with a hole drilled near the edge, from southeastern Tennessee, North Carolina, and southeastern Arkansas. One of these, from Caldwell county, North Carolina, is of banded slate; the others are of clay slate or sandstone. Two of them have straight and zigzag lines on both faces, and notches around the edge.

Allied to these are a number of pieces of flat stone from southeastern Tennessee, Kanawha valley, and North Carolina, with the faces partially rubbed down smooth, the edges being untouched. They are of slate, talc, or argillite.

From southeastern Tennessee and North Carolina there are several pieces of steatite, which may have been for sinkers. Some have a hole 119 near one end, others a hole at each end, while still others are not perforated. All have been worked over the entire surface, and some of them are well polished. One of these is represented in figure 131.

B. Gorgets with two holes. Of these there are several subdivisions, differing more or less widely in form. They are as follows:

1. Thick, with both the sides and the ends incurved or reel-shape; faces flat or slightly convex. This form is represented by the specimen shown in figure 132, from a mound, Knox county, Ohio. There is another from the same place, a third from Kanawha valley, and a fourth from Butler county, Ohio; all of green slate.

Fig. 132.—Gorget, reel-shape.

2. Rectangular, or with sides or ends, or both, slightly curved, either convex or concave; faces flat. Shown in figure 133 (green slate, from a grave in Kanawha valley, West Virginia).

Fig. 133.—Gorget.

3. Widest at middle, with single or double curve from end to end; very thin; both sides flat.

120

4. Same outline but thicker; one face flat, the other convex. Represented by figure 134 (shale, from Jackson county, Illinois). The distribution of the form is as follows:

Fig. 134.—Gorget.

5. Same outline, but quite thick, approaching the “boat-shape” stones in form. In some the flat side is slightly hollowed out. A majority of them are not perforated. The type (figure 135) is of sandstone, from a mound at Adelphi, Ohio.

There are also, from Butler county, Ohio, Kanawha valley, West Virginia, and Savannah, Georgia, one each of slate; from Ross county, Ohio, two, and from Kanawha valley, and Cocke county, Tennessee, one each, all of sandstone. There are two (of sandstone and slate) from Kanawha valley, which differ from the others in having the sides parallel, giving them a semicylindrical form.

The pattern of the specimen illustrated in figure 136 (striped slate, from Butler county, Ohio, of which a number have been found in that state), may be classed between the gorgets and the boat-shape stones. The shorter end of the object has, sometimes, a projection or enlargement at the top, apparently for suspension, although no perforated examples have been found.

Banner Stones.

Under the head of “banner stones” are placed ornaments having the ends at right angles to the perforation. The hole is drilled in a midrib, from which the faces slope by either straight or curved lines to the edges. The two halves of the stone are symmetrical. In most specimens one face is flatter than the other, even plane in some cases. Some specimens are finished to a high 121 polish, before the hole is started; others have the hole completed with the exterior more or less unfinished. The specimens in the Bureau collection may be classified as follows:

A. Rectangular or trapezoidal, with sides and ends sometimes slightly curved inward or outward.

B. Reel-shape.

C. Crescentic.

D. Butterfly pattern.

Fig. 135.—Gorget, boat-shape.

Fig. 136.—Gorget resembling boat-shape stone.

The last three varieties may be considered as only modifications of the simple rectangular banner stones. By rounding off the corners of the articles or dressing them to sharp points, by cutting away portions from the sides or by trimming away the central portions at either or both ends of the perforations, all these different forms may be produced.

Fig. 137.—Banner stone.

Fig. 138.—Banner stone.

A. A typical specimen is illustrated in figure 137. It is of slate, and was taken from a mound in Kanawha valley, West Virginia. Another good example, shown in figure 138, is of sandy slate, from a grave in Monroe county, Tennessee. The geographic range of this type is wide, though the objects are not abundant. 122

B. The reel-shape banner stones are somewhat variable, but are fairly illustrated in figure 139, representing a specimen of argillite from Sevier county, Tennessee.

A related form has the middle cut from one end, leaving two horn-like projections extending parallel with the hole. An example of this form, shown in figure 140, is of banded slate, from a mound in Kanawha valley, West Virginia.

Fig. 139.—Banner stone, reel-shape.

Fig. 140.—Banner stone, with horn-like projections.

Fig. 141.—Banner stone, crescent-shape.

Fig. 142.—Banner stone, crescent-shape

C. The crescentic banner stones might better be termed “semilunar,” since most of them are flat at one end and curved at the other. Occasionally one has both ends curved and parallel, the sides also slightly curved, making the article reniform. Others have the ends straight and parallel, with the sides curved or like the zone of a circle. Two have a midrib for the hole, with the sides dressed down quite thin, as with the butterfly gorgets. All were finished in form before the drilling was done, though some had not received their final polish. The type is illustrated in figures 141 (steatite, from northwestern North Carolina), 142 (pagodite, from Rhea county, Tennessee), and 143 (sandstone, 123 from Jefferson county, Tennessee). The last form is sometimes called a perforated ax, but the material and fragile make exclude it from every class except the ceremonial stones.

D. The “butterfly” gorgets are so named from their resemblance to a butterfly with expanded wings. The sides or wings are usually quite thin, either semicircular or like a spherical triangle in outline. The perforated mid-rib is shorter than the wings and carefully worked. A good example, shown in figure 144, is of ferruginous quartz from Monongahela, Pennsylvania, and that illustrated in figure 145 is of banded slate from Kanawha valley. There is also one of the latter material from Lewis county, Kentucky.

Fig. 143.—Banner stone, crescent-shape.

Fig. 144.—Butterfly banner stone.

Fig. 145.—Butterfly banner stone.

Fig. 146.—Banner stone.

An aberrant form is elliptical in section at the middle, round or nearly so at the ends, the sides expanding rapidly from end to middle by 124 double curves. It is represented by figure 146 (ferruginous quartz, from Kanawha valley, West Virginia), and by a specimen of quartzite from Union county, Mississippi.

Boat-shape Stones.

There are two types of relics, perhaps ceremonial, for which no use has been determined, and which are named from their general resemblance to the form of a boat. They are as follows:107

A. With flat face more or less hollowed, sides triangular and parallel. A number are not perforated. The type is shown in figure 147 (striped slate, from Davidson county, North Carolina).

Fig. 147.—Boat-shape stone.

Fig. 148.—Boat-shape stone.

B. Coming to a point at each end; flat side, deeply hollowed; perforations near the ends, with a groove between them in which the suspending cord rested. Some have a flattened projection in which the groove is made. The type (figure 148) is of steatite, from a grave in Sullivan county, Tennessee. The distribution is as follows: 125

Picks.

Fig. 149.—Pendant.

Fig. 150.—Pick.

The relics known as picks from their form and not at all from their function vary considerably in size. Not all are perforated. A good example, shown in figure 150, is of striped slate, from Knox county, Ohio. There are also in the collection, from Union county, Mississippi, one specimen of greenstone; from Jackson county, North Carolina, one of slate, and from Montgomery county, North Carolina, one each of steatite and slate. The last named is the half of a larger one that was broken at the part drilled, and has had a hole drilled near the larger end of this fragment, which has not been reworked.

Spool-shape Ornaments.

Fig. 151.—Spool-shape ornament.

Relics of spool shape, probably ornamental rather than industrially useful, are not uncommon in copper, though very rare in stone. The specimen shown in figure 151 is of sandstone, from Jackson county, Arkansas. There are also, from Prairie and Lonoke counties, one each of sandstone, and from Jackson county two of the same material; from Clark county there is one of pinkish slate, with the stem drilled between and parallel to the faces, the others with stems drilled lengthwise.

Bird-shape Stones.

Stone relics of bird form are quite common north of the Ohio river, but are exceedingly rare south of that stream. A good example, shown 126 in figure 152, is of granite, from Vernon county, Wisconsin, and the collection embraces another specimen, of sandstone, from Kanawha valley, West Virginia.

Fig. 152.—Bird-shape stone.

According to Gillman, bird-shape stones were worn on the head by the Indian women, but only after marriage.108 Abbott109 quotes Col. Charles Whittlesey to the effect that they were worn by Indian women to denote pregnancy, and from William Penn that when squaws were ready to marry they wore something on their heads to indicate the fact. Jones110 quotes from De Bry that the conjurers among the Virginia Indians wore a small, black bird above one of their ears as a badge of their office.

Shaft Rubbers.

The shaft of an arrow is straightened by wetting and immersing it in hot sand and ashes, and bringing into shape by the hand and eye. To reduce the short crooks and knobs it is drawn between two rough grit stones, each of which has a slight groove in it; coarse sand is also used to increase the friction.111

Again, a rock has a groove cut into it as wide as the shaft and two or three times as deep. Into this the crooked part of the shaft is forced, and by heating or steaming becomes flexible and can be easily made straight, which shape it will retain when dry.112

A somewhat different device for the same purpose appears in the Bureau collection. It is illustrated in figure 153 (of fine sandstone); there was another part to correspond with that shown. The specimen is from Monongahela, Pennsylvania.

Tubes.

As the use of stone tubes by the Indians has given rise to considerable discussion, the following references to the various ways in which they have been employed may help to settle it. 127

Schoolcraft observed that the Dakota Indians used a horn tube in bleeding; one end was set over the cut, and the other vigorously sucked.113 Powers says that the Klamath Indians use tubes for smoking,114 while H. H. Bancroft says that the Acaxees of Mexico employ “blowing through a hollow tube” for the cure of disease,115 and also that the Indians of southern California inhale smoke of certain herbs through a tube to produce intoxication.116 According to C. C. Jones the Florida and Virginia Indians used reeds in treating diseases by sucking or blowing through them, and also used them in cauterizing; and he observes that the Indians of Lower California employed similar processes, using stone tubes117 instead of reeds. Hoffman illustrates the removal of disease through the agency of a tube of bone by a Jĕs´sakīd´ or medicine-man of the Ojibwa.118 Read calls attention to the fact that the old Spanish writers describe a forked wooden tube, the prongs being inserted in the nostrils, while the other end was held over smoldering herbs, and suggests that the Indians may have used stone tubes in the same way.119

Fig. 153.—Shaft rubber.

The Indian mode of inhaling smoke would produce the same result, whether drawn through the mouth or into the nostrils.

The use of stone tubes for astronomical purposes, which has been discovered by some imaginative writers, is, of course, absurd; nevertheless they are useful in viewing distant objects on a bright day, especially when looking toward the sun.

Nearly all of the tubes made of soft material with tapering perforation seem to have been gouged rather than drilled. Schumacher observes that the California Indians drilled their tubes from both ends and enlarged the hole from one end by scraping, the mouthpiece being made of a bird bone stuck on with asphaltum.120

There are five classes of stone tubes in the collection of the Bureau, as follows:

A. One end flattened and expanding into a wing on either side. This class is illustrated by figure 154 (from Kanawha valley, West Virginia). The corners of this specimen have been trimmed off; the typical form is indicated by the dotted lines. There are also from the 128 same locality one of quartzite, and from Ross county, Ohio, one of sandstone.

B. Conical; the bore more tapering than the exterior. Represented by the specimen shown in figure 155, of sandstone, from a mound in Kanawha valley, West Virginia.

C. Hour-glass shape, usually but not always with a narrow ring or projection around the smallest part. Exterior with gently curving outlines; the perforation is usually in the form of a double cone, with the points at the smallest part of the tube, which may or may not be midway between the ends. A good specimen, illustrated in figure 156, is of steatite, from Sevier county, Tennessee.

D. Of nearly uniform diameter inside and out; section circular, elliptical, or flattened on one side. This form is exemplified by figure 157, a specimen from North Carolina. There are also one each from Caldwell, Haywood, and Montgomery counties, North Carolina, all of slate.

Fig. 154.—Tube, one end flattened.

Fig. 155.—Tube, conical.

E. Round or elliptical in section, ¾ to 2½ inches long; probably beads. The collection includes specimens from Bradley county, Tennessee, of steatite; from Savannah, Georgia, of ferruginous sandstone; and from Union county, Mississippi, of jasper.

Pipes.

So much has been written concerning pipes that few references seem necessary, and none will be given except from Col. R. I. Dodge, who, after an experience of many years among the Plains Indians, says that the latter have different pipes for different occasions, as the medicine 129 pipe, peace pipe, council pipe, and a pipe for common use. Each is sacred to its own purpose.121

Fig. 156.—Tube, hour-glass form.

In an article so highly prized by its owner, great pains would be expended to give an ornamental appearance to one which would be used on important ceremonial occasions; and it would be carved or worked in a manner gratifying to its maker or the one for whom it was intended. This fact, and the statement quoted above, will explain the great variety in form from a limited area. Still, in some sections of the country there are certain types that prevail, and may be in some cases peculiar to these localities; such, for instance, are the long stemmed pipes from western North Carolina and eastern Tennessee.

In many pipes of soft stone the bowl is gouged out instead of drilled.

Fig. 157.—Tube, cylindrical.

The pipes in the Bureau collection embrace the following classes:

A. Stem with an elliptical or somewhat triangular section; the bowl near one end, leaving a projection in front; stem hole in long end. The form is shown in figure 158. From Caldwell county, North Carolina there are two similar pipes of steatite. Another, from Preston county, West Virginia, differs only in having the stem hole in the short end.

Fig. 158.—Pipe, flat base.

B. Same form of stem; no projection in front, the bottom of the stem curving up gradually into the front of the bowl. This type is represented by figure 159 (of steatite, from a mound in Loudon county, Tennessee). There are also, from Kanawha valley, West Virginia, an example of talcose slate, and from Caldwell county, North Carolina, one of steatite.

C. Stem having a midrib in which the hole is bored. One of steatite, from Caldwell county, North Carolina, has a prow; the others have not. Another of steatite from Loudon county, Tennessee, has a slender projection below the bowl, as if for a handle. The axis of the 130 bowl and that of the stem meet at any angle between 100° and 170°. Figure 160 represents a typical specimen, of steatite, from a mound in Sullivan county, Tennessee. There are also, from Caldwell county, North Carolina, and Kanawha and Preston counties, West Virginia, one each, and from Sullivan county, Tennessee, two, all of steatite; and there is an example from Kanawha valley, West Virginia, of material not identified.

Fig. 159.—Pipe.

Fig. 160.—Pipe.

D. With bowls and stems either round or square; very large. A good example (figure 161) is of red sandstone, from southeastern Missouri; it is the only pipe in the entire collection of the Bureau on which is shown any attempt at ornamentation. From Jefferson county, Tennessee, and Savannah, Georgia, there are one each, of steatite.

Fig. 161.—Pipe, ornamented.

Fig. 162.—Pipe.

E. Cylindrical bowl, with a square-edged groove around it near the middle, below which the bottom has a somewhat celt like form, with stem hole in one side. A small hole is drilled near the edge at the 131 bottom, probably for the purpose of suspending feathers or other ornaments. The type is represented by figure 162 (of limestone, from Crawford county, Wisconsin). Pipes of the same form are found also in central Ohio.

Fig. 163.—Pipe, long-stemmed.

F. Round stem from one-half inch to 10 inches long; bowl at extreme end, set on at various angles from nearly a right angle to almost a straight line. Good examples are illustrated in figure 163 (steatite, from Caldwell county, North Carolina) and 164 (also of steatite, from a mound in Monroe county, Tennessee). The other specimens in the collection are distributed as shown in the table:

Fig. 164.—Pipe, short-stemmed.

G. Same form of stem, short, with flange around the top of the bowl. Represented by one of sandstone, from a mound in Monroe county, Tennessee (figure 165), and three of sandstone and two of marble from eastern Tennessee.

Fig. 165.—Pipe.

Fig. 166.—Pipe.

H. Small, stem more or less squared, bowl upright. There are two examples of this class from Monroe county, Tennessee, each having a flat projection or ridge on top of the stem, which is perforated for attachment of ornaments. The type, represented in figure 132 166, is of clay slate, from Monroe county, Tennessee. It will appear from the following table that the distribution of this form is limited:

I. Egg-shape bowl, stem hole in the side. One from Bradley county, Tennessee, of argillaceous limestone, has a hole drilled from end to end, but no stem hole. It may have been made so intentionally, or the drilling may have been carried too far and the specimen left unfinished. The type is of barite, from Sevier county, Tennessee (shown in figure 167). Another specimen, from McMinn county, Tennessee, is of argillaceous limestone.

Fig. 167.—Pipe.

J. Form like last, with a flange around the top of the bowl. A typical specimen, shown in figure 168, is of steatite, from Loudon county, Tennessee. There are, also, from Preston county, West Virginia, one of sandstone, and from Caldwell county, North Carolina, two of steatite.

K. Bowls egg-shape, but quite long and sometimes rather pointed at the bottom; stem hole in the side. This class includes the following: From Savannah, Georgia; Roane county, Tennessee; and Adams county, Ohio, one each of sandstone; from Holt county, Missouri, one of micaceous sandstone; from Kanawha valley, West Virginia, one of indurated red clay, possibly catlinite; and from Caldwell county, North Carolina, three of steatite.

Fig. 168.—Pipe.

Chipped Stone Articles.

Materials and Manufacture.

The chipped implements in the Bureau collection, are nearly always made of some form of flint or similar chalcedonic rock, as it is easily chipped and can be brought to a keen edge or point. Sometimes quartz, quartzite, argillite, or even a more granular rock is used; but this is infrequent, and is due to the scarcity of the more desirable material. 133

In the spades and hoes first to be considered the flaking seems to have been by percussion mainly, if not entirely; the same method appears to have been employed in obtaining flakes from blocks, to work into the smaller implements. Some of the processes used in making them will be hereinafter described.

Spades.

It must be admitted that most Indians depended largely on agriculture for subsistence; some historical works that represent them as barbarous hunters, depending entirely on the chase, will, on the same page perhaps, relate how Virginia and New England pioneers were saved from starvation by supplies of corn, beans, and pumpkins obtained from the Indians. This being the case, some method of cultivation was necessary.

It is not to be inferred that “cultivation” implies all that is now meant by the term; the Indian seems merely to have worked the hill in which his corn was planted and not the whole surface of the field, a shallow hole being scooped out in which the grain was dropped, and as the stalk became larger the dirt was heaped up around it. The remains of many “Indian old fields” in various parts of the country show this, there being no long ridges as in cornfields of the present day, but only a great number of these detached hills. The great scarcity of implements suitable for such work argues nothing, for in most parts of the country stone easily worked and adapted to the purpose is unobtainable.

There are a few flint deposits found in southern Illinois in which the material occurs in nodules that can be made with even less work than a piece of wood into suitable implements; and in the country which may be considered as belonging to this archeologic district the flint hoes and spades are tolerably abundant. In other portions of the country, wood, the shoulder blades of large animals, and musselshells perforated for attachment to a handle, were formerly used; the shells are frequently found, but the other materials have long since disappeared.

Early observations on the industries of the aborigines are significant. Thus, according to De Forest, the Connecticut Indians used spades rudely constructed of wood, or of a large shell fastened to a wooden handle;122 and Palmer123 figures a hoe made of horn, 14 by 5 by one-fourth inches, in a wooden handle 5 feet long, which is split and slipped over the smaller end; such, with others of wood and stone, were used among the Utah Indians before iron was introduced. Dawson holds that they were probably prepared in large numbers for the planting time, when the whole tribe mustered to till the fields, and that when the work was over they were gathered and hidden in some safe place until the next 134 season.124 This may have been the case to some extent, but the specimens found in these hiding places seldom have marks of use, and it is more probable that they were the property either of persons living at a distance or of an individual manufacturer in some particular village, being thus concealed for safe-keeping until there was a demand for them or, perhaps, to await a convenient time for transportation. A sedentary tribe would have no more reason for hiding this than any other kind of property.

Fig. 169.—Chipped spade with pointed ends.

Fig. 170.—Chipped spade with rounded ends.

The chipped implements known as spades are frequently found buried in large numbers. Two caches were disclosed by high water in 1884, near Caseyville, Kentucky, containing, respectively, 57 and 75 specimens from 6 to 13 inches long.

The most common form is that having an oval or elliptical outline, with the ends either coming to a point or rounded. Long use of those having pointed ends would wear them off until they approached the others in form; but so many of both patterns show no evidence of use 135 that this distinction must be considered intentional. The principal varieties are as follows:

A. Those with pointed ends. Figure 169 represents a typical specimen of yellow flint, from Union county, Illinois.

B. Those with the ends rounded. Represented by figure 170 (yellow flint, from Union county, Illinois).

A specimen from Jackson county, Illinois, has had a portion of the edge broken squarely. The polish over this fractured surface shows that it was long used after breaking without being rechipped to a sharp edge. This indicates usage only in loose ground, as it evidently would be quite difficult to force the square, broken part into a hard soil or tough sod.

The specimens from Polk county, Tennessee, are pecked or chipped, or both, and are quite roughly made. They are neither scratched nor polished, and may be unfinished implements of some other class, though agreeing closely with the flint spades in shape and size.

C. A modification of the last form has the upper portion chipped away along the sides until it is ovoid, with a blunt point, leaving the lower part a regular curve. An example, shown in figure 171, is of grayish brown flint, from Scott county, Missouri. There are also one each from Mississippi county, Missouri, and Hopkins county, Kentucky, of the same material.

D. Like the above, but much shorter in ratio to the width, and with a flatter curve. The type, figure 172, is of yellow flint, from a mound in Obion county, Tennessee. There are also three from Union county, Illinois, one of them with almost the same dimensions. 136

E. Semicircular outline, with sides notched for securing the handle, as in arrowpoints and spearheads. Represented by figure 173, showing a specimen of gray flint from a mound in Mississippi county, Arkansas. There are four additional specimens, all from Union county, Illinois.

F. A related form, also notched for attachment of handle. Figure 174 represents an example of yellow flint, from Poinsett county, Arkansas, the only one of this shape in the collection.

From Jackson county, Illinois, there is a series beginning with a small scraper and a small scraper-like celt, and passing gradually into the large spades or digging-tools, there being a number of intermediate forms and sizes. Two specimens, only 6 inches long, have the glazed surface so characteristic of these implements, which could have been produced only by long-continued use in digging.

From a workshop at Mill creek, Union county, Illinois, there are a large number of pieces in every stage of work. Among them can be made series of all the different types here given, from the nodule in its natural state to the completed implement. Near by is a flint deposit showing extensive aboriginal quarrying.

Fig. 171.—Chipped spade, ovoid.

Dawson,125 in speaking of these implements, says: “The rudest of all rude implements, similar to the paleoliths of Europe, were used by the more settled and civilized agricultural nations.” While the majority of them are rude, simply because there was no necessity for elaborate work or fine finish in tools of this class, yet there are many specimens (as, for example, the one shown in figure 171) which in symmetry and workmanship will compare favorably with the larger specimens of other types, due regard being had to the fact that the coarse flint of which they are usually made does not admit of the most delicate execution.

Turtlebacks.

The singular name “turtleback” is suggested instantly on seeing a specimen of the class so designated by Abbott and others. As commonly used, it refers to rude or unfinished leaf-shape implements of any size, which may be found in great abundance almost anywhere. 137 It is used here, however, to denote more especially the disks or almond-shaped pieces of flint or chert sometimes found cached in considerable numbers.

Perkins126 records the discovery of such caches in Vermont; an exceptional case, as they are seldom found outside of the Mississippi valley. The southern portion of Illinois has furnished more than any other section; those found there are almost invariably made from nodules of bluish gray hornstone, the concentric lines being strongly marked.127

Fig. 172.—Chipped spade.

The Bureau has secured a large number from southern Illinois, ranging from 3½ to 7½ inches in length, some nearly circular, others having a length nearly twice the breadth. All have secondary chipping around the edges. Many of the larger ones and most of the smaller have the edges more or less worn or polished in such manner as would result from use as knives or scrapers. A typical specimen is shown in figure 175. 138

Fig. 173.—Chipped spade, showing handle notches.

Fig. 174.—Chipped spade.

Fig. 175.—Chipped disk, or “turtleback.”

Stevens128 denies in strongest terms that these relics are unfinished implements, saying it is the worst possible form into which flint could be chipped for carrying or for future work. On the other hand, Cheever129 says the Indians of California usually carry a pouch of treasures, consisting of unfinished arrowheads or unworked stones, to be slowly wrought out when they are industriously inclined. Catlin, too, observed that the Apache sometimes carry bowlders of hornstone a long distance to obtain material for arrowheads;130 and according to 139 im Thurn, the various Indian tribes of Guiana have each their special manufacture and exchange with other tribes.131 Tylor says:

Till lately the Patagonians, when they came on their journeys to a place where suitable flint or obsidian was to be found, would load themselves with a supply of lumps to chip into these primitive currier’s scrapers.132

Both Jewitt133 and Evans134 say that stones of this character were used as sling-stones; but there is no evidence that North American Indians ever used slings. Speaking of similar stones, Tylor remarks:

They were used either as knives or scrapers; with the curved side upward (or out) there would be no danger of cutting a hide in skinning game, and they could be used to cut up the flesh; while by putting the pointed end in the handle they could be used as scrapers.135

The smoothed edge in so many specimens substantiates the last statement, while the theory that they are unfinished implements finds support in the fact that nearly all the nodules from which they are made have an ellipsoid form, and the present shape of the implement would result from chipping away the useless weathered surface to lessen the weight.

Smaller chipped Implements.

Materials and Modes of Manufacture.

In the remaining portion of this paper, which will treat of the smaller chipped implements, a plan somewhat different from that of the preceding part will be followed.

As already stated, these specimens are almost invariably made of some form of flint; this term including chalcedony, basanite, jasper, chert, hornstone, and similar rocks. So common is its use that the term “flints” is gradually being adopted as a name for all the different classes of arrowheads, knives, drills, etc. The exceptions are not numerous enough to justify separate classification, so no tables of material will be used. Further, the great abundance of such relics in all portions of the country makes useless any allusion to the number from any particular locality; about the only limitation to their discovery is the amount of time and care which one chooses to give.

Before entering on the description, some quotations may be given in regard to methods of making these chipped implements.

According to Evans, the Mexican Indians take a piece of obsidian in the left hand and press it firmly against the point of a small goathorn held in the right, and by moving it gently in different directions they chip off small flakes until the arrow is complete;136 they also 140 cut a notch in the end of a bone, into which the edge of the flake is inserted and a chip broken off by a sideways blow.137 According to the same author, the Eskimo sometimes set the flake in a piece of split wood. The arrow is roughly chipped by blows with a hammer, either direct or with a punch interposed, and is then finished by pressing off fine chips with a point of antler set in an ivory handle.138 Not only leaf-shape barbed arrows, but also ones either with or without the stem, can be produced by pressure with a point of antler; the former, however, are the more easily made, and were probably earlier in use.139

The Plains Indians lay the flat side of a flake of obsidian on a blanket, or other yielding substance, and with a knife nick off the edges rapidly. In their primitive state they probably used buckskin instead of the blankets, and pointed bone or horn instead of the knife.140

The Apache holds the flake or flint in his left hand, places his punch at the point where the chip is to be broken off, and it is struck by an assistant, thus knocking a chip from the under side; the flake is then turned and the process repeated, until the arrow is complete. The stone is held in the hand, as it can not be chipped on a hard substance.141 A punch observed by Catlin in use by these Indians was a whale tooth 6 or 7 inches long, with one round and two flat sides. The Fuegians, according to the same authority, use a similar process and make as fine implements.142

The Eskimo make a spoon-shaped cavity in a log, lay the flake over it, and press along the margin, first on one side and then on the other, like setting a saw, until they form two sharp serrated edges. The working tool is a point of antler firmly bound into a piece of ivory. The same plan is used by widely separated peoples.143

Nilsson, in chipping out gun flints with a stone hammer, found it necessary to have the point operated on lie immediately above a point that rested on the rock “anvil” which he used.144

The Veeard or Wiyot of California used a pair of buck-horn pincers tied together with a thong at the point; they first hammered out the arrowhead in the rough, and then with these pincers carefully nipped off one tiny fragment after another.145 The Klamath cover the hand with a piece of buckskin to keep it from being cut, and lay a flake along the ball of the thumb, holding it firmly with the fingers. With a point of antler from 4 to 6 inches long, they press against the edge, thus removing scales from the opposite side; they turn the flake around and over frequently, to preserve symmetry.146 141

The Shasta Indian lays a stone anvil on his knee, holds the edge of the flake against it, and with his stone hammer chips off flakes, finishing the base first, and gently chipping the whole arrow into shape. Both obsidian and glass are used.147 The Shoshoni Indians used the same process.148

A Pit River Indian has been seen to make a very sharp and piercing arrow from a piece of quartz, with only a piece of round bone, one end of which was hemispherical with a small crease in it (as if made by a thread) one-sixteenth of an inch deep. The arrow was made by pressing off flakes by main strength, the crease being to prevent the bone from slipping, and affording no leverage.149 John Smith (1607) says of the Powhatan Indian:

His arrowhead he maketh quickly, with a little bone, of any splint of stone or glass.150

The Cloud River Indian used two deer prongs, one much smaller than the other, the points ground to the form of a square, sharp-pointed file. He had also some pieces of iron wire tied to sticks and ground in the same manner; these were better than the deer horn, because harder, and not needing to be sharpened so often. The flake was held firmly in the left hand, guarded by a piece of buckskin; he pressed off chips with the larger tool, turning the arrow end-for-end when done on one side, so as to keep the edge opposite the middle line. The notches for barbs were worked out in a similar manner with the smaller tool.151

Some of the California Indians prefer agate and obsidian for their implements, as the close grain admits more careful working. They use a tool with its working edge shaped like a glazier’s diamond (apparently a piece of bone or antler with a square-cut notch on the side); the flake is held in the left hand, while the nick in the side of the tool is used to chip small fragments.152 Peale makes similar statements, and adds that the notches are of different sizes to suit the different stages of work.153

The Klamath Indians, according to Schumacher, have a slender stick 1½ feet long, with a piece of sea-lion tooth, or antler, fastened to the end of it. Holding one end under the arm to steady it, they take a flake in the left hand, wrapped in a piece of buckskin so as to leave only the edge exposed, and by pressure with the point of the tool break off flakes as large as necessary, the last being quite fine, to give sharp edges to the arrow. The notches are worked out by means of a point of bone 4 or 5 inches long, without a shaft.154 Chase gives a similar account, but says that iron points have now taken the place of the bone or horn points formerly used.155 142

It may not be out of place in this connection to give a few quotations in regard to the length of time required for making an arrowhead.

According to the Marquis de Nadaillac, the Mexicans could turn out a hundred flint knives (probably only unworked obsidian flakes) an hour,156 while Crook says that the Plains Indians with only a knife for nicking off the edges, will make from fifty to one hundred arrows in the same period.157 Chase found that a Klamath Indian required five minutes to complete a perfect arrowhead;158 though Stevens observes that a Shasta Indian spent an hour in chipping one from a flake of obsidian,159 and Lubbock states that the most skillful Indian workmen can not hope to complete more than a single arrow in a day’s hard work.160 Powers also speaks of the aborigines of California as "using that infinite patience which is characteristic of the Indian, spending days, perhaps weeks, upon a single piece;161 and Tylor notes “that utter disregard of time that lets the Indian spend a month in making an arrow.”162

The last two references are probably to the large and finely worked pieces used for ceremonial or ornamental purposes.

Classification of the Implements.

The only practicable division of the greater part of the smaller flints is into stemmed and stemless, the former having a prolongation at the base for firmer attachment to a shaft or handle, the latter being of a triangular or oval shape. The stemmed implements may be barbed or not, and the stem either narrower or broader toward the end.

The name “arrowhead” so commonly applied, fits only the minority of specimens, as none but the smaller ones could be so used; the larger are too heavy. The longest stone arrowpoint in the extensive collection of arrows in the National Museum measures two and five-eighths inches in length and is narrow and thin. An arrowpoint two inches in length is seldom seen. The larger specimens were probably knives and spearheads; but it would be difficult to assign any certain use for a particular type, the markings on so many indicate usage for which their shape would seem to render them unsuitable. It is probable that a single specimen served a variety of purposes.

Wood, bone, and shell were also used to a considerable extent, in the manufacture of implements for which flint would seem much better adapted. Thus for fish spears the southern Indians used canes, sharp-pointed, barbed, and hardened in the fire,163 while knives were formerly made of flint or cane; these are still used when the hunting knife has 143 been lost.164 The California Indians had arrows tipped with hard-pointed wood for common use, and with agate or obsidian for war.165

The accompanying diagram (figure 176) will render plain the different terms used in the following descriptions:

Fig. 176—Diagram, explaining terms.

a Point. b Edge. c Face. d Bevel.166 e Blade. f Tang. g Stem. h Base. i Notch. k Neck. m Barb, or shoulder.

The only difference between barb and shoulder is that the barb is prolonged toward the base. The shoulder is called squared or rounded according to whether the edge of the implement makes an angle or a curve where drawn in to form the stem.

In the stemless specimens the base is the end opposite the point.

A tapering stem means one narrowing toward the base; straight, one whose sides are parallel; and expanding, one which is widest at the base.

Stemless Flints.

CHARACTERS AND USES.

The stemless flints are triangular or oval in outline. For convenience they will be divided into those small enough for arrowpoints (not above 2½ inches long) and those which are too large for such purpose. The latter reach to the length of 7½ inches. They are chipped to a sharp edge all around. The ratio of width to length varies from 1:4 to 4:5.

These objects were mostly for use as knives, scrapers or spearheads. Some of the thicker ones were spikes for clubs. Abbott167 mentions three triangular jasper implements 3 to 4 inches long from graves, associated with fragments of large bones which showed plainly that they had been used for clubs, and the Iroquois are known to have used a club with a sharp-pointed deer-horn about four inches long inserted in the lower side. Schoolcraft168 illustrates a pointed stone with a square 144 section (apparently of the class usually called “picks”), mounted in a club which is curved at the end to let the spike set in the end at a right angle to the handle; and Brickell observes that the North Carolina Indians used clubs or long poles, in the ends of which were fastened artificially sharpened stones, or horns of animals.169 Morgan also notes that among the Iroquois rows of arrow-shaped chert heads about two feet in extent have been found lying side by side. They were set in a frame and fastened with thongs, forming a species of sword.170 According to Tylor the Mexicans had a similar sword, with obsidian teeth gummed in holes in a war club,171 and Bourke observed at Taos pueblo a similar weapon with iron teeth.172 But the number of specimens found mounted indicates that most of them were used as knives or scrapers.

LARGER IMPLEMENTS.

A. With base and edges straight or slightly convex; corners square. The type illustrated in figure 177 is from Montgomery county, North Carolina. Similar forms come also from eastern Tennessee; central and western North Carolina; southwestern Illinois; Miami and Scioto valleys, and central Ohio; southwestern Wisconsin; northeastern and southwestern Arkansas; northeastern and northwestern Alabama, and Coosa valley in the same state; Kanawha valley, West Virginia; northeastern and central Kentucky; and Savannah, Georgia.

Fig. 177.—Triangular chipped flint.

B. Base straight or nearly so; edges parallel most of the length, curving abruptly to a point; usually with one face less convex than the other, even quite flat, giving a plano-convex section; medium size. The specimen shown in figure 178, from Kanawha valley, West Virginia, is representative. Other examples come from eastern Tennessee; central North Carolina; northwestern Alabama; Kanawha valley; and southwestern Illinois.

Fig. 178.—Chipped flint.

C. Base straight or nearly so; corners square or slightly rounded; edges convex, curving gradually and regularly to the point; usually widest about one-third of the way above the base; varying much in width, and in length from 6½ inches down to the arrowpoint. A few of the largest have the edges slightly expanding at their junction with the base, for firmer attachment to a handle. The type is 145 figure 179 (from Loudon county, Tennessee). Other specimens are from eastern Tennessee; central and western North Carolina; Kanawha valley; Keokuk, Iowa; Miami and Scioto valleys, and central Ohio; eastern, southern, and southwestern Wisconsin; northeastern Arkansas; central and northeastern Kentucky; northwestern Georgia, and Savannah; southwestern Illinois; and Coosa valley, Alabama.

Fig. 179.—Chipped flint.

Fig. 180.—Chipped flint, somewhat bell-shape.