Anthropology Library

MEM. CONN. ACAD., VOL. Ill

PLATE I

Plate I

(Frontispiece.)

Interior of a shallow chalice artistically decorated with alligator motives, the central
figure being that of the alligator-god : a gem of aboriginal art and symbolism.
Three-fifths. Page 159. (See PI. XLV.)

MEM. CONN. ACAD., Vol. III.

MEMOIRS OF THE

CONNECTICUT ACADEMY OF ARTS AND SCIENCES

INCORPORATED A. D. 1799

VOLUME IK MARCH, 1911

A Study
of Chiriquian Antiquities

BY

GEORGE GRANT MAC CURDY, Ph.D.

ASSISTANT PROFESSOR OF ARCHEOLOGY, AND CURATOR

OF THE ANTHROPOLOGICAL COLLECTION

YALE UNIVERSITY

NEW HAVEN, CONNECTICUT
YALE UNIVERSITY PRESS

IQI I

LIBRARY

Add'l

GIFT

6-if: •

A' lia^rr eTT

WEIMAR : PRINTED BY R. WAGNER SOHN

PREFACE.

This volume is chiefly the result of a careful study of the unparalleled collec-
. tions of Chiriquian antiquities belonging to Yale University. These, and especially
the gold ornaments, have been supplemented by other collections, public and
private, notably those of Mr. George G. Heye and of Mr. Minor C. Keith, both
of New York. If the work has merits, they are due in no small measure to the
character and extent of the collections examined. Its defects should be ascribed
in part, at least, to the fact that, having been prevented by force of circumstances
from visiting the region in question, my investigations have been limited to
museums and libraries. For this reason the choice of title : " A Study of Chir-
iquian Antiquities," would seem to be particularly appropriate. Unless otherwise
stated in their description, the illustrations are from specimens in the Yale
collection.

From the view-point of conventionalism, the art of Chiriqui is so instructive that
much space is given to the evolution of decorative motives from animal forms.
In this connection, I wish to acknowledge my indebtedness to two previous writers
on the subject. Professor F. W. Putnam of Cambridge. Mass.. and Professor
William H. Holmes of Washington, D.C. The nature of the study has necessitated
the free use of illustrations, in the preparation of which much care has been
bestowed. With the exception of the drawings from tracings, the illustration from
de Bry (fig. 324). and the five cuts (figs. 277, 278, 331. 332. 365), for the use of
which my thanks are due to the Bureau of American Ethnology, they were all
made for this volume by Mr. Rudolph Weber of New York, and his assistant,
Mr. William Baake. The cost of the drawings, both pen-and-ink and water-color,
has been borne by Mrs. Evelyn MacCurdy Salisbury of New Haven, without whose
generous cooperation the work could not have been prepared.

Among others whom I desire to thank for cooperation in various ways should
be mentioned Professor Charles Schuchert for facilities at the Museum. Professor
Joseph Barrell for determining the nature of the materials from which the stone
objects were made, Mr. Max Dessauer for testing the powers of the musical in-
struments. Professor J. W. D. Ingersoll for the translation of a sixteenth century
text, and Professor B. B. Boltwood, Dr. C. H. Mathewson, Dr. George F. Kunz of
New York, and Mr. Harry N. Ray, for measurements, tests and observations bearing
on the metal work.

In New York also where my studies led me, are a number to whom I am
much indebted. Mr. George G. Heye and Mr. Minor C. Keith both placed their
valuable Chiriquian (and Costa Rican) collections at my disposal, as did also
Mr. Edwin Lamson. Dr. Wm. f. Lamson, the Lenox Branch of the New York Public
Library through Mr. Wilberforce Eames and Mr. W. A. Elliot, and the Metropolitan
Museum of Art through Sir Purdon Clarke and Mr. John H. Buck. To Professor
M. H. Saville of Columbia University. I am likewise under obligations for access
to comparative material and for valuable suggestions.

VI PREFACE.

Although the late Professor Othniel C. Marsh died nearly seven years before
I began the study of Chiriquian antiquities, it was his foresight, generosity and
consummate skill as a collector that brought together the series on which this
volume is primarily based. Mention in this place of the part he played is but
an inadequate expression of the tribute I would pay to his memory. In the reading
of copy and proofs I gratefully acknowledge the valuable assistance of Miss Lucy
P. Bush, for sixteen years assistant and private secretary to Professor Marsh, of
Dr. Katharine J. Bush, and Miss Mary S. Gillette.

In thanking Yale University and the Connecticut Academy of Arts and Sciences,
through the Committee on Publication acting on behalf of both, for means of
diffusing the results of this study, it is most fitting that I should recall the debt
our science owes to the American Ethnological Society, which published nearly
all the early literature, now exeedingly rare, on the subject of Chiriquian an-
tiquities, and which at that time numbered among its members such distinguished
Yale professors as President Theodore Dwight Woolsey, Professors Edward
Elbridge Salisbury and J. Willard Gibbs, the elder and S. Wells Williams, who
later became professor of Chinese at Yale.

YALE UNIVERSITY, NEW HAVEN, CONN.,
May 26, 1909.

TABLE OF CONTENTS.

P»ge

PREFACE . V

LIST OF ILLUSTRATIONS . IX

INTRODUCTION . 1

History .............. 1

Cemeteries and tombs 8

Human remains ............. 14

People ... 14

COLLECTIONS . . . 19

STONE , 21

Arrow-points 21

Spear-points 22

Celts 22

Polishing stones 26

Metates 26

Rubbing or hand stones . . . 33

Stools 35

Images 36

Ornaments 39

Petroglyphs 43

POTTERY 45

Classification .............. 46

(For revised classification, see page 65)

Unpainted ware 47

Armadillo group (armadillo motive, p. 56) 48

Salmon-colored sub-group 71

Black incised or serpent group 72

Handled group, unpainted 73

Painted ware 75

Handled group, painted .76

Tripod or Fish group 80

Red Line group 92

Chocolate incised group 95

Scarified group 96

Maroon group 100

White Line group . . 102

Lost Color group 103

Alligator group 125

Polychrome group 151

Various objects of clay 161

Stools 161

Spindle-whorls and stamps 163

Needle-cases 164

Figurines 165

Musical instruments 169

CONTENTS.

Page

Rattles . . -169

Drums . . -169

Wind instruments ... .169

METAL ... -189

Alloys of gold and copper . .194

Casting ... -194

Articles of use . . • 197
Ornaments .
Figurines of animals .

Human figurines . .210
Figurines with mixed attributes

Masks .... .... 223

Plaques . • 224

CONCLUSIONS .... .227

BIBLIOGRAPHY .... .235

INDEX 239

LIST OF ILLUSTRATIONS.

MAP. P»Re

The Province of Chiriqui. ....... To face page 18

PLATES.
(Its own list to be found facing each Plate.)

TEXT-FIGURES.
Figures 1—3. — Arrow-points. '/• . . . . . . . .21

Figures 4, 5. — Spear-points. */• . . . . . . . . 22

Figure 6. — Cylindrical celt with thick blunt pole. '/> . . . . . .24

Figure 7. — Large polished celt with pointed chipped pole; from Gualaca. '/> . . 24

Figure 8. — Diminutive triangular celt with chipped pole. '/• .... 24

Figure 9. — Octagonal celt with polished and faceted blade and chipped pole. */• . 25
Figure 10. — Polished hexagonal chisel with chipped base. 'I' . . .25

Figure 11. — Polished cylindrical celt with distinct shoulder. 1' . . . .25

Figure 12. — Polished and chipped chisel. */« . . . . . . .25

Figure 13. — Grooved celt. *l' ......... 26

Figures 14, 15. — Jasper polishing stones. '/« . . . . .26

Figure 16. — Chalcedony polishing stone. '/' . . . . . . .26

Figure 17.— Metate with three legs, and hand stone. '/• . . . .27

Figure 18. — Three-legged metate of the Nicoyan type; from Gualaca. "• . . 28

Figure 19.— Prototype of Chiriquian metate. '/• . . . . .28

Figure 20. — Metate with openings in the hollow stand ; from Vivala. ''' . . 28

Figure 21. — Metate with base open at the ends, the sides being decorated with

incised panels. '/» . . . . . . . . .29

Figure 22. — Modern wooden seat used by the Indians of Chiriqui. ''» . . . . 29

Figure 23. — Metate of the four-legged type, ornamented with ten animal heads attached

to the margin of the plate. '/• . . . . . . . .29

Figure 24. — Metate representing anterior half of two animals. '/• ... 30

Figure 25. — Metate representing a single zoomorphic unit. '/> . . .30

Figure 26. — Jaguar metate with rectangular top. '/» . . . . . .30

Figure 27. — Jaguar metate the legs of which are decorated with stars in champleve. '/• 31
Figure 28. — Jaguar metate with guilloche ornamentation. '/« . . . .31

Figure 29.— Metate with animal head at each end. '<« . . . . .32

Figure 30. — Small crudely shaped metate with rectangular plate. 'I* . . 33

Figure 31. — Rectangular rubbing stone. 'I' ....... 34

Figure 32.— Pestle. '/• 34

Figure 33. — Sub-spherical pestle streaked with gold. ''' . . . .34

Figure 34.— Image representing crude animal form. '/• . . . . .37

Figure 35. — Small image of a quadruped. '/• . . . . . .37

Figure 36. — Image seated on cylindrical pestle ; may represent man or monkey. '/• 37
Figure 37. — Large crude human image with " skil " ornament on the head ; from

Bugavita. '/•._. . . . . . . . . .37

Figure 38. — Crude human image ; from Bugavita. ''• ..... 37

Figure 39. — Rough statue in standing posture; from San Carlos. '/» . . 38

MEMOIRS CONN. ACAD., Vol. III. n

X A STUDY OF CHIRIQUIAN ANTIQUITIES.

Page

Figure 40.— Large statue of a female— the so-called Panama Venus; a, front view;

b, profile view. V« ..... .39

Figures 41, 42.— Highly polished agate beads. '/• . . 40

Figure 43.— Green jasper bead. '/« . • 40

Figure 44. — Small agate pendant. '/> . 40

Figure 45.— Highly polished amulet of pale green jade. '/• . . 40

Figure 46.— Large translucent amulet of jade ; from Miravalles, Costa Rica. '/« . 40

Figure 47.— Small celt-shaped amulet. (Lamson collection.) '/> . . 41

Figure 48.— Ornithomorphic celt-shaped amulet of jade. '/• . 41

Figure 49.— Ornithomorphic amulet of green translucent jade. 'A . . 42

Figure 50.— Zoomorphic amulet of agate. '/•..... .42

Figure 51. — Amulet made from a shark's tooth; from Divala. (Lamson collection). 'A 42
Figure 52. — Southwest face of the piedra pintal or pictured rock at Caldera.

(After Holmes.) .... . .43

Figure 53.— Northeast face ot the piedra pintal, from a water-color drawing made

by Gentil .43

Figure 54.— Tracing of a petroglyph. (After Bateman.) . . 44

Figure 55. — Primitive rectangular bowl ; from El Banco. Armadillo ware. '/> . 49

Figure 56.— Primitive elongated bowl with small handle; from Divala. Armadillo ware. '/> 49
Figure 57.— Calabash type of cup with diminutive handle; from near Bugavita.

Armadillo ware. 'I* ......... 49

Figure 58. — Hemispherical bowl with spoutlike projection of rim ; from near Bugavita.

Armadillo ware. '/«.... .49

Figure 59.— Calabash type of bowl with no projection of rim. Armadillo ware. '/> 49
Figure 60. — Tripod bowl with plain hollow supports. Armadillo ware. '/> . 50

Figure 61. — Tripod bowl with plain hollow supports, apparently inflated. Armadillo ware. 'A 50
Figure 62. — Tripod with zoomorphic supports. Armadillo ware. '/> . . 52

Figure 63.— Tripod bowl; from Bugavita. Armadillo ware. '/> . . . 52

Figure 64.— Tripod with froglike supports ; from Bugavita. Armadillo ware. '/> . 52

Figure 65.— Tripod of classic form, with froglike supports, and rim ornamented with

armadillos. Armadillo ware. '/>.... .52

Figure 66.— Tripod bowl of superior modeling, supported by three grotesque apelike

heads ; from near Bugavita. Armadillo ware. V« . . . . 53

Figure 67. — Tripod representing a primitive zoomorphic type. Armadillo ware. '/* . 53
Figure 68.— Vessel of zoomorphic type, with four legs ; from Divala. Armadillo ware. V> 54
Figure 69.— Tripod bowl showing sculptor's use of shadows in modeling the eye;

from Bugavita. Armadillo ware. 'I1 ...... 54

Figure 70.— Tripod bowl exhibiting further use of shadows in modeling the eye.

Armadillo ware. 'A ...... .54

Figure 71. — Tripod bowl with solid fishlike supports. Armadillo ware. V» . . 55

Figure 72. — Bowl with annular base; from El Banco. Armadillo ware. '" . 55

Figure 73.— Goblet-shaped vessel; from Bugavita. Armadillo ware. '/» . . 55

Figure 74. — Bowl supported by annular base and human figures. Armadillo ware. 'A 55
Figure 75. — Bowl supported by three realistic armadillos. Armadillo ware. 'A . 58

Figure 76.— Bowl with similar tripod supports, the extremities of the armadillo being

reduced to feet only. Armadillo ware. ''• . . . .58

Figure 77. — Tripod bowl in which the hindlegs and feet of the armadillo are elimi-
nated from the supports ; from Escaria. Armadillo ware. 'A . . 58
Figure 78. — Tripod, the supports of which are ornamented with carapace motives only ;

from Escaria. Armadillo ware. V« . . 58

LIST OF ILLUSTRATIONS. XI

Page

Figure 79.— Vase with shoulder ornament, in which the carapace motive is present on

ventral surface only ; from Vivala. Armadillo ware. '/> . . . .59

Figure 80.— Vase with shoulder ornament showing but two divisions of the ventrally

placed carapace. Armadillo ware. ''« ...... 59

Figure 81.— Vase with shoulder ornament and only one division of the ventrally placed

carapace. Armadillo ware. '/> . . . . . . . .59

Figure 82.— Cup with shoulder ornament, on which an unbroken band represents the

ventrally placed carapace motive. Armadillo ware. '/« . . . .59

Figure 83.— Tripod vase with shoulder ornament, to which carapace motives are applied

both dorsally and ventrally. Armadillo ware. '/> . . . .60

Figure 84.— Large vase with armadillo shoulder ornaments, from which the carapace

motive is absent. Armadillo ware. '/« ...... 60

Figure 85.— Neck of vase ornamented with the tail motive. Armadillo ware. '/» . 61
Figure 86.— Another example of the tail motive. Armadillo ware. '/• . . 61

Figure 87.— Neck of vase ornamented with alternating groups of tail- and eye-motives ;

from Vivala. Armadillo ware. '/'....... 61

Figure 88.— Neck of vase decorated with tail motive in meander, and an eye symbol

in each angle. Armadillo ware, 'i' . . . . . . .61

Figure 89. — Neck and shoulder ornament of a vase proving the identity of the cara-
pace motive. Armadillo ware. */• . . . . . . .62

Figure 90. — Meander of carapace symbols accompanied by foot symbols. Armadillo ware. '/« 62
Figure 91.— Double neck of vase, ornamented with carapace- and eye-symbols.

Armadillo ware. '/> . . . . . . . . .62

Figure 92.— Neck of vase showing foot- and eye-symbols in alternation. Armadillo ware. '/• 63
Figure 93.— Vase, the single handle of which is ornamented with carapace motives.

Armadillo ware. '/' ......... 63

Figure 94. — Vase with two undecorated handles. Armadillo ware. '/• • . 63

Figure 95. — Vase with single handle forming an arch over the aperture. Armadillo

ware. 'I* ........... 65

Figure 96. — Cup with single shoulder ornament representing the armadillo. Armadillo

ware. '/»........... 65

Figure 97. — Cup with single flat handle showing possibly a conventionalized form of

the armadillo ; from Divala. Armadillo ware. V» . . . .65

Figure 98. — Vase with shoulder ornaments representing the frog, probably the Surinam

Toad (Pipa americana). Armadillo ware. '/> . . . . .66

Figure 99. — Vase with frog shoulder ornaments; from Vivala (not Divala). Armadillo ware '/» 67
Figure 100.— Frog illustrating the ceramic type of technique. Armadillo ware. '/> . 67
Figure 101. — Vase with shoulder ornaments representing a frog in the tadpole stage.

Armadillo ware. V» ......... 68

Figure 102.— Vase with zoomorphic shoulder ornaments. Armadillo ware. ^ . . 68

Figure 103.— Vase with grotesque zoomorphic shoulder ornament; from Bugavita.

Armadillo ware. '/» ......... 68

Figure 104. — Fantastic zoomorphic shoulder decoration. Armadillo ware. '/> . . 68

Figure 105. —Unique flat-bottomed vase; from Bugavita. Armadillo ware. '/» . . 69

Figure 106. — Vase with handles represented by the human figure. Armadillo ware. '/« 69
Figure 107. — Small vase with anthropomorphic shoulder ornaments ; neck decorated

with carapace meander combined with foot symbols. Armadillo ware. 'I* . 69
Figure 108. — Cup with zoomorphic handle decoration. Armadillo ware. '/' . . 70

Figure 109. — Cup with decoration on handle probably representing a human head.

Armadillo ware. '/• 70

XII A STUDY OF CHIRIQUIAN ANTIQUITIES.

Page

Figure 110.— Example of anthropomorphic handle decoration. Armadillo ware. '/• . 70

Figure 11 1.— Cup with grotesque human figure serving as handle. Armadillo ware. 3/' 70

Figure 112. — Another treatment of the same motive, the nose being developed into

a proboscis. Armadillo ware. '/'....... 70

Figure 113. — Cup with handle decoration suggesting the human form with nose

developed into a proboscis. Armadillo ware. '!• .70

Figure 114.— Cup with human effigy for handle. Armadillo ware. '/• . .71

Figure 115. — Cup with froglike figure serving as handle. Armadillo ware. '/• . 71

Figure 116; a. — Vase decorated with parallel zigzag incised lines. ''", b. — Detail of in-
cised pattern decorating the bottom of the vase shown in figure a. Serpent
ware. '/« ........... 72

Figure 117.— Small vase with shoulder decoration consisting of a triangular treatment

of the serpent motive. Serpent ware. ''• . . . . . .73

Figure 118.— Tripod bowl with supports representing a fish; from Divala. Serpent ware. Vt 73

Figure 119.— Vase bearing life symbols on handles and incised X-shaped ornamen-
tations on shoulder. Unpainted handled ware. '/» . 74

Figure 120.— Vase with horizontally placed handles representing animal forms. Un-
painted handled ware. '/» ...... .74

Figure 121. — Vase with carapace symbol ornamenting base of handles. Painted

handled ware. 'I' ........ .77

Figure 122. — Vase whose handle decorations represent either fin or carapace motives.

Painted handled ware. ''• ........ 77

Figure 123. — Vase with same life form on handles as in preceding example ; from Jacu.
Painted handled ware. '/• .......

Figure 124. — Vase with carapace symbol decorating the handles. Painted handled ware. ''" 78

Figure 125. — Vase with twisted handles composed of two strands ; from Divala. Painted

handled ware. '/« . . . . . . . 78

Figure 126. — Detail of shoulder ornamentation on vase shown in Plate XXI, figure h.

Painted handled ware. ''• ...... .78

Figure 127.— Vase with life forms, probably human, serving as handles. Painted

handled ware. '/« . . . . . . . . .78

Figure 128. — Vase with single handle spanning the aperture. Painted handled ware. 'I' 79

Figure 129.— Vase with grotesque forms serving as handles; from Caklera. Painted

handled ware. '/> . . . . . . . . . .79

Figure 130. — Vase with plastic decoration suggesting the carapace motive ; from Divala.

Painted handled ware. '/• ........ 79

Figure 131.— Vase with well-developed collar but without handles; fromVivala. Painted

handled ware. '/• .80

Figure 132. — Vase without handles. Painted handled ware. '/» . . .80

Figure 133. — Tripod with long solid supports and with fillets applied obliquely to the

handles ; from Bugavita. Fish ware. *'• .81

Figure 134.— Tripod with twisted handles and with supports resembling the body of

a fish. Fish ware. 'i> ......... 81

Figure 135.— Tripod, the legs of which are realistic representations of the fish. The

half-human monsters serving as handles may represent the Jaguar-god. Fish ware. '/> 82

Figure 136. — Tripod with plain handles and fishlike supports. Fish ware. '/« . 82

Figure 137. — Tripod with twisted handles and supports in which the anal tins are

present. Fish ware. '/• . . . . . . . .83

Figure 138. — Tripod in which the dorsal fin appears as a lump on the nose of the fish.

Fish ware. '/• 83

LIST OK ILLUSTRATIONS. XIII

Page

Figure 139. — Tripod with gracefully modeled supports representing the fish. Fish ware. '/> 84
Figure 140.— Typical example of handled ware with legs, the. latter exhibiting pec-
toral fins only. Fish ware. 'I' . . . . . . .84

Figure 141. — Tripod support, the only ornamentation being a single dorsal fin. Fish

ware. '/> ........... 85

Figure 142. — Tripod, in which the supports embody a fusion of the fish and the human

form. Fish ware. *l* ......... 87

Figure 143. — Tripod, in which the supports combine the fish with the head and feet

of the owl. Fish ware. '/> .... 87

Figure 144. — Tripod, in which the legs combine characters of the fish with those of

some carnivorous animal. Fish ware. '/» . . . . .88

Figure 145. — Ornate tripod with twisted handles representing animal forms. Fish ware. ''- 88
Figure 146. — Tripod, in which the supports are modified to represent the human form.

Fish ware. '/» . 89

Figure 147. — Tripod with supports representing the human form. Fish ware. '/J . 89
Figure 148.— Tripod support representing the human form. Fish ware. V> . . 90

Figure 149. — Similar example from another tripod. Fish ware. '/»... 90
Figure 150.— Tripod support exhibiting similar technique ; but the head of the figure is

scarcely human. Fish ware. '/« ....... 90

Figure 151. — Tripod, the shallow bowl of which is decorated with crab- and fin-motives;

from El Banco. Fish ware. 'I* ...... .91

Figure 152. —Vase with annular base; from El Banco. Fish ware. '/• . . 91

Figure 153.— Tripod with elongated shallow bowl, and handles suggesting life forms;

from Bugavita. Fish ware. ll> ...... .91

Figure 154. —Another tripod similar in type, unpainted ; from Bugavita. Fish ware. '/> 91
Figure 155.— Tripod embodying animal forms in the supports; from Caldera. Probably

of Costa Rican origin. '/» ..... 92

Figure 156. —Tripod with short looplike supports ; from El Banco. Red line ware. V» 93
Figure 157. — Bowl with low annular base ; from Gualaca. Red line ware. '/> . 94

Figure 158.— Bowl with perforated annular base. Red line ware. '/« . . 94

Figure 159.— Shallow bowl with perforated annular bell-shaped base; from El Banco.

Red line ware. '/«... .94

Figure 160.— Globular vase. Red line ware. '/« • 94

Figure 161.— Tripod vase with incised shoulder and short hollow supports supplied with

pellets but not slit; from Divala. Chocolate incised ware. '/• . . 95

Figure 162.— Tripod with incised panels on shoulder and short supports representing

animal heads; from Divala. Chocolate incised ware. V» . 96

Figure 163.— Very small tripod with unbroken incised pattern around shoulder, and

short solid supports. Chocolate incised ware. ''. . . . 96

Figure 164.— Non-scarified vase with handles crudely representing animal forms; from

Caldera. Scarified ware. V» • 96

Figure 165.— Tripod bowl with handles and supports resembling head, wings and tail

of a bird ; from Caldera. Scarified ware. '/« 97

Figure 166.— Non-scarified tripod, heart-shaped in horizontal section. Scarified

ware. '/«..... 97

Figure 167.— Tripod in which shoulder ornaments and supports represent the head and

arms, and the legs respectively of the human figure; from El Banco.

Scarified ware. ''»
Figure 168.— Open-mouthed bowl with three short supports placed close together.

Scarified ware. ''» ......... 98

XIV A STUDY OF CHIRIQUIAN ANTIQUITIES.

Page

Figure 169.— Boat-shaped tripod with rim extended to a point at the ends, each point

being accompanied by two nodes. Scarified ware. V» . . .99

Figure 170.— Vase with loop handles strongly resembling that shown in figure 164;

from Caldera. Maroon ware. '/>.... .100

Figure 171.— Vase ornamented with scarifications alternating with pairs of vertical

punctated ribs. Maroon ware. '/>.... . 100

Figure 172.— Small tripod completely encircled by a series of plain vertical ribs ;

from El Banco. Maroon ware. 'I* ....... 101

Figure 173.— Bowl with ornamentation in relief ; from Vivala (not Divala). Maroonware. V> 101
Figure 174. — Open-mouthed bowl with annular base and animal features in high relief;

from near David. Maroon ware. 'I* ....... 101

Figure 175. — Vase with single handle, opposite which is a relief ornament consisting

of a pair of eyes, prominent eyebrows and a nose ; from Gualaca ; Maroon

ware. '/« . 101

Figure 176.— Vase decorated with the alligator dorsal-view motive in white. White

line ware. '/> .......... 102

Figure 177. — Tripod, the shoulder of which is decorated with the dorsal- view alligator

motive in white. White line ware. '/> ...... 102

Figure 178; a. — Small tripod with shoulder ornamentation suggestive of hieroglyphics,

and legs representing animal forms ; from Bugavita. '/» ; b. — Detail of shoulder

ornamentation. White line ware. '/' . . . . . . .102

Figure 179. — Narrow-necked vase broken in plane of greatest diameter, illustrating

carelessness in finish of interior; from Divala. Lost color ware. '/» . . 104

Figure 180. — Vase in red and black with large aperture, and two arched panels on

shoulder. Lost color ware. '/> ....... 106

Figure 181. — Vase in red and black with narrow aperture, showing life forms in low

relief alternating with two panel decorations ; from Bugavita. Lost color ware. '/' 110
Figure 182. — Vase in red, light cream and black, the three panels of the shoulder zone

being decorated with alligator motives. Lost color ware. V> . .110

Figure 183. — Vase in red, light cream and black, ornamented with bilaterally sym-
metrical figures each covering about half of the upper zone. Lost color ware. '/> 111
Figure 184. — Vase, in which the peripheral band and those tangent to the neck are

painted white. Lost color ware. '/> . . . . . .113

'Figure 185. — Vase, in which the peripheral band and those tangent to the neck are

painted white and are crossed by the lost color pattern. Lost color ware. '/> 113
Figure 186.— Vase ornamented with four four-sided panels. Lost color ware. '/> . 114
Figure 187. — Globular vase ornamented with a design suggesting a conventionalized

alligator. Lost color ware. V» . . . . . . . .114

Figure 188.— Globular vase decorated with the spool-shaped motive. Lost color ware. '/» 115
Figure 189.— Wide-mouthed vase. Lost color ware. '/> . . . . .115

Figure 190. — Small vase with rosette-like ornamentation. Lost color ware. '/' . 115

Figure 191.— Compound vase decorated with three quadrangular panels in each story.

Lost color ware. '/2 . . . . . . . . .115

Figure 192. — Vase showing unusual disposition of read and light colors of the original

ground. Lost color ware. '/« . . . . . . .116

Figure 193. — Unique vase with large long neck sourmounted by a rectangular lip, and

with tuberosities on the body. Lost color ware. '/» . . . .116

Figure 194. — Small vase with radiate ornamentation. Lost color ware. J/J . . 118

Figure 195. — Small vase with radiate ornamentation, the elements of which may be

traced to an alligator motive. Lost color ware. 'I' . . . .118

LIST OF ILLUSTRATIONS. XV

Page

Figure 196. — Vase in pale yellow and black, with zonal decoration ; from Divala. Lost

color ware. '/« . . . . . . . . . .118

Figure 197. — Vase in light cream and black, decorated with vertical panels each en-
closing a faulted meander. Lost color ware. */» . . . .118

Figure 198.— Small vase decorated with parallel coiled bands; from Vivala (not Divala).

Lost color ware. */« . - . . . . . . . .119

Figure 199. — Vase with handles and with frondlike ornamentation. Lost color ware. '/» 119
Figure 200. — Vase with primitive human figures as shoulder ornaments. Lost color ware. V« 120
Figure 201. — Very small vase with relief figures on shoulder representing the human

head and arms. Lost color ware. !/» . . . . . . . 120

Figure 202.— Vase ornamented with bird's head and tail in relief, the latter resembling

a horizontal loop handle. Lost color ware. '/>..... 121
Figure 203. — Vase with vertical handle, opposite which is a figure with human body and

the beak of a bird, presumably the parrot-god ; from Divala. Lost color ware. */' 121
Figure 204.— Small double vessel with single arched handle; from El Banco. Lost

color ware. 'I* .......... 121

Figure 205. — Double vessel with two outer handles and a single short arched central

handle. Lost color ware. */'........ 121

Figure 206. — Detail showing favorite representation of alligator in profile. '/> . . 126

Figure 207.— Similar treatment of the alligator. '/• . . . . .126

Figure 208. — Alligator in profile with in verted scale-group motives and nuchal appendage. '/> 127
Figure 209. — Detail showing fusion of two alligators in profile. '/» ... 127

Figure 210.— Alligator profile showing exaggeration of jaws and nuchal appendage. *'» 128
Figure 211. — Alligator profile adapted to fit the space at hand. '/•... 128
Figure 212. — Profile of alligator with head turned backward, the long lower jaw being

balanced by the well developed nuchal appendage. */> . . . . 129

Figure 213. — Profile of alligator in which only the head and nuchal crest remain

prominent. */» . . . . . . . . . .129

Figure 214.— Well-balanced but angular and stylistic representation of the alligator. */• 130
Figure 215. — Highly conventionalized two-headed alligator. */> . . . 130

Figure 216.— Double-headed alligator with nuchal crest on the head at the left. '/> 130
Figure 217. — Two-headed alligator showing transposition of parts, both nuchal crests

being attached to the same neck. !/> . . . . . . . 130

Figure 218. — Two-headed alligator with nuchal crests absent. '/> ... 130

Figure 219«— d. — Series of motives, each representing the double-headed alligator. '/« 131
Figure 220. — Multiple alligator motive. '/» ........ 131

Figure 221.— Simplified multiple alligator motive. V» ..... 131

Figure 222. — Detail probably representing two body-lines, each with a single scale symbol. '/» 131
Figure 223. — Sigmoid scroll representing two body-curves. '/« . . . 131

Figure 224.— Conventionalized alligator in profile; hieroglyphic stage. 'I' . . 131

Figure 225.— Wholesale reduction and simplification of the alligator in profile. '/« . 131
Figure 226a— c. — Conventionalized alligator designs. '/« . . . . .133

Figure 227. — Vase, whose shoulder zone includes six arched panels filled with scale

symbols; from Divala. Alligator ware. '/» . . . . . .133

Figure 228. — Vase, whose three arched panels are associated with scale motives.

Alligator ware. '/» ......... 133

Figure 229. — Vase having but two arched panels, in which scale- and spine-motives

are grouped ; from Divala. Alligator ware. '/« . . . .134

Figure 230. — Double-necked vase with the dorsal-view motive as a panel decoration;

from Bugavita. Alligator ware. 1* . . . . . . .134

XVI A STUDY OF CHIKIQU1AN ANTIQUITIES.

Page

Figure 231.— Vase, in which the dorsal-view motive alternates with quadrangular panels.

Alligator ware. '!>... .135

Figure 232.— Vase with alligator motives on lip and shoulder. Alligator ware. 'I' . 135
Figure 233.— Vase in which the alligator motives are obscure. Alligator ware. '/' 135
Figure 234.— Vase with obscure alligator motives on lip and shoulder. Alligator ware. ''• 135
Figure 235.— Vase with alligator motives on lip, and with multiple body-line as a panel

decoration. Alligator ware. '/> . . 136

Figure 236.— Vase with simple, body-line motives on shoulder and lip. Alligator ware. 2/» 130
Figure 237.— Vase with shoulder panels showing obscure multiple body-line, accompan-
ied by body-markings. Alligator ware. '/« ... . 137
Figure 238.— Vase with small vertical loop-handles and the alligator motive as a shoulder

decoration. Alligator ware. '/> . .137

Figure 239.— Vase without neck, the relief ornaments being two opposed animal heads.

Alligator ware. V«
Figure 240. — Neckless vase with dorsal-view alligator motive as a panel decoration.

Alligator ware. '/• . . .138

Figure 241 a, b. — Designs representing conventionalized alligators. ''* . . 140

Figure 242.— Elaborate vase supported by human figures; end view (see Plate XLII,

fig. c). Alligator ware. '/»... . 142

Figure 243. —One of the alligator figures decorating the neck of the preceding vase. '/» 142
Figure 244 a, 6.— Jaguar-god ornamented with alligator motives and serving both as

rattle and receptacle. Alligator ware. ''' . . 143

Figure 245.— Zoomorphic vessel representing the tapir and ornamented with alligator

motives. Alligator ware. '/».... .145

Figure 246. — Globular vessel representing a bird, a low annular support taking the

place of feet ; from Vivala (not Divala). Alligator ware. 'I' . . 146

Figure 247.— Ornithomorphic vessel decorated with scale-group and spine motives.

Alligator ware. '/«.... . 146

Figure 248.— Bird form with realistic removable head. Alligator ware. '/» . 147

Figure 249; a.— Elaborate bird form with removable head, and ornamented with

dorsal-view motives. Alligator ware. 'I'; b. — Diagram showing mode of

stringing neck and body for suspension. ... . 148

Figure 250a, b.— Vase showing elaborate and highly conventionalized use of alligator

motive ; from Divala. Alligator ware. V« • 149

Figure 251.— Vase with problematical decoration. Alligator ware. '/> . . . 150

Figure 252.— Vase combining features of alligator and lost color groups, the dorsal-
view motive occurring on the peripheral band and those tangent to the neck.

Alligator ware. '/> . . . . . . . . . .150

Figure 253.— Sigmoid scroll in which the alligator motive appears as simply cut out

the black band where it is broadest (see Plate XLIV, fig. a). '/• . .152

Figure 254. — Branching scroll in which two of the alligator motives (notches) are

partially differentiated (see Plate XLIV, fig. a). '/' . . . .153

Figure 255. — Vase with elaborate branching scroll from which the alligator motives

are partially detached and accompanied by dots and circles representing

body-markings. Polychrome ware. ''• . . . . . .153

Figure 256. — Unique type of vase, the zonal decorations being alligator motives.

Polychrome ware. '/>......... 155

Figure 257. — The largest vase in the collection, unique in form and decorated with

alligator motives. Polychrome ware. '/• . . . . . .150

LIST OF ILLUSTRATIONS. XVII

Page

Figure 258.— Conventionalized alligator designs illustrating the process of sparing the

pattern out of the field (see fig. 257). ''• 157

Figure 259.— Large fine clay stool, the seat supported by three legs alternating with

strange half-human forms. Armadillo ware. '/• . . . 163

Figure 260. — Spindle-whorl consisting of a plain perforated disk. ''• . .164

Figure 261. — Spindle-whorl decorated with heads and arms in relief. '/» . 164

Figure 262.— Bell-shaped stamp; a, side view; b, view of lower end. Scarified ware. '/• 164
Figure 263. — Needle-case. Lost color ware. '/> ...... 164

Figure 264.— Needle-case ; from Divala. Lost color ware. '!• . 164

Figure 265. — Human figurine representing mother and infant. Alligator ware. '/> . 165
Figure 266.— Human figurine. Alligator ware. 'I' ...... 165

Figure 267 ; a. — Seated human image holding a shallow bowl. Alligator ware. *'«

£.— Detail of double-headed alligator decorating the right arm; c, a?.— Detail

of one end of the panel decorating the legs. '/• . . . . 166

Figure 268.— Human figure seated on a stool or metate. Alligator ware. */» . . 167

Figure 269. — Standing masculine figure; from Jacu. Alligator ware. 'I' . . 167

Figure 270. — Figurine representing the armadillo; a. — lateral-view; b.— ventral-view. '/> 168
Figure 271. — Perforated figurine representing the peccary (Dicotyles). Alligator ware. '/> 168
Figure 272.— Grotesque figurine representing the peccary. Alligator ware. '/» . . 168

Figure 273. — Gourd-shaped rattle. Lost color ware. '/>... . 169

Figure 274.— Double whistle. '/• .170

Figure 275.— Reed-shaped whistle decorated with alligator motives. Alligator ware. '/' 171
Figure 276.— Top-shaped whistle with two finger-holes; from El Banco. '/' . . 172

Figure 277.— Top-shaped whistle with three finger-holes. Alligator ware. (After

Holmes.) '/. .172

Figure 278 a, b. — End views of instrument shown in preceding figure. (After Holmes.) 172
Figure 279. — Whistle combining drum and bird form. Alligator ware. '/• • 173

Figure 280.— Drum whistle surmounted by grotesque biped life form. Alligator ware. '/• 173
Figure 281.— Drum surmounted by half-human figure combined with serpent and bird

forms. Alligator ware. '/' . . . . . . . .174

Figure 282. — Bird-shaped whistle. Alligator ware. '/'... . 175

Figure 283.— Bird-shaped whistle decorated with scale-group symbols. Alligator ware. 'I' 175
Figure 284.— Whistle representing an owl. Alligator ware. '/« . 175

Figure 285.— Owl-shaped whistle decorated with multiple body-line motive. Alligator

ware. V« .......... i . 175

Figure 286.— Whistle representing a crested partridge. Alligator ware. '/• . . 176

Figure 287.— Bird-shaped whistle; from Divala. Alligator ware. '/• . . 176

Figure 288.— Whistle representing a duck. Alligator ware. '/« . 177

Figure 289.— Whistle representing a bird in the act of alighting. '/• . . 177

Figure 290.— Simplified bird form of whistle. Lost color ware. '/• . . . 177

Figure 291. — Composite bird form of whistle. Lost color ware. *'• . . 177

Figure 292.— Small bird-shaped whistle. Lost color ware. '/• . . . .177

Figure 293.— Small bird-shaped whistle. Lost color ware. '''.... 177

Figure 294.— Whistle representing the crab, the back being decorated with alligator

motives. Alligator ware. '/•..... .178

Figure 295. — Whistle representing a crab. Alligator ware. '/< . . .178

Figure 296. — Whistle representing a round-bodied crab. Alligator ware. ''• . . 178

Figure 297.— Whistle representing a scorpion with alligator motives decorating the

sides. Alligator ware. '/« ..... .178

Figure 298.— Whistle in the form of a coiled snake. Lost color ware. */• . . 179

XVIII A STUDY OF CHIRIQUIAN ANTIQUITIES.

Page

Figure 299. — Composite form consisting of the iguana and snake, the sides being de-
corated with the multiple alligator motive. '/' . . . . .179

Figure 300. — Whistle in the shape of a quadruped. Lost color ware. '/• . . 180
Figure 301.— Whistle, in which the head has the appearance of being removable.

Lost color ware. '!• ......... 180

Figure 302.— Figurine serving as both whistle and rattle. Armadillo ware. '/• . 180

Figure 303. — Whistle representing the armadillo. Alligator ware. '/• . . 180
Figure 304.— Whistle representing a squirrel feeding, the sides being decorated with

the multiple alligator motive. Alligator ware. '/' .... 181

Figure 305.— Whistle representing a squirrel in motion; the alligator motive appears

in the panels decorating the sides. Alligator ware. '/« . . . 181
Figure 306. — Whistle suggesting the ground-squirrel, though the head is catlike. Alli-
gator ware. '/« . . . . . .. . . . 181

Figure 307.— Whistle with tapir-like head. Alligator ware. '/« . . . 182

Figure 308. — Whistle evidently representing the young of Tapirus bairdii. Alligatorware. '/' 182
Figure 309.— Whistle representing a species of tiger-cat, the head also serving as a

rattle. Alligator ware. '/« ........ 182

Figure 310.— Whistle representing a double-headed jaguar, both necks being perforated,

the heads serving as rattles. Alligator ware. '/« . . . . .183

Figure 311.— Jaguar-shaped whistle with ocellated markings of the jaguar. Alligatorware. 'I' 184

Figure 312. — Whistle in the shape of a deer. Alligator ware. "'« ... 184

Figure 313.— Whistle representing a deer apparently making an attack. Alligatorware.3/' 184

Figure 314.— Monkey-shaped whistle with three finger holes. '/• - . . . . 184

Figure 315.— Figurine serving as a whistle and representing a mythical form with

mixed attributes. Alligator ware. '/• . . . . . . .185

Figure 316.— Figurine serving as a whistle. The pose is human, the right arm con-
verted into a mouth-piece. Alligator ware. '/< . . . .186

Fjgure 317.— Whistle presumably representing a monkey, although the pose is avian.

Alligator ware. '/• . . . . . . . . .186

Figure 318.— Whistle figurine representing a three-headed monster with human attributes ;

the multiple alligator motive encircles the body. Alligator ware. ''• . .186
Figure 319.— Whistle representing a carefully modeled human head. Alligator ware. '/« 187
Figure 320.— Whistle representing the human head. Alligator ware. '/« . . 187
Figure 321.— Diminutive tripod with bird-shaped whistle attached to the side. Arma-
dillo ware. >u .......... 187

Figure 322.— Small cup with false whistle attached to the shoulder. '/• . . . 188

Figure 323.— Figure of a frog carved in resin; from Divala. Lamson collection. '/• 195

Figure 324.— Illustration showing primitive process of casting gold images. (After de Bry.) 196

Figure 325.— Needle of nearly pure copper. '/> ...... 197

Figure 326.— Copper tweezers perforated for suspension. '/« .... 198

Figure 327.— Small copper bell fashioned like the modern sleigh-bell. '/• . . 198

Figure 328.— Base metal bell representing the human head. Lamson collection. '/• . 198
Figure 329.— Base metal casting representing a frog, the head of which is adorned

with two conventionalized alligator heads. Lamson collection. '/> . . 198
Figure 330.— Figurine in base metal, to each knee of which is attached a conventional-
ized alligator head. (After Holmes.) . . . . . . .199

Figure 331.— Animal figure in base metal plated with gold, from the mouth of which

project two conventionalized alligators. (After Holmes.) . . . 199
Figure 332.— Gold figurine ornamented with alligator motives. Metropolitan Museum,

New York. '/• . . 200

LIST OF ILLUSTRATIONS. XIX

Page

Figure 333. — Bead evidently made from a gold nugget. '/' .... 201

Figure 334.— Gold bead. '/• 201

Figure 335. — Cylindrical bead made of sheet gold. 'A ..... 201

Figures 336-338.— Cylindrical gold beads. 'A 201

Figure 339. — Gold ornament, to be worn as a labret or earring. Heye collection. 'A 201
Figure 340. — Fine gold bell surmounted by an animal form ; the pellet inside is a

gold nugget. 'A .......... 201

Figure 341. — Gold figurine of a fish. Metropolitan Museum, New York. '/< . . 202

Figure 342. — Figurine of a deer in base metal plated with low-grade gold. Heye

collection. 'A .......... 202

Figure 343. — Small gold figurine of the armadillo. Heye collection. 'A . . 202

Figure 344.— Gold figurine of a frog; a. — dorsal view; b. — profile view. 'A . . 202

Figure 345. — Two united images of the frog in base metal plated with gold ; from

El Banco. 'A 203

Figure 346. — Two gold figurines of the frog united by bands of twisted wire. 'A . 203

Figure 347. — Gold image of a frog; ceramic type. 'A ..... 203

Figure 348. — Gold image of a frog. Metropolitan Museum, New York. '/« . . 204

Figure 349. — Simplified alligator figurine in low-grade gold. 'A . . . 204

Figure 350.— Gold image of an alligator holding in its mouth a human leg; from

Pueblo Viejo. Heye collection. 'A ....... 204

Figure 351. — Large massive gold image of an alligator. New York Public Library

(Lenox Foundation). 'A ......... 205

Figure 352. — Bat-shaped image in fine gold. New York Public Library (Lenox

Foundation). '/• 205

Figure 353. — Large image of a bird in an alloy of copper, gold and silver. '/« . 206

Figure 354.— Gold figure of a bird, the head surmounted by snake heads. '/• . 207

Figure 355.— Gold figure of a bird, the tufts being made of coiled wire ; a.— ventral

view; b.— profile view. 'A ........ 207

Figure 356. — Gold image of a bird with large wing-shaped tufts, holding a small rep-
tile in its beak. Heye collection. 'A . . . . . . . 208

Figure 357. — Gold figure of a bird with a fish held in its beak and with conventional-
ized alligator heads serving as tufts. Heye collection. 'A ... 208
Figure 358.— Small bird figurine in gold, its head adorned with two conventionalized

alligator heads. Heye collection. 'A ....... 208

Figure 359. — Small gold figurine of a bird ; a. — dorsal view ; b. — profile view. 'A . 209

Figure 360.— Gold figure of the jaguar. 'A ....... 209

Figure 361. — Gold image of an apelike monster with two heads. 'A . . 209

Figure 362. — Small gold image of an ape. Heye collection. 'A . . . . 209

Figure 363. — Seated human figure fashioned from a gold nugget. 'A . . 213

Figure 364. — Standing human figure fashioned from a gold nugget. 'A . . . 213

Figure 365.— Image of "pure gold" representing the alligator-god. (After Holmes.) 214

Figure 366. — Gold figure of the alligator-god. Heye collection. 'A ... 215

Figure 367.— Image of gold representing the alligator-god; from the Huacal de los

Reyes, Rio General, near Terraba, Costa Rica. Keith collection. '/• . . 215

Figure 368.— Gold figure of the alligator-god; from the Huacal de los Reyes, Rio

General, near Terraba, Costa Rica. Keith collection. '/• ... 215

Figure 369.— Gold image representing the parrot-god. Heye collection. 'A . . 217

Figure 370. —Gold figure of the parrot-god. Heye collection. 'A ... 218

Figure 371. — Gold figure of the parrot-god. Heye collection. 'A ... 218

XX A STUDY OF CHIRIQUIAN ANTIQUITIES.

Page

Figure 372. — Gold image of the parrot-god with alligator motives at the head and

feet. Keith collection. '/• ........ 219

Figure 373. — Image of the parrot-god in gold, ornamented with alligator motives at

the head and feet. Keith collection. '/» . . . . . .219

Figure 374. — Gold image presumably of the jaguar-god. Heye collection. '/» . 221

Figure 375. — Gold image of the jaguar-god, the perforated bars at the head and feet
being alligator motives ; from the Huacal de los Reyes, Rio General, Costa
Rica. Keith collection. '/• ........ 222

Figure 376.— Gold figurine representing the jaguar-god with an alligator motive serving

as head dress ; from Rio General, Costa Rica. Keith collection. '/« . . 222

Figure 377. — Gold figure of the jaguar-god with the alligator motive serving as head-
dress; from Mercedes, Costa Rica. Keith collection. '/' ... 222
Figure 378.— Rare gold mask representing the human face. '/' . . . 224
Figure 379.— Large gold plaque with five embossments representing the human breast

Heye collection. V» ......... 225

Figure 380.— Gold plaque. Heye collection. '/»... . . 226

Figure 381.— Dorsal-view (alligator) motives from vases of the lost color ware. . 230

Figure 382. — Dorsal-view motives as seen on the white line ware. . . . 230

Figure 383. — Dorsal-view motives from vases of the alligator ware. . . . 230

Figure 384. — Dorsal-view motives from vases of the polychrome ware. . . . 231

INTRODUCTION.

History. — Chiriqui is the name of a lagoon, river, volcano and province, the
latter being the westernmost province of the new Republic of Panama ; the term
therefore has no connection whatever with Cherokee with which it is sometimes
confused. The word is evidently of Indian origin, the meaning of which I have
been unable to trace. It appears in the works of Oviedo1 as Cheriqui and has
since suffered little change in spelling. In this respect it has fared better than
many early geographic names. The same may be said of Panama, a word that
means "abounding in fish," and that has been retained as the name of a town
since 1519, also of a bay and of the Isthmian region, although the political fortunes
of the latter have been checkered indeed. Among its early governors were Diego
de Nicuesa, Vasco Nunez de Balboa, who was the first white man to cross the
Isthmus (1513), and Pedro Arias de Avila, commonly known as Pedrarias Davila,
who founded the city of Panama (1519). The Isthmus was incorporated under
the vice-royalty of New Granada in 1718 and, with the exception of the period
of independence from 1859-61, was a part of that South American country under
its changing titles, first as the New Kingdom of Granada (1719-1810), then as
Republic of New Granada, United States of Colombia, and Republic of Colombia,
respectively, until November, 1903, when it became an independent republic.

Panama was the first region in continental America to be settled by Europeans,
and since 1535 has had a special importance because of the possibilities of a
trade route across the Isthmus. The value of such a route was seen by Oviedo,
the historian, who accompanied Pedrarias Davila to the Isthmus in 1514. The
following is taken from The [natural] hystorie of the vveste Indies, by Oviedo y
Valdes : 3

"And if therfore this nauigation may be founde in the South sea for the trade of
spices (as we trust in God) to bee brought from thense to the sayde porte of
Panama (as is possible enough) they may afterwarde easly passe to the Northe
sea notwithstandynge the difficultie of the waye of the XX. leaques aforesayde.
Whiche thynge I affirme as a man well trauayled in these regions, hauynge twyse
on my feate passed ouer this strayght in the yeare. 1521. as I haue sayde. It is
furthermore to be vnderstode, that it is a maruelous facilitie to bryng spices by
this way which I wil now declare. From Panama to the ryuer of Chagre, are
foure leaques of good and faire way by which cartes may passe at pleasure by
reason that the mountaynes are but few and lyttle, and that the greateste parte
of these foure leaques is a playne grounde voyde of trees. And when the cartes
are coomme to the sayde ryuer, the spices may be caryed in barkes and pinnesses.
For this ryuer entereth into the North sea fyue or. VI leaques lower than the port

1 Gonzalo Fernandez de Oviedo y Valdes. Historia general y natural de las Indias,
IV, 108, 1855.

2 The first three books on America; transl. by R. Eden, from the Latin of Peter Martyr
of Anghierra; ed. by Edward Arber, 234, Birmingham, 1885.

MKMOIRS CONN. ACAD , Vol. III. 1

2 A STUDY OF CHIRIQUIAN ANTIQUITIES.

of Nomen dei, and emptieth it selfe in the sea nere vnto an Hand cauled Bastimento.
where is a very good and safe port. Yowr maiestie may now therfore consyder howe
great a thynge and what commoditie it may be to conuey spices this way, foras-
much as the ryuer of Chagre hauyng his original! only two leaques from the South
sea, contineweth his course and emptieth it selfe into the other North sea. This
ryuer runneth fast and is very greate, and so commodious for this purpose as may
be thowght or desyred. . . . But to returne to speake sumwhat of the conueying
of spices, I say that when it shal please almighty god that this nauigation afore-
sayde shal bee founde by the good fortune of yowre maiestie, and that the spices
of the Ilandes of the South sea (which may also bee otherwyse cauled the Ocean
of the East India in whiche are the Ilandes of Molucca) shalbe browght to the sayd
coaste and the porte of Panama, and bee conueyd from thense (as we haue sayde)
by the firme lande with cartes unto the ryuer Chagre, and from thense into this
owr other sea of the North, from whense they may afterward bee browght into
Spayne, I say that by this meanes the vyage shall bee shortened more then seuen
thousande leaques."

It was, in fact, the belief in the existence of a short trade route to the Far
East that led to the discovery of America. Although called the discoverer of
the New World, Columbus saw very little of the mainland — only a short stretch
of the South American coast near Trinidad island on his third voyage ; and
Central America and Panama, on his fourth and last voyage. It was while searching
for the straits that might lead him to that part of East India already known to
geographers that Columbus entered the bay which forms a part of Chiriqui lagoon
and which was named for the great Admiral (Bahia del Almirante'). He had fol-
lowed the coast from a point off northern Honduras, opposite the island of Guanaja,
stopping at a number of places before reaching the coast of Chiriqui. The story
of what he did and saw is full of interest, as being the first historic account of
a land whose antiquities inspired the present study. On the 25th of September,
1502, " the expedition cast anchor off a certain island called by the natives Quiriviri,
by Columbus Husita, at more than one and a half leagues distance from Cariay,
a population situated along the banks of a great river (probably Sari Juan de
Nicaragua). The natives came running in great numbers attracted by the strangers.
They were armed with bows, arrows and darts made of black palm and pointed
with strong fish spines. Others carried long lances and macanas [wooden weapons
generally edged with sharp flint], and all were nude except about the loins which
were wrapped in white and red cotton cloth. The men with long hair tied at the
back of the head, and the women with the hair trimmed (cortado}. Some wore
plates of low-grade gold (guanin) and others w7ore jewels of the same metal sus-
pended about the neck."1

Before leaving this region, now called the Mosquito Coast, Columbus took on
board two of the natives as guides. Sailing on the 5th of October, he soon reached
what is now known as Boca del Toro, one of the entrances to Almirante bay.
" In one of the ports of these islands in Almirante bay were anchored twenty
canoes. The natives went about nude with their bodies painted in colors, wearing

1 Transl. from J. Acosta. Compendio historico, etc., 2, Bogota, 1901.

HISTORY. 3

plates of fine gold suspended from the neck. These were the first examples of
pure gold the Spaniards had seen along that coast, and which also caused them
to commit the injustice of taking from two natives the ornaments which the latter
had refused to exchange for Spanish trinkets, a fact which proves that traffic by
force had some early precedents in the history of America."

Diego de Porras,1 who accompanied Columbus, is quoted as saying : " The number
of pieces of gold which we obtained, great and small, was 220, weighing alto-
gether 9 marcos, 3 ochavas, 7 tomines, and 1 grain (equal to about 80 ounces).
There were also 12 pieces of guani (gold much alloyed), weighing one mark and
six ounces. Two pieces were not weighed - - one a large tube, and the other a
plate like a mirror." Specimens of the latter class were worn suspended on the
breast by chiefs and important personages, who refused to part with them.

All the inhabitants of the islands as well as of the mainland assured the
Spaniards that gold was to be found at a number of localities inland, one of
which was called Veragua. This name seems to have become the synonym of
wealth in the minds of the explorers, and has ever since been linked with the
fortunes of the house of Columbus. His grandson Luis Columbus, after having
been forced in 1536 to give up all claims to the title of viceroy, received in return,
among other things, an estate of twenty-five leagues square in Veragua and the
title Duke of Veragua. The land grant was finally given up after several
unsuccessful attempts to found settlements ; but the dukedom of Veragua still
exists, the title having come down through Francesca, a niece of Luis Columbus.
During the colonial period, Veragua (or Veraguas) corresponded to the western
part of the Isthmus, including the territory of Chiriqui (as well as of Burica).

The 17th of October, Columbus continued his voyage as far as to the mouth
of the Guaiga river. Hundreds of natives rushed to the shore brandishing their
arms and threatening to prevent the Spaniards from landing. Finally assured of
the peaceful mission of the latter, they became quiet and somewhat reluctantly
consented to exchange their gold plaques, etc., for Spanish trinkets.

Columbus next anchored in another river a few leagues away, called Cateba.
The inhabitants here were also warlike at first, but later bartered away their gold
ornaments for hawks'-bells, beads, etc. The Spaniards observed two things that
struck them as being worthy of record : One was that the chief, who here as
well as almost everywhere along the coast was called Quibi, was distinguished
from the other natives only by the fact that he protected himself from the con-
stantly falling rain (it was in October) by a great leaf of a tree; the other was
a piece of mortar, the first they had seen in America.

Continuing his voyage, Columbus touched at the mouths of the Cobrara and
Cubiga rivers. According to the Cariay Indians, no gold was to be found beyond
the Cubiga. A few leagues further on, Columbus entered a port to which on
account of its beauty he gave the name Puerto Bello. This is but a short
distance east of Colon. They pushed still further eastward reaching in turn
Puerto de los Bastimientos and Puerto del Betrete. Finally deciding to return to

1 E. G. Squier. More about the gold discoveries of the Isthmus. Harper's weekly,
Aug. 20th, 1859.

4 A STUDY OF CHIRIQUIAN ANTIQUITIES.

Veragua in search of gold, they retraced their course, reaching the mouth of the
Belen river in January, 1503. Ascending this river and the Veragua and Urira
rivers nearby, they exchanged Spanish trinkets for the gold ornaments of the
natives. Gold mines were found on the head-waters of the Urira. The brother
of the Admiral extended his explorations in the direction of Boca del Toro as
far as Cateba. He returned with so much gold obtained from the natives that
Columbus wrote to his rulers that he had seen more gold on the coast of Veragua
in two days than on the island of Cuba or Haiti in four years. Columbus and
his brother decided to found a colony on the. Belen river near its mouth. The
place was soon abandoned, however, on account of the hostility of the natives
and, after following the coast as far eastward as the Gulf of Darien, Columbus
set sail for Haiti. More than four centuries have elapsed since then, and his
dream of a channel leading to the mainland of Asia is soon to be realized in the
completed Panama Canal.

The glitter of gold on the Isthmian shores was not long in attracting other
colonists. The Atrato river, flowing northward into the Gulf of Darien, was made
the boundary line between two provinces - - New Andalusia on the east and
Castilla del Oro on the west. The latter included not only the Isthmus but also
a considerable portion of the Central American region. Nicuesa, the first governor
of Castilla del Oro, attempted to found a settlement at various points earlier
visited by Columbus — Belen on the coast of Veragua, Puerto Bello, and Puerto
de los Bastimientos, renamed by Nicuesa, Nombre de Dios; but his plans were
everywhere frustrated by the pestilential climate and by the Indians, whose ear-
lier experience with the white man had tended to increase their hostility. In the
meantime, colonists from New Andalusia had formed a settlement on the western
coast of the Gulf of Darien, in Nicuesa's territory. Nicuesa, having been rescued
from Nombre de Dios, was invited to govern at this new settlement, but proved
unwelcome, was deported and was never heard from again. Vasco Nunez de Balboa
was elevated to the command. His exploits during the next few years form one
of the most thrilling chapters in the early history of Spanish America. Chief
among them was the crossing of the Isthmus and the discovery of the Pacific
Ocean in 1513. His outward as well as return journey was a triumph not only
of discovery, but also of diplomacy and conquest. The native chieftains every-
where paid him tribute, either voluntary or forced, in the form of gold treasure,
provisions, guides and laborers. After subduing two caciques, Careta and Ponca,
he next made a friendly visit to the territory of Comagre, ruled over by a cacique
of the same name. The dwelling of Comagre, " surpassed anything they had yet
seen for magnitude, and for the skill and solidity of the architecture. It was one
hundred and fifty paces in length, and eighty in breadth, founded upon great
logs, surrounded with a stone wall ; while the upper part was of woodwork, cu-
riously interwoven and wrought with such beauty as to cause surprise and ad-
miration. It contained many commodious apartments. There were store-rooms
also ; one filled with bread, with venison and other provisions ; another with va-
rious spirituous beverages, which the Indians made from maize, from a species of
palm, and from roots of different kinds. There was also a great hall, in a retired
and secret part of the building, wherein Comagre preserved the bodies of his

HISTORY. 5

ancestors and relatives. These had been dried by fire, so as to free them from
corruption, and afterwards wrapped in mantles of cotton, richly wrought, and inter-
woven with pearls and jewels of gold, and with certain stones held precious by
the natives. They were hung about the hall with cords of cotton and regarded
with great reverence, if not with religious devotion." * The son of Comagre gave
Balboa '• 4,000 ounces of gold, wrought into various ornaments " ; 2 from him, also,
Balboa first received intimation of the Pacific Ocean and the riches of its shores.
Before starting on his journey across the Isthmus, Balboa remitted to the King
15,000 crowns of gold, i. e., one-fifth of all collected to date under his jurisdiction.

The incidents of Balboa's expedition across the Isthmus in search of the southern
sea are replete with interest. In two days he was at the headquarters of Ponca,
who assured him of the truth of what Comagre's son had told Balboa concerning
a great sea beyond the mountains, and gave him several curiously wrought orna-
ments of fine gold which came from the shores of that sea. Setting out anew
on a toilsome journey of four days, during which time the expedition covered but
ten leagues, they encountered a warlike cacique, Quaraqua, the ransacking of
whose village yielded a golden booty. From this village an ascent of a few hours
brought Balboa to the crest of the mountains, from which he beheld the broad
waters of the Pacific.

Having called upon his followers to witness that he took possession of that
sea, and the lands bathed by its waters, in the name of his sovereigns, Balboa
began the descent to its shores. He soon encountered the valiant cacique,
Chiapes, whose courage vanished at the sight of bloodhounds and of the havoc
produced by the mysterious Spanish firearms. Heeding the advice of their Indian
guides, Chiapes " came trembling to the Spaniards, bringing with him five hundred
pounds' weight of wrought gold as a peace-offering, for he had already learnt the
value they set upon that metal. Vasco Nunez received him with great kindness
and graciously accepted his gold, for which he gave him beads, hawks'-bells and
looking-glasses, making him in his own conceit the richest potentate on that side
of the mountains."3

Accompanied by Chiapes, the Spaniards finally reached the salt waters of the
Pacific in a bay to which Balboa gave the name of San Miguel. The theatrical
and impressive ceremony of taking possession of " these seas and lands and coasts
and ports and islands of the south, and all thereunto annexed," being over, Balboa
and a company of picked men, in nine canoes manned by Indians, attempted to
explore the neighboring gulf (now called San Miguel). After being rescued from
the boisterous seas by taking refuge on a small island, they succeeded in reach-
ing the mainland near the dwelling of a cacique named Tiimaco. A midnight
attack on the village with guns and bloodhounds was successful, the fleeing In-
dians leaving behind an abundance of provisions, much gold and many pearls.
Through the mediation of Tumaco's son and a '' mutual exchange of presents, a
friendly intercourse was soon established. Among other things the Cacique gave

1 Irving. Life and voyages of Christopher Columbus, III, 118, 1892.

2 Ibid.

;l Ibid., 141.

6 A STUDY OF CHIRIQUIAN ANTIQUITIES.

Vasco Nunez jewels of gold weighing six hundred and fourteen crowns, and two
hundred pearls of great size and beauty." Tumaco also told Balboa " that far to
the south there was a country abounding in gold, where the inhabitants made use
of certain quadrupeds to carry burdens. He moulded a figure of clay to represent
these animals, which some of the Spaniards supposed to be a deer, others a
camel, others a tapir; for as yet they knew nothing of the lama, the native beast
of burden of South America."1

The last cacique on the Pacific coast to pay tribute to Balboa was Teaochan,
from whom he received gold, pearls and an abundant supply of provisions. Re-
crossing the Isthmus to the Gulf of Darien, the Spaniards passed through the
territory of a rich and powerful chief, called Poncra, whose village was ransacked
and gold obtained to the value of three thousand crowns. Here they remained
thirty days until rejoined by a detachment that had been left at the village of
Chiapes. From the cacique who accompanied this party presents were received
to the value of two thousand crowns in gold.

Balboa and his men next faced the redoubtable Tubanama, the most powerful
of the mountain chieftains and the one of whom the son of Comagre had spoken.
In a midnight attack, Tubanama was captured. The price of his ransom was gold
ornaments to the value of nine thousand crowns. Nothing, however, would induce
him to tell where the mines that produced this treasure were located. Balboa
instituted a secret survey which disclosed the presence of gold in such quantities
that he planned to found two settlements in the neighborhood.

Balboa's brilliant career was soon cut off by the arrival of the newly appointed
governor of Castilla del Oro, Pedrarias Davila, who did not cease to thwart and
persecute his gifted and highly successful predecessor, and who finally had him
beheaded on a false charge of treason ; but not until Balboa with incredible energy
and resourcefulness had caused to be transported over the mountains materials
and stores for two brigantines in which he was to sail over the waters of the
Pacific to the land of the Incas about which Tumaco had informed him.

After founding Panama in 1519, Pedrarias sent Espinosa to explore the Pacific
coast to the westward, placing at his command the very ships that had been
constructed under almost insuperable difficulties by Balboa. Arriving at the prov-
ince of Burica, west of the present Chiriquian boundary, Espinosa set out on his
return journey by land and was presumably the first Spaniard to cross that part
of Chiriqui now celebrated for its antiquities. One ship was sent farther up the
coast to a gulf called San Lucar, in Nicaragua. Espinosa first traversed the prov-
ince of Huista, evidently on Chiriquian territory. Here the Spaniards remained
for some time loading their ships with maize before sending them back to Panama.
These may have been anchored in the present port of David. They observed
that the people of " this province and of that of Burica, were almost exactly the
same in the fashion of their clothes, and in their customs. The women wore a
truss round their loins, as their clothing ; and the men were naked. The country
is fertile, with plentiful supplies of fish, and a great quantity of swine,2 which

1 Op. cit., 148.

'- They mistook peccaries for swine.

HISTORY; 7

were caught with large nets of stuff like hemp, called by the Indians nequen, the
meshes being a finger in breadth. These nets were fastened at the entrance of
a wood where there was a herd of swine, who came against the nets and were
unable to get through the meshes. Then the people called out, the nets fell over
the swine, and they were killed with lances, so that none escaped of those that
fell into the nets."1

Gil Gonzalez de Avila also coasted along these shores as far as the Gulf of
San Lucar. About the same time Francisco Hernandez de Cordova was sent by
Pedrarias to subdue and settle Nicaragua. He founded the cities of Leon and
Granada. Two of his captains, Francisco Campanon and Soto, objecting to his
acts, left overland for Panama where they arrived without horses and barefooted.
" They had passed the villages of the Indians at night, and taken provisions from
them. Thus they had reached the province of Chiriqui, which is between Burica
and Nisca, where there was a settlement which had been made by Captain Benito
Hurtado, by order of Pedrarias, called the city of Fonseca. . . . After these ten
Spaniards had passed through this city of Fonseca, the captain, with some followers,
set out in the direction of Nicaragua, whence the others had come. Thus the
settlement was abandoned ; for those who remained, seeing that their comrades
did not return, went after them to the Gulf of San Lucar." 2 Fonseca, presumably the
first settlement made on the Pacific coast of Chiriqui and so soon to be abandoned,
was probably at or near the present San Lorenzo on the Rio Fonseca (see map).

With Pedrarias there came Oviedo, the " first chronicler of the New World,"
and surveyor of the royal foundries, from whom we get a glimpse of the Pacific
coast tribes of the region from which our antiquities came. Passages from his
Historia General y Natural de las Indias? in which Chiriqui is first mentioned,
are as follows : u En la costa del Sur, en el golpho de Orotina, comien9a la lengua
de Nicaragua, e de alii discurre ha9ia Poniente ; e mas adelante 9inco leguas hay
un grand pueblo de chorotegas a la parte del Levante ; e ocho leguas al Poniente
de la dicha Orotina hay otro que se llama Coribia. E son los dos indios de otra
lengua apartada de todas las que se han dicho en esta historia : e alii traen las
mugeres bragas, e todo lo demas traen desnudo, e tambien en la provin9ia de
Cheriqui y en Judea ; pero Cheriqui ni Judea no son desta goberna9ion, sino en
la costa desde el golpho de Orotina al Oriente ha9ia Panama. En las islas del
golpho de Nicaragua 6 de Orotina todas las mugeres traen bragas ; e son choro-
tegas e lo mismo los de Nicoya, como esta dicho.

" Desde Nicoya a la parte del Oriente ha9ia Panama e Castilla del Oro e lo
demas son los ca9iques senores : e de alii abaxo al Poniente ha9ia Nicaragua son
behetrias e comunidades, e son elegidos los que mandan las repiiblicas. .

" La provin9ia de los Cabiores es a veynte 6 veynte e 9inco leguas de Cheriqui,
al Poniente en la costa del Sur ; e la provin9ia de Durucaca es junto a la de
Cabiores. En estas dos provin9ias hilan los hornbres como mugeres, e lo tienen
por cosa e offi9io ordinario para ellos.

1 Pascual de Andagoya. Narrative of the proceedings of Pedrarias Davila ; transl. by
Clements R. Markham, 24, London, 1865.

2 Op. cit., 37.

l! IV, 108, Madrid, 1855.

g A STUDY OF CHIRIQUIAN ANTIQUITIES.

" La provincia que los espanoles llaman Judea, llaman los indfos Barecla, la qua!
confina con Cheriqui y esta en la mesma costa del Sur, seys leguas mas al Poniente
de la dicha Cheriqui : llamaronla Judea, porque es la gente de alii muy vil e sugia
e para poco."

Southern Chiriqui, the region which has furnished practically all the antiquities,
is to-day almost as difficult of access as it was in the time of Espinosa, Campanon
and Oviedo. In order to reach it one must still cross the Isthmus and go by boat
from Panama, a distance of 300 miles around the peninsula of Azuero, to the port
of David. With the completion of the Panama Canal and the development of the
more direct routes over the mountains from Bocas del Toro and by way of Costa
Rica this region will become better known commercially as well as archeologically.1
It first became prominent as a field for archeological research with the discovery
of the golden treasure at Bugavita in 1858-9.

Cemeteries and Tombs. — A mountain chain divides Chiriqui into two nearly equal
parts. As has been intimated, the ancient cemeteries or huacals, as they are called,
in Spanish America are practically confined to the southern half, that is to say,
the Pacific watershed. In fact, Dr. Merritt, at one time director of a gold mine in
Veragua, had never " heard of any such burial grounds on the northern side of
the Isthmus, from the lagoons of Chiriqui to the valley of the Chagres — and where
they would have been discovered by the gold-seeker, who has been ransacking
this section for more than 300 years." Nevertheless, it is probable that this region
will yet yield a rich archeological harvest when it is explored as thoroughly as a
part of the Atlantic coast of Costa Rica has been. On the other hand, ancient
cemeteries are met with everywhere on the Pacific slope, from near the mountain
tops to within a short distance of the sea. An idea of their number and distri-
bution may be had from the account of Thomas F. Meagher,2 who crossed the
Isthmus from David to Bocas del Toro fifty years ago :

" A mile outside Dolega the party stopped at the house of Don Roberto Soes,
the discoverer of the golden relics in the Indian graves of Chiriqui. All the way
from David we had ridden through thousands of these disemboweled and ransacked
graves, and in every direction, for leagues and leagues, from Terraba and Boruca
to Santiago de Veragua, we might have seen tens of thousands more."

One of the best known huacals is that of Bugavita (near Bugaba), where so
many gold ornaments were found in 1858-9. Dr. Merritt's description of this
cemetery is quoted at length because of its excellence and of the fact that copies
of his paper are extremely rare : *

" The Huacal of Bugaba embraced an area of twelve acres, but was divided

into two sections — by a slight depression extending in an east-and-west direction

- in which not a single grave has been encountered. This depression of the

surface varied in width from eighteen to ten yards, toward the east. The two

1 A railroad is now being built from Panama to David.
! The new route through Chiriqui. Harper's mag., XXII, 198, 1861.

3 J. King Merritt. Report on the huacals, or ancient graveyards, of Chiriqui ; publ. by
the Amer. ethnol. soc. previous to vol. I of its Bulletins.

CEMETERIES AND TOMBS. 9

sections were respectively five and seven acres, and were located on slightly
elevated eminences, about four hundred yards from a small river, the course of
which is northeast by east at this point. The rise from the river banks to the
Huacal is very gradual, except at the northern boundary, which is somewhat
abrupt, and around which the river turns toward the east. The general direction
of the Huacal is north and south; and the greater portion of the graves were
found on the western and southern slopes. There did not appear a general reg-
ularity in the position of the Huacas, or graves, but frequently there would occur
several side by side. The distance between the grave-pits varied from nine to
fifteen inches at the more crowded portions of the Huacal. The universal direction
of the quadrangular Huacas is north and south by the polar star.

" There are two forms of Huacas or graves - - the oval and quadrangular ; and
their mode of construction is an interesting matter for consideration. I, with a
small party of peons, examined carefully and systematically one of the former and
several of the latter description; and besides, saw a large number of both varieties
opened by parties in the vicinity. The material uniformly employed in the con-
struction of the sepulcher proper was flat and rounded river stone. The oval
grave-pits were from four and a half to six feet deep, and from three to four feet
in their largest diameters. A wall of the rounded river stone, two and a half to
three feet high, lined the grave-pit at the bottom, after the manner of a modern
well. From the top of this wall to the surface the entire area of the grave-pit
was closely packed with rounded river stone. Within the limits of the wall, which
seemed to be the tomb proper, were found principally the golden figures, and the
vessels of pottery, etc. The greater portion of the oval or circular Huacas were
located in the northern and western sections of the burial-ground ; and, as a
general rule, yielded the most figures of gold and the finest specimens of pottery. '
The relics in these were found usually at the eastern and northern sides ; and the
gold figures sometimes were located, it is said, in the crevices of the wall — but
in no instance in the earthen jars associated with them. The circular graves,
being confined to the slopes of the Huacal, were more or less covered by the
wash from the elevated sections of the Huacal, so that the top stones of the
package were in many cases nearly a foot beneath the surface. No vestige of
the human body was discovered in the oval Huacas ; but a black loam occupied
the spaces between the relics and the stone package. Occasionally earthen vessels
were found in the stone package near the surface. The quadrangular Huacas
were constructed in two modes. In one case the grave-pit was lined by walls
of rounded river stone about one third the distance to the surface ; and from the
top of these walls the entire area of the grave-pit was closely packed with river
stone, as in the oval grave. Within the limits of these walls, and in close proxim-
ity to the bottom of the grave-pit, were found most of the relics contained in
them. These Huacas were larger and yielded more gold images and finer pottery
than the other variety of the quadrangular grave, and were in juxtaposition with
the oval graves, occurring interspersed with them in the northern and western
sections of the Huacal, but abounding principally in the southern portion of
the ground. Some of these Huacas were nearly six feet deep, especially those
situated in the depressed sections of the Huacal, and the area of the grave-pit
MEMOIRS CONN. ACAD., Vol. III. 2

10 A STUDY OF CHIRIQUIAN ANTIQUITIES.

measured frequently seven by four and a half feet. In these the relics were
usually found near the bottom, at the northern and southern extremities, and
more or less on the eastern side. The gold figures most frequently were located
about one fourth of the distance from the head to the foot of the grave-pit, and
in the medial line. In these Huacas, also, earthen vessels were encountered in
the stone package near the surface, and generally at either extremity.

" The other variety of the quadrangular Huaca, although poor in relics, was
more artistically and carefully constructed, and in a better state of preservation ;
for in many of them everything was encountered ' in situ."1 In these a vault existed,
which was formed of flat river stone, and was of the requisite size to contain a
human body in the supine position, so far as the length and breadth are considered,
but in the height giving ample space for the introduction of the earthen relics.
These Huacas were located mostly on the more elevated portion of the Huacal,
and in the southern and eastern sections. A grave-pit had been sunk about three
feet deep and six and a half by four feet in area ; and then a lesser pit, by eight
to ten inches on all sides, was farther sunk to the additional depth of about two
feet. This smaller pit was lined by flat stones placed edgewise, which were held
in position by other flat stones resting flatwise upon the upper edges of these
and the surface of the recess in the sides of the grave-pit. The floor of the vault
was not paved, but presented a hard, pebbly clay surface, with two depressions
frequently, one at either extremity, and corresponding with the probable positions
of the occiput and the heels of the body. The cover to the vault was composed
of flat stones, quite closely adjusted, and sufficiently overlapping the sides to be
firm, and not liable to cave in from the variations of position of the sides, consequent
upon subsequent settlings of the earth and stone from natural causes. From the
cover of the vault to the surface, the entire area of the grave-pit was closely
packed with river stone, somewhat larger than ordinary paving-stone. In these
Huacas the relics were also found mostly in the vault, at the head and foot, and
on the east side. So far as my experience goes, earthen-ware only was found in
these, although I was told that a few had contained the smaller gold figures.
Specimens of pottery were encountered also in these, at either extremity of the
stone package, immediately above the cover of the vault. The location of these
vaulted Huacas, being the more elevated section of the burial-ground, and the
vaults also preserving their outlines, the stone package of most of them was
slightly raised above the general surface. A black loamy earth occupied all the
graves proper, or the original position of the body in it.

" It would seem, from the facts I have stated, that the gold-bearing graves were
those which were ruder in their construction, and which occupied, to some degree,
a particular section of the Huacal, although the limits of this were not well defined.
The golden ornaments were not found in many successive Huacas, even at the
richest points of the Huacal. Pottery, however, was encountered more or less in
every Huaca. It is reported that in other Huacals, in the vicinity of Bugaba, traces
of human hair have been discovered ; but in those of Bugaba, which I explored
carefully, no such evidence of the body was encountered. I have, however,
the enamel of a molar tooth, taken from a grave in a Huacal, near that of
Bugaba.''

CEMETERIES AND TOMBS. 11

Mr. John F. Bateman. a companion of Dr. Merritt, explored a number of huacals
in the highlands near the Volcano of Chiriqui. Some of these were on the head-
waters of the river Caldera, east of the Volcano. At one place about three hundred
circular graves were found. Nearby was a huacal some twelve acres in extent and
completely covering the elevated portion of a pasture (potrero). The graves were
so close together " that in excavating one we would open three or four others.
These were all regularly built sepulchres ; the body having been laid on the hard-
pan or clay, the sides formed of flat stones, and these covered with large flat
stones, many of which would measure a yard square. In these graves, and those
in the adjoining forest, which vary in depth from three to four feet, are found the
same river stones. In the forest are found additional stones, — quadrates, of four
inches by twenty inches in length. These were placed vertically, thirty inches
apart, around the edge of the quadrangular graves.

u In this locality I witnessed the opening of a large grave about ten feet in
depth, marked by five round pillars of stone, of fourteen inches in diameter and
from five to six feet in length, — three to four feet of which were in the ground.
The pillars were placed to represent a square with one in the center. Under this
one, on the clay, was found a plate of gold, four inches in diameter, and the
small figure of an anteater. No pottery was found in this grave, and this was the
only one in that locality containing gold. Stone hatchets were found here, but
no musical instruments. The pottery was all small, and rude in shape and material.
Small basins standing on three feet, each [foot] of which contained a small pellet
of clay. None of the pottery was either glazed or painted." l

Bateman also found a huacal north of the Volcano, presumably near the source
of Rio Chiriqui Viejo. Following the ridge for a mile, he came upon the graves
he had long wished to see — " those marked with pillars of basalt, mossgrown and
bearing marks of extreme age." The grave opened by Bateman was marked by
four pillars in a line ; the one at the south end was large, and the other three,
at intervals of thirty inches, were smaller. Large quantities of river stone of all
sizes were found in this grave, also great quantities of broken pottery, different
in quality from any previously discovered. " It was very thick and finely glazed2
on both sides." The condition of the clay at a depth of five feet pointed to a
double burial, the position of the bodies being marked by black loamy earth alone.

De Zeltner, from whom the Yale Museum obtained some of the finest specimens
of pottery, speaks of six types of Chiriquian graves: (1) The oval cist; (2) quad-
rangular grave, with walls and roofing of stone, one meter wide by about two
and a half meters long ; (3) so-called fortified tombs, deeper than the two preceding,
quadrangular in form, with a square stone pillar at each corner and a fifth stone
pillar of smaller size at the center; (4) tombs with roofing of flag stones and
provided with four pillars one and a half meters high. Between the pillars the
walls are built up of rounded stones ; (5) tombs with roofing of earth, according
to de Zeltner,3 the most common type of all and described by him at length :

1 J. F. Bateman. Bull. Amer. ethnol. soc., I, 28, 1860--61.

2 He evidently employs the term glaze in the sense of slip.

3 A. de Zeltner. Note surles sepultures indiennes du departement de Chiriqui, Panama, 1866.

12 A STUDY OF CHIRIQUIAN ANTIQUITIES.

" Apres avoir ecarte les obstacles accumules par le temps et la vegetation, on
decouvre au niveau du sol un pavage, en forme de carre long, de trois a quatre
metres de long sur une largeur de deux ou trois, et de quatre-vingts centimetres
d'epaisseur en general; a chaque angle, un pilier carre en pierres, de 1.60 de
hauteur et de 0.25 au plus d'epaisseur. Au centre, a distance egale des quatre
angles, se trouve un autre pavage carre, correspondant au premier. Celui-ci
recouvre une voute en forme de citerne, d'un metre de long et de quatre-vingts
centimetres a peu pres de large. Le sol est empierre, et c'est la veritable entree
de la guaca. A deux metres a peu pres de profondeur se trouve enfin le tombeau,
dont la distance du premier pavage est en definitive de quatre a cinq metres.
Cette sorte de chambre a pres de trois metres de long, deux de large ; et sa hauteur
n'est guere que d'un metre et demi. Elle est creusee dans le sol sans etre revetue
d'aucune espece de ma9onnerie. Sa forme est pyramidale, et la pente commence
a 1' entree pour aller en diminuant jusqu'au fond " ; (6) canal tombs, recognized
by a surface packing of river stones, two meters long by one meter in width.
Beneath this the tomb is sunk in the earth and is not lined with stones of any
kind.

Mr. J. A. McNiel, who collected most of the specimens figured in this work,
witnessed the opening of many graves. He says there are seldom surface indi-
cations to attract the inexperienced eye, the location of the grave being ascer-
tained by thrusting a light iron bar into the ground till it strikes the stone roofing
over the vault. The latter varies in depth from one to five meters. Even after
allowing for surface wash and fill, there must have been considerable variation in
the original depth of the graves, as some of the deepest ones were so situated as
to preclude the possibility of subsequent fill. According to McNiel, the flat roofing
stones rest upon round ones which form the walls of the vault. These were
evidently brought from river beds, and in many cases from long distances, as they
are more water-worn and much smoother than the stones of the locality in question.
He found no apparent orientation in a given direction, the graves, about a thou-
sand, in all the groups examined by him being " promiscuously strewn at much
cost of space."

M. A. L. Pinart,1 a later observer, states that the oval cists are the most nu-
merous ; and mentions a rectangular type of grave with walls and roofing of flat
stones. The human bones were placed on the floor of the chamber with no ap-
parent order, generally near the walls. The gold ornaments were with the bones,
while the pottery and stone objects were found at the center. He also speaks of
another class with rudely built chamber, whereas '• sur les cotes on avait creuse
dans la parois des niches, parfaitement garnies de dalles dans lesquelles etaient
deposes les cadavres ; chaque niche se fermait par une autre dalle." In such
tombs the artifacts of stone and pottery were placed in the chamber, the gold
ornaments, however, occurring only in the niches. Pinart insists that there are
but two kinds of graves in Chiriqui (and in all Panama), and is equally sure that
the bones were cleaned before being interred.

1 Chiriqui : Bocas del Toro, Valle Miranda. Bull. Soc. de geogr. de Paris, 7C ser., VI,
433, 1885.

CEMETERIES AND TOMBS. 13

In comparing the description of the graves as given by various observers, the
latter appear to agree in certain respects and differ in others. For example, all
agree that there were at least two generalized forms of graves — the oval and the
quadrangular. There is also agreement as to their dimensions and depth beneath
the surface. That one type of quadrangular grave was walled with flat stones
placed edgewise and covered with one or more flat stones, there seems to be
common accord. The floor of all the graves, both oval and rectangular, was
simply earth or hard-pan. Any attempt to further harmonize the descriptions of
the various observers, however, results in the accentuation of numerous contra-
dictions and discrepancies. Merritt describes a quadrangular type of tomb, with
walls of rounded river stone, resembling in every respect the oval cist except in
horizontal section. The only other author who mentions this kind of lining for
any but the oval graves is de Zeltner, yet he adds features not noted by Merritt,
such as a pillar at each corner and a roofing of flag stones. Bateman was pres-
ent at the opening of a grave, at each corner of which was a round pillar and
a fifth pillar in the center. This may have been the grave on which de Zeltner
bases his class number three, but he speaks of the pillars as being square instead
of round. Curiously enough, de Zeltner's fifth class, the one he calls the largest
in point of numbers, is not mentioned by any other writer. McNiel could discover
no apparent attempt at orientation. On the other hand, Merritt says the universal
direction of the quadrangular graves at Bugaba (Bugavita) is north and south by
the polar star. Bateman, also, was satisfied that the bodies were all placed north
and south.

Merritt says that, as a general rule, the circular graves at Bugavita yielded the
most figures of gold and the finest specimens of pottery ; and that the quadran-
gular cist lined with rounded stones contained more gold images and finer pottery
than the vaults built of flat stones, the latter, although more artistically and care-
fully constructed and in a better state of preservation, being poor in relics. In
referring to the same cemetery Bateman states that gold is found in some of the
graves, " while others in close vicinity, although containing more pottery, and
that of a higher order, contained no gold, — the richest graves having the least
pottery." According to McNiel, the oval grave-pits had very few artifacts of any
kind. His workmen informed him that where stone images were found, " it was
looked upon as indicating a rich grave in pottery and probably gold."

Mr. C. V. Hartman1 gives a careful description of graves examined by him in
the highlands and on the east coast of Costa Rica where the same two generalized
types abound — the oval and the quadrangular, the latter being the more abundant.
Both classes were often found in the same group; as was the case at Bugavita.
The two kinds resemble those of Chiriqui except that Hartman mentions a floor
of cobble-stones for the oval pits and of flat stones for some, at least, of the
quadrangular cists. The walls of the latter were either of flat stones set on end
or of cobble — stones, depending on nature's supply. As many of these rectangular
cists are shorter than the average human body reclining horizontally at full length,
the disposition of the remains in such graves points to the custom of interring

1 Archaeological researches in Costa Rica, Stockholm, 1901.

14 A STUDY OF CHIRIQUIAN ANTIQUITIES.

the bones or only a part of them instead of the corpse. Pinart believes that this
custom also prevailed in Chiriqui.

Human Remains. — The rarity of human remains in Chiriquian graves has often
been emphasized as indicating either the practise of cremation or a great antiquity
for the graves. Evidences of cremation, however, are lacking. That it would
require a long period of time for bones buried in a region of great rainfall to
decay completely, is by no means certain. Dr. Merritt points out that a black
loamy earth marked the position of the body in all the graves at Bugavita. No
human hair was found in these, but the enamel of a molar tooth, as well as
human hair, were obtained from neighboring cemeteries. In the double burial
noted by Bateman, " there were no signs of human remains, — only the black loamy
earth showing the original position of the body." He also mentions having seen,
" taken from a grave of the quadrangular kind, three teeth, a small piece of bone,
apparently a rib, and three pieces of the skull of a human body, but so fragile
that they crumbled at the touch and by exposure." Seemann states that human
skeletons are sometimes met with in the graves, but that they crumble into dust
on being removed. McNiel, who opened more than a thousand graves, found
human bones in some of them, but did not succeed in obtaining a single skull
even approximately complete. The disposition of the human bones in the graves
was such as to convince Pinart that the bones were buried after the flesh had
been removed. Another peculiarity recorded by the latter author is that gold,
when present, was always associated with the bones, whether the latter were
placed near the walls of the vault or in adjacent niches, while the artifacts of
stone and bone were not.

De Zeltner states that very friable fragments of human remains are sometimes
found in graves of the second, fourth and fifth types, as described by him, but
no trace of human bones is found in the other three types. He secured one
cranium sufficiently well preserved to be cast. A copy of this in plaster was
presented to the American Ethnological Society at its April meeting in 1860, by
Dr. J. P. Kluge and Mr. William Nelson l of Panama. The cranium is described
as " entire except the upper jaw, small for an adult, and rather broad in the
middle and flat behind." All efforts to trace this cast or its original have proved
unavailing.

Another Chiriquian human skull, collected by Dr. E. Menard,2 a physician in
the employ of the French Panama Canal Company, was given by him in 1890
to the School of Anthropology, Paris. It is described as having " un front bas
et retreci mais droit, avec bosse frontale saillante, un indice cephalique. 78.5."
Dr. Menard also speaks of an artificial deformation that would seem to indicate
some connection with ancient Peruvian skulls.

People. — The discoverers of the Isthmus all testify to its relatively large Indian
population. A century later (1606), the missionary, Melchor Hernandez, found as
many as six distinct languages spoken on and near the shores of the Chiriqui
lagoon by ten different tribes, as follows : Borisques, Bugabaes, Chagres, Cothos,

1 Hist, mag., IX, 158, 1865.

2 Les poteries des sepultures indiennes du Chiriqui ; chez Pichat a Chatillon-sur-Seine, 1881.

PEOPLE. 15

Dolegas, Dorasques, Dures, Utelaes, Zaribas and Zunes. The reduction has gone
steadily on since the advent of the European. According to Gabb, l " A strange
fatality seems to hang over these Isthmian Indians. Even when not brought into
contact with the debasing influences of civilization, the tribes are visibly dimin-
ishing." To other obliterating agencies must be added intermixture with the
blood of both whites and blacks.

Brinton2 draws the ethnographic boundary line between North and South
America at the mountain chain which separates Nicaragua from Costa Rica, and
the head-waters of the Rio Frio from those of the more southern and eastern
streams. " Beyond it we come upon tribes whose linguistic affinities point towards
the southern continent." Fernandez states that at the time of the discovery the
outposts of Nahuatl civilization did not reach farther south than Chiriqui lagoon,
while Uhle places the northern limit of Peruvian culture at Pasto. The region
between includes the present republics of Colombia and Panama, and forms a
linguistic and archeological barrier between the great civilizations of Mexico and
Peru. In this culture zone the dominant factor is Chibchan. The original home
of the Chibchas was on the plains of Bogota and Tunja. Brinton believes their
language to have been " much more widely disseminated throughout New Granada
at the time of the discovery than later writers have appreciated." Dr. Max Uhle's 3
important researches serve to confirm this view. The dialectic evidence points
to attrition and gradual loss of the original form as one proceeds from Colombia
through the Isthmus into Costa Rica, making it clear that the invasion was from
South America into North America, and not the reverse. Thus, even in Costa
Rica, the only tribes whose language shows no affinity with the Chibcha are the
Guetares and Orotinans, both belonging to the Chapanec linguistic stock of Chia-
pas, and the Guatusos, which, judging from the vocabularies already collected,
are an independent stock related neither to the Nahuatl nor the Chibcha. As
for Panama, the only possible breaks in the Chibcha linguistic chain of influence
are to be found among the Cunas or Coibas, who at the time of the discovery
occupied the territory from the Gulf of Darien and the Atrato river on the east
to the river Chagres on the west, and the Changuina-Dorasque stock of Chiriqui.
Uhle points out certain verbal similarities between the Cuna and Chibcha, while
Pinart, who has published extensively on the Cuna, notes affiliations with the
Carib. The present state of our knowledge would not warrant its classification
with either of these linguistic stocks.

The tribes occupying the province of Chiriqui in recent times are the Guaymis
and Dorasques. The Guaymis inhabit both slopes of the Cordillera and are
divided into three sub-tribes, each speaking a distinct dialect: (1) The Muois, of
whom only three were living in 1880, (2) the Moves or Valientes, and (3) Murires
or Sabaneros. The generic name Guaymi is from the Muoi dialect and means man.

1 Wm. M. Gabb. On the Indian tribes and languages of Costa Rica. Proc. Amer. philos.
soc., XIV, 491, 1875.

2 D. G. Brinton. The American race, 164, 1901.

8 Verwandtschaften und Wanderungen der Tschibtscha. C. R. Congres intern, des Ameri-
canistes, 466, Berlin, 1888.

16 A STUDY OF CHIRIQUIAN ANTIQUITIES.

Pinart * has every reason to believe that the Guaymis are the descendants of
the race that constructed the ancient huacals from which our Chiriquian an-
tiquities came. They have a tradition to the effect that before the arrival of the
Spaniards and even for a certain period after that event, they manufactured pot-
tery, but by reason of the greater durability of the iron pots and the ease with
which they could be procured, the art of pottery making was lost by degrees.
They were also metal workers in gold, copper and their alloys. On the occasion
of Pinart's visit the natives still possessed a number of gold ornaments which they
claimed to have inherited from their ancestors and which differed in no respect
from those found in the graves. In this connection it is of inferest to recall a
similar condition of affairs among the Tiribis of Costa Rica, as related by Gabb : 8
" The chiefs on great occasions wear gold ornaments, similar to those now found
in the Huacas of Chiriqui. Whether these have been recovered from some of
these graves, or whether they have been handed down from time immemorial, is
not known." Of the four or five seen by Gabb, two belonged to the reigning
chief and three represented birds, one of which was double, no doubt similar to
those in the Keith collection, recently found at Mercedes, Costa Rica.

Pinart describes the Guaymis as living in separate houses, the walls of which
are of bamboo or reeds, and the roofs, of palm leaves. The interior is divided
into small rooms by bamboo partitions, each member of the family having his
own room. The furnishings are simple - - a few rude hammocks and wooden
blocks for seats (see fig. 22). Kitchen utensils include metal pots of European
origin, a flat stone serving as a species of metate on which to grind cocoa and
maize, calabash plates and saucers, gourds for conserving water, a wooden mortar
and pestle for decorticating rice and other grains. Their arms consist of bows and
arrows, lances with points of hard wood, also a lance with several points for
fishing, and the inevitable machete. Formerly they used a small shield made of
tapir skin.

Their costume is simple. They paint the body. The male wears a simple
loin cloth made of bark (tiumf) ; the female, a band somewhat larger that descends
to the knees. When it rains, both sexes wear a large sleeveless mantle of bark
reaching below the knees. Ornaments include necklaces and bracelets of animal
teeth or of glass. During the grand ceremonies the chiefs wear diadems com-
posed of showy feathers, those of the quetzal being the most esteemed. Each com-
munity recognizes a hereditary chief.

According to Pinart, the Guaymis are of small stature with a tendency to cor-
pulence, and of a robust constitution ; color, yellowish brown to dark brown ; hair
black, stiff and glossy ; the head large in proportion to the body, long oval ; the
face particularly flat and broad between the zygomatic arches ; nose prominent,
often thick at the base ; mouth large and lips thick ; beard almost wanting and
the body devoid of hair. The same author speaks of artificial deformations of
the teeth, the canines being faceted so as to resemble saw-teeth ; also the ab-

1 A. L. Pinart. Bocas del Toro, Valle Miranda (avec carte dans le texte). Bull. Soc. de
geogr. de Paris, 7= ser., VI, 433, 1885.

2 Op. cit., 519.

PEOPLE. 17

sence among women of the upper left canine, which is knocked out at the time
of the first menstruation.

Among the Guaymis are found manifest traces of totemism, each tribe, family
and individual having its tutelary animal. Like American Indians in general, they
believe in spirits and animism, employing magicians (siikia) and making offerings
to appease evil spirits. The dead are carried far into the forest and placed on
a scaffold. After a year has elapsed an official goes to the place, cleans the
bones, binds them in a package and transports them to the family sepulcher.
Formerly they deposited with the dead all his possessions. The Talamancas of
Costa Rica dispose of their dead in a similar manner.

The Changuina-Dorasque stock formerly occupied the greater part of the prov-
ince of Chiriqui. By the middle of the XVIlIth century their limits were con-
fined to the plains of Chiriqui. In 1887 their number was reduced to thirteen
or fourteen persons of pure blood, living chiefly near Bugaba, Caldera and Dolega
(see map). As to the Dorasque tribe proper, the last member died between the
years 1882 and 1887. They were said to be lighter in color than the Guaymis
and also less cultured.

As between the Guaymis and the Dorasques. Pinart believes the former to be
the descendants of the ancient Chiriquians. Dr. Berendt1 would give the credit
to the Cunas or Coibas. but as he apparently included the Guaymis among the
Coibas, his conclusion does not differ materially from that of Pinart. Linguistic-
ally the Guaymis are more closely related to the Talamancas, the Terrabas and
the Borucas (or Bruncas) than to the Dorasques. Curiously enough the antiquities
from Boruca and Terraba have many points in common with those from Chiriqui.
That there was intercourse between the territory of the Talamancas (Mercedes)
and Chiriqui, is also attested by archeological evidence. The latter therefore
supports the claim of the Guaymis, who with the Borucas, Terrabas and Talamancas
belong to the Chibcha linguistic stock.

If. then, the language was influenced by migrations from the south, did these
also bring the dominant forces that molded the art of Chiriqui? Pinart thinks
not, but that the ancient art of Chiriqui was influenced more by Mexico than by
the South. The northern impulse might have been transmitted along the Pacific
Coast. It could have also come down the Atlantic side ; for when in 1564, Coronado
subdued the Guaymis and Talamancas, he also encountered, in the valley of Coaza
(Robalo ?), a stream that flows into the Almirante bay, the Chichimecs, now extinct.
In order to communicate with the chief of this tribe it was necessary to employ
a Mexican interpreter. The art of Nicoya, of eastern Costa Rica and of Chiriqui,
is certainly superior to that of the Isthmian provinces to the east of Chiriqui.

1 Dr. C. H. Berendt. Geographical distribution of the ancient Central American civiliza-
tion. Jour. Amer. geogr. soc., VIII, 132, 1876.

MEMOIRS CONN. ACAU., Vol. III.

MEM. CONN. ACAD., Vol. Ill

MAP

PROVINCE OF CHIRIQUI

s ° «> 3,° 3.Q 4,o s,°A///e.r

THE COLLECTIONS.

The Chiriquian antiquities belonging to Yale University number several thousand
specimens. The choicest of these were collected from 1859-1866 by M. A. de
Zeltner,1 French Consul at Panama, whose good fortune it was to be living there
at the time of the first discoveries. The major part of the collection, however,
was made by Mr. J. A. McNiel a few years later. The de Zeltner collection was
bought in 1872, after it had been shipped to Paris and a part of it already for-
warded to Heidelberg, where de Zeltner was soon to go as his country's repre-
sentative in the consular service. The purchase was made by Professor Othniel
C. Marsh, who also bought the McNiel collection in 1878. These collections
include gold objects, stone implements, metates, stools of stone and earthenware,
and a series of pottery representing various kinds of ware and unsurpassed in the
number of its rare and valuable specimens.

These treasures have been in storage for about thirty years, there being no
facilities for their exhibition, and only recently have they been accessible for pur-
poses of study. They form the central feature of one of the three principal col-
lections given to the University by Professor Marsh in 1898, viz., the " Collection
of American Archeology and Ethnology." 3

Of the priority of stone art over ceramic art, there can be no question, and it
is also generally admitted that the manufacture of pottery antedates a knowledge
of the use of metals. In discussing the three classes of artifacts, therefore, it has
been thought best to follow the foregoing sequence, although no attempt is here
made to fix the relative age of individual specimens, which would presuppose a
much more thorough knowledge of the various types of Chiriquian graves, as well
as of the character of their contents and that of the surrounding soil. It may
be worth while, however, to call attention to certain phylogenetic relations which
are traceable, not only through the various groups of a given class, but which also
bind the product of the stone worker to that of the potter and the artificer in metal.

That these phylogenetic ties lie between stone and pottery on the one hand
and pottery and metal objects on the other, rather than between stone on the
one hand and metal on the other, is significant as bearing on the general phylo-
genetic trend in the development of Chiriquian art as a whole. The subject
will be discussed further in describing the various specimens in the three classes
that best illustrate the influence of technique in one medium over that in another.
It is also significant that these ties which bind the art of all three classes
together are centered in the group of unpainted pottery, called by Holmes " terra
cotta" or "biscuit" ware, a group which probably stands for an early stage in
the development of Isthmian ceramic art. The question of the relative ages of
the various groups of potterv will be taken up in more detail when treating of
the " biscuit " ware, or armadillo ware as I prefer to call it.

1 Note sur les sepultures indiennes du departement de Chiriqui (Etat de Panama), Panama, 1866.

2 Yale alumni weekly, VII, Jan. 20, 1898.

STONE.

With the exception of its architectural manifestations, which are relatively in-
significant, the stone art of Chiriqui compares favorably with that of the regions
either to the northwest or southeast. It includes arrow-points, spear-points, celts,
polishing stones, pestles, metates or mealing stones, stools, images, ornaments,
petroglyphs and sculptured columns.

McNiel states that stone axes and other implements were found in every local-
ity visited— from Gualaca east of David to Divala on the west and El Banco
and Jacii on the north. As to their mode of occurrence, he says : " I have not
been able to discover any particular difference in the class and style of graves
in which the stone implements were found. Having obtained them from every
locality visited, I can say that the majority of those in my collection are from
Bugavita. But as a large portion of the pottery is also from that locality, I have
not thought the fact important. The three or four large stone axes of a different
form are from the extreme easterly limit of my work, Gualaca. These were brought
in to me by my assistant, with about the same number of cantaros (pottery vessels).

4> The stone images, statuettes, etc., are less frequently found ; and are gener-
ally found either on the surface, above the graves or a short distance below.
As I have no practical knowledge of my own in regard to these, having at no
time been present when any of those in my collection were found, I can only
state from information, which I believe to be in the main reliable. My inquiries
in that direction also elicited the reply that where stone images of any kind were
found it was looked upon as indicating a rich grave in pottery and probably gold."

Arrow-points. — The distinguishing charac-
ters of Chiriquian arrow-points first attracted
the attention of archeologists years ago. As
early as 1862, Dr. Merritt l " spoke of the
arrow-heads found by him in the Chiriqui
graves differing from all others he had seen ;
those from Chiriqui being pyramidal, having
four cutting edges converging to the point.
Some of them appear to have been designed
to set into the end of the shaft without fasten-
in order to remain in the wound."

mg,

Figs. 1—3. — Arrow-points.

The fact is that these rudely shaped arrow-
points are almost always triangular in section. The flat nuclear or inner surface,
slightly concave longitudinally, is left untouched, while the other two surfaces
are chipped irregularly, producing toothed edges, especially those bordering the
nuclear surface (fig. 2). A few have a fourth surface roughly parallel to the
inner surface, but narrower, and are therefore trapezoidal in section (fig. 3).

1 Dr. J. King Merritt. Report of Amer. ethnol. soc. meeting. Hist, mag., VI, 154, 1862.

22

A STUDY OF CHIRIQUIAN ANTIQUITIES.

The stems are triangular in section and slope toward the base, lending weight
to Dr. Merritt's supposition that some might have been so designed as to become
free from the shaft and remain in the wound. The material is a flinty jasper.

The only specimen in the collection that looks like true flint is given in figure 1.
It is lighter in color than the two arrow-points, and different in shape, there
being no well-defined stem. It may have been a goldsmith's tool rather than an
arrow-point. A similar specimen was figured by L. Simonin ' as coming from
the tomb of an ancient Chiriquian goldsmith. Traces of gold were left on its
lateral margins, as if it had served as a touchstone. The grave was opened by
M. J. Thevenet in 1859, who also found in it a celt similar to figure c (PI. II),
two polishing stones, etc. These were all said to have been of flint, which is

probably an error, the possible ex-
ception being the one with traces of
gold. One of the most interesting
stone implements in the Yale collec-
tion is a sub-cubical pestle-like ob-
ject, polished over its entire surface
and showing almost everywhere traces
of gold (see fig. 33).

Spear-points. — The implements that
may be classed as spear-points are
made of a velvet-black flinty quartz
resembling basanite. In workmanship
they are not unlike the arrow-points,
the differences being chiefly depen-
dent on the character of the materials
employed and the size of the flakes.
The edges also are not so serrated.
The stem is insignificant in com-
parison with the length of the blade.
The inner surface is often chipped
at the point and for the length of
the stem (figs. 4 and 5). The point
is sometimes ground and polished.
Celts. — While Chiriquian arrow-points and spear-heads are very few in number,
celts have been found in comparatively large quantities. They may be grouped
into several well-defined types, for the most part hatchets and chisels. The adze
and gouge forms are practically unknown. The same may be said of the grooved
celt, there being but one in the United States National Museum and one in the
Yale Museum.

The same variety and homogeneity that characterize the ceramic art of Chiriqui
are also stamped upon the stone art. It is almost wholly the product of a small
compact culture area and an uninterrupted culture period. The workmen were
limited to only a few varieties of stone, and these were not of a kind to encourage

Figs. 4, 5. — Spear-points. '/'

1 La vie souterraine ou les mines et les mineurs, 486, Paris (L. Hachette), 1867.

CELTS. 23

a high degree of development in the art of chipping. True flint and obsidian,
for example, seem to have been practically unknown.

The technique included chipping, pecking, grinding and polishing. The col-
lection of celts comprises specimens in various stages of development, from which
the processes of working and the resultant shapes may be determined. These
seem to vary in a measure with the character of the material. For example, a
certain type of celt is usually made of a black fine-grained volcanic tufa, ranging
in hardness from that of hornfels on the one hand to basanite or touchstone on
the other. The implements of this group are always chipped and not pecked.
On the other hand those made of silicified volcanic ash and of andesitic lava take
other forms and are usually pecked and not chipped.

The chipped types of celt are shown in Plate II (figs, a and h). The compara-
tively slender and graceful shape in figure a is due entirely to chipping, the only
polished areas being the facets that meet to form the edge. The materials of this
group are the black fine-grained volcanic tufas resembling hornfels and basanite.
The surface of the specimens is covered uniformly with a whitish patina. In
figure h the polishing reaches almost the entire length of the blade and the two
polished faces form each a single convex surface. Celts similar in shape to these
were found by Hartman in the highland plains of Costa Rica, Province of Cartago.

Sometimes the shape is varied by the production of three facets, the bevels
that produce the edge, however, never reaching more than about half-way to the
base (fig. c). A specimen resembling this was described by Simonin * as being
found in the grave of an ancient Chiriquian goldsmith.

Early stages in the pecked type of implement are reproduced in figures d and e.
The larger looks as if it had been made from an oblong, flattened, water-worn
pebble, tapering toward one end. The latter becomes the base or pole (called
by Sir John Evans, " butt-end "). The polishing extends backward along a median
line to the very tip of the pole, a feature which distinguishes the pecked type
from the polished type. The lateral facets are left unpolished. In the smaller
specimen the polished bevels at the edge extend but a short distance toward the
pole, a single exception proving the rule.

Figure / is a typical example. The lines are all gracefully wrought out, those
bounding the polished facets converging artistically and extending from the corners
of the cutting edge to the very tip of the pole. The lateral facets are coarsely
ground. These also are sometimes polished as in figure <?, which represents a
high degree of artistic skill and finish.

Two variants from the general type are given in figures b and i. In the former
there is a marked constriction in the blade immediately above the corners of the
edge. In the latter the edge is comparatively straight and long. The length,
however, has been somewhat reduced by polishing off the corners.

Figure 6 represents a class with rounded, even cylindrical, section and thick
blunt pole. Specimens of this group seem to have been shaped by pecking,
grinding and polishing. In some examples the entire surface is polished. The
material is generally a compact volcanic tufa, reddish to black in color.

1 Op. cit., fig. 139.

24

A STUDY OF CHIRIQUIAN ANTIQUITIES.

There is a small group of chipped and polished implements including alike the
largest and smallest celts in the collection. The former are made of andesitic
lava and all come from Gualaca. One of the handsomest is reproduced in figure 7.

Fig. 6.

Fig. 6. — Cylindrical celt with
thick blunt pole. '/>

Fig. 7. — Large polished celt
with pointed chipped pole ;
from Gualaca. '/«

Fig. 7-

Its length is 23.2 centimeters and greatest breadth 11.5 centimeters. The broad edge
forms a sweeping curve, the ends of which are carried along the lateral margins more
than half the length of the implement. The pointed pole is not
polished. This type is not constant ; for as the lateral margins become
straighter, the curve of the edge flattens somewhat, tending to
produce more or less triangular outlines. The diminutive celts of
which figure 8 is an example take the latter form. They are
chipped from a dark compact material, probably silicified ash or
tuff.

Chisels, which are relatively rare, are characterized by being
broadest at the base and tapering gradually to the edge. The
materials used are porphyritic andesite, tuff, and a black volcanic
tufa resembling hornfels and basanite.

chipped pole. '/« In figure 9, the base or pole is chipped only. The blade is finely
polished and faceted so as to produce an octagonal section. The
faceting of the lateral margins is of rare occurrence among Chiriquian stone imple-
ments. Of similar workmanship is the chisel shown in figure 10, except that the

CELTS.

25

marginal facets are more pronounced and the bevel at the edge is not continued
to form a median facet. This specimen, therefore, is hexagonal in section.

A handsome implement deviating in form from both of the foregoing is given
in figure 11. A distinct shoulder separates the shapely, but unpolished, pointed

I

Fig. 9.

Fig. 10.

Fig. ii.

Fig. 9- — Octagonal celt with polished and faceted

blade and chipped pole. */•
Fig. 10. — Polished hexagonal chisel with chipped

base. '/•

Fig. II. — Polished cylindrical

shoulder. */>
Fig. 12. — Polished and chipped chisel. '/«

Fig. 12.
celt with distinct

pole from the polished cylindrical blade. The only facets are the bevels at the
edge. These tools compare favorably with those of the same kind found in any
part of the world.

Some of the chisels are made of the same material as the hatchets reproduced
in figures a, h and c (PI. II), a black fine-grained volcanic tufa weathering white
on the surface. Figure 12 is an example. Its lateral margins are not faceted and
the relative shortness of the blade may be due to repeated resharpening of the
edge.

The only grooved celt in the collection is reproduced in figure 13. It is of
pale greenish gray quartzite. The groove is continuous, but shallow, and very
MEMOIRS CONN. ACAD., Vol. III. 4

A STUDY OF CHIRIQUIAN ANTIQUITIES.

near the base of the implement. Holmes figures a single grooved celt from
Chiriqui, which is, however, different in type from this one. Both may have been
importations.

Polishing Stones. — The uses of the potter's polishing stone
are well known. For such purposes preference was given
to pebbles of jasper (figs. 14 and 15) and chalcedony (fig. 16).
The entire surface of some stones is completely altered by
long-continued wear, while others are only slightly faceted.
Metates. — The traffic in ancient mealing stones by those
inhabiting Chiriqui during the historic period has reduced
the original number materially. In a letter to Professor Marsh,
McNiel leads one to infer that these hand-mills were found
either on the surface above the graves or immediately below
the surface, which made their subsequent removal by the
natives comparatively easy. Dr. J. King Merritt speaks of
the corn-grinders as " frequently occurring in the huacas."
We have more definite knowledge of the disposition of me-
tates in the ancient grave's of the peninsula of Nicoya, Costa
Rica, where Hartman l found as many as three metates in a
single bunched burial. In two cases the metate was placed
immediately over a skull, thus protecting it from the pressure
of the soil. This was in the burial ground at Las Guacas,
where as many as two thousand metates were found, a number
far exceeding that from any single locality in Panama.

Chiriquian metates are not only less numerous than those of Nicoya, but are
also distinctly different in aspect. They are made to assume animal forms, i. e.,
each specimen represents a complete zoomorphic unit, the subject chosen being
the jaguar. In these respects they resemble the metates of the ancient Guetares
culture of the Costa Rican highlands. On the other hand the metates of the
Nicoyan peninsula are three-legged and are more nearly related to Mexican than

13.— Grooved celt. '/'

Fig. 14. Fig. 15. Fig. 16.

Figs. 14, 15.— Jasper polishing stones. '/' Fig. 1 6. —Chalcedony polishing stone. '/«

1 C. V. Hartman. Archeological researches on the Pacific coast of Costa Rica. Mem.
Carnegie museum, III, no. 1, 16 and PI. XLVII, 1907.

METATES. 27

Chiriquian culture. While many of these are richly ornamented with biomorphic
motives, the specimen as a whole is not a complete zoomorphic unit. The shape
of the plate is also different, being flat in the direction of the transverse diameter
and concave longitudinally. It is also approximately rectangular in outline.

The metates of Chiriqui are made of a volcanic rock, chiefly andesitic lava.
The diversity of form and finish led Holmes to suggest that the metates of Chiriqui
might "represent distinct peoples or different grades of culture." Nothing could
be more homogeneous than the large group representing the jaguar. That this
is genetically related to the more primitive types, there can be little question.

Fig. 17. — Mctate with three legs, and hand stone. '/<

The simplest kind of mealing stone is a flattened river boulder such as is still
in use among the Talamanca Indians of Costa Rica and Chiriqui. In order to
lessen the weight of the metate and render it more easily transportable, as much
as possible of the base was removed. This was accomplished in two ways, viz.,
(1) by hollowing out the base and reducing its external dimensions; and (2) by
cutting away all but three points of contact — the smallest number that will give
stability. From the former were developed not only the zoomorphic types of
metate, but also the so-called stools. The tripod type was not further elaborated
in Chiriqui, but reached a high state of development in the province of Nicoya,
Costa Rica. Only two examples of this type are illustrated. The rubbing surface
of one of these (fig. 17) is an elongated oval and concave in the direction of both
its length and breadth. The rim is chipped away at the end over the unpaired
leg. One of the sub-cylindrical rubbing stones exactly fits this metate. Both are
of the same material, but unfortunately the collector's notes do not reveal whether
they were found together. The grinding surface of the other three-legged metate,
figure 18, is quite different in shape, being concave in the direction of its length
only and flat transversely, thus belonging very distinctly to the Nicoyan type. A
rim proper does not exist, but is suggested by an incised line near the margin.
The lateral margins are slightly concave, while the margins at the ends are
markedly convex. The under surface is left in the rough, and the three legs are
very short. This specimen is from Gualaca.

The prototype of practically all the Chiriquian forms of metate as well as of
stool is to be found in figure 19. The shape of the top is intermediate between
oblong and oval. The dishing or hollowing out is very slight and the rim low.
The base is cut away on all sides and deeply excavated in the bottom, leaving

A STUDY OF CHIRIQUIAN ANTIQUITIES.

the dish-shaped plate to rest on a hollow stand that is somewhat larger at the
bottom than at the top.

The hollow stand could be made still lighter by cutting windows in its sides
and ends. This is what is done in figure 20, which represents a metate of nearly the

same size and shape
as the foregoing. The
outlines are a little
more nearly rectan-
gular. This is partic-
ularly true of the
stand, which is also
smaller at the bottom
than at the top. There
are eight large open-
ings, three on a side
and one at either end.
The plate at the top
has a low rim and is
thick enough at the
margins to admit of
a peripheral incised
ornament, consisting
of two parallel hori-
zontal lines connected
at short intervals by
transverse lines. We
have in this metate
all the essential con-
structional features to
be found in the so-
called stools, except
that the latter have
a circular top and
stand instead of oval
to rectangular ones.
The metate shown
in figure 21 is the gift
of Mr. Edwin Lamson
of Summit, New Jer-
sey. It resembles
the preceding in its
roughly rectangular
top and its sloping base. But the latter is hollow and entirely cut away at the
ends, so that it does not present a continuous contact support. The sides are
decorated with incised panels. The design is a faulted meander with branching
incised lines filling the angular spaces. This metate and one of the pottery stools

Fig. 19.

Fig. 20.

Fig. 1 8. — Three-legged metate of the Nicoyan type; from Gualaca.

Fig. 19. — Prototype of Chiriquian metate. */'

Fig. 20. — Metate with openings in the hollow stand. ''•

METATES.

29

Fig. 21.

(see PI. XLVI, fig. a) form a connecting link between mealing stones on the one
hand and stools on the other.

In the Keith collection may be
seen two stone metates from Mer-
cedes. Costa Rica, like the Lamson
specimen in shape, only they are not
incised on the sides. The wooden
stools or seats in use among the
present-day Indians of Chiriqui have
approximately the same shape, as
may be seen by consulting figure 22.
Dr. J. Walter Fewkes describes a re-
markable clay figurine from Santo
Domingo that is seated on a metate-
shaped clay stool.1

By cutting a block from the center
of each side of the hollow base, as
well as from both ends, the four-
legged type of metate is produced
as seen in figure 23. Animal heads
are placed at each end to increase
the zoomorphic effect. In the present
instance eight diminutive heads are
added, none of them projecting be-
yond the margin of the plate.

A somewhat similar specimen is
reproduced in figure 24. The in-slop-
ing legs are more highly finished. Each pair represents the forelegs of an animal.
The head going with one pair is that of the jaguar; the other looks more like that
of a mastiff with overhanging upper lips. The outer surface of the legs and the

Fig. 22.

Fig. 21. — Metate with base open at the ends, the sides being
decorated with incised panels. '/»

Fig. 22.— Modern wooden seat used by the Indians of Chi-
riqui. '/•

Fig. 23. — Metate of the four-legged type, ornamented with ten animal heads
attached to the margin of the plate. '/'

margin of the plate are decorated with incised patterns. The top of the plate is
considerably worn, the wear extending to the low rim, particularly at the ends.

1 Twenty-fifth ann. rept., Bur. Amer. ethnol., PI. LXXXII, 1903-04.

30

A STUDY OF CHIRIQUIAN ANTIQUITIES.

A single zoomorphic unit is represented in figure 25. The head is clumsy,
not the typical jaguar head, probably that of some canine. The tail, which is
broken, curved downward and laterally till it joined the left leg at the knee. The

Fig. 24. — Metate representing anterior half of two animals. '/<

outer surface of the legs is cross-hatched with rather deep incisions, as was also
the tail. The margins of the plate representing the animal body are marked in
a similar manner. There is a pair of bands in relief at each knee.

Fig 25. — Metate representing a single zoomorphic unit.

We have now followed the various steps in the development of the complete
zoomorphic unit from the commonplace mealing stone. Once established, this
unit becomes constant. The jaguar appears to be chosen to the exclusion of all

other forms. It would be interesting
to know the reasons for such a
choice. It may be that, as in Mexi-
co, the jaguar was esteemed one of
the primeval gods, its name being
given to the Earth. It would be
appropriately associated, therefore,
with the sources from which the

Fig. 26. — Jaguar metate with rectangular top. '/» grain and nuts COme. The skill with

which the majestic bearing of this

animal is rendered borders on the marvelous, The hand of the sculptor was
guided not only by artistic skill but also by reverence for the subject.

Among the jaguar metates there are two types of mealing plate ; (1) the rec-

METATES.

31

tangular and (2) the oval. The latter far outnumber the former. An example of
the rectangular top is given in figure 26, where the borders are flat and ornamented
with lozenge-shaped incisions. The top is hollowed out, thus leaving a compara-
tively high rim on all sides. Within are red and yellow stains as if the metate
had been used as a mortar for mixing paints. The few specimens noted of this
type have a longitudinal furrow in the tail, with slanting lateral incisions, as if
to indicate the parting of the hair.

Fig. 27. — Jaguar metate the legs of which are decorated with stars in champleve. '/«

Some new decorative elements are introduced in figure 27. On the outer sur-
face of each Iftg are from one to two stars in champleve. The number of points
to the star varies from six to nine. Each star is enclosed in a circle. This motive
does not occur on any other metate in the collection. The guilloche pattern on
the head and tail, however, is characteristic of both Chiriquian and Costa Rican
metates, it and the lozenge-shaped design very cleverly representing the rosette
markings of the jaguar. The low rim of the mealing plate has been almost ground

Fig. 28. — Jaguar metate with guilloche ornamentation. '/«

away. The neck of the jaguar is very short, bringing the ears so close to the
rim of the plate that they, too, are ground at the tips. One of the rubbing stones
(fig. 31) fits this metate perfectly.

A more characteristic jaguar metate is illustrated in figure 28. The border of
the oval plate is plain, with the exception of two parallel horizontal incised lines.
Elsewhere the markings of the jaguar coat are indicated by somewhat involved

32

A STUDY OF CHIRIQUIAN ANTIQUITIES.

guilloches. These are confined to the outer more easily visible surfaces, as in
nature. The joints and muscles of the legs are worked out in a somewhat con-
ventional, though effective fashion. The neck and root of the tail encroach upon
the mealing surface of the plate. The muzzle of the jaguar is relatively short
and the jaws are wide open, a characteristic feature of all representations of the
jaguar, whether in stone, clay, or gold. The border of the plate, as well as the
head, legs and tail, are decorated with guilloches.

The finest and largest Chiriquian metates are grouped together in Plate III.
The one reproduced in figure a is particularly graceful and lifelike. The head
and neck are beautifully modeled, and the incised patterns on legs and tail are
original. The other two are exactly of a height (37 cm). Figure b has the wider
plate, being but little longer than broad. Contrary to the general rule, the tail
is relatively short and is not carried downward and laterally to unite with one of

the feet. The largest
of all the metates is
shown in figure c. It
measures 1.17 meters
in length by one-half
meter in breadth. The
legs and tail are plain,
but the borders of the
plate are incised as is
also the forehead. The
two rows of teeth and
the overlapping of the
canines are distinctly
shown.

There is a small
metate in the collec-
tion (fig. 29) resembling one from Chircot in the highlands of Costa Rica, figured
by Hartman. The plate is nearly round and perfectly flat, with incised borders.
There is an animal head at either end. The legs are short, thick and roughly
triangular in section. The smallest of the metates is given in figure 30, its length
being 16.5 centimeters. The plate is flat and rectangular, the short legs are round
in section, the animal head is only roughly blocked out.

One of the most elaborately carved metates is in the collection (cat. no. 8250)
of Mr. George G. Heye of New York. The border, which is not faceted,
is divided midway on either side by two raised vertical lines. The angular
guilloche pattern on the front half is composed of four intertwining bands, while
that on the rear half consists of only two. A similar pattern is carried down the
foreleg, while the hindleg is ornamented with a series of squares which re-
semble still more closely the ocellated markings of the jaguar.

Much pains was spent on the muscles and joints of the legs. The toes are
indicated and even the pads on the soles of the feet are carefully worked out.
The incised design on the top of the head is a marked variation from the usual,
the central feature being a rosette ; surrounding this is a series of raised triangles.

Fig. 29. — Metate with animal head at either end. '/»

RUBBING OR HAND STONES. 33

The same motive repeated, except that the triangles point outward, is half cut
away by the deep dishing of the mealing plate. The bottom of the latter is quite
concave, as might be expected from the rather highly convex ventral curve ; and
is highly polished, particularly in the middle. It is also stained black, the stains
reaching nearly to the top of the high rim. The tail has two deep longitudinal
grooves, the sides of which are
ornamented with rows of trian-
gles in relief, their bases meeting
at the bottom of each groove.
There is also a row of these
triangles on either side of the
tail. The teeth are faithfully
rendered, the artist taking care
to make the lower canines close
in in front of the upper, as is

the Case in nature. This Spec- Fig- 3°- — Small crudely shaped raetate with rectangular plate. V»

imen was collected by Mr. F.

D. Utley in the St. Andres mountains, near Bugavita. He obtained from the same
locality a very handsome jaguar metate similar to figure 28, the body, however,
being relatively longer and natter.

The Heye collection includes a large metate similar to figure 24, except that
the heads at either end are much flattened and the short sloping legs are per-
fectly plain. Of his Costa Rican metates, one has been figured by Hartman.1
Two others are worthy of special mention. One of these (cat. no. 9622) is very
similar to the ordinary Chiriquian jaguar metate: by the addition of the head at
one end and the lower extremities from the pubic arch downward at the other,
the plate is converted into the body of a human female facing upward. The
human legs are curved downward and laterally till they unite with the metate
legs exactly as if they were a pair of jaguar tails instead. The other metate (cat.
no. 1872) with a plain rectangular plate and high rim and flat bottom is supported
on the backs of two jaguars, each with the head turned so as to face outward
and with nose on the ground. The tail is curved upward so as to take some of
the load. Each jaguar is represented with a single foreleg and hindleg.

Rubbing or Hand Stones. — Very little general interest attaches to the compara-
tively insignificant upper millstones, which accounts in part, at least, for the rela-
tively small number to be found in the collections. According to Hartman, the
metates of Las Guacas far outnumbered the rubbing stones. During his " ex-
cavations on the spot, where about fifty complete metates were exhumed, not a
single complete rubbing stone was discovered. Only a couple of small frag-
ments were brought to light." In his previous excavations at Las Casitas, how-
ever, he did find " las manos " with several metates. The characteristic Nicoyan
grinding stone was sub-cylindrical, " and so much longer than the breadth of the
metate that the hands of the women when grinding could comfortably grasp both

1 C. V. Hartman. Archeol. researches on the Pacific coast of Costa Rica. Mem. Carnegie
museum, III, no. 1, figs. 64 and 65, 1907.

MEMOIRS CONN. ACAD., Vol. III. 6

34 A STUDY OF CHIRIQUIAN ANTIQUITIES.

ends of the grinder on either side." This was made possible because the
metate's grinding surface was flat in the transverse direction and not bounded
by a raised rim.

The Chiriquian grinding stone had to fit a very different sort of grinding sur-
face - - one concave in both directions and bounded by a low rim. The upper
stone, therefore, is relatively short. An example is reproduced on its metate
(see fig. 17). It resembles a flattened cylinder, rounded at the ends. Another type
is nearly rectangular in section, with flattened ends. The stone reproduced in
figure 31 is a good example. It fits one of the metates in the collection (see
fig. 27), but may not have been found with it.

There is a single specimen that would seem to be better adapted for use in a
mortar than on a metate (fig. 32). It may have served as a grinding stone for
the small metates with high square rim (see fig. 26) or the so-called stone stools
with circular tops (see PI. IV, fig. e). In shape it may be compared to an inverted
toadstool. A similar type of pestle has been found in southern Indiana and Ohio,
as well as on the northwest coast of America.

Fig. 31. Fig. 32.

Fig. 31. — Rectangular rubbing stone. V«
Fig. 32.— Pestle. '/»

Fig. 33-
Fig. 33. — Sub-spherical pestle streaked with gold.

A cylindrical hammer stone with tapering and battered ends should be mentioned.
One of the most interesting stone implements of the collection is a sub-spherical
pestle with polished surfaces and almost everywhere streaked with gold (fig. 33).
De Zeltner, from whom the specimen was bought, believed it to have been used
by the natives to reduce the gold to powder preliminary to the process of fusion.
It was entered in his catalogue as a " pierre de porphyre, forme de boulet un
peu equarre ayant servi de marteau d'or, on voit encore les traces du metal."
This is the specimen referred to by Gabriel de Mortillet1 as "ayant servi a triturer
le metal afin d'en faciliter la fusion." The stone is of volcanic origin, probably
a well-cemented ash or tuff.

1 Materiaux pour 1'histoire primitive et philosophique de I'homme, IV, 65, J868.

STOOLS. 35

Stools.— There is a small group of stone carvings to which the names circular
metates, mortars or stools might apply almost equally well. Their kinship to the
metates has already been mentioned. If the sculptor were to begin with a block of
stone whose length was greater than its breadth or thickness he would end with a
metate. If on the other hand, all three dimensions of the original block were
about equal, the final product would be the so-called stool. Both groups have
a plate showing the effects of use. In either group this plate may be supported
by a solid column, a hollow stand in the sides of which openings are cut, or by
four legs. Even the guilloche ornament so common to the metates is also found
on some of the stools. The latter might well have served as seats, in which case
continuous use would have smoothed the surface of the plate. While the wear
on the stools is unmistakable, it is not quite so marked as on the metates.

In speculating on the use to which these objects were put it should be noted
that similar stools were also made of clay. That the latter were used either as
mortars or stools, is highly improbable. Holmes suggests that they may have been
employed as supports for articles, such as vases or idols, or possibly as altars.

One difference between metates and stools, which may point to a difference in
their use, will suggest itself by consulting the illustrations, viz., the association
of the jaguar with the metate and that of the monkey and man with the stool.
It is true that the heads suspended from the plate of one of the stools (see PI. IV,
fig. b) may possibly be referred to the jaguar. Holmes also describes a clay stool
supported by " two rudelv modeled ocelots and two monkey- like figures," but
these are rare exceptions that prove the rule.

A series of five stools is given in Plate IV. One of the ruder forms is shown
in figure a. The top is concave; its margin is decorated with nine animal heads.
Four vertical openings are cut in the sides of the hollow bell-shaped support.
A somewhat similar but more finished stool is reproduced in figure b. There is
a more highly developed border to the plate, from which hang four jaguar-like
heads. The peripheral ornamentation consists of a series of disconnected sigmoid
scrolls for three-fourths of the way and of a simple guilloche pattern for one-
fourth. In one of its stages the guilloche becomes a series of linked S's — a proof
that it and the scroll have a common origin, as pointed out by Holmes. The
sides of the hollow stand are nearly vertical, the spread being confined almost
wholly to the continuous foot. A stool similar to the two foregoing, but of more
finished workmanship and about three times as large, was found recently at Mer-
cedes, Costa Rica, and is now in the Keith collection. The tallest of the seats
(fig. c) is supported by four round legs : near the top of each there is a human
head in relief. Just above these, encircling the plate, are thirty small animal heads.

The support for the seat becomes frankly zoomorphic in figure d. Four monkey-
like figures stand on the slender basal ring, carrying the circular seat-plate on
their heads and uplifted hands. While all four monkeys are equidistant from each
other, they are also grouped in pairs by bringing the tips of the tails and the
two adjacent elbows into contact, forming in this way a delicate and continuous
tracery reaching half-way round the specimen. The same thing is repeated on
the opposite side. The only lack of bilateral symmetry in the two halves is due
to an accident that happened to the piece, presumably while it was still in the

36 A STUDY OF CHIRIQUIAN ANTIQUITIES.

sculptor's hands. The right arm of one of the monkeys was broken off. The
ingenuity with which the mutilation was concealed is seen in figure d'. All traces
of the broken member were removed, even to the fingers holding the plate, and
a necessarily abbreviated arm in relief was chiseled on the monkey's breast. A
small crack in the plate where the hand previously rested may have been caused
by the same fall or blow that carried away the outstretched arm. In other
respects the piece is perfect. The margin of the plate is ornamented with scroll
work. Stools similar to this one are found as far north as Mercedes, Costa Rica,
from which place Mr. Minor C. Keith obtained the largest and finest specimen of
the type in question I have ever seen.

The top in figure e is more like a mortar than a seat. It is dished to a depth
of 3.7 centimeters and the inside is stained a dark color, as if it might have served
to mix paints. The border is undecorated and curved instead of being flat, giving
the whole the appearance of a shallow bowl. The latter is held by four monkeys,
two upright and two, alternating, on their heads. One of these is tailless. This
stool resembles the one, also from Chiriqui, exhibited by Captain J. M. Dow at a
meeting of the American Ethnological Society in 1860, and described as " a
circular dish about ten inches in diameter, with a rim supported by five human
figures, standing on a narrow circular foot, all carved in stone and hollow, prob-
ably a chafing dish for sacrifice." *

One of the stone stools in the collection is so much like that represented in
figure 13 of Holmes's work that it is not reproduced here. There is a similar
example in the Heye collection (cat. no. 7054) with this difference that the base
also carries a row of heads, inverted and smaller than the upper series.

Mr. Heye possesses another Chiriquian stool (cat. no. 8286) of sandstone, the
hollow bell-shaped stand of which is curiously grooved as if it had been used as
a polisher. The grooves are vertical and form two series ; those of the upper
series are the larger and are cut through the sides of the stand, while those of
the lower group are smaller and do not go through. This piece was collected at
Bugaba by Mr. Utley.

Images. — This class of objects is not very large, the majority being represen-
tations of the human form. Some of these are fine examples of the art of prim-
itive sculpture ; others are quite crude and unfinished in appearance. The original
shape of the stone in figure 34 suggested the lines of some animal so that little
was done save to round the body and neck and bring out the eyes and nose a
trifle. A small image of a quadruped, with head raised and neck twisted half-
way round, is seen in figure 35. It is without merit either in conception or exe-
cution. This cannot be said, however, of the next illustration (fig. 36), which
may be described as representing a man or a monkey seated on the end of a
cylindrical pestle and holding something to his mouth. The remarkable head-dress
should be compared with that of the figure standing on one of the drum whistles
(see fig. 280).

One of the large crude human images (fig. 37) has the appearance of being
perched on the end of a stone pillar. The head is relatively large, the hair or

1 Bull. Amer. ethnol. soc., I, 11, New York, 1860-61.

IMAGES.

37

head covering, represented both in relief and by means of parallel incised lines,
running from front to back. At the vertex there is the stump of what once might

frig- 34- — Image representing crude animal form. '/'

Fig. 35. — Small image of a quadruped. '/<

have been a cylindrical shaft. One of the Chiriquian gold figurines in the Heye
collection has a similar head-gear, only much taller, resembling the high Tlingit

v.'.-----.:-.-'-?; '*.-»• .-••••• f
Bj:SS2fe5 ' ":.!-; •'•>&! --

^^&iL-: ^«&

Fig. 3°- Fig- 37- Fig- 3»-

Fig. 36. — Image seated on cylindrical pestle ; may represent man or monkey. '/•
F'g- 37' — Large crude human image with " skil " ornament on the head; from Bugavita. '/•
Fig. 38. — Crude human image ; from Bugavita. '/«

hats built up of superimposed disks (skil). The body is short. The arms are
indicated in relief, being bent at the elbow and again at the wrist, with the hands
against the sides of the face. The legs are drawn as if in a sitting posture and

38

A STUDY OF CHIRIQUIAN ANTIQUITIES.

hug the pillar. Figure 38 is in the same general class, the differences being
limited to minor details. Both these pieces are from Bugavita. One represents
a male, the other a female.

There are three standing statues in the collection, one of which has the appear-
ance of being merely blocked out (fig. 39), although some striking character-
istics are already visible : the long nose, for instance, with the line of the bridge

straight and continuous with the forehead ; the top of
the head, shaped like those of the clay figurines described
on page 165 ; the loin-cloth and the short legs bent a little
at the knee. The feet have not been cut apart. The
body is long. The arms are in relief, the right hand is
on the right breast and the left on the abdomen. This
piece is from San Carlos.

Statues of the female sex predominate. The finest
of these (figs. 40 a and 406) was, according to the
finder from whom Professor Marsh bought it, " dug
up in Chiriqui in 1871," and appropriately named the
" Panama Venus." It undoubtedly does represent some
familiar goddess, as there is in the collection another
copy, not so elaborately finished, however, and less than
half as tall as the one reproduced. The latter is 78.5
centimeters in height and stands securely on its own
feet by virtue of their breadth and the projecting heels.
The toes are indicated and the malleoli, both median
and lateral, are prominent. The relatively short legs
are slightly flexed at the knee. The incised ornament
on the loin-cloth and the head-dress, the only articles of
apparel, resembles somewhat that on the borders of the
mealing stones already described. The body is round
dorsally and flat ventrally, the flatness being emphasized
by the square shoulders and stiff angular pose of the
arms and hands. The loin-cloth is also flattened in front
to be in keeping with the surfaces above. The neck,
on the other hand, is round in front and flattened behind.

The hair, represented in relief, reaches to the shoulders, suggesting the same
style of treatment as seen in the clay figurines. The flatness of the crown of
the head may be intended as a feature of the head-dress only. The face is
pointed, the nose long and straight, and the mouth small. The two examples
of ''Panama Venus" in the Yale collection are both in a perfect state of preservation
and resemble each other much more closely than either does the one figured by
Holmes. All three however undoubtedly refer to the same personage.

In 1860 Mr. Totten and Mr. Center, engineers of the Panama Railroad Company,
gave to the American Ethnological Society three stone statues of an unusual type,
described in the record of the meeting ' as " being about two feet high, cut from

Fig. 39. — Rough statue in stand-
ing posture; from San Carlos. '/'

1 Hist, mag., IV, 144.

ORNAMENTS.

39

hard, dark-colored stone, and represent the human form and features distorted,
and with legs bent. Two of them have square tapering pedestals, about two feet
long, apparently designed to be stuck upright in the ground; and the third may
have had a similar one. They are said to be the only objects of the kind found
in the graves, though
it has been reported
that a number of such
specimens are stand-
ing in one of the grave-
yards in a forest at
Chiriqui." The three
statues in question
were from Panama,
presumably the prov-
ince of Chiriqui.

Ornaments. — Under
this head are grouped
beads and amulets.
Both classes are rare
in Chiriqui. Two large
highly polished agate
beads are reproduced
in figures 41 and 42.
The conical borings
are made with precis-
ion at both ends and
stop at the center,
leaving only room
for the passage of a
slender thread. There
is but a single green
jasper bead (fig. 43)
in the collection. A
small agate pendant,
with a vertical and

a horizontal boring Fig. 40 a. Fig. 40*.

meeting at a point
near the upper end,
is seen in figure 44.

Because of their resemblance to those from Nicoya, the few amulets to be found
in Chiriquian collections are presumably of Costa Rican origin, a presumption
that is further strengthened by their great abundance at Las Guacas. Hartman
states that '" of all the objects found in the burial ground of Las Guacas none
are more numerous than the amulets. Several thousand specimens have been
unearthed. The commonest are those which have the shape of a celt or a grooved
axe, which has been divided lengthwise, the convex side serving as the front."

Fig. 40. — Large statue of a female — the so-called Panama Venus ; a front
view ; 1> profile view. '/•

40

A STUDY OF CHIRIQUIAN ANTIQUITIES.

This fits the description of the specimens found in Chiriqui. A plain amulet
of this type is reproduced in figure 45. The flat back is perfectly plain, its median
ridge or scar having been ground down and polished. The convex front has a
central transverse groove. Near the top there is a single hole for suspension. The
periphery is not faceted, that part of the margin forming the edge being sharper

Fig. 41. Fig. 42.

Figs. 41, 42.— Highly polished agate beads. '/«

Fig. 45. — Highly polished amulet
of pale green jade. */•

Fig. 46. — Large translucent amulet
of jade ; from Miravalles, Costa
Rica. '/'

Fig. 45-

Fig. 43-

Fig. 44-

Fig. 43 — Green jasper bead.
Fig. 44. — Small agate pendant.

V.

Fig. 46.

than the rest. The surfaces are everywhere highly polished. The material is a pale
green jade, translucent only near the thin edges.

On some of the pieces described by Hartman there remained the median scar
at the back, produced by sawing into the parent block from two sides until it
was possible to pry loose the piece intended for the amulet. In the manufacture
of these objects large pebbles were sawed lengthwise a number of times until only

ORNAMENTS.

41

Fig- 47-

a thin central flat slab was left. Even this was perforated and also used as an
amulet if the material was precious. There is one such slab in the Yale Museum.
It is from Miravalles, Costa Rica, but is appropriately reproduced here (fig. 46).
Its length is twenty centimeters. The thickness varies, one half being only
two millimeters in thickness, and the other three, so that the piece is translucent
at all points, as it is cut from a mass of excep-
tionally fine jade. The longitudinal scars produced
by the failure of the saw-cuts on the opposite
sides to meet in a common plane (and by the
prying-loose process) are visible on both back
and front. A slice was removed from one side
of this slab. The other lateral margin and the
ends seem to follow rather closely the contours
of the original pebble or mass. Sections of the
margin at each end show, in fact, the original
crust, while the rest of the specimen is polished
artificially. Near the smaller end are two holes
bored from one side only, one of them being a
second attempt at boring, after a hole had been
started too close to the margin.

The small celt-shaped amulet shown in figure 47
serves as a connecting link between the preced-
ing figure and the one that follows. It was a
piece retained by the Lamson Brothers when
they sold their (the McNiel) collections to Professor
Marsh. It is flat on the back and has the groove
across the front, but the lateral margins and pole
are faceted and there are two holes for suspen-
sion instead of one. They also represent the
eyes of the life form, the head of which is further
suggested by the incised lines above and below.
This may be either anthropomorphic or ornitho-
morphic. There are two additional holes made
by boring diagonally into the lateral facets and
the back, respectively. These may have been
for the attachment of ornament or apparel. The
material is dark green and opaque, probably a
variety of hornstone or basanite.

A fine ornithomorphic celt-shaped amulet of
jade is shown in figure 48, one that compares
favorably with the best specimens from Las

Guacas figured by Hartman. It does not seem to have been formed by splitting
a celt in two. The back is convex, as was also the front until its lower half was
cut away to make the blade of the celt and the tail of the bird as well. Then
comes the body of the bird with wings in relief folded on the breast. A trans-
verse hole for suspension is bored through the constriction at the neck. The
MEMOIRS CONN. ACAD., Vol. III. 6

Fig. 48.

(Lam-

Fig. 47. — Small celt-shaped amulet.

son collection.) '/'
Fig. 48 — Ornithomorphic celt-shaped amulet

of jade. '/•

A STUDY OF CHIRIQUIAN ANTIQUITIES.

prominent beak is in low relief and reaches well down on the breast. The eyes
are scarcely visible — two little shallow depressions sunk by boring. Above are
the two ear-tufts so characteristic of the gold and clay parrots described in other
chapters. If this resembles the stone amulets from Nicoya, it also suggests the
gold amulets of Chiriqui, and may not be an importation. The surface is every-
where highly polished except at the edge.

Another ornithomorphic amulet, but not celt-shaped, is given in figure 49. The
beak is long, straight and hooked at the end — evidently that of the parrot. The

Fig. 49. — Ornithomorphic amulet of green translucent jade. '/' Fig. 50. — Zoomorphic amulet of agate. '/«

head and body combined are not so long as the beak, although more massive.
In the place of ear-tufts, there is a prominence at the base of the beak. The
region of the neck is indicated by a peripheral incision, at the level of which is
placed the hole for suspension. The base of the tail is marked by a horizontal
incised line across the back. The feet and toes project very little beyond the
contour line of the body and resemble very closely the feet of the well-known gold
parrots. In fact, the entire figure seems to have been inspired by the work of the
Chiriquian goldsmith. On the other hand, it is so different from the bird forms among
the amulets found at Las Guacas that to assume it to be of Nicoyan origin would be

hazardous. The specimen is of the finest quality of
green translucent jade and in perfect condition.

The abundance of jade ornaments found at Las
Guacas, together with " worked blocks of the crude
stone," leads Hartman to conclude that the amulets
were manufactured there and that a mine of the
mineral may yet be discovered in the neighborhood.
It is highly probable, therefore, that there was a com-
mon source of jade supply for the regions in question,
and that it was in Costa Rica.

Figure 50 represents a zoomorphic amulet in a
simple yet effective manner. It is a slab from an
agate pebble. Shallow notches cut out of the periph-
ery leave muzzle, forelegs and hindlegs, respect-
ively ; while the entire dorsal contour, from the back
of the head to the tip of the tail, is produced by

collection) '/•

forefeet are perforated for suspension.

PKTROGLYPHS.

43

Chiriquian ornaments of stone are comparatively rare. Those of gold were
relatively numerous. Ornaments of less durable materials, such as bone, teeth,
shell, etc., may have been used. The historian Gonzalo Fernandez de Oviedo y
Valdes, who visited Nicoya in 1529, says that the Indians wore necklaces of sea
shells. Similar perishable ornaments may have been used by the prehistoric races
of Nicoya, as well as of Chiriqui. In the Lamson collection there is a single
shark's tooth (fig. 51), the neck of which is trimmed and perforated. It was found
by McNiel in a grave at Divala and was evidently worn in the same fashion as
were the celt-shaped amulets of stone.

Petroglyphs. — Seemann ' speaks of finding in western Veraguas (Chiriqui) remains
of a numerous tribe, which he calls Dorachos. These remains consist of " tombs,
monuments and columns of different sizes, covered with fantastic figures, or re-
presentations of natural objects, differing entirely from either the hieroglyphics of
Mexico or those of Central America." Seemann was perhaps the first (1848) to
give a detailed description of a granite block at Caldera, north of David, known
in the vicinity as piedra pintal, or painted stone. The sides and even the flat
top of this huge boulder
(or outcrop ?) are covered
with incised figures. The
characters are from a half
to one inch deep except
on the weather side, where
they are nearly effaced,
a proof of their great an-
tiquity. The work is as-
cribed to the Dorachos.
Seemann made a drawing
of the piedra pintal, which
was later reproduced by
W. Bollaert.3 Holmes
published a sketch by
McNiel of the southwest
face (fig. 52), showing
approximately the same
number of glyphs as in
Seemann's drawing. Trac-
ings by M. A. L. Pinart
are said to reveal at least
forty glyphs on the same
face.

Fig. 52. — Southwest face of the piedra pintal or pictured rock at Caldera.
(After Holmes.)

Fig- 53- — Northeast face of the piedra pintal, from a water-color drawing
made by Gentil.

Seemann was struck by the similarity of the Chiriquian and certain Northumbrian
and Scottish petroglyphs. Charles Rau,3 however, was unable to discover any

1 Berthold Seemann. Narrative of the voyage of H. M. S. Herald, I, 312, London, 1853.

2 Antiquarian, ethnological and other researches in New Granada, etc., 30, London, 1860.

3 Observations on cup-shaped and other lapidarian sculptures in the Old World and in Amer-
ica. U.S. geogr. and geol. surv. of the Rocky Mt. region. Contr. to Amer. eth., V, 69, 1881.

44 A STUDY OF CHIRIQU1AN ANTIQUITIES.

figures on the piedra pintal that are identical in shape with European stone sculp-
tures, excepting concentric circles and a few carvings resembling wheels with
four spokes. He very justly points out that simple devices like these, when
found in different countries, are no proof of ethnic affinity or contact.

Dr. J. Walter Fewkes 1 notes the similarity between Porto Rican petroglyphs
and those of Chiriqui. The chief motive on the southwest side of the piedra
pintal seems to represent a front view of the human or other face. It is repeated
with variations six or eight times. A motive not unlike this characterizes Porto
Rican petroglyphs, and is also found in Colombia and Venezuela.

A water-color drawing of the northeast face of the piedra pintal was made by
Gentil for Father Heyde. Through the kindness of Professor M. H. Saville of
Columbia University I am able to reproduce a tracing of this sketch (fig. 53),
showing its close agreement with Bollaert's statement that " the top and other
sides have signs of a circular and oval form, crossed by lines."

Other Chiriquian petroglyphs have been noted by Mr. J. F. Bateman in a
letter to the American Ethnological Society, accompanied by a tracing which is
reproduced in figure 54. According to the letter : " The hieroglyphic enclosed
was found on a large surface boulder of lava, in the parish of San Miguel,

district of Chiriqui. The tracing is perfectly correct, two
feet long. The boulder is irregular in shape, and appears
once to have been covered with figures or hieroglyphics ;
but the others are so indistinct as to prohibit being copied.
The outlines which I send you are on an angle of the
boulder- facing the west. The position of the rock is
eight miles nearly due west of the mountain called El
Fig. 54.— Tracing of a pctro- Volcan, which is said to have an altitude of thirteen thou-

glyph. (After Bateman.) ;

sand feet."

None of the petroglyphs seem to bear any relation to the many decorative
motives used on the pottery found in the graves. This fact suggests the prob-
ability of their belonging to a different age and civilization. At the time of
Seemann's voyage, ancient sculptured columns were used for building purposes
in the town of David. The characters on these, however, were raised instead
of incised and were considerably smaller than those of the piedra pintal.

1 The Aborigines of Porto Rico and neighboring islands. Twenty-fifth ann. rept, Bur. Amer.
ethnol., 151, 1903-1904.

2 Bull. Amer. ethnol. soc., I, 12, New York, 1860-1861.

POTTERY.

In the early stages of human culture, the association of ceramic art with stone
art is well-nigh universal. Both stone and clay have archeological possibilities of a
high order. They are alike almost indestructible. In the case of stone, the
ingredients are already mixed and the firing is done by Nature's hand. It occurs,
therefore, in a more utilizable and available state — one that even invites use, and
was the first to appeal and yield to nascent human mentality.

The conditions which gave rise to the discovery of the uses of baked clay and
the appreciation of its economic and artistic importance were very different from
those surrounding the first race of tool-users. De Mortillet believed pottery to
be an invention of the neolithic period; according to Rutot it dates back to the
cavern epochs of the paleolithic period. When necessity knocks hard enough,
the door of invention opens. The need of vessels must have been felt at a rela-
tively early stage in the history of primitive man ; and as the habits of the race
became more and more sedentary, the fragility of clay vessels, their weakest
point, counted less and less against them. The very fact that fictile products
are not easily transportable adds to their usefulness as a criterion for locating
culture areas.

Great as is the value of ceramics from a practical standpoint, the interest of
the archeologist is centered rather in the role it has played in the development
of art. The plasticity of clay is particularly well calculated to stimulate the imagi-
nation. It yields readily to any form that fancy may dictate. These forms,
whether useful or ornamental, in themselves present surfaces that admit of further
decoration by means of engravings or painted designs, thus bringing into play
the great realm of art, from sculpture in the round and relief to engraving and
painting.

The place of aboriginal American pottery could scarcely be better given than
in the words of William H. Holmes:1

" It is hardly possible to find within the whole range of products of human
handicraft a more attractive field of investigation than that offered by aboriginal
American ceramics, and probably no one that affords such excellent opportunities
for the study of early stages in the evolution of art and especially of the esthetic
in art. The early ware of Mediterranean countries has a wider interest in
many ways, but it does not cover the same ground. It represents mainly the
level of the wheel, of pictorial art, and of writing, while American pottery is
entirely below this level, and thus illustrates the substratum out of which the
higher phases spring. But it should be noted that not merely the beginnings of
the story are represented in the native work. The culture range is quite wide,
and opportunities of tracing progress upward to the very verge of civilization are

1 Twentieth annual report, Bur. Amer. ethnol., 19, 1898-1899.

46 A STUDY OF CHIRIQUIAN ANTIQUITIES.

afforded. Between the groups of products belonging to the inferior tribes scattered
over the continent from Point Barrow to Terra del Fuego, and those representing
the advanced cultures of Central America and Peru, there is a long vista of
progress."

The fictile products of Chiriqui may be divided into a number of well-defined
groups. With the exception of certain fugitive pieces, these are bound together
by common though sometimes rather slender threads. The pottery as a whole
may be looked upon as a more or less consistent unit. There are traces of
northern influence and some evidence of contact with the peoples to the south,
but these combined were not sufficient to overcome the art-molding forces from
within. The latter were strengthened by the solidarity that comes with large
numbers dwelling in contiguous communities over a restricted and somewhat
isolated area. The abundance and size of the ancient cemeteries (huacals) scat-
tered over nearly every part of the province are proof that the region was populous
and the period of continuous occupation relatively long. There was time for
tradition to make itself felt, for the development of centers where art crystallized
about the beliefs and customs of the people. Thanks to the mildness of the
climate, the care with which the graves were constructed and the practice of
burying with the dead the objects most prized in life, the material for the study
of Chiriquian ceramic art is not only most plentiful, but also especially well
preserved.

CLASSIFICATION.

Holmes classifies Chiriquian pottery under two general heads, each with sub-
divisions, as follows :

a. Unpainted ware :

1. Terra cotta or biscuit group.

2. Black incised group.

b. Painted ware :

1. Scarified group.

2. Handled group.

3. Tripod group.

4. Maroon group.

5. Red line group.

6. White line group.

7. Lost color group.

8. Alligator group.

9. Polychrome group.

It is admitted that " the characters upon which the classification is based are
somewhat heterogeneous and include material, color, shape, finish, ornamentation,
method of manufacture, and evidences of use. No single character and no one
group of characters can be relied upon to distinguish the different groups. We
must depend, therefore, upon an assemblage of characters or upon one character
in one place and another in another place." The foregoing classification was used
" mainly as a means of facilitating description."

UNPAINTED WARE. 47

It would be difficult to devise anything essentially different that would be better
than the classification proposed by Holmes ; yet it is by no means perfect. There
are more tripods, for example, in his " biscuit group " than in the " tripod group,"
so-called. And the tripod group has many characters in common with the
" handled group." Again, he places the tripod group with the painted ware, but
many of them were never painted at all. As a large proportion of the unpainted
tripods that come under the head of " painted ware " had been used over the
fire, they cannot be looked upon as unfinished specimens. There was evidently
no thought of painting the pieces in question.

Since the " white line " group is such a small one and resembles so closely
some of the pottery from Costa Rica, it might have been well to eliminate that
group altogether from the discussion of Chiriquian pottery. The same may be
said of the " red line " ware, which also resembles a variety of Costa Rican
pottery. But specimens of these two groups have undoubtedly been found in
Chiriquian graves. Holmes's monograph is not only so well done but also so
well known that to make serious changes in his classification would lead to un-
necessary confusion. We shall, therefore, retain the white line and the red line
groups.

Quite as distinct as either of these is another small group in the Yale collection,
which is not mentioned by Holmes. It consists of a series of chocolate-colored
tripods with incised ornamentation ; this will be more fully discussed under the
head of the "chocolate" ware. It, also, has marked Costa Rican affinities.

There is still another series of vessels that does not seem to have been touched
upon by Holmes. In form it resembles somewhat the handled group ; and, like
the latter, almost all the pieces had served as cooking utensils. None, however,
were painted. The paste, too, differs from that of the handled group, being darker
in color and averaging somewhat coarser. In some instances there is a distinct
salmon-colored slip, reminding one of the slip in one variety of the biscuit group.
This series evidently stands between the handled group and the biscuit group.
None are mounted on tripods. It may, for convenience, be styled the unpainted
variety of the handled group.

I think that three important changes should be made in the classification of
Holmes. On subsequent pages, I shall give my reasons for proposing the name
armadillo group in place of " terra cotta " or " biscuit " group ; serpent group instead
of "• black incised " group, and fish group instead of " tripod " group.

UNPAINTED WARE.

As a matter of convenience Chiriquian pottery may be classified as belonging
to two grand divisions depending on the use or non-use of paint. The line of
separation, however, is not so distinct as those defining the limits of the groups
composing each division. We find, for example, that one variety of handled ware
is painted, while another is unpainted. On the other hand, some examples of the
tripod or fish ware, one of the groups of painted pottery, do not seem to have
ever been painted. With these exceptions, the various groups fall wholly either
within the painted class or the unpainted class. Biscuit or armadillo ware, for

48 A STUDY OF CHIKIQUIAN ANTIQUITIES.

example, is always unpainted, and alligator ware is always painted. It is worthy
of mention that the largest group of unpainted ware probably served ceremonial
purposes, while vessels of one group of painted ware were especially employed
in ways utilitarian.

THE ARMADILLO GROUP.

This is the large group called by Holmes " terra cotta " or " biscuit " ware. It
includes " only the pale grayish yellow and reddish tints of the burned clay."
This group is represented in the Yale collection by over sixteen hundred pieces.
A majority of the specimens are tripods. Almost as many are round-bottomed
vases. Sixteen have annular bases. The handled and tripod groups combined
number but six hundred and sixty- three, only two hundred and seventy-three of
these being tripods. If the collection as a whole may be considered a represen-
tative one, and there is reason to believe that such is the case, then the tripod
variety of the armadillo group has the first claim to consideration as the tripod
group, since its numbers are far greater, both absolutely and relatively, than those
of the tripod variety of the handled group.

In beauty of form and ornament, the specimens of the armadillo group are
inferior to none. Perhaps better than any other class they typify the plastic skill
and unerring taste of the ancient Chiriquian potter. There are only a few generic
forms, but an almost bewildering number of individual variations. In point of
size, also, much latitude was allowed. The smallest vessels are to be found among
the cups. The vases average the largest, while the tripods occupy an inter-
mediate place. The smallest cup in the collection has a capacity of but six cubic
centimeters ; the tripods vary in capacity from twenty cubic centimers to about two
liters (2000 "); and the largest vase holds 6.6 liters (6600 cc). The walls are care-
fully finished, both inside and outside, and are of uniform thickness. The walls
even of the largest vases are quite thin, averaging not more than four-tenths of a
centimeter in thickness. Flat bottoms are practically wanting. Handles are com-
paratively rare, being either single or paired and of the vertical loop type.

The paste used in the armadillo group may be likened to that of our modern
porous flower pot, but in color it is not quite so pronounced a red. A cup-
shaped vessel with legs, holding 335 cubic centimeters, was filled with water and
left in a room at a temperature of seventy degrees Fahrenheit. It was emptied
by combined evaporation and filtration in seventy-two hours.

A question arises as to the function of these artistic vessels. Holmes thinks
that it could not have been of a domestic nature, " as they show no evidences
of discoloration or wear." This is not strictly true of the specimens in the Yale
collection. Out of a total of 1620 pieces, eighty-one or exactly five per cent
bear marks of having seen service. These marks generally consist of a greasy
smoky surface over both the exterior and interior. The use, however, was not
necessarily domestic ; it may have been ceremonial. A probable key to the nature
of such ceremonial use may possibly be found in Seemann's * description of the
rites attending the death of a chief among the Indians of Panama at the time of

1 Op. cit., I, 316.

THE ARMADILLO GROUP.

49

the discovery : " The corpse after being enclosed in the best blankets (mantas),
and decorated with golden ornaments, was suspended over a fire, and the grease
dropping out, carefully collected into earthen vessels; when dry, the body was
interred, or, in some districts, preserved above ground."

Beginning with the more primitive forms, figure 55, from El Banco, resembles
in shape an oblong wooden cup with rectangular rim. Another elongated form
of equally rude workmanship is given in figure 56. The wooden angularity of
outline, however, has disappeared, the rim being oblong and developed at one
end into a diminutive handle which suggests the stem of a fruit, of which the
vessel itself would represent one half. It "comes from Divala, twenty-five miles
west-northwest of David.

Fig- 55-

Fig. 56.

Fig- 57-

Fig. 58.

Fig. 59-

Fig- 55- — Primitive rectangular bowl ; from Kl Banco. Armadillo ware. '/'

Fig. 56. — Primitive elongated bowl with small handle; from Divala. Armadillo ware. '/>

Fig 57- — Calabash type of cup with diminutive handle; from near Bugavita. Armadillo ware. '/'

tig. 5^. — Hemispherical bowl with spoutlike projection of rim ; from near Bugavita. Armadillo ware. '/>

Fig 59- — Calabash type of bowl with no projection of rim. Armadillo ware. V*

The hemispherical cup reproduced in figure 57 is related to the foregoing type,
although of superior workmanship. Its prototype was the vessel made from part
of a gourd or from the fruit of the calabash tree (Crescentia cujete). A slight pro-
jection at one point on the rim may represent the stem of the fruit, as well as a
handle that is practically functionless because of its small size. In figure 58, the
rim projection is not only much increased but is grooved so as to form a sort
of spout instead of handle. These last two pieces are from the same locality (six
mill's northwest of Bugavita). Figure 59 is an example of the calabash type without
any rim protuberances.

MEMOIRS CONN. ACAD., Vol. III. 7

50

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Fig. 60.

Variations from the calabash type are seen in Plate V. One of these may be
brought about by the incurving of the rim (fig. a), thus restricting the area of the
mouth opening. Another step is taken in a vessel from Escaria (fig. 6), which is
depressed vertically. The incurving of the rim leads quite naturally to angular
outlines. Another method of producing angular outlines is by carrying the sides
of the vessel up vertically, as shown in a bowl from Bugavita (fig. c).

If after building the sides of the vessel
past its greatest horizontal diameter, the
rim be carried up more or less vertically,
there is obtained the bowl-shaped vase seen
in figure d. The neck may make a variety
of angles with the shoulder. It may be in-
sloping, vertical, or flaring. It may be long
or short. The relatively tall vase with
pointed bottom is an example of the flaring
neck (fig. e). It and the foregoing are both
from Divala. A long flaring neck, with a
pronounced lip as an accompaniment, is
given in figure /. Sometimes four slight
protuberances appear at the corners of the
mouth. These maybe developed horizontally
to form more or less square lips about a
circular opening (fig. g).

Again the neck may have two stories,
the lower in-sloping and the upper flaring,
as in figure /* (from Divala). The lower

Fig. 60.— Tripod bowl with plain hollow supports, story, which hereafter will be called the

collar, is often chosen as a field for incised
ornamentation.

As we proceed, it will be seen that with

one exception the entire ceramic art of the ancient Chiriquians is but an elabo-
ration of the foregoing elementary forms — forms that had their origin in vessels
made of wood, the gourd, the calabash, etc. The simplest elaboration is the adding
of three legs to form a tripod. These began, no doubt, as short pegs. They soon
grew in dimensions and were made hollow. They were then supplied with earthen
pellets as rattles, and slit so that the sound might not be muffled (fig. 60).

Tripods of the armadillo group, as has been noted, are very numerous and
about sixty-five per cent are of the calabash type, i. e., hemispherical in shape.
Only about fifteen per cent are angular in outline, and the remainder (twenty per
cent) are supplied with necks. Collars are rare. Only two or three of the tripods
have an oblong bowl with angular rim, and perhaps as many are so altered in the
equatorial zone by ornaments in relief as to disguise somewhat the essentially
spherical nature of the bowl, thereby suggesting some animal form ; the crab, for
example.

Sometimes the legs have the appearance of being inflated and thus stand out
from the body of the vessel in such a manner as to be quite conspicuous, even

Armadillo ware. '/«
Fig. 61. — Tripod bowl with plain hollow supports,
apparently inflated. Armadillo ware. '/>

THE ARMADILLO GROUP. 51

when the latter is viewed from above (fig. 61). Again, they may end in a blunt
point. In some cases the tripod bowl is open and shallow, resembling a halved
calabash, in others it is almost spherical. An example of the rarer angular bowl
is given in Plate VI (fig. a). Tripods with a single median slit in the leg are
even more numerous than those with two lateral slits in each leg. To this type
belongs a rare double tripod (fig. J), the bowls being united by a sub-cylindrical
tube, slit, and carrying a single ball of clay. The shoulder ornaments probably
represent the armadillo. In figure c each leg has four vertical slits, two lateral
and two median. This brings us to a rather small but interesting group of tripods
with multiple slits in each leg, the slits being usually short and seldom vertical
(fig. d).

The pellets, etc., with which the hollow legs of tripods are supplied, are as a
rule little balls of clay, numbering from one to half a dozen or more in each leg.
In rare instances, calcareous concretions are used in place of clay balls. Mr.
McNiel sent the following note with the collections he sold to Yale University
Museum :

David, Chiriqui, U. S. C. May, 1879.

Herewith I send samples of what seems a curious formation, found imbedded in soft friable
rock which outcrops near steamer's landing, 3 miles from David. I am not certain whether
all the pebbles found in the legs, etc. of the pottery from the ancient graves are of this
material, but I found, on examining broken specimens, their identity. I believe the rock
in which they are found to be tufaceous, being much honey-combed by irregular cells.

J. A. McNiel.

As far as the Yale collection is concerned, the use of these concretions as
pellets for tripod legs, etc., is very rare indeed, occurring perhaps in a single
case only (fig. e). The tripod in question, which is of high artistic merit, is much
discolored by grease and smoke.

The rattles in the legs of some of the tripods are not pellets at all ; but simply
little masses of clay, irregular in shape, that were pushed into the hollow of the
legs when the latter were slit or punctured. This is particularly true of the group
of tripods last referred to, in which each leg bears multiple punctures in the form
of short slits, crosses, etc. In these examples, the leg was first attached to the
body of the tripod, then punctured. As a rule, the character of the margins of
the slits would seem to indicate that the punctures were made after the paste
had become dry and hard, perhaps after baking (fig. /). With a mallet and small
punch, both of wood, I was able to produce similar punctures on baked specimens
without shattering the hollow leg. The slits in figure (j could have been made
in the same way. They were lengthened by three or four successive light taps,
after which the end punctures were made with the same instrument. The foregoing
examples are both from Bugavita. Another tripod vase in which the leg rattle
is supplied only with irregular fragments of burnt clay is reproduced in figure Ji.
Here some of the punctures take the form of a cross. This piece comes from
Escaria.

It did not require a wide stretch of the imagination to arrive at the zoomorphic
possibilities of the plain tripod leg. By the application of nodes and fillets of
clay to the hollow tripod supports they immediately assume animal forms, as

52

A STUDY OK CHIRIQUIAN ANTIQUITIKS.

illustrated in figure 62, which suggests man or the ape. Something similar is
seen in an example from Bugavita (fig. 63). The long arms are those of the
monkey, but where the head should be there squats a little round-bodied animal
with long tail.

But in all probability the first efforts to convert the ordinary tripod leg into an
animal form were much more realistic in their results. Take, for example, figure 64

Fig. 62. — Tripod with zoomorphic supports.
Armadillo ware. '/»

Fig. 63. — Tripod bowl ; from Bugavita.
Armadillo ware. V>

(also from Bugavita). A few nodes and fillets, plain and incised, added to the
usual type of support bring out the form of a frog, complete in every detail.
Somewhat similar and equally good results are produced by the same means in

Fig. 64. — Tripod with froglike supports ;
from Bugavita. Armadillo ware. '/>

Fig. 65. — Tripod of classic form, with
froglike supports, and rim ornamented
with armadillos. Armadillo ware. '/»

figure 65. The bowl of the latter is classic in form. The rim is ornamented with
two armadillos. The frog is a favorite shoulder ornament on vases of the biscuit
or armadillo group.

As the plain leg became more and more disguised by the application of nodes
and fillets, its identity seems to have been lost sight of, as seen in Plate VII,
where it is a caricature of the human form. In figures «, d and e, there is very
little alteration of the plain tripod support. A bifurcation is only hinted at in

THE ARMADILLO GROUP. 53

figures &, c and /;, with a suggestion of toes in the last two. The supports in figures/
and g are frankly human, the representation of the loin-cloth removing all doubt
on this point. The bowl in the former is exquisitely turned, resembling in its
outline the Venetian renaissance goblet.

The identity of the primitive leg form is completely lost in figure 66. Each
support is an apelike head with four projecting tongues that fill and even greatly
distend the mouth. If the tripod were inverted it would show a drooping left ear
on each head. The potter's sense of humor shows to good advantage here as well
as in many succeeding illustrations. From the view-point of modeling, this piece
has no superior in the whole collection. The diameter of the cup is everywhere
equal, as is that of its rim, and the distance separating the grotesque heads is
everywhere the same. Other specimens from this locality, eight leagues west-
northwest of Bugavita on the Acoo river, are also beautifully modeled.

Grotesque heads are often attached to the rim of tripods, as illustrated in
Plate VIII, figures a and b. These heads occur singly or in twos, one on either

Fig. 66.— Tripod bowl of superior modeling, sup- Fig. 67. — Tripod representing a primitive zoo-

ported by three grotesque apelike heads ; from near morphic type. Armadillo ware. '/>

Bugavita. Armadillo ware. '/'

side. Sometimes a tail takes the place of one of the heads. In that case there
may be four feet with toes pointing in the direction of the head, instead of three,
as in figure c. The animal head on the rim in this piece resembles that of the
lama, which probably means either Peruvian influence or formerly a wider geo-
graphic range north. It will also be seen that the whole body of the animal,
instead of the head only, may appear on the rim (figures rf, e and /) ; and that
zoomorphic features may be shifted to the equatorial zone, altering somewhat the
more or less spherical form of the bowl (figures g and h\

A further step toward the conversion of the phytomorphic into the animal type
is taken in figures 67 and 68. The sculptor's treatment of the eye is an interesting
study. It is sometimes a plain node. More often the node is flattened and marked
with an annular indentation such as a straw or hollow reed would make. But
the most common method of all is to represent the eyelids by an incised line
across the center of the flattened node. When the latter is comparatively large
and set in a shallow cavity, it may look exactly like the head of a screw.

There is a curious little group of tripods, showing a pair of prominent screw-

54

A STUDY OF CHIRIQUIAN ANTIQUITIES.

head eyes on each support. Practically all of them come from one locality.
Divala. In some of these, the flattened node is dispensed with and a straight
incision made in the shallow socket. This leads to a new phase of the sculptor's

art, where the effect of
relief is produced by
shadows only, as wit-
ness the spacious empty
sockets in figure 69. The
deeper shadow of a little
hole sunk in each may
represent the pupil. An-
other way of represent-
ing theeyeballsbymeans
of shadow is shown in
figure 70. The horizontal
incisions on each sup-
port are probably in-
tended to imitate the
carapace of the arma-
dillo.

The fish which, as will be seen, is such an important factor in the so-called
tripod group, is seldom reproduced in the armadillo ware. An example is shown
in figure 71, which illustrates a pair of pectoral and ventral fins, respectively, and
two dorsal fins, serving to transform a plain solid support into the body of a fish.

Fig. 68. — Vessel of zoomorphic type, with four legs.
Armadillo ware. '/'

Fig. 69. — Tripod bowl showing sculptor's use of shadows in
modeling the eye. Armadillo ware. '/>

Fig. 70. — Tripod bowl exhibiting further
use of shadows in modeling the eye.
Armadillo ware. '/>

The foregoing tripods all belong to what Holmes l calls the terra cotta or biscuit
group of ware. This group also comprises vases, bowls and cups. The characteristic
types of tripod are well rendered in the preceding illustrations, the ornaments
being confined chiefly to the supports. Among the vases, on the other hand, the

1 W. H. Holmes. Ancient art of the province of Chiriqui. Sixth aim. rept., Bur. Amer.
ethnol., 67, 1884-85.

THE ARMADILLO GROUP.

55

decoration is shifted to the shoulder, neck and handles, there being no supports
to decorate. In shape, the body of the vase is often identical with that of the
tripod. To convert a tripod into a vase, one has but to remove the applied
supports. To make the kinship more apparent, shoulder ornaments or handles
are sometimes added to the tripods (see figs. 65 and 83). Such additions, especially
of handles, are quite rare.

Fig. 71. — Tripod bowl with solid fishlike supports.
Armadillo ware. '/>

Fig. 72. — Bowl with annular base ; from El Banco.
Armadillo ware. '/«

Another form of support suggested itself in the annular base or foot, as seen
in a specimen (fig. 72) from El Banco. This form of foot, like the feet of the
tripod, increased in size, was sometimes made hollow, slit, and supplied with pellets,
as in figure 73, a handsome goblet-shaped vessel from Bugavita. The latter specimen
has a greasy smoky exterior, the blackest part being the concave bottom of the base.

Fig. 73. — Goblet-shaped vessel ; from Bugavita.
Armadillo ware. '/>

'g. 74- — Bowl supported by annular base
and human figures. Armadillo ware. '/'

In the specimen exhibited in figure 74, the bowl is supported not only by an
annular base but also by three fantastic human forms. There is a hand at either
end of each long arm, making four in all — two resting on the head and two on
the stomach. The thorax is eliminated. Two animal heads adorn the rim, and
pellets convert the hollow base into a rattle.

Legs and annular bases alike were put on after the bowl was shaped but
before the application of the slip and the polish. Handles, plain and ornamental,

56 A STUDY OF CHIRIQUIAN ANTIQUITIES.

were also added, but even these did not suffice to obliterate evidence of the fact
that the vessel had passed through the calabash stage. All such vessels we
shall call pftytomorphic. Almost ninety-eight per cent of the whole Chiriquian
collection, painted and unpainted, belong to this class. The zoomorphic class is
manifestly of later development. It includes all pieces that, taken as a whole,
form each a complete zoomorphic unit. Animal forms appear, of course, in the
phytomorphic class, but only as supports, handles or ornamental features.

In studying the elaboration of simple forms, it is often possible to determine
which is the earlier and which the derived. In comparing one group with another,
however, the evidence is not so abundant, yet it may be possible to determine
which group is the earlier and which the later.

But when it comes to individual specimens, it would be impossible, in the
present state of our 'knowledge, to say that any or all of the specimens of terra
cotta ware antedate any or all of the examples of the alligator group, for example.
Nevertheless, my personal belief is that the ancient inhabitants of Chiriqui were
making biscuit ware of a high order of excellence before they ever learned to
paint figures of the alligator that occur so often in the group of that name.

The armadillo group certainly stands for a high degree of perfection in the
potter's art. But a high grade of excellence may appear at an early stage in
any great art-awakening, as is attested by the paleolithic sculpture, engraving and
frescoes of the Dordogne, France, the Phidian age in Greek sculpture, and the
Italian paintings of the XVth and XVIth centuries.

Perhaps the chief reason for believing that the armadillo group represents an
early stage of excellence lies in the fact that individuality is stamped on almost
every piece. The fundamental forms are few ; but the variations in the execution
of these forms are multifarious. The modeling seldom fails to please, and the
interior is finished with the same care as the exterior, with the exception of the
zone nearest a constricted mouth that is too small to admit the hand of the
modeler. The ornamentation is confined to figures in the round, and patterns
incised and in relief; but is always as tasteful as it is rich in variety.

Contrast these manifold variations in form and ornamentation with the monotony
of the handled group, for instance, and the latter's shop-made character becomes
most striking. With the addition of the new element of color, one would expect
to see progress in ornament if not in form, yet such is not the case. The same
unattractive form is repeated again and again, the only variable factor being in
the treatment of the handles.

From the view-point of form, the lost color group and the alligator group are
also more monotonous than is the armadillo group, and less attention is paid to
modeling. This is particularly true of the interior of the vessels, which is left
in the rough, thus causing the sides to be thick in some places and thin in others.
The bottom is generally quite thick. The latter condition may have been intentional
in order to secure more stable equilibrium. Such work may be compared to
that of an artist who is in too great haste to give up drawing for the palette
and brush.

The Armadillo Motive. — Among the animals represented in this group, the arma-
dillo is easily predominant. In speaking of the American zoological provinces

THE ARMADILLO GROUP. 57

north and south of the Mexican table-land, Seemann1 says: "The armadillo, for in-
stance, which indisputably belongs to South America, is found in no part of Panama ;
but again appears in the neighborhood of Mazatlan, in latitude 23° 12' 0" north."

This is probably incorrect. A recent letter from Dr. J. A. Allen, Curator of the
Department of Mammalogy and Ornithology, American Museum of Natural History,
states that two species belonging to different genera are found on the Isthmus
of Panama at the present time, one known as Cabassous centralist, the other being
the wide-ranging nine-banded armadillo, Tatu novemcinctum. The armadillo was
evidently an object familiar to the ancient potters of Chiriqui. As it now inhabits
the Isthmus, it did so, no doubt, on the occasion of Seemann's voyage in 1853.
One of the two species above mentioned, Tatu novemcinctum, which has " an ex-
tensive range from Texas to Paraguay," 2 is recognized in the ceramic art of
Chiriqui by the treatment of the carapace, which is composed of three quite
distinct regions, the banded region differing in structure from the anterior and
posterior sections. This triple division of the carapace is faithfully reproduced in
many instances. It may be done by three incised fillets (see figs. 77 and 79). A
more exact delineation, however, is arrived at in painted ware, when the banded
region is represented by transverse parallel lines and the anterior and posterior
division by cross lines or spots (see PI. XLI, fig. a). In one instance (see fig. 270),
the middle section is composed of three bands in relief, painted red ; while over
the shoulders and the hips, the smooth surface of the carapace is painted black.
This specimen more nearly resembles the small Tatu tricinctum of South America
than it does the larger nine-banded species.

That Cabassous centralis also served the ancient potter as a model is highly
probable. The three regions of the carapace are almost as distinct as in the nine-
banded armadillo, but Cabassous is not quite so large and its head, ears and tail
are proportionately shorter.

In the development of art, it would be difficult to estimate how far realism lies
from the very first steps toward imitation on the one hand, and from the last
stages of conventionalism on the other. The stages at either end of the line
may often appear to be identical. It is therefore hazardous to decide whether a
given representation be a first attempt at copying a model or the work of a man
with a background of art inheritance and training, who can suggest the whole
by a skilful elimination of non-essentials.

If the line of art development were plotted, it would probably be found to rise
rather suddenly to the acme of realism, and then drop slowly to about its original
level. The accompanying series of illustrations, however, does not begin at the
beginning but rather at the crest of the realistic wave, and descends gradually
to the trough, probably that one lying on the conventional side ; yet some of
the stages shown might just as well be steps in the ascending, as in the descend-
ing scale. In other words, a definite chronological sequence has not yet been
established. There are reasons, however, for the belief that a representation of
the entire armadillo came first, after which certain parts, such as the carapace,

1 Op. cit, I, 261.

2 Flower and Lydekker. An introduction to the study of mammals, 201, London, 1891.
MKMOIRS CONN. ACAD., Vol. III. 8

58

A STUDY OF CHIKIQUIAN ANTIQUITIES.

were seized upon to take the place of the whole. Finally a single band of the
carapace was chosen, not only to stand for the whole animal, but also to be used
as an ornamental motive pure and simple.

Three realistic armadillos support the tripod shown in figure 75. The work
has an unfinished appearance, but the parts are all represented, even to the three
regions of the carapace. In figure 76, the parts are all present, but the legs of
the armadillo are reduced to feet only. In a specimen from Escaria (fig. 77) the

Fig. 75. — Bowl supported by three realistic armadillos,
dillo ware. '/>

Arma-

Fig. 76. — Howl with similar tripod supports
the extremities of the armadillo being
reduced to feet only. Armadillo ware. V1

reduction is carried further by the total elimination of the hindlegs and hindfeet.
The one essential feature of the armadillo is the carapace. This persists after all
other parts have been eliminated, as shown in figure 78, where it is represented
by horizontally arranged incisions.

A more detailed series showing the gradual elimination of parts is given in
Plate IX. The armadillo is represented in the round in figure a. The simple

g' 77- — Tripod bowl in which the hindlegs and feet of the
armadillo are eliminated from the supports. Armadillo
ware. '/»

Fig 78. — Tripod the supports of which are
ornamented with carapace motives only. Ar-
madillo ware. V»

tripod support in figure b is incised to represent the carapace ; an indifferent head
is added and the hindfeet are brought forward close to the forefeet. In figure c,
the hindfeet have entirely disappeared and the forefeet are conventionalized, a
flattened node with parallel incisions representing toes. This becomes eventually
a decorative motive that is used independently of the armadillo or any other animal
form. The head is considerably reduced and simplified in figures d and e, the

THE ARMADILLO GROUP.

59

legs and feet having disappeared completely. In the latter figure, the carapace is
suggested by the nail marks of what appears to be a female hand. In figure </,
the head and forefeet reappear as plain nodes, while the carapace is wanting.
The series ends (fig. f) with simple tripod supports horizontally incised to represent
the carapace, the armadillo motive par excellence.

Fig. 79. — Vase with shoulder ornament in which the
carapace motive is present on ventral surface only.
Armadillo ware. '/"

Fig. 80. — Vase with shoulder ornament showing
but two divisions of the ventrally placed cara-
pace. Armadillo ware. '/«

We have seen that many tripod supports were fashioned to resemble the
armadillo, sometimes highly realistic and sometimes conventionalized. In eighty-
one of the biscuit vases, the head of the armadillo is employed as a shoulder
ornament, the back of the head always being turned toward the aperture of the
vase. The carapace, there-
fore, if it were visible at
all, would have to be shifted
from the dorsal to the ventral
side. Such is the case in
figure 79, where the three
regions of the carapace are
indicated by the indented
horizontal fillets, each being
interrupted for a space in
the median ventral line,
which is only a concession Fi&- 8l--Vase ^ith sho«!del: °rna-

7 ment and only one division of

to realism, since the bands the ventrally placed carapace.

Armadillo ware. '/'

Fig. 82. — Cup with shoulder
ornament on which an un-
broken band represents the
ventrally placed carapace mo-
tive. Armadillo ware. '/'

do not reach the median

line in nature. A ventral

view would include the margins of the carapace on either side. It is important

to note, however, that no attempt is made to represent the dorsal part of the

carapace where it would not be easily visible. Only two divisions of the cara-

60

A STUDY OF CHIRIQU1AN ANTIQUITIES.

pace are present in figure 80, while in figure 81, but a single section remains.
The eyes and ears still persist.

Sometimes the carapace is carried across the ventral surface without a break
at the median line. An example of this is seen in figure 82, where a single band
represents the entire carapace. This band does not appear on the dorsal surface.
The identity of these shoulder ornaments might be in doubt were it not for the
occasional ventrally placed carapace motives. One of the connecting links be-
tween the tripods and vases (fig. 83) belongs to this type. A single element of
the carapace is placed on the ventral side, the ends of the two incised fillets almost
touching the median ventral line ; while another carapace element is applied longi-
tudinally over the vertebral column. The eyes, ears and nose are all prominent.

Fig. 83.

Fig. 84.

I'ig- 83. — -Tripod vase with shoulder ornament to
which carapace motives are applied both dor-
sally and ventrally. Armadillo ware. V>

Fig. 84. — Large vase with armadillo shoulder or-
naments from which the carapace motive is
absent. Armadillo ware. '/»

In a majority of cases the carapace is entirely eliminated, the forelegs
usually and quite naturally taking its place (PI. X, figs, a and b}. The
animal foot reduced to its lowest terms is seen in figure c, it being a flattened
node with parallel incisions. This conventionalized foot will appear purely as an
ornament in a group of vases to be described later. Even the conventionalized
foot does not always accompany the armadillo head (fig. d). The reduction in
parts and in size of the head itself reaches the limit in figures e and /.

Two of the largest vases belonging to the armadillo ware have armadillo heads
in the round for shoulder ornaments. One of these is reproduced in figure 84.
The shape is graceful and the modeling so perfect as to compare favorably with
wheel-work. The ears of the armadillo are placed low, the eyes are wanting,
and the nose is much depressed. Vases of the armadillo ware even finer and
larger than this one have been found at Boruca, south of Terraba, Costa Rica.
Mr. Minor C. Keith has one splendid specimen that came from a grave at Mercedes
on the Atlantic coast of Costa Rica. It was evidently an importation, but the

THE ARMADILLO GROUP.

61

geographic distribution of this
ware must be extended at least
to take in the region bordering
on the Gulf of Dulce in Costa
Rica.

The entire armadillo is often
used, also, as a shoulder de-
coration on vases of this group.
Some good examples are given
in Plate XL In figure a the
parts, even to the three regions
of the carapace, are all fairly
well indicated. The forefeet rest
on the rim, the hindfeet on the
shoulder of the vessel. A small
vase from Escaria (fig. 6) shows
the armadillo placed somewhat
lower and in an attitude as if in
the act of burrowing with its
unusually long nose. In fact,
the burrowing attitude is char-
acteristic and is very satisfactorily
shown in figure c. Note that
the feet here are flattened nodes
with parallel incisions. In other
words, they are foot symbols
that may be used entirely alone,
independent of the animal form
to which the foot originally be-
longed, as will be seen in sub-
sequent illustrations (see fig. 90).
The tail is incised with a so-
called herring-bone pattern, the
lines running, however, in the
opposite direction from what one
would expect. The flaring rim
of the vase is set so as to pro-
duce a diaphragm where the
neck is most constricted. The
use of the foot symbol as an
independent ornament is shown

Fig. 85.

Fig. 86.

Fig. 87.

Fig. 88.

in fiffUre d, where it OCCUrS in Fig. 85.— Neck of vase ornamented with the tail motive. Arma-

,.,. a/

pairs on the neck of the vase. Rg. ^.— Another example of the tail motive. Armadillo ware. ''•

By a comparison of the arma- Fig. 87.— Neck of vase ornamented with alternating groups of tail-

,.,, ., . r. '4.1 *i and eve-motives. Armadillo ware. */'

dlllo tails in figure e With the F|g gg j_Neck of vasc decorated with tail motive in meander, and

Ornament on the neck of the an eye symbol in each angle. Armadillo ware. ''«

62

A STUDY OF CHIRIQUIAN ANTIQUITIES.

vase in figure /, it will be seen that the tail was used apart from the animal as

a decorative feature.

The tail motive is often repeated in a series that completely surrounds the

neck of a vase otherwise devoid of plastic animal ornamentation (figs. 85 and 86).

In figure 87 groups of beautifully executed tail motives alternate with groups of

a motive derived apparently from the. eye. Another effective pattern is obtained

by carrying the tail symbols in meander around the neck of a vase, each angular

space being filled in by an eye symbol, as in figure 88.

The most characteristic feature
of the armadillo is its carapace.
It ranks first among armadillo
motives. We have already noted
how one band of the carapace
was made to stand for all three
divisions. Why could it not
have been used independently
to represent not only the whole
carapace but also the whole
animal? A careful comparison
of the incised fillets that adorn
the neck of the vase shown in
figure 89, with the incised fillet
that represents the carapace of
the armadillo on the shoulder
below, proves them to be identi-
cal. We have then a carapace
symbol as well as a foot symbol
and a tail symbol. There remain
the flattened nodes with annular

Fig. 89.

Fig. 90.

indentations seen on the neck
of the same vase. Their coun-
terpart is somewhat sparingly
used to represent the eye, par-
ticularly of owl-like and monkey-
like heads. They are, no doubt,
eye symbols, but they are not
confined to Chiriquian art. On
the contrary, this particular or-
nament is common to the art
of various times and peoples. The carapace symbol like the tail symbol makes
an attractive meander in combination with either foot symbols (fig. 90) or eye
symbols (fig. 91). With due regard to their mode of occurrence in nature the
foot- and eye-symbols alternate in groups of four and two, respectively, forming
an original pattern (fig. 92), but one not so pleasing as the foregoing meanders.
Foot, tail, eye and carapace, all may become purely decorative elements used
independently of the whole animal and independently of each other ; separately

Fig. 91.

Fig. 89. — Neck and shoulder ornament of a vase proving the
identity of the carapace motive. Armadillo ware. '/'

Fig. 90. — Meander of 'carapace symbols accompanied by foot
symbols. Armadillo ware. '/•

Fig. 91. — Double neck of vase, ornamented with carapace- and
eye-symbols. Armadillo ware. '/>

THE ARMADILLO GROUP.

63

or in combination one with another, producing patterns as pleasing to the eye as
they are original. In the figures here cited, the eye symbol is one of the com-
mon forms of representing the eye in Chiriquian art, viz., a flattened node with
annular indentations. Another type will be described on a subsequent page.

Vases, are often supplied with
a raised zone or collar below
the neck, as has already been
pointed out. The collar is
usually adorned, but not al-
ways. When it is decorated
the neck proper is generally
plain. The collar ornament
always consists of incised lines

Fie. Q2. — Neck of vase showing foot- and eye-symbols in alter- • j ,• T-I,

nation. Armadillo ware. V or indentations. These are

sometimes combined in such

a way as to produce effective patterns. Figures a and b of Plate XII show ex-
amples of collared vases, the latter being provided with an unusually large aperture.

In the Yale collection, there are several vases with two mouths. With two
exceptions, these belong to the armadillo ware. Three are reproduced in Plate XII.
In figure c, which represents an excellent example, a bridge connects the flange or
lip of one mouth with that of the other. Over this bridge two armadillos face each
other. The identity of the carapace with the carapace symbols that adorn both
necks is beyond question. Alternating with these symbols are eye ornaments,
after the fashion of the prevailing eye form in Chiriquian art.

Fig. 92.

ig- 93- — Vase, the single handle of which is
ornamented with carapace motives. Arma-
dillo ware. '/'

Fig. 94. — Vase with two undecorated handles. Ar-
madillo ware. '/>

Sometimes the paired necks are perfectly plain. Occasionally they are set
rather far apart, as in figure d, or fused as in figure e. In the latter case, a
vertical partition wall in the plane of fusion divides the interior into two equal
parts, thus making a double vessel.

64 A STUDY OF CHIRIQUIAN ANTIQUITIHS.

A study of the foregoing illustrations will bring out the interesting fact that
where either the neck or the collar is decorated there is always a flange or lip
at the mouth opening. On the other hand, where the neck (or collar) is unadorned
there is almost never a flange (see PI. X). The latter seems to be a necessary
accompaniment of the ornamented neck or collar to satisfy the sense not only
of protection for the ornament but also of appropriateness in outline.

Reference has been made to tripods with features peculiar to vases, viz., shoulder
ornaments consisting either of the armadillo head (see fig. 83) or of the entire
animal. Examples of the latter are given in figure /' (see also fig. 65). In one
instance the nose and tail are long and slender; in the other, owing to the
exposed position of the head, the nose is almost eliminated to avoid breaking.

It has been shown how the simple tripod support may be transformed into the
armadillo, conventional as well as realistic ; also how the armadillo, as a whole
Or in part, was used as a shoulder ornament, and the tail-, foot-, eye- and carapace-
symbols, as ornamental motives for the necks of the handles of vases.

The handles of the armadillo group are always placed vertically, and are usually
paired, connecting the lip with the shoulder. Of single handle there are two
types; (1) connecting the lip with the shoulder, as in the paired handles, and
(2) the handle starting from opposite points on the lip and forming an arch over
the mouth opening.

These handles are generally ornamented, as illustrated in Plate XIII. The paired
handles shown in figure a are not only adorned with three horizontally applied
incised fillets representing the three sections of the armadillo carapace, but are
also surmounted each by a pair of ears to heighten the zoomorphic effect. Those
in figure b are entirely covered with carapace motives, leaving no doubt as to
the significance of the decoration. Along the zone of union between the pro-
longed lip and the ascending ramus of the handles in figure c, nodes and fillets
are applied representing the ears, feet and carapace of the armadillo. One has
but to refer back to figure a (PI. XI) in order to arrive at the artist's point of
view, namely, that the handles illustrated in figures a to e are, to all intents and
purposes, armadillos facing each other and peeping into the mouth of the vase.
The armadillo motive in figure d is similar to that in figure c, except that it is
much more simplified.

Instead of applied incised fillets, the handle itself may be simply incised to
represent the armadillo carapace, as in figure e. From the horizontally applied
fillets and the horizontally incised handles, it is but a step to fillets and incisions
running obliquely ; and a second step to the twisted handle so common in the
tripod group, so-called, as will be seen by comparing figures b and e with sub-
sequent illustrations (see fig. 125). In figure 145, the eyes are actually represented,
so that the twisted handle evidently means a life form, probably the armadillo.

An example of one type of single handle is given in figure /. A well-developed
flange protects the applied tail- and eye-symbols about the neck of the vase,
illustrating again the sense of protection for the decorations in relief and of ap-
propriateness in outline. On the crest of the handle, three eye ornaments in a
row occupy a field bounded on two sides by carapace motives.

The armadillo motive in figure 93 is limited to a pair of incised fillets applied

REVISED CLASSIFICATION.

65

horizontally to the single lateral handle. The paired handles in figure 94 are
unadorned. A vase with angular outlines and single handle forming an arch over
the aperture and attached to the rim at either side is represented in figure 95.

An unusual treatment of the armadillo is seen in figure 96. The head barely
protrudes beyond the carapace and the latter almost encircles the body of the
animal. A long fillet encircling a head on the flat handle in figure 97 may be
a conventionalized form of the armadillo. A similar motive appears in figure a
of Plate XVI (see also fig. 130).

Fig. 96.

Fig. 95-

ig- 97-

Fig. 95. — Vase with single handle forming an arch over aperture. Armadillo ware. '/»
Fig. 96. — Cup with single shoulder ornament representing the armadillo. Armadillo ware. V>
Fig. 97. — ('up with single flat handle showing possibly a conventionalized form of the ar-
madillo. Armadillo ware. V»

Revised Classification. — From the foregoing illustrations it is quite evident that
the armadillo must have played an important decorative as well as symbolic role
among the makers of the biscuit ware. The frog is not a very close second.
After these come a number of animal forms more or less fantastically treated and
not always determinable.

Of the 3668 pieces of pottery in the Yale collection, 1620 belong to the arma-
dillo group. Of these about one-fifth are decorated with the armadillo motive in
one or several of its many forms, the occurrence of other animal representations
being rare in comparison. Among the makers of this ware, the preponderating
influence of the armadillo suggests the totemic potency of this animal. Why not,
then, call this group the armadillo group?

Such a terminology, together with the change from " black incised " to serpent
group, as I have suggested on page 47, would tend toward uniformity. For sim-
ilar reasons, the "tripod'' group of Holmes might well be called the fish group.1

1 According to Seemann, the quantity of fish, especially in the Bay of Panama, early
gave rise to the name of Panama, or place where fish abound. Fish are also abundant in
the rivers.

MEMOIRS CONN. ACAU., Vol. III. 9

66

A STUDY OF CHIRIQUIAN ANTIQUITIES.

All these changes would be in line with Holmes's choice of the name alligator
for one of the important groups of painted ware. The names of life forms would
then be given to four groups, instead of to one only.

By incorporating the proposed changes into the classification of Holmes,1 the
result is as follows :

a) Unpainted ware :

1. Armadillo group (instead of "terra cotta" or "biscuit'' group).

2. Salmon-colored group.

3. Serpent group (instead of "black incised" group).

4. Handled group.

b) Painted ware:

1. Handled group.

2. Fish group (instead of " tripod " group).

3. Chocolate incised group.

4. Scarified group.

5. Maroon group.

6. Red line group.

7. White line group.

8. Lost color group.

9. Alligator group.
10. Polychrome group.

The armadillo still enters largely into the life of certain tribes. According to
Dr. A. Hrdlicka,2 it is a cherished article of food among the Indians of western

Mexico, and medicinal pro-
perties are ascribed to certain
parts. The Cora Indians, for
instance, save the carapace
and administer decoctions of
it as a cure for stomach dis-
orders. In this connection
it is of interest to recall that
the nine-banded armadillo ap-
pears in ancient Mayan picture
writing.

As has been stated, the de-
corative features of the arma-
dillo group are limited to
figures in the round and to
patterns either incised or
in relief. They are also
generally limited in location

Fig. 98. — Vase with shoulder ornaments representing the frog, probably
the Surinam Toad (Pipa americana). Armadillo ware. 'I*

1 Op. cit., 67, 80 and 84.

a Personal letter, dated Oct. 4, 1906.

to the shoulders, collars,
necks, feet and handles of

THE ARMADILLO GROUP. 67

the vessels. Shoulder ornaments are always in the round. The frog is a favorite
subject with the potters, a number of species being represented. A rather long-
nosed variety, with eyes placed far back and a number of cup-shaped markings
scattered over the back, is given in figure 98. It is probably the Surinam Toad
(Pip a americana) and the markings may represent the eggs that are placed on the
back of the female by the male, each egg sinking into a little pouch of skin
covered by a gelatinous film. The embryos develop in these little pouches. Or
it may be the u very minute species, beautifully spotted with black and red," *
that is said to be used bv the Indians to poison arrows. The most logical way
to suggest spots would be by markings of this sort, as paint was not used in this
group of ware. A short-bodied species is represented in a vase from Divala (fig. 99).
The eyes are indicated by indentations instead of by the usual incised nodes.

Fig. 100.— Frog illustrating the cc-

Fig. 99- — Vase with frog shoulder ornaments; ramie type of technique. Armadillo

from Divala. Armadillo ware. '/> ware. V1

Other varieties are brought together in Plate XIV. Figure a shows a charac-
teristic example of the adaptation of ornament to form. The noses of the frogs
project above the level of the rim. This gracefully turned vase is from Jacu.
Mention has already been made of the influence of technique in one medium over
that in another. A good illustration of this is to be found in figure b, where the
unduly flattened and attenuated hindfeet are exactly analogous in treatment to
the hindfoot that is characteristic of the frog in metal. Evidently the potter has
here taken his cue from the goldsmith (see fig. 344). In addition to the shoulder
ornaments in the round, the collar bears a neatly incised pattern.

Another example of metal technique executed in clay is to be seen in figure c,
where the hindfeet of each pair of frogs are represented by a long flat band of
clay. One has but to compare this with its prototype in metal (see fig. 345).
In the latter case, however, only the median legs of the pair are actually united.

As a rule, figures in the round used as ornaments are attached to the finished
surface of the vessels. But in figure d, a perfectly modeled vase, each frog covers
an oval horizontally flattened opening on the shoulder of the vase. The margins
of the holes are neatly finished and, in connection with the mouth openings of
the frogs, may have served as spouts. The feet of the frogs are of the flattened
metallic type. This attenuated tvpe may have originated in an effort to represent

1 B. Seemann. Op. cit., I., 265.

68

A STUDY OF CHIRIQUIAN ANTIQUITIES.

the web-foot of the frog, something that would hardly have been attempted
originally in clay.

The goldsmith, on the other hand, sometimes borrowed forms that are essentially
ceramic in character, as when figure 100 was copied more or less faithfully in
figure 347, which represents in metal a short round-bodied frog with bulging eyes
and a median dorsal band reaching from the nose to the end of the spinal cord.
In the terra cotta frog, the body is hollow and supplied with a clay pellet; the
eyes are solid lumps of clay. In the metal frog, the bell-shaped eyes, provided
with pellets of copper, function as rattles. The ornaments on the shoulder of the
vase reproduced in figure 101 resemble the frog in the tadpole stage.

Fig. 101. — Vase with shoulder ornaments re-
presenting frog in tadpole stage. Armadillo
ware. '/>

Fig. 102. — Vase with zobmorphic shoulder orna-
ment. Armadillo ware. 'I'

The frog was abundant, especially during the wet season, January to April,
and must have been an important totemic animal. In speaking of the abundance
of toads about Porto Bello, Seemann quotes Lloyd, as follows : " So prodigious is

Fig. 103. — Vase with grotesque zoomorphic shoulder
ornament. Armadillo ware. '/»

Fig. 104. — Fantastic zoomorphic shoulder decora-
tion. Armadillo ware. /•

their number after rain, that the popular prejudice is that the rain-drops are changed
into toads (' de cada gota viene un sapo'); and even the more learned maintain
that the eggs of this animal are raised with the vapor from the adjoining swamps,
and, being conveyed to the city by the rains, are there hatched. The large size

THE ARMADILLO GROUP.

of the animals however — many of them being from four to six inches in breadth
— sufficiently attests their mature growth in more favourable circumstances. After
a night of rain the streets are almost covered with them, and it is impossible to
walk without crushing some."

The armadillo and the frog by no means mono-
polized the attention of the makers of this biscuit
or armadillo ware. Many of their efforts cannot be
referred definitely to any particular genus. Some are
droll, some fantastic, all are interesting. Figures 102
to 104 give some indication of the resourcefulness
of the potter in posing his figures. Flat-bottomed
vessels are. exceedingly rare in the armadillo group,
there being but a single example in the Yale collec-
tion (fig. 105).

The distinctively human form is not often met with
and is, as a rule, not executed so satisfactorily as
are the forms of many of the lower animals. This
holds true for the painted as well as the unpainted

. • c. j-rr? ii j- i- • i ii i • A j j Fig. IOS. — Unique flat-bottomed vase.

ware. It is often difficult to distinguish that intended Armadillo ware. '/»

for man from that designed for monkey. The most

characteristic feature of the human head is the flattening between the frontal

protuberances and the occiput. This character is well illustrated in figure 106,

Fig. 106. — Vase with handles represented by Fig. 107. — Small vase with anthropomorphic shoulder orna-
the human figure. Armadillo ware. '/' ments ; neck decorated with carapace meander combined

with foot symbols. Armadillo ware. *l*

where the standing human form takes the place of the ascending ramus of the
ordinary flattened paired handles. The human form as a shoulder ornament also
appears in figure 107. The neck of this small vase is decorated with a carapace
meander and a foot motive as an accompaniment.

A variety of plastic shoulder ornaments are grouped together in Plate XV, which

70

A STUDY OF CHIRIQUIAN ANTIQUITIES.

also affords an indication of the diversity in shape and dimensions of the rim.
Two monkeys with hands on shoulders and with the long tail brought forward
and across the chest sit on the shoulder of the vase in figure a. The green
lizard is treated more or less realistically in figure b. Two catlike creatures, each
chasing the tail of its mate, are reproduced in figure c. A remarkable form of
vase may be seen in figure d. The diameter of the body is only two-thirds that
of the rim. The latter is bell-shaped and finished in a perimetric roll. A similar
rolling of the rim upon itself is shown in figure e. The plastic figure on the
shoulder may be the armadillo, since carapace motives are placed both transversely
and longitudinally on its back. The shoulder ornament in figure / probably
represents a bird.

A number of small cups with single handle are included in the biscuit or armadillo
group. The smallest of these cups has a capacity of only six cubic centimeters,

Fig. 108.

Fig. 109.

Fig. no.

Fig. in.

Fig. 112.

Fig. 113.

Fig. 108.— Cup with zobmorphic handle decoration. Armadillo ware. '/«

Fig. 109. — Cup with decoration on handle probably representing a human head. Armadillo ware. '/«
Fig. no. — Example of anthropomorphic handle decoration. Armadillo ware. '/'
Fig. III. — Cup with grotesque human figure serving as handle. Armadillo ware. '/<

Pig. 112. — Another treatment of the same motive, the nose being developed into a proboscis. Armadillo ware. '/»
Fig- 113' — Cup with handle decoration suggesting the human form with nose developed into a proboscis. Ar-
madillo ware. '/«

THE SALMON-COLORED SUB-GROUP.

71

ig. 115. — Cup with froglike
figure serving as handle.
Armadillo ware. '/'

while the capacity of the largest vases in the armadillo group exceeds 6000 cubic
centimeters. The largest of the Chiriquian vases, however, belong to the so-called
polychrome group, one vase in the Yale collection holding twelve liters (see fig. 257).
The single handle is usually a vertical loop connecting shoulder with rim and
so altered by the application of plastic features as to suggest some animal form
(figs. 108-115). Those in figures 109 and 110 are probably human. More often
a single effigy, human or otherwise, attached to the shoulder of the cup takes
the place of a handle (figs. 111-115). The human effigy in figure 111 is a
striking example of the ancient potter's sense of humor and his skill in giving
material expression to the same. The figure sits with elbows on knees. The
right hand, raised to the nose,
partially obscures abroad smile;
while the left rests on the rim
of the cup. The cone-shaped
hat, with four nodes grouped
about the apex and secured by
a fillet that passes beneath the
chin, scarcely hides a short
" pig-tail '' behind. The conical
head-dress may be also seen in .

* Fig. 114. — Cup with human effigy

figure 112 (as well as in the for handle. Armadillo ware. '/«

handled ware). In the former,

both hands are held to an abnormally long nose that might well be styled
a proboscis. A similar feature is shown in figure 113, but the cone-shaped hat is
wanting. These are comparable to the god with a long proboscis illustrated in
ancient Mayan codices.

THE SALMON-COLORED SUB-GROUP.

There is a variety of armadillo ware, usually highly polished and with a salmon-
colored slip, the form and technique exhibited in it also differing somewhat from
the ordinary biscuit pottery. Some characteristic examples of this variety are
brought together in Plate XVI. Mention has already been made of the armadillo
represented as rolled up for defense or repose. The head projecting from the
shoulder of a salmon-colored vase (fig. a) and encircled by two incised fillets may
be, perhaps, a conventionalized rendering of the same idea. This might be taken
with propriety for a coiled serpent, were it not for the fact that representations of
the serpent on Chiriquian pottery are almost wholly confined to one small group, the
so-called black incised ware of Holmes, where the serpent motive is so all-pervading
as to justify a change of the name from " black incised " to the serpent group.

Two vases of unusual shape should be described in this connection. The
broken one is illustrated (fig. b) instead of the complete one, because of its being
the more eccentric in form. While the greatest diameter is twenty-six centimeters, the
height to the mouth opening is only eleven centimeters. The latter is oval in shape
and so small as to admit only two fingers. The specimen not figured is higher
and has a somewhat larger mouth opening, but its greatest diameter is less ; the
handles, also, are not so conspicuous as in the one figured. Each handle is

72 A STUDY OF CHIRIQUIAN ANTIQUITIES.

ornamented with a lizard-like form in relief. The bottom of each vessel is flat,
quite thin, and smoked by use over the fire. This type of vessel may have
served to heat water, on account of the comparatively great area of the surface
exposed to fire, from which it could have been easily removed by means of the
prominent handles. To empty such a utensil of its contents, however, after
its removal from the fire, would seem to have been an awkward undertaking.
The type might have served as a foot-warmer had it come from a colder
climate. The workmanship in both examples is crude and the paste coarse.
Holmes 1 figures a specimen similar to these, except that he describes the bottom
as being concave. He also states that the type is common in Peru and occasion-
ally met with in Central America.

THE BLACK INCISED OR SERPENT GROUP.

In point of numbers this is a small group and, with few exceptions, strikingly
uniform in style and decoration. The color, where not leached out, is black to
brown. With the exception of a large tripod with loop-shaped supports — the
largest tripod in the entire collection, with a capacity of 5.6 liters (5600 cc)
— the series is remarkably uniform in respect to size, the average capacity being
about 125 cubic centimeters. The walls are practically impervious to water.

Fig. lid a. Fig. n6A.

Fig. ll6a. — Vase decorated with parallel zigzag incised lines. '/» ; b. — Detail of incised pattern decorating the

bottqm of the vase shown in figure a. Serpent ware. V'

Some typical examples of this interesting group of ware are illustrated in
Plate XVII. A characteristic specimen from Divala is shown in figure a. Where
they merge into the lip, the vertical handles are crossed by an incised fillet
resembling the carapace symbol so common to the armadillo ware. Each side
of the vase is decorated by an incised pattern evidently intended to represent a
two-headed serpent with forked tongue. The body of the serpent is folded on
itself in such a manner as to produce geometric outlines and thoroughly cover
the field to be decorated. The deep incisions were made before the paste
hardened and were filled with a white substance, probably infusorial earth, white
clay or chalk, that stands out in bold contrast to the black ground.

1 Op. cit., fig. 88.

THE HANDLED GROUP, UNPA1NTED. 73

In figure i, the two-headed serpent is passed entirely around the vessel almost
three times. The serpent symbol also survives in the geometric decorations shown
in figure c (from the same locality), as well as in figures d and e.

Not all specimens of the black incised group are incised. One of these ex-
ceptions is reproduced in figure /. It is a vase with large plain handles, which
appears to have been used as a receptacle for the white substance that fills the in-
cisions on vessels of this group. The interior of a vase of about the same capacity,
belonging to the lost color group, is also smeared with the same white paste.

A vase recalling a Costa Rican (Tres Rios) variety of ware is reproduced in
figure 116 a. Two parallel incised lines, forming a band, describe the circumference
in a zigzag course, crossing what might be called the equator at twelve points,
and where they meet the lines are sometimes broken. This band is perfectly
plain, the punctate areas being the triangular fields each bounded by the band
and the equator, six of them above the latter and six below. The two elements
in the realistic serpent are here beginning to separate. The same thing is
taking place about the neck of the vessel, where there are two plain horizontal
bands and a dotted zone above them. Contrary to the general rule, an inter-
esting incised pattern adorns the bottom of this vase (fig. 1166). Two plain
bands bound a rectangular field. On each of the long sides of the rectangle are
three triangles, each bounded on two sides by a plain zigzag band. Only the
enclosures, rectangular and triangular, are dotted with incisions. The breaking up of
the elements that went to make up a realistic serpent (or serpents) results not only
in economy of labor but also in more thoroughly covering the area to be decorated I
hence conventionalism has ever been as much the child of economy as of ritual.

The breaking up of the elements that enter into the realistic representation of the
serpent is seen also in figure 117, where the shoulder of a small vase is ornamented

Fig. 117. — Small vase with shoulder

decoration consisting of a trian- p;g. ng.—Tripod bowl with supports re-

gular treatment of the serpent mo- presenting a fish. Serpent ware. '/»

tive. Serpent ware. */•

with a series of triangles formed by the body motive with the body-markings appear-
ing only in the enclosed spaces. This series of illustrations ends with a tripod vase
(fig. 118), the body of which is plain. The supports are representations of the fish.

THE HANDLED GROUP, UNPAINTED.

There is a comparatively large group of unpainted ware that is related on the
one hand to the salmon-colored variety of armadillo ware, and on the other, to
MEMOIRS CONN. ACAD., Vol. III. 10

74 A STUDY OF CHIRIQUIAN ANTIQUITIES.

the handled group that is painted. The paste used in its manufacture may be
either coarse grained or fine grained, but is always tenacious. The exterior of
nearly every piece is blackened by smoke and soot. The vessels were not sus-
pended over the fire, the convex bottom over a considerable area being generally
not discolored, as if it had rested on ashes or sand while the fire burned on all sides.

The vessel illustrated in figure 119 is coated with smoke, soot and grease, in-
side and out, which may account for its remaining absolutely impervious to water
after a test of twenty hours' duration. Life symbols adorn the otherwise plain
vertical handles, and the shoulder is ornamented with four neatly incised, X-shaped
patterns, two on either side. This vessel has a capacity of 575 cubic centimeters.
Another type of handle, one characteristic of the painted handled ware, is shown
in figure 120. It consists of a loop horizontally applied to the shoulder. Two
plain nodes for eyes or ears indicate a life form.

A series typical of those that are related to the armadillo ware may be seen
in Plate XVIII, and appropriately enough each example is ornamented with arma-
dillo motives. The paired handles in figure a form raised horizontal loops, fitting
close against the neck of the vase. By the application of nodes and fillets the

Fig. 119. — Vase bearing life symbols

on handles and incised X-shaped or- Fig. 120. — Vase with horizontally placed

namentations on shoulder. Unpainted handles representing animal forms. Un-

handled ware. '/' painted handled ware. '/•

handle is transformed into the head, shoulders and forelegs of the armadillo. To
make sure of its identity, the carapace symbol is freely used.

Figure b shows a similar arrangement, except that the nose is wanting. The
handle is reduced to a mere knob in figure c, but the eyes above and the carapace
symbol at the base suffice to convey the idea intended. In other words, it shows
a ventrally placed carapace symbol resembling those in figures 82 and 83. In
figure d the horizontal loop handle is reduced in size and somewhat removed
from the neck of the vessel. Carapace symbols ornament its summit and bases.
The same is true of figure e, except that three single fillets are used instead of
three pairs, the handles being also reduced in size and changed so as to occupy
a vertical position. In figure / the vertical handles unite shoulder and rim. They
are simply plastered over with carapace symbols. Two perforated nodes near
the top represent the eyes. On the opposite handle, a median hole just below
the eyes may be intended for a mouth or nose.

The remaining vessels of the unpainted handled group are related rather to the

PAINTED WARE. 75

painted handled group than to the armadillo ware. Examples are given in Plate XIX.
Figure a represents one type. The looped handles placed horizontally are so treated
as to indicate life forms. Three incised nodes represent the mouth and eyes, the
latter being quite far apart, and incisions at the bases suggest either feet or the
carapace. The same style of handle with slight variations is repeated in figure b.

What appears to be two pairs of handles in figure c is in reality but a variation
of the type of handle shown in two preceding figures (see PI. XVIII, figs, a and
i). The head is an incised node on the rim. The forelegs are set rather far
apart, so as to look like a pair of handles. The eyes are placed on the shoulders
and the knees are marked by carapace symbols. The same form of handle is
repeated in other specimens, with gradual elimination of life symbols.

The decorative element on the handles shown in figures d and e may represent
either man or monkey. The body, also, of the vessel in figure d is ornamented
by a row of flattened pellets, forming an irregular circle at the level of the bases
of the handles. This circle is interrupted at one handle but drops suddenly and
passes beneath the handle on the opposite side. These flattened pellets probably
represent the body-markings of the alligator. A similar motive is found on
potter}' from the highlands of Costa Rica, ancient pueblo of Curridabat, where
representations of the alligator or parts thereof predominate. The bottom of the
vessel is flattened over a limited area.

The vase reproduced in figure / has a similar flattening of the bottom. The
base of each vertical handle is bifurcated, making a combination handle, conform-
ing to both the horizontal and vertical patterns. One of the manlike or apelike
figures holds his hands at the top of his head, while the hands of the other are
held to his chin.

PAINTED WARE.

The painted pottery found in Chiriqui is divisible into about ten more or less
distinct groups (see p. 66). While some of these classes are large, examples of
others are so rarely met with as to suggest the possibility of their being im-
portations. While the basis of classification generally rests on an ensemble of
characters rather than on some one distinctive feature, the lines of demarcation
are easily traceable by any one at all familiar with Chiriquian ceramic art.

As might be anticipated, the introduction of color as an ornamental factor often
tended to minimize the importance of incised and plastic features as well as of
form and finish in general.

As a result, decorative motives derived from the armadillo and the serpent,
which have given their names to two important groups of unpainted ware, are
practically never executed in paint; while incised or plastic motives derived from
the alligator, whose name is applied to a characteristic group of painted ware,
are equally rare. It is likewise true that no class of ware surpasses the armadillo
group in general excellence of modeling, while the poorest examples in this
respect are to be found among the lost color vases, which are particularly lacking
as regards interior finish.

A comparison of the various classes of painted ware among themselves brings
out some striking differences that apply even to the manner in which a given

76 A STUDY OF CHIRIQUIAN ANTIQUITIES.

decorative motive is to be expressed. The motive derived from the dorsal view
of the alligator, for example, is employed extensively in both the alligator and
the lost color groups and likewise occurs in the white line group; but the artist's
expression of it is always consistent with the class in which he is working, and
it thus becomes not only a distinguishing badge but also one of kinship. The
same thing may be said for the motive derived from the profile view of the alli-
gator, its particular guise depending on whether it is intended for the polychrome
or the alligator ware.

The various groups of painted ware differ among themselves, also, in respect
to the quality of the paste, the prevailing vase forms, the nature of the slip or
the ground color, the method of producing the designs and the number and kind
of colors thus employed. In groups with but a single ground color and no delineating
color, there is usually a supplementary incised or plastic ornamentation. The handled
ware owes its identity to the absence of legs and the form of lip and handles rather
than to the presence of a paint that is also employed on tripods of the fish group.

Various shades of pale yellow and of red were the prevailing ground tints.
While the ground is generally uniform for a given vessel, in the lost color group
it often consists of two colors, the neck and bottom being red and the shoulder
a pale orange. The delineating colors include various shades of red, two kinds
of black, a white and a dull purple, the last two being particularly rare colors.
Red occurs in at least three hues and is always a permanent color, serving as a
ground and also as a delineating color. Black of two kinds was freely employed
both as a ground and as a delineating color, the latter permanent, the former
thin and fugitive. The rare and permanent purple probably comes from a non-
ferruginous metallic oxide. The ancient potters of Chiriqui were masters of the
brush in three distinct methods of painting: (1) The usual method of producing
a figure or pattern was by the direct application of one or more delineating
colors — red, black, white, and in one variety of ware, an additional color, purple.
The red and purple were frequently employed as mass colors outlined in black ;
(2) The lost color process was confined to a single large group. It consisted of
tracing the design in wax over the ground color, the application of a solid coat
of black over the area to be ornamented, and immersing the vessel in hot water
which melted the wax, removing with it the black from the design and thus
leaving the latter in the color of the original ground ; (3) Sparing the figure out
of the ground was seldom employed, occurring only in the alligator and the
kindred polychrome ware. The few examples of this technique evince much skill
and ingenuity.

THE HANDLED GROUP OF PAINTED WARE.

This is called the handled group by Holmes, and yet in the Yale collection
thirty out of the three hundred and ninety vessels belonging to it have no handles.
While all bear paint marks, the surface is never completely covered with paint.
With an ample brush, red paint is applied, usually in horizontal encircling bands,
from which two or three loops or triangles rise to the neck on each side of the
paired handles ; or the paint may be put on in a series of daubs or short streaks.
In rare instances it completely covers the exterior up to the level of the collar

THE HANDLED GROUP OF PAINTED WARE. 77

or neck. Seldom is the outer surface of the flange or lip painted, but its inner surface
is always painted either wholly or in part. The handles are entirely free from paint.

The character of the unpainted areas would seem to indicate that a slip was
rarely and sparingly used. The finish of the interior is even more neglected than
that of the exterior. The walls of the vessels are generally much thicker at the
bottom than elsewhere, a character serving at least one useful purpose, since it
affords a certain degree of stable equilibrium to a form that would otherwise be
unstable. The bases are more or less rounded and never supplied with annular
supports. It is true that many vessels of similar material, form and finish are
mounted as tripods ; but these are placed for convenience with the tripod or fish
group. It will be seen later that those mounted as tripods are much smaller than
their counterparts in the handled group, and the type of handle is also different.

The handled ware proper may be divided into sub-groups depending on the
style of handle. In one of these the paired handles are placed vertically, uniting

Kig. 121. — Vase with carapace symbol Fig. 122. — Vase whose handle decorations re-

ornamenting base of handles. Painted present either fin or carapace motives,

handled ware. '/• Painted handled ware. */•

lip with shoulder, a form that is common also to the armadillo ware. In another
the handles are attached to the shoulder of the vessel alone and in a horizontal
position, a type that is common to the unpainted handled ware. Single handles,
either of the vertical type or that spanning the orifice from lip to lip, are extremely
rare. Animal forms attached to the shoulder of the vessel sometimes take the
place of handles and, as has already been said, handles may be entirely wanting.

The orifice of the vessel is always round, but the character of the lip or flange
surrounding it depends upon the type of handle. Where the handle is attached
to the shoulder alone, or where there are no handles, the lip is circular also,
and every point on its margin is in one and the same horizontal plane. On the
other hand, an oval lip carried upward at both ends goes with the vertical handle.

Hitherto the ornamentation has been the work of the sculptor or engraver.
We now have to reckon with a new force in the domain of decoration, viz.,
color or, rather, paint. Whether or not the summary fashion in which the paint
is here applied, represents the beginning of a new art era or a decadent phase
of the same, it is not within the province of the present study to determine. It
is, however, worth while to note the effect of the painter's work on that of the
sculptor and engraver, where all three are combined in one whole. In the first

78 A STUDY OF CHIRIQUIAN ANTIQUITIES.

place they do not overlap. The work of the sculptor is confined to handles.
These are never painted. The engraver has chosen the shoulder and neck or
collar for his field of operations. The incised areas are never painted.

A study of the handle ornaments of this group reveals the fact that the arma-
dillo motive is one of the principal threads that bind together practically all

Fig. 123.— Vase with same life form on handles as in Fig. 124.— Vase with carapace symbol decorating

preceding example. Painted handled ware. '/« the handles. Painted handled ware. '/»

classes of Chiriquian pottery, serving an important ornamental role full of symbolic
and mythologic meaning. It demonstrates, also, that many purely decorative
motives had their origin in some life form or in elements thereof. In Egypt this

Fig. 125. Fig. 126. Fig. 127.

Fig 125. — Vase with twisted handles composed of two strands. Painted handled ware. ''»

Fig. 126. — Detail of shoulder ornamentation on vase shown in Plate XXI, figure h. Painted handled ware. 'I*

Fig. 127. — Vase with life forms, probably human, serving as handles. Painted handled ware. '/>

form seems to have been the lotus ; in Chiriqui it was the" armadillo and, as we
shall see later, the alligator.

The familiar carapace symbol appears in figure 121. The life forms in figures 122
and 123 are probably identical, although the nodes may be fin motives instead
of carapace motives. The vase reproduced in the latter figure is one of the
largest vessels of the handled group, having a capacity of three liters. An un-

THE HANDLED GROUP OF PAINTED WARE.

79

mistakable carapace symbol appears in figure 124, which represents a large vase
having a uniform coat of red paint covering the entire exterior with the exception
of the handles and lip. The inner surface of the latter leading down to the
orifice, however, is painted.

The two prevailing types of handle are shown in Plate XX. In one case, the
handle is vertical, connecting shoulder and lip (figs, a-d) ; in the other it is placed
horizontally, both ends being attached to the
shoulder (figs. e-h\ In both types, it is almost
always ornamented with nodes and fillets — life
forms or elements of the same. The character-
istic method of applying the red paint in bands,
loops and arches is also well illustrated in this
plate. The horizontal type of handle ac-
companies a circular lip. Where the neck is
of sufficient length it is generally adorned with
incised patterns, as seen in figures f and h.
When the neck is short the shoulder may be
incised instead (fig. </).

Two strips of clay are twisted on each other
to form the handles in figure 125. This type
of handle is common to the tripod group (see
Pis. XXII-XXIV). It was evidently derived
from a life form, probably the armadillo, as
pointed out on page 64.

The vertical handles in Plate XXI are converted into human or apelike forms
by the addition of plastic features. The proboscis noted in certain examples of
the armadillo ware (see figs. 112 and 113) reappears in figure a. Each handle in

Fig. 128. — Vase with single handle spanning
the aperture. Painted handled ware. ' '/'

Fig. 129. — Vase with grotesque forms serving
as handles. Painted handled ware. '/'

Fig. 130. — Vase with plastic decoration suggesting the
carapace motive. Painted handled ware. '/•

figure b represents twins. Two grotesque forms attached to the shoulder take
the place of handles in figure g. Here again the shape of the lip, not being
affected by connection with a handle, is circular and the shoulder is ornamented
with an incised pattern (see PI. XX, fig. g). In figure h, both hands cover the

80 A STUDY OF CHIRIQUIAN ANTIQUITIES.

mouth. The figure on the opposite side (fig. 126) holds the right hand to the
chin and supports it there by grasping the wrist with the left hand.

The life forms in figures 127 and 128 are probably human also. The handle
in the latter vase is a single loop spanning the orifice. In figure 129, the two
grotesque forms attached to the shoulder take the place of handles. They are
unlike. One is the armadillo with tip of nose hidden behind the forefeet, and
with two incised fillets representing the carapace, one applied longitudinally be-
ginning at the top of the head, and the other, horizontally about the back of the
neck. The other figure is a cross between the human and the ape. The plastic
decorative motive in figure 130 is similar to one already noted in the armadillo
group of ware (see fig. 97 and PI. XVI, fig. a) — a somewhat reduced and simpli-
fied head surrounded by an incised fillet, suggesting the carapace motive. The
surface below the collar of this vase is smoked and sooty, a condition common
to this group, at least seventy-five per cent of which have been used over the fire.

As has been already stated, not all vases of the handled ware have handles,
figure 131 being an example. It has a well-developed collar. The body is given

Fig. 131.— Vase with well-developed collar but Fig. 132.— Vase without handles. Painted

without handles. Painted handled ware. '/• handled ware. '/'

a solid coat of red paint, while neck and collar were left untouched. Another
vase without handles is reproduced in figure 132. The arched panels on the
sides are not unlike those seen in the lost color ware.

THE TRIPOD OR FISH GROUP.

Mention has already been made of the affinity between this group and the handled
ware, which consists chiefly in a similarity of paste, scant use of slip, the
kind of paint used and the method of applying it. and the fact that a large
majority of vessels in both groups bear sooty incrustations.

There are also some fairly well-marked differences to be noted. The bowls of
the tripods average smaller than those of the handled group. With few excep-
tions, the type of rim or lip is quite different, the typical tripod rim being at all
points in the same horizontal plane and recurved outward upon itself. The descent
from the outer edge of the rim to the orifice is a gentle convex curve. The

THE TRIPOD OR FISH GROUP.

81

Fig. «33-

orifice is proportionately greater than in the handled ware, with a diameter but
slightly smaller than the greatest inside diameter of the bowl, the latter usually
being quite shallow.

While in the handled group the in-
terior of the bowl is never painted, in
the tripod ware this portion is more often
painted than any other part, which is
explained by the fact that the relatively
large mouth opening and shallow bowl
render the interior visible at all points.
Many of the tripods are not painted at
all. These may be said to have their
counterpart in the unpainted variety of
the handled group. On consulting the
illustrations, it will be seen that the most
satisfactory and elaborately wrought tri-
pods are among those that are not painted.

Twisted vertically-placed handles are
the rule among the tripods. Only two
examples of the twisted handle were to
be found in the handled group. On the
contrary, there are very few tripod handles
wrought into realistic animal forms, a
style of handle very common indeed in
the handled group. In the latter ware,
neck ornaments are rare. Among the
tripods, especially those unpainted, the
neck is almost always decorated with
fillets or incised patterns, the usual orna-
mentation being two long incised fillets,
each reaching half-way around the neck.
The ends of these fillets do not touch,
leaving a blank space underneath each
handle. The handles were put on first,
then the fillets, and lastly the feet.

The tripod supports are very different
from those of the armadillo ware. They
are much longer and, on account of their
length, are spreading, so as to prevent
the vessel from being easily overturned. Fig. 133.— Tripod with long solid supports and with

The sunnorts are rarelv solid Ficrnrp 133 fi"ets applied obliquely to the handles. Fi.sh ware. '/«

Id. IMgUI id, Fig ,34._Tripod with twisted handles and with sup.

therefore, represents an exception tO the Ports resembling the body of a fish. Fish ware. '/»

rule, not only in regard to the supports,

but also in the shape of the rim. The fillets applied obliquely to the handles
are life motives, and form a link in the series that includes handles with horizon-
tally applied fillets on the one hand, and those that are obliquely incised or
MEMOIRS CONN. ACAD., Vol. III. 11

Fig. 134-

82

A STUDY OF CHIRIQUIAN ANTIQUITIES.

twisted on the other (see PL XIII, figs, a, b and e; PI. XXII, figs, e and /). As
far as Chiriquian art is concerned, all may have been derived from the armadillo
carapace.

The hollow tripod supports are generally provided with long median slits,
through which may be seen the movable pellets serving as rattles, one to each

support being the rule. The
shape of the plain leg is often
such as to resemble the body
of a fish. A pointed hip
(PI. XXII, fig. d) or a lateral
flattening at the free end
(PI. XXII, fig. e) makes the
resemblance even more strik-
ing. To aid the imagination
further, incisions are made on
the pointed hip to represent
the mouth of the fish (fig. 134).
Two twisted strands form each
handle. A fillet incised hori-
zontally and broken at the
handles ornaments the neck.
With the addition of eyes
and fins, the image is complete,
as is the case in figure 135,
where pectoral, ventral, cau-
dal and dorsal fins are all
present. To save one dorsal fin, the median
slit is bridged, while the other dorsal fin is
pushed forward to a point almost between
the eyes. The mouth is slightly open, show-
ing teeth. Each handle is skilfully fashioned
into a monster with human body and head
of a bear or jaguar. This may be the jaguar-
god, which is described in the chapter deal-
ing with the alligator ware and of which
there are fine examples among the gold
ornaments. The hands support the lower
jaw. The bowl of this tripod is typical
for the group, with its shallow bottom, large
orifice, and recurved rim, every portion of
which is in the same horizontal plane -
the urn-shaped bowl par excellence.

The urn-shaped bowl is repeated in figure 136. The handles, however, are
plain. The long median slit leaves room for but a single dorsal fin, which is
placed forward between the eyes. Teeth are represented by short incisions at
right angles to the slightly open mouth. This vessel is not painted, neither are

Fig. 135. — Tripod, the legs of which are realistic representations of
the fish. The half-human monsters serving as handles may re-
present the jaguar-god. Fish ware '/J

Fig. 136. —Tripod with plain handles and fishlike
supports. Fish ware. '/>

THE TRII'OD OR FISH GROUP.

83

the two preceding it, but all three are highly polished, particularly within, and
are more or less smoke-blackened on the outside.

By comparing the last
three illustrations with the
two that follow, one sees at
a glance the superior form
and finish of the unpainted
ware. The tripod shown in
figure 137 is painted red,
both inside and outside, with
the exception of the neck
and twisted handles. The
single median slit is extend-
ed to the tip of the nose,
leaving no room for even
a single dorsal fin. The
rather prominent cat-fish
mouth is crossed from lip
to lip by a number of fillets.
The pectoral and ventral
fins are present, as is the
unpaired anal fin, which is
placed exactly opposite the
median dorsal slit on each
leg, and almost as far re-
moved from the tail as are
the ventral lateral fins. This
is the only occurrence of

the anal fin in the entire collection. Professor
F. W. Putnam 1 did not find it in the series he
studied, and it is not mentioned by Holmes.

The dorsal fin reappears in figure 138 simply
as an incised lump on the nose of the fish. The
pectoral and caudal fins are represented, but the
mouth is wanting. An incised band surrounds
the neck, being carried under the handles with-
out a break — an exception to the rule. The
handles are incised diagonally, giving them the
appearance of being twisted (see fig. 133). The
interior and exterior are both smeared with red
paint, which, however, does not completely hide
the slip at all points.

The fish's mouth in figure 139 is elaborately finished. In the muzzle, on
either side of a median plane, there is a pair of openings that widen toward the

Fig. 137. — Tripod with twisted handles, and supports in which the anal
fins are present. Fish ware. V»

Fig. 138. — Tripod in which the dorsal fin
appears as a lump on the nose of the
fish. Fish ware. '/»

1 Conventionalism in ancient American art. Bull. Essex inst., XVIII, 162, 1886.

84

A STUDY OF CHIRIQUIAN ANTIQUITIES.

corners of the mouth. Zigzag fillets span these openings and a pair of fillets
surrounds the mouth, the whole of which is easily visible because of the pro-
truding lower jaw. The pectorals and one dorsal fin are represented. The eyes
are left out. The neck and handles are plain, but there are traces of paint
splotches within and without that were spread over the slip in streaks by rubbing.

The characteristic tripod rim has completely disappeared in figure 140, which
is a typical example of the handled ware plus the legs. The paint is also
applied exactly as in that group. The inner surface of the rim leading to the
orifice is painted, but the interior of the bowl is not, neither are the handles.
A pair of pectoral fins remain to give a fishlike aspect to the legs. The tail is
slightly flattened laterally, but not incised.

The vase of which one foot is shown in figure 141 is also of the handled type,
differing only from the latter in its comparatively large mouth opening and

Fig. 139. — Tripod with gracefully modeled supports
representing the fish. Fish ware. '/>

Fig. 140. — Typical example of handled ware with
legs, the latter exhibiting pectoral fins only.
Fish ware. V"

painted interior. The wide-spreading legs are clumsy and ponderous. Nothing
of the fish remains except a single dorsal fin at the upper end of the long
median slit.

The painters of the tripod group possessed an interesting technique. The red
paint was applied to the slip in spots or bands and rubbed down while in the
process of drying, thus producing the effect of floating clouds or the flecked
surface of the eggs of certain birds. This is seen to best advantage over the
surface of the interior, where the red was sometimes simply spattered on and
then rubbed in. This spattering treatment minus the rubbing in was practised
by the ancient Tusayan potters, fine examples of it being found by Fewkes 1 at
the ruined pueblo of Sikyatki.

1 Seventeenth Ann. rept, Bur. Amer. ethnol,, Pt. 2, 650, 1895—1896.

THE TRIPOD OR FISH GROUP.

85

Specimens of this technique are reproduced in Plate XXII. A characteristic
interior for the tripod group in shape and painting is shown in figure a. The
feet are ornamented with the eyes, mouth and tail fin of the fish. The same
technique is seen in figure b, the best effects being at the bottom of the interior.
The supports are relatively long and spreading, with only pectoral fins represented.
In figure c, the effect of cross-rubbing the red bands on the legs is visible. At
the hip are the head, beak and eyes of a bird, probably the hawk. Just beneath
the head are two incised nodes representing the feet. A good specimen of the
flecked exterior is seen in figure d. Here the fish form of the foot is faithfully
reproduced without the help of incised lines or relief ornament. The red bands
in figure e were not rubbed while in process of drying, and the vessel is figured
here by way of contrast; the interior, however, is characteristically flecked. The
unpainted feet represent the fish again, the only feature emphasized being the
tail fin. A good example of the diagonally incised handle that may be looked
upon as a connecting link between the horizontally
incised and the twisted handle is reproduced in fig-
ure /.

A series of tripods with fish supports is given in
Plate XXIII. All are blackened by soot and smoke,
and only one (fig. /) bears any trace of ever having
been painted. Mouth and fins are indicated in all.
Each fish in figure a has both a dorsal and a ventral
median slit. By this means, light is let in, making the
movable pellet more conspicuous. The weight is also
reduced. Although large, they are of graceful outline
and, by the addition of incised nodes and fillets, rep-
resent the fish in truly realistic fashion. The teeth
are reproduced by zigzag incisions. The neck of this
graceful urn-shaped vase is adorned with a series of
six scrolls, below which runs an incised fillet, broken
at the handles. Each of the handles is formed by
three twisted strands.

The dorsal fins are generally sacrificed to the dorsal slit, but this is sometimes
avoided by bridging the slit, as in the foregoing case. There is no such bridge
in figure 6, so that the second dorsal fin is placed forward between the eyes,
and the first dorsal rests on the tip of the nose. The mouth is represented in
a summary way by means of two parallel incisions. Pectoral, ventral and caudal
fins are present. The ornaments on the neck consist of foot symbols and an
incised fillet.

Especially noteworthy are the highly curved, fishlike legs of figure c. Both
dorsal fins are sacrificed here, but pectoral and ventral fins are present. The
caudal fin, though distinctly forked, is not incised. The slightly open mouth is
formed by fillets, and into the opening are stuck pellets for teeth. The incised
band about the neck of the vase resembles the carapace symbol. Each handle
is made of three twisted strands. The interior of each hollow leg is covered by
a beautiful grooved pattern.

Fig. 141. — Tripod support, the only
ornamentation being a (single
dorsal fin. Fish ware. V1

86 A STUDY OF CHIRIQUIAN ANTIQUITIES.

An unusual type of tripod is reproduced in figure d. The paste and exterior
are those of the armadillo ware, but the shape is unique. The long neck is
cone-shaped and the plain handles are angular. It is placed here because of the
fish supports. Only the pectorals and the caudal fin remain. The eyes are omitted,
the mouth being the feature emphasized. The longitudinally incised fillets about
the mouth opening are in parallel planes, so far apart that the mouth has four
corners instead of two. The teeth are indicated by incised parallel lines. It will
be noted that one foot is immediately below the handle, an exception to the
rule, which is that one leg divides the space between the handles about equally
on one side, while the other two legs are attached near the handles on the opposite
side. Rarely is there a deviation from this symmetrical arrangement of handles
and supports, as will become evident from a study of the illustrations. There is
another vase in the collection that is even more like the tripods of the armadillo
ware as regards paste and form, but it has handles and the legs are long. The
eyes reappear in figure e, so does the dorsal fin, which is slightly forward of the
eyes. The mouth is an incised fillet.

There seem to be certain general rules governing the use of life forms as tripod
supports. One is that all three should represent one and the same animal.
Figure / shows one of the very few exceptions that prove this rule. Here one
leg represents the fish, while the other two are highly realistic alligator forms.
The alligator nostrils, eyes and ears are conspicuous. The osseous dorsal plates
are represented by annular indentations and there are five cone-shaped prominences
along the vertebral column. The saurian attitude of the limbs is well rendered.
The second illustration brings the mouth and teeth of the alligator, as well as of
the fish, into view. The nose of both alligators is pressed against the rim of the
vase, while that of the fish falls considerably short of the same. Spots of red
paint on the legs and on the interior of the vase have been spread in a way
characteristic of this group, viz., by rubbing. The outer surface of the bowl is
so blackened by smoke and soot as to make it impossible to determine whether
it was flecked or given a uniform coat of paint. The two alligator supports are
slit ventrally and supplied with movable balls of clay. The fish support is slit
both ventrally and dorsally and also supplied with a ball of burnt clay.

From his study of Chiriquian pottery, Holmes concludes that " there is a general
consistency in the use of life forms. " He says : " The fish and other creatures
used, although variously conceived and treated, are never confused. When the
fish is employed, no features suggesting other animals appear and when the heads
of other creatures occupy the upper extremity of the leg all the details refer to
these creatures with uniform consistency."

There are some noteworthy exceptions to this rule in the Yale collection, as
the accompanying illustrations will show. The legs seen in figure 142 embody
a complete fusion of the fish and the human form, the same pair of eyes answering
equally well for each. The prominent fish mouth serves as a head-dress for the
human figure. The hands rest on the lower part of the chest. The incised pattern
below stands for a garment or wrapping. They may be referred equally well to
the fish, as the fishlike legs of certain other tripods are incised in a somewhat
similar manner. The likeness of this tripod bowl to vessels of the handled group

THE TRIPOD OR FISH GROUP.

87

Fig. 142.

is worthy of note. The legs, handles and neck are not painted, neither is the
interior below the orifice.

There is also a fusion or confusion of forms in figure 143. The mouth and the
pectoral and caudal fins are faithfully rendered. On the nose, however, and pro-
jecting some distance farther forward than its tip, is built up the head of an owl
with prominent eyes made of coiled fillets.
In a median line, just below the eyes,
is a projecting cone, which from its position
and shape might answer to the require-
ments of a dorsal fin as well as of the
owl's beak. It is probably intended for
the beak only, because of the single in-
cision on either side. The owl's feet
appear a little lower down. Each handle
of the vase is marked by a median in-
cision, to meet which, parallel incised lines
are carried down from either side. On
each side, also, and reaching from handle
to handle, a broad fillet is attached to the
neck and cut by a median incised line
running horizontally. But before the
horizontal incision was made, the fillet
was incised with numerous parallel slant-
ing lines. A series of eye symbols, six
in all, is applied just above the fillet, three
on either side. None of the relief embel-
lishments are painted, the color being
applied only to smooth surfaces. The
interior, as usual, is flecked with paint
and more highly polished than the ex-
terior.

A vase (fig. 144) that may not be of
native Chiriquian workmanship is intro-
duced here because the legs combine fish
characters with those of some carnivorous
animal. For the fish, there are two pairs
of lateral fins and one dorsal fin, but no
eyes nor mouth. The other life form is
represented by the head and forelegs. This

,. . ,, Fig 142. — Tripod in which the supports embody a

5 resembles a class of pottery m the sfusion oftheFfish and the human {££ Fishware! •/•

Keith Collection recently found at PaSO F'g- H3- — Tripod in which the supports combine the
TJ^«I /-« TT L r r ,1 /~«i • • fish with the head and feet of the owl Fish ware. '/»

Keal, Costa Rica, not far from the Chiri-
quian frontier.

The supports are of uniform diameter throughout the greater part of their length.
They are not curved, neither are they spreading. The three rows of slits in each
look as if they might have been made after the paste had hardened. There are

Fig- 143-

88

A STUDY OF CHIRIQUIAN ANTIQUITIES.

no balls within. One does find, however, in each leg a quantity of angular
fragments of burnt clay, apparently punched from the slits as the latter were
being made. They answer all the purposes of the ordinary movable ball. A
similar phenomenon has already been noted as occurring in a series of tripods
belonging to the armadillo group (PI. VI, figs. /, g and K). The neck is embellished
with an incised band on each side extending from handle to handle. At the ends
and in the middle of each band, there is a fin symbol.

Two short bands of clay are applied horizontally to each diminutive handle.
The bowl is pointed at the bottom and relatively high, with a distinct shoulder.

Fig. 144.

Fig- 145-

Fig. 144. — Tripod in which the legs combine characters
of the fish with those of some carnivorous animal.
Fish ware. V1

Fig. 145. — Ornate tripod with twisted handles re-
presenting animal forms. Fish ware. '/«

A solid coat of red paint covers the inner surface of the neck leading down to
the orifice, and the outer surface of the bowl from the shoulder down. The legs
and outer surface of the neck and shoulder are pale chocolate — the color of the
slip. The paste is reddish brown.

Two other exceptions to the general rule of consistency may be found in
Plate XXIV. That the tripod legs in figure a are intended to represent fish is
indicated by a pointed nose and pectoral fins, but grotesque apelike forms are
seated on the nose of each fish. This is the largest tripod in the collection and
the only one of this group not provided with handles. The interior is painted
with more care than the exterior. In figure 6, the legs are all alike except that
the embellishment at the hip is different in each case — an alligator and two
species of bird, one being the owl.

THE TRIPOD OR FISH GROUP.

89

The hip ornament however is, as a rule, uniform for each tripod. In figure e,
it is a long-tailed beast in the attitude of repose, while in figure d it is a bird
(the hawk or owl) with coils for eyes. Other coiled fillets adorn the neck of the
vase, the scroll pattern being the same as that in figure a (PI. XXIII), only
inverted. A very effective representation of the owl, with prominent eyes made
of coiled fillets and set in a large head, is shown in figure e. The wings and
feet are also well rendered. The frog, which was so prominent in the armadillo
ware, reappears in figure /. Only the eyes, mouth and forelegs are visible.

A vase uniting a number of interesting characters is reproduced in figure 145.
The hip ornament is a long-tailed animal with prominent muzzle and teeth, pre-
sumably the jaguar. A series of four animal heads adorns the recurved lip of the
vase, reminding one of the head ornaments about the margin of the seats in the

Fig. 146. — Tripod in which the supports are modi-
fied to represent the human form. Fish ware. '/»

Fig. 147. — Tripod with supports representing the human
form. Fish ware. '/>

stone and terra cotta stools, so called (see Pis. IV and XLVI). The decoration
on the neck consists of two incised bands on either side (from handle to handle),
with a row of eye symbols between. The twisted handle is also supplied with
eyes, apparently to emphasize the fact that a twisted handle is a life form (see
PI. XIII, figs, a, 6, e; and figs. 125, 133). The Chiriquian artist was seldom guilty
of making too free use of decorative features. He rather overstepped the limits
of good taste, however, in this one instance.

The human form does not seem to have been a particularly favorite subject
with the ancient potters of Chiriqui. Judging from the few examples at hand,
the results were often humorous but seldom realistic or artistic, and in their work
it is frequently impossible to distinguish man from the monkey. There is no
doubt, however, as to the meaning of the tripod supports in figure 146. All three
are alike — short legs, long body, arms flexed at the elbows bringing the hands
to the chest, large face and nose and small cranial capacity. The vessel is flecked
MEMOIRS CONN. ACAD.', Vol. III. 12

90

A STUDY OF CHIRIQUIAN ANTIQUITIES.

with paint both inside and outside. The hollow bodies may once have contained
movable balls.

The tripod supports shown in figure 147 are solid. The human legs are indi-
cated by a slit, which, however, does not separate the feet. An indented band
surrounds the loins. The arms are rudely fashioned and variously disposed in
the three figures. The paste of this tripod is black, but bleached on the surface,
which is faintly streaked with red.

Figures 148 and 149 represent supports from two different tripods which are
apparently the work of the same potter. The fusion of the feet, the short legs,
long body and occipital flattening of the head are all characteristic. In each.

Fig. 148.

Fig. 149-

Fig. 150.

Fig. 148. — Tripod support representing the human form. Fish ware. '/»

Fig. 149. — Similar example from another tripod. Fish ware. '/'

Fig. 150. — Tripod support exhibiting similar technique, but head of the figure is scarcely human. Fish ware. '/>

there is a praiseworthy attempt at rendering the human ear. The nose is prom-
inent; the arms, with one exception, are molded with freedom, being clear of
the body for the greater part of their length ; the number of fingers and toes
ranges from seven to four, the artist making no effort to be exact in this respect,
but striving rather to produce the effect of the foot or hand as a whole. In one
case, the knees are marked by short bands of clay. A somewhat similar tech-
nique is exhibited in another tripod support (fig. 150), except that the arms and
legs are longer and the head can scarcely be called human. One hand rests
against the chin. The upper part of one arm has been lost.

One small tripod and a small vessel with annular base are reproduced here,
although quite different in shape from the characteristic tripods of this group.
The paste and painting, however, are common to this ware. In figure 151, the
legs are short and solid. The bowl is shallow, with paint marks inside and out-
side, but is not provided with neck or handles. Fish and crab motives extend

THE TRIPOD OR FISH GROUP.

91

half-way round the bowl just below the rim. Another vessel in the series is
fundamentally the same, but with variations in the details of execution. The legs
are curved and set closer together, the bowl is angular in outline and not painted
inside. The life form on one side of it is probably intended to represent the
monkey. The interior of the urn-shaped vessel shown in figure 152 is given a
solid coat of red paint ; of the exterior, only the neck and the concave bottom

Fig. 151. — Tripod, the shallow bowl of which is decorated
with crab and fin motives. Fish ware. V>

Fig. 152. — Vase with annular base.
Fish ware. '/•

are painted. The zone between shoulder and base bears relief ornaments and is
finished in a salmon-colored slip.

The fish group includes a number of vessels with shallow basins that would
seem to be strangers here but for the identity of paste and paint. In these

'53- — Tripod with elongated shallow bowl, and
handles suggesting life forms. Fish ware. '/>

Fig. 154. — Another tripod similar in type, unpainted ;
from Bugavita. Fish ware. V>

respects they help to link the fish group with the handled ware. One has been
led to associate tripods with vertical handles ; but in figures 153 and 154, the
handles being modifications of the rim, belong to the horizontal type. Some are
twisted, others are accompanied by foot symbols, and still others are perfectly
plain. Figure 154, an unusual type, shows the only one that does not bear marks
of red paint. The legs in all are rudely finished.

There are two tripods in the collection that cannot be referred to any of the
groups into which the Chiriquian pottery has been divided. Although found in

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Chiriqui, they may be importations, probably from Costa Rica, as the tripod

supports resemble certain pieces found by Hartman at Orosi. The example shown

in figure 155 is from Caldera : another,

not figured, is from El Banco. They

are alike in every respect. The paste

is friable and rather coarse. The neck

ornaments are identical, and the same

animal form is embodied in the legs of

each. The slip in the one figured is

yellow, and large triangular areas painted

red alternate with the three tripod legs.

The surface of the other is so completely

altered by smoke as to obliterate all

traces of slip or paint.

THE RED LINE GROUP.

The red line group is represented by
thirty-one vessels with a wide range of
form but with certain uniform characters.
Faulty firing is indicated by discolored
spots on the surface and by cracking of
the paste. The paste and slip are of
the same color— a light-gray orange.
The delineating color is a bright sienna
red, resembling the red of the handled
and tripod groups.

An idea of the variety of form within this comparatively small group may be
gained by consulting Plate XXV. The handles when present are always vertical
loops, either paired or single. The supports of tripods are always looped bands
of clay. Annular supports, either solid or perforated, are quite frequently em-
ployed. At least a dozen pieces bear marks of use.

Figure a represents a graceful type of vase, with globular body, from which
large, gently curving loop handles rise to fuse with the prolonged lip on either
side. Each handle is decorated with an indented fillet applied in the region of
the sharpest curve. Above and below each fillet is a band of red that is also
carried along the margins of the lip and handles. There is likewise a painted
panel on either side of the body, reaching from the neck to the level of the
attachment of the handles with which the panels alternate. The paint is care-
lessly applied, being allowed to run at some of the corners. The surface is every-
where marked by evidences of handling and use previous to burial. The round
spot of paint found on the bottom of a majority of the vases of the red line
group is lacking in this one.

By removing one handle and the prolongation of the lip on the same side,
one arrives at the form seen in figure b. the outlines of which are particularly
pleasing. There is the same graceful curve to the handle, which is also supplied

Fig. 155. — Tripod embodying animal forms in the
supports ; from Caldera. Probably of Costa Rican
origin. '/!

THE RED LINE GROUP.

93

with an indented horizontal fillet representing some life form. The paint is
applied to the handle and the rim as in the preceding, with a painted panel on
either side. Between these panels and opposite the single handle are two plain
nodes of clay. A large black spot on the outer surface is due to faulty firing.

A different effect is produced by the specimen shown in figure c, where the
neck is smaller and the lip is prolonged almost horizontally to meet the ascending
ramus of the single handle. On the neck, opposite the handle and protected by
the projecting lip, are nodes and fillets applied so as to reproduce the human
features — ears, eyes and nose. At the base of the neck there is a slight beading.
A pattern painted in red lines covers the upper half of the body on either side,
and the bottom is painted red.

A representation of the vases without handles or supports is given in figure d.
The margin of the rim is marked by radiating, straight-line incisions, inside of
which is a band of red paint. Another band of color encircles the body in the
region of its greatest diameter, and the customary large round spot of red adorns
the bottom. The walls are thick and not carefully worked over the inner sur-
face.

The characteristic looped tripod foot is shown in figure e, a specimen from
Caldera. Indeterminate life forms are applied on opposite sides of the body, the
upper half of which is also adorned with a ladder-like zigzag pattern in red
encircling the vessel. The walls are thin and badly cracked.

The tall ewer-shaped vessel re-
produced in figure / is unique. It
may have served as a drum. A drum-
head could have been stretched over
the aperture and made fast to the
prominences that surround the neck.
The bell-shaped base is not perforated.
A slip covers the interior of the mouth
opening to a depth of three centi-
meters and on it two red bands are
traced horizontally ; a third marks the
rim, and two more surround the neck
just above the relief ornament. Four
horizontal bands encircle the body, one
of them being a series of connected

triangular areas. The base also is marked by longitudinal and transverse
bands.

The large tripod reproduced in figure 156 comes from El Banco. A red band
encircles the rim, while groups of parallel red lines alternating with small trian-
gular painted areas adorn the shoulder. The walls are thick and firm. A bowl
from Gualaca with unusual characters is illustrated in figure 157. The walls are
ten millimeters thick. The annular base is low and not perforated ; its rim and
outer surface are painted red. Three broad bands of red are carried upward from
the annular base to the rim at the aperture, which is also adorned with a band
of the same color. The three alternating fields are partially filled in with red

Fig. 156. — Tripod with short looplike supports; from El
Banco. Red line ware. ll>

94

A STUDY OF CHIRIQUIAN ANTIQUITIES.

lines, to some of which comblike patterns are attached. Contrary to custom, the
interior of this shallow open-mouthed bowl is undecorated, except by a spattering
of paint, which may or may not be intentional.

When the vessels are shallow, the painted decoration is confined to the upper
rim and the interior, where it would be the most easily visible. Figure 158 is

Fig- 157-

Fig. 158.

Fig- 157- — Bowl with low annular base; from

Gualaca. Red line ware. V«
Fig. 158. — Bowl with perforated annular base.

Red line ware. '/»

Fig. 159.

Fig. 160.

Fig. 159. — Shallow bowl with perforated
annular bell-shaped base ; from El
Banco. Red line ware. */'

Fig. 1 60. — Globular vase. Red line ware.1/'

an example. The annular base or foot is pierced by four unsymmetrical holes.
On the interior, a gridiron-like pattern reaches from margin to margin, passing
by the center and provided with two lateral projections near each end. The
annular foot in figure 159 is taller, bell-shaped and provided with two large sym-

THE CHOCOLATE INCISED GROUP.

95

metrical perforations. A pattern composed of groups of parallel lines meeting, at
an angle and forming a cross completely covers the interior. This specimen,
which comes from El Banco, is slightly discolored by use. Some of the more
simple forms consist of globular bodies with neck and rim ; the handles, feet and
other plastic accessories being absent. Figure 160 is an example.

THE CHOCOLATE INCISED GROUP.

This is a small group numbering only eight specimens, all tripods. It is not
mentioned by Holmes. Reference to the illustrations gives one a good idea as
to the homogeneity of the group. The walls are comparatively thin and carefully
formed, both as to exterior and
interior. The paste and firing are
of such a character as to make
the vessels practically impervious
to water. The chocolate colored
coating applied uniformly over the
entire outer surface and the visible
portion of the aperture is probably
in the nature of a slip. The paste
is somewhat lighter in color than
the slip.

Another characteristic feature is
the incised ornament completely
encircling the shoulder on each
tripod. The incisions were made
after the application of the color-
ing matter and probably after the
firing, the instrument used evidently
being either of stone or metal.
The incisions are narrow and

shallow in some cases and broad and deep in others. The patterns usually consist
of three quadrangular fields filled in wholly or partially by cross lines, with groups
of parallel vertical lines at both ends of each field (fig. 161). The panels are some-
what lengthened in figure 162 and not accompanied by the groups of parallel
vertical lines, while in figure 163 a single unbroken pattern completes the circuit.
The latter is one of the smallest tripods in the entire collection, its height being
less than four centimeters.

The tripod supports are generally solid, short, straight, pointed pegs. In
figure 161, they are slightly curved, hollow and provided with pellets, but not
slit. Figure 162 differs from the others in several respects. The body represented
is a shallow open-mouthed bowl. The relatively large legs are incised, hollow,
perforated, supplied with pellets and so fashioned as to resemble an animal
head.

The chocolate incised group resembles a certain class of Costa Rican pottery
from Tres Rios. Figures 161 and 162 illustrate ware from a locality twenty-five

Fig. 161. — Tripod vase with incised shoulder and short hollow
supports supplied with pellets but not slit ; from Divala.
Chocolate incised ware. Vi

96

A STUDY OF CHIRIQUIAN ANTIQUITIES.

miles west-northwest of David. A single specimen from the environs of David,
somewhat similar to the former, is figured by Alphonse Pinart.1

Fig. 162. — Tripod with incised panels on shoulder
and short supports representing animal heads ; from
Divala. Chocolate incised ware. V"

Fig. 163.- — Very small tripod with unbroken
incised pattern around shoulder, and short
solid supports. Chocolate incised ware. '/'

THE SCARIFIED GROUP.

The scarified group numbers twenty specimens in all. Sixteen are tripods,
three have convex bottoms, and one is flat-bottomed. A slightly raised marginal
flange in the last-named has preserved the flat bottom from wear. All are covered
with a uniform coat of maroon paint, which in most cases is spread over the
surface of the interior also. The paint was applied after the scarifications had
been made and, 'as a rule, the scratched areas were not painted. The region
bounded by the feet of the tripods is always left smooth.

A majority of the pieces come from a
locality twenty-five miles northwest of David,
at the head-waters of a stream called in
McNiel's notes, Rio Chiriagua. Seven or
eight are from Caldera about twenty miles
northeast of David, and the large flat-bottomed
vessel (see PI. XXVI, fig. a) came from a
locality noted by McNiel as being at lati-
tude 8° 34' N. and longitude 82° 26' W. of
Greenwich, i. e., about eight or ten miles
north of David. The pottery of this same
group described by Holmes came from the
graves of the two other localities, " one near
C. E. Taylor's hacienda, north of David, on
the slopes of Mount Chiriqui, and the other
at Alanje, southwest of David." Specimens

of this ware are therefore fairly well distributed over the province of Chiriqui,
although found in relatively small numbers.

Fig. 164.— Non-scarified vase with handles crudely
representing animal forms ; from Caldera
Scarified ware. '/>

1 Les indiens de 1'Etat de Panama. Rev. d'ethnogr., 125, fig. 20, 1887.

THE SCARIFIED GROUP.

97

Holmes speaks of the group as being a variety of ware standing " so entirely
alone that had it arrived unlabeled no one would have recognized its affinities
with Chiriquian art/' The group does possess a number of distinct characters.
On the other hand, the affinity of certain pieces with specimens of the maroon
group was so marked that great difficulty was experienced in classifying the
specimens in question. The piece shown in figure 164, for example, resembles
very closely one of the maroon vases (see fig. 170). Both are from the same
locality, Caldera, and are alike in shape and size. Even the handles are of the

Fig. 165. — Tripod bowl with handles and supports resembling head, wings and
tail of a bird (or man) ; from Caldera. Scarified ware. V"

same type, except that in one case they stop short of making a complete loop,
and end in a crude pattern suggesting the ornament found on the loop handles
of the other at relatively the same zone. The ornament in both instances is
intended to convey the same idea, that is still better expressed in figure 165.
Here the handles, three in number, are probably intended to represent a bird
perched on the shoulder of the vessel. Each is directly over one of the three
feet. These suggest more than anything else
the spreading tail of a bird and are provided
with an incised band at the tip. It would seem
as if a section of the bird had disappeared in
the body of the vessel that spans the distance
between shoulder ornament and foot. This
specimen is from Caldera also.

It will be seen from figure 164 that scarifica-
tions are not always present. Another example
without scarification is given in figure 166, which

is heart-shaped in horizontal section, the form Fig. 166.— Non-scarified tripod, heart-shaped
•being produced by a vertical indentation on one in horizontal section. Scarified ware. '/«
side.

Another example of shoulder ornament and tripod leg being made to represent

the upper and lower extremities, respectively, of one and the same creature is

seen in figure 167, where the human form is treated rather fantastically. Mouth,

eyes and long hair are indicated. The hands rest on the chin. The section

MEMOIRS CONN. ACAD., Vol. III. 13

98

A STUDY OF CHIRIQUIAN ANTIQUITIES.

from arm-pits to knees is lost in the body of the vessel. Only one shoulder orna-
ment remains, the other two having disappeared, one of them evidently while the
vessel was still in use, as the place where it stood had been rubbed down and
painted before the tripod was buried — an example of prehistoric pottery mending
that should not escape notice. This specimen is from El Banco.

One marked character of the
group is the tendency in the
tripods toward short legs placed
close together. The latter were
so arranged in order that they
might be made as short as pos-
sible, and they were made short
because long legs of such coarse,
friable, poorly baked paste would
not long endure. They are gen-
erally either blunt pegs or are
slightly spread at the extremity,
so as to suggest a' three-toed
foot, probably that of the tapir.
The feet of the flat-bottomed
tripods have completely dis-
appeared. The open-mouthed
bowl represented in figure 168
is a typical example of the
bunching of three short legs.

Two of the tripods are ob-
long basins, with the rim carried
up to a point at the ends, near
each of which are two rim nodes.
Figure 169 is an example. There
are four scarified areas on each
vessel — two on a side, an upper
and a lower, separated by a
narrow, horizontal, smooth band.
The areas of one side are sep-
arated from those of the other
by a wide, vertical, smooth band
at both ends and along the
bottom, where the band spreads

fig. 1 68. — Open-mouthed bowl with three short supports placed

close together. Scarified ware. '/« into a field in which the three

stump legs are set. A thin coat

of maroon paint covers the entire surface, both inside and outside, including the
incised patterns. In another boat-shaped tripod vase, the paint was applied to
the smooth surfaces only. That it was applied after the incised patterns were
made, is evidenced at numerous points by the careless way in which the paint
was allowed to fill the incisions that border on smooth fields. This is best seen

Fig. 167. — Tripod in which shoulder ornaments and supports rep-
resent the head and arms, and the legs of the human figure,
respectively; from El Banco. Scarified ware. '/»

THE SCARIFIED GROUP. 99

in the upper areas, where small groups of parallel incisions alternate with narrow
painted bands. The scarifications in the lower fields are the exact counterpart
of a geologist's drawing to represent cross-bedding in section.

Some of the finest specimens of the scarified ware are reproduced in Plate XXVI.
The most remarkable member of the group is the large jar with flat bottom
and flaring rim (fig. a). Holmes figures a jar of somewhat similar make, but
smaller and less attractive in form and finish. The scarified zone reaches from
near the base to the neck, and therefore covers the surface of a truncated cone.
A careful study of the incised lines reveals the relative order in which they were
made. The circumferences at neck and base were first described. The twelve
groups of vertical lines, three in a group, followed and were drawn from neck
toward base. The twelve alternating, rather broad bands were then scarified

Kig. 169. — Boat-shaped tripod with rim extended to a point at either end, each point being accompanied by
two nodes. Scarified ware. Va

obliquely, the lines being drawn always from above downward — to the right in
one field and to the left in that adjoining. Many of the vertical lines are almost
as straight as if they had followed a ruler. But all the work is free-hand and
has the air of having been done by a skilled hand working rapidly. The entire
surface, both within and without, is painted. The jar is 26.5 centimeters high.
The walls at the rim are one centimeter thick and they grow gradually thicker
in the direction of the base, which is itself quite ample. Thus the vessel, though
empty, could be overturned with difficulty. Empty, it weighs 3000 grams, and
yet its capacity (sand) when filled to the brim is only 1810 cubic centimeters.
Its serviceableness as a containing vessel was relatively small. On the contrary,
it would hold weighty substances without danger of bursting and may have seen
service in that capacity. It does not seem to have been used over the fire, but
a majority of the vessels in the scarified group were so employed.

The convex bottom of another vessel not supplied with legs is completely
covered with scarifications (fig. b). This is a two-storied vase. The upper story
being left without incisions while the lower is entirely covered by them, the
effect is the same as if a smaller bowl had been set into a larger one. It is always
desirable, therefore, that illustrations of such specimens should be labeled as

100 A STUDY OF CIIIRIQUIAN ANTIQUITIES.

composite vessels in order to avoid mistaking them for illustrations of bowls that
are stacked for the purpose of saving space.

One vase in this group is mounted on four legs (fig. c~). To complete the
zoomorphic unit, a head and tail are applied on opposite sides below the rim.
The carapace motive on either side proves that the animal represented is the
armadillo. The artist, however, took some liberty with his motive, making the
lines in the anterior and posterior sections of the carapace run longitudinally
instead of transversely. A graceful form of tripod is to be seen in figure d. The
legs are rather long for this group, but are solid like all the others. They probably
represent the armadillo, the head of which is indicated by the prominence at
the hip.

The bell-shaped, flat-bottomed tripod vase with legs missing (fig. e) is an inter-
esting type. The bottom is smooth. The sides are divided into two unequal zones
of scarifications, an upper and a lower, by a horizontal painted band near the
top. The lower zone is itself divided into three fields by means of three vertical
painted bands. The scarified surfaces were left unpainted. These give the effect
of having been produced by the impress of some textile fabric. Two lizard-like
forms in relief, on opposite sides, complete the decoration.

The entire surface above the feet of one small tripod (fig. /) is scarified in a
manner that suggests the warp and weft of basketry. Figure g is interesting in
that one of the vertical scarified bands which encompass the body of the vessel
was left unfinished. And yet this tripod had seen considerable service over
the fire.

THE MAROON GROUP.

Holmes called one group of Chiriquian pottery the maroon group " for want
of a better name." He figures four out of a total of not more than a dozen
pieces in the National Museum. The collection in the Yale University Museum

Kig. 170. — Vase with loop handles strongly resembling Fig. 171. — Vase ornamented with scarifications

that shown in figure 164; from Caldera. Maroon alternating with pairs of vertical punctate

ware. '/> ribs. Maroon ware. */»

numbers perhaps no more. Mention has already been made of the difficulty
experienced in deciding whether certain specimens belonged to this or to the
scarified group, and in that connection figure 170 was fully described (p. 97).
The paste of which the piece represented in figure 171 is composed, resembles
very closely that in the scarified ware, except that it is somewhat finer. Moreover,

THE MAROON GROUP.

101

this specimen, which comes from thirty-five miles north-northwest of David, is not
only actually scarified, but alternating with the scarifications are pairs of vertical,
punctate ribs suggestive of the long jointed lizard's tail shown in figure e
(PI. XXVI).

In a small tripod from El Banco (fig. 172), the body of the vessel is completely
encircled by a series of plain vertical ribs. The entire rib-bearing surface was
never polished and was simply stained a maroon color. The paste is relatively

Fig. 172. — Small tripod completely encircled
by a series of plain vertical ribs ; from El
Banco. Maroon ware. V»

ig- 173. — Bowl with ornamentation in relief;
from Divala. Maroon ware. '/«

fine and very tenacious. The ribs are applied fillets, triangular in section. Of
similar paste is a shallow bowl from Divala (fig. 173). The unpolished band
encircling the vessel just under the rim is decorated with a circle of eye ornaments
in relief, the circle being broken by the figure of a frog at each pole.

The open-mouthed bowl with annular base and animal features in high relief
(fig. 174) is the most characteristic piece in the lot. The rim and inner surface

Fig. 174.

Fig. 1 74. - Open-mouthed bowl with annular base and
animal features in high relief; from near David.
Maroon ware. '/«

Fig- '75' — Vase with single handle, opposite which
is a relief^ ornament consisting of a pair of eyes,
prominent eyebrows and a nose ; from Gualaca.
Maroon ware. '/>

are highly polished and painted a rich purplish maroon color. This bowl comes
from near David (Lat. 8° 31' N. ; Long. 82° 27' W. of Greenwich).

A vase from Gualaca, with single vertically placed handle (fig. 175), completes
our list of illustrated specimens for this group. Its only relief ornament consists
of a pair of eyes, prominent eyebrows and a nose, opposite the handle. The
weathered outer surface and the aperture were once coated with maroon paint.
The paste is coarse, but relatively durable.

102

A STUDY OK CHIRIQUIAN ANTIQUITIES.

THE WHITE LINE GROUP.

Only four specimens can be referred with certainty to this group. Three others
have all the characters except that white paint was not used. Six of the seven
are tripods with narrow necks. The piece without legs (fig. 176) is the largest
of the group and differs from the tripods in having a comparatively large orifice.

The shoulder meets the body proper at
a rather pronounced angle, below which
are numerous striae produced by an un-
usually rough polishing instrument. The
ground tint is a dull red. Two indifferent
animal heads, in relief, are used as shoulder
ornaments. The shoulder is also decorated
with six groups of vertical lines in white,
that reach down to the peripheral angle.
Each line is accompanied by one or two
rows of white spots, the spots being mar-
ginal and in actual contact with the line.
This motive repeated in the next figure
is similar to one that is often employed
with variations in the lost color and
alligator groups, and is evidently an al-
ligator motive, derived from a dorsal view

of the alligator (see Pis. XXXI, XXXVI). The vessel is badly smoked from use
over the fire.

Similar decorative motives are employed on a tripod (fig. 177) with gracefully
curving outlines and narrow orifice. The ground tint is the same — a dull red.
Indented nodes applied to the shoulder on either side represent life forms. The

Fig. 176. — Vase decorated with the alligator dorsal-
view motive in white. White line ware. Vi

ig- ' 77- — Tripod, the shoulder of which is decorated
with the dorsal-view alligator motive in white.
White line ware. V"

Fig. 1780. -Small tripod with shoulder ornamen-
tation suggestive of hieroglyphics, and legs re-
presenting animal forms ; from Bugavita. '/'

rim ends square against a horizontal plane. Near its outer margin begin six groups
of white lines, three lines in a group, which are carried downward in vertical
planes to about the level of the greatest horizontal circumference of the body.

THE WHITE LINE GROUP. 103

There is a row of white spots on both sides of each group of white lines, the
spots however seldom touching the lines. This again is the dorsal-view motive
derived from the alligator, the lines and spots representing the rows of spines and
the scales on the animal's back. This motive is described in a subsequent chapter.
Each leg is marked transversely by three parallel white lines, suggesting the
carapace motive of the armadillo. Sooty incrustations cover the lower half of the
outer surface of this tripod.

A small tripod from Bugavita (fig. 178 «) presents some remarkable features in
the way of decoration, suggesting that the ancient Chiriquians might have made
use of a hieroglyphic or phonetic system of writing. This tripod differs from the
preceding in having a rounded recurved rim. The inner surface of the orifice
and the bottom are painted red. The rest of the outer surface, including the legs,
is finished in a warm reddish gray slip. The legs are provided with a head and
ears in relief that remind one of the characteristic armadillo attitude.

The white line decorations on the shoulder of the vessel are arranged in three
groups, certain symbols being repeated in each. These can be studied to better
advantage in the tracings, which show the whole series at a glance (fig. 178 b}.
Reading from left to right the first symbol is not unlike a lower-case d ; the second

Fig. 178*. — Detail of shoulder ornamentation on figure 17801. White line ware. V'

may be compared to a capital 7, and the third to a capital B. The second group
begins in the same way with the d, the top of which, however, is connected with
the base of the succeeding I by a stroke of the brush that has no counterpart
in the first group. The last symbol is slightly damaged at the base, but is
evidently the initial d symbol inverted. With one or two minor exceptions, the
third group is the negative of the second. The stroke, for example, that is carried
downward from the top of the negative d does not connect with the base of
the I, and there is a slight break in the loop of the remaining symbol. The
resemblance of these symbols to certain letters of our alphabet is of course
fortuitous.

The white line ware resembles a certain type of Costa Rican pottery more
closely than it does any group of Chiriquian pottery. The fact that comparatively
few specimens of the white line ware have been found in Chiriquian graves is
another reason for supposing them to be importations.

THE LOST COLOR GROUP.

In point of numbers the lost color group ranks next to the armadillo group.
Its chief distinguishing feature, as the name given by Holmes implies, is the
method by which the decorative motives are produced in color, or rather by the
removal of coloring matter. Aside from this universal character, the vessels present

104

A STUDY OF CHIRIQUIAN ANTIQUITIES.

a wide range in point of form and size. The vast majority may be classed as
bottle-shaped vases with globular bodies. Handles are comparatively rare. A
number of open shallow bowls are mounted as tripods. Life motives in relief are
sparingly used. In only a few instances are these emphasized sufficiently to stamp
the specimen as a zoomorphic unit. To the latter class belong a small number
of figurines that represent quadrupeds, birds and serpents and serve as whistles.
Other forms, including gourd-shaped rattles, are phytomorphic. There are also
miscellaneous forms, such as cylindrical needle-cases and double cups with single
connecting arched handles.

The paste ranges from yellowish gray to pale red. The outer surface is care-
fully formed and polished. No attention is paid to the inner surface, especially

of the narrow-necked bottles, the result being
that the walls are quite thick in some places
and thin in others. This carelessness in the
finish of the interior is seen in a bottle broken
in the plane of its greatest diameter (fig. 179).
The inner surface is covered with what appear
to be thumb-marks, a fact which, taken in
connection with the position of the break, leads
one to conclude that the vessel was made in
two pieces. After bringing the two pieces
together, the contact irregularities were removed
by inserting a stick or pointed implement through
the aperture. Marks of this stick are seen
along the line of cleavage on both halves.
Such an interior was not suited to domestic
purposes, an opinion also supported by the
wealth of exterior ornamentation and the ab-
sence of sooty incrustations. These vessels
were valued for esthetic and symbolic reasons
and not for their storage capacity or as utensils.
The making of narrow-necked vessels in two
or more parts has been reported from Peru.
Dr. Davis exhibited specimens of this sort at a meeting of the American Ethnological
Society.1 December 15, 1859. One of these " of spherical form had separated
itself into two hemispherical halves, by an even, horizontal fracture."

Sometimes but a single ground color is employed, either a pronounced red
pigment or a light to salmon-colored slip. Frequently the two ground tints
appear in pleasing combination, the lower half of the body, for example, being in
red, the upper half light, and the neck red. Again the red is the true ground,
on or in which appear bands or fields of white ; or the ground may be white and
marked by bands of red. The upper light zone is frequently crossed by bands
of red tangent to the neck. But whatever the combination, the red and the light
are each and both only the ground on which the artist works out his design.
This brings us to the secret of the lost color process.
1 Hist, mag., IV, 48.

179. —Narrow-necked vase broken in
plane of greatest diameter, illustrating care-
lessness in finish of interior. Lost color
ware. '/'

THE LOST COLOR GROUP. 105

The secret was unknown to Holmes when he so aptly named the group. He
supposed that " the actual patterns, so varied and interesting, were worked out
in a pigment or fluid now totally lost, but which has left traces of its former
existence through its effect upon the ground colors. In beginning the decoration,
a thin black color, probably of vegetal character, was carried over the area to. be
treated, and upon this the figures were traced in the lost color. When this color
(if it was indeed a pigment, and not merely an acid or 'taking out' medium)
disappeared, it carried with it the black tint beneath, exposing the light gray and
red tints of the ground and leaving the interstices in black. The interstitial
figures thus formed are often of such a character as to be taken for the true
design. In examining the decoration of this ware, it is essential that this fact
should be kept in mind, as otherwise great confusion will result."

Holmes declared that the nature of the materials employed could not be de-
termined. The foregoing quotation is given to show what a really good guess
Holmes made. His errors were only two. The black was not applied before the
' taking out ' medium, and his so-called ' taking out ' medium was neither a pigment
nor an acid.

While on an expedition to Central America in 1896 and 1897, Professor C. V.
Hartman l visited the Aztec village of Izalco, Salvador. There he was able to
observe a method of ornamenting calabash vessels by means of what has been
called •' negative painting." The work is done by women only. The shell is
opened with a knife and the seeds and soft interior removed. The skin is taken
off by means of a bivalve (Area) shell. When sufficiently dry, the outer surface
is decorated in the following manner : The designs are traced with a small paint
brush, the medium being beeswax blackened by smoke and kept in a fluid con-
dition by heating in a clay vessel over a charcoal fire. The second step begins
with the preparation of a black adhesive solution, consisting of sugar or honey,
powdered charcoal and the pod of a leguminous plant (Papilionaceae). The mix-
ture is placed in a large earthen vessel and allowed to boil over a fire. Its
adhesiveness comes from the sugar, its color from the charcoal, and the pulverized
pods give the varnish-like finish. The solution is rubbed on the outer surface of
the calabash vessel with a rag and left to dry. The vessel is finally immersed in
a bath of hot water, which melts the wax and removes with it the color from the
waxed portions, leaving the pattern in the color of the original ground. The hot
water does not injure in the slightest the black that remains ; so that instead of
a white vessel with dark ornamentation there results a black surface with designs
in white.2

What explanation could be more simple, effective and plausible? The so-called
'taking out' medium used by the ancient Chiriquian was therefore evidently wax,
and it was applied before instead of after the coating of black, portions of which

1 Die Baumkalabasse im tropischen Amerika. Boas anniversary volume, 1906. New era
printing company, Lancaster, Pa.

* Professor Pittier, who quite recently visited Salvador and bought various calabash
vessels at Izalco, states that the ancient negative method of decorating is no longer practised
there. (Note to Professor Hartman.)

MEMOIRS CONN. ACAD., Vol. III. 14

106

A STUDY OF CHIRIQUIAN ANTIQUITIES.

it removed on being heated, leaving the desired pattern in the ground color and
the interstices in black : the latter is therefore not a true delineating color. The
name lost color, however, is still as appropriate as it was when first given to the
group more than twenty years ago. Hartman and others call it " negative paint-
ing." The same process was known to the ancient inhabitants of Ecuador, Pro-
fessor M. H. Saville having found some fine examples of it around Rio Bamba.
It also occurs in northern Ecuador and southern Colombia. I have just discovered
in the Keith Collection of ancient pottery from Mercedes, Costa Rica, a vase
painted in this manner. Hartman cites recent examples from Guatemala, and it
is also met with in the native art of Java and Hawaii.

Much of the richness in contrast between the black interspaces and the patterns
in the original ground colors is lost, owing to the ease with which the black rubs
off. When new, the ware must have been highly effective. So much of the black
pigment has disappeared through usage before burial, and especially from long
contact with the earth in a region of relatively great rainfall, that the original
ground colors show everywhere through the black, and in many places the pattern
is completely lost because of the absence of the black.

Beginning with examples in
which the entire original ground
is red, a vase from Divala with
linear decorations is reproduced
in Plate XXVII (fig. a). It took
a high degree of skill to arrive
at such perfect proportions as
are here exhibited. The pattern
was first laid down in wax over
the red, the entire outer surface
then being coated with black.
After the vessel was passed
through the hot-water bath, the
pattern appeared in the red of
the original ground ; what is
left of the black on the inter-
spaces becomes what might be
called the final ground, and what
was originally the ground takes
the place of the delineating color.
The framework of the pattern
consists of two horizontal bands,
one broad and one narrow, just below the greatest circumference of the body,
together with two sets of similar bands tangent to the neck on either side and
in nearly vertical planes. These divide the upper zone into two arched panels
and two alternating upright panels. Two bands in a vertical plane mark each
upright panel. The outer surface of the neck is marked by bands in a similar
sense, while the lip and inner surface of the orifice are left in the original ground
tint. The arched panels are filled in with groups of parallel bands that form a

Fig. 180.— Vase in red and black with large aperture, and two
arched panels on shoulder. Lost color ware. '/»

THE LOST COLOR GROUP. 107

sort of compressed or faulted meander. The lower zone is in black except for
four groups of radiating bands passing from the lower horizontal band down
about half-way to a median point on the bottom. This is a favorite treatment
for the lower zone.

The modeling in figure 180 is much inferior to that of the preceding and the
aperture is unusually large for this group of ware. The lip is red and recurved
(the pattern seldom encroaches on the lip and never on the inner surface of the
aperture). There is a similar framework of horizontal peripheral bands dividing
the, body into an upper and a lower zone, and of bands tangent to the neck on
either side, dividing the upper zone into four pauels, two being arched and two
upright. The decoration of the upright panels is more felicitous than it was in
the colored figure a. In both instances, the endeavor was to break up the broad
expanse of black. Here, on the side turned toward the observer, it is very
successfully done by a plantlike form with two pairs of lateral fronds. The
opposite panel is treated differently. There is once more the faulted meander
filling the arched panels. The treatment of the lower zone is characteristic for
the group.

Other examples of vases with solid red ground and linear ornaments are given
in Plate XXVIII. The same method of treatment is repeated in figure a, except
that the two main panels of the upper zone are quadrangular instead of arched,
and the faulted meander is cut in two by a series of vertical bands. The upright
panels being narrow are not decorated. The neck is ornamented with vertical
bands that encroach on the lip. The frequently employed framework of horizontal
peripheral bands and others tangent to the neck on two opposing sides is found in
figure b. The arched panels thus formed are ornamented in a singularly pleasing,
happy-go-lucky fashion. Bands arranged in groups of two or singly meet at various
angles and the angular black interstices, when large enough, are marked by one or
several spots, some of which are set in small circles or rudely triangular spaces.
The lower zone is decorated with eight series of bands in alternating groups of
two and three, converging toward the bottom.

The straight-line or banded motive is continued in figure e, but the effect is
wholly different. There is the same separation of the body into two zones by
two horizontal peripheral bands. The decoration of the upper zone consists of
ten groups of lines or narrow bands (the red, not the black) radiating from the
neck and alternating with triangular spaces. The upper part of the lower zone
is marked by seven horizontal bands and a single horizontal series of narrow
elongated quadrangular panels, each enclosing a single row of spots. The lower
part or bottom is left in black, as is also the outer surface of the neck. The ab-
sence here of visible fields in black might easily lead one to mistake the black
linear interspaces for the real pattern, the radiating triangular spaces in the color
of the original ground tending to strengthen the deception ; but the black is
never a true delineating color in the lost color group. The polisher used on
this vase being rather coarse, the method of using it can be easily detected. The
strokes were in straight lines. Those on the upper zone were tangent to the
neck and on four sides, the stria? on opposite sides being roughly parallel to
each other. The striae on the bottom are all in one direction, as if the position

108 A STUDY OF CHIRIQUIAN ANTIQUITIES.

of the vessel relative to the workman had not been changed throughout the
process. It is worth while to note also that the polishing seems to have been
done after the application of the final coat of paint (black), as the paint has
disappeared along the lines of the striae. Such, however, was not the sequence
of events. On closer observation the black paint is found to have been removed
from the ridges of the striae and not from the intervening grooves ; while the
underlying red paint has not been worn even from the ridges. The wearing of
the black paint from the ridges is therefore due to its exposed position and the
non-adhesive character of the coloring matter employed.

With but slight variations, what has been said of the preceding is also true
of figure rf, except that, in place of the series of elongated dotted panels on the
lower zone, there is a horizontal band of triangles with apices pointing upward,
alternating, of course, with triangular black interspaces. The upper zone is rather
sharply depressed, giving to the vase angular outlines.

An especially pleasing arrangement of groups of parallel bands, meeting each
other at various angles with a variety of intervening triangles, quadrangles and
pentagons, is given in figure e. One band never cuts another. When two groups
meet, one disappears beneath the other. At one place four groups of bands meet
and there is the same orderly overlapping. Three narrow bands are carried round
the shoulder just below the neck. The latter is short, with vertical sides, while
flaring necks are usual for this group of ware. The entire lower zone is left in
the original red ground tint. Here again one might easily take the black bands
to be the pattern against a background of angular fields in red. This vase is
from Divala.

The upper zone in figure / is traversed by a meandering group of narrow
parallel bands, the four lower corners of the meander being cut by a group of
narrow horizontal peripheral bands. With the exception of a single horizontal
band near its upper margin, the lower zone is left in the original ground color.
The. lip is flaring and rectangular in outline, there being but two other examples
of this type of lip in the lost color group.

In Plate XXIX are shown some specimens with handles and some with orna-
ments in relief. The original ground color is red. The handles in figure a are
small vertical loops, with black and red cross bands. The collar from which
they spring to meet the prolonged lip on either side is decorated with groups of
narrow parallel bands, forming a broken meander. The remainder of the upper
zone of the body is marked by horizontal bands and panels, each panel with a
row of spots in a rectangular field. The lower zone is left in red.

The type of neck is somewhat different in figure 6, and the lip is recurved,
but the cross-banded handles are practically the same as in the preceding. The
pattern is confined to the upper half of the body and consists of four triangles
with the rather large intervening black spaces relieved by groups of lines forming
compound and simple curves. The hollow of each curve in the sigmoid scroll
as well as the simple curve is marked by a spot. This spot represents the body-
markings of the alligator. The sigmoid scroll is therefore composed of two
alligator bodies linked together, while the simple scroll on the right is the body-
line of a single alligator. This is one of the two predominant alligator motives

THE LOST COLOR GROUP. 109

so well exemplified in the alligator and the polychrome ware (see PI. XL, fig. /;
and text-fig. 256). Its occurrence in the lost color group is one of the many
links that bind the ceramic art of Chiriqui into one consistent whole.

The narrow neck of a bottle-shaped vase from Divala (fig. c) was never polished
and is surrounded by a series of eye ornaments in relief. The contact between
the upper and lower zone is slightly angular. Groups of radiating bands with
intervening triangles mark the upper half. Some of the broader black interspaces
are relieved by rows of spots. With the exception of a small round area on the
bottom, which is in solid black, horizontal bands and panels completely cover
the lower zone.

Reference has already been made to the lack of adhesiveness in the final coating
of black paint, which is supposed to be of a vegetal nature. Nowhere is it com-
pletely preserved. In many cases very little remains ; in others it has disappeared
entirely.

In figure d, a specimen from Divala, only the upper half received the wax
treatment. Above the three horizontal equatorial bands, a diamond-shaped panel
is repeated six times. In each panel are series of parallel bands meeting at an
angle and grouped around a central cross, the motive being similar to that on
the inner surface of an open shallow bowl of the red line group.

A wide-mouthed vase with recurved lip is reproduced in figure e. The pattern
is confined to the upper half of the body, the final coating of black encroaching
but a short distance on the lower red zone, which is separated from the upper
bv horizontal bands. Similar bands are carried around the shoulder just below
the neck. The three series of vertical bands accompanied by marginal serrations
are motives derived from the dorsal aspect of the alligator. These bands alone may
be looked upon also as completing the formation of three quadrangular panels,
all being treated in the same way — a nest of triangles in the center, with the
remaining space covered by a triangular piece-work of red and black.

Thus far the body of the vessels of this group has always been divided into
an upper and lower zone. A distinct departure is taken in figure /. Under the
recurved lip there is a pronounced collar, with a life form in relief — a prominent
beaklike nose, eyes set wide apart, and a tail opposite the nose. The framework
of the design on the body consists of three series of longitudinal bands. Each
of the three panels thus formed is subdivided into three panels by two groups of
horizontal bands. These panels are all traversed by faulted meanders. The wider
of the intervening black bands are generally marked by a single row of spots
in the color of the original ground. This style of ornament recalls the serpent
motive, which characterizes the group by that name.

One of the vases with original ground of red (fig. 181) had seen much service,
the recurved lip having disappeared piecemeal. The broken surfaces are aged
and smoked and the tone of the entire red ground is deepened. On opposite
sides of the body are life forms in low relief. Alternating with these on the
upper zone are two four-sided panels, each filled by a painted design that retains
only slight traces of a life form. The elements of the design are the same on
either side, but the combination of these elements is such as to produce two
somewhat different results. The motive resembles one that reappears on a number

110

A STUDY OF CHIRIQUIAN ANTIQUITIES.

of vases (see PI. XLII). There are also in this group several small red bottles,
the ornaments on which consist of plain and scalloped horizontal bands.

As a rule, in vessels of the lost color ware, the original ground consists of
two colors — red and a light cream color that in rare cases is replaced by salmon.
The two ground colors are usually applied in zones, figure b (PI. XXVII) being
a good example, reproduced in color. The wide neck and bottom are red. The
middle zone is a rich cream color, which is now visible only as the pattern on
an ultimate field of black. Between the two bands about the neck and the two
peripheral bands (the lower one of which is in the red zone) is a row of ten
monkey-like figures, all facing the same way and in precisely the same attitude.
Each is a conventionalized view in absolute profile, of a monkey sitting bolt

Fig. 181. — Vase in red and black with narrow aperture,
showing life forms in low relief alternating with
two panel decorations. Lost color ware. '/>

Fig. 182. — Vase in red, light cream, and black, the
three panels of the shoulder zone being decorated
with alligator motives. Lost color ware, '/i

upright, its long erect tail reaching as high as the head and recurved at the
tip. Both eyes are represented on the side toward the observer, but no attempt
is made to show more than two of the four feet. These are conspicuously large
and two-toed, the only exception being those of the tenth monkey, which is
restricted to a smaller space than the others. The feet here are not spread,
the body is more slender, and the eyes are placed in a vertical row instead of
horizontally. The whole effect is such as to reveal in a striking way the methods
employed in the lost color process. A tripod of this group is decorated with a
similar monkey design. It should be recalled that long-tailed monkeys abound
in the American tropics, the Capuchin monkeys (Cebus) being the most numerous.
Other genera represented are: Mycetes, Ateles and Nyctipithecus.

The same disposition of the two ground colors is seen in figure 182, where
the three panels of the cream zone are filled with highly conventionalized life
motives attached to the banded framework of the design and sharply recurved at
the free end. The first figure in each row is attached to the band above, as is
the attachment of every alternate figure to the band below. Each is accompanied

THE LOST COLOR GROUP. Ill

by one or two rows of dots. The design is referable to the spines and scales
of the alligator. The width of the panels being greater than that of the light
upper zone, the lower part of the design in each appears in the red ground of
the lower zone.

In figure 183 very little black is used, just enough to serve as a setting for
the design. It is, in fact, one of the few examples in which the black becomes
a component part of the design in two colors — black and light on a light ground.
The same may be said of the black and red designs on a red ground (see
PL XXIX, fig. d). It is not a perfect example, however, for the black frame-
work of the design is fused with the solid coat
of black on the neck. Two series of gracefully
curving bands, some continuous and some broken,
with a horizontal connection at their bases, form
a bilaterally symmetrical figure that covers almost
half the upper zone. This is repeated on the op-
posite side.

The original ground colors of the series brought
together in Plate XXX are also light above and red
below. Figure a presents what might be considered
the vestiges of a square lip. The otherwise
circular margin of the rim is interrupted at four Fig l83._vase in red, light cream and

points by Slight prominences, tWO of which have black, ornamented with bilaterally

~- T-. • . i symmetrical figures each covering

been broken off. Painted eye ornaments surround a'bout half of Je uppcr zone Los6t
the neck, and the shoulder is adorned by two panels color ware. V»
with a design of plain and scalloped horizontal bands.

Life forms in relief are comparatively rare in the lost color group. They
usually consist of an animal head projecting from the body of the vase on one
side and a tail on the other. Beyond this the globular form of the body is not
interfered with. In figure £>, two heads appear on one side and a single tail on
the other. The heads are alike and reptilian in character, with eyes represented
in color. The tail was cut off short in a vertical plane that is not at right angles
to the axis of the common body. The two panels that reach laterally from head
to tail may be intended to supplement the representation in relief, thus completing
the connection between the double head and the tail.

The animal head and tail in the round on the shoulder of the vase reproduced
in figure c are those of the racoon. Although the head is conventionalized,
there is no mistaking the genus in the aspect of the pointed nose and the position
of the eyes. The latter are represented both in the round and in color. The
tail is cross-banded but much shortened to avoid breaking. The well-known
Procyon lotor does not range farther south than Costa Rica. Procyon cancrivorus,
a crab-eating racoon, that belongs in Colombia and Guiana, is also found as
far north as Chiriqui and is probably the species here represented. Quadrangular
panels decorated with dorsal-view (alligator) motives, some of them incomplete,
reach from head to tail on either side. A series of eye ornaments surrounds the
neck of the vase.

The design in figure d consists of radiating lines enclosing rows of spots.

112 A STUDY OF CH1RIQUIAN ANTIQUITIES.

These bands are repeated in groups of two ; triangular spaces alternating with
each group. This recalls the prevailing motive in the serpent ware. In figure e
a similar serpent motive is also seen decorating the arched panels as well as the
alternating fields.

The cream zone in figure / does not reach quite to the neck. Above and
below this ornamented zone the ground is red. A single vertical handle unites
rim and shoulder. Meeting the latter at a point within the light area, it is also
light instead of red. Of the four panels formed by series of vertical lines, two
are narrow and two wide. The decorative motive for the latter is in the shape
of a labret or spool. It is distributed in such a way as to make both horizontal
and vertical rows. There is a certain system in the irregularity of the everywhere
connecting black interstices, that causes the eye to be fixed upon them rather
than upon the rows of labret-shaped figures forming the real design. The figure
in the upper left-hand corner of one of the panels is just half the size of the
others, the space left over when finishing the decoration being too small to
accommodate one of the regular size, and too large to be left unbroken. A like
attempt to utilize the left-over space occurs on the opposite panel, where half the
motive is employed instead of a whole one reduced in size (see also fig. 196). The
spool or labret ornament is found on other vases, sometimes being the center
toward which radiating bands converge (see fig. 195).

Thus far the oft-recurring framework of bands, peripheral as well as tangent
to the neck on either side, has been in the color of the original through which
they pass. Where the whole ground was red these bands were red also (Pis. XXVII
and XXVIII), and where the upper half was light the bands cutting it tangent to
the neck were light (PI. XXX). In Plate XLII (fig. a) all the original ground is
red with the exception of the peripheral band and those tangent to the neck,
which are white. They were painted on first; then the bottom and the upper
panels were colored red ; finally came the waxing process and the coat of black,
which was eventually removed from the design by melting the wax. This specimen
is from Divala, and is not only beautifully modeled and painted but also one of
the best preserved of the entire group.

The outline is softly angular, due to the sloping shoulders and slightly pointed
bottom. The equatorial band cuts the body into two almost identical halves, the
upper being surmounted by a narrow neck and projecting but well-formed lip.
The framework of bands is a thick white paste that has received a high polish.
The red field encroaches somewhat on the margins of the bands, and the black,
the last to be applied, narrows them still more. The lower half of the vessel is
in two zones, the nether of red and the upper of black, with groups of narrow
longitudinal bands. The neck and lip are also banded. Of the four panels
surrounding the neck, the vertical ones are decorated with triangles. The design
on the arched panels, alike on both sides, may or may not be construed as a
life form. The diamond-shaped center may represent the body, and the six
appendages, the head, tail and legs. These are all composed of the same ele-
ments — a short straight band accompanied by two longer curving parallel bands.
The four short appendages are attached to the center of the four sides of the
square ; the two long ones, at opposite corners. The whole is a piece of decorative

THE LOST COLOR GROUP.

113

work admirably conceived and adapted to the space at the artist's disposal. The
design may have no significance other than this.

Sometimes the horizontal band is placed much below the plane of the greatest
horizontal diameter of the vase, as in figure 184. The long vertical panels are
decorated with a serpentine design. The design on the relatively large arched
panels is made up of a different arrangement of the same elements — narrow bands
and rows of spots. Groups of these cross at varying angles, producing an effective
bit of decoration. The lip is delicately modeled.

The vase reproduced in figure 185 is exceptional from several points of view.
The equatorial band and the two tangent to the «eck are simply those parts of
the cream-colored slip not covered by the red, instead of the thick white paste
employed in delineating the fundamental bands of the preceding figures. In the
latter also these were retraced in wax, so that the ultimate design never crossed

Fig. 184 — Vase in which the peripheral
band and those tangent to the neck
are painted white. Lost color ware. '/>

Fig. 185. — Vase in which the peripheral band
and those tangent to the neck are painted
white and are crossed by the lost color
pattern. Lost color ware. 'I'

them, but was limited to the panels and the lower half of the body. Here,
however, the design is absolutely independent of the white bands crossing and
recrossing them in every direction. The foundation of the intricate design seems
to be a series of parallel bands carried in waves about the body of the vessel,
rising almost to the neck three times and as many times dropping a little below
the light equatorial band. Above and below these waves are loops, cross bands,
spots, circles, triangles, etc. The bottom was discolored (black) in the process
of firing.

There is still another series of vases in which the peripheral bands and those
tangent to the neck are red on a light ground, as illustrated in Plate XLII (fig. 6).
The upper light zone is crossed by two red bands tangent to the neck. The
bottom, lip and inner surface of the orifice are also red. The neck is ornamented
with a series of composite eye ornaments. The vertical panels are marked by
bands and by small triangular and oval to circular spotted fields. A disjointed
meander traverses each arched panel.

MEMOIRS CONN. ACAD , Vol. III. 15

114

A STUDY OF CHIRIQUIAN ANTIQUITIES.

In figure 186 the red bands rise vertically from the lower zone, cross the light
upper zone and fuse with a red band that encircles the short neck. Four-sided
panels, four in number, are thus produced. The details of the decoration on each
cannot be determined, owing to the almost complete loss of the black paint.

A new element entering into the technique of the decorator is shown in figure
187. The light zone in the color of the slip includes the neck, and is considerably
larger than the red zone below. It is separated into two panels by vertical
bands in the color of the slip instead of in red, while the red is used as a true
delineating color on each panel. Having been polished down before the paint

Fig. 186. — Vase ornamented with four four-sided panels.
Lost color ware. '/>

Fig. 187. — Globular vase ornamented with a design
suggesting a conventionalized alligator. Lost color
ware. '/>

was thoroughly dry, the outlines are blurred in places and some of the red has
been carried out onto the light ground. The design, suggesting a conventionalized
alligator, was never wholly finished in red and is not quite the same on the two
sides. When being retraced in wax, the missing parts were added and now
appear in the light ground color. The whole figure was surrounded by a narrow
band (in light) now partially effaced in places from exposure or handling. The
black interspacial areas are reduced in size by dashes of the brush that left light
bands of varying lengths.

There are four distinct zones on the globular vase shown in figure 188 ; a red
zone reaching from the rim almost half-way to the plane of greatest diameter, a
rather broad light zone, a narrow red zone representing also the peripheral band,
and the bottom light again. The broad light zone (now black except the design)
is decorated with the labret- or spool-shaped motive already mentioned in con-
nection with Plate XXX (fig. /). The artist repeated the motive in a horizontal
series by adapting the size and number of figures to the space at hand. The
result is that the last one to be made is considerably reduced in size. The narrow
red zone below is ornamented with a row of lozenge-shaped figures.

THE LOST COLOR GROUP.

115

The tendency to break up light surfaces with bands of red is well illustrated
in figure 189, representing a wide-mouthed vase. The ground of the principal
zone is light. The red ground of the bottom invades the light upper zone at
four points, each ascending tongue, forming the center of a design in the color
of the light ground, is composed of two looped bands, the outer margin of the
outer one being scalloped.

Fig. 189.

Fig. 190.

Fig. 191.

I'ic;. 1 88. — Globular vase decorated with the spool-shaped motive. Lost color ware. '/>

Fig. 189. — Wide-mouthed vase. Lost color ware. '/>

I' it;- 190. — Small vase with rosette-like ornamentation. Lost color ware. '/»

Fig. 191. — Compound vase decorated with three quadrangular panels in each story. Lost color ware. */»

There is only one equatorial band of red on the vase represented in figure 190.
The bands tangent to the neck on two sides, so frequently met with, are here
continuous, forming a vertical circle. The radiate ornament is repeated below
the equatorial band, so that the complete design is in the form of a rosette. At
its center is a labret-shaped figure. The red band bears no relation whatever to
the ultimate decoration of the surface.

A STUDY OF CHIRIQUIAN ANTIQUITIES.

A compound vase is reproduced in figure 191. The two neck constrictions,
the repeated curves of the outlines, and the decoration, all combine to give the
appearance of a small vase resting on the rim of a larger one. Of the two
original ground colors, the red as a rule predominates. Here the order is reversed,
the red being confined to three rather narrow horizontal bands at the periphery,
the lower neck constriction, and the rim, respectively. The rest is in light salmon
color. Three quadrangular panels envelop the upper zone of the first story and
the upper story is decorated by three similar smaller panels enclosing a spotted
field. The bottom is in the original ground (salmon), with the exception of a
single black band a short distance below the peripheral band of red.

An unusual disposition of the light and red colors of the original ground is to
be seen in figure 192. There is no zonal division. The red field on the bottom
is oval, instead of circular in shape, from either end of which a broad red band

Fig. 192. — Vase showing unusual disposition Fig. 193. — Unique vase with large long neck

of red and light colors of the original surmounted by a rectangular lip, and with

ground. Lost color ware. '/» tuberosities on the body. Lost color

ware. '/«

is carried vertically upward to the neck, where it divides and forms a sort of
collar. The two vertical bands are also united on either side by a red band, each
drawn in a plane slightly inclined to a horizontal plane. There are thus formed
two panels on a side, somewhat irregular in shape. The design on these in light
and black is almost wholly lost. The red paint was applied in a careless manner,
especially along the margins, an irregularity that was corrected in the waxing
process and covered by the final coating of black, the latter being worn off.

A vase of unique shape is now introduced (fig. 193) because of the treatment
of the two colors of the original ground. The lip is here prolonged on either
side until it becomes rectangular in outline. Its upper surface is painted red.
The neck is comparatively long and large. The relatively small body would be
globular but for the tuberosities on the shoulder, each emphasized by an encircling
band of red (the only red ground on the body). Of the decoration on the light
ground, including the tuberosities, very few traces remain, but enough to identify
the vessel as belonging to the lost color group.

THE LOST COLOR GROUP. 117

While the lost color group differs from all others in respect to form as well as
method of ornamentation, the decorative motives employed are often akin to those
that characterize other groups. For example, one of the distinguishing motives
of the alligator ware consists of a series of parallel lines or bands, the outer ones
being serrated along their distal margins. These lines represent the longitudinal
markings on the back of the alligator, while the marginal serrations are added in
order to emphasize the dorsal tuberosities or spines. The serrations sometimes
assume the form of triangles that are filled with dots to indicate the scales. In
other words, this motive is a conventionalized rendering of the alligator's dorsal
aspect. I have called it, therefore, the dorsal-view motive. This dorsal-view motive
is not confined to the alligator group alone. Fine examples of it occur on at
least five of the six lost color vases reproduced in Plate XXXI. Both types of
lateral serrations are seen in figure a : the triangle type on the crdam ground of
the shoulder zone, and the spine type on the red ground of the neck. The pat-
terns on the latter are marred, however, the black color of the field having dis-
appeared in part.

The original ground of the shoulder zone of the vase reproduced in figure b
is also a light cream color. But the light zone is traversed by three red bands
tangent to the neck, the red being the same as that covering the bottom and
the neck, respectively. The arched panels and the field surrounding the neck are
completely covered by a series of dorsal-view motives. Here the lateral serrations
are plain spines. The use of three arched panels, instead of two, is typical of a
series of vases belonging to the alligator ware (see PL XXXVI.)

The vase, in shape like a tea-caddy (fig. c\ is from twenty-one miles northwest of
•David. The bottom is almost flat (in another specimen the bottom is perfectly
flat). In addition to the red lip and neck, red bands encircle the body at the
shoulder and at the margin of the bottom. From the upper band, other red bands,
three in number, are carried vertically downward, then horizontally, and finally
vertically again to meet the band at the bottom, thus forming three overlapping
panels. Each design consists of an enclosure in the shape of the panel, with a
series of sessile figures otherwise suggesting the quadruped form. The zone
between the angles at the neck and the shoulder is also divided into three panels,
not by red bands, but by a dorsal-view motive in the light color of the original
ground. A single sketchy quadruped figure occupies each of the shoulder panels.

In a vase from Bugavita (fig. d), the pattern is confined to the shoulder zone
and is of a pronounced salmon color. It consists of the dorsal-view motive placed
vertically and repeated seven times. The colors are well preserved in the example
shown in figure e. It represents a vase in which all the original ground is red,
and the bands surrounding the panels are white. The coat of black, when first
applied, covered the entire visible surface except the recurved lip and inner surface
of the orifice. The two arched panels are relatively small, the pattern being alike
in both — two dorsal-view motives placed obliquely.

The original ground in figure / consists of two colors, the lip and the bottom
being red and the shoulder cream. The lip looks as if it had been cut out of
the base of a globular bowl. As it was built upward and outward, it was also
slightly incurved. The orifice is oval. The margin being four-sided, quite

118

A STUDY OF CHIRIQUIAN ANTIQUITIES.

naturally takes the form not of a square but of a rectangle, the long sides of
which are cut lower than their opposites. Patterns cover the entire outer surface.
That on the neck and lower zone consists of converging bands. The upper light
zone is divided into three panels, the pattern on all being the same. It includes
all the elements of the dorsal-view motive. The spots (scale motive), however,
instead of being placed in the marginal triangles, are bunched together in a
median field.

Fig. 194. — Small vase with radiate orna-
mentation. Lost color ware. */•

Fig. 195. — Small vase with radiate orna-
mentation, the elements of which may
be traced to an alligator motive. Lost
color ware. */>

Fig. 196. — Vase in pale yellow and black, with
zonal decoration. Lost color ware. '/"

Fig. 197. — Vase in light cream and black, deco-
rated with vertical panels each enclosing a faulted
meander. Lost color ware. */«

The lost color ware includes a series of small vases, all the original ground color
of which is a rather light cream to salmon-colored slip. The radiate ornament
already noted (see fig. 190) abounds in this series. Another example is given in
figure 194. The same ornament is repeated in the circular panel on the opposite
side. Some of the elements of this pattern suggest those that make up the dorsal-
view motive. They may have been derived from another source, however. An-
other variety of the radiate ornament or rosette is seen in figure 195. Here again
are present the elements that form the dorsal- view motive, but the radiate arrange-
ment of the marginal triangles gives a different aspect to the motive as a whole.

The zonal decoration recurs in figure 196. The bottom is the color of the

THE LOST COLOR GROUP.

119

salmon slip. Above are alternating bands of black and salmon color. The two
broad black bands are broken up by a succession of diamond-shaped figures.
Red occurs only on the rim and the inner surface of the orifice. The panels in
figure 197 are vertical, reaching from the collar to the bottom, and bounded by
groups of parallel lines. Each panel holds a faulted meander, itself composed of
parallel lines.

Zonal and panel decoration is characteristic of the lost color group. In very
rare instances is the panel idea lost sight of. Figure 198, representing a small
vase from Divala, is a case in point. The original ground tint was cream-white,
the design executed in wax being therefore of that color. It consists of two

Fig. 198. — Small vase decorated with parallel coiled
bands ; from Divala. Lost color ware. '/»

Fig. 199. — Vase with handles and with frondlike
ornamentation. Lost color ware. '/'

parallel bands, one solid and one dotted, curved in the shape of a ram's horn
that describes a vertical circle twice, the base of the horn being at the neck and
the tip in the plane of the equatorial diameter. The black field within the outer
coil is decorated with a white pattern in the shape of a four-pointed starfish.
The design is repeated on the opposite side of the vessel and in the same sense,
i. e., each represents the right horn. A similar design is found on a small vase
from Bugavita. In the latter case, the coil is carried round three times and is
therefore closer, leaving no room for the starfish pattern.

The vase reproduced in figure 199 is provided with a pair of handles uniting
shoulder with prolonged lip. The handles having been applied before the polishing
took place, their under surfaces as well as the neck and the parts of the shoulder
underneath the handles are left in the rough. The lip and the outer surface of
the handles are painted red. The original ground of the neck and entire body
is a pale salmon slip. The black is so nearly gone that the decoration is much
dimmed. A frondlike motive is repeated over and over again. A median point
on the bottom is the center. Through it pass four slender vertical parallel bands
reaching from a position half-way between the handles on one side to a like
position on the opposite side. Beginning at the bottom and ascending on either
side to the neck are successive pairs of opposite fronds. Each frond consists of
three slender parallel bands rather sharply curved at the tips, with a single row
of spots adjacent and parallel to the longer upper band. In some cases this row

120 A STUDY OF CHIRIQUIAN ANTIQUITIES.

of dots is carried up the stem to the base of the succeeding frond. In the large
field below the base of each handle there is a design with diamond-shaped center
and six frondlike appendages, recalling the design in the arched panels of the
vase reproduced in Plate XLII (fig. a).

As has been already pointed out, representations of the human form are prim-
itive in their conception and execution. This is true of the two human figures
in relief on the shoulders of the vase reproduced in figure 200. The attitude in
both cases is striking. The human figure on one side seems to be in the act
of descending feet foremost and with back against the surface, to which it clings
with outstretched arms. The treatment of the head is especially interesting. The

Fig. 200. — Vase with primitive human figures as shoulder

ornaments. Lost color ware. '/>

Fig. 201. — Very small vase with relief figures on shoulder
representing the human head and arms. Lost color
Fig. 200. ware. '/»

neck is not indicated ; the chin receding ; the mouth open ; nose and retreating
forehead are in a straight line; and the occipital region is flattened, meeting the
frontal portion at a rather sharp angle. This type of head is to be seen in
figure 201 and on some of the figurines and whistles (see fig. 266). It also recalls
the primitive figurines with birdlike heads, from Argos, that date back to 1200 B. C.
Black circles of paint represent the eyes. The other human figure is the same
in every respect save the attitude. It seems to have fallen backward while in
the act of ascending face foremost. The legs are sharply flexed at the knee
and the arms are extended to prevent further descent. The upper zone in which
the relief ornaments occur is light cream, while the bottom, the margin of the lip
and inner surface of the orifice are red. The black paint has almost entirely dis-
appeared, so that the design cannot be determined beyond the series of painted
eye ornaments surrounding the neck.

In figure 201 is reproduced one of the smallest examples of the lost color ware.
The original ground is partly red and partly light cream. The relief figures on
the shoulder represent the human head and arms. The head belongs to the prim-
itive bird type and there is a distinct coiffure in relief, accentuated on the rather
flat top by bands of light and black. The eyes are in relief that is heightened
by the use of color. The upper zone is ornamented with the dorsal-view (alligator)
motive.

THE LOST COLOR GROUP.

121

The black paint has entirely disappeared from the vase represented in figure 202.
The orifice, rim, and bottom are red and the shoulder zone, light. The relief
ornament consists of a birdlike head, opposite which is a horizontally applied loop
handle taking the place of a tail. A small vessel from Divala, 25 miles west-
northwest of David, is shown in figure 203. A single vertical handle connecting
rim and shoulder on one side is balanced on the other by a figure in relief, with
human body and a beak of a bird, presumably the parrot-god so well represented
among the gold figurines (see figs. 369, 373).

The group includes a number of small double vessels with single arched handle
connecting the inner margins of two lips. One of these from El Banco, is shown

Fig. 202.— Vase ornamented with bird's head and
tail in relief, the latter resembling a horizontal
loop handle. Lost color ware. Vt

Fig. 203. — Vase with vertical handle, opposite which
is a figure with human body and the beak of a
bird, presumably the parrot-god ; from Divala.
Lost color ware. */»

Fig. 204. — Small double vessel with single arched
handle ; from El Banco. Lost color ware. V"

Fig. 205. — Double vessel with two outer handles
and a single short arched central handle. Lost
color ware. '/'

in figure 204. The original ground is light cream. No traces of black paint
remain. The outer surface is blackened by smoke. The crest of the arched
handle is marked, by a fillet. There is no communication between the interior
of one bowl and that of the other.

The double vessel reproduced in figure 205 differs in detail from the preceding.
The original ground is of a rich salmon color. Each of the two vessels has a
vertical handle connecting the outer margin of the lip with the shoulder. A single
short arched handle connects the adjacent margins of the lips. This handle in
common is strengthened by a vertical prop. The two bowls are placed close
MEMOIRS CONN. ACAD., Vol. III. 16

122 . A STUDY OF CHIR1QUIAN ANTIQUITIES.

together, their interiors communicating by means of a foramen large enough to
admit the little finger.

A comparatively small number of vessels belonging to the lost color group are
mounted as tripods. A selection from these to show the variations in form and
ornamentation as well as the character of the tripod legs, is seen in Plate XXXII.
The body of a tripod vase from Divala (fig. a) is so altered in the equatorial
region as to resemble the crab. The mouth is indicated by a node with one
horizontal and three vertical incisions. Next to the mouth is the first pair of
legs, flexed, with incisions to indicate claws. Back of the legs and making the
complete circuit of the body are six pairs of spines. The third pair, being much
larger than the others, gives to the cephalothorax the appearance of being broader
than long. The crablike shape is thus rendered without lengthening the equatorial
diameter of the interior in the direction of the large spines. The proportionately
great breadth of the cephalothorax is similarly indicated in several crablike tripods
of the armadillo group, with this exception — that in some cases the greatest
horizontal diameter of the interior is in line with the pair of great lateral spines,
thus further emphasizing the breadth of the cephalothorax. The neck of the
vase in question is red ; the shoulder, i. e., the dorsal part of the crab's body, is
light cream ; the ventral surface or portion below the spines is red, with the ex-
ception of a modified trefoil or trilobed pattern on the bottom and between the
the tripod supports, which is light cream. The solid tripod supports are deeply
incised, painted red, and cross-banded by the usual lost color process. The most
interesting part of the painted decoration is reserved for the back of the crab or,
in other words, the shoulder of the vessel, which is divided into four panels by
labret- or spool-shaped ornaments, each one being accompanied by two eyes. This
particular combination of the two motives evidently represents the animal head.
One of these is therefore very appropriately placed directly over the crab's mouth.
The same design unaccompanied by eye ornaments was noted in Plate XXX
(fig. /) and in figure 190. Each panel is decorated with a row of monkeys sitting
upright, with long recurved tail and extended legs. All are facing in the same
direction, i. e., to the right, just as in the vase from Bugavita (see Plate XXVII,
fig. 6). In both examples the monkey figures are so nearly alike that they might
be considered the work of one school of artists, if not, indeed, of the same artist.

Figure b, reproducing a shallow tripod bowl, is an interesting example of con-
verting the body of the vessel into the the body of the animal without essentially
altering the simplicity of the original phytomorphic outlines. In the region of
greatest horizontal diameter of the body, the walls are suddenly drawn in for a
short distance and then turned upward and slightly outward at the margin. This
mere suggestion of a neck is excuse enough for the artist to leave the interior
unpainted. The life elements are in relief and applied to the exterior in the plane
of the angle between the narrow shoulder and the body proper. Three flattened
nodes in a horizontal plane represent the tail and wings. The neck rises almost
vertically, the head which is turned to one side being that of the turkey-buzzard
(Cathartes aura). Its baldness is emphasized by an incised fillet about the neck,
separating the feathered from the unfeathered portion ; by eyes in high relief,
nodes with annular indentations, and by absence of paint. On the other hand,

THE LOST COLOR GROUP.

the neck of the buzzard, as well as the rim and external surface of the vessel
— that is to say, the feathered neck and body of the bird — were originally painted
red, with the exception of a triangular area on the bottom bounded by the tripod
supports, which like the buzzard's head and the interior of the vessel is pale brick-
red, the color of the paste after firing. The painted portions are everywhere
decorated with groups of parallel straight bands that meet at various angles. The
supports are banded horizontally.

The lost color group is characterized by a prodigality of external ornamentation,
even the bottom of the vase, which would be invisible under ordinary circum-
stances, coming in for its share. One would expect to find a decorative design
on the easily visible interior of the open shallow bowl reproduced in figure c.
The margin of the rim and the interior are given a uniform coat of red, there
being absolutely no traces of black over a waxed pattern. With the exception
of the legs, the whole outer surface, however, including the bottom, is elaborately
decorated. The original ground of the sloping sides is a light cream slip ; that
of the bottom is a red paint. The spool-shaped ornament accompanied by a pair
of eyes is repeated so as to divide the light zone into five panels. These are
each crossed vertically by three groups of parallel bands, the group on the left
in each case being somewhat smaller than the other two groups. The decorative
unit, therefore, is the spool-head and eyes and three groups of bands, in other
words, the head and body of an animal. If an animal form, then the most plau-
sible interpretation of the banded body is that it represents the carapace of the
armadillo. The pattern in red on the bottom consists of series of curved bands
grouped about the bases of the tripod supports, and might be described as a
modified trefoil.

The neck in figure d being only slightly constricted is sufficient reason for not
painting the interior. The rim and bottom, including the legs, are red and the
shoulder is finished in a light salmon slip still visible through the black inter-
spaces in the design. Six groups of narrow vertical bands alternate with as many
labret- or spool-shaped ornaments. Three of these are on end and three placed
horizontally. Each is accompanied by a pair of eye ornaments. The spool-shaped
figure thus suggests the head and nose of some animal. There is a group of
bands for each head. Therefore, the whole is intended to represent an animal,
presumably the armadillo (see fig. c).

That the interior of the shallow tripod bowl shown in figure e was treated by
the lost color process, there can be no doubt, because of the preservation of the
black paint. The design is pleasing and bilaterally symmetrical. Traces of black
are also found on the rim, which was originally red. When painted at all, the
interiors of the lost color group are always colored red. The modified trefoil
pattern on the red bottom is similar to that shown in figure c. It may be that
the painted interiors of all the shallow bowls of this group were treated to wax
and black paint, as was the case in the last bowl described. There are very
faint traces of black bands on the interior of a large shallow tripod bowl from
Escaria.

In figure /, the interior of the tripod bowl, being easily visible, is painted (red)
and highly polished. If it was ever decorated with a pattern of black over wax,

124 A STUDY OF CHIRIQUIAN ANTIQUITIES.

the black has so completely disappeared as to leave no traces of it. The outer
surface, including legs and bottom, is all decorated ; the sides with black on light
cream, the legs and bottom with black on red. The sides are divided into four
panels by means of four groups of vertical bands, two of these groups being
themselves subdivided into three groups, the outer ones differing from the inner
in such a manner at to suggest the anterior, middle, and posterior regions of
the carapace. In the four panels are figures, apparently representing rolled-up
animals.

A deviation from both the bottle-shaped vase and the shallow tripod bowl is
noted in figure g. The sloping, flattened upper half of the body and the collar
rising sharply above produce a new type of vase. The bottom and tripod supports
are red ; the upper stories, light cream. The design on the bottom is the modified
trefoil. The legs are marked by horizontal bands.

With figure h, we drop back to the plain primitive calabash type for the body
of the tripod bowl. The rim and interior, being easily visible, are painted red
but were not otherwise decorated so far as can be detected at present. The
outer surface of the body is divided into two zones, an upper in the light color
of the slip and a lower in red. Both were decorated, faint traces of the black
paint being still visible. The legs are of special interest. They are rather long
and provided with wide lateral slits. Near the attachment of each tripod support
are two protuberances for eyes, the representation being further accentuated by
a circle of black paint surrounding a black spot. Between the eyes is the spool-
shaped figure, also in black, and completing the head symbol, as shown in previous
illustrations (see figs, a, c and d). The identity of the head motive, therefore, is
established beyond the shadow of a doubt. Below the head are three black
horizontal bands (the carapace motive) so that each tripod support is a life form,
evidently the armadillo.

Another interesting feature about this specimen is that, from the view-point of
the technique involved in the decoration, it serves as a connecting link between
the lost color group and the so-called alligator group that succeeds it. The lost
color technique is everywhere followed in ornamenting the body of the vessel,
as regards both exterior and interior. The interior, for instance, of shallow bowls
of this type is always painted red, while in the alligator group it is coated with
a light cream slip. The exterior was treated to the usual coat of non-adhesive
black over a waxed pattern — the very essence of the lost color process. On
the other hand, the design on the tripod supports is produced directly by the
application of black paint, the black being the design, instead of the interspace,
and also being a much more permanent black than that employed in the lost
color process. It is apparently the same black that is used as a delineating color
in the alligator group. This is another evidence of the homogeneity of Chiriquian
art, although the various groups may be quite distinct as a whole.

A number of small figurines, whistles, needle-cases and rattles, belonging to the
lost color group will be described in a subsequent chapter.

125

THE ALLIGATOR GROUP.

One of the most interesting classes of Chiriquian pottery is that appropriately
named by Holmes, the alligator group. In point of numbers, it is inferior to the
lost color group. It is like the latter in that a majority of the vessels may be
classed as bottle-shaped vases with globular bodies ; and in the comparative rarity
of features in relief, such as handles, legs and plastic shoulder decorations. The
average size is also about the same in the two groups. On the other hand, there
are striking differences to be noted. The line of demarcation between neck and
shoulder is not so sharp and the width of neck is proportionately greater. The
interiors are more carefully finished, although much inferior in this respect to the
armadillo group. The forms are usually pleasing and are perhaps more varied
than in the lost color group.

The chief distinction between the two groups lies in the processes of orna-
mentation in color; for, while telling its story, each depends primarily on color
to attract the attention. The lost color process has already been described. The
original ground was usually in two colors, a red pigment and a light cream slip
in alternating zones. Frequently the entire original ground was red, sometimes
light cream to salmon. The designs were always in one or both of these colors.
In the alligator group the designs are always on a pale yellow slip, which is
almost always applied to the entire surface. The lip is often red and in a
number of cases the same pigment covers the bottom, but designs are never
worked out upon the red surfaces. Both red and black are delineating colors,
the black playing perhaps the more important role. The black is a fast color in
comparison to the black of the lost color group.

The motives are either distinctly life forms or else derivatives that are traceable
to such forms. The alligator is the favorite subject : hence the name given to the
group. Among the bottle-shaped vases, the decorative motives are confined to
the shoulder and are not always set in panels. Frequently the shoulder zone is
divided into two or three arched panels, calling to mind the arched panels of the
lost color group. In the latter there are generally but two such panels ; on the
other hand, in the alligator group, the prevailing number is three. Again, the
motives are not always set in a delimited zone or in panels, but are often painted
on the shoulder without a setting.

The motives derived from the armadillo as well as from the fish, the frog, the
bird and the jaguar are almost always plastic. Plastic forms of the alligator are
not often met with except among gold ornaments. Painted forms of the alligator
and derivatives therefrom are on the contrary very abundant, especially in two
of the groups of pottery.

While these painted figures of some saurian were referred by Holmes to the
alligator, it is by no means certain whether the ancient Chiriquian artist may not
have had in mind the crocodile instead. Both are found in Chiriquian waters.
According to Sumichrast * the alligator (Alligator punctulatus) is nocturnal and
solitary, never inhabiting running or brackish water, but only the bottom of shallow

1 Quoted by A. Giinther in Biologia Centrali- Americana, reptilia and batrachia, 18, 1885-1902.

126

A STUDY OF CHIRIQUIAN ANTIQUITIES.

or muddy streams in plains or stagnant pools in woods. The habits of the cro-
codile are quite different. They frequently live in numerous herds and are found
in both salt and fresh waters — estuaries, rivers and lakes. Seemann, therefore,
probably meant the crocodile when he wrote : " Alligators are numerous on the
mouths of rivers, where they are found sunning themselves on the muddy banks."
Armand Reclus1 might have had the crocodile instead of the alligator in mind
when he said : " The caymans sleep with the mouth open, the upper jaw almost
vertical." Such scenes must have been familiar to the ancient Chiriquians, and
may account for the characteristic way in which the jaws were represented —
open mouth and upturned snout. The length of the jaw is often exaggerated,
as if the artist had in mind Crocodilus americanus, var. acutus, instead of the shorter
muzzled alligator. It was the crocodile also that figured as cipactli in the ancient
Mexican picture-writing. Notwithstanding all these reasons for believing the cro-
codile to be the saurian depicted, it does not seem best to change the name
chosen by Holmes. It is retained, therefore, with the proviso that it be given a
liberal interpretation.

A favorite representation is that showing the alligator in absolute profile (fig. 206).
This is one of the more realistic forms, but contains several of the elements that

have become independent decorative
motives ; hence it may be regarded
as a sort of key to much that follows.
Note especially the open mouth with
teeth, the prolonged and upturned
snout, and the dorsal markings on
the head, back and tail. The last
consists of groups of spots in a
triangular and a semicircular field,
and a series of spines. These
represent the body-markings of the
alligator (or crocodile, as the case
may be). As is well known, the
dorsal scutes or scales form longi-
tudinal series. The keels or spines
of the scutes may be easily seen
in profile and therefore give the
artists little trouble. The scales
are not so conspicuous. The diffi-
culty of reproducing them is happily
obviated by placing them in tri-
angular or semicircular fields that

stand out above, the dorsal line. I shall call these scale-group motives. The
other markings are of course spine motives.

A similar treatment of the alligator is noted in figure 207 — open mouth with
teeth, and upturned snout. Here however there are no spines on the back, the

Fig. 206. — Detail showing favorite J representation of alligator
in profile. '/>

Fig. 207. — Similar treatment of the alligator.

1 Panama et Darien, 149, 1881.

THE ALLIGATOR GROUP. 127

body-markings being represented by only one type of scale-group motive. This
is repeated three times on the head and five times on the tail.

A variation of the foregoing is seen in figure 208, which is a tracing from the
largest alligator vase in the collection, the gift of Mr. Edwin Lamson of Summit,
N. J. Here, the triangular scale-group motives on the head and neck rest on the

Fig. 208. — Alligator in profile with inverted scale-group motives and nuchal appendage. '/>

apex instead of on the base of the triangle. The scales on the body and tail
are grouped, but are not enclosed. The same motive is used for the teeth, the
latter being confined to the upper jaw only. This illustrates one of the processes
employed in conventionalism, viz., the substitution of one part for another. Atten-
tion is called to the appearance of a long crest that is attached to the back of
the neck, the meaning of which is not clear; also to the downward sweep of the
body curve, which is a marked feature of alligator representations in color.

There are a number of processes, taken either separately or in combination,
that may lead to conventionalism. Some of these are : (1) The reduplication,
exaggeration, elimination and fusion of parts ; (2) transposition, shifting and sub-
stitution : (3) isolation of parts, and their use independently of the whole ;
(4) wholesale reduction and simplification ; and (5) adapting the figure to fit a

Fig. 209. — Detail showing fusion of two alligators in profile.

given space. To give balance, for example, a second head may be added, as in
figure 209, which is in reality a fusion of two alligators, also carrying with it the
elimination of the tails and hindlegs. Elimination may arise simply from lack of
space, as exemplified in this case by the loss of the nuchal crest from one of the
necks. Triangular scale-group motives are recognized on the head and neck at
the left. Those on the right are also triangular, but stand on their apices and

128

A STUDY OF CHIRIQUIAN ANTIQUITIES.

are filled in solid instead of being open dotted fields. This type was noted in
the lost color ware (see PI. XXXI, fig. /).

The exaggeration of parts is well shown in figure 210, where the jaws, partic-
ularly the lower one, are exceptionally long. Both are recurved. The nuchal

Fig. 210. — Alligator profile showing exaggeration of jaws and nuchal appendage. '/•

crest also reaches its maximum, and the downward swing of the body-line is
typical. Another characteristic of the alligator profile is seen to good advantage
here, namely, the longitudinal rise of the fronto-nasal region.

That figure 211 came from a panel is evident from the disposition of parts.
The head is turned so as to fill the hollow of the body curve. The shortness
of the panel made it expedient, also, to flex the tail sharply and shorten the
nuchal crest. The top of the head is adorned with four typical spine motives.

The same relative position of head, body, legs and
tail is seen in figure 212, except that here the lower
jaw is longer than the body and tail combined and
the nuchal crest unduly developed, apparently as a
counterpoise to the extreme length of jaw. In figure213.
it is the body, legs and tail that suffer from reduction,
while the head and nuchal crest remain prominent.

Some of the vases from which several of the forego-
ing tracings came are illustrated in Plate XXXIII.
These vase forms are typical for the group, with the
possible exception of figure a, in which the setting of

the neck on the shoulder deviates from the type. The slip is a pronounced salmon
color. The unusually broad decorated zone is divided into four panels, two
long and two short. The long panels are each adorned with the figure of an alli-
gator. The two form a procession, i. e., each when turned toward the observer is
found to be facing toward the right. The framework is in red. The outlines and
superficial markings representing the scales, spines, toes, and even the teeth, are
in black. A short stroke of black also denotes the central part of the body and
the eye. Practically all the features common to the conventional representation of
the alligator are present here — the wide-open mouth, with pronounced jaws and
upturned snout ; the dotted protuberances (scales) on head and back ; the spines

Fig. 211. — Alligator profile adapted
to fit the space at hand. V'

THE ALLIGATOR GROUP.

129

on head, neck and tail : and the curve of the back emphasized by the uplifted
head and tail.

In both form and finish, the piece shown in figure b is more characteristic of
the alligator ware. The line of the shoulder passes without a break into the line
of the neck. The latter is slightly flaring and ends bluntly without any marked
lip differentiation. The entire outer surface, except the lip, is finished in a pale
yellow slip. There is no paneling about the two alligators on the shoulder, the
head of each reaching almost to
the tail of the other. Each figure
is in black and red, the outline
being a slender band of black and
the interior, red. The shape is
influenced by the contour of the
surface to be decorated, which in
a measure at least would account
for the upturned snout, the very
long lower jaw curving downward
and backward on itself, the curve
of the long tail, and the enormous

f 212— Profile of alligator with head turned backward, the
long lower jaw being balanced by the well developed nuchal

appendage.

'/•

development Of the Occipital Or

nuchal appendage — a sort of ex-
aggerated spine, with a smaller

spine in black near its base. The characteristic swaying downward of the back,
noted by Holmes, may also be traced to the same source. Details such as teeth
and toes are left out. The profile is, as usual, absolute. Very similar to the
preceding is figure c, reproducing a vase discolored by smoke. The shape of the
vase is practically the same, as is also the treatment of the alligator, except that
the secondary spine is missing and the hanging lower jaw curves upward into
the wide-open mouth, instead of downward.

The design on the shoulder of a double-necked vase from Bugavita (fig. d)
looks at first glance like a meaningless bundle of waving arms. The groundwork
is red, as usual, and the outlines are in black. The
whole is a conventional treatment of the alligator, the
neck being turned so as to project the exaggerated jaws
backward over the entire length of body and tail. The
lower jaw, which is turned downward at the end, follows
rather closely the curve of the back and tail. The snout
is upturned. There is a long curved nuchal appendage
acting as a balance to the prolonged mandible on the
left. The space between and below is amply filled by

a relatively small trunk, tail and legs. The alligator on the opposite side is re-
produced in figure 212.

The lines are much easier to follow in figure e, where the jaws are comparatively

short and straight. The nuchal appendage, though reduced in size, is still prom-

inent. Somewhat more sketchy is the representation of the alligator in figure /.

The head proper is the most prominent part, the red foundation enclosing a

MEMOIRS CONN. ACAD., Vol. 111. 17

Fig. 213. — Profile of alligator
in which only the head and
nuchal crest remain prom-
inent. '/'

130

A STUDY OF CHIRIQUIAN ANTIQUITIES.

triangular space in which the eye is set. With this head triangle as a center,
the boundary lines are simply continued to form the other parts. The legs are
eliminated. The body and tail are straight and much shorter than the nuchal
appendage. These are balanced on the right by an upturned snout and drooping
lower jaw.

A well-balanced but angular and stylistic representation of the alligator is shown
in figure 214. Conventionalism is also somewhat heightened by the breaks in
the black outline at the ends of the feet and the jaws. The balance would

Fig. 214. — Well-balanced but angular and stylistic
representation of the alligator. 2/'

Fig. 215. — Highly conventionalized two-headed al-
ligator. '/>

be complete if the tail were replaced by a head. This is what has been done
in figure 215, which is genetically related to the foregoing and which presents
a highly conventionalized two-headed alligator. The black outlines are still further
broken up and dots enough to answer for eyes as well as teeth are placed in
each of the wide-open mouths. Another double-headed alligator is seen in figure 216,

Fig. 216.

Fig. 216. — Double-head alligator, with nuchal crest on the

head at the left. '/>
Fig. 217. — Two-headed alligator showing transposition of

parts, both nuchal crests being attached to the same neck. '/«
Fig. 218. — Two-headed alligator with nuchal crests absent. '/»

Fig. 218.

the neck of the head on the left being supplied with a nuchal crest. Here the
dots representing eyes and teeth are placed where space invites, rather than
where they belong.

A good example of the transposition of parts is given in figure 217, both nuchal
crests being placed on the same neck in order to avoid interference and to har-
monize with the paired jaws at each end of the motive. The outlining black
color is entirely wanting. In the next illustration (fig. 218) both nuchal crests
are lacking, but the setting of black reappears. The teeth in each mouth are in-
dicated in a highly diagrammatic fashion.

THE ALLIGATOR GROUP.

131

A series of motives, each representig the double-headed alligator, is seen in
figure 219. The unmistakable elements are : the downward swing of the body-
line, which is angular instead of curved : scale motives in the first two ; and
spine motives in all, those in a being particularly characteristic. These four tracings
are from one vase, filling a narrow zone on its shoulder.

a ted

Fig. 2 19 a— d. — Series of motives, each representing the double-headed alligator. V*

In figure 220, there is a separate body-curve for each alligator; in the hollow
of each curve is a dash of paint to represent the body-markings, distinct scale
motives being attached to two of the three. The design in figure 221 is similar
but more simplified. The four strokes of the brush at each bend in the compound
body-line seem to suggest the union of four separate alligator motives, just as
the preceding figure suggests three.

Fig. 220. — Multiple alligator motive. 'I*

Fig. 22:. — Simplified multiple alligator
motive. '/'

Figure 222 probably represents two body-lines combined, each being accompanied
by a single scale symbol. The body-curves of two alligators are often joined in
such a manner as to form a sigmoid curve, as in figure 223, a simple type in
which the scale motives are left out.

In figure 224 there is a simple horizontal body-curve, with a distinct differentia-
tion of head and tail, and with a spot in the hollow of the curve to indicate the
dermal markings — in other words, the alligator is in profile. This is an excellent
example of conventionalism by means of a wholesale reduction and simplification
of parts — conventionalism that has reached the hieroglyphic stage. How easy
it would have been to make use of this readily executed symbol in a system of

Fig. 225.

Fig. 222.

Fig. 223.

Fig. 224.

Fig. 222. — Detail probably representing two body-lines, each with a single scale symbol. V"

Fig. 223. — Sigmoid curve representing two body-curves. */'

Fig. 224. — Conventionalized alligator in profile ; hieroglyphic stage. !'>

Fig. 225. — Wholesale reduction and simplification of the alligator in profile. 'I'

writing as did the ancient Egyptians, for example. It could have stood for the
word alligator or for some attribute of that animal. There is however no evidence
that the ancient Chiriquians made any such use of it. They were content ap-
parently to employ this symbol in a decorative and ceremonial sense only. The
spine motive and the scale-group motive, both of which have already been described.

132 A STUDY OF CHIRIQUIAN ANTIQUITIES.

could also have been employed as hieroglyphs combined with or in place of the
motive representing the whole animal.

Wholesale reduction and simplification of the profile view of the alligator is
seen in figure 225 : nothing being left but the body-line, which is produced by a
short stroke of the brush. This form of the alligator motive usually occurs in
series, the alternate units being inverted.

Familiarity with the foregoing illustrations from tracings will make it possible
to follow the rapidly progressive steps in the process of conventionalization
portrayed in Plate XXXIV. Both duplication and elimination are noted in
figure a. There is a head, for example, at either end of the common body,
and, for lack of space, only one is provided with a nuchal appendage. The
eyes and teeth are placed in an angle below each lower jaw. The space on the
right, being larger, is filled in with three black spots instead of one. The
black outlines are broken at all the ends of the red framework, while in places
they are doubled. The design on the opposite side is practically the same except
that the nuchal appendage is more conspicuous. A similar conception of the
alligator motive may be seen in figure i, except that it is treated in a still more
summary fashion. The nuchal appendage has disappeared entirely, but the curve
of the back that is common to both heads is quite characteristic. This small
vase is slightly angular at the periphery.

The vase reproduced in figure c is divided into zones. The lower is red. The
upper is outlined by two red bands. Within these are two black bands, which
in turn limit the field containing the symbols of life forms, the latter also in black
on a pale yellow paste. The neck is unadorned, while the lip and inner surface
of the aperture are finished in a red pigment. The characteristic curve of the back
is to be noted in the alligator motives. Fore- and hind-feet point in the same
direction ; otherwise it would be impossible to distinguish the head from the tail.
In one instance, the body-line of the alligator takes the form of a simple scroll.
Two dorsal-view motives serve to divide this narrow zone into two panels.

An interesting vase from Divala is shown in figure d. The lower zone is red.
The black and red bands and the alligator motives (in black) of the upper zone
are on a pale yellow slip, which is continuous over the neck. The lip and inner
surface of the aperture are finished in red. A series of alligator motives, all
essentially alike, but illustrating progressive reduction and elimination of parts
is carried around the shoulder in a continuous panel (see fig. 219).

In figure e, the horizontal plane of greatest diameter is situated much nearer
the neck than is usual for this group. The broad lower zone is red. The black
and red bands of the upper zone enclose two panels. Each of these is filled by
a row of alligator symbols in black. The ends of the body-line are folded back
until they almost meet. Within are spots (two to four) representing the dermal
markings. The alternate motives are inverted. This is also true of figure /, where
simplification and general reduction reach their limit, the abbreviated body not
being accompanied by markings of any kind.

The significance of some of the zonal shoulder decorations in Plate XXXV is
more or less problematical. The alligator motive is recognizable in figure a.
Two or three body-lines are united, the free ends being treated as heads. Sec-

THE ALLIGATOR GROUP.

133

ondary curved lines fill in the remaining spaces between the upper and lower
boundary lines. This design is repeated three times with but slight variations.
A similar motive is seen in figure 6, where there is an additional body-curve in
each group, and the design itself is simplified.

The elaborate motive on the shoulder of the vase reproduced in figure c is
referable to the alligator, the spine symbols being easily recognizable. It is
repeated on the opposite side with some modification (fig. 226 a). Alternating
with these are much smaller and simpler alligator motives (fig. 2266 and c). The

a be

Fig. 226 a— c. — Conventionalized alligator designs. ''«

narrow shoulder zone in figure d is divided into two panels, each being ornamented
with the dorsal-view motive — three parallel lines, the outer ones bearing spines along
their external margins. The decorative motive in figure e may be a variant of the
same thing. The shoulder of one vase (fig. _/) is decorated with concentric rings.
These are variously connected so as to form pairs ; in one instance there is a single
group, attached to the outer ring of which are two spirals obviously representing the
second group. The relationship between spirals and concentric rings in Egyptian
art has been pointed out by Goodyear.1

Kig. 227. — Vase whose shoulder zone includes six
arched panels filled with scale symbols ; from
Divala. Alligator ware, '/a

Fig. 228. — Vase whose three arched panels arc asso-
ciated with scale motives. Alligator ware. '/»

It has been noted that the curved band representing the body-line of the al-
ligator is usually accompanied by spots, used as scale symbols. This combination
is seen in figure 227, a vase from Divala. The narrow shoulder zone includes six
small arched panels, each filled with spots. The alternating spaces are undecorated.
A similar motive is seen in other specimens, where the arched panels are

1 Wm. H. Goodyear. Grammar of the lotus, PI. VIII. London, 1891.

134

A STUDY OF CHIRIQU1AN ANTIQUITIES.

reduced to three, alternating with wider intervening open spaces. The same
though somewhat more elaborate motive is reproduced in figure 228. Here the
intimate association of the scale symbols with the curved bands representing the body
is suggestive. There is plenty of room for the spots in the center of these enclosed
fields, but they cling everywhere so closely to the boundaries as to appear like half-disks.
Three arched panels with decorated intervening areas also occur in figure a of
Plate XXXVI, the design in each arched panel being the dorsal-view motive.
Freedom in the treatment of the arched panel runs through the entire series
represented in this plate. The arched panels are quite large in figure 6, the spaces
between them and surrounding the neck being in the color of the slip. Each panel
enclosed two groups of concentric triangles, the angles of the inner ones being
marked with dotted dentals ; the whole panel decoration is a variation of the dorsal-
view motive. The suggestive association of spots and crossed zigzag bands is shown
in figure c, each band being a multiple body-line accompanied by dermal markings.
There is a multiplication of the arched bands tangent to the neck in figure d, a

Fig. 229. — Vase having but two arched panels in which
scale- and spine-motives are grouped ; from Divala.
Alligator ware. *'•

Fig. 230. — Double-necked vase with the dorsal-view
motive as a panel decoration ; from Bugavita. Al-
ligator ware. '/•

vase from Divala. What remains of the enclosed field is crossed by six short
vertical bands, the outer ones bearing lateral spines — a typical example of the
dorsal-view motive. This vase and the three preceding (a, b and c) are highly
characteristic of the alligator ware in regard to paste, slip, form and decoration.
The two vases (both from Divala) represented in figures e and / are perhaps not
so typical, except for the character of the black and red paint used and the treat-
ment of the arched panels. The slip is thinner and of a duller hue. Both the
neck and the lower zone are treated to a coat of red, leaving the slip to show
only on the shoulder. The short vertical bands crossing each panel in figure e
are accompanied by dotted dentals (the scale-group motive). This combination
represents a section of the alligator's body viewed from above, as first noted in
the lost color ware (see PI. XXXI). The panel decorations in figures a, b. f and
e are variants of the same motive.

THE ALLIGATOR GROUP.

135

The number of arched panels is reduced to two in figure 229, reproducing a
vase from Divala. The three divisions of each panel are ornamented with spine-
and scale-symbols. One of the two characteristic types of dorsal-view motives
is seen in the two panel decorations on the shoulder of a double-necked vase
from Bugavita (fig. 230) ; the other type is the design on the shoulder zone
in figure 231, which alternates with quadrangular panels in red. The decorative

Fig. 231. — Vase in which the dorsal-view motive
alternates with quadrangular panels. Alligator
ware. '/>

Fig. 232. — Vase with alligator motives on lip and
shoulder. Alligator ware. 'I*

Fig. 233. — Vase in which the alligator motives are
obscure. Alligator ware, 'h

Fig. 234. — Vase with obscure alligator motives on
lip and shoulder. Alligator ware. ''«

motives running through a series of small vases (figs. 232-236) consist of the
diagrammatic use of the elements of these two types, as well as the profile view
of the body-line, either compound (fig. 235) or simple (fig. 236). The latter motive
is often repeated three times on the lip (figs. 232, 235, 236).

There is a series of vases belonging to the alligator ware in which the decora-
tion consists of four rosettes painted on the shoulder and unaccompanied by arched

136

A STUDY OF CHIRIQUIAN ANTIQUITIES.

panels or horizontal bands (PL XXXVII, with the exception of fig. a). The
rosettes are usually composed of triangles, spots, and a cross, in varying combination,
all of which may be referred to elements of a life form, presumably the alligator.
The painted designs are placed on slightly raised nodes apparently produced by
pressure from within of the middle finger tip.

In a small vase from Bugativa (fig. i), the outer circle of the design is red ;
the inner, black. Within are two spotted triangular fields, separated by a spool-
shaped space in red. On the lip, directly above each rosette, is an alligator
motive consisting of the body device of curved and red bands accompanied by spots
on the concave side only. This style of lip decoration is confined to the smaller
vases. The rosettes of figure c are each bounded by three circles, the middle
one only being red. The field within is almost completely filled by three triangles

Fig. 235. — Vase with alligator motives on lip and
multiple body-line as a panel decoration. Alli-
gator ware. */»

Fig. 236. — Vase with simple body-line motives on
shoulder and lip. Alligator ware. !/»

bearing spots (the scale-group symbol). The triangles are disengaged from the
inner circle in figure d, and the interspace is marked by a cross. The nodes are
scarcely perceptible in figure e, and the number of concentric rings is reduced
to two, the inner one enclosing a cross surrounded by four triangles. The nodes
are very much exaggerated in figure /, and are not produced by pressure from
within, but are applied to the outer surface. The inner circle is again marked
by a cross and the intervening spaces by triangles. The various stages in the
transition from the rosette to the cross are represented in the collection. In one
vase the cross alone remains, a typical Greek cross with foundation in red and
outlines in black. The same disposition of the two colors has already been noted
in the more elaborate representations of the alligator.

Plastic features applied to the shoulders of vases belonging to the alligator
group are comparatively rare. A representative series is given in Plate XXXVIII.
Painted ornament accompanies the relief, but the two do not necessarily refer to
the same animal. The plastic elements in figure a, a vase from Divala, consist
of a head frankly human, balanced on the opposite side by a short upturned
animal tail. Alternating with these features on either side is the painted representa-

THE ALLIGATOR GROUP. 137

tion of an alligator in black and red, with its characteristic hooked snout, bunched
dermal markings and pronounced body-curve ; toes, teeth and eye are all indicated.
The recurved lip and angular outlines of this vase are unusual for this group.

The alligator type of ware is followed closely in the outlines of figure b. The
two relief ornaments are placed on a level with the union of neck and shoulder.
Both are marked by two cross bands of black and are alike in shape and size,
probably representing two tails instead of two heads. Alternating with these, but
lower on the shoulder, are two reptilian forms, probably the alligator, executed in red
and black. The treatment is unique, the view being dorsal or ventral instead of the
lateral aspect or profile. All four legs are represented, but there is some doubt
as to which is the head and which the tail. The designs on the opposite sides
are practically identical. The painting was done on a polished yellow surface.
Paste of a lighter color was afterwards spread over the entire surface of the vase,
except these painted portions.

The two plastic heads are unmistakable in figure c. The outlines of the alli-
gator painted on the shoulder of the vase are angular, with the exception of the

Fig- 237.— Vase with shoulder panels showing Fig. 238.— Vase with small vertical loop-handles

obscure multiple body-line, accompanied by and the alligator motive as a shoulder decoration,

body-markings. Alligator ware. '/» Alligator ware. '/>

dorsal curve. Another step in the diagrammatic treatment of the alligator form, a
common body with a head at each end, is shown in figure «!, where the black con-
tour lines are broken at the angles. A similar angular motive consisting only of
the red foundation occurs in figure e. The two prominent nuchal appendages are
both placed on the same neck so as to balance the pair of jaws at either end of
the figure. Scale- and spine-symbols are wanting. The animal head and tail, as
relief features, are reduced to mere nodes in figure /. Reaching from head to tail
on either side, is a panel that contains a multiple body-line in red. Accompanying
this line are the usual dermal markings in the hollow of each body-curve.

The identity of the body-line is somewhat obscure in figure 237. It takes a
zigzag course through the center of the shoulder panel, is broken at the angles
and everywhere accompanied by dotted dentals representing body-markings.
Vertical loop-handles, as seen here, are • of rare occurrence in the alligator group.
In figure 238 the two small handles represent an animal head and neck. The
MEMOIRS CONN. ACAD., Vol. III. 18

138

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Fig 239. — Vase without neck, the relief ornaments being
two opposed animal heads. Alligator ware. '/«

nose is slightly upturned and teeth are visible in the open mouth. The two
panels on the shoulder of the vase are filled in with scale-group symbols ' seg-
ments of circles enclosing dots.' Above each panel is a row of similar markings.
There are a few vessels without necks. Such, for example, is the one rep-
resented in figure 239. Two animal heads on opposite sides constitute the relief
ornament. The rows of dots in the small open fields of the alternating panels
are all that is left to suggest the alligator. In another vessel of similar type (fig. 240),

the panels are arched and between them
appear in relief the head and tail of
some reptilian form, the tail being
turned to one side.

Vessels of this group mounted as
tripods are comparatively rare. The
tripod supports are usually short, solid,
pointed cones that are marked j)y
horizontal black bands, as seen in
Plate XXXIX, where the variations in
the form of the body are also fairly
well indicated. In figure a, a vase from
El Banco, the neck is quite short. The
shoulder decoration consists of three
panels, each bearing a series of scale-
group symbols. Each of the two ho-
rizontal panels in figure b are double.
The four smaller black panels in a
framework of red are all treated in prac-
tically the same manner — a triangular
(or trapezoidal) space in the middle and
a trapezoid at either end, the three open
fields thus formed being decorated with
scale- and spine-symbols.

Only spine symbols decorate the
neck of a wide-mouthed tripod vase
from Jacu (fig. c). These are sharply
recurved and distributed in pairs, rising from the upper one of three black
bands that surround the neck. Between the second and third band is a broad
red band. Even less easily interpreted are the spine symbols in figure d, a shal-
low tripod bowl from Divala. The visible interiors of shallow bowls are usually
painted, this one being an exception. It is somewhat more highly polished
than the exterior, and the slip, if used at all, must have been very thin. The rim
is painted red. Below it, on the outer surface, are two black bands. Next comes
a red band, hanging from which is a series of spine symbols in black. The latter
would be inexplicable but for the use of the same motive in other connections.
The legs are banded, as usual, and probably represent life forms.

The bowl of a tripod from Divala (fig: e) is not so shallow as the preceding,
and the rim, being gently incurved, does not mark the level of the greatest hori-

Fig. 240. — Necklessvase with dorsal-view alligator motive
as a panel decoration. Alligator ware. */'

THE ALLIGATOR GROUP. 139

zontal diameter. Nevertheless the interior is painted red. The lip and outer sur-
face, except the delineating colors, are finished in a thick coat of cream slip. The
zone between the rim and the base of the supports is banded horizontally and
divided into three panels, each of which contains scroll patterns, whole as well
as broken. In figure / we have once more the attractive shoulder decoration
consisting of three arched panels (see PI. XXXVI). The well-preserved red and
black delineating colors have an effective background in the generous use of a
pale yellow slip.

A number of the tripods combine plastic and relief decoration, as may be seen
in Plate XL. The head and tail are applied to the shoulder of the vase without
affecting the general globular form of the body, which is nevertheless intended
to take the place of the animal body. The identity of the animal representation
in the round can not always be determined with certainty. It may or may not
refer to the same animal as the accompanying forms executed in color. This is
probably the case in figure a. The life form painted on either side is the alligator,
with the characteristic attitude of the jaws, nuchal appendage and downward curve
of the body.

Somewhat similar in general outline as well as decoration is the tripod vase
represented in figure b. The life form in the round may be the alligator, partic-
ularly since there are two painted scale symbols on the back of the head. This
view is strengthened by the fact that the two alligators painted on the sides are
oriented in such a manner as to correspond to the form in relief, i. e., all three heads
are pointing in the same direction. The tendency of the artist if not restrained by
other considerations is to paint an animal form on one side, and after revolving
the vase until the opposite side appears, paint the second figure exactly as he
did the first. The two thus form a procession. This is true in almost every in-
stance, but is not in case of this specimen where the two painted life forms, if
they could move forward would meet each other, instead of falling one behind
the other. The typical features of the alligator are emphasized in both.

There is a peculiar combination of relief and painted decoration in figure c.
The relief portions of this vase consist of a simplified head with longitudinal black
bands ; eyes (or breasts) set far apart and surrounded by black circles ; long
curved fillets of clay representing eyebrows, or arms, edged with black bands and
also marked by cross bands; and, on the opposite side, a similar fillet to represent
the tail. This is straight, being carried from the neck of the vase downward to
the level of its greatest horizontal circumference. The head and eyes in relief
are supplemented by painted representations of toothed jaws on either side of
a median plane. The throat is painted red ; which color is also carried downward
on the shoulder of the vase like a bib or apron. Between the tail and the curved
eyebrow (or arm) on each side are panels, both containing representations in
black of the alligator. One of these (fig. 241 a) is unmistakable, with its upturned
snout and long nuchal appendage which, like the head, back and tail, carries a
series of scale symbols. The lower jaw serves as a foreleg. The other (fig. 241 b')
could not be recognized but for its association with this particular class of ware.

Figure d presents no original features. The sculptured head and tail (the latter
bent sidewise) are not distinctive. The two panels are decorated with scale-group

140 A STUDY OF CHIRIQUIAN ANTIQUITIES.

symbols. The smallest vessel of this group (tig. e] is but 45 millimeters in height,
with a capacity of less than 20 cubic centimeters. A compound tripod with two
of the six legs eliminated is shown in figure /. There is no connection between
the interior of the two vessels. The most interesting part of their decoration con-
sists of a series of alligator motives about the two rims — the downward curving
body-line with a single dot above.

A number of unusual forms are brought together in Plate XLI. Figure a comes
very near to being a complete zoomorphic unit. It is a tripod instead of a
quadruped, with two legs in front and but one leg behind, which is in a median
plane. The tail forms a vertically placed loop-handle, emerging from underneath
the carapace to fuse with the hindleg. The legs are marked with parallel trans-
verse lines in black, suggesting the carapace motive, in which case each leg
would represent an armadillo. The head is well defined, the nose being pointed
downward as if to reach the ground. The ears are represented in the round as
well as by means of black paint ; the eyes, by paint alone — a black circle on
cream slip surrounding a black spot. The carapace, like the ears, is represented

a b

Figure 241 a, b. — Designs representing conventionalized alligators. '/'

both in the round and in color. Its anterior and posterior regions are marked
by black spots on the cream slip. The banded character of the middle region is
indicated by transverse parallel lines in alternating groups of black and red. The
circular aperture cuts all three regions of the carapace ; its margin or rim is
painted red. The black spots on top of the head also represent dermal armor.
The vessel is covered inside and outside by a uniform slip over which the delin-
eating colors are applied, except on the interior and the outer ventral surface.
The alligator motive is entirely wanting.

There is nothing to suggest the alligator in figure b, unless it is the dotted
dentals on the tripod supports. The latter represent animal heads, are hollow and
supplied with pellets. The periphery of the vase is decorated with eight prominent
crablike spines and what is probably intended for a head.

A broken vase from San Carlos, eighteen miles northwest of David, is repro-
duced in figure c. With the broken part has disappeared the tail and body of
a large painting of an alligator that occupied the rear panel, leaving only the
head and nuchal appendage, with spine symbols. A fragment carrying a short
tail in the round to match the head opposite has been preserved also among the
fragments. In front is an apron-shaped field covered with ocellated markings
resembling the coat of the jaguar. At the top, this apron is cut so as to fit
about the neck of an animal, probably the jaguar, with wide-open mouth showing

THE ALLIGATOR GROUP. 141

canine teeth, and at the sides it is so cut as to leave free the arms that
are set akimbo. The whole evidently represents the jaguar-man or, to be more
exact, the jaguar-god (see also fig. 244). The two tripod supports in front are
alike, but differ in shape from the one behind. A similar somewhat smaller spec-
imen from Mercedes, Costa Rica, belongs to Mr. Minor C. Keith.

Another vase not at all typical of this group is reproduced in figure d. The
life features in the round consist of a head with wide-open mouth, probably the
jaguar : below which is a semicircular apron-like pattern, a drooping tail and rather
long arms that hang free from the body, each hand resting on the base of a
tripod support. One arm is lost. Much of the paint has disappeared ; the highly
original design on the neck however is still practically complete.

A thick-walled and thick-lipped bowl from Jacu (fig. e) is provided with an
annular base- — the only example of this type in the group. The base and the
interior are painted red. The lip and outer surface of the body are finished in a
pale yellow slip. A series of dotted triangles, or scale-group symbols, in black
encircles the vessel.

The form reproduced in figure / is unique. A small shallow cup with vertical
walls is supported by three grotesque half-human forms, whose feet rest on a
ring-shaped base and whose extended arms help to bear the weight of the cup.
Three panels decorated with alligator motives alternate with the heads of the
supporting figures. The kinship between this form and the so-called stools made
of stone (see PL IV), as well as modeled in clay (see PL XL VI), is striking, the
main difference being that the cup is replaced by a slightly concave seat plate.

Another conception of the use of engaged, plastic life forms as supports for
vessels is illustrated in Plate XLII (fig. c). This vase is one of the gems of the
de Zeltner collection and was given to him by Senor Obaldia, formerly Vice-Presi-
dent of New Granada. It was first inadequately figured and described by de Zeltner l
and later mentioned by Holmes when describing a piece somewhat similar, though
neither so large nor so perfect. The Yale specimen is 39.3 centimeters long,
28.5 centimeters in height and 18.5 centimeters in breadth.

The bottom is perfectly flat, in outline like an ellipse flattened a little at the
sides. The latter rise almost vertically to meet the flattened shoulders at an angle.
The neck is long, round in section and gracefully flaring. The supporting forms
at either end are apparently human and of the female sex. Their bodies are an
integral part of the body of the vase, the wall being pushed out to form a prom-
inent abdominal protuberance. The head, rising from the angle at the shoulder
and gently inclined forward, is crowned and hollow, the long transverse slit at
the top communicating with the interior of the vase. The features are all ren-
dered plastically and emphasized by paint. The outline of the rim of the ear is
characteristically human. The bulging eyes are protected by prominent brows.
The nose resembles the beak of a bird, which was the Chiriquian's favorite model
for the human nose, as noted on a preceding page. Some of the features are
seen to better advantage in figure 242. The parted lips, oval in outline, reveal
three long incisor teeth. The feet are three-toed and in the sole of each there

1 Note sur les indiens du departement de Chiriqui, 8, 1866.

142

A STUDY OF CHIRIQUIAN ANTIQUITIES.

is a round perforation communicating with the interior of the hollow leg. Atten-
tion is called to the painted, tongue-shaped breast ornament, reaching down well
over the abdominal protuberance. A similar ornament is noted elsewhere in this
group (see fig. 244).

The long panels on either side are ornamented
with a superb example of the two-headed al-
ligator. The dermal markings on one head and
neck being the same as a decorative motive
on the crown of each female head (fig. 242).
The neck of the vessel, resembling in shape
an inverted bell, is surrounded by four panels,
each filled with an excellent figure of the al-
ligator, one of which is reproduced in figure 243.
The execution of the painted designs is equal
to the skill displayed in the modeling.

The rare specimen reproduced in figure 244
is a fantastic combination of feline, serpentine
and human characters, and serves as a rattle
or a receptacle at will. It is complete barring
a small chip off one heel, revealing the nature
and color of the paste, which is red throughout,
the hue deepening from the surface inward.
The slip is the usual pale yellow.

The body, legs and arms are human, excepting
the hands, which are converted into serpent heads
by means of a terminal slit for the mouth and
black spots for eyes. The feet are nondescript,
the prolongation of the heels being equal to that
of the. toes. The same type of foot is used
for birds, which makes possible the erect posture
for biped figurines. For that reason this par-
ticular specimen was catalogued by de Zeltner
as figure debout simply. But it is in more stable
equilibrium when lying on its back, in which
position the greatly distended stomach and chest,
in the shape of a hollow, truncated cone, serves
as a vessel.

The head is apparently that of the jaguar,
Felis onca; or perhaps the ocelot (Mex. ocelotlj,
Felis pardalis. The mouth is open, the lips

Fig. 242.

Fig- 243'

Fig. 242. — Elaborate vase supported by human
figures ; end view (see Plate XLII, fig. c).
Alligator ware. '/>

Fig. 243. — One of the alligator figures dec-
orating the neck of the preceding vase. */>

being drawn back to reveal the teeth. The
upper and lower rows are separated by hori-
zontal openings in the region of the molars and of the incisors. The overlapping
of the long canines is admirably indicated by two teeth on either side that reach
from the upper to the lower alveoles. Perforations representing the nostrils and
the external auditory openings also communicate with the hollow interior of the

THE ALLIGATOR GROUP.

143

head, which is supplied with pellets. The head therefore functions as a rattle.
The neck is perforated transversely as if for suspension.

The peculiar throat and chest decoration in red, and reaching down to the
level of the mammae, has already been noted (see fig. 242). It may mean an
element of ceremonial apparel. This is probably the jaguar-god (see also PI. XLI,
fig. c). The now familiar scale-group motive of dotted triangles, and also me-
anders accompanied by dots, is distributed over the various parts. The decorator

Fig. 244 a, b. —Jaguar-god ornamented with alligator motives and serving both as rattle and receptacle.
Alligator ware. V"

may or may not have had the alligator in mind when finishing this particular
piece. The origin of the motive however is beyond question, whether the artist
was conscious of it or not.

The happy fusion of vase and animal form is shown in Plate XLIII (figs, a
and b). The legs differ in no way from the simple tripod support, except that
they are four in number. This is more a concession to the maintenance of the
zoomorphic idea than to any thought of achieving a satisfactory state of equi-
librium. To the primitive workman, the three-point support has much in its favor,
no delicate adjustment of leg length being necessary. All three feet touch the
supporting surface immediately and firmly, whether it be plane or undulating.
On the other hand, when four legs are employed, one is almost always worse
than useless because too short to take its portion of the weight, which but for

144

A STUDY OF CHIRIQUIAN ANTIQUITIES.

the presence of the worthless member might have been easily distributed among
the other three. This probably accounts for the prevalence of the tripod.

The globular body of the vase becomes the trunk of a jaguar by the addition
of the head and tail of that animal. They are both hollow, thus adding much
to the capacity of the vessel as a whole. The external contour lines are exceed-
ingly graphic. The surface of the interior is everywhere smoothed down with
care, especial attention being given to contact lines at the base of neck and of
tail. The features of the head are in relief, which is strengthened by the judicious
use of paint. The thick lips are parted, both rows of teeth being visible. The
four overlapping canines seem to reach from the upper to the lower jaw as in
other representations of the dentition of the jaguar (PI. XLI, fig. c; also text-fig.
244). The fields of black spots about the eyes and on the throat evidently have
reference to the markings of the jaguar's skin. This is not true however of the
remarkable series of patterns that cover the top of the head, the neck, sides, and
tail to its tip, except the under surface, which like the belly is unmarked by
delineating colors.

There are 89 panels, no two being alike in size and shape. They approximate
squares, rectangles, trapezoids, trapezia and triangles. The outlines of the pattern
follow those of the containing panel. The markings of the inner field include
lines that meet or cross each other, sometimes forming simple or compressed
meanders. These lines and meanders are always accompanied by dots on one
or both sides. In a few small panels, dots only are to be found. There are no
circles or dotted circles. In other words, these panel patterns have no reference
to the markings on the jaguar's skin. They are repetitions, with variations, of the
well-known alligator motive. The panels themselves may be a concession to the
ocellated patches on the jaguar's coat, but there the comparison ends.

In an outline drawing, Holmes l figures a vase of the alligator group, with
shoulder zone divided into panels that enclose somewhat similar patterns, except
that dots are used sparingly and in a few instances the characteristic plumelike
spine symbol is recognized. These devices originated in the alligator motive.
Holmes did not refer them to the alligator, although he believed them to have
" features suggesting a pictorial original and doubtless derived from one." The
same writer raises the question whether these ancient peoples might not have
been " supplied, through the conventionalizing agencies of the art. with devices
that could have been employed as ideograms and letters."

A dorsal view (fig. b) of this interesting piece impressses one not only with
the variety of the patterns, but also with the skill displayed in the arrangement
of the panels. Luckily for the artistic effect of the whole, bilateral symmetry was
not thought of or else was studiously avoided.

The bowl of the vessel is completely lost in the body of the animal, when we
come to such examples as the one reproduced in figure c, where only the neck
and rim emerge above the animal's back. The tapir is chosen for representation,
probably Tapirus bairdii, since this species is known to exist in Panama. The
specimen is carefully finished. The short tail and legs, as well as the long pro-

1 Op. cit, fig. 284 (see also fig. 285).

THE ALLIGATOR GROUP.

145

boscis, are characteristic. The mouth is open enough to reveal two rows of short
teeth, all of the same pattern. Perforations for the external auditory openings
communicate with the hollow interior of the head. The neck is solid. The use
of the delineating colors is out of the ordinary, black instead of red being the
more freely employed, especially for the broader bands and areas. The painted
designs have nothing to do with the skin markings of the tapir. The origin of
the motives is traceable to the alligator, although the choice of them in this in-
stance may have been without any special significance other than decorative.
There is no decoration on the ventral surface, from the chin to the end of the tail.

The insignificant tail and
projecting buttocks of the
tapir are well characterized
in figure 245. The end of
the proboscis has been
broken off. The hollow
head and neck add con-
siderably to the capacity
of the vessel, the rim of
which rises from the back
of the animal figure. The
free use of black, as in
the preceding example, is
to be noted in the decora-
tion. Longitudinal panels
reaching from the head to
the tail enclose the alli-
gator motives. Attached
to this specimen was a note in the handwriting of the collector as follows : -

Fig. 245. — Zoomorphic vessel representing the tapir and ornamented with
alligator motives. Alligator ware. '/»

Jan. 1878. ' Vivala! This huacal (waucal) seems to be continuous for miles along river
of same name. Graves vary from 3 to 12 feet deep, a very few having gold. Many
contain nothing desirable; others, 1 to 15 or 16 pieces, mainly pottery. J. A. McNiel.

The Yale collection includes a number of rare bird forms that present some
very interesting characters. As far as I have been able to ascertain, complete
examples of this kind are not to be found in any other collection. The largest
of these (fig. 246) is from Divala. It is first of all a vessel, the globular form of
the body not being even masked by the addition of ornithomorphic characters.
Instead of two legs, there is a low annular support approximating in size that of the
circular rim at the mouth opening, the latter being gently flaring and painted red.

The bird's head, which unfortunately is broken off, was placed rather low on
the shoulder at a short distance above the plane of greatest horizontal diameter of
the vase. Its shape and size can not be determined beyond the certainty that it
could not well have been relatively large without a greater distention of the sur-
face contour of the body at its base. The wings and tail are in nearly the same
horizontal plane as was the head, the wing attachment reaching from near the
MEMOIRS CONN. ACAD., Vol. III. 19

146

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Fig. 246.

head almost to the root of the tail. The length of the short tail is the same as

the spread of the wings.

The base of the missing head is surrounded by two black circles that are

cut at the top by the lower of
two horizontal black circles.
Below these are black and red
zonal bands, interrupted only
by the head space. The tail
and wing feathers are indicated
by black parallel bands that
are partially lost, owing to
the crumbling away of the
underlying slip. The breast
is decorated with an original
design in black that evinces
brush-work skill of a rather
high order. The walls of the
vase are thick and the rela-
tively fine-grained, rich red
paste is tenacious. An at-
tempt to mend the break that
caused the loss of the bird's
head was made by using a
black substance that may
possibly be the same material
as the body of the black paint
employed in one of the de-
lineating colors. On analysis,
the substance proves to be
pine pitch, which might well
be one, at least, of the ingre-
dients in the black paint be-
cause of its adhesive qualities
as well as of its color. Other
possible blacks are oxide of
manganese, charcoal and soot.
According to Mr. C. V. Hart-
man,1 the native women of
Izalco, Salvador, decorate
their calabash vessels with a
paint, the black color of which
is derived from powdered
charcoal.
The bird form is somewhat more pronounced in figure 247, although the ad-

Fig. 247.

Fig. 246. — Globular vessel representing a bird, a low annular support
taking the place of feet ; from Divala. Alligator ware. '/'

Fig. 247. — Ornithomorphic vessel decorated with scale-group and spine
motives. Alligator ware. '/«

1 Die Baumkalebasse im tropischen Amerika. Boas anniversary volume, 203, 1906.

THE ALLIGATOR GROUP.

147

dition of avian characters, such as two feet, wings and horizontal spreading tail,
does not alter the globular form of the vessel. The legs are quite short and so
placed as to combine with the relatively large flat-bottomed feet in giving to the
figure a certain degree of equilibrium. The feet are spread equally at the heel and
toe, the number of the toes not being indicated. This is the typical bipedal form of
foot and is found also in human figurines. The spread of the tail is about equal
to that of the wings. Both are more highly differentiated than in the preceding
vase, and are directed upward as if in flight. The wings are only partially ex-
tended as if the bird were in the act of alighting, and the backward pointing tips
are truncate. The head, for-
tunately preserved in this case,
is a mere lump of clay situated
on the periphery and propor-
tionately smaller than the head
in the foregoing example.

The neck of the vessel is
much more constricted than in
the former example. Its walls
are vertical and cut square at the
top. The rim is painted red
and there are two black hor-
izontal bands about the neck.
Had it been left undecorated
and had there been no other
provision for a head, one might
be led to suppose that the
latter was a separate piece made
to fit over the neck. But such
was not the case. The whole
is complete as it stands. On
the shoulder is a narrow zone
decorated with scale-group (al-
ligator) motives. A black paint
was smeared on the upper sur-
face of wings and tail. The
head protuberance is painted

red and surrounded by a black band, outside of which is an irregular circle of
scale- and spine-symbols. The short legs are marked on the outside by two hor-
izontal black bands. The rest of the outer surface is finished in a cream-colored
slip. This form is of special significance as being a connecting link between the
foregoing bird form (with annular base, indifferent wings and tail, and rather wide
aperture with flaring lip) and those that are to follow.

The next step leads to the disappearance of the painted lump on the breast
that served as a head and the appearance of a more or less realistic, removable
head that fits over the aperture and undecorated neck at the summit of the ve-
sicular body (fig. 248). The latter is spheroidal as in the last figure, but the wings

Fig. 248. — Bird form with realistic removable head,
ware. '/»

Alligator

148

A STUDY OF CHIRIQUIAN ANTIQUITIES.

and tail are treated in a different manner. The wings are fused into the tail and
are only slightly in relief, being indicated in a large measure by means of paint.

The chief delineating
color here, as also on
head and legs, is black ;
the red is used for mas-
ses (on feet, ear-tufts
and beak) and for fill-
ing in. The scale-
group motive is freely
employed on head and
neck, wings and tail.
The ear-tufts suggest
the owl, but the beak
and the short square
tail are more like those
of the parrot. When
compared with the
head ornaments of
gold and stone images
of the parrot, these ear-
tufts are straightway
clothed with new signif-
icance the explanation
of which is suggested
on page 221. Among
the whistles in the shape
of birds, a majority rep-
resent the parrot, rarely
however with ear-tufts.
It is probable that the
bird forms with remo-
vable heads also rep-
resent the parrot. The
separation of the lower
mandible from the up-
per is generally in-
dicated by a slit. The
eyes are in rather low
relief, outlined in black
paint. The neck near
its base is perforated
transversely (see fig.
2496). There are al-

Fig. 249 ; a. — Elaborate bird form with removable head and ornamented with go fWQ cOrrCSDOndinp"
dorsal-view motives. Alligator ware. '/' ; b. — Diagram showing mode of
stringing neck and body for suspension. holes in the top of the

THE ALLIGATOR GROUP.

149

hollow head in front of the ear-tufts and somewhat nearer the median plane, making
it possible to tie the head fast to the neck and if need be to suspend the com-
pleted bird form. The highly polished cream-colored slip on the throat, neck
and body is nowhere marked by delineating colors.

A similar, and in some respects superior, type is reproduced in figure 249 a.
The beautifully turned body is somewhat longer than broad and thus just fails
of being a spheroid. The tail, beak
and the ear-tufts are longer; and the
wings are in higher relief, their tips
uniting in a median plane over the
base of the tail; the carpal angles stand
out prominently. The long sharply
curved beak has been broken at the end.
The eyes as in the foregoing are in
rather low relief and outlined in black
paint. There is a similar set of holes
at the base of the neck and in the top
of the head. Three short toes are in-
dicated on each foot. The flat sole is
pierced by a hole that extends vertically
upward for a distance of 15 millimeters
into the short solid leg. On the
contrary, the rather long legs of the
preceding figure were hollow and slit
vertically on the inside.

The paste in all these bird forms is
alike — a deep red, in which the white
grains of the tempering material are
more or less conspicuous. The slip is
a light cream. The latter, in the spec-
imen under consideration, is not quite
adhesive enough to hold firmly the
delineating colors, the result being that
the decoration on the head, wings, feet
and upper surface of the tail is marred
by the scaling off of the paint. Here
again black is the chief delineating color,
the characteristic alligator motive being
executed in it alone; while the red was

used for masses and filling in. The design in black on the wings is the dorsal-
view alligator motive similar to that on the shoulder of the vase reproduced in
figure 230. Practically the same motive is found on the head and upper surface
of the tail.

The use of these bird forms with removable heads is problematical. The
broad flat feet give comparative stability. On the other hand, they may have
been suspended by means of a cord that passed through the four holes in the

Fig. 25050:. — Vase showing elaborate and highly conven-
tionalized use of alligator motive ;fromDivala. Alligator
ware. ')2 ; 6. — Panel design on the opposite side.

150

A STUDY OF CHIRIQUIAN ANTIQUITIES.

neck and head. In the specimen belonging to the Yale collection, these holes
do not show any distinct trace of wear. By passing a cord in through one of
the neck holes, up and out through the hole in the head on the same side, down
through the other head hole and out through the remaining neck hole (fig. 249 6),

the free ends can be tied together,
thus making the head fast to the
body ; and if long enough they can
be tied to a bracket, the head and
trunk being held securely together
by the weight of the latter.

Of these bird forms, the first two
described were collected by McNiel
and the last two by de Zeltner.
With the latter's collection there
came also a third head but no cor-
responding body. As has been said,
the paste in all is a deep red similar
to that of the polychrome group to
be described later. The character
of brush-work in the largest (see
fig. 246) is not unlike that in the
polychrome group.

The discussion of the alligator
ware ends with a description of three
vases having characters that deviate
somewhat from the general type in
point of form and material, as well
as in the general character of the
painted designs. But they are, never-
theless, more nearly related to the
typical alligator ware than to any
other.

In figure 250 a, reproducing a vase
from Divala, the paste is coarse and
the modeling crude. There is a
rather sharp angle between neck
and shoulder, instead of the blend-
ing of the contour lines so charac-
teristic of the alligator group. The
usual pale yellow slip does not seem
to have been employed. On the
other hand, the ground color is every-
where red. The delineating colors
are white and black, the white being
applied as a thick layer. It has scaled off in places, leaving the paste of the
walls bare. The black seems to have been used for outlining and for filling

Fig. 251.

Fig. 252.

decoration. Alligator

Fig. 251. — Vase with problematical
ware. '/>

Fig. 252. — Vase combining features of alligator and lost color
groups, the dorsal-view motive occurring on the peripheral
band and those tangent to the neck. Alligator ware. '/>

THE POLYCHROME GROUP. 151

in small interspaces. The red was employed for a like purpose in addition to its
use as a ground color, pure and simple. The order of application was evidently
white, black and red.

The shoulder zone is divided into two panels by means of alternating, vertical
black and white bands. The design in each 'panel is reptilian and presumably
represents one and the same animal, although to one is given four legs, each
terminating in three toes, while the other is legless (fig. 250 b). The body is
serpentine in both, and the head more or less sharply defined.

The second of these vases (fig. 251) is in every respect more characteristic
of the alligator ware. The design is executed in black and red on a pale yellow
slip. Its meaning is problematical. The lines of the drawing are somewhat
similar to those on the preceding vase. The neck and lower half of the body
are finished in red.

The last vase (fig. 252) combines certain features of the alligator- and lost color-
groups, respectively. It may be recalled that in the lost color ware an oft-repeated
scheme of decoration was the division of the body into two zones by a horizontal
peripheral band and a subdivision of the upper zone into panels by means of
bands tangent to the neck on either side. The same scheme is followed here,
but the colors and their method of application are those of the alligator ware.
The bands are pale yellow (slip) slightly tinged with red, while the lower zone
and the panels of the upper are red. The bands are everywhere outlined in
black and are crossed by parallel slanting black lines in groups of three, rep-
resenting the body of the alligator, with spines projecting on either side — in
other words, the dorsal-view motive. This specimen is from Corredor.

THE POLYCHROME GROUP.

To the Yale collection belong six specimens of the polychrome group, called
by Holmes " the most artistic of the wares of Chiriqui." This ware is remarkable
for its rarity as well as its refinement and beauty of ornamentation. The National
Museum possesses only three examples and from one of these the purple color,
the distinguishing character of the group, has been left out. In the Peabody
Museum of Harvard University, there is a single specimen of what might be called
polychrome ware, but here also the purple color is lacking. On the other hand,
purple is one of the delineating colors in all six of the Yale specimens.

While perfectly distinct as a group, the polychrome ware is more closely akin
to the alligator ware than to any other. It may be recalled that in the alligator
group the delineating colors were black and red, black being used for contours
and red more especially for filling in. This is also true of the polychrome ware.
The slip in both is practically the same, except that it has a more distinctly
salmon tint in some polychrome pieces. In the latter, the paste is a dark red,
similar to that in the bird forms of the alligator group already noted (pp. 145-49)
and also to that in a class of whistles and figurines to be described later, though
belonging to the alligator ware. Certain decorative motives are also common to
both. In point of form however there is a distinct divergence from the tripods

152 A STUDY OF CHIRIQU1AN ANTIQUITIES.

and round-bottomed vases, especially the latter, so common in the alligator group.
With two exceptions, these are entirely wanting, and in their stead appear highly
differentiated forms with annular bases that are sometimes developed into tall
hollow stands.

The series begins with a pitcher-shaped vase from Gualaca (PI. XLIV, fig. a).
The body, which is flattened uniformly above and below, is supported by a low
annular base and surmounted by a neck that is gently flaring and not provided
with a spout. The pronounced and horizontally flattened lip is damaged by
weathering and chipping. The handle, which was attached to it and to the shoulder
below, is entirely gone. The ascending ramus of the handle was pegged to the
shoulder, as indicated by an empty hole. To be in harmony with the main out-
lines of the vessel, the horizontal ramus of the handle must have met the ascend-
ing branch at an angle of less than 90°. This is true of the handle of a pitcher

(cat. no. j^g) in the Peabody Museum of Harvard Uni-
versity, that belongs evidently to the polycrome group,
although no purple was used in the decoration of the
vessel. Neither is the form of the latter so pleasing.

The Yale specimen is somewhat damaged from weather-
ing. The paste is much lighter in color than that of
the other specimens, a fact due partly to bleaching. The
entire surface was coated with a pale yellow slip, over
which the designs were executed in black, red and
purple. A broad red band encircles the annular base
and the neck constriction. These are joined by four
broad longitudinal bands, alternating in purple and red,
Fig. 253.— Sigmoid scroll in one being in line with the handle. A narrow purple
which the alligator motive fe d surrounds the base of the handle. In the four

appears as simply cut out of

the black band where it is panels thus formed and on the neck, black alone is
broadest ^(see Plate XLIV, employed. Three bands, two black and one red, dec-
orate the lip. The orifice is unpainted.

The attention is at once fixed upon the ornate scrolls of the panels (fig. 253),
all of which are approximately the same, except that of the two facing the handle ;
one is of necessity left-handed and both are slightly modified to make room for
the base of the handle. The ingenuity displayed in adapting the motive to the
space at hand, the training of the eye and the delicacy of touch are all
marvelous. The scroll ornament is carried horizontally about the neck, one and
one-half units of the motive being required to fill the space to be decorated
(fig. 254). The oval notches cut from the body of the scroll at intervals where
the black band would be broadest evidently mean something more than a mere
effort to relieve the pattern of broad black areas. They are highly conventionalized
alligator motives of the profile variety, with the dermal markings not represented.
In the last illustration (fig. 254), two of the motives are not completely fused with
the mass of the scroll; the body-line is visibile for a short distance near its
center. The specimen in the Harvard University collection is decorated with
a similar branching scroll that completely encircles the body of the vessel.
Beginning at a point beneath the handle, it is developed in both directions until

THE POLYCHROME GROUP.

153

Fig- 254-

the two arms unequal in length meet and fuse. The alligator motives represented
by oval notches are distributed precisely as in the Yale example.

A round-bottomed vase is reproduced in figure 255. The nearly spherical lower
half is given a uniform coat of red. The orifice is also painted red. The over-
hanging lip and short neck are banded with red and black. The decoration on
the flattened upper zone consists of an animal head and tail in relief and painted
panels on either side. The panel motive is precisely the same as that on the
neck of the pitcher
from Gualaca (see fig.
254), except that the
details are more elab-
orate and the execution
is of a higher order,
three colors also being
used instead of one.
Here again it takes one
and one-half units of
the motive to fill the
long panel. The ter-
minal half -units are
bilaterally symmetrical,
the end of the scroll
in each case being
purple outlined in
black. The somewhat
abbreviated central
half-unit terminates in
a red coil instead of
a purple. This half
taken with the one on
the right completes a
sigmoid scroll. In other
words, the growth of
the scroll complex is

Fig. 254. — Branching scroll in which two of the alligator motives (notches)
are partially differentiated (see Plate XLIV, fig. a), '/«

Fig. 255. — Vase with elaborate branching scroll from which the alligator mo-
tives are partially detached and accompanied by dots and circles representing
body-markings. Polychrome ware. 'I'

from right to left, as
it was of that on the
neck of the pitcher. In
this instance, the oval
notches of figure 254

become so much more highly differentiated as scarcely to be recognizable.
Fortunately the two upper median ones serve as a connecting link between the
simpler form as seen on the pitcher and the partially detached form employed at
the lower corners, for example. Each one represents the body-line of the alligator,
and the accompanying dotted ovals stand for the dermal markings. The panel
decoration on the opposite side is almost completely weathered away. Judging
from what remains, it was the exact counterpart of the one illustrated. In con-
MEMOIRS CONN. ACAD., Vol. III. 20

154 A STUDY OF CHIRIQUIAN ANTIQUITIES.

ception, execution, balance and symbolism, this scroll is indeed a masterpiece,
one that baffles the skill of a modern artist to copy faithfully.

The next vase (PL XLIV, fig. &) in the series is in perfect condition. The form
though eccentric is graceful. The usual annular base is present. The lower half
of the oval body is finished in what might be called a red (light maroon) slip.
The same finish is given to the orifice and neck. Below and encircling the latter
is a broad band of a brighter red ; the same that is used as a delineating color.
On the upper zone and at each end of the oval body is a bovine' head in relief,
turned to the right at an angle of 90°. The hollow interiors of the two heads
communicate with that of the main body of the vase. Alternating with these
heads are two long panels painted over a salmon-colored slip. The black frames
for the panels are bounded above and below by continuous black bands and
beneath the lower of these is a bright red band bordering on a dark red slip.
The colors therefore consist of two shades of red, pale salmon, black and purple.

The panel design is the same as in the two preceding — a highly ornamental
running scroll, four half-units being employed on one side and six on the
other. The growth of the scroll however is from left to right in both panels ;
the reverse of that noted in the foregoing examples. Beginning also with a
sigmoid scroll, it grows in a natural way by budding or branching instead' of
by a series of interlocked S's — in other words, it might be called the direct or
branching scroll as opposed to the interlocked or reverse type, otherwise known as
the Vitruvian scroll. The volutes are tipped with purple and red alternately. The
characteristic alligator motive is woven into the design as was noted in the preced-
ing examples. The decorative motive on the body of the previously mentioned
pitcher belonging to the Peabody Museum of Harvard University is also the
branching scroll, similar to that seen in figure 254.

The features of the two heads in relief are emphasized by the skilful use of red
and black colors over a salmon slip. Of special interest is the treatment of the
black bands over the eyes, giving to the slip the effect of a delineating color. At-
tention should also be called to the shading of the muzzle by means of black lines.

The use of the scroll or spiral ornament is not confined to any particular land
or people. It has developed in various centers and at various times and has been
spread over the face of the earth both by borrowing contemporaries^ and by
descendants. Many attempts have been made to trace its origin. In Egypt, it
is supposed to have come from the curling sepals of the lotus1 and to have been
carried with the Egyptian civilization into Europe. In our own country it reached
a high degree of development in the lower Mississippi valley and in the pueblo
region of the Southwest. Holmes2 would trace the origin of the scroll to the
coiled fillet of clay with which the potter began his vessel ; to the basket-maker's
platted or twisted splints of wood ; to the spire of a conch-shell, or to the linear
representation of waves of water.

Of the two types of scrollwork (each with its variants), the classic Vitruvian
or reverse scroll and the direct or branching scroll, the latter would seem to

1 Op. cit.

2 W. H. Holmes, Origin and development of form and ornament in ceramic art. Fourth
ann. rept., Bur. ethnol., 456, 1882- '83.

THE POLYCHROME GROUP.

155

be the more primitive and appears to have been the favorite among the ancient
Chiriquian potters. The single S was used, to be sure, but not a series of inter-
locked S's to form a current scroll of the reverse type.

The branching scroll is apparently of much less frequent occurrence than the
reverse scroll. With the exception of Chiriqui, it is rarely seen in the art of the
Western hemisphere. Riegl l figures
examples of it from Rhodes, Greece,
Pompeii and Syria. These are practically
all outspoken representations from the
plant world. On the other hand, the
Chiriquian branching scroll bears no re-
semblance to any plant except in the
manner of its growth, as suggested by
its name. The motive bears the stamp
of originality in all its details and may
justly be called Chiriquian.

A rare form of vase is illustrated in
figure 256. It was one of 'the few selected
by deZeltner2 for illustration; his photo-
graph, however, is not only very small
but is also lacking in detail. Using de
Zeltner's diminutive photograph as a
guide, Holmes3 attempted to reproduce
it by means of a drawing, which is un-
avoidably inexact. At that time Holmes
had not seen the original. In shape, it
is like an elongated bell resting on a
shallow bowl. As usual, the support is
annular and the lip projects but slightly
above the flat angular shoulder. The
decorated zones are bounded by hori-
zontal bands applied in groups of three,

consisting of a red band between two black ones of unequal breadth. The upper
zone is divided into four panels by means of two pairs of triple vertical bands.
The two larger panels are ornamented with a sort of fretwork, the design being
approximately the same in each. The outlines are in narrow bands of black.
The filling is done in purple, except at the ends, where it is red. The derivation
of this motive is presumably the same as that of the sigmoid scroll, i. e., from
the united body-lines of two alligators.

The chief interest attaches to the motive already familiar by reason of its as-
sociation with the foregoing examples of branching scrollwork (see text-figs. 253-255
and PI. XLIV). Here it is repeated eight times in order to fill the lower zone,
and is completely detached from the branching scoll, which is left to be supplied

1 Alois Riegl, Stilfragen, figs. 76, 96, 128, 130 and 158, Berlin, 1893.

2 Op. cit.

3 Op. cit., fig. 213.

Fig. 256.— Unique type ot vase, the zonal decorations
being alligator motives. Polychrome ware. '/«

156

A STUDY OF CHIRIQUIAN ANTIQUITIES.

by the imagination. The design represents the body-line of the alligator (see
PI. XL, fig. /; and text-fig. 224), and the concentric rings placed in the center
of the dorsal concavity are the body-markings. The multiple motive is thus
broken up into its constituent units, each representing a whole alligator. An
excellent example of this motive is figured by Holmes.1 The Yale Museum is
particularly fortunate in having examples of it in various stages of its development
and showing its connection with the branching scroll.

The largest vase of the polychrome group and indeed of the entire collection
is reproduced in figure 257. In shape it is unique and in modeling, perfect.
Beginning with the simple rounded bottom, the walls are carried upward and

Fig. 257. — The largest vase in the collection, unique in form and decorated with alligator motives.
Polychrome ware. '/>

rapidly outward till the greatest diameter is reached at the high and sharply
angular shoulder. There is a further flattening at the collar approximating a
horizontal plane. From this rises a short neck with angular lip. Rather far out
upon the shoulder and on opposite sides are two raised highly conventional animal
forms, which also resemble functionless handles. Similar forms occur on one of
the polychrome vases belonging to the United States National Museum.

The red slip of the lower zone and of the aperture differs in tone from the delineat-
ing red, and is approximately of the same shade as the paste. The ground color
of the upper zone, including collar and neck, is the usual cream slip, on which
excepting the collar, appear the delineating colors, black, red and purple. The

1 Op. cit, fig. 211, 212.

THE POLYCHROME GROUP. 157

highly skilful and artistic panel ornaments are broken in the middle by the sharp
angle at the shoulder. The four panels are grouped so that two opposing ones
form a pair, the smaller pair being in line with the relief forms. The designs in
the large panels are similar, but not identical. The same is true of those in the
small panels. These designs all seem to rest on the broad black band that forms
the bottom of the common panel frame, but do not quite reach the black band
at the top of the frame.

The purple is used very sparingly, being found only in the two small panels.
Here, the principal element of the design resembles the false door (Scheinthiir) of
an Egyptian tomb. The narrow longitudinal central framework is filled in with
purple, one margin of which is thrown into a series of wave-like curves. The
crosspiece at the top is also purple. Above this comes the life form in relief and
above the latter, a much flattened oval ring of purple outlined in black. The details
of the panel opposite are practically the same, except that red is employed, instead
of purple, in the central framework.

The designs of the two large panels are each built up around a central red
oval. About this are grouped the four arms of an X-shaped design, each contain-
ing a small panel, decorated with the false door pattern resembling somewhat
that already described. In the upper and lower angles of the X is an elaborate
rendering of the alligator motive in black — the body-line of the animal and a
small concentric ring placed over the dorsal concavity (compare with fig. 256).
Within the lateral angles are quadrangular fields, each enclosing a more or less
realistic representation of the alligator. These are reproduced in figure 258.

Fig. 258. — Conventionalized alligator designs illustrating the process of sparing the pattern out of the field (see
fig. 257). V»

The technique here is different from that which obtains elsewhere throughout
this group. The design is produced not by delineating colors, but by filling in
the field around the figure in black, leaving the figure in the color of the slip.
The skill and ingenuity with which this work is executed are simply amazing.
They make the ancient Chiriquian a worthy rival of those who boast the tra-
ditions of Greece and of the renaissance of Italy. He proved himself master of
the brush in three distinct systems: (1) The production of the figures by direct
application of delineating colors ; (2) the lost color process, and (3) by sparing
the figure out of the ground (ausgespartes Ornament).

The series ends with the gem of the polychrome group and the finest work
of ancient Chiriquian decorative art known to the author. In originality of design,
richness of detail and skill in execution, it stands alone (PI. XLV and frontis-
piece). Holmes l attempted to reproduce this piece also, from de Zeltner's miniature

1 Op. cit, figs. 214, 215.

158 A STUDY OF CHIRIQUIAN ANTIQUITIES.

photograph, but his two figures are of necessity full of slight inaccuracies. Like
the rest of the polychrome series, except the first specimen,1 this came to Yale with
the de Zeltner collection.

The shallow bowl, which has a diameter of over twenty-seven centimeters, is
mounted on a hollow perforated stand that gives to the whole a height of about
nineteen centimeters. The thickness of the walls varies. The rim of the bowl
is ten millimeters thick, but this thickness is soon reduced by half as the center
is approached. The same is true of the stand, the walls of which are thickest
at the base and grow thinner rapidly toward the top. Of the four longitudinal
slits like elongated triangles that pierce the walls of the stand, the alternating
ones point in opposite directions. These openings were cut before the surface
was polished or painted. The stand is in excellent condition, but the bowl had
been not only severed from its support but also broken into several pieces.
These had been put together in an indifferent manner while still in de Zeltner's
hands, and the beauty of the designs had been much marred by repainting. By
means of alcohol, which luckily does not affect any of the original colors (cream,
black, red and purple), I have removed all such painted restorations.

This is the only vessel in the Yale or any similar collection, where every
visible part has been carefully decorated, the interior of the hollow stand only
excepted. The ground is a cream-colored slip. Black is the color chosen for all
the outlines and for the minuter details. It is everywhere applied with a very
fine-pointed brush or instrument, the lines, whether straight or curved, being drawn
with precision. Red and purple are used alternately as mass colors, always on
spaces that are outlined in black.

The ornamentation of the base consists of a lower horinzontal red band contin-
uous with two spirally ascending red band, and a horizontal purple band at the
top continuous with two spirally descending purple bands alternating with the
ascending red bands. These four bands pass directly over the four openings.
The designs on the intervening spiral panels resemble the herring-bone pattern,
two pointing upward and two alternating with these, pointing downward. Thus,
the balance in the use of color as in the design is everywhere maintained.

The exterior of the bowl (fig. b) is decorated as follows : A short distance above
the top of the stand, there is a red band and above this a broad zone limited
above and below by black bands. Within this zone are what appear at first sight
to be four panels or compartments, each enclosing an alligator's head. This, in
fact, is the interpretation de Zeltner2 put upon them when he said: — "La partie
inferieure de la coupe est divisee en quatre compartiments, dont chacuri renferme
un dragon paint en noir et rouge sur fond blanc ; les encadrements sont tantot
rouges, tantot violets." A more minute observation proves that they are not panels
(four) at all. The entire design resolves itself into two units of the classic fret.
The ends of each fret are linked with alligator heads, which face each other, one
being of necessity inverted. Holmes has already pointed out the derivation of
the fret from the body-line of the alligator. In the present instance, we have a

1 This pitcher with broken handle formed a part of the McNiel collection.
9 Op. cit., 9, 10,

THE POLYCHROME GROUP. 159

realistic verification, showing that a unit of the fret involves the body-lines of
two alligators. The upturned snout and the spines on the back of the head are
easily recognizable. The bands composing the fret are delicately outlined in
black and also divided into long slender compartments, that are filled in with red
and purple alternately. The fret with accompanying heads on the opposite side
is similar to the one figured. This is one of the most elaborate examples of the
association of a life form with the fret or scroll.

The conventionalized treatment of the alligator by the ancient Chiriquian artists
suggests a comparison with that of the crocodile (cipactli) and of the blue feather
snake (xiuhcouatl) of the ancient Mexicans. The conventional head of a bird worn
on the forehead of so many Mexican deities (Ginteotl, Xochipilli, Tonacatecutli,
Quetzalcouatl, TonatiuK) also reminds one of the alligator head with its recurved
jaws. In this connection, it is interesting to note that life forms with a head at
either end of a common body are incised on some of the metates from Las
Guacas, province of Nicoya, Costa Rica, recently described by Hartman.1 In one
case, the common body is a guilloche pattern and in the other it has the ap-
pearance of being tied in a knot, bringing the two heads closer together. The
latter do not seem to have been noticed by the author. They look very much
like the Chiriquian alligator head. In Peruvian art, also, there is a reptilian motive
akin to the so-called alligator motive of Chiriqui, even to the spine- and scale-
symbols, the nuchal crest and hooked snout.

The most extraordinary design is the one inside the cup or chalice (PL I)2 — a
human body and extremities surmounted by the alligator's head with all its
characteristic traits (the suspended lower jaw, recurved snout and a frontal as
well as a nuchal crest). This is the same mythical creature, excellent examples
of which are to be recognized among the gold figurines (see PI. XLVIII, fig. g ;
and text-figs. 365-368), to which I have given the name alligator-god. The artist-
ically executed spines with alternating red and purple are attached to the crests
instead of to the head proper. Within the field back of the eye and leading
down to the shoulders are three alligator motives — the curve of the body ac-
companied by the symbol for the body-markings placed in the dorsal concavity.
In the upper and the lower motive there is a dorsal angle instead of a dorsal
curve. The central figure, however, is very similar to the motive as it appears
in preceding illustrations of polychrome ware (see fig. 256). The same design is
employed to fill angular spaces on opposite sides of the field. Something resem-
bling a tail branches off from the body on either side in the region of the hips.
Accompanying each of the four extremities there is a design composed of series
of parallel lines meeting at an angle. Not an inch of space is left undecorated ;
spines and teeth are to be seen everywhere. Both de Zeltner and Holmes speak
of the resemblance this interior decoration bears to Chinese art, particularly to

1 Archeological researches on the Pacific coast of Costa Rica. Mem. Carnegie Museum,
III, no. 1, PI. VIII, fig. 2, and PI. XVIII, fig. 2, 1907.

2 The original water-color drawings for the five chromolithographic plates (I, XXVII,
XLII, XLIV and XLVIII) in this volume were destroyed by fire before the first proofs of
the same had been corrected.

A STUDY OF CHIRIQUIAN ANTIQUITIES.

that of the earliest known period ; but neither ventures to claim an identity of
origin.

The source from which the purple was obtained remains a mystery. That its
origin was known only to a few, is evident from its rare use. It does not seem
to have been known outside the province of Chiriqui, from which only eight
specimens have been reported that include purple as a delineating color. Six
are in the Yale Museum and two in the United States National Museum. The
color probably comes from a non-ferruginous metallic oxide and was apparently
applied before firing. Had it been applied after the firing, a vegetable or animal
dye might have been used ; that from a mollusk (Purpura patula) for instance, which,
according to Dr. H. Pittier de Fabrega, may be found clinging to the rocks between
high and low tide levels on both the Atlantic and the Pacific coast of Panama.

The purple industry is practised by the natives of Cano Island, off the mouth
of the Diquis river ; also by those living at Golfo Dulce,1 on the southwest coast
of Costa Rica. In Nicoya, the Indians color their cotton thread in a primitive
manner by simply drawing it over the shell opening of the snail, thereby dampen-
ing it in the escaping purple liquid. The color, which is at first greenish yellow,
becomes violet on drying. The shell is reddish about the opening, remind-
ing one of P. hcemastoma Linne of the Mediterranean, with which the fishermen
of Minorca still mark their linen. On the other hand, the purple of the ancient
Phenicians and Greeks came from Murex trunculus and M. brandaris Linne.

On page 181 of his Conchyliologie (1742), M. Degallier d'Argenville states that
the " Conque Persique" is made use of both in Panama and Guatemala to color
cotton thread and, on that account, is called " Pourpre de Panama." According to
van Maartens, the " Conque Persique" of d'Argenville is the Purpura persica of
Brugiere and Lamarck, a species very similar to Purpura patula and its representa-
tive in the Indian Ocean.

As early as 1744, Don Antonio de Ulloa saw at S. Elena, in what is now
Ecuador, and also at Nicoya (Costa Rica), purple color produced from a snail,
and speaks of it as being pronounced and durable, so that it lost nothing either
from frequent washing or from long use.

Thomas Gage 2 was a still earlier observer, his account being as follows :

" About Chira, Golfo de Salinas, and Nicoya, there are some farms of Spaniards, few and
very small Indian Townes, who are all like slaves employed by the Alcalde Maior, to make
him a kind of thred called Pita, which is a very rich Commodity in Spain, especially of
that colour wherewith it is dyed in these parts of Nicoya, which is a purple colour ; for the
which the Indians are here much charged to work about the Sea shore, and there to finde
out certain shels wherewith they make this purple dye."

In this connection, the observations of Mr. C. V. Hartman are interesting. On
one of his recent expeditions to the Pacific coast of Costa Rica, he visited Guana-

1 Joaquin Bernardo Calvo. Administracion Soto, Repiiblica de Costa Rica, Apuntamientos
geograficos, estadisticos e historicos. S. Jose de Costa Rica, 1886.

2 The English-American, his Travail by Sea and Land : or, A New Survey of the West-
Indies, Containing A Journall of Three thousand and Three hundred Miles within the main
Land of America, 192, 193. London (R. Cotes), 1648.

STOOLS. 161

costa, where he saw an Indian woman from Chiriqui wading in the water in search
of Purpura. She would put the shell to her mouth and blow into it, causing the
snail to discharge a greenish yellow fluid, which she applied to white cotton thread.
The fluid in drying turns to purple.

This industry was met with recently in Tehuantepec, southwestern Mexico, by
Professor Eduard Seler, who while there obtained a petticoat or tunic (Span, enaguas),
which the Zapotecan women wear only on special occasions and which but few can
afford to own. The costliness of the garment is explained by the quantity of
snails that would be required to color it. The snails in question are not very
plentiful. They are taken from the water alive and spit upon (man bespucke sie).
The fluid emitted by the animal on account of this unusual treatment is collected
and the creature thrown back into the water. This is certainly a more econom-
ical process than the removal of the soft parts from the shell, customary in Ulloa's
time. Mrs. Zelia Nuttall * notes a survival of the purple industry in Mexico.

The use of purple in the decoration of pottery is not only exceedingly rare in
Chiriqui, but seems to be confined to that culture area. The nearest approach
to it that I have been able to discover is seen on a vase from Mercedes (Costa
Rica) belonging to the Keith collection. Curiously enough the purple is employed
as a ground color on the piece in question, the pattern being produced by the
lost color process. The design is therefore purple, while the interspaces are
covered with apparently the same non-adhesive black that was used for a like
purpose on the lost color ware of Chiriqui. In the character of the paste and
the modeling, however, the specimen resembles neither of the two Chiriquian
wares, lost color and polychrome.

VARIOUS OBJECTS OF CLAY.

The ancient artificers of Chiriqui left practically no architectural monuments. Their
fame must rest upon the achievements of the potter, the goldsmith and the sculptor.
In the mastery of clay they had few equals and no superiors on the Western
Hemisphere. This is shown in the infinite variety of form and perfection of the
modeling, in the painted as well as plastic decoration and the variety of uses to
which the product was put. The pieces already described were vessels either
for domestic use or for ceremonial and decorative purposes. In addition to this
general class, there are various objects of clay, such as stools, similar to those
executed in stone ; spindle-whorls ; stamps ; small receptacles with removable covers,
that may have served as needle- or jewel-cases ; figurines or statuettes and musical
instruments, the latter including rattles and whistles.

STOOLS.

The genetic relationship between metates and stools was referred to in the
chapter on objects of stone. There is a series of clay stools that follow rather
closely the types executed in stone, except that they are more ornate, as might

1 Putnam anniversary volume, 368-384, 1909.

MEMOIRS CONN. ACAD., Vol. III. 21

162 A STUDY OF CHIRIQUIAN ANTIQUITIES.

be expected from the ease with which clay may be modeled. They all belong
to the terra cotta ware.

A series of clay stools is represented in Plate XLVI. One of the more elementary
forms is shown in figure a. The top is nearly square, concave in both directions
and without decoration. There are four legs, the two on each side connected
by a basal crosspiece, leaving the ends open. A figure, half-human, half-monkey,
stands between each pair of legs and helps to support the seat or plate. This
piece resembles one stage in the development of the metate (see fig. 21) on the
one hand and, on the other, the wooden seats (see fig. 22) in use among the
Chiriquians of today.

Among the clay stools the elements of the prevailing type are seen to good
advantage in figure b. The seat is round instead of square and the four legs rest
on a ring at the bottom. The whole structure is strengthened by two opposite com-
paratively large grotesque figures, with monkey heads and human bodies. These
are hollow and serve as rattles. Similar figures are to be found among the gold
objects. In fact the metal-worker's technique seems to have been borrowed by
the makers of the clay stools. Contrary to the general rule, the top in figure c
is gently convex instead of concave. The four flat legs to which grotesque figures
are applied rest on a connecting ring at the bottom. The four interspaces are
almost completely filled by four supporting figures.

An interesting type of stool is shown in figure d. The three legs or pillars are
spread laterally till they almost meet. In other words, the top or seat rests on
a hollow stand whose walls are continuous except for three narrow vertical slits.
Three fantastic forms are applied to the supporting walls. These are slit on
the hollow bodies of the figures outside. To a marked degree, the flattened or
beaten character of the arms of the latter resembles work in metal. There are
nineteen additional small grotesque heads applied to the margin of the seat. An
incised braidlike fillet encircles the base ; a similar one is placed beneath the series
of small heads. Incised zigzag patterns relieve the monotony of what would other-
wise be plain interspaces. They are probably serpent motives and are so placed
as to appear to be held in the hands of the three applied figures, recalling the
attitude of the alligator-god holding a snake in either hand (see PI. XLVIII, fig. g).

A low stool with convex seat is reproduced in figure e. The circular wall of
the hollow stand is broken at irregular intervals where openings of various shapes
and sizes are so made as not to interfere with, but on the contrary to heighten the
effect of the numerous applied figures of men and monkeys. The bodies of all
these are uniformly marked with diminutive circular indentations, which occasion-
ally appear also on the supporting wall. Three of the figures have human bodies
and monkey heads ; two of them are inverted, the other is upright. The remaining
ten are long-tailed monkeys, apparently both old and young ; all of them in pictures-
que attitudes, some playful, some demure.

In figure /, the monkey forms that support the seat are similar to those seen
in the stone stools (see PI. IV, fig. d). The animals seem to be swinging round
a circle holding each other's hands. The tail of each is fused with the left elbow
of the figure on its right and is also supported by a tiny monkey sitting on the
basal ring below it. The margin of the seat is battered in places, as if the piece

SPINDLE-WHORLS.

163

had seen service in some capacity. The seat of one rather tall stool of crude
workmanship is supported on the heads of four caryatids, whose arms are disposed
after the fashion of the " Panama Venus " (see fig. 40). The seat of the smallest
clay stool in the collection rests on the uplifted heads and tails of four animal figures.
The largest and in many respects the finest piece of this kind is reproduced
in figure 259. It measures 30 centimeters across the top and is 17 centimeters high.
There are but three legs alternating with strange half-human forms that stand on
the connecting ring below and help to support the concave seat above. Except
for the loin-cloth and absence of tail, there is little to suggest the human
figure. Here again the arms are fashioned as if of hammered gold. The bodies
are hollow, slit down the back and each carries a clay ball. The grotesque head

Fig. 259. — Large fine clay stool, the seat supported by three legs alternating with
strange half-human forms. Armadillo ware. */"

is not human. It may be that of an ape or an alligator, the horizontal flattening
of the jaws suggesting the latter. In some respects, the entire figure is analogous
to certain gold figurines of the alligator-god. There are nineteen additional small
heads surrounding the rim. The seat is concave and highly polished for ware
of this kind.

SPINDLE-WHORLS AND STAMPS.

The spindle-whorl is one of those links that bind the archeology of region to
region and of age to age. The weaver's art seems to have developed in widely
separated parts of the world and in some places earlier than in others. Spindle-
whorls were found by Schliemann at Troy, by Keller in the Swiss lake-dwellings,
and by many archeologists in various parts of Europe and America. There is often
a striking similarity between those found in regions remote from each other; as,
for example, Troy and Mexico. The collection of spindle-whorls from Chiriqui

164

A STUDY OF CHIRIQUIAN ANTIQUITIES.

is small, consisting mainly of plain perforated disks made from potsherds (fig. 260).
In one specimen (fig. 261), the axis is somewhat prolonged in one direction and
about it are two opposed figures, probably human, each indicated by a head and
arms in relief. The head and arms only are indicated (fig. 261). This and the
preceeding are both in the natural color of the baked clay.

Specimens belonging to this class are extremely rare. Holmes figures none
whatever. The only one in the Yale collection is bell-shaped and made of the
same paste as the pottery of the scarified group (fig. 262). The paint is also the
same. There is an attempt at a perforation through the part serving as a handle,

Fig. 260.

Fig. 261.

Fig. 262.

Fig. 260. — Spindle-whorl consisting of a plain perforated disk. '/«

Fig. 261. — Spindle-whorl decorated with heads and arms in relief. '/>

Fig. 262. — Bell-shaped stamp; a, side view; b, view of lower end. Scarified ware. */«

but the two holes do not seem to have become continuous at a central point.
The small stamp at the top resembles an eye with branching rays. The large
stamp at the mouth of the bell is of the same nature and design; but instead
of a single slit across the eye, there are two forming a cross, outside of which are
two concentric rings of rays.

Fig. 263.

Fig. 264.

Fig. 263. — Needle-case. Lost color ware. */»
Fig. 264. — Needle-case ; from Divala. Lost
color ware. */•

NEEDLE-CASES.

Under this head is grouped a series of
small oblong to oval receptacles, the use
of which is more or less problematical. They
were called by Holmes needle-cases, but he
also suggested that they could have served
to contain " any other small article of domestic
use or of the toilet." They belong for the
most part to the lost color ware. They
consist of two parts, body and lid. Both are
perforated near their margins of contact, in
such a manner as to make the holes of one
coincide with those of the other. The two
specimens 'illustrated (figs. 263-264) belong to
this lost color group of ware ; the lid or cover
is missing from each.

165

FIGURINES.

The majority of objects in clay that may be classed primarily as figurines rep-
resent man. On the other hand, plastic life forms as ornaments or supports for
vases rarely have human attributes. The primitive artist everywhere has familiarized
himself with the animal forms about him more than with that of his own kind.
This is somewhat unexpected, as the human model would seem to be the most
available from every point of view. The cave-dwellers of Europe have left many
and excellent engravings and frescoes of species now extinct. They took little
pains however to leave portraits of their own race. The early human figurines
of Greece were extremely primitive, particularly as regards the physiognomy. The
faces were more birdlike than human. The heads of the human figurines from
Chiriqui are perhaps more rodentlike than avian. The Chiriquian artist seems

Fig. 265. — Human figurine representing mother and
infant. Alligator ware. '/>

Fig. 266. — Human figurine. Alligator
ware. '!•

to have treated the human form as a joke, the result being generally grotesque
if not even ludicrous. When it came to modeling an owl or a peccary however
his work was more faithful to life.

Figurines that are first of all musical instruments are placed under the latter head.
On the contrary, those that contain pellets are retained here, since the typical
gourd-shaped rattles make a class by themselves. Every figurine in the series
as constituted belongs to the so-called alligator ware with its characteristic paste,
slip, delineating colors and decorative motives. All are perforated transversely
through the neck, evidently for suspension. As has been said, a majority are
represented as human, or perhaps more accurately, as divine. Figure 265, for
example, reminds one of Isis and Horus ; a mother seated, her right hand resting
on the knee and her left holding an infant to her breast. The nose, the most
prominent feature, is fused with the forehead by means of a high bridge ; the

166

A STUDY OF CHIRIQUIAN ANTIQUITIES.

mouth is hidden beneath the overhanging nose and the chin is wanting. This
type of profile is repeated over and over again.

The hair, represented by black stripes, reaches to the shoulders. This character
together with the sitting posture and spreading legs is common to all the
female figurines. The base in each of these is pierced by three holes or slits that
communicate with the hollow interior of both legs and body. These seated
figurines may all be images of one and the same mother goddess. They
have their counterpart not only in Isis and Horus, but also in Bachue and

her son of the Chibchas ; in the Mexican
Goddess of Flowers, Xochiquetzal, with
Macuilxochitl-Xochipilli on her arm; and in
Ciuacouatl of the Valley of Mexico.

A similar statuette is shown in figure 266,
evidently the same mythical character with-
out the infant. Both hands rest on the
thighs, the fingers being indicated by black
lines. The usual conelike termination of
tlie legs is here recurved to represent feet.
In addition to the neck perforation, there
is another marking the external auditory
opening.

In the case of figure 267 a, it would be
difficult to determine which sex is intended.
The image holds in its hands a shallow
bowl, which is also securely attached to
it at three points, the abdomen and the
knees. The painted decoration is an im-
portant feature; representations of the alli-
gator in black cover both arms and fill a
panel that extends the length of the two
legs from ankle to ankle, passing across
the rump. Of the alligators that can not
be seen in the general view, three are
selected as samples. The one with a
double head (fig. 2675) is from the right
arm. The artist has contrived to show both

nuchal appendages, but the feet of only one alligator. In the figure (267 d)
at the end of the panel, a different technique is employed. Here it is the field left
unpainted that makes up the figure of the alligator; in other words, with the
single exception of the eye, the black determines the form represented without
being a part of it. It is the double-headed form of the alligator and evinces a
high degree of imaginative skill, the head on the right being inverted. Two such
seated figurines holding bowls are in the collection of Dr. W. J. Lamson of
Summit, N. J. They also belong to the alligator group, but are not ornamented
with figures of that animal.

In figure 268, is shown what appears to be a human form seated on a stool or

Fig. 267; a. — Seated human image holding a shallow
bowl. Alligator ware. '/» ; b.— Detail of double-
headed alligator decorating the right arm ; c,
d. — Detail of one end of the panel decorating
the legs. '/'

FIGURINES.

167

Fig. 268. — Human figure seated on a stool
or metate. Alligator ware. */>

metate. The hands rest on the shins ; and the feet, which projected over the
edge of the seat, have been broken off. The seat itself has four feet. This
specimen recalls one from Santo Domingo, illustrated by Fewkes, and points to a
relationship between metates and stools.

A different type of image is reproduced in
figure 269, which exhibits a standing masculine
figure with short legs, flattened body and elongated
face, on which the nose, broken in this instance,
continues to be the dominant feature. On the
head is the characteristic conical cap met with in
the armadillo, as well as in the handled, group
of ware. The body and head are both hollow,
the latter being supplied with pellets. The ex-
ternal auditory opening is marked by a perforation.
This specimen is from Jacu and probably represents
a water-carrier, as witness the jug borne on the
back. The painted decoration is injured by sur-
face disintegration due to weathering. There is
a red field on the throat and breast reaching down
over the abdomen.

Figurines of animal forms other than those adapted as
whistles are not very numerous. Figure 270 represents the
armadillo. The attitude suggests the act of burrowing or
possibly an attempt to roll up for- defense, which has been
frustrated by the interposition of a thick bar placed trans-
versely across (or through) the stomach and projecting .some
distance beyond the carapace on either side. The head is
depressed, the nose pointing slightly backward and the chin
pressed against the chest. The tail, indicated both in the
round and by means of a band of black paint, is brought for-
ward underneath, applied to the ventral surface. The extremi-
ties are sharply flexed, the toes of the forefeet pointed for-
ward and those of the hindfeet turned backward (fig. 270 6).
The three regions of the slightly raised carapace are quite
distinct. The smooth surface of the anterior and posterior
sections is given a uniform coat of black paint, while the
central region is painted red and grooved so as to bring out
in relief three rather wide bands. The transverse bar pass-
ing through (or across) the stomach is also painted red. One F'g- 269.— standing mas
of the projecting ends of this rod was evidently broken be-
fore the red paint was applied, as a coat of it is carried
over the uneven surface of the break. Or the break may have occurred while
the object was in use and was afterward painted over to match the sur-
rounding color. It is, at all events, a case of prehistoric repairing. The part
broken off may have been a head for which the opposite end could have
served as a tail, since the diameter of the latter is not so great as that of the

culine figure ; from Jacu.
Alligator ware. '/>

168

A STUDY OF CHIRIQUIAN ANTIQUITIES.

missing protuberance. There is a ventral perforation that communicates with the
interior of the hollow body.

It is interesting to note that the disposition of
colors on the carapace of the armadillo whistle (see
fig. 303) and of this armadillo figurine is the same.
In each case, the central banded region is in red,
while the rest of the carapace is in black. In both
cases also, a red median band, bounded by two
parallel lateral black bands, is carried between the
ears and forward toward the tip of the nose. The
treatment of the eyes and nostrils is also identical.
Two examples of the peccary (Dicotyles) are given.
On account of its stripes, one of these (fig. 271)
might be intended for a young tapir (Tapirus bairdif).
Black is the prevailing color, red occurring only on
the snout and as a single transverse band on each
leg. The mouth is partly open, exposing teeth ; the
lower jaw is much shorter than the upper. There
is the customary neck perforation, and in the hollow
body are a number of pellets.

Figure 272 leans to the grotesque, the body being
excessively short. Holes for nostrils are punched
in the large cylindrical snout, the end of which is
painted red. There are also two narrow transverse
bands of red on each leg. Black is employed in

mass to cover practically the entire body. In discussing the fauna of the Isthmus,
Seemann1 says: "Pigs wander in herds about the forest and are dreaded by the
natives who, when they see them, seek safety in flight or by climbing a tree."

Fig. 270. — Figurine representing the ar-
madillo, a. — lateral view ; A. — ven-
tral view. V»

Fig. 271.

Fig. 272.

Fig. 271. — Perforated figurine representing the peccary (Dicotyles). Alligator ware. '/»
Fig. 272. — Grotesque figurine representing the peccary. Alligator ware. '/'

As the colored peccary (Dicotyles tajacu) is not savage and does not attack man
or other animals, the variety referred to by Seemann is presumably the Warree or
white-lipped peccary (D. labiatus).

1 Op. cit, 262.

169

MUSICAL INSTRUMENTS.

In treating of this subject we are limited to musical instruments of clay and of
metal, none made of bone, gourd, wood, etc., if ever in existence, having been preserved
to us. Rattles, drums, bells and whistles are included in the list. In addition to
these specialized instruments, the ancient potters were fond of dropping pellets
into everything hollow that might serve to give forth a rattling sound. The legs
of practically all tripods were thus treated, as were many plastic life forms that
adorn the shoulders of vases. By the application of the same principle to the gold
ornaments discussed in another chapter, the bell with free clapper instead of the
rattle proper was evolved.

Battles. — All objects of clay employed primarily as rattles take the form of the
gourd, which, when dried, makes a complete rattle, the prototype of those made
of clay. One example (fig. 273) belonging to the
lost color ware is chosen for illustration. It is
a typical specimen with its globular body and
rather long straight neck or handle, similar in shape
to the rattles held in the right hand of a number
of gold figurines (see PI. XLVIII, fig. e). The
upper part of the handle is a hollow cylinder open
at the end ; its base is solid, with the exception
of a transverse perforation that serves as a means
of suspension. In the upper part of the body, at
a given level, are four slits that communicate with
the hollow interior. The latter does not connect
in any way with the hollow of the handle or its
transverse perforation. This specimen comes from,
Divala. Another specimen, unfortunately broken, is
particularly graceful in form, and has four cruciform
slits in the peripheral zone. The scar left by the
removal of the handle shows that there was no
communication through the handle between the
interior and the outside. The dozen or more oval
pellets of clay inside are worn smooth by use.

Drums. — This type of percussion instrument has
no representative in the collection, unless it is

the specimen figured in Plate XXV (fig. /). A number of the whistles however
have drum-shaped bodies, which fact goes to prove that, among the ancient
Chiriquians, the drum was a familiar object.

Wind-instruments. — Under this head may be grouped by far the largest number
of Chiriquian musical instruments of clay. According to Pinart,1 the musical in-
struments of the present natives (Guaymis) of the region are limited chiefly to the
bone flute and the marine conch-shell. To these he adds the drum made by
hollowing out the trunk of a tree and covering one end with skin. Pinart, who
believes the Guaymis to be the descendants of the builders of the ancient huacals,

1 Alphonse Pinart. Les indiens de 1'Etat de Panama. Rev. d'ethnog., VI, 33, 117, 1887.
MEMOIRS CONN. ACAD., Vol. III. 22

Fig. 273. — Gourd-shaped rattle,
color ware. 'I'

Lost

170 A STUDY OF CHIRIQUIAN ANTIQUITIES.

describes one of their ceremonies, the balza, in which the conch-shell plays an
important role. I quote the author's own words :

Quand un village a decide de donner une balzaria et que 1'epoque en a etc fixee, Ton
expedie des messagers prevenir les autres villages et faire les invitations . . . On invite tout le
monde, hommes et femmes, jeunes et vieux. Suivant les distances a parcourir, chacun par
groupe de famille se met en route de maniere a arriver au lieu du rendez-vous deux jours
avant le commencement des ceremonies . . . Durant le trajet, les invites soufflent de temps
en temps dans de grosses conques pour que leur son fasse connaitre a toute personne
habitant pres du chemin, leur passage et le but de leur voyage.

If these Guaymis are descendants of the ancient race, it is not strange that
the latter should have left behind them so many examples of their wind-instru-
ments. The makers were not satisfied with stopping at the plain flute form, or
even with the ocarina. Their love for the plastic life form is nowhere better
exemplified than in the series of whistles in which the bird form is quite natur-
ally the favorite embodiment. On the other hand, some animals are represented
whose cries bear no resemblance to the sound of a whistle, as may be inferred
from the following list : Man, monkey, puma, deer, tapir, squirrel, ground-squirrel,
iguana, armadillo, crab, scorpion, parrot, owl, duck, partridge, several species of
small birds and other animal forms that can not be definitely determined.

The tail or a leg is usually adapted as a mouthpiece and there is always
provision for suspending the instrument, chiefly by means of a transverse per-
foration through the neck. The finger-holes, generally two
^ in*. jn number, have no fixed position except in the bird forms,

where both are almost invariably on the breast.

While the power and range of these whistles and flutes are
limited, the quality of the tone is often melodious. There are
generally three whole tones, each of the two intervals forming a
major second. Sometimes the first interval is equal to one and
a half whole tones, i. e.. a minor third, making the compass
from the lowest to the highest tone equal to the first four
notes of the scale instead of the first three. The tones are
Fig.277-Doubiewhistie.»/. noted according to high pitch. The lowest tone is produced

with both finger-holes stopped, the middle tone with one

hole open and the highest with both holes open. The holes are usually so
nearly of the same size that it makes no difference which is opened first. The
pitch can be made to vary with the force of the breath. By making judicious
selections, a number of instruments may be played in unison.

In the construction of the whistling apparatus one and the same principle is
always involved, viz., the directing of a stream of air against a cutting edge at
the mouth of a chamber. The working out of this principle in one of its simpler
forms, except that the parts are in duplicate, is exemplified in figure 274. There
is a fusion of two gourd-shaped whistles at the points of contact between the
bodies and at the termination of the necks ; the latter form a two-holed mouth-
piece, thus making it convenient to blow both whistles at the same time. As
these are not of the same pitch, the result is a shrill ear-splitting sound that can
be heard for a great distance. On the other hand, the two tones can be produced

MUSICAL INSTRUMENTS.

171

separately by covering first one and then the other vent-hole. Dr. W. J. Lamson
has about a dozen of these double whistles.

In order to increase the range, it was necessary to pierce the air-chamber by
means of finger-holes. All the other instruments have from two to four of
them, the number of tones producible being always one more than the number
of holes. No account is taken of the tones
produced by increasing the force of the breath.
In the reed-shaped instruments, of which there
are two in the collection, additional range- is
produced by leaving the cylindrical chamber
open at the distal end. Figure 275 is an example.
Its range with the end of the chamber first complete-
ly closed and then entirely opened is expressed
in the scale accompanying the figure.

The distal hole is so near the end as to be of
no use when the latter is open ; otherwise, two
more tones could be added, one with the distal
hole open and one with both open. Every pos-
sible tone between the lowest and highest in this
scale may be produced by only a partial closing
of the end, the pitch depending on the degree
of closing. This instrument is admirable in its
way, but it could have been improved ; first
by placing the distal finger-hole a little farther
from the end, and second, by adding a third
finger-hole nearer the mouthpiece. The colors,
black and red on a cream slip, are perfectly
preserved. The designs are grouped in two
zones and represent the oft-repeated alligator
motives, body-lines and body-markings.

The collection includes three instruments that
differ from the flute type only in the form of the
chamber, which is top-shaped instead of cylindrical,
and in the size of the terminal opening, which
is relatively small. One of these from El Banco
is reproduced in figure 276. It is unpainted and
incised. The surface of the fine-grained paste is
blackened by handling. The vent-hole is at the

base of a solid conical tip, on one side of which is attached a looped fillet of
clay, enclosing for only a part of its way the air-passage leading to the vent.
On the other side is a tiny frog in relief. In addition to the hole at the opposite
end, there are two finger-holes near the periphery, by means of which three
tones may be produced with the end-hole closed and three with the end-hole
open. Were it not for the eccentric position of the vent and the fillet enclosing
the air-passage of the mouthpiece, this specimen might be easily mistaken for a
spindle-whorl.

Fig. 275. — Reed-shaped whistle decorated
with alligator motives. Alligator
ware. '!<•

172

A STUDY OF CHIRIQUIAN ANTIQUITIES.

sed £rut ofien

Fig. 276.

a Fig. 278. b

Fig. 276. — Top-shaped whistle with two finger-
holes ; from El Banco. '/•„

Fig. 277. — Top-shaped whistle with three finger-
holes. Alligator ware. (After Holmes.) '/>

Fig. 278. — End views of instrument shown in pre-
ceding figure. (After Holmes.)

1 Op. cit, figs. 245-246.

Of the other two specimens belonging
to this general type one is unpainted and
incised like the foregoing. It also has
but two finger-holes. The easily break-
able part of the mouthpiece is lost, thus
increasing the general resemblance to a
spindle-whorl. The third instrument be-
longs to the alligator group of ware. In
general shape and in the character of the
painted designs, it is so much like the
one reproduced by Holmes,1 that I have
borrowed, his illustrations (figs. 277 and
278). The number of tones that can be
produced is the same for both. The range
of the Yale specimen is given in the ac-
companying staff, and is exactly the first
half of the diatonic scale.

End closed- End o/^en

With possibly a single exception, there
are no drums in the collection ; but
there are a number of drum-shaped whist-
les. One of the simplest of these is of
plain biscuit ware. The chamber is not
quite cylindrical, the diameter of one end
being slightly larger than that of the other.
The smaller end is surmounted by a plastic
animal figure.

The combination of drum and bird is
seen in figure 279. Here the drum is more
nearly conical than cylindrical. Projecting
from its convex surface are the head and
tail of a bird, the tail serving as a mouth-
piece. The neck of the bird is perforated
for suspension. Two finger-holes are
on the breast of the bird, this number
being constant unless otherwise stated.
The gently concave base is handsomely
decorated and alligator motives cover the
ventral surface of the bird.

Sometimes the drum is surmounted by
grotesque life forms or combinations of the

MUSICAL INSTRUMENTS.

173

same, as in figure 280. In this case, the diameters of
the drumheads are equal and somewhat larger than
the central diameter. The end serving as a base is
not decorated. The whole serves as a pedestal for a
biped that is neither man nor ape nor bird, but in
some respects resembling all three. An elaborate head-
dress, painted and in relief, extends the length of the
life form, downward and backward till it comes in
contact with the margin of the drumhead forming the
mouthpiece of the whistle. The perforation in the
breast has no connection with the whistling apparatus.
This instrument is exactly an octave lower in tone
than the preceding.

The same general scheme, but with additional
life forms, is repeated in figure 281. The undecorated
end on which the drum rests is perfectly flat and is
perforated in the center by one of the two finger-
holes. The panels on the sides are covered with
familiar alligator motives, as in the preceding. The
principal figure in the group on top is half man, half
beast. Its arms extend back till the hands touch the
arched body of a two-headed serpent. Attached to this
body near the center of the arch are two small heads,
bird or reptile. There is also a prop extending from
the mouthpiece to the center of the arch against which
the standing figure leans.

In by far the greater number of cases the air-chamber
of these wind-instruments is but the hollow body of
some life form, the bird being the favorite. Among
birds, a number of varieties are met with, the one
most frequently represented being the parrot. Pinart
found that the totemic figure the most often used
among the Guaymis is that of a small species of
parrot, in honor of which he heard many chants.
Lionel Wafer,1 one of the early writers, in discussing
animal life on the Isthmus says : " They have Parrots
good store, some Blue, and some Green, for Shape
and Size like the generality of the Parrots we have
from Jamaica. There is here great variety of them,
and they are very good Meat."

Attention is again called to the series of bird forms
already noted in the discussion of the alligator group
of ware. This avian character originated in a dis-
tinctly globular vase (see fig. 246) having aperture

Fig. 279.

Fig. 280.

Fig. 279. — Whistle combining drum
and bird form. Alligator ware. '!>

fig. 280. — Drum whistle surmounted
by grotesque biped life form. Alli-
gator ware. l>

1 A new voyage and description of the Isthmus of America. 2d ed., 91, London, 1704.

174

A STUDY OF CHIRIQU1AN ANTIQUITIES.

and rim ; the tail and wings of a bird in relief about its periphery ; and the head, a
mere protuberance on the side opposite the tail. Then followed a further differentia-
tion, especially in the change from annular base to the bipedal condition (see fig. 247).
Lastly the painted protuberance on the breast disappeared and a realistic removable
head was made to fit snugly over the rim on top (see figs. 248, 249, 249 a). These
birds represented parrots. We have among the whistles the same parrot form
(PI. XL VII), but much smaller in size and with non-removable head. A closer study

of figure a reveals the fact that the head, though
firmly united with the trunk, has the air of being
detachable. This impression is heightened by the
separation of the painted designs ; that on the head
and neck from that on the body. The genetic re-
lationship goes back a step further, as witness the
red spot on the pointed breast (compare with figs. 246
and 247).

That these are what might be called phylogenetic
rather than fortuitous characters, is proved by their
repetition. In figure 6, one sees again the vestigial
head on the breast as well as the removable looking
head on the neck. Attention is also called to a
painted design on the back that resembles the letter
T. This same design also appears on other spec-
imens. The tones produced by this whistle are the
same as those of the preceding example.

In figure c, the head has the appearance not only
of being removable but also of having been revolved
on its axis nearly 180°, as if to emphasize this feature.
There is also the red spot on the breast and
the T on the back. Figure d is very similar to
the two preceding, except that the red spot has
disappeared from the breast, the latter at the same
time losing its pointed character. This instrument
produces the same tones as those represented in
figures 279, 298 and 301, being just an octave higher
than that illustrated in figure 280.

The head still sits lightly on the body in figure e, a specimen from Jacu. Here
again we find the pointed breast, marked this time by a circle of black paint.
The wings are outstretched as if ready for, or in the act of, flight. Instead of
the T on the back, there is an attractive design in black representing some life
form, a motive found also on Nicaraguan pottery. A bird of a somewhat different
type is represented in figure /. The neck is shorter and the relatively longer
beak rests against the breast. All these bird-whistles can be made to stand firmly
on their feet without toppling over.

A species of water-bird is probably intended in figures 282 and 283. In the
latter the beak is bent abruptly to one side and is in contact with the breast

Fig, 281. — Drum surmounted by half-
human figure combined with serpent
and bird forms. Alligator ware. '/•

MUSICAL INSTRUMENTS.

175

for its whole length. In the former the entire head is turned to one side, the
end of the beak resting on top of the right wing.

The characteristic features of the owl are well rendered in figure 284. Although
the head is proportionately too large, the effect on its anatomy of twisting the
neck to one side speaks in every line. One almost forgets that it is made of

Fig. 282. — Bird-shaped whistle. Alligator ware. '/•

Fig. 283. — Bird-shaped whistle decorated with
scale-group symbols. Alligator ware. '/«

Fig. 284. — Whistle representing an owl.
ware. V«

Alligator Fig. 285. — Owl-shaped whistle decorated with mul-

tiple body-line motive. Alligator ware. '/«

anything so unyielding as clay ; yet there is the sharp demarcation between head
and trunk in respect to the disposition of the delineating colors.

The owl also comes in for treatment in figure 285. A rear view is given in
order to show the multiple alligator motive on the back. The head is turned to

176

A STUDY OF CHIRIQUIAN ANTIQUITIES.

the right as in the preceding, but it is a head of clay set on the shoulders
rather than growing there. It should be mentioned here that nearly all the whistles
figured belong to the alligator group of ware.

A species of crested partridge is represented in figure 286. The head and neck
are continuous with the body from the standpoint both of modeling and painted

decoration; one of the rare ex-
ceptions that prove the rule. This
whistle produces exactly the same
tones as those produced by the
whistles illustrated in Plate XLVII
(figs, a and b). It would be difficult
to determine the kind of bird rep-
resented in figure 287, an example
of alligator ware from Divala, in
which smoke and much handling
have almost completely obscured
the delineating colors. In it the
two reversionary characters already
noted are once more expressed.

A species of duck, probably the
teal, is reproduced in figure 288.
Note the general attitude of the
head, which is turned so that the
long bill may have the protection
of continuous contact with shoulder
and wing. Fig. 289 is not a very
effective whistle but a most inter-
esting piece of modeling, as it
represents a bird in the act of
alighting. This is told in the at-
titude of feet, wings, neck and
head, as well as the tail. The
vent-hole is placed dorsally instead
ofventrally. The ware is unpainted ;
the wing feathers are indicated by
incised lines and the body is marked
by numerous small annular inden-
tations. Two of these, a little larger
than the others, communicate with
the hollow interior, thus forming
finger-holes. They are so small
however as to be of little use.

There is a single specimen of armadillo ware among the bird forms. The
modeling is indifferent. The head and feet are reduced in size, and the wings
are entirely wanting. The finger-holes are placed on the sides, an exception to
the rule. The three tones are full of melody. The simplification of parts is

Fig. 286. — Whistle representing a crested partridge,
ware. '/'

Alligator

Fig. 287. — Bird-shaped whistle ; from Divala. Alligator ware. '/'

MUSICAL INSTRUMENTS.

177

marked in the remainder of the series. A good example of this is shown in
figure 290, where the danger of breaking is reduced to a minimum. The range
is greater than usual, the first interval being a perfect fourth and the second, a
major third. On the neck of a duck without head, feet or wings, there is
perched a much smaller duck having a pair of heads (fig. 291). This grotesque

Fig. 288. — Whistle representing a duck. Alligator
ware. '/«

Fig. 289. — Whistle representing a bird in the ac t
of alighting. '/'

Fig. 290.— Simplified bird form of whistle. Alli-
gator ware. '/•

Fig. 291. — Composite bird form of whistle. Lost
color ware. '/•

Fig. 292. — Small bird-shaped whistle,
ware. '/'

Lost color

Fig. 293. — Small bird-shaped whistle,
ware. '/>

Lost color

specimen belongs to the lost color ware. The small whistles reproduced in
figures 292 and 293 also belong to the lost color group. The smallest bird-
whistle in the series has a maximum length of only 3.8 centimeters.

The crab form seems to have appealed to these ancient whistle makers. The
forms are varied enough to suggest familiarity with several varieties. One of the
finest examples is shown in figure 294. The designs painted on the back, in-
MEMOIRS CONN. ACAD., Vol. III. 23

178

A STUDY OF CHIRIQUIAN ANTIQUITIES.

eluding the multiple body-line of the alligator, with accompanying dermal mark-
ings, are in a fine state of preservation. The left (and this is true of the whole
crab series) foreleg is converted into a mouthpiece.

A somewhat different type of crab is given in figure 295, which represents one
of the very few whistles not provided with some means of suspension. No attempt
was made to account for the exact number of legs in any of the crustacean forms.

Fig. 294. — Whistle representing the crab, the back
being decorated with alligator motives. Alligator
ware. '/«

Fig- 29S- — Whistle representing the crab. Alli-
gator ware. '/«

Kig. 296. — Whistle representing a round-bodied crab.
Alligator ware. '/«

Fig. 297. — Whistle representing a scorpion, with
alligator motives decorating the sides. Alligator
ware. */«

They are all so fashioned as to stand on four legs like a quadruped, but also
have an additional first pair armed with large claws like chelipeds and serving as
weapons rather than as supports. The tones produced in these two crab-whistles
are exactly the same. The artist was often very successful in giving concrete
expression to an idea. This is seen in the representation (fig. 296) of a round-
bodied crab just ready to strike. The left foreleg, converted into a mouthpiece,

MUSICAL INSTRUMENTS.

179

Fig. 298. — Whistle in the form of a coiled snake.
Lost color ware. '/'

emulated so closely the head and beak of a bird that eyes have been painted on,
thus heightening the resemblance. The chelipeds are both perforated for suspen-
sion.

The scorpion was likewise given only four legs in addition to the first pair
(fig. 297). The long tail is flexed sharply till it comes in contact with the back,
forming in this manner a loophole by which the figurine may be suspended.
The sides are decorated with the multiple body-line of the alligator and its scale
symbols.

Among reptiles, there is a good represen-
tation of a coiled snake with laterally
flattened head (fig. 298). The mouthpiece
is built on the lower ring of the coil, the
end of the tail being hidden underneath
the mass. The specimen belongs to the
lost color ware, and the decoration has been
almost completely lost through the rubbing
off of the black paint.

There is a composite form representing
an iguana and snake, the flat-headed snake
forming a crest the entire length of the
iguana (fig. 299). The neck of the snake
is arched, forming a loophole for suspen-
sion ; its body and tail fuse with the
crest of spines along the back and tail of
the iguana. The dewlap is conspicuous.
The right forefoot is converted into the
mouthpiece of the whistle, which is capable
of producing three clear and musical tones.
A panel on each side of the whistle is or-
namented with the oft-recurring multiple
alligator motive.

Various quadrupeds are included in the
present series. The form can usually be de-
termined by the successful rendering of some
distinctive character. When doubt exists, it
is as much due to our lack of knowledge of
isthmian fauna as to any shortcoming on

the part of the artist. Figure 300 is an example of one of the doubtful cases.
In figure 301, more attention is given to detail, especially in the modeling of the
head, which is partially separated from the body by a constriction. The most
plausible explanation of this is seen in the series of birds with removable heads.
This specimen, like the preceding, belongs to the lost color ware.

One is impressed by the sincerity of the attempt to represent some tardigrade
species (fig. 302). This instrument is apparently the work of an apprentice unaided
by any knowledge, technical smartness or trickery, and as such is full of interest.
It belongs to the armadillo ware. There is a single clay pellet in each hollow

Fig. 299.— Composite form consistig of the iguana
and snake, the sides being decorated with the
multiple alligator motive. '/«

180

A STUDY OF CHIRIQUIAN ANTIQUITIES.

leg, also one in the cavity common to both head and neck, so that we have to
do with a rattle as well as a whistle. The representation of the armadillo (fig. 303).

Fig. 300.— Whistle in the shape of a quadruped.
Lost color ware. '/«

Fig. 301. — Whistle in which the head has the ap-
pearance of being removable. Lost color ware. '/<

although not more realistic than the foregoing, evinces a higher degree of pro-
fessional skill. The characteristics of the carapace are brought out by relief as

well as by the red and black
colors common to all alligator
ware. The neck is perforated
for suspension.

The primitive artists may
or may not have been divided
into incipient schools of
thought and expression. If
they were, the influence of
academic canons must 'have
been strong. The animal in
repose, that is to say not in
action, was the chosen model.
Once in a while however an
artist breaks the bonds of
classic restraint and produces
new and admirable results,
as in the case of the bird
alighting. Another notable
example may be seen in the
squirrel (fig. 304) feeding
on a nut that is held be-
tween the uplifted forefeet,
the body lifted and supported
by the hindlegs. The mul-
tiple alligator motive is paint-
303.— Whistle representing the armadillo. Alligator ware. '/« ed On both'sides of the body.

Fig. 302. — Figurine serving as both whistle and rattle. Armadillo ware.'/*

MUSICAL INSTRUMENTS.

181

It will be seen from the accompanying staff that the tones are the same as in
two of the crab-whistles (see figs. 294 and 295).

Rather infrequently, locomotion is expressed by extending the legs both forward
and backward, as in figure 305, which also is evidently intended for a squirrel ; but
the sense of motion is not imparted to the rest of the body. The characteristic

Fig. 304. — Whistle representing a squirrel feeding,
the sides being decorated with the multiple alligator
motive. Alligator ware. '/'

Fig. 305. — Whistle representing a squirrel in motion,
the alligator motive appearing in the panels decorat-
ing the sides. Alligator ware. '/•

sharply recurved tail extends forward almost to the back of the head. One finger-
hole is placed ventrally and the other on the left side. Attention is called to the
alligator motive in the panel that covers each side. At least four species of squirrel
are found in Panama.

An effort to avoid the stereotyped form is sometimes seen in the turning of
the head, as if to look back or to one side. This is true of the striped creature
reproduced in figure 306. The mark-
ings of the body suggest the ground-
squirrel, while the head and mouth are
more like those of the cat tribe. This
instrument agrees in tone with that in
figure 299. Similar longitudinal stripes
are present in figure 307 ; but in this
instance the nose and mouth are more
like those of the tapir. The two finger-
holes are placed on the shoulders. The
tones of this whistle are the same as
those in figures 300 and 305 and an
octave lower than those in figure 292.

There is no mistaking the intent of
the one who modeled figure 308, al-
though the end of the tapir's nose has been broken. The short lower jaw is
indicated by a slit on either side. The skill in modeling is surpassed only by that
in the application of the paint. With the exception of a red spot under each

Fig. 306.- — Whistle suggesting the ground-squirrel, though
the head is catlike. Alligator ware. '/'

182

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Fig. 307. — Whistle with tapir-like head. Alligator ware. 'I'

ear, black alone is used. The series of lines on the head, back and legs are
fine, clean-cut, straight and parallel. This is particularly true of the five longitudinal

lines along the back that seem
almost too perfect to have been
done by free hand. The char-
acter of the painted decoration
on this figurine may be taken
as a clue to the particular species
of tapir with which the ancient
Chiriquian was most familiar.
Tapirus dowii, when full-grown,
is a nearly uniform blackish
brown. It is said to be unspotted
when young. On the other hand,
Tapirus bairdii, when young, is
irregularly marked with white
spots and stripes. The whistle
in question, therefore, evidently
represents the young of Tapirus
bairdii. The larger vesicular forms
of the tapir (see PI. XLI1I, fig. c
and text-fig. 245) may also be
referred to the same species and

§, j J~r/;«||| | MP' \ im to the adult stage after the coat-

' * ^^^liJF ^9 ing nas become a uniform color.

The right foreleg forms the mouth-
piece to the whistle. The two
finger-holes are placed at the
anal opening and on the left
shoulder, respectively.

A species of tiger-cat is rep-
resented in figure 309. It is ap-
parently in motion and the at-
titude of the head, which is hollow
and supplied with a clay pellet,
adds interest to the composition.
The decoration is unusually well
preserved. The range from the
lowest to the highest tone is much
less than in the preceding whistle.

Peculiar interest attaches to
the frequent representations of

Fig- 3°9- — Whistle representing a species of tiger-cat, the head ., , ,-,, .

also serving as a rattle. Alligator ware. '/• the Jaguar> "Ot Only among Chl-

riquian whistles, but also as plastic

ornaments on zoormorphic forms of vessels (see PI. XLI, figs, c and d; PL XLIII,
figs, a and b ; text-fig. 244). In Mexican mythology, the jaguar is intimately

Fig. 308. — Whistle evidently representing the young of Tapirus
bairdii. Alligator ware. V*

MUSICAL INSTRUMENTS.

183

associated with musical instruments. It appears in the form of Tepeyolldtli, the
"Heart of the mountains," the eighth of the nine lords of the night hours.1
Tepeyollotli is also met with as lord of the third day-count and of the third
Tonalamatl division, where he is represented in the form of a jaguar and is ex-
plained as the "Echo in the mountains" (el retumbo de la voz, cuando retumba
en un voile de un cerro al otro). The roar of the jaguar is thus brought into
direct association with the "Echo in the mountains." In a Codex Borgia
figure, Tepeyollotli blows a shell-horn. In other cases the shell is worn as a
breast ornament, or is otherwise associated with the ged. Tepeyollotli is also
the name of th'e Mexican god that is combined with the drum, just as we found
deities associated with some Chiriquian drum-whistles (see figs. 280, 281). Equally
significant is the double-headed jaguar shown in figure 310. In each mouth, there
is what might well be considered a musical instrument held to the lips by means
of the forefeet. Holmes, who figured a whistle with four jaguar heads, spoke of

Fig. 310. — Whistle representing a double-headed jaguar, both necks being perforated,
the heads serving as rattles. Alligator ware. '/«

the objects held to the mouths, as tongues. Such is not the case with the Yale
specimen, because the object extends into the mouth only as far as the front
teeth. Moreover, each instrument is held in place by the fingerlike toes of the
forefeet. We have here therefore the counterpart of the Mexican Tepeyollotli,
symbolizing the " Echo in the mountains." The two finger-holes are both on the
opposite side of the figure. The tones are the same as in the preceding whistle
and are full of melody. The paint near the finger-holes is much worn, evidently
from use. Both necks are perforated. The ear-holes lead to the head cavities,
each of which is supplied with clay pellets.

Another smaller specimen of the jaguar, with mouth open as if uttering a cry,
is shown in figure 311. No instrument therefore is necessary to suggest the
" Echo in the mountains." It should be noted in this connection that the Chiri-

1 Codex Borgia; and Codex Vaticanus, no. 3773, Seler's elucidation, 103.

184

A STUDY OF CHIRIQUIAN ANTIQUITIES.

quians almost always represented the jaguar as in the act of roaring, a fact of
special significance, when viewed in the light of our knowledge of the role played
by the jaguar in Mexican and Mayan mythology. According to Dr. W. Stempell,1
it is figured in all four of the known Mayan codices. The tail of the specimen
in question is curved to one side the end touching the hip. The patterns in
black and red on zoomorphic forms belonging to the alligator ware are, as a rule,
either purely geometric or referable to the alligator motives rather than to the

fig. 311.— Jaguar-shaped whistle with ocellated mark- Fig. 312. — Whistle in the shape of a deer. Alligator

ings of the jaguar. Alligator ware. '/« ware. '/•

Fig. 313. — Whistle representing a deer apparently making
an attack. Alligator ware. */«

Fig. 314. — Monkey-shaped whistle with three
finger-holes. '/'

body-markings of the animal under consideration. But in these two jaguar-whistles,
the markings are true to nature — rings enclosing spots. This and the following
whistle are alike in pitch.

Among ruminants Seemann found the Venado, a species of deer, roaming in
herds over the savannas. This animal is probably represented in figures 312 and
313. Cariacus virginianus is the species found in Panama. Special attention is

1 Die Tierbilder der Mayahandschriften. Zeitschr. f. Ethnol., XL, 704, 1908.

MUSICAL INSTRUMENTS.

185

given to the head in each. The teeth are rather conspicuous. In both figures
one finger-hole is placed on the belly and the other on the side, the left side
in the smaller of the two. The larger one seems to be making an attack, the
head being turned so as to bring the horns into place. The tones of this whistle
are particularly pleasing.

One of the most realistic forms is that of the monkey shown in figure 314.
The paste of which it is composed is coarse and dark, the surface being covered
for the most part by a red slip on which there are no delineating colors. The
piece therefore does not belong to the alligator ware as is the rule for the
whistles;. The legs are broken. The long fingers of the two hands grasp either
end of a cylindrical bar, the middle part of which is held in the mouth. The
bar is perforated transversely at a median point, as if it were intended to represent
a whistle. There are three finger-holes to the air-chamber ; two dorsal and one
ventral, so that four tones ma}' be
produced without resort to more
than a partial stopping of any of
the holes by means of special finger-
ing.

FigureSlo represents an exception-
ally well preserved piece. The at-
titude of the body is birdlike. The
feet are characteristic for both bird
and man or monkey. The arms
are those of either man or monkey,
and the head is apelike. Something
resembling a fish is held in the hands.
The mouth is open, revealing the
teeth. There is a motive similar to
this in Peruvian art — a bird hold-
ing a fish — found especially at lea.
An analogous idea is expressed in

one of the gold ornaments (see PI. XLIX, fig. a), where the object held to the
mouth may be a fish or a conch-shell. One of the finger-holes in our specimen
is ventral, the other on the left shoulder. The tones are exceedingly clear and
musical. Although the painted designs are broadly similar on the two sides, there
is no striving after exact bilateral symmetry in this respect. Both red and black
are employed on the right side of the head and neck, while on the left side
black only is used and with slight variations in the design.

Another combination of man and beast is shown in figure 316. There being
no tail, the right arm is converted into a mouthpiece. One hand or something
held in it is pressed to the mouth. The left arm is akimbo with the hand held
against the breast near a small protuberance resembling a nipple. The two finger-
holes are placed dorsally on the periphery. The back of the head and neck is
marked by black lines that reach to the shoulders — the characteristic way of
representing human tresses.

The form reproduced in figure 317 is presumably that of a monkey, although
MEMOIRS CONN. ACAD., Vol. III. 24

Fig- 3l5-~~ Figurine serving as a whistle and representing a
mythical form with mixed attributes. Alligator ware. ''•

186

A STUDY OF CHIRIQUIAN ANTIQUITIES.

the pose is avian. Both arms are bent sharply backward, the right hand grasping
the end of the tail and the left resting against the back. The black color is
used much more extensively than the red, the latter appearing only on the throat
and as three longitudinal bands on the breast. There are duplicates of this form
in the collection.

One of the most interesting of the whistle figurines is a three-headed monster
with human attributes (fig. 318). The central head is the largest; the others,
alike in size, are situated at the angle of the shoulders. The group suggests the

?ig. 316. - Figurine serving as a Fig. 317. — Whistle presumably rep-
whistle. The pose is human ; the resenting a monkey, although
right arm is converted into a the pose is avian. Alligator
mouthpiece. Alligator ware. '/• ware. '/•

Peruvian pottery figurine of Tunapa placed between

his two sisters. The legs are short, body long and

attitude erect. The two finger-holes are in front, and

the mouthpiece stands out from the back at nearly

right angles. Human tresses are represented in black

on all three heads, and in each of the three mouths

is an object resembling a protruding tongue, or an

instrument of music similar to that held in the mouths of the double-headed

jaguar (see fig. 310). A multiple alligator motive encircles the common body.

It has been already mentioned that the ancient Chiriquian artist was not very
successful in his portrayal of the human form and features. Figure 319 is an
exception that proves this rule. It consists of the head only, with the hair so
arranged as to form a projection on the back for the mouthpiece to the whistle.
One finger-hole is in front of the right ear and the other on the throat. The
modeling is done with a high degree of skill and knowledge of facial topography,

Fig. 3 1 8. — Whistle figurine represent-
ing a three-headed monster with
human attributes ; the multiple al-
ligator motive encircles the body.
Alligator ware. */<

MUSICAL INSTRUMENTS.

187

except that the size of the mouth is somewhat exaggerated. The lips are parted,
revealing the teeth, which are closed. The region about the mouth is painted
red. Alternating groups of black and red longitudinal lines cover the face from
the hair down to the level of the nostrils. There is a black pattern about the eyes
not unlike a pair of goggles, except that there is no connection over the bridge
of the nose. The black hair is combed
back and brought together forming
the mouthpiece to the whistle. Strange
to say, no effort was made at model-
ing -the ears, which stand out at
right angles to the sides of the head.
The left ear is perforated for sus-
pension.

In figure 320 the artist reverts once
more to the primitive stereotyped
form resembling the head of a ro-
dent or of a bird almost as much as ^^^KBb^^Sii W^ 0
it does that of man. The mouth-
piece at the back is treated as if it
were the tail of an animal. One
finger-hole is in front of the right
ear and the other near the top of
the head on the same side. As in

the preceding figure, the face is marked by alternating groups of black and red
lines, which in this case are wavey instead of straight. Much ingenuity was
exercised in representing the eye by means of intaglio.

Kigurc 319. -Whistle representing a carefully modeled human
head. Alligator ware. V*

gird

ig. 320. — Whistle representing the human head.
Alligator ware. '/«

Fig. 321. — Diminutive tripod with bird-shaped whistle
attached to one side. Armadillo ware. '/«

The series of whistles ends with a diminutive tripod bowl (fig. 321), to one side
of which is attached an air-chamber with mouthpiece. This whistling apparatus
takes the form of a bird, the head of which has been broken off. The very
small finger-holes are placed on either side of the missing head. The tones

188

A STUDY OF CHIRIQUIAN ANTIQUITIES.

are the highest of any in the entire group of whistles, being just an octave
higher than those of the instrument reproduced in figure 318. The piece, which

comes from Escaria, belongs to the armadillo ware.
A small cup (fig. 322) is described here because
there is attached to its shoulder a make-believe
whistle. The latter is shaped like a fish, except
that head and tail are both lifted ; the end of the
latter is slit to resemble a mouthpiece, and near its
base is another incision where the vent ought to be.
This cup, the surface of which is discolored by

Fie. 322. — Small cup with false 111 • i j i e

whistle attached to the shoulder. •/• smoke, belongs to the unpamted class of ware.

METAL.

When America was dicovered, its races were still on the. threshold of the metal
age. In regions favored by the presence of gold, silver, copper and their alloys,
however, certain peoples had already achieved remarkable results in metal work.
This is particularly true of the region extending from Mexico on the north to
_ Peru on the south. If the thirst for geographic knowledge impelled Columbus
to make his voyages of discovery, hunger for gold was the mainspring of the
Spanish Conquest which followed. Barbaric wealth of gold and the barbaric point
of view as to its relative value served to turn the heads of the early explorers,
leading eventually to the practical extermination of the conquered and the ultimate
downfall of the conquerors.

Curiously enough, Columbus saw very little of the mainland of the Western
Hemisphere — only a part of the northern coast of South America, which he ex-
plored on his third voyage ; and the coast of Central America from Guanaja
Island, off Honduras, to a point about half-way between Colon and the Gulf of
Darien, which he discovered on his fourth and last voyage. It was Columbus,
therefore, who discovered Chiriqui and who was the first to observe the gold
and copper ornaments worn by the natives. When on September 25th, 1502, the
expedition cast anchor near the mouth of San Juan de Nicaragua, Columbus noted
that some of the natives wore plates of low-grade gold (guanin) and others wore
jewels of the same metal suspended about the neck.

The coast of Chiriqui was reached in October, the ships of Columbus entering
Almirante Bay (see map) through Bocas del Toro. At a port on one of the islands
in the bay, the explorers saw twenty canoes anchored. According to Acosta's1
account, " the natives wore plates of fine gold suspended from the neck." As
these were the first examples of pure gold the Spaniards had seen along that
coast, they took by force from two of the natives the ornaments which the latter
had refused to exchange for Spanish trinkets. The gold ornaments taken from one
weighed twenty-two ducats ; those taken from the other, fourteen. The inhabitants
assured the Spaniards that gold was to be found at a number of localities to the
south and west, one of which was called Veragua. This name seems to have
taken a firm hold on the imagination of the explorers, becoming in their minds
the synonym of wealth. Columbus called this coast Costa de los Contrastes and
afterward Costa Rica and Costa de Veragua. A few years later the name Castilla del
Oro was given to the entire isthmian region.

On their way eastward, after leaving Almirante Bay, the explorers entered a
number of rivers, where they were generally successful in exchanging worthless
trinkets for gold ornaments. At Puerto del Betrete, Columbus decided to return to
Veragua in search of gold. Ascending the Urira river, gold mines were found at
its head waters. The explorations were extended in the direction of Bocas del
Toro as far as Cateba, and netted much golden booty obtained from the Indians.

1 Joaquin Acosta. Compendio historico, segunda ed., 4, Bogota, 1901.

190 A STUDY OF CHIRIQUIAN ANTIQUITIES.

In his explorations on the Isthmus a few years later, Vasco Nunez de Balboa
was even more successful in wresting from the native chiefs their golden treasures.
From Chiapes he received " five hundred pounds weight of wrought gold," and
from another cacique, Tumaco, "jewels of gold weighing six hundred and fourteen
crowns." l On his return trip across the Isthmus, Balboa's net booty from three
chiefs and their followers amounted to 14,000 crowns.

Seemann speaks of copper and gold being found all over Panama. He states
that as long as the Spaniards retained possession of the country, the extraction
of gold was carried on to some extent, certain mines being very productive.
" The most important were, and still are, those of the Mineral de Veraguas. The
gold is found there on the plains, and large pieces are also obtained from the
beds of rivers and rivulets." The mines of Estrella in Chiriqui are mentioned as
having been celebrated and as still holding a place in the traditions of the country.

J. H. Smith,2 a contemporary of Seemann, states that tradition bears witness
to the auriferous richness of the lands sloping from the volcano of Chiriqui toward
Punta Burica. He locates the gold mine of Tisingal, one of the richest ever
worked by the Spaniards, as being "behind the volcano of Chiriqui and among
the range of the spurs of the Cordillera, forming the northern limits of Burica."
According to the same author, gold is found at Guanavano, Charco Azul and in
an extensive quartz formation at Las Brenas that is visited regularly by the Indians
of Terrora, who grind the rock and extract the gold. This locality is on the road
leading from the territory of Burica to Costa Rica.

In a recent article Mr. Charles Melville Brown8 also discusses Tisingal. He
says that somewhere on the immense slopes of the volcano of Chiriqui, probably
in the district of Bugaba, " lies the lost mine of the Indians, ' Tisingal,' known to
and worked by the early Spanish settlers, who changed its name to ' La Estrella.' "
Brown also states that during the year 1833-'34, some old official documents
relating to this mine were found in the archives at Cartago, Costa Rica. He
gives a translation from one of these documents, which will throw some light on
the possibilities of placer mining by the Indians of Chiriqui:

" In 1587 Don Diego de Sojo, Captain of the Guard of the Governor of Veraguas,
set out from Santiago de Veraguas to explore this little-known part of the King's
province, and in his report to the Governor of Veraguas he says :

The quantity of gold that abounds here is great and of good karat, as can be
seen from the plates of gold the Indians beat out, it not being alloyed with other
metals. The rivers abound with gold, besides there being other precious metals
in all the ranges of the Province extending over an area of 20 leagues on the
shores of Almirante Bay, called thus as it was discovered by Admiral Colon
(Christopher Columbus) ; being on the very coast of Veraguas, a distance of some
15 or 20 leagues from the Escudo de Veraguas (a small island near the mouth
of the Chiriqui Lagoon on the Atlantic).

1 Irving. Life and voyages of Christopher Columbus, III, 147, 1892.

2 Observations on the territory of Burica, in the province of Chiriqui, Isthmus of Panama.
Jour. Roy. geographic soc., XXIV, 257, 1854.

3 Tisingal, The lost mine of Panama. Bull. Inter, bur. Amer. republics, XXX, 424, 1910.

METAL.

But the greatest quantity of gold exists in the hills of Corotapa, on the shores
of the same bay near the banks of the Rio de la Estrella (River of the Star — now
supposed to be the Changuinola River), a prodigious river and the richest in the
world, whose sands are of gold ; defended and guarded by a bellicose nation that
lives along its banks at its mouth at the place known as Horobaros . . .

And the Indians extract gold with calabashes in very large grains, and a cacique
of the same town named Ucani works it into the said pieces . . .

From these same hills Captain Munoz, Sergeant-Major of Don Perafan de Ribera,
Governor-General and Captain that he was of Costa Rica — took from the tombs
of the dead, which he found one league inland from the coast, such a great
quantity of gold as to swell two large chests of the kind in which shoes and
nails for the cavalry are brought over from Castile. And being covetous of more
treasure he started inland with 60 men he had with him, leaving the two chests
buried at the foot of a ceiba tree, well locked and nailed, and started inland in
search of the Indian village. But after having traversed hardly a league he was
attacked by such a number of natives that some of his companions were killed,
he turning and fleeing pursued by the natives to the very waters of the sea; and
with difficulty did he embark in his frigate and escape, leaving his heart buried
at the foot of the ceiba tree where he had left the chests of gold, and where
they remain to this day."

According to Brown the search for the golden treasure of Tisingal still continues.
" During the dry season of 1909 there was discovered about four days ' journey
from the present town of Bugaba a l guacal ' or burying ground, from which there
has been taken over $25,000 worth of gold up to the present time. This ' guacal '
is located in the district of Corregidor, and as many as 400 natives have been
at work in it at one time. Don Antonio Anguizola, Governor of Chiriqui, has
sent out a party of 20 men in search of other ' guacales,' and it is not improbable
that during the present dry season others will be located . . . The gold ornaments
are usually found in the bottom of the grave, arranged as though they had been
on the breast of the body at the time of burial. Sometimes the ornaments are
found in one of the ' cantaros ' or little drinking jars, and in several graves recently
opened it was found under the second slab. A few years since an Englishman
in Boquete opened a grave from which he took over $2,000 worth of bullion."

In his report to Captain F. Engle, Dr. John Evans,1 geologist of the Chiriqui
Commission, states that on the tributaries of the Cricamola river several panfuls
of earth were washed by one of his men, and in every instance gold was found.
The aggregate value of the gold in three panfuls was about one dollar, which is
considered a large yield. Cricamola is a stream that flows from the Valle Miranda
northward into the Chiriqui lagoon. In other localities, Dr. Evans found ores
of iron, copper and platinum. Judging from the geology of the country and the
discoveries previously made, he believed it to be rich in minerals.

M. A. L. Pinart 2 believes the Guaymis who inhabit Valle Miranda to be
descendants of the Indians that constructed the ancient cemeteries (huacals)
found everywhere over Chiriqui, Veragua, Azuero and Cocle. They have a

1 New York Herald, Dec. 8, 1860.

2 Bull. Soc. de geog. 7° ser., VI, 433, Paris, 1885.

192 A STUDY OF CHIRIQUIAN ANTIQUITIES.

tradition that before the arrival of the Spaniards and even for a certain period
after that event, they worked in metal — gold, copper and their alloys. The
Guaymis of Valle Miranda still possess a number of ornaments in these alloys,
which they claim to have inherited from their ancestors — ornaments that differ in
no respect from those found in the huacals of Chiriqui. Pinart also believes that
the Guaymis still bury their dead in these ancient huacals. Gabb observed
similar gold ornaments in the possession of a Tiribi chief and one of his principal
warriors, but was not able to ascertain whether these highly prized images had
been taken from the ancient graves or handed down as heirlooms (see p. 16).

Interest in the ancient burying-grounds of Chiriqui may be said to date from
1858, when rich discoveries of gold were made at Bugavita. While wandering
through the forests in the vicinity of his cabin, a native of Bugavita, in the district
of Boqueron, encountered a tree that had been uprooted by a recent tempest,
revealing a small earthen vessel. " Upon examination this proved to contain,
wrapped in a swathing of half-decayed cloth, divers images of curious and fantastic
shape, and of so yellow and shining a metal that he at once suspected them to
be of gold." 1 Authors differ as to the details of the discovery. Dr. Otis says it
occurred in June and that in less than a fortnight over a thousand people were
at work, having dug up 225 pounds weight of images, most of which proved to
be of the finest gold. According to the report of Dr. J. King Merritt,2 who was
director of a gold mine in Veragua, it was in the autumn of 1858, while gathering
their crop of corn, that two Spanish Creole farmers, Ignacio Guerra and Victorio
Pitti, " accidentally discovered a golden image, which had been exposed by the
uprooting of a plant. They cautiously and secretly made farther search by sink-
ing a pit at this point, and were successful in obtaining more. With this en-
couragement they determined to explore the Huacal, the existence of which had
been well known for years. They proceeded in this work, not continuously, but
as circumstances would permit, until the first of May, 1859, when their concealed
operations became known to the inhabitants of the vicinity. The consequence
was that by the middle of May more than a thousand persons were engaged in
ransacking the graves of this Huacal." This cemetery covered an area of only
twelve acres. According to Thomas Francis Meagher, the discoverer of the golden
relics was Don Roberto Soes.3 It is estimated by Dr. Merritt that the metallic
value of the images found at Bugavita alone amounted to $50,000. C. W. Liiders 4
estimated their value at nearly a million dollars. The largest gold figurine of
a high degree of purity found here is said to have weighed between eleven and
twelve ounces. Much of the pottery illustrated in this work came from Bugavita.

The major part of these gold images from Bugavita is said to have been melted
at Panama for their bullion value. Of those that were brought to the United
States only a few found their way into museums ; the plaque and the vampire

' F. M. Otis, M. D., in Harper's weekly, Aug. 6th, 1859.

2 Report on the huacals, or ancient graveyards of Chiriqui. Publ. by Amer. ethnol. soc.
previous to vol. I of its Bulletins.

3 The new route through Chiriqui. Harper's mag., XXII, 198, 1861.

4 Jahrbuch der Hamburgischen wiss. Anstalten, VI, 22, 1888.

METAL. 193

(see fig. 352) at the Lenox Library, New York, probably being two of the rare
and fortunate exceptions. What became of the gold ornaments to the value of
$70,000, exhibited at a meeting of the American Ethnological Society, is not known.1
The part that was sent to London has not been traced. Much of it presumably was
consigned to the melting pot. A few specimens are preserved in the Hamburg
Museum. It is highly probable that some of the pieces in the Yale collection
and in that of Mr. George G. Heye formed a part of the Bugavita treasure. This
is also true of the nine specimens exhibited by Mr. Alfred B. Taylor, at a meeting
of the Numismatic and Antiquarian Society of Philadelphia, October 5th, 1865, but
their present whereabouts is unknown. Chiriquian gold ornaments are rare in
European museums with the possible exception of that at Madrid, which I have
not yet seen.

Although the discoveries of 1858-59 attracted the attention of the world to the
archeological possibilities of Chiriqui, it does not follow that the presence of gold
ornaments in Chiriquian graves had been entirely unknown prior to that time. Squier,
the well-known archeologist, is authority for the statement that " large quantities
have been taken from time to time for many years past ; and I was informed by the
late Governor of the Bank of England, that several thousand pounds worth were
annually remitted from the Isthmus as bullion, to that establishment." It is to be
deplored that some of these ornaments did not go to the British Museum instead,
where the collection of isthmian gold figurines is still quite small ; under the wise
management of Mr. C. H. Read however it is being increased as opportunity
offers. In 1907 Mr. Read purchased a splendid specimen, the metal value of
which is £25. It consists of a group of eight figurines set in a rectangular panel,
the dimensions of the panel being 9.5 by 6 centimeters. The motive and workman-
ship are similar to those of the much smaller piece reproduced in Plate XLIX (fig. a).

In a given huacal the number of graves containing gold is small in comparison
with the total number. Some of the more elaborately constructed graves are said
to contain nothing at all. Those in which metal objects are found are usually
rich in pottery and stone implements. Dr. Merritt describes two kinds of graves
found at Bugavita, the oval and the quadrangular (see pp. 9 and 10). The oval
grave-pits were lined with rounded river stone. This type occurred principally
in the northern and western sections of the burial-ground ; " and as a general
rule yielded the most figures of gold and the finest specimens of pottery." It is
said that some of the gold figures were located in the crevices of the wall, " but
in no instance in earthen jars associated with them." The latter statement con-
flicts with that of Dr. Otis, previously cited.

There were two distinct types of the quadrangular grave. The lining of one
kind was of rounded river stone resembling in this respect the oval graves. They
were also in juxtaposition to the latter and like them in yielding more gold images
and finer pottery than the second type of quadrangular grave. This variety,
" although poor in relics, was more artistically and carefully constructed, and in
a better state of preservation." The walls of the vault were lined with flat stones
set upright. The cover of the vault was also composed of flat stones carefully fitted.

1 Hist, mag., IX, 158, 1865.

MEMOIRS CONN. ACAD., Vol. III. 25

194

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Pinart's description of the graves and the disposition of the gold objects differs
somewhat from the foregoing (see p. 12). He speaks of certain graves with walls
made of flat slabs set upright ; and the cover, a great stone slab. In these the
pottery and stone objects were found toward the center, while the human bones
were placed without apparent order along the walls, the gold, if there was any,
always being with the bones. The other type mentioned by Pinart consisted of
a more rudely constructed vault, into the walls of which were built recesses or
niches, each one lined with and covered by flat slabs of stone. Here the pottery
and stone objects were placed in the vault, and the gold in the niches.

The gold images of Chiriqui are a source of wonder and admiration to every
beholder. A careful study of them serves to throw new light on the mythology
and artistic skill of the ancient race, without lessening appreciably the mystery
that envelopes the methods they employed in order to arrive at such splendid
results. For the most part, they seem to have used alloys of gold with copper,
ranging practically from pure gold on the one hand to pure copper on the other.
A small percentage of silver is present in some specimens, one being the large
figure of a parrot illustrated in figure 353. Dr. C. H. Mathewson of the Hammond
Metallurgical Laboratory at Yale analyzed two of the specimens illustrated in this
work (see PI. XLIX, fig. g and text-fig. 370) and found them to be nearly pure
gold, the only alloy being copper. The alloys often appear to be natural. In
some cases, very effective figurines were produced by only slight alterations in
the shape of nuggets or masses of the native metal (see figs. 363 and 364). A
great majority of the pieces however were cast, either wholly or in part. Some
are constructed from a number of separate castings which were afterwards welded
together. Others are cast as a unit, certain parts of which — as tail, wings or
feet — being altered later by hammering. This class includes perhaps the largest
number of specimens. The accessible surfaces, especially those on the half intended
to* be seen, are carefully burnished. On the opposite half however hammer marks
are often distinctly visible. The subspherical pestle illustrated in figure 33 could
have been used both as hammer and burnisher. Its surfaces are streaked with
gold that in places is driven into the grain of the rock.

The gold seems to be concentrated on the surfaces, forming a layer of brighter
yellow than the interior. This is true even where there is but little alloy.
Sometimes this outer coating is but little more than a superficial bloom. When
susceptible of measurement, it is found to be thicker in some places than in
others. A possible explanation of this rich coating may be found in a state-
ment by Acosta to the effect that the Indians of Santa Marta, on the north coast
of Colombia, " had much gold and copper, also gilt copper — and the copper
was gilt by the use of the juice of a plant rubbed over it, then put into the fire,
when it took the gold color." Oviedo, " surveyor of the melting shops of King
Ferdinand" and Charles V, was the first to mention this process, Eden's trans-
lation l of his words being as follows :

" And forasmuche as I haue spoken of golde, I haue thought good to declare
sumewhat howe the Indians can very excellently gylte suche vesselles of copper and

1 The first three English books on America ; from the Latin of Peter Martyr of Anghiera ;
ed. by Edward Arber, 213, Birmingham, 1885.

METAL. 195

base gold as they make. For they can gyue them so fayre and floryshynge a
coloure, that all the masse whiche they gylte, appearethe as though it were golde
of .xxii. caractes and better. This colour they gyue with a certeyne herbe as
thoughe it were wrought by the arte of any goldesmyth of Spayne or Italic, and
wold of them be esteemed as a thynge of greate ryches, and a secreate maner
of gyldynge." As far as the laws of metallurgy are concerned, this statement
can be considered worthy of entire credence.

Another possible method of producing similar results would be to coat the wax
or resin model with gold before it was encased in the mold. In melting the
model, the gold would adhere as a shell to the interior of the mold, which could
then be filled with alloy. On removing the mold and burnishing the surface,
the process would be complete. This is the suggestion of Dr. George F. Kunz.

There are several specimens in the Yale Collection (see fig. 342) with a foundation
of metal apparently devoid of gold, over which is laid sheet gold, giving the
exact appearance of being plated. As the base metal underneath decays, the gold
comes off in scales.

The precise nature of the molds employed in casting is still a mystery, no
specimens or even fragments of them having been preserved. Sand or clay might
have been used. Modern goldsmiths make use of cuttlefish bone in casting small
objects not more than five by ten centimeters in dimensions. It requires two
for the halves of the mold, and they give a comparatively smooth surface to
the casting. The cuttlefish is said to be found in Chiriquian waters. The sand
box either undivided or in halves might have been in use. The character of
some unburnished surfaces suggests sand casting. One fine large figure of
the alligator-god (see PI. XLVIII, fig. g) seems to have been cast as one piece
in a sand box of two halves, because of the impression left by a short rod
inserted in order to produce the opening in the ring at the back. The same
results could not have been secured by an
undivided mold. In order to save metal and
reduce the weight of the figurines, the latter
were cast hollow, the figure being open for
all or part of its length either on the dorsal
or ventral aspect, but always on the one not
intended to be seen. The frequent crystalliza-
tion of the gold on the hollow surfaces, and
other superficial characters indicate that resin
or a similar substance must have been em-

„!„ ,-j en- -MT •»»• /-• v -,i Fig. 323- — Figure of a frog carved in resin;

ployed as a filling. Mr. Minor C. Keith efro^ 3DivalaB Lamson coKflcction. .A
possesses an interesting figurine from Rio

General, near Terraba, Costa Rica, in the hollow head of which the resin plug
is still retained (see fig. 375).

The Lamson collection includes a frog carved out of resin, which is mentioned
here because of its probable bearing on the art of casting in metal (fig. 323).
The ring on the throat for suspension is partially broken away ; the feet, which
were presumably flattened, are lacking. We evidently have to do here with the
metallic type of frog. The resin where freshly broken is of a rich reddish

196

A STUDY OF CHIRIQUIAN ANTIQUITIES.

color and the piece is perfectly translucent, even in its thickest parts, resembling
in this respect the familiar commercial rosin, only of a finer quality and different
smell when burning. The image came from Divala.

Sir Walter Raleigh l gives an interesting account of primitive metallurgy as prac-
tised by the Indians of Guiana in 1595 : " I after asked the maner how the Epuremei
wrought those plates of golde, and howe they could melt it out of the stone ;
hee tolde mee that the most of the golde which they made in plates and images,
was not severed from the stone, but that on the lake of Manoa, and in a multi-
tude of other rivers they gathered it in graines of perfect gold and in peeces as

Fig. 324. — Illustration showing primitive process of casting gold images. (After de Bry.)

bigge as small stones, and that they put it to a part of copper otherwise they
could not worke it, and that they used a great earthern pot with holes round
about it, and when they had mingled the gold and copper together, they fastened
canes to the holes, and so with the breath of men they increased the fire till the
metall ran, & then they cast it into moulds of stone and clay, and so make
those plates and images."

De Bry's illustration showing how the natives of Guiana cast their gold images
is reproduced in figure 324. Judging from the text accompanying his figure,

1 Richard Hakluyt. The principal navigations, voyages, traffiques & discoveries of the
English nation, etc. Reprint of 2d ed., X, 414, 415, 1904.

METAL. 197

the latter may have been inspired by the above narrative of one of Raleigh's men.
The dates of these two accounts would not necessarily conflict with such a view.
De Bry's text is as follows :

" Incolse regni Guiana statuas & imagines suas, plerumque ex paruis auri granulis
fundunt, quze in quodam lacu, non piocul a regia ciuitate Manoa, & in aliis
fluminibus, quas sese in lacum istum exonerant, colligunt. Ad grana ista aurea
paululum seris assumunt, ut auru tractari facilius possit, & postea vasi alicui fictili
includunt, quod multa habet foramina, ad quae fistulae quaeda siue calami aptantur,
ita ut in typos promineant, qui ad ignem sub vase isto in hunc usum collocati
sunt, ut aurum habitu oris liquefactum & ex vase fictili promanans excipiant." l

The fact that an alloy of gold and copper fuses at a lower temperature than
either metal alone accounts in part at least for the prevalent use of the alloy
among the natives of America. That they were able however to cast pure copper
and probably copper alloyed with tin, which has a higher melting point than gold,
is attested by the existence of figurines of that metal.

In describing and illustrating gold ornaments, great caution has been exercised
in eliminating all pieces, the authenticity of which could be questioned. In the
early days, collectors often had difficulty in disposing of originals at little if any
more than their bullion value. Now that the demand is great and the supply
small, they often command two and three times the gold value. Nearly all the

Fig. 325. — Needle of nearly pure copper. V1

specimens here figured were collected thirty to forty years ago, which fact of
itself creates a strong presumption in favor of their being genuine. In case of
more recent acquisitions, reliance is placed largely on a thorough comparison
with pieces that are undoubtedly authentic. I have spent months in searching
for criteria, by means of which even the most skilful frauds might be detected,
but am not yet prepared to announce complete success.

The number of metal objects in which gold plays no part is relatively small.
These are presumably either nearly pure copper or bronze. Among them may
be mentioned a needle (fig. 325), the eye of which is produced by slitting one
end, spreading the halves apart and then bringing their tips together. The
needle is perfectly round and of uniform diameter except in the region of the
tapering point.

Toilet articles are represented by bronze or copper tweezers, evidently intended
for plucking out hair or beard. Similar ones have been found in Peru. Squier
speaks of finding a pair of bronze tweezers suspended by a thread around the
neck of a mummified fisherman from a tomb at Pachacamac. The specimen
reproduced in figure 326 is perforated for suspension. One of the arms is broken
off. The type is more refined than that figured by Squier.

1 J. T. and J. I. de Bry. Collectiones peregrinationum in Indiam orientalem et Indiam
occidentalem, part VIII, pi. XVII, Francofurti ad Moenum, 1599.

198

A STUDY OF CHIRIQUIAN ANTIQUITIES.

The ancient Chiriquians were very fond of rattles. It is not surprising that
some of these should have taken the form of the common sleigh-bell of the north.
The simplest form consists of a hollow globular body slit underneath and provided
with a loop at the top and a relatively large metal pellet within. It seems to be
genetically related to the bell-shaped projecting eyes of the frog (see PI. XLVIII,
fig. i, and text-fig. 347). These bells are made of copper washed with gold, also
of practically pure gold. A tiny example in copper1 is given in figure 327. This

Fig. 326.

Fig. 327.

Fig. 328.

Fig. 326. — Copper tweezers perforated for suspension. V1

Fig. 327. — Small copper bell fashioned like the modern sleigh-bell. '(>

Fig. 328. — Base metal bell representing the human head. Lamson collection. '/»

simple type may be varied so as to represent the human head, as shown in
figure 328, a specimen belonging to Dr. Wm. J. Lamson of Summit, N. J. The
relatively large pellet inside resembles a dropping of copper. The prominent ears
are represented by coils of wire. A similar (bronze) bell from the Stearns collection
was figured by Holmes. In both cases the human features are inverted when
the bell is suspended from its loop. The jaguar's head is also employed as a bell,

one of this kind from Rio General, Costa Rica, having
recently been obtained by Mr. Minor C. Keith.

The most remarkable of the base metal figurines be-
longs to Dr. Lamson and is reproduced in figure 329. It
represents a frog and is fashioned with much skill, the
whole being cast. The exaggerated flattening of the hind-
feet is due in part to hammering and is almost constant
among the metal figurines of the frog. Its significance
will be noted later. The forefeet are plain rings for
suspension. The ancient Chiriquians were not always
satisfied with the realistic representation of the animal
form as a complete and single zoomorphic unit. They
sometimes combined parts of distinctly different animals
into one. More often they would ornament an otherwise
complete animal figure with motives derived from a wholly
different kind of animal or from parts thereof. Such is

the case here, where the frog's head is decorated with two conventionalized

alligator heads, each recognized by an eye, upper jaw with recurved snout and

lower jaw, the latter in each case being curved downward and backward to fuse

1 Dr. Mathewson finds that this specimen contains 20 per cent gold and a trace of silver.

Fig. 329. — Base metal casting
representing a frog, the head
of which is adorned with two
conventionalized alligator
heads. Lamson collection. 'A

METAL.

199

with the nose of the frog, on each side of the median line. The thinness of the cast-
ing and the delicacy of form and finish lead one to conclude that base alloys were
highly prized by the ancient Chiriquians ; this belief is strengthened by the fact
that objects of this class average quite as small as those of
gold, every part suggesting economy in the use of the alloy.
The use of the conventionalized alligator or alligator
head as a decorative and perhaps symbolic motive on
metal figurines does not seem to have been detected by
previous writers. Bollaert,1 for example, reproduces a gold
figurine (see fig. 365) that has a human body and an
alligator head. He not only did not recognize the head
as being that of an alligator, but also did not see the
three additional diagrammatic alligator heads and bodies
that are woven into the setting of the central figure. A sub-
sequent writer copied this illustration, calling it simply a
" grotesque human figure in gold, from Bollaert." He also
reproduced two other specimens 2 in which the alligator-
head motive is used, but apparently did not recognize
it. In one instance (fig. 330) this motive is attached to the knees of a " quad-
ruped with grotesque face ; " and in the other (fig. 331), it is repeated on each
side of the head of what Holmes thinks may be a crayfish " with complicated yet

33°- — Figurine in base
metal, to each knee of which
is attached a conventionalized
alligator head. (AfterHoImes.)

O;

Fig. 331. — Animal figure in base metal plated with gold, from the mouth of which
project two conventionalized alligators. (After Holmes.)

graceful antenna-like appendages, made of wire neatly coiled and welded together
by pressure or hammering.'' Liiders differs from Holmes as to the meaning of

1 William Bollaert. Antiquarian, ethnol. and other researches in New Granada, 32, fig. 3, 1860.

2 Ancient art of the province of Chiriqui, Sixth ami. rept. Bur. ethnol., figs. 35, 40,
1884-'85.

200

A STUDY OF CHIRIQUIAN ANTIQUITIES.

this figurine, insisting that it is not a crayfish but a cuttlefish (Tinten-fiscli). I differ
from both these writers and believe it to be a composite figure, with alligator
head and body and bird wings and tail. Two pairs of the "antenna" coils re-
present the curling ends of alligator jaws, between which the teeth are plainly
visible, the lower as well as the upper jaw being coiled on itself. These are alli-
gator-head motives very much conventionalized. Each is attached to what Holmes
calls an antenna, which in fact is a conventionalized alligator body projecting
from the mouth of the figurine. The avian characters are perfectly distinct. The
alligator is recognizable in the longitudinal raised lines and nodes on the back,
the projecting eyes and sudden rise of the naso-frontal region, the curve of the
back, and long muzzle surmounted by a suspension ring that takes the place of the
recurved snout (see profile). Had it not been thought necessary to emphasize the
latter feature, this ring would have been placed ventrally, where, if intended only

as a means for suspension, it rightfully belongs.
Subsequent illustrations will bear me out in these
conclusions. Figures with mixed attributes were
the favorites, whether intended for ornamental or
ceremonial purposes.

A gold figurine in the Metropolitan Museum
(fig. 332) belongs to the same class. The figure
as a whole, like the preceding, is a conventional-
ized alligator. The tail however is forked, and in
place of wings there are four legs, each terminating
in a conventionalized alligator head instead of a foot.
From the mouth protrude two alligators, their dorsal
spines resembling bird beaks. The heads of these
two alligators are highly characteristic, except that
the teeth are represented by pellets. The two bodies
resemble the Mexican feather-snake, as illustrated in
the ancient codices. The prominence in the region
of the eye is very marked in all six of these con-
ventionalized heads. Both of these figurines should be compared with three
specimens from the valley of Rio General, Costa Rica, two of which form part
of a collection (gift of Frau Dr. Mertens) recently acquired by the Royal Ethno-
graphical Museum of Berlin.1 The larger one resembles figure 331, except that
the hindlegs of the alligator are retained, the wings replacing the forelegs only.
The other is comparable to figure 332, but no conventionalized alligators protrude
from the mouth ; and the two forks of the tail are differentiated into alligators,
which, like the back of the parent figure, are ornamented with spines and scales.
The most perfect link in this chain of alligator forms is a specimen (also from
Rio General) in the Keith collection — a realistic alligator with two conventionalized
alligators protruding from the mouth and two other similar ones taking the place
of the tail. The parent figure and the four attached to it are all decorated with

Fig. 332. — Gold figurine ornamented
with alligator motives. Metropolitan
Museum, New York. '/'

1 E. Seler. Zeitschr. f. Ethnol., XLI, Taf. VI (bottom row at the center), 1909.

METAL.

201

raised spine- and scale-motives. The piece in the Metropolitan Museum is classed
as one of fourteen Mexican antiquities, the gift in 1896 of Mr. Audinet Gibert. If
found in Mexico it was evidently an importation from the south, since it fits so

Fig- 333- Fig. 334- Fig. 335.

Fig. 336.

Fig. 337-

Fig. 333 — Bead evidently made from a gold nugget. '/'

Fig. 334. — Gold bead. '/'

Fig. 335. — Cylindrical bead made of sheet gold. '/'

Fig. 336—338. — Cylindrical gold beads. '/'

Fig. 338-

Fig. 339-

perfectly into the series from Chiriqui and the adjoining district in Costa Rica,
and belongs to the Central American type of goldsmith's art.

In the Yale collections there are a number of gold beads of various shapes and
sizes (figs. 333-338). Some of them are evidently cast ; others are irregular in
shape, and seem to have been made of gold nuggets. One
bead is made of sheet gold rolled into the form of a cylinder.
Three cylindrical beads are reproduced, on which the enlarge-
ments at the ends seem to be a part of the casting rather
than fillets applied subsequently.

A small gold object (fig. 339) in the possession of Mr.
George G. Heye was apparently intended to be worn as a
labret or perhaps earring, and hence belongs in the general
class of articles of personal adornment. There is a similar
but larger specimen in the British Museum, which has, in
place of the four horizontal slits, four rows of horizontally
arranged triangular perforations. Most of the gold objects
described here however are evidently for ceremonial purposes
as well as for adornment. A much more elaborate example
than either of the foregoing was reproduced by Herr Liiders
as part of the celebrated Bugavita treasure.

The gold bell illustrated in figure 340 is one of the gems
of the Yale collection. It is surmounted by an animal form.
The upper story, which is shaped like a church bell, is or-
namented at the top and bottom by raised horizontal bands.
Between these are groups of bands forming x-shaped figures.
These features in relief appear to have been cast with the

Fig. 340.

surmounted by an ani-
mal form; the pellet in-
side is a nugget of gold.

Fig. 339. — Goldornament,
to be worn as a labret
or earring. Heye col-

body of the bell, the lower half of which is hemispherical Fi lectio0nL.Fine old bell
and slit vertically. The pellet inside is a gold nugget. The
entire piece is of fine gold.

Among the rare animal forms represented in gold may be
mentioned the fish and the crayfish. Examples of these were

illustrated in an article by F. M. Otis in Harper's Weekly. I have been unable
to trace the originals ; but have found in the Metropolitan Museum, New York,
a fish (fig. 341) similar in type to the ones figured by Otis. According to the
MEMOIRS CONN. ACAD., Vol. III. 26

202

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Museum records it came from a Chiriquian tomb, having been collected in 1859
by Mr. D. M. Corwing, while on a trip to Panama. It was afterwards given to the
Museum by Mr. Meredith Howland.

Representations of Cervidae are also rare. A specimen belonging to the Heye
collection is shown in figure 342. The neck and body are hollow, the latter being

Fig. 341.— Gold figurine of a fish. Metropolitan Museum,
New York. '/«

Fig. 342. — Figurine of a deer, in base metal
plated with low-grade gold. Heyc col-
lection. '/'

Fig- 343-

open ventrally. It is of base metal coated with a thin layer of low-grade gold,
which has disappeared from parts of the legs on account of the disintegration of
the baser foundation. The only gold figurine of the armadillo that I have seen
is the small one also belonging to Mr. Heye (fig. 343).

The frog was a favorite with both
potter and goldsmith of ancient Chiriqui.
Those fashioned out of metal were almost
universally identical in type, especially in
regard to the flattening of the hindfeet,
as seen in figure 344. The suggestion
may have come originally from the web-
foot of the frog, but other reasons must
be sought to explain why it should be
characteristic of the metal frog only.
Gold is precious and of an attractive
color. The attractiveness and apparent
size of the figure can be almost doubled
by the flattening of the hindfeet. As
there are always rings for suspension,
usually in the forefeet or between them,
these objects were evidently worn as
charms or ornaments. The flattening of
the hindfeet would help to steady the
figurine and keep it from rolling as the
wearer moved. The clay frogs, on the
other hand, being merely ornamental and symbolic features on vases and being
made of a common and non-precious material, there was no occasion to flatten

a b

Fig. 344-

Fig. 343. — Small gold figurine of the armadillo. Heye

collection. '/'
Fig. 344. — Gold figurine of a frog; a. — dorsal view,

6. — profile view. '''

METAL.

203

the extremities. Moreover flattening suggests hammering, which plays no part
in ceramic technique, but which does belong to metal technique. It is interesting
to note therefore where the metal technique (or rather the effects of it) has been
borrowed by the potter (see PI. XIV, figs. 6, c and d).

A figurine of two frogs united by the fusion of two adjacent hindfeet and also
by a tiny rod connecting the heads, is shown in figure 345. Here all the forefeet

Fig. 345- — Two united images of the frog in base metal
plated with gold ; from El Banco. '/'

Fig. 346. — Two gold figurines of the frog united by
bands of twisted wire. V'

are simple rings for suspension. The group was cast in base metal, then entirely
covered by a thin layer of gold, which has been removed in places from the edges
by flaking or by the disintegration of the base metal. This specimen is from
El Banco, Mount Chiriqui. In figure 346, the two frogs are united by two bands
of twisted wire, which may have been cast separately from the frogs themselves.
The rings that take the place of the two adjacent forelegs are very much worn
by the thread from which the group was once suspended.

Two of the gold figurines of frogs are reproduced in color (PI. XL VIII. figs, a
and b). The head of the smaller of these (fig. a) is ornamented with what appear
to be two gold wires, each coiled on itself at both ends, a stylistic way of rep-
resenting the snake's head (see fig. / and text-fig. 354).
The coils however are not visible from the ventral surface
and were evidently cast with the rest as one piece. The larger
specimen (fig. 6) has large eyes in the shape of sleigh-bells,
each supplied with a small ball of metal, apparently of copper
or a low-grade alloy of copper and gold. This is without
doubt the frog mentioned by Bollaert,1 as it agrees with his
description even to color and weight.

All the foregoing may be said to belong to the metallic
type of frog. An exception that proves the rule is seen in

figure 347, which is a ceramic type of frog (compare with fig. 100). The body
is short and high and there is no flattening of the hindfeet. The eyes in this
case are also prominent and bell-like. The ancient American bell may have been
derived from such a rendering of the animal eye.

In the Meredith Rowland collection at the Metropolitan Museum, there is a
gold frog which like the preceding has not the flattened hindfeet (fig. 348). It is
a fine specimen with a nose decoration of spirals, a herring-bone pattern down the

Fig. 347. — Gold image of
a frog ; ceramic type. '/»

1 Op. cit., 31 and 32.

204

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Fig. 348. —Gold image
of a frog. Metro-
politan Museum. New
York. '/•

back, and eyes resembling foot symbols. On the bottom of each forefoot there is a
ring for suspension. The Museum possesses another small gold figurine of a frog of
the type figured by Bollaert,1 and probably the identical specimen.

Figurines of the alligator are not so numerous as those of
the frog. One of the simplified forms is reproduced in figure 349.
This specimen, which was purchased from de Zeltner, was called
in his catalogue " une sorte de poisson " ; but it is an alligator,
as indicated by the prominence between the eyes, and the long
jaws showing teeth. The usual characteristic recurving of the
snout is represented here by a simple ring at the end of the
upper jaw (compare with fig. 331).

Mr. George G. Heye of New York has a large collection of
ancient Chiriquian gold ornaments, among them the representa-
tion of an alligator (fig. 350). Mr. Utley obtained this inter-
esting specimen at Pueblo Viejo. The attitude is one of motion ;
spreading legs, head raised and tail curved to one side. A
single row of long spines reaches from between the eyes to near
the tip of the tail. Under each forefoot there is a ring for sus-
pension. The prominence between the eyes and the upturned snout

are both characteristic. It holds
in its mouth a part of a human
leg (from the knee down). The
latter is recognized as human by
the flat foot and ankle-band. The
crab-god reproduced in color (see
PI. XLVIII, fig. h*) also holds in
its mouth the lower half of a
human leg.

The New York Public Library
(Lenox Foundation) is the for-
tunate possessor of some of the
gold ornaments brought to New
York by Mr. J. F. Bateman of
Panama, who accompanied the
first exploring party to Chiriqui,
in August, 1858. Mr. Bateman
exhibited his collection at a meet-
ing of the American Ethnological
Society, in October, 1860. It is
stated in the Proceedings2 that
" The articles exhibited differed
very much in size and form, a cricket, frog, a sea-shell, and a man, each from
one to two inches in length, and an ounce or two in weight ; a tiger or jaguar,
five ounces, and an alligator, eight and a half ounces, (the heaviest and largest

1 Op. cit, 32.

2 Bull. Amer. ethn. soc., I, 21, 1861.

Fig. 349-

Fig. 35°-

Fig. 349- - Simplified alligator figurine in low-grade gold. '/'
Fig. 350. — Gold image of an alligator holding in its mouth a
human leg ; from Pueblo Viejo. Heye collection. '/'

METAL.

205

figure yet discovered) and a circular piece of sheet gold, about six inches in diameter,
weighing one and a half ounces, with two holes, apparently for suspension."

The large alligator and jaguar, the circular piece of sheet gold, human figurine
and frog were afterwards bought by Mr. Robert L. Stuart and presented to the
Lenox Library. The alligator is reproduced in figure 351. According to Mr. Elliott

Fig. 351.— Large massive gold image of an alligator. New York Public Library (Lenox Foundation). '/•

of the Library, its weight is eight ounces and 24.6 dwts., or 250.42 grams. The
length somewhat reduced by the curving tail is fifteen centimeters. Although
head, body and tail (the latter two open ventrally) are hollow, the walls are every-
where massive. The nostrils and eyes are prominent; but contrary to the rule,
there are no dorsal scales and spines.
On the breast there is a single ring for
suspension. The gold has the appear-
ance of being about 20 carats fine. It is
no wonder that the country from whence
this specimen came was called Castillo,
del Oro.

There may also be seen with the Stuart
collection a curious figurine given by
Mr. Lenox himself and purporting to be
from Chiriqui (fig. 352). This was evi-
dently one of the early discoveries, prob-
ably the idealistic image that Bollaert
calls : " A bat, with outspread wings and
legs, having a dragon-like head, sur-
mounted by four horns, curling inwards,
of the purest gold and weighed six
ounces." l The same specimen was
figured in Harper's Weekly, August 6,
1859 ; the illustration however was a
negative instead of a positive, in which
the head was so inclined as to give

Fig- 352- — Bat-shaped image in fine gold. New York
Public Library (Lenox Foundation). '/•

a very different appearance from the

original. Liiders also reproduced this piece as part of the Bugavita treasure of
1 Op. cit., 31.

206

A STUDY OF CHIRIQUIAN ANTIQUITIES.

1858-59. It may represent the vampire (Vampyrus spectrum}. Peter Martyr (An-
ghiera), writing at the time of the Spanish Conquest, said that on the Isthmus of
Darien the lives of men as well as of cattle were in danger from blood-sucking
bats. Alston, on the contrary, later declared the blood-sucking propensity of
these creatures to be mythical.

Among Chiriquian gold figurines, representations of the bird dispute first place
with those of the frog. They are variations of one and the same kind of bird,
probably the sacred parrot that played an important role in the ceramic art of
this ancient people. It is characterized by a long hooked beak and tufts on the
head. The expanded wings and spreading tail suggest flight. For two reasons,

the bird in flight
would be an excel-
lent subject from the
standpoint of the gold-
smith, viz.; its adapta-
bility as a breast
ornament and the op-
portunity it gave to
produce an apparent-
ly large figurine out
of a relatively small
amount of gold.

The ornaments " ru-

dely shaped like ea-
gles" that Columbus
saw among thelndians
of theVeraguan coast
were probably rep-
resentations of this
same macaw or parrot.
One of the Indians is
said to have had a gold eagle (parrot) worth twenty-two ducats.

The largest of these bird forms (fig. 353) has a total spread of wing of 18.8 centi-
meters. The head, body and tail appear to have been cast in one piece. The
head ornaments and the wings were cast independently and welded on afterwards,
the effects of welding being plainly visible on the dorsal side. There is a broad
band about the neck made to resemble a series of wire strands, that bears an
incised zigzag pattern. One ring for suspension is situated on the beak and an-
other on the back of the neck. The head, neck and body are hollow and open along
the back, the figurine presenting an appearance of completeness only when seen
from the front. It is an alloy of copper and gold (six carats) with a little silver, and
is richer in some parts than in others. A similar piece is shown in figure 354 ; here
the tufts on the head each consists of the head and neck of a snake.

In figure 355, the tufts are formed of three pairs of gold wires coiled at their
free ends. They show irregularities in casting. The wings and tail are beaten
out to a thinness of less than half a millimeter; the marks of hammering

Fig. 353. — Large image of a bird in an alloy of copper, gold and silver

METAL.

207

Fig- 354-

are visible on the back but have been almost completely removed from the front by
a process of burnishing that has left numerous fine striae, which cross in all directions.

The Heye collection in-
cludes a bird figurine with
plain wing-shaped tufts al-
most as large as the wings
themselves (fig. 356). A small
reptile is held in its beak.
A similar example, also from
the Heye collection, is shown
in figure 357. In this case,
a fish is held in the beak ;
and the tufts on the head,
instead of being plain, are
alligator heads, as indicated
by the curving jaws and teeth ;
the latter are represented
by pellets, as was the case
in one painting of the alli-
gator. The piece is massive
and of fine gold. The hammer
marks are seen to excellent
advantage on the back of the
wings and tail. Both of these
specimens are from the valley
of Rio Chiriqui Viejo.

A small specimen (fig. 358)
in the Heye collection is
made up of pieces of sheet
gold somewhat thicker than
that in the average plaque.
The neck, body and tail are
one piece. The wings are
welded to the breast, while

the head with its ornaments

is made up of five parts. The

jaws and nuchal crest, so

characteristic in paintings of

the alligator, are gold wires

laid on pieces of sheet gold

that compose not only the

bird's head but also the necks

of the two alligators. The nuchal crests are so coiled as to represent the eyes

of the bird, thus serving two purposes. One small gold image of a bird differs

from all the foregoing in that it is intended to be seen from the back instead

of the front (fig. 359). The wings also are differently placed. The ring for

Fig- 355-

Fig. 354. — Gold figure of a bird, the head surmounted by snake heads. '/'
Fig. 355. — Gold figure of bird, the tufts being made of coiled wire;
a. — ventral view, b. — profile view. V'

208

A STUDY OF CHIRIQUIAN ANTIQUITIES.

suspension is formed half by the long hooked beak and half by a short gold wire
attached to the breast.

A highly conventionalized double bird form is reproduced in color (PI. XLVIII,
fig. c). The two birds are united at the end of their tails, by one wing in
common and by a bar connecting the heads in the region of the eyes. Contrary

Fig. 356. — Gold image of a bird, with large wing-
shaped tufts, holding a small reptile in its beak.
Heye collection. '/'

Fig. 357. — Gold figure of a bird, with a fish held in its
beak and with conventionalized alligator heads serving
as tufts. Heye collection. '/'

to the rule, wings and tails are burnished on both dorsal and ventral surfaces.
The gold is particularly pure. This type of double bird is frequently met with
in Costa Rica, Mr. Keith alone possessing sixteen examples of it from Mercedes.

We find representations of the jaguar not only in stone
and pottery but also in gold. They all agree in having
a wide-open mouth revealing teeth. The animal illustrated
in figure 360 is a departure from the conventional method of
rendering life forms in repose. Here, there is action in every
member, a very strong contrast to the formal pose of the
jaguar image reproduced in color (PI. XLVIII, fig. <£). The
head, neck and tail of the latter are cast in the round,
although hollow.

Gold objects representing the human form as a whole or
in combination with other animal forms comprise some of
the most interesting pieces of the series. For convenience,
the apelike forms are also placed in this group, as it is often
difficult to determine whether ape or man was intended. Figure 361 is frankly
apelike. It is a two-headed monster. The two long monkey tails ending in snake
heads are curved upward on either side and are held in the single pair of hands.
The body is so constructed as to form an elongated bell or rattle, inside of which is a

fig- 3S8-— Small bird fig-
urine in gold, its head
adorned with two con-
ventionalized alligator
heads. Heye collection. '/'

METAL.

209

subangular pellet, apparently of copper. The piece, which seems to have been
cast as a whole, has lost much of its head ornamentation of coiled wire, presum-
ably during its use in recent times as a belt ornament. When thus employed it
was held in place by a flat tongue attached to the back, the free end of which
could be inserted under the belt ; but this work of the modern jeweler has been
removed. Two similar specimens have been found in Costa Rica (Mercedes and
Rio General) and are now in the Keith collection.

Fig. 359-

Fig. 360.

Fig. 361.

Fig. 359. — Small gold figurine of a bird; a, dorsal view; b, profile view. V«

Fig. 360. — Gold figure of the jaguar. '/'

Fig. 361. — Gold image of an apelike monster with two heads. Vi

In the Heye collection is a small ape figurine (fig. 362) that is similar in attitude
to the representations of the monkey on the stone and clay stools, except that
the tail is longer, being arched over the head and held in the hands as a rope
jumper holds his rope. The pose is exactly like that in the repousse figure of a
monkey on a circular gold plaque from Cartago, Costa Rica, which is now in the
National Museum at San Jose. It should also be compared with the
type reproduced in figure 361 as well as with a specimen illustrated
by Holmes.1

In point of casting, one of the most interesting and delicate
pieces is seen in Plate XLIX (fig. a). Although of intricate design,
there is no external evidence of the joining of parts. The new sur-
face bloom everywhere gives to the group the appearance of being
fresh from the molds. There is no burnishing and no hammering
except two blows on the left foot of the left figure. The braid
of gold wire that forms the hollow square, within which the two
figures are set, is so deftly cast that in one place the light shows
between the strands. The same effect is produced by plaiting with three wires.
The two human figures with elbows touching are holding to their mouths some-
thing that resembles a conch-shell or a fish. A large group from Chiriqui similar
to this was recently purchased by the British Museum. The specimen is 9.5 centi-

Fig. 362.— Small
gold image of an
ape. Heye col-
lection. '/'

1 Ancient art of the province of Chiriqui, fig. 29.
MEMOIRS CONN. ACAD., Vol. III.

27

210 A STUDY OF CHIRIQUIAN ANTIQUITIES.

meters broad by 6 centimeters high and contains gold to the value of about $125.
Instead of two human figures, there are eight, all of them holding the conch-shell
to the lips. The two central figures are large and their bodies are converted into
bells; about them are grouped the six small figures. Among the clay whistles is
one (see fig. 315) that represents a monkey holding a fish before its mouth. It
would be interesting to know whether these representations all refer to the same
myth and what connection, if any, they might have with a similar theme in
Peruvian art, viz., a bird holding a fish.

A fine example of the human figurine is reproduced in color (PI. XLVIII, fig. e).
The eyes, mouth, nose and ears are all prominent, the latter resembling coils of
wire. The arms are half extended, the right hand holding what may be inter-
preted as a rattle (see fig. 273) and the left, a long rod, probably a flute, one end
of which is in the mouth and the other, enlarged at the end by a double coil of
wire, is free. The long fingers are executed with care. The brow is adorned
with what appears to be a three-ply braid, to each end of which is attached a
snake's head. The only articles of apparel consist of a loin-girdle terminating in
a coil at either side, a small square apron covering the pubis, and knee-bands.
The whole is cast as one piece, with the head, body and legs hollow and left
open at the back. This specimen was purchased from de Zeltner, whose relatively
small collection contained many choice examples of ceramic as well as the gold-
smith's art. His description differs from my own in respect to the head-dress
and the object held in the right hand. He also thought that the flattening of the
feet was intended to suggest a kind of footwear, while in my opinion it means
nothing more than does the characteristic flattening of the hindfeet of the frog.
It has already been suggested that the lateral flattening of the frog's feet gave
greater stability, when suspended from the neck of the wearer. A like increase
of stability is here provided for in the lateral extension of the arms and in the ob-
jects held in the hands. There is therefore no need of an exaggerated flattening
of the feet. Wherever there is but a single median ring for suspension, the
extremities are flattened and extended laterally to give balance ; and where
they are not so expanded, the desired state of equilibrium is attained in
another way ; i. e., by two lateral rings for suspension. It has been suggested
that the feet were flattened in order that they might be inserted in a crevice,
thus supporting the figurine in an erect position. There is some evidence tending
to prove that the larger pieces, especially those mounted on a vertically flattened
horizontal bar, may have been thus placed or else simply rested against a wall
or other suitable object ; for in the large figurine of the alligator-god (see
PI. XLVIII, fig. g}. the ring for suspension at the back bears no marks of wear,
while the lower edge of the bar at the feet is worn much more than is the
upper edge of the bar at the head.

This same deity is represented by numerous gold figurines, which however
are usually much smaller and simpler than the foregoing. It may have been a
god similar to Pan of Greek mythology. Another example is given in Plate XLIX
(fig. 6), where the attitude is the same, the rattle being nearly always in the
right hand and the flute or stick in the left. There is a simple fillet of gold
about the loins, and the knee-bands have dropped till they might be called anklets.

METAL. 211

The sex is masculine. The large ears are delicate sigmoid scrolls, one coil of
which is much greater than the other. The casting is quite thin. This same
type is found as far north as Mercedes, Costa Rica.

A representation of the same deity, but with variations in the details of handling,
is shown in Plate XLIX (figs, c and c'). It is immediately identified by the rattle
in the right hand, and the flute held to the mouth by the left. The features are
carefully wrought. The end of the nose seems to be pierced by a short hori-
zontal rod. The face is set in a sort of frame, the lower end of which is missing.
From behind this frame protrudes the head-dress adorned with a number of
blunt horns. The median fluted piece at the top is a flattened and curved
loop that served as a means of suspension after the two scapula rings had
worn through. The figurine, which is thick although hollow and open at
the back, bears evidence of much wear. Since there was a ring for suspension
back of each shoulder, it was not thought necessary to extend the arms laterally
and flatten the feet. The sex is male. This may be the specimen of which
Mr. J. F. Bateman speaks in a communication to the American Ethnological
Society.1

A small specimen from the Lamson collection (fig. d) is reproduced here be-
cause it is apparently a kindred deity. The figure holds a rattle in each hand.
In this case, as well as in the three foregoing, it will be noted that the at-
titude of the arms suggests that the rattles are in action. The head is adorned
with two pairs of plumes. The band about the loins is conspicuous ; it seems
to have been applied to the figure and made fast by hammering down the free
ends at the back. The legs are flattened bars of gold, grooved both in front and
at the back, and are probably welded to the trunk. The rest of the figurine
appears to have been cast as one piece, with the possible exception of the loin-
girdle, the ring at the back and the rattles. The latter two may have been cast
first and added to the resin core previous to running the final mold. This spec-
imen has the outward appearance of being about 18 carats fine ; in reality it
is considerably less. The processes by which the surface was finished in a finer
quality of gold than the foundation is not definitely known.

Belonging to the Heye collection is a figurine with features and head-dress
not unlike those of the foregoing ; it is reproduced in figure e. Here there is
but a single rattle, which contrary to custom is held in the left hand, the right
being extended downward almost parallel with the body. Another human figurine
in the Heye collection (fig. /) has a remarkable head-dress resembling the skil
of the Tlingit Indians. A similar head ornament, unfortunately broken, occurs
on one of the stone figurines of the Yale collection (see fig. 37), also on a gold
figurine from Mercedes, Costa Rica, recently collected by Mr. Keith. Both these
gold figurines are from Divala and are of rich yellow gold, the former being
approximately 20 carats fine.

An unusual group consisting of two human figurines is in the Howland col-
lection at the Metropolitan Museum (fig. </). It is said to have been found in a
Chiriquian tomb, but looks more like the art of Colombia. The two images are

1 Bulletin, I, 21, 1860-'61.

212 A STUDY OF CHIRIQUIAN ANTIQUITIES.

exactly alike, forming a group that is bilaterally symmetrical. In the outer hand
of each figure is something resembling a canoe paddle ; while in each inner hand
a staff is held vertically One of these seems to be set with a tomahawk blade
and the two are bound together at three or four points. In addition to the
head-dress, the central feature of which resembles the blade of a paddle, there
is an elaborate necklace and loin-band from which there hangs a short apron.
The long pointed noses are turned sharply upward (retrousse). The ear orna-
ments, the cylindrical ends of which are seen above each ear, are characteristic-
ally Colombian.

The ancient artificers of Chiriqui understood the value of a frame to a picture.
One example has already been given in figure a. A second instance is noted
in figures h and h\ which is also a fine illustration of skill in hammering and
uniting castings. The hammer marks are visible on both the front and back
of the frame, but the welding shows only on the back. The human figurine
was molded in a single piece ; the two sides and bottom of the frame were
cast in three pieces. The dorsal view shows how their thinness was increased
by hammering and how the various elements were united — the ears and
hands to the sides of the frame, and the latter and the feet to the bottom. Two
slender grooved vertical bars, are attached to the elbows and to the lower
part of the frame, dividing the space that separates the sides of the latter
from the legs. To these bars and to the sides of the frame are attached six
small triangular castings, three on each side. They are similar in shape to some
of the spine- or scale-motives painted on pottery of the alligator and lost color
groups. Two rods of gold, welded at points between the shoulders and the frame,
are carried some distance above the head-dress and end in recurved loops that
droop forward.

The fact that a repetition of the human figure with similar attributes occurs not
only in metal but also in clay and stone tends to increase our interest in ancient
Chiriquian mythology, which abounds in original elements and is so highly
developed. If only its history could be known and names be given to these
distinct groups of deities, as has been done for Egypt, Greece and Mexico !

Two groups have already been cited : the man holding a conch or fish to his
mouth and the man with flute and rattle. A third is represented by the small
figurine in color (PI. XLVIII, fig. /). This is characterized by the head-dress and
by the act of pulling two snakes from its wide-open mouth, one held in each
hand by a grip about the neck. Each head is formed by two wires coiled at
the ends, while the body of each is composed of two twisted wires. The head
of the snake, in the right hand, has been broken off. The foundation of the head-
dress is low and flat. Its chief ornamental feature consists of two snake heads
similar to the foregoing, but with bodies made of wires that are not twisted.
The three projections on the front of the head-dress would be meaningless,
were it not for other representations of the same deity, wherein these ele-
ments are treated in a more realistic fashion. They are bird heads, an illustra-
tion from the work of Holmes (fig. 30) leaving no doubt on this point. The
snake heads in Holmes's figure are also more realistic, resembling more the
serpent heads with forked tongues on the black incised pottery (serpent ware).

METAL.

213

There are other minor non-essential differences of detail. The Yale specimen,
for instance, has a loin-girdle and knee-bands, while the one described by
Holmes has nothing about the loins. On the other hand, the latter example
has an additional snake head attached to each foot and is not only larger but is
also framed in at the top and bottom by the customary broad flattened bars
of gold.

The ingenuity of the ancient Chiriquian goldsmith was expended not only on
these skilfully executed pieces but also on masses of native gold that gave sugges-
tion of the human form or parts of it. Figure 363 is an example. It represents a
seated figure with outstretched arms, clothed in a broad loin-cloth and cap. The
cap was evidently welded on ; the loin-cloth and arms may have been also. The
unevenness of the surface, particularly on the back, has been removed by grind-
ing. A somewhat similar piece is repro-
duced in figure 364.

Perhaps the most important and re-
markable divinity of this series is the
specimen with a human body and alli-
gator head, illustrated in Plate XLVIII
(fig. g). Although the figure is large and
in an elaborate setting it is made of rich
yellow gold. As far as the eye is able
to discern, the whole was run in a
single mold. The alligator-god standing
erect in the center is framed in by a
broad thin bar at the top and the bottom,
by a snake on each side and an alli-
gator head at each of the four corners.
The conception is original and skilfully
executed. The ears are serpent heads. The mouth is wide open, revealing teeth,
and the snout is coiled downward on a level with the mouth, presumably to
protect it from breaking. In a previous chapter it was noted that the snout of the
alligator as painted on pottery was invariably recurved upward. The goldsmith's
favorite way of indicating the eye, viz., an elongated pellet surrounded by a
fillet or wire, the two ends being brought together at the outer angle of the eye,
is shown in the central figure. The same is true of the eyes in the four alligator
heads at the corners ; one of these is somewhat disfigured through faulty casting.
The teeth in each of these four heads consist of a series of four pellets resem-
bling scales rather than teeth. Attention should also be called to the fact that the
wires forming the snouts are coiled upward as is the rule in representations of
the alligator. The snake on each side bridges the distance from one alli-
gator head to the other, its tongue touching the snout of the upper and its tail
that of the lower. The anterior half of the snake on the right has been lost.
The group is intended to be seen from the front only ; hence the braiding of
three wires which form the snake's body does not appear on the back ; neither
do the pellets and fillets that represent eyes, teeth and jaws. There is also
no reducing of the irregularities due to casting, except for the hammering

Fig. 363.

Fig. 364.

Fig. 363. — Seated human figure fashioned from a gold

nugget, '/i
Fig. 364. — Standing human figure fashioned from a

gold nugget. '/'

214

A STUDY OF CHIRIQUIAN ANTIQUITIES.

of the flat bars at the top and bottom. This idol may have graced a shrine
instead of being a personal ornament. The ring for suspension is not worn ; on
the other hand, the bottom of the lower bar is worn more than the top of the
upper one ; as if the weight of the figure had rested there. The very weight
of the piece would militate against its being suspended as a neck ornament.

An alligator-god not so large as the foregoing and without the elaborate setting
is in the British Museum (Cat. No. 4536). It is entered as a " monstrous standing
figure with horned head, one horn partly broken off, the other horn terminating
in a serpent-like head. At the back of the neck is a loop. Height 31/,, inches ;
width 3 inches ; weight 303 dwts. 7 gr." This figure stands in the same atti-
tude as the Yale specimen, but the feet do not rest on a flattened bar and
nothing is held in the hands. The so-called horns project laterally in the same
plane as the half-extended arms and end in inverted alligator heads each with

wide-open jaws, the upper one being recurved.
An eye, a short nuchal crest and a tooth leave
no doubt as to the identity of these heads.

Bollaert * illustrates another example of this deity
which is not so large as either of the foregoing,
but has features common to both (fig. 365). It
stands on the inverted body of an alligator, which
in turn is supported by the customary horizontal bar
at the base, the points of contact being the spines on
the back of the alligator, whose head is recognized
by an eye, upturned snout and open mouth with
teeth. The attitude of this alligator-god is the same
as in the foregoing examples, resembling more the
British Museum specimen in the head and head-
dress, as well as in the fact that nothing is held in
the hands. The snout ends in two upturned points.
Instead of horns, there are two highly conventional-
ized inverted alligators, their coiled tails being

attached to the sides of the head, and their heads resting on the shoulders of
the alligator-god. Bollaert, as well as a subsequent writer, who copied his il-
lustration, failed to detect the alligator motive in these ornamental and presum-
ably symbolic features. Bollaert described the specimen as " An idol of
hideous and obscene conception, with legs and arms extended : the head flat,
having a fan-like crown at the back, a wide open mouth, and a hooked nose."
There is a fine specimen of the alligator-god (fig. 366) in the Heye collection.
The head resembles that of the example in the British Museum, while the flat-
tened bar at the head and the feet suggest the Yale specimen, as do also the
four attached stylistic heads of the alligator ; these heads however are exactly
reversed in position.

In the Keith collection are two splendid alligator-gods from the Huacal de los
Reyes, Rio General, near Terraba, which fact seems to indicate that the domain

Fig. 365. — Image of " pure gold " rep-
resenting the alligator-god. (After
Holmes.)

1 Op. cit, 32,

METAL.

215

of this deity extended well into Costa Rica. In figure 367 there is the characteristic
prominence in the region of the eyes and the application of a wire coil to the
end of the nose to represent the hooked snout. The body of a snake issues from

Fig. 366. — Gold figure of the alligator-god,
collection. '/'

Heye

Fig- 367. — Image of gold representing the alligator-
god ; from the Huacal de los Reyes, Rio General,
near Terraba, Costa Rica. Keith collection. '/•

the mouth, its head reaching to the abdomen. Three of the conventionalized
alligator heads have the curving snout broken off, only the one on the left-hand
side of the head being complete in this respect. The jaws are wide open, and
across them run bars for teeth. The piece is consider-
ably worn. The bright yellow color of the gold is only
slightly tinged with copper. The other example (fig. 368)
has a similar nose piece, but the fronto-nasal prominence
is not so well accentuated. Instead of the oft-recurring
head-dress composed of a flattened bar with attached
conventionalized alligator heads, there is a framework
enclosing five sigmoid designs in the shape of a figure
eight, each representing a multiple alligator motive. The
body and legs are human. At the shoulders and hips,
however, are four additional extremities, as if to give
greater weight to the reptilian attributes, each of them
ending in an alligator head instead of a foot. What
seems to be a double alligator-god from this same
region has just been figured by Professor Eduard Seler.1
but the half-tone is so small as to leave one in doubt
concerning the distinguishing features. One of the nine
Chiriquian gold images exhibited by Mr. Alfred B. Taylor2
at a meeting of the Numismatic and Antiquarian Society

1 Zeitschr. f. Ethnol., XLI, Taf. VI (third row from bottom), 1909.

2 Golden relics from Chiriqui (paper read Oct. 5, 1865).

Fig. 368. — Gold figure of the
alligator-god ; from the Hua-
cal de los Reyes, Rio Gen-
eral, near Terraba, Costa
Rica. Keith collection. */>

216 A STUDY OF CHIRIQUIAN ANTIQUITIES.

of Philadelphia was presumably the alligator-god. It is described as a " Human
figure with head of a monster ; the mouth open and horns projecting from the
end of the nostrils ; the head surmounted by a sort of crown, projections from
which on either side form an ornamental framework around the whole figure."
The nine specimens were a part of the famous Bugavita treasure.

Were the evidence furnished by the gold figurines in question not sufficient to
establish the existence of an alligator-god in ancient Chiriquian mythology, the
pedigree of this deity could still be based on the remarkable painting in the
chalice reproduced in Plate I.

As a rule, metal figurines of man, monkey and bird are so constructed as to
present the front or ventral surface to view ; while the reverse is true of quadrupeds,
reptiles, fishes, crustaceans, etc., the back or dorsal view being the one exposed.
This rule is so nearly universal (I recall but one exception, fig. 359, and this
may be an importation from Colombia) that it is often helpful in determining the
nature of figurines with mixed attributes. A specimen illustrated by Dr. Max
Uhle1 and belonging to the National Museum of Costa Rica furnishes a case in
point. Its human attributes are almost entirely wanting but for the fact that it
is intended to be seen from the front or ventral surface ; the only other likeness
to man is in the region of the neck, shoulders and chest. On the other hand,
the alligator attributes are everywhere emphasized. The long tail is brought
forward in a median line along the belly, where it can be seen. The hindlegs
are brought forward in order to. make room for two conventionalized alligator
heads, one attached to each hip. The head of the figure is supplied with sigmoid
scrolls of wire for ears and is surmounted by an elaborate head-dress in a vertical,
transverse plane. The upper margin of this head-gear is serrated to represent
the dermal markings of the alligator, while at each side is a stylistic alligator
head with all its most characteristic features, even to the triangular scale-group
symbols on the back of the head and neck. The hands or forefeet, as the case
may be, are extended outward and upward till they touch the two decorative
heads that form a part of the crown of this alligator-god.

Uhle does not seem to have understood the true meaning of this figure. He
described it as half man and half beast to be sure, but mistook the decorative
and symbolic alligator heads attached to the hips for the lower extremities, which
he thought ended in snake-like heads. The hindlegs proper, which had been
shifted forward, he thought might be " a third and fourth or a fifth and sixth leg."
He described the head-ornament as a two-headed snake.

There was not only an alligator-god but also a crab-god. The one shown in
Plate XLVIII (fig. K) is an alloy of gold and copper, with a very thin surface
wash of gold that has disappeared from the more exposed portions. The specimen,
which is a single casting, is a most ingenious fusion of man and crab. The
combination presents the dorsal view of the crab and the ventral view of the
man, with the result that both appear to be complete. The crustacean carapace is
particularly true to nature. There are however but eight crab legs, one pair not
being represented, unless the human legs were intended to take the place of those

1 Globus, LX, 164, fig. 7, 1891.

METAL.

217

missing. The chelipeds or front legs armed with large claws also function as
the man's arms and are lifted as if to seize the ears. In order to emphasize their
human attributes, they are each supplied with two arm-bands or bracelets. The
only other articles of adornment or apparel are the crown and the anklets. The
correct number of toes may be counted on the feet, which are hammered rather
thin and to which serpent heads are attached. The human features are large,
the nose being characteristically so. This crab-god holds in his mouth the lower
half of a human leg severed at the knee — an interesting fact, the significance of
which can only be surmised (see also fig. 350). The crab's body being hollow and
supplied with a small ball of copper serves as a bell. The dorsal view reveals
a slit or opening, in shape like a horseshoe, the tongue of metal outlined by it
looking for all the world like the turned-under tail of a crab, but it is fastened
to the anterior end, perhaps in order to emphasize the fundamental structure of
the bell.

Other deities with mixed attributes also occur, the bird-god being one. An
example from the Heye collection is seen in figure 369. The body is human ;
the head that of a bird. There is the customary flat-
tened bar at the top and bottom. To emphasize the
bird attributes, six additional bird heads are attached
to the figure, two of these taking the place of hands
and the other four attached to the bars at the head and
feet, respectively, as was previously seen in two groups
of the alligator-god (see PI. XLVIII, fig. g ; and text-
fig. 366).

Another figurine in the Heye collection probably refers
to the same deity (fig. 370). Here however the human
attributes are minimized. In what would otherwise be a
complete bird form, human arms and hands simply take
the place of wings. There is the usual wide-spreading,
slightly forked tail. The head and its ornament are the
same as in the preceding figurine, but the two additional
inverted bird heads are not worked out in detail.

In these Chiriquian deities with human attributes, it is generally the latter that
dominate. That is so say, the body and extremities are usually human, and the
head, animal; in other words, a man with an animal mask, and with ornaments
representing parts of the animal in question or of some other. The reverse is
true in one specimen belonging to the Heye collection (fig. 371). Here the head,
breast and arms are human, and the body and lower extremities avian. The
tail being much reduced in size, the bird characters are not evident at first glance.
In order to further emphasize these, a bird foot is the central feature of the elaborate
head-dress and the human hands are replaced by bird feet. Two conventional-
ized bird heads are also placed at the sides of the body and serve as supports
for the elbows.

Among the Chiriquian antiquities exhibited by Captain Dow before the American
Ethnological Society, nearly fifty years ago, was a gold image with attributes
suggesting the foregoing. It was " in the form of a man, holding a bird in each
MEMOIRS CONN. ACAD., Vol. III.

Fig. 369. — Gold image rep-
resenting the parrot-god.
Heye collection. '/'

218

A STUDY OF CHIRIQUIAN ANTIQUITIES.

hand, sustaining one on his forehead." 1 It should be noted that here the entire bird
takes the place of the bird foot present in figure 371 — evidently another example
of the bird-god. The Imperial Museum of Natural History, Vienna, possesses a

small bird-god of gold — an avian head
and wings and a human body and lower
extremities. Decorative alligator heads
are attached to the calves, one on each
side, and a small animal is held in the
beak of this deity. If God created man
in His own image, man on the other hand
in making to himself graven images of
his god or gods would quite naturally give
them human attributes. This rule seems
to have held good in Chiriqui as well as
elsewhere. It is not strange therefore that
the pre-Columbian Chiriquians should have
chosen to represent their gods, many of
them at least, with human attributes.
In fact the parallelism between these
Chiriquian deities and certain gods of the
Hindu pantheon is most suggestive. I need
only mention Ganesha, the god of prudence
and policy, represented with human body
and elephant's head ; or some of the many
incarnations (avataras) of Vishnu, as: 1)
Matsya, part fish and part man ; 2) Kurma,
part tortoise and part man ; 3) Varaha with
human body and the head of a boar. There
is also Hanuman, one of the lesser deities,
with a monkey's head on a human body.
The evident esteem in which the parrot
was held, both among the Chibchas and
the Chiriquians, as well as the particular
type of avian characters seen in the bird-
god, leads me to conclude that the latter
might be given the specific name of parrot-
god instead. The beaks are always parrot-
like. Two of the finest examples of what
we shall henceforth call the parrot-god
were recently acquired by Mr. Keith and
like the two alligator-gods in the Keith
collection form part of the golden treasure of the Huacal de los Reyes in the
valley of Rio General, Costa Rica, discovered some three years ago, and

1 Bulletin Amer. ethnol. soc., I, 12, 1860-'61.

2 Edward Moor. The Hindu pantheon, Pis. 1, 48, 53, 54, 1810.

Fig. 370. — Gold
collection. '/'

figure of the parrot-god. Heye

Fig. 371.— Gold
collection. V"

figure of the parrot-god. Heye

METAL.

219

almost rivaling in richness the Chiriquian huacal of Bugavita. These two parrot-
gods are similar except as to details. The body and legs are human. The head
and the forked wings that replace the human arms are the only avian characters.
The figures are strengthened by the characteristic flattened bars at the head and
feet, to which and to the bars are attached conventionalized alligator heads. The
eyes are large and projecting. Each image is supplied with knee- and loin-bands.
A number of differences are to be noted. In figure 372, the eyes are bell-shaped,
hollow, slit and provided with pellets ; and a small reptile is held in the beak.
The alligator heads are made of wires. The image reproduced in figure 373
holds a fish in its beak. There is a ring for suspension at the back as well as
on the beak. In both of these examples, certain parts, as the wings and alligator
heads, have the appearance of being cast separately and then attached to the

Fig. 372.— Gold image of the parrot-god, with alli-
gator motives at the head and feet. Keith collec-
tion. V«

Fig. 373. — Image of the parrot-god in gold, or-
namented with alligator motives at the head
and feet. Keith collection. '/•

central figure by fusion and pressure. On the other hand, there are some features
which can be accounted for only by the supposition that while the model was
composed of a number of parts, the gold figure itself was cast as a whole.

Frau Dr. Alice Mertens has recently given to the Royal Ethnographical Museum,
Berlin, a valuable series of Costa Rican gold ornaments also from the valley of
Rio General. Judging from the published photographic reproductions,1 one of
these is a double parrot-god, with two complete human bodies, each having
two arms and a single wing, the latter attached to the distal shoulder in both.
One head is missing; the other is that of the parrot. Another double image
in the same collection resembles the foregoing in every essential feature with
the possible exception of the heads, the smallness of the half-tone rendering
it impossible to determine whether they are avian. If not, then we have to

1 Eduard Seler. Vorlage einer neu eingegangenen Sammlung von Goldaltertiimern aus
Costa Rica. Zeitschr. f. Ethnol., XLI, 463, 1909.

220 A STUDY OF CHIRIQUIAN ANTIQUITIES.

do with a mixture of three instead of two forms. In this connection it should
also be noted that human attributes do not always constitute one of the ele-
ments in composite gold images. For example, the wings and tail of a bird are
sometimes attached to the head and body of an alligator. One image of this
type is illustrated by Liiders,1 who also figures a deity with body and tail of a
bird, human arms with a rattle in each hand, and an alligator's head with horns
representing two conventionalized headless alligators. Where the figure is part
human I have suggested that it be called a god. Whether images composed of
two or more animal forms can be classed as gods is a question.

Representations of the parrot-god are confined neither to the gold figurines nor
to Chiriqui. At Mercedes, near the Atlantic coast of Costa Rica, Mr. Keith has
found a number of stone slabs of various sizes resembling somewhat the grave-
stones in the colonial cemeteries of New England. Instead of being used as
headstones however they are said to be found at the bottom of the graves.
They are skilfully carved out of volcanic rock and are generally ornamented with
figures in relief, or in the round. Two of the smaller and simpler slabs are each
adorned with a parrot-god. It stands at the centre of one end, with wings extended
outward and downward until they touch the corners of the slab. The entire figure
is thus in the round. The body and legs are human, to which are added a bird's
head and wings. In one instance the wings terminate in jaguar heads. The
Central American parrot-god recalls the Assyrian god of fecundity, which is re-
presented as having the head and wings of an eagle.2

Many of the celt-shaped amulets from Las Guacas, Nicoya (Costa Rica), des-
cribed by Hartman3 as "anthropomorphic," are also ornithomorphic, i. e., have
mixed attributes. These are in all probability representations of the same parrot-
god of which such fine examples in gold have just been noted. That the blade
of the celt is also the tail of a bird, there can be no question. Hartman states
that : " Often the lower part of the face, including the nose, is represented enclosed
by a square incision, having its upper corner at the root of the nose, its opposite
at the apex of the chin." Now the upper half of this diamond delineates either
the human nose or the slanting lines that begin near the base of the nose and
extend past the corners of the mouth. The lower half of this diamond is some-
times the outline of the human chin but is often that of a bird's beak instead.
Thus the "chin" which, according to Hartman, "is in numerous specimens tri-
angular, beak-shaped, prolonged downwardly on the breast," is in reality the
parrot's beak. In other words, the head is both human and avian, the figure as
a whole being analogous to that executed in gold, the differences being practically
confined to such as are due to the nature of the medium. Even the head-dress is
the same although not so elaborate. Instead of the relatively large gold bar
representing a common animal body (usually that of the alligator) with conventional-
ized heads at each end, there is a reduction of the whole, differing in degree,

1 C. W. Liiders. Jahrbuch der Hamburgischen wissenschaftl. Anstalten, VI, Taf. Ill,
Abb. 10, 11, 1888.

2 L'Anthropologie, XX, fig. 48, 1909.

3 Archeol. researches on the Pacific coast of Costa Rica. Mem. Carnegie mus., Pittsburgh,
III, no. 1, Pis. XXX1II-XXXV, XLIV, 1907.

METAL.

221

from the two opposing heads brought close together to two simple projections
resembling the ear-tufts. In fact the two tufts characteristic also of parrot figures
in clay may be but the highly conventionalized symbolic head-dress composed of
alligator heads — a supposition strengthened by the fact that these same images
are decorated with alligator motives in color (see figs. 248-49). Hartman rightly
construed the meaning of the head-dress on a number of the celt-shaped amulets
from Nicoya as being two alligator heads united by a short common body. It
is interesting to note that tiny anthropomorphic and ornithomorphic celt-shaped
amulets of gold have been found in Colombia. The Field Museum, Chicago,
possesses a number of these, .although no single specimen that I have seen com-
bines both human and avian characters.

The gem of the Heye collection, representing a creature part human and part
jaguar, presumably a jaguar-god, is reproduced in figure 374. The central feature
is the human body surmounted
by the jaguar's head. The hands
are replaced by jaguar heads.
There is an extra pair of arms
attached to the sides of the body,
which also terminate in jaguar
heads. Finally, there are bars
projecting laterally from the ankles
that end in jaguar heads, the bars
themselves serving as bodies for
these. The loin-band is composed
of the bodies of two snakes. As
in the case of the crab-god, the
human body is so constructed as
to form a bell or rattle, the metal
ball within being of about the same
grade of alloy as the figure itself.
The entire group is set in a square
frame of plaited work, on the outer
margin of which and at the back
is attached a series of spirals. This
fine specimen, which was collec-
ted by Mr. Utley at Pueblo Viejo, should be compared with the jaguar deity as
exemplified in Chiriquian ceramic art (see PI. XLI, and text-fig. 244).

The Keith collection of gold ornaments includes a number of jaguar-gods. The
one reproduced in figure 375 is from the Huacal de los Reyes, Rio General, Costa
Rica. The bars at the head and feet are each altered by a row of triangular
perforations, while the four alligator heads are so highly conventionalized as to
become simply curved extensions of the flattened bars. There is no differentiation
of jaws, teeth and eyes. In fact, each bar with its two curved extensions is a
multiple alligator motive — a common body with a head at each end, the triangular
perforations representing alligator spines or scale groups. The feet and hands of
this image are represented by very fine wires. The right foot had been broken

Fig. 374. — Gold image presumably of the jaguar-god,
collection. '/'

Heye

222

A STUDY OF CHIRIQUIAN ANTIQUITIES.

off and afterwards mended, as is indicated by a carefully drilled hole through
the foot and another through the bar directly underneath. This is the figurine

in the hollow head of which the resin plug
is still retained ; it is of special interest in
throwing light on the process of casting. In
figure 376, reproducing a jaguar-god also from
Rio General, it is quite evident that the bar
and its curved extensions represent the mul-
tiple alligator motive, each head being dif-
ferentiated. The spines on the common body
are placed ventrally, as in the preceding
figure. The bar with the alligator heads at
the feet is entirely wanting in this example,
which should be compared with two images
reproduced by Liiders.1

A third jaguar-god in the Keith collection
(fig. 377) is from Mercedes on the Atlantic
side of Costa Rica. The treatment is similar
to that in the preceding figure, except that
a rattle resembling a dumb-bell is held in
each hand ; and instead of the horizontal bar
representing a body common to the two con-
ventionalized alligator heads, the two bodies
are set like horns in the head of the jaguar-
god. In both figurines however the alligator motive is alike, in that the spines
and scales are placed ventrally instead of dorsally. In figure 375, the same result

Fig. 375. — Gold image of the jaguar-god, the
perforated bars at the head and feet being
alligator motives ; from the Huacal de los
Reyes, Rio General, Costa Rica. Keith col-
lection. V<

Fig. 376. — Gold figurine representing the jaguar-god
with an alligator motive serving as head-dress ; from
Rio General, Costa Rica. Keith collection. '/•

Fig. 377- — Gold figure of the jaguar-god, with
the alligator motive serving as head-dress ; from
Mercedes, Costa Rica. Keith collection, '/i

was arrived at by means of a row of triangular perforations. The three figures
taken together furnish the key to the meaning of the bars that form the head-

1 Op. cit, Taf. II, Abb. 4, 5.

METAL. 223

and foot-piece to so many Chiriquian gold figurines, this type of setting for such
images being in fact one of the characteristic features of Chiriquian art. These
bars are derived from, or at least merge into animal forms — two heads and a
common body. A majority of them represent the alligator, a rare exception to
this rule being given in figure 369, where bird heads are employed. The bar in
that case is presumably an avian body, common to the two heads which it con-
nects. The present northern limit of the use of these flattened symbolic bars at
the head and feet of gold images is Nicaragua, west of the lakes. A good
example from this region is in the American Museum of Natural History. This
image, which is small and of low-grade gold, resembles that reproduced in
figure 375. The head however is apelike ; there are loin- and knee-bands ; and
the bars at the head and one at the feet are not even perforated. Similar gold
figurines from Colombia are to be seen in the Field Museum, Chicago.

Like the parrot-god, the jaguar-god also occurs on the monumental stone slabs
as well as in the form of independent stone statues, examples of both having
been found recently at Mercedes, Costa Rica, by Mr. Keith. One of these slabs
is about two meters high by fifty-nine centimeters wide and ten centimeters thick.
The front is plain. Resting on the top is a group of three jaguar-gods carved in
the round. The bodies and extremities are human. While the head in each case
is that of the jaguar, it is adorned with long human hair reaching down to the
lumbar region. In addition, the larger, central figure wears a crown. The grouping
is admirable. The chief god rests on both knees with arms extended and hands
on the shoulders of the figures at the sides. These two smaller gods are of equal
size, the one on the right kneeling on the right knee and that on the left kneeling
on the left knee ; the group as a whole is thus bilaterally symmetrical. The
two lateral margins at the back of the slab are decorated with eighteen small
figures of the jaguar, nine on each side, with heads all directed toward the group
of jaguar-gods. Beginning at the upper corners, these jaguar figures are distrib-
uted at equal distances, the last ones being situated at least thirty centimeters
from the foot of the slab.

Equally remarkable is the great stone statue representing the jaguar-god and
found also at Mercedes by Mr. Keith. It is of stocky proportions, with a height
equal to that of a short man. It wears a cap or crown, cylindrical ear-plugs and
a sash carried over the left shoulder and reaching down to the left hip. The
right arm is missing. In the left hand is held a human head, the long coiled
hair of which is brought up over the right shoulder, as if to balance the sash on
the left. This statue has certain points in common with one also from Mercedes,
illustrated byHartman,1 although the head and body of the latter are both human.
The discovery of these fine examples in stone of the jaguar-god not only serves
as a confirmation of my belief in the existence of that deity based on the gold and
earthenware specimens already described, but also extends the boundaries of the
cult half-way across Costa Rica.

The Yale collection includes a single rare, plaque-like gold mask representing
the human face (fig. 378). It is a squarish sheet of gold-leaf, somewhat thicker

1 Archaeological researches in Costa Rica, PI. 3, fig. 1, 1901.

224

A STUDY OF CHIRIQUIAN ANTIQUITIES.

than ordinary tin-foil. The margin is plain and near each of the four rounded
corners is a perforation, the puncture having been made from the front ; one of
these has been almost completely obliterated by a break, but the traces of it are
distinct enough to prove its existence. It is not likely therefore that this piece
was worn as a breast ornament suspended from the neck, but rather as a mask.
As it is much too small to cover an adult face, it may have served to cover the
face of an idol. The human features — eyes, nose and mouth — are slightly
raised by repoussage. There are also three elongated lumps produced in the same
way. One of these is placed horizontally on the forehead and in line with the
nose and mouth ; the other two are paired and placed vertically under the eyes
and a little above the level of the mouth. They may possibly represent tattooing.
A gold mask from Agua Caliente, near Cartago, Costa Rica, marked in almost
the same manner has been figured by Uhle as belonging to the Troyo collection

of the National Museum at San Jose.
Similar gold masks were found by
Schliemann on the faces of the dead
in certain tombs of Mycenae.

Dr. George F. Kunz 1 describes a
somewhat similar " plaque ornament "
that was found on the banks of the
Mingindo river, a tributary of the
Atrato, in the state of Cauca, United
States of Colombia. This however is
circular and the mask-like characters
are not so pronounced. There are but
two perforations, their position being
in the upper part between the eyes,
suggesting that the piece was used as
a breast ornament. It belongs to Mr.
S. L. M. Barlow, who states that a

banker of South America had purchased a hundred of these shield-shaped orna-
ments simply for their bullion value. They were melted down, no description of
them having been kept. The same author2 describes a circular gold plaque
(from a Florida mound), the surface of which is decorated with raised work
resembling beads and pendants. These are near the margin, while in the center
is a circular raised portion. This piece is not perforated.

Bollaert mentions that Messrs. Pixley and Company of London allowed him
(about 1860) to examine five of their circular gold plates from Chiriqui. The
largest was TU inches in diameter and had seven circular embossments. In des-
cribing the voyage of Columbus along Costa Rica and Veragua, Irving 3 says : " Here,
for the first time on the coast, the Spaniards met with specimens of pure gold, the
natives wearing large plates of it suspended round their necks by cotton cords."

1 American antiquarian, Sept., 1887.

2 Op. cit., July, 1887.

3 Life and voyages of Christopher Columbus, II, 314, 1892.

Fig. 378. — Rare gold mask representing the human face. */>

METAL.

225

There are a number of splendid gold plaques in the Heye collection, the most
interesting being illustrated in figure 379. It is circular and made of relatively
thick sheet gold. All the repousse work on it seems to have been done from
the back. Near the margin are two circles of raised points. Within this field are
five large embossments, each surrounded by a single row of raised points. The
disk is supplied with two pairs of perforations for suspension, the outer pair having
been made after a crack had penetrated to one of the original holes. The two
small holes near the margin are the result of mending the break. The large emboss-
ments are realistic representations of the female breast, even the nipple being
faithfully rendered. A girl before marriage is supposed to wear a plaque with

Fig. 379. — Large gold plaque with five embossments representing the human
breast. Heye collection. '/'

a single breast, while after marriage a plaque with a pair of mammae may be
worn. Mr. Heye has one small piece of sheet gold with but a single breast
design.

The Heye collection includes two other gold plaques, almost identical in size
with that in the foregoing figure. One has two circles of raised points about
the margin ; the other, which is exceedingly thin, has a circular indentation near
the margin but no embossments. Both are convex on the front, while the plaque
with the five mammae is perfectly flat save for the repousse work. The piece
reproduced in figure 380 also belongs to the Heye collection, and is of special
interest because the repousse work is not all done from the same side, it being
sunken except the circle near the margin which is raised. The central field is
convex.

MEMOIRS CONN. ACAD., Vol. III. 29

226

A STUDY OF CHIRIQUIAN ANTIQUITIES.

The use of metal disks or plaques as breast ornaments was widespread in pre-
Columbian times. They are reported by Ambrosetti from the Calchaqui region

of Argentina. A mummy found by Uhle
near Casabindo, wore a copper disk on
its breast. The specimen is now in the
Museum fur Volkerkunde, Berlin. Bollaert
figures a gold plaque from Cuenca, Ecua-
dor ; and a half-dozen from Angel, Prov-
ince of Carchi, are to be seen in the
Museum of the University of Quito. Saville
describes copper disks, also from Ecuador,
each with an embossed face in the center.
Specimens from the United States of Co-
lombia and from Florida, described by
Kunz, have already been noted. Clarence
B. Moore illustrates a copper plaque from
Mt. Royal, Florida, and points out its re-
Fig. 380.- Gold plaque. Heye collection. '/« semblance to those shown by Le Moyne

on king Outina. Gold plaques have been

found in various tombs of Mexico, in some cases on the breastbone. They are
the symbol of Tezcatlipoca, " shining mirror," and are represented in the ancient
codices as worn on the breast of the priests of Tezcatlipoca.

CONCLUSIONS.

Before American archeology can become a science, it must be studied region-
ally and intensively. Each region may be likened to one of the blocks in a
picture puzzle, with this difference, that the seriousness of the problem is due
more to the loss and indistinctness of parts than to their disarrangement. Not
only is the picture on many of the blocks indistinct and incomplete, but often
whole blocks are missing. It behooves the archeologist therefore to study the
more minutely every existing fragment for clues that may lead to the reconstruc-
tion of that which has vanished.

Chiriqui as a region for intensive study has many points in its favor. It lies
at the gateway of two continents, midway between the classic fields of Mexico
and Peru. That part from which the vast majority of antiquities came is of small
extent, bounded on the north by a rugged mountain chain and on the south by
the Pacific Ocean, a veritable garden spot, sloping gently from the foot-hills of
the Cordilleras to the sea and abundantly watered by scores of mountain streams.
The soil is exceedingly rich, capable of supporting a large population, and the
climate is agreeable and healthy. On such a soil and amid such an environment
a splendid civilization might well be nurtured and grow to maturity.

My reason for choosing Chiriqui as a field for intensive study is that the most
important collection of Chiriquian antiquities known, the one belonging to Yale
University, is in my charge. What I know of Chiriquian archeology has been
learned from it, from other collections both public and private, and from libraries.
What I have gleaned is offered to students of American archeology as a con-
tribution that invites a supplement from some one who may be so fortunate as to
have the time, inclination and opportunity for a personal investigation of this
region. It may not yet be too late to gather much information regarding the various
types of graves, for example, and the association of artifacts within them. Until
this is done, our knowledge of the time element and of the various steps in the
development of Chiriquian culture must remain largely conjectural.

In freely acknowledging my indebtedness to previous writers on the archeology
of Chiriqui, the chief among them being Professor W. H. Holmes, it is fitting
that I should close with a presentation of some of the fruits that have ripened
primarily in the light of my own researches.

The art products of Chiriqui comprise implements, statuettes, metates, stools,
petroglyphs, sculptured columns, amulets and ornaments of stone ; pottery in the
form of vases, tripods, bowls, cups, figurines, whistles, rattles, spindle-whorls and
stools ; and metal objects of gold and copper, alloys of these, and bronze.
A single shark's tooth, perforated for suspension, has been noted, and one figurine
of rosin.

Attention is called to the general phylogenetic trend in the development of
Chiriquian art as a whole. The ties that bind the product of the stone worker to
that of the potter and the artificer in metal are more noticeable between stone
and pottery on the one hand and pottery and metal objects on the other, than

A STUDY OF CHIRIQUIAN ANTIQUITIES.

between stone on the one hand and metal on the other. It is significant that
these ties binding together the art in all mediums seem to be centered in the
group of unpainted pottery, to which I have given the name armadillo ware.

With the exception of architecture, the stone art of Chiriqui compares favorably
with that of Mexico or Peru. The chipping and polishing of stone implements
was carried to a high degree of perfection in view of the serious handicap due
to the apparent lack of flint and obsidian, which have been everywhere the chief
heritage of stone-age culture. Use was made of local stone almost exclusively,
the choice of materials being confined principally to those of volcanic origin.
The overcoming of technical difficulties inherent in the raw materials is nowhere
better exemplified than in the series of celts, metates and stools. Jade ornaments
are rare; the character of the material and the workmanship point to a kinship
with the numerous amulets of jade that have been found in Nicoya, or at least
to a common source of jade supply located presumably somewhere in Costa Rica.

Fictile products form the great bulk of Chiriquian antiquities. The major part
of these are simply vessels in the form of vases with round bottoms or mounted
as tripods. Their prototype was presumably the calabash. Aside from this the
plant world had practically no influence on the elaboration of form and ornament
in clay, the plasticity of which is so well calculated to stimulate the imagination,
clay yielding readily to any form that fancy may dictate. These forms, whether
useful or ornamental, present surfaces that invite further decoration by means
either of engravings or painted designs, thus bringing into play the whole realm
of art, from sculpture in the round and relief to engraving and painting.

Chiriquian pottery admits of classification into : (1) Unpainted ware and (2) painted
ware, each main division comprising a number of subdivisions. My reasons are
given for proposing several changes in the nomenclature of Holmes. Instead of
the "terra cotta" or "biscuit" group the name armadillo is recommended for the
great group of unpainted ware, and the serpent group is substituted for the " black
incised " group. In the division of painted ware, a change of one group name
is suggested, viz., fish group instead of " tripod " group. These changes are all
in line with Holmes's choice of the name "alligator" for one of the important
groups of painted ware. The names of animal forms are thus applied to four
groups instead of only one, as was the case in the earlier classification. The
changes therefore are justified by precedent, as well as on the ground of appro-
priateness and uniformity, and by the fact that the ancient Chiriquian potter drew
his motives wholly from the animal world.

I have given the name of chocolate incised ware to a small group not hitherto
noted, with distinctly Costa Rican affinities. Some of the larger groups admit of
further differentiation ; for example, the salmon-colored variety of the armadillo
ware. The handled group also comprises two more or less distinct varieties, one
painted and the other unpainted. On the other hand, the polychrome ware and
the alligator ware have so many points of similarity that they might well be
considered as varieties of one group, to all of which the name alligator applies
with equal fitness. Alligator motives predominate in both, but those in each group
are distinctive in character. This fact supported by the presence of characters that
are common to the polychrome ware alone — such as the use of an additional

CONCLUSIONS. 229

delineating color (purple) and the occurrence of the branching scroll — offers
sufficient reason for retaining the old name, although the latter is liable to lead
to confusion, having already been applied by other writers to totally different
kinds of pottery. In fact the name polychrome might equally well denote any
class of pottery in the decoration of which the requisite number of colors were
employed.

In making animal motives a basis for classification, therefore, it must not be
inferred that a motive derived from a given animal is confined wholly to a given
group of ware. While distinctive of the armadillo group, motives derived from
the armadillo or parts thereof appear sparingly in other groups, serving as ties
that bind together a series of related groups. Motives derived from the armadillo
are everywhere plastic, presumably because they originated in a class of unpainted
ware that depended on sculpture and relief for ornament. When transferred to
painted ware, their plastic origin still asserted itself. The development of a whole
series of motives derived from the armadillo is first noted in this work. This evolution
was accomplished by a process of elimination, by wholesale reduction and simplifi-
cation, also by the isolation of parts and their use as symbolic or decorative
motives independent of the animal as a whole. We have thus eye, foot, tail and
carapace motives. These are employed in series, either separately or in com-
binations, to produce highly original, significant and decorative patterns — such
as, for example, a meander encircling the neck of a vase, composed of a series
of tails or of carapace bands, with an eye or a foot symbol filling each angular
space.

The alligator was also a great favorite with the ancient Chiriquian potter.
Motives derived from it however are executed in color instead of in relief. They
characterize a group of ware that depends on color for ornament rather than
on sculpture, and when carried over into other groups they appear consistently
as painted forms, but with an individuality somewhat altered by the technique
of the group of ware in question (see figs. 381-384).

Realistic painted representations of the alligator are in profile. By this means, the
peculiarities of its anatomy and pose are easily indicated. A number of processes
set in action lead to conventionalism. Sorrfe of these are : (1) The reduplication,
exaggeration, elimination, or fusion of parts or units ; (2) transposition, shifting and
substitution ; (3) isolation of parts and their use independently of the whole ; (4)
wholesale reduction and simplification, and (5) adaptation to fit a given space. The
profile figure is eventually reduced to a mere body-line with a spot in the hollow
of the dorsal curve to represent scales and spines on the animal's back. This
becomes a decorative motive and when repeated in a series forms a pleasing
pattern. Groups of two, one of them being inverted, form a unit of the sigmoid
scroll. A number of these motives are often combined in such a way as to pro-
duce the branching scroll that characterizes the ornamentation of the polychrome
ware. A single profile motive reduced to its lowest terms would have made a
very convenient hieroglyph in a system of writing, but there is no evidence to
prove that it was used as such. The same can be said of the spine motive and
the scale-group motive, both of which become differentiated from the profile
motive.

230

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Another favorite alligator motive is that derived from the dorsal view and
frequently employed as a panel decoration. The rows of spines and scales on
the back of the alligator are represented by a number of parallel lines, the outer

Fig. 381. — Dorsal-view (alligator) motives from vases of the lost color ware.

ones alone bearing spine, scale, or scale-group symbols, and these only along
their outer margins. This motive is common not only to the alligator ware but
is also frequently met with in the lost color group, where its appear-
ance is somewhat altered on account of the lost color technique.
A comparison of figures 381 and 383 discloses these differences
at a glance. Figure 382 is taken from a tripod vase of the white
line group and differs from the same motive in both the alligator
and lost color groups. That the profile motive also undergoes changes
when passing from the alligator group to the polychrome group, is
evident from a comparison of figure 384 with figure 224.

„ The ancient potter of Chiriqui was master of the brush in three

sal-view mo- distinct systems of painting: (1) The usual method of producing a
figure or pattern by the direct application of one or more deline-
ware. ating colors — red, black, white, and in one variety of ware, an ad-

ditional color, purple ; (2) the lost color process, which was confined
to a single rather large group of ware. It consisted of tracing the design in wax
on the ground color, the application of a solid coat of black over the area to
be ornamented, and immersing the vessel in hot water which melted the wax,

Fig. 383.— Dorsal-view motives from vases of the alligator ware.

removing with it the black from the design and thus leaving the latter in the
color of the original ground ; (3) Sparing the figure out of the ground, which was
seldom employed, occurring only in the alligator, and the kindred polychrome
group. The few examples of this technique evince much skill and ingenuity.

CONCLUSIONS. 231

Primitive metallurgy is well exemplified in the art of Chiriqui. It was still
practised with presumably similar results at the time of the Spanish conquest.
The first explorers were much impressed by the richness and abundance of the
gold plaques, figurines, etc., used by the natives as amulets and ornaments.
De Bry figures a group of native metal-workers in the act of casting gold images.
The gold ornaments seen by Columbus in the Chiriqui lagoon were evidently
similar to those taken from the graves during the past half century. The latter
vary in fineness from nearly pure gold on the one hand to pure copper on the
other. A small percentage of silver is present in some specimens. The alloys
often appear to be natural. In some instances very effective figures were produced
by slightly altering the natural shape of nuggets or masses of the native metal.
A great majority of the pieces were cast, either wholly or in part. Some appear
to have been constructed from a number of castings which were afterwards joined
together. It is probable however that in some cases, at least, this appearance
is alone due to the fact that the model was built up of a number of parts, the
casting itself being in reality a unit. Others were cast as a unit, certain parts
of which — as tail, wings or feet — being further flattened by hammering.

Fig. 384. — Dorsal-view motives from vases of the polychrome ware.

The secret of concentrating the gold on the surfaces in order to form a layer
of brighter yellow than the interior, so well illustrated in specimens from the
graves, seems to have been known also to the natives with whom the Europeans
first came in contact, for Acosta says the Indians were able to produce a gold
color by rubbing a low-grade alloy with the juice of a plant. Acosta probably
got his information from Oviedo, the first inspector of " gold foundries " 1 in the
New World, who refers to the same process.

As a vehicle of mythological and artistic expression, the gold images of Chiriqui
are a worthy rival of the splendid series of pottery. The forms represented are
from the animal world, more or less faithful representations of the bird, frog,
jaguar, alligator, monkey and man being frequently met with.

The majority however are composite in character, possessing attributes of various
animals, and are generally described by earlier writers as monsters. My endeavor
to unravel the apparent tangle of mixed attributes has, I believe, led to the
identification of a number of important Chiriquian deities.

As might be expected, some human attribute is common to practically all such
divinities. One of the most important of these consists of a human body and
extremities surmounted by an alligator's head. In order to emphasize the reptilian
character, additional conventionalized alligators or alligator heads are attached to
various parts of the figure or of its setting. To such an image I have given the

1 Irving. The life and voyages of Christopher Columbus, III, 359, 1892.

232 A STUDY OF CHIRIQUIAN ANTIQUITIES.

name alligator-god. The conventionalized alligator head is often used, also, as a
decorative and symbolic motive on other figures than those of the alligator-god
I have demonstrated that the flattened horizontal bars, with their projecting attach-
ments, placed at the head and feet of so many Chiriquian gold images, are
traceable to multiple life forms — a common body with a head at each end (generally
the alligator).

There is likewise a mythical image, part human and part bird, which I have
named the parrot-god ; to two other forms I have given the names jaguar-god and
crab-god. A number of these deities are reproduced in earthenware, one of the
most interesting being the jaguar-god. It should be recalled here that the finest
example of ceramic decoration in the collection is a painted figure of the alligator-
god. The gold images are all supplied with one or more rings for suspension.
It is worthy of note that all human figurines and those with human attributes are
so constructed as to present the front or ventral surface to view when suspended.
This is true also of bird and monkey figurines, while the reverse is true of those
representing all other animals, the dorsal view being the one exposed.

It is still too early to determine the precise geographical boundaries of Chiriquian
culture. That these surpassed the present political boundaries of the province of
Chiriqui, particularly in the direction of Costa Rica, seems to be an established
fact. Numerous specimens of alligator ware, so characteristic of Chiriqui, have
been found in the region of Terraba ; and fine examples of armadillo ware, also
characteristic of Chiriqui, are noted from Boruca and even as far north as Mer-
cedes, where Mr. Keith also obtained a jaguar-god of alligator ware. The al-
ligator motives so characteristic of certain classes of Chiriquian pottery are found
on Costa Rican pottery that otherwise bears little resemblance to Chiriquian ware.
Thus we find the dorsal-view motive, the scale-group and the spine motives, as
well as the sigmoid scroll with a spot in the hollow of each curve, the last three
occurring as far north as Nicoya. The genetic relationship between Chiriquian
and Nicoyan culture is further emphasized by the evident kinship of the celt-
shaped jade amulets from Las Guacas to the gold parrot-gods from the valleys
of Rio General and Chiriqui Viejo.

The recent discovery near Terraba and at Mercedes of gold images represent-
ing the alligator- and jaguar-gods indicates that the cult of those deities once
extended even beyond the waters of the Gulf of Dulce. Gold images in the
Chiriquian technique occur, though rarely, even as far north as Nicaragua. The
conventionalized treatment of the alligator points to a possible kinship between
Chiriquian and Mexican mythology. Several favorite Chiriquian totemic animals
appear in ancient Mayan codices. The characteristic jaguar metates and the two
prevailing types of stone stools have been found in Costa Rican graves. There
are also many points of resemblance in the graves themselves.

Archeological evidences of contact with and influence by South American
civilization might be even more plentiful were the archeology of Colombia and
Ecuador as well-known as that of Mexico and Central America. The lost color
process, one of the characteristic Chiriquian methods of ceramic decoration, has
been found on ancient pottery from Rio Bamba, Ecuador. Petroglyphs not unlike
those of the piedra pintal are reported from Colombia and Venezuela (as well as

CONCLUSIONS. 233

from Porto Rico). Balboa's interview with Tumaco, who molded a figure of a
lama in clay, should be recalled here as suggesting Isthmian intercourse with the
far south. Of the ties that bind Chiriqui to the southern continent, the strongest
hitherto revealed are linguistic.

With due allowance for the influences radiating from the great civilizations to
the north and the south, the results of the present study point to the forces from
within rather than to those from without, as being the chief factors in the devel-
opment of Chiriquian culture, which contains many elements of fundamental
importance to a complete history of primitive art.

MKMUIKS CONN. ACAD., Vol. III. 30

BIBLIOGRAPHY.

ACOSTA, JOAQUIN. Compendio historico del descubrimiento y colonizacion de la Nueva

Granada en el siglo decimosexto, segunda ed. Bogota, Colombia, 1901.
ANDAGOYA, PASCUAL DE. Narrative of the proceedings of Pedrarias Davila in the provinces

of Tierra Firme or Castilla del Oro, and of the discovery of the South Sea and the

coasts of Peru and Nicaragua; transl. and ed. by Clements R. Markham. London, 1865.
ARBER, EDWARD (Editor). The first three English books on America; transl. by R. Eden,

from the Latin of Peter Martyr of Anghiera. Birmingham, 1885.
BATEMAN, JOHN F. Account of a visit to the huacas, or ancient graveyards of Chiriqui.

Bull. Amer. ethnol. soc., I, New York, 1860-61.
BERENDT, C. H. Geographical distribution of the ancient Central American civilization. Jour.

Amer. geogr. soc., VIII, New York, 1876.
Be ILLAERT, WILLIAM. Antiquarian, ethnological and other researches in New Granada, Equador,

Peru and Chile, etc. London, 1860.
BRY, J. T. and J. I. DE. Collectiones peregrinationum in Indiam orientalem et Indiam oc-

cidentalem, part. VIII. Francofurti ad Moenum, 1599.
BRINTON, DANIEL G. The American race. Philadelphia, 1901.
BROWN, CHARLES MELVILLE. Tisingal, the lost mine of Panama. Bull. Inter, bur. Amer. republics,

XXX. Washington, 1910.

Bull. Amer. ethnol., soc. I, 11, 12, 21, New York, 1860-61. (Author's name not given.)
CALVO, JOAQUIN BERNARDO. Administracion Soto, Republica de Costa Rica, Apuntamientos

geograficos, estadisticos e historicos. S. Jose de Costa Rica, 1886.
EDEN. R. The first three English books on America, from the Latin of Peter Martyr of

Anghiera; ed. by Edward Arber. Translation. Birmingham, 1885.
EVANS, JOHN. The Chiriqui survey. New York Herald (daily), Dec. 8, 1860.
FEWKES, J. WALTER. Archeological expedition to Arizona in 1895. Seventeenth Ann. rept,

Bur. Amer. ethnol., Pt. 2, Washington, 1895-96.

The Aborigines of Porto Rico and neighboring islands. Twenty-fifth ann. rept,

Bur. Amer. ethnol., Washington, 1903-1904.
FISCHER, H. Bericht iiber eine Anzahl Steinsculpturen aus Costarica. Abh. herausgegeben

vom naturwissenschaftl. Vereine zu Bremen, VII, 1882.
FLOWER and LYDEK.KER. An introduction to the study of mammals living and extinct.

London, 1891.
GABB, WM. M. On the Indian tribes and languages of Costa Rica. Proc. Amer. philos. soc.,

XIV, Philadelphia, 1875.
GAGE, THOMAS. The English- American, his travail by sea and land: or, a new Survey of the

West-India's, containing a journall of three thousand and three hundred miles within

the main land of America. London (R. Cotes), 1648.
GOODYEAR, WM. H. Grammar of the lotus. London, 1891.
GUNTHER, ALBERT C. L. G. Reptilia and Batrachia. Biologia Centrali-Americana, London,

1885-1902.
HAKLUYT, RICHARD. The principal navigations, voyages, traffiques & discoveries of the

English nation, etc., X. Reprint of 2nd. ed., Glasgow, 1904.
HARTMAN, C. V. Archaeological researches in Costa Rica. Stockholm, 1901.

Die Baumkalabasse im tropischen Amerika. Boas anniversary volume. New era

printing company, Lancaster, Pa., 1906.

236 A STUDY OF CHIRIQUIAN ANTIQUITIES.

Archeological researches on the Pacific coast of Costa Rica. Mem. Carnegie

museum, III. No. 1, Pittsburgh, 1907.
Hist, mag., IV, 47, 48, 144, 1860; VI, 154, 1862; IX, 158, 1865, Boston. (Author's name

not given.)
HOLMES, W. H. Origin and development of form and ornament in ceramic art. Fourth ann.

rept. Bur. of ethnol., Washington, 1882—83. (Name changed to Bureau of American

ethnology with sixteenth annual report.)

Ancient art of the province of Chiriqui. Sixth ann. rept., Bur. of ethnol., Washington,

1884-85.

Aboriginal pottery of the eastern United States. Twentieth annual report., Bur.

Amer. ethnol., Washington, 1898-99.
HRDLICKA, ALES. Personal letter, dated Oct. 4, 1906.
IRVING, WASHINGTON. The life and voyages of Christopher Columbus and the voyages and

discoveries of the Companions of Columbus. New York, 1892.
KUNZ, GEORGE F. Gold ornaments from United States of Colombia. American antiquarian,

Sept. 1887.
LUDERS, C. W. Der grofie Goldfund in Chiriqui im Jahre 1859. Jahrbuch der Hamburg-

ischen wissenschaftl. Anstalten, VI, Hamburg, 1888.
LYDEKKER (and FLOWER). An introduction to the study of Mammals living and extinct.

London, 1891.
MAHOUDEAU, P. G. Sur un crane precolombien de la province du Chiriqui (Etats-Unis de

Colombie). Bull. Soc. d'anthr. de Paris, 4e ser., I, Paris. 1890.
MARKHAM, CLEMENTS R. Narrative of the proceedings of Pedrarias Davila in the provinces

of Tierra Firme or Castilla del Oro, and of the discovery of the South Sea and the

coast of Peru and Nicaragua; by Pascual de Andagoya. Translation. London, 1865.
MARSH, O. C. Yale alumni weekly, VII, Jan. 20, 1898.
MARTENS, E. VON. Purpur-Farberei in Central-Amerika. Verhandl. Berliner Ges. fur Anthr.,

Ethnol., und Urgeschichte, Berlin, 1898.
MEAGHER, Thomas F. The new route through Chiriqui. Harper's magazine XXII, New

York, 1861.
MENARD, E. Les poteries des sepultures indiennes du Chiriqui. Chez Pichat a Chatillon-sur-

Seine, 1888.
MERRITT, J. KING. Report on the huacals, or ancient graveyards of Chiriqui. Publ. by the

Amer. ethnol. soc. (New York) previous to vol. I of its Bulletins.
MOOR, EDWARD. The Hindu pantheon. London, 1810.
MORTILLET, GABRIEL DE. Materiaux pour 1'histoire primitive et philosophique de 1'homme,

IV, 65, 1868.
NUTTALL, ZELIA. A curious survival in Mexico of the use of the Purpura shell-fish for dyeing.

Putnam anniversary volume, New-York, 1909.
OTIS, F. M. The new gold discoveries on the Isthmus of Panama. Harper's Weekly, Aug. 6th,

New York, 1859.
OVIEDO Y VALDES, GONZALO FERNANDEZ DE. Historia general y natural de las Indias, islas

y Tierra-Firme del mar oceano. Madrid, 1855.

The first three English books on America (from the Latin of Peter Martyr of

Anghiera) ; ed. by Edward Arber. Birmingham, 1885.
PINART, A. L. Chiriqui: Bocas del Toro, Valle Miranda (avec carte dans le texte). Bull.

Soc. de geogr. de Paris, 7= ser., VI, Paris, 1885.

Les indiens de 1'Etat de Panama. Rev. d'ethnog., VI, Paris, 1887.
PUTNAM, F. W. Conventionalism in ancient American art. Bull. Essex inst, XVIII, Salem,

Mass., 1886.

BIBLIOGRAPHY. 237

RAU, CHARLES. Observations on cup-shaped and other lapidarian sculptures in the Old World

and in America. U. S. geogr. and geol. surv. of the Rocky Mt. region, Contr. to

North Amer. ethnol., V, Washington, 1881.
RECLUS, ARMAND. Panama et Darien. Paris, 1881.

RIEGL, ALOIS. Stilfragen. Grundlegungen zu einer Geschichte der Ornamentik. Berlin, 1893.
SEEMANN, BERTHOLD. Narrative of the voyage of H. M. S. Herald, I. London, 1853.
SELER, EDUARD. Elucidation of Codex Vaticanus no. 3773 (Codex Vaticanus B). Berlin and

London, 1902-03.

Codex Borgia, eine altmexicanische bilderschrift der Bibliothek der Congregatio

de propaganda fide: hersg. auf kosten seiner Excellenz des Herzogs von Loubat.

Erlauterung. Berlin, 1904-09.

Vorlage einer neu eingegangenen Sammlung von Goldaltertiimern aus Costa Rica.

Zeitschr. f. Ethnol., XLI, Berlin, 1909.

SIMONIN, L. La vie souterraine ou les mines et les mineurs. Paris (L. Hachette), 1867.
SIRET, Louis. Les cassiterides et 1'empire colonial des Pheniciens. L'anthropologie, XX,

Paris, 1909.
SMITH, J. H. Observations on the territory of Burica, in the province of Chiriqui, Isthmus

of Panama. Jour. Roy. geogr. soc., XXIV, London, 1854.
SQUIER, E. G. More about the gold discoveries of the Isthmus. Harper's weekly, Aug. 20th.,

New York, 1859.

STEMPELL, W. Die Tierbilder der Mayahandschriften. Zeitschr. f. Ethnol., XL, Berlin, 1908.
SUMICHRAST, F. Contribution a 1'histoire naturelle du Mexique. Bull. Soc. zool. de France, V,

Paris, 1880.
TAYLOR, ALFRED B. Golden Relics from Chiriqui. Proc. Numismatic and antiquarian soc.

of Philadelphia, 75-80, Philadelphia, 1867 (read in 1865).
UHLE, MAX. Verwandtschaften und Wanderungen der Tschibtscha. C. R. Congres intern.

des Americanistes, Berlin, 1888.

Costaricanische Schmuckgerate aus Gold und Kupfer. Globus, LX, Braunschweig, 1891.
WAFER, LIONEL. A new voyage and description of the Isthmus of America, 2nd. ed.

London, 1704.
ZELTNER, A. DE. Note sur les sepultures indiennes du departement de Chiriqui (Etat de

Panama). Panama, 1866.

INDEX.

Acoo river, 53

Acosta, Joaquin, 2, 189, 194, 231

Adze, 22

Agate, 39, 40, 42; beads, 39, 40; pendant,

39, 40

Agua Caliente, 224
Alanje, 96
Allen, J. A., 57
Alligator, 56, 75, 76, 78, 86, 88, 102, 103, 108,

109, 111, 114, 117, 120, 125, 126-133,

136-140, 142-145, 151, 153, 155-159, 162,

166, 178, 179, 198-200, 204, 205, 207, 208,

213, 214, 216, 220, 223, 228-232
Alligator design, conventionalized, 114, 133
Alligator figurine, 204
Alligator-god, 162, 163, 172, 176, 195, 210,

213, 214-218, 232
Alligator group, 46, 56, 66, 102, 117, 125-152,

173,212,230
Alligator head, 158, 159, 198-200, 207, 208,

213-216, 218-222, 231, 232
Alligator motive, 102, 108, 110, 118, 131, 132,

135-138, 140, 141, 143-145, 147, 149, 152,

153, 156, 157, 159, 171-173, 175, 178-181,

184, 186, 214, 215, 219, 221, 222, 228, 230,

232

Alligator punctnlatus, 125
Alligator supports, 86
Alligator symbols, 132
Alligator vase, 127
Alligator ware, 48,82, 109, 117, 129, 133-138,

142, 143, 165-168, 171-175, 177, 178,

180-187, 230, 232
Alloys, 16, 189, 192, 194, 195, 197, 199, 203,

206, 216, 221, 231; natural, 231
Almirante Bay, 2, 17, 189, 190
Alston, Edward K., 206
Altars, 35

Ambrosetti, Juan B., 226
America, 34, 163, 189
America, natives of, 197
American archeology, 19, 227
American ceramics, 45
American Ethnological Society, 36, 38, 44,

104, 193, 204, 211, 217

American Indians, 17

American Museum of Natural History, 57, 223

American pottery, 45

American Zoological provinces, 56

Amulets, 39-43, 220, 221, 231 ; of jade, 228, 232

Andesite, porphyritic, 24

Andesitic lava, 23, 24, 27

Anghiera, 206

Anguizola, Don Antonio, 191

Animal heads, 35, 55, 89, 102, 122, 138

Anteater, 11

Anthropology, Paris School of, 14

Anthropomorphic amulet, 41, 220, 221

Anthropomorphic handle decoration, 70

Anthropomorphic shoulder ornaments, 69

Antiquarian and Numismatic Society of Phila-
delphia, 193, 215

Ape, 52, 209

Apelike figure, 75 ; form, 79, 88 ; head, 223 ;
monster, 209 ; supports, 53

Arched panels, 106, 107, 112, 113, 117, 120,
125, 133-135, 138, 139

Archeology, American, 19, 227

Argentina, 226

Argos, 120

Armadillo, 51, 52, 54, 56-58, 60, 61, 64-66,
69-71, 74, 78-80, 82, 100, 103, 124, 125,
140, 168, 170, 180, 202, 228, 229

Armadillo group, 47-71, 88, 103,125,229

Armadillo motive, 56—65, 75, 78 ; everywhere
plastic, 229

Armadillo supports, 58

Armadillo ware, 19, 47, 49, 50, 52-55, 58-63,
65-70, 73-77, 79, 81, 85, 86, 163, 167, 176,
179, 180, 187, 188, 228

Armadillo whistle, 168

Arrow-points, 21, 22

Asia, 4

Assyrian god, 220

Ateles, 110

Atlantic, 8, 17, 60, 190, 220, 222

Atrato river, 4, 224

Ausgespartes Ornament, 157

Avataras, 218

Avila, Gil Gonzales de, 7

240

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Avila, Pedro arias de (see Davila)

Axe, 39

Aztec village, 105

Azuero, 8

Bachue, 166

Bahia del Almirante, 2

Balboa, Vasco Nufiez de, 1, 4-6, 190, 233

Balza, 170

Balzaria, 170

Barlow, S. L. M., 224

Basanite, 22, 23, 41

Bat, 205, 206

Bateman, John F., 11, 13, 14, 44, 204, 211

Beads, 39

Bear, 82

Belen river, 4

Bell, 169, 198, 201, 203, 208, 217, 221

Berendt, C. H., 17

Berlin, Royal Ethnographical museum of, 200,

219, 226
Bird, 16, 104, 121, 123, 125, 141, 142, 145-150,

176, 177, 179, 180, 185, 187, 200, 201,

206-210, 216-218, 220, 231
Bird figurine, 207, 232
Bird-god, 217, 218
Bird-whistle, 77, 148, 174, 175
Biscuit group, 46, 47, 52, 54, 66, 70, 228
Biscuit ware, 19, 47, 48, 56, 65, 69, 172
Blach incised group, 46, 47, 65, 66, 72, 73
Black incised ware, 71, 212
Blue feather snake, 159
Boar, 218

Boca del Toro, 2, 4, 189
Bocas del Toro 8
Body-line, 108, 128, 140, 153, 155, 156, 158,

159, 171, 178, 179, 229
Body-line motive, 136

Body-markings,73,75, 108,153, 156,159, 171,184
Bogota, 75
Bollaert, William, 43, 44, 199, 203-205, 214,

224, 226
Boquete, 191

Boqueron, district of, 192
Boruca, 8, 17, 60, 232
Borisques, 14
Bottle, 110
Brinton, D. G., 15
British Museum, 193, 201, 209, 214
Bronze, 197, 227

Bronze tweezers, 197

Brown, Charles Melville, 190

Brugiere, 160

Bruncas, 17

Bry, J. T. and J. I, de, 196, 197, 231

Bugaba, 8, 10, 13, 14, 17, 36, 190, 191

Bugavita, 8, 13, 14, 21, 33, 37, 49, 50-53, 55,
91, 102, 103, 117, 119, 122, 129, 134-136,
192, 193, 218; discoveries of gold images
at 192

Bugavita cemetery, size of, 192

Bugavita treasure, 193, 201, 205, 216

Burica, 3, 6, 7, 190

Burnisher, 194

Cabassous centra/is, 57

Calabash, 16, 49-51, 56, 105, 146, 191, 228

Calcareous concretions, 51

Calchaqui region, 226

Caldera, 92, 93, 96, 97, 100

Caldera river, 11, 17, 43

Campanon, Captain Francisco, 7, 8

Canine, 30

Cano Island, 160

Cantaros, 21, 191

Capuchin monkey, 110

Carapace, 57-66, 75. 140, 167, 168, 216 (crus-
tacean)

Carapace meander, 69

Carapace motive, 58-64, 70, 77-80, 100, 103,
124, 140, 180, 229

Carapace symbol, 62-64, 72, 74, 75, 77-79, 85

Careta, a cacique, 4

Cariacus virginianus, 184

Cariay, 2

Cariay Indians, 3

Carib, 15

Cartago,Costa Rica, 1 90,209,224 ; province of,23

Casabindo, 226

Castilla del Oro, 4, 6, 189, 205

Castile, 191

Cateba river, 3, 4, 189

Cathartes aura, 122

Cat, tiger, 182

Cauca, Colombia, 224

Cebus, 110.

Ceiba tree, 191

Celts, 21-24, 26, 29, 220, 228

Cemeteries, 191, 192, 220

Cemeteries and Tombs, 8-14, 46, 192

INDEX.

241

Center, Mr., 38

Central America, 2, 4, 43, 46, 72, 105, 189,
202, 220, 232

Ceramic art, 19, 22, 51, 57, 75, 109, 206, 210,
221; form, 68; technique, 203; type, 67

Cervidae, 202

Chagres river, 8, 14

Chalcedony, 26

Chalice, 159

Changuina-Dorasque stock, 17

Changuinola River, 191

Chapanec stock, 15

Charcoal, 146

Charco Azul, 190

Charles V., 194

Cheriqui, 1, 8

Chiapes, 5, 6, 15, 190

Chibchas, 15, 17, 166, 218

Chicago, Field museum of, 221, 223

Chichimecs, 17

Chinese art, 159

Chircot, 32

Chiriqui, 1, 2, 3, 6-8, 13, 14, 16, 26, 38, 40,
42-44, 92, 111, 155, 161, 163, 165, 201,
204, 205, 212, 218, 220

Chiriqui, archeological possibilities of, 193;
as a region of intensive study, 227

Chiriqui, burying grounds of, 192 ; Commis-
sion, 191 ; fictile products of, 46 ; forest
of, 38; gold images of, 192, 194; gold
plate from, 224 ; Governor of, 191 ; graves,
21, 103, 193; huacalsof, 192: Indians of,
27, 29, 190 ; inhabitants of, 26, 56 ; lagoon,
14, 15, 190, 191, 231; metates of, 27;
Mount, 96, 203 ; mythology of, 194; plains
of, 17; province of, 17, 38, 78, 96, 160,
190, 191; volcano of, 11, 190

Chiriquian art 62, 63, 82, 97, 124, 223 ; cera-
mic, 22, 46, 51, 57, 75, 109, 221

Chiriquian artist, 125, 159, 165, 186; amulets,
39; arrow-points, 22

Cliiriquian culture, boundaries of, 232

Chiriquian deities, 217. 218, 231 ; frontier, 87 ;
gold figurines, 37, 206, 213, 215, 231, 232 ;
gold ornaments, 193. 204 ; goldsmith, 22,
23, 42, 202, 213 ; graves, 19, 193 ; grinding
stone, 34 ; metates, 26, 27, 31-33 ; mytho-
logy, 212, 216, 232; petroglyphs, 43,44;
spear-points, 22 ; totemic animals, 232 ;
waters, 125, 195; whistles, 182, 183
MEMOIRS CONN. ACAD., Vol. III.

Chiriquians, 17, 103, 105, 126, 131, 162, 169,

182, 198, 199
Chisels, 22, 24

Chocolate incised group 47, 66, 95, 96
Chocolate incised ware, 95, 96, 228
Cinteotl, 159
Cipactli, 126, 159
Cist, 11, 13
Ciuacouatl, 166
Classification of Chiriquian pottery, 46, 47,

228; revised classification, 65, 66
Coaza, valley of, 17
Cobiga river, 3
Cobrara river, 3
Code, 191

Codices, Mayan, 71, 184, 232
Coibas, 15, 17
Colombia, 1, 15, 106, 111, 194, 211, 216, 221,

223, 224, 226, 232
Colon, 3, 189

Columbus, 2, 3, 4, 189, 190, 206
Columbus, Luis, 3
Comagre, 4, 5, 6
Compressed meander, 107
Conch, 212

Conch-shell, 169, 170, 185
Conque Persique, 160

Conventionalism,processesof, 127, 130,131,229,
Copper, 16, 189-192, 194, 197, 198, 203, 215,

217, 231
Copper bell, 198; disk, 226 ; figurine, 197 ;

needle, 197 ; ornaments, 189 ; pellets, 68 ;

tweezers, 197, 198
Cora Indians, 66
Cordillera, 15, 190, 227
Cordova, Francisco Hermandez de, 7
Corredor, 151
Corregidor, 191
Corotapa, hills of, 191
Corwing, D. M., 202
Coronado, 17

Costa de las Contrastes, 189
Costa de Veragua, 189
Costa Rica, 8, 13, 15-17, 23, 26, 27, 29, 35,

39, 40-42, 60, 61, 73, 87, 92, 96, 111, 141,

159-161, 189-191, 195, 198, 200, 201, 209,

211, 215, 216, 218, 220, 221, 223, 224,

228, 232
Costa Rica, Atlantic coast of, 17, 220, 222;

Pacific coast of, 160

31

242

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Costa Rican graves, 232 ; highlands, 26, 32 ;

Indians, 27; metates, 31, 33; pottery, 95,

103, 232
Cothos, 14

Crab, 50, 122, 170, 177, 178, 21G, 217
Crab-god, 204, 216, 217, 221, 232
Crab motive, 90, 91 ; whistle, 178, 181
Crayfish, 199-201
Cricamola river, 191, 192
Cricket, gold figure of, 204
Crocodile, 125, 126, 159
Crocodilus americanus yar. acntus, 126
Cuba, 4

Cuenca, Ecuador, 226
Cunas, 15, 17

Cup, 49, 59, 65, 70, 71, 104, 141, 188
Cup-shaped markings, 67
Curridabat (ancient pueblo), 75
Cuttle fish, 195, 200

Darien, Gulf of, 4, 6, 15, 189 ; Isthmus of, 206

David, 6, 8, 21, 43, 44, 49, 51, 96, 101, 117, 140

Davila, Pedrarias, 1, 6, 7

Davis, E. H., 104

Deer, 170, 184, 202

Dezallier d'Argenville, Antoine Joseph (name

misspelled in text), 160

Dicotyles, 168 ; D. labiatus 168 ; D. tajacu, 168
Diquis river, 160
Divala, 21, 43, 49, 50, 54, 67, 72, 95, 101,

106, 108, 109, 112, 119, 121, 122, 132-136,

138, 145, 146, 149, 150, 164, 169, 176, 195,21 1
Dolega, 8, 15, 17
Dorachos, 43
Dorasques, 15, 17
Dordogne, France, 56
Dorsal-view motive, 102, 103, 109, 111, 117,

118, 120, 132-135, 138, 148, 150, 151, 232
Dow, Captain J. M., 36, 217
Drinking jars, 191
Drums, 169, 172
Drum-whistle, 173, 174, 183
Dulce, Gulf of, 61, 160, 232
Duck, 170, 176, 177, 204
Dures, 15

Eagle, 206, 220

Earring, 201

East India, 2

Ecuador, 105, 160, 226, 232

Eden's translation, 194

Effigy, human, 71

Egypt, 78, 212

Egyptian art, 133 ; civilization, 154 ; tomb, 157

Egyptians, 131

El Banco, 21, 49, 55, 92-95, 98, 101, 121, 138,

171, 172, 203
Elliott, W. A., 205
El Volcan, 44
England, Bank of, 193
Enaguas, 161
Engle, Captain F., 191
Escaria, 50, 51, 58, 61, 123, 188
Escudo de Veraguas, 190
Espinosa, 6, 8
Estrella, mines of, 190
Estrella, Rio de la, 191
Ethnographical museum of Berlin, Royal

200, 219, 226

Ethnological society, American, 14, 36, 38,

44, 104, 193, 204, 211, 217
Europe, 154, 163; cave-dwellers of, 165
European museum, 193
Evans, Sir John, 23, 191
Eye motive, 61, 229

Eye ornaments, 63, 101, 109, 111, 113, 122, 123
Eye symbols, 61-64, 87, 89

Faulted meander, 107, 109, 113, 119

Feather-snake, Mexican, 200

Felts onca, 142

felis pardalis, 142

Ferdinand, King, 194

Fernandez, 15

Fewkes, J. Walter, 29, 44, 84, 167

Field Museum, Chicago, 221, 223

Fin motive, 77. 78, 90, 91

Fish, 54, 73, 81, 82, 85, 86-88, 125, 185, 188,

201, 202, 207, 208, 210, 212, 216, 218, 219
Fish group, 47, 65, 66, 76, 80-87, 89, 228
Fish motive, 90; supports, 55, 73, 81-88
Fish ware, 47, 81-91

Flint, 22, 23, 228

Flute, 169-171, 210-212

Florida, 224, 226

Fonseca, city and river, 7

Foot motive, 69, 229

Foot symbols, 61-64, 69, 85, 91, 204

F'oot-warmer, 72

France, paleolithic sculpture of Dordogne, 56

INDEX.

243

Francesca, niece of Luis Columbus, 3

French Panama Canal Company, 14

Fret, 158, 159

Frio, Rio, 15

Frog, 66-69, 89, 101, 125, 171, 195, 198,

202-206, 210, 231 ; ceramic type of, 203;

metallic type of, 195, 203
Froglike supports, 51, 52

Gabb, Wm. M., 15, 16, 192

Gage, Thomas, 160

Ganesha, 218

Gentil, water-color drawing of piedra pintal
by, 43, 44

Gibert, Audinet, 201

Gilt copper, 194

Goblet, 53

God 71 ; of Hindu Pantheon, 218

Gold, 16, 22, 32, 34, 145, 189, 190-192, 194,
195, 197-199, 203, 206, 208, 213, 215, 231

Gold amulet, 42; beads, 201 ; bell, 201, bird,
201 ; eagle, 206 ; figures, 9, 10, 13, 193, 199,
207, 208; figurines, 37, 121, 159, 163, 169,
192, 193, 202-204, 206, 231 ; images, 148,
192^ 194, 196, 203-205, 207-209, 215, 223 ;
masks, 223, 224; mines, Tisingal, 190; mines,
Veragua, 192; nugget, 213; objects, 19,
194, 201, 208; ornaments, 12, 16, 43, 82,
125, 169, 185, 189, 191-193, 197, 201, 204;
parrot, 42 ; plaque, 192, 209, 224-226, 231 ;
plate, 11, 189, 190, 224

Goldsmith, 22-24, 67, 68, 195, 201, 202, 206, 213

Golden ornaments, 12, 16, 49

Golden treasure, 190 ; Huacal de los Reyes,
218, 221; Tisingal, 191

Golfo de Salinas, 160

Goodyear, Wm. H., 133

Gouge, 22

Gourd, 16, 49, 50, 169

Granada, 7

Graves, 11, 13, 19, 21, 103, 145, 193, 194

Greece, branching scroll from, 155; deity
names of, 212; early human figurines
from, 165; traditions of, 157

Greek mythology, 210

Greeks, purple of ancient, 160

Grinding stone, 33, 34

Ground-squirrel, 170, 181

Guaiga river, 3

Guacal, 191

Gualaca, 21, 24, 27, 93, 94, 101, 152; pitcher

from, 153

Guanacosta, 160, 161
Guanaja, island of, 2, 189
Guana vano, 190
Guanin, 2, 3, 189
Guatemala, 106, 160
Guatusos, 15

Guaymis, 15-17, 169, 170, 173, 191, 192
Guerra, Ignacio (creole farmer), 192
Guetares, 15 ; culture of, 26
Guiana, 111; Indians of, 196
Guilloche ornament, 32, 35; pattern, .31, 32,

35, 159

Gulf of Darien, 4, 6, 15, 189
Gulf of Dulce, 61, 160, 232
Gulf of San Lucar, 6, 7

Haiti, 4

Hamburg Museum, 193

Hammond metallurgical Laboratory of Yale

University, 194
Handled group, 47, 48, 78, 80, 81, 86, 92,

167, 228 ; painted, 46-48, 56, 66, 74-80 ;

tripod variety of, 48; unpainted, 47, 66,

72-74, 81

Handled ware, 47, 71, 74, 76-81, 84, 91
Handles, 55, 63-65, 69-72, 74-85, 87-89,

91-93, 96, 97, 99-101, 108, 112, 119, 121,

124, 137, 140, 152

Hanuman, a lesser Hindu deity, 218
Harpers Weekly, 201, 205
Hartman, C. V., 13, 23, 26, 32, 33, 39-42, 92,

105, 106, 146, 159, 160, 220, 221, 223
Harvard University, Peabody museum of,

151, 152, 154
Hatchets, 11, 22, 25
Hawaii, 106
Hawk, 85, 89
Head motive, 124
Head symbols, 124

Heads, grotesque, 53; removable, 147—149
Heidelberg, 19
Hernandez, Melchor, 14
Herring-bone pattern, 158
Heyde, Father, 44

Heye, George G., 32, 193, 201, 204, 229
Heye collection, 33, 36, 37, 201, 202, 204,

207-209, 211, 214, 215, 217, 218, 221,

225

244

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Hieroglyphics, 43, 44, 102, 103, 229

Hindu Pantheon, god of, 218

History, 1-8

Holmes, William H., 19, 26, 27, 35, 36, 38, 43,
45-48, 65, 66, 71, 72, 76, 83, 86, 95-97, 99,
100, 103, 105, 125, 126, 129, 141, 144, 151,
155-159, 164, 172, 183, 198-200, 209, 212,
213, 227, 228

Honduras, 2, 189

Hornfels, 23

Hornstone, 41

Horobaros, 191

Horus, 165, 166

Howland, Meredith, 202

Hrdlicka, Ales, 66

Huacal de los Reyes, 214, 218, 221

Huacals, 8-11, 16, 145, 169, 191-193

Huacas, 9-11, 16, 26, 46

Huista, province of, 6

Human attributes, 186; body, 121, 198, 199;
bones, 14, 194; effigy, 71; figure, 55, 67,
70, 98, 120, 142, 213, 216; figurine, 147,
165-167, 205, 211, 212, 232; form, 36,
52, 55, 67, 70, 78-80, 86, 89, 90, 97, 120,
208; hair, 14; head, 35, 69, 70, 187, 188;
remains, 14; skull, 14; supports, 87, 89,
90, 97, 98

Hurtado, Captain Benito, 7

Husita, island, of, 2

lea, 185

Idols, 35

Iguana, 170, 179

Images, 21, 36-39, 196, 203, 206, 208

Imperial museum of Natural History, Vienna,
218

Implements, 19, 22-25, 34

Incas, 6

Incised areas, 78 ; band, 88, 89, 97 ; design
32 ; group, 72, 95 ; lines, 27, 72-74, 85-87,
95, 96, 99 : motive, 75 ; ornament, 38, 76,
95; pattern, 29, 32, 72, 79, 81, 86, 96, 98

Indian Ocean, 160

Indiana, 34

Indians, 160, 189, 191, 206; American, 17;
Chiriqui, 27, 29, 190; Cora, 66; Costa
Rica, 27; Guiana, 196; Isthmian, 15; Pa-
nama, 48 ; Santa Marta, 194 ; Talamanca,
27; Terrora, 190; Tlingit, 211

Instruments, musical, 11, 177-188

Iron, 191

Irving, Washington, 5, 224

Isis, 165, 166

Isthmus, 1, 4-6, 8, 14, 15, 57, 190, 193, 206;

fauna of, 168
Isthmian Indians, 15; gold figurines, 193;

provinces, art of, 17; region, 189
Italian paintings, 56
Italy, 157
Izalco, Salvador, 105, 146

Jacu, 21, 67, 138, 141, 167, 174

Jade, 40—42 ; celt-shaped amulets of, 232 ;

ornaments, 228; supply, 228
Jaguar, 26, 27, 29-33, 35, 82, 89, 140-142,

144, 182-184, 186, 204, 205, 208, 209, 223,

231

Jaguar-god, 82, 141, 143, 221-223, 232
Jaguar metate, 30, 31, 33
Jaguar- whistle 83, 184
Jamaica, 173

Jasper, 22 ; bead, 39, 40 ; polishing stone, 26
Java, 106

Keith, Minor C., 36, 60, 141, 195, 198, 208,

211, 218, 220, 223, 232
Keith collection, 16, 29, 35, 87, 106, 161,

200, 209, 214, 218, 219, 221, 222
Keller, Ferdinand, 163
Kluge, J. P., 14

Kunz, George F., 195, 224, 226
Kurma, incarnation of Vishnu, 218

Labret, 112, 201

Labret ornament, 112, 122

Labret-shaped figure, 115; motive, 114

Lama, 53, 233

Lamarck, 160

Lamson, Edwin, 28, 29, 127

Lamson, William J., 198

Lamson Brothers, 41

Lamson collection, 41-43, 166, 171, 195, 198,

211

Las Brenas, 190
Las Casitas, 33
Las Guacas, Nicoya, 26, 33, 39, 41,42, 220;

celt-shaped jade amulets from, 232 ; met-

ates from, 159
Lava, andesitic, 23, 24, 27
Le Moyne, Jacques, 226

INDEX.

Lenox, James, 205

Lenox Library, New York, 193, 205

Leon, the city of, 7

Life form, 64, 66, 74, 75, 78-80, 86, 87, 89,

91, 102, 109-112, 124, 125, 136, 138, 139,

141, 208; motive, 81, 104, 110; symbols,

74, 75
Linen, 160
Lizard, 70, 101
Loin-cloth, 53, 163, 213
Lost color group, 46, 56, 66, 73, 75, 76, 102,

103-120, 122-125, 150, 151, 212, 230
Lost color process, 104, 122, 124, 125, 230, 232
Lost color technique, 230
Lost color ware, 80, 104, 106, 110, 111, 113-

116, 118-121, 128, 134, 151, 161, 164, 169
Lotus, 78, 154
Luders, C. W., 192, 199, 201, 205, 220, 222

Macanas, 2

Macaw, 206

Machete, 16

Macuilxochitl-Xochipilli, 166

Madrid, Museum of, 193

Magnesia, oxide of, 146

Man, 35, 36, 52, 75, 89, 170, 185, 204, 208,

212, 216-218
Mantas, 49

Maroon group, 46, 66, 97, 100, 101
Maroon ware, 100, 101
Marsh, Othniel C., 19, 26, 38
Martens, E. von (name misspelled in the

text), 160
Martyr, Peter, 206
Masks, gold, 223, 224
Mathewson, C. H., 194
Matsya, an incarnation of Vishnu, 218
Mayan codices, 71, 184, 232; mythology,

182, 184; picture writing, 66
Mazatlan, 57

McNielJ. A., 12-14, 19, 21, 26, 43,51,96, 145, 150
McNiel collection, 41
Meagher, Thomas F., 8, 192
Mealing plate, 31 ; stones, 21, 26-30, 38
Meander, 107-109, 113, 118, 119
Mediterranean, 160
Menard, E., 14
Mercedes, 16, 17, 29, 35, 36, 60, 106, 141, 161,

208, 209, 211, 220, 222, 223, 232
Meredith Howland collection, 203, 211

Merritt, J. King, 8, 11, 13, 14, 21, 22, 26, 192, 193

Mertens, Frau Alice, 200, 219

Metal, 19, 68, 189-226

Metal figurine, 199 ; frog, 68, 201 ; instru-
ments, 95; objects, 193, 197; pellet, 198,
200; technique, 23, 67

Metallurgy, 231

Metates, 16, 19, 21, 26-35, 159, 167, 227, 228

Metropolitan Museum, 200-204, 211

Mexico, 15, 17, 30, 43, 212, 161, 163, 189, 201,
212, 226-228

Mexican antiquities, 201; deities, 159; feather-
snake, 200; god, 183; Indians, 66 ; inter-
preter, 1 7 ; picture writing, 126; table-land,57

Mineral de Veraguas, 190

Mines of Estrella, 190 ; Tisingal, 190 ; Veragua,
192

Mingindo river, 224

Minorca, fishermen of, 160

Miravalles, Costa Rica, 40, 41

Mississippi valley, scroll ornament in, 154

Model, resin, 195; wax, 195

Molar teeth, human, 14

Monkey, 35, 36, 52, 75, 89, 91, 110, 122, 162,
170, 185, 208-210. 218, 231

Monkey heads, 62 ; figurines, 232

Monster, 82, 186, 231

Monuments, stone, 43

Moore, Clarence B., 226

Mortar, 16, 35, 36

Mortillet, Gabriel de, 34, 45

Mosquito coast, 2

Mt. Chiriqui, 96, 203

Mt. Royal, Florida, 226

Moves, dialect, 15

Multiple motive, 156

Munoz, Captain, 191

Muois dialect, 15

Murex brandarius\Jm\\e, lQO;M.(runcu/us,i60

Museum, British, 193, 201, 209, 214; Field
Chicago, 221, 223; European, 193; Ham-
burg, 193; American, Natural History, 57,
223 ; Imperial Natural History, Vienna, 218;
Madrid, 193; Metropolitan, 200-204, 211;
Peabody, of Harvard University, 151, 152,
154; Royal Ethnographical, Berlin, 200,
219, 226; San Jose, 209, 216. 224; U.S.
National, 22, 100, 151, 156, 160; University,
of Quito, 226; Yale, 22, 160

Musical instruments, 11, 165, 169, 177-188

246

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Mycenae, 224
Mycetes, 110

Mythology. Chiriquian, 212, 216, 232, Greek,
210, Mayan, 182, 184; Mexican, 232

Nahuatl, 15

National Museum, (U.S.), 22, 100, 151, 156,
160; of San Jose, 209, 216, 224

Needle, 197

Needle-cases, 104, 124, 164

Nelson, William, 14

Neolithic period, 45

Nequen, a coarse net for catching swine, 7

New Andalusia, 4

New England, cemeteries of, 220

New Granada, 1, 15, 141

New York, 32, 193, 200-202, 204, 205

Nicaragua, 2, 6, 7, 15, 189, 223, 232

Nicaraguan pottery, 174

Nicoya, 17, 26, 42, 43, 159, 160, 220, 221,
228, 232 ; peninsula of, 26 ; province of, 27.

Nicoyan amulets, 39 ; culture, 232 ; grinding
stone, 33 ; type of metate, 27

Nicuesa, Diego de, 1, 4

Nisca, 7

Nombre de Dios, 4

North America, 15

Northumbrian petroglyphs, 43

Numi, 16

Numismatic and antiquarian Society of Phila-
delphia, 215

Nufiez, Vasco (see Balboa)

Nuttall, Zelia, 161

Nyctipilhecus, 110

Obaldia, Senor, 141

Obsidian, 23, 228

Ocarina, 170

Ocelots, 35

Ohio, 34

Ornithomorphic amulet, 41, 42, 220, 221;

character, 145; vessel, 146
Oro, Castilla del, 4, 6, 189, 205
Orosi, Costa Rica, 92
Orotiftans, 15
Otis, F. M., 192, 193
Outina, 226
Oviedo y Valdes, Gonzalo Fernandez de, 1,

7, 8, 43, 194, 231
Owl, 87-89, 148, 170

Owl supports, 87
Owl-like heads, 62

Pachacamac, tomb at, 197

Pacific Coast, 17

Pacific Ocean, 4, 5, 6, 8, 227

Painted Handled group, 76-80

Painted Pandled ware, 74, 77-80

Painted stone, 43

Painting, systems of, 157, 230

Paleolithic engraving, 56 , frescoes, 56 ; period,

45 ; sculpture, 56
Pan, god, 210
Panama, 1, 2, 6-8, 14, 15, 26, 38, 57, 144,

160, 181, 184, 190, 192, 202, 204
Panama Canal, 4, 8; French Company, 14
Panama Indians, 48
Panama Railroad Company, 38
Panama Venus, 38, 162
Pantheon, gods of Hindu, 218
Paraguay, 57
Paris, 19

Parrot, 148, 170, 173, 174, 194, 206, 218, 219, 221
Parrot-god, 121, 217-220, 223, 232
Parrot, gold, 42
Partridge, 170, 176
Paso Real, 87
Pasto, 15
Peabody Museum of Harvard University, 151,

152, 154
Peccary, 168
Pedrarias (see Davila)
Pellets,50,51,55,68,75,82,85,95,143,l78,182,183,

219; calcareous concretions as, 51 ; clay, 51,

68; copper, 68; gold, 201, 207; metal, 198,

200

Peru, 15, 46, 72, 104, 189, 197, 227, 228
Peruvian art, 159, 185, 210; influence, 53;

skulls, 14; pottery figurine, 186
Pestle, 16, 21, 34, 194
Petroglyphs, 21, 43, 44, 227, 232; Colombia,

44 ; Northumbrian and Scottish, 43 ; Porto

Rican, 44 ; Venezuela, 44
Phenicians, purple of ancient, 160
Phidian age, 56
Philadelphia, 216
Philadelphia, Numismatic and Antiquarian

Society of, 193, 215
Phytomorphic type 53 ; vessels, 56
Picture writing, Mayan, 66 ; Mexican, 126

INDEX.

247

Piedra pintal, at Caldera, 43, 44, 232
Pinart, Alphonse, L., 12, 17, 43, 96, 169, 173,

191, 192, 194
Pine pitch, 146
Pipa americana, 66, 67
Pita, a kind of thread, 160
Pitcher, 152, 153

Pitti, Victorio (creole farmer), 192
Pittier de Fabrega, H., 160
Plaques, gold, 192, 207, 209, 224
Plates, 16, 31; gold; 189, 190
Platinum, 191
Point Barrow, 46
Polychrome group, 46, 66, 71, 150, 151-161,

230
Polychrome ware, 76, 109, 151, 152, 159, 161,

228, 229

Polishing stones, 22, 26
Pompeii, branching scroll from, 155
Ponca, 4, 5
Poncra, 6

Porphyritic andesite, 24
Porras, Diego, de, 3
Porto Bello, 68
Porto Rico, 233
Porto Rican petroglyphs, 44
Potrero, 11

Pourpre de Panama, 160
Prehistoric repairing, 167
Procyon cancrivorus, 111; P. lotor, 111
Prototype, 67

Public Library of New York, 204
Pueblo, 75, 85, 154
Pueblo Viejo, 204, 221
Puerto de los Bastimientos, 3, 4
Puerto Bello (same as Porto Bello), 3, 4
Puerto del Retrete, 3, 189
Puma, 170
Punta Burica, 190
Purple industry, 160
Purpura, 161; P. hcemastonia Linne, 160;

P. patula, 160; P. per ska, 160
Putnam, F. W., 83

Quaraqua, 5
Quartzite, 25
Quetzal, 16
Quetzalcouatl, 159
Quibi, 3
Quiriviri, 2

Racoon, 111

Raleigh, Sir Walter, 196

Rattles, 50, 51, 68, 82, 104, 124, 142, 143, 165,

169, 198, 210-212, 220-222
Rau, Charles, 43
Reclus, Armand, 126
Red line group, 46, 47, 66, 92-95, 101
Red line ware, 47, 93, 94
Read, C. H., 193
Removable head, 147-149
Repouse, 225
Reptiles, 179, 216
Resin, 195
Resin frog, 195 ; model, 195 ; plug, in the

head of a gold figurine, 195, 222
Reyes, Huacal de los, 214, 218, 221
Rhodes, branching scroll from, 155
Ribera, Don Perafan de, 192
Riegl, Alois, 155
Rio Bamba, Ecuador, 106, 232
Rio Chiriagua, 96
Rio Chiriqui Viejo, 11, 207, 232
Rio de la Estrella, 191
Rio Fonseca, 7
Rio Frio, 15

Rio General, 195, 198, 200, 209, 214, 218, 221, 222
Robalo, 17
Rodent, 187

Rosette-like ornamentation, 115, 118
Royal Ethnographical museum of Berlin, 200,

219, 226

Rubbing stone, 27, 31, 33
Ruminant, 184

Sabaneros dialect, 15

Sand, used in casting, 198

Sandstone, stool of, 36

Salmon colored sub-group, 66, 71, 73

Salvador, 105, 106

San Carlos, 38, 140

San Jose, National Museum, 209, 224

San Juan de Nicaragua, 2, 189

San Lorenzo, 7

San Lucar, Gulf of, 6, 7

San Miguel, 5, 44

St. Andres mountains, 33

S. Elena, 160

Santa Marta, Indians of, 194

Santiago de Veragua, 8, 190

Santo Domingo, 29, 167

248

A STUDY OF CHIRIQUIAN ANTIQUITIES.

Saville, Marshall H., 44, 106, 226
Scale-group motive, 126, 127, 131, 134, 146-148,

229
Scale-group symbols, 136, 138, 139, 143, 175,

216, 221, 230, 232

Scale motive, 131, 133, 134, 200, 212
Scalesymbols,131,133-135,137-139,147,179;230
Scarified group, 46, 66, 96-100
Scarified ware, 96-99, 164
Schliemann, 163, 224
School of Anthropology, Paris, 14
Scorpion, 170, 178, 179
Scottish petroglyphs, 43
Scroll, 89, 108, 139, 152, 154, 159 ; branching,

152,154, 155; current, 155; direct, 154;

reverse, 154, 155 ; sigmoid, 153-155, 229,

232; Vitruvian, 154
Sculpture, Phidian age of Greek, 56
Sculptured columns, 21, 44, 227
Sea shells, 204; necklaces of, 43
Seats, stone, 29, 35, 36
Seemann, Berthold, 14, 43, 44, 48, 57, 68,

126, 168, 184, 190
Seler Eduard, 161, 215
Serpent group, 47, 65, 66, 71, 72, 228
Serpent motive, 71, 73, 75, 109, 112
Serpent ware, 72, 73, 112, 212
Shark's tooth, 42, 43 ; perforated for sus-
pension, 227
Sheet gold, 195, 205
Sikyatki, ruined pueblo of, 84
Silver, 194, 206, 231
Simonin, L., 22, 23
Single handle, 63, 65, 70, 71, 77, 79, 80, 82,

93, 100, 101, 112, 121, 152
Skil, of Tlingit hats, 37
Slip, 47,55,71, 76,77,91-93, 95,104, 118,139,141
Smith, J. H., 190
Society, American Ethnological, 14, 36, 38, 44,

104, 193, 204, 211, 217; Numismatic and

Antiquarian, of Philadelphia, 193, 215
Soes, Don Roberto, 8, 192
Sojo, Don Diego de, 190
Soto, 7

South America, 1, 2, 6, 15, 57, 189, 224
Spear-points, 21, 22
Spain, 160

Spaniards, 3, 5-7, 16, 160, 189, 190, 192
Spanish America, 4, 8
Spanish Conquest, 189, 206, 231 ; trinkets, 189

Spindle-whorls, 163, 164, 171, 172

Spine motive, 126, 128, 131, 134, 136, 200,
212, 229, 232

Spinesymbols,133,135,137, 138,144,147,221,230

Spool, 112, 122

Spool-shaped motive, 114, 115

Squier, E. G., 3, 193, 197

Squirrel, 170, 180, 181

Stamp, 163, 164

Starfish, 119

Statues, 38, 223

Statuettes, 21, 227

Stearns Collection, 198

Stempell, W., 184

Stools, 19, 21, 27-29, 35, 36, 162, 163, 166,
167, 209, 227, 228; stone, 19, 34, 36, 89,
141, 209; terra cotta, 89; tripod variety
of, 27; wooden, 29

Stone, 19, 21, 32, 34, 43, 208, 212; grinding,
33, 34; hand, 27; mealing, 21, 26, 27, 29,
30, 38; painted, 43; polishing, 21, 26;
rubbing, 27, 31, 33

Stone amulet, 42; art, 19, 21, 22; axes, 21 ;
carvings, 35; figurine, 211; hatchets, 11;
images, 21, 148; implements, 19, 21, 22,
24, 34, 193 ; metates, 29 ; monuments, 43 ;
seats, 29, 35, 36; slabs, 223; statues, 38,
223; stools, 19, 34, 36, 89, 141, 209

Stuart (Robert L.) collection, 205

Sukia, magicians employed by the Guaymis, 17

Sumichrast, F., 125

Summit, N. J., 28, 127, 166, 198

Surinam Toad, 66, 67

Swiss lake-dwellings, 163

Syria, branching scroll from, 155

Tail symbols, 62, 64

Tail motive, 61, 62, 229

Talamanca Indians, 27

Talamancas, 17

Tapir, 98, 144, 145, 168, 170, 181, 182

Tapirus bairdii, 144, 168, 182 ; T. dowii, 182

Tain novemcitictum, 57 ; T. Iricinclum, 57

Taylor, Alfred B., 193, 215

Taylor, C. E., 96

Tea-caddy, vase in shape of, 117

Teaochan, 6

Tehuantepec, 161

Tepeyollotli, 183

Terraba, 8, 17, 60, 195, 214, 232

INDEX.

249

Terra Cotta group, 46, 47, 54, 66 ; stools, 89
Terra Cotta ware, 19, 46, 56
Terra del Fuego, 46
Terrora, Indians of, 190
Texas, 57

TezcatHpoca, priests of, symbol of, 226
Thevenet, J. 22
Tiger, 204
Tiger cat, 182
Tin, 197

Tiribi, 16; chief of, 192
Tisingal, 190, 191
Tinten-fisch, 200
Tlingit hats, 37; Indians, 211
Tombs and cemeteries, 8—14, 43
Tonacatecutli, 159
Tonalamatl, 183
Tonatiuh, 159
Tortoise, 218
Totten, Mr., 38
Touchstone, 22, 23

Treasure, of Bugavita, 192, 193,216; ofHuacal
de los Reyes, 218, 221 ; of Tisingal, 191
Trefoil pattern, 123, 124
Tres Rios, 73, 95
Trinidad, 2

Tripodgroup, 46-48,54,64-66,79-87,89,92,228
Tripod ware, 47, 87
Troy, 163

Troyo collection, 224
Tubanama, 6
Tufa, 23, 24
Tuff, volcanic, 34
Tumaco, 5, 6, 190, 233
Tunic, of Zapotecan women, 161
Tunja, 15

Tusayan potters, 84
Tweezers, 197, 198

Uhle, Max, 15, 216, 226

Ulloa, Don Antonio de, 160, 161

United States National Museum, 22, 100, 151,

156, 160

University of Quito, Museum, 226
Unpainted Handled ware, 74
Urira river, 4, 189
Urn-shaped vessel, 91
Utelaes, 15
Utley, F. D., 33, 204, 221

Valientes dialect, 15

Valle Miranda, 191, 192

Vampire, 192, 206

Vampyrus spectrum, 206

Varaha, incarnation of Vishnu, 218

Venado, 184

Venezuela, 232 ; petroglyphs, 44

Venetian renaissance goblet, 53

Veragua, 3, 4, 8, 189-191, 224; Costa de,

189; Duke de, 3; Francesca de, 3; gold

mines of, 192 ; river, 4 ; Santiago de, 8
Veraguas, 3, 43 ; Governor of, 190
Vienna, Imperial Museum of Natural History,

218

Vishnu, 218
Vivala, huacal, 145
Volcanic ash, 23, 24, 34; rock, metates of,

27; tufa, 23, 24; tuff, 34

Wafer, Lionel, 173

Warree, 168

Wax model, use of in casting, 195

White line group, 46, 47, 66, 102, 103

White line ware, 102, 103

Whistles, 104, 120, 124, 148, 151, 169-188, 210

Wind-instruments, 169, 170, 173

Wooden mortar and pestle, 16

Wooden stools, 29

Xiuhcouatl, 159

Xochipilli, 159

Xochiquetzal, Mexican goddess of flowers, 166

Yale, 151-153, 194, 213, 214

Yale University collection, 19, 22,38,47,48
51, 63, 65, 69, 71, 76, 86, 145, 150, 151,
156, 158, 163, 172, 193, 195, 201, 211, 223

Yale Museum, 22, 160

Zapotecan women, 161

Zaribas, 15

Zeltner, A. de, 11, 13, 14, 19, 34, 142, 150,
155, 157-159, 211 ; collection, 19, 141, 158

ZoO morphic amulet, 42 ; class, 56 ; effect, 29,
64 ; features, 53 ; handle, 70 ; shoulder
ornaments, 68; supports, 35, 51, 52; type,
27, 53; unit, 26, 27, 30, 56, 100, 104, 139,
198; vessel, 54, 145, 182, 184

Zunes, 15.

Plate II

Celts made of black fine-grained volcanic tufa. Page 23.

Figure a. — Chipped celt, the only polished areas being the facets forming the
edge. 2/3

Figure b. — Variant form of celt with marked constriction in the blade. 2/3

Figure c. — Celt with three polished facets on each side. 2/3

Figure d. — Large celt of the pecked type, in which the polishing extends along
a median line from edge to pole. 2/3

Figure e. — Small celt of the pecked type, only the bevels at the edge being pol-
ished. 2/3

Figure /. — Typical celt with distinct demarcation between the pecked and the
polished facets. 2/3

Figure g. — Highly artistic celt with all facets completely polished. 2/3

Figure h. — Celt in which the chipping has not been entirely removed by pol-
ishing. 2/3

Figure i. — Variant form of celt with long straight edge. 2/3

MEM. CONN. ACAD., VOL. Ill

PLATE

d

Plate III

Large fine jaguar metates of stone. Page 32.

Figure a. — Graceful metate with well modeled head and with original patterns on
legs and tail. l/4

Figure b.— Large metate with broad oval plate and relatively short tail. J/e
Figure e. — Large metate with plain legs. V?

MEM. CONN ACAD., VOL. Ill

PLATE I!

cl

Plate IV

Series of stone stools. Page 35.

Figure a. — Rude stool with concave top, the margin being decorated with nine

animal heads. J/4

Figure b. — Stool with decorated border from which hang four jaguar-like heads. !/J
Figure c. — Tall stool supported by four round legs, near the top of which is a human

head in relief. Thirty small animal heads encircle the seat ; from Vivala. '/i
Figure d. — Low stool supported by four monkey-like figures. J/i
Figure d'. — Another view of the same stool, showing concealment of a mutilation. */4
Figure e. — Stool with mortar-like top supported by monkey and human forms. J/4

MEM. CONN. ACAD., VOL. Ill

PLATEJjIV

/ -S ,-

d

d'

Plate V

Variations of the calabash type of vessel. Armadillo ware. Page 50.

Figure a. — Bowl with incurved rim. */a

Figure b. — Bowl depressed vertically, with angular outlines: from Escaria. 2/»

Figure c. — Bowl with vertical sides ; from Bugavita. '/2

Figure d. — Bowl-shaped vase with vertical rim ; from Divala. Vs

Figure e. — Tall vase with pointed base and flaring rim ; from Divala. '/a

Figure /. — Vase with flaring rim and recurved lip. J/a

Figure g. — Vase with squarish lip around circular opening. Va

Figure h. — Vase with collar; from Divala. Vs

MEM. CONN. ACAD., VOL. Ill

PLATE V

a

S

f

Plate VI

Series of tripods. Armadillo ware. Page 51.

Figure a. — Tripod with angular bowl and supports slit laterally. J/a

Figure b. — Rare double tripod, the connection serving as a rattle. 2/6

Figure c. — Tripod, the supports of which show each four vertical slits; from Buga

vita. Va

Figure d. — Tripod with multiple slits in supports. */a
Figure e. — Artistic tripod much discolored by grease and smoke. Va
Figure /. — Tripod with animal head on the shoulder and multiple punctures in

supports ; from Bugavita. '/a

Figure g. — Angular bowl with multiple punctures in supports ; from Bugavita. l/a
Figure h. — Tripod with multiple punctures in supports ; from Escaria. Va

MEM. CONN. ACAD., VOL. Ill

PLATE VI

a

g

Plate VII

Series of tripods in which the supports suggest the human form.
Armadillo ware. Pages 52, 53.

Figure a. — Tripod showing only slight modification of an otherwise plain support, '/a

Figure b. — Tripod with bifurcated supports representing human legs, '/a

Figure c. — Tripod, the supports of which are decorated with the human head and

extremities. J/a
Figure d. — Tripod, each support of which is decorated with two human heads and

arms. Va

Figure e. — Tripod with supports but slightly altered to suggest the human form. J/a
Figure /. — Artistic tripod in which the human form of support is recognized by

the loin-cloth. Va
Figure g. — Tripod in which the loin-cloth is the distinguishing feature of the

human form of support; from El Banco. Va
Figure h. — Tripod whose supports are grotesque representations of the human

form; from Bugavita. */a

MEM. CONN. ACAD., VOL. Ill

PLATE VI!

a

Plate VIII

Tripods with grotesque heads serving as shoulder ornaments.
Armadillo ware. Page 53.

Figure a. — Tripod with a single grotesque figure attached to the rim. Va

Figure b. — Tripod with two animal heads attached to the rim. Va

Figure c. — Bowl with animal head and tail decorating the rim and with four

supports, the whole representing a zoomorphic unit. 2/5
Figure d. — Tripod with animal figures resting on extensions of the rim ; from

Bugavita. 2/6

5

Figure e. — Tripod with entire animal forms as shoulder ornaments. 2/
Figure /. — Tripod with the frog as shoulder ornaments ; from Bugavita. Va
Figure r/. — Tripod, the body of which is somewhat altered by the representation

of animal heads at each end. Va

Figure, h. — Tripod with zoomorphic features encircling the equatorial zone : from
Bugavita. Va

MEM. CONN. ACAD., VOL. Ill

PLATE VIII

S

f

a

Plate IX

Series of tripods, the supports of which represent the armadillo.
Armadillo ware. Pages 58, 59.

Figure a. — Angular bowl with supports representing the armadillo in the round. */a
Figure b. — Angular bowl with supports incised to represent the carapace, the head

and feet being added. Va
Figure c. — Bowl in whose supports the armadillo is recognized by incisions for

the carapace, and incised nodes for the forefeet. 1/a
Figure d. — Bell-shaped bowl supported by conventionalized forms of the armadillo,

recognized by head and carapace. 1/a
Figure e. — Tripod, the human nailprints on the supports of which represent the

carapace of the armadillo. Va

Figure /. — Angular bowl with supports representing the armadillo. Va
Figure #.— Tripod in whose supports the armadillo is recognized in the plain nodes

representing head and forefeet. 1/a
Figure h. — Tripod bowl in whose supports the armadillo is recognized by a much

reduced head only ; from Divala. Va

Figure i. — Tripod with simple supports horizontally incised to represent the cara-
pace. Va

MEM. CONN. ACAD., VOL. Ill

PLATE IX

a

Plate X

Series of vases in which the armadillo head is employed as a shoulder ornament.

Armadillo ware. Page 60.

Figure a. — Vase with the armadillo head and forelegs as shoulder ornaments. *l&

Figure b. — Vase with armadillo heads as shoulder ornaments, the forelegs being
reduced to mere nodes ; from Bugavita. J/z

Figure c. — Vase showing the armadillo head and conventionalized foot as shoulder
ornaments. Va

Figure d. — Vase with large simplified armadillo heads as shoulder ornaments, the
conventionalized foot being absent ; from Bugavita. l/a

Figure e. — Vase in which the armadillo head and feet employed as shoulder orna-
ments are much reduced ; from Bugavita. 2/6

Figure /. — Vase showing extremely reduced and simplified armadillo heads as
shoulder ornaments. V»

MEM. CONN. ACAD., VOL. Ill

PLATE X

a

Plate XI

Series of vases in which the entire armadillo is employed as a shoulder decoration.

Armadillo ware. Page 61.

Figure a. — Vase, the armadillo serving as shoulder ornaments, all three divisions

of the carapace being represented. '/a
Figure b. — Small vase in which the armadillo is placed lower on the shoulder and

in a burrowing attitude ; from Escaria. */»

Figure c.— Vase showing the characteristic burrowing attitude of the armadillo. Va
Figure d. — Upper portion of vase showing that the armadillo foot and the incised

nodes on the neck of the vase are identical. Va
Figure e. — Vase in which the shoulder ornaments represent the armadillo with

characteristic incised tail. Va
Figure /. — Vase with the armadillo as shoulder ornaments and with the tail motive

as a neck ornament. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XI

Plate XII

Series of collared and double-mouthed vases. Armadillo ware. Page 63.

Figure a. — Collared vase with incised zone, '/a

Figure I. — Collared vase. Va

Figure c. — Double-mouthed vase showing that the eyes and carapace of the arma-
dillo are identical with the decorative motives on the necks of the vase :
from Bugavita.

Figure d.— Plain double-mouthed vase, in which the paired necks are set rather
far apart, '/a

Figure e.— Double vessel in which the necks are fused. \/2

Figure /. — Tripod vase with shoulder ornaments representing the entire arma-
dillo; from Escaria. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XII

a

Plate XIII

Series of vases in which the handles are decorated with armadillo motives.

Armadillo ware. Page 64.

Figure a. — Vase, the paired handles of which represent the armadillo. Va
Figure b.— Vase with paired handles entirely covered with carapace motives. l/*
Figure c. — -Vase in which the paired handles are ornament with nodes and fillets

representing armadillo motives. Va
Figure d. — Vase with paired handles, on which the armadillo motives are much

simplified. J/a

Figure e. — Vase with paired handles simply incised to represent the carapace, '/a
Figure /. — Vase with a single handle, the neck and the handle being decorated

with applied tail symbols and eye symbols. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XIII

Plate XIV

Series of vases in which the frog is employed as shoulder ornaments.

Armadillo ware. Page 67.

Figure a. — Vase showing admirable adaptation of ornament to form ; from Jacu. x/a

Figure b. — Vase on which the representation of the frog, especially its flattened
feet, suggests the technique in metal rather than in clay. */a

Figure c. — Vase with a pair of frogs on each shoulder, the hind feet being re-
presented by a flat band of clay, '/a

Figure d. — Beautifully modeled vase in which the metal-like frogs serving as
shoulder ornaments cover oval openings in the wall of the vase. J/2

Figure e. — Vase with elongated neck, and shoulder ornaments representing the
frog. V«

Figure /. — Small vase with shoulder ornaments representing a ceramic type of
frog. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XIV

a

Plate XV

Series of vases showing a variety of plastic shoulder ornaments.
Armadillo ware. Pages 69, 70.

Figure a. — Vase with monkeys serving as shoulder ornaments, '/a

Figure b. — Vase with shoulder ornaments representing the lizard. lls

Figure c. — Vase with two catlike creatures figuring as shoulder ornaments. J/a

Figure d. — Remarkable form of vase in which the diameter of the rim is much

greater than that of the body. Va

Figure e. — Vase, the shoulder ornaments of which resemble the armadillo. ]/2
Figure /'. — Vase with squarish lip and with shoulder ornaments representing a form

of bird. 2/6

MEM. CONN. ACAD., VOL. Ill

PLATE XV

a

Plate XVI

Series of vase forms selected from the salmon-colored variety of
armadillo ware. Page 71.

Figure a. — Vase with shoulder ornaments that may represent the armadillo head

encircled by the carapace motive. Va

Figure b. — Rare vase with loop handles and flat bottom. 1k
Figure c. — Vase with elongated neck and plastic shoulder ornaments ; from Buga-

vita. */a
Figure d. — Vase with bulging collar on which the plastic shoulder ornaments rest;

from Vivala. Va

Figure e. — Vase with incised collar and no shoulder ornaments. 2/6
Figure f. — Vase in which the plastic figures serve as handles connecting the rim

and shoulder, '/a

MEM. CONN. ACAD., VOL. Ill

PLATE XVI

a

Plate XVII

Typical examples of serpent ware. Pages 72, 73.

Figure a. — Characteristic vase with handles, the incised decoration representing
a two-headed serpent with forked tongue ; from Vivala (not Divala). 3/4

Figure b. — Vase in which the incised two-headed serpent encircles the body nearly
three times. ZU

Figure c. — Vase on which the serpent motive survives as a purely geometric

pattern ; from Divala. 3U

Figure d. — Vase ornamented with geometric serpent symbols ; from Divala. 3/.j
Figure e. — Vase with incised geometric decorations derived from the serpent. s/4
Figure /. — Vase with no incised decorations. */a

MEM CONN. ACAD., VOL. Ill

PLATE XVII

f

d

a

Plate XVIII

Series of typical unpainted handled vases in which armadillo
motives predominate. Page 74.

Figure a. — Vase in which the handles represent the armadillo. 2/6

Figure b. — Vase, the handles of which are decorated with armadillo motives. 2/s

Figure c. — Vase in which the handle is reduced to a mere knob, only the eyes

and the ventrally placed carapace of the armadillo being represented. 2/6
Figure d. — Vase in which the reduced horizontal loop-handles are decorated with

carapace symbols. 2/a
Figure e. — Vase in which the much reduced vertical handles are ornamented with

carapace motives. 2/6
Figure f. — Vase in which the vertical handles unite shoulder and rim and are

entirely covered with carapace symbols. 2/6

MEM. CONN. ACAD., VOL. Ill

PLATE XVIII

a

Plate XIX

Vases belonging to the unpainted handled ware. Page 75.

Figure a. — Vase with horizontal loop-handles so treated as to indicate an animal
form. 2/B

Figure b. — Vase with animal motives ornamenting handles and shoulder; from
Divala. 2/6

Figure c. — Vase decorated with plastic animal forms, connecting shoulder and rim
and so arranged as to suggest double handles ; from Gualaca. 2/5

Figure d. — Vase with plastic human figure decorating the handles ; the flattened
nodes resembling rivet heads that encircle the body of the vessel prob-
ably represent the body-markings of the alligator. */2

Figure e. — Flat-bottomed vase in which each horizontal handle is composed of
two united figures of either man or monkey. Va

Figure /. — Flat-bottomed vase in which both horizontal and vertical type of handle
are combined in one. 2/B

MEM. CONN. ACAD., VOL. Ill

PLATE XIX

a

Plate XX

Series of vases showing two prevailing types of rim and handles, the latter de-
corated with animal motives. Painted handled ware. Page 79.

Figure a. — Vase with vertical paired handles connecting shoulder and lip, the
three nodes on each handle representing an animal form. 2/6

Figure b. — Vase with graceful vertical handles, on each of which is a pair of nodes
connected by a fillet. 2/6

Figure c. — Vase with two nodes on each handle representing the ears or eyes of
some animal form. 2/s

Figure d. — Vase with but a single node on each handle. 2/6

Figure e. — Vase with horizontal handles that do not affect the shape of the rim ;
from Bugavita. 2/6

Figure /. — Vase with horizontal handles, the neck being adorned with incised

patterns. 2/6
Figure g. — Vase with incised shoulder, and handles that represent animal forms;

from El Banco. 2/6
Figure h. — Vase in which the two plain nodes on each handle represent the ears

or eyes of some life form. 2/6

MEM. CONN. ACAD., VOL. Ill

PLATE XX

d

Plate XXI

Series of vases in which the vertical handles are converted into human or apelike
forms by the addition of plastic features. Painted handled ware. Page 79.

Figure a. — Small vase in which the handles resemble the human figure with pro-
boscis. Va

Figure b. — Vase in which each handle is a twin figure ; from Bugavita. '/a
Figure c. — Vase in which each handle is represented by two figures ; from Divala. Va
Figure d. — Vase in which each handle represents a grotesque human figure. ','2
Figure e. — Small low vase with flaring rim with handles representing the human

form. Va

Figure /. — Small vase with sloping rim and grotesque figures serving as handles. Va
Figure g. — Vase with incised shoulder ornamentation, two grotesque figures taking

the place of handles. Va

Figure A. — Vase in which the human figures serving as handles are deftly modeled
(see text-fig. 126); from Bugavita. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXI

d

Plate XXII

Series of tripod vases in which the red paint is smeared by being rubbed before

it was dry. Fish ware. Page 85.

Figure a. — Vase in which the supports are ornamented with the eyes, mouth and
tail fin of the fish ; from Bugavita. Va

Figure b. — Vase with spreading supports bearing pectoral fins only. Va

Figure c. — Vase in which the supports are made to resemble a bird, probably the
hawk or parrot. Va

Figure d, — Vase with plain supports except for the projection at the hip to suggest
the nose of the fish. Va

Figure e. — Large vase with plain legs except for a flattening at the foot to in-
dicate the caudal fin. Va

Figure /. — Vase with diagonally incised handles and plain supports. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXI!

a

d

Plate XXIII

Series of large tripod vases. Fish ware. Pages 85, 86.

Figure a. — Graceful urn-shaped vase, the neck being adorned with a series of
scrolls and an incised fillet, and the tripod supports showing realistic
representations of the fish, with all except the anal fins present. ljs

Figure b. — Large vase, the dorsal fins on the supports being placed forward on
the tip of the nose and between the eyes. ljs

Figure c.— Vase with highly curved fishlike supports on which only pectoral, ven-
tral and caudal fins are present. 1J3

Figure d. — Unique vase with long cone-shaped neck and plain angular handles,
only the pectoral and caudal fins remaining on the fish-shaped sup-
ports. Vs

Figure e. — Vase, the supports of which are plain except for a hip decoration re-
presenting the mouth, eyes and dorsal fin of the fish, '/a

Figure /. — Large vase, one of whose supports represents the fish, while the other
two are highly realistic alligator forms. lla

MEM. CONN. ACAD., VOL. Ill

PLATE XXIII

a

f

Plate XXIV

Series of tripod vases in which the fish attributes are overshadowed by the ad-
dition of other animal forms. Fish ware. Pages 88, 89.

Figure a. — Large vase without handles, the fishlike supports showing grotesque

apelike figures on the nose of each fish ; from El Banco. 1U
Figure b.— Large vase with twisted handles, the hip ornaments on the legs being

different in each case : from El Banco. Vs

Figure c. — Vase with a long-tailed animal reposing on the hip of each support. Vs
Figure d. — Vase in which the hip ornament resembles the owl. Vs
Figure e. — Vase, the supports of which bear a realistic representation of the owl

as hip ornaments ; from Divala. 1J3
Figure /. — Vase in which the frog appears as hip ornament on the supports, '/a

MEM. CONN. ACAD., VOL. Ill

PLATE XXIV

Plate XXV

Series of vases exhibiting a variety of forms. Red line ware. Pages 92, 93.

Figure a.— Large globular vase with gently curving paired handles. 1I3

Figure b. — Small vase with single handle, opposite which are two plain nodes of

clay. 1I»
Figure c. — Vase with single handle, the relief ornaments on the neck representing

the human features. 2/B
Figure d. — Vase without handles, the margin of the rim being marked by radiating

incisions ; from Divala. '/2
Figure e. — Tripod with characteristic loop foot, paired indeterminate life forms

decorating the body of the vase ; from Caldera. 2/r,
Figure /. — Unique vase with projecting collar and annular foot : may have served

as a drum. 1J3

MEM. CONN. ACAD., VOL. Ill

PLATE XXV

Plate XXVI

Series of vessels of varied form and decoration. Scarified ware. Pages 99, 100.

Figure a. — Fine large jar with flat bottom and flaring rim, the scarified zone reach-
ing from near the base to the neck. 1J3

Figure b. — Two-storied vase ; the upper story plain, the lower story entirely cov-
ered with scarifications. :/a

Figure c. — Vase with four legs, the applied head and tail completing the zoomorphic
unit. The scarifications represent the armadillo carapace. l/s

Figure d. — Graceful tripod with long solid legs probably representing the arma-
dillo, as indicated by the prominence at the hip. 2/6

Figure e. — Bell-shaped, flat-bottomed tripod, the legs of which are missing. The
scarified surfaces give the effect of having been produced by some
textile fabric ; from Caldera. 2/6

Figure /. — Tripod bowl with scarifications suggesting the warp and weft of bas-
ketry. a/3

Figure (/.—Tripod scarified in vertical bands, one being left unfinished. Vs

Figure h. — Vase with flaring rim, the bowl being practically covered with scari-
fications. */3

MEM. CONN. ACAD., VOL. Ill

PLATE XXVI

Plate XXVII

Vases representing two distinct types of lost color ware. Pages 106, 110.

Figure a. — Globular vase with linear decorations covering body and neck, the
entire original ground being red in color. s/6

Figure b. — Vase, the original ground of which consists of red and cream zones,
the latter being decorated with a series of monkey forms; from Buga-
vita. :'/6

MEM. CONN. ACAD., VOL. Ill

PLATE XXVI I

Plate XXVIII

Series of vases with original ground of red and with linear decorations.
Lost color ware. Pages 107, 108.

Figure a, — Vase with quadrangular panels in upper zone. Va

Figure b. — Vase with ornamented arched panels in upper zone. Va

Figure c. — Vase in which the upper zonal decorations, consist of groups of lines

radiating from the neck, the lower zone being banded horizontally. Va
Figure d. — Vase in which the upper zone is decorated longitudinally and the lower

zone latitudinally ; from El Banco. Va
Figure e. — Vase, the upper zone of which is decorated with groups of parallel

bands that meet and disappear; from Vivala (not Divala). Va
Figure /. — Vase with flaring rectangular lip, the upper zone being traversed by

a meandering group of parallel lines ; from Vivala. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXVIII

Plate XXIX

Series of vases, some with handles or ornaments in relief, the original ground
color being red. Lost color ware. Pages 108, 109.

Figure a. — Vase with cross-banded handles and with linear and panel decorations. 2/B
Figure b. — Vase with cross-banded handles, the decorations near the base of each

handle consisting of alligator motives. 2/B
Figure c. — Vase, the unpolished neck of which is ornamented with eye symbols

in relief, the painted decoration of the body being longitudinal in the

upper zone and latitudinal in the lower zone ; from Vivala (notDivala). 2/6
Figure d. — Vase, the shoulder of which is decorated with six diamond-shaped

panels : from Divala. 2/6
Figure e. — Wide-mouthed vase with recurved lip, the shoulder being decorated

with dorsal-view motives derived from the alligator; from Divala. 2/5
Figure /. — Vase, the collar of which is modified to represent an animal form. 2/5

MEM. CONN. ACAD., VOL. Ill

PLATE XXIX

b

a

Plate XXX

Series of vases in which the ground colors are light above and red below, either

alligator motives or serpent motives appearing in the designs.

Lost color ware. Pages 111, 112.

Figure a. — Vase whose circular rim bears vestiges of a square lip ; the neck is

ornamented with painted eye ornaments. Va
Figure b. — Vase with two animal heads in "relief on the equatorial zone, a single

tail appearing on the opposite side. Va
Figure c. — Vase whose shoulder is decorated with the head and tail of the

racoon, the panel patterns of the intervening spaces being alligator

motives, Va

Figure d. — Vase decorated with design resembling serpent motives. Va
Figure e. — Vase in which the panel designs resemble the serpent motive. Va
Figure /. — Vase with single vertical handle, the shoulder decoration being a spool

shaped motive. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXX

a

d

Plate XXXI

Series of vases, the decorations on which are derived from the dorsal view of
the alligator. Lost color ware. Pages 117, 118.

Figure a. — Vase with two types of dorsal-view motive derived from the alligator,
one on the neck and one on the shoulder; from Bugavita. 2/6

Figure b. — Vase in which the arched panels and the field surrounding the neck
are covered by series of dorsal-view motives, the lateral serrations being
plain spines. 2/6

Figure c. — Flat-bottomed vase like a tea-caddy in shape ; the body is de-
corated with three overlapping panels : the three panels on the shoulder
are separated by dorsal-view motives. Va

Figure d. — Vase in which the shoulder zone is decorated with the dorsal-view
motive repeated seven times ; from Bugavita. 2/B

Figure e. — Vase in which the two small arched panels each contain two dorsal-
view motives placed obliquely. Va

Figure /. — Vase with rectangular lip, the panel decorations on the shoulder re-
sembling the dorsal-view motive placed horizontally. lla

MEM. CONN. ACAD., VOL. Ill

PLATE XXXI

a

Plate XXXII

Series of tripods showing variations in form and ornamentation.
Lost color ware. Pages 122 — 124.

Figure a. — Tripod vase with the body altered in the equatorial region to resemble
the crab ; the shoulder is divided into four panels by spool-shaped
ornaments, each panel showing a series of monkeys ; from Vivala (not
Divala). 2/5

Figure b. — Shallow tripod bowl representing the turkey-buzzard, the painted dec-
orations consisting of groups of parallel lines meeting at various
angles ; from Divala. %

Figure c. — Open shallow tripod bowl the outer surface of which, except the legs,
is elaborately decorated ; from Divala. 2/6

Figure d. — Tripod vase showing a design of six groups of vertical bands alter-
nating with spool-shaped ornaments, accompanied by eyes, '/a

Figure e. — Tripod bowl showing painted interior decorations ; from Divala. Va

Figure f. — Tripod vase whose entire outer surface is completely covered with
decorations, the sides being divided into panels by groups of vertical
bands ; from Divala. 2/3

Figure g. — Rare type of tripod vase with sloping flattened upper zone and steeply
sloping collar. 2/B

Figure h. — Calabash type of tripod bowl ; spool-shaped ornaments accompanied by
eyes convert the supports into an animal form, the cross-banding sug-
gesting the armadillo carapace ; from Divala. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXXII

d

Plate XXXIII

A typical series of vases. Alligator ware. Pages 128 — 130.

Figure a. — Vase in which the two large panels are each decorated with a realistic
representation of the alligator, combining all its characteristic features. 3/s

Figures b and c. — Characteristic vases, the shoulder of each being ornamented
with two alligators ; c is from Vivala. 2/5

Figure d. — Double-necked vase with representation of the alligator in which the
head is turned, the long jaws extending over the back and tail ; from
Bugavita. '/a

Figure e. — Vase illustrating conventional treatment of the alligator; from Buga-
vita. */,

Figure /. — Vase whose shoulder decorations represent the alligator with much
reduced body and extremities. 3/5

MEM. CONN. ACAD., VOL. Ill

PLATE XXXIII

a

Plate XXXIV

Series of vases showing progressive steps in the process of conventionalizing the
alligator form. Alligator ware. Page 132.

Figure a. — Vase in which the decoration represents the alligator with two heads
but only one nuchal appendage - - an example of both reduplication
and elimination. 3/ft

Figure b. — Vase ornamented with the double-headed type of alligator motive. 3/.-,
Figure c. — Vase, the shoulder zone of which is adorned with alligator motives. 3j-n
Figure d. — Vase in which the lower zone is red and the upper zone shows a series

of alligator motives ; from Vivala (not Divala). 1/2

Figure e. — Vase in which the panel decoration on the shoulder consists of alli-
gator motives. */j

Figure f. — Vase in which simplification and general reduction of the alligator
form reach their limit. 1/»

MEM. CONN. ACAD., VOL. Ill

PLATE XXXIV

Plate XXXV

Series of vases in which the meaning of the zonal shoulder decoration is some-
what obscure. Alligator ware. Page 132, 133.

Figure a. — Vase in which the alligator motive is recognizable in the zonal shoulder

decoration, the design being repeated three times. 3/5
Figure b. — Vase in which the zonal decoration is the multiple body-line of the

alligator with body-markings in the hollow of each curve. 3/r>
Figure c. — Vase of which the elaborate motive on the shoulder is referable to

the alligator (see text-fig. 226). V2
Figure d. — Vase in which the shoulder zone is divided into three panels, each

ornamented with the dorsal-view motive. 1/2
Figure e. — Vase in which the shoulder decoration is probably a variant of the

dorsal-view motive. l/.,
Figure /. — Vase in which the shoulder ornamentation consists of paired concentric

rings. Va

MEM. CONN. ACAD., VOL. Ill

PLATE XXXV

a

Plate XXXVI

Series of vases with three arched panels, illustrating freedom in the treatment of
alligator motives. Alligator ware. Page 134.

Figure a. — Characteristic vase in which the decoration of the arched panels con-
sists of the dorsal-view motive. 1/2

Figure b. — Vase whose panel design is a variation of the dorsal-view motive. */2

Figure c. — Vase in which the association of spots and zigzag lines suggests
a multiple body-line with dermal markings; from Divala. 1J2

Figure d. — Vase, the panel design of which is a typical dorsal-view motive ; from
Divala. V2

Figure e. — Vase in which the short vertical bands crossing each panel are accom-

1;

panied by scale-group symbols ; from Divala. '/•>
Figure /. — Vase whose panel design is a variant of the dorsal-view motive ; from
Divala. *»

MEM. CONN. ACAD., VOL. Ill

PLATE XXXVI

Plate XXXVII

Series of vases in which the shoulder decoration consisting of four rosette patterns
is the dominant feature. Alligator ware. Page 136.

Figure a. — Vase with alligator motives ornamenting lip and shoulder. 2/s
Figure b.— Small vase with alligator motives on the lip, also on the raised nodes
of the shoulder ; from Bugavita. 2/.<

Figure c.— Vase in which the design is confined to the rosettes on the raised

nodes, the central feature of each being three scale-group symbols. 1/0
Figure d. — Vase in which the scale-group motives are disengaged from the inner

circle of the rosette ; the interspace is marked by a cross ; from Divala. */-
Figure e. — Vase in which the center of the rosette is a cross accompanied by

four scale-group motives ; from Divala. '/a
Figure f. — Vase in which the accentuated nodes are applied to the outer surface,

the painted decoration being a variant of that in figure e. ljs

MEM. CONN. ACAD., VOL. Ill

PLATE XXXVII

a

d

Plate XXXVIII

Representative series of vases with plastic features applied to the shoulder.

Alligator ware. Pages 136, 137.

Figure a. — Vase ornamented with a human head and short upturned tail alternating
with painted representations of the alligator ; from Vivala (not Divala). llt

Figure b. — Vase with unique form of painted alligator motive. llz

Figure c. — Vase with two opposed plastic heads on the shoulder, the painted
alligator form showing stylistic tendencies. 1/2

Figure d. — Vase illustrating a diagrammatic treatment of the double-headed alli-
gator between the plastic shoulder ornaments. ljt

Figure e. — Small vase in which the painted ornament represents the double-headed
alligator with both nuchal crests applied to the same neck. '/2

Figure /. — Vase in which the relief ornaments are reduced to mere nodes, a panel
containing a multiple body-line accompanied by dermal markings being
placed on each side. J/«

MEM. CONN. ACAD., VOL. Ill

PLATE XXXVIII

d

Plate XXXIX

Series of rare tripod vases. Alligator ware. Pages 138, 139.

Figure a. — Tripod of which each panel decoration is formed by a series of scale-
group motives ; from El Banco. 1/2
Figure b. — Tripod in which the two double panels are decorated with scale and

spine symbols ; from Jacu. 1/.>
Figure c. — Wide-mouthed tripod with spine symbols arranged in groups of two ;

from Jacu. '/2
Figure d. — Shallow tripod bowl with horizontal bands decorating the rim. from

the lowest of which hangs a series of spine symbols in black ; from

Divala. >/2
Figure e. — Tripod bowl with gently incurved rim ; the central zone is divided

into three panels, each containing sigmoid patterns ; from Vivala (not

Divala). ^
Figure /. — Tripod vase in which the shoulder decoration consists of three arched

panels ; the designs in these are probably variants of the dorsal-view

motive. 1/0

MEM. CONN. ACAD., VOL. Ill

PLATE XXXIX

a

Plate XL

Series of vases combining plastic and painted decoration.
Alligator ware. Pages 139, 140.

Figure a. — Tripod vase with shoulder ornaments representing the animal head and
tail in the round, also with painted figures of the alligator on each
side. */2

Figure b. — Tripod in which both plastic and painted ornaments are probably
referable to the alligator. 1/.,

Figure c. — Small tripod illustrating the combination of relief and painted dec-
oration. '/«

Figure d. — Vase with plastic shoulder ornaments, the panel designs being the
scale-group motive. 1/2

Figure e. — Small tripod in which the painted ornament supplements the plastic. 1/1

Figure /. — Compound tripod with a series of alligator motives about the two
rims. V

MEM. CONN. ACAD., VOL. Ill

PLATE XL

a

Plate XLI

Series of vases representing a variety of unusual forms.
Alligator ware. Pages 140, 141.

Figure a. — Tripod vase resembling the armadillo. 1/a

Figure b. — Tripod vase i-eseinbling the crab, and with animal heads for supports. l/::
Figure c. — Tripod representing the jaguar-god, the arms of which serve as handles
to the vessel ; on the panel at the back is a painted figure of the alli-
gator; from San Carlos. 1/.,

Figure d. — Tripod probably representing the jaguar-god. 1/2

Figure e. — Unique thick-walled vase with annular base, and a series of scale-
group symbols encircling the bowl ; from Jacu. 1J2

Figure f. — Unique form of vase supported by three half-human forms standing
on a ring-shaped base ; the panel designs consist of the alligator
motive. 2/g

MEM. CONN. ACAD., VOL. Ill

PLATE XLI

a

Plate XLII

Lost color ware and alligator ware. Pages 112, 141, 142.

Figure a. — Small vase beautifully modeled and painted, illustrating the use of
white bands crossing the original ground of red ; from Divala. Lost
color ware. 3/5

Figure b. — Vase in which the peripheral band and those tangent to the neck are
red on a light ground ; from Divala. Lost color ware. :l/5

Figure c. — Rare form of vase illustrating the use of engaged plastic life forms as
supports, the figures being apparently those of the human female ; re-
presentations of the alligator adorn the panels of the neck and sides
(see text- fig. 242). Alligator ware. 2'

ii

MEM. CONN. ACAD., VOL. Ill

PLATEXLII

Plate XLIII

Distinct animal forms with vesicular bodies. Alligator ware. Pages 143 — 145.

Figure a. — Fusion of vase and animal form representing the jaguar ; jaguar-like
markings appear about the eyes and on the throat, while the neck,
sides and tail are decorated with a series of 89 panels, all enclosing
alligator motives. x/8

Figure b. — Dorsal view of the foregoing.

Figure c. — Vase representing the tapir, the painted designs being traceable to the
alligator. 1/2

MEM. CONN. ACAD., VOL. Ill

PLATE XLIII

Plate XLIV

Polychrome ware. Pages 152 — 154.

Figure a. — Pitcher-shaped vase with horizontally flattened lip ; the handle is broken
off; the notches cut from the scrolls on the neck and panels represent
engaged alligator motives (see text-rigs. 253 and 254) ; from Gualaca. 2/3

Figure b. — Vase of eccentric form, the chief ornamental feature being the elabo-
rate branching scroll filling each of the two shoulder panels : here the
egaged alligator motives are more easily distinguishable than in figure a. ll?

MEM. CONN. ACAD., VOL. Ill

PLATE XLIV

Plate XLV

Superb example of ancient Chiriquian decorative art.
Polychrome ware. Pages 157, 158.

Figure a. — Unique vase consisting of a shallow bowl mounted on a hollow per-
forated stand, ever}' visible part being ornamented in black, red and
purple on a cream ground. The design on the outer surface of the
bowl resolves itself into two units of the classic fret, while that on the
interior (see frontispiece) is a representation of the alligator-god accom-
panied by alligator symbols. 1/2

Figure b. — Detail (from fig. a) of the classic fret composed of a common body-
line, each end of which terminates in an alligator head. */2

MEM. CONN. ACAD., VOL. Ill

PLATE XLV

a

Plate XLVI

Series of clay stools. Armadillo ware. Page 162.

Figure a. — Elementary type of stool resembling one class of metate as well as
the recent Chiriquian wooden stools. */«

Figure b. — Stool illustrating the transition from the square type to the prevailing
round type. 1/g

Figure c. — Stool with gently convex top. '/»

Figure d. — Stool with seat that rests on a hollow stand, whose walls are con-
tinuous except for narrow vertical slits. The flattened or beaten char-
acter of the supporting figures resembles work in metal. '/»

Figure e. — Low stool with convex seat resting on a hollow perforated stand, to
which are applied figures of men and monkeys. 1/s

Figure /. — Stool, the seat of which is supported by figures of monkeys standing
on a ring-shaped base. The same type is seen among stone stools. 1/8

MEM. CONN. ACAD., VOL III

PLATE XLVI

Plate XLVII

Series of whistles in the form of birds. Alligator ware. Page 174.

Figure a.— Parrot-shaped whistle with vestigial head on the breast and a real head
that has the appearance of being removable. •/„

Figure b.— Bird-shaped whistle with vestige of a head on the pointed breast and
head on the shoulders that has the air of being detachable. 2/.,

Figure c.— Bird-shaped whistle with head turned so as to heighten the resem-
blance to a former adjustable state. The painted spot on the breast
represents a vestigial head. 2/3

Figure d. — Bird-shaped whistle. 2/8

Figure e. — Bird-shaped whistle with primitive vestigial head on the breast ; from
Tacu. 2/,

•J to

Figure /. — Bird-shaped whistle. 2/8

MEM. CONN. ACAD., VOL. Ill

PLATE XLVII

Plate XLVIII

Series of gold images. Pages 203-217.

Figure a. — Image of fine gold in the shape of a frog with flattened feet and with

serpent motive ornamenting the head. V,
Figure b. — Image of low-grade gold representing a frog with eyes resembling

sleigh-bells ; the metal ball in each is practically pure copper. Vi
Figure c. — Image of fine gold representing a double bird, '/i
Figure d. — Realistic jaguar figurine of fine gold. */i
Figure e. — Human figurine of high-grade gold, holding a rattle in the right hand

and something resembling a flute in the left (see PI. XLIX, figs, b and c) ;

the apparel is represented by head ornament, loin-girdle, apron and

knee-bands, '/i
Figure /. — Human image of low-grade gold represented as in the act of pulling

two snakes from its wide-open mouth. The head-dress consists of two

serpent heads. Vi
Figure g. — Image of fine gold representing the alligator-god, the flattened bars at

the head and feet each being a common body for the two alligator

heads attached. Vi

Figure h. — Figurine of low-grade gold representing the crab-god holding in his

mouth a human foot and leg severed at the knee. J/i
Figure i. — Animal figurine of fine gold. 1jl

MEM. CONN. ACAD., VOL. HI

PLATE XLVI1I

Plate XLIX

Series of gold images approximating the human form. Pages 209 212.

Figure a. — Group of two figurines evincing skill and taste in the art of casting. '/,
Figure b. — Human figurine with rattle in the right hand and a reed-shaped object

held to the mouth by the left hand. */,

Figure c. — Human image probably representing a musician or god of music: c'.—
Side view of same, showing effects of having seen much service, '/i
Figure d. — Human image with arms terminating in rattles. Lamson collection. '/,
Figure e. — Human image holding a rattle in the left hand. Heye collection. '/,
Figure /. — Human figurine with head-dress resembling the " skil " of the Tlingit

Indians. Heye collection. Vi

Figure g. — Group of two human figurines suggesting the art of Colombia. Metro-
politan Museum. 2/:j

Figure h. — Human image set in a frame ; h'. — Side view of same, illustrating skill
in hammering separate castings, '/i

MEM. CONN. ACAD., VOL. Ill

PLATE XLIX

a

s

7 DAY USE

RETURN TO DESK FROM WHICH BORROWED

ANTHROPOLOGY LIBRARY

This publication is due on the LAST DATE
and HOUR stamped below.

General Library

University of Californi

Berkeley

RB 17-BOm-3, 63
(D58088lO)4188

U.C. BERKELEY LIBR; ES

T