
SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY
VOLUME 6







Rocafuerte Painted (red and black-on-white) square basin of Form 14, painted in Technique 2. The unusual double snake motif is executed
in low relief on the rim and in painting on the interior. This vessel was assembled from sherds in the sample from N-P-2, Cut 1 and
Broadside A (U.S. National Museum).



Archeological Investigations
on the Rio Napo,
Eastern Ecuador
Clifford Evans and Betty J. Meggers
7
SMITHSONIAN INSTITUTION PRESS
Washington 1968


         A Publication of the
SMITHSONIAN INSTITUTION
United States National Museum
               LIBRARY OF CONGRESS 67-60067
UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON, 1968
For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. 10401 - Price $4.50

Preface
    The eastern slopes of the Andes attracted our attention in 1950, when it became
probable that the Marajoara Phase on the island of Marajo at the mouth of the Amazon
was derived from northwestern South America. Our first opportunity to investigate the
possibilities for archeological fieldwork came when we were in Guayaquil, Ecuador, in
1954 and met Coronel Jorge V. Gortoire, who had served for a period as commandant
of the Ecuadorian Army Post at Tiputini. Conversation with him reawakened our latent
interest in the area, and we began to make specific plans. In October 1956, having been
awarded Grant No. 2012 from the Penrose Fund by the American Philosophical Society
and granted official detail by the Smithsonian Institution, we returned to Ecuador to
undertake the fieldwork.
    Through the courtesy of Coronel Rafael Andrade Ochoa, at that time Commander-
general of the Fuerza Aerea Ecuatoriana, we received authorization to fly from Quito
to Tiputini in an Ecuadorian Air Force DC-3 transport plane. However, almost daily
rains maintained the airstrip in unsuitable condition for landing and after several weeks
of waiting in Quito for the weather to break, we gave up and arranged to fly by commercial
airline in a Junkers Tri-Motor to Shell-Mera and then in a single engine Norseman to
Tena. A day on horseback brought us to Latas, where we secured a dugout canoe manned
by Quechua-speaking Indians to take us downriver. Although the trip was longer and
more difficult than it would have been by air, it gave us invaluable first-hand experience
with conditions along the Rio Napo (pis. 1-5). We were able to follow our hourly progress
on U.S. Air Force Preliminary Base Map 950A (Scale 1:500,000), which perfectly repro-
duced every bend and island. By the afternoon of the fifth day, when we arrived at Tiputini,
we were well prepared to appreciate the comments of Orellana's men, who preceded us
by 415 years.
    When we stepped on shore at Tiputini, the military post that was to be our base of
operations, we were delighted to discover not only that there was an archeological site
on the spot, but that the pottery included incised and excised techniques of decoration
diagnostic of the Marajoara Phase, although only painted vessels had been previously
reported from the Rio Napo. With the cooperation of army personnel and local residents,
we were able to investigate a number of sites particularly along the portion of the river
between Tiputini and the mouth of the Rio Yasuni, which marks the boundary between
Ecuador and Peru. We also checked the lower Rio Tiputini. During our stay, the river
was unusually low, and extensive sand bars reduced the channel in places to a slender
meandering stream (pi. 4b). Giant trees temporarily resting on beaches (pi. 3b) attested
to the force of the current at other times of the year, lending credence to descriptions by
Orellana's companions (see pp. 106-107), who had the misfortune to encounter higher
water than we did.
    
PREFACE
    At the conclusion of the survey, we had accumulated several tons of specimens and
were sufficiently familiar with the river to look forward to returning to Quito by air. As
was the case in October, intermittent rain kept the airstrip soft, but we were prepared to
wait as long as necessary this time, since going by river would have taken at least two
weeks. An Ecuadorian Air Force DC-3 finally came on December 15, and two hours after
takeoff we were in Quito—by every standard of comparison, another world.
    We left behind us in the Province of Napo-Pastaza many friends never to be seen
again, and memories still fresh as we write this ten years later. Sr. Jose Bernardo Crespo
Pando made us his guests while we worked at Nueva Armenia, and allowed us to use his
home as a base from which to visit nearby sites. Philosopher, businessman, and astute
observer of the world from afar, he was an invaluable promoter of our cause as^ well as
an entertaining host. Several pleasant days were also spent at the home of Sr. Jose Rafael
Urvina on the Rio Tiputini, where we received all possible cooperation and courtesy.
Other land owners who granted us permission to work on their property and to whom
we offer our thanks are Sr. Juan Francisco Buitron (Hacienda San Juan, Cotacocha),
Sr. Osvaldo Bijarini Aridi (Florencia), and Sr. Alfonso Antonio Cox Vega (Bello Hori-
zonte). Sr. Pedro Jamn, at that time Jefe Politico of the Junta del Canton Aguarico,
kindly allowed us to dig a few holes in the main street of Nuevo Rocafuerte.
    Since we were unable to arrange for a trip to the Rio Aguarico, we are particularly
indebted to Rene Alberto Hinoyosa Carrera, then a second lieutenant stationed at Tipu-
tini, who collected sherds for us from Cabo Minacho on the Rio Giiepi and Panacocha
on the Rio Cuyabeno (fig. 3). Other young officers at Tiputini, who provided us not
only with various kinds of assistance but also with pleasant companionship, include Solo-
man Hernandez V., Augustin Carvalho V., Raul Costales, and Fausto Bustamonte. We
are indebted to the commandant at that time, then Major J. Gonzalo Ramos Sevilla,
for permitting us to use Tiputini as our base, and providing us with quarters and other
kinds of help.
    During our negotiations to enter the Oriente by air, we were aided in numerous ways
by Jorge V. Gortaire V., then a colonel and director of the Colegio Militar "Eloy Alfaro"
in Quito. Other kinds of help and guidance were provided by Enrique Martinez Q.,
manager of the Compania General de Comercio y Mandato in Quito, and his assistant
Francisco Punina Y. To these and other military and governmental officials whose names
escape us after a decade, we wish to express our appreciation for the many favors, large
and small, that we have not forgotten, and which helped to make our visit memorable
as well as scientifically fruitful.
    Finally, we wish to record our indebtedness to the late Emilio Estrada, who while
teasing us for persisting in our "whim" to go to the Rio Napo, exercised his considerable
influence to help us secure the necessary permissions from military authorities. Although
his interventions were often unobtrusive, it is probable that they were instrumental in
making it possible for us to carry out the work described in the present report.
    Other obligations have been incurred during efforts to work out the affiliations of
Rio Napo archeological complexes. Our ability to trace the downriver movement of the
Polychrome Horizon Style (fig. 68) stems from permission granted by the Ethnographical
Museum in Goteborg, Sweden, to take detailed notes and photographs during the summer
of 1960 of sherd samples collected in the 1920's by Curt Nimuendajii. This museum work
was supported financially by Grant No. 2664 from the Penrose Fund of the American
Philosophical Society. Peter Paul Hilbert, who is responsible for what little stratigraphic
information is available from the middle and upper Amazon, has again generously made
available unpublished data. Donald Lathrap, whose chronological sequence in eastern
Peru is one of the rare reliable reference points, has kindly allowed us to consult his un-
published doctoral dissertation at Harvard University, which supplements data secured
by examination of the collections from his 1956 fieldwork. We wish also to record our
appreciation to the American Museum of Natural History, the Museum of the American
Indian (Heye Foundation), the Brooklyn Museum, the Museum of Primitive Art, the
    
PREFACE	vii
Peabody Museum of Archaeology and Ethnology (Harvard University), the Musee de
l'Homme (Paris), the Museu Paraense Emilio Goeldi (Belem), the Museu Paulista (Sao
Paulo), the Instituto Geografico e Historico do Amazonas (Manaus), the Museo Victor
Emilio Estrada (Guayaquil), the Museo Arqueologico del Banco Central del Ecuador
(Quito), and the Casa de la Cultura Ecuatoriana (Quito), all of which have granted us
permission to examine and photograph specimens or have provided us with photographs
for publication. Several Napo Phase anthropomorphic urns have passed into the hands of
private collectors, who have allowed us to include them in our illustrations. To Thomas P.
Flannery, Alan C. Lapiner, Jay C. Leff and Howard S. Strouth, we take this opportunity
to offer public thanks.
    Carbon-14 dates have been furnished by the Smithsonian Institution and the Univer-
sity of Pennsylvania. We are indebted to the Creole Foundation for a grant to assist in
obtaining the first series of dates from the latter laboratory.
    In conclusion, it is a pleasure to record once again our indebtedness to members of the
Smithsonian Office of Anthropology Processing Lab staff, especially Willie Mae Pelham
and Robert C. Jenkins, for their careful preparation of the sherd collections for study.
George Robert Lewis, scientific illustrator in the Smithsonian Office of Anthropology, has
produced his customarily excellent drawings from badly eroded pottery, poor illustrations
in published sources or photographs, as well as the maps and diagrams. We apologize for
delaying so many years to provide them with captions. The plates owe their clarity to the
high quality enlargements furnished by the staff of the Smithsonian Institution Museum
of Natural History Photo Lab. The manuscript was efficiently typed by Anne M. Lewis,
Smithsonian Office of Anthropology.
    For scholars interested in consulting the illustrated material, some clarification of the
symbols and legends may be useful. All specimens not otherwise credited are in the United
States National Museum, where a large type collection of sherds has been deposited.
Specimens in other collections are so identified, and addition of the word "courtesy"
indicates that the photographs were supplied by the individual or institution named.
A key has been employed in figures showing rim profiles, permitting rapid recognition of
association between form and presence or absence of red slip or decoration, explained in
each caption. The relative frequency of rims, shown in black, white, or hachure, approxi-
mates the relative popularity of the form with each type of surface treatment.
CE
BJM
Smithsonian Institution
Washington, D.C.
December 13, 1966
POSTSCRIPT
    After leaving our hands in mid-December of 1966, this manuscript pursued a devious
route through reviewers and editors, and was finally ready for press in late May of 1967.
By that time, however, all publication funds had been committed and typesetting had to
be delayed until the beginning of the next fiscal period. On July 12, 1967, a messenger
from the Government Printing Office (GPO) picked up the nine assorted packages con-
taining our edited text, figures, cl arts, tables, and plates; two packages containing an
edited manuscript and illustrations from another department; and eight skids loaded
with cartons of surplus publications for condemnation and destruction. On August 16,
    
viii	PREFACE
1967, our editor phoned the GPO to inquire the date that galley proof might be expected
(having been advised that typesetting would take about six weeks), and learned to her
dismay that the manuscript and accompanying packages in question had not been re-
ceived. A thorough inquiry on the part of persons in all agencies involved over a period
of several weeks finally reconstructed the incredible series of events that resulted in the
two manuscripts and their illustrative materials accompanying the condemnation material
through spot checks and transfers from one vehicle to another to a waste paper salvage
company and thence to the city dump, where they were presumably burned (not being
suitable for salvage as waste paper). A carbon of the unedited text, original draft copies
of maps, diagrams and tables, and a miscellaneous group of photographic enlargements
of individual negatives were the only things remaining in our office. The most heart-
breaking aspect of the loss was the destruction of the original drawings, which had con-
sumed more than a year of painstaking labor on the part of Scientific Illustrator George
Robert Lewis.
    In the reconstitution of illustrative material, we have had the cooperation of the
entire Smithsonian Institution on a priority basis, a concession that we deeply appreciate.
To shorten the time required to reproduce the figures, we also have had the assistance of
Scientific Illustrator Marcia Bakry, who has redrawn all the maps and rim profile dia-
grams. George Robert Lewis has recreated most of the line drawings of specimens; a few
have been eliminated because of insufficient time. We wish to record our gratitude to
both of these individuals for tackling the job conscientiously and rapidly, and producing
results that are as good as or better than the first attempts. The Museum of Natural
History Photographic Laboratory has provided quality service with close to "polaroid"
speed in reprinting to specific scale, contrast, and tone the numerous negatives from
which the final plates were again assembled and remounted. Joan Horn of the Smithsonian
Institution Press, who fortunately saved all her editorial notes, has accorded us priority
treatment and tackled the job with her usual enthusiasm, fervor, and accuracy the second
time around. Anders Richter, Director of the Smithsonian Institution Press, and the
staff of the Government Printing Office have collaborated to assure that publication
time was as brief as possible. To these and other individuals who have sympathized with
our predicament and resupplied lost illustrations, we extend our heartfelt thanks. To
our contemporary colleagues at the Smithsonian Institution who may be fearful for the
fate of their manuscripts, we offer the observation that the rate of frequency for this type
of occurrence appears to be only once every hundred years.
CE
BJM
February 29, 1968

Contents
Page
GEOGRAPHICAL DESCRIPTION		1
THE YASUNI PHASE		7
Description of sites and excavations		7
N-P-10: Puerto Miranda Hill		7
N-P-ll: Puerto Miranda Bank     		8
Data from other investigations		8
Analysis of materials		8
Stone artifacts		8
Abrader		8
Ax fragments		8
Hammerstone     		9
Ceramic classification     		9
Reconstructed vessel forms		9
Common forms		9
Rare forms		12
Base forms		12
Pottery type descriptions     		12
Miranda Modeled     		12
Miranda Plain		13
Yasuni Incised		14
Yasuni Incised and Punctate		15
Yasuni Nicked		15
Yasuni Plain		15
Yasuni Red     		16
Yasuni Zoned Hachure     		17
The seriated sequence and its implications		17
Diagnostic features of the Yasuni Phase     		17
THE TIVACUNDO PHASE		19
Description of sites and excavations		19
N-P-7: ChacraAlfaro		19
N-P-8: Barranco Alfaro     		19
Data from other investigations		19
Analysis of materials		20
Stone artifacts		20
Pottery artifacts		20
Abraders     		20
Potrests		20
Ceramic classification     		20
Reconstructed vessel forms		21
Common forms		21
Rare forms		21
Base forms		24
Pottery type descriptions     		24
Alfaro Plain		24
Chacra Plain		25
IX

X	CONTENTS
Page
Tivacundo Incised and Zoned Red		26
Tivacundo Plain		27
Tivacundo Red Painted		28
Unclassified Cariape-tempered Plain		28
Unclassified Sand-tempered Plain		29
Unclassified Decorated		29
The seriated sequence and its implications		30
Diagnostic features of the Tivacundo Phase		31
THE NAPO PHASE     		32
Description of sites and excavations		32
N-P-l: Tiputini		32
N-P-2: Nueva Armenia		33
N-P-3: Nuevo Rocafuerte     		34
N-P-4: Bello Horizonte		35
N-P-5: Florencia		35
N-P-6: Puerto Alfaro     		35
N-P-9: Cotacocha		36
N-P-l 3: Panacocha		36
Data from other investigations		36
Rio Aguarico		36
Eden		36
Rio Indillama		36
Oasis		37
Providencia		37
Rumi-tuni		37
Rio Yasuni		37
Unspecified location on Rio Napo     		37
Analysis of materials	       .   .	38
Stone artifacts		33
Abraders     		33
Axes		39
Chisel		40
Cores		40
Grinding stones		40
Hammerstones		40
Natural pebbles		40
Paint stone		40
Spalls		40
Pottery artifacts		42
41
Disk
Potrests		42
Roller stamps     		40
Ceramic classification     		40
Reconstructed vessel forms		43
Common forms		43
Rare forms		59
Base forms		52
Pottery types     		52
Armenia Plain		52
Armenia White-on-red		53
Napo Negative		53
Napo Plain     		r n
Napo Plain Excised		59
Napo Plain Incised		59
Napo Red		CQ
Napo Red Excised     		gj
Napo Red Excised, White Retouched	'A.   'AAA.	62
Napo Red Incised		52
Napo White Excised		53

CONTENTS	XI
Page
Napo White Incised		63
Rocafuerte Incised		63
Rocafuerte Painted		65
Tiputini Plain		72
Tiputini Plain Excised		72
Tiputini Plain Incised		74
Tiputini Red Excised		75
Tiputini Red Excised, White Retouched		76
Tiputini Red Incised		76
Tiputini White Excised     		76
Tiputini White Incised     		77
Unclassified decorated		77
Trade pottery		77
The seriated sequence and its implications		78
Diagnostic features of the Napo Phase		81
THE COTACOCHA PHASE		83
Description of sites and excavations		83
N-P-3: Nuevo Rocafuerte		83
N-P-9: Cotacocha		83
N-P-14: Latas		83
N-P-l 5: Tiputini Road		83
Analysis of materials		83
Ceramic classification     		84
Reconstructed vessel forms		84
Common forms		84
Minor rim forms		84
Base forms		84
Pottery type descriptions     		84
Cotacocha Plain		84
Cotacocha Red		85
Latas Plain     		86
Unclassified Cariape-tempered Plain		86
The seriated sequence and its implications		86
Diagnostic features of the Cotacocha Phase		87
THE RIO NAPO CULTURAL SEQUENCE AND ITS IMPLICATIONS		88
The Yasuni Phase		88
The Tivacundo Phase		93
The Napo Phase		93
Beruri, Rio Purus     		95
Boca do Xavier, Rio Urubu		95
Coari Region		95
Guarita Phase (Manaus)     		96
Ilha dos Muras     		96
Itacoatiara     		97
Rio Japura		97
Lago Araga Region		98
Rio Madeira     		98
Manacapuru     		99
Marajoara Phase (Maraj6)     		100
Miracanguera		100
Mocajatuba, Rio Urubu     		101
Nazare dos Patos      		101
RioPadauiri     		102
Paura      		102
Pogo do Jaburu		102
Ponta Negra, Rio Negro     		102
Pontao, Rio Urubu		102
Santarem Area		102
Tauaquera, Rio Aniba		103

XII	CONTENTS
Page
Tefe Region		}°?
Significance of the distributional pattern		J04
European discovery and the Cotacocha Phase		|JJ6
Hypothetical reconstruction of Amazonian prehistory      		j°J
LITERATURE CITED	
APPENDIX: Tables 1-11	
PLATES    	
Tables
TEXT
A.	Characteristics of Napo Phase anthropomorphic urns		53
B.	Frequency of vessel shapes in decorated variants of Rocafuerte Painted	
C.	Occurrence of Yasuni Phase decorative elements in Tutishcainyo Phase pottery   ...	90
D.	Frequency of temper varieties in a sherd collection from the Rio Giiepi		94
APPENDIX
1. Frequency of vessel forms of the Yasuni Phase		115
2. Frequency of pottery types at sites of the Yasuni Phase		116
3. Frequency of vessel forms of Cotacocha Phase pottery		116
4. Frequency of vessel forms of the Tivacundo Phase		116
5. Frequency of pottery types at sites of the Cotacocha Phase	■   •	116
6. Frequency of pottery types in surface collections and stratigraphic excavations at sites
of the Tivacundo Phase	       117
7. Frequency of stone artifacts at sites of the Napo Phase	•       117
8. Frequency of pottery types in surface collections and stratigraphic excavations at sites
of the Napo Phase	       118
9.	Frequency of Napo Phase plain and decorated pottery classified  by temper,  and
frequency of incised and excised decoration classified by single- and double-line
techniques	       121
10.	Frequency of vessel forms from surface collections and stratigraphic excavations of
sites used in the seriated sequence of the Napo Phase	       124
11.	Frequency of technique variants of Armenia White-on-red, Rocafuerte  Incised  and
Rocafuerte Painted	       126
Illustrations
FIGURES
1.	Map of the Province of Napo-Pastaza, showing principal landmarks and distances
covered during descent of the Rio Napo		2
2.	Principal routes of communication between the highland basins and the eastern low-
lands of Ecuador	*		5
3.	Rio Napo and its tributaries, showing the location of archeological sites and isolated
finds		6
4. Sketch map of the Yasuni Phase site of N-P-l 0		7
5. Yasuni Phase stone abrader		8
6. Yasuni Phase ax fragment		9
7. Rim profiles and reconstructed shapes of the Yasuni Phase, Common Forms 1-5    .   .	10
8. Rim profiles and reconstructed vessel shapes of the Yasuni Phase, Common Forms 6-9
and Base Forms A-C		11
9.	Rim profiles of Yasuni Phase Rare Forms 1-10		12

CONTENTS	Xni
Page
10. Type sherds of Miranda Modeled		13
11. Type sherds of Yasuni Incised		14
12. Type sherds of Yasuni Incised and Punctate		15
13. Type sherds of Yasuni Nicked		16
14. Type sherds of Yasuni Zoned Hachure		17
15. Sketch map showing the locations of sites N-P-6 of the Napo Phase and N-P17 and
N-P-8 of the Tivacundo Phase on the left bank of the Rio Tiputini		18
16. Tivacundo Phase polishing pebble		20
17. Base sherd used as an abrader, Tivacundo Phase		20
18. Rim profiles and reconstructed vessel shapes of the Tivacundo Phase, Common Forms
1-4     		22
19.	Rim profiles and reconstructed vessel shapes of the Tivacundo Phase, Common Forms
5 and 6		23
20. Rim profiles of Tivacundo Phase Rare Forms		24
21. Base forms of Tivacundo Phase plain pottery		25
22. Type sherds of Tivacundo Incised and Zoned Red		26
23. Rolled out design on exterior of Tivacundo Incised and Zoned Red oval bowl (cf. pi. 9).	27
24. Type sherds of Tivacundo Red Painted		28*
25. Unclassified decorated sherds from the Tivacundo Phase		29
26. Seriated sequence of Tivacundo Phase sites based on changes in pottery type frequency .	30
27. Changes in relative frequency of Tivacundo Phase Common Forms 1-6		30
28. Sketch map of the Napo Phase site of N-P-l      		32
29. Sketch map of the Napo Phase site of N-P-2     		33
30. Sketch map of the Napo Phase site of N-P-3      		35
31. Napo Phase stone axes		38
32. Napo Phase stone tools		39
33. Napo Phase hammerstones		41
34. Decorated potrest from the Napo Phase		42
35. Roller stamp probably of Napo Phase origin		43
36. Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 1-4.	44
37. Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 5-6.	45
38. Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 7-9.	46
39. Shoulder corner contour of carinated vessels of Napo Phase Common Forms 7-10,
showing decorative notching		47
40.	Rim profiles and reconstructed vessel shapes of the Napo Phase,  Common Forms
10-13		48
41.	Rim profiles and reconstructed vessel shapes of the Napo Phase,  Common Forms
14-16		50
42.	Rim profiles and reconstructed vessel shapes of the Napo Phase,  Common Forms
17-19		51
43. Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Form 20.   .	52
44. Rim profiles of Napo Phase Rare Vessel Forms 1-5		56
45. Base forms of Napo Phase pottery		57
46. Napo Negative bowl interior		58
47. Reconstructed design of Napo Red Excised bowl exterior		61
48. Type sherds of Rocafuerte Incised, combined with red and black-on-white painting.	64
49. Fragment of Rocafuerte Incised bowl		66
50. Type sherds of Rocafuerte Painted, red-on-white		67
51. Decorative techniques of Rocafuerte Painted, black-on-white		67
52. Decorative techniques of Rocafuerte Painted, red and black-on-white		68
53. Type sherd of Rocafuerte Painted, red and black-on-white, Technique 2		69
54. Rolled-out design of Rocafuerte Painted jar of Form 6 with anthropomorphic face on
rim		70
55.	Rolled-out design of Rocafuerte Painted, red and black-on-white, Technique 3 on
rim and black-on-white, Technique 3 on body		71
56. Rolled-out design of Rocafuerte Painted, red and black-on-white		72
57. Type sherds of Rocafuerte Painted, red and black-on-white		73
58. Reconstructed Rocafuerte Painted vessel with anthropomorphic treatment		74
59. Rim profiles and reconstructed vessel shape of sherds of probable trade origin from
Napo Phase sites		78

XIV	CONTENTS
Page
60.	Seriated sequence of Napo Phase sites derived from classification of both plain and
decorated sherds by temper		78
61.	Seriated sequence of Napo Phase sites derived from classification of incised and excised
sherds by technique of incision		79
62.	Correlation of seriated sequences of Napo Phase sites based on temper and on technique
of incision and excision		80
63. Seriated sequence of Napo Phase sites based on changes in pottery type frequency.   .	80
64. Changes in relative frequency of Napo Phase Common Forms 1-20		80
65. Rim profiles and reconstructed vessel shapes of the Cotacocha Phase     		85
66. Occurrence of Yasuni Phase ceramic traits of decoration and vessel shape in other
complexes representing the Zoned Hachure Horizon Style		89
67.	Map of northern South America, showing the location of sites representing the Zoned
Hachure Horizon Style		91
68.	Map of the Amazon region showing the location of sites representing the Polychrome
Horizon Style		94
69. Rim and body profiles of sherds from the Coari region		95
70. Rim and body profiles of sherds from the Guarita Phase		96
71. Excised vessel from Ilha dos Muras		96
72. Rim and flange profiles of decorated sherds from Itacoatiara, Guajara		97
73. Rim and body profiles of decorated sherds from the Lago Araga region		98
74. Rim and body profiles of sherds from Manacapurii		99
75. Anthropomorphic urns with painted decoration from Miracanguera		100
76. Anthropomorphic urn from Miracanguera		101
77. Zoomorphic vessel from Miracanguera		101
78. Rim and body profiles of sherds from Ponta Negra, Rio Negro		103
79. Occurrence of Napo Phase ceramic traits of decoration and vessel shape in Colombia
and Brazil	       105
80.	Location of regions with multiple phase and single phase occupations in the Amazon
basin and periphery, and some postulated routes of migration and diffusion   ....       109
PLATES
FRONTISPIECE: Rocafuerte Painted (red and black-on-white) square basin of Form 14, painted in
Technique 2.
1. Air views of the Rio Napo.
2. Views of the upper Rio Napo.
3. Views of the Rio Napo.
4. Decline in water level of the Rio Napo opposite N-P-2 during a six week period in November-
December 1956.
5. Views of the Rio Napo, showing varieties of modern water transportation.
6. Yasuni Phase sites.
7. Views of the Rio Tiputini.
8. Tivacundo Phase pottery types and artifacts.
9. Tivacundo Incised and Zoned Red.

10. Type sherds of Tivacundo Plain.
11. Tivacundo Phase pottery types.
12. Sites of the Napo Phase.
13. N-P-2: Nueva Armenia, a site of the Napo Phase.
14. Napo Phase site of N-P-2: Nueva Armenia.
15. Sites of the Napo Phase.
16. Views of the Rio Tiputini.
17. Potrest fragments from Napo Phase sites.
18. Type sherds of Armenia Plain.
19. Type sherds of Armenia White-on-red.
20. Type sherds and complete vessel of Armenia White-on-red.
21. Type sherds of Napo Plain.
22. Complete vessels of Napo Plain.
23. Type sherds of Napo Plain Excised.
24. Type sherds of Napo Plain Excised.
25. Type sherds of Napo Phase pottery types.
10. 
CONTENTS	XV
26. Complete vessels of Napo Plain Incised.
27. Type sherds of Napo Plain Incised.
28. Type sherds of Napo Plain Incised.
29. Type sherds of Napo Plain Incised.
30. Type sherds of Napo Plain Incised.
31. Type sherds of Napo Plain Incised showing rare combination of single-line and double-line
techniques.
32. Type sherd and bowl of Napo Red Excised.
33. Type sherds of Napo Red Excised.
34. Type sherds of Napo Red Excised, White Retouched.
35. Type sherds of Napo Red Incised.
36. Type sherds of Napo White Excised.
37. Type sherds of Napo White Incised.
38. Type sherds of Napo White Incised.
39. Type sherds of Napo White Incised.
40. Type sherds of Napo White Incised.
41. Bowls of Rocafuerte Incised combined with red and black-on-white painting.
42. Basin of Form 14 with Rocafuerte Incised decoration along one side and red and black-on-
white painting in Technique 2 on the remainder of the interior.
43. Type sherds of Rocafuerte Incised.
44. Type sherds of Rocafuerte Painted, red-on-white.
45. Type sherds of Rocafuerte Painted, black-on-white.
46. Rocafuerte Painted vessel of Form 6 from a site on the Rio Aguarico.
47. Type sherds of Rocafuerte Painted, red and black-on-white.
48. Type sherds of Rocafuerte Painted, red and black-on-white, Technique 2.
49. Type sherds of Rocafuerte Painted, red and black-on-white, Technique 5.
50. Vessels of Rocafuerte Painted, black-on-white.
51. Partially restored bowls of Napo Phase pottery types.
52. Rocafuerte  Painted   (red  and  black-on-white)  vessel of Form   11   with anthropomorphic
treatment.
53.	Rocafuerte  Painted  (red  and  black-on-white)  vessel of Form   11  with antoropomorphic
treatment.
54. Vessels of Rocafuerte Painted, Form 17.
55. Rocafuerte Painted anthropomorphic vessel Type A.
56. Rocafuerte Painted anthropomorphic vessel Type A.
57. Rocafuerte Painted anthropomorphic vessel Type B.
58. Rocafuerte Painted anthropomorphic vessels.
59. Rocafuerte Painted anthropomorphic vessel Type B.
60. Rocafuerte Painted anthropomorphic vessel Type B.
61. Rocafuerte Painted anthropomorphic vessels.
62. Rocafuerte Painted anthropomorphic vessels.
63. Rocafuerte Painted anthropomorphic vessel Type C.
64. Rocafuerte Painted anthropomorphic vessel Type C.
65. Rocafuerte Painted anthropomorphic vessel Type C.
66. Tiputini Plain Excised bowl of Form 5.
67. Type sherds of Tiputini Plain Excised.
68. Type sherds of Tiputini Plain Excised.
69. Type sherds of Tiputini Plain Excised with white retouch in incisions and excisions.
70. Type sherds of Tiputini Plain Incised.
71. Type sherds of Tiputini Red Excised.
72. Type sherds of Tiputini Red Excised.
73. Napo Phase pottery types.
74. Napo Phase pottery types.
75. Type sherds of Tipituini White Incised.
76. Pottery from Napo Phase sites.
77. Cotacocha Phase pottery types.
78. Sherds from Peruvian complexes representing the Zoned Hachure Horizon Style.
79. Decorated sherds from Coari.
80. Type sherds of the Guarita Phase.
81. Decorated sherds from Itacoatiara.
54. 
XVI	CONTENTS
82. Decorated sherds from Manacapuru.
83. Vessels from Manacapuru.
84. Decorated sherds of Marajoara Phase pottery types.
85. Decorated sherds of Marajoara Phase pottery types.
86. Marajoara Phase bowls with squarish form.
87. Type sherds of Joannes Painted, Marajoara Phase.
88. Anthropomorphic vessels from middle and lower Amazon sites.
89. Anthropomorphic vessels from middle Amazon sites.
90. Decorated sherds from Nazare dos Patos.
91. Excised and incised sherds from middle Amazon sites.
92. Decorated sherds from the vicinity of Santarem.
93. Decorated sherds from the Tefe region.
94. Decorated sherds from Loreto, Rio Suno, Ecuador.
82. 
293-822  O - 68 - 2

Archeological Investigations on the Rio Napo,
Eastern Ecuador



Geographical Description

The eastern Ecuadorian lowlands are a vast level
carpet of tropical rain forest drained by myriad rivers
and streams that merge one by one to form major
tributaries of the Amazon (fig. 1). Their nearly
adjacent, generally parallel courses dissect the terrain
when the water level is low, and spill over the inter-
vening land when it is high, so that " . . one can
not travel because of the great amount of water and
the marshes and creeks that there are there" (Pizarro,
in Heaton, 1934, p. 250). In 1541, this was "a great
uninhabited region" (op. cit., p. 247), transversed by
Pizarro and Orellana and their forces with hardship
and privation, and the archeological record indicates
that along the banks of the Rio Napo at least this
label was applicable during long intervals in pre-
historic times.
  The eastern face of the Andes is abrupt in Ecuador,
and the rivers cascade in narrow ribbons down steep-
walled canyons. At Tena, only 90 kilometers east of
Latacunga in the central highlands, the elevation
drops to 512 meters. Between these towns the eastern
cordillera rises above 4000 meters and snow-capped
Cotopaxi, 75 kilometers northwest of Tena, reaches
5897 meters. The eastern slope is clothed in dense
vegetation, the product of warmth and moisture.
Rainfall records over a 7-8 year period at Tena
show an average of 218.4 millimeters in January (the
dryest month) and 502.9 millimeters in June (the
wettest month), with a yearly average of 3810.0
millimeters. Although there is fluctuation over a 24-
hour period, average monthly temperature is nearly
constant. Records at Tena for a 4—5 year period
show averages of 22.3°C for the coldest month (July)
and 23.9°C for the warmest month (November),
with an annual average of 23.3°C (Ferdon, 1950, pp.
69, 75).
  
Daily temperature fluctuation is considerable, how-
ever, as a few observations made during our stay
at Tiputini clearly show. From a 6 a.m. low of 21.7°C,
the temperature may rise to 27.7°C at 10 a.m.,
30.0°C at midday, 31.7°C at 2 p.m., falling to 26.7°C
at 8 p.m. On a rainy day, the midday reading may
be only 25.5°C, rising to 26.7°C at 2 p.m. and
falling back to 25.8°C at 5:30 p.m. These are shade
records; one experiment in the sun pushed the maxi-
mum to 47.2°C. In human terms, there is great
variation in sensation of heat at the same tempera-
ture depending on the strength of the breeze and
the presence or absence of sun or rain. Although
21.7°C was the lowest reading we achieved, our most
vivid memories are of being chilly rather than too
hot.
  October, November, and December are generally
the dryest months, but rain can be expected any
day of the year. Showers are briefer in this "dry
season," and the lessened precipitation is reflected in
the lowering of the rivers. The Napo in December of
1956 was filled with large sandbars between which
the water meandered in narrow channels (pi. \a).
Since even the dryest month receives more rain than
falls in many other parts of the world over a whole
year, vegetation remains lush and green.
  The fauna is Amazonian and includes a variety
of birds and mammals suitable for sustenance of
man, although wildlife is rarely seen by a modern
traveler. Orellana's expedition was furnished with
"meats, partridges, turkeys and fish of many sorts"
(Carvajal, in Heaton, 1934, p. 175) and something
called "monkey-cat" (op. cit., p. 411) by Indians
living near the mouth of the Rio Curaray. Peccary,
agoutis, and monkeys are still caught and eaten in
the area. Of cultivated plants, the Spaniards were
offered  "maize and yuca and  sweet potatoes  and
  
SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6






FIGURE 1.—Map of the Province of Napo-Pastaza, showing principal landmarks and distances covered during descent of the Rio Napo.

other food products" (Ortiguera, in Heaton, 1934,
p. 314) raised in slash and burn clearings (pi. \b).
A more vivid impression of the stretch of river
between Latas (pi. 2a) and Tiputini as it appeared
during five days of November is provided by ob-
servations from the journal kept during our trip.
The daily distance traversed (fig. 1) provides a scale
against which the progress of Orellana's expedition
can be measured (see pp. 106-107).
  NOVEMBER 9—FIRST DAY : The river between Latas
and the Rio Misahualli has rock outcrops. The current
was strong and we floated most of the way. The
paddlers don't work much, but we seemed to be
travelling at a good pace. Following the map (fig.
1), we saw nothing at Vargas Torres. Farther down

there was a settlement with a church and several
houses near where Ahuano is shown on the map.
Passed there at 1 p.m. For the next 2 hours, we could
see the Cordilleira Galeras to the north and west
(15 km. from the Napo). Along the river were small
steep-sided hills, sometimes sliced through making
sheer 15 meter cliffs with strata of large pebbles like
on the present beaches. Between the hills were low
areas and sandbars covered with tall fan-leafed cane.
During the afternoon, the land became generally
lower, with wide sandbars often covered with large
rounded pebbles and littered with whole trees left by
the last high water (pi. 2b, 3b). Down here the river
seems to rise less than they said it did at Latas, but
it is hard to tell how much from the appearance of

GEOGRAPHICAL   DESCRIPTION

the vegetation. It is not a smooth flowing river, but
full of eddies, back washes, and turbulent areas.
When we cross one of the latter, it sounds as though
the bottom of the boat was being sand blasted.
  Stopped at 2:30 p.m. for 20 minutes at the settle-
ment of some gold panners near where Santa Rosa
is shown on the map. It was a low bank and must
flood in June. The paddlers swapped oranges for
some corn meal cakes and one-quarter of a roasted
monkey. The sun came out hot and bright and we
were glad it had been shady before. The river is so
wide, 500-1000 meters, that there is a lot of reflection
and it doesn't take long to sunburn. There are in-
numerable islands and other places that become is-
lands in high water. This is the widest river we have
travelled on—it makes the Essequibo in British Guiana
look like a creek. It must be fantastic in the rainy
season when it overflows the present banks.
  Except near Ahuano, we passed no settlement all
day and only 3 or 4 times did we see a house being
lived in. The land is too low or too steep for agri-
culture and the river is too rough with turbulence
to be very good fishing. A few kapok trees, some purple
flowered vines, otherwise the usual palm and ever-
green vegetation. Very few birds. Beautiful blue sky
and white clouds,  white tree trunks,  green leaves.
  When we stopped it was 6 p.m. and almost dark.
There was a wide bar on the left bank and we had
to walk several hundred meters back to where there
was a grove of trees to hang the hammocks. Slow
sprinkle off and on throughout the night.
  NOVEMBER IO—SECOND DAY: Awoke at 5:30 a.m.
as the sun was rising and got away at 6:05 a.m. The
morning was sunny and the river had dropped about
a meter so the current was less swift. Passed the mouth
of the Suno on the left at 8:05 a.m. and got caught in
the cross waters until the men had sense enough to
get farther out in the Napo. These river tributaries
are difficult to detect because their mouths look like
those of branches of the Napo coming from around
islands. The places shown on the map above and
below the Suno no longer exist. The islands are
larger, the banks are generally low and by noon we
had passed out of the area of pebbly sandbars. In the
vicinity of the Suno we could see a range of moun-
tains to the northwest, including the cone of the
volcano Sumaco (pi. 3a) with clouds hanging below
the peaks (50 km. from the Napo). About 11:15 a.m.
we passed the 300 meter contour line. Passed Armenia
Vieja at 11:35 a.m. The river has slowed down con-
siderably and by 2 p.m. the men were having to
paddle. It is remarkable that the water flows at all
when one thinks that it drops less than 300 meters in
3218 kilometers, under 15 centimeters per 1.6 kilo-
meters.
  
At 1 p.m. we stopped at the mouth of a small creek
entering from the left bank above the Payamino to
dynamite fish. One of the men cut a stick of dynamite
into thirds, wrapped one-third in a leaf, tied a pebble
to the bundle to make it sink, added the fuse and cap,
borrowed our matches and headed up the bank. We
waited with 2 men in the boat below the mouth.
They threw in a few hunks of clay and then the
charge. Two men dived in immediately and started
grabbing fish so fast they had to hold some of them
in their mouths. One started drifting downstream and
our men took after it, wasting a lot of energy because
they didn't catch it. Loaded the catch (half a dozen
large fish and a few more small ones) and continued
with only a half hour lost. Stopped at 1:40 p.m. at
the mouth of the Payamino for 15 minutes while the
men borrowed a pan from the residents to cook some
of the fish for lunch. High bank, no sherds.
  Passed the mouth of the Coca at 2:30 p.m. and
found one farm on the upstream side with some cows.
The Coca has its mouth divided by an island. Below,
the Napo is wider with banks from 1-6 meters high
showing signs of erosion during the last high water.
Most abundant wild life is butterflies, several of which
flitted around the boat, as well as some large bees. A
few more birds than yesterday—half a dozen white
herons, one blue egret, a tree full of paraquets, oriole
nests, a couple of kingfishers. A few can be heard
singing in the forest.
  A high, 6-meter bank appeared on the left at 4 p.m.
and was occupied by the most extensive settlement
we have seen since the village of Napo. The Indians
said it was Yana Yacu. There was a school near the
bank and a motor launch at the dock. Stopped at
5:05 p.m. on a sandbar on the left bank and started
looking for a place to sleep. The trees were farther
from the shore than last night and separated from
the beach by a shallow lagoon. Judging from its
dampness, this bar was mostly under water yesterday.
The Indians started cutting bamboo from among the
driftwood and putting poles in the ground to support
our waterproof nylon tarps as a tent. The poles
seemed secure so we started to cook supper as it
was getting dark. All was well until 3 a.m. when we
had a shower. Then from 4:30 a.m. until dawn we
were not sure we would escape catastrophe. A strong
wind rippled and ballooned the tarps and we dis-
covered that several places were not well tied. The
rain blew in from the north end and sprinkled in
from the east, but fortunately not enough to get us
really wet. The rain was accompanied by lightning
and thunder until dawn.
  NOVEMBER 11 —THIRD DAY : Cooked coffee and oat-
meal at 6 a.m. as the rain beat steadily down. The
area beneath the tarp was invaded by termites, which
  
SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6



were dispatched with DDT.  Inquired at 7 a.m. if
we went or waited, and the men said wait.
  9	a.m.—clouds beginning to break. Spent the past
3 hours absorbing the atmosphere of rain on the Napo
and contemplating the view to the west, upriver,
with trees increasingly gray with distance, dull gray
sky, eroded and polished tree trunk and driftwood
immediately beside the tent, sandbar continuing
beyond and also on the other shore—a dismal and
dreary scene. No one living for miles around; nothing
but sand, low cane, and trees. Howling monkeys in
the distance. We are learning why no one lived here.
This is the dry season too! We feel for Orellana in
1541 without nylon tarps and primus stoves, but
remembering the cold rainy highlands, perhaps he
didn't find it much different—maybe better since it
is warmer here.
  10	a.m.—Stopped raining so broke camp and left.
Sprinkled for a couple of hours. Passed the mouth
of the Jivino at 2:20 p.m. and arrived at Providencia
at 6:10 p.m. just after sunset. The river is very
capricious now. The sandbars change form constantly
and are very extensive in this area. Out in the middle
of a stretch a kilometer and a half wide it is so shallow
our dugout cannot get across. We had to back off
once this afternoon. As a result of having to zigzag,
we make less distance than we otherwise could. In
the current we make good time.
  NOVEMBER 12—FOURTH DAY: Didn't get away
until 7:40 a.m. It was clear when we left but clouds
began to build up during the morning, some white,
others black. Protection from the sun was welcome.
A couple of light showers. Came to San Roque at
1:00 p.m. and had a hard time getting to the house
because of a large sandbar in front. Left at 2:10 p.m.,
the sky cloudy and the breeze cool. More detours
around sandbars not yet above water. We are told
this is the best time to travel. Next month when the
water is lower, it flows swiftly through a narrow
channel and is dangerous. In June when it is high, it
is a menace, very wide, full of foam, and carrying
along debris from twigs to huge trees.
  We can follow our progress well with the U.S. Air
Force 1:500,000 Preliminary Base Map #950A, al-
though a few of the islands seem to have changed
since the air photos were taken in 1943. Came to a
high red bank at 4:15 p.m. with half a dozen aban-
doned houses, which we deduce to be an abandoned
oil camp at the beginning of the trail to the Aguarico.
Decided to stop, remembering two nights ago when
we traveled another hour below Yana Yacu and got
caught on a sandbar in the rain. First mosquitoes
came out at dusk so had to dig out the nets.
  NOVEMBER 13—FIFTH DAY: About 5 a.m. it started
to sprinkle and we were afraid we were stuck for most

of the day. However, it stopped and the boys said we
had to get started if we would make Tiputini tonight.
Shoved off at 5:40 a.m. Pushed along all day with
one stop to dynamite fish (no luck), one for the men
to bum some food (half a smoked agouti with head,
and 6 small bananas) and one to cut cane for poles to
be used on their return trip. Banks 1-3 meters high,
houses separated by long uninhabited stretches from
1 to 2 hours apart. Channel broken by many large
islands but the river has fewer bends so we are making
faster progress on the map (fig. 1). Trees not dense,
shores fringed with cane growing in strips 100-200
meters long; 3 main leaf types: fan-leafed, banana-
leafed, and asparagus-leafed. More trees with brown
and gray trunks, less than half are white. Occasional
hanging vines with red, pink, or white flowers. Clouds
building up all day to the east but dissolving with no
rain. Thunder in the distance also. Eroded treetrunks
and branches of temporarily stationary driftwood pro-
trude, especially near the banks. This river is very
dirty. The surface has fine sticks and the water is so
heavily laden with silt that a cupful has a brownish
tint.
  Were pleasantly surprised to round a bend and see
Tiputini in the distance. Looked like a city compared
with other settlements along the river. Pulled in at
5:15 p.m.
  In climate and topography, eastern Ecuador is part
of a zone extending up to an elevation of about 600
meters along the eastern Andes for about 8 degrees
north and south of the equator (Tosi, 1960, p. 236).
In this region, the silt-laden rivers overflow annually
inundating most of the countryside. Because silt depo-
sition is heaviest in the riverbed and along the margins,
the larger rivers gradually increase their elevation and
build up bordering dikes. The amount of alluvium
deposited decreases with increasing distance from the
river channel, producing a decline in elevation. As a
consequence, the annual fall of the rivers (pi. 4) is not
accompanied by drying of the hinterland, where
drainage is impeded by impermeability of the soil and
lack of access to all but the small and less elevated
streams. Evaporation is slow and extensive areas re-
main mucky or inundated throughout the year. Only
the dikes remain permanently above water or rapidly
emerge as the water goes down.
  To these disadvantageous factors is added limited
soil fertility. The situation has been summarized as
follows by Tosi (1960, pp. 243-244; translated from
Spanish):
  . . The climax soil of this formation, which is not common
and does not occupy extensive areas, is so called Low-Humic
Latosol ("latosol humico dulce"), the product of the cumulative
effect of climate and natural vegetation on the basic rocks of
  
GEOGRAPHICAL   DESCRIPTION




FIGURE 2.—Principal routes of communication between the highland basins and the eastern lowlands of Ecuador.

igneous or metamorphic origin and moderate slope. The pH of
5.0 to 6.5 indicates a certain degree of natural fertility under
forest cover and a good capacity for maintained or even in-
creased productivity with fertilization and other special
agricultural techniques . . More extensive geographically
and much more common is a complex of residual soils derived
from ancient marine, lake or alluvial sediments, which were
already very deficient in certain basic minerals at the time of
their deposition
  In part, this degree of senility . is due to the absence of
relief and of slope of the land: the rhythm of its geological
erosion is not and has not been as rapid as the rhythm of the
influence of atmospheric phenomena on the subsoil or of the
removal of bases from the upper horizons. This is the enigma:
the most level residual soils and those consequently best suited
for mechanized cultivation are among those of lowest natural
fertility. The majority are useless for permanent cultivation
even with abundant and frequent fertilization because of their
very low capacity for alteration and the toxicity of the iron and
aluminum they contain     .  .
  These lands are clearly of little or no value for development
of an agricultural or pastoral economy. Beneath the exuberant
and dense natural forest cover exists a very delicate ecological
equilibrium. If this primary equilibrium is interrupted by
cutting, burning, open cultivation or pasturing, the inevitable
result is loss of the already limited natural fertility of these
soils, which leads to their rapid abandonment.
  Although the limited potentiality of the eastern
lowlands has inhibited construction of modern roads,
numerous ancient trails provide communication with

the highlands (Wolf, 1933, pp. 225-226; Porras, 1961,
pp. 113-117). The two most travelled routes are that
from Quito to Papallacta and Baeza, with a northern
branch leading to the Rio Coca and a southern one
to Tena and the Rio Napo (fig. 2-1), and that from
Ambato via Bafios and Mera to Puyo and the Rio
Pastaza or the Rio Napo (fig. 2-2). Other trails lead
from the Riobamba basin to Macas on the Rio Upano
or to the Rio Pastaza (fig. 2-3); from the Cuenca
basin to the Rio Santiago (fig. 2-4); and from the
Loja basin to the Rio Zamora (fig. 2-5). Two more
routes originate in southern Colombia: one from
Tulcan to the headwaters of the Rio Aguarico (fig.
2-6), and the other from the Pasto region of the Rio
Putumayo (fig. 2-7). Other trails link rivers either
at the edge of the foothills or where their courses
run in closest proximity. Once in the lowlands,
abundance of navigable rivers, the flatness of the
terrain and the uniformity of environment facilitate
movement in almost any direction (pi. 5).
  Human settlement of the easten Ecuadorian low-
lands cannot be understood except in the context of
the topography, climate, natural resources, and agri-
cultural potential of the region. Combining limited
subsistence possibilities with almost unlimited flexi-
bility of riverine movement, the environment not
only channelled but in large measure has determined
the character of the archeological record along the
Rio Napo.
  
SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



  No evidence of preceramic inhabitants has been
reported, and in view of the nature of the terrain
and vegetation, it is doubtful that remains can be
found if preserved. Four pottery-making phases have
been recognized, each of independent origin, and all

but the last apparently of relatively short duration
in the area. These will be described separately in
detail, preliminary to a reconstruction of the pre-
history of this segment of eastern lowland Ecuador.




▲    LOCATION  DEFINITE
+   LOCATION  APPROXIMATE


FIGURE 3.—Rio Napo and its tributaries, showing the location of archeological sites and isolated finds.

The Yasuni Phase
DESCRIPTION OF SITES AND EXCAVATIONS

Two habitation sites of the Yasuni Phase were en-
countered in close proximity on the right bank of the
Rio Napo about 3 kilometers below the mouth of the
Rio Tiputini (fig. 3). Neither provided sufficient depth
for stratigraphic excavation.

N-P-l0: Puerto Miranda Hill
A small grass covered hill rises 16 meters above the
low water level on the right bank of the Rio Napo.
The summit is relatively level, and the slope precip-


MODERN
BUILDINGS

....     ,^.	    ■	,^-     -	..,.:~—   C\
LIMIT OF SITE
NAPO

FIGURE 4.—Sketch map of the Yasuni Phase site of N-P-10.

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



itous only on the north side, where the river has
exercised its cutting action (pi. 3c, 6a). A small creek
drains into the river about 120 meters upstream, and
the surface from here eastward and over the hill has
been cleared of its natural forest for pasture.
  A small habitation site is located on the northwest
portion of the summit and probably formerly extended
into the area now lost by erosion (fig. 4). Sherds occur
over an approximately circular area 18 meters in
diameter to a depth of 20-30 cm., beneath a sterile
zone 5 cm. thick occupied by dense grass roots. Soil
in the habitation zone was grayish brown, in contrast
to the reddish color of the sterile clay. Small rounded
pebbles were frequent, the larger ones broken.
Sherds were abundant but small and very badly
eroded. The hardness of the soil and shallowness of
the deposit made stratigraphic testing impossible and
the majority of the sherd collection came from a 2 by
1 meter area along the exposed margin facing the
river. Four sherds were encountered near a modern
house beside the creek on the level land to the west.
            
N-P-l 1: Puerto Miranda Bank
The artificial grassland extends for about half a
kilometer along the bank of the Rio Napo upstream
from the creek west of N-P-l0 (fig. 3). The elevation
of the almost vertical bank (pi. 6b) rises gradually
from 4 meters above low water level in the vicinity
of the creek to 6 meters at the vicinity of the site,
toward the western part of the clearing. Testing in
the area of highest elevation revealed sherds for a
distance of about 145 meters along the bank and 15
meters inland. The existence of a number at the
water's edge, fallen from the upper edge of the bank,
suggests that part of the site had been lost by erosion.
The refuse deposit extended to a depth of 25-30 cm.,
with sherds sparse in the upper 10 cm. Soil was
brownish gray sandy clay. The sherd sample was
collected from the summit as well as the river bank.
Data from Other Investigations
No other sites or artifacts attributable to Yasuni
Phase origin have been reported from the Rio Napo
or its tributaries.

ANALYSIS OF MATERIALS

Stone Artifacts
Close examination of rocks from Yasuni Phase sites
revealed the majority to be either concretions nat-
urally present in the soil or small water worn pebbles.
Only four showed traces of shaping for use.
Abrader
FIGURE 5
A thin rectanguloid piece of coarse quartzite with
metamorphosed surfaces has edges apparently rounded
by use. One surface is slightly concave and the other
correspondingly convex. Length is 5.5 cm., width
4.2 cm., thickness 1.2 cm. From Site N-P-10.
Ax Fragments
FIGURE 6
A piece of quartzite has been shaped by percussion
into what appears to be the butt of an ax. Outline
is not symmetrical. The sides are slightly concave
and have been smoothed to facilitate halting. The
implement is broken immediately below this region.
Cross-section is roughly ovoid. Surfaces are extremely
uneven and there is no evidence of pecking or polish-
ing. Width at the lower end of the notches is 5.2 cm.,
thickness 2.0 cm. Existing length is 4.2 cm. From
Site N-P-l 1.
  
A spall of metamorphosed tuffaceous sandstone has
a smooth to polished surface suggesting that it may
have come from an ax. However, this identification
remains tentative since no other polished axes have


0       I       2       3 CM
FIGURE 5.—Yasuni Phase stone abrader. Worn edge is between
arrows.

THE   YASUNI   PHASE



L
0       1        2       3 CM
FIGURE 6.—Yasuni Phase ax fragment.
been collected from Yasuni Phase sites. From Site
N-P-10.
Hammerstone
A waterworn pebble of sandstone conglomerate shows
slight battering on one end. It is 3.0 by 3.0 by 2.5
cm. From Site N-P-10.
Ceramic Classification
The pottery types of the Yasuni Phase are based on
classification of 1365 sherds from N-P-10 and 35
rim and diagnostic body sherds from N-P-l 1. Both
samples are badly eroded, typically removing both
surfaces and any decoration that might have been
applied to them. Consequently, the 4-percent occur-
rence of decorated sherds probably is an inaccurately
low representation of the original   frequency.
  Yasuni Phase pottery is now britde, crumbly and
soft, all characteristics for which erosion seems largely
responsible. Many sherds have been "dissolved" down
to the core, producing an abnormally thin wall (a
factor that has been taken into consideration in
measurement of body wall thickness). Only deeply
incised or punctated decoration and modeling re-
main visible under these circumstances. Extreme ero-
sion also causes rounding of broken edges,  making

intentional  lobing  difficult  to  distinguish   at  times
from nature's handiwork.
  Two plain types have been recognized on the basis
of temper: Miranda Plain tempered with sand, and
Yasuni Plain tempered with ash. The small number
of decorated sherds includes six distinct techniques,
each of them represented by very few examples.
Such small samples would normally be treated as
"unclassified decorated," but in this case ease of
reference and the probability that the techniques
represented were more frequent than surviving evi-
dence implies have served to justify the recognition of
six decorated pottery types. Decoration, apparently
applied to a plain surface, is by incision, punctation,
zoned hachure, a row of nicks, and simple adornos.
A few sherds show traces of red wash or slip. The
majority of lobed and nicked rims have been included
in the plain types in the absence of other evidence
of decoration. However, many, if not all, may belong
to vessels originally additionally ornamented by in-
cision or zoned hachure on the walls.
  No complete or restorable vessels exist for the
Yasuni Phase. All forms have been reconstructed
from evidence provided by rim, body, and base
sherds, including orientation, contour, and diameter.
Although all are treated as circular, some fragments
exhibit a degree of flatness suggestive of ovoid form.
Diameters at the large end of the size range may
represent such cases, and may thus be misleading.
The presence of square vessels seems improbable in
the absence of sherds from corners, which are abun-
dant in Napo Phase pottery where square vessels are
common.
  Pottery type descriptions have been arranged in
alphabetical order, following descriptions of vessel
shape characteristics. Provenience and frequency of
both pottery types and vessel shapes are given on
Appendix tables 1 and 2.
Reconstructed Vessel Forms
Common Forms
1.	Shallow bowl or platter (fig. 7-1):
Rim: Profiles are unstandardized. The principal
variants are interior thickening to produce a broad,
level or slightly insloping surface or, less commonly,
upturned 1.0-1.5 cm. below the lip. Some of the
latter appear to have a horizontal incision on the
exterior about 5 mm. below the lip. Exterior rim
diameter 22-48 cm.; majority 28-40 cm.
Lip:   Tapered or rounded.
Body wall thickness:    5-10 mm.
Base:    Probably rounded or slightiy flattened.
2.	Rounded bowl (fig. 7-2):
Rim:   Direct, outslanting at an angle between 50 and

10

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



70 degrees. Rim diameter varies between 8-36 cm.;
majority 24—36 cm.
Lip:   Tapered, rounded, or flattened.
Body wall thickness:   5-7 mm.
     Base:   Probably rounded or slightiy flattened.
3.    Small rounded bowl with slightly everted rim (fig. 7-3):
Rim: Sharply everted 0.5-1.0 cm. below the lip and
typically slightly thickened on the interior at the region
of eversion, producing a flat horizontal or slightly in-
sloping top. Interior rim diameter 6-20 cm.; majority
6-8 cm.
Lip:    Tapered or rounded; occasionally undulating.
Body wall thickness:   2.0-3.5 mm.
Base:    Probably rounded.
4-.    Large  rounded  bowl  with  slightly  everted   rim   (fig.
7-4):
Rim: Everted and thickened 1-2 cm. below the lip,
producing a trianguloid profile. Rim top is flat and
level or outsloping. Interior rim diameter 18-34 cm.;
majority 28-32 cm.

Lip:   Rounded or rarely, flattened; occasionally em-
bellished with notches or miniature lobes.
Body wall thickness:   3.0-8.5 mm.
Base:    Probably rounded or slightly flattened.
5.	Large bowl with short vertical wall (fig. 7-5):
Rim: Outsloping walls turn upward 1.5-3.0 cm. below
the lip producing a rounded shoulder, above which
the wall is usually thickened. Slight to pronounced
eversion occurs about 1 cm. below the lip, often
combined with interior thickness producing a flat,
usually horizontal top. Interior rim diameter: 14-38
cm.; majority 26-36 cm.
Lip:   Rounded
Body wall thickness:    4.5-7.5 mm.
Base:   Rounded or flattened.
6.	Large,   deep,   carinated   bowl   with   trianguloid   rim
(fig. 8-6):
Rim:    Slightly outflaring, everted, and thickened 1.0-
1.5 cm. below the lip producing a trianguloid cross-







////III
I   i   i   i   i   I   i
0     4      8     12 CM
VESSEL   SCALE
i  . i  ■ i  ■ i
0      I      2      3 CM
RIM    SCALE


  yjrssrryr
>f/H!J
             
FIGURE 7.—Rim profiles and reconstructed vessel shapes of the Yasuni Phase, Common Forms 1-5.  (Black=undecorated, white=
decorated.)

THE   YASUNI   PHASE

11



rt
i-rt r r
IT""

>m)<


u   '  '   I   '  '
0     4     8     12 CM
VESSEL   SCALE

'   i   i   i   i   ■   i
0      1       2      3   CM
RIM   SCALE

/ X

FIGURE 8.—Rim profiles and reconstructed vessel shapes of the Yasuni Phase, Common Forms 6-9 and Base Forms A-C.  (Black=
undecorated, white = decorated.)

12

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



section. The top is typically concave and sloping
downward toward the exterior. Interior rim diameter
22-46 cm.; majority 30-40 cm.
Lip:   Tapered or rounded.
Body wall thickness:   4.5-10.0 mm.
Base:   Rounded or flattened.
7.	Large carina ted bowl with broad flange rim (fig. 8-7):
Rim:    Sharply everted to produce a broad horizontal
or slighdy upslanting or downslanting flange, 2.0-
3.5 cm. wide; angle of eversion is more rounded on
the exterior than the interior. Interior rim diameter
16-36 cm.
Lip: Rounded or slighdy flattened; typically embel-
lished with lobes and notches.
Body wall thickness:   4.5-6.5 mm.
Base:   Rounded or flattened.
8.	Carinated bowl with everted rim (fig. 8-8):
Rim: Sharply everted to produce a narrow flange
(width about 1.5 cm.), with horizontal, upslanting
or downslanting orientation. Angle of eversion is
typically more rounded on the exterior. Interior rim
diameter 22-42 cm.; majority 26-36 cm.
Lip: Tapered, rounded, or slighdy flattened; often
embellished with notches or undulating.
Body wall thickness:   4-7 mm.
Base:    Probably rounded or flattened.
9.	Deep bowl with cambered rim (fig. 8-9):
Rim: Outflaring to about 1 cm. below the lip, where
it is inturned at an angle approximately 90 degrees
to the previous direction, producing an inslanting
camber with a slightly concave exterior. Rim diam-
eter 16-26 cm.
Lip:   Rounded.
Body wall thickness:    3-5 mm.
Base:    Probably rounded or slightly flattened.
Rare Forms
1. Deep bowl with a slightly incurving upper wall
turning outward 2.5 cm. below the rounded lip.
Rim diameter 28 cm. (fig. 9a).
2. Vessel with slightly outslanting upper wall with a
flange attached on the exterior between 1 and 2
cm. below the rounded lip. The flange projects 1 cm.
above the exterior surface, and is ornamented by
vertical nicks 5-7 mm. apart. Rim diameter about
28 cm. (fig. 9*).
3. Deep bowl with nearly vertical wall slighdy thickened
on the exterior 2 cm. below the lip and tapering
upward from that point. Rim diameter about 30 cm
(fig- 9c).
4. Vessel with slightly constricted mouth. Rim turns
upward 1 cm. below the rounded lip. Mouth diameter
22 cm. (fig. 9d).
5. Jar with upper wall incurving and then upturned 2
cm. below the lip. Exteriorly thickened for 1 cm.
below the rounded lip. Rim diameter 20 cm. (fig. 9e).
6. Jar with incurving upper wall, everted to produce a
1 cm. wide slightly outsloping collar. Tapered lip.
Mouth diameter 16 cm. (fig. 9/).


7. Vessel with convex insloping upper wall joining
lower wall at a 100 degree angle, exteriorly thickened
rim, rounded lip. Mouth diameter about 12 cm. (fig.
9g).
8. Vessel with constricted mouth, direct rim, and rounded
lip. Mouth diameter 14 cm. (fig. .9h).
9. Open bowl with a raised interior border produced
by increasing the thickness of the wall by 2 mm.,
3 cm. below the lip (fig. 90- A slight change in
curvature of the exterior wall occurs at the location
of the increase in thickness on the interior. The lip
is rounded and embellished by lobing and nicks.
Two body sherds show a similar treatment on the
interior.

i   i   i   ■   i   ■  i
0       I       2      3 CM
10. Open bowl with outslanting wall, exteriorly thickened
rim with triangular cross-section. Diameter at rim
top, 18 cm. (fig. 9/").
FIGURE 9.—Rim profiles  of Yasuni  Phase  Rare  Forms   1-10.
Base Forms
Four distinct base forms are represented in the sherds.
Since there are no complete vessels, their association with
the nine reconstructed vessel forms is hypothetical.
A.	Rounded, unthickened, with a curved or slightly
angular junction with the body wall. (fig. 8-A).
B.	Flat, joining the body wall at an angle between 55
and 65 degrees. Junction more curved on interior
than exterior as a result of slight thickening, (fig. 8-B).
C.	Flat, joining the body wall at an angle of 90 to 110
degrees, and thickened on the exterior at the point of
junction, forming a heel. Diameter 16-18 cm. (fig.
8-C).
D.	Annular, tapering from point of attachment to lower
edge and slighdy outflaring. Identification is based
on orientation and curvature of two small badly eroded
fragments. Diameter 18 cm.
Pottery Type Descriptions
Miranda Modeled
PASTE AND SURFACE:   All on Miranda Plain (p.  13); see
that type description for details.

THE   YASUNI   PHASE

13



FORM:
Rim:   Direct or slightly exteriorly thickened with tapered
  or flattened lip.
Body wall thickness:   3-5 mm., increasing to 9 mm. at
  carination.
Base:   No direct evidence.
Reconstructed vessel shapes:   The only rim represents Rare
Form 7; other sherds are from the shoulder of carinated
    vessels.
DECORATION (fig. 10):
or low relief
surface and
One rib has
Technique:    High relief nubbins (fig. 10 b, c
ribs   (fig.   10a)   applied   to   the   exterior
smoothed over to obliterate the junction.
sloughed off leaving a slight depression. Ribs are 3 cm.
long, approximately 1 cm. wide at the surface, and 3



0       1       2       3 CM
FIGURE 10.—Type sherds of Miranda Modeled.

mm. wide at the flattened top. Elevation is about 3 mm.
Nubbins are prominent and the only complete example
is smaller than average, measuring 1.5 cm. along the
shoulder and projecting 5 mm.
Motif: Ribs run vertically about 4 cm. apart on the rim
exterior; nubbins project from angle of carination (their
frequency cannot be described because of the absence
of sherds with more than one nubbin per sherd).
TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE :   Yasuni Phase.
Miranda Plain
PASTE:
Temper: Waterworn sand, principally clear and white
quartz. Grains rarely exceed 1 mm., and typically are
smaller than 0.5 mm. Temper is well distributed and
abundant enough to give eroded surfaces a sandpaper
texture. A minority have very litde sand temper.
Texture: Abrasive like fine-grained sandpaper; compact
without air pockets. Breaks with a jagged edge, which
is not friable.
Color: About 50 percent are completely oxidized, light
orange to bright orange through the cross-section; the
the remainder are dark gray throughout or oxidized on
the surface leaving a medium to dark gray core.
  Method of manufacture:    Coiling; coil line breaks are rare.
SURFACE :
Color: Light tan to light brown most typical, with varia-
tions toward light orange and gray brown. Occasional
small medium gray irregularly shaped fire clouds. In-
terior rarely a uniform black.
Treatment: Less than 5 percent have original surfaces
intact. These were smoothed sufficiendy to depress
temper grains, producing a compact, fine-grained tex-
ture smooth and even to touch although with imperfec-
tions visible. One sherd has broad (1-2 mm.) horizontal
smoothing tracks.
  Hardness:   2.5.
FORM:
Rim: Exteriorly thickened, interiorly thickened, everted,
cambered or direct, with rounded, flattened, or tapered
lip.
Body wall thickness: Range 2-9 mm.; majority 4-6
mm.
Base: Flat (Form B, fig. 8-B); rounded, curving gradu-
ally to side walls (Form A, fig. 8-A).
Reconstructed common vessel shapes (figs. 7, 8):
Form 8: 18.2 percent.
Form 4: 13.2 percent.
Form 2: 12.1 percent.
Form 1: 12.1 percent.
Form 3: 7.7 percent.
Form 5: 7.7 percent.
Form 6:   6.6 percent.
Minor vessel shapes (frequency less than  5  percent; figs.
8, 9):   Forms 7 and 9; Rare Forms 3, 8, and 9.
OCCASIONAL DECORATION:   Vessels of Common Forms 3, 4,
7, and 8 often have nicked or lobed lips.

14

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE:   Yasuni Phase.
Yasuni Incised
PASTE AND SURFACE:   Predominantly on Yasuni Plain (p.
15),  rarely on  Miranda  Plain   (p.   13);  see those  type
descriptions for details.
FORM:
Rim:   Direct or flange,  with rounded  lip;  large  lobes
  occur on open bowls of Form 2.
Body wall thickness:    5-8 mm.
Base:   Annular (Form D) or flattened (Forms B and C).
Reconstructed common vessel shapes (figs. 7, 8):
Form 2: 4 rims; decoration on interior.
Form 6: 1 rim; decoration on exterior.
Form 7:  3 rims; decoration on rim top and exterior
above waist.

DECORATION (fig. 11):
Technique: Incised lines on a plain surface. Execution
varies from thin sharp marks resembling knife cuts
(width 0.1-0.2 mm.) to lines 0.5-1.0 mm. wide and
about 0.5 mm. deep. Wider lines may have very ir-
regular margins, but the extent of surface erosion makes
it impossible to determine whether this reflects sloppy
technique of incision or damage subsequent to burial.
(Where arrangement of pits was sufficiently regular to
appear intentional, the sherd was classified as Yasuni
Incised and Punctate.) Well preserved incisions are
straight and sharply defined, although sometimes in-
terrupted where they cross small surface defects. Inter-
sections often undershot. Straight lines tend to be paral-
lel but not evenly spaced. Separation 0.6-1.2 cm.
Motif: Patterns are composed principally of arrange-
ments of parallel straight lines. The most frequent are
parallel horizontal lines spaced 5-12 mm. apart, creating
bands   broken   by  vertical   incisions  into  rectangular







J
5 CM







FIGURE 11.—Type sherds of Yasuni Incised.

THE   YASUNI   PHASE

15



zones  (fig.   lla-b,  d-e), and parallel horizontal lines
displaced by several steps at intervals (fig.  11/). One
design features a squared spiral (fig. lie).
Associated techniques:   A row of small vertical nicks may
occur  along  the  lower  edge  of the  decorated  zone
(fig. Ua-b).
TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE:   Yasuni Phase.
Yasuni Incised and Punctate
PASTE  AND  SURFACE:   All on Yasuni  Plain  (p.   15);  see
  that type description for details.
FORM:


Rim:    Flange with flattened lip ornamented with small
broad lobes and notches.


                     0      1       2      3  CM
FIGURE  12.—Type sherds of Yasuni Incised and Punctate.
Body wall thickness:    3-5 mm.
Base:   No direct evidence.
Reconstructed vessel shapes (figs. 8, 9):
Form 7: 1 rim; decoration on rim top.
Rare Form 10: 1 rim; decoration on interior.
DECORATION (fig. 12):
Technique: Relatively broad (about 1.5 mm.), deep
(1 mm.), U-shaped incisions terminating in, or accom-
panied by, circular punctates about 2 mm. diameter
at the surface. Punctates are considerably deeper than
incisions and have a conical or rounded contour.
Although erosion obscures details of execution, several
sherds give the appearance of drag-and-jab or "dotted"
lines creating parts of the design (fig. 12 a-b).
Motif: Parallel and intersecting lines on the rim top,
symmetrically accented by single punctates or termi-
nating in a punctate, or parallel incisions on the surface,
some of which terminate in a punctation or have
"dotted" segments.

TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE:   Yasuni Phase.
Yasuni Nicked
PASTE AND SURFACE:   Four on Yasuni Plain (p. 15), one on
Miranda Plain (p.  13); see those type descriptions for
details.
FORM:
Rim:   Exteriorly thickened at or below the lip.
Body wall thickness:    3-9 mm.
Base:   No direct evidence.
Resconstructed vessel shapes (figs. 7, 9):
Form 5:   1 rim; decoration on vertical exterior wall.
Rare Form 2:  1 rim; decoration on lip of flange.
Rare Form 5:  1 rim; decoration on upper surface of rim
thickening.
DECORATION (fig. 13):
Technique:    Parallel vertical cuts or nicks,  in one case
probably made with fingernail, in others by a pointed
tool. Length 5-7 mm., width about 1 mm., separation
1-11  mm., with little variation on a single example.
Depth less than 1 mm.
Motif:   Single row adjacent to lip, on lip of flange rim or
on exterior wall above carination. (Nicked lips, common
on vessels of Form 8, have been considered occasional
embellishments of plain types.)
Associated technique:   Two sherds have small rim lobes.
TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL DISTRIBUTION OF THE TYPE :   Yasuni Phase.
Yasuni Plain
PASTE :
Temper: Black ash, the larger particles elongated and
showing the cellular structure characteristic of cariape.
Size ranges from minute specks to particles 4 mm. long;
the majority are intermediate between these extremes.
Glossy lumps of irregular form are characteristic; black
color contrasts sharply with gray paste producing a
speckled appearance. Light gray siliceous particles rep-
resenting the typical appearance of cariape (Meggers
and Evans, 1957, p. 81) occur in about 5 percent of the
sherds. Waterworn sand is characteristically also pres-
ent, and often abundant.
Texture: Extremely friable, probably in part because of
the badly deteriorated condition of the sherds. Sand
temper gives a sandy texture, but the paste is less abra-
sive to the touch than that of Miranda Plain. Air
pockets are absent, but holes left by leached organic
temper occur.
Color: Typically medium to light gray throughout the
cross-section; about 20 percent are light orange or light
tan throughout the cross-section.
Method of manufacture:    Coiling; rare fractures along coil
junctions show overlapping union.
SURFACE:
Texture: Observation is limited to about 5 percent of
the sherds with traces of the original surface. These

293-822  O - 68

16

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6







L
0       1        2       3 CM
FIGURE  13.—Type sherds of Yasuni Nicked.

are smooth to the touch and generally even although
some temper grains remain visible. One sherd shows
polishing striations on interior and exterior. Some have
a low gloss.
Color: Typically light tan to grayish tan, shading to
whitish or light orange. Small medium gray fire clouds
probably are characteristic, since they are common on
the few sherds with the surface preserved. Dark gray
to black surfaces are excessively rare.
Hardness:   2 (in view of the extent of erosion, this softness
may not accurately reflect the original condition).
FORM:
Rim: Exteriorly thickened, ipteriorly thickened, everted,
cambered, flange or direct, with rounded, flattened, or
tapered lip.
Body wall thickness: Range 2-8 mm.; majority 5-6 mm.
Loss of surfaces suggests that these measurements may
be about 1 mm. too small.
Base: Thickened (Form C, fig. 8-C); flat or slighdy
flattened (Form B, fig. 8-B). Two small fragments may
represent annular bases, but their eroded condition
makes broken surfaces impossible to distinguish con-
clusively from finished ones so that their present
appearance may be deceptive.
Reconstructed common vessel shapes (figs.
7,8)
Form 6
Form 8
Form 5
Form 4
Form 7
Form 1
36.2 percent.
14.7 percent.
13.7 percent.
10.7 percent.

8.8 percent.
5.8 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
7-9): Form 2, 3, 9, Rare Form 6.
OCCASIONAL DECORATION: Vessels of Forms 4, 7, and 8
often have nicked or lobed lips.
TEMPORAL DIFFERENCES WITHIN THE TYPE: None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE :   Yasuni Phase.
Yasuni Red
PASTE AND  SURFACE:   All on Yasuni Plain  (p.   15);  see
that type description for details.
FORM:
Rim:   Flange (lip missing).
Body wall thickness:   4-7 mm.
Base:   No direct evidence.
Reconstructed common vessel shapes (fig. 7) :
Form 5 with flange rim; paint on interior and exterior.
Carinated body sherd; paint on exterior.
DECORATION :
Technique:   Rich red slip, well preserved on one sherd,
faint trace on the remainder. Fine grained, even, thin.
Motif:   Completely covering the surface or applied to
the rim and body wall above shoulder.
TEMPORAL     DIFFERENCES     WITHIN     THE     TYPE:    None
  observable.
CHRONOLOGICAL POSITION OF THE TYPE:   Yasuni Phase.
  
THE   YASUNI   PHASE

17



Yasuni Zoned Hachure
PASTE AND SURFACE:   On Yasuni Plain (p. 15); see that
type description for details.
FORM:
Rim:   Direct or exteriorly thickened with rounded lip.
Body wall thickness:   4—6 mm.
Base: No direct evidence.
Reconstructed vessel shapes (figs. 7, 9):
Form 2: 1 rim; decoration on interior.
Form 4: 1 rim; decoration on interior.
Rare Form 4: 1 rim; decoration on exterior.
  Rare Form 10: 1 rim; decoration on interior.
DECORATION (fig. 14):
Technique:   Incised   lines   bounding    bands   and   zones
textured   with  fine   hachure.   Incisions   0.5-1.0   mm.
wide, straight, generally parallel, and evenly spaced.
Intersections may be overshot or fall short of junction.
Hachure composed of very fine, shallow, closely spaced
lines typically perpendicular to the longest axis of the
zone,  or of more haphazardly arranged and widely
  spaced fine lines, or crossed lines.
Motif:   Parallel  incisions  6-10   mm.   apart   appear   to
define   bands   or   rectangular   zones   filled   with  fine
parallel or cross hachure;  alternatively, straight and
stepped incisions form more irregularly shaped zones
containing fine hachure.  Severe erosion makes most
patterns impossible to reconstruct.

TEMPORAL     DIFFERENCES     WITHIN     THE     TYPE:   None
  observable.
CHRONOLOGICAL POSITION OF THE TYPE:   Yasuni Phase.

J
3 CM
FIGURE 14.—Type sherd of Yasuni Zoned Hachure.

THE SERIATED SEQUENCE AND ITS IMPLICATIONS

The shallowness of the refuse deposit at both of the
habitation sites assigned to the Yasuni Phase
obviated the possibility of securing stratigraphic
information on trends of ceramic change. Conse-
quently, the phase characteristics must be viewed as
representing a single point in time, However, pro-
nounced differences in the relative frequency of vessel
shapes at the two sites (Appendix table 1) suggest
that they  are  not contemporary,  and  that  further

work in the area will bring to light additional
sites that may permit establishment of a seriated
sequence.
  A single carbon-14 date was obtained by extracting
the organic temper from 2 kilograms of Yasuni Plain
sherds from N-P-10. The date of 2000 ±90 years
ago, or 50 B.C. (SI-300) seems acceptable in view of
the prehistoric sequence on the Rio Napo and the
probable affiliations of the Yasuni Phase.
  
DIAGNOSTIC FEATURES OF THE YASUNI PHASE

The Yasuni Phase is defined from investigation of
two sites on the right bank of the Rio Napo, neither
with sufficient depth of refuse to permit stratigraphic
excavation. One of the sites is situated on an elevated
section of the river bank, the location favored by all
of the other phases. The other is unique in occupying
the summit of a high hill adjacent to the river. Both

have suffered badly from erosion making the original
dimensions difficult to reconstruct. However, the
smaller site on the hilltop was estimated at 20-30
meters in diameter, while the other appeared to extend
about 145 meters along the bank and 15 meters inland.
No evidence was found of disposal of the dead.
There is no evidence bearing on subsistence pattern.

18

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Hunting, fishing, and gathering resources must have
been exploited, and slash-and-burn agriculture can
also be assumed to have been practiced. However,
the absence of griddle fragments implies that bitter
manioc was absent or not processed for consumption
in the manner typically employed in later times.
Although rim forms and body contours of pottery
vessels are distinct from those of other phases identified
in the area, the differences do not seem to reflect
significant differences in function.
  The pottery of the Yasuni Phase has been classified
into two plain and six decorated types. Miranda
Plain is sand tempered, abrasive in texture, and
incompletely to completely oxidized in firing. Yasuni
Plain is tempered with particles of charcoal and
cariape, producing a very friable texture, and pre-
dominantly incompletely oxidized in firing. Both
types occur with approximately equal frequency at
the point in time represented by the unselected
sherd sample from N-P-10.
  With the exception of red slipping covering the
entire surface (Yasuni Red), decoration is by incision
or modeling. It must be kept in mind, however,
that these techniques are most resistant to obliteration
and the surfaces of all sherds are in extremely poor
condition. The extreme rarity of all the decorated
pottery types probably also reflects the condition of

the sample rather than the original frequency of
decoration. Incision may occur alone (Yasuni Incised),
in conjunction with punctation (Yasuni Incised and
Punctate) or be used to delimit rectilinear zones
textured with fine parallel or cross hachure (Yasuni
Zoned Hachure). Other forms of embellishment are
a row of nicks along the lip or exterior surface (Yasuni
Nicked), and simple applique (Miranda Modeled).
  Vessel shapes emphasize open containers, wide in
comparison to depth. Rims are typically everted to
produce a narrow to broad horizontal to sloping
flange, often thickened on the exterior at the angle
of eversion and frequently ornamented with lobes or
nicks. Direct and short cambered rims also occur.
In addition to rounded or flat bases, an unusual
"heeled" form occurs. Low annular bases may occur.
  The only shaped stone tool is a fragment of a
percussion chipped ax. A few natural stones showed
wear from use in pounding or abrading. No pottery
artifacts were found.
  No pottery of trade origin was identified from
either of the Yasuni Phase sites. This factor, added
to the relatively greater degree of deterioration of
the pottery surfaces, argues in favor of an early
position in the regional sequence. This inference is
supported by the single carbon-14 date, which places
the occupation of N-P-10 at 2000 ±90 years ago,
or 50 B.C. (SI-300).
  

  

100   M


FIGURE 15.—Sketch map showing the locations of sites N-P-6 of the Napo Phase and N-P-7 and N-P-8 of the Tivacundo Phase
the left bank of the Rio Tiputini.

The Tivacundo Phase
DESCRIPTION OF SITES AND EXCAVATIONS

Two habitation sites of the Tivacundo Phase were
found in close proximity on the left bank of the Rio
Tiputini, a tributary entering the right bank of the
Rio Napo (fig. 3, pi. 7).
N-P-7: Chacra Alfaro
At the time of our visit in November 1956, a small
area about 100 meters inward from the left bank of the
Rio Tiputini was cleared and planted in manioc,
plantain, banana, and papaya. Although grass cov-
ered most of the surface, sherds and potrest fragments
were abundant over an area 30-35 meters in diameter
toward the southeast edge (fig. 15). Concentration
was greatest near the center, becoming sparser toward
the margins. The soil throughout the clearing was
medium to dark brown with a high proportion of clay,
causing it to bake hard with exposure to the sun. A
surface collection was made, which was supplemented
with material derived from scattered tests.
  A 1 by 1 meter stratigraphic test pit was excavated
northwest of the center of the site, where the refuse
accumulation appeared to reach maximum depth.
Levels were controlled in arbitrary depths of 8 cm.
Level 0-8 cm. was largely occupied by grass roots.
Soil was contaminated with ash of the recently burned
clearing and contained relatively few sherds. In Level
8-16 cm. the soil was looser and slightly darker
brownish tan; sherds were abundant. In Level 16-24
cm. sherds became sparser, disappearing as the brown
loamy clay of the natural soil was reached.

N-P-8: Barranco Alfaro
A little upstream from N-P-7, the Rio Tiputini makes
a 180 degree bend. The force of the current in the
rainy season has eaten into the bank, causing slumping
and exposing fragments of pottery for a distance of
about 2.5 meters (fig. 15). Tests revealed nothing,
however, suggesting that all but a remnant of the site
has been washed away. As a result of displacement of
the soil, sherds extended from 75 cm. below the surface
down to the water's edge (a depth of 4 meters). While
the original depth and thickness of the refuse deposit
could not be clearly ascertained, it was probably
relatively shallow.
  Repeated exposure to percolating rainwater and to
river action had produced an advanced degree of
deterioration in the pottery, especially the surfaces,
which often adhered to the sticky clay soil when the
attempt was made to remove them. Moisture has
also made sherds soft and easily fractured. In spite
of the difficulties, a relatively large sample was ob-
tained. Because of the impossibility of stratigraphic
control, all materials were combined into a single
collection.
Data from Other Investigations
No other sites or vessels, either complete or fragmen-
tary, of Tivacundo Phase pottery types have been
recorded from the Rio Napo or the Rio Tiputini.
19

20

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY
ANALYSIS OF MATERIALS

VOLUME   6



Stone Artifacts

The only stone artifact from the Tivacundo Phase is a
small irregularly shaped andesite pebble with several
flat slick surfaces resulting from use as a polishing
tool (fig. 16). Elongated with a trianguloid cross-
section, it measures 3.5 cm. long, 1.5 cm. wide at the
base, 0.5 cm. wide at the top, and 2.2 cm. in maximum
height. Provenience is N-P-7, Cut 1, Level 16-24 cm.

L
J
J_
0	I	2 CM
FIGURE 16.—Tivacundo Phase polishing pebble.
Pottery Artifacts
In terms of their origin, Tivacundo Phase pottery
artifacts fall into two categories: manufactured ob-
jects and reused sherds. Examples of the former are
limited to potrests; the latter consist only of abraders.
Abraders
FIGURE 17
Three sherds of Chacra Plain containing a large
amount of sand in addition to the charcoal particles
diagnostic of this pottery type show two or more deep
grooves on the exterior surface produced by abrasion.
Width has a different range of variation on each sherd,
probably reflecting differences in the diameter of the
shaft being abraded, since depth on all three examples
is 1-2 mm. The ranges of variation are 3-4 mm.,
4-5 mm., and 5-6 mm. Grooves are generally parallel
to the rim, but may be horizontal or diagonal in
orientation. The sherds include a rim of Form 2, a
base of Form B, and a body sherd. All are from N-P-7,
Surface.
Potrests
PLATE 8k-i
Forty-three irregularly shaped hunks of fired clay
represent an unknown number of potrests. The ma-
jority have very sandy paste and are fired dark to
bright orange throughout. Rare variations in com-

position include almost pure clay, and fine sand and
charcoal temper. This lack of consistency suggests
that potrests were made from leftover material for
pottery manufacture, or from whatever clay was at
hand. They were constructed by shaping one large
or several small lumps into a cylinder with flattened
ends. Their poor state of preservation suggests that
what firing they received resulted from use to sup-
port a cooking vessel.
  Surfaces typically remain uneven, retaining pits,
scars, and high spots. A few show vertical grooves,
probably finger smoothing marks. Form is cylindrical,
with slightly concave sides, flattened ends, and rounded
corners. The absence of complete examples makes
height unmeasurable, but fragments suggest that it
was about 15 cm. Base diameters range from 8-12
cm.; one upper end has a diameter of 9 cm.
Ceramic Classification
The pottery types of the Tivacundo Phase are based
on classification of 2241 sherds. If the selected sample
from   N-P-8   is   eliminated   from   the   calculation,

J
1
J-
0       I       2       3 CM
FIGURE  17.—Base sherd used as an abrader, Tivacundo Phase.

THE   TIVACUNDO   PHASE

21



decoration occurs on 5.6 percent of the sherds. The
extensive erosion of the surfaces, which would re-
move painted designs, may have distorted the fre-
quency to some extent, but there is no reason to
assume that it was originally markedly higher.
  Tivacundo Phase pottery in general is characterized
by sandy texture and oxidized surfaces. Cross-sections
sometimes show a laminated structure but paste is
typically compact. All surfaces have been badly dam-
aged by exposure to acid soil, making their original
condition difficult to describe. In general, they
appear to have been even, but neither floated suffi-
ciendy to eliminate defects nor polished. A thin red
wash has been tentatively identified on a few sherds,
but red or white slips were not observed.
  Three plain pottery types have been recognized
on the basis of temper and firing differences: Alfaro
Plain, sand tempered and incompletely oxidized;
Tivacundo Plain, sand tempered and completely
oxidized; and Chacra Plain, tempered with charcoal
particles. Two additional varieties of temper are
represented in rare sherds: These are cariape and
fine muscovite sand. Decorative techniques are re-
stricted to fine incision and red painting, which are
typically combined in zoned patterns. Vessel shapes
are few and simple; ovoid as well as circular outline
occurs. Symmetry is good.
  With the exception of Form 3, represented by a
complete ovoid bowl of Tivacundo Incised and Zoned
Red, all vessel shapes have been reconstructed from
evidence provided by sherds. Association of the
different base forms with body and rim profiles
has been inferred from considerations of diameter
and inclination, except in Forms 1-3 where large
sherds extend from lip to bottom. Worthy of special
note are several sherds representing flat bottoms
riddled with perforations made when the clay was
wet.
  Pottery type descriptions have been arranged in
alphabetical order, following description of vessel
shape characteristics. Provenience and frequency of
both pottery types and vessel shapes are given on
Appendix tables 4 and 6.
Reconstructed Vessel Forms
Common Forms
1.	Rounded bowl with direct rim (fig. 18-1):
Rim:    Outcurving to upcurving, unthickened. Diameter
  14-36 cm.; majority 24-32 cm.
Lip:   Rounded, rarely slighdy flattened.
Body wall thickness:   4—8 mm.
Base:    Probably rounded or slightly flattened.
2.	Rounded bowl with slightly thickened rim (fig. 18-2):
Rim:    Outcurving to nearly vertical, with small thick-

ening or beading on the upper 5 mm. of the exterior
not exceeding the body wall thickness by more than
1.5 mm. A similar effect is occasionally produced on
unthickened rims by a narrow groove 2-6 mm. below
the lip on the exterior. Diameter 16-30 cm.; some
vessels may be ovoid rather than circular.
Lip:   Rounded or flattened at center.
Body wall thickness:   5-7 mm.
Base:    Probably rounded or slighdy flattened.
3.	Depressed globular bowl with slightly thickened rim
(fig. 18-3):
Rim: Incurving, with slight thickening or beading on
the upper 3-5 mm. of the exterior not exceeding the
body wall thickness by more than 1 mm. The effect
may be accomplished by slight eversion rather than
thickening. Rim diameter 14-36 cm., majority 20-30
cm.; ovoid as well as circular examples occur.
Lip:   Rounded; rarely slightly flattened at center.
Body wall thickness:    3.5-6.0 mm.
Base:   Probably rounded or slightly flattened.
4.	Depressed globular bowl with everted, thickened rim
(fig. 18-4):
Rim: Sharply everted (angle between 50 and 90
degrees) 1.0-2.5 cm. below the lip, producing a flat
insloping top, and thickened on the exterior producing
a trianguloid cross-section. The lower edge of the
thickening is unerased and demarcated in cross-
section by 1-2 mm. increase in thickness. Interior
rim diameter 12-36 cm.; majority 16-26 cm.
Lip:   Tapering or flattened.
Body wall thickness:    4—9 mm.
Base:   Probably pedestal (Form A) or flat (Form B).
5.	Collared jar (fig. 19-5):
Rim: Upper 1.5-4.0 cm. of the incurving wall is bent
upward producing a nearly vertical or insloping
collar. Diameter 20-30 cm.; majority 22-26 cm.
Lip:   Rounded or slightly flattened at the center.
Body: A number of body sherds showing an angular
change in contour on the exterior may be associated
with this form (fig. 20g).
Body wall thickness:   4—7 mm.
Base:   Probably flat (Form B) or concave (Form C).
6.	Large globular jar with slightly thickened rim  (fig.
19-6):
Rim: Upper 0.8-1.6 cm. of incurving wall is bent
upward, producing a narrow beading or collar, which
may be accentuated by slight thickening. The lower
edge may be a step-like elevation or a narrow groove.
Mouth diameter 18-46 cm.; majority 30-42 cm. On
several sherds interior junctions are unobliterated on
the last three coils adjacent to the lip.
Lip: Rounded or flattened, resulting in a circular or
rectanguloid cross-section.
Body wall thickness:   0.8-1.9 cm.
Base:    Probably pedestal (Form A) or flat (Form B).
Rare Forms
1.    Flat platter or griddle,  1.2-2.2 em. thick, with direct
rim and rounded or tapering lip. Three examples with

22

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6







W7j)j//mn



I   ■   l   i   I   i   I
0      I      2      3 CM
RIM   SCALE
■ i i ' ' ' i
O      4     8      12 CM
VESSEL  SCALE



FIGURE 18.—Rim profiles and reconstructed vessel shapes of the Tivacundo Phase, Common Forms 1-4. (Black=undecorated, white=
decorated.)

THE   TIVACUNDO   PHASE

23


V
<WR


n«wv


	
O     4     8     12 CM
VESSEL   SCALE


l   ■   I   ■   l   ■   '
0      I      2      3 CM
RIM     SCALE

FIGURE 19.—Rim profiles and reconstructed vessel shapes of the Tivacundo  Phase, Common Forms 5 and 6. (Black=undecorated
white = decorated.)

24

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



diameters of 24-26 cm. may represent griddles, but
two with diameters of 14 and 18 cm. are smaller than
the usual range of griddle size (fig. 20a).

2. Globular jar with constricted mouth and direct or ex-
panding rim, rounded lip. Mouth diameter 14 cm.
(fig. 20b-c).
3. Rounded bowl with expanding rim, rounded lip. In-
terior rim diameter 26 cm. (fig. 20d).
4. Ovoid bowl with rounded shoulder and slightly everted
rim accompanied by slight thickening of the interior
wall (fig. 20*).
5. Large shallow vessel with concave bottom, nearly ver-
tical curved walls and broad horizontal flange rim (lip
missing). Depth 7.5 cm.; base diameter 24 cm. (fig. 20/).


B.
C.
D.

Flat, typically joining the body wall at an angle of
40-50 degrees, but occasionally falling above or below
this range. Junction is typically angular on the ex-
terior and rounded on the interior, and 2-5 mm.
thicker than the body wall. Several examples are rid-
dled with unequally spaced (2-10 mm. apart) perfora-
tions 3-5 mm. in diameter, punched from the exterior
when clay was very wet before firing and creating a
ridge about 1 mm. elevation around the opening on the
interior (pi. Sf-g). Diameter of perforated examples
8-12 cm.; diameter of unperforated examples 18-24
cm.   (fig.   21-B).
Concave, with an angular or rounded junction to the
body wall. A concavity of 2-5 mm. is produced by
thinning of the wall and does not result in convexity on
the interior. Diameter 12-18 cm. (fig. 21-C).
Incipient annular base, resulting from combining the
pedestal of Form A with the concavity of Form C.
Diameter  16 cm.   (fig.  21-D).
Rounded, continuation of curvature and thickness of
the body wall (pi. 9).




0        I        2       3 CM
FIGURE 20.—Rim profiles of Tivacundo Phase Rare Forms.
a, Rare Form 1. b-c, Rare Form 2. d, Rare Form 3. e, Rare
Form 4. /, Rare Form 5. g, Typical shoulder treatment.
(Black = undecorated, white = decorated.)
Base Forms
Five base forms can be distinguished in the sherd sample.
The rarity of complete vessels makes association of base and
rim forms dependent upon correlation of dimensions and
general body contour. Only the occurrence of rounded or
slightly flattened bases (Form E) with rims of Vessel Forms
1-3 is confirmed by complete vessels.
A. Flat and sharply demarcated from the body wall on the
exterior by a nearly vertical to outsloping "pedestal"
5-10 mm. high, sometimes reflected by a convexity on
the interior. Diameter 10-28 cm. (fig. 21-A).

Pottery Type Descriptions
Alfaro Plain
PASTE :
Temper: Abundant angular grains of multicolored (pink,
white, black) sand, estimated to comprise about 50
percent of the volume. Size rarely exceeds 1 mm.
Occasional sherds have additional rare grains of black
charcoal.
Texture: Very sandy and friable; eroded surfaces and
broken edges crumble into sand when rubbed, prob-
ably in part because of deterioration of paste in the
ground. Compact.
Color: Typically medium to dark gray throughout the
cross-section; occasionally oxidized in a band adjacent
to one or both surfaces leaving a gray core. Rarely, the
the gray core is displaced to the inner half of the cross
section, oxidation penetrating from the exterior surface
to the center.
   Method of manufacture:    Coiling.
SURFACE :
Color: Brown, grayish brown, or brownish orange.
Darker hues tend to occur on the interior. A minority
of sherds show a whitish to light orange tint. Dark gray
fire clouds occur.
Treatment: Few sherds retain their original surface;
these are even but not smooth or polished, and retain
minor defects and visible temper grains.
   Hardness:    3.
FORM:
Rim: Exteriorly thickened, everted or expanded, with
rounded or tapered lip.
Body wall thickness:    Range 4-10 mm.; majority 4-7 mm.
Base: Rounded or slighdy flattened (Form E); flat
(Form B, fig. 21-B); pedestal (Form A, fig. 21-A);
concave (Form C, fig. 21-C).

THE   TIVACUNDO   PHASE

25



Reconstructed common vessel shapes (figs. 18, 19):
Form 2: 27.0 percent.
Form 5: 23.1 percent.
Form 4: 21.2 percent.
Form 3: 11.5 percent.
  Form 6:   9.7 percent.
Minor vessel shapes (frequency less than 5 percent,  figs.
  18, 20): Form 1; Rare forms 2 and 3.
TEMPORAL DIFFERENCES WITHIN THE TYPE: Whitish to
light orange surfaces are most frequent at the earliest
site (N-P-8). Vessel Form 5 is restricted to this same
portion of the sequence (Appendix table 4).
CHRONOLOGICAL POSITION OF THE TYPE: A common plain
type throughout the seriated sequence, with a trend of
slightly declining frequency (fig. 26).
                      Chacra Plain
PASTE :
Temper:    Small   irregularly   shaped   lumps   of  charcoal




'  ■   i  ■  '
3 CM
FIGURE 21.—Base forms of Tivacundo Phase plain pottery. A,
Flat, "pedestal". B, Flat, angular junction with body wall.
C, Concave. D, Incipient annular.

and angular grains of multicolored sand; either may
predominate or they may be equally abundant (when
charcoal is rare, i.e., only one or two particles could be
detected, the sherd was classified as Alfaro Plain or
Tivacundo Plain). Charcoal particles range from pin-
point size to about 1 mm. diameter, rarely reaching
3 mm. in length. Elongated hunks may resemble cariape
but the striated cellular structure of the latter is absent.
Carbon may also take the form of irregular small
black or medium gray splotches, sometimes visible only
under 5X magnification. Sand grains are under 1 mm.
diameter. Distribution of charcoal particles is often
uneven, but the gaps are occupied by less prominent
sand particles with the result that temper is well
distributed.
Texture: Compact, sandy but not friable. Air pockets
are not characteristic but small cavities may occur
between coil junctions where kneading is not complete.
Occasional examples have a laminated appearance.
Color: Dark gray or medium gray throughout cross-
section; rarely, light brown. Carbonizing effect of
charcoal temper causes blotchy appearance in some
light brown pastes. Oxidation seldom penetrates
beyond the surface, but occasional sherds have an
oxidized layer of relatively even width extending to
1 mm. inward.
Method of manufacture: Coiling, exhibited in unobliterated
junctions on some interior surfaces, and occasionally in
cross-section.
SURFACE:
Color: Light orange to buff to light brown of relatively
even hue except where marred by small (2-3 cm. dia.)
dark gray fire clouds. Oxidation sometimes produces a
whitish surface film. Interior surfaces are typically
less completely oxidized than the exterior and occa-
sionally have a grayish hue.
Treatment: Few sherds retain their original surfaces;
these show smoothing sufficient to obliterate most sand
and charcoal temper particles but leaving scattered
irregular scars up to 1 by 3 mm. in size. Even but
slighdy abrasive to the touch.
Hardness:   2.5-3.
FORM:
Rim:   Direct,   exteriorly   thickened   or   everted,   with
  rounded, tapered or flattened lip (pi. 8a-,?).
Body wall thickness:   Range 0.4-1.5 cm.; majority 5-8 mm.
Base:   Rounded   or   slighdy  flattened   (Form   E);   flat
(Form   B,   fig.   21-B,   pi.   Sf-g);   pedestal   (Form  A,
  fig. 21-A); incipient annular (Form D, fig. 21-D).
Reconstructed common vessel shapes (figs. 18, 19):
Form 4: 28.0 percent.
Form 3: 21.5 percent.
Form 6: 15.9 percent.
Form 1: 14.0 percent.
Form 2: 11.2 percent.
  Form 5:   5.6 percent.
Minor vessel shapes (frequency less than 5 percent; fig. 20):
  Rare form 1.
TEMPORAL   DIFFERENCES   WITHIN   THE   TYPE:  Vessels   of
Form 5 are common only at the earliest site, while Form
2 is absent (Appendix table 4).
  
26

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



CHRONOLOGICAL POSITION OF THE TYPE: Present through-
out the seriated sequence, increasing from 12.3 to 34.8
percent and then declining to 23.3 percent (fig. 26).
Tivacundo Incised and Zoned Red

i	i	i	i
0       I       2      3  CM
PASTE: Decoration was applied to all of the plain types,
with the following frequencies: Alfaro Plain (p. 24),
46.5 percent; Tivacundo Plain (p. 27), 39.1 percent;
Chacra Plain (p. 25), 8.6 percent; Unclassified sand-
tempered Plain (p. 29), 0.5 percent; see those type
descriptions for details.


FIGURE 22.—Type sherds of Tivacundo Incised and Zoned Red.
a-b, Technique  1.   c, Technique 2.

SURFACE :
Color:   Pale orange to buff.
Treatment: All suffer from some degree of erosion. Those
in good condition are even and smooth to the touch,
although some temper grains remain visible and fine
horizontal scratches left by the smoothing tool can be
seen under 5X magnification.
FORM:
Rim:   Exteriorly   thickened   or   slightly   everted   with
rounded flattened or tapered lip.
Body wall thickness:   Range 0.4-1.4 cm.
Base:   Rounded or slightly flattened (Form E); pedestal
(Form A, fig. 21-A).
Reconstructed common vessel shapes (figs. 18, 19):
Form 3: 50.0 percent.
Form 6:  19.3 percent.
Form 2:  11.5 percent.
Form 4:    7.8 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
18, 20):   Form 1; Rare Form 5.
DECORATION (figs. 22, 23; pi. 9):
Technique: Fine incised lines superimposed with a red
painted line, or demarcating zones filled with red.
Incisions 0.2-0.5 mm. wide, straight, and even in
appearance to the unaided eye. Magnification to 5X
shows irregularities in depth and discontinuities. Mar-
gins may be even or jagged. Long lines are straight
although parallelism is imperfect especially when
separation exceeds 1 cm. Overshot or undershot
junctions are typical. Faint or broken lines occur in
some areas unrelated to the final pattern, suggesting
that guidelines were sketched onto the surface to orient
the artist. Red coloring is applied in two ways:
1. In bands and zones bounded by incisions (the most
common technique). Rich, dark, even color subject
to flaking off especially when the underlying
surface is well smoothed inhibiting penetration.
Paint may fill the zone or fall short of reaching the
boundary leaving an unpainted strip 0.5-1.0 mm.
wide. None of the examples show overlap of paint to
surface outside the boundary incision (figs. 22a-b, 23;
pi. 9b-d).
2. Covering fine incisions (less common and used in
conjunction with Technique 1). Rich red band, 1.5-
2.5 mm. wide applied so as to evenly straddle an
incised line, overlapping the surface on both sides
(fig. 22c; pi. 9e).
Motif: Complicated all-over patterns on the exterior,
composed of small symmetrical or asymmetrical zones
and broad bands, are characteristic of Technique 1.
Zones typically combine straight and curved lines. The
most distinctive element is a small concave-sided equi-
lateral triangle. Scale is correlated with vessel size:
zones tend to be small on small vessels and large on
large ones, the constituent elements remaining similar.
Technique 2, also restricted to the exterior, is applied
to parallel lines drawn independently or connecting
red zones.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observable.

THE   TIVACUNDO   PHASE

27






FIGURE 23.—Rolled out design on exterior of Tivacundo Incised and Zoned Red oval bowl (cf. pi. 9).

CHRONOLOGICAL POSITION OF THE TYPE :   Limited to N-P-8,
the earliest site in the seriated sequence (fig. 26).
                       Tivacundo Plain
PASTE :
Temper: Angular grains of multicolored sand, diameter
typically less than 0.5 mm. but sometimes larger than
1 mm. Abundant with a tendancy toward clustering.
Some sherds also have rare grains of charcoal.
Texture: Sandy and friable, crumbling into sand on
eroded surfaces and broken edges. Tendency to lamina-
tion, with air pockets occurring between layers.
Color:   Bright orange, reddish orange, or rarely brownish

orange  through  cross-section.   A few  examples hav
orange core and gray band adjacent to the surface.
Method of manufacture:   Coiling, evident in occasional coil
line fractures and incompletely obliterated junctions on
some jar rim interiors.
SURFACE :
Color: Light orange, reddish orange, brownish orange,
buff or tan, with the tone generally uniform on small
sherds. Rarely medium to dark gray, probably localized
fire clouding. Exterior and interior of a single sherd
generally have a similar hue.
Treatment: Few sherds retain their original surface; these
are  even  and  smoothed  sufficiendy  to create  slight

28

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



floating but leaving some temper grains exposed and
small irregularities.
Hardness:    3 to 3.5.
FORM:
Rim:   Direct,    exteriorly   thickened   or   everted,   with
  tapered, rounded or flattened lip (pi. \0a-h).
Body wall thickness:   Range 0.3-1.5 cm.; majority 5-9 mm.
Base:   Rounded   or   slighdy   flattened   (Form   E);   flat
(Form B, fig. 21-B), sometimes perforated  (Form B,
pi.   IQi-j);   pedestal   (Form   A,   fig.   21-A);   concave
  (Form C, fig. 21-C).
Reconstructed common vessel shapes (figs. 18, 19):
Form 2: 25.0 percent.
Form 4: 19.8 percent.
Form 1: 16.7 percent.
Form 3: 16.7 percent.
  Form 6: 14.6 percent.
Minor  vessel shapes  (frequency less  than 5 percent, fig.
  19): Form 5.
TEMPORAL DIFFERENCES WITHIN THE TYPE: Surface color
of examples from N-P-8 is lighter orange than on
examples from N-P-7. Vessels of Form 5 occur only at
N-P-8.
CHRONOLOGICAL POSITION OF THE TYPE: Tivacundo Plain
shows a pronounced increase in frequency and constitutes
the most abundant type in the latest levels of the seriated
sequence (fig. 26).
Tivacundo Red Painted
PASTE:   Decoration occurs on the following plain types:
Chacra Plain (p. 25), 41.8 percent; Alfaro Plain (p. 24),
32.7  percent;  Tivacundo  Plain   (p.   27),   25.5  percent.
When segregated by levels, frequencies generally duplicate
the  trends of popularity exhibited  in  the plain types,
implying that selection was not practiced in vessels for
decoration.   See  plain   type  descriptions  for  details  of
paste.
SURFACE:   All  surfaces  are  badly  eroded,   but  evidence
suggests their original condition was similar to that of
Tivacundo Incised and Zoned Red.
FORM:
Rim:    Direct  or  exteriorly  thickened  with  rounded  or
(Form   E);  flat,
, 19):
  flattened lip.
Body wall thickness:   Range 4-10 mm.
Base:   Rounded   or  slightly  flattened
  sometimes perforated (Form B).
Reconstructed common vessel shapes (figs, lc
Form 2: 42.2 percent.
Form 6: 36.8 percent.
Form 1: 10.5 percent.
    Form 3: 10.5 percent.
DECORATION:    (fig. 24; pi. l\a-e):
Technique:   Rich dark red paint applied in narrow (2-3
mm.) or wide (5-10 mm.) bands or irregularly shaped
areas.  Extensive damage by erosion makes details of

execution impossible to observe.
Motif:   Patterns  of straight  and  curved  parallel  lines
on the exterior. Only two sherds (both from N-P-8)
have complicated designs in the style associated with
Tivacundo Incised and Zoned Red.  Rim exterior is
painted red on larger jars.
TEMPORAL DIFFERENCES WITHIN THE TYPE:   Complicated
patterns are restricted to N-P-8, the only site producing
Tivacundo Incised and Zoned Red.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence as the most common decorated
type (fig. 26).

0       I       2       3  CM
FIGURE 24.—Type sherds of Tivacundo Red Painted.
Unclassified Cariape-tempered Plain
PASTE:
Temper: Cariape; particles typically 1-2 mm. in length,
occasionally 5 mm. Color predominandy white, but
gray and black particles also occur, as well as some
black charcoal lumps like the tempering of Chacra
Plain. Orientation is random. Rare sherds contain sand
in addition to cariape.
Texture:    Compact; no lamination or air pockets.
Color: Typically medium to dark gray core leaving
oxidized band up to 2 mm. wide adjacent to one or
both surfaces; rarely light orange through cross-section.
  Method of manufacture:    Coiling.
SURFACE :
Color: Light orange to whitish on exterior; light orange
or medium gray on interior.
Treatment:   All  badly  eroded;  rare patches of original

THE   TIVACUNDO   PHASE

29



FIGURE 25.
surface show smoothing sufficient to obliterate most
temper particles but not removing all pits and defects.
Hardness:    2.5.
FORM:
Rim:   Direct or upturned and "beaded," with rounded
  lip (pi. 11/).
Body wall thickness:   Range 0.5-2.0 cm.
Base:   Flat (Form B, fig. 21-B).
Reconstructed common  vessel  shapes   (fig.   19):   The  single
rim sherd represents Form 6.
TEMPORAL     DIFFERENCES     WITHIN     THE     TYPE:     None
observable.
CHRONOLOGICAL   POSITION   OF   THE    TYPE:   Limited    to
N-P-8, the earliest site in the seriated sequence (fig. 26).
Unclassified Sand-tempered Plain
PASTE:
Temper:    Very fine sand grains up to 0.5 mm. diameter
and smaller particles of muscovite that glitter as specks
  on eroded surfaces.
Texture:   Very compact, fine grained.
Color:    Cream  or  buff throughout  the cross-section  or
adjacent to exterior surface leaving a pale gray band
along die interior; rarely, light gray throughout the
  cross-section.
Method of manufacture:    Coiling;   coil junction  fracture
relatively common, and showing steep angle of overlap.
SURFACE:
Color: Pale cream or buff of even tone on both interior
and exterior; rarely pinkish orange. Medium gray dis-
coloration results from fire clouding.
Treatment: No sherds retain their original surfaces; pres-
ent condition suggests only that it was relatively even.
Hardness:   3.5.
FORM:
Rim:   Direct,    exteriorly   thickened   or   everted,   with
  rounded or tapered lip.
Body wall thickness:   5-7 mm.
Base:   Probably rounded or slighdy flattened (Form E).
Reconstructed common vessel shapes (figs. 18, 20):   Forms 2,
4, and Rare Form 2.
TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE:   Limited to the
early part of the seriated sequence (fig. 26).
Unclassified Decorated
  INCISED (fig. 25a).—Two body sherds have broad
(2-3 mm.) shallow (less than 1 mm.) incisions, curved
in cross-section. Incisions are straight and parallel.
Execution is poor so that width and depth are not
consistently maintained. One on Chacra Plain, one
on Tivacundo Plain; both from N-P-7, Cut 1, Level
0-8 cm.
  
One fragment, a constricted mouthed vessel with
direct rim, has fine incision on the exterior, similar in
technique to that of Tivacundo Incised and Zoned
Red, although executed with less precision (fig. 25a).
Paste is Tivacundo Plain. The last two coils adjacent
to the lip are unobliterated on the interior. N-P-7,
Cut 1, Level 8-16 cm.
  NICKED SHOULDER (fig. 256).—One sherd has a row
of nicks or notches of variable length (4-7 mm.) and
separation (4-6 mm.) along a slightly angular
shoulder. A narrow horizontal red-painted line is
faintly visible just above. Paste is Chacra Plain.
N-P-7, Cut 1, Level 8-16 cm.
  RED SLIP.—Several sherds have a reddish film on
the exterior or both surfaces that may be the remnant
of a wash or slip. Color is a dull reddish brown rather
than the rich red of painted decoration. Five on Chacra
Plain, three on Tivacundo Plain, one on Alfaro Plain;
Vessel shapes 4, 5, and 6 are represented. All from
N-P-8.


0       I       2       3 CM
-Unclassified decorated sherds from the Tivacundo
Phase,   a, Incised,   b, Nicked shoulder.

30

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY
THE SERIATED SEQUENCE AND ITS IMPLICATIONS

VOLUME   6



The seriated sequence of the Tivacundo Phase is
based on tendencies exhibited by plain pottery types
in successive levels of the stratigraphic excavation at
N-P-7 (fig. 26). Two of the types show marked and
regular changes in frequency, Chacra Plain declining
from 31 percent to 22 percent, and Tivacundo Plain
increasing from 32 percent to 45 percent during the
period of time represented by the refuse accumula-
tion. Alfaro Plain shows a smaller decline in fre-
quency, from 27 to 25 percent. Tivacundo Red
Painted, the only decorated type present throughout
the sequence, shows no consistent pattern of change.
The relatively frequencies of pottery types in the
sample from N-P-8 appear to fit the trends shown by
the plain types best when seriated below the earliest
level at N-P-7. If the distorting influence on plain
type frequencies resulting from the abnormal amount
of Tivacundo Incised and Zoned Red is overcome by
classifying the decorated sherds in terms of the plain
types on which the decoration was applied, the best
agreement is still achieved by placing N-P-8 at the
lower end of the sequence. A single carbon-14 date,
obtained by extracting the organic material from 2
kilograms of sherds of Chacra Plain from N-P-7,
Cut 1, Level 0-8 cm., places the latter portion of the
seriated sequence at 1440 ±70 years ago, or A.D.
510 (SI-330).
  
The relative chronological position of the two
sites is of importance because of several marked
differences in the occurrence of minor plain and
decorated types, which are of potential utility in
tracing the affiliations of the Tivacundo Phase. A
small number of undecorated sherds with cariape
temper were identified in the sample from N-P-8,
while none were found in the larger sample from
N-P-7. The other unclassified plain type, tempered
with fine muscovite sand, is more frequent at N-P-8
than at N-P-7, where it occurs only in the earliest
level of the stratigraphic excavation. Of the decorated
types, Tivacundo Incised and Zoned Red is limited
to N-P-8 (fig. 26).
  Differences between the relative frequencies of most
of the pottery types suggest that there is a gap between
the occupation of N-P-8 and N-P-7. This inference
is strengthened by examination of differences in
vessel shape popularity. Since the upper and lower
levels of the excavation at N-P-7 produced too few
rim sherds to constitute a reliable sample (Appendix
table 4), the rims from all three levels have been
combined. This procedure results in rim samples of
nearly equal size from the two sites, minimizing
errors that might be attributed to differences in size
of the samples compared.
When the relative popularity of the six common



N-P-7,   CUT   I, 0-8 CM
8-16
16-24

□
D
a



N-P-8, SURFACE
0    10   20   30  40   50%

ALFARO
PLAIN

TIVACUNDO
PLAIN

CHACRA
PLAIN

■	■
I	I
UNCL.       UNCL
PLAIN       PLAIN
(CARIAPE) (SAND)

I
TIVACUNDO
INCISED a
ZONED   RED

I
TIVACUNDO
RED
PAINTED

FIGURE 26.—Seriated sequence of Tivacundo Phase sites based on changes in pottery type frequency (see appendix table 6).


i i i i i i i i i
0     8     16   24   32 CM

y^y



N-P-7, CUT I,  0-24 CM

l	l	C

N-P-8,  SURFACE
    3                                         4
VESSEL     FORMS   	
FIGURE 27.—Changes in relative frequency of Tivacundo Phase Common Forms 1-6. Sites are arranged in seriated order.

THE   TIVACUNDO   PHASE

31



vessel shapes at N-P-8 and N-P-7 is compared,
marked differences are evident in two: Forms 2 and 5
(fig. 27). The pronounced increase in popularity of
Form 2 can be accounted for to a limited extent by
the corresponding decrease in the frequency of Form
3, from which it differs only in inclination of the
rim (fig. 18). Since Form 3 is the one most often
chosen for incised and red zoned decoration, its
greater frequency at N-P-8 may reflect selection of

the sample for decorated sherds. However, an even
sharper contrast in the popularity of Form 5 at the
two sites cannot be dismissed on such grounds, since
this form is always undecorated, as far as can be
determined from the present condition of sherd
surfaces. The only example of an incipient annular
base (fig. 21-D) comes from N-P-8, but the general
infrequency of base sherds makes it possible to attribute
this restriction to the inadequacy of the available
sample.

DIAGNOSTIC FEATURES OF THE TIVACUNDO PHASE

The archeological description of the Tivacundo Phase
is based on evidence from two habitation sites occu-
pying adjacent locations on the left bank of the Rio
Tiputini where the elevation is above flood level.
Only one is intact, the other having been nearly
destroyed by erosion of the river bank. The former
has a diameter of 30-35 meters, and the refuse
accumulation does not exceed 24 cm. This is the only
phase not encountered on the Rio Napo, suggesting
that smaller streams may have been the preferred
location. No evidence was found of burial practices.
  Although no direct evidence of subsistence pattern
was encountered, it can be assumed that fishing,
hunting, and slash-and-burn agriculture were prac-
ticed. A few sherds of flat, circular pottery griddles
were identified, but their rarity as well as the rela-
tively small reconstructed diameter make it appear
improbable that they were utilized for preparation
of bitter manioc. Also of problematic significance
are vessels with flat bottoms riddled with small
perforations, which would have permitted draining
or steaming of the contents.
  The pottery of the Tivacundo Phase has been
classified into three plain and two decorated types.
Alfaro Plain, an incompletely oxidized sand-tempered

ware, declines slightly in frequency, while Tivacundo
Plain, sand tempered and completely oxidized, shows
a considerable increase from early to late. Chacra
Plain, tempered with ash, attains maximum popu-
larity in the middle of the seriated sequence. Two
minor plain types, one tempered with cariape and
the other with fine sand, are restricted to the early
levels. The most striking decorated type, Tivacundo
Incised and Zoned Red, is also an early diagnostic,
while red painted designs not bounded by incision
(Tivacundo Red Painted) occur with similar fre-
quency throughout the sequence. Vessel shapes are
rounded, open or globular, and typically have a
slightly modified "beaded" rim. Some are ovoid
rather than circular. A conical-based vertical-necked
jar is an early form.
  Pottery artifacts are limited to sherds used as
abraders and large solid, cylindrical potrests. No
shaped stone tools have been reported, and the only
stone artifact found was a small polishing stone.
  The only carbon-14 date available (see p. 93)
indicates that the Tivacundo Phase occupied the
region subsequent to the end of the Yasuni Phase
and prior to the arrival of the Napo Phase, or around
the middle of the first millenium A.D.
  
293-822  O - 68 - 1

The Napo Phase
DESCRIPTION OF SITES AND EXCAVATIONS

The best represented of the archeological phases along
the Rio Napo and its tributaries is the Napo Phase.
Seven habitation sites were investigated, but although
the majority were extensive in area the refuse deposit
was too shallow or too disturbed by modern occupa-
tion for stratigraphic excavation except in the case of
N-P-2. Information on disposal of the dead comes
from numerous accidental finds of burial urns in the
region.
N-P-l: Tiputini
The main headquarters of the Shell Petroleum Com-
pany during exploration in the eastern Ecuadorian

lowlands was on the left bank of the Rio Napo a few
kilometers above the mouth of the Rio Tiputini (fig.
3). The buildings subsequently were taken over by
the Ecuadorian Army and have remained in use.
This location was probably selected for modern occu-
pation because the land from the river bank inward
for a considerable distance maintains an elevation
above flood level, providing not only a large area for
habitation but also conditions suitable for an airstrip.
The elevation also made it attractive as a village site
to earlier immigrants, and the segment paralleling the
river bank is occupied by one of the principal sites of
the Napo Phase.
The area of Napo Phase habitation coincides almost


FIGURE 28.—Sketch map of the Napo Phase site of N-P-l, showing the relationship of the archeological remains to modern buildings and
the arbitrary divisions made for surface collection.
32

THE   NAPO   PHASE

33



exacdy with that on which modern buildings have
been erected (fig. 28), extending some 650 meters
along the river bank and increasing from about
40 meters wide at the northwest end to about 120
meters in maximum width in the southeastern half.
The southeastern portion occupies the summit of a
peninsula-like elevation, the edges of which drop off
rapidly on the east and more gradually on the south-
east. Here, and at the northwest end of the site, eleva-
tion reaches about 7 meters above low water level,
while the center is depressed slightly to 6 meters above
water level. The vertical river bank becomes more
sloping as the summit narrows, and with descending
elevation becomes low and subject to flooding im-
mediately below the southeastern edge of the site. The
Napo is very wide here and a low island lying opposite
Tiputini (pi. la) is inundated during the rainy season.
Modern occupation of the site has done extensive
damage to the archeological deposit. The surface
around most of the buildings is kept clean of vegeta-
tion, with the result that up to a meter of the soil has
been removed, along with sherds and other kinds of
cultural evidence (pi. 12a). It was consequently im-
possible to find a spot with sufficient depth for strati-
graphic testing, and only surface collections could be
made. Since it was important to determine whether
the site area represented a large population for a
short period of time or a small population successively
occupying various portions of the area over a longer
period of time, the surface material was collected in
four spatial units (fig. 28). Comparison of the percent-

age frequency of the various pottery types in the four
collections permits evaluation of the nature of the
occupation at the site (see pp. 78-80).
  With the exception that sherds were less abundant
in Area 1, no significant differences were observed in
occurrence or distribution of the archeological re-
mains. Most of the sample was collected from eroding
banks, drainage ditches, and less disturbed portions
of the surface. Where all vegetation had been elimi-
nated, the soil was hard packed and baked and sherds
were badly eroded and easily broken. The circular
bodies of several badly fractured vessels were exposed
in the path in front of the mess hall and two vessels
attributed to Tiputini have been preserved in the
Colegio Militar in Quito (pis. 54a, 64). Fragments
of stone tools were collected from Area 4.
  Two other polychrome anthropomorphic jars are
said to have come from N-P-l  (pis. 58a, 63).
N-P-2: Nueva Armenia
A group of low rolling hills occupies the right bank
of the Rio Napo about halfway between the Rio
Tiputini and the Rio Yasuni (fig. 3). Much of the
area has been cleared of forest and converted into
grassland for cattle raising, making the contour of
the terrain clearly visible (pi. 13). A cluster of houses
and other modern buildings occupies a section along
the river bank where the hills retreat inland leaving
a relatively level area about 4 meters above the low


FIGURE 29.—Sketch map of the Napo Phase site of N-P-2, showing modern occupation, arbitrary divisions used for surface collection,
and the location of excavations.

34

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



water level (pis. \3a-b, 14a). The surface, including
much of the present occupation but continuing
farther to the southeast, shows evidence of habitation
by people of the Napo Phase (fig. 29).
  The habitation area, recognized by the presence
of sherds, is bisected by the eastern arm of a small
creek flowing down from the hills to the west. The
greatest concentration of pottery occurred adjacent
to the river, extending along the bank of the Rio
Napo for approximately 250 meters. Inland extension
varied from 30 to 65 meters. Absence of surface
vegetation around the buildings and the sandy nature
of the soil has facilitated erosion, but damage to the
archeological deposit is much less than what has taken
place at N-P-1. The sandy condition of the soil,
permitting drainage, has also maintained the sherds
in far better condition than at any other site. Sherd
refuse continues on the downriver side of the creek
for at least 250 meters. It may extend farther but the
thick grass and hard soil made verification impossible.
  Depth of accumulation was sufficient on the highest
part of the site, on the west side of the house (fig. 29),
to permit stratigraphic excavation. The 3 by 3 meter
cut was divided into two artificial 15 cm. levels.
The soil was dark brown sandy loam, becoming
darker gray toward the southwest, where sherds were
most abundant. The soft loose composition of the soil
made digging with a trowel possible, and sherds and
other inclusions easy to detect (pi. \4b). At 22-25 cm.
in depth, the soil color changed sharply from grayish
brown of the occupation layer to light yellowish
brown sterile sand. The surface of the sterile deposit
was uneven, especially in the south corner, where
sherds occurred in a pocket 50 cm. in depth. Testing
to a depth of 2 meters showed the deposit to continue
sterile, sandy loam giving way to light gray clay inter-
mixed with red streaks.
The abundance, large size, and excellent surface
condition of the sherds in the vicinity of Cut 1 made
it desirable to increase the size of the sample. This
was done by extending the excavation 3 meters to
the southeast. Conditions were similar to those in
Cut 1, and only extremely small plain body sherds
were discarded. Most of the pottery came from the
upper 15 cm. of the deposit. A pocket 25 cm. in
diameter and 50 cm. in depth was filled with frag-
ments of many different vessels. Material from this
extension of the Cut 1 was designated as Broadside
A and cataloged as a separate unit. A second large
sample, including two small complete vessels, was
obtained from an irregular area about 3 meters in
diameter south of Cut 1 by the women of the house-
hold. For identification, it was labelled Broadside B.
For general surface collection, the site was divided
into two areas. Area 1, from the store eastward to the

creek (fig. 29), presented characteristics similar to
those encountered in the excavation, except that
sherds were sparser. The surface distribution tended
toward a clustering pattern, with areas of relatively
great concentration separated by nearly sterile regions.
Testing showed the depth of the refuse to repeat that
found east of the house. A smaller sample was obtained
from the surface of Area 2, west of the store. Sherds
collected by local children from a subarea within
Area 2 adjacent to the north side of the house, were
kept as a separate classificatory unit (Area 2a).
  A complete anthropomorphic vessel with poly-
chrome decoration (pi. 52) was excavated prior to
our visit from immediately in front of the house. A
second of similar construction (pi. 53) and a restorable
bowl with Rocafuerte Incised decoration on one side
of the rim (pi. 42) have since been collected and
deposited in the Museo Arqueologico del Banco
Central, Quito. Several stone axes (fig. 31 a-c) were
also attributed by local residents to the site.
N-P-3: Nuevo Rocafuerte
The modern village of Nuevo Rocafuerte occupies
the right bank of the Rio Napo about 2 kilometers
above the mouth of the Rio Yasuni and the Ecua-
dorian-Peruvian border (fig. 3). Most of the houses
are lined in a row paralleling the shore (pi. \2b) and
separated from the bank by a double avenue broken
down the center by a series of rectangular grass
plots (fig. 30). The surface of the avenue and around
most of the buildings is kept free of vegetation, and
consequent erosion by annual rains has increased
surface indications of aboriginal occupation. Sherds
are scattered along the bank for more than half a
kilometer, extending both to the east and west of the
modern village. Maximum elevation is at the central
portion of the site, where the vertical bank rises to
about 4 meters above low water level. The western
portion is about a meter lower. Because of irregulari-
ties in the contour of the river bank, inland extension
varies from 50 to 100 meters.
  Random testing of the deposit produced sherds to
a depth of 12 cm. or less, but showed no concentra-
tion except along the bank opposite the easternmost
of the modern buildings. Test A, 2 by 1 meters,
revealed medium gray clay with only a few sherds,
which were added to the general surface collection
for study. Test B, a few meters to the west (fig. 30),
was more rewarding. In this 1.5 meter square area,
sherds were closely packed in black sandy loam to a
depth of 15-30 cm. Extension of Test B to the south-
east showed the dense accumulation to occupy an
ovoid area about 75 by 100 cm. in diameter and 1
  
THE   NAPO   PHASE

35






FIGURE 30.—Sketch map of the Napo Phase site of N-P-3, showing the portion of the site occupied by the modern village of Nuevo
Rocafuerte and the location of test excavations.

meter in depth. The presence of numerous fragments
from a single large painted vessel suggests that the
latter may have been interred with the mouth at
ground level and after its original function had
ceased, it may have been used as a refuse depository.
The cavity was surrounded by the sterile light tan
clay, which cccurred elsewhere in the site at a depth
typically less than 10 cm. below the existing surface.
Although no anthropomorphic urns were encoun-
tered during our visit, residents stated that several
had been found in the past and were taken to the
Jijon y Caamafio Museum in Quito.
N-P-4: Bello Horizonte
For about 3 kilometers upriver from N-P-3, the bank
is a rolling succession of hills and depressions. Sur-
face sherds are scattered throughout this area but
no concentration was observed until N-P-4, where
the highest elevation is attained by the bank (fig. 3;
pi. 15a). Sherds were most abundant adjacent to the
nearly vertical river bank, where three small sum-
mits reach an elevation of about 6 meters above low
water level. A modern house had been built behind
this slight ridge, and the surface of the western two-
thirds of the site had been kept clear of vegetation
since the construction of the house some 30 years
prior to our visit. The owner contended that the
river bank was then much farther to the north.
  Sherds were most abundant in the area of highest
elevation, extending about 100 meters along the
bank and 20 meters inland. Testing of spots with
greatest concentration of surface sherds produced
little or nothing. Soil was sandy, slightly grayer in

the refuse deposit than below. Sherds were typically
restricted to the upper 5 cm., occasionally occurring
to a depth of 10 cm. The sample was collected mainly
from the 90 by 40 meter area east and north of the
house, where the surface was free from debris and
vegetation.
  The owner reported that an anthropomorphic urn
with modeled legs was found at the site some years
before and sold to a passing American.
N-P-5: Florencia
A cluster of high hills about half way between N-P-2
and N-P-4 embraces a small level area (fig. 3; pi.
\5b). A creek flowing northward near the western
edge serves as the source of fresh water to the modern
residents during the dry season, when the Rio Napo
channel retreats northward behind a 500-800 meter
wide sandbar. The entire area has been cleared of
forest and is covered with pasture, resembling the
situation at N-P-2. Remains of Napo Phase habita-
tion were traced over an area about 75 meters along
the bank by 30 meters inland, where the elevation
was 2 to 4 meters above low water level. The refuse
deposit was very thin, rarely exceeding 5 cm. The
soil over most of the site consisted of red-orange sandy
clay with occasional spots of grayish hue, representing
remnants of the habitation layer.
N-P-6: Puerto Alfaro
A relatively small Napo Phase habitation site is located
on the high left bank of the Rio Tiputini about 100
meters upstream from the Tivacundo Phase site of

36

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



N-P-8 (fig. 15). Sherds occur over an area about 150
meters long by 50 meters inland. Most of the surface
lies within the dooryard of a modern house, which is
kept free of vegetation (pi. 16a). The opposite bank is
low (pi. 166). The light tan soil bakes hard in dry
weather and is reduced to mud during the rainy
season, with detrimental effects on the pottery. Testing
showed sherds to be restricted to the surface. Several
stone tools were also collected (fig. 31/).
N-P-9: Cotacocha
About 8 kilometers above N-P-l (fig. 3), the left
bank of the Rio Napo attains an elevation of 7 meters
above low water. A modern house occupies the eastern
quarter of a cleared area extending about 25 meters
along the bank and a slightly greater distance inland.
Beyond its limits, the ground was obscured by grass
and banana trees. Within the cleared area, the soil
was light tan to brown, hard-baked clay. Sherds were
scattered on the surface and extended to a depth of
5 cm. The majority were classified as belonging to the
Cotacocha Phase, but 26 were of Napo Phase origin,
only one of which was decorated.
N-P-l 3: Pafiacocha
A fragmentary Rocafuerte Painted (red-on-white)
vessel of Form 6 was collected by one of the soldiers
from Pafiacocha, said to be near the mouth of the
Rio Cuyabeno, a tributary on the left bank of the
Rio Aguarico (fig. 3). The paste is sand tempered.
Data From Other Investigations
The exotic and often beautifully painted anthro-
pomorphic burial urns of the Napo Phase have found
their way into a remarkably large number of mu-
seums, considering the remoteness of the area of their
origin. Unfortunately, few have exact provenience
identification. Where a location is specified it is
sometimes unidentifiable because the name is no
longer used or distances cannot be computed with
confidence on a modern map. In spite of these
drawbacks, the data enlarge the geographical distri-
bution of Napo Phase pottery.
Rio Aguarico
Three complete vessels with characteristic Napo
Phase form and decoration have been found on the
Rio Aguarico. An Armenia White-on-red jar of
Form 17 (pi. 20c), in the Peabody Museum, Harvard
University, was collected by John Gillin (1936,
p. 469) who gives the following description of its
provenience:

The specimen was presented to me by two Aguarico Indians
who happened to share our camping place .... I went next
morning to the spot where the specimen was alleged to have
been found. A hole was found in the side of a mud bank with
evidences of considerable digging by sticks and cutlasses. The
writer is convinced that the urn was found by the Indians in
situ, four feet six inches deep in alluvial, gray clay. The surface
of the ground was covered to a depth of about nine inches with
vegetable mould and a matting of roots from the forest growing
there . . We had no tools for digging, but a day spent in
working with cutlasses revealed several sherds of gray, un-
derrated pottery and a hearth of charcoal at a depth of five
feet nine inches from the surface. The creek in question, which
is nameless, is a right tributary of the Aguarico which joins the
latter an estimated forty-five miles above its mouth. The site is
located on the left bank of the creek about one and a half miles
above its mouth.
An unnamed tributary of the Rio Aguarico in the
approximate location specified by Gillin lies almost
due north of N-P-9. Overland distance from the
Rio Napo here is only about 20 kilometers (fig. 3).
  The second vessel, a Rocafuerte Painted (red and
black-on-white) jar of Form 6 with an anthropomor-
phic face on a rim lobe (pi. 46), was acquired during
our survey from a local resident who said it had
been found on the right bank of the Rio Aguarico a
short distance above the mouth of the Rio Cuyabeno
(fig. 3). The paste is sand tempered. No associated
material was observed.
  An anthropomorphic urn of type B, with the orifice
in the base, is said by Alan C. Lapiner to have come
from the Rio Aguarico at the mouth of the Rio Eno,
located approximately at 76 degrees 25 minutes west
longitude (fig. 3), which places it farther up river than
other documented finds of Napo Phase ceramics.
The vessel has both polychrome and grooved decora-
tion, and painted designs surround both eyes (pi.
62b). Details are given on Table A.
Eden
A Napo Plain Incised bowl (USNM Archeology
No. 385647), a side-notched stone ax (USNM Arche-
ology No. 385649; fig. 3\d) and a T-shaped stone ax
(USNM Archeology No. 385648; fig. 3\e) are attrib-
uted to the vicinity of Eden. Recent maps show
Eden on the right bank of the Rio Napo, above the
mouth of the Rio Dumbique, or about 80 kilometers
upstream from N-P-l (fig. 3).
Rio Indillama
A Rocafuerte Painted (black-on-white) jar of Form 17
(pi. 54b) is said to have come from the right bank of
the Rio Napo a littie below the mouth of the Rio
Indillama (fig. 3).

THE   NAPO   PHASE

37



Oasis
A collection of 40 sherds, 37 of them decorated, in the
Museum of the American Indian, Heye Foundation,
New York, are from Oasis on the right bank of the
Rio Napo, a few kilometers upstream from Eden (fig.
3). The range of decorated techniques and vessel
shapes is comparable to that at other Napo Phase
habitation sites. Although selected for decorated
sherds, the attempt has been made to seriate this
collection in the Napo Phase sequence (figs. 61-63)
after correcting it by allowing for a 61.8 percent
frequency of plain sherds, which is the average figure
for unselected Napo Phase samples. (Appendix table
8.)
Providencia
An excised bowl (pi. 32b) was found by Howard S.
Strouth near Providencia on the left bank of the Rio
Napo (fig. 3). Diameter is 28 cm.
Rumi-tuni
Two pottery vessels of Napo Phase types in the Amer-
ican Museum of Natural History, New York (AMNH,
41.1/3985 and 41.0/9141), are attributed to "Rumi-
tuni", located on the right bank, 315 kilometers above
the mouth of the Rio Napo. Measurement of this
distance on the most accurate map available places it
in the vicinity of the mouth of the Rio Curaray, in
what is now Peruvian territory (fig. 3). The specimens
consist of a Napo Red Incised bowl of Form 5 and a
Rocafuerte Painted (red and black-on-white) semi-
cylindrical anthropomorphic jar lacking the head (pi.
65).
Rio Yasuni
A Napo Phase site has been reported by Pedro I.
Porras (personal communication) near Puerto Vargas,
on the left bank of the Rio Yasuni about 4 kilometers
from the mouth (fig. 3). The bank here rises 3 meters
above low water level. A 2 by 2 meter stratigraphic
excavation showed the Napo Phase refuse to be over-
lain by 20 cm. of humus, which, contained fragments
of a modern painted vessel. Below 20 cm., Napo Phase
sherds were sparse and neither of the two 10 cm.
levels produced a sample of 100 sherds. Sterile clay
appeared at about 37 cm.
  
Classification of the sherd sample by temper resulted
in 93.6 percent sand, 3.8 percent cariape, and 2.6
percent charcoal, a combination that seriates between
N-P-l, Area 4, and Oasis, in the upper third of the
sequence (fig. 60). Tabulation of single-line and
double-line incision shows 45 percent frequency of
single-line and 55 percent frequency of double-line
technique, equivalent to that of N-P-3, Surface and
Test A, which seriates early (fig. 61). The inconsistency
of these results may stem from insufficient sample size.
The presence of all decorated types except Napo
Negative and Tiputini White Excised favors the earlier
position, since many are absent from the late portion
of the sequence (fig. 63).
  A large anthropomorphic figure is reported to have
been found about 80 meters upstream from the loca-
tion of the excavation. The figure is seated, with
abbreviated legs, a cylindrical body expanding slightly
at rounded shoulders, a constructed neck and globular
head. Ears, eyes, nose and mouth are in relief; the
mouth is shown open. Treatment is not like that of
any of the other anthropomorphic urns, but wide
variation in rendition of anthropomorphic features is
characteristic of Napo Phase art.
Unspecified Location on Rio Napo
An anthropomorphic urn in the possession of a Rio
Napo resident in 1956 (now owned by Thomas Flan-
nery) is of interest because of the circumstances re-
ported by the finder concerning its discovery. The
vessel is of type B with the orifice at the bottom, and
has arms and legs in full round (see Table A for de-
tails). It was encountered during excavation for a
house post, buried upside down. The opening was
covered with a lid of unspecified type. The arms and
legs had been broken off and placed inside. No
remnants of bones or teeth were reported.
  Other vessels are illustrated in various publications
or preserved in museum collections. They are either
attributed because of their style to the Rio Napo or
have only this generalized provenience identification.
Anthropomorphic examples are described on Table A.
Rocafuerte Painted vessels illustrated by Jijon y
Caamano (1951) include jars of Forms 16 and 17, the
latter with a bowl-shaped lid (op. cit., figs. 497-8,
502) and a basin of Rare Form 5 (op. cit., fig. 503).
  
38

SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY
ANALYSIS OF MATERIALS

VOLUME   6



Stone Artifacts
The soil along the Rio Napo is devoid of large
stones at the longitude of the Napo Phase sites.
However, above the mouth of the Rio Suno (fig. 1),
bars of rounded cobbles and pebbles provide a
ready source of this raw material. Both shaped and
polished tools and unshaped rock fragments showing

use were collected from most of the Napo  Phase
sites. Proveniences are indicated on Appendix table 7.
Abraders
Five small stones from various sites (Appendix table 7)
have facets worn on parts of the surface from use as
abraders. Rock materials are fine sandstone or pumice.
One quartz pebble appears to be polished from use.














0    1     2    3    4    5  CM







,, . *\









FIGURE 31.—Napo Phase stone axes,   a-d, Side notched,   e-f, T-shaped.

THE   NAPO   PHASE

39








/■■£
Form is irregular and size varies from 5.0 by 3.5 by
3.0 cm. to 3.5 by 2.5 by 1.0 cm.
Axes
FIGURE 31
The majority of the stone axes, represented by four
complete and two broken examples with the blades
missing, are side-notched, fashioned from granite,
andesite, greenstone, and slightly metamorphosed
tuffaceous sandstone. The butt is nearly flat, and
sides tend to be generally parallel, tapering gradually
toward the convex blade. One example (fig. 3\d) is
atypical in form, the sides diverging below the notch
to an expanded blade. Notches are shallow, with a
maximum depth of 3-4 mm. except for one unusually
large sample where depth reaches 7 mm. The notched
area is even but not polished, and pecking may
extend inward along the surface to produce a slight
depression (fig. 31a). Surface polish may be nearly
complete, leaving a few pits and flaws especially
near the butt. More typically, however, polishing is
best adjacent to the blade and may also occur at
both edges or on the face. The butt edge is battered
principally from use, but to some extent from shaping
by pecking without subsequent polishing.
  Vertical sections show the two faces to be parallel
except on one example, flattened at the butt, and
curving to a sharp edge at the blade. The horizontal
section varies from the butt, which tends to be rec-
tanguloid, to the region immediately below the
notches, where the edges become rounded. About
half the distance between the notches and the blade,
the cross-section becomes ovoid with blunt to sharp
ends.
  Length of complete examples varies from 9.5 to
11.5 cm., with one fragment of considerably larger
size. Width at the butt is 6.4 to 7.4 cm., at the blade
about 5 cm., except for one example in which the
butt is 4.5 cm. wide and the blade 5.7 cm. Maximum
thickness is 2.3 to 2.7 cm., except for the above
mentioned unusually large fragment with a thickness
of 3.7 cm.
  Two axes have a T-shaped or "eared" form, but
one has parallel sides (fig. 31/) while the other expands
toward the blade (fig. 31*). The former is tuffaceous
sandstone, polished adjacent to the blade and on the
end of the butt, with the remainder of the surface
evenly pecked and unpolished. Thickness increases
from 1.3 cm. at the flat butt to 2.8 cm. toward the
blade. The sides are slightly flattened. Maximum
width at the butt is 10.8 cm., width below the ears is
8 cm. Length is 9.5 cm. The second example, fashioned
from silicified volcanic rock, has a completely polished
surface leaving two small flaws adjacent to the butt.

FIGURE   32.—Napo

Phase   stone   tools.
b, Chisel.

J
3   CM
a,    Grinding    stone.

40

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



The blade shows slight battering from use. Width at
the ears is 4.6 cm., maximum width 5.6 cm.; total
length is 10.3 cm. Thickness increases slightly from
1.3 cm. at the flat butt to 1.8 cm. toward the blade.
Chisel
                            FIGURE 32b
The only tool in this category comes from the surface
at N-P-6. It is fine-grained slate, incompletely shaped
from an elongated waterworn pebble, leaving the
original contour unmodified at the naturally rounded
butt end. The sides are slighdy convex, and the faces
taper toward a well sharpened convex blade. As a
result, the cross-section is nearly circular at the middle,
becoming almond-shaped at the blade. The surface
is even and polished to a slick finish except adjacent
to the butt. Total length is 12.2 cm., maximum
diameter 2.5 cm. at the center, blade width 1.7 cm.
Cores
The only two examples are from N-P-l. Both are
chert. Dimensions are 3.5 by 2.5 by 1.8 cm. and 2 by
2 by 1 cm.
Grinding Stones
FIGURE 32a
Two fragments, both from N-P-l, Area 4, show use
polish on one or more surfaces. One is a thick fragment
of dark grayish green andesite that has been altered
to reddish tan on the worn surfaces. The fragment
is at present a rectanguloid block (fig. 32a). One
surface is slightly convex and the other slightly concave
when viewed in the long axis. In cross-section, the
former is slightly convex and the latter markedly so.
Projection of the contours of the most strongly curved
faces suggests that the original form may have re-
sembled that of a large mano. Thickness ranges from
4.2 to 5.5 cm., existing length is about 11 cm.
  The other fragment is also grayish green andesite.
Asymmetrical use polish occurs on the end, extending
about 3 cm. backward on one face and 2 cm. on the
other. The remaining surface is roughly shaped by
pecking. Cross-section is ovoid. Existing length is
10 cm., thickness 3.5 cm., and reconstructed width
about 6 cm.
Hammerstones
FIGURE 33
Hammerstones may be divided into two categories:
those that were intentionally shaped and natural
rocks that show traces of use. The first category in-
cludes two specimens, one disk-shaped and the other
rectanguloid.
  The disk-shaped hammerstone (fig. 33b), from the
surface of N-P-2, is grayish tan andesite with black
pyroxene crystals. It has an even, but not polished

surface, with numerous shallow to deep pits. Shape
is not perfectly circular, so that diameter varies
from 7.3 to 7.5 cm., with a maximum thickness of 2 cm.
The faces are parallel and slighdy convex, tapering
to rounded edges.
  The rectanguloid example (fig. 33a) is grayish
white granite. It comes from N-P-l, Area 3. Both
ends and one side are severely battered. The outline
resembles that of the notched axes, with flat butt,
parallel sides, and a convex lower end, suggesting
that it may be an unfinished ax. Cross-section is
ovoid with rounded edges. Most of one face and half
of the other are even but not smooth to the touch,
with deep pits and irregularities unobliterated.
Length is 11.5 cm., width 5.6 cm. and thickness
2.1 cm. The upper portion of a similar object has an
even surface and a considerably battered butt. The
rock is dark gray porphyry. Provenience is N-P-l,
Area 4. Width is 7.1 cm., maximum thickness 2.9 cm.,
existing length 6.3 cm.
  Eleven natural stones show traces of battering on
one or more edges. Rock materials represented are
andesite, quartz, quartzite, porphyry, and sandstone.
The majority are fragments of waterworn pebbles
with existing dimensions between 7.5 by 7.5 by 2.5 cm.
and 4.5 by 3.0 by 1.5 cm. Their condition suggests
temporary use and there is no evidence of intentional
shaping.
  Two pebbles, both from N-P-l, Area 4, show
evidence of more extended use (fig. 33 c, d). One is
quartz, cylindroid with rounded ends, one of which is
severely worn. A smaller area of similar wear occurs
on the other end. Length is 5.8 cm., width 4.8 cm.,
and thickness 4.2 cm. The other is a naturally triangu-
loid andesite pebble with one flat and one convex
face. The effects of battering are visible in two areas
along the edge and at the center of the flat face.
Diameter is 6.5 by 7.0 cm., maximum thickness
4.3 cm.
Natural Pebbles
Complete or fragmentary pebbles of chert, quartz,
sandstone, quartzite, and granite were collected from
most sites. Since these types of stone were all rep-
resented in objects showing use, the presence of
unworked stones may be of cultural origin. Size
varies between 6.5 by 4.0 by 1.8 cm. and 4.0 by
2.8 by 22 cm.
Paint Stone
An irregularly shaped hunk of limonite from N-P-3,
Test B Extension shows striations from rubbing.
Dimensions are 6.5 by 5.5 by 3.0 cm.
                               Spalls
Andesite, quartzite, and sandstone spalls from water-
                         
THE   NAPO   PHASE

41



worn pebbles or tools with worked surfaces (axes?)
appear to be by-products of manufacture or use of
stones for pounding. None show evidence of use in
their present condition.
Pottery Artifacts
All Napo Phase pottery artifacts are objects of pri-
mary manufacture. No sherds were observed to show
sierns of use or reworkiner.
signs of use or reworking.

Disk
One fragment of a small pottery disk may represent
a spindle whorl. The paste is fine sand tempered
and fired tannish orange. Surfaces are even and
smooth but not polished or floated sufficiently to
obliterate temper particles. Outline is not perfectly
circular, and one surface is flat while the other is
slightly convex. Diameter is about 3 cm., thickness
at the center 3.5 mm. Perforation diameter 3 mm.
Provenience is N-P-2, Broadside B.








4     5  CM








FIGURE 33.—Napo Phase hammerstones.   a-b, Shaped stones,    c-d, Natural pebbles.

42

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6





0       I        2      3 CM
FIGURE 34.—Decorated potrest from the Napo Phase.

Potrests
FIGURE 34; PLATE 17
More than 200 fragments of potrests were collected,
representing all of the Napo Phase sites. A number
of upper or lower ends are included, but the majority
are irregularly shaped body fragments, 2-6 cm. in
diameter. Form is not symmetrical, making diameter
measurement of fragments unreliable. The majority
of diameter measurements ranged from 8-12 cm. at
the upper end and 10-18 cm. at the base. The only
example complete enough to provide both height
and diameter was 9 cm. in diameter and 13 cm. tall.
Another measured 15 cm. with part of the height
missing.
  Most of the fragments are made of untempered
clay. A minority have sandy paste, perhaps being
made of clay left over from vessel manufacture.
Color is light orange to pinkish orange, or occasion-
ally pale tan or grayish. A layer 2 to 10 mm. wide
adjacent to the surface may be paler buff or orange
than the core. Texture is laminated with abundant
air pockets. Characteristic fracture into irregular
hunks (pi. 17 c, e) suggests construction by aggrega-
tion of lumps of clay until the desired volume was
achieved. A number of surface fragments have ir-
regular inner walls suggesting that the interior either
was hollow or was never sufficiendy well fired to
be preserved under the destructive action of weather-
ing. Others were clearly of solid construction.
  Surface is brown, tan or rarely pale orange, often
with small irregularly shaped fire clouds. Treatment
varies from smoothed to produce an even surface to
irregular and uneven, especially on the ends. Hor-
izontal or vertical smoothing marks may occur, made
by the finger or possibly a pebble tool.
  Only one example is decorated on the upper end
with incision and excision (fig. 34). It came from
N-P-2, Broadside A.
Roller Stamps
FIGURE 35
Although no roller stamps or fragments were col-
lected at any of the Napo Phase sites included in the
survey, one example in the Museo Victor Emilio
Estrada, Guayaquil, is attibuted to the Rio Napo. The
roller surface bears a deeply excised pattern resembling
that on pottery vessels, suggesting that the artifact is
of Napo Phase origin. The design area consumes less
than half the total length, the remainder being divided
between two tapering handles that give the object
a rolling pin form. Total length is 15 cm., length of
the design portion 6 cm., and maximum diameter
about 6.6 cm.

THE   NAPO   PHASE

43




i—i	i	i
0    I     2    3 CM
FIGURE   35.—Roller   stamp   probably   of  Napo   Phase   origin.
Ceramic Classification
The pottery types of the Napo Phase are based on
classification of 13,896 sherds from 8 habitation sites,
supplemented by a number of complete vessels in
private collections and museums. Decoration is more
frequent than in the other archeological phases on
the Rio Napo, varying between 34 to 47 percent of
the sherds where the sample seemed adequate.
  Although Napo Phase pottery frequendy exhibits
surface erosion, preservation is much better than in the
Tivacundo and Yasuni Phases. Surfaces are often intact,
showing traces of smoothing and in some cases pol-
ishing, but rarely achieving a slick or lustrous finish.
Both red and white slips occur in association with
painting, incision and excision, and more complicated
treatments involve the application of red or white
pigment to incised lines. In a few cases, more than
one technique was applied to a single vessel, usually
a combination of incised or excised exteriors with
painted interiors.
  The undecorated pottery was classified on the basis
of temper differences into three plain types: cariape-
tempered Armenia Plain, sand-tempered Napo Plain,
and charcoal-tempered Tiputini Plain. An unusually
large number of decorated types—19 in all—has
been established after much deliberation. Incised
and excised designs occur in two varieties, one done
with a single pointed tool and the other with a double-
or multiple-pointed tool. Each technique occurs with
plain, red and white slipped surfaces, so that sub-
division of the incised types by technique doubles
the number of types. Differences in vessel shape and
temporal position presented by the single-line and
double-line techniques in the Napo Phase ceramic
complex are sufficient to warrant their separation
into distinct types, and this procedure is further
justified by the differences in geographical distribution

of the two techniques in the Amazon Basin. Other
characteristic Napo Phase decorated techniques are
red-on-white, black-on-white, and red and black-
on-white painting, white-on-red painting, nega-
tive painting, and red slipping without further
embellishment.
  Twenty-one common shapes have been recognized
for Napo Phase pottery, nearly all represented by
one or more complete vessels. The outstanding
characteristic is avoidance of circular form. The
majority of all vessels are squarish when viewed from
above, and many have concave sides and projecting
corners. Even the most nearly circular examples often
have slightly flattened sides, although truly circular
oudine also occurs. Annular bases, channeled and
cambered rims, and several kinds of anthropomorphic
treatment are additional features setting Napo Phase
pottery apart from that of other complexes in the
area.
  Since most rim sherds did not provide measurable
diameters, vessel shapes have been drawn to the
size and proportions of a complete specimen, in
contrast to the procedure followed for illustrations
of pottery types in other phases. In several cases,
the same rim profile is associated with two distinct
body shapes; where this is known to occur, both
forms have been shown. Rim profiles have been
selected to illustrate both the range of variation
exhibited by a form and the association of the form
with plain and decorated pottery types. Plain rims
are drawn in all figures in solid black, while red
slipped ones are hachured; decorated (incised, excised,
painted) rims are shown in outline. Decorated portions
of the surface are usually identified by arrows or
brackets.
  Pottery type descriptions have been arranged in
alphabetical order following the description of vessel
and base forms. Provenience and frequency of both
pottery types and vessel shapes can be found in
Appendix tables 8 and 10.
Reconstructed Vessel Forms
Common Forms
1.    Open bowl with exteriorly thickened rim (fig. 36-1):
Rim: Upcurving or incurving and thickened on the
exterior to produce an elevated band 2-4 (rarely, 5)
cm. wide adjacent to the lip. Thickness is 2-3 mm.
greater at the center than the body wall, and typi-
cally decreases slightly both above and below. Non-
circular.
Lip:    Flattened; very rarely, rounded.
Body wall thickness: Range 3-7 mm.; majority 3-5
mm.
Base:    Rounded or slightly flattened (Form C).

44

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6
CA




I   ■   I   i    I   i   I
0       I       2      3 CM
RIM   SCALE





VESSEL  SCALE



FIGURE   36.—Rim  profiles  and  reconstructed  vessel  shapes  of the  Napo  Phase,   Common   Forms   1—4.   Arrows  and  brackets
designate decorated zone.

THE   NAPO  PHASE

45



Deep bowl with outslanting wall (fig. 36-2):
Rim: Outslanting or slightly everted, turning upward
1.5-2.5 cm. below lip often assuming a nearly verti-
cal orientation. Upturned section is convex on the
exterior, either because of thickening or correspond-
ing concavity on the interior. Lower edge of up-
turned section may blend into the wall or change
direction abrupdy. Noncircular.
Lip:   Tapered and flattened.
Body wall thickness: Range 4.0-9.5 mm.; variable on
a single vessel.

Base:   Annular (Form D), or flattened (Form A).
Bowl with curved cambered rim (fig. 36-3):
Rim: Outcurving wall, concave on exterior, turns
upward producing a convex exterior contour 3.0-5.5
cm. below the lip. This curved camber typically
results from gradual change in direction and is
rarely accompanied by thickening of the wall. Non-
circular.
Lip:   Flattened.
Body wall thickness:   Range 4.5-7.0 mm.
Base:   Flattened (Form A).






0     l     2
RIM   SCALE

FIGURE 37.—Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 5-6.   Brackets designate decorated zone.

46

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6




0      I      2      3 CM
RIM   SCALE
VESSEL SCALE




FIGURE 38.—Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 7-9.   Arrows and brackets designate
decorated zones.

THE   NAPO   PHASE

47



4.	Open bowl with "channel" rim (fig. 36-4):
Rim: Curvature of the body wall is markedly in-
creased 2-3 mm. below the lip, producing an effect
on the exterior similar to that of Form 1 both in
contour and orientation. This is accomplished with-
out thickening, with the result that a channel or
marked concavity is produced on the corresponding
portion of the interior. Slight thickening of the wall
increases the abruptness of the change of direction
on the interior, and it typically appears as a sharp
ridge. Change of direction on the exterior may be
gradual or abrupt and step-like. Noncircular.
Lip:   Flattened.
Body wall thickness: Range 4-6 mm.; thickened to
6-9 mm. at lower border of channel.
Base:   Flattened (Form A), or rounded (Form C).
5.	Open bowl widi broad "channel" rim (fig. 37-5):
Rim:   Construction   similar   to   Form   4,   but   with
  with channel width 3-5 cm. Noncircular.
Lip:   Flattened.
Body wall thickness:   Range 4—6  mm.;  thickened  to
  6-10 mm. at lower border of channel.
Base:   Flattened (Form A), or rounded (Form C).
6.	Large deep bowl or jar with curved cambered rim
(fig. 37-6):
Rim:    Construction similar to Form 3, but with change
in  curvature  beginning  9-11   cm.   below  the  lip.
Change in direction on the exterior may be con-
tinuous and gradual or abrupt, the latter condition
produced by slight thickening. Diameter may reach
50 cm.
Lip:   Flattened.
Body wall thickness:   Range 9-13 mm.
Base:   Flat (Form A, B, or E).
7.	Bowl  or jar with everted,  exteriorly thickened  rim
(fig. 38-7):
Rim: Slighdy to strongly everted and thickened
abrupdy 1-3 cm. below the lip, tapering from a
maximum thickness at this point toward the lip.
Lower edge of thickening is smoothed over typically
producing a concave zone of junction on the ex-
terior. The interior surface continues the curvature
of the wall eversion. Thickened band is generally
convex on the exterior surface; rarely, it is concave.
Eversion is frequendy greater at corners than at
sides.
Lip:   Flattened.
Body wall thickness: Range 4—9 mm.; lower edge of
exterior rim thickening, 6.5-12.0 mm.; thickness
at carination 1.0-2.7 cm.
Body form: Two distinct shoulder forms occur with
this rim construction (also with Forms 8, 9, and 10).
A.	Carinated and typically thickened to increase
the prominence of the angle (fig. 38, bottom).
Thickening is often greater at the corners than
at the sides; when this occurs, the sides are con-
cave when the vessel is seen from above. Corners
vary from rounded to projecting (fig. 39).
B.	Rounded and unthickened or slightly thickened.

Noncircular form when viewed from above is
much less pronounced, but flattening of the sides
is always perceptible.
Base:   Rounded (Form C).
8.    Bowl or jar with everted, exteriorly thickened tapered
rim (fig. 38-8):
Rim:    Everted and exteriorly thickened  1-3 cm. be-
low the lip, tapering from that point to the lip.
Curvature of inner wall and of convex surface of
tapering outer wall often meet at the midline to
form a symmetrical cross-section. Noncircular.

I	L
0       I      2      3      4     5 CM
FIGURE 39.—Shoulder corner contour of carinated vessels of Napo
Phase Common Forms 7-10, showing decorative notching.
Lip:    Tapering and slighdy rounded.
Body wall thickness:    Range 4—6 mm.
Body form:    See Form 7.
     Base:    Rounded (Form C).
9.    Bowl or jar with everted "channel" rim (fig. 38-9):
Rim: Everted and bent abrupdy 1.2-2.5 cm. below
the lip creating an external appearance similar to
that of Form 7 but with a clearly defined channel
on the interior produced either by a combination of
thickening at the lower edge and thinning toward
the lip, or by allowing the inner wall contour to
parallel that of the outer wall. Noncircular.

293-822 O - 6B

48

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6




'■■•'■'
0      I      2      3 CM
RIM   SCALE
VESSEL   SCALE




FIGURE 40.—Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 10-13. Arrows designate decorated
(Black = undecora ted, white=decora ted, hachure=red slipped.)

THE   NAPO   PHASE

49



Lip:    Flat.
Body wall thickness:    Range 4—9 mm.; majority 4-7
  mm.
Body form:    See Form 7.
Base:    Rounded (Form C).
10.	Bowl or jar with cambered rim (fig. 40-10) :
Rim: Cambered, curved, or angular on the exterior
but typically angular on the interior, frequentiy
changing direction at an angle approaching 90
degrees. This form is distinguished from the
"channel" rim of Form 9 both by its greater length
(height) and by greater uniformity of thickness of
the cross-section, although there is a continuum from
one to the other. Noncircular.
Lip:    Flat.
Body wall thickness: Range 2-9 mm.; majority 4-6
mm.
Body form:    See Form 7.
Base:    Rounded (Form C).
11.	Constricted mouth jar (fig. 40-11, pis. 52, 53).
Rim:    Direct and incurving.
Lip:    Flat or rounded.
Body wall thickness:    Range 6-13 mm.
Body form:    Some complete vessels have a strongly
  curved shoulder, others may be more globular.
Base:    Flat (Form B).
12.	Shallow bowl with direct rim (fig. 40-12):
Rim: The outslanting wall curves upward about 3-5
cm. below the lip, depending on the diameter of
die bowl. Change of direction is gradual but
sufficient so that the rim orientation approaches
vertical and the lip is typically horizontal. Rim
form is usually direct but rarely expanded or
tapered. Diameter 14—42 cm.
Lip:    Flat; rarely rounded.
Body wall thickness:    Range 4—9 mm.
Base:    Flat (Form B) or rounded (Form C).
13.	Large deep bowl with direct rim (fig. 40-13):
Rim: Direct, continuing the outward slope of the
body wall or curving upward about 5-9 cm. below
the lip to assume a vertical or slighdy incurving
orientation. Change of direction is typically gradual
but occasionally perceptible on the exterior as a
rounded angle. Diameter 28^6 cm.
Lip:   Flat or rarely rounded.
Body wall thickness:   Range 6-12 mm.
Base:   Flat (Form A) or rounded (Form C).
14.	Large  shallow   basin   (fig.   41-14;   frontispiece;   pis.
42, 51a).
Rim: Strongly outsloping and slightly everted to
produce a convex interior contour, sometimes
enhanced by slight thickening of the wall. Non-
circular.
Lip: Flat and usually lobed or ornamented with an
elaborate pattern of steps and notches.
Body wall thickness:   Range 5-13 mm.
Base:   Rounded (Form C).
15.	Flange rim jar (fig. 41-15):
Rim:   Everted   at   an   angle   of   approximately   90

degrees and thickened to produce an angular contour
on the interior and a curved one on the exterior.
The flattened or slighdy convex top 1.3-1.5 cm.
wide, is usually level, but occasionally insloping
or outsloping. Interior diameter of two specimens
measures 20-24 cm., but others are probably
noncircular.
Lip:   Tapering or flattened.
Body wall thickness:   Range 3-6 mm.
Base:   Probably rounded (Form C).
16.	Large collared jar (fig. 41-16):
Rim: The effect of a broad collar is produced either
by thickening of the wall or by alteration of curvature
or both. Collar height ranges from 4.5-14.5 cm.,
with the majority about 9-12 cm. high. Thickening
is usually 2-4 mm. greater than the lower body wall.
The collar surface is convex on the exterior, and
the upper portion is slighdy incurved. Diameters
of 36-42 cm. were measured, but the majority are
probably noncircular.
Lip:   Flat, rarely curved.
Body wall thickness: Range 4-10 mm.; majority
4—6 mm.
Base:   Probably flat (Form A).
17.	Semicylindrical jar (fig. 42-17; pi. 54):
Rim: Slighdy everted, vertical or insloping; wall
thickness may remain uniform, increase or decrease
in the upper 3 cm. Diameter 17—48 cm.
Lip:   Flat and generally horizontal.
Body wall thickness:   Range 7-13 mm.
Body form: Range of variation from a tall jar to a
broader, more open vessel results from different
combinations of diameter and height. A pronounced
to gendy rounded shoulder occurs at the junction
between the outflaring lower wall and the more
vertical upper one, and diameter is equal or greater
than that of the rim.
Base:   Flat (Form A) or rounded (Form C).
18.	Globular jar with constricted neck (fig. 42-18):
Rim: Everted to produce an outsloping neck. Junc-
tion with incurving body wall may be curved or
slightly angular. One example is thickened on die
exterior 5 cm. below the lip, producing a collar-
like band.
Lip:    Flat.
Body wall thickness:    Range 4—6 mm.
Base:    Probably rounded (Form C).
19.	Constricted mouth jar with exteriorly thickened rim
  (fig. 42-19):
Rim:    Incurving wall is upturned and thickened 1.0-
1.5 cm. below the lip producing a narrow "beaded"
  edge. Interior diameter 28-46 cm.
Lip:    Rounded.
Body wall thickness:    Range 6-12 mm.
Base:    Probably flattened (Form A).
20.	Large shallow basin (fig. 43-20):
Rim: Upslanting and direct or narrowing slightly
beginning 2 cm. below the lip. Diameter of one
example, 42 cm.

50

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6




RIM   SCALE





l   I   I   I   I   i   I
0     4      8     12   CM
VESSEL   SCALE


FIGURE 41.—Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 14-16. Arrows designate decorated zones.
(Black = undecorated, white=decorated.)

THE   NAPO  PHASE
n

51





i i i i i i i
  0     4     8    12 CM
VESSEL    SCALE
I   i   I   .   I   .   I
0      I      2     3 CM
RIM   SCALE


■AStmSgSF1}M7m./>£■■ 'H18y /^£y.:■Z ,,;.;.:-. -*;-0.'.y      ."^y^- ."  ■■:Oi-~
&M7MFIGURE 42.—Rim profiles and reconstructed vessel shapes of the Napo Phase, Common Forms 17-19. Arrows and brackets designate
decorated zone. (Black=undecorated, white=decorated, hachure = red slipped.)

52

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Lip:    Rounded or slighdy flattened; one example is
  ornamented with nubbins.
Body wall thickness:    Range 1.2-2.0 mm.
Base:    Flat (Form A).
■S^ftw&to&ftfo&n ■■■— >M	OOiM
0      4      8      12 CM
VESSEL    SCALE
	
0       I       2      3 CM
RIM   SCALE
20

FIGURE 43.—Rim profiles and reconstructed vessel shapes of the
Napo   Phase,   Common   Form   20.    (Black=undecora ted.)
21.   Anthropomorphic vessels (table A; pis. 55-65):
Three general types of anthropomorphic vessels (not
including common forms occasionally receiving minor
anthropomorphic treatment) can be distinguished on
the basis of location of the orifice. Elevation of relief,
style of facial features, and degree of realism or
stylization seem to be independent variables.
A.	Orifice at top of head (pis. 55-56).
The face occupies one side of a collared rim (cf.
Form 16); body is cylindroid or rounded, bottom
flat or slightly concave. Limbs project free of the
body or are partly in relief. Height of one ex-
ample is 38 cm.
B.	Orifice at the bottom (pis. 57-61, 62b).
The face occupies one side of a hemispherical
head, separated from the rounded body by a
narrow constriction. The lower edges of the body
wall curve inward to a direct rim with flattened
lip leaving the central bottom open. Limbs may
be absent, shown as relief, or project free. Height
of two examples is 23.5 and 48.0 cm.
C.	Orifice at neck, removable head (pis. 62a, 63-65).
No heads are represented. The rounded body
wall curves inward at the neck and is cut back
about 2 cm. below the lip producing a channel

into which the lid forming the head would fit.
Limbs are shown as discontinuous relief or project
free; legs may be vestigial. Seated and kneeling
positions are shown. Heights of three examples
are 15.0, 34.5, and 43.0 cm. from base to rim at
neck.
Rare Forms
1. Rounded jar widi constricted mouth and everted rim,
tibickened at the angle of eversion and tapering to a
rounded lip. Mouth diameter 14-28 cm. (fig. 44-a).
2. Griddle widi slighdy raised and expanded rim, rounded
lip. Diameter 40 cm. (fig. 44-6).
3. Open bowl with flaring rim thickened on the exterior
about 11 cm. below the lip, from 2-4 mm. greater than
die diickness of the body wall, forming a raised band.
Flat lip. Rim diameter 28 cm. (fig. 44-c).
4. Shallow rounded bowl with sharply everted rim broken
into large flat insloping lobes. Noncircular (fig. 44-d).
5. Bowl with outflaring walls and sharply everted rim widi
an outsloping to nearly vertical flat top, flat or tapered,
and notched or lobed lip. Noncircular (fig. 44-*).
Base Forms
Five base forms are represented in Napo Phase pottery (see
also Appendix table 10). The existence of complete vessels
and large fragments makes it possible to associate most of
diem with one or more of me rim variations.
A.	Flat, curving junction with the wall; undiickened to
slighdy tiiickened either at bottom or at curve. Diam-
eter 6-20 cm. (fig. 45-A).
B.	Flat, angular junction with the wall. Junction diickened
to produce a gradual curve on the interior. Bottom
sometimes thickened. Exterior may have leaf impres-
sion (pis. 18 i-l, 2\h). Diameter 10-18 cm. (fig. 45-B).
C.	Rounded and unthickened or slighdy diickened at
center (fig. 45-C).
D.	Annular, vertical or flaring ring, varying greatiy in
height and profile. Height 0.5-3.0 cm.; diameter 8-16
cm. (fig. 45-D).
E.	Flat, joining die wall at an angle approximating 90
degrees; typically thickened at angle to produce a
slighdy less abrupt transition on die interior. Bottom
may be of greater or less thickness than wall. Non-
circular (fig. 45-E).
Pottery Types
Armenia Plain
PASTE:
Method of manufacture: Coiling; fracture along coil j unc-
tions common; widtiis of 1.8, 2.0, and 2.2 cm. observed.
Temper: Cariap6, varying from small scattered white
siliceous particles to large and easily observed cellular
"bundles." More finely ground cariape typically as-
sociated with moderate to large amounts of fine sand
containing hematite particles up to 3 mm. diameter.

                 THE   NAPO   PHASE
TABLE A.—Characteristics of Napo Phase anthropomorphic urns

53


CharacteristicsType AType BSexFemaleMaleMale?Male?Male?TemperSandSandForm:ArmsFull round (missing)Full round
Bulging lower armFull round
Bulging lower armFull roundFull roundLegsReliefReliefFull roundRelief except feetFull roundFull roundBulging calfBulging calfBulging calfBulging calfDecoration:HeadHair blackHair blackR/WHair blackFaceUnpaintedRed(eroded)PaintedPaintedBodyR &B/WR & B/WR/WR &B/WTechs. 2 & 3Technique 2Technique 2Dimensions:Total height38 cm.42.5 cm.38 cm.34 cm.Body height26 cm.25 cm.Max. diameter38 cm.21.5 cm.Head diameter18.5 cm.
(orifice)21.9 cm.
(orifice)23 cm.Base diameter25.5 cm.20 cm.16.5 cm.Body wall th.1.5 cm.7-8 mm.Special featuresEar lobes perforatedShield held with bothShield in left handEar lobes perforatedEar lobes perforatedHair comes to pointhandsConical depression inElongated object heldLong pigtail downat middle of back.  right hand
Legs bent so soles
abut at center frontby both hands  back
"Rings" around
upper arm and calfProvenienceRio Napo 20 leagues
above Rio AguaricoRio NapoRio NapoN-P-3?Rio NapoCollectionUhle, 1921, Lam. 3-4Jay C. LefT, on loan
to Brooklyn
MuseumAMNH 41.0/9183Jijon y Caanafio,
1951, fig. 499Thomas FlanneryPlate555661b5754
   
SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY
TABLE A.—Characteristics of Napo Phase anthropomorphic urns.—Continued

VOLUME   6


CharacteristicsType BSexMaleMaleMaleMaleNoneTemperSandSandForm:ArmsFull round
Bulging lower armNoneRelief, bulging
lower armFull roundNoneLegsFull roundNoneReliefFull roundNoneBulging calfBulging calfBulging calfDecoration:HeadR & B/WB/WR & Brown/WHair blackR/WFace(eroded)PaintedPaintedR &B/WPaintedBodyR & B/WB/WR & Brown/WR &B/WR &B/WTechnique 2Technique 4and groovingTechnique 2Technique 3Dimensions:Total height34 cm.36.8 cm.23.5 cm.54 cm.48 cm.Body height13 cm.Max. diameter19 cm.30 cm.Head diameter20.3 cm.Base diameter15.5 cm.16 cm.
(orifice)Body wall th.7 mm.Special featuresHands joinedShield or disk withShield held in bothPossibly once heldpanpipe held inhandsshieldboth handsTriangular relief on
  chest
Long pigtail down
backProvenienceN-P-lOriente regionNapo AreaRio Napo 20 leagues
above Rio AguaricoRio NapoCollectionMuseum ofMAI-HFCasa de laUhle, 1921,Musee de 1'HommePrimitive Art, 56.386/1723Cultura, QuitoLam. 1-208.22.79Plate58a586596061a                        THE   NAPO   PHASE
TABLE A.—Characteristics of Napo Phase anthropomorphic urns.—Continued

55


CharacteristicsType BType CSexMaleFemaleNoneMaleNoneTemperSandSandSandSandForm:ArmsReliefReliefFull roundRelief and full roundReliefLegsReliefReliefFull roundRelief and full roundReliefBulging calfBulging calfBulging calfBulging calfDecoration:HeadR & B/W(missing)(missing)(missing)(missing)FacePaintedBodyR &B/WR & B/WR &B/WR &B/WR & B/WTechnique 3?Techniques 2 & 3Technique 2Techs. 2 & 3Techs. 2  & 3Dimensions:Total height29 cm.Body height15 cm.19.3 cm.34.5 cm.43 cm.Max. diameter20 cm.20.6 cm.27 cm.26 cm.Head diameterBase diameter14 cm.23 cm.21-23 cm.Body wall th.10 mm.7—9 mm.Special featuresTriangular relief onKneelingUpper edge recessedArms bent backwardUpper edge recessedchestCenter of bottom
  broken out
Upper edge broken
  off
Hands and feet
have 4 digitsto receive lid  to receive lid
Center of bottom
broken outProvenienceRio AguaricoRio NapoN-P-lN-P-lRumi-tumi; right
bank of Rio Napo
315 km. above mouthCollectionAlan C. LapinerAMNH 41.0/9184Museo Victor
Emilio EstradaColegio Militar,
QuitoAMNH 41.1/3985Plate62b62a63646556

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Black ash also common, differing from charcoal temper
in having elongated form resembling that of the siliceous
cariape. Evenly distributed; large particles tend to be
aligned parallel to the surface,  but there are many
exceptions.
Texture:   Majority sandy. Large airpockets (1.5 by 3.0
mm.) typical, some resulting from organic matter in
the  clay  destroyed  during  firing,   others  from  poor
kneading of coil junctions.
Color:    Cross-section ranges from solid  orange to solid
gray, with the majority incompletely oxidized leaving
a gray core.
SURFACE :
Color: Tan, dull brownish orange to grayish brown;
color variable over small areas due to poorly controlled
firing. Medium gray fire clouds frequent on interior
and exterior. Exterior may be blackened, possibly from
use in cooking.
Treatment (pi. 18 c-h): Few surfaces are smoothed suf-
ficiendy to eliminate flaws. Majority relatively even
but marred especially on the exterior by scratches
caused by scraping with a rough-edged tool parallel
to tlae rim. Where scratches are absent, pits and flaws
common.  A minority are  better smoothed,  floating

finer particles of clay to produce a fine textured finish.
Broad (2-3 mm.) concave smoothing tracks visible on
   some, but not typical.
Hardness:    3.
FORM:
Rim:    Direct with flat lip (pi. 18 a-b). Size of sample:
  33.
Body wall thickness:    Range 0.6-1.9 cm.;  majority 8-10
  mm.
Base:    Flat (Forms A, B, and E); two of Form B have
leaf impressions on the exterior (pi. 18 i-l). One an-
nular (Form D) fragment with perforations.
Reconstructed common vessel shapes (figs. 40—43):
Form 12: 36.1 percent.
Form 17:  18.0 percent.
Form 20:  12.0 percent.
Form 16:    9.0 percent.
  Form 13:    6.0 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
40, 44):    Form 11, Rare Form 4.
TEMPORAL DIFFERENCES WITHIN THE TYPE :    None observed.
CHRONOLOGICAL POSITION OF THE TYPE:    Present at max-
imum frequency (12.3 percent) in the early part of the
seriated sequence and declining thereafter; absent from
the latest site (fig., 63).











ca

FIGURE 44.—Rim profiles of Napo Phase Rare Vessel Forms 1-5. Arrows designate decorated zone. (Black = undecorated, white = deco-
rated, hachure = red slipped.)

THE   NAPO  PHASE

57










B


FIGURE 45.—Base forms of Napo Phase pottery. A, Flat, curving junction with body wall. B, Flat, angular junction with body wall-
C, Rounded. D, Annular. E, Flat, joining body wall at angle of 90 degrees or more. (Black=undecorated, white=decorated>
hachure=red slipped.)

58

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Armenia White-on-red
PASTE :    On Napo Plain (p. 58) or Armenia Plain (p. 52);
  see tiiose type descriptions for details.
SURFACE:
Color:    The undecorated interior is tan, brown, grayish
brown, widi light to dark gray fire clouds. A few ex-
amples are white slipped.
Treatment:    Even   but   widi   pits,   striations,   exposed
temper  grains and  occasionally  scratches  or  larger
flaws. Slighdy to markedly less well finished dian the
decorated exterior.
Hardness:    2.5.
FORM:
Rim:    Direct,   everted,   cambered   or   channel   (rare),
with flat lip. Size of sample: 14.
Body wall thickness:    Range 4—10 mm.
Base:    Probably rounded (Form C).
Reconstructed common vessel shapes (figs. 36-37, 40-42, 44):
Forms 3, 4, 6, 11, 12, 14, 17, Rare Form 3.
DECORATION (pis. 19-20; Appendix table 11):
Technique: White painted designs on a red slipped
exterior surface (continuing over the lip). Red slip
varies from a coat diick enough to be of even color
to a tiiin wash with a tendency to fire orange. Incom-
plete polishing may produce a striated effect because
polishing tracks tend to fire deeper red than intervening
unpolished streaks. White paint varies in diickness,
thicker coatings providing more color contrast, but
also tending to crackle and chip off. Thinner coatings
have a streaky appearance resulting from nonuniform
coverage of die underlying red slip. Execution is
extremely variable, ranging from a few carefully done
to a few crude and sloppy; the majority are inter-
mediate. Bands often are of irregular width, but tend
to be straight and parallel.
  A rare variant (pi. 20 a-b) has the design laid out
in narrow black lines,   1-2 mm. wide and  1-3 cm.
apart.   White   paint   applied   between   die   oudine,
sometimes   overlapping   die   black   lines   but   rarely
obliterating them.
Motif:   Few sherds are large enough to show more than
one or two parallel lines; however, coils, undulating
bands and stepped elements occur.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence with no consistent trend of
changing frequency (fig. 63).
Napo Negative
PASTE :   All on Napo Plain (p. 58); see that type description
  for details.
SURFACE:
Color:   Buff to pale tan or light orange.
Treatment: Floated sufficientiy to produce even, compact
finish; slick to touch but not glossy.
  Hardness:   3.
FORM:
Rim: Direct, exteriorly diickened or everted with flat
or tapered (rare) lip. Size of sample: 3.
Body wall thickness:   Range 4-8 mm.

Base:   Probably rounded (Form C).
Reconstructed vessel shapes (figs. 36, 40, 44):   Forms 1, 12,
Rare Form 4.
DECORATION (fig. 46; pi. 25 d,f):
Technique:   Resist painting forming gray to black bands
and zones of irregular form on the interior of open
bowls. One well preserved example (pi. 25/) has black
bands 5-10 mm. wide, separated by narrow unpainted
bands of more uniform width. Margins are sharply
defined.
Motif:   Parallel   bands,   interlocking  coils,   and  asym-
metrical zones.
ASSOCIATED   TECHNIQUES:   The   exterior   may   be   Napo
Plain  Excised,   Napo   Red  Excised,   or Napo   Red
Excised, White Retouched.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Very rare, but
probably   present   throughout   the   seriated   sequence
(fig. 63).

FIGURE  46.—Napo Negative  bowl interior.
Napo Plain
PASTE:
Method of manufacture: Coiling. Coil line fractures rela-
tively common, showing coil widths of 1.3, 2.0, and
2.5 cm. (pi. 21 i-k). Junctions do not overlap strongly;
edges tend to be slighdy concavo-convex. Coil junctions
occasionally incompletely obliterated on surface.
Temper: Sand grains usually less than 0.5 mm. in
diameter, occasionally 0.5-1.0 mm. or, in tiiicker
sherds 1.0-1.5 mm. diameter. White particles contrast
sharply with gray paste, glossy black ones with orange
paste (botii colors generally present). Well distributed
and sufficientiy abundant to give fine sandpaper feel
to eroded surfaces. Occasional large (up to 8 mm.
diameter) red hematite inclusions, probably natural
components of the clay.
Texture: Typically compact; breaks evenly, edges not
friable. Occasional long thin airpockets.
Color: Complete variation from orange tiiroughout the
cross-section to gray throughout the cross-section.
About half are completely oxidized. Orange usually
uniform in shade throughout; light orange, light tan,
or buff most typical; reddish orange or rusty brown

THE   NAPO  PHASE

59



less common. Gray varies from dark (black) to pale,
with litde variation on a single sherd. Dark gray core
is rare. Incompletely oxidized examples may show a
gray band along one surface instead of at the center
of the core.
SURFACE:
Color: Typically light tan or light orange, including
buff or cream; occasionally bright orange, reddish
orange, gray-brown, or dark gray. Gray hues often
result from fire clouding, making abrupt transition
from orange color. Dark or light colored surfaces may
occur with eitiier gray or orange cross-section.
Treatment: Smoothed sufficientiy to submerge larger
temper grains and produce a fine-grained "floated"
layer, but leaving smootiiing tracks, defects, and irreg-
ularities. Both broad and fine smoothing tracks occur,
but not on the same sherd; orientation is principally
parallel to rim. Smoothing or scraping may drag
temper particles leaving a scratch about 1 cm. long.
Unevenness is common especially on thicker sherds,
and reflected in unequal thickness of body wall. Some
bowl interiors have polishing striatums, but finish is
never slick or lustrous. Smoothing less complete on jar
interiors, leaving rough zones.
  Hardness:   3-3.5.
FORM:
Rim: Direct with flat lip (pi. 21a, c-d); rarely, exteriorly
thickened or channel with flat or rounded lip (pi. 21 b,
e-f)\ occasionally thickened lobes along a rounded lip
(pi. 2\g). Size of sample: 352.
Body wall thickness: Range 4-16 mm.; majority 8—10
mm.; variable on a single sherd.
Base: Flat (Forms A and B), rounded (Form C), or
annular (Form D). Flat bases may show leaf impressions
on exterior (pi. 2\h).
Reconstructed common vessel shapes (figs. 36, 40-42; pi. 22a):
Form 16: 18. 2 percent.
Form 13: 13.4 percent.
Form 17: 11.7 percent.
Form 12: 10.2 percent.
Form 11: 8. 8 percent.
Form 15: 7. 4 percent.
Form 18: 5. 7 percent.
Form   1:    5. 4 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
36, 42-44; pi. 22 b-c):   Forms 4, 19, 20; Rare Forms
1,2.
TEMPORAL     DIFFERENCES     WITHIN     THE     TYPE:    None
  observed.
CHRONOLOGICAL  POSITION  OF  THE  TYPE:   The  principal
plain type throughout the duration of the Napo Phase
(fig. 63).
Napo Plain Excised
PASTE: Predominantly on Napo Plain (p. 58), occasionally
on Armenia Plain (p. 52), rarely on Tiputini Plain
(p.  72); see those type descriptions for details.
SURFACE:
Color:    Pale buff, light orange, light tan, brown to grayish

brown, with frequent large fire clouds on both interior
and exterior.
Treatment: Variation from even and smooth with fine-
grained texture and flaws obliterated, to sandy leaving
broad smoothing tracks and pits visible. Interior of
bowls typically better smoothed than exterior.
  Hardness:   2.5-3.
FORM:
Rim: Collared, direct, channel, exteriorly thickened or
cambered with flat or rounded lip. Size of sample: 34.
Body wall thickness: Range 5-12 mm. (not including
carination).
Base:   Probably rounded (Form C).
Reconstructed common vessel shapes (figs. 36, 38, 40, 41):
Form 16: 29. 4 percent.
Form 12: 26. 4 percent.
Form 4: 8. 8 percent.
Form 2: 5. 9 percent.
Form 7: 5. 9 percent.
Form 13:   5. 9 percent
Minor   vessel   shapes   (frequency   less   than   5   percent;
figs. 36-38, 41, 43):   Forms 1, 3, 5, 9, 14, 20.
DECORATION (pis. 23-24, 25 a-c, e):
Technique: Double- or multiple-pointed tool used to
produce lines and textured zones on an unslipped
surface. Lines 2-5 mm. wide, and shallow (depth less
than 1 mm.); typically poorly defined because of
dragged edges and execution with successive strokes
the ends of which do not coincide. Terminations often
pushed up. Lines crooked and unevenly parallel.
  Excised zones produced by multiple, often overlap-
ping strokes of the incising tool, creating grooves and
ridges parallel to the long axis of the zone. Depth is
similar to that of incised lines on the same vessel.
Excision is typically restricted to small zones such as
widened intersections and corners. A rare treatment is
addition of white pigment to the incisions and excised
zones.
Motif:   Complicated overall patterns sometimes divided
into panels or bands by straight vertical or horizontal
lines. Squared coils occur singly or interlocking, often
with a short dash at the center. Patterns are rectilinear,
softened by rounding of junctions and corners.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the duration of the seriated sequence (fig. 63).
Napo Plain Incised
PASTE:    Predominantiy on Napo Plain (p. 58), rarely on
Armenia Plain  (p.  52)  or Tiputini Plain  (p.   72);  see
  those type descriptions for details.
SURFACE :
Color: All shades from buff to black resulting from poorly
controlled firing, fireclouding, and possibly retiring
during domestic use.
Treatment: Incompletely smoothed, often showing pits,
flaws, roughness, horizontal smoothing tracks and
fine brush-like striations. Coil junction may remain
incompletely   obliterated   on   interior   of   carination.

60

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Slight floating eliminates abrasive texture from most
surfaces, but temper grains may remain visible. Best
smoothing occurs on interior of open bowls. Decorated
areas do not receive better treatment than undecorated
ones.
  Hardness:   2.5-3.
FORM:
Rim: Direct, exteriorly thickened, channel, everted or
collared, with flattened or rarely rounded or tapered
lip. Size of sample:  165.
Body wall thickness: Range 4—14 mm., reaching 2.4 cm.
at carination.
Base:   All forms.
Reconstructed common vessel shapes (figs. 36, 38-41; pi. 26):
Form 16: 21.2 percent.
Form 7: 15.3 percent.
Form 9: 11.5 percent.
Form 1: 11.5 percent.
Form 8: 9.7 percent.
Form 10: 8.5 percent.
Form 12:    7.9 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
36-37, 40-42, 44):   Forms 2, 3, 4, 5, 13, 14, 15, 17, 18,
Rare Form 5.
DECORATION (pis. 26-31):
Technique: Incision with a double- or triple-pointed tool
on an unslipped surface, producing a mark with parallel
striations in the bed. Rare examples show a multi-
pointed tool. Incisions wide (2-6 mm.; majority 3-5
mm.) and typically about 1 mm. deep, but may vary
from superficial to 2 mm. in depth. Width is often
uniform on a single sherd, but may be erratic because of
inconsistent orientation and pressure of the tool. In-
cisions with a double-pointed tool typically show par-
allel smooth tracks with a central ridge of lower elevation
than the surface of the sherd. A rough texture and
sporadic ridging is more characteristic of marks made
with tri- or multi-pointed tools. Straight lines may be
executed in successive strokes, the ends of which may
not be perfecdy superimposed. Lines are not evenly
spaced or perfecdy parallel. Corners are not overshot,
but overlap is frequentiy visible. Ends are frequentiy
pushed up, leaving a lump of clay 1 mm. in elevation.
Dragged margins are typical of incisions with a rough
trough. Short strokes (length 5 mm.) may occur, but
punctates are absent.
  Two rare variants occur: (1) the use of double lines
in combination with lines executed with a single-
pointed tool (pi. 31), and (2) addition of white pigment
to the incisions. In the first variant, single and double
lines may be interdigitated or may be used on different
parts of the vessel, in which case they may reflect
accidental or careless picking up of the wrong tool. The
second variant, white fill, is difficult to distinguish from
white slipping subsequent to incision, which often
erodes the more exposed portions of the surface while
remaining in the more protected beds of the incisions.
Motif: Complicated filling of space with interlocking
vertical   strokes  attached   to  longer  hroizontal   ones,

interlocking angular coils, and concentric rectanguloid
elements. Predominantiy straight lines are softened by
rounding of corners and junctions. Even filling of space
gives the impression of symmetry, but few patterns are
completely symmetrical.
ASSOCIATED TECHNIQUES :   Negative painting may occur on
the interior.
TEMPORAL DIFFERENCES WITHIN THE TYPE : Examples com-
bining double- and single-line techniques are restricted
to the early half of the seriated sequence.
CHRONOLOGICAL POSITION OF THE TYPE: The second most
common decorated type, present throughout the seriated
sequence without consistent change in frequency (fig. 63).
Napo Red
PASTE:    Red slip is applied to all varieties of paste; see
type descriptions of Armenia Plain (p. 52), Napo Plain
(p. 58) and Tiputini Plain (p. 72) for details.
SURFACE:
Color:
Unslipped surfaces:    Same  range   as  Armenia   Plain,
Napo Plain, and Tiputini Plain.
Slipped surfaces: Variation from deep rich red through
red-orange to tile orange depending on thickness
and firing; readily visible when paste is buff but
difficult to detect when both slip and paste are
orange. Dark gray fire clouds typical.
Treatment:
Unslipped   surfaces:    Irregular    to    even    but    rarely
smooth; texture remains granular on sand-tempered
sherds; pits and flaws characteristic on interior of
narrow mouthed vessels. Minor fluctuation in wall
thickness  evident in  slight  undulation  of surface.
Polishing striations sometimes appear adjacent to
the rim on die interior.
Slipped surfaces:    Thicker slip forms an even coating,
smootii to touch and with low luster.  Horizontal
polishing striations may occur. Applied to the ex-
terior,   usually   above   the   region   of   maximum
diameter.
Hardness:    2.5.
FORM:
Rim:    Direct   with   flattened   lip;   rarely   rounded   or
tapered hp. Size of sample: 86.
Body wall thickness:    Range 5-14 mm.; majority 6-9 mm.
Base:    No examples since slip does not extend to base;
probably flat (Forms A and B) or rounded (Form C).
Reconstructed common vessel shapes (figs. 40, 42):
Form 13: 33.8 percent.
Form 11: 32.6 percent.
Form 17:  17.4 percent.
Form 12:  10.5 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
41, 44):    Form 15, Rare Form 5.
TEMPORAL DIFFERENCES WITHIN THE TYPE :    None observed.
CHRONOLOGICAL POSITION OF THE TYPE:    Napo Red occurs
throughout the seriated sequence in a frequency not ex-
ceeding 12 percent (fig. 63).

THE   NAPO  PHASE

61



Napo Red Excised
PASTE: Predominately on Napo Plain (p. 58), occasionally
on Armenia Plain (p. 52), rarely on Tiputini Plain
(p. 72); see those pottery type descriptions for details.
SURFACE:   Like Napo Red Incised (p. 62); see that type

description for details.
FORM:
Rim:   Direct, exteriorly thickened, channel or cambered,
  with flat or rounded lip. Size of sample: 35.
Body wall thickness:   Range 4—8 mm.
Base:    Probably rounded (Form C).
  

'     y     '	i	i	i
0      I      2      3     4     5 CM
FIGURE 47.—Reconstructed design of Napo Red Excised bowl exterior.

62

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Reconstructed common vessel shapes (figs. 36, 38, 40; pi. 32b):
Form 12: 28.6 percent.
Form 4: 25.7 percent.
Form  2: 20.0 percent.
Form   1:    5.7 percent.
Form  3:    5.7 percent.
Form  7:    5.7 percent.
Minor vessel shapes (frequency less than 5 percent; figs. 37,
40):   Forms 6, 10, 13.
DECORATION (fig. 47; pis. 32-33):
Technique: Incision and excision with a double- or
multi-pointed tool prior to application of red coating
to the surface. Incisions broad (2-5 mm.), shallow
(generally less than 1 mm. deep; rarely 2 mm. deep),
with square or tapered ends. Margins vary from
sharply defined to ragged, and terminations may be
pushed up. Trough texture is granular from exposed
temper particles, which are typically suppressed by
slight floating of the surface. Workmanship tends to
be sloppy. Excised zones small, often restricted to
widening of intersections and corners; unobtrusive.
  Red slip was typically applied after incision, filling
the cuts. The surface was not subsequentiy polished, so
that finish depends upon treatment prior to decoration.
Discoloration by firing may make slip invisible on part
of a vessel, while other areas remain a rich dark red.
Motif:   Similar to Napo Red Incised (p. 62); see that
type description for details.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence, with the possible exception of
the latest levels (fig. 63).
Napo Red Excised, White Retouched
PASTE:   Exclusively on Napo Plain (p. 58); see that type
description for details.
SURFACE:   Similar to Tiputini Red Excised, White Re-
touched (p. 76); see that type description for details.
FORM:
Rim:    Channel, direct or exteriorly thickened, with flat
or rounded lip. Size of sample: 9.
Body wall thickness:   Range 4—12 mm.
Base:   Rounded (Form C) or annular (Form D).
Reconstructed common vessel shapes (figs. 36 40):   Forms 1,
2, 12.
DECORATION (pi. 34):
Technique: Incision and excision on a red slipped surface
with a double- or multi-pointed tool. Incisions wide
(3y> mm.), shallow (less than 1 mm.), uniform in
width and depth on a single sherd. Margins may be
dragged or irregular; terminations square and some-
times pushed up.
  Excision produced by repeated application of incising
tool to widen area. Depth is similar to that of incised
lines; floor is often uneven.
  Red coating typically a thin wash contrasting poorly
with the reddish orange natural surface, always applied
prior to decoration.
  
White coating applied to cuts varies from a thick
coating showing fine crackle to a thin wash slopping
over onto the adjacent surface.
Motif: Complicated patterns of short straight lines often
interdigitated so as to create a negative effect in which
the motifs are formed by die intervening surfaces
rather than the incisions.
ASSOCIATED  TECHNIQUES:   Negative  painting  may  occur
on the interior.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Sporadically rep-
resented but present throughout the seriated sequence
except at the latest site (fig. 63).
Napo Red Incised
PASTE:    Predominantly on Napo Plain (p. 58), rarely on
Armenia Plain  (p.  52)  or Tiputini Plain  (p.  72);   see
  those type descriptions for details.
SURFACE:
Color: Light buff, tan, orange, brown to gray; widely
variable over a small area from fire clouding and poorly
controlled firing.
Treatment: Smoothed sufficientiy to produce an even
surface and sometimes floated but more commonly
remaining slightly granular in texture. Fine smoothing
striations may occur parallel to the rim. Interior gen-
erally slightly better smootiied than exterior, which
bears the decoration.
   Hardness:    2.5.
FORM:
Rim: Channel, exteriorly thickened, cambered or
direct with flat or rarely rounded lip. Size of sample:
43.
Body wall thickness:    Range 3-9 mm.; majority 5-6 mm.
Base:    Rounded (Form C); possibly annular (Form D).
Reconstructed common vessel shapes (figs. 36, 40):
Form 12: 23.3 percent.
Form 4: 18.5 percent.
Form 1: 16.3 percent.
Form 13: 11.6 percent.
Form   2:    7.0 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
36-38, 41, 44):    Forms 3, 5, 8, 9, 15, Rare Form 4.
DECORATION (pi. 35):
Technique: Incision with a double- or multi-pointed tool
before or after red slipping of the surface. Incisions
wide (3-5 mm.) and relatively shallow (less than 1
mm.). The trough typically has a granular appearance
from exposed temper grains, providing strong textural
contrast with the surface especially when done sub-
sequent to slipping. A double-pointed tool is most
typical, leaving two parallel grooves with a slight but
clearly visible ridge between; occasionally the ridge is
off center. Terminations are square or tapered; square
ends may have a thrown up margin. When incised
prior to slipping, the slip fills and rounds over some
roughness. Incisions may have a ragged or even margin;
the former is more common. Margins are rarely
thrown up.  Spacing is a little more equal than on

THE   NAPO  PHASE

63



Reconstructed common vessel shapes (figs. 38—40):
Tiputini Red Incised, lines are less crooked and more
evenly parallel.
  Red slip varies from a good coating like that typical
of Tiputini Red Incised to a thin wash that fires
brownish or blends into the unslipped surface. Thin
wash is more typical of Napo Red Incised.
  On some examples, a white coating was added to the
trough subsequent to red slipping. Thickness varies
from a solid coating widi fine crackle to a faint wash
that may become blended into the surface by firing
discoloration. White is often slopped over edges or
ends of incisions onto the adjacent red surface.
Motif:    Complicated   arrangements  of concentric   and
interlocking vertical and horizontal parallel lines.
ASSOCIATED   TECHNIQUES:    One   bowl   with   Napo   Red
Incised, white retouched decoration on the exterior is
polychrome painted on the interior.
TEMPORAL DIFFERENCES WITHIN THE TYPE :    None observed.
CHRONOLOGICAL POSITION OF THE TYPE:    Present through-
out the seriated sequence in slightly increasing frequency
(fig. 63).
Napo White Excised
PASTE AND SURFACE:   On Napo Plain  (p. 58); see that
type description for details.
FORM:   Only body sherds are represented; they appear to
correspond to carinated and rounded bodies of vessel
Forms 7-10 (figs. 38-40).
DECORATION (pi. 36):
Technique:   Incision  and  excision with a  multi-pointed
tool prior to the application of a white coating to the
surface. Technique and slight use of excision resemble
Napo Plain Excised (see p. 59).
Motif:   Straight vertical and horizontal lines in compli-
cated arrangements (cf. Napo Plain Excised).
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Very rare (fig. 63)
but given status as a pottery type to facilitate comparison
of the ceramic complex with that of other areas.


Form 9
Form 7
Form 8
Form 10
51.8 percent.
20.7 percent.
17.3 percent.
6.9 percent.
Minor vessel shapes (frequency less than 5 percent;  fig.
41):   Form 16.
DECORATION (pis. 37-40):
Technique: Incision widi a double- or multiple- (rare)
pointed tool on a plain surface subsequentiy coated with
a thin white slip or wash. Incisions broad (2.5-6.0 mm.;
majority 3-5 mm.), usually 0.5-1.0 mm. deep, with
ragged, irregular margins. Pushed up terminations are
common but thrown up margins are rare. Incisions
typically are bifurcated by a fine ridge, whose straight-
ness contrasts with the irregularity of the walls. Rarely,
this ridge is at or near surface elevation. Execution is
not careful; lines are unequally spaced, unevenly par-
allel and crooked.
  Small circular applique nubbins may occur on lobes.
Diameter is 0.5-1.0 mm., elevation 1-3 mm. A central
punctate is typical.
  A thin wash was applied to the decorated zone subse-
quent to incision. Color is typically cream rather than
white. Present condition shows no sheen or evidence of
polish, suggesting application by dipping or brushing
without subsequent smoothing. Paper thin but tends to
minimize the visibility of incision when lines are
shallow. Eroded examples of this type are difficult to
distinguish from Napo Plain Incised sherds with white
fill in the incisions.
  The carination may be expanded to broad horizontal
lobes, which continue the incised decoration on their
upper surfaces.
Motif:   Complicated interlocking arrangements of hori-
zontal and vertical parallel lines, sometimes forming
symmetrical or asymmetrical panels.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Possibly absent
from the latest levels of the seriated sequence (fig. 63).
  

  
Napo White Incised
PASTE:   Predominantly on Napo Plain (p. 58), rarely on
Armenia Plain  (p.  52)  or Tiputini Plain  (p.  72);  see
those type descriptions for details.
SURFACE:
Color:    Orange,   tan,   reddish  orange,   buff;  dark  gray
  where fire clouded.
Treatment:   Even but not smooth, retaining a granular
texture   and   leaving   pits   and   flaws   unobliterated.
Smoothing  striations  parallel  to  the  rim  vary from
  fine brush-like marks to deep cuts.
Hardness:   2.5.
FORM:
Rim:    Channel,  exteriorly thickened or collared  (rare)
with flattened, tapered or rounded (rare) lip. Size of
  sample: 29.
Body wall thickness:   Range 3.5-8.5 mm., increasing to
  3.3 cm. at carination (corner).
Base:   Probably rounded (Form C).

Rocafuerte Incised
PASTE:   Predominantiy on Napo Plain (p. 58), rarely on
Armenia Plain (p. 52); see those type descriptions for
details.
SURFACE :
Color:   Buff to light tan.
Treatment:   Smoothed insufficiently to remove all flaws
  and eliminate granular texture of sandy paste.
Hardness:    3.
FORM:
Rim:    Everted or direct with flat or rounded (rare) lip.
Size of sample:   26.
Body wall thickness:   Range 5-10 mm.
Base:   Rounded (Form C) or flat (Form A).
Reconstructed vessel shapes  (figs.  40-41,   44;   frontispiece;
pis. 41-42):
Form 14:       88.5 percent.
Form 11:	7.8 percent.
Rare Form 3: 3.7 percent.

293-822  O - 68

VOLUME   6
SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

cV<U1V-Qwa2mco•a(AttbnC^.MoOJ3caw42bnII•5CrtrtXIO,,<!>TJ-whII
bncCojgJiaRj•dX!-—*andign.T)V4JTJCJc.5rtioa
ooa<J.—,.>*T3ctiseerlu>coHHiUTiuH1pamoTIrta<*H       rt

5y" \c/>CDV-V        |a>.W:  1*?-:    I1?     1CO»;:■' *    1■*9Pa

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY
VOLUME   6

THE   NAPO  PHASE

65



DECORATION (figs. 48-49; frontispiece; pis. 41^13):
Technique: Incision widi a double-pointed tool prior
to white slipping; incisions painted red subsequent to
slipping of the surface. Incised lines broad (4-6 mm.),
shallow to deep (less tiian 1 mm. to 2 mm.), crudely
executed, with very irregular margin, and marked
variation in width and deptii. Trough very uneven
and showing parallel ridges left between tool points.
Small areas excised by adjacent tool strokes in a zone
previously oudined by incision.
  Cream to off-white slip applied after incision covering
surface and incised decoration. Some vessels have a
smooth, even coating; others are uneven partly because
of poor leveling of underlying surface. Fine crackle
may occur.
  A red coating was applied after slipping to incisions
and excisions, following die irregular margins and
rarely sloping slighdy over onto the adjacent surface
(pi. 43 a-b, e). Rich red to reddish orange color provides
strong contrast with white slip. About 50 percent have
decoration limited to red-retouched incisions.
  Polychrome painting of the surface surrounding the
incision (pi. 43 c-d, f) occurs on about 50 percent of
the vessels. In such cases, the red retouch may extend
beyond the incision to produce a red zone bordered by
a narrow black line. This variant may not have all
incisions retouched with red.
Motif: Rocafuerte Incised occurs almost exclusively on
the upper interior surface of open basins, generally
associated with complicated lobing of the lip. Typical
elements are straight and hooked lines, many of which
terminate in a notch in the lip.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed;
examples    with    and    without    associated   polychrome
painting  occur  with  equal  frequency  throughout  the
seriated sequence (Appendix table 11).
CHRONOLOGICAL POSITION OF THE TYPE:   Present in minor
frequency except at the latest sites (fig. 63).
Rocafuerte Painted
PASTE:   Predominantiy on Napo Plain (p. 58), occasionally
on Armenia Plain  (p.  52)  or Tiputini Plain  (p.   72);
see those type descriptions for details.
SURFACE:
Unslipped:   Like Napo Plain, Armenia Plain, and Tiputini
Plain; see those type descriptions for details.
White slipped:
Color:   Off-white to cream, dirty white, or buff. Some
correlation with thickness since thinner coatings are
more   transparent  and   incompletely   obliterate   the
underlying buff to orange surface.
Treatment:   Even, frequentiy smooth, and never slick
or glossy.   Paper  thin  coating often  clearly visible
in cross-section because of sharp contrast with paste
color; this type provides the whitest and most even
surface. A thinner application does not overcome the
granular   texture   of  the   underlying  surface.   Fine
crackle may occur, but is not typical.
Hardness:   3.
FORM:
Rim:   Direct, everted, cambered, or exteriorly thickened;
rarely,  channel or  collared;   flat   or   rounded   (rare)
lip. Size of sample: 186.

Body wall thickness:   Range   5-18   mm.;   majority   6-10
  mm.
Base:   Flat  (Forms A,  B,  E),  rounded  (Form  C),  or
  annular (Form D).
Reconstructed common vessel shapes (figs. 36-42):   Although
the principal vessel shapes are almost the   same   in
the three varieties of painting, there is some difference
in emphasis (table B).
TABLE B.—Frequency of vessel shapes in decorated variants of Rocafuerte
Painted

Decorated VariantsVessel shapesRed and black-
on-whiteBlack-on-whiteRed-on-whiteForm     PercentForm     PercentFormPercent17        32.714        30.41755.514        23. 117        23.91111.2Common6          8.76        13. 168.413          6.77        10.9148.411          5.811         10.9155.612          4.85          4.4132.87          3.812          4.4162.83          3.813          4.416          3.8Minor5          2.9
9          1.9
1          0.9
21          0.0DECORATION (figs. 50-58; pis. 44-65): Although painted
decoration can be separated on the basis of paint color
into tiiree distinct varieties (red-on-white, black-on-
white, and red plus black-on-white), the eroded condition
and small size of most sherds makes recognition of tiiese
categories possible in less than 25 percent of die sherds
classified as Rocafuerte Painted. The majority were
placed in this type because they had a white slipped surface,
sometimes with a trace of red or black paint. Even
where preservation is good, subdivision is not always
accurate. Reliability is greatest for red-on-white, since
this typically occurs alone. Differentiation between black-
on-white and polychrome may be difficult, however, since
many polychrome designs make limited use of red so
that small sherds may represent only black-on-white
portions of the design. Examination of larger sherds
and complete vessels shows several distinctive types of
decoration within each of the three varieties.
Technique:
A. Red paint on a white slipped surface (fig. 50; pis. 44,
616). Paint rich red to light orange, depending on
thickness and firing. Line width not uniform because
of uneven flow and use of multiple strokes to produce
longer lines. Application may be in the form of thin
lines   (width   1.5-4.0   mm.),   wider   bands   (width

66

SMITHSONIAN  CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6


g
o
'5
a
'v
u
a M
II
■M
M II
SP"!
a rt
IS
rt T)
•a te
u
rt
3 e
bo
rt "O
^ii
« -5
<*-
,+J    to
OH   JJ
3 8.
X> ,_
o J2
U    s-.
p
j
.a s
•3-a
u   rt
H   TJ
U
U
ed   o
O —<

ta
D
C
ta



66

SMITHSONIAN  CONTRIBUTIONS  TO  ANTHROPOLOGY

VOLUME   6






THE   NAPO  PHASE

67



0.7-1.7 cm.), or solid zones. Solid zones may intervene
between lines or occur independentiy; in botii cases
they are oudined with narrow lines and then filled
in. Dots occur rarely either singly or in a row.
Lines and bands are not straight or evenly parallel;
intersections are often overshot. Designs generally
incorporate broad and narrow lines and solid areas,
although relative frequency varies considerably.

often remain visible. Corners are widened, suggesting
splaying out of brush on curve; corners may be
angular on the inner side and curved on the outer one.
Solid black zones and very wide bands are outiined
with narrow lines and then filled in. If well done,
fill blends with and renders invisible the outline; if
sloppy, filling is a lighter gray and streaky from
unequal thickness of the paint.





3 CM










FIGURE 50.—Type sherds of Rocafuerte Painted, red-on-white.
B. Black paint on a white slipped surface (fig. 51;
pis. 45, 46-rim, 50, 54b, 58b). Line width varies but
generally falls into two classes: (1) narrow (0.5-2.0
mm.) lines, and (2) wide (majority 1.0-1.5 cm.,
some 2.0 cm.) bands. Narrow lines may be consistent
in width or vary throughout the range over a short
distance because of unequal pressure or difference in
the amount of paint on the applicator. Consecutive
strokes often overlap or produce abrupt changes in
line width. Intersections may be overshot (fig. 51
b-c). Wider bands may be done with a wide appli-
cator or by repeated parallel strokes with a narrow
one.  Striations parallel to the length of the band

FIGURE 51.—Decorative techniques of Rocafuerte Painted, black-
on-white, a, Technique 1.  b-c, Technique 2. d, Technique 3.
  Dotted lines are carelessly executed, a single line
being composed of "dots" of varying shape (circular,
triangular, elongated, or shapeless "blobs"), size,
and orientation. Spacing is unequal, varying from
adjacent or even overlapping to 3 mm. apart. Failure
to lift the applicator may produce a "drag-and-jab"
effect.
  Lines, bands, dots, and solid areas are employed
in three principle ways, rarely combined on a single
vessel:
1.   Bands  delimiting  fields  containing  patterns of
dotted lines (fig. 51a; pis. 45 g-h, 50a).

68

SMITHSONIAN   CONTRIBUTIONS  TO  ANTHROPOLOGY

VOLUME   6



2. Narrow solid lines and dotted lines interspersed,
or narrow solid lines defining fields occupied by
dotted lines (fig. 51 b-c; pis. 45 i-j, 54b).
3. Pseudo-negative patterns produced by drawing
the design in narrow lines and filling in the
background, leaving the unpainted surface to
carry the predominating motifs (figs. 51^, 54a,
55-body, 51 d; pis. 45 a-f, 46-rim, 50 b-d, 58b).
C. Black and red paint combined on a white slipped
surface (pis. 46-49, 51-65). Solid and dotted lines,
bands and zones are of die same varieties and con-

struction   as   in   red-on-white   and   black-on-white
designs. Red and black painting is combined in five
principal ways, several of which may occur on a
single vessel:
1. Broad (1.0-1.5 cm.) red bands bordered by
narrow (1 mm.) black lines, which may also
occur independentiy. Black border may overlap
the edge of the red band partiy or completely,
or may be adjacent (fig. 52a; pi. 46-body,
47 a-b).








FIGURE 52.—Decorative techniques of Rocafuerte Painted, red and black-on-white. (Stippling=red; black=black.)   a, Technique 1
b, Technique 2.   c, Technique 3.   d, Technique 4.   e, Technique 5.

THE   NAPO   PHASE

69



Broad black bands and narrow black solid or
dotted lines used to create the design, portions of
which are accented with narrow red lines or
filled with red (figs. 52b, 53, 54b, 56-body, 57
a-c; pis. 48, 55a, 56-57, 58a, 60, 62a-side and
rear, 63, 64-body, 65-body).
Narrow black solid or dotted lines used to form
the design, portions of which are filled with red
(figs. 52c, 55-rim, 56-rim, 58; pis. 47 c-d, 54a,
55b, 62a-front, 626, 64-arms and legs, 65-arms
and legs).
Thin paired red and black lines, each about
1 mm. wide, the inner edges abutting (fig. 52a";
pi. 47*).

5.   Pseudo-negative patterns produced by drawing
the design in narrow black lines and filling in
the background  (cf.  black-on-white Technique
3). Red is used for accent, filling small areas or
bordering larger ones (fig. 52e; pis. 49, 51a).
Motif:   Painted designs are difficult to describe because
of their frequentiy complicated arrangement of asym-
metrical  elements  drawn  with  straight  and  curved
lines. Coils, hooks, S-shaped figures, stepped elements,
and parallel lines are repeatedly employed.
TEMPORAL DIFFERENCES WITHIN THE TYPE:   All techniques
occur throughout the seriated sequence, with the possible
exception of polychrome Technique 1, which is absent
from the upper third (Appendix table 11).


FIGURE 53.—Type sherd of Rocafuerte Painted, red and black-on-white, Technique 2. (Stippling=red, black=black.)

70

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6


FIGURE   54.—Rolled-out design  of Rocafuerte  Painted jar  of   Form 6 with anthropomorphic face on rim.  (Stippling = red,
black=black.)   a,  Black-on-white  Technique   3   on   rim.   b,   Red and black-on-white Technique 2 on body. (cf. pi. 46).

THE   NAPO  PHASE

71


I	I	l	I	I	I
0      I      2     3     4    5 CM
FIGURE 55.—Rolled-out design of Rocafuerte Painted, red and black-on-white, Technique 3 on rim and black-on-white, Technique 3 on
body. (Stippling=red, black=black.)

72

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6




5  CM




FIGURE 56.

-Rolled-out design of Rocafuerte Painted, red and black-on-white. (Stippling=red, black=black.) Technique 3 on rim and
Technique 2 on body.



CHRONOLOGICAL DISTRIBUTION OF THE TYPE : The principal
decorated type, constituting between 10 and 15 percent
of the pottery in most levels, throughout the duration of
the seriated sequence (fig. 63).
Tiputini Plain
PASTE:
Method of manufacture: Coiling; occasional fractures ex-
pose coils 1.5-2.0 cm. wide.
Temper: Charcoal, varying in size from minute up to 2
mm., with rare particles up to 7 mm. Relatively abun-
dant and easily visible in oxidized paste. Also fine sand
containing mica flecks. Charcoal is typically rounded
and does not have cellular structure of cariape. A few
sherds also have sparse amounts of cariape.
Texture: Sandy, friable. Removal of charcoal by erosion
leaves pits, weakening the sherds.
Color:    Grayish tan to dark gray core, fired tan to light
orange in a band 0.5-1.0 mm. wide along both surfaces;
firing variable on a single sherd.
SURFACE :
Color: Buff, light tan, light orange, reddish tan, grayish
tan to gray; considerable variation on a single sherd.
Gray patches result from fire clouding.

Treatment: Smoothed with hand and possibly a scraping
tool but leaving unevenness and irregularities. Smooth-
ing may drag large temper particles, producing a short
scratch. Never slick to touch.
  Hardness:    3.5-4.
FORM:
Rim:    Direct widi flat lip. Size of sample: 26.
Body wall thickness: Range 0.7-1.7 cm.; majority 10-12
mm.
Base: Flat (Forms A and B), rounded (Form C); one
annular (Form D).
Reconstructed common vessel shapes (figs. 40-42):
Form 16: 54.0 percent.
Form 12:  19.3 percent.
Form 11:    7.7 percent.
Form 17:    7.7 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
40, 42, 44):    Forms 13, 18, and Rare Form 2.
TEMPORAL DIFFERENCES WITHIN THE TYPE :    None observed.
CHRONOLOGICAL POSITION OF THE TYPE:    Occurs through-
out the Napo Phase sequence in decreasing frequency
(fig. 63).
Tiputini Plain Excised
PASTE: Predominantiy on Napo Plain (p. 58), occasionally
on Armenia Plain (p. 52), rarely on Tiputini Plain (p. 72);
see those type descriptions for details.

THE   NAPO   PHASE

73









2    3    4   5 CM

FIGURE 57.—Type sherds of Rocafuerte Painted, red and black-on-white. (Stippling=red, black=black.)   a-c, Red and black-on-
Technique 2.   d, Black-on-white, Technique 3 with red on upper border.

white



SURFACE :   Like Napo Plain Excised (p. 59); see that type
description for details.
FORM:
Rim:   Direct, cambered, channel, or exteriorly thickened
  with flattened or rounded (rare) lip.
Body wall thickness:   Range 4—9 mm.
Base:   Probably rounded (Form C).
Reconstructed common vessel shapes (figs. 36—37, 40; pi. 66):
Form 12: 49.2 percent.
Form 33: 16.9 percent.
Form   5: 15.4 percent.
  Form   4:   6.1 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
37-38, 40-41):   Forms 6, 7, 9, 13, 15.

DECORATION (pis. 66-69):
Technique: Incisions and excised areas executed widi
a single-pointed tool on an unslipped surface. More
carefully done but with greater variation in width
and depth than on Napo Plain Excised. Incisions 1-5
mm. wide, up to 2 mm. deep, with width and depth
relatively uniform on a single vessel; rectanguloid
cross-section. Edges and terminations sometimes thrown
up. Intersections not overshot. Strokes straighter than
in Napo Plain Excised, but not evenly parallel.
  Excised zones outlined by incision and cut back to
a depth equal or greater than incisions on the same
vessel. When excision goes deeper than the outiine,
margins tend to be ragged. Bed of excision varies from
  
74

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6




0   I    2   3   4   5 CM


FIGURE 58.—Reconstructed Rocafuerte Painted vessel with anthropomorphic treatment. (Stippling in zones=red, black=black.)

level but not smooth to uneven, leaving distinct parallel
gouge marks and intervening ridges.
  Small circular punctates may be associated; depth
is similar to that of incisions on the same vessel.
  A few examples (pi. 69) have thick white paint applied
to wide incisions and excised areas, often slopping
over onto the margin of the adjacent surface. This
inlay tends to flake off, but can usually be distinguished
from white slipping subsequent to decoration by its
greater thickness, as well as absence from the well
preserved portions of the surface.
Motif: Design area may be divided into panels by an
undulating line. Excision is employed to cut back
trianguloid or stepped zones, frequently leaving a
circle in relief at the center. Excision may be limited to

occasional accent in incised designs but more typically
is employed to produce 50 percent or more of the
pattern.
TEMPORAL    DIFFERENCES    WITHIN    THE    TYPE:     None
observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence except possibly in the most
recent levels (fig. 63).
Tiputini Plain Incised
PASTE: Principally on Napo Plain (p. 58); rarely on
Armenia Plain (p. 52) or Tiputini Plain (p. 72); see
those type descriptions for details.
SURFACE :
Color:   Buff, tan, orange, reddish orange, grayish brown;

THE   NAPO  PHASE

75



uniform or with little variation in tone except where
fire clouded.
Treatment: Variation from even and compact but not
slick, to irregular with pits, flaws, striations, and pro-
truding temper grains. Decorated surfaces (exterior)
may be more poorly smoothed than undecorated
interiors.
  Hardness:    2.5-3.
FORM:
Rim: Exteriorly thickened, cambered, channel, direct,
everted or collared, with flattened or rarely rounded or
tapered lip. Size of sample: 63.
Body wall thickness: Range 4—10 mm.; majority 4—6 mm.
(except at carination, which may reach 1.8 cm.).
Base: Flat (Forms A, B, and E), rounded (Form C);
probably also annular (Form D).
Reconstructed common vessel shapes (figs. 36, 38, 40-41):
Form 12: 36.5 percent.
Form 7: 9.5 percent.
Form 8: 7.9 percent.
Form 1: 7.9 percent.
Form 4: 6.4 percent.
Form 14:    6.4 percent.
Minor vessel shapes (frequency less than 5 percent; figs.
36-41, 44):   Forms 3, 5, 9, 10,  11, 13,  15, 16; Rare
Form 4.
DECORATION (pi. 70):
Technique: Incision with a single-pointed tool on an un-
slipped surface. Wide variation in width and depth of
lines and in quality of design. Incision width typically
1-2 mm., rarely up to 5 mm.; depth 1-3 mm. with
vertical walls. Execution typically sloppy, with thrown
up edges and ends, ragged edges, or a "plug" of clay
left in the trough at the end of a stroke. Lines are
crooked, unequally spaced and unevenly parallel, but
junctions are not overshot. Punctates may be associated,
either circular (diameter 2 mm.) or rectanguloid (4 by
2 mm. to 4 by 5 mm.). Depth is the same as the asso-
ciated incised lines.
  A few sherds preserve traces of white pigment in the
bed of the incision, applied prior to firing. This tech-
nique is difficult to distinguish from white slipping in
the eroded condition characteristic of most sherds, and
may have been more common than evidence now
suggests. Sherds combining single- and double-line
techniques have been classified under Napo Plain
Incised (see p. 59).
Motif:    Complicated combinations of straight and curved
lines, including coils, concentric triangles, and undulat-
ing lines. One example is diagonal checkerboard (pi.
IQh). A face also occurs only once (pi. 70a).
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence in slightly declining frequency
(fig. 63).
Tiputini Red Excised
PASTE:   Principally on  Napo  Plain (p. 58), occasionally
on Armenia Plain (p. 52), rarely on Tiputini Plain (p.
  
72); see those type descriptions for details.
SURFACE:
Color: Cream, light tan, light brown to reddish brown.
Fire clouding frequent.
Treatment: Generally even, varying from smoothed leav-
ing flaws, pits, and scraping striations to compact and
smooth to touch, with the latter in the minority.
Decorated surface (exterior) slighdy better finished but
may retain striations, small pits, and visible temper
grains.
  Hardness:   2.5.
FORM :
Rim: Direct, cambered, channel or exteriorly thickened
(rare), with flat or rounded lip. Size of sample: 64.
Body wall thickness:   Range 3-12 mm.; majority 4—9 mm.
Base:    Probably rounded (Form C).
Reconstructed common vessel shapes (figs. 36-37, 40):
Form 12: 50.0 percent.
Form   3: 22.2 percent.
Form   5:    6.7 percent.
Form   4:    5.5 percent.
Minor vessel shapes (frequency less than 5 percent;   figs
36-37, 40):   Forms 1, 6, 11, 13.
DECORATION (pis. 5\b, 71-72):
Technique: Incisions and excised zones executed with a
single-pointed tool before or after the application of a
red slip. Incisions are sharply defined, vertical walled,
rectangular in cross-section. Width varies from 1-5
mm., depth from 0.5-3.0 mm., but variation is slight
on a single example. Narrower lines are more even
in width than wider ones, which often have irregular
margins partly resulting from failure of successive tool
strokes to coincide. Margins and terminations may be
slighdy thrown up. Corners are usually rounded; inter-
sections are not overshot. Lines are not equally spaced
or evenly parallel.
  Excised areas are defined by incised lines, which
often remain visible; a few examples are so irregular
as to suggest outlining may have been omitted. Ex-
cision is gouged out, leaving a very uneven floor in
most cases. Depth varies from 1-4 mm. Excised seg-
ments are not uniform in size or shape, even when
they constitute repeating elements' of the design.
  Small circular to ovoid punctates are frequentiy
employed.
  Eighty-five percent of the examples were red slipped
on the exterior prior to incision, the remainder subse-
quendy. Color varies from rich red to reddish orange,
often in the form of a thin wash incompletely oblit-
erating the underlying surface.
Motif:    Parallel vertical or diagonal lines define panels
filled   with   stepped   elements,   interlocking   or   single
coils   or   asymmetrical   figures.   Larger   excised   areas
often contain a circle with a central punctate.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Present through-
out the seriated sequence except possibly in the latest
levels (fig. 63).

76

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



Tiputini Red Excised, White Retouched
PASTE:    Predominantiy on Napo Plain (p. 58), rarely on
Armenia Plain (p. 52) or Tiputini Plain (p. 72); see those
type descriptions for details.
SURFACE:
Color:    Buff, light tan, grayish tan, orange to reddish
orange;  relatively uniform  shade  except  where  fire
clouded.
Treatment:    Smoothed, obliterating flaws; sufficient float-
ing to bring enough fine clay particles to surfaces to
produce  an  even,   compact  surface.   Fine   brush-like
horizontal   striations   may   remain   visible.   Majority
even and smooth to touch.
Hardness:    2.5.
FORM:    Same as Tiputini Red Excised (see p. 75). Size of
  rim sample: 26.
DECORATION (pi. 73 a-l):
Technique: Incision, excision and punctation with a
single-pointed tool before or after addition of red slip,
subsequendy white coated. Incision variable in widtii
(1-6 mm.) and depth (0.5-3.0 mm.), but relatively
uniform on a single vessel. Lines bold, sharply defined,
vertical walled, but not straight or evenly parallel.
Edges not dragged or thrown up; intersections not
overshot.
  Excised zones outiined by incision and then gouged
out with short strokes parallel to the longest axis of the
zone. Bed is typically left rough, but better done ex-
amples have parallel scoring of even depth. Boundary
may be jagged or nonsymmetrical. Depth varies from
1-3 mm., and may be equal to or greater than the depth
of incised lines on die same vessel.
  Circular punctations are 1-4 mm. in diameter, 1-3
mm. deep.
  Thin red slip or wash was applied before or after
incision to the exterior, including the lip, and some-
times slopping slightly over onto the interior. Rarely,
red was applied only to the decorated area, leaving the
lower exterior plain. Color varies from rich red to
reddish orange depending on thickness and firing.
The final step in decoration was the addition of a
white wash or slip to the incisions and excised zones.
The  material  varies from a  thick coating with fine
crackle lines to a thin wash, which often slops unevenly
onto the adjacent surface.
Motif:    Similar to Tiputini Red Excised (see pis. 71-72).
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:    Present through-
out   the seriated sequence except possibly in the upper
third (fig. 63).
Tiputini Red Incised
PASTE:   Predominantly on Napo Plain (p. 58); rarely on
Armenia Plain  (p.  52)  or Tiputini Plain  (p.   72);  see
those type descriptions for details.
SURFACE :
Color   (undecorated   surfaces):   Cream,   light  tan,   light
brown, grayish brown, reddish orange; variable over
small area because of poorly controlled firing and fire

clouding. Dark gray fire clouds common.
Treatment: Even but rarely smoothed sufficiently to
eliminate abrasive texture derived from sand temper.
Pits and defects rare. Horizontal smoothing striations
may be accompanied by a slight ridge.
  Hardness:   2.5.
FORM:
Rim: Direct, exteriorly thickened, cambered, channel or
everted, with flattened, rounded, or tapered lip. Size
of sample: 19.
Body wall thickness:   Range 4-10 mm.
Base:   Rounded (Form C), possibly also flat (Form A).
Reconstructed common vessel shapes (figs. 36, 40—41):
Form 12: 52.8 percent.
Form 4: 21.1 percent.
Form 1: 10.5 percent.
Form 3: 10.5 percent.
Form 15: 5.3 percent.
DECORATION (pi. 74 a-i):
Technique: Medium to broad incisions executed with a
single-pointed tool before or after application of a red
slip. Width varies considerably (1-4 mm.) within the
type, but is relatively uniform on a single vessel. Depth
is between 1-2 mm.; also with litde variation on a
single example. Depth is not correlated with width,
and the widest lines may be at either end of the range.
Circular or elongated punctations may be associated.
Execution is typically sloppy, with ragged or thrown
up margins, unequal spacing and poor parallelism.
Corners tend toward rounding. Intersections are not
overshot.
  A red slip was applied typically after incision but
occasionally before. Where well preserved, the color
is rich red to reddish orange. Streakiness may result
from uneven application. Where applied after incision,
beds of incisions are evenly coated suggesting a rela-
tively liquid condition. Red may slop unevenly over
the lip onto the interior of the rim, but does not cover
the interior.
  Tiputini Red Incised, white retouched variety occurs
on a few sherds, where a white coating was applied
to the incisions, often slopping over onto the adjacent
surface along the margin (pi. 14 j-o). When applied
thickly, the pigment develops fine crackle and tends
to flake off especially when applied over red slip.
Motif:    Parallel lines changing direction at 90 degree
angles   to   produce   hooks   and   rectanguloid   figures,
sometimes with a central dash or punctation.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL DISTRFBUTION OF THE TYPE:   Absent from
the upper third of the seriated sequence (fig. 63).
Tiputini White Excised
PASTE AND SURFACE :   On Napo Plain (p. 58); see that type
  description for details.
FORM:
Rim:   Cambered with flat lip. Size of sample: 3.
Body wall thickness:    5 mm.
Base:   Rounded (Form C).

THE   NAPO  PHASE

77



  Reconstructed common vessel shapes (fig. 36):   Form 3.
DECORATION (pi. 73 m-n):
Technique:   Incision and excision with a single-pointed
tool prior  to coating of the decorated area with a
white wash. Execution similar to that of Tiputini Plain
Excised (see p. 72).
Motif:   Undulating line divides the decorated area into
semicircular panels containing a curved stepped line
and a circle with a central punctate.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL POSITION OF THE TYPE:   Very rare (fig.
63) but given type status to facilitate comparison of the
ceramic complex with that of other areas.
Tiputini White Incised
PASTE :   All on Napo Plain (p. 58); see that type description
  for details.
SURFACE :
Color: Orange, tan, reddish brown; dark gray fire
clouds frequent.
Treatment: Superficially smoothed leaving tracks parallel
to the rim, scratches and other flaws. Sandy texture
resulting from protruding temper grains.
  Hardness:   3.
FORM:
Rim: Channel or exteriorly thickened, with flat or
tapered lip. Size of sample: 5.
Body wall thickness: Range 3.5-7.0 mm., increasing to
2.7 cm. at carination.
Base:   Probably rounded (Form C).
Reconstructed common  vessel shapes  (figs.   36,   38):   Forms
4, 7, 8, 9.
DECORATION (pi. 75):
Technique: Incision with a single-pointed tool on a plain
surface, subsequently covered with a thin white wash.
Incising tool unstandardized so that almost every
example differs in width, depth and form of lines.
Width 1.5-5.0 mm., typically with little variation on
a single sherd; however, one example has lines 1.5 mm.
wide adjacent to a zone containing 4 mm. wide grooves.
Depth is less than 1 mm. Wide lines tend to be shallow
and concave in cross-section, resulting in poor visibility
when the surface is uneven. Bed may be smooth or
filled with fine parallel striations. Terminations are
typically pushed up. Not equally spaced or evenly
parallel.
  Subsequent to incision, a white coating was applied
to the decorated zone. Erosion has reduced evidence
to scattered traces on most examples, suggesting that
it was never thick or evenly applied. This type is
difficult to distinguish from white filled incisions of
Tiputini Plain Incised (see p. —).
Motif:   Most examples feature a single or interlocking
squared coil, sometimes with a punctate at the center.
TEMPORAL DIFFERENCES WITHIN THE TYPE :   None observed.
CHRONOLOGICAL  DISTRIBUTION OF THE TYPE:   Absent from
the upper third of the seriated sequence (fig. 63).

Unclassified Decorated
  BLACK-ON-RED PAINTING (pi. 16a).—One body sherd
is decorated on the exterior with small black figures
on a thinly red slipped surface. Open spaces enclosed
by black lines were subsequentiy retouched with red
in the style of normal polychrome designs. Paste is
Napo Plain. One example from N-P-2, Broadside B.
  FINGER-PRESSED LIP.—A large (diameter 36 cm.)
deep bowl with outslanting walls has die lip decorated
by pressing with the finger, producing a continuous
series of deep impressions separated by high ridges
and with an undulating or scalloped border resulting
from displacement of the clay to the sides. Paste is
Napo Plain. One vessel from N-P-3, Test B extension.
  FINGER-PRESSED RIM (pi. 76 b-g).—Several rim
sherds from large vertical walled bowls have decora-
tion on the exterior rim thickening produced by
pressing with the finger from left to right, creating a
row of shallow, fingertip sized depressions separated
by high (5 mm.) curved ridges. The pottery type is
Napo Plain or Armenia Plain. Examples from N-P-2,
Area 2a and N-P-3, Surface and Test A, and Test B
extension.
  IMPRESSED RING (pi. 16h).—One fragment of an
open bowl widi rectanguloid rim lobes has a large
ring impressed from the exterior, leaving a conical
center. Exterior diameter is 4.3 cm., deptii 5-7 mm.
There is no distortion of die interior surface because
of thickening of the wall in this region. The pottery
type is Armenia Plain. From N-P-2, Broadside A.
  INCISED LINE (pi. 76o).—One sherd of a bowl widi in-
teriorly thickened rim, tapered to a rounded lip and
everted. Mouth diameter is 20 cm. Fine sand temper
with light orange-red surfaces and traces of a rudi-
mentary polish. Exterior decoration of fine incised
lines, 0.5 mm. wide, crooked but evenly spaced widi
intersections overshot. From N-P-3, Surface.
Trade Pottery
Several sherds from N-P-2 and N-P-3 are distinct in
paste, vessel shape and decoration from the ceramic
complex of the Napo Phase. They appear to be con-
temporary with the Napo Phase pottery, rather tlian
to result from later intrusion; if so, they must have
originated from trade.
  The paste is tempered with fine sand sometimes
containing abundant mica, visible as pinpoint
sparkles on the surface. Otiier sherds lack the mica.
One contains black ash. Surface finish is best on die
rim top and upper interior, poorest on die exterior,
which remains slighdy uneven.
Two principal kinds of decoration are represented:

78

SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6



(1) corrugation, beginning at the rim and extending
down the exterior of the neck in a series of fine, hor-
izontal, slighdy overlapping coils (pi. 76 i-m), or
occupying the waist ramer man the neck; and (2)
parallel, closely spaced incised lines forming bands and
zones alternately left plain and filled with trianguloid
punctations of distinctive form (pi. 16n). Both tech-
niques may occur on different parts of the same
vessel. In one case the corrugations are embellished
with shallow ovoid punctates (pi. 76m); another has
diagonal scoring (pi. 76/).
  Vessel shape is a wide mouthed jar with a prominent
rounded or angular shoulder, insloping, vertical or
outsloping neck and sharply everted rim with a flat
sloping top tapering to a flat or rounded lip. Interior
rim diameter is 14—24 cm., with one example having
a square rather than circular form (fig. 59).


I   i    I
I    I   I
0       1       2      3 CM
RIM    SCALE
■ i i i i i i
0      4      8      12 CM
VESSEL    SCALE
FIGURE 59.—Rim profiles and reconstructed vessel shape of sherds
of probable trade origin from Napo Phase sites.

THE SERIATED SEQUENCE AND ITS IMPLICATIONS

The seriated sequence of the Napo Phase is based on
trends in popularity of the plain pottery types as
indicated by their relative frequencies in the two
levels of N-P-2, Cut 1, the only stratigraphic excava-
tion representing the Phase. Analysis showed an in-
crease in the frequency of sand-tempered Napo Plain

and decreases in the frequencies of the other two
plain types, cariape-tempered Armenia Plain and
charcoal-tempered Tiputini Plain. Fourteen other col-
lections, representing the surface and tests at N-P-2
and five other Napo Phase sites were sufficiendy large
and unselected to be useful for seriation. The attempt



[
CARIAPE
N-P-l, AREA I
N-P-l, AREA 3
N-P-l, AREA  2
N-P-6, SURFACE
OASIS
N-P-l, AREA   4
N-P-5,   SURFACE
N-P-2, AREA   2A
N-P-3, TEST   B
N-P-2, BROADSIDE  A
N-P-2,   BROADSIDE   B
N-P-4,  SURFACE
N-P-3,  SURFACE  8r  TESTA
N-P-3, TEST   B   EXTENSION
N-P-2,  AREA I
N-P-2, CUT I, 0-15 CM
15-30
I   '   I   '   I   i   i   '   I   i   i
0    10   20   30  40   50%
   
I
SAND


0
□
CD
I
0
□
D
□
(=1
IZZ)

1      11	J1—11        1nII                                r     —i1               1i	iCHARCOAL

FIGURE 60.—Seriated sequence of Napo Phase sites derived from classification of both plain and decorated sherds by temper.

THE   NAPO  PHASE

79



to interdigitate them into the sequence, however,
brought to light several problems. First, the large
number of decorated pottery types into which each
sample was classified resulted in very small represen-
tation of most of the types, making trends difficult to
observe. Second, differences in frequency between
samples are relatively small, probably because a short
length of time is involved. As a consequence, an
attempt at seriation did not produce one ordering
that was clearly preferable to other possibilities. Many
of the samples appeared to fit equally well in several
alternative positions.
  Two supplemental methods of analysis were em-
ployed to evaluate the alternative seriated sequences
and determine which was most likely to be reliable.
One was the classification of all sherds by temper,
regardless of the surface finish, and calculation of the
relative frequency of the use of sand, cariape, and
charcoal (Appendix table 9). When the strips were
arranged in the order indicated by the percentage
distribution in the two levels of N-P-2, Cut 1, namely
increasing popularity of sand temper and decreasing
popularity of both cariape temper and charcoal tem-
per, the result is a relatively smooth pattern of change
in all three types (fig. 60). A second basis for analysis
was provided by the presence of two techniques of
incision, one executed with a single- and the other a
double- or multiple-pointed tool. All incised and ex-
cised sherds were sorted in terms of tool type employed,
and both relative and absolute frequency of the two
types was calculated for each sample (Appendix table
9). Seriation following the trend exhibited by the two
levels of N-P-2, Cut 1, again produced another rela-
tively smooth pattern of change (fig. 61), but resulted

in a different arrangement from that based on temper.
Since the reliability of different kinds of traits and
size of samples for chronological inference has been
debated (e.g. Lathrap, 1964; Evans and Meggers,
1964), this conflicting result is worth analyzing in
some detail.
  Correlation of the two seriated sequences (fig. 62)
shows most of the displacements to be minor. Four
collections occupy the same relative position in both
sequences (N-P-l, Area 4; N-P-2, Broadside B;
N-P-4, Surface; and N-P-2, Cut 1, Level 15-30
cm.), five are shifted upward or downward by only
one unit (N-P-l, Area 1; N-P-l, Area 3; Oasis;
N-P-2, Broadside A; N-P-3, Test B Extension), and
three are shifted upward or downward by two units
(N-P-l, Area 2; N-P-2, Area 2A; N-P-3, Surface
and Test A). One sample is displaced by three units
(N-P-6). These represent more than two-thirds of
the collections, and considering that the differences in
frequencies determining seriated position are often
less than one percent, these discrepancies are
insignificant.
  The four remaining collections are displaced mar-
kedly from one seriated sequence to the other. The
practice of seriation is based on the premise that
cultural traits change in frequency through time in
a systematic manner approximating a bell curve, and
all classes of traits should theoretically be expected
to produce a similar chronological result. In reality,
however, traits differ in their suitability for chrono-
logical studies. Some change too slowly to be useful,
others are too rare to appear in small samples or to
show a clear trend of changing frequency.
Insufficient  sample   size   may   explain   the   larger



N-P-l, AREA 2
N-P-l, AREA I
N-P-l, AREA 3
OASIS
N-P-2, CUT I, 0-15 CM
N-P-l, AREA 4
N-P-6, SURFACE
N-P-2, AREA I
N-P-2, BROADSIDE A
N-P-2, AREA 2A
N-P-2, BROADSIDE B
N-P-4, SURFACE
N-P-3, TEST B EXTENSION
N-P-3, TEST B
N-P-3, SURFACE  ft TESTA
N-P-5, SURFACE
N-P-2, CUT I,  15-30 CM

0
0
D
□
□
D
CD
CD
n

CD



SINGLE
LINE
0     10   20 30 40   50%

DOUBLE
LINE
ABSOLUTE    FREQUENCY

SINGLE
LINE

RELATIVE   FREQUENCY

DOUBLE
LINE

FIGURE 61.—Seriated sequence of Napo Phase sites derived from classification of incised and excised sherds by technique of incision.
293-822 O - 68

80

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



discrepancies in the Napo Phase seriation attempts.
In all the collections used for seriation, the total
sherds available for classification by temper exceed
250 except for one sample totalling 167 sherds (Ap-
pendix table 9). Since unselected samples of 100 or
more have been shown to be dependable for seriation
purposes (Ford, 1949, pp. 35-36), the arrangement
based on temper satisfies the requirement of adequate
sample size. By contrast, the total number of incised
and excised sherds exceeds 100 in only 9 of the 17
seriated collections. While most of the undersized
samples agree closely with the result obtained from
the larger number of sherds classified by temper, two
differ markedly: N-P-2, Area 1, represented by only
57 incised and excised sherds, and N-P-5, with only
38. In these cases, inadequate sample size is the
orobable explanation for the lack of correspondence
TEMPERING
MATERIAL
 TECHNIQUE
OF   INCISION
N-P-l, AREA I
N-P-l, AREA 3
N-P-l, AREA 2
N-P-6, SURFACE
OASIS
N-P-l, AREA   4
N-P-5, SURFACE
N-P-2, AREA  2A
N-P-3, TEST  B
N-P-2, BROADSIDE A
N-P-2, BROADSIDE B
N-P-4, SURFACE
N-P-3, SURF 8 TESTA
N-P-3, TEST B EXT.
N-P-2,  AREA  I
N-P-2, CUT I, 0-15
15-30

N-P-l, AREA 2
N-P-l, AREA I
N-P-l, AREA 3
OASIS
N-P-2, CUT I, 0-15
N-P-l, AREA  4
N-P-6, SURFACE
N-P-2, AREA I
N-P-2, BROADSIDE A
N-P-2, AREA 2A
N-P-2, BROADSIDE B
N-P-4, SURFACE
N-P-3, TEST B EXT.
N-P-3, TEST B
N-P-3, SURF. 8. TESTA
N-P-5, SURFACE
N-P-2,CUT I, 15-30
FIGURE 62.—Correlation of seriated sequences of Napo Phase sites
based on temper and on technique of incision and excision.
and the result obtained from the larger sample should
be the more reliable. The two remaining collections
(N-P-2, Cut 1, Level 0-15 cm. and N-P-3, Test B)
have sufficiently large samples in both classifications
to be considered dependable, so that a choice be-
tween them must be based on other considerations.
The relative validity of the two arrangements can
be assessed by construction of a third sequence
utilizing the pottery types rather than isolated modes
of paste or decoration. When this is done, the best
fit is achieved when the order indicated by the chang-

ing popularity of temper varieties is followed. Trends
in the plain types generally parallel those of the
temper seriation (cf. figs. 60 and 63), and the two
samples showing poorest fit (N-P-5 and N-P-6)
have been distorted by surface erosion. In each case,
the proportion of Napo Plain is larger and that of
Rocafuerte Painted is smaller than expected (fig. 63),
indicating that some sherds originally having
painted decoration are now plain.
  Few of the decorated types show notable or con-
sistent changes in frequency. In all cases, however,
the single-line technique represented by the Tiputini
Incised and Excised types dies out earlier than the
double-line technique represented by the Napo
Incised and Excised types. Only one pair of types
approaches the ideal pattern of successive dominance,
namely, Tiputini Red Incised, which expands and
then declines as Napo Red Incised grows more pop-
ular. In excised decoration, all slip variants tend to
disappear, leaving double-line excision on a plain
surface (Napo Plain Excised). Rocafuerte Incised,
characterized by incisions painted red subsequent to
white slipping of the surface, is absent from the
upper third of the existing sequence. Armenia White-
on-red and Napo Negative are present throughout,
as is Rocafuerte Painted, which is always the most
popular decorated type.
  All of the plain types are present and relatively
common at the beginning of the sequence. At the end,
sand temper (Napo Plain) occurs in 99 percent of
the sherds and charcoal temper (Tiputini Plain) is
represented by the remaining 1 percent (fig. 60).
Cariape tempering ceases to exist shortly before the
final levels.
  Only one collection made by previous investigators
was large enough to be classified and inserted into the
seriated sequence. This is the sample of 40 sherds
from Oasis in the Museum of the American Indian,
Heye Foundation. Analyses on the basis of paste
(fig. 60) and technique of incision (fig. 61) place it
in a late relative position. The selectivity for decorated
sherds made it necessary to correct the sample before
it could be seriated into the pottery type sequence. This
was done by computing the average frequency of
decorated sherds in unselected collections and in-
creasing the total used for percentage calculation to
include the equivalent number of plain sherds
(Appendix table 8). The result fits rather well into
the upper portion of the sequence (fig. 63).
  When the attempt is made to detect changes in
popularity of vessel shapes, the small size of rim
samples must be taken into consideration. Only
four collections included more than 100 rims, five
had between 50 and 100 rims, and six had less than
50 rims. As a consequence, when the relative frequency
  
N-P-l, AREA  I
N-P-l, AREA 3
N-P-l, AREA 2
N-P-6,   SURFACE
OASIS
N-P-l, AREA  4
N-P-5,  SURFACE
N-P-2, AREA  2A
N-P-3, TEST B
N-P-2,   BROADSIDE  A
N-P-2,   BROADSIDE   B
N-P-4,  SURFACE
N-P-3, SURFACE 8 TEST A
N-P-3, TEST  B  EXT.
N-P-2,  AREA   I
N-P-2, CUT 1, 0-15 CM
15-30

[

□
D
D
□
a
□
□
CCD

a
CDD
a
□
D
CD
1=1

I	I
D
CD
  CD
'	1
'	l

CD
D

I	—I


CD
□
D
D
D
0
□
□


0DD1D□
D1
Da
0CD
01DDDD1Da1□11aCDDDDDi□□
1□D
D101aDD010i111D□aD0
a
i
i
s
D
0
I
CD
□
D
□



	I.I.I010   20  30 40   50%
Scale A1,1.1,1,1,105     10    15   20  25%
Scale B- o
H X
=> UJ

Q
Q UJ
UJX
°S
xo
UJ l-
O X
UJ
CC UJ

=>  UJ

UJ
< tr

FIGURE 63.—Seriated sequence of Napo Phase sites based on changes in pottery type frequency (See Appendix table 8).





 L_
01I   i   i   i   I   i   I816   24   32 40 CM1   .   i   .   i   i  i   .   i   .   i01020  30 40   50%
N-P-l, AREA  3I00Da
D
0i	1D
D□
□  0
CDa
□
a□N-P-l, AREA 41=1DN-P-2, AREA 2ADai      ia□CDD□DaI	1N-P-3, TEST BCD]□aaaDCZDDDDN-P-2, BROADSIDE A
N-P-2, BROADSIDE  BoD□□00CD□□DCDD□
Da
CDDD1	1
CD□
CD□
□N-P-4,SURFACEDaN-P-3, SURF. 8 TEST A
N-P-3, TEST B   EXT.aaDaD□DCDai—ic=i
■i□
■■1	l
11	11	1
CD1	1
a
inDN-P-2, AREA 1CD1              1i    iN-P-2, CUT 1, 0-15 CM     p■15-30           Li1i11
I|■|    1
12345678910
VESSELII
FORMSI	—I
0
CD

CD
CD
0
CD
13


□CDDa□CDDCD»D0CDD14

D
CD
17

i
I
20

FIGURE 64.—Changes in relative frequency of Napo Phase Common Forms 1-20.   Sites are arranged in seriated order.



THE   NAPO  PHASE

81



of the 20 common forms is calculated and the levels
are arranged in seriated sequence, few trends emerge
(fig. 64). Most of the forms are present with similar
frequency throughout, or exhibit occasional erratic
fluctuations attributable to inadequacy of the sample.
All forms are present initially and remain in use at
the end of the sequence. Possibly significant increases
in popularity are exhibited by Forms 1 and 15, but
sampling error cannot be ruled out since the two
samples occupying the latest positions contain less
than 100 rims each.
  It should be kept in mind that evidence of chron-
ological orientation for the seriated sequence is limited
to the trends exhibited by the two levels of N-P-2,
Cut 1. The shallowness of this excavation makes its
reliability as a basis for determining the direction of
change open to question, and the possibility exists
that early and late portions are inverted. Inferences
derivable from changes in artifact types or relative
antiquity of sites are therefore tentative unless sup-
ported by other kinds of evidence.
  The three carbon-14 dates do not clearly validate
or refute the seriated sequence. Two come from
N-P-2 and are not in seriated order, although the
difference between them is only 11 years so that
their inversion is not likely to be significant. The
earliest, from N-P-2, Broadside B, is 782 ±53 years
or A.D. 1168 (P-347); the more recent, from N-P-2,
Cut 1, Level 0-15 cm., is 771 ±51 years or A.D.
1179 (P-269). The third date is from N-P-3, Test B
Extension, which seriates intermediate between the
other two collections. However, it is considerably
more recent than either at 470 ±180 years or A.D.
1480 (SI-299). All determinations were made from
organic carbon extracted from charcoal-tempered
sherds of Tiputini Plain, so that there is no obvious
basis for rejection of any date on the ground of prob-
able contamination or lack of association with the
Napo Phase. The fact that the two earlier dates were
furnished by the University of Pennsylvania and the
late one by the Smithsonian Institution introduces
a possibility that different laboratory procedures
may be responsible for the discrepancy. Numerous
independendy seriated collections show N-P-2 and
N-P-3 to be contemporary, so that either the single
N-P-3 date or the two N-P-2 dates must be rejected.

In view of the ethnohistorical evidence that the
region was uninhabited in 1521, a date of A.D. 1480
seems too recent to correspond to the lower portion
of the sequence, in spite of the short duration implied
by the relatively small amount of ceramic change
that took place. The earlier dates also agree better
with the date from Coari on the Solimoes (see pp. 94-95)
where ceramic resemblances to the Napo Phase occur.
  Several interesting observations can be made when
the composition and geographical location of the
sites are examined in terms of the seriated sequence.
First, quantitative results obtained for samples repre-
senting different portions of the same site are so
similar as to suggest that the entire area was occupied
simultaneously (figs. 28-29, 63). Taken in conjunction
with the extremely shallow nature of the refuse
deposits, this situation implies the existence of rela-
tively large villages for a brief span of time. Although
the site sample is too small to be conclusive, the fact
that the two smallest sites (N-P-5 and N-P-6)
occupy a relatively late position in the seriated
sequence suggests a diminution in village size.
  When the geographical location of the individual
sites is compared with their seriated position, no
clear pattern is evident. When contemporary occupa-
tions are considered, however, it can be seen that
the three earliest sites (N-P-2, N-P-3, N-P-4) are
farther downriver than the three later ones (N-P-5,
N-P-6, N-P-l) (fig. 3). The Oasis sample, which
seriates late, is the farthest upstream. On the other
hand, the small number of sites investigated allows
the possibility that this patterning may be fortuitous,
and the ambiguous result of attempts to seriate a
sherd sample from the Rio Yasuni (see p. 37) lends
credence to this view.
  Burial urn types and nonceramic artifacts are too
rare and accompanied by insufficient provenience
data to be analyzed in chronological terms.
  Pottery of possible trade origin occurred in sherd
collections from three locations at two sites: N-P-2,
Broadside A and Area 2, and N-P-3, Test B extension.
These collections occupy seriated positions in the
early half of the sequence, implying that communica-
tion with the makers of the alien ceramics dates
from the time of arrival of the Napo Phase in the
area.
  
DIAGNOSTIC FEATURES OF THE NAPO PHASE

Seven habitation sites in the 1956 survey represent
the Napo Phase. Six are on the Rio Napo and one on
the Rio Tiputini, where the elevation of the bank is
sufficient to escape flooding. Occupation refuse is dis-

tributed in a relatively narrow strip along the bank,
typically varying between 20 and 65 meters wide,
erosion by the river having in some cases reduced the
original width. Two sites are more than 500 meters

82

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



in length and the majority exceed 100 meters. All
are very shallow and disturbed by recent habitation,
so that the seriated sequence is based on trends of
ceramic change exhibited by only one stratigraphic
excavation of two 15 cm. levels.
  There is no direct evidence of subsistence pattern.
Numerous potrest fragments relate to cooking prac-
tices. Large griddles, some 40 cm. in diameter, suggest
the Napo Phase processed bitter manioc according
to the same methods used today by South American
Tropical Forest Indians. Shallow, flat bottomed ba-
sins of similar size perhaps had a related function.
Two fragments of what appear to be large manos, if
correctly identified, imply the processing of grain.
Consideration of the probable highland origin of the
Napo Phase suggests that this may have been maize.
Fishing and hunting must have contributed impor-
tantly to the diet, as well as the gathering of wild
plants.
  A number of anthropomorphic urns have been col-
lected from habitation sites, while others appear to
have been buried in isolated spots. Finds have been
accidental and few details are recorded. Bones have
been reported in some, and the small size implies
that the burial was secondary, although incomplete
cremation is also a possibility. The vessels vary in the
completeness and realism of anthropomorphic execu-
tion, and display considerable variety in rendition of
facial features. The surface is painted, sometimes com-
bined with broad incision.
  Information on dress and ornament is provided by
the anthropomorphic urns (pis. 55-65). Face painting
is characteristic, surrounding the eyes and extending
over the cheek and in some cases around the mouth.
Although similar, the patterns on right and left are
never exact mirror images. Ear lobes are sometimes
perforated. Hair may be gathered into a long queue
down the back. Both arms and legs have bulging areas
between constrictions suggesting use of ornamental
ligatures. A triangular relief at the center of the chest
may represent an ornament. Bodies are nude, but
three figures hold small circular shields. One of the
shields has geometrical ornamentation, while another
is painted with a panpipe having the tubes graduated
from both sides toward the center.
  The pottery of the Napo Phase has been classified
into 4 undecorated and 18 decorated types. Napo
Plain, tempered with sand, is by far the most abun-
dant type, constituting 35 percent or more of the
sherds in unselected samples. Two minor plain types
are Armenia Plain, tempered with cariape, and Tipu-
tini Plain, tempered with fine charcoal. Napo Red,
in which the surface is covered with a fine red slip,

has also been included among undecorated types.
  Decoration is by incision, excision, painting, and
rarely modeling on a plain surface, frequently com-
bined with a red or white slip. In the most elaborate
types, incisions or excisions have been colored to
contrast with the surface or slip. Examples include
Rocafuerte Incised, in which a white slip was applied
after incision and the incision subsequently colored
red, and Napo Red Excised, White Retouched, in
which the process was reversed so that the excisions
and incisions are accented with white after the sur-
face was slipped red. Two easily differentiated vari-
eties of incision reflect use of two kinds of tools, one
with a single and the other with a double or multiple
end. Different combinations of these two techniques
with different kinds of surface treatment account for
most of the proliferation of decorated types. The
remainder represent several kinds of painting: white-
on-red, negative, red-on-white, black-on-white, and
polychrome (red and black-on-white). The former
two are very rare, while the latter three are variants
of Rocafuerte Painted, which is the most popular
decorated type.
  Napo Phase vessel shapes include several varieties
of channel-rimmed bowls, large basins and small
open to constricted mouthed vessels with rounded or
markedly carinated shoulders. Larger vessels have
collared or channel rims. Bases may be flat, rounded
or annular. Outline is typically square. All forms and
decorative techniques abound in habitation sites.
The only form clearly not of domestic use is the
anthropomorphic  burial  urn.
  Pottery and stone artifacts are rare. Solid cylin-
drical potrests, occasionally ornamented with incised
and excised patterns, are the most common. The
design style on roller stamps of rolling-pin shape
suggests that they are probably associated with the
Napo Phase, although this has not been confirmed.
One small perforated pottery disk may be a spindle
whorl. Stone artifacts include pecked and incom-
pletely polished side-notched axes and rare eared
axes, chisels, grinding stone fragments, hammerstones,
and unshaped rocks used for abrading, hammering,
and cutting.
  The absolute chronological position of the Napo
Phase is attested by three carbon-14 dates. Two
place it in the latter half of the 12th century A.D.
The third, A.D. 1480, seems too recent to be accept-
able. The existence of pottery with zoned punctate
decorations and non-Napo Phase vessel shapes at
several sites implies trade relations with a group in
the Andean foothills to the west with a different
ceramic tradition.
  
The Cotacocha Phase
DESCRIPTION OF SITES AND EXCAVATIONS

This Phase is represented by a small sample of sherds
from four sites, most of which were known at the time
of collection in the field to be of recent origin. No
effort was made to locate more sites, or to accumulate
ethnographic material.
N-P-3: Nuevo Rocafuerte
An annular-based bowl with red-banded decoration
and cariape-tempered paste came from the surface,
where it had probably been thrown by one of the
modern residents after it broke. Its inclusion in the
Napo Phase sample was inadvertent, and no effort
was made to collect recent pottery from this modern
settlement (see pp. 34-35 for detailed description).
N-P-9: Cotacocha
In addition to the Napo Phase material from this site
(see p. 36), a large number of Cotacocha Phase
sherds were collected. Residents of the house occupy-
ing the site were absent at the time of our visit, so that
it could not be ascertained whether the pottery was
used by them or by prior occupants. The slight amount
of surface erosion in comparison with that exhibited

by the Napo Phase sherds is circumstancial evidence
of recency.
N-P-l4: Latas
A modern house is located on the left bank of the Rio
Napo, just below the first rapid (fig. 1). The spot is
strategic in river commerce, since portage is required
for all cargo intransit upriver (pi. 2d). The land rises
considerably a little in from the river bank, and the
slopes and summit are under cultivation. Two small
sherd concentrations were encountered on the top of
the first shelf of terrace, each about 5 meters in diam-
eter. Depth was no greater than 5 cm.
N-P-l 5: Tiputini Road
An area beside the road connecting Tiputini with the
airstrip has been used since the time of the Shell
Petroleum Company as a trash dump (same location
as N-P-l; see fig. 3). Among the cans and bottles was
a broken vessel of Form 1 with red slip on the rim and
exterior below the neck, probably discarded by one
of the Indian families in the employ of the Company
or of the Army base (see pp. 32-33).

ANALYSIS OF MATERIALS

No stone artifacts and only one possible pottery arti-
fact are associated with the Cotacocha Phase. The

latter  is  a small spool-shaped object from N-P-9,
about 4 cm. in diameter at the ends and 2.8 cm. in

83

84

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



diameter at the constricted waist. Both faces are
concave. The paste is cariape tempered.
Ceramic Classification
Cotacocha Phase pottery types are based on material
from only four sites, two of which produced a single
vessel each. In the absence of stratigraphic control, it
was not possible to test the relative value of ceramic
features for revealing temporal differences. Both tem-
per and firing were consequently used in the classifica-
tion of plain sherds. This results in two principal
sand-tempered plain types: completely oxidized Co-
tacocha Plain and incompletely oxidized Latas Plain.
A few cariape-tempered sherds have been segregated
into an unclassified plain type. Walls are thin and
surfaces are not even or polished. Decoration is simple
and largely a by-product of construction, consisting
as it does of unobliterated coil junctions and fingernail
marks left in the process of joining coils. Zoned red
slip or red bands also occur. Vessel shapes are few and
simple; only circular outlines were recognized.
  Type descriptions have been arranged in alphabeti-
cal order following description of vessel and base
forms. Information on provenience and frequency is
provided in Appendix tables 3 and 5.
Reconstructed Vessel Forms
Common Forms
1.	Jar with constricted neck and everted rim (fig. 65-1):
Rim: Insloping to nearly vertical or slighdy out-
sloping and mildly to sharply everted between 1.0
and 2.5 cm. below the lip. Interior of the bend may
be rounded or angular; exterior is rounded because
of thickening at the bend. Lower edge of the thicken-
ing is unobliterated, remaining as a line or elevation.
Exterior rim diameter, 18-36 cm.
Lip:   Rounded or flattened.
Body wall thickness:   Range 4-6 mm.
Base:   Pedestal (Form A) or flat (Form B).
2.	Wide mouthed jar with vertical rim (fig. 65-2):
Rim: Insloping, becoming approximately vertical 2.5-
3.0 cm. below the lip. Change in direction is angular
on the interior and observable as a slight step on the
exterior. The effect may be produced with or without
thickening. Rim diameter about 32 cm.
Lip:   Rounded.
Body wall thickness:   Range 3.5-5.5 mm.
Base:   Probably flat (Form B) or concave (Form C).
3.	Shallow bowl with annular base (fig. 65-3):
Rim:    Upcurving and direct. Diameter 22-24 cm.
Lip:   Rounded.
Body wall thickness:   Range 4—6 mm.
Base:   Annular (Form D).

Minor Rim Forms
1. Nearly vertical rim possibly representing a jar similar
to Form 2, with unsmoothed coils on the exterior.
Rim diameter 32 cm. (fig. 65-4). The upper coil is
decorated with fingertip impressions.
2. Insloping rim, everted 4 cm. below the rounded lip
producing a slightly outsloping collar. Rim diameter
32 cm. (fig. 65-5).
Base Forms
Four base forms are distinguishable. Two are correlated
with rim forms on complete vessels; the others may be
alternatives to any of the three principal forms.
A.	Pedestal, the flat bottom making a vertical, steplike
junction with the exterior. Pedestal height about
8 mm. Diameter 12 cm. (fig. 65-A).
B.	Flat, making an angular junction with the outsloping
body wall. The interior is a smooth curve. Diameter
12 cm. (fig. 65-B).
C.	Slightly concave and making a rounded junction with
the outsloping body wall. A slight convexity is de-
tectible on the interior. Diameter about 10 cm. (fig.
65-C).
D.	Annular, flaring, and thickened so as to create a
smoothly curving surface on the exterior. A small
concavity occurs on the interior. Diameter 6.5 cm.,
height 2 cm. (fig. 65-3).
Pottery Type Descriptions
Cotacocha Plain
PASTE:
Method of manufacture:    Coiling.
Temper: Fine angular sand, grain diameter typically
less than 0.5 mm. Color predominantly black, with
scattered white quartz and golden iron pyrites particles.
Abundant, giving speckled appearance to cross-section.
Texture: Fine grained, often laminated with fine cre-
vasses parallel or diagonal to surface.
Color:    Buff, light orange to tan throughout the cross-
section; rarely grayish brown.
SURFACE :
Color: Light buff, tan, grayish brown, often widi a
dusty hue. Dark gray to black surfaces rare, although
gray spots from fire clouding are typical.
Treatment: Best finished examples retain pits and un-
evenness; range like that of Latas Plain. Interior often
better smoothed than exterior.
  Hardness:    4.
FORM:
Rim: Everted and slighdy thickened, with rounded or
flattened lip. Size of sample: 6.
Body wall thickness:    Range 4-7 mm.
Base: Flat (Form B) and concave (Form C); probably
also pedestal (Form A).
Reconstructed common vessel shapes (fig. 65): Form 1: 100
percent.

THE   COTACOCHA  PHASE

85



OCCASIONAL DECORATION:    Most vessels have minor em-
bellishment of one or more of the following kinds:
1. Unsmoothed coils on neck, either left plain or orna-
mented on the lower half by a row of fingernail
marks (pi. 77 d-e).
2. Fingernail marks on the lower edge of exterior rim
thickening, overlapping slighdy onto adjacent sur-
face. Rarely, fingertip marks are substituted (pi. 77
a-c).
3. Horizontal row of punctates at base of neck, made
with stick or fingernail.
TEMPORAL DIFFERENCES WITHIN THE TYPE : None observable.
CHRONOLOGICAL POSITION OF THE TYPE:   No evidence.
Cotacocha Red
PASTE AND SURFACE :   On Latas Plain (p. 86) or Unclassified
Cariape-tempered Plain (p. 86); see those descriptions
  
for details.
FORM:
Rim: Direct or everted and exteriorly thickened;
rounded lip. Size of sample: 3.
Body wall thickness:   Range 4—6 mm.
Base:   Pedestal (Form A) or annular (Form D).
Reconstructed common vessel shapes (fig. 65):
Form 3: 66.6 percent (2 vessels).
Form 1: 33.3 percent (1 vessel).
DECORATION :
Technique: Application of a thin red coating to all or
part of the surface. Wide color variation from orange
to red to reddish brown because of unequal thickness
and poorly controlled firing. Application may be
before or after striated polishing; in the latter case,
slip is duller than adjacent surface.
Motif:   Two principal categories occur:








0    4     8    12 CM
VESSEL SCALE
0      I      2     3 CM
RIM 8i BASE SCALE

FIGURE 65.—Rim profiles and reconstructed vessel shapes of the Cotacocha Phase. 1-3, Common Forms 1-3. 4—5, Minor Forms 1-2.
A-C, Base Forms. Arrows and brackets designate decorated zones. (Black=undecorated, white=decorated, hachure=red slipped.)

86

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



1. Slipping of portions of the surface, such as the
body and rim leaving the neck bare, characteristic
of jars.
2. Curved red bands of finger width filling quadrants,
found on bowl interiors.
TEMPORAL DIFFERENCES WITHIN THE TYPE: None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE: Absent from
N-P-l 4.
Latas Plain
PASTE :
Method of manufacture: Coiling; coils often unobliterated
on the exterior for ornamental effect.
Temper: Fine angular sand, grain diameter typically
less than 0.5 mm. Color predominantly black, with
scattered white quartz and golden iron pyrites par-
ticles. Abundant giving speckled appearance to cross-
section.
Texture: Fine grained, often laminated with fine cre-
vasses parallel or diagonal to surface.
Color: Typically medium to dark gray throughout the
cross-section; rarely light gray. Oxidation may be
limited to the surface or penetrate to 2 mm. inward
in a band of uniform or fluctuating width along both
surfaces.
SURFACE :
Color: Light buff (whitish), light tan, light orange,
brownish orange, grayish brown, dark gray to black;
color may vary gready over a small area because of
fire clouding. Interior and exterior may be similar
shades or at opposite ends of the range.
Treatment: Smoothed superficially leaving unevenness
and small defects; sometimes striated polished leaving
parallel horizontal smoothing tracks and producing a
more compact surface; never polished enough to
create luster. Interior tends to be better smoothed
than exterior.
Hardness: 4.
FORM:
Rim:    Everted  and  slightly thickened  on the  exterior,
  
with flattened or rounded lip. Size of sample: 9.
Body wall thickness:   Range 3.5-5.0 mm.
Base:   Probably pedestal  (Form A), flat (Form B), or
  concave (Form C).
Reconstructed vessel shapes (fig. 65):
Form 2:	55.6 percent.
Form 1:	22.2 percent.
Minor Form 1: 11.1 percent.
  Minor Form 2: 11.1 percent.
OCCASIONAL DECORATION (pi. 77/-/):   Unobliterated coils
and fingernail marks are the same as on Cotacocha Plain
(p. 84); see that type description for details.
TEMPORAL DIFFERENCE WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL POSITION OF THE TYPE :   No evidence.
Unclassified Cariape-tempered Plain
PASTE:
Method of manufacture:    Coiling.
Temper:    Cariape, in the form of whitish "bundles" up to
5 mm. long, visible on the surface or in cross-section.
Black ash and sand also occur,  the latter in smaller
proportion than in Cotacocha Plain and Latas Plain.
Texture:   Abundant fine air pockets, often occurring at
ends of longer temper particles; orientation parallel or
perpendicular to surface. Breaks with a very uneven
edge.
Color:   Dark gray or brown throughout the cross-section,
or pardy gray and partiy brown.
SURFACE :
Color:   Tan to dark gray; large black fire clouds.
Treatment:    Smoothed leaving broad (about 2 mm. wide)
  tracks, but with large pits and defects remaining.
Hardness:    2.5.
FORM:    Only body sherds are included in the sample; wall
thickness 5-10 mm.
TEMPORAL DIFFERENCES WITHIN THE TYPE:   None observ-
able.
CHRONOLOGICAL   POSITION   OF   THE   TYPE:   Absent  from
N-P-l 4.

THE SERIATED SEQUENCE AND ITS IMPLICATIONS

The small size and nonstratigraphic origin of
pottery samples from Cotacocha Phase sites makes
diachronic analysis of the archeological remains
impossible (Appendix tables 3 and 5). However,
both distributional and historical evidence suggest
that the easternmost sites reflect the recent spread
of Quechua speaking acculturated Indians down the
Napo in very recent times.
  Pottery resembling  that of the  Cotacocha  Phase
both in vessel shape and in embellishment by un-

smoothed coils, fingernail marks and zoned red slip
occurs archeologically in the vicinity of Cotundo in
the Andean foothills (fig. 1). In this area these
elements have some time depth, perhaps extending
backward into the pre-European period.
  The recent spread of the Quechua language down
the Napo has been noted by Ferdon (1950, pp. 4-5).
When Alfred Simson travelled in eastern Ecuador
before 1886, he found Quechua speaking Indians
only as far east as the mouth of the Rio Coca (ibid.).
  
THE   COTACOCHA  PHASE

87



Since then their area of distribution has extended to
the Peruvian border and possibly beyond. Ferdon
questions   whether   this   represents   peopling   of   an

uninhabited area or displacement of aboriginal pagan
groups by Christianized Quechua speakers. Existing
archeological evidence favors the former alternative.

DIAGNOSTIC FEATURES OF THE COTACOCHA PHASE

Pottery classified as belonging to the Cotacocha
Phase was collected at four locations along the
Rio Napo, from Latas to near the Peruvian border,
and equates with the historic settlement of the region.
The sparsity of sherds and small number of vessels
represented reflect both the scattered homestead
pattern of modern settlement and the adoption of
nonceramic types of containers.
  The ceramic complex includes two principal sand-
tempered plain types: Cotacocha Plain with com-
pletely   oxidized   firing,   and   Latas   Plain   with   in-

completely oxidized firing. A few cariape-tempered
sherds are associated. Unsmoothed coils and finger-
nail marks occur on the exterior of most plain vessels.
Walls are relatively thin but surfaces are not even
or polished. Decoration is restricted to red bands or
zones unbounded by incision. Vessel shapes are
simple utilitarian forms, including open bowls some-
times with annular base, and globular-bodied jars
with insloping to vertical neck and everted rim.
  The only artifact associated with the Cotacocha
Phase is a spool-shaped object of cariape-tempered
pottery of unknown use.
  
The Rio Napo Cultural Sequence and Its Implications

An attempt to trace the origin and affiliations of the
archeological phases identified on the Rio Napo and
its tributaries is a fascinating and frustrating task. It
is fascinating because two of the four phases can be
equated with complexes that are widely dispersed over
northern South America; it is frustrating because so
little is known of vast regions, and hypotheses re-
garding paths of diffusion must be erected on the
flimsiest of evidence. Even in the present limited
state of knowledge, however, a picture can be sketched
that may serve to focus attention on areas where
archeological investigation is urgentiy needed.
  
The time that man first set foot on the eastern
Ecuadorian lowlands may always be a matter of in-
ference from evidence in the adjacent highlands
rather than local finds. The combination of dense
vegetation, intermittently flooded terrain, riverbank
erosion, sediment deposition and dike formation, and
availability of perishable material for tools suggests
that the probability of preceramic sites being pre-
served, much less discovered, is infinitesimal. However,
it seems reasonable to assume that hunters explored
the lowlands along with the highlands as they peopled
the continent at least 10,000 years ago and perhaps
much earlier.
  
THE YASUNI PHASE

The archeological record on the Rio Napo begins
with the introduction of pottery making. The earliest
known ceramic complex belongs to the Yasuni Phase,
and although negative evidence is never conclusive,
its affiliations to early Formative complexes in the
Andean region make it likely that it represents the
first sedentary pottery-making immigrant group to
settle on the banks of the Rio Napo. A single carbon-14
date places this occupation at 2000 ±90 years ago, or
50 B.C. (SI-300). Two habitation sites were investi-
gated, both shallow and relatively small in area.
A considerable number of distinctive elements char-
acterize the ceramic complex and are useful for tracing
the affiliations of the Yasuni Phase. In decoration,
these consist of zoned incision (subdivided into fine
cross hachure, broad cross hachure, fine parallel ha-
chure, broad outiine, patterns of narrow bands or
88

irregular zones) on the flat rim top or exterior wall,
broad-line incision, zoned punctation, dot-ended line,
dotted line or drag-and-jab punctation, rim lobing,
applique rib, nicked rim, and zoned red. Vessel shape
details include Yasuni Phase Common Forms 1-9
and features such as angular carination, rounded
shoulder, sublabial flange, and annular base.
  Examination of the geographical occurrence of
these diagnostic Yasuni Phase traits of decoration
and vessel shape reveals a remarkable pattern.
Complexes possessing a large number of the traits in
question (fig. 66) are Waira-jirca and Tutishcainyo,
located in the highlands and eastern lowlands of
Peru, respectively; Macas in eastern Ecuador; Puerto
Hormiga on the Caribbean coast of Colombia;
El Mayal and Rio Guapo styles on the Venezuelan
coast, and the Jauari and Ananatuba Phases on the
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

89




CERAMIC TRAITSCOMPLEXESPuerto Hormiga
Waira-jirca
Jauari
Ananatuba
Rio Guapo
El MayalTutishcainyo
Macas
YasuniDECORATION
Applique rib
Broad-line incision
Dot ended line
Dotted line
Excision
Nicked rim
Nicked shoulder
Rim adornos
Rim lobing
Zoned incision:
Fine cross hachure
Broad cross hachure
Fine parallel hachure
Narrow bands
Irregular zones
Broad-line border
On flat rim top
Zoned punctation
Zoned redxxxxxx
x    ?                         x
X
X      X
XXX
X                X                XX
X                                    XX
            XXXX
X                          X
X       X       X       X       X
xxx           ?
xxxxxx
xxxxxx
x           x
X      X
XX               X
XXX
x           x
X                X
XXX
X                X
X                X
X      X
X      X
X                X
XX?
XXX
XXX
XX?
X       XVESSEL SHAPE
Yasuni Phase Forms
1
2
3
4
5
6
7
8
          9
Bottle
"Tecomate"
Angular shoulder
Rounded shoulder
Waist flange
Sublabial flange            X
X                XX
X
?
X                          XX
xxxxxx
X                                    X
xxxxX               X
X                  X
            X
X                X
            X
XXX
X                X
X                X
X      X
X       ?
X      ?
X               X
X
X               XOTHER FEATURES
Shell temper
Tubular pipe
Shell midden            X             x
X                X
X                 X                XXXFIGURE 66.—Occurrence of Yasuni Phase ceramic traits of decora-
tion and vessel shape in other complexes representing the Zoned
Hachure Horizon Style (X = present; ? = identification uncertain).
Information is derived from the following sources: Macas,
Bushnell, 1946; Tutishcainyo, Lathrap, 1962; Waira-jirca,
Izumi and Sono, 1963; Puerto Hormiga, Reichel-Dolmatoff,
1965b; Rio Guapo and El Mayal, Cruxent and Rouse, 1958;
Jauari, Hilbert, 1959a; Ananatuba, Meggers and Evans, 1957.

lower Amazon. Only Macas, in the southeastern
Ecuadorian highlands, is in relatively close geo-
graphical proximity (fig. 67). Except on the Vene-
zuelan coast, each complex is believed to be the
earliest in the local ceramic sequence, and where
carbon-14 dates have been obtained, they support
this placement. The greatest antiquity has been
established at Puerto Hormiga, Colombia, with five
dates ranging from 4502 ±250 (1-1123) to 5040 ±70
(SI-153) years ago (Reichel-Dolmatoff, 1965b, p. 53)
or 2552 to 3090 B.C. Next in antiquity is the Waira-
jirca Phase at Kotosh, Peru dated between 3180 ± 130
(N-69-2) and 3800±110 (GAK-262) years ago or
1230 to 1850 B.C. A charcoal sample from an
Ananatuba Phase site, collected by Mario F. Simoes
(n.d.), has been dated at 2930±200 (SI-385) years
or 980 B.C. The two Venezuelan occurrences, El
Mayal and Rio Guapo, are placed by Cruxent and
Rouse (1958, pp. 101, 121) in late Period II on stylistic
grounds. A carbon-14 date for El Mayal gave 1795±
80 (Y-297) years ago or A.D. 155 (op. cit. p. 15).
  In 1961, in an attempt to reconstruct on a broad
scale the prehistory of the Tropical Forest Area, we
proposed the existence of a Zoned Hachure Horizon
Style represented by the Tutishcainyo, Yasuni,
Jauari, and Ananatuba complexes (Meggers and
Evans, 1961, pp. 375-378). In the absence of carbon-
14 dates, and unaware of the existence of the Waira-
jirca and Puerto Hormiga complexes, we postulated
an Andean origin for this style. More recent informa-
tion now permits a reexamination of its possible
origin and dissemination.
  If these complexes are derived from the north coast
of South America, as the priority of the Puerto Hor-
miga dates implies, their wide geographical and chron-
ological separation would be expected to result in
stylistic differences. This is indeed the case. However,
the patterning of these differences does not coincide
with the obvious natural routes connecting known
sites. A brief review will illustrate the problem.
  Puerto Hormiga pottery, which is dated so much
earlier than the other members of this group that it
constitutes a potential source of the tradition, com-
bines zoned hachure and broad-line incision with
adornos of an incipient Barrancoid style. This com-
bination of traits makes Puerto Hormiga (or a complex
derived from it) a logical predecessor for both the
southern and eastern representatives. On the Ven-
ezuelan coast by the time of El Mayal, zoned hachure
is a minor decorated technique and modeling has
become dominant (Cruxent and Rouse, 1958, pi. 44).
The minimum lapse of some 2700 years between El
Mayal and Puerto Hormiga allows sufficient time for
much alteration both through cultural drift and
through   influence   from  local   Venezuelan   ceramic
  
90

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



styles, so that the survival of zoned hachure and other
traits relating to the Yasuni Phase even to a minor
degree is rather remarkable. This tradition continues
into the lesser Antilles, where zoned hachure and
adornos are major elements in the earliest ceramic
complex, appearing about the beginning of the
Christian era (Bullen, 1965, p. 240).
  Diffusion toward the south and the Peruvian high-
lands appears to have been accompanied by cultural
drift in the opposite direction to that along the Ven-
ezuelan coast. In the Waira-jirca Phase, modeling is
rudimentary or absent and zoned hachure has become
the dominant decorative technique (pi. 78 a-l). The
Puerto Hormiga trait of applying red pigment to
incisions subsequent to firing is frequently used to
enhance zones of hachure. Excision has also been re-
tained as a minor decorative technique. Typical
Waira-jirca Phase vessel shapes are rounded jars with
constricted mouths like those of Puerto Hormiga.
  When the effort is made to fit the Tropical Forest
occurrences into this dichotomy, the pattern becomes
less distinct. Along the lower Amazon, for example,
the Jauari Phase combines bulbous-eyed rim adornos
with zoned hachure, aligning it with the Caribbean
complexes. This affiliation is strengthened by the fact
that the site is a shell midden (as are Puerto Hormiga,
Rio Guapo, and El Mayal), and by the presence of
tubular pipes, which also occur at Rio Guapo and
El Mayal. By contrast, the Ananatuba Phase on
Marajo Island at the mouth of the Amazon lacks
both modeling and tubular pipes, although the
zoned hachure decoration is identical in technique
and motif to that of Jauari Phase ceramics (cf. Meggers
and Evans, 1957, pis. 38-41, and Hilbert, 1959a,
pp. 13, 15). This situation suggests that the Jauari
Phase is earlier and that the absence of certain traits
from the Ananatuba Phase is the result of simplifica-
tion during diffusion, a frequently observed effect.
These traits set the lower Amazonian occurrences
apart from those in the west, and suggest that their
relationship may be an indirect one, stemming from
remote common ancestry, rather than a direct one
resulting from west to east diffusion down the Amazon.
If the carbon-14 dates are accepted, the greater
antiquity of the Ananatuba Phase also rules out
derivation from the western lowland complexes
(Tutishcunyo and Yasuni).
  The three eastern Andean complexes form a some-
what more consistent unit. Although the Tutishcainyo
and Yasuni Phases are at least 850 kilometers apart
in a direct line, and much more widely separated if
natural river routes between them are considered
(fig. 67), resemblances in both vessel shape and
decoration are numerous. Similarities in decorative
elements   are   particularly   striking,   considering  the

poor surface condition of the Yasuni Phase material.
Duplications include parallel stepped lines, drag-
and-jab ("dotted") lines, incision terminating in
punctation, a row of nicks at the lower edge of the
decorated zone, sublabial flange with nicked edge,
squared coils, and of course zoned hachure (pi.
78 m-t; table C; Lathrap, 1958, fig. 1). Vessel shapes
are more divergent since the strongly carinated
shoulders and broad flanges characteristic of the
Tutishcainyo complex do not appear to occur in the
Yasuni Phase. However, Yasuni Phase Forms 1, 2, 4,
and 8 (and possibly 6 and 7) resemble Tutishcainyo
vessel shapes, and rims of Yasuni Phase Rare Forms
1, 2, 5, 6, and 8 are represented in the Early
Tutishcainyo   sample   (Lathrap,    1962,   table   142).
TABLE C.—Occurrence of Yasuni Phase decorative elements in Tutishcainyo
Phase pottery (figures refer to Lathrap, 1962)

PhasesTraitsYasuniEarlyLateTutishcainyoTutishcainyoParallel stepped linesXFig. 31 f-hFig. 50cDrag and jab linesXFig. 23 e, 25Fig. 43/, hSquared coilXFig. 29cFig. 4l£Elongated rectangular zonesXFig. 25created by vertical connectionsbetween straight parallel linesIncision terminating in punctateXFig. 28 d-eHorizontal parallel lines on rimXFig. 414/exteriorNicks on lower rim thickeningXFig. 41<fRow of nicks at lower edge ofXFig. 44£decorated zoneVertical applique ribsXFig. 44 d-fSublabial flange with nickedXFig. 46/decorationProtuberances at shoulderXFig. 41 h-jWhen the chronological distribution of the shared
traits in the Tutishcainyo sequence is examined
(table C), the closest resemblance is with the Late
Tutishcainyo complex. Although there are no carbon-
14 dates available for Late Tutishcainyo, its antiquity
has been estimated by Lathrap (1965) as prior to
1000 B.C. The Yasuni Phase date of 50 B.C. (SI-300)
is considerably more recent than might be expected
in view of the numerous similarities between the
pottery of the two complexes.
  The most accessible source for the Tutishcainyo
ceramic complex is the Waira-jirca Phase, the earliest
component of the stratigraphic sequence established
at Kotosh in the central Peruvian highlands (fig. 67).
However, although numerous detailed correspond-
ences  can  be  observed  in the zoned-hachure tech-
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

91



V


.1        J         1/       ^*4y>     l,iv , #:J.1*                              CM                             O(^yV\lfe%   "^ssg^oCO
■a--cto.  SUIfUDOOl ^VNSv           '^S^teO*';/""V** ^/7<o|/	JV     |5s5	                     »             ^V/,.'--b_/    f w;z as ft^J    \   NVA< CTV«|   y%^-y>n^^r\  ^A'"%      )yA    y.    H5 «:vS//   ^vxA'''.''^•r A           j—.                  .^vCM	'       ^—%5 y?" -.y'A         \<.      \\f-             /^                \/—v. c"    7fii*           \>\^/>-«   "iiii?*^^"'^)\ *';''—_^^s      ^p^s*     7jS*   —\ '^^-^L-^—v^*.<^I**«.   or:     /w-*^""   '^'■■^„.y<      #■&..,  \M..,I(L^    \?'
iiT:,..      '■>_••>  \>f$"* <Tw      *"V>«w	                             /-^     'V      ^i                                                   .          „v-^^\                  ^jlMn.'i.                                      f-N.        ic-!,„._                                                                   •>..„••>*x       ID"■■y    4„,y        " J-*yy
Ayy y?        ^^A00
in^-ylfF'ffe x  ^ ,\^y^^     ^^^^^^..'"c-y,„.4 ^"'"y       -'"..*v> '•■.*■:.;-* k*vfe   -*. -^jSsV      -s;""           \(    \\                h''    %«OL ^-v5'""i ,,y"""'"A,,■FV.                      i,u.t^'rv         \            11%	»!7D
O

/OV^
Xt

'j*1*; §'"*, >,.

JHriE


^""i      A      O^OII.,,   m^be"\i^!,""C%o^J ^H'X.'»i«Wi,J3^^^^Nfca.CO&..         ?.f«■sy,,^c  ;^^y-Jc"t*Americl^*"rSsoocc

y-.

•5^
3 I J I 3 ¥

/;

V?i:

92

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



nique and motif, vessel shapes are less similar. Also,
characteristic Tutishcainyo Phase decorative tech-
niques, such as ornamental rim treatment and
drag-and-jab or "dotted" lines substituting for in-
cision, are absent from Waira-jirca Phase pottery. A
pronounced difference in application of zoned hachure
exists in the emphasis on discrete rectanguloid panels
on Tutishcainyo pottery, while Waira-jirca designs
are dominated by continuous narrow bands (Izumi
and Sono, 1963, pis. 81-83). Characteristic features
of Tutishcainyo Phase vessel shapes, such as broad
horizontal rims and flanges (Lathrap, 1962, figs.
20-29), are not represented in Waira-jirca Phase
ceramics. These differences imply the existence of
a time gap between the two phases. The Waira-
jirca Phase has been dated by carbon-14 between
3180 ±130 (N-69-2) and 3800±110 (GAK-262)
years ago, or 1230-1850 B.C. One date for a complex
at site PAC-14 on the Rio Nazaratequi, originally
said to have affinities with Late Tutishcainyo, is
1346±110 years or A.D. 604 (P-995), while another
sample from 30 centimeters deeper in the same
excavation gave 2620 ±100 years of 670 B.C. (Y-
1546). Lathrap (pers. commun.) subsequently re-
jected these dates as relevant to Tutishcainyo, and
continues to support an antiquity of some 4000 years
(Lathrap, 1967 p. 17) for the latter. Although specula-
tion will continue until dates for the Tutishcainyo Phase
are obtained, the amount of divergence from Waira-
jirca suggests that it may be more recent than has
been supposed.
  The affiliations between Yasuni Phase and Macas
pottery are more difficult to evaluate because of de-
ficiencies in the size and condition of both samples.
A majority of the Macas sherds is decorated with
zoned red, which cannot be identified on Yasuni
Phase pottery because of the eroded condition of
the surfaces. However, many decorative features are
shared and Macas vessel shapes also emphasize
everted or exteriorly thickened flat-topped rims,
often with lobed lips (Bushnell, 1946, fig. 1 a-o;
Collier and Murra, 1943, pi. 7).
  An attempt to trace affiliations of Yasuni Phase
pottery with other Ecuadorian complexes is inhibited
by the vacuum that exists in the data from the high-
lands for the early time period. On the Ecuadorian
Pacific coast, similarities can be noted between some
of the Yasuni Phase rim forms and those of the
Machalilla Phase. The most significant may be
Yasuni Phase Form 9, which has an unusual cam-
bered rim like that of Machalilla Phase Generalized
Form   14  (Meggers,   Evans,  and  Estrada,   1965,  p.

142). However, the more characteristic flat-topped
and exteriorly thickened Yasuni Phase rim varieties
are not present on the coast in this early period, and
decoration is also different. Zoned hachure occurs
in coastal Formative complexes, it is true, but it is
executed with a single width of line, setting it apart
from the broad border and fine hachure diagnostic
of the Yasuni Phase and of the Zoned Hachure
Horizon Style in general. Zoning of the latter kind
appears in the pottery of Period D of the Valdivia
Phase, also on the Ecuadorian coast (Meggers, Evans,
and Estrada, 1965, pis. 113 1-s; 114 a-p), but the
relatively late date in comparison to that on the
north coast of Colombia suggests that this may be
derived from a highland representative of the Zoned
Hachure Horizon Style rather than serving as a
possible source.
  This review of the distribution of  decoration by
zoned  hachure,   in  which  the  border   incisions  are
wider than those employed for texturing, raises more
problems than it resolves.  Clusterings of associated
elements, particularly vessel shape, as well as carbon-
14 dates now available,  permit recognition of two
paths of diffusion from a  north Colombian center.
One of these moved southward, probably through the
Andean   highlands,   reaching   the   Huallaga   valley
after a lapse of about a millenium (fig. 67). Another
moved more slowly eastward along the Venezuelan
coast, surviving in places along the eastern coast and
in the Lesser Antilles after more than two and a half
millenia.   Postulation   of  a   third   emination   either
southeastward from the north Colombian center or
splitting off from one of the other paths seems nec-
essary to account for the relatively great age of the
lower Amazonian Ananatuba Phase,  as well as its
trait inventory, which deviates from those of the more
recent representatives of the two principal dissemina-
tions. Although one can speculate that its introduc-
tion may have proceeded down the Japura or the
Negro, such hypotheses are essentially meaningless.
It is obvious that the distribution of the Zoned Ha-
chure Horizon Style cannot be interpreted satisfac-
torily   until   additional   evidence   has   accumulated
from portions of northern South America intervening
between known occurrences, and a larger number of
carbon-14  dates  becomes  available.   The  fact  that
no   pottery   of   this   horizon   style   is   included   in
Nimuendajii's sherd samples from more than 75 sites
on the middle and lower Amazon, however, suggests
that filling of the distributional gaps may not be an
easy task.
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS
THE TIVACUNDO PHASE

93



A gap of five centuries separates the carbon-14 date
for the Yasuni Phase occupation of the Rio Napo
from that of the next pottery-making group identified
by our survey. This is the Tivacundo Phase, rep-
resented at two small sites on the Rio Tiputini. One
of these, N-P-8, has been reduced to a remnant by
cutting of the river bank, although a site of the Napo
Phase (N-P-6) a few meters away is still largely intact
(fig. 15). Before carbon-14 dates were obtained, this
circumstance was interpreted as evidence that the
Tivacundo Phase occupation was the earlier of the
two, an inference also supported by the greater degree
of deterioration of Tivacundo Phase pottery surfaces.
This relative chronology has been confirmed by a
carbon-14 date of 1440±70 years or A.D. 510
(SI-330) obtained for N-P-7, the more recent of the
two Tivacundo Phase sites in the seriated sequence.
  The affiliations of the Tivacundo Phase are obscure.
The zoned red decoration is similar in technique to

that reported from Macas and the southern Ecua-
dorian highlands, but details of execution and motif
are different. The relatively broad-line incision used
to define red zones on Macas sherds contrasts with
the extremely fine incision characteristic of Tivacundo
Phase decoration. The painting of a red band on top
of some fine incisions is a Tivacundo Phase char-
acteristic conceptually similar to the red-retouched
incision of the Napo Phase, but in the latter case red
was applied to the bed of broader lines. Vessel shapes
are unrelated to those of any other complex so far
described in the region. Simple rounded bowls and
neckless jars are typical, as is a slight modification of
the rim giving the effect of beading. Oval as well
as circular outiine occurs. In spite of the distinctive
character of both vessel shape and decoration, no
comparable material can be cited either from the
highlands or from the lowlands.

THE NAPO PHASE

Identification of the Napo Phase as the third or next
to latest in the relative chronological sequence on the
Rio Napo is based on circumstantial evidence and
carbon-14 dating. The former consists of the physical
location of N-P-6 with relation to N-P-8, which
has just been reviewed, and the lesser degree of
deterioration exhibited by sherd surfaces in comparison
to samples of Yasuni and Tivacundo Phase pottery,
implying a briefer exposure to the deleterious effects
of alternately soaked and baked acid soils. The carbon-
14 dates, obtained from carbon extracted from sherds
of Tiputini Plain (Evans and Meggers, 1962). are in
agreement with this relative position. The two most
acceptable are 782 ±53 years or A.D. 1168 (P-347)
and 771 ±51 years or A.D. 1179 (P-269). The third,
which seems too recent (see p. 81), is 470 ± 180 years
or A.D. 1480 (SI-299).
  The seriated sequence of Napo Phase sites indicates
that the ceramic complex was introduced into eastern
lowland Ecuador in a fully developed condition.
Since all of the decorative variants and vessel shapes
are present in the earliest levels (figs. 63-64), all can
be used for tracing the origin of the Phase. However,
the task is complicated by lack of evidence from the
adjacent  lowlands.   No  archeological  remains  have

been described between eastern Ecuador and the
vicinity of Tefe, Brazil, some 1900 kilometers down-
river or about 1200 kilometers in a straight line
(fig. 68), although Lathrap (1967) has reported
discovery of several sites around an ox-bow lake on
the right bank of the Amazon a littie upstream from
the Colombian town of Letitia. A similar void extends
850 kilometers to the south, as far as Pucallpa on the
Rio Ucayali in the eastern Peruvian lowlands. The
entire eastern portion of Colombia is also virtually
unknown archeologically.
  The highland situation is not much better. While
the north highlands of Peru are poorly reported, what
has been described fads to include diagnostic Napo
Phase elements. Highland Ecuador has also been
superficially investigated, but Jijon y Caamano who
had a lifetime of familiarity with the archeology once
observed (1951, p. 377) that "the civilization on the
upper Napo, near the junction with the Curaray, is
completely distinct from those of the Ecuadorian high-
lands and coast." By contrast, although the highlands
of Colombia are also poorly known, existing data
attest the presence of many Napo Phase character-
istics. These are best represented in the Quimbaya
region of the upper Rio Cauca (figs. 68, 79; Cubillos
  
94

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6



and Bedoya, 1954; Duque, 1963a). A lesser number
occur in the San Agustin area. Traits include sand
temper, details of vessel shape such as thickened
carination, square outiine, annular base, anthropo-
morphic treatment, and possibly channel rim, and
decorative techniques such as excision, white retouch,
single-line incision, red and black-on-white painting,
negative painting and white-on-red painting. Urn
burial becomes a typical method of disposal of the
dead throughout Colombia by 1000 A.D. (Reichel-
Dolmatoff, 1965a, p. 136), and is reported several
centuries earlier in the San Agustin region (Duque,
1963b, fig. opp. p. 106). Anthropomorphic urns often
have limbs modeled in the round to show swellings
resulting from ligatures like those of Napo Phase
examples. A vessel from Manizales (Uhle, 1889, pi. 2,
fig. 9) resembles Napo Phase Form 6. Unfortunately,
these traits are poorly defined both temporally and
spacially, although present knowledge suggests that
they are all extant at a sufficientiy early time to
antedate the Napo Phase. Their area of distribution
does not appear to extend to the southern Colombian
highlands.
  By the time of its arrival in eastern Ecuador, the
Napo Phase had acquired several ceramic features
not reported in the Colombian highlands. Most dis-
tinctive of these is tempering with cariape, obtained
by burning the siliceous bark of certain species of
trees. Other additions of possible non-highland origin
include collared rims, lobed rims, grooving, double-
line incision, and the production of pseudo-negative
designs by positive painting. The application of a
white slip or wash subsequent to incision is also un-
reported in the highlands, along with red retouch of
incisions on a previously white slipped surface. Several
of these traits are widespread in the Amazon basin;
others are of limited distribution, if present evidence
can be relied upon (fig. 68). Their incorporation into
the Napo Phase complex is-significant in that it sug-
gests an indirect route from the highlands to the Rio
Napo, an implication also inherent in the northerly
distribution of Napo-like traits in the Colombian
highlands.
  Tracing of this route is impeded by the absence of
information from eastern Colombia and the upper
Amazon. Most of the Napo Phase traits have been
reported on the middle Amazon between the Rio
Japura' and the Rio Negro, but here two features
absent in the Napo Phase are characteristically as-
sociated—adornos and flanges. A small collection of
sherds from the Rio Guepi, a tributary of the Rio
Putumayo (fig. 3), incorporates the temper variants
characteristic of the Napo Phase (cariape*, black ash,
and sand) as well as sherd temper, which makes its
appearance  on  the Amazon  at Manacapuru.  Un-

fortunately, the only decorated technique represented
is red slip (table D). Although the chronological posi-
tion of this collection is unidentified, it demonstrates
that cariape tempering was employed in south central
Colombia. Whether the Napo Phase incised and
excised techniques also occur on the upper Putumayo
has not yet been established.
TABLE D.—Frequency of temper varieties in a sherd collection from the Rio
Guepi.

TemperSurface treatmentPlainRed slippedCariape
Charcoal
Sand
Sherd74
13
5
16
4TOTAL9310  Most of the ceramic traits diagnostic of the Napo
Phase occur along the middle Amazon between the
Rio Japura and the Rio Tapajoz and on the island of
Marajo (fig. 68), a distribution that led to the form-
ulation of the Polychrome Horizon Style (Meggers
and Evans, 1961, pp. 379-381). Except for limited
excavation by Hilbert, documentation is principally
in the form of complete vessels and decorated sherds
collected over the past century and deposited in mu-
seums around the world. Provenience identification
is often vague. Where a specific place name is pro-
vided, it may be impossible to find on a map. In some
cases, several collections with the same provenience
are so different in content that they must represent
different sites or archeological phases. Other collec-
tions incorporate a range of decorative techniques
and vessel shapes indicative of mixture, but whether
this is the result of reoccupation of the site, amalgama-
tion of different cultural traditions, or careless methods
of collection in which samples were mixed by the col-
lector cannot be determined without systematic field-
work. These inadequacies in contextural data make
it necessary to restrict this review to the occurrence of
Napo Phase traits, ignoring associated features.
  Evidence of the relative or absolute chronological
position of these complexes along the Amazon is min-
imal. Hilbert's work in the Manaus area and that of
Meggers and Evans on Marajo, places the Guarita
and Marajoara Phases late but prior to European
contact (Meggers and Evans, 1961, pp. 379-381). The
only available carbon-14 date derives from Hilbert's
work at Coari (Evans and Meggers, 1962, p. 244).
Organic temper, in this case spicules of fresh-water
sponge or cauixi, extracted from a sample of sherds,
  

FIGURE    68.—Map of the Amazon region showing the location of sites representing the Polychrome Horizon Style.



CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

95



gave an age of 800±47 years or A.D. 1150 (P-373).
This is nearly contemporary with the earliest Napo
Phase date of A.D. 1168 (P-347).
  Major locations from which Napo Phase (Poly-
chrome Horizon Style) traits have been reported will
be reviewed in alphabetical order. Locations are shown
on figure 68 and traits are summarized on figure 79,
where the arrangement is in geographical order from
west to east.
Beruri, Rio Purus
  ARCHEOLOGICAL REMAINS.—Two vessels in the cus-
tody of the Instituto Geografico e Historico do Ama-
zonas, Manaus, are from Beruri near the mouth of
the Rio Purus. One is an anthropomorphic urn with
a flat bottom, rounded shoulders, tall concave-walled
neck, and slightly everted rim. The surface is badly
eroded and retains no traces of slip or ornamentation.
The lid has the form of a rounded bowl whose mouth
diameter coincides with that of the jar rim. One side
bears an anthropomorphic face framed by a relief
band; eyes and nose are also relief (mouth area is
broken off). The lid surface retains traces of white
slip. Due to inability to remove specimens from the
exhibit case, the measurements are approximate. Jar
height is about 38 cm.; lid height about 10 cm.
  The second vessel is a deep bowl mounted on a tall
ring base. The sides of the bowl are flattened pro-
ducing a squarish outline. The exterior was decorated
with broad (5-8 mm. wide) grooves prior to the addi-
tion of a white slip. Traces of black painting remain
between the grooves. The pattern includes a stylized

face with eyes and nose in low relief. Total height
is about 24.5 cm.
REFERENCES.—None.
Boca do Xavier, Rio Urubu
  ARCHEOLOGICAL REMAINS.—A collection of about
100 sherds from the lower Rio Urubu, a small tributary
of the left bank of the Amazon (fig. 68), was deposited
by Nimuendajii in the Goteborg Museum. A variety
of decorative techniques of non-Napo Phase affiliation
are represented, but a few sherds represent waist
flanges with grooved decoration like that reported
from other middle Amazonian locations. Surfaces
are typically red or white slipped subsequent to inci-
sion. Incision with a single- or double-pointed tool
also occurs, as do small adornos. Broad everted rims
and flanges may be lobed. Temper is cariape or
cauixi.
REFERENCES.—None.
Coari Region
  ARCHEOLOGICAL REMAINS.—The ceramic complex
in the vicinity of Coari, on the right bank of the
Amazon (fig. 68), has been defined by Hilbert (n.d.).
Detailed analysis of sherd samples from stratigraphic
tests reveals many similarities to Napo Phase pottery.
Cariape, cauixi, and charcoal temper occur; sand
was not employed. Decoration is predominantly
grooved (pi. 79 b-f, h-i), with the surface sometimes
subsequently white (pi. 79a) or red slipped. Other
techniques include single-line incision (pi. 79 l-m),
excision and painting (red-on-white, black-on-white,
  

FIGURE 69.—Rim and body profiles of sherds from the Coari region. (Courtesy Peter Paul Hilbert; black = undecorated, white = decorated,
hachure = red slipped.)
293-822  O - 68 - B

96

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



red and orange-on-white, red and black-on-white;
79 g, j-k), Adornos are rare. Painted designs are
poorly preserved but several suggest the pseudo-
negative technique.
  A number of features of vessel shape duplicate
those of the Napo Phase. Among these are channel
rims (fig. 69a; Napo Phase Form 10), collared rims
(fig. 69c; Napo Phase Form 16), shallow bowls with
direct rim (fig. 696; Napo Phase Form 12) and vessels
with a thickened carination or, more typically, a
waist flange (fig. 69 e; Napo Phase Forms 7-10),
usually with a notched edge. Exteriorly thickened
rims (fig. 69d), probably associated with the flanges,
resemble Napo Phase Form 7. The flat lip characteris-
tic of vessels from the Coari region is also paralleled
in the Napo Phase. Base forms include all Napo
Phase varieties. Although the small size of most sherds
makes contour difficult to recognize with certainty,
one  base  suggests   a  square  outline.
  Other sources add little to the details furnished by
Hilbert's work. However, a small globular vessel
illustrated by Cruls (1942, pi. 12, lower left) is of
interest because it possesses a square orifice. The
exterior is covered with grooved decoration.
  REFERENCES.—Cruls, 1942, p. 213 and pi. 12;
Hanke, 1959, pp. 43-50 and figs. 15-27; Hilbert,
n.d.

FIGURE 70.—Rim and body profiles of sherds from the Guarita
Phase. (Courtesy Peter Paul Hilbert; black=undecorated,
white = decorated.)
Guarita Phase (Manaus)
  ARCHEOLOGICAL REMAINS.—Construction of an oil
refinery near Manaus has destroyed a site belonging
to the Polychrome Horizon Style.  Hilbert  (1959b)

was able to make two small stratigraphic tests in
a remnant of the site, and a surface collection. The
site area was reported to have been 200-300 meters
in diameter. Guarita Phase pottery is cauixi or
cariape tempered. Decoration is by grooving (pi. 80
a-j) prior to addition of a white slip, polychrome
painting (red and black-on-white, red and orange-
on-white), red-on-white painting, and less commonly,
incision with a single- or double-pointed tool. Poly-
chrome painting may embellish grooved designs.
Adornos are rare.
  Vessel shape characteristics include broad everted
rims (pi. 80&), sometimes with lobes, and prominent
waist flanges with notched or plain edges (fig. lOd;
pi. 80 l-m). Thickened carination is rare. Some sherds
appear to represent vessels with flattened sides, but
truly square examples have not been noted. Shallow
bowls with flat lip (fig. 70c) and rims of Napo Phase
Forms 7 and 8 can be recognized (fig. 70 a-b).
  REFERENCES.—Hilbert, 1959b; Metraux, 1930, pp.
154-165.

FIGURE 71.—Excised vessel from Ilha dos Muras.   (After Barbosa
Rodrigues, 1891, pi. 8-1.)
Ilha dos Muras
  ARCHEOLOGICAL REMAINS.—Two complete vessels
described and illustrated in the last century by Bar-
bosa Rodrigues combine typical Napo Phase charac-
teristics with alien elements. They originate from
Ilha dos Muras, just above the mouth of the Rio
Negro (fig. 68). One is a square basin with a tall
pedestal base, the exterior of which is covered with
excised decoration more reminiscent of the Marajoara
Phase than of the Napo Phase examples. The de-
scription suggests it may be red slipped and white
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

97



retouched. The second is a square vessel with a
channel rim rising to peaks at the corners, a flange
or thickened carination, and a pointed bottom (fig. 71).
The exterior of the collar is covered with excision re-
sembling that on Napo Phase examples (e.g., fig. 47).
REFERENCES.—Barbosa Rodrigues, 1891-1892, pp.
28-32 and pi. 8, figs. 1 and 2.

FIGURE 72.—Rim and flange profiles of decorated  sherds from
Itacoatiara, Guajara (Museu Paulista, Sao Paulo).
Itacoatiara
  ARCHEOLOGICAL REMAINS.—Hilbert has investigated
a large habitation site underlying the western half of
the town of Itacoatiara, on the left bank of the Amazon
about 100 kilometers below Miracanguera (fig. 68).
The aboriginal occupation area extends for about 800
meters along the bank and about 200 meters inland.
The pottery is cauixi tempered, and decorated with
single-line (pi. 81 j-k) and double-line incision (pi.
81 l-p), grooving (pi. 81 q-r), excision (pi. 81 a-f),
painting (red and black or red and orange-on-white)
and modeling in the form of small zoomorphic or
geometric adornos. Some examples are white slipped
after incision (pi. 81 g—i). Fine incisions may be filled
with red or yellow pigment. Vessel shapes have little
in common with Napo Phase forms, but broad everted
rims, sometimes lobed, occur (fig. 12b), along with
vestigial channel forms (fig. 72a). A single hollow rim
represents the westernmost reported occurrence of
this form. Waist flanges are common (fig. 72c).
  An anthropomorphic urn illustrated by Netto (1885,
pi. 5A, fig. 3)  resembles examples from the.nearby

site of Miracanguera. It has a low pedestal base, wide
rounded shoulders and slightly concave walls. A
specially made bowl-like lid covers the orifice. A
stylized face on the upper side of the neck is the only
anthropomorphic detail. The exterior is white slipped
and probably was polychrome painted.
  Two large sherd collections were made at Itacoatiara
by Harald Schultz in 1955 and deposited in the Museu
Paulista, Sao Paulo, Brazil. One, from Barrio Colonia,
consists predominantly of adornos and sherds decor-
ated by finely incised lines, although certain traits of
the Polychrome Horizon Style are represented (such
as excision, double-line incision, and white slipping).
The other collection, labeled "Itacoatiara, Guajara'",
includes grooving, excision, single-line and double-line
incision, red-on-white painting, and white slipping
subsequent to grooving, as well as ring impression,
dentate stamping, and punctation. In both samples,
tempering is cauixi.
  REFERENCES.—Hilbert, 1959b; Netto, 1885, p. 548
and pi. 5A, fig. 3.
Rio Japura
  ARCHEOLOGICAL REMAINS.—Potsherds and complete
vessels are reported to occur at a number of locations
along the Rio Japura, but none is identified with
sufficient precision to permit location on a map.
Hanke visited several sites between the mouth and
the Colombian border. Most of the illustrated sherds
are geometric and zoomorphic rim adornos. However,
a number of Napo Phase characteristics are associated,
including grooved decoration, lobed rim treatment,
waist flanges, black-on-white painting, single-line in-
cision and possibly double-line incision (although the
drawings are not clear). Broad everted rims are
typical.
  An anthropomorphic urn from the lower Japura
above Mapari has been described by Metraux. The
stylized face resembles that of some Napo Phase urns,
as does the general vessel form (cf. fig. 68).* Limbs
are absent. The orifice is covered with a bowl-shaped
lid made for the purpose. The jar exterior is white
slipped and a broad red band encircles the waist.
Height is 42 cm., maximum diameter 37.5 cm., and
body wall thickness 1 cm. Hilbert (1962b, p. 465),
who later visited the Mapari site, reports the ceramic
complex to be "a cauixi tempered ware with painting
in red and black on white, grooving and flanges
around the area of maximum body diameter by open
bowls. Modeling is rare and seems to be limited to an-
   * The Amazonian distribution of this style of anthropomorphic
vessel has been reviewed by Imbelloni (1950, p. 148, footnote 83),
who suggests that the diffusion proceeded from east to west.
  
98

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



thropomorphic burial urns." No other sites of this
tradition were found by Hilbert in a survey extending
upstream for 350 kilometers, although he found it
characteristic "in most of the sites around the mouth
of the Japura as well as around Tefe" (ibid.).
  Another collection of sheids from various locations
along the Rio Japura was made by A. Melchoir and
deposited in the Koninklijk Instituut voor de Tropen
in Holland (Feriz, 1963, pp. 150-151). Particularly
interesting is a rim lobe with anthropomorphic eye
treatment, and grooved and polychrome decoration
reminiscent of Rocafuerte Incised designs on the
flaring walls of basins of Form 14 (frontispiece; pi. 42;
op. cit., fig. 13). This sherd was obtained 350 kilo-
meteis above the mouth of the Japura (op. cit., p. 168).
Other fragments exhibit such Napo Phase elements
as square outline (op. cit., figs. 3, 9) and thickened
carination (op. cit., figs. 9-10). One location spe-
cifically mentioned is a site on the bank of Lago
Amana, some distance inland from the left bank of
the Japura near the mouth (fig. 68). Burial urns
were observed eroding out of the ground (op. cit.,
fig. 1), and a fragment of face (op. cit., fig. 11) sug-
gests that some may have been anthropomorphic.
A rim sherd with excised decoration on the exterior
and a "not quite circular opening" was obtained
in the vicinity (op. cit., p. 165 and fig. 5). Adornos
also occur (op. cit., figs. 19-22).
  REFERENCES.—Feriz, 1963; Hanke, 1959, pp. 51-60
and figs. 28^-3; Hilbert, 1962b; MeVaux, 1930,
pp. 165-166 and fig. 35.
Lago Ara^a Region
  ARCHEOLOGICAL REMAINS.—A surface collection
made by Hilbert at Lago Araga, about 80 kilometers
west of Codajas on the left bank of the Amazon
(fig. 68), incorporates a number of traits allying it
with the complex at Coari, and more remotely with
the Napo Phase. Cauixi temper predominates over
cariape, while a few sherds contain charcoal or clay
(crushed sherd?). Decorated techniques include
double-line incision, excision, grooving, and single-
line incision. Slipping is not reported and red painting
is rare.
  Among Napo Phase elements of vessel shape,
channel rims (fig. 13a) and collared rims (fig. 73c)
are rare. A few rims resemble Napo Phase Forms 7
and 8 (fig. 13b) and are probably associated with
body sherds showing thickened carination (fig. 13d)
or waist flanges (fig. 13d). Numerous sherds represent
square vessels, some with rounded rather than flat
sides, but one exhibits the pointed corners and concave
sides of some Napo Phase examples. Grooved decora-

tion frequently extends onto the flange, terminating
in notches on the edge.
  A small rounded jar with a slightly everted rim,
decorated on the exterior with incision and excision,
has been reported by Barbosa Rodrigues from "above
Cudayas," possibly from the same site investigated
by Hilbert. It is tempered with cauixi.
  REFERENCES.—Barbosa Rodrigues, 1877; Hilbert,
n.d.

FIGURE 73.—Rim and body profiles of decorated sherds from the
Lago Araga region (Courtesy Peter Paul Hilbert). Brackets
designate decorated zones.
Rio Madeira
  ARCHEOLOGICAL REMAINS.—Two small pottery sam-
ples collected by Nimuendaju and separately cata-
loged in the Goteborg Museum both seem from the
description to be from Guajara on the lower Rio
Madeira (fig. 68). One sample contains six sherds,
all cauixi tempered, exhibiting flanges, a large rim
lobe and grooved decoration. The other sample totals
ten sherds, principally zoomorphic and anthropo-
morphic adornos, sometimes accompanied by incision.
Sherd tempering is characteristic. A fragment of a
cylindrical stamp is included.
  Several anthropomorphic urns of the style associ-
ated with Miracanguera but attributed to sites on
the lower Rio Madeira are in the collection of the
Instituto Geografico e Historico do Amazonas,
Manaus. One, from the Rio Ipixuna, has a flaring
pedestal base, angular shoulder and insloping walls
free of anatomical detail (pi. 89c). The head forms
the lid, and bears an anthropomorphic face framed
by a relief band. Nose and mouth are low relief; eyes
are narrow horizontal incisions. Eyes and mouth are
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

99



outlined with paint, which also ornaments the rest
of the face. The vessel is covered with a polished
white slip and painted with black lines in the pseudo-
negative technique. Jar height is about 21 cm., lid
height about 7 cm. Another with similar painted
decoration (pi. 89d) is of unknown provenience.
  Three other urns in the same collection are from
Nova Olinda. One is very similar both in form and
anthropomorphic treatment to that from the Rio
Ipixuna. Another has a flaring base, low rounded
shoulders, and a tall vertical-walled neck (pi. 89b).
The surface of this neck was decorated with broad
(4-5 mm. wide) grooves before the application of a
thick white slip, which filled and rendered almost
invisible the grooved decoration. Its presence is now
revealed by the eroded condition of the surface. The
head, which forms the lid, has an elaborate relief
band framing the face. The glass exhibit case could
not be opened so measurements are approximate. Jar
height is about 45 cm., lid height about 17 cm. The
third vessel has lost the head-lid. Anthropomorphic
detail on the body is relatively realistic, including low
relief arms, high relief lower legs and free modeled
feet. Breasts are low nubbins, navel a raised ring with
a depressed center, and genitalia a triangular relief.
Traces of a complicated pattern in fine and medium
reddish brown lines remain on the white slipped
surface. Relief is bordered by broad red bands. Jar
height is about 54 cm.
REFERENCES.—None.
Manacapuru
  ARCHEOLOGICAL REMAINS.—The modern settlement
of Manacapuru is on the left bank of the SolimSes
above the mouth of the Rio Negro (fig. 68). Hilbert
undertook a small stratigraphic test here, which
produced more than 2000 sherds. Cauixi is the
principal temper, but cariape and crushed sherd
also occur. Decoration includes grooving, single-line
and double-line incision, and red-on-white (rarely,
red and orange-on-white) painting. Some examples
were red slipped prior to incision. Vessel shapes are
simple and not reminiscent of Napo Phase forms.
  Other collections from the Manacapuru region,
however, show closer affiliations to the Napo Phase.
Hanke investigated four locations, which are not
separately identified in the sherd sample deposited
in the Museu Paulista. This sample includes grooved,
excised, single-line incised and painted (red and
black-on-white) decoration, as well as lobed rims,
and zoomorphic and anthropomorphic adornos. Red
or white slipping occurs subsequent to grooving.
Temper   is   sherd,   cariape,   or   cauixi.   Thickened

carinations and flanges with notched ornamentation
are typical. Some vessels are noncircular. Hanke
(1959, p. 41) reports the existence of secondary urn
burial.

3 CM
  Additional evidence comes from decorated sherds
deposited by Harald Schultz and Geraldo Pinheiro
in the Museu Paulista. Included are fragments with
typical Napo Phase decoration, both in technique and
motif, among them single-line and double-line incision
(pi. 82 g-h, j ), excision (pi. 82 c-d, f), black-on-white
FIGURE 74.—Rim and body profiles of sherds from Manacapuru
(Museu Paulista, Sao Paulo). Brackets designate decorated
zones. (White = decorated, hachure = red slipped.)
(pi. 82/), and polychrome (red and black-on-white)
painting (pi. 82k). Other Napo Phase characteristics
include red or white slipping subsequent to grooving
or incision (pis. 82 a—b, e, i, 83a), and a tendency to
square shape (pi. 83b). Several specimens have promi-
nent flanges (fig. 74c; pis. 82 b, e, 83a). Flat and
cylindrical stamps with deeply grooved or excised
designs are associated. A number of features of vessel
shape duplicate those of the Napo Phase. Among
these are channel rims (fig. 74a), shallow bowls with
direct rims (fig. 14b), exteriorly thickened rims
(fig. 74c) associated with waist flanges, and both plain
and lobed waist flanges (fig. 74c; pis. 82 b, e, 83a).
REFERENCES.—Hanke, 1959, pp. 37-43 and figs.
1-14; Hilbert, n.d.

100

SMITHSONIAN  CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6





FIGURE 75.—Anthropomorphic urns with painted decoration from
Miracanguera. Measurements not provided on original
drawings, a, Anuiropomorphic urn with face on neck and
with a lid. b, Jar with face on flange and with a pedestal base.
(After Barbosa Rodrigues, 1891, pis. 3-1 and 3-4.)

Marajoara Phase (Marajo)
  ARCHEOLOGICAL REMAINS.—Numerous sites on the
eastern half of the island of Marajo at the mouth of
the Amazon (fig. 68) have produced a ceramic com-
plex that has long commanded attention (Meggers
and Evans, 1957, pp. 259-324) because of the variety
and complexity of decoration (pis. 84-87, 88b). The
existence of large samples from stratigraphic testing
(ibid.; Hilbert, 1952; Figueiredo and Simoes, 1963),
as well as numerous surface sherds and complete
vessels, permits more accurate identification of Napo
Phase traits than is possible in any other part of the
Amazon basin. Few of the features selected for dis-
tributional analysis are absent (fig. 79). These are
decoration by double-line incision,* grooving, nega-
tive painting, white-on-red painting, and white slip-
ping subsequent to incision; channel and collared
rims, waist flanges, and anthropomorphic urns with
the separable head serving as the lid. Tempering is
by crushed sherd. Although cariape has been reported
by Mordini (1947, p. 640) in some sherds from
Pacoval, this identification has not been verified by
other observers. It is interesting because of the as-
sociation of cariape with the Polychrome Horizon
Style farther up the Amazon, and because of the
relatively early position of Pacoval in the Marajoara
Phase seriated sequence (Meggers and Evans, 1957,
fig. 141). Complete vessels from Pacoval exhibit other
survivals of traits more common upriver, such as
flattening of the sides of circular bowls (pi. 86)
and modeling of limbs with swollen calves on anthro-
pomorphic jars (pi. 88b).
  REFERENCES.—Figueiredo and Simoes, 1963; Hil-
bert, 1952; Meggers and Evans, 1957, pp. 259-404;
Mordini, 1947; Palmatary, 1950.
Miracanguera
  ARCHEOLOGICAL REMAINS.—One of the earliest sites
to be reported from the middle Amazon is Miracan-
guera, on the left bank approximately opposite the
mouth of the Rio Madeira (fig. 68), where Barbosa
Rodrigues collected numerous burial urns nearly a
century ago. At that time the cemetery was inundated
during the rainy season, and when the water receded
"on the beach, thousands of potsherds attest the
great number of burial urns destroyed by the collapse
of the bank, to be carried in fragments and interred
in the bottom of the river" (Barbosa Rodriques, 1892,
p.  2). The urns are typically large jars with flat,
  * Guajara. Incised, a late Marajoara Phase decorated type,
employs a different kind of double-line incision from that charac-
terizing the Napo Phase.
 
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

101



annular or pedestal base, low rounded shoulders and
insloping neck (figs. 75-76; pi. 89a). Anthropomorphic
details include stylized arms and legs, often partly in
the round, and a face either on the neck (fig. 75a), on
the lid (pi. 89<z) or on a lobe extending vertically from
the rim (fig. 15b). The surface is white slipped and
painted with red and black.  One example in the

FIGURE 76.—Anthropomorphic urn from Miracanguera.
Measurement not provided on original. (After Barbosa
Rodrigues, 1891, pi. 2-8.)
Museu Paulista (pi. 88a) of slightly different style, is
painted in the pseudo-negative technique typical of the
Napo Phase. Other decorative techniques include
grooving, with the surface subsequently white slipped
and the grooves painted with red, single-line
incision, and adornos. Napo Phase vessel shapes in-
clude shallow bowls with flat lip, channel rims, and
small vessels with exteriorly thickened rims and waist
flanges or thickened carination. Ovoid or boat shapes
occur, but square vessels have not been reported.
Zoomorphic forms combine painting and modeled
animal heads and tails (fig. 77).
  REFERENCES.—Barbosa Rodrigues, 1891-1892; Nor-
denskiold, 1930, pp. 22-3 and pi. xxvi, f.


FIGURE 77.—Zoomorphic vessel from Miracanguera with red and
black-on-white painted decoration. Length from nose to tail
circa 24 cm. (After Barbosa Rodrigues, 1891, pi. 3-7.)
Mocajatuba, Rio Urubu
  ARCHEOLOGICAL REMAINS.—Nimuendaju collected
157 sherds from an island in Lago Saraca on the lower
course of the Rio Urubii (fig. 68), which are now in
the collections of the Goteborg Museum. Although
the complex generally relates more closely to that at
Itacoatiara than to more typical members of the
Polychrome Horizon Style, it incorporates such ele-
ments as excision, single-line and double-line incision,
red and orange-on-white and red-on-white painting,
red-retouched incision on a white slipped surface, and
small adornos. Some fine incisions are white filled.
Broad horizontal rims, sometimes lobed, occur but
flanges are unrepresented. There is one annular base.
Several rim adornos appear to be from the corners of
square bowls. Temper is cariape or cauixi.
REFERENCES.—None.
Nazare dos Patos
  ARCHEOLOGICAL REMAINS.—Samples in the Goteborg
Museum, collected by Nimuendaju from three sites
on the lower Rio Tocantins, contain a number of
Polychrome Horizon Style features. The largest
sample, some 350 sherds, is from Nazare dos Patos,
on the right bank (fig. 68). A remarkably large
number of Napo Phase decorative techniques are
present, including grooving typically done with a
double-pointed or serrated tool (pi. 90 a-h) and ex-
cision (pi. 90 i-j) associated with plain, white or
red slipped surfaces, and sometimes red or white
retouched. Surface erosion sometimes makes it un-
certain whether a white slip was applied after grooving
(pi. 90/, h) or the cuts were white filled. Grooving
is rare, as is red-on-white painting. Adornos are
typically small. Temper is predominantiy sand, and
rarely cariape. Elements of vessel shape include broad
  
102

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



everted rims with lobed or irregular lip, hollow rims,
and slightly thickened carinations.
REFERENCES.—None.
Rio Padauiri
  ARCHEOLOGICAL REMAINS.—One fragmentary vessel
in the collection of the Museu Paraense Emilio Goeldi
is attributed to the Rio Padauiri, an affluent of the left
bank of the Rio Negro (fig. 68). It has a prominently
thickened carination and grooved decoration covers
the upper wall. This is the northernmost reported
occurrence of these traits in the Amazon basin.
REFERENCES.—None.
Paura
  ARCHEOLOGICAL REMAINS.—A collection of 84 sherds
in the Goteborg Museum, acquired by Nimuendaju
from Paura on the left bank of the Amazon below
the mouth of the Rio Uatuma (fig. 68), is of interest
principally for the presence of several broad everted
rims with excised decoration (pi. 91 a-c). Single-line
incision prior to red slipping of the surface, and small
adornos are also typical. Some vessels have annular
bases. Tempering is cauixi or cariape.
RE FERENCE S.—None.

have exceeded  10 cm. The soil is white sand, dis-
colored gray in the occupation layer.
  A large surface collection was made. Temper is
sand or cariape. Decorative techniques and vessel
shapes indicate this site to belong to the Guarita Phase.
Decoration is by single-line incision, painting (red-on-
white, black-on-white, or red and black-on-white),
or grooving, with the latter predominant. Grooved
designs were often subsequently white or red slipped,
and sometimes embellished with black-on-white
painted decoration. Vessel shapes include Napo Phase
Forms 1 (fig. 18a), 1 (fig. 18b), 8 (fig. 78c), 10 (fig.
78*), 11 (fig. 78/), 12 (fig.78^), and 14 (fig. 78i),
and Rare Forms 2 and 4. A common rim may have
evolved from Napo Phase Form 9, with a flat upper
surface replacing the channel (fig. 18d). Other simi-
larities to Napo Phase vessel shapes include the an-
nular base, lobed rim, noncircular outline, thickened
carination and waist flange (fig. 78/z).
  Although the Polychrome Horizon Style has been
guess-dated prior to European contact (Meggers and
Evans, 1961, fig. 7), several blue and white glass "seed"
beads were collected from the surface of Ponta Negra.
Association is not conclusive since fragments of Euro-
pean china, tile, and glass were also collected from
the surface,   indicating reoccupation in recent times.
REFERENCES.—None.



Pogo do Jaburu
  ARCHEOLOGICAL REMAINS.—A small sample of dec-
orated sherds in the Museu Paulista was collected
from Poco do Jaburu, tentatively placed in the vi-
cinity of Anama on the left bank of the Solimoes
below the mouth of the Rio Purus (fig. 68). All are
cauixi tempered. Decorative techniques include ex-
cision, grooving, single-line incision on a plain or
red slipped surface, double-line incision and red-on-
white painting. Thickened carination and flanges are
typical.
REFERENCES.—None.

Pontao, Rio Urubu
  ARCHEOLOGICAL REMAINS.—The Goteborg Museum
has 59 sherds collected by Nimuendaju from a site on
the north shore of Lago Saraca on the lower Rio
Urubu (fig. 68), probably representing the same com-
plex as at Mocajutuba. Temper is cauixi. Decoration
includes excision (pi. 91/), single-line and double-line
incision (pi. 91c), grooving (pi. 9\g), red and black-
on-white painting, and small adornos. Vessel shape
details include broad everted rims, often lobed,
flanges, and possibly square outiine.
REFERENCES.—None.



Ponta Negra, Rio Negro
  ARCHEOLOGICAL REMAINS.—While in Manaus in
1966, we were able to visit a site on the left bank of
the Rio Negro just below the mouth of the Igarape
Taruma (fig. 68). Surface sherds extended along the
summit of the bank for about 650 meters, and inland
to a maximum of 150 meters. Elevation of the bank
above river level in June, at the beginning of the dry
season, was about 15 meters. The habitation deposit
has been badly disturbed by excavation of sand for
road construction, but original depth appears not to

Santarem Area
  ARCHEOLOGICAL REMAINS.—The Goteborg Museum
collections made by Nimuendaju, represent a number
of localities around Santarem, which produced typical
Santarem style ceramics. However, a few sherds
appear to belong to the Polychrome Horizon, suggest-
ing that there may be sites of this occupation in the
area. Diagnostic features include single-line (pi.
92 h-l) and double-line incision (pi. 92 a-g), red and
black-on-white painting, broad everted rims, flanges,
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

103



thickened carinations, and annular bases. One lobed
rim has an incised design with a touch of excision,
Adornos are abundant. All types of temper are present:
sand, crushed sherd, cariape*, and cauixi. A fragment
of a vessel with excised decoration and a flange is
illustrated by Feriz (1963, fig. 14).
REFERENCES.—Feriz, 1963.
Tauaquera, Rio Aniba
  ARCHEOLOGICAL REMAINS.—Two small collections
deposited by Nimuendaju in the Goteborg Museum
are from Tauaquera, on the left bank of the Rio
Aniba near its mouth (fig. 68). Although cataloged
separately, they may be from the same location.
Each contains 35 sherds. Temper is sand, sherd,
cariape, or cauixi. Decoration includes excision
(pi. 9ld), grooving, incision with a single-pointed
tool, and adornos. Broad everted rims and one
hollow rim occur. Other sherds are decorated with
relief and punctation, which do not form part of the
Polychrome Horizon Style.
REFERENCES.—None.

Tefe Region
  ARCHEOLOGICAL REMAINS.—Several kinds of dec-
orated sherds are reported from the vicinity of the
confluence of the Rio Tefe" and the right bank of the
Amazon (fig. 68). An anthropomorphic head about
11 cm. high, broken off at the neck, is interpreted
by Metraux (1930, pp. 146-147 and fig. 1) as a
figurine because of the absence of an opening in
the top. Its size suggests, however, that it may repre-
sent an urn with the orifice in the bottom like many
Napo Phase examples. The surface is white slipped
but no traces of painting are noted except touches of
red on the mouth and ears. Other pottery from this
region described by Metraux consists of anthropo-
morphic and zoomorphic adornos.
  Two vessels, one complete and one fragmentary,
oiiginate from Lago Macupy, located by Hebert
(1907, p. 185) "near the Japura" and by Metraux
(1930, pi 152) on the Rio Tefe. The fragment repre-
sents an anthropomorphic seated vessel with limbs
modeled in the round and hands on the knees. The
head is superficially demarcated from the rounded
  


I  i l i I i I
0       I      2      3 CM







FIGURE 78.—Rim and body profiles of sherds from Ponta Negra, Rio Negro. (Black = undecorated, white = decorated.)

104

SMITHSONIAN   CONTRIBUTIONS   TO  ANTHROPOLOGY

VOLUME   6



body, and features are crudely modeled. Sex is male.
The other vessel is an ovoid jar with constricted neck
and everted rim. The shoulder is decorated with a
pattern of concentric diamonds and coils executed by
broad deep grooves. Small zones of punctation are
associated, possibly previously excised (Metraux,
1930, pi. II).
  A number of sherds from the Tefe region in the
Koninklijk Instituut vor de Tropen in Holland have
been described by Feriz (1963). Napo Phase features
include excision (op. cit., pp. 170-171 and figs. 26-7),
red-on-white painting (op. cit., p. 171 and fig. 28),
and an anthropomoiphic face framed by an elevated
band (op. cit., fig. 12). Several different sites are
repiesented, some of which also produced adornos
(op. cit., figs. 17-18).
  Two collections made by Harald Schultz from the
Rio Tefe are in the Museu Paulista. One, from
Abial, Lago Tefe, includes white slipped, excised
(pi. 93 a-c) and grooved (pi. 93 d-h) decorative
techniques. The other, from Boca de Tefe, includes
single- and double-line incision, excision, grooving,
red-on-white painting, and rim adornos. All sherds are
cauixi tempeied.
  REFERENCES.—Feriz, 1963; Hanke, 1959, p. 51;
Hebert, 1907; Hilbert, 1962a; Metraux, 1930.
Significance of the Distributional Pattern
When the occurrences of the principal Napo Phase
ceramic features of temper, decoration, and form are
arranged in geographical ordei (fig. 79), a pattern of
west to east diffusion emerges. In order to show
this more clearly, the traits have been divided into
three geneial groups: (1) those present in highland
Colombia and in the Napo Phase; (2) those not
reported for highland Colombia but present in the
Napo Phase; and (3) those absent in the Napo Phase
but associated with the Polychrome Horizon Style
along the Amazon. Within each of these groups,
traits have been arranged in order of increasing
downriver occurrence.
  Evaluation of the distributional pattern must take
into consideration the incompleteness of the data,
which probably accounts for the spotty occurrence
of many traits between the Rio Napo and the island
of Marajo. At these two extremes, the ceramic
complex is known from large and systematically
collected and analyzed samples, whereas the inter-
vening sites are often represented by a few complete
vessels or a handful of selected sherds, and even these
may not be described in detail. Decorative effects
such as white slipping subsequent to incision, red
retouch   of  incisions,   white   retouch   of  incision   or

excision, or color of painted designs often cannot be
detected on illustrations and are rarely described.
In consideration of such factors, it is probably safe
to assume that most, if not all, of the traits would
have less broken distributions if the evidence were
more complete. It is also predictable that many
traits now in Group 2 will be reported from highland
Colombia when the pottery there is better known
and has been better described.
  In spite of the distance separating the Rio Napo from
the island of Marajo in the mouth of the Amazon,
resemblances between Napo Phase and Marajoara
Phase ceramics are numerous and striking. Similari-
ties extend beyond duplication of nearly all decorative
techniques, to details of motif (Meggers and Evans,
1958, figs. 3-5). The correspondence is heightened by
the absence or extreme rarity in these two complexes
of two prominent features associated with the Poly-
chrome Horizon Style throughout the middle Ama-
zon: adornos and flanges. The obtrusive grooved
decoration of the intervening area is also little in
evidence.
  Marajoara Phase pottery, although employing the
same range of decorative techniques, is far more
elaborate than that of the Napo Phase. This is evident
not only in greater complexity of execution and more
complete coverage of the vessel surface, but also in
the more frequent juxtaposition of different techniques
on the same vessel. In the Napo Phase, even painting
is rarely combined with other techniques, whereas
Marajoara Phase bowls with incised or excised ex-
teriors are typically painted on the interior. Other
Marajoara Phase vessels may have one type of decora-
tion on the rim and another on the body. Napo Phase
pottery is primitive by comparison.
  Vessel shapes diverge more than does decoration.
Although several of the common Napo Phase rim
profiles occur on Marajoara Phase vessels (such as
Forms 1, 7, and 8), they are not typical. The Napo
Phase channel and collared rims are absent, and a
rounded lip is characteristic. While exaggerated
thickening of the carination is a distinctive Napo
Phase specialization, thickening of the angle on Mara-
joara Phase shouldered vessels is relatively slight.
Another Napo Phase diagnostic, the square shape, is
so weakly manifested by Marajoara Phase bowls as
to be easily overlooked unless attention is called to it
(pi. 86). The anthropomorphic urns of the two phases
are noncomparable in most respects, although one
Marajoara Phase example (pi. 88b) is reminiscent of
the Napo Phase style. Marajoara Phase specializations,
such as stools and tangas, have not been reported in
eastern Ecuador.
  Even at the present incomplete stage of investiga-
tion,   the  distributional  evidence  indicates  that the
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

105

ARCHEOLOGICAL SITES AND REGIONS

HIGH-
LANDS

LOWLANDS

CERAMIC TRAITS

Si     ™
0    OH
O    o
:2  'So ^
ft      60      «
(4  &   ai   H
S  2:
a

O
o  .2   .o
►5 3     22    -
Z *     3    «
OH
-n      &    3
I

U<M^S0HOO


■O  o
6   13
(rt     inj     "rt     J3
PH    C/3
m h   2



GROUP 1
Negative painting
White-on-red painting
Anthropomorphic urn, Type B
Channel rim
Grooving
Sand temper
Square outline
Anthropomorphic urn, limbs in full round
Thickened carination
Annular base
Incision, white retouched
Red and black-on-white painting
Red-on-white painting
Single-line incision
Excision
Excision, white retouched
Broad everted rim bowl
Anthropomorphic urn, limbs absent or stylized

XXXX

GROUP 2
Black ash temper
"Collared" rim
Anthropomorphic urn, head-lid
Double-line incision
Cariape temper
White slipped after incision
Lobed rim
"Pseudo-negative'' painting
Red retouched incision on white slip
GROUP 3
Body flange
Adornos
Cauixi temper
Red and orange-on-white painting
Sherd temper
Hollow rim
FIGURE 79.—Occurrence of Napo Phase ceramic traits of decoration and vessel shape in Colombia and Brazil. Sites are arranged in
geographical order, from west to east. (X = present, ? = identification uncertain.) Information is derived principally from Museum
collections, see pp. 95-104 for sources.

106

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



spread of the Polychrome Horizon Style cannot be
interpreted as a simple down-river movement. Com-
binations of decorative technique and vessel shape are
different in eastern Ecuador, on the middle Amazon,
and on Marajo Island, implying either complicated
diffusion patterns or temporal differences or both.
While some degree of error may be introduced by
mixed surface collections, other variations between
sherd samples are clearly the result of amalgamation
of the Polychrome Horizon Style with preexisting
or subsequent local or neighboring ceramic styles.
The variation in frequency and elaboration of an-
thropomorphic and zoomorphic adornos may be one
effect of such differential acculturation. Other dif-
ferences may reflect multiple routes of diffusion. The
fact that the headwater tributaries of the Japura
(Caqueta) and lea (Putumayo) tend to converge in
the southeastern mountains of Colombia (fig. 68)
makes them all potentially accessible for movements
out of the southern highlands. Several are known to
have been used by recent colonists (Chaves, 1945,
map facing p. 582), and there is some evidence of
earlier intrusions into the lowlands (Silva Celis, 1963).
Variations in speed of transmittal and amount of
acculturation from previously existing pottery-making
groups along the way would be expected to create
complexes of variable content by the time the Amazon
was reached. Detailed sequences of ceramic change

are needed throughout the northwestern lowlands to
evaluate this theoretical possibility.
  The fate of the Napo /Phase in eastern Ecuador is
also a matter of speculation at present. It might be
inferred on the basis of the sequence derived from the
seriation of pottery types that the group was moving
upriver (see pp. 78-81), and that continuation in this
direction brought closer proximity and perhaps ul-
timately amalgamation with groups living in the
Andean foothills. A square vessel from Avila (Porras,
1961, fig. 17-13) and a few diagnostic Napo Phase
sherds collected by Pedro I. Porras from Loreto on
the Rio Suno (fig. 1) could be explained in this way
(pi. 94 j-k). On the other hand, the Loreto material
includes sherds of a type also encountered in a few
Napo Phase sites, where it was identified as of trade
origin (cf. pis. 94 a-i and 76 i-n). The evidence is
thus not conclusive as it bears on the fate of the Napo
Phase.
  The report by Donald Lathrap (1965, p. 10) of a
ceramic complex in eastern Peru "closely related to
the well known style of the Rio Napo in Ecuador,"
suggests another possibility. This complex, known as
the Gaimito Phase, was identified at several large
habitation sites on the borders of Lago Imariacocha,
southeast of Pucallpa (fig. 68). A carbon-14 date of
A.D. 1320±60 (Y-1544) places these remains suffi-
ciently later than those of the Napo Phase to suggest
a derivation from eastern Ecuador.
  
EUROPEAN DISCOVERY AND THE COTACOCHA PHASE

In the year 1541, when the Rio Napo moved into
the sphere of history, it was uninhabited. This situa-
tion wrought much hardship on the men who accom-
panied Orellana on his inadvertent voyage of explo-
ration, which became the first descent of the Amazon.
Although the exact route of the expedition and locus
of particular events are subject to differences of
interpretation (e.g., Heaton, 1934; Millares, 1955;
Reyes y Reyes, 1942; Rodriguez, 1955), the general
picture is clear: no settlements were encountered
between the mouth of the Rio Coca and one day's
travel above the mouth of the Rio Curaray.
  Along the Rio Coca,   the Spaniards found Indian
villages (Ortiguera, in Heaton,  1934, p. 314), where
thanks to trinkets and to trade-offerings of salt, which a mong
them is prized highly, and to iron axes and iron machetes,
they began to bring . food in the form of many kinds of
fish and maize and yucca and sweet potatoes and other food
products  that were  to  be had in  the land.	they  [the
Spaniards] went out to fish the river, where they caught a

great deal of fish, because it was very well stocked with it, an
they killed with their arquebuses many turkeys and ducks.
The appearance of the Indians resembled that of
more recent groups throughout the eastern montafia
(Ortiguera, in Heaton, 1934, p. 315):
The people dwelling along this river have faces with very
regular features and have good physiques, and are dressed,
wearing cloaks and tunics handpainted in various designs and
colors, and the women have clothes with the same painted
designs. Among the men there were some who wore gold
medal-like pieces on their breasts, and the women [wore]
ear-hoops of the same [metal] hanging from their ears, and
other pieces hanging from their noses and throats. The weapons
which they have are "macanas," which are sticks cut from
black palm-trees, long like broadswords, having born edges and
a point, which [weapons] they handle prettily and nimbly,
and spears to be thrown.
Continuing downriver, the Spaniards entered the
uninhabited area and began to suffer from lack of
provisions. On the basis of information that a village
was to be encountered at a distance of four day's

CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

107



travel, Pizarro decided to send a party for supplies.
The plight in which failure to find the village placed
the search party is vividly described by one of its
members  (Robles et al., in Heaton,   1934, p. 258):
. . the expedition  of  exploration	left .     . to   go   in
search of maize down the river at the river junction concerning
which information was at hand, which junction everybody, and
the Governor [Pizarro] in particular, said might be at a distance
of something like a four days' journey at the most; and we,
coming in search of the said maize [undertook this journey]
without food or supplies, eating roots, herbs, [and] very
dangerous unknown fruits, and harassed by this shortage [of
food] we journeyed on for nine days, all through uninhabited
territory, and at the end of them God Our Lord having pity
[on us], He saw fit to place in our path a village where [upon
investigation] in it, we found a certain supply of maize; and from
the great hunger thus endured there died several Spaniards,
and those of us who were spared became quite ill from the said
suffering, because  .	it was very great, not only as a con-
sequence of not eating but also because of the constant rowing
from sun to sun, for this alone was sufficient to kill us.  .  .
Regarding the starvation, Carvajal {in Heaton, 1934,
p. 408) is more explicit:
... we were in great danger of death because of excessive
hunger that we were enduring.	lacking other victuals,
we were eating leather from the seats and bows of saddles, and
also the leather from game of the chests or hampers whose
covering was made out of it, in which we were transporting the
little clothing and bedding that we had, and a few tapir skins,
not to mention the soles and [even whole] shoes that could be
found among the members of the party; and, though there was
no sauce other than hunger itself, this latter created in them a
taste. .
  The village that offered the Spaniards salvation
from their desperate situation was located a little
upstream from the mouth of the Rio Curaray. This
identification derives from Carvajal's description (in
Heaton, 1934, p. 412) of a large tributary passed on
the first day's journey after their departure, which
"emptied in with such a great onrush and force . .
we barely escaped drowning in getting past the junc-
tion of the rivers in a great jam of logs which the
current had brought down." Only the Curaray is
large enough to create these conditions.
The lower Rio Napo produced a few small villages,

but the unsavory reputations of the Spaniards pre-
ceded them and the houses were usually found deserted
or even burned by the Indians. Starvation was again
becoming a serious threat when they emerged from
the Napo into the Amazon.
  In 1651, the banks of the Napo were still unpopu-
lated above its junction with the Aguarico (de la
Cruz, 1900, p. 35). Subsequently, they have been
settled by Quechua-speaking Indians, the makers of
the Cotacocha Phase pottery. Whether these colonists
came from the highlands, where similar pottery has
been reported in the Loja basin (Jijon y Caamano,
1951, pp. 231-233, figs. 254-6), or the eastern foothills,
where ceramic resemblances also occur (Porras, 1961,
pi. 15), has not been determined. It is possible, as
Ferdon has noted (1950, pp. 4-5), that the diffusion of
the Quechua language and the ceramic style reflects
progressive Christianization of the Indians rather than
population movement; on the other hand, the gap
that seems to exist in the archeological sequence and
the testimony of Orellana's men argue for intrusion
of people in post-Spanish times.
  Houses of the modern residents are rypically
scattered at wide intervals along the rivet banks,
while their garden clearings occupy inland locations
visible from the air (pi. \b). Toward the Peruvian
border, some of the land has been cleared for pasture
(pis. 6, 13). Rice and cotton have been introduced
as cash crops. Nuevo Rocafuerte, an administrative
center founded after the 1942 border crisis, and the
army base at Tiputini are the two principal settle-
ments. Both are supported by the Ecuadorian national
government. A mission at Nuevo Rocafuerte, also
with the aid of external resources, propagates the
Catholic religion. This latest culture to invade the
Rio Napo thus differs from earlier ones in being
independent of local resources for survival. Incorpora-
tion of the Rio Napo region into the political frame-
work of a modern nation may have broken the pattern
of intermittent occupancy that appears to have charac-
terized it throughout prehistoric time. Only the
future will tell.
  
HYPOTHETICAL RECONSTRUCTION OF AMAZONIAN PREHISTORY

The pattern of intermittent occupancy exhibited by
pottery-making groups on the Rio Napo is of interest
because it contrasts strongly with the evolutionary
continuum characteristic during the prehistoric period
in  the  highland   and   coastal  portions  of Ecuador.

There, although influences passed to and fro, and the
interplay of culture and environment shaped the
products in different ways, the thread of continuity
can be detected. Continuity is also the dominant
pattern elsewhere in the Andean highlands. By con-

108

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6



trast, discontinuity (i.e., a sequence of intrusive
complexes) has been reported from other parts of
the Amazonian lowlands, including Marajo Island
(Meggers and Evans, 1957), along the middle Amazon
(Hilbert, n.d.) and to judge from preliminary ac-
counts, in eastern Peru (Lathrap, 1965).
  A case can be made for the existence of a correlation
between these contrasting highland Andean and
lowland Amazonian patterns of cultural development
and the principal geographical characteristics of the
two areas. It has been pointed out (e.g., Reichel-
Dolmatoff, 1965, pp. 28, 82) that the geographical
diversity of the Andes created multiple ecological
niches requiring subsistence specialization for most
efficient use. The effect was centripetal in the sense
that increasingly efficient exploitation of a local
environment correspondingly diminished the possi-
bilities of successful transplantation to a different
ecological niche. The centripetal effect was enhanced,
or at least not offset, by the existence of mountain
ranges, deserts, swamps, and other kinds of natural
barriers between regions.
  In the Amazonian lowlands, the situation is reversed,
and a centrifugal effect develops. Differences in
climate and topography are minimal over nearly
half a continent, producing a relatively high degree
of uniformity in soil, vegetation, and fauna. Similar
subsistence techniques are almost equally effective in
the Guianas, the upper Xingu or eastern Peru, and
local resources requiring specialized technology are
few. Sedentariness is actively discouraged by rapidly
impoverished soils, while the vast network of navigable
rivers encourages mobility. The consequence, ap-
parently, was a semisedentary way of life, charac-
terized not only by frequent resettlement, but by
migration, which spread languages and cultural
elements from one edge of the lowland tropical
forest to the other.
  Although few complete sequences are available
for the tropical lowlands, existing data suggest two
kinds of archeological situations, and their geo-
graphical distribution can be explained by the
centrifugal hypothesis. One is the already mentioned
succession of distinct ceramic complexes of non-
homogeneous origin. The other is the presence of a
single ceramic complex of long or short duration.
When the known occurrences of these two kinds of
situations are plotted on a map (fig. 80), it is evident
that the areas in which a sequence of archeological
phases has been reported are adjacent to the eastern
slope of the Andes in Ecuador and Peru, and along
the middle and lower Amazon. On the Ucayali
and its tributaries, Lathrap (1965, p. 12) reports
twelve ceramic complexes, which he places in chrono-
logical  order on stratigraphic  evidence.  Our work

has identified a sequence with four independent
phases along the Rio Napo in Ecuador. Five successive
phases have been recognized on the island of Marajd
(Meggers and Evans, 1957; Simoes, 1966). Although
investigation has been less intensive along the middle
and lower Amazon, Hilbert's (1955, 1959 a, b),
stratigraphic work and surface collections made by
Nimuendaju and others from a large sample of sites
attest to the existence of local chronologies composed
of several phases of differing content, implying
diversified origin.
  The presence of a single archeological ceramic
phase has been reported from several other parts of
the lowlands, all peripheral to the Amazon River. The
Rupununi Phase of the interior savanna of Guyana
(formerly British Guiana) and adjacent Brazil, and
the Taruma Phase of the upper Essequibo river are
post-European arrivals in their respective areas
(Evans and Meggers, 1960). A simflar situation is
described by Simoes (1967) on the upper Xingu,
where pottery also appears to be of relatively recent
introduction. Hilbert (1962b, p. 465) encountered
remains of a single phase in a 350 km. segment of
the Rio Japura. Survey of parts of the upper Orinoco,
Ventuari, and Manipiare rivers in southern Vene-
zuela produced evidence of two geographically
isolated archeological phases, each accounting for
all the sites discovered in their respective areas
(Evans, Meggers, and Cruxent, 1960).
  The greater antiquity of pottery making, and the
higher level of general cultural development character-
istic of the Andean area throughout the aboriginal
period, make it seem probable that the primary direc-
tion of diffusion was from west to east. The occurrence
of multiple phases adjacent to the Andes and along
the main course of the Amazon, and single phase
occupations in peripheral regions seems consistent
with such a hypothesis. With the passage of time, new
immigrant groups would have found on the banks of
the Amazon more settiements representing greater
cultural variety. The ceramic evidence indicates that
amalgamation, acculturation, assimilation, displace-
ment, or even extinction were alternative results at
different times and places. As population density in-
creased, the centrifugal effect of the lowland environ-
ment would be expected to manifest itself in the
dispersal of pottery-making groups up the tributaries
to the north and south. The diminishing time depth
for ceramic-using phases in these marginal areas and
relative rarity of multiphase sequences is what would
be expected to result from such a process.
  Although this reconstruction of Amazonian pre-
history is supported by relatively little archeological
evidence, it is interesting to note that a generally
similar conclusion has been proposed by Noble on the
  
CULTURAL   SEQUENCE   AND   ITS   IMPLICATIONS

109



.1
CM.1
oj
COJ                J                 J
CO                        *t                         CM 1         |f
°      /CM      ; y y *       (,■"?,-    1                                     J^^^0~?-i                    ;.„.A^^ji^:;."-,CD
<3-/wX                            *   w -€"x/x/li^s.                  '"y T. 5?woCyPS^x^suiiuoooi ^J!>^N3-.,,1' \^iK";.._b
inow[^7  »                                               :1(,'.,>\>.-      *4'c'"'WSST"'^N. *:,•;,:-..b	KJ
IT)y^C         -a—M^t.,,f'^^^r-ky                       ?,,../fc-^7           _„„_ ^J'r;"sy^r      w_J/V^/          v"f ~yj "hs.        Cf1'n?-vin\A—^I    —   ■*•. sr     iy   ^3          v^™%^7            *3tylAAA/^^^fK                  W -.°co
in^V         y.y~'*•-
/        #.y   *.^A^    i ^-\CO
in'"*'■%•      J9«   1^51            yJi                       C^y                     -*-*'_inAiycA^^"'^   j^^y    VJ&\           ^ \^A^S, ^vCD_	           y    Y~\   --' "V    ^Bali^^r^irr—-^'i-^^81^•■■■''/'<- $'W&Aj- •*"" f"A""     <A^            ,-,\< ^' ■»„09
_L^        >YV*<?K;    ^   *■'->-?■h»      i,..tb_
CD,U.         ^»"'                                       V_^    *.,»„?          ^■•■'^.  £*•>        ^v       *f*''til'^T^A          o\y~" -CM
CD,|W'           .!--" ** .*, yy''<& JJ^^-i^^J^j ,'"V'"i,*y "OJ _
CDl-'J-OOUDjg^j               ^V       "*■!<"-,f y«.':.■S"\      ~—& ^  -<.iklA'                       O"«.,,:J.?X=                                 U J i. ..•'.-f ,1^                   ^•ivf..-"£                °\2"'#'   j •<*                    yID:£i ^ -^vy -y
ia        %   - °f"---<.  4F"^   si\ ^"--■■',|4yyiifC c* ^.y   °TI^y                     V,  A'                   a5-;0                   *"#         jKA'"''y% 2FCO
CO/3\ f] (A>	''"^RL/'cy   ^^y>4„.y  iCD
CD1      wI         \L£-J^.j.             KT         ^^I /i   /               -^!k^^-W^v                                                                         J^'rtjfW/ / /   \                "* \CD
COI          ^               ^1A^\±M*O£X      7 NsW \,       '^-xA       fy/v-?^ cola
\   ^?'-#y ^ "%^vAy* *        °-
/£'■■:■-::*     t            "=!;
).,;--_ vs.      ■■■.„           fpAn^-;ipjMSX^¥'.f m   C=\    or !
- V    ^     L'<-•°\       X      "*?) ''*'■''-:     W4^   't?j     \j&C^     Vb _-h%V'j       .:.&i     °o',..    \"*\     "  =      ?      ¥§     \1       # /           y^                      ^N —\ A~if,        y""4   i"7A\#      (       \                         ^"'&*-  \4.                                      i\ J        """"'-....Ml,., .	3»"''\   l-\'Af' ii >l***uaJ-icH*M*>-^'\    1 >vI          lot"Si      ''-■"-in	„                	 	1   =<JF   ^              /           A„              4(l^_ v/      \  y^yAA o.
yq\v'.if       / y.     	'•■VAV*
7\             /    ^-O*'               -H\^//^/My jf    ikA•'-'-^ "   j^jaA^^^V _'.-%.    *■	't           Y"	JJfcv^   XKO    A'"-J/i       y   Vfj^jriw^s^^^y^^co^>-Ay       ^A**~\fif!/lA^«    / ^ '\_y^A^_ CD.Mf#j*$A7\       fs~\    A77A"-7Z^7^^  ^^^X^CD _	o
CO%f7^y.ft .::''<i;    .,iiii""'/'	    yA-f/7^b_■ \i 7ff\DI JI avdivy^y1 iO-\<f       y    •'">' V'CDCM
1-     <7f    A            «
i   y- y         i^-                                  CM
1                                     1o
1CVJ                                    V                                     ^>
1                                         1                                         1CO                                  2
1            1o    .
•£ c
rt   o
a -55
2*1
a, ra
-fi "5
a, c
bO   eo
.s a
-o a
rt "O
H    ^
_g a
u   S
^^
a  -
2 ^
3 _g
S a
^

110

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6



basis of linguistic analysis. His reconstruction of the
origin and differentiation of Arawakan languages
leads him to conclude that "this most likely occurred
near the headwaters of the Ucayali and Madre de
Dios in what is now southeast Peru some 3500 to 5000
years ago" (Noble, 1965, p. 107). In Noble's view,
"the most divergent Tupian languages are also dis-
tributed in such a way as to suggest that they de-
veloped here" (op. cit., p. 109). Unfortunately, even
the minimal estimate for the Arawakan dispersal is
earlier than any known ceramic complex, including
the Zoned Hachure Horizon Style. Even if the date
were sufficiently recent, however, the reconstructed
phylogeny of the Arawakan family (op. cit., p. 108)
and the modern distribution of Arawakan speakers
(op. cit., map) fails to coincide with known occurrences
of any pottery style or complex of ceramic traits.
  The relatively small amount of work that has been
done, both in archeology and linguistics, makes it
likely that even the most generalized reconstructions
of Amazonian prehistory will be modified by future
work. A few basic themes seem visible, but most of the
variations are unknown. If the foregoing hypothesis
about the centrifugal effect of the ecological setting is
correct, the archeological picture is likely to become
more complicated with further investigations. As a
consequence, it will probably be many decades before
we begin to achieve full understanding of the manner
in which natural factors affected colonization of the
Amazonian lowlands by man.
  In the meantime, attempts to generalize are justi-
fiable principally because they offer  guidelines for

research. Along the Rio Napo, the combination of
poverty of the soil and widespread flooding creates
a habitat with little subsistence potential either for
gatherers or for cultivators. How widespread is this
ecological situation over the Amazonian lowlands?
On the Rio Napo, archeological evidence has led to
the inference that the area was colonized by at least
four groups probably representing varying levels of
sociopolitical complexity. The briefness of their resi-
dence suggests that whatever the initial expectations
may have  been,   the  subsistence  potential  did  not
meet them. The only solution was to move on. Al-
though   the   word   "determinism"   has   long   been
avoided  by  anthropologists,   present evidence  from
archeology,   geography,   soil   science,   and   ecology
suggests that the environment of the Rio Napo de-
termined an intermittent pattern of prehistoric human
settlement along its margins. Whether the introduc-
tion of new ingredients, such as crops of Old World
origin or future technological advances, can disrupt
the present ecological balance and create conditions
more favorable for human exploitation is yet to be
explored. For the past and present, it can be stated
with reasonable confidence that,  given the natural
conditions and the native American food plants, the
effect on human habitation was predictable, which is
to say inevitable. Whether this combination of factors
is common or rare over the lowlands as a whole is
the crucial question. The vastness of the area involved
and the paucity of investigations in all fields of science
underway or in prospect suggest that the answer may
be long in coming.

Literature Cited
BARBOSA RODRIGUES, JOAO
1877.   Arte cerarnica.    Antiguidades do Amazonas, vol. 2, pp. 2-23.    Riode Janeiro.
1891-1892.   Antiguidades do Amazonas.    Vellosia, vol. 2 (1885-1888), pp. 1-40 and vol. 4
(1885-1888), plates.
BULLEN, RIPLEY P.
    1965.    Archaeological chronology of Grenada.    American Antiquity, vol. 31, pp. 237-241.
BUSHNELL, G. H. S.
1946.   An archaeological collection from Macas, on the eastern slopes of the Ecuadorian Andes.
Man, 1946, no. 2, pp. 2-6.
CARVAJAL, GASPAR DE
    See Heaton, H. C; Millares, Jorge; Reyes y Reyes, Raul.
CHAVES CH., MILCIADES
1945.    La colonization de la Comisaria del Putumayo: un problema etno-economico-geografico
de importancia nacional.    Bol. Arqueologia, vol. 1, pp. 567-598.    Bogota.
COLLIER, DONALD, AND MURRA, JOHN V.
1943.   Survey and excavations in southern Ecuador.    Field Museum of Natural  History,
Anthropological Series, vol. 35.    Chicago.
CRULS, GASTAO
1942.   Arqueologia amazonica.    Rev. Servigo do Patrimonio Historico e Artistico Nacional,
vol. 6, pp. 169-220.    Rio de Janeiro.
CRUXENT, JOSE M., and ROUSE, IRVING
1958.   An archeological chronology of Venezuela.    Social Science Monographs VI.    Pan
American Union, Washington, D.C.
CUBILLOS, JULIO CESAR, and BEDOYA, VICTOR A.
1954.   Arqueologia de las riberas del Rio Magdalena, Espinal-Tolima.    Rev. Colombiana de
Antropologia, vol. 2, no. 2, pp. 115-144.    Bogota.
DE LA CRUZ, LAUREANO
1900.   Nuevo descubrimiento del Rio de Maranon llamado de las Amazonas, hecho por la
religion de San Francisco, afio de 1651.    Madrid.
DUQUE GOMEZ, LUIS
1963a. Los Quimbayas; Resefia etno-historica y arqueologica.    Historia de Pereira.    Bogota.
1963b. Resefia arqueologica de San Agustin.    Instituto Colombiano de Antropologia.    Bogota.
EVANS, CLIFFORD; and MEGGERS, BETTY J.
1960.   Archeological investigations  in  British  Guiana.    Bur.  American  Ethnol.  Bull.   177.
          Washington, D.C.
1962.    Use of organic temper for carbon  14 dating in lowland South America.    American
Antiquity, vol. 28, pp. 243-245.
1964.   British Guiana archaeology: a return to the original interpretation. American Antiquity,
vol. 30, pp. 83-84.
EVANS, CLIFFORD; MEGGERS, BETTY J.; and CRUXENT, JOSE M.
1960. Preliminary results of archeological investigations along the Orinoco and Ventuari
Rivers, Venezuela. Actas del 33 Congreso Internacional de Americanistas, vol. 2,
pp. 359-369.    San Jose, Costa Rica.
ill
293-822 O - 68 - 9

112	SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY	VOLUME   6
FERDON, EDWIN N., Jr.
1950.    Studies in Ecuadorian geography.   Monographs of the School of American Research,
no. 15.    Santa Fe.
FERIZ, H.
1963.   The ceramics of Tefe-Amana; a contribution to the archaeology of the Amazon.    Ethnos
1963, nos. 2-4, pp. 147-176.
FIGUEIREDO, NAPOLEAO, and SIMSES, MARIO F.
1963.   Contribuicao a arqueologia da fase marajoara.    Rev.  Museu Paulista, vol.   14,  pp.
455-470.    San Paulo.
FORD, JAMES A.
1949.	Cultural dating  of prehistoric  sites  in  Viru  Valley,   Peru.    American  Museum  of
              Natural History Anthropological Papers, vol. 43, part 1, no. 2.    New York.
GILLIN, JOHN
1936.   An urn from the  Rio Aguarico,  eastern  Ecuador.    American Anthropologist,  vol.
              38, pp. 469-470.
HANKE, WANDA
1959.   Archaologische  Funde  im  oberen  Amazonasgebiet.    Archiv  fur  Volkerkunde,   vol.
              14, pp. 31-66.
HEATON, H. D., Editor
1934.   The Discovery of the Amazon, according to the account of Friar Gaspar de Carvajal
and other documents.    American Geographical Society Special Publ. no. 17.    New
              York.
HEBERT, J.
1907.   Survivances decoratives au Bresil.   Journ.  Soc. Americanistes de Paris, vol.  4,  pp.
              185-191.
HILBERT, PETER PAUL
1952.   Contribuicao a arqueologia da ilha de Marajo.    Instituto de Antropologia e Etnologia
         do Para, publ. 5.   Belem.
1955.   A   ceramica   arqueologica   do   regiao   do   Oriximina.    Instituto   de   Antropologia   e
         Etnologia do Para, publ. 9.   Belem.
1959a. Achados arqueologicos num sambaqui do baixo Amazonas.   Instituto de Antropologia e
         Etnologia do Para, publ. 10.   Belem.
1959b. Preliminary results of archeological investigations in the vicinity of the mouth of the
Rio  Negro,  Amazonas.   Actas  del  33   Congreso  International  de  Americanistas,
         Tomo II, pp. 370-377.   San Jose, Costa Rica.
1962a. New stratigraphic evidence of culture change on the Middle Amazon (Solimoes).   Akten
         des 34 Internationalen Amerikanisten-kongresses, Wien,  1960, pp. 471-476.
1962b. Preliminary results of archeological research on the Japura River, Middle Amazon.
         Akten des 34 Internationalen Amerikanisten-kongresses, Wien,  1960, pp. 465-470.
n.d.      Archaologische   Untersuchungen   am  mittleren  Amazonas.   Marburger   Studien  zur
             Volkerkunde, Band 1, Dietrich Reimer Verlag, Berlin (in press).
IMBELLONI, J.
1950.	La extrana terracota de Rurrenabaque (noreste de Bolivia) en la arqueologia de Sud-
             america.   Runa, vol. 3, parts 1-2, pp. 71-169.   Buenos Aires.
IZUMI, SEIICHI, and SONO, TOSHIHIKO
    1963.	Andes 2: Excavations at Kotosh, Peru, 1960.   Tokyo.
JIJON Y CAAMANO, JACINTO
    1951.	Antropologia prehispanica del Ecuador, Resumen, 1945.   Quito.
LATHRAP, DONALD W.
1958.   The cultural sequence at Yarinacocha, eastern Peru.   American Antiquity, vol. 23,
         pp. 379-388.
1962.   Yarinacocha: Stratigraphic excavations in the Peruvian Montana.    Unpublished Ph.D.
Dissertation, Harvard University.
1964.	An alternative seriation of the Mabaruma Phase, northwestern British Guiana.   Amer-
ican Antiquity, vol. 29, pp. 353-359.
1965.	Investigaciones en la selva peruana, 1964-5.   Bol. Museu Nacional de Antropologia y
Arqueologia, ano 1, no. 4, pp. 9-12.    Pueblo Libre, Lima.
1967.   Report on the continuing program of research on the culture history of the upper
Amazon basin.   Dept. Anthropology, Univ. Illinois, Urbana (multilithed).

LITERATURE   CITED	113
MEGGERS, BETTY J. and EVANS, CLIFFORD
1957.	Archeological investigations at the mouth of the Amazon.   Bur. American Ethnology
Bull. 167.   Washington, D.C.
1958.	Archeological evidence of a prehistoric migration from the Rio Napo to the mouth of the
Amazon.   In Migrations in New World Culture History, Raymond H. Thompson,
Editor, University of Arizona Social Science Bull. 27, pp. 9-19.   Tucson.
1961.   An experimental formulation of horizon styles in the Tropical Forest Area of South
America.   In Essays in Pre-Columbian Art and Archaeology, Samuel K. Lothrop
and others, pp. 372-388.   Cambridge, Mass.
MEGGERS, BETTY J.; EVANS, CLIFFORD; and ESTRADA, EMILIO
1965.   Early Formative period of coastal Ecuador: The Valdivia and Machalilla phases.   Smith-
sonian Contrib. Anthropology, vol. 1.   Washington, D.C.
METRAUX, ALFRED
1930.   Contribution a l'etude de l'archeologie du cours superieur et moyen de 1'Amazone.   Rev.
Museo de la Plata, vol. 32, pp. 145-185.
MILLARES, JORGE HERNANDEZ, editor
1955.   Relation del nuevo descubrimiento del famoso rio Grande de las Amazonas, por Fray
Gaspar de Carvajal, O.P. Edition, introduction y notas de Jorge Hernandez Millares.
Fondo de Cultura Economica, Mexico-Buenos Aires.
MORDINI, ANTONIO
1947.   L'ile de Marajo (bas Amason): un probleme archeologique a resoudre.    Actes du 28
Congres International des Americanistes, pp. 637-642.   Paris.
NETTO, LADISLAU
1885.   Investigates sobre a arqueologia brazileira.   Archivos do Museu Nacional do Rio de
Janeiro, vol. 6, pp. 257-554.
NOBLE, G. KINGSLEY
1965.	Proto-Arawakan and its descendants.    International Journ. American Linguistics, vol.
             31, no. 3, part II.
NORDENSKIOLD, ERLAND
    1930.   L'archeologie du bassin de 1'Amazone.   Ars Americana, vol. 1.   Paris.
PALMATARY, HELEN C.
1950.   The pottery of Marajo Island, Brazil.   Trans. American Philosophical Soc, n.s., vol.
39, part 3.   Philadelphia.
PORRAS G., PEDRO I.
1961.    Contribution al estudio de la arqueologia e historia de los valles Quijos y Misagualli
(Alto Napo) en la region oriental del Ecuador, S.A.   Quito.
REICHEL-DOLMATOFF, GERARDO
1965a. Colombia: Ancient Peoples and Places, vol. 44.   London and New York.
1965b. Excavaciones arqueologicas en Puerto Hormiga  (Departamento de Bolivar).   Publ.
Universidad de Los Andes, Antropologia 2.   Bogota.
REYES Y REYES, RAUL, editor
1942.   Relation del nuevo descubrimiento del famosa Rio Grande que descubrio por muy
gran ventura el Capitan Francisco de Orellana, por P. Gaspar de Carvajal.   Biblioteca
Amazonas, vol. 1.   Quito.
RODRIGUEZ FABREGAT, E.
    1955.    Pasion e cronica del Amazonas.   Buenos Aires.
SILVA CELIS, ELIECER
1963.   Movimiento de la civilization agustiniana por el alto Amazonas.   Rev. Colombiana
de Antropologia, vol. 12, pp. 389-399.   Bogota.
SIMOES, MARIO F.
1966.	Resultados preliminares de uma prospecgao arqueologica na regiao dos Rios Goiapi
e Camara (Ilha de Marajo).   Atas do Simposio sobre a Biota Amazonica, vol. 2,
pp. 207-224.   Rio de Janeiro.
1967.	Considerac5es preliminares sobre a arqueologia do Alto Xingu, Mato Grosso. Programa
Nacional de Pesquisas Arqueologicas:  Resultados Preliminares  do   Primeiro Ano
1965-66.   Museu Paraense Emilio Goeldi, PublicacSes Avulsas No. 6, pp. 129-151.
Belem.
n.d.      The Castanheira site: New evidence on the antiquity and history of the Ananatuba
phase (Marajo Island, Brazil).   American Antiquity (in press).

114	SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY	VOLUME   6
Tosi, JOSEPH A., JR.
1960.   Zonas de vida natural en el Peru.    Instituto Interamericano de Ciencias Agricolas de
la OEA, Zona Andina, Bol. Tecnico no. 5.   Lima.
UHLE, MAX
1889.   Kultur und Industrie Sudamerikanischer Volker.   Vol. 1, Alte Zeit.   Berlin.
1921.   Los principios de la civilization en la sierra peruana.   Bol. Academia Nacional de
Historia, vol. 1 (1920) pp. 44-56.   Quito.
WOLF, THEODORE
1933.   Geography and geology of Ecuador.   Toronto.

APPENDIX
TABLES 1-11
TABLE 1.—Frequency of vessel farms of Yasuni Phase

N-P-10: Puerto Miranda HillN-P-l 1: Puerto
Miranda BankVessel
FormsMiranda
PlainYasuni
PlainMiranda
ModeledYasuni
IncisedYasuni
Incised
  and
PunctateYasuni
NickedYasuni
RedYasuni
Zoned
HachureMiranda
PlainYasuni
Plain Totals of
each FormNo.     %No.     %No.     %No.     %No.     %No.   %No.     %No.     %No.     %No.     %No.     %Common
Form
1
2
3
4
5
6
7
8
   9
Rare
Form
1
2
3
4
5
6
7
8
9
10
Unclass. 8    10.2
11     14.1
7	9.0
8	10.2
6      7.7
5	6.4
3      3.8
17    21.8
3      3.9
3      3.9
1       1.3
6	7.75      5.1
1       1.0
1       1.0
11	11.2
12	12.2
37    37.8
9      9.2
15    15.3
5      5.2
1       1.0
1       1.01   100.04    50.0
1     12.5
3    37.51    50.0
1    50.01    33.3
1    33.3
1    33.31   100.01    20.0
1    20.0
1    20.0
1    20.0
1    20.03	23.1
4	30.7
1      7.7
1      7.7
1      7.7
3    23.11    25.0
1	25.0
2	50.017      8.0
17      8.0
8      3.8
24    11.4
22    10.4
44    20.8
16      7.6
32    15.2
8      3.8
1      0.5
1       0.5
3       1.4
1      0.5
1      0.5
1      0.5
1      0.5
1      0.5
1	0.5
2	0.9
10      4.7TOTALS78  100.098 100.01   100.08 100.02  100.03    99.91   100.05  100.013  100.04 100.0211   100.0Base
Forms
A
B
C
D-1	16.6
3    50.0
2	33.42    50.0
2    50.0-      -2	20.0
3	30.0
3    30.0
2    20.0TOTALS6  100.0-   --   - j   -   - |   -  -4 100.0-      -10 100.0115

116

SMITHSONIAN   CONTRIBUTIONS  TO   ANTHROPOLOGY

VOLUME   6



TABLE 2.—Frequency of pottery types at sites of the Yasuni Phase

Surface and TestsPottery TypesN-P-10:
Puerto
Miranda
HillN-P-l 1:
Puerto
Miranda
BankNo.         %No.       %PLAINMiranda Plain	
Yasuni Plain	621      45.5
688      50. 327      77.2
8      12.8DECORATEDMiranda Modeled      	
Yasuni Incised and Punctate
Yasuni Nicked	
Yasuni Red	
Yasuni Zoned Hachure     .   .   . 5        0.4
27        2.0
5        0.4
5        0.4
7        0.5
7        0.5-GRAND TOTALS     ....
SUBTOTALS:
Plain types   .    .    .    ...
Decorated types .    .   .1365    100.0
1309      95.9
56        4. 135    100.0
35    100.0
0TABLE
3.—Frequency  of vessel forms  of  Cotacocha   Phase pottery

Vessel FormsN-P-3:
Nuevo
Rocafuerte-
SurfaceN-P-9:
Cotacocha-
SurfaceN-P-14:
Latas-
SurfaceN-P-l 5:
Tiputini
Road-
SurfaceNo.    %No.       %No.    %No.    %Common Form
1
2
      3
Minor Form
1
21  100.09      75.0
1	8.3
2	16.71    25.0
1	25.0
2	50.01     100.0TOTALS1  100.012    100.04 100.01     100.0Base Forms
A
B
C
D1  100.01      50.0
1      50.0-     -1    100.0TOTALS1  100.02    100.0-      -1    100.0Total Sherds1    -27592      -1
TABLE 4.—Frequency of vessel forms of ihe  Tivacundo PhaseN-P-8:N--P-7:  Chacra AlfaroBarranco
AlfaroTotalsVessel FormsSurfaceCut 1: Com-
bined LevelsSurface and
Testsof each
FormNo.%No.       %No.       %Common Form11914.69      10.75        5.23323627.723      27.47        7.36632015.416      19. 124      25.06043829.213      15.513      13.5645_-1        1.221      21.92261410.817      20.220      20. 751Rare Form132.31        1.21         1. 152---1         1. 113_---04---1         1. 115_--1         1. 11Unclass.--4       4.72        2.06TOTALS130100.084    100.096    100.0310Base FormsA215.43      25. 13      23. 18B969.27      58.27      53.823B: Perforated17.72      16.7-3C17.7-2      15.43D--„1        7.71E----        -0TOTALS13100.012    100.013    100.038TABLE 5.—Frequency of pottery types at sites of the Cotacocha Phase
N-P-l5:
Tiputini
Road-
Surface
No.    %
N-P-14:
Latas-
Surface
N-P-9:
Cotacocha-
Surface
N-P-3:
Nuevo
Pottery Types
Rocafuerte-
Surface
%
No.
No.    %
 5      5.4
87    94.6
134     48.7
1        0.3
129      47.0
Cotacocha Plain .
Cotacocha Red
Latas Plain  .   .   .
Unclassified
Cariape-
tempered
Plain	
No.    %
1    100.0
4.0
11
1  100.0
1    100.0
92 100.0
275    100.0
TOTALS
1  100.0

APPENDIX
TABLE 6.—Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Tivacundo Phase

117


Pottery TypesN-P-7: Chacra AlfaroN-P-8:
BarrancoSurfaceCut 1Alfaro0-8cm.8-16 cm.16-24 cm.Surface  &
TestsNo.%No.%No.       %No.%No.           %ATEDAlfaro Plain     	206
273
278
4224.1
32.0
32.5
4.925
23
4525.2
23.2
45.5157      26.9
166      28.5
237      40.642
55
51
126.6
34.8
32.3
0.6178        32.667        12.3Pi
0
nTivacundo Plain     	65        11.9Unclassified Cariape-tempered Plain	14          2.6PUnclassified Sand-tempered Plain	16          2.9fcTivacundo Incised and Zoned Red	 2
540.2
6.34
24.1
2.022        3.8
2        0.295.7187        34.3i—i
<Tivacundo Red Painted	10          1.8jUnclassified Decorated	9          1.6GRAND TOTALS	855
799
56100.0
93.4
6.699
93
6100.0
94.0
6.0584    100.0
560     96.0
24       4.0158
148
10100.0
93.6
6.4546      100.0SUBTOTALS:
Plain types	340        62.2Decorated types	206        37.8TABLE 7.—Frequency of stone artifacts at sites of the Napo Phase

Type of Stone Artifact

N-P-l
Area 2

N-P-l
Area 3

N-P-l
Area 4

N-P-2
Areas
1 &2

N-P-2
Broad-
side A

N-P-2
Broad-
side B

N-P-2
Cut 1
0-15
cm.

N-P-2
Cut 1
15-30
cm.

N-P-3
Test B

N-P-3
Test B
  Ex-
tension

N-P-4

N-P-6

Eden

Rio
Yasuni

Total
of each
type



Abrader     	
Ax (Notched) ....
Ax (T-shaped or eared)
Chisel	
Core	
Grinding Stone . . .
Hammerstone ....
Natural Pebble     .   .   .
Paint Stone	
Spall	

5
9
2
1
3
2
5
22
1
5

118

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6

TABLE 8.—Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Napo Phase

N-P-l: Tiputini

N-P-2: Nueva Armenia



Pottery Types

Surface and Tests

Surface and Tests



Area    1

Area 2

Area 3

Area 4

Area 1

Area 2



No.	%

No.	%

No.	%

No.	%

No.	%

No.	%



7
2
3
124
22
1
3
106
6
50
117
0
389
275
0
168
100
0
152
99
0
Armenia Plain	
Napo Plain     	
Napo Red	
Tiputini Plain	
Armenia White-on-Red    .   .
Napo Negative	
Napo Plain Excised ....
Napo Plain Incised ....
Napo Red Excised ....
Napo Red Excised, White Re
touched   	
Napo Red Incised	
Napo White Excised ....
Napo White Incised ....
Rocafuerte Incised ....
Rocafuerte Painted ....
Tiputini Plain Excised . . .
Tiputini Plain Incised . . .
Tiputini Red Excised . . .
Tiputini Red Excised, White
Retouched	
Tiputini Red Incised . . .
Tiputini White Excised . .
Tiputini White Incised . .
Unclassified decorated  .   .   .
Trade pottery     	
GRAND TOTALS     .   .   .
SUBTOTALS:
Plain types      .   .   .
Decorated types
Trade pottery    .   .


45
5
26.9
3.0
0.6
 1
17
0.6
10.2
1.2
94        56.3
2	1.2
167      100.0
30.0
70.0
0.0


359
27
3
54.2
4.1
0.4
1.1
0.3
0.4
18.7
3.3
0.1
0.4
16.0
0.9
0.1
664      100.0
58.5
41.5
0.0


142
25
1
53.1
9.3
0.4
1.1
 3
36
 1.1
13.4
14	5.2
37        13.8
6	2.2
0.4
268      100.0
62.7
37.3
0.0

42
537
97
10
1
117
35
14
141
9
26
6
 3
10
1073
686
387
0

3.9
50.0
9.1
0.9
0.1
0.7
10.9
0.7
3.3
0.7
1.3
13.1
0.8
2.4
0.6
0.3
0.9
0.3
100.0
63.9
36.1
0.0


19
90
29
14
 7.5
35.9
11.5
5.5
 6
31
2.4
12.4
1.2
2
2
39
7
3
0.8
0.8
15.6
2.8
1.2
0.8
1.2
0.4
251       100.0
60.6
19.4
0.0

13
1
2
2
17
2
1
16
4
2
68
16
51
1

19.1
1.5
2.9
5.9
 2.9
25.1
2.9
1.5
23.5
5.9
2.9
2.9
1.5
1.5
100.0
23.5
75.0
1.5

APPENDIX
TABLE 8.—Frequency of pottery types in surf ace collections and stratigraphic excavations at sites of the Napo Phase.—Continued

119



N-P-2: Nueva Armenia—Continued

N-P-3: Nuevo
Rocafuerte



Pottery Types

Surface and
Tests

Broadside A

Broadside B

Cut 1

Surface and
Test A



Area 2a

No.

%

No.

%

0-15 cm.

15-30 cm.



No.	%

No.	%

No.	%

No.

%



Armenia Plain
Napo Plain
Napo Red   .   .
Tiputini Plain
 
15
279
44
14
 
2.7
51.1
8.1
2.6

104
1325
259
207

3.6
45.5
8.9
7.1
 
87
682
132
86

4.7
36.8
7.1
4.6

235
827
118
79

12.3
43.4
6.2
4.1

148
203
42
96

23.3
32.0
6.6
15.6

47
172
51
12

10.1
36.9
10.9
2.6



88
6
7
11
3
2
Armenia White-on-Red    .   .
Napo Negative	
Napo Plain Excised   ....
Napo Plain Incised    ....
Napo Red Excised     ....
Napo Red Excised, White Re-
touched    	
Napo Red Incised	
Napo White Excised ....
Napo White Incised  ....
Rocafuerte Incised     ....
Rocafuerte Painted    ....
Tiputini Plain Excised  .   .   .
Tiputini Plain Incised   .   .   .
Tiputini Red Excised    .   .   .
Tiputini Red Excised, White
Retouched	
Tiputini Red Incised     .   .   .
Tiputini White Excised     .   .
Tiputini White Incised     .   .
Unclassified decorated  .   .   .

6	1.1
9	1.6
41	7.5
1	0.2
0.2
1.1
1.5
16.1
1.1
1.3
2.0
0.5
0.4
0.9

21
1
21
292
25
11
27
92
8
315
43
76
41
14
10
10
3

0.7
0.1
0.7
10.1
0.8
0.4
0.9
3.2
0.3
10.8
1.5
2.6
1.4
0.5
0.3
0.3
0.1

58
29
194
10
4
15
 2
51
3
326
44
78
14
13
18
2
6
4

3.1
1.6
10.4
0.5
0.2
0.9
0.1
2.7
0.2
17.6
2.3
4.2
0.8
0.7
0.9
0.1
0.3
0.2

5
5
 25
353
17
1
28
6
115
33
29
29

0.3
0.3
1.3
18.5
0.9
0.1
1.4
0.4
6.0
1.7
1.5
1.5
0.1

3
1
7
23
1
15
4
47
15
13

0.4
0.1
1.1
3.6
0.1
2.4
0.6
7.4
2.4
0.2

1
7
46
3
 1
10
1
 2
51
 8
19
5
 2
20
1
7

0.2
1.5
9.9
0.6
0.2
2.1
0.2
0.4
10.9
1.7
4.2
1.1
0.4
4.3



Trade pottery

0.2



1895
1010
6
489
146
0
352
194
0
1259
648
0
GRAND TOTALS
SUBTOTALS :
Plain types
Decorated types
Trade pottery

546      100.0
64.6
35.4
0.0

2911       100.0
65.1
34.7
0.2

1858      100.0
987
871
0
53.1
46.9
0.0

1907      100.0
66.0
34.0
0.0

635      100.0
77.0
23.0
0.0

466
282
184
0

100.0
60.5
39.5
0.0

120

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY	VOLUME   6
TABLE 8.—Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Napo Phase.    Continued



    N-P-3: Nuevo
Rocafuerte—Continued

N-P-4: Bello
Horizonte
 
N-P-5:
Florencia

N-P-6: Puerto
Alfaro

Oasis



Pottery Types

Test B

Test B—
Extension

Surface and
Tests

Surface and
Tests

Surface and
Tests

Surface



No.

%

No.

%

No.

%

No.

%

No.

%

No.

%



11
185
10
11
21
173
48
19
59
310
10
62
26
232
27
16
3
1
5
3
203
2
27
3
2
1
1
3
1.5
0.3
2.0
301
185
0
441
299
1
261
193
0
Armenia Plain	
Napo Plain     	
Napo Red	
Tiputini Plain	
Armenia White-on-Red    .   .
Napo Negative	
Napo Plain Excised ....
Napo Plain Incised ....
Napo Red Excised ....
Napo Red Excised, White Re
touched    	
Napo Red Incised	
Napo White Excised ....
Napo White Incised ....
Rocafuerte Incised ....
Rocafuerte Painted ....
Tiputini Plain Excised . . .
Tiputini Plain Incised . . .
Tiputini Red Excised . . .
Tiputini Red Excised, White
Retouched	
Tiputini Red Incised . . .
Tiputini White Excised . .
Tiputini White Incised . .
Unclassified decorated  .   .   .
Trade pottery     	
GRAND TOTALS     .   .   .
SUBTOTALS:
Plain types     .   .   .
Decorated types
Trade pottery    .   .
 
5.4
47.8
5.6
3.3
0.6
44
3
9. 1
0.6
0.2
1.6
7
 2
65
 3
31
4
10
1.4
0.4
13.4
0.6
6.4
0.8
2.0
0.4
0.4
486      100.0
62.0
38.0
0.0

8.0
41.9
1.3
8.4
1. 1
 1
36
0. 1
4.9
0.4
0. 1
0.7
0.4
27.4
0.3
3.6
0.4
0. 1
0.4
0. 1
741       100. 0
59.5
40.4
0. 1
 
4.6
38. 1
10.6
4.2
2.0
11
51
2.4
11.2
13
2.9
3	0.6
49	10.8
2	0.5
26	5.7
6	1.3
23
5. 1
454      100.0
57.5
42.5
0.0

3.9
67. 1
3.6
 1
15
5
1
19
13
3.9
0.4
5.5
1.8
0.4
6.9
4.7
0.7
0.7
0.4
276	100.0
217	78.6
59	21.4
0	0.0

6	0.6
775	79.3
51	5.2
19	2.0
0.2
2
53
4
0.2
5.4
0.4
23
2.4
15
3
19
0. 1
0.4
977	100.0
851	87. 1
126	12. 9
0	0.0

58
1
16
5
1
3
1
94
58
36
0

61.8
1. 1
17.0
5.3
1.1
3. 1
1. 1
4.2
2. 1
1. 1
2. 1
100.0
61.8
38.2
0.0

APPENDIX

121

TABLE 9.    Frequency of Napo Phase plain and decorated pottery classified by temper and frequency of incised and excised decoration classified by single- and double-
line techniques

N-P-l:TiputiniN--P-2: Nueva ArmeniaSurface and TestsSurface and TestsArea 1Area 2Area 3Area 4Area 1Area 2No.%No.%No.         %No.%No.%No.         %Temper of all plain and decorated pottery
types:167100.0  6
658 1.0
99.0   1          0.3
267       99.722
70
9812.0
6.5
91.520
46
1858.0
18.3
73.72         2.9
7        10.0
59        87.1Sand	Total sherds per site unit used for percen-
tage calculations	167100.0664100.0268      100.01073100.0251100.068        100.0Motifs of all incised and excised decorated
types:
Single-line	 2
20 9.1
90.9  6
147 3.9
96.1 6        10.2
53        89.853
16324.5
75.515
4226.5
73.58          26.5
22          73.5Double-line      	Total sherds used for percentage calcula-
tions     	22100.0153100.059      100.0216100.057100 030        100 0122

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY

VOLUME   6

TABLE 9.—Frequency of Napo Phase plain and decorated pottery classified by temper and frequency of incised and excised decoration classified by single- and double-
line techniques.—Continued

N-P-2: Nueva Armenia—ContinuedN-P-3: Nuevo
RocafuerteSurface and
Tests—Con.Broadside ABroadside BCut 1Surface and
Test AArea 2a0-5 cm.15-30 cm.No.         %No.%No.%No.%No.%No.        %T emper of all plain and decorated pottery
types:
Black ash     	38          6.9
33          6.0
475        87. 1303
215
239310.4
7.4
82.2153
204
15018.3
10.9
80.8126
466
13156.6
24.6
68.8103
234
29816.2
36.4
47.426         5.6Cariape	75        16 1Sand     	365        78 3Total sherds per site unit used for percent-
age calculations	546      100.02911100.01858100.01907100.0635100.0466      100 0Motifs of all incised and excised decorated
types:
Single-line	29        33.2
58        66.8194
44830.2
69.8175
30536.5
63.591
42417.8
82.243
4648.2
51.855        44.8
68       55.2Double-line	Total sherds used for percentage calcula-
tions     	87      100.0642100.0480100.0515100.089100.0123      100.0APPENDIX

123

TABLE 9.—Frequency of Napo Phase plain and decorated pottery classified by temper and frequency of incised and excised decoration classified by single- and double-
line techniques.—Continued

N-P-3:Nuevo Rocafuerte--Con.N-P-4: Bello
HorizonteN-P-5FlorenciaN-P-6: Puerto
AlfaroTestBTest B—ExtensionSurface and TestsSurface and TestsSurface and TestsNo.%No.%No.           %No.%No.           %Temper  of all  plain  and  decorated  pottery
types:
Black ash	34
40
4127.0
8.3
84.7100
84
55713.5
11.3
75.250            10.6
46              9.5
358           79.911
24
2413.9
8.7
87.427              2.7Cariape	10              1.0Sand	940            96. 3Total sherds per site unit used for percentage
calculations	486100.0741100.0454          100.0276100.0977          100.0Motifs  of all  incised  and  excised  decorated
types:50
6344.3
55.736
4643.9
56. 157            43.2
75            56.819
2147.5
52.523            22.0Double-line	82            78.0Total sherds used for percentage calculations .113100.082100.0132          100.040100.0105          100.0293-822 O - 68 - 10

124	SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY	VOLUME   6
TABLE 10.-—Frequency of vessel forms and base forms from surface collections and stratigraphic excavations of the sites used in the seriated sequence of the Napo Phase

N-P-l: TiputiniN-P-2: Nueva ArmeniaVessel FormsSurface and TestsSurface and TestsBroadside ABroadside BArea 2Area 3Area 4Area 1Area 2aNo.%No.         %No.%No.%No.         %No.         %No.%Common Form11417.67        28.0118.426.91           1.415          4.241.82---21.5--2         2.86          1.710.43---21.513.49        12.59         2.5104.6411.2-75.3--2         2.820         5.662.7511.21          4.021.5--4         5.63         0.952.36--2          8.0--517.4-4          1. 120.8722.5-43. 126.97         9.732          9. 1167.3822.5-64.513.43          4. 116         4.5125.5911.2-96.926.91           1.416          4.53716.91033.81          4.043. 1--2          2.88          2.2104.61122.51          4.043. 126.98        11. 118          5. 1135.9121316.31          4.01914.526.96          8.340        11.33516.21356.35        20.0107.626.97         9.729          8.3177.81445.0-139.913.44          5.619          5.462.7151518.82          8.043. 1---4          1. 131.31667.52          8.0139.9413.95          6.952        14. 7104.61767.51          4.0129.213.49        12.546        13.02913. 41811.21          4.010.813.4-          -5          1.4--1911.2------          --          ---20----13.4-3          0.9--Rare Form1--_          _____20.82------_-          --          ---3--------          -1          0.3--411.21          4.0----1           1.41          0.3--5---10.8___          __          ___Unclassified22.5-75.326.91           1.47          1.910.4TOTAL80100.025      100.0131100.029100.072      100.0355      100.0219100.0Base FormsA2-11_2_41312_B
C_:—1-1-12
31-D7-25-1_15__E——-1-2--61-TOTAL9-38-6-62914-Total Sherds664-2681073-251-54629111858-APPENDIX

125

TABLE 10.    Frequency of vessel forms and base forms from surface collections and stratigraphic excavations of the sites used in the seriated sequence of the Napo
Phase.—Continued

N-P-2Nueva Armenia--Con.N-P-3: Nuevo RocafuerteN-P-4: Bello
HorizonteVessel FormsCut 1Surface and
Test ATest BTest B—
ExtensionSurface and0-15cm.15-30cm.TestsNo.%No.%No.           %No.           %No.           %No.           %Common Form142.411.52            3.4_           __            _6          13.0231.8--1            1.7---           -363.734.42            3.45            8.8-1            2.1453.1--2            3.41            1.73            4.13            6.4521.211.5-            -1            1.76            8.3-642.411.51            1.71            1.75            6.9-7169.834.43            5.13            5.35            6.94           8.5810.657.32            3.42            3.55            6.93            6.49159.1913.25            8.58          14.116          22.32            4.210106.123.01            1.71            1.71            1.42            4.211137.9710.36          10.21            1.72            2.81            2.1124225.71522.016          27.012          21.11            1.415          32.01353.111.53            5.19          16.01            1.4-1453.123.01            1.74            7.04            5.64            8.515--11.5-1            1.7-1            2.11674.257.38          13.6-12          16.72            4.21774.2--4            6.86          10.59          12.5-1863.734.4---1            2.119----2            3.4---20--11.5--1            1.4-Rare Form
 1
2
3
4
5.._„     .1            1.4.     .10.61
11.5
1.5---Unclassified127.368.7-2            3.5-2            4.2TOTAL164100.068100.059        100.057        100.072        100.047        100.0Base FormsA10_8_5172B
C
D2
1
1-1--112___3_            _1E3-----1-            -TOTAL17-9-8295Total Sherds1907-635-466486741454126

SMITHSONIAN   CONTRIBUTIONS   TO   ANTHROPOLOGY	VOLUME   6
TABLE  11.—Frequency of technique variants of Armenia White-on-red, Rocafuerte Incised, and Rocafuerte Painted


N-P-l: TiputiniN-P-2: Nueva ArmeniaPottery Types andSurface and TestsSurface and TestsBroadside
ABroadsideTechnique VariantsArea 1Area 2Area 3Area 4Area 1Area 2Area 2aBNo.%No.      %No.      %No.      %No.      %No.      %No.      %No.      %No.      %ArmeniaWhite-on-redWhite-on-red10.67      1.13      1.11    0.1_4      5.96     1.121    0.758    3.1Black and white-on-red .-    —Rocafuerte IncisedWhite incised, redretouched   .----      -9    0.8_     _--    -3    0.12    0.15White incised and redand black-on-white  .5    0.52      0.8--    -5    0.21    0.05Rocafuerte PaintedRed and black-on-white9254.867    10.235    13.1116 10.830    12.013    19.154     9.9231    7.9257 13.8Red-on-white21.228      4.22      0.78    0.7-     -1       1.510     1.848    1.731    1.6Black-on-white--11       1.6-17    1.69      3.62      2.924     4.436    1.238    2.1Total sherds per siteunit used for per-centage calculations .167—664     -268     -1073    -251      -68      -546     -2911    -1858   -APPENDIX
TABLE  11.—Frequency of technique variants of Armenia White-on-red, Rocafuerte Incised, and Rocafuerte Painted.—Continued

127


   N-P-2: Nueva
Armenia—ContinuedN-P-3: Nuevo RocafuerteN-P-4:
Bello
HorizonteN-P-5:
FlorenciaN-P-6:
Puerto
AlfaroOasis Pottery Types and
Technique VariantsCut 1Surface
and Test
ATestBTest B—
ExtensionSurface
and TestsSurface
and TestsSurface
and TestsSurface0-15 cm.15-30 cm.No.      %No.      %No.      %No.      %No.      %No.      %No.      %No.      %No.      %  Armenia
White-on-red
White-on-red
Black and white-on-
red5    0.33     0.41      0.22      0.4
1      0.27      1.0
1      0.19     2.01     0.42     0.2-     -Rocafuerte Incised
White incised, red
  retouched   .
White incised and red
and black-on-white   .1    0.1
5    0.34     0.62      0.41      0.2
1      0.22      0.3
1      0.13     0.6Rocafuerte Painted
Red and black-on-
white
Red-on-white
Black-on-white88    4.6
27    1.4'39     6.1
1     0.2
7     1.148    10.3
2      0.4
1      0.262    12.8
2      0.4
1      0.2196    26.5
3	0.4
4	0.542     9.3
1     0.2
6     1.319     6.915     1.54     4.2Total sherds per site
unit used for per-
centage calculations .1907    -635     -466     -486     -741      -454    -276     -977     -94     -

Plates 1-94




r.
&
y^:'^ A A>£>*. > <
~r;*  'y^y—v	A*^€5*1


PLATE 1
Air views of the Rio Napo. a, Large sand bars exposed in December. Roofs at N-P-l: Tiputini show as bright spots at the center left;
grass covered hills at N-P-2 at center top. b, Upper Rio Napo, with modern garden clearings in the foreground.
293-822 O - 68 - 11






s8¥   "XisfilMfi^s^


PLATE 2
Views of the upper Rio Napo.   a, Looking from Latas toward the right bank across a small rapid,   b, Low left bank near Santa Rosa
strewn with drift trees.




*fc»


a






^-****«*
k^MKU»^4&

PLATE 3
Views of the Rio Napo. a, Left bank near the mouth of the Rio Suno, with Sumaco volcano in the distance, b, Low right bank above
Armenia Vieja, showing large tree deposited by flood waters, c, Looking downstream just below the mouth of the Rio Tiputini,
with the cleared hill at N-P-10 visible at the center.



                                               PLATE 4
Decline in water level of the Rio Napo opposite N-P-2 during a six week period in November-December 1956.
                                           

"y-■•* <
fir >T







PLATE 5

c, Small inboard motor
Views of the Rio Napo, showing varieties of modern water transportation,   a, Poling dugout opposite N-P-2.  b Port at Nuevo Rocafuerte.
Sm^l inboard motor launch at the junction of the Rio Napo (right) with the Rio Yasuni (foreground).






                                                                 PLATE 6
Yasuni Phase sites,   a,  N-P-10 on the summit of a low hill,   b, N-P-l 1  (beyond buildings) seen from the hill occupied by N-P-10.
                                                    
smMmW^WQEm

W§&»,

•i





AAJr
5^ 1/6
•% 'Tf 'Jv A-■""-'«•-


                                                      PLATE 7
Views of the Rio Tiputini.   a, Looking downstream from N-P-6.   b, Mouth, with left bank of the Rio Napo in the distance.
                                                


■VJASA







M^'!> ^-A:"*T-^
rf










"t$07fm*
> A?- 1       ?•>;' ^J£-d*1
*. Thyfrfad,



I	L

J	L
2     3     4     5  CM

                                                          PLATE 8
Tivacundo Phase pottery types and artifacts,   a-e, Chacra Plain rims,  f-g, Chacra Plain perforated bases,   h-i, Potrest fragments.
                                                    







-A


/
.yt viTV^'A,-"''l^r^^y1 ." *
.4- i"^.- ".; ■•'j y: S.r""*
■':  &   . - A,^ AA     .
Id	M' ■ '    ; .-•
b-*iOfrS'i^k:.':S)<F"'>-'     ■ i'ziP'*1"~-~i"V"-1- V>^
IBA I y^ y !.;$? ftjMt?-   «x
.^Ay£yy >A>.
' ^y-y"%«* $ .A,

I	1	1	I	L_J
0    I    2   3   4   5 CM
b-e

                                                            PLATE 9
Tivacundo Incised and Zoned Red.   a, Oval vessel (cf. fig. 23).   b-d, Type sherds, Technique 1.   e, Type sherds, Technique 2.
                                                


?M&
$M07f^f< ' -v^y y

^yyfA.A,,yy >      '   .
:s-y y>\»Hr
■/■■










0     I     2     3    4     5 CM

                             PLATE 10
Type sherds of Tivacundo Plain,   a-h, Rims,   i-j, Perforated bases.
                          


•v   i  «     ■ >.•: i.   - ^
xOir *■-■!.■■■*,
1> '
t<r

:■  J




.
I	I	I	I	I	I
0      1       2      3      4      5 CM



T^-<"~

^%I^^^A:S^A^y"-A.y:-.-;,y y-
-'•^^^-v■'"'*' 't ■-.   -~,   . •-.   r&l    :.*-  ■: : --

^

•>-y.:.>--- i*-*=,

"Sfe^aftj^S
                                                 PLATE 11
Tivacundo Phase pottery types,    a-e, Tivacundo Red Painted.  /, Unclassified cariape-tempered plain.
                                       


                                                                 PLATE 12
Sites of the Napo Phase,   a, N-P-l: Tiputini, with the original surface elevation visible beneath the porches,   b, N-P-3: Nuevo Rocafuerte.
                                                    





a


Wma-s

PLATE 13
N-P-2: Nueva Armenia, a site of the Napo Phase,   a, Looking from slightly downstream toward the site (right),   b, Site area covered
with modern buildings, seen from the hill behind.

tfiimlii

                                               PLATE 14
Napo Phase site of N-P-2: Nueva Armenia,   u, Northeast end of the site (Area 2).   b, Excavation of Cut 1.
                                          


PLATE 15
Sites of the Napo Phase,   a, N-P-4: Bello Horizonte on the high bank at the center,   b, N-P-5: Florencia, on the level area to the right
of the hill.


PLATE 16
Views of the Rio Tiputini.   a, Bank occupied by the Napo Phase site of N-P-6.   b, Lower bank on the opposite side of the river.


I	I	I	I	L_
0    I    2    3    4    5 CM


PLATE 17
Potrest fragments from Napo Phase sites,   a-b, d,f-h, Examples showing exterior surface treatment,   c, e, Interior texture and manner of
fracture.
293-822 O - 68 - 12




-• '. ; '" "
'■)  **%     "■'■'■■;
&^JjBrffSE» f


«sw





-   "    V
•- I-'  *

I	I	I	'	'	'
0     I     2    3    4     5 CM

PLATE 18
Type sherds of Armenia Plain,   a-b, Rims,    e-h, Surface treatment,   i-l, Leaf impressions on exterior of bases.


PLATE 19
Type sherds of Armenia White-on-red.







I	I	I	L
(Mill
0	5 CM
0    I    2   3   4   5 CM

PLATE 20
Type sherds and complete vessel of Armenia White-on-red.   a-b, Rare variant painted black and white-on-red.   c, Vessel of Form 17,
from the Rio Aguarico (Peabody Museum of Archaeology and Ethnology, Harvard University).









I     I    '	I	I	I
0    I    2   3    4   5 CM



                                                                PLATE 21
Type sherds of Napo Plain,   a-g, Rim sherds,   i-k, Typical fracture along coil junctions,   h, Leaf impression on exterior of base.
                                                   







■


PLATE 22
Complete  vessels of Napo Plain,   a, Form 8, rim diameter 23 cm.    b-c, Rare Form 1, rim diameter 12 cm. (U.S. National Museum).


           PLATE 23
Type sherds of Napo Plain Excised.
          

0    I    2   3   4   5 CM


           PLATE 24
Type sherds of Napo Plain Excised.
          

-^p^*^C











. W-l"'t^mj^Qp*»
a r"Mi»iimni^iinL^
I	L
j	1
0      I     2     3     4 CM

dUHi|

                                              PLATE 25
Type sherds of Napo Phase pottery types,   a-c, e, Napo Plain Excised,   d, f, Napo Negative.
                                          







L5CM


                                             PLATE 26
Complete vessels of Napo Plain Incised,   a-b, Form 9.   c-d, Form 8.   e-f, Form 7. (U.S. National Museum)
                                        







f ^#«ife^i^i
■ -i 1
I    1   V   f:  ''   '#y-

i    V c

i^S**



I	L

J	I

0    I    2    3    4   5 CM


••'' <>    t.
S^::l.   A
-yy ■. *.
■'■'A:'"-■'"■'■'       is .'M-
s** «■	\\     \\  v       .
%

♦ y^Ss*?5*^?

           PLATE 27
Type sherds of Napo Plain Incised.
          
i   v     f V --yy'l . -^   Uv:y$»E
0    I    2   3   4   5 CM
•    4T-   ■■•-. ._

PLATE 28
Type sherds of Napo Plain Incised.


y--   ■3ry>
"VV.
lyteyy y^y 4.•■•
Ai<tA.-y^-:A^l;vt.,->\ • y,h:«-.   «^\-.    s


           PLATE 29
Type sherds of Napo Plain Incised.
          

"^'/MJ


           PLATE 30
Type sherds of Napo Plain Incised.
          
-AT.-.-

0     1      2     3     4     5 CM
'

                                                                 PLATE 31
Type sherds of Napo Plain Incised showing rare combination of single-line and double-line techniques (a-b,  d-e, Heye  Foundation).
                                                    

                                         PLATE 32
Type sherd and bowl of Napo Red Excised,   a, Heye Foundation,   b, Courtesy Howard S. Strouth.
                                     

I     t	I	I	i—I
0     I    2    3    4   5 CM




293-822 O - 68 - 13
          
PLATE 33
Type sherds of Napo Red Excised.
          


•r  .;  AyA-alv
& -■    w,y^ 2$
K:	7^,^-7










I	I	J	I	!	I
0      I      2     3     4     5 CM

^sapBSS^

                    PLATE 34
Type sherds of Napo Red Excised, White Retouched.
                  

I	i	I	I	L
0     I     2    3    4    5  CM


          PLATE 35
Type sherds of Napo Red Incised.
         
^•n^"'	i   \   f   -A*^A'    ' yJ9Sr/
:




&5'AA$A'





"mirOF





a







Ar
AAA-Ay
o.mmssg»
A  j    «£

•™'9w r'1H$:   fF^f^w fnS
i—i—i—i	i	i
0   1    2   3   4   5 CM

            PLATE 36
Type sherds of Napo White Excised.
          

r^.-y
AyA



\ r^-'7-r-9*
9   ^>.   .    •      ^-^A-       .-     r;y.
'M.f
    •rag*-., v
•y     --. ^*r*v«-r>
%sr
J	I
I	I	L
0   1    2   3   4   5 CM

            PLATE 37
Type sherds of Napo White Incised.
          


11 "










*A#M
I	L
J	L
0    I     2    3    4    5 CM

           PLATE 38
Type sherds of Napo White Incised.
          




3«





1     11i1_J8%20     12345 CMP           PLATE 39
Type sherds of Napo White Incised.
          





I—I	1	1    I    '
0    I    2    3   4   5 CM

I	1	1	I	L_J
0   I   2 3 4 5 CM
PLATE 40
Type sherds of Napo White Incised.


I	I	I	I	I	I
0     I     2    3    4    5 CM

PLATE 41
Bowls of Rocafuerte Incised combined with red and black-on-white painting,   a, N-P-3, Test B (U.S. National Museum),   b, Unidentified
provenience (American Museum of Natural History).



PLATE 42
Basin of Form 14 with Rocafuerte Incised decoration along one side and red and black-on-white painting in Technique 2 on the remainder
of the interior. From N-P-2. (Courtesy Museo Arqueologico del Banco Central del Ecuador, Quito.)



r. &*   .: w  ' -\y (
_J	I	I	I	I
0    I    2    3   4   5 CM
a,b,e
I	I	I	I	I	I
0     I     2     3     4    5 CM
c,d,f

                                                                 PLATE 43
Type sherds of Rocafuerte Incised,   a-b, e, Combined with red and black-on-white painting.   c-d,f, Red paint applied only to incisions.
                                                    







■   -J. i\


i  i
0    11	1     I     1
2	3    4   5
CMfy*q                    PLATE 44
Type sherds of Rocafuerte Painted, red-on-white.
                

                                                      PLATE 45
Type sherds of Rocafuerte Painted, black-on-white,   a-f, Technique 3.   g-h, Technique 1.   i-j, Technique 2.
                                           

PLATE 46
Rocafuerte Painted vessel of Form 6 from a site on the Rio Aguarico. Painting is red and black-on-white, Technique 1 on
the body and black-on white, Technique 3 on the rim (cf. fig. 54).   Height 34 cm. (U.S. National Museum).







0     I     2    3    4     5 CM
i:«i^rat^>fs*r^hf!^m ...» -f    ~~AA

I       I	I	L

                                                       PLATE 47
Type sherds of Rocafuerte Painted, red and black-on-white,   a-b, Technique 1.   c-d, Technique 3.   e, Technique 4.
                                            

 I     I     I     I     I     I
0   I   2  3 4   5CM
a-c, e


PLATE 48
Type sherds of Rocafuerte Painted, red and black-on-white, Technique 2.


yfj*y   i

293-822  O - 68 - 14
                          
PLATE 49
Type sherds of Rocafuerte Painted, red and black-on-white, Technique 5.
                          

PLATE 50
Vessels of Rocafuerte Painted, black-on-white, a, Jar of Form 16 painted in Technique 1. b, d, Jars of Form 17 painted in Technique 3
c far of Form 17 painted in Technique 3 and supplied with a lid of Form 12 (in Jijon y Caamafio collection, Quito, courtesy of Pedro
I. Porras).


I	I	I	I	I	I
0	5 CM


PLATE 51
Partially restored bowls of Napo Phase pottery types,   a, Rocafuerte Painted, red and black-on-white, Technique 5. b, Tiputini Red
Excised.







I	I	I'll	I	I	I	I	I
0	5
0 CM


PLATE 52
Rocafuerte Painted  (red and black-on-white) vessel of Form  11 with anthropomorphic treatment (Museo Victor   Emilio   Estrada,
Guayaquil).


or   "



PLATE 53
Rocafuerte Painted (red and black-on-white) vessel of Form 11 with anthropomorphic treatment. (Courtesy of Museo Arqueologico del
Banco Central del Ecuador, Quito).


&
0
'En
UJ
"o
U
O    u
■*  p
CO  ^


CO
H
<
.t! co
A +>
£ A
<As
9 A
-M .„
1,3 A
-rt   U
0 a
IH
0
>


P-i a
<M -a
'S <
H «

-O  O
-rt  CM


H
<
OH

'ifl <
^~, CO
—1 CO
CM    ^
£ -^
O    (U
BE
< J
OJ C
&. O
o
y
a, c
o --1
A   «
c ^
rt  <:
Ti  «
£j  o
£ -4
2il
^ Ti
<5 ^
U     UJ
O     SH
fC


PLATE 56
Rocafuerte Painted anthropomorphic vessel Type A, painted red and black-on-white in Technique 2. (Courtesy Jay C. Leffand Brooklyn
Museum.)   See table A for detailed description.











PLATE 57
Rocafuerte Painted  anthropomorphic vessel Type B, painted red and black-on-white in Technique 2. (Courtesy Thomas Flannery.)
See table A for detailed description.




52   '> '      4

£

■I   %	i	-
«   '   #     k *...   ■ W  .    .        ■■','.'»
J^*l
   *y.p..,
=«y iy


SH    .3
3 t:0 JSS"S3-•   0S feO     urtH        ^^ffi•S >*be §3-d  3.„ 0CM  OV.2* uG  CO-S 20 coH^c  S°•- 'JJoj AA «> u7 ^c  2*? '3Jj x0 0«a i>3HcTJ —G   ^cort .tilO_, xl«Hu   C<73   Oj£ rt5sPHG   u
.a rtrt -2S.^^T5«   Jj«.aCL rt>. CXHM"wi  f-H
>  M
OJ   o
<;   u
;>   rt
'&      JJ
.S     SH
u ,0
PH
PH t«
§ Si


rt
rt
s-
5
a
rt
rt
O
XI
G
-rt    O
T3    -^
IS


P5
a



CM      .
si
a
4

A   Q
o    W
v   uj
H T3
fl ^
.H JJ
8 ^
!■§
O  *"
rtH<C
rt Ji
X)   +J
G   v
-rt W
UJ
SH   /-N
X3 T3
JJ     UJ
5 3
'^  ft
PH    JJ
PP."^
JJ   rt
PH   O
H CH
^H   rt
cn    be
§3 2
>  o
X   ^
a



>■
H
^G
„   O
V  '£
S3
§1
a I
""■ X3
x) rt
y

O     W
1      u-^ «3ort     .3y	,      SHX)   o5 to<£>IHH
x) artSH       rt<SHX!    3
<U    rtjrt   aPH.5 Zrt t_PH    OFQ §3v   <u*9ucn     SH•—I   4>I £<y <> *•—-o   v11118	&
-£	x)
c	2
03	S
X)	'rt
<u	PH
G
'rt
rt



*v.£iy
CM
IP
rt
H
<
rt
U
rt

-a &
rt rt
§3d
CM  <J
Ti ^
rt a
y—-    SH
G   O
CO
s.u.
CO
V   3
3   cr1
.ET'S
-■ H
H   G
c —
•-<   UJ
CD .-^
H3|
G   O
o   rt
*S
-rt ^3
Ti c
C   rt
rt
x>TJ
t>      SH
SH
-a "^
G.e
PH ^

«J cr
% 2  ■
g.a G
offi.9
X	-3    PH
o a a
CL,^  TJ
•5 a'3
C	3	<j
rt	v	TJ
	.	VI
XI	3	SH
8   ^   s9
'3 gy
Cu   rt   <u
„ O
3 i«
<x < y
rt O^c/3
C/0
c_j
PH1



rt
rt
<
PH

rt
rt
>

sCM
u3O"'3Xo«jrtC
P.3'-Pa,«j'u^cjXcn
CJ£XIcXIoV-i'rtort<L>3X)SHXI
Gc2rt<-aJJCjSH3rtTJcuVcOJC/J.'rtPHduil,OJcrart>>>>rt rt
3"o0cnOJ>uXauO£op.oSHXGPH
P<


Ml •——&


PLATE 64
Rocafuerte Painted anthropomorphic vessel Type C, painted red and black-on-white in Techniques 2 (body) and 3 (arm and legs).
(Colegio Militar, Quito.)    See table A for detailed description.


XJ
G
rt
y
X!
3   u
FT" U
Cx)
O  x)
4J     g
h rt

A£$<G
O<UJiArtHQV<V)T)G.	,
aj   O
SH    +JJ
X) hH
UJ   HH
£ ~rt
'rt  h
PH  B
« rt
u£
si0
H  B
IS
o   5.
•fi .y
PH    U
U      V
M
p o
PH

293-822  O - 68 - 15


                                                   PLATE 66
Tiputini Plain Excised bowl of Form 5; height 8 cm., exterior mouth diameters 23 by 25 cm. (U.S. National Museum).
                                              

I	I	I	I	L_
0    I    2   3  4   5 CM


             PLATE 67
Type sherds of Tiputini Plain Excised.
           


Mr *'SLvNBkKW"n1     1l          1i      >10     123    45 CM             PLATE 68
Type sherds of Tiputini Plain Excised.
           


                                  PLATE 69
Type sherds of Tiputini Plain Excised with white retouch in incisions and excisions.
                              

j&tflr ~* ■


PLATE 70
Type sherds of Tiputini Plain Incised.






PLATE 71
Type sherds of Tiputini Red Excised.




-f^j-f^j^ljjirnr*









I—I—I     I
I     2    3    4    5 CM


            PLATE 72
Type sherds of Tiputini Red Excised.
           



a




J	i	\	i	i
I    2   3   4   5 CM


                                                PLATE 73
Napo Phase pottery types,   a-l, Tiputini Red Excised, White Retouched,   m-n, Tiputini White Excised.
                                      







m



i	i	i	i	i	i
0     1     2    3    4    5 CM

PLATE 74
Napo Phase pottery types,   a-i, Type sherds of Tiputini Red Incised,   j-o, Type sherds of Tiputini Red Incised, white retouched
variety.



'A)A '. 'A ■-. 7






J	L
o

J	I	I
I      2      3      4     5 CM

             PLATE 75
Type sherds of Tiputini White Incised.
            


**ti&$&r

             0%M
I	I	I	I	I	I
0   I   2   3 4  5 CM


0    I    2    3   4   5 CM


                                          PLATE 76
Pottery from Napo Phase sites,   a-h, o, Unclassified decorated,   i-n, Sherds of probable trade origin.
                                      


j*CT: -j



L
0


J	I
J	I	L
I     2     3     4     5  CM














*M


                             PLATE 77
Cotacocha Phase pottery types,   a-e, Cotacocha Plain.  /-/, Latas Plain.
                          

■<PH	IfimMmiy*   ■jr:tm*m* ■
\  V'V(.,i'',y.   '"^--.    '■- *»










|y • Ay'AyV-^

m
^0<:A^

PLATE 78
Sherds from Peruvian complexes represented in the Zoned Hachure Horizon Style,   a-l, Waira-jirca Phase (after Izumi and Sono, 1963,
pi. 836).   m-t, Tutishcainyo Phase.


PLATE 79
Decorated sherds from Coari.   a, White slipped after incision,   b-f, h-i, Grooved, sometimes with a double-pointed or serrated tool.
g, j,   Red-on-white,    k, Black-on-white,   l-m, Single-line incised. (Courtesy Peter Paul Hilbert.)


PLATE 80
Type sherds of the Guarita Phase from Refinaria Site, Manaus. a-j, Guarita Grooved, k, Everted rim with grooved and polychrome
(red and black-on-white) decoration, l-m, Flanges with grooved and polychrome painted (red and black-on-white) decoration.
(Courtesy Peter Paul Hilbert.)








a

V
■■■-
/


/
'7;ro
h












<f

A
B


♦fefl1?
J	I
I	i	L
0      I      2      3     4     5 CM
J	I
I	I	L
0   I   2   3  4  5 CM

PLATE 81
Decorated sherds from Itacoatiara.   a-f, Excised,   g-i, White slipped after incision,  j-k, Single-line incision,   l-p, Double-line incision.
q-r, Grooved,    {a-b, d,f, i, k-m,p, r, Courtesy Peter Paul Hilbert, Scale A; c, e, g-h,j, n-o, q, Museu Paulista, Scale B).
















«***
# *^

*

-*fc-,.

•

*>*.

I	I	I	!	'	'
0   1    2   3   4   5 CM

PLATE 82
Decorated sherds from Manacapuru.   a-b, e, i, White-slipped after grooving.   c-d,f, Excised,   g-h, j, Double-line incised.   £, Red and
black-on-white.   /, Black-on-white.    (Museu Paulista.)



PLATE 83
Vessels from Manacapuru.   a, Flanged vessel white slipped after grooving,   b, White slipped vessel with tendency to square outline.
(Museu Paulista, Sao Paulo)




y>





L
J	I
0       I        2       3        4       5 CM

PLATE 84
Decorated sherds of Marajoara Phase pottery types,   a-c, Arari Plain Excised,   d, Arari Plain Excised, White Retouched,   e, Arari Red
Excised.  /, Arari Double-slipped Excised. (American Museum of Natural History)



yyy

^





>A



i	i	i	i	i	i
0     I     2     3    4    5 CM

PLATE 85
Decorated sherds of Marajoara Phase pottery types,  a, Arari Plain Excised,  b, Anajas White Incised,  c, e-f, Pacoval Incised, d, Arari Red
Excised. (American Museum of Natural History)










PLATE 86
Marajoara Phase bowls with squarish form,   a-b, Pacoval Incised,   c, Arari Red Excised, White Retouched, (a, Heye Foundation, b-c,
American Museum of Natural History)



cH

I	1	1	I	'      '
0     I     2     3    4    5  CM

PLATE 87
Type sherds of Joannes Painted, Marajoara Phase,   a, c, Red-on-white,   b, Red and black-on-white. (American Museum of Natura
History)



irt
in
3  3
v ^
3 G
bn >.
rt -£
£ rt
<y o
a^
<
~" X
X
PH
SH
P
3
o
PH
P



V '''                 mm      -H%^M* .;
J

u

w
rt
<
rt

rt
rt
|3
^ P
CD    vO
O    +3
3	O
rt	<->
§	o
rt	0
rt	z.
" 5
cn   "3
3   «
ss
3 CS-
he
G
rt   d
rt   5
.3    P
rt    G
«   3
5  >
< "*
rt S















&
-<FW\1 S P
£►.   tar- ..■■.*'   '
0     I

4     5 CM

*yAs^

NEW**!


PLATE 90
Decorated sherds from Nazare dos Patos.   a-h, Grooved, typically with a double-pointed or serrated tool.  /, h, White slipped after
grooving,   i-j, Excised. (Goteborg Museum, Sweden)














L.
0
J	L
I       2       3       4 CM


                                                       PLATE 91
Excised and incised sherds from middle Amazon sites,    a-c, Paura.   d, Tauaquera.   e-g, Pontao. (Goteborg Museum, Sweden)
                                                 









i^-S*^^

Xv-y;-|^i^A-:^

I	L
J	L	I
0     I     2    3    4    5 CM

4 • ^^pq^-yit^^wrr" *
v      if

PLATE 92
Decorated sherds from the vicinity of Santarem.   a-g, Double-line incised,   h-l, Single-line incised, (a-b, d-e, Goteborg Museum; c, f-l,
Heye Foundation)







I	I	I	I	■	I
0     I     2    3    4     5 CM
                                            PLATE 93
Decorated sherds from the Tefe region,   a-c, Excised,   d-h, Broad grooved. (Museu Paulista)
                                   


7 :'v
e














I	L

J	I	I
2     3     4     5 CM

PLATE 94
Decorated sherds from Loreto, Rio Suno, Ecuador,   a-d, Zoned punctate,   e-i, Unzoned punctate,   j-k, Possible Napo Phase trade.
(Courtesy Pedro I. Porras)
U.S.  GOVERNMENT  PRINTING OFFICE : 1968 O - 293-822