TRANSACTIONS OF THE 
 
 CONNECTICUT ACADEMY OF ARTS AND SCIENCES 
 
 Incorporated A. D. 1799 
 
 YOLUME 21, PAGES 315-442 MAY, 1917 
 
 A Survey 
 
 of 
 
 Ancient Peruvian Art 
 
 BY 
 
 PHILIP AINSWORTH MEANS 
 
 YALE UNIVERSITY PRESS 
 
 NEW HAVEN, CONNECTICUT 
 1917 
 
TRANSACTIONS OF THE 
 
 CONNECTICUT ACADEMY OF ARTS AND SCIENCES 
 
 Incorporated A. D. 1799 
 
 VOLUME 21, PAGES 315-442 MAY, 1917 
 
 A Survey 
 
 of 
 
 Ancient Peruvian Art 
 
 BY 
 
 PHILIP AINSWORTH MEANS 
 
 YALE UNIVERSITY PRESS 
 
 NEW HAVEN, CONNECTICUT 
 
 1917 
 
THE TUTTLE, MOREHOUSE St TAYLOR COMPANY, NEW HAVEN, CONN. 
 
TABLE OF CONTENTS 
 
 PAGE 
 
 I. INTRODUCTION 3 T 9 
 
 Sketch Map of Peru .... facing 321 
 
 II. THE CULTURE PERIODS OF PERUVIAN 
 
 ART 3 22 
 
 1. The Proto-Chimu and Proto-Nasca Cul- 
 
 tures 3 22 
 
 2. The Culture known as Tiahuanaco I. . 324 
 
 3. The Culture called Tiahuanaco II. . . 326 
 
 4. The Red-white-black and Epigonal Cul- 
 
 tures 328 
 
 5. The Chimu and Nasca Cultures . . 329 
 
 6. The Colla-Chulpa Culture . . . 330 
 
 7. Early Inca Culture 333 
 
 8. The Inca Culture at its Height . . 336 
 
 III. ANALYSES OF THE ARTS OR CULTURES . 339 
 
 1. A Critical Analysis of Proto-Chimu and 
 
 Proto-Nasca Art 339 
 
 2. A Critical Analysis of Tiahuanaco II Art 351 
 
 3. Relations between Proto-Chimu, Proto- 
 
 Nasca and Tiahuanaco II 363 
 
 4. A Critical Analysis of tlie Epigonal and 
 
 Red-white-black Cultures . . . 369 
 
 5. A Critical Analysis of the Chimu and 
 
 Nasca Cultures 372 
 
 6. A Critical Analysis of Late Inca or Cuzco 
 
 Art 377 
 
3 iS Table of Contents. 
 
 PAGE 
 
 IV. THE QUESTION OF CHRONOLOGY AND 
 
 DATES IN EARLY PERUVIAN ART . 383 
 
 APPENDIX I: THE ARCHAIC TYPE . . .390 
 
 APPENDIX II : CHRONOLOGICAL ORDER OF THE 
 
 EARLY PERUVIAN CULTURES . . 393 
 
 BIBLIOGRAPHY 394 
 
 LIST OF ILLUSTRATIONS 403 
 
I. INTRODUCTION. 
 
 It will be the writer's endeavor to present in this paper a brief 
 review of the various types of art to be found in pre-Columbian 
 Peru. The work is the fruit of some four years' study, two 
 years of that period having been devoted to a systematic collec- 
 tion of data in various places and under the direction of various 
 people. As the main purpose will be to establish a basis for 
 the classification of Peruvian art-objects, the study will be con- 
 fined to those regions where the form and stratigraphic relations 
 of the various art-types that make up the sequence of cultures 
 have been determined with a reasonable degree of precision. The 
 reader is urged carefully to bear in mind the fact that many of 
 the various types are to be found in regions far removed from 
 those here to be specified. But in those regions which are far 
 from the source of an art-type or culture new environmental 
 and psychological conditions almost inevitably exercise an influ- 
 ence which results in profound modifications of the original type. 
 The writer hopes that this paper will help to link certain of the 
 Peruvian arts or cultures with certain types of objects from 
 such regions as Ecuador, Eastern Bolivia, North-western Argen- 
 tina and Northern Chile. It may even be possible in time to 
 gather material evidence which will conclusively prove the basic 
 unity of all the more advanced types of art in aboriginal America. 
 
 In the writer's opinion it is still too early to attempt, with 
 any likelihood of success, to read or interpret the inner signifi- 
 cance of- the various designs that we shall study. Attempts of 
 this nature have been made by Berthon, (1911), Joyce, (1913b), 
 Posnansky, (1914), and others, but still it seems to be unavoid- 
 able under the present limitations of our information that all 
 speculations of this sort should lack an atmosphere of conclu- 
 siveness. In this day, with our present incomplete knowledge 
 of these ancient peoples, we should not attempt to read into their 
 exotic designs a set of significances expressed in terms of our 
 own experience. Rather, the investigator should seek painstak- 
 ingly to analyze the various component parts of each pre- 
 Columbian art or culture, as well in Brazil, Argentina, Chile, 
 Ecuador, Colombia, Panama, and Middle America as in Peru 
 
320 Philip Ainsworth Means, 
 
 and Bolivia, with a view first to finding out the distribution of 
 each and every element, and ultimately to arriving at some safe 
 and permanently tenable opinion as to the cultural ancestry of 
 each of the cultures that have flourished in the several regions. 
 
 The writer also believes that it is time for a serious attempt 
 to be made to construct for the various cultures of pre-Columbian 
 Peruvian art a chronology, supplied with approximate dates, simi- 
 lar to the one already established for the Maya area. In order to 
 arrive at any permanently valuable opinion as to the cultural 
 position and cultural ancestry of these Peruvian art-types, it 
 will be necessary first to know, at least approximately, when and 
 how long they flourished. For many years it has been the 
 fashion for South American archaeologists to look askance at all 
 efforts to construct a chronology. The recent researches of 
 Dr. Uhle, of the late Sir Clements Markham, of Sr. Arturo 
 Posnansky, of the late Dr. Gonzalez de la Rosa and of others 
 have, however, afforded material that seems to justify a formal 
 undertaking of the construction of a date-chronology for the 
 various Peruvian cultures. The author has already made a 
 tentative effort in this direction, 1 and the reception it has met 
 with has encouraged him to pursue the matter further. It is 
 inevitable that discussion of this important matter should finally 
 result in the establishment of a reasonably correct date-chronol- 
 ogy. Accordingly, in the hope of bringing that desideratum of 
 Peruvian archaeology nearer, he has ventured to insert at the 
 end of this study a tentative date-chronology of the various art- 
 periods or cultures of early Peru. 
 
 The author is greatly indebted to many people for the aid, 
 of various sorts, that they have given him during the preparation 
 of this paper. Chief among these are the following: Dr. Roland 
 B. Dixon, of Harvard University; Dr. Alfred M. Tozzer, of 
 Harvard University ; Dr. George F. Eaton, of Yale University ; 
 Professor George Grant MacCurdy, of Yale University ; Dr. 
 Herbert J. Spinden, of the American Museum of Natural His- 
 tory ; Mr. Charles W. Mead, of the American Museum of 
 Natural History ; Professor Marshall H. Saville, of the Museum 
 of the American Indian ; Mr. Sylvanus Griswold Morley, of the 
 Carnegie Institution ; Dr. Ales Hrdlicka, of the United States 
 
 1 Means, 1917. 
 
Lombayehi 
 Chfc/^. , 
 
 r-^acosmayo ' ))» 
 Truxillo ^f> 
 
 SKETCH MAP OF PERU 
 
 Showing the locations of the chief sites mentioned in the body 
 
 of the paper. 
 
A Survey of Ancient Peruvian Art. 321 
 
 National Museum ; Mr. F. W. Hodge, of the Smithsonian Insti- 
 tution ; and, Mr. Thomas A. Joyce, of the British Museum. To 
 all these gentlemen the writer wishes to extend his thanks for 
 their help. 
 
 Acknowledgments are also due to the authorities of the 
 Peabody Museum, Cambridge, Mass., to those of the Museum of 
 Fine Arts, Boston, and to those of the American Museum of 
 Natural History for permission to figure various objects in their 
 collections. Mr. Guernsey of the Peabody Museum, Cambridge, 
 was so kind as to help the writer in taking some of the pictures 
 that accompany the paper, and Dr. Denman Waldo Ross was so 
 good as to spend a long time discussing the aesthetic side of the 
 designs on several of, the textiles here illustrated. 
 
II. THE CULTURE PERIODS OF PERUVIAN ART. 
 
 Before proceeding to a detailed analysis or to any endeavor 
 to coordinate the various cultures, it will be best for us to state 
 as briefly as possible what the periods of culture are and where 
 each is found at its highest development. Appendix II shows 
 their chronological position with respect to one another, and the 
 accompanying map shows the location of the chief sites connected 
 with each of the cultures. It remains for us to summarize the 
 outstanding features of the various types. 
 
 i. THE PROTO-CHIMU AND PROTO-NASCA CULTURES. 
 
 One may conveniently distinguish between the two subdivi- 
 sions of this earliest coast culture-period by remembering that 
 the Proto-Chimu flourished all along the northern half of the 
 Peruvian littoral and the Proto-Nasca along the southern half. 1 
 This subdivision is arbitrary, being based on the form of arts 
 prevailing in the two regions. It is not a wholly satisfactory 
 classification, and it may ultimately have to be modified. For 
 example, it may sometime become desirable to delimit at Pachaca- 
 mac a style which should be called "Proto-Pachacamac." Our 
 information is, however, too scanty to justify such a course as 
 yet, and it is better for the present to rely upon the classification 
 here offered, which does preserve and emphasize the main lines 
 of differentiation between the major varieties of the earliest 
 coast art. 
 
 The art of the region around Chan Chan and Moche 2 in the 
 modern department of Libertad is characterized by features that 
 
 1 The terms "Proto-Chimu" and "Proto-Nasca" were adopted by 
 Dr. Uhle after he discovered that the objects belonging to them did not 
 belong to the Chimu and Nasca cultures. The name by which the early 
 but highly gifted people called themselves is unknown. 
 
 2 Here again, the nomenclature must be commented upon. The two 
 places just named are near Trujillo and they are the chief sites for Proto- 
 Chimu ware. The name Chimu is used for the same sites at a later period, 
 when the Chimu culture was flourishing. Chimu is derived from the 
 Mochica place-name Chimorr or Chamorr ; Moche is the Hispanicised 
 form for Muchik; Chiclayo was formerly Chajaep; Lambayeque is 
 derived from Nampajek. Cf. Middendorf, 1892, p. 64. 
 
A Survey of Ancient Peruvian Art. 323 
 
 set it in sharp contrast to other Peruvian art-types. As a rule, 
 the other Peruvian cultures are marked by conventionalization. 
 The Proto-Chimu, on the contrary, is comparatively free from 
 conventionalization and is marked by strong realism, especially 
 in the animal forms, "portraits" and "landscapes." In close 
 association with the elaborate modelling in the round went 
 painted decorations of a type always easy to identify. These 
 paintings were usually in dark reddish brown on a cream-colored 
 slip. In a few cases such colors as light red, orange and buff 
 were used in the vase-paintings. The outlines of the figures 
 are marked by a grace that is unusual in Peruvian art, and in 
 the grouping of the various scenes a striking command of the 
 principles of composition and grouping is displayed. Some of 
 the vase-paintings of this period partake of the nature of genre 
 paintings, and they help us in no slight degree to reconstruct the 
 material culture and customs of the people whom they depict. 
 
 It is but right to say here a word or two regarding the reasons 
 that have led Uhle, Joyce and several others to believe that the 
 Proto-Chimu and the Proto-Nasca are the earliest Peruvian arts. 
 The architecture associated with remains of this culture takes 
 the form of massive walls built up of large balls of clay placed 
 in position while still wet and allowed to dry in such a manner 
 that they partly ran together, thereby forming a solid mass 
 of material. Stratigraphic evidence proves that this architecture, 
 of which only a little is left, is the oldest. 3 
 
 Reserving further comments on Proto-Chimu art for a later 
 page, we will now run over the outstanding features of Proto- 
 Nasca art, always bearing in mind the fact that it was probably 
 not only contemporaneous with Proto-Chimu but also closely 
 associated with it on ethnic grounds. 
 
 Undoubtedly Proto-Nasca will, in time, serve more truly to 
 explain certain problems than will Proto-Chimu. At the same 
 time, regarded merely as an art, it is not so remarkable. It is 
 more like other Peruvian arts, for reasons that will later appear. 
 Unlike the Proto-Chimu, Proto-Nasca is not characterized by 
 graceful modelling and graceful painting. Rather, it sacrifices 
 both the form of the vessels and the lines of the paintings to a 
 remarkable wealth of coloration. To the novice, it is true, the 
 
 3 Joyce, 1912, p. 179; Uhle. 1913, pp. 102-103; Means, 1917. 
 
324 Philip Ainsworth Means, 
 
 Proto-Nasca vessels appear sombre enough, but the more one 
 studies them the more he becomes impressed with the wonderful 
 richness and variety of their tints. The mere fact that most of 
 them are from the dark side of the color-scale does not impair 
 the effect of subdued richness. If, then, we never find in Proto- 
 Nasca the astonishingly good modelling that excites wonder, 
 and sometimes amusement, at the Proto-Chimu art, the lack is 
 in part made up for by the presence of sumptuous color combina- 
 tions that may well give valuable hints to modern artists. 
 
 It is the opinion of Mr. Joyce that no textiles of this period 
 have survived to the present time. 4 But for reasons to be 
 enlarged upon later, the present writer ventures to hold the con- 
 trary opinion on this point. 
 
 Though profoundly different, as has been shown, the Proto- 
 Chimu and Proto-Nasca arts have similarities to one another 
 that are quite as significant as their divergences. The similarities 
 are to be found in the subject-matter of the two arts rather than 
 in the details of their execution. In both, the use of headdresses 
 decorated with animal-faces is apparent; in both, the use of 
 various sorts of masks and of eye-painting is noticeable; and 
 in both the centipede-like tail ending in a human face is often 
 found. An important article by Mr. Joyce affords the material 
 for forming these opinions. 5 
 
 2. THE CULTURE KNOWN AS TIAHUANACO I. 
 
 The researches of Posnansky, Uhle, Gonzalez de la Rosa and 
 others have established the fact that the remains at and around 
 Tiahuanaco in Bolivia represent two sharply differentiated 
 cultures. Of these, the cruder was the earlier. Posnansky, to 
 whom the subdividing is chiefly to be credited, calls this first 
 and simpler epoch "Tiahuanaco Primitivo." The writer, in 
 
 4 Joyce, 1912, p. 200. 
 
 s Joyce, 1913b. 
 
 ,: Though we shall fall in with modern usage and employ the name 
 Tiahuanaco, it is to be noted that the early name for the place appears 
 to have been Taypicala. This, according to Cobo (IV, p. 65) and Ban- 
 delier (1911, pp. 222 and 243), has the meaning of "Stone-in-the-Center 
 (of the Universe)." The word appears to be derived from the "Aymara" 
 (correctly, Colla) terms taipiri, center, and ccala, worked stone. (Cf. 
 Vocabulario poliglota incaico, 1905.) 
 
A Survey of Ancient Peruvian Art. 325 
 
 seeking for a good English equivalent for this term, decided to 
 adopt one that was suggested by Aegean archaeology — hence 
 "Tiahuanaco I," and also "Tiahuanaco II." 
 
 The architecture of Tiahuanaco I was true megalithic masonry. 
 In building a wall, the early Tiahuanaco people adopted the 
 simple but effective method of setting up at intervals large 
 vertical oblong monoliths. In the edges of these nearest to the 
 next pillar grooves were often cut from the base to the top and 
 into them the builders fitted other blocks of stone by means of 
 which a wall of comparatively small stones was made between 
 the large ones. 
 
 In all probability Tiahuanaco I was contemporary, at least 
 in part, with Proto-Chimu and Proto-Nasca. Nevertheless, as 
 will be developed later, there is no trace of the early coast types 
 to be found associated with Tiahuanaco I deposits. It is, in 
 the writer's opinion, impossible to say with accuracy whether 
 or not any pottery or textiles have survived from the Tiahuanaco 
 I period. Posnansky, however, figures two rude stone heads used, 
 apparently, as wall-ornaments, dating from this period. 7 
 
 Even a brief study of Tiahuanaco I reveals the fact that it 
 is totally unlike either of the probably contemporaneous coast- 
 cultures. What, then, is it like? Is it an indigenous and autoch- 
 thonous culture? The whole trend of modern investigation into 
 the ancient cultures of America discourages belief in the autoch- 
 thonous nature of the Tiahuanaco I culture. It must, therefore, 
 have been derived from some other region. As it obviously is 
 not connected, even remotely, with any of the other cultures 
 in South America that can possibly have been contemporary with 
 it, save for one possible exception, we must study, however 
 briefly, the strands of evidence that bind it to the group of cul- 
 tures which constitute that exception. It is, then, suggested that 
 the erectors of the Tiahuanaco I culture were related to, or even 
 members of, the great Arawakan stock of Brazil. This is as 
 yet but a theory. Facts, however, lend it a certain color of truth. 
 These facts we will briefly outline. 
 
 Far to the south-east of Lake Titicaca, in the Bolivian province 
 of Santa Cruz, is a site called Samaipata which yields cut rocks 
 very suggestive of the stone-work of Tiahuanaco I. We owe 
 
 Posnansky, 191 1, p. 33. 
 
326 Philip Ainsworth Means, 
 
 our knowledge of this place to Baron Nordenskjold, and it is 
 his opinion that the remains at Samaipata are associated with 
 Arawakan builders. 8 Archaeology, then, offers a slender thread 
 with which to bind the Tiahuanaco I culture with the Arawakan 
 stock at Samaipata. But this is not all the evidence afforded by 
 archaeology. The island of Marajo, at the mouth of the Amazon, 
 yields evidences of occupation by a people who had a stone 
 technique of a grade similar to that of the Tiahuanaco I people. 
 Finally the characteristic feature of the better sort of Tiahuan- 
 aco I stone-carvings is the continuity of the eyebrows and nose 
 so as to form a T-shaped figure. 9 This feature is also found in 
 some of the pottery heads from Marajo in the Peabody Museum, 
 Cambridge, and likewise it is observable on the secondary decora- 
 tions of the Weeping God figure at Tiahuanaco. (See Plate 
 VII.) 
 
 Furthermore, linguistics and a study of migrations seem to 
 throw some light on the situation. Haddon indicates roughly 
 that there was a shift of peoples from north-eastern South 
 America toward the Titicaca and Samaipata regions. Chamber- 
 lain and others indicate that members of the Arawakan linguistic 
 stock are to be found far over toward the Andes at the latitude 
 of Lake Titicaca. 10 
 
 On the whole, then, there is a certain justification for sug- 
 gesting that the first high-cultured dwellers at Tiahuanaco were 
 derived from stock belonging to the eastern half of the continent. 
 The reader is reminded, however, that this whole point is in 
 an embryonic state of discussion. Only long and systematic 
 work will definitely establish the Arawakan derivation of the 
 Tiahuanaco I people and their culture. 
 
 3. THE CULTURE CALLED TIAHUANACO II. 
 
 If Tiahuanaco I was probably contemporary with the Proto- 
 Chimu and Proto-Nasca cultures of the coast, Tiahuanaco II 
 is no less probably derived, at least in part, from the latter of 
 those two coast cultures. This will be enlarged upon later on. 
 
 s Nordenskjold, 1902, 1906, 1906b. 
 
 9 See Posnansky, 1914, Plate XXXX. 
 
 10 Haddon, 1912; Chamberlain, 1913b, p. 474 ff. 
 
A Survey of Ancient Peruvian Art. 327 
 
 There can be but little doubt that the culture which we call Tia- 
 huanaco II was the most highly developed in South America. 
 It even rivals the Maya culture of the "Old Empire" cities in 
 the southern part of Yucatan. 
 
 It may be true that it is dangerous to measure the actual spread 
 of a culture by the boundaries of the territory within which 
 remains of its distinctive products are to be found. Political 
 affinities, of course, cannot be determined by any such evidence ; 
 but, nevertheless, the fact that Tiahuanaca II objects are found 
 from Colombia to Argentina is a proof that the cultural domi- 
 nance of Tiahuanaco II was exceedingly widespread. 
 
 As we have seen, there was a shift, in the transition from 
 Proto-Chimu to Proto-Nasca, from a light-toned art enriched by 
 good modelling to a dark-toned art characterized by poor and 
 slight modelling. For reasons to be brought out later it is but 
 natural that we should find the characteristics of Proto-Nasca art 
 carried on to their logical development in the art of Tiahuanaco 
 II. This natural state of affairs is found to exist. 
 
 In Plate VII we see an important portion of the largest mono- 
 lithic gateway at Tiahuanaco. It may safely be said to be an 
 epitome of Tiahuanaco II art. Its characteristics, from our 
 point of view, may be listed thus : ( 1 ) A headdress decorated 
 with ray-like tabs. (2) Square-headed chief figure with round 
 eyes from which run down the "tears." (3) A short stout body 
 with a necklace and a short, skirt-like garment held up by bands 
 that run over the shoulders. (4) Four-digit hands holding cere- 
 monial staves. All these elements will, of course, be analyzed 
 in full later on. 
 
 In general terms, one may say that Tiahuanaco II art, whether 
 in stone, pottery, textiles or bronze, is the most elaborate we have 
 yet seen. Birds with human bodies, pumas, fishes and other ani- 
 mal forms combine with almost innumerable conventionalized 
 decorations to form an art of surpassing complexity. In the 
 pottery of this period we find a sacrifice of coloration to a per- 
 fection of the almost glaze-like finish. In other cases, however, 
 Tiahuanaco II pottery has neither rich coloration nor fine finish. 
 Red and black are the chief colors employed, though sometimes 
 white is found as well. The textiles, however, naturally preserve 
 a wider range of tint. Unfortunately, most of those that have 
 survived into our day come from the coast and so do not repre- 
 
328 Philip Ainsworth Means, 
 
 sent Tiahuanaco art as having the austere elaborateness that 
 marks it in the highlands. Indeed, this characteristic of the 
 textiles of Tiahuanaco II on the coast may have been found also 
 on the pottery from that region and period. It may well have 
 been a heritage from the rich-tinted Proto-Nasca period. 
 
 In many ways the civilization of the Tiahuanaco II "Empire" 
 was the highest that ever flourished in pre-Columbian America. 
 As has been said, it may not have been wholly a political 
 "empire," but it is probable that all through the wide area where 
 Tiahuanaco II objects are found there was a constant interchange 
 of ideas and merchandise. This opinion is borne out by the 
 fact that all the chief edifices at Tiahuanaco itself were of mas- 
 sive stone. On the coast, however, where the earlier people 
 (Proto-Chimu and Proto-Nasca) had used adobe and where 
 stone was not easily obtainable, the Tiahuanaco II people adapted 
 the old clay-ball architecture of their predecessors to their needs 
 by modifying the clay balls into real bricks of sun-dried clay. 
 These bricks, or adobes, ranged in size from seven or eight 
 inches in length to three feet or more. Different sized adobes 
 were used for different needs, just as different sized stones were 
 used in the similar circumstances. 
 
 4. THE RED-WHITE-BLACK AND EPIGONAL CULTURES. 
 
 In general, it may be said that the red-white-black ware fol- 
 lowed the Tiahuanaco II period of the north part of the coast, 
 and that "epigonal" ware was distinctive of the southern part 
 of the coast. Both were the successors of Tiahuanaco II, and 
 both, especially the "epigonal," were influenced by it and by 
 the earliest cultures. In this period the architecture remained 
 much the same as in the preceding one, and the only radical 
 difference that exists between Proto-Chimu and red-white-black 
 on the one hand and Proto-Nasca or Tiahuanaco II and "epi- 
 gonal" on the other is that neither of the later types were as 
 technically admirable as the earlier ones. 
 
 Leaving for a later page the discussion of the details of this 
 art-period, we will mention the only hint we possess of who the 
 makers of the red-white-black ware were. It seems that the 
 dynasty of Chimu was preceded in a portion of the north part 
 of the coast by another dynasty called Naymlap whose chief seat 
 
A Survey of Ancient^ Peruvian Art. 329 
 
 was Lambayeque. 11 The Naymlap people came from the north 
 by sea, and they built up a state that was apparently conquered 
 by the chief Chimu. All this, however, is as yet mere unsub- 
 stantiated theory. 
 
 5. THE CHIMU AND NASCA CULTURES. 
 
 With this period one begins to get some hint of the political, 
 social and ethnological conditions under which the people lived. 
 Several authors, ancient as well as modern, give valuable infor- 
 mation on this head. 12 
 
 All that, interesting though it is, lies without the scope of the 
 present paper. We will therefore content ourselves with observ- 
 ing that in the period which we are now considering the north- 
 ern portion of the coast, from Tumbez down to Casma, was under 
 the sway of a great chief known to the Incas as Chimu Capac 
 (Great Chimu). The valleys of Rimac, Pachacamac and Chan- 
 cay were ruled by another great chief called Cuismancu. Runa- 
 huanac, Huarcu, Mala and Chilca were ruled by Chuquimancu. 
 Ica and Pisco- (and perhaps Nasca) were ruled by the powerful 
 chief Chincha. 
 
 The different valleys being so divided from one another in 
 political ways, it is not surprising that we find considerable local 
 differences in art-types as well. Yet we have no grounds for 
 assuming that the coast peoples were not rather closely related on 
 ethnic lines, which explains, no doubt, certain widespread resem- 
 blances between the arts of the various regions. 13 
 
 The architecture of this period, perhaps because of the influ- 
 ence of Tiahuanaco II, was very elaborate. Adobe continued to 
 be the chief material, but it was used in more complex ways. 
 Palaces, workshops, reservoirs, aqueducts and many other elab- 
 orate works were constructed. The custom of using stucco reliefs 
 on walls became fairly common. 14 
 
 11 See Markham, 1912, p. 222; Joyce, 1912, pp. 50 ff. ; Beuchat, 1912, 
 pp. 584 ff. ; Means, 1917; Garcia Rosell, 1903. 
 
 12 See Cieza, 1864, pp. 233 ff., 1883, pp. 185-193; Garcilarsso, II, pp. 181- 
 201; Cobo, 1892, IV, pp. 47-54; Markham, 1912, pp. 200-239; Joyce, 1912, 
 pp. 95 ff. ; Garcia Rosell, 1903; Arriaga, 1621. 
 
 13 Hrdlicka. 1914, pp. 41 ff., and pp. 52 ff. 
 
 "Middendorf, 1894-95, II, p. 375 ; Squier, 1877, pp. 136 ff., 150 ff. ; 
 Joyce, 1912, pp. 150 ff. 
 
33° Philip Ainsworth Means, 
 
 The artifacts of the period under consideration are chiefly in 
 the form of pottery, albeit textiles are also present to a consid- 
 erable extent. In general, designs on Nasca textiles may be said 
 to take the form of rather simple, but by no means crude, geo- 
 metric patterns, perhaps with a slight and conventionalized zoo- 
 morphic element, such as those in Uhle, 1913b, Figures 3, 7, and 
 9. On both pottery and textiles of this region and period the 
 colors were much less numerous and splendid than they were in 
 either the Proto-Nasca period or the Tiahuanaco II period. If, 
 then, Nasca art can be said to preserve an echo of the color tra- 
 ditions of its predecessors, and also of their geometric tendencies, 
 (for some of its chief motifs are derived directly from some of 
 their minor ornamental details), so, in no less degree, can the 
 black modelled ware of the Chimu period be said to preserve the 
 realistic tendencies, as well as some of the decorative motifs, of 
 Proto-Chimu art. 
 
 6. THE COLLA-CHULPA CULTURE. 
 
 The name chosen to distinguish this period is made up of the 
 two names applied by various writers to the people who lived 
 in it. 15 
 
 As the general culture-level was so low, it is but natural that 
 the pottery of this period should be poor. The best collection of 
 it is that made by Bandelier which is now to be seen in the Ameri- 
 
 "Joyce, 1912, p. 75, Markham, 1912, p. 186, Beuchat, 1912, p. 576, and 
 others use the term Collas. Bandelier, 1910, pp. 184 ff., calls them Chullpa. 
 (The double 11 is without justification.) The term Aymard, often applied 
 to these people by writers, and even by such first-rank authorities as 
 Bandelier (1910, pp. 63 ct passim), Hrdlicka (1911, p. 1) and others, is 
 entirely misleading. The people who lived in the Titicaca basin between 
 the time of Tiahuanaco II and the Inca conquests were the Collas. It 
 was they who produced the culture here to be discussed and who built 
 the chulpas or burial-towers. The name Aymara was first given to these 
 people by the Jesuits of Juliaca some time before 1590, and it was estab- 
 lished in usage by Bertonio (1603) and Torres Rubio (1616). All this 
 has been emphasized by Markham (1912, p. 192) and Joyce (1912, p. 75) 
 but it cannot be dwelt upon too often. The mistake of the Jesuits is 
 accounted for by the fact that the Aymaras, whose original home was 
 between Cuzco and the continental divide, were conquered by the Inca 
 Pachacutec and were moved, by him, to Lake Titicaca as mitiniaes. 
 (Sarmiento, 1907, p. 108; Garcilasso, II, p. 50.) 
 
A Survey of Ancient Peruvian Art. 331 
 
 can Museum of Natural History, New York City. It has not 
 seemed to the writer to be worth while to include pictures of this 
 type, so a description will be given in order that some idea of the 
 type may be formed. 
 
 In the Bandelier collection are a number of jars from Sillus- 
 tani, a place that was probably the site of important activities 
 during the Colla-Chulpa period. 16 The vessels are made in two 
 styles. One is a small type of vessel of white clay, rather coarse 
 and undecorated; the other type is made up of red ware, also 
 coarse, with designs in black upon it. Other specimens, doubt- 
 less from this period, are a class of rather coarse and clumsy 
 bowls with design suggestive of the "epigonal" of the coast. 
 (See Bandelier, 1910, Plate XXI.) Coarse bottles of dark red 
 clay, sometimes decorated with black lines, and gray bottles with 
 incised rectilinear spirals seem to exhaust the artistic repertory 
 of the Colla-Chulpa potter. In bronze work, however, the Colla- 
 Chulpa folk were much more advanced, as is evidenced by the 
 archaeology of the region where the chulpas abound. 17 
 
 It would be a mistake to close our study of this intermediate 
 period without a brief study of the unusual architectural form 
 that peculiarly marks it. The chulpas are strictly speaking stone 
 towers, either circular or rectangular in plan. They vary greatly 
 in size and neither their use nor their distribution is yet definitely 
 settled. Even with our present limited information, however, it 
 is possible to distinguish several types of chulpa. Sir Clements 
 Markham long ago suggested that the cruder types might have 
 been adopted later by the Incas who evolved from them the less 
 crude types. 18 Without formally accepting this theory, we will 
 discuss each of the types in the order of their apparent antiquity, 
 bearing in mind the possibility that appearances may be deceptive. 
 The most primitive form of chulpa, then, is that which is found 
 at Ouellenata and Ullulloma. 19 The former of these places is 
 close to the north-western end of Lake Titicaca ; the latter is about 
 fifty miles north-west of there. Primitive chulpas also occur at 
 
 16 Bandelier, 1910, pp. 184 ff. ; Bandelier, 1905; Squier, 1877, pp. 376-384; 
 Markham, 1912, pp. 186 ff. 
 "Beuchat, 1912, pp. 580 ff. ; Nordenskjold, 1906, 1906b. 
 
 18 Markham, 1871, p. 308. 
 
 19 Squier, 1877, pp. 386 ff. 
 
 Trans. Conn. Acad., Vol. XXI 23 1917 
 
33 2 Philip Ainsworth Means, 
 
 Sillustani, on the west of the northern end of Lake Titicaca, at 
 Kalaki on the eastern shore, and at Coni and Curahuara far to 
 the south-east of the Lake. 20 It will be seen that this type of 
 chulpa wa^ built over a wide area. Speaking in general terms, 
 it is a round stone tower which is smaller at the bottom than at 
 the top. The stones are uncut, and had some binding material. 
 In some cases the roof is flat; in others it is a truncated cone. 
 Stone was the sole material. The edifices of this type belong to 
 the fourth period of Posnansky's culture-sequence. He calls it 
 the "epoch of edifices of adobe and pirca (uncut stone)." 21 
 This reckoning would place the style just prior to Inca times. 
 The second type was, in outline, the same as the first. It tended, 
 however, to be larger, and the stone was carefully cut so as to 
 make a beautifully built wall. Sillustani, Coni and Kalaki are 
 the chief sites for this type. The third and final type was some- 
 what the same as type two in regard to the material, but it dif- 
 fered from the other two in being rectangular in plan and very 
 large, sometimes as much as thirty feet in height. 22 Unlike the 
 other two types, which had but one interior chamber, this third 
 type sometimes had two chambers, one above the other. It is to 
 be noted that this type is the only one which occurs outside the 
 Titicaca drainage. There is an interesting example of it at 
 Palca, not far from Tacna in northern Chile. 23 
 
 The question of who the Collas really were is a complex and 
 
 20 Bandelier, 1910, pp. 243 ff., 1910, p. 186; von Tschudi, 1868, V, 
 pp. 202 ff. 
 
 21 Posnansky, 1911b, p. 17. 
 
 " Squier, 1877, pp. 352 ff., 372 ff. 
 
 23 Squier, 1877, pp. 242 ff. 
 
 The whole question of the distribution of the chulpa-type of building 
 is a highly important one, in all probability. The type has prototypes 
 over a very large area. The writer has found it in the region of Ollan- 
 taytambo. It exists in the neighborhood of Oroya (see Dr. William C. 
 Farrabee's photographs in the Peabody Museum) and something strik- 
 ingly like it is found at Cuelap and other sites in the region of Chacha- 
 poyas. (Bandelier, 1907.) Again, in the district of Huarochiri, buildings 
 of the chulpa type are found in the middle portion of Peru and fairly 
 near the coast. (Hrdlicka, 1914, Plates 3 and 4.) At present the evidence 
 is rather tantalizing than illuminating. One can only say that over this 
 wide area there seems to have been a material culture of the same 
 general level as.that of the Colla in immediately pre-Inca times. 
 
A Survey of Ancient Peruvian Art. 333 
 
 important one. In considering it one must not forget the presence 
 in the Titicaca basin of another and much lower-cultured stock 
 called Urus or Uros. The general trend of the evidence at hand 
 regarding the Urus shows them to be very low-cultured and 
 quite widely distributed. In fact, their area at the time we are 
 considering extended from Titicaca down to Lake Poopo or 
 Aullagas. It may have extended westward to the Chilean coast. 
 The stock was probably an old one. Boman (1908, I, p. 72) 
 suggests that the Urus were vestiges of the earliest pre-Yunca 
 (i. e. pre-Proto-Chimu) population, and that they were driven 
 south and east by the earliest high-cultured invaders. At the 
 same time, we must remember that, in the same general area, 
 the higher-cultured Collas had a culture which was similar to that 
 found in the north-western parts of Argentina. It might be 
 suggested that one of these racial elements represents the inhabi- 
 tants of the Tiahuanaco II "empire" and that the other repre- 
 sents the invading race which may have helped to bring it to a 
 close. But which is which, and if this is the truth, we cannot 
 surely tell. 2,1 To some it may seem more satisfactory to assume 
 that there were two strata of population — Collas and Uros — who 
 were mutually aloof. Such a state of affairs has been known to 
 exist in Asia, Oceania and elsewhere. Certainly the Titicaca 
 basin is spacious enough to permit isolated groups of Uros to 
 dwell wholly apart from the surrounding Colla communities. 
 
 7. EARLY INCA CULTURE. 
 
 As has been said before, the culture of the mountain regions 
 away from the sea suffered a general and marked subsidence 
 after the Tiahuanaco II period, a subsidence which we have 
 studied under the name of Colla-Chulpa culture. Therefore, 
 when that gens of the valley of Cuzco which later became the 
 Inca dynasty began to raise its own culture-level and that of 
 the surrounding tribes it had not much artistic tradition on which 
 to establish its own art. 25 
 
 24 Cf. Chamberlain, 1910, 1910b, 1911, 1913; Boman, 1908; Garcilasso, 
 II, pp. 223-227. 
 
 25 It seems to the writer that the character of the Inca gens has never 
 been properly appreciated, -save, in a measure, by the late Sir Clements 
 Markham. According to Sarmiento (1907, pp. 37 ff.), the people in imme- 
 
334 Philip Ainsworth Means, 
 
 Difficult though it sometimes is to distinguish between early 
 and late Inca pottery forms, it is, in the writer's opinion, possible 
 to establish a series of vessels from Machu Picchu 26 that will 
 serve to throw some light upon the development of the most 
 typical form of Inca (or Cuzco) pottery — the aryballus. But 
 the reader should take care to bear in mind that the simpler and 
 cruder forms, forms probably longer in use than the more 
 advanced types, undoubtedly continued to be employed by the 
 very late generations of the Incas' subjects as cooking utensils, 
 etc., while the finer products of the potter were reserved for less 
 heavy work. In spite of this, however, the fact remains that the 
 cruder types, being very like the Colla-Chulpa pottery both in 
 form and in material, were probably older types of vessels than 
 the decorated and graceful forms. The reader is urged, then, 
 
 diately pre-Inca times lived without governmental organization of any 
 sort except that in times of danger a military officer with the title of 
 Sinchi was chosen. Besides this, in the opinion of Sir Clements Markham 
 (1912, pp. 159 ff.), there was a social organization based upon the family 
 at the head of which stood the puric. Several purics combined together 
 into an ayllu or lineage. This system was carefully studied by Sir 
 Clements Markham, and we have to thank him for showing us what the 
 social conditions in the highlands before the rise of the Inca ayllu were. 
 He did not, however, lay stress upon the historical significance of all this. 
 Sarmiento (1907, pp. 40 ff.) tells us that just before the rise of the 
 Incas, there were, in the valley of Cuzco, six ayllus in the possession of 
 the region. Three of these, whose names he did not know, were native ; 
 three others, the Alcabisa, the Copalimayta, and the Culunchima, came 
 and settled amicably among them. Later on, the Inca also came from 
 not far off and settled at Cuzco. Strife arose between them and the other 
 families which was not finally subdued for some time. Like the heads 
 of all the other ayllus, the chief of the Incas bore the title sinchi. Hence 
 we get the name Sinchi Rocca, borne by the first historic Inca. 
 
 26 The name "Machu Picchu" is the one given to this site by Dr. Hiram 
 Bingham, who visited it for the first time in 191 1. Although the name is 
 not a wholly satisfactory one, it has been thought best to continue its use 
 here because the site has already become well known under it, and because 
 the name Vilcabamba-the-Old (or Vilcabamba viejo) is rather clumsy, a 
 fault which outweighs its greater historical accuracy. In any case, "Machu 
 Picchu" is preferable to the "Matcho Picho," "Macho Piccho" and so on 
 of such writers as Sartiges, 1851, and Wiener, 1880. The phrase machu 
 pichu means "old ridge." The late Sir Clements Markham was of the 
 opinion that the cc in the name "Machu Picchu" was a mistake. The 
 name is pronounced Pi-chu, not Pic-chu. 
 
A Survey of Ancient Peruvian Art. 
 
 335 
 
 to turn his attention to Dr. George F. Eaton's work on the osteo- 
 logical material from Machu Picchu, and to Dr. Hiram Bing- 
 ham's "Types of Machu Picchu Pottery." 27 In the first place, 
 we are justified in assuming that the delicately formed, well 
 decorated aryballus of the type shown in our Plate XIII was one 
 of the ultimate forms of Cuzco pottery by the fact that it is 
 this type of vessel that is found most widespread, even in regions 
 like Ecuador, Chile and Argentina where Inca influence did not 
 arrive until very late. It will be our task therefore to show 
 in a series the forms that led up to the final aryballus type. This 
 we will now do. The series proposed by the writer is made up 
 as follows : 
 
 First Step. 
 
 Rough, undecorated ware. Ea- 
 ton, Plate XIV, Figure 4. Bing- 
 ham, Fig. 48, No. 7a. Also see 
 Eaton, Plate IX, Figs. 3 and 4 for 
 a variation of the First Step. 
 
 Third Step. 
 
 Still coarse ware with more pro- 
 nounced neck. Sometimes deco- 
 rated(?). Handles small and 
 moved down from the lip. Eaton, 
 Plate XIV, Fig. 5. Bingham, Fig. 
 48. No. 8a. 
 
 Second Step. 
 
 Slightly finer ware, sometimes 
 decorated in colors, with enlarged 
 handles and more pronounced neck. 
 Eaton, Plate XIII, Figs. 1 and 2. 
 Bingham, Fig. 47, No. 6a. 
 
 Fourth Step. 
 
 The well-known tvpe of Cuzco 
 Aryballus. Our Plate XIII. Ea- 
 ton, Plates V, VII, & X. Bamps, 
 1879. Oyarzun, 1910. Joyce, 1912, 
 Plate XXII, and p. 229. 
 
 It will, perhaps, be well to say again that the particular crude 
 specimens above referred to are not, in all likelihood, themselves 
 older than the more refined specimens. For example, in the 
 same grave with the specimen representative of the first step, 
 Dr. Eaton found skeletal remains of the coast type, which implies 
 that the pot belonged to people who had come up from the coast 
 at some time subsequent to the Inca conquest of the littoral. 28 
 The point of the series presented, however, does not dwell in 
 the antiquity of the specimens, but rather in the relative antiquity 
 of the types of form. 
 
 To sum up, then, our impressions of the early Inca culture 
 we will say that the time in which the Inca ayllu was extending 
 its ascendancy over the other Quichua tribes in the neighborhood 
 of Cuzco, the people of the Cuzco region were gradually evolv- 
 
 " 7 Eaton, 1916, Plates V-XIV; Bingham, 1915b, entire. 
 " s Eaton, 1916, p. 45 ff . 
 
336 Philip Ainsworth Means, 
 
 ing from the simple pottery-types of their ancestors a new kind 
 of pottery which was to find its fullest florescence under the 
 last five Inca rulers. Because of the lack of detailed knowledge 
 of the early Inca period, we shall not touch upon that culture 
 again in this paper. 
 
 8. THE INCA CULTURE AT ITS HEIGHT. 
 
 As the Inca culture is the nearest to us historically it is but 
 natural that we should know more about it than we do of the 
 rest. It is even possible to draw up a fairly complete and reliable 
 history of the Inca dynasty, especially of the last six rulers. 
 For a long time it was customary to assign all evidences of pre- 
 Columbian culture in Peru to the Incas ; indeed, that is still done, 
 unfortunately, by some writers. They disregard the growing 
 evidence which points more and more clearly to the inferiority 
 in many respects of the Incas to their various predecessors. 
 
 The Incas were, nevertheless, wonderful people. They had a 
 real genius for government and their state was the only truly 
 socialistic monarchy that has ever existed. The individual was 
 nothing; the state, that is the Inca himself and his blood-rela- 
 tives, was supreme in all things. It is not surprising that, in 
 a state like this, strongly centralized, autocratic, theocratic and 
 all-controlling, the art of outlying regions should all tend to 
 approximate that of the capital of the dominion ruled by the 
 Inca from Cuzco. This is, in the writer's opinion, the psycho- 
 logical explanation of the fact that from Quito to Chile and 
 from the Pacific to the Brazilian wilderness, vessels, architecture, 
 weapons, textiles and language all conform, with varying degrees 
 of closeness, to the fashion or example set by the people of Cuzco. 
 Typical Cuzco pottery is found wherever the Inca conquerors 
 penetrated ; Quichua dialects prevail to-day over the same areas. 
 
 As far as shape is concerned, the vessels made by the sub- 
 jects of the Incas of the later generations are the most graceful 
 in Peru. The aryballus, the beaker, the bowl and many other 
 forms, all very attractive, are found. Dr. Hiram Bingham, whose 
 trips to Peru have resulted in the publication of many valuable 
 pictures of Inca sites and products, has given a resume of the 
 commoner Inca forms. 29 Machu Picchu, the site from which 
 
 Bingham, 1915b. 
 
A Survey of Ancient Peruvian Art. 337 
 
 most of these objects come, was thought by Dr. Bingham to be 
 Tampu Tocco, the "cradle" of 'the Incas. He also presented 
 convincing evidence pointing to the fact that the city was Vilca- 
 bamba-the-Old, a celebrated sacerdotal establishment of the Incas 
 in post-conquest times. 30 It should be noted that the work of 
 Dr. Eaton has left very little doubt as to the modernity of this 
 site as compared with that of Tiahuanaco or Chimu. Every 
 class of object found there, every bit of osteological evidence, 
 points to the fact that Machu Picchu was built at some time 
 after the Incas had conquered the coast of Peru and had had 
 time enough to be affected by the influence of coast art. 31 Wc 
 are indebted, therefore, to the Yale Expedition for the unveiling 
 of a city which, though known to travelers for many years, has 
 never, until recently, been photographed and adequately described. 
 Machu Picchu is undoubtedly the most valuable site in the Cuzco 
 region, for it presents an epitome of all that the Incas knew of 
 art, architecture and engineering at a time when they were at 
 the zenith of their power. We shall, therefore, consider Machu 
 Picchu pottery to be representative of all that the pottery of 
 late Inca Peru was, and we shall study it accordingly, assuming 
 Machu Picchu pottery to be all that Cuzco pottery was in the 
 last part of the Inca period. 
 
 The characteristics, then, of Inca pottery as shown by the 
 Machu Picchu collections, are : ( 1 ) The predominance of almost 
 classically graceful shapes such as aryballi, pelikai, dishes, bowls 
 and so on. (2) The widespread and often-repeated use of cer- 
 tain fixed and definite geometric decorations. (3) The general 
 scarcity of anthropomorphic decoration. (4) The occasional 
 association of perfectly recognizable Cuzco shapes and decora- 
 tions with some element introduced from the coast, such as 
 modelled anthropomorphic handles on dishes or life-like butter- 
 flies painted on the bottoms of shallow bowls. (5) Cuzco pottery 
 is, in general, lighter in tone than either Tiahuanaco II or Proto- 
 Nasca. As has been said, pottery of a pure Cuzco type is found 
 from Ecuador to Chile. 32 In all this huge area a surprising 
 
 30 Bingham, 1915, pp. 180 ff. 
 
 31 Cf. Bingham, 1913, 1915, 1915b, 1916; Eaton, 1916; Dorsey, 1901, PI. 
 XLII ; Joyce, 1912, p. 198; Uhle, 1903; Hrdlicka, 1916c. 
 
 32 Cf. Bamps, 1879, Atlas; Saville, 1907-1910; Rivet, 1912; Oyarzun, 
 1910. 
 
33& Philip Ainsivorth Means, 
 
 steadfastness to the original type is to be observed ; but, never- 
 theless, local sub-types do develop in several cases. Such a one 
 is the Inca style on the islands of Titicaca and Coati in Lake 
 Titicaca. There, though still perfectly definitely related to the 
 usual Cuzco types, the pottery is marked by a tendency to break 
 the decorated surface up into very small geometric areas which 
 are made prominent by the contrasting of dark brown with cream 
 color. This is noticeable in the collections from Titicaca now in 
 the American Museum of Natural History, New York. 
 
 As is usually the case, the textiles, though showing affinities 
 of design with the pottery, are richer in the variety of their 
 colors. The Incas' subjects were as good weavers as any in 
 aboriginal Peru. That their art was strong and flourishing at 
 the time of the Spanish conquest is proved by the fact that shortly 
 after the conquest there were produced some of the finest speci- 
 mens of Inca tapestry that we have. 
 
 Unlike their predecessors, the people of Tiahuanaco II, the 
 Incas' subjects, though admirable architects, did not decorate 
 their buildings with any great amount of carving. To replace 
 the decoration applied to the walls of huge stones by the carv- 
 ings of Tiahuanaco II type, the subjects of the Incas evolved a 
 new type of architecture. It takes the form of exquisite walls 
 made of reasonably large stones laid in courses of quite aston- 
 ishing accuracy. Often the lowest course would be made of 
 stones of say a foot high ; the next course would be slightly 
 lower, and so on to the top. The effect of this technique was a 
 wall of wonderful symmetry and beauty. Such a wall needed no 
 carving to make it sightly. Dr. Bingham gives an excellent pic- 
 ture of this late Inca type of wall. 33 
 
 This Inca culture, then, was the last of the long series of pre- 
 Columbian Peruvian cultures. With our brief review of the 
 chief features of those cultures thus brought to a conclusion, we 
 will now turn to a detailed analysis of the Plates which accom- 
 pany this paper and which have been chosen with a view to 
 setting forth the more prominent characteristics of the principal 
 culture-types. 
 
 33 Bingham, 1913, p. 488. 
 
III. ANALYSES OF THE ARTS OR CULTURES. 
 
 i. A CRITICAL ANALYSIS OF PROTO-CHIMU AND 
 PROTO-NASCA ART. 
 
 Aside from examining the Plates that accompany this paper, 
 the reader is urged to examine those that are to be found in the 
 works referred to in the footnote. 1 It is hoped, however, that 
 the examples of the two very early types of art herewith pre- 
 sented will prove sufficient material for those who cannot seek 
 further for it. 
 
 Plate I shows five specimens of Proto-Chimu art, all to be 
 found in the Peabody Museum at Cambridge, Mass. -It will 
 readily be observed that two characteristics hold true for all the 
 specimens given ; these are : light coloration, and grace of line. 
 In Figures 1-4, the chief source of admiration on the student's 
 part is the wonderfully life-like modelling. Figures 1 and 2 are 
 especially remarkable in this respect. In Figure 1 we see a man 
 attacking a deer with a massive club. His small dog looks on. 
 With the exception of the deer's body and the man's feet the 
 modelling is far better than that in some of the early Egyptian 
 and Cretan figures. The man's clothes seem to consist of a 
 loose-fitting shirt with sleeves and of a hat or helmet adorned 
 with two rosette-like protuberances and a sort of frontal ridge. 
 From the helmet proper a strip of cloth runs down to and under 
 the man's chin. The nose of the man is large and somewhat of 
 the Semitic type. The chin is somewhat receding. The dog 
 on this vase is probably one of those which the early people kept 
 for use in the chase. 2 On the body of the vase is to be seen 
 a composition that is very typical of Proto-Chimu art. It is 
 painted in dark brown on the white slip of the vase and, like the 
 modelled group above, represents a hunting scene. It should 
 be noted that the costumes of the figures in the painted part of 
 the decoration differ considerably from that of the modelled man. 
 
 1 Cf. Uhle, 1908. 1910, 1912, 1913, 1913b, 1914 ; cf. Reiss und Stubel, 1880-87 ; 
 Baessler, 1902-03 ; Putnam, 1914 ; Theresa von Bayern, 1907 ; Joyce, 1912, 
 1913b; Beuchat, 1912 ; Mead, 1915; Squier, 1877; Berthon, 1911 ; Rivero 
 and von Tschudi, 185 1 ; and many other works. 
 
 2 Cf. Joyce, 1912, p. 125. 
 
34° Philip Ainsworth Means, 
 
 The former, for example, have the black "stockings" that are 
 so frequently seen in Proto-Chimu vase-paintings; also, the 
 painted men have a very different headdress from the modelled 
 man. But most important of all is the fact that the painted men 
 appear to be either wearing masks or else to be adorned with 
 face-paint. Indeed, if the latter is the case, the "stockings," 
 "knee-caps" and "sleeves" must be assumed to be nothing less 
 than body-painting. From all of these elements of decoration 
 the modelled man is entirely free. Figure 2 is another type of 
 modelled vase from the Proto-Chimu period. It shows a per- 
 sonage, apparently masked to represent a fox or some such ani- 
 mal, sitting facing a semicircle of five foxes. The personage's 
 headdress, though different from that in Figure 1, is, neverthe- 
 less of the same general type. The striking features about this 
 figure are the headdress and the fangs, to both of which we shall 
 refer later. Again, the back of the middle fox is adorned with 
 a design which Posnansky calls signo escalonado — stairsign. 3 To 
 this also, we shall refer, in another connection. Around the base 
 of the vessel, in the region analogous to that occupied by the 
 painted hunting-scene in Figure 1, we see a landscape. The trees 
 and plants are shown by means of shallow lines engraved, appar- 
 ently with a blunt stick while the clay was still moist, in the 
 reddish slip of the vessel. This landscape is full of charm 
 because of its quaint realism. It is even possible to attempt to 
 identify the tree as an algaroba and the smaller plants as cactus. 
 This sort of vessel sometimes leads students into attempting an 
 "interpretation" of the scene. While the modelled portion of 
 the vessel undoubtedly represents some sort of ceremony 
 employed by the people of that period, it is, nevertheless, danger- 
 ous to reconstruct, let us say, a totemic clan organization, from 
 such evidence as this. 
 
 As the vases shown in Figures 1 and 2 represent a very large 
 and important sub-type of the Proto-Chimu pottery, it will be 
 well to summarize briefly our impressions of them before going 
 on to an examination of the other sub-types. 
 
 We see that the vessels of this sub-type comprise two separate 
 areas of decoration, each marked by a distinct technique. In 
 both the painted (or engraved) area and the modelled area of 
 
 3 Posnansky, 1913. 
 
A Survey of Ancient Peruvian Art. 341 
 
 the two vessels we observe the following features : ( 1 ) A marked 
 tendency toward realism of representation; (2) A decided lack 
 of rich and varied coloration (dark brown, red and cream color 
 being the tints found) ; (3) A gracefulness of line which is 
 not constricted by any sort of conventionalization; (4) The use, 
 by several of the human figures, of masks or face-painting, of 
 fangs and of an easily recognizable type of headdress. If the 
 reader will examine the Plates in some of the works already 
 referred to he will see further examples of these characteristics, 
 as well as some others that occur in Proto-Chimu pottery of this 
 sub-type. For example, look at the scenes from vessels shown 
 by Mr. Joyce. 4 These show new forms of the headdress, the use 
 of face-painting and of masks, the presence of fangs, and also 
 a curious use of girdles ending in aerpent heads. The black 
 "stockings" also are found in these figures, as well as in Figure 
 1 of our Plate I. Also, the use of peculiar fluted wings is rather 
 often met with. 5 
 
 So much, then, for the modelled and painted sub-type of Proto- 
 Chimu vases. There are still several other types for us to con- 
 sider. Look, for example, at Plate I, Figure 3. This specimen, 
 the original of which is in the Peabody Museum, Cambridge, 
 comes from Chimu (Trujillo) and shows, in addition to the 
 typically Proto-Chimu fangs, a further development of the head- 
 dress. The latter seems to be composed of the stiff ridge or 
 core found in the headdresses of Figures 1 and 2 with the addi- 
 tion of ornaments that may be intended for feathers. These 
 feathers are important, and they will be referred to in connection 
 with our analysis of Proto-Nasca art and of the Chavin Stone. 
 
 Plate I, Figure 4, shows an example of a type of vessel that 
 has always excited admiration in students of ceramics. 6 It is 
 called the "Portrait type." Possibly this particular example is 
 not really Proto-Chimu, but similar "portraits" have been found 
 that have painted on them unmistakable Proto-Chimu decorations. 
 
 4 Joyce, 1912, pp. 126, 127. 
 
 5 See Joyce, 1912, p. 155, for an admirable specimen of Proto-Chimu 
 vase-painting with fluted wings, serpent-tail and fanged masks. Also 
 examine plates in Reiss and Stubel, 1880-87, and in Baessler, 1902-03. 
 
 8 See, for example, Jacquemart, 1873, pp. 190 ff . ; and Young, 1879, pp. 
 404 ff. ; Squier, 1877, pp. 180 ff. 
 
34 2 Philip Ainsworth Means, 
 
 As the specimen here shown has a headdress that has several 
 points in common with that of the chief figure in Figure i, it is 
 assumed for the nonce that this "portrait" is Proto-Chimu. No 
 one who has studied a series of these human-faced vessels and 
 has noticed the wide differences and unfailing individualization 
 that characterize each one of them can fail to lean toward the 
 belief that this type of vase is indeed a "portrait type." There 
 is absolutely nothing of inherent impossibility about the idea that 
 a people so highly gifted with plastic skill as the Proto-Chimu 
 people may have developed the habit of employing their vessels 
 as a medium whereby to perpetuate the likenesses of their great 
 men. In any case, empirical evidence leads us to believe that 
 some such habit did prevail, for every good specimen of the 
 "portrait type" portrays an individual, not a type. And it should 
 be noted that realistic portrait-making is in entire accord with the 
 marked realistic tendencies of the Proto-Chimu culture pottery. 
 Nor are "portrait types" lacking in other parts of America. 7 
 
 Last of all, in the matter of Proto-Chimu sub-types, comes 
 that variety which is represented by Plate I, Figure 5. In this 
 division come almost innumerable stylistic decorations which, 
 though they may show slight conventionalization, never show 
 geometrical tendencies to the exclusion of all curves. The pres- 
 ent specimen, in the writer's opinion, is intended to represent a 
 starfish. s In this type also occur many variations of the "stair- 
 sign" (signo escalonado) often in conjunction with the starfish 
 (or octopus) motif. 
 
 Passing over for the present the numerous forms of pottery 
 which may some day be definitely assigned to this period (a 
 passing-over process which will have to be repeated many times 
 in the present state of our information), we will endeavor to 
 draw up a tentative classification of the Proto-Chimu sub-types. 
 
 Sub-type I Landscapes. Vessels having modelled scenes as 
 
 well as painted or engraved ones. Usual 
 colors : white or cream slip, dark brown and 
 red. 
 
 7 Spinden (1916b) claims them for Central America, and Holmes 
 (1916b) shows an excellent example of aboriginal portraiture from 
 Quirigua. 
 
 s It is the opinion of Prof. MacCurdy that the design here mentioned is 
 derived from the octopus, not from the starfish. This, of course, may 
 well be the case. 
 
A Survey of Ancient Peruvian Art. 343 
 
 Sub-type II Portraits. The faces of the portraits often have 
 
 features in common with Sub-type I and Sub- 
 type III, (such as headdress, formal inci- 
 dental decoration motifs, etc.). 
 
 Sub-type III Partly conventionalized decorations. Even these, 
 
 however, are seldom rectilinear entirely. 
 Cream and red are the more usual colors. 
 
 Sub-type IV Numerous miscellaneous types not yet decided 
 
 upon. 
 
 Having completed our study of the distinguishing elements of 
 Proto-Chimu art, we will now examine into the traits of Proto- 
 Nasca art. 
 
 We have seen that realism, grace of line and light coloring 
 were three of the chief characteristics of Proto-Chimu art. We 
 find in Proto-Nasca art an almost complete reversal of these fea- 
 tures. There is, to be sure, an apparent attempt at realism in 
 some of the Proto-Nasca sub-types, but it is an unsuccessful 
 one in most cases. Look, for instance, at Plate II, Figure 2, and 
 at Plate II, Figure 1. In both of these we have a survival of 
 the wonderful modelling that marks out Proto-Chimu art from 
 all the rest. Both of these specimens preserve a certain degree 
 of realism. The former, to note the most prominent feature in 
 each case, holds in his left hand a spear-thrower almost identical 
 with those found in Peru by Dr. Uhle. 9 It would be hard to 
 find a better representation of an object than this one. Then, 
 too, the hands on the other specimen mentioned are absolutely 
 realistic. Their realism consists above all in this : That they 
 are shown in the natural closed position and the finger-nails of 
 the fingers are not shown. These two specimens, therefore, both 
 with five-digit hands and fairly well-modelled heads, may be said 
 to represent a survival of the Proto-Chimu art-tradition in the 
 Proto-Nasca type and, at the same time, to constitute the nearest 
 approach of Proto-Nasca art to realism. 
 
 Wares of this type were not, however, the most character- 
 istic expression of Proto-Nasca art. Far more common and far 
 more typical were such productions as those that appear on Plate 
 II, Figures 3-6, and Plates III and IV. Excellent examples of 
 Proto-Nasca plastic art are given by numerous writers, to whose 
 
 9 Uhle, 1909. 
 
344 Philip Ainsworth Means, 
 
 work the reader is urged to refer. 10 By study of the Plates that 
 accompany this article and those that go with the works here 
 referred to, it will be seen that in the matter of form the Proto- 
 Nasca pottery was not so diverse as the Proto-Chimu. A tenta- 
 tive division into sub-types will, as in the case of Proto-Chimu, 
 be offered for the Proto-Nasca art. At present we will limit 
 ourselves to a consideration of the decoration. 
 
 Color is indubitably the "strong point" of Proto-Nasca art. 
 For example, Plate II, Figures 3, 4 and 6 are all of remarkably 
 rich tonality. Red, brown, gray, yellow and black, as well as 
 cream-color, are the tints most frequently met with. The finish 
 of some of the Proto-Nasca pots is so lustrous as almost to 
 suggest a glaze. As for the subject-matter of Proto-Nasca art, 
 it cannot be so easily described as that of Proto-Chimu, although 
 the two have much in common in that respect. Proto-Nasca vase- 
 paintings mostly concern themselves with the portrayal of a few 
 personages who, being few in number, occur again and again 
 in the vase paintings. These paintings were no doubt supposed 
 to represent deities or mythical persons ; at all events, there is 
 absolutely nothing realistic about them ; they are merely elaborate 
 and formal portrayals of putative objects of veneration. The 
 chief personages of Proto-Nasca art seem to be two in number. 
 Each occurs in several variations. We will describe them in turn, 
 applying arbitrary names for the sake of ease of identification. 
 
 The "Centipede God." See Plate II, Figures 3 and 4, Plate 
 III, Figures 1 and 2, and Plate IV, Figure 2. The name chosen 
 is suggested by the fact that this 'god' is usually shown as 
 having a long body at right angles to his face and fringed with 
 spike-like objects that are evidently conventionalized legs. Some- 
 times he has a series of subsidiary human faces where the legs 
 ought to be; sometimes both legs and faces occur (as in Plate 
 III, Figure 2). Again, the "Centipede God" is shown as a man, 
 strongly conventionalized to be sure, who has centipede attributes 
 such as the girdle shown in Plate II, Figure 4. It is very inter- 
 esting to note certain well nigh invariable features that mark 
 the portrayal of the "Centipede God," whether that 'god' is 
 
 10 Joyce, 1912, Plate I, Joyce, 1913b; Therese von Bayern, 1907; Reiss 
 u. Stiibel; Baessler,, 1902-03; Berthon, 1911, Plates I-VI ; Uhle, 1913b, 
 p. 358 ff. 
 
A Survey of Ancient Peruvian Art. 345 
 
 the chief portion of the design or merely a comparatively insig- 
 nificant adjunct to the design. These features are: (1) The use 
 of a very distinctive mouth-mask; (2) The predominance of 
 hands with less than the true number of digits, usually with four 
 digits; (3) The frequency with which the tongue is shown 
 sticking out of the mouth ; (4) The almost invariable presence of 
 a broad flat headdress in the form of a rather highly conven- 
 tionalized human face; (5) The frequent appearance of cere- 
 monial staffs held in the hands. We will say a few words about 
 each of these features in turn. 
 
 (1) The Mouth Mask. Plate II, Figure 3, and Plate III, Fig- 
 ures 1 and 3, show very typical forms of this element. In Plate 
 III, Figure 1, it is seen to consist of a central portion with mouth- 
 and nostril-holes and of two wing-like portions, one on each 
 side of the mouth. These wings are marked by lines of a con- 
 ventional nature that may be a survival of the curling-feather- 
 like rays that distinguish the mouth-mask of Figure 3. These 
 rays are perhaps related to certain elements of decoration that 
 occur in later arts, as well as in other sub-types of Proto-Nasca 
 art. 
 
 (2) Four-digit Hands. Plate II, Figure 3 ; Plates III, Fig- 
 ures 2 and 3, and IV, Figure 2, all show typical examples of the 
 four-digit hands that so often accompany, as in all these 
 instances, one or more of the several criteria that mark this 
 "Centipede God" motif. The development from natural five- 
 digit hands to these very artificial conventionalized four-digit 
 hands is a matter of great importance, as will be shown in con- 
 nection with Tiahuanaco II art. 
 
 (3) The Protruding Tongue. The Plates already mentioned 
 show this feature. In the pottery with the "Centipede God" 
 motif the protruding tongue is not nearly so widely developed 
 as it is in some other cases, especially in that of the textiles. 
 But even in the "Centipede God" figure on Plate III, Figure 2, 
 the tongue shows the beginnings of decoration on its upper sur- 
 face. The element of tongue-decoration becomes very prominent 
 in other types of Proto-Nasca pottery. 
 
 (4) The Broad Flat Headdress. Plate III, Figure 1, shows 
 a standard form of the "Centipede God's" headdress. The brim 
 almost always consists of at least two layers separated by a line. 
 In the center, over the eyes of the 'god,' is a conventionalized 
 
346 Philip. Ainsworth Means, 
 
 human face. Typical forms of this headdress are shown on 
 Plate II, Figure 3, on Plates III and IV. 
 
 (5) The Ceremonial Staffs. The Plates already mentioned 
 show good examples of the staffs. It is to be noted that in vase- 
 paintings where the mouth-mask, headdress and hands preserve 
 the greatest amount of naturalism the staff most closely approxi- 
 mates the spear-thrower shown in Plate II, Figure 2, though at 
 no time is the resemblance very strong. In the more conven- 
 tionalized designs, however, the staffs (here usually two in num- 
 ber and so arranged as to be bilaterally symmetrical) are 
 themselves so conventionalized as to be scarcely definable in 
 regard to their use. 
 
 Bearing in mind the well-known principles that apply to dec- 
 orative arts, the principles of elimination and simplification which 
 will be spoken of later, the writer ventures to suggest that of 
 the two groups of pottery that we have been studying, that exem- 
 plified by Plate II, Figures 1, 2 and 3, is the older, and that 
 the "Centipede Gods" on Plates II, III and IV were a later 
 style. So much, then, for the modelled ware and for the "Centi- 
 pede God" motif. 
 
 We will now examine another motif which may be called, for 
 the sake of convenience, the "Multiple-headed God." Our Plate 
 II, Figure 5, shows an excellent specimen of this motif. Another 
 is shown by Joyce (1912, Plate I). In this motif the heads of 
 the personage consist of hardly more than eyes and mouth and 
 tongue. In some cases, the body of the 'god' has a chief head 
 in approximately the correct position. Then, running out from 
 the shoulders, are a lot of subsidiary heads attached to the body 
 by their run-out tongues. The subsidiary heads are decorated 
 by feather-like rays reminiscent of the decorations on the mouth 
 mask seen on Plate III, Figure 3. Sometimes, as in Joyce's 
 Plate I, the chief head has a headdress of the type associated with 
 the "Centipede God." Also, the "Multiple-headed God" and 
 the "Centipede God" have other points in common, notably: 
 (1) The occasional presence of a centipede-like girdle with the 
 tongue sticking out (see Joyce, 1912, Plate I) ; (2) The pres- 
 ence of four-digit hands (though five-digit hands sometimes 
 appear in both) ; (3) The presence of the minor decoration, 
 seen in our Plate II, Figure 3, and in Joyce's Plate I, made up 
 
A Survey of Ancient Peruvian Art. 347 
 
 of two thick rings with a tassel or tassels hanging from them : 
 (4) The beginnings of a marked tendency toward bilateral sym- 
 metry, both of line and of color: (5) The continuance in the 
 "Multiple-headed God" motif of the rich coloration found in the 
 "Centipede God" motif. (Joyce's Plate I shows the presence 
 of buff, blue, yellow, purplish-red, pink, white and black.) The 
 mouth-mask and ceremonial staff usually do not appear in the 
 "Multiple-headed God" motif designs. 
 
 The "Centipede God" and the "Multiple-headed God" appear 
 to be the chief personages of Proto-Nasca vase-painting. They 
 do not, however, by any means include all the forms that go 
 to make up this complex art. Space permits us to mention only 
 one other constantly recurring feature. This is the human face 
 which is to be seen in our Plate IV, Figure 2, and in Joyce's 
 Plate I, at the base of the vessels. When this face appears 
 thus, painted, not modelled, it strongly suggests the modelled 
 faces that appear in Plate II, Figure 1. The manner in which 
 the eyes are shown, the hair-dressing, the nose and the mouth 
 are all strikingly alike in both the modelled and the painted 
 versions of the motif. At the same time, it should be noticed 
 that very often lines suggestive of tears run down a short dis- 
 tance from the eyes of the painted forms, but not from those 
 of the modelled ones. 
 
 What has been said of Proto-Nasca art is, of course, very 
 far from beginning to be an exhaustive study of that subject. 
 It is, however, enough to give a fair idea of the chief features 
 of that culture. It is but right to say, nevertheless, that aside 
 from the vases bearing decorations more or less anthropomor- 
 phic or zoomorphic, whether modelled and painted or merely 
 painted, there is another class of Proto-Nasca vessels which, 
 though having the rich coloration and the same general technique 
 of the other classes, is merely decorated with such patterns as 
 dots, lines and so on like those which appear in some of Ber- 
 thon's Plates (1911). 
 
 We will now attempt to draw up a classification of the sub- 
 types of Proto-Nasca pottery. Then we will take up the ques- 
 tion of Proto-Nasca textiles. 
 
 Although all Proto-Nasca pottery may be said to be distin- 
 guished by a subordination of form to color and of realism to 
 
 Trans. Conn. Acad., Vol. XXI 24 1917 
 
34 8 Philip Ainsworth Means, 
 
 complexity, it is not enough for us to content ourselves with 
 this general sort of statement. We must look further with a 
 view to establishing various sub-types of Proto-Nasca art, for 
 it must necessarily be assumed that the people who produced the 
 art flourished for at least two or three centuries and that they 
 developed in that time a number of modifications which appear 
 in their productions. Before we do this, however, we must 
 definitely assure ourselves as to whether we have been correct in 
 assuming that Proto-Nasca art was indeed related to or descended 
 from Proto-Chimu art. For the present we shall content our- 
 selves with examining into the relationship of the two without 
 attempting to prove the descent of one from the other. The 
 Plates in this article, those in Joyce's article on the Clan-Ancestor 
 (Joyce, 1913b), those in Berthon (1911), and in the articles by 
 H. R. H. Prinzessin Therese von Bayern (1907), and Uhle 
 (1914), afford ample material for a comparison. An exami- 
 nation of the two arts brings out the following points of con- 
 tact : ( 1 ) The use of eye-painting and masking ; (2) The presence 
 of feather-like ornaments; (3) The use, in connection with the 
 costume, of various appendages and adornments derived from 
 or suggested by animals or parts of animals (i. e. such ele- 
 ments as the centipede girdles) ; (4) The gradual transition from 
 realistic, modelled, five-fingered Proto-Chimu art to partly real- 
 istic, modelled five-fingered or four-fingered Proto-Nasca art, 
 together with the apparently contemporaneous rise of non-mod- 
 elled, constantly more conventionalized forms of vase-painting. 
 
 From the foregoing it will be seen that a very real underlying 
 similarity of subject-matter binds Proto-Chimu art to Proto- 
 Nasca. 
 
 Reserving for another place the critical consideration as to 
 the descent of Proto-Nasca art, we will now present a tentative 
 classification into sub-types on a combined basis of form and 
 decoration. 11 
 
 11 The author wishes to call attention to the very able study of Nasca 
 pottery by Edward K. Putnam (1914), and to say that he departs from 
 the classification of Proto-Nasca pottery offered by Mr. Putnam only 
 because it is too detailed for his present purposes and because it does not 
 emphasize the points he wishes to bring out. 
 
A Survey of Ancient Peruvian Art. 349 
 
 CLASSES REMARKS 
 
 I Modelled and painted ware . The class most like Proto-Chimu. 
 
 a Semi-realistic That is, having fair modelling in 
 
 combination with five-digit 
 hands, 
 b Non-realistic That is, poor modelling com- 
 bined with four-digit hands. 
 II Painted ware— not modelled The predominant Proto-Nasca 
 
 type. 
 a "Centipede God" 
 
 Motif Perhaps derived in part from the 
 
 Proto-Chimu habit of masking, 
 b "Multiple-headed 
 
 God" Motif Linked to "Centipede God" in 
 
 several ways (see above) and to 
 Proto-Chimu by use of feather- 
 like ornaments. 
 c Painted human 
 
 face motif Found usually on the same ves- 
 sels as the two foregoing types, 
 it is, at the same time, strongly 
 like Class I, a, and Class I, b. 
 d Miscellaneous ..... Forms made up of all sorts of 
 
 elements borrowed from the 
 foregoing types. 
 
 Having reviewed the distinguishing marks of Proto-Nasca art 
 as represented by the pottery, we have now arrived at the impor- 
 tant question of Proto-Nasca textiles. It has been said by good 
 authorities that there were no textiles dating from a time prior 
 to the rise of the culture of Tiahuanaco II. 12 
 
 It is, however, the opinion of the present writer that this 
 belief is a mistaken one. It will, no doubt, be granted by any- 
 one that if Proto-Nasca textiles do survive to the present day, 
 they will have the same or similar designs upon them as do the 
 pottery remains. We shall endeavor to show that such designs 
 do survive in textiles. Before doing so, however, it will be well 
 to remind the reader that there is no class of textiles that can 
 safely be assigned to the Proto-Chimu culture. 
 
 The reader's attention is called to Plate IV, Figures 1 and 3. 
 The first shows a woven cloth from lea now in the Museum of 
 
 12 Joyce, 1912, p. 200; Beuchat, 1912, p. 574. 
 
35° Philip Ainsworth Means, 
 
 Fine Arts, Boston. The second shows a border from an lea 
 shawl in the American Museum of Natural History, New York. 
 We will begin with an examination of Plate IV, Figure 3. In 
 it we discover a number of striking analogies with Proto-Nasca 
 pottery designs. Like the vase-paintings of the "Centipede- 
 God" type, this design has : ( I ) A mouth-mask which combines 
 wing-like side ornaments of the same type as those on Plate III, 
 Figures I and 3, with a pair of feather-like ornaments reminis- 
 cent of those in Plate III, Figure 3. (2) The hands of the figure 
 (as far as one can tell) and its feet have less than the true 
 number of fingers and toes. (3) The tongue is run far out 
 and is highly decorated, a tendency already shown in the pottery. 
 (4) The headdress is broad and flat; it has a brim made up of 
 two layers and there is a conventionalized human face in the 
 center. Compare it with the headdresses on Plate II, Figure 3 ; 
 on Plate III, Figure 1, and on many other Proto-Nasca vase- 
 paintings of the "Centipede God" type. Also remark that in 
 this textile design, as in some examples of the "Centipede God" 
 pots, the centipede element is preserved by the girdle-like append- 
 age. The tongue of the figure likewise reminds one of the centi- 
 pede motif. In other words, of the five criteria that we found 
 to be distinctive of the very important "Centipede God" motif 
 Proto-Nasca vessels, four are present in the textile design which 
 we have been studying. Does not this suggest that the textile 
 and the vase-paintings in question have a common source which 
 accounts for their similarities in subject-matter? Again, Plates 
 V and VI seem to have several points in common with Proto- 
 Nasca pottery, although, on account of the comparative complex- 
 ity of their embroidered designs, it is hard to know whether to 
 compare the personages they portray with "Centipede God" or 
 with the "Multiple-headed God." For this reason, therefore, 
 it will be best for us to content ourselves with comparing these 
 textile designs with Proto-Nasca vase-paintings in general. The 
 following features, then, may be observed in both the textiles 
 in question and in various specimens of Proto-Nasca pottery: 
 (1) The mouth-mask with wing-like side ornaments; (2) The 
 protruding tongue, highly decorated and endowed with centi- 
 pede-like attributes ; (3) The broad flat headdress decorated with 
 a conventionalized human face; (4) The color-scheme is very 
 suggestive of Proto-Nasca pottery. 
 
A Survey of Ancient Peruvian Art. 351 
 
 The foregoing remarks are meant to convey the impression 
 that Proto-Nasca designs do occur on both pottery and embroid- 
 ered textiles. Since this is so, the writer finds it impossible to 
 imagine how anyone can assume, as some have done, that they 
 were not made by the same people. Differences between the tex- 
 tile designs and the pottery designs do exist, of course, but in the 
 writer's opinion, they may all be explained by the difference in 
 medium, the technique of pottery decoration not unnaturally 
 causing results divergent from those produced by textile 
 embroidery. 
 
 Without pausing at present to discuss the transition from 
 Proto-Nasca art to Tiahuanaco II art, we will now turn our 
 attention to the region of Lake Titicaca and study the early 
 cultures in that area. 
 
 2. A CRITICAL ANALYSIS OF TIAHUANACO II ART. 
 
 Hitherto in our study we have had to deal mainly with pottery 
 and textiles. In the case of Tiahuanaco II art stone adds itself 
 to the other two as an important art-medium. If it were our 
 purpose to follow Tiahuanaco II art in all its ramifications, we 
 should have to consider the bronze work of northern Argentina 
 as well. The chief media, then, for the art of the exceedingly 
 important period we are about to study are: in the highlands, 
 stone and pottery ; in the coast-regions, pottery and textiles. 
 
 We will first examine the Tiahuanaco II art with a view to 
 setting forth its content and characteristics. The reader is urged 
 to turn to Plate VII which shows the chief figure of the great 
 monolithic gateway at Tiahuanaco. The Plates in Posnansky's 
 work on this site should likewise be consulted. 13 For want of a 
 better term we have referred to this figure as the "Weeping- 
 God." Variations of it occur over a huge area, and in stone, 
 pottery, textiles and bronze. Sometimes the "tears" are lack- 
 ing, but there is always some other feature to identify the several 
 variations. We will now minutely examine the Weeping God 
 on the great monolithic gateway. He is a short stocky personage 
 with a large head which is almost square. Around the head is 
 a sort of frame; the inner band of the frame is adorned with a 
 series of the _Jj~ sign which we shall find often later on. 
 
 Posnansky, 1914, Plates LXV-LXXXIV. 
 
35 2 Philip Ainsworth Means, 
 
 Standing out from the inner band are twenty-four ray-like tabs 
 or tassels. These tassels, all of them conventionalized, fall into 
 three groups or classes, (i) The puma-headed tassels, six in 
 number and much conventionalized; (2) Tassels, seventeen in 
 number, composed of what look like ribbons ended off by stone 
 rings, but which cannot well be described with accuracy because 
 of their conventional nature: (3) One anthropomorphic tassel 
 showing the conventionalized face of a man with eyebrows and 
 nose shown continuous. The face of the Weeping God has been 
 destroyed by time to a deplorable extent, but two large round 
 eyes, deeply sunk, remain almost unharmed. From the eyes 
 hang two bands ended off with puma-heads. On each of the 
 bands are two sunken dots suggestive of "tears." The nose of 
 the figure has been shattered, but it was probably once quite 
 prominent. At present it is squarish and rather broad and long. 
 The whole face is covered over with traces of secondary orna- 
 mentation. The body is not separated from the head by any 
 definite neck. The mouth is a mere rectangular slit sunk in the 
 face, totally lacking in any true modelling, the body is short and 
 chunky, and the legs are much too short to be in proportion, 
 unless, indeed, a kneeling posture is indicated. The garment of 
 the figure is a short fringed skirt held up by shoulder-bands. The 
 top of the skirt is marked with rectangular decoration of a type 
 to be observed elsewhere on the carving, and by two puma-heads 
 similar to those on the headdress and elsewhere. The fringe 
 of the skirt is made up of six human faces of the same type 
 as that noted on the headdress. The shoulder-bands are adorned 
 with a conventionalized figure alternated with conventionalized 
 bird-heads. A large breast ornament hangs between the shoulder- 
 bands. It has the form of a fish in semi-lunar posture with his 
 head to the left and turned upward and his tail, to the right, also 
 turned upward. The face of the fish recalls the conventionalized 
 human faces already noted. Just below the fish is a repetition of 
 the conventionalized figure that appears on the shoulder-bands 
 and two other examples of the bird-heads that also appear there. 
 The arms of the Weeping God, though not at all true to nature, 
 are the best modelled parts of the figure. At each elbow are 
 two puma-heads, one above the other. From the two lower puma- 
 heads hang two more conventionalized human faces. The hands 
 of the figure have but four digits. In the right hand is a large 
 
A Survey of Ancient Peruvian Art. 353 
 
 ceremonial staff. The upper half of it bears a rectangular decora- 
 tion just like that on some of the tabs of the headdress. It is 
 surmounted by an indeterminate object. The lower half of the 
 staff is decorated in much the same way save for the fact that 
 the central panel is sunk as it is on the upper border of the skirt. 
 The base of the staff consists of a conventionalized bird-head. 
 In the case of the staff in the left hand of the Weeping God we 
 find the lower half identical with the one just described. The 
 upper half, however, is bifurcated and the two prongs are topped 
 by bird-heads similar to those already seen on the breast-orna- 
 ment. 
 
 Having enumerated in detail the features of the Weeping God, 
 it will be well for us to note in general terms some of its char- 
 acteristics. In the Weeping God, then, we have a highly con- 
 ventionalized bas-relief in stone which shows considerable artistic 
 advance. For one thing, the tendency toward bilateral symmetry 
 noted in connection with Proto-Nasca art appears again here, and 
 it has gained considerably in strength. Save for the staffs and 
 the breast-ornament, the Weeping God is bilaterally symmetrical, 
 and the exceptions to that symmetry do not in the least interfere 
 with the impression of perfect bilateral balance. Moreover, the 
 constant re-statement of three or four motifs of decoration in 
 various combinations is eloquent of conventionalization that has 
 been long in developing. Lastly, the technique of the bas-relief 
 is of that square-edged type which would naturally develop out 
 of a round-surfaced stone technique after conventionalization 
 had set in. 
 
 On the same gateway with the Weeping God are forty-eight 
 secondary figures in relief of the same type. There are twenty- 
 four on each side of the central figure. Here again, the tendency 
 toward bilateral symmetry is observable, for all the figures face 
 toward the Weeping God. These secondary personages fall into 
 two classes : ( 1 ) Those with bird-like bodies and human faces ; 
 and, (2) Those with bird-like bodies and bird faces. Each of 
 them bears before him a staff which approximates in form to 
 one or the other of those held by the Weeping God. All the fig- 
 ures of both classes have four-digit hands, tears and tear-lines, 
 and a constant repetition of the If"" sign, and of the puma-, 
 fish- and bird-head motifs. The wings of the figures are, in part, 
 almost realistic, and they recall the fluted wings we noted in con- 
 
354 Philip Ainsworth Means, 
 
 nection with the Proto-Chimu art. All the figures are represented 
 as running toward the Weeping God, and the speed of their 
 motion is well indicated by their cloak-like garments which are 
 streaming out behind them. 
 
 Repetition and re-statement of decorative motifs and themes, 
 together with the tendency toward symmetry, may be said to be 
 the underlying principle of the conventionalities of Tiahuanaco 
 II art as embodied in the monolithic gateway. It is especially 
 noted in the frieze which runs the whole length of the gateway 
 just below the Weeping God and just above the doorway. 
 Thoughout that whole composition fragments and portions of 
 motifs already noted can be picked out. 14 
 
 Aside from the typical Tiahuanaco II decorations on the several 
 gateways at Tiahuanaco (the others are unimportant), the same 
 or similar motifs appear on the pottery from that vicinity. The 
 American Museum of Natural History has a fine collection of 
 Tiahuanaco II pottery from Copacabana and Tiahuanaco. In 
 general the tonality is rather sombre, red and black being the 
 most frequent colors. Sometimes, however, white and orange also 
 appear. In the Peabody Museum at Harvard University there is 
 a small but excellent cup of this period decorated with the face 
 of the Weeping God. Sometimes, as in the case of some of the 
 New York specimens, the Weeping God appears in the pottery 
 without his tears ; at other times the decoration takes the form 
 of parts of the secondary motifs, such as puma- or bird-heads 
 in the Tiahuanaco style, or variations of the second type of tab 
 on the Weeping God's headdress (i. e. the "ribbon-and-stone- 
 ring" motif). Cups, bowls, ollas and vessels with spouts like 
 those on teapots are the commoner forms. One of the New- 
 York specimens measures almost a foot across although it is but 
 a fragment. Modelled puma-heads in clay also occur. In short, 
 the plastic art of the Tiahuanaco II period, although it is none 
 too plentifully represented in our museums, is richly diversified. 
 
 Our Plate VIII, Figure i, shows a poncho from Tiahuanaco 
 now in the American Museum of Natural History, New York. 
 The writer believes that, although it bears none of the motifs 
 so far shown to have been typical of Tiahuanaco II art, it does 
 bear a swastika-like motif on its border, and is therefore to be 
 
 14 See Posnansky, 1914, Plates LXXIII-LXXXI. 
 
A Survey of Ancient Peruvian Art. 355 
 
 connected with a cup with the same motif shown by Joyce (1912, 
 p. 207). This cup, both because of the puma-heads and because 
 of the general technique, is obviously Tiahuanaco II. The swas- 
 tika is a motif which is excessively rare in Peruvian art. The 
 ones in this specimen are not perfect in form. The reader is 
 warned that this garment may not be Tiahuanaco II after all, 
 though the writer now believes that it is. 
 
 Plate VIII, Figure 2, shows a fine piece of cloth from the 
 Nasca region. It has affinities with both the Proto-Nasca and 
 Tiahuanaco II styles as follows: With Proto-Nasca, face-paint- 
 ing (or masking), centipede element, and coloration; with Tia- 
 huanaco II, tear-lines, eyebrows and nose in T form, three-digit 
 hands. The specimen is in the Museum of Fine Arts, Boston, 
 and is the gift of Dr. Denman Waldo Ross. 
 
 Though it may at first seem illogical, we will now turn our 
 attention to the Tiahuanaco II art of the coast; then we will 
 study it in another region of the highlands. The reason for 
 this course will become apparent later. 
 
 The writer regrets that he has not been able to obtain any 
 satisfactory pictures of coast Tiahuanaco II art. Much material 
 is readily accessible to the student, however, and the following 
 works should be consulted: Baessler, 1902-03; Reiss und Stub el, 
 1880-87; Holmes, 1889; Oyarzun, 1910; Uhle, 1901, 1902, 1003, 
 1908, 1910, 1910b, 1910c, 1912, 1913, 1913b, 1914; Putnam, 
 F. K., 191 4; Therese von Bayern, 1907; Beuchat, 1912; Joyce, 
 1912, 1913b; Bamps, 1879. (The reader is especially referred 
 to the works in italics.) 
 
 It will be remembered that in the Tiahuanaco II art of the 
 interior two things were very noticeable: the tendency toward 
 bilateral symmetry in the design, and the comparative poorness 
 of coloration. Of these characteristics only the former appears 
 on the coast. As in the case of Proto-Nasca art, coloration on 
 both vessels and textiles was extremely rich. For example, look 
 at Plate 134, Figure 373, in Baessler. The design that appears at 
 that place shows two birds with squarish heads. The design 
 comes from Pachacamac. A detailed description of it may be of 
 use to the reader. The two birds, whose heads alone appear, 
 face one another. They have hawk-like beaks, darkened eye- 
 areas and headdresses adorned with tabs ending off in three 
 fringes just like those on the minor figures of the monolithic gate- 
 
35 6 Philip Ainsworth Means, 
 
 way at Tiahuanaco. The angularity so noticeable in the art of 
 this period at Tiahuanaco itself is here preserved to a consid- 
 erable degree. Between the two bird-heads and around each of 
 them is a frame or border adorned with repetitions of the IT 
 sign. Although the Plate in question is not in color, several 
 tints are indicated. Again, Baessler, Plate 144, Figure 403, 
 shows a wonderful specimen of coast Tiahuanaco II art. It is 
 a goblet from Pachamac adorned with a very beautiful design. 
 The colors are cream, purple, gray, brown, red and black. The 
 finish is lustrous and the arrangement of the color-areas is mas- 
 terly. The decoration resolves itself into several bands. At 
 the top is a band of the stair-sign motif ; it is gray with purple 
 borders. Attached to the outer edge of the borders are a num- 
 ber of conventionalized puma-heads in purple. They are remin- 
 iscent of those on the monolithic gateway. Those on the top 
 of the band face to the reader's right; the ones at the bottom 
 face to the left. On the gray central stripe of the stair-sign 
 band are a number of conventionalized three-digit bands in black 
 and gray alternated with similar feet in brown, cream and black. 
 Below this band of decoration comes a narrower one made up of 
 IT" signs in red on a cream ground. Below that, in turn, 
 comes a wide band of black on which is painted an almost 
 bilaterally symmetrical square-headed Weeping God. A slight 
 difference in the two ends of his mouth is the sole exception to 
 symmetry. His eyes are in cream and black and, like those 
 of the Weeping God on the monolithic gateway, are large 
 and round with a band of "tears" running down from 
 each of them. The face is red, the nose, gray in color, is 
 broad and squarish like that of the Weeping God at Tia- 
 huanaco. His gray lips form a rectangular mouth contain- 
 ing three groups of rectangular teeth and two groups of fangs, 
 the order being, from left to right, teeth- fangs-teeth- fangs- 
 teeth. The teeth are cream-colored. Finally, at the bottom of 
 the design, comes a band containing twelve oblong rectangles on 
 each of which are two small disks of color with a dot in the 
 center. These rectangles are arranged in double file, six in a 
 row. They are arranged in the manner here approximately indi- 
 cated, and they may be said to be a sort of study in color-arrange- 
 ment. Numbers 1, 3, 5, 8, 10 and 12 are red with cream disks; 
 2 is cream with purple disks; 4, 9 and 11 are gray with red, 
 
A Survey of Ancient Peruvian Art. 
 
 357 
 
 purple, red disks respectively ; 6 and 7 are purple with cream 
 disks. What this design can have been intended to represent it 
 is difficult to imagine. The only thing it seems to bear the 
 slightest resemblance to is the group of finds on the island of 
 La Plata, Ecuador, which Dorsey called "Perforated and 
 engraved stones." These objects are small rectangular oblongs 
 upon which are engraved circles with a dot, the number varying 
 from three circles up to eight. Dorsey suggests that perhaps the 
 stones in question were used in some game. 15 
 
 1 
 
 
 2 
 
 
 3 
 
 
 A 
 
 
 5 
 
 
 6 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © 
 
 
 
 
 
 
 
 
 
 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 
 © © 
 
 10 
 
 11 
 
 M 
 
 Fk 
 
 The decoration composed of small circles with a dot in the 
 centre occurs also on some objects from Machu Picchu. Its 
 occurrence there may mean one of several things : ( 1 ) That 
 some subjects of the Tiahuanaco II "empire" were once at 
 Vilcabamba and left these objects behind them; (2) That the 
 subjects of the Inca who dwelt at Vilcabamba happened to see 
 the motif on some remains of the former period and copied it; 
 (3) That the design, which is essentially simple, was "invented" 
 twice, first by the coast Tiahuanaco II people, secondly by the 
 Inca's subjects. The writer inclines to the belief that the last 
 is the correct explanation. It is, however, so simple a design 
 that it has been "invented" again and again in various parts of 
 the world. 
 
 One more example of Tiahuanaco II art in connection with 
 the pottery of the coast will serve to round out our present brief 
 account of the matter. It is found in Baessler, Plate 140, Figure 
 392, and it shows another variation of the square-headed Weep- 
 ing God motif. The colors are red, cream, brown and white. 
 The figure is shown at full length. As before, the resemblances 
 to the Weeping God of Tiahuanaco, despite the divergences, are 
 
 Dorsey, 1901, p. 262, and Plate LVII. 
 
35 8 Philip Ainsworth Means, 
 
 very marked. The face of this coast Weeping God, then, is sur- 
 rounded by a frame strikingly like that of its prototype. From 
 the inner band decorated with dots spring eleven tab-like decora- 
 tions which fall into three groups on the basis of form: (i) 
 Three straight tassels ending in a fringe of three pieces; (2) 
 Six tabs reminiscent of the ribbon-and-stone-ring tassels of the 
 Weeping God of Tiahuanaco; (3) Two long tabs ending off in 
 an affair similar to the fringed tabs of the first type. All these 
 are arranged about the face in such a manner as to result in 
 absolute symmetry. In fact, the whole figure is absolutely sym- 
 metrical save for the arrangement of the color-areas. The eyes 
 are cream and brown and are large; the "tears" are indicated 
 merely by two lines, one running down from each eye. The nose 
 is broad and square ; the mouth rectangular with eight square 
 teeth and no fangs. Two hands with four digits grasp two 
 staffs which are exactly alike except for color and which are 
 arranged symmetrically. They suggest tremendously conven- 
 tionalized bows, and are adorned with repetitions of the fringed 
 tab element. In short, this figure, besides presenting several 
 very close resemblances to the Weeping God of Tiahuanaco in 
 its details, resembles it in more general terms also. We find in 
 the Tiahuanaco figure a strong tendency toward bilateral sym- 
 metry, a symmetry which is fully attained in this coast figure. 
 More than that, we observe that the two have another significant 
 characteristic in common, namely, the constant re-statement of 
 minor decorative elements (such as the fringed tab). These 
 resemblances are extremely significant. 
 
 So much, then, for Tiahuanaco II designs on the pottery of 
 the coast. The material relating to Tiahuanaco II designs in 
 coast textiles is no less ample, and the evidence it presents points 
 just the same way as that offered by the pottery. We will, 
 therefore, consider only one example of Tiahuanaco II coast 
 textile-design. It is shown by Reiss and Stiibel, vol. II, Plate 49. 
 It is a rich garment from Ancon. There are two variations 
 either of the Weeping God himself or of the two types of minor 
 figures on the monolithic gateway. We will enumerate the analo- 
 gies between this design and other arts that we have examined. 
 One of the two variations has : ( 1 ) A human face and a head- 
 dress suggestive of the first type of minor figure on the mono- 
 lithic gateway (i. e. human face with bird body) ; (2) Tears 
 and tear-lines; (3) Four-digit hands; (4) Two staves; (5) 
 
A Survey of Ancient Peruvian Art. 359 
 
 Shoulder-bands; (6) Fish and bird attendants; and, (7) A 
 mouth similar in shape to those on the monolithic gateway. The 
 other variation has : ( 1 ) Fluted wings recalling both those found 
 in Proto-Chimu art and those found on the minor figures of the 
 monolithic gateway; (2) Five-digit hands reminiscent of those 
 on some of statues at Tiahuanaco and of Proto-Chimu art, 
 also; (3) Tears and tear-lines; (4) One staff; (5) A decora- 
 tion on the headdress suggestive of the "ribbon-and-stone-ring" 
 motif of the Weeping God both at Tiahuanaco and on the pot- 
 tery just reviewed. The colors in this tapestry are by far the 
 richest we have yet come across, and they are not likely to be 
 surpassed. They are yellow, light yellow-brown, dark yellow- 
 brown, red, pink, pale green, purple, black and white. The effect 
 is one of great richness, and also of a generally light tonality, 
 wherein, perhaps, we may see the influence of Proto-Chimu 
 art. To sum up our impressions of coast Tiahuanaco II art we 
 will say that it derives its minor motifs and its tendency toward 
 symmetry, or rather its marked indulgence in symmetry, directly 
 from Tiahuanaco which also provided most of the subject-matter. 
 The rich coloring, however, came from Proto-Nasca. 
 
 Having now completed our survey of Tiahuanaco II art in 
 the Titicaca drainage and on the coast, we will examine its 
 manifestations in another part of the highlands. Before doing 
 so, however, we will mention in passing the fact that save for a 
 vestige here and there Tiahuanaco II art does not appear promi- 
 nently in the Cuzco region. One exception to this rule is a 
 pottery vessel adorned with an anthropomorphic puma having 
 four digits, fangs and tab-like head ornaments. Its provenance 
 is Cuzco, and it is shown by Seler. 16 
 
 It is the Tiahuanaco II art at Chavin de Huantar, however, 
 that claims the major part of our attention. 
 
 The chief example of ancient art at Chavin is the famous 
 greater Chavin monolith. This wonderful piece of stone-carving 
 is in the Museo Nacional at Lima. It is about six feet long and 
 two broad. 17 Probably no other single artifact from Peru helps 
 more than this in the study of the relations between Proto- 
 
 16 Seler, 1893, Plate VII, Fig. 8. 
 
 17 The writer has seen and examined the original stone. Both Sir 
 Clements Markham and Mr. Joyce are mistaken in thinking the stone to 
 be twenty-five feet long. Markham, 1912, p. 34; Joyce, 1912, p. 176. 
 
360 Philip Ainsworth Means, 
 
 Chimu, Proto-Nasca and Tiahuanaco II arts. Several able 
 studies of the stone have appeared, chief among which are two 
 by Markham and that by Polo. is With the aid of our Plate IX 
 we will now examine this stone and its bearing upon our subject. 
 The characteristic of the stone which first strikes the beholder 
 is the tremendous elaboration of the design. One has to study 
 it carefully before it resolves itself into its component parts. 
 When this is done, it becomes apparent that the design falls into 
 halves, the lower of which shows a personage holding two staves, 
 and the upper of which is made up of a mass of inverted faces 
 with their secondary decorations. We will study the halves in 
 that order. The personage is unquestionably derived in part 
 from the Weeping God motif. The face is square and is edged 
 with serpent-heads faintly analogous to the tab-like ornaments 
 of the Weeping God. The face, on the other hand, is utterly 
 different in both content and treatment from that of the Weeping 
 God. Indeed, it is very difficult to decide just which of the 
 numerous complex features belong to the face of the personage. 
 One may assume, if he chooses, that the two upper dots are his 
 eyes and the involutions just above them are conventionalized 
 eyebrows while the two dots below are nostrils. This is, per- 
 haps, the most satisfactory interpretation. 19 The mouth which, 
 from one aspect, looks like an adaptation of the toothed and 
 fanged rectangular mouth seen in coast Tiahuanaco II, again 
 presents difficulties because, on turning the Plate upside down, 
 it turns out that the mouth is formed by two fanged puma-heads 
 set nose to nose and lip to lip. It may be suggested that in the 
 group of details formed by the puma-heads and the twined ser- 
 pent-heads just behind each of them we see a faint survival 
 and tremendous conventionalization of the mouth-mask of Proto- 
 Nasca art. As in the case of the Weeping God on the mono- 
 lithic gateway, the body is short and square. There are no 
 
 18 Markham, 1904 and 1908; Polo, 1899. 
 
 19 Prof. MacCurdy's interpretation of the plate differs from the writer's, 
 for he thinks the two upper dots to be the nostrils of an inverted face 
 like those on the upper half of the stone. There is a good deal to be said 
 in support of this view. But an examination of our Plate IX, or, still 
 better, the large one in Polo, 1899, will show that the writer's interpreta- 
 tion is also valid. We may say, therefore, that the two dots in question 
 serve, in one position, as eyes for the face of the chief head of the design, 
 and, when reversed, act as nostrils for an inverted subsidiary head. 
 
A Survey of Ancient Peruvian Art. 361 
 
 "tears." There is an area of ornamentation on the breast made 
 up of a new variation of the _TJ~" sign edged with feather- 
 like ornaments reminiscent of Proto-Chimu art and Proto-Nasca 
 art. (See Plate I, Figure 3, and Plate III, Figure 3.) This 
 feather-motif occurs many times on the stone. The garment of 
 the personage reminds us of that on the Weeping God of Tia- 
 huanaco in that it is a short skirt-like affair. The puma-heads 
 that adorn the upper edge of the Tiahuanaco figure's skirt have 
 here become so conventionalized that it is nearly impossible to 
 recognize them. The fringe of human faces on the Tiahuanaco 
 skirt has become mere unadorned rectangles. The arms, it is 
 well to note, are'in exactly the same position and much the same 
 in shape both here and on the monolithic gateway. But a marked 
 difference is found in the hands. At Tiahuanaco we found the 
 hands of the Weeping God were fairly close to nature in their 
 modelling despite the fact that they had but four digits. Here, 
 on the other hand, we find a wider departure from realism in 
 the drop to but three digits and in the elaboration of the finger 
 nails into a decorative element. In the two staves we discover 
 a still wider departure from the original theme. The staves are 
 almost exactly alike, which is in itself a significant matter. They 
 have been widened so as to make room for the immensely elab- 
 orate ornamentation with which they are encrusted. So complex, 
 in fact, is the overlaid design that it is nearly impossible longer 
 to distinguish any of the features that we perceived in the staves 
 held by the Weeping God of Tiahuanaco. Some may be able to 
 discover in the formalized faces at the base of the two staves 
 a faint echo of the bird-heads that are found at the bottoms of 
 the Tiahuanaco staves. 
 
 So much, then, for the lower half of the design on the Chavin 
 stone. In order properly to study the upper half it will be neces- 
 sary to reverse the Plate. On doing so we find three grotesque 
 faces proceeding from one another's mouths and each with its 
 tongue protruding and highly decorated. These faces all have 
 fangs, but otherwise they are unlike one another, although the 
 last two from the center do resemble each other closely. The 
 nose of the first face is adorned with a combination of the. 
 feather-motif, fang-motif and serpent-head motif. The noses of 
 the other two are much simpler and are marked only by an odd 
 but simple comb-like figure. On each side of the central band 
 of decoration formed by those three faces is a fringe of alternated 
 
362 Philip Ainsworth Means, 
 
 serpent-heads and feather-motifs. The tongue' of the last head 
 is likewise encrusted with the two. 
 
 A word about the general features of this, the greater Chavin 
 stone, should be said before we go on to compare it with other 
 artifacts. It is a bas-relief of the same technique as the Tia- 
 huanaco frieze. The work is finer because the stone lends itself 
 more readily to the cutter's tools. 
 
 At Chavin is another remarkable stone carving, the lesser 
 Chavin stone. It is described by Polo and by Enock. 20 It was 
 found in an underground chamber ; indeed, according to Enock, 
 much of the work and many of the chambers in the "castle of 
 Chavin" are subterranean. This feature is reminiscent of Tia- 
 huanaco itself. The lesser stone is at once similar to and differ- 
 ent from the greater. The chief points of likeness are the 
 profuse use of fangs and serpents as decorative motifs, and 
 the constant re-statement of these motifs recalls not only the 
 greater Chavin stone, but also the monolithic gateway. The 
 differences are chiefly these : lack of any trace of comprehensible 
 composition, lack of bilateral symmetry and considerable modi- 
 fication of technique. 
 
 As our description has proceeded we have made occasional 
 references to resemblances between the Chavin stones and other 
 objects. It will now be our task to systematize these resemblances. 
 Each of the elements which constitute the resemblances will be 
 found in the following table in its appropriate column : 
 
 Affiliations Between Chavin and Other Arts. 
 
 proto-chimu proto-nasca tiahuanaco 
 
 Fangs Fangs 
 
 Feathers Feathers 
 
 Multiple inverted heads. 
 
 Puma-heads 
 Staves Staves 
 
 Symmetry 
 Mouth mask ( ?) 
 
 Too few digits Too few digits 
 
 Skirt 
 JJ~ sign 
 
 Repetition of motifs 
 in many parts 
 of the design. 
 
 20 Enock, 1907, p. 72 ff. ; and Polo, 1899. 
 
A Survey of Ancient Peruvian Art. 363 
 
 The table makes clear, perhaps, the three-fold source of the 
 art found in the Chavin stones. We now find ourselves brought 
 to the important question of the historic, artistic and ethnic rela- 
 tions between the three great arts we have studied. 
 
 3. RELATIONS BETWEEN PROTO-CHIMU, PROTO-NASCA 
 AND TIAHUANACO II. 
 
 We have now studied three ancient Peruvian cultures. It is 
 obvious that, from both the artistic and the archaeological points 
 of view, they form a group. We must now endeavor to 
 answer the question, How are these cultures connected? 
 
 Already we have pointed out the basic similarity in subject 
 matter of Proto-Chimu and Proto-Nasca. From one of those 
 cultures the other in all probability was derived. But which 
 was the elder is only revealed by minute analysis. In the Proto- 
 Chimu we find an art which is of a distinctly advanced nature. 
 It has, so far as we know, no introductory manifestations, cruder 
 in type than itself in its own locality. Inasmuch as advanced 
 arts do not suddenly spring into being from nothingness, it can 
 only be supposed that. Proto-Chimu art was introduced into the 
 region with which we associate it from some other region. The 
 same may be said of the Proto-Nasca art. Uhle and Joyce seem 
 to incline to the belief that this art is the elder of the two, and 
 Uhle believes it to have had an origin in the north, perhaps in 
 Middle America. 21 
 
 Let us see, then, if Proto-Nasca can really be justly considered 
 older than Proto-Chimu. In doing this we must first determine 
 from what area or areas it could have been derived (assuming 
 that it ivas derived from some source outside of the Andean 
 area). A rapid survey of the whole field of American civiliza- 
 tions assures us that only from one area could such cultures as 
 the Proto-Nasca and Proto-Chimu have been derived — Middle 
 America. There is much evidence that seems to point toward 
 all the South American cultures as having been derived from the 
 region to the north, but unfortunately this is not the place to 
 examine that evidence. We will assume, therefore, that if, as 
 seems probable, the Proto-Nasca and Proto-Chimu cultures were 
 
 21 Uhle, 1913, pp. 341 ff. ; 1914, pp. 15 ff. ; Joyce, 1912, pp. 178 ff.; Means, 
 1917. 
 
 Trans. Conn. Acad., Vol. XXI 25 1917 
 
364 Philip Ainsworth Means, 
 
 the result of a cultural migration, that migration must have had 
 its origin in Middle America. From a geographical standpoint, 
 then, it would be difficult to explain why the migrants, on their 
 way south, passed the region of Chimu and went first to that 
 of Nasca where they developed the Proto-Nasca art after their 
 arrival, and then gradually spread north along the coast, in time 
 reaching Chimu where the Proto-Chimu culture was perfected. 
 This theory is almost impossible to svipport on geographical 
 grounds. But geographical objections are not the only ones. 
 Other and more serious drawbacks to the theory present them- 
 selves. 
 
 These drawbacks we will now enumerate. In Proto-Chimu 
 art we have a realistic art which has so thoroughly outlived the 
 preparatory stages characteristic of all arts that there is hardly 
 a trace left of the crudenesses that mark the infancy of all sub- 
 civilized or high primitive arts. 22 It is a decorative art that has 
 reached so high a level as to combine no small degree of repre- 
 sentation with its decorative purpose. In other words, Proto- 
 Chimu art is one ripe for the influence of several principles- of 
 art-development. These principles all find their natural resultant 
 in Proto-Nasca art. 
 
 A few quotations from Dr. Spinden's work, "A Study of 
 Maya Art," will make clear this point. The mere fact that Dr. 
 Spinden is speaking of Middle American art does not alter the 
 fact that what he says applies equally well to Peruvian art or to 
 any other art of similar rank. 23 
 
 "In the imaginative modification of any given natural figure, 
 for the purposes of decorative art, there are a number of rather 
 definite processes. Each of these is amenable to the fundamental 
 principles of design, such as balance, rhythm and harmony, as 
 these terms have besn elucidated by Dr. Ross. 24 Each process 
 may show, moreover, the phases of conscious and unconscious 
 manipulation of the subject matter. Lastly, these processes of 
 intensive development of a design motive, .... work both 
 singly and in combination. It is possible to detect much of the 
 counterplay. 
 
 12 The general remarks made here are intended to apply solely to 
 decorative, as contrasted with representative, art. 
 
 23 Spinden, 1913, p. 38 ff. 
 
 24 Ross, 1901, 1907. 
 
A Survey of Ancient Peruvian Art. 365 
 
 "The processes are: 1. Simplification. 2. Elaboration. 3. 
 Elimination. 4. Substitution. 
 
 "Careful analysis of one group of designs after another, dur- 
 ing which special attention is paid to the changes in homologous 
 parts, makes pretty clear the matter in which the imagination 
 works. In the first place, imagination does not create, it merely 
 reshapes and recombines, taking suggestions and material from 
 
 any thing lying within the field of experience 
 
 "Simplification. . . . Dr. Harrison Allen discusses the rela- 
 tions between natural forms and art forms. He finds that the 
 tendencies of conventional art are : 
 
 1 st, to repeat the normal lines of the model; 
 
 2nd, to diminish the normal lines of the model ; 
 
 3rd, to modify according to a symbol; 
 
 4th, to modify according to mystic or religious ideas. . . . 
 
 ''Elaboration-. Of less real worth in the development of art, 
 but of more common occurrence, is the process of elaboration. 
 This process amplifies rather than reduces and by means of 
 adventitious ornament renders the original form more com- 
 plex. . . . 
 
 "Elimination. Elimination of one feature after another of a 
 natural motive till only one or two survive is a common phenom- 
 enon the world over in decorative art. In Maya art the process 
 is frequently observed in the case of the serpent. Very often 
 the entire lower jaw is omitted. . . . 
 
 "Substitution. The process of substitution likewise plays a 
 great part in all highly developed art, whether barbaric or 
 civilized. The substitution of new and striking details for old 
 and commonplace ones — even at the cost of the first meaning 
 of the design — is one of the simplest and most natural ways by 
 which imagination can reconstruct and revivify worn-out sub- 
 jects. . . . Especially in decorative art, details of a composi- 
 tion realistic or geometric may be progressively replaced by 
 other quite different details until in the end only a trace of 
 the original setting remains." 25 
 
 Let us now seek carefully to apply the principles outlined 
 by Dr. Spinden to Proto-Chimu, Proto-Nasca and Tiahuanaco 
 II arts. We will take up the principles in order. We find, in 
 the transition from Proto-Chimu to Proto-Nasca and to Tiahu- 
 anaco II (at Tiahuanaco, on the coast, and at Chavin), that an 
 important part is played by the second aspect of Simplification. 
 We have in Proto-Chimu a highly realistic decorative art in 
 which both modelling and painting seek to approach as nearly 
 
 25 Cf. Allen, 1881 ; Batchelder, 1910; MacCurdy, 191 1, 127. 
 
366 Philip Ainsworth Means, 
 
 as possible to the outlines of natural models. The outcome of 
 this seeking is skillful modelling and the correct number of such 
 parts as fingers and toes. In Proto-Nasca art, on the other 
 hand, the principle of simplification finds reaction in the diminish- 
 ing of the tendency toward modelled representation and of the 
 habit of carefully representing the correct number of ringers and 
 toes. The result is a simplification of the form of the vessels 
 and of the outlines of the hands at the expense of truth. The 
 principle of simplification makes itself still more felt in the tran- 
 sition from Proto-Nasca to Tiahuanaco II. In the former it 
 had only begun to make itself apparent; modelling, of a simpli- 
 fied sort, to be sure, still survived, as did also five-fingered hands. 
 In Tiahuanaco II, however, pottery with modelled forms of men 
 or animals is more rare, though modelling in stone is still found, 
 probably on account of the influence of Tiahuanaco I. But five- 
 fingered hands are here in the minority to a decided degree. 
 Simplification has caused the vanishing of realistic hands from 
 the decoration on the monolithic gateway. The status remains 
 the same in coast Tiahuanaco II. In the greater Chavin stone, 
 however, we find the last result of the influence of the principle 
 of simplification. In this stone the modelling is at its lowest 
 ebb, and the hands, with but three digits here, have lost nearly 
 all semblance to reality, and have become mere elements in the 
 scheme of decoration. Thus we see that simplification leads us, 
 step by step, down the line of the arts of this period — Proto- 
 Chimu, Proto-Nasca, Tiahuanaco II — in the order named, with 
 the Chavin stone as the culmination of its influence. 
 
 Let us now find out what application the principle of Elabora- 
 tion has to these arts. In Proto-Chimu we found that fangs, 
 eye-painting, animal-masks, animal-disguises and other similar 
 features were represented. Each of these is acted upon by the 
 principle of elaboration. Fangs, it is true, are not a prominent 
 element of Proto-Nasca art. But they appear with great prom- 
 inence in Tiahuanaco art, both of the mountains and of the coast, 
 and on the Chavin stones they often form an element of decora- 
 tion of the highest importance. Eye-painting, animal-masks and 
 animal-disguises all survive in Proto-Nasca art and are all more 
 elaborate there than in Proto-Chimu. So much so, in fact, that 
 their development can go but little further, and they almost dis- 
 appear in Tiahuanaco II art. It may be suggested, however, 
 
A Survey of Ancient Peruvian Art. 367 
 
 that eye-painting is elaborated into the tears and tear-lines of 
 Tiahuanaco II art while masking finds a faint elaborate revival 
 in the puma-heads at the mouth of the chief figure on the larger 
 Chavin stone. To show one more manifestation of the influence 
 of elaboration we will mention the "Multiple-headed God" motif 
 of Proto-Nasca art. The God is always distinguished by the 
 manner in which his subsidiary inverted heads proceed from one 
 another's mouths and by the presence of feather-like secondary 
 ornaments. Elaboration results in the multiple-headed figure on 
 the larger Chavin stone. That figure, like its Proto-Nasca proto- 
 type, has several inverted heads proceeding from one another's 
 mouths and it is marked by elaborate secondary decoration in 
 the form of feather-like ornaments. We must note here that as 
 nothing of the sort is to be observed at Tiahuanaco the trans- 
 mission from Proto-Nasca to Chavin must have been direct, and 
 that the two were at least partly contemporary. Likewise, as we 
 have pointed out, there are a number of Proto-Chimu elements 
 found on the Chavin stone. One more evidence of the influence 
 of elaboration should not be ignored. That is the contrast 
 between the staffs found in Proto-Nasca with those in Tiahuanaco 
 II and, above all, in Chavin. The contrast needs no comments, 
 save that here, again, the culmination of the process is found at 
 Chavin. 
 
 Nor do we lack for signs of the presence of influences on the 
 part of the principle of Elimination. As we have noted, the 
 Proto-Chimu art shows full realistic representation of the whole 
 of men and animals. Between Proto-Chimu and Proto-Nasca 
 we find an elimination of most of the body parts by the latter 
 art. In Tiahuanaco II, however, again probably on account of 
 the influence of Tiahuanaco I, the habit of showing the body is 
 revived, but some of the lines and curves of nature are markedly 
 absent, both in Tiahuanaco II art proper and at Chavin. 
 
 Finally, the principle of Substitution is readily seen to have been 
 at work. The eye-painting of Proto-Chimu and Proto-Nasca is 
 substituted by the tears and tear-lines of Tiahuanaco II. The 
 puma-head and ribbon-and-stone-ring tabs on the Weeping God's 
 headdress at Tiahuanaco are replaced by the serpent-heads that 
 occupy analogous places on the larger Chavin stone as well as 
 on the lesser one. Again, the fish-like breast ornament of the 
 Tiahuanaco figure finds a substitute at Chavin in the conventional 
 
368 Philip Ainsworth Means, 
 
 breast ornament of the figure on the larger stone. Once more 
 the culmination of the process is at Chavin. Indeed, in the lesser 
 Chavin stone one may see an excellent example of what Mac- 
 Curdy describes as "transposition." 26 It is to be observed in 
 the breaking up of the hitherto harmonious and comprehensible 
 design into a chaotic melee of component parts and ill-assorted 
 decorative motifs. One would be but reasonable in thinking the 
 lesser Chavin stone to represent the art-stream, which we have 
 watched so long, at its vanishing point. 
 
 Such, then, in very broad outline, is the general trend of the 
 evidence afforded by a study of the application of the four great 
 principles to Proto-Chimu, Proto-Nasca and Tiahuanaco II art. 27 
 We must now endeavor to interpret the evidence in terms of 
 probable cultural migration. There is not space here to go into 
 a detailed comparative analysis of the minor decorative motifs 
 in Middle American and South American cultures, but the writer 
 is convinced by careful study that the evidence of such an 
 analysis would not differ from that afforded by the broader lines 
 of modification. 28 
 
 To sum up the whole matter briefly, we find that a series of 
 closely related arts is associated in turn with Chimu, Nasca, 
 Tiahuanaco (mountains and coast) and Chavin. We find the 
 art a little older step by step as we go from one of these regions 
 
 26 MacCurdy, 1911, p. 127. 
 
 27 One piece of pottery, reported on by Uhle (1913b, p. 363). almost con- 
 stitutes in and of itself a proof of the blending or fusion of Proto-Nasca 
 art into Tiahuanaco II art. The vessel in question is a shallow bowl 
 from Tiahuanaco. On the broad rim is painted, in easily recognizable 
 Proto-Nasca style, a serpent, the head of which is strikingly like the 
 puma heads so often found in Tiahuanaco II art. The fact that the vessel 
 comes from Tiahuanaco proves that Proto-Nasca art was carried thither, 
 and the association of it with Tiahuanaco II art on the same vessel proves 
 their close relationship. 
 
 28 The reader's attention is here called to the art of Chiriqui. In many 
 ways strikingly similar both in form and in content to the three early 
 Peruvian arts, the art of Chiriqui is also similar to them in the matter of 
 its development toward conventionalism from realism. It may well be 
 that some day a close connection will be proved between the earliest 
 (realistic) forms of Chiriqui art and the earliest (realistic) forms of 
 Peruvian art. The reader is urged to consult the following works : 
 MacCurdy, 1906, 1911, pp. 127 ff., 1913 ; Holmes, 1885, 1887; Joyce, 1916, 
 pp. 144 ff. 
 
A Survey of Ancient Peruvian Art. 369 
 
 to the next, a little older, that is, in point of development ; the 
 age in point of time from our own day decreases as we go up 
 the series of sites. This does not mean, however, that one site 
 was abandoned before the next began to flourish. In fact, the 
 evidence proves that the first and last steps have much in com- 
 mon, and that they must have been at least partly contemporary. 
 The opinion of the writer is that one should conceive of the 
 slowly ageing art as the result of a slow spread of related peo- 
 ples in several directions during a long time. While the spread 
 was going on new sites were founded and new phases of the 
 common art-ideal developed, but neither the old settlements nor 
 the old phases of art were thereby at once robbed of vigor. 
 What the political status of these people was we shall never 
 know. We must remain content with what evidence we can 
 wring from the vestiges of their culture. 
 
 4. A CRITICAL ANALYSIS OF THE EPIGONAL AND 
 RED-WHITE-BLACK CULTURES. 
 
 So far, we have studied three cultures which we have just 
 seen to be intimately linked together by lines of cultural descent. 
 We have hitherto considered a cultural series that spread from 
 the coast to the mountains. We have now come to a fork in 
 that stream. 
 
 It is clear from the evidence presented by the architectural 
 remains and by the artifacts that the three cultures so far con- 
 sidered were of a high order. What brought the last of them, 
 the Tiahuanaco II culture, to a close we can but guess at this 
 distant date. It is clear enough, however, that at the end of 
 the Tiahuanaco II period something happened which checked 
 the development of civilization in both mountains and coast. 
 Vague whispers of the cataclysm persisted in the folklore of the 
 country down to Spanish, and even into our own, times. The 
 early chroniclers report the traditions of the event in various 
 ways, none of which needs to be particularized here. In the 
 nature of the case, the character of the catastrophe must have 
 been gigantic in order to bring about the far-reaching results 
 that it did. Whether it was a terrible earthquake, an invasion 
 of savage peoples or some great epidemic of disease or a com- 
 bination of these things we cannot tell. We only know that in 
 
370 Philip Ainszvorth Means, 
 
 the Titicaca drainage the result was a sudden and very marked 
 lowering of the culture-level, while on the coast and in other 
 regions remote from Lake Titicaca the subsidence in culture, 
 though noticeable, was not so marked. One more thing seems 
 to be disclosed by known facts. As we have seen, Tiahuanaco 
 II art spread far from Tiahuanaco itself. As we shall see, a 
 decadent form of Tiahuanaco II art lingers on around the edges 
 of the old Tiahuanaco "Empire." It is chiefly at Tiahuanaco 
 itself and in the region between Lake Titicaca and Cuzco that 
 the drop in culture is most noticeable. This would seem to 
 indicate that the cataclysm, whatever it was, took place in the 
 mountain regions. The divergence in culture-level that thus 
 sprang up between the mountain regions and the coast resulted 
 in a wide breach between the later arts of the two regions. 29 
 
 The cultures which we are to consider in this section are both 
 coast cultures. The "Epigonal" art is mainly identified with 
 the southern parts of the coast — Pachacamac, Nasca and lea — 
 where the influence of the Tiahuanaco II period had been 
 strongest. Uhle is the scientist to whom the most credit for 
 
 29 The author thinks that it is only fair to warn his reader here that 
 the explanation offered to account for the marked lack of connection 
 between Tiahuanaco II art and Inca art is open to a number of objections. 
 In the first place, if Tiahuanaco II influence did spread into the Cuzco 
 region, it must inevitably have left its stamp upon the art of that region. 
 Archaeology does not permit us to deny that Tiahuanaco II art did spread 
 to Cuzco, — and far beyond it. Why, then, is there so little of Tiahuanaco 
 influence in Cuzco or Inca art? Why is there not at Cuzco, as at 
 Titicaca, Koati and Tiahuanaco, an intermediate type of art which, 
 though much lower in grade than Tiahuanaco II art, still preserves some 
 vestiges of the old tradition? If the forces that brought the Tiahuanaco 
 II art in the Titicaca drainage to an end .were unable completely to 
 obliterate the older style of that region, why were they so much stronger 
 at Cuzco than at Tiahuanaco that they were able to wipe out completely 
 the older art? An answer to these three questions, which were suggested 
 by Dr. Roland B. Dixon, may perhaps be found in the study of the dis- 
 tribution in Peru of the type of culture represented by the Colla-Chulpa 
 type. An examination of this distribution shows that Colla-Chulpa culture, 
 or something very like it, is found throughout the Peruvian highlands 
 from Bolivia to Cuelap in Chachapoyas. It is not like the coast cultures 
 of the time (that is, the period just before the rise of the Incas). Place- 
 names, it is true, have a character remotely suggestive of the coast, but 
 this may have been the result of Inca mitimaes (transferred colonies). 
 
A Survey of Ancient Peruvian Art. 371 
 
 the study of this period should be given, and the reader is urged 
 to examine some of his Plates. 30 When we compare Tiahu- 
 anaco II art with "Epigonal" we at once see wherein the dif- 
 ference lies. The latter is but an unskillful and decadent attempt 
 to continue the traditions of the former. Again and again it 
 is possible to recognize portions of well-remembered Tiahuanaco 
 II motifs on "Epigonal" artifacts, but always the latter are 
 far inferior to their prototype in both line and color. So close 
 is the resemblance sometimes that one would be tempted to 
 say that the "Epigonal" things were indeed made in the Tiahu- 
 anaco II period, but by unskilled artists. This, however, is inter- 
 dicted by the irrefutable stratigraphic evidence. The "Epigonal" 
 wares and textiles occur in later strata than the Tiahuanaco II 
 artifacts. The Weeping God, the puma-heads, the bird-figures 
 and many other Tiahuanaco II motifs occur again in "Epigonal" 
 art. 
 
 Closely associated with the "Epigonal" art is another art- 
 type which, for want of a better name, we call red-white-black 
 ware after the colors in which it is painted. This type is asso- 
 ciated with the coast from Pachacamac northwards to Trujillo 
 (Chimu) and even beyond; it occurs in the same strata with 
 "Epigonal" at Pachacamac, which proves the approximate con- 
 temporaneity of the two. Our Plate XI, Figures 1 and 2, shows 
 two excellent examples of this ware. The originals, in the 
 Peabody Museum, Cambridge, Mass., came from Recuay, north- 
 ern coast region. The colors are red, white and black. In 
 Figure 1 we see a dragon-like figure that is very distinctive of 
 this site. As Joyce points out, 31 it is very similar to a motif 
 found on some Proto-Chimu vases, and a crude derivative of 
 it appears in northwestern Argentina. In the face that adorns 
 the front of the vessel's neck we perceive a very strong tinge 
 of Proto-Chimu tradition. The ear-plugs and headdress are 
 both reminiscent of the analogous portions of the vase shown 
 in Plate I, Figure 2. A great deal less skill in modelling is 
 shown, however. 
 
 To sum up the features of the "Epigonal" and red-white-black 
 arts, we may say that each flourished in the area in which the 
 
 30 Uhle, 1903, Plates V and VI. 
 81 Joyce, 1912, p. 183. 
 
37 2 Philip Ainsworth Means, 
 
 previous culture from which it derived most of its characteristics 
 had flourished. This explains why "Epigonal" art, which differs 
 from Tiahuanaco II only in its imperfection, throve in the region 
 where Tiahuanaco II had been at its best, and why red- white- 
 black art, similar in many respects to Proto-Chimu, existed in 
 the same territory as the latter. On the whole, this period was 
 one of stagnation. At any rate, nothing appears to have been 
 done to advance the development of art in Peru. 
 
 Of what went on in the mountains during this period we know 
 absolutely nothing. Perhaps the shock caused by the putative 
 cataclysm had been so great as to result in a state of affairs 
 almost verging on savagery. There is a possibility that it was 
 at the beginning of this period that the very low-cultured Urus 
 entered the Titicaca basin. They came from the south. 32 
 
 5. CRITICAL ANALYSIS OF THE CHIMU AND NASCA 
 
 CULTURES. 
 
 As we have already noted, we know something definite in 
 regard to the political, social and ethnological aspects of the 
 people of this period. It will be our task in the present section 
 to study their art, and in doing so we shall observe several close 
 similarities between this coast-culture and the Proto-Chimu and 
 Proto-Nasca cultures. We can but hope that the close artistic 
 correspondence between the two is a token of social corre- 
 spondence. 
 
 The distinctive ware of the Chimu period is the black ware 
 that comes from the northern half of the coast and from various 
 regions here and there in the highlands. Though the ware in 
 question has a wide distribution, one may generalize by saying 
 that it is especially distinctive of the northern half of the coast. 
 
 In Plate XI, Figures 3, 4, 5, and 6, are shown four very good 
 specimens of the type. The originals are all in the Peabody 
 Museum, Cambridge, Mass. Besides the fact that the vessels 
 are made of black clay, another new and distinctive feature 
 presents itself. This is the important part played by bas-relief 
 in the decoration of the vessels. In every case, the technique 
 
 32 Chamberlain, 1910, pp. 417 ff. ; Bandelier, 1910, pp. 36 ff. ; Garcilasso, 
 II, pp. 224 ff. ; Balboa, 1840, Chap. XI; Boman, 1908, p. 71 ; Polo, 1901. 
 
A Survey of Ancient Peruvian Art. 373 
 
 is the same square-edged variety that is to be noted in the bas- 
 reliefs on the monolithic gateway. This similarity may or may 
 not be significant. In the case of Figure 5 modelling in the 
 round also plays an important part, and the human head with 
 its large fan-shaped headdress recalls some of the figures of 
 Proto-Chimu art. A further development of this headdress is 
 seen on some of the specimens in the American Museum of 
 Natural History. The development takes the form of the addi- 
 tion of great plume-like ornaments that rise in a curve from the 
 headdress and fall down on both sides of the wearer's face. In 
 decorations with this motif there is to be observed a very marked 
 residue of the old Tiahuanaco II tendency toward bilateral 
 symmetry, and also a number of other criteria typical of that 
 period. For example, one vessel in the New York collection 
 shows a personage with a perfectly symmetrical plumed helmet 
 who is holding two staves or weapons in his symmetrically 
 arranged four-digit hands. Indeed, four-digit hands are by 
 no means uncommon in this period. It was a tendency inherited 
 from the previous periods. A great many vessels however, 
 like Figure 4, show no such symmetry and lack entirely any 
 seeming resemblance to Tiahuanaco II art. Indeed, the anthro- 
 pomorphic figure on that flask seems to be in the clutches of a 
 creature more closely resembling the dragon-like animal we 
 noticed in connection with the red-white-black ware than any- 
 thing else. Again, there is a large class of black ware vessels 
 like Figures 3 and 6 totally devoid of either anthropomorphic or 
 zoomorphic decoration. 
 
 If skill in modelling is one of the strong points of Proto- 
 Chimu art, it is so of Chimu art as well. Evidence of this is 
 given by the large class of "portraits" in both black ware 
 and in red ware. Above all, the modelled vessels representing 
 peanuts, potatoes, guanacates, squash, paltas, and other vegeta- 
 bles and fruits are especially eloquent of the high artistic capa- 
 bilities of the potters in the Chimu period. These vessels are 
 adorned with modelled forms which, except in the matter of 
 color, are absolutely true to nature. 
 
 All this does not mean to imply that the Chimu people used 
 solely this black clay for their vessels. The black is emphasized 
 merely because it is the most predominant and characteristic. 
 Red clay painted in white slip was used, but it lacked the excel- 
 
374 Philip Ainsworth Means, 
 
 leiice and the diversity, as well as some of the distinguishing 
 motifs, of the Proto-Chimu pottery. As we have said, "por- 
 traits" continued to be produced in this period, and we find 
 them in both black ware and red. It is often difficult definitely 
 to assign a "portrait" to one or the other of the two possible 
 periods. 
 
 Still other striking products of this period were the textiles 
 and the stucco wall-decorations derived from them. In Plate 
 XII, Figures I and 2, we see reproductions of textiles of this 
 period. Brighter colored cloths with animal and human figures 
 combined with conventional ones were also fairly common. 
 There is, however, nothing especially new about them, and we 
 would better take up the very remarkable architectural achieve- 
 ments of the period. Only by referring to Rivero and von 
 Tschudi and to Squier can one get a really adequate view of the 
 wonderful city of the Chimu kings. 33 Great walls thirty feet in 
 height and ten feet thick at base by five feet thick near the top 
 are distinguishing features of one type of ruins of the Chimu 
 period. Another type does not have a tapering cross-section. 
 Adobe is the usual material, of course, and it was one which 
 lent itself admirably to the construction of a huge city of dwell- 
 ings, canals, reservoirs, gardens and palaces. The interior sur- 
 faces of some of the walls are decorated with arabesques in 
 stucco which arouse hearty admiration in the beholder. Squier 
 gives numerous pictures of the various specimens of arabesque. 
 We will content ourselves with noting three main classes of 
 arabesque. The simplest type is that of the three specimens 
 shown by Squier (1877, P- 154 Q, as consisting of lozenge- 
 shaped depressions, or square ones, let into the surface of the 
 wall in such a way as to form a lattice or checker-board pattern. 
 In the same class, but a bit more elaborate, is the design which 
 consists of a raised pattern in the form of a double line of 
 stair-sign design. 34 The second type, while still largely geo- 
 metric, is obviously derived directly from textiles of the type 
 shown in Plate XII, Figure 1. The technique, as in the case 
 of the simplest type, is of the square-edged variety. 35 It com- 
 bines, like the textile-type with which it is related, a mingling 
 
 Rivero y von Tschudi, 1851, pp. 268 ff. ; Squier, 1877, pp. 135-192. 
 Cf. Middendorf, 1894-95, H, pp. 375 ff. 
 Squier, 1877, p. 137. 
 
A Survey of Ancient Peruvian Art. 375 
 
 of geometric with zoomorphic elements. The third and final 
 type might be described as curvilinear on account of the 
 predominance of curved lines. In this type zoomorphic and 
 anthropomorphic elements play a very important part. One 
 decoration of this final type seems to be of a simpler nature than 
 one other. It is made up of a series of large hollow squares 
 in stucco relief. Below them are some extraordinary figures 
 resembling conventionalized tapirs. These figures have their 
 "probosces" down and their "legs" to the observer's right with 
 their arched "backs" on the left. There are two of them under 
 each square. 36 One is at a loss to explain this combination of 
 motifs and likewise the motifs themselves. More comprehen- 
 sible is the other specimen of this type. It is distinguished by a 
 very rich composition (still in the square-edged technique) made 
 up of conventionalized men, birds, fishes, crabs, lobsters and other 
 such things. It is plainly the work of a people who were closely 
 in contact with the sea. Two things are very interesting in 
 connection with the human figures, namely, that they wear pre- 
 cisely the same headdress as the figures already described as 
 occurring on the pottery of this period, and that, like those 
 figures, they have less than the real number of digits. The 
 crabs and lobsters in the design are almost life-like. Interwoven 
 with these elements is another one which is like nothing else in 
 Peruvian art. It is a curving device not unlike a W on which 
 are shown some of the animals referred to. The reader is 
 urged to turn to Plate XVI in Joyce (1912) for an adequate 
 presentation of this remarkable design. 
 
 To sum up, then, our impressions of Chimu art, we will say 
 that it bears a general and marked resemblance to the Proto- 
 Chimu, both in the subject-matter and in the treatment. As is 
 only natural, there are accretions from the intervening arts, new 
 motifs and a new tendency to use dark-colored clay for vessels. 
 Likewise, it is not difficult to see in the remarkable wall-decora- 
 tions of this period an attempt to continue the tradition of richly 
 carved ornament found to be so prominent in Tiahuanaco II art. 
 The choice of material — stucco — is easy to explain on the ground 
 that the coast people were already used to stucco as a wall- 
 coating and that suitable stone for the purpose of carving into 
 bas-reliefs was scarce on the coast. 
 
 See Squier, 1877, p. 154. 
 
376 Philip Ainsworth Means, 
 
 Let us now turn to a brief examination of the same period 
 further down the coast, designated by the name of Nasca. The 
 reader is urged to consult Uhle, 1913b. The tradition of rich 
 coloring noticed hitherto in the southern coast-region did not 
 die out with the Tiahuanaco period. As we have already seen, 
 the Epigonal period carried on the forms of Tiahuanaco art to 
 the point where they were on the verge of falling to pieces 
 as the direct result of too-long repetition. The last pre-Inca 
 period of the southern coast exhibits an art which derives its 
 color from both the Proto-Nasca and the Tiahuanaco periods 
 and which still preserves a few of the motifs that mark the 
 latter art. Look, for example, at Plate X, Figure 9 of Uhle, 
 1913b. On the vessel there shown the reader will notice a bird- 
 figure which is considerably like the bird-figures in Tiahuanaco 
 art or in Epigonal. All the other motifs on the vessel, however, 
 are new, and they are distinctive of the period we are now 
 studying. At the same time, the matter of pottery forms is an 
 interesting one. Besides the more usual bowls and dishes, Nasca 
 art shows a new pottery form, namely, the large globular vessel 
 with a fairly wide flaring neck. In most cases, it should be 
 noted, the body of the vessel has a slight tendency to be oval 
 rather than spherical. In the Inca period this tendency becomes 
 emphatic, in the Nasca region, as we shall see. The textiles 
 of this period are practically all adorned with geometric designs. 
 Our old friend the "stair-sign" is a motif that is often found. 
 Color in the textiles becomes duller. 
 
 To sum up the period just before the Inca period on the coast 
 in one sentence we may say that the northern half of the littoral 
 preserved the old tendency toward modelled forms in pottery and 
 toward animal forms in textile-designs, and, at the same time, 
 that the southern half of the coast continued to make many- 
 colored pottery although both the pottery and the textiles show 
 a preponderance of geometric forms over life forms. In both 
 parts of the coast it was essentially a period in which creative 
 forces of the race's imagination were at a low ebb. This may be 
 indicative of the state of affairs in other branches of human 
 activity at that time. The old culture of Tiahuanaco had died 
 away from some shock at the centre and the communities on 
 the coast that had been dependent on it for artistic stimulation 
 fell into a period of stagnation which was only brought to a 
 close by the Inca invasion. 
 
A Survey of Ancient Peruvian Art. 377 
 
 6. A CRITICAL ANALYSIS OF LATE INCA OR CUZCO ART. 
 
 In Inca art we come to the last phase of aboriginal art in 
 Peru, Ecuador, Bolivia, Chile and Argentina. As the type from 
 which all variants of the Inca types were derived was peculiar 
 to Cuzco and its region, we will examine the art of that district 
 before tracing its spread over the wide area it eventually covered. 
 As we have noted before, the collection of pottery and other 
 artifacts made by the various Yale Peruvian Expeditions in the 
 Cuzco region is the most representative collection of Cuzco pot- 
 tery now in this country. The articles by Dr. Bingham show 
 excellently well the nature of the site in which most of these 
 things were found. 37 Important also for our purposes, is the 
 recent publication by Dr. Eaton. 38 The evidence presented by 
 him proves conclusively that most of the burials at Machu 
 Picchu are relatively recent, probably dating not farther back 
 than sixty or eighty years before the Conquest. Since this is 
 so, we must assume that the artifacts from there are also recent. 
 None have been found that are pre-Inca. 
 
 Besides the Yale collection, that in Berlin and that of Dr. 
 Caparo Muhiz at Cuzco are the best two. 39 It will be well to 
 note that the late Inca period which we are now to discuss 
 includes the reigns of the last three unmolested Incas : Pachacutec, 
 Tupac Yupanqui and Huayna Capac. The period began, prob- 
 ably, somewhat after 1400. When Inca Pachacutec assumed 
 the red fringe of sovereignty the Inca dominion already included 
 most of the territory between Chincha and Huanuco on the 
 north and Arica and Tucuman on the south. It was extended 
 by Pachacutec and his successors so as to include all the territory 
 between the northern part of the modern Ecuador and the River 
 Maule in Chile and between the ocean and the montaha or forest- 
 region. In the last days just before the Spanish conquest, when 
 the ill-fated Atahualpa was Inca, Quito, Cajamarca, Cuzco and 
 the island of Titicaca were the chief centres of importance. 
 Cuzco still remained the capital. 40 
 
 "Bingham, 1913, 1915, 1915b, 1916. 
 
 38 Eaton, 1916. 
 
 39 Seler, 1893. 
 
 40 Cf. Means, 1917; Pedro Sancho, 1840. 
 
378 Philip Ainsworth Means, 
 
 We will first discuss the matter of forms in Cuzco (i. e. Inca) 
 pottery before taking up that of decorative motifs. First comes 
 the stately aryballus, at once the most typical Cuzco form and 
 the most universally adopted one wherever Inca power pene- 
 trated. Our Plate XIII shows two good examples of this type. 
 There are several sub-types of aryballi. A tentative classifica- 
 tion is to be offered later. Next in order of frequency of occur- 
 rence come the beaker type, shown in Plate XV, Figure 2, the 
 pelike type, Plate XIV, Figure 1, the bowl, dish and numerous 
 other forms. 
 
 In the matter of decoration we find that the geometric figures 
 are in a large majority over anthropomorphic or zoomorphic 
 ones. At the same time, modelled ornament, save for the uni- 
 versal cat's-head nubbin, is found to be essentially foreign to 
 Inca pottery. It does occur, of course, but it is an extraneous 
 element. (Plate XIV, Figure 3; Plate XV, Figures 3 and 4.) 
 By far the greater part of Inca pottery decorations are made 
 up of combinations of a comparatively small number of motifs. 
 We will describe several of these. One of the most widespread 
 is that seen in Plate XIII, Figure 1. An old Indian at Cuzco 
 told the writer that the design represented a conventionalized 
 quipu or knot-record and that the design was applied particularly 
 to the vessels of the quipucamayoc who looked after the quipus. 
 Without accepting this as an absolute fact, we will call this 
 design the "quipu-motif." Another frequently seen motif is the 
 meander (Plate XIII, Figure 2). A third is the lines-and- 
 cross motif (Plate XIV, Figure 1). A fourth we will call the 
 "diamond motif" (Plate XIV, Figure 3). A fifth might be 
 described as the "saw-tooth motif" (Plate XIV, Figure 3). 
 There are numerous other motifs that might be enumerated if 
 space permitted, but the five named are the commonest and one 
 rarely finds a vessel of Inca type that has not at least one of 
 them upon it. In regard to color the Inca or Cuzco type is 
 rather sombre. Black, dark brown, light brown, red and some 
 white are the usual tints. 
 
 Cuzco types tend to vary but little from the original model. 
 Nevertheless, local variations do occur in several regions, and 
 in the Inca pottery at Cuzco itself marked influences from the 
 arts of subjugated peoples are to be seen. We shall take up 
 
A Survey of Ancient Peruvian Art. 379 
 
 in turn our consideration of these departures from the usual 
 type. 
 
 It may be said that the Inca dominion spread first south then 
 north. The Inca artifacts found in northern sites are, on that 
 account, likely to be more recent than those found in the southern 
 sites. In Argentina and Chile Inca vessels are frequently met 
 with. Boman (1908, I, Plate X) shows two aryballi from Lapaya 
 in north-western Argentina. The shape of the vessels and the 
 arrangement of handles and nubbins are exactly the same as in 
 vessels from Cuzco or Machu Picchu. The pattern on the better 
 of the two pots is divided into two motifs which are the "diamond 
 motif" in two forms, and a debased form of the "saw-tooth 
 motif." Boman's Plates XI and XIV (vol. I) show other 
 Cuzco-type vessels from Lapaya which do not call for special 
 mention. His Plate XVIII (vol. I) shows two aryballi from the 
 Argentine site of Lerma. One shows the "saw-tooth motif" 
 and the "diamond motif." The other combines a perfect Cuzco 
 shape with a well-modelled snake whose head is near the neck 
 of the vessel and slightly raised as if to strike. In general, then, 
 these designs, though obviously derived directly from Cuzco 
 prototypes and totally unrelated to any other Peruvian art, are 
 marred in some cases by a crudeness and uncertainty of execu- 
 tion that may, perhaps, be attributed to a lack of skill on the 
 part of local makers. An examination of Cuzco pottery from 
 Chilean sites reveals a similar situation. Oyarzun (1910, p. 
 363 ff.) shows six Inca or Cuzco aryballi from places in northern 
 Chile. In three cases both shape and decoration are of the best 
 Cuzco style, but in the other three the designs, though derived 
 directly from Inca ones, are crude in point of execution. Turn- 
 ing our attention to Ecuadorian sites we find that the state of 
 affairs is much the same as in the far south of the Inca dominion. 
 Dorsey (1901, Plate XLII) shows a fine Inca aryballus from 
 the island of La Plata in the Bay of Guayaquil. It is exactly 
 of the same shape as the Cuzco of Machu Picchu vessels and 
 it is adorned with the "quipu motif." Bamps (1879, Plates II, 
 III, and IV) shows many Inca vessels from points further north 
 and east in Ecuador. Again, both in "shape and in the execu- 
 tion of the designs, these vessels could not be told apart from simi- 
 lar ones from Cuzco or Machu Picchu. So far as archaeological 
 
 Trans. Conn. Acad., Vol. XXI 26 
 
 1917 
 
380 Philip Ainsworth Means, 
 
 work has thus far shown, the potters of the north were more 
 successful in their attempts to copy the Cuzco style than were 
 those of the south. We should bear in mind, however, the likeli- 
 hood that cruder specimens of vessels of the Inca type have not 
 been reported on. A vessel from Ibarra, Ecuador, is noteworthy 
 in this connection. It is shown by Seler ( 1893, Plate 48, Figure 
 20). It is an aryballus, but the graceful shape of the prototype 
 is not preserved in this copy; the flowing line that, in the Cuzco 
 vessels, merges the neck with the body is here broken by a 
 pronounced shoulder. The decoration, however, combines the 
 "quipu motif" with the "diamond motif." 
 
 It may seem odd at first that the widest divergences from 
 the Cuzco standard do not occur in the regions furthest from 
 Cuzco. Pachacamac and lea are the two sites which show the 
 most strongly localized arts. The reader is urged to consult 
 Uhle's publications on this point. 41 In the period that preceded 
 the Inca period at Pachacamac, as we have seen, the people made 
 a great number of black clay vessels with one-handled globular 
 bodies and necks adorned with rather coarsely modelled human 
 faces. The combination of this art with Inca vessels of the 
 aryballus type resulted at Pachacamac in giving two handles to 
 the vessels and in adding paint to the modelled face. We should 
 not fail to note that in many cases where the hands appear in 
 the Inca vessels they have five fingers. This emphasizes the 
 breaking away from the old Tiahuanaco tradition. At lea, as 
 we have previously observed, large vessels of a slightly oval 
 shape were made in the last pre-Inca period. These develop 
 into a definitely egg-shaped or cask-shaped type decorated some- 
 times with Inca motifs and sometimes with Nasca motifs. 
 
 This brings us to the consideration of the other type of varia- 
 tion from the Cuzco standard. It is the class of variation which 
 consists in a manifestation of the influence of local pre-Inca arts 
 on the Cuzco type. The reader has just seen the effect that Inca 
 art had upon the modelled black ware of the coast. He is now 
 asked to turn his attention to its corollary, the type which shows 
 the influence of the black ware of the coast upon the Cuzco 
 types. In Plate XV, Figure 4, we have an excellent specimen of 
 this class. Though both come from Machu Picchu, Figure 3 may 
 
 41 Uhle, 1903, Plates VIII and XIII, 1913b, Plate X. 
 
A Survey of Ancient Peruvian Art.. 381 
 
 well be the coast-form which served as a model for the other. 
 In both examples there is but one handle, and the general shape 
 is the same in both. Figure 4, however, is definitely associated 
 with Inca art by the "quipu motif" on its body. 
 
 We will now draw up a classification of Inca pottery on a 
 basis of form and decoration : 
 
 Type I Large open-necked vessels (often painted with geo- 
 metric designs). 
 
 a Deep bowls without handles (Seler, 1893, 
 
 Plate I), 
 b Various types with handles gradually ap- 
 proaching aryballus. 
 Type II Aryballus type. Narrow neck, two handles and 
 nubbin. 
 
 a With geometric designs only. 
 
 b With painted designs both geometric and 
 
 animal, 
 c With modelled anthropomorphic element and 
 
 painted design in combination, 
 d Miscellaneous sub-types. 
 
 Type III Plates, bowls, braziers, cups, etc. 
 
 a With geometric designs. 
 
 b With animal or human designs. 
 
 c With both. 
 Type IV Miscellaneous beakers, bottles and pots. 
 
 We must now turn our attention to the question of Inca or 
 Cuzco textiles. To the modern eye they appear the most beau- 
 tiful of all Peruvian textiles. As we shall see, however, they 
 are not technically so wonderful as the Proto-Nasca embroideries. 
 Plates XVI and XVII show four typical Inca textiles. A glance 
 will show the reader that those on Plate XVI are of a very dif- 
 ferent type from the other two. They come from the island of 
 Titicaca, and the originals are in the American Museum of 
 Natural History, New York. In Plate XVI it is seen that the 
 decorative tendency is to break the surface up into small patches 
 of color. This same tendency may be remarked on Inca pottery 
 from the same site. The number of decorative motifs is too 
 great to dwell upon at length ; we shall have to content ourselves 
 with noting that the motif which consists in a slanting band 
 ended off by two squares each containing a dot, which squares 
 are repeated on each side of the band, occurs on an Inca cloth 
 
382 Philip Ainsworth Means, 
 
 from lea. 42 Variants of the "saw-tooth motif" and of the 
 "diamond motif" are present in each of these ponchos, recalling 
 the Inca pottery. Both also show the frequent use of rectilinear 
 spirals. Plate XVII, Figure 1, is also a poncho from the island 
 of Titicaca. It is in several shades of red and has a white 
 cruciform figure much like that on the cask-shaped vessel from 
 lea shown by Uhle (1913b, Plate X, A). Figure 2 comes from 
 the coast and shows a slight influence, in the form of cat-like 
 figures, from the Chimu period. 
 
 We must now summarize our impressions of Inca art. We 
 may do so by saying that geometric decoration has a great pre- 
 ponderance over animal or human motifs. While Inca pottery 
 derives most of its charm from its graceful form, it is by no 
 means to be despised because it has not a great range of color. 
 The designs are usually simple but pleasing, and in most cases 
 they are peculiar to Inca art. In the textiles the same tendency 
 toward geometric designs is to be noted, although here again 
 other elements do occasionally play a part. In general, the color 
 of the textiles is brighter and more various than that of the 
 pottery. 
 
 Uhle, 1913b, p. 344. 
 
IV. THE QUESTION OF CHRONOLOGY AND DATES 
 IN EARLY PERUVIAN ART 
 
 As was said at the beginning of the paper, the writer, after 
 surveying the development of art in ancient Peru, wishes to 
 present a date-chronology of the various cultures. The dates 
 here to be presented are only approximate. In the nature of 
 things, we must be prepared to allow for an error of a century 
 or more in the remoter epochs. 
 
 It is necessary that a word should be said as to the methods 
 employed in drawing up the chronology. In the total lack of 
 all written records of any sort we have to meet a great obstacle. 
 This is partly overcome by certain things which we will speak 
 of soon. Moreover, tradition, which sometimes does much to 
 aid in the establishment of an approximate chronology, is here 
 limited almost wholly to the Inca period. These are the chief 
 disadvantages to be met with. We will now examine the 
 conditions which are more favorable to our end. 
 
 In trying to construct a date-chronology for the various higher 
 cultures of the Andean region, one must bear in mind that it 
 is inherently improbable that the cultures of South America 
 possess an antiquity greater than those of Middle America. The 
 researches of Dr. Hrdlicka have clearly shown this improbability. 
 He has shown four very important truths: (i) Man is zoolog- 
 ically a newcomer in this hemisphere; (2) Man, when he arrived 
 on this continent, was in a stage of culture "superior to that of 
 the late Pleistocene" ; (3) Man, since arriving in this hemisphere, 
 has inevitably undergone certain secondary modifications as to 
 physical type and culture ; (4) There exists to-day in north- 
 eastern Asia a racial element that is descended from the same 
 ancestors as the American Indians. 1 
 
 Since, from the point of view of the zoologist, Man is an Old 
 World animal that reached America by way of Siberia and the 
 Aleutian Islands, it must be assumed that the northern parts of 
 the continent were peopled sooner than the southern parts. This 
 supposition applies to any tribes, no matter what their cultural 
 grade may be. Nor is mythology lacking in indications of the 
 
 1 Cf. Hrdlicka, 1912, 1912b, 1912c, 1912c!. 
 
384 Philip Ainsworth Means, 
 
 southwardly shift of the high-cultured people of the west coast 
 of South America. In the face of all this, then, the onus pro- 
 bandi rests upon him who would maintain that the South Ameri- 
 can populations are older than the North American or Middle 
 American. 2 
 
 Let us, then, assume for the purposes of the present discussion 
 that Man entered America from the north and slowly spread 
 southward. The primary migrations of Man in America have 
 a southward trend. His secondary migrations often do not. 
 In the Middle American region (Mexico, Yucatan, Guatemala, 
 Honduras, Salvador, Nicaragua, Costa Rica and Panama) we 
 have a number of very high cultures. Those of Mexico and 
 Yucatan are, in many respects, as high or higher than those we 
 have been studying. Up to about 752 A. D. all is vague and 
 uncertain as to cultural events in Mexico. In or about that year, 
 however, the Toltecs founded Tula. 3 More important for us 
 is the cultural type described by Tozzer as "archaic." It is 
 much older than the Toltec culture and much more widespread. 
 Indeed, we may say that the archaic type occurs scatteringly 
 from the valley of Mexico down to Panama. 4 It will perhaps 
 be proved to be the ancestor of most of the later high cultures 
 of Middle and South America. At any rate, the meager seven 
 centuries from the founding of Tula to the Spanish conquest 
 is obviously not long enough to account for the development and 
 wide distribution of the calendar-system and the various related 
 dialects in Middle America. We must assume that the people 
 of the archaic period flourished long before the time when the 
 earliest high cultures of Middle America began to develop their 
 own peculiarities, peculiarities which, however, never succeeded 
 in blotting out the fact that all the cultures had a common origin. 5 
 
 2 This is not the place to go into the question of geologically ancient man 
 in America. Those .who wish to do so are urged to read Hrdlicka, 1912, 
 and the numerous works listed in the Bibliography of that publication. 
 All that it is necessary to say here is that Hrdlicka has shown the extreme 
 unlikelihood of the existence of any of the morphologically primitive types 
 of men in America. 
 
 3 Tozzer, 1916, p. 464. 
 
 4 Tozzer, 1916, p. 466; Spinden, 1915; see Appendix for discussion of 
 "archaic type." 
 
 5 Means, 1917. 
 
A Survey of Ancient Peruvian Art. 385 
 
 In Yucatan we can fairly carry the beginning of protohistory 
 back many centuries. This is largely due to the work of Mr. 
 Bowditch and to that of Mr. Morley. 6 As the present writer 
 has explained elsewhere, the difference between the chronologies 
 of these two authors is neither serious nor great. The earliest 
 dated Maya remains are, respectively, the Tuxtla statuette and 
 the Leyden plate. The former bears the Maya date 8.6.2.4.17 
 (about 100 B. C.) ; the date on the latter is 8.14.3. 1. 12 (about 
 40 A. D.). 7 In spite of the fact that these inscriptions are so 
 early, the system in which the dates are set down is absolutely 
 the same as that in which those of the "Old Empire" cities 
 in southern Yucatan are written. The significance of this is, 
 of course, that even so early as 100 B. C. the Mayas had gone 
 through the centuries-long process of evolving their calendar 
 system. We must postulate, in Mr. Morley's opinion, at least a 
 thousand years of preparatory development. 8 This period of 
 development should be understood to include the migrations of 
 the various branches of the original stock to the place in which 
 they are found in later eras. From about the time of Christ 
 to the end of the seventh century the "Old Empire" of the 
 Mayas was running its course. From then to the middle of 
 the fifteenth century the "Transitional Period" and "New 
 Empire" rose and fell. 
 
 We will now summarize the chronological conditions known to 
 be true of Middle America. For at least eleven centuries before 
 Christ various migrations (mainly southward) were accom- 
 panied by the steady development of individual cultures, all 
 variants of a common origin, albeit influenced by environmental 
 and psychological conditions. By the time of Christ, the high 
 cultures of Middle America had almost crystallized into their 
 final forms. 
 
 6 Bowditch, 1901 ; Morley, 1910, 1915; Means, 1917b, p. 3. 
 
 'While on his most recent expedition for the Carnegie Institution of 
 Washington, Mr. Morley discovered an important site in northern Guate- 
 mala. He gave it the name of Uaxactun — Eight-stone — because he found 
 there a large stela bearing the Maya date, 8.14.10.13.15, equal to about 
 50 A. D. Another inscription at that site may possibly be eighth cycle, 
 also. We have, consequently, at least three inscriptions dating from 50 
 A. D. or earlier. (Information given by Mr. Morley to the writer.) 
 
 s Morley, 1915, p. 194-196; Holmes, 1916. 
 
386 Philip Ainsworth Means, 
 
 In South America, what do we find? We find a series of cul- 
 tures following one another in logical succession. We find that 
 the earliest are the most like the Middle American cultures. 
 We find, besides, two independent criteria which enable us to 
 build up an approximate chronology. Each will be described in 
 turn. 
 
 The list of "kings" of Peru given by Fernando Montesinos 
 on the authority of Bias Valera has only lately begun to receive 
 the attention it merits. 9 While it emphatically cannot be accepted 
 as real history, it is, nevertheless, important as indicating that 
 popular legend in the time of the Incas preserved the memory 
 of many generations of rulers. Counting the Incas, the "kings" 
 on the list number 102. Markham, an accomplished historian in 
 other fields as well as in the Peruvian, considers that 27 years is 
 a fair average for the length of a reign. Accepting this in its 
 totality for the nonce, we find that the list of rulers is thought 
 by Montesinos to cover a period of 2,754 years, or, in other 
 words, that the first ruler flourished about 1224 B. C. (1530 A. D. 
 minus 2,754). This date, then, is the very earliest that even 
 Montesinos is willing to accept. Everyone will agree that this 
 date is hardly tenable. As Markham says (1912, p. 41), we 
 must allow for repetitions, overlappings and other errors. Let 
 us, then, be conservative and consider that there were but seventy 
 reigns. This gives us about 1,900 years as the period covered 
 by the list, and it puts the earliest ruler about 350 B. C. Sir 
 Clements Markham (loc. cit.) prefers the initial date 200 B. C. 
 We may say, then, that in all probability, the earliest "king" of 
 Tiahuanaco I (it was of the mountain races that Montesinos 
 wrote) flourished about 200 B. C. Probably, however, culture 
 was low and local for many generations. We find that the 
 "first dynasty" of Montesinos is frequently marked by the name 
 Pirua. It consists of eighteen rulers. Let us call it fifteen; 
 15 X 2j = 405 years; or, in other words, the Pirua "dynasty" 
 came to a close about 200 A. D. Was not this perhaps the end 
 of the Tiahuanaco I period? The next "dynasty" is marked by 
 the name Amauta in many cases. Montesinos gives it forty-five 
 rulers. Let us call it thirty ; 30 X 27 = 810 ; this brings us 
 
 9 Montesinos, 1840, 1882; Markham, 1912, p. 303 ff. 
 
A Survey of Ancient Peruvian Art. 387 
 
 up to about 1000 A. D. This date, however, does not fit well with 
 known historical facts. Let us, then, say that the Amauta 
 "dynasty" (perhaps of Tiahuanaco II) flourished from about 
 200-900 A. D. Montesinos calls the dark period that followed the 
 Amauta "dynasty" the "Tampu Tocco period." In it we may 
 see our Colla-Chulpa period. He gives it twenty-seven rulers. 
 Let us call it ten ; 10 X 27 = 270 years ; or, to put it differently, 
 the dark period began to draw to a close about 11 70 A. D. This 
 brings us to the threshold of the Inca period. The late Dr. 
 Gonzalez de la Rosa constructed a date-chronology of the Inca 
 period which seems to the writer wholly acceptable. A modified 
 version of it is given here. 10 
 
 Reigns of the Incas, According to 
 Dr. Gonzalez de la Rosa. 
 
 Sinchi Rocca 1 134- 1 1 97 
 
 Lloque Yupanqui 1 197-1246 
 
 Mayta Capac 1246-1276 
 
 Capac Yupanqui 1276-1321 
 
 Inca Rocca 1321-1348 
 
 Yahuar Huaccac 1348-1370 
 
 Viracocha 1370-1425 
 
 Pachacutec 1425-1478 
 
 Tupac Yupanqui 1478-1488 
 
 Huayna Capac 1488-1525 
 
 It may be more satisfactory to some to reduce the thing to 
 round numbers, thus : Viracocha, 1370-1420; Pachacutec, 1420- 
 1480; Tupac Yupanqui, 1480-1490; Huayna Capac, 1490-1525. 
 Either step will result in a fairly accurate basis on which to fix 
 one's idea of the reign-periods. 
 
 So much, then, for one of our two criteria. It has been noted 
 that this one concerns the mountain region primarily. The other 
 is important for the coast cultures. It is unfortunate that it 
 has not yet been fully studied. 
 
 The islands off the coast of Peru have long been famous for 
 their deposits of guano. These lie in masses of enormous thick- 
 ness. Markham says that two and one-half feet a century is 
 approximately the rate of accumulation. The rate no doubt 
 fluctuated slightly, but the careful investigations made by Mark- 
 
 Gonzalez de la Rosa, 1909; Means, 1917, p. 244. 
 
3 88 
 
 Philip Ainsworth Means, 
 
 ham have led him to accept the above rate as a fair average. 
 According to Gonzalez de la Rosa, antiquities occur in the guano 
 at depths varying from nine feet to forty or more. 11 This means 
 that in 1870 (at which date the investigations were made) the 
 antiquities presumably varied in age from about four centuries 
 (i. e. 9 feet gives a date of about 1450) to about sixteen cen- 
 turies (i. e. 40 feet gives a date roughly equal to 200 A. D.). 
 Perhaps future work will yield more detailed information as to 
 which cultures are found at various depths in the guano. At 
 all events, it seems possible that for want of a better criterion 
 we must bear the evidence of the guano deposits in mind. 
 
 It is now well for us to summarize and tabulate the general 
 results of the evidence brought out by the foregoing discussion. 
 Once again the reader is asked to remember that the dates here 
 offered claim to be no more than roughly approximate guides to 
 the imagination. 
 
 An Approximate Chronology of the Early Cultures 
 
 of Peru. 12 
 
 Mountain Regions 
 
 Primary Migrations 
 Tiahuanaco I 
 
 Coast Regions 
 Primary 
 Migrations and 
 Proto-Chimu and 
 Proto-Nasca 
 
 Dates 
 ? -200 A. D. 
 
 Tiahuanaco II 
 
 Coast Tiahuanaco 
 II, followed by 
 "Epigonal" and 
 red-white-black 
 
 200-900 
 
 wares 
 
 Colla-Chulpa period 
 (called "Tampu Tocco" 
 by Montesinos) 
 
 Continuance of 
 above styles 
 
 900-1 100 
 
 Early Inca 
 
 Chimu and Nasca 
 
 1 100-1400 
 
 Late Inca dominion approaching its zenith 
 
 1400-1530 
 
 11 Gonzalez de la Rosa, 1908. 
 
 12 The reader is particularly reminded that there is much evidence to 
 show that Proto-Chimu, Proto-Nasca and Tiahuanaco I all contributed 
 
A Survey of Ancient Peruvian Art. 389 
 
 This brings us to the end of our subject. When, in 1531, the 
 Spanish conquest of Peru began, the Inca dominion — Ttahua- 
 ntin-suyu — was being torn to pieces by a civil war between the 
 legitimate ruler, Huascar and the usurper Atahualpa. Subse- 
 quent evolution in Peruvian Art lies beyond the scope of the 
 present work. 
 
 towards the formation of Tiahuanaco II. Moreover, as Tiahuanaco II 
 art grew older it became more and more complex, spreading, at the same 
 time, into regions very far away from Tiahuanaco itself. The fact that 
 the specimens of Tiahuanaco II art from the more distant regions often 
 show the admixture of elements taken over directly and bodily from 
 Proto-Chimu and Proto-Nasca art, shows that, even when Tiahuanaco II 
 was approaching its end, the two early coast arts were still vigorous. 
 The dates on the above table, therefore, should not be regarded as the 
 terminal dates of the culture periods, but as the approximate dates at 
 ■which each was at its strongest development. 
 
APPENDIX I : THE ARCHAIC TYPE. 
 
 Dr. Herbert J. Spinden kindly wrote at the writer's request 
 this summary of his views as to the significance of the "archaic 
 type." 
 
 "An archaic culture allied to that of Mexico and Central 
 America seems once to have spread across Colombia and Ecuador 
 to the coast of Peru. In Peru the culture has not been isolated 
 in pure form — if we may use this chemical phrase in archae- 
 ology — unless it should prove to be that which Uhle briefly 
 describes from the earliest shell-heap remains at Ancon. He 
 figures several heads that resemble very closely those of the 
 lowermost horizon in Mexico and he finds associated with them 
 pottery characterized by incised and plastic decoration. 1 It need 
 hardly be pointed out that the pottery of the Archaic horizon 
 in the north is also characterized by plastic decoration and that 
 when incised or painted decorations occur the designs are 
 exceedingly simple. Highly "conventional" designs based upon 
 an animal motive are not found in the truly archaic, but are 
 characteristic of the second crop of cultures after religion and 
 ceremony had developed to the point that it could react strongly 
 upon art. 
 
 "But in the absence of other data we may be permitted to rest 
 our theory upon the presence in the coastal region of Peru of 
 figurines presumably related to those of the Archaic horizon 
 although found among the products of a later time. At Ancon, 
 and at other sites as well, are found nude female figurines with 
 the short stubby arms that are so characteristic of the products 
 of the Archaic horizon from Mexico to Colombia. These 
 figurines are usually moulded rather than modeled and it seems 
 unlikely that moulds came into use until the upper archaic or 
 even later. The standing pose is more common than the sitting 
 one. In the American Museum collections there are perhaps 
 twenty-five examples of these figurines, and others are reproduced 
 by Putnam. 2 
 
 1 Uhle, 1912, pp. 22-45. 
 
 2 Putnam, 1914, Plate XIX. 
 
A Survey of Ancient Peruvian Art. 391 
 
 "In addition to female figurines there are many examples of 
 pottery vessels from Ancon, Trujillo, etc., in which a human 
 figure is represented in a fashion that harks back to the archaic, 
 namely with the elbows and knees both flexed and the former 
 directly over the latter. Of course, in the cases of both the 
 figurines and the vessels the qualities peculiar to Peruvian art 
 had already become set. 
 
 "The theoretical considerations that connect the spread of 
 archaic ceramic art with the spread of agriculture are very 
 strong. No one can get away from the fact that maize, beans and 
 squashes constitute four species (Zea mays, Phaseolus vulgaris, 
 Cucurbita maxima, and C. pepo) wherever agriculture is found in 
 America. The Lima bean (Phaseolus lunatus) had a more 
 restricted use. 
 
 "It seems not unreasonable to suppose that careful research will 
 bring to light more evidence on the occurrence of figurines of 
 early type. These objects may have been neglected in favor of 
 those of greater artistic interest. For instance, Dorsey, in 
 discussing the finds on the Island of La Plata, says : — 
 
 " 'Practically all this pottery was in fragments, only two pieces 
 were found in perfect condition. With the exception of not 
 more than a dozen pieces, all the fragments were parts of small 
 images in the form of human figures. . . . From fragments 
 representing perhaps a thousand images not more than half a 
 dozen pieces were found which bore any trace of paint. . . . 
 All the pottery, with a very few exceptions, is hand made; that 
 is, it was not made in a mould, which was commonly employed 
 on the mainland of Ecuador and throughout a large extent of 
 Peru.' 3 
 
 "Many of the fragments figured by Dorsey are distinctly 
 archaic in treatment. Of course it might be argued that the 
 archaism is absolute rather than relative but a comparison of 
 special features gives ample evidence of transitions from one 
 region to the next." 
 
 To these remarks by Dr. Spinden the writer would like to add 
 a few of his own. 
 
 As has been said, the "archaic type" is stratigraphically the 
 earliest in Middle America. Therefore, if it does occur in South 
 
 3 Dorsey, G. A., 1901, pp. 266-267. 
 
39 2 Philip Ainsworth Means, 
 
 America it must be expected to be the earliest there also. If 
 one is to believe that the "archaic type" was a very early type 
 which spread all over the northern half of Latin America, must 
 not one also assume that the various later cultures were devel- 
 oped from it in the several regions involved? Such a develop- 
 ment would occur after religion and ceremony had gained 
 considerably in strength, as Dr. Spinden says. On the other 
 hand, if the "archaic type" is looked upon as a cultural land- 
 mark rather than as a culture in itself, the finding of it in the 
 wide area mentioned does not prove much. In other words, if 
 we are to believe that all art at some time or other passes through 
 a stage wherein it shows "archaic type" characters, the mere 
 fact that art with archaic characters is found in both Middle 
 America and South America does not mean much. The writer, 
 however, finds that the former interpretation is the better. There 
 can be but little doubt as to the absolute priority in point of time 
 of the archaic culture of the Peruvian shell-heaps. The work of 
 Uhle has shown that in Peru, as in Middle America, the earliest 
 culture of all was the archaic type, and we now know that this 
 type was uniform throughout Middle America and on the Peru- 
 vian coast. It is the foundation whereon all other cultures were 
 built. 
 
APPENDIX II. 
 
 A TABLE TO SHOW ROUGHLY THE CHRONOLOGICAL ORDER 
 OF THE EARLY PERUVIAN CULTURE PERIODS. 
 
 Names. 
 Proto-Chimu. 
 
 Areas. 
 The coast from Tum- 
 bez to Ancon. 
 
 Remarks. 
 Characterized by realism 
 and light tonality. 
 
 Proto-Nasca. The coast from Pacha- Distinguished by conven- 
 
 camac to Arica. tionalization and rich color- 
 
 ing. 
 
 Tiahuanaco I. 
 
 In the mountains, from 
 Samaipata to Cuzco, and 
 especially about Lake 
 Titicaca. 
 
 A culture rich in architec- 
 tural remains. Endowed with 
 a stone technique. Not like 
 P-C or P-N, possibly Ara- 
 wakan. 
 
 Tiahuanaco II. 
 
 "Epigonal" and 
 Red-white-black. 
 
 In mountains and on Probably a complex of the 
 coast, from Colombia to three foregoing cultures. 
 Argentina and Chile. 
 
 On the Peruvian coast. Decadent forms of Tiahu- 
 anaco II culture. 
 
 Chimu and Nasca. On the Peruvian coast. 
 
 Revival of some of the 
 features of Proto-Chimu and 
 Proto-Nasca. 
 
 Colla-Chulpa. 
 
 Around Lake Titicaca. 
 
 Low culture with faint 
 traces of Tiahuanaco II in- 
 fluence. 
 
 Early Inca. 
 
 Cuzco region. 
 
 Beginning of new period 
 in the mountains. 
 
 Inca "Empire." 
 
 From Ancasmayo in The last pre-Columbian 
 Ecuador to Maule in culture. Graceful forms, re- 
 Chile, strained coloring. 
 
 Separates contemporaneous cultures. 
 
 — — — Separates partly contemporaneous cultures. 
 
 Separates non-contemporaneous cultures. 
 
BIBLIOGRAPHY OF WORKS REFERRED TO IN 
 
 THIS PAPER. 
 
 ABBREVIATIONS 
 AA American Anthropologist. 
 
 AASP American Antiquarian Society Proceedings. 
 AJA American Journal of Archaeology. 
 
 AJS American Journal of Science. 
 
 AMJ American Museum Journal. 
 
 APAMNH Anthropological Papers American Museum of Natural History. 
 
 BAE Bureau of American Ethnology. 
 
 BGA Berliner Gesellschaft fiir Anthropologic 
 
 BSGL Boletin de la Sociedad Geografica de Lima. 
 
 BSGLP Boletin de la Sociedad Geografica de La Paz. 
 
 CAAS Connecticut Academy of Arts and Sciences. 
 
 CIA Congres internationale des americanistes. (See also, ICA.) 
 
 CIAAP Congres internationale d'anthropolgie et d'archeologie pre- 
 
 historiques. 
 
 FCMP Field Columbian Museum Publications. 
 
 FFLSA Facultad de Filosofia y Letras, Seccion Antropologica (Buenos 
 
 Aires). 
 
 HS Hakluyt Society. 
 
 ICA International Congress of Americanists. (See also, CIA.) 
 
 JAP Journal de la Societe des Americanists de Paris. 
 
 JRGS Journal of the Royal Geographical Society. 
 
 MFAB Museum of Fine Arts Bulletin (Boston). 
 
 NAMS Nouvelles Archives des Missions Scientifiques. 
 
 NGM National Geographic Magazine. 
 
 PAAAS Proceedings of the American Academy of Arts and Sciences. 
 
 RBAE Report of the Bureau of American Ethnology. 
 
 RH Revista Historica (Lima). 
 
 SMP Smithsonian Miscellaneous Publications. 
 
 TCCC Trabajo del Cuarto Congreso Cientifico. 
 
 ZE Zeitschrift fiir Ethnologic 
 
 Allen, Harrison. 
 
 1881. An Analysis of the Life Form in Art. Transactions 
 of the American Philosophical Society, XV (n. s.) 
 PP- 271-351. 
 Arriaga, Pablo Jose de. 
 
 1621. Extirpacion de la idolatria en el Peru. Lima. 
 1910. Extirpacion de la idolatria en el Peru. Facsimile of 
 1st Ed. Buenos Ayres. 
 
A Survey of Ancient Peruvian Art. 395 
 
 Baessler, Arthur. 
 
 1902-03. Ancient Peruvian Art. Berlin and New York, 3 
 vols. 
 Balboa, Miguel Cabello. 
 
 1840. Histoire du Perou. Ternaux-Compans. Paris. 
 Bamps, Anatole. 
 
 1879. Les antiquites equatoriennes du Musee Royale d'Anti- 
 quites de Bruxelles. CIA, Third Session, vol. II, 
 p. 47 and Atlas. 
 Bandelier, Adolph Francis. 
 
 1905. The Aboriginal Ruins of Sillustani, Peru. AA, (n. s.) 
 
 VII, pp. 49-68. 
 1907. The Indians and Aboriginal Ruins near Chachapoyas, 
 Northern Peru. New York. 
 
 1910. The Islands of Titicaca and Koati. New York. 
 
 191 1. The Ruins at Tiahuanaco. AASP, XXI, pp. 218-265. 
 Barnett, Mme. Anna. 
 
 1909. Etude technologique d'un tissu peruvien. . . JAP, 
 
 VI, pp. 265-268. 
 Batchelder, E. A. 
 
 1910. Design in Theory and Practice. New York. 
 von Bayern, Therese, S. K. H. Prinzessin. 
 
 1907. Einige Worter uber die Kulturentwicklung im Vor- 
 
 spanischen Peru. Mimchener Jahrbuch d. Bildenden 
 Kunst., 1907, pp. 1-7. 
 Berthon, Paul. 
 
 191 1. Etude sur le precolombien du Bas-Perou. NAMS, 
 
 Fascicule 4. 
 Beuchat, Henri. 
 
 1912. Manuel d'archeologie americaine. Paris. 
 Bingham, Hiram. 
 
 1913. In the Wonderland of Peru. NGM, April, 1913. 
 
 1915. The Story of Machu Picchu. NGM, February, 1915. 
 1915b. Types of Machu Picchu Pottery. AA, (n. s.) XVII, 
 
 pp. 257 ff. 
 
 1916. Further Explorations in the Land of the Incas. NGM, 
 
 May, 1916. 
 Boman, Eric. 
 
 1908. Antiquites de la region andine. . . . Paris, 2 vols. 
 Trans. Conn. Acad., Vol. XXI 27 1917 
 
396 Philip Ainsworth Means, 
 
 Bowditch, Charles P. 
 
 1901. Memoranda on the Maya Calendars used in the Books 
 of Chilan Balam. AA, (n. s.) III. 
 Chamberlain, Alexander F. 
 
 1910. The Uran, a new South American Linguistic Stock. 
 
 AA, (n. s.) XII, pp. 4I7-424- 
 1910b. Sur quelques families linguistiques . . . de l'Ame- 
 rique du sud. JAP, VII, pp. 179-202. 
 
 191 1. On the Puelchean and .... Atacamerian .... 
 
 Linguistic Stocks. AA, (n. s.) XIII, pp. 458-500. 
 
 1913. Linguistic Stocks of South American Indians. AA, 
 (n. s.) XV, pp. 236-247. 
 
 1913b. Principal Tribes and Subtribes of the Arawakan Lin- 
 guistic Stock of South America. JAP, X, pp. 
 
 473-496. 
 Cieza de Leon, Pedro de. 
 
 1864. Travels. Edited by Clements R. Markham. HS, 1864. 
 1883. Second Part of the Chronicle of Peru. Edited by 
 Clements R. Markham, HS, 1883. 
 Cobo, Bernabe. 
 
 1892. Historia del Nuevo Mundo. Edited by Marcos Jimenez 
 de la Espada, Seville, 1892, 4 vols. 
 Debenedetti, Salvador. 
 
 1912. Influencias de la cultura de Tiahuanaco. FFLSA, 1912, 
 
 p. 27 ff. 
 1912b. Influencias de la cultura de Tiahuanaco en ... . 
 el noroeste argentine CIA, Eighteenth Session, I, 
 pp. 298-300. 
 Dorsey, George A. 
 
 1 901. Archaeological Investigations on the Island of La Plata, 
 Ecuador. FCMP, No. 56. 
 Eaton, George F. 
 
 1916. The Collection of Osteological Material from Machu 
 Picchu. CAAS, Mem. V. 
 Enock, C. Reginald. 
 
 1907. The Andes and the Amazon. London. 
 Falies, Louis. 
 
 1915. Etudes sur les civilizations americaines. Paris. 
 Flint, Sarah G. 
 
 1916. Peruvian Textiles. MFAB, October, 1916. 
 
A Survey of Ancient Peruvian Art. 397 
 
 Garcia Rosell, Ricardo. 
 
 1903. Departamento de Piura. BSGL, XIII, pp. 193-224. 
 Garcilasso de la Vega, el Ynca. 
 
 1869, 1871. Royal Commentaries of the Yncas. Edited by 
 Clements R. Markham. HS, 2 vols. 
 Gonzalez de la Rosa, Manuel. 
 
 1908. Estudio de las antigiiedades halladas bajo el huano. 
 
 RH, III, pp. 39-45- 
 1908c. Les deux Tiahuanaco. . . . ICA, Sixteenth Session, 
 pp. 405-428. 
 
 1909. Ensayos de cronologia incana. RH, IV, trimestres I, 2. 
 Haddon, A. C. 
 
 1912. Wanderings of Peoples. New York. 
 Holmes, W. H. 
 
 1885. Ancient Art of the Province of Chiriqui. RBAE, 
 
 1884-85, pp. 3-187. 
 1885b. A Study of the Textile Art. RBAE, 1884-85, pp. 
 
 189-252. 
 1887. The Use of Gold and Other Metals among the Ancient 
 
 Inhabitants of Chiriqui. BAE, 1887. 
 1889. Textile Fabrics of Ancient Peru. BAE, 1889. 
 1916. The Oldest Dated American Monument. Art and 
 
 Archaeology, III, pp. 275-278. 
 1916b. A Ouirigua Mystery. Art and Archaeology, IV, p. 341. 
 Hrdlicka, Ales. 
 
 191 1. Some Results of Recent Anthropological Exploration 
 
 of Peru. SMP, Vol. 56, No. 16. 
 
 1912. Early Man in America. AJS, XXXIV, pp. 543-554. 
 1912b. Restes dans l'Asie orientale de la race qui a peuple 
 
 l'Amerique. CIAAP, Fourteenth Session, pp. 
 
 409-414. 
 1912c. Early Man in South America. BAE, Bulletin 52. 
 191 2d. Derivation and Probable Place of Origin of the North 
 
 American Indian. ICA, Eighteenth Session, pp. 
 
 58-62. 
 1914. Anthropological Work in Peru in 1913. SMP, Vol. 
 
 - 61, No. 18. 
 1916c. Review of Eaton, 1916. AJS, September, 1916, p. 281. 
 Jacquemart, Albert. 
 
 1873. History of Ceramic Art. London. 
 
398 Philip Ainsworth Means, 
 
 Joyce, T. A. 
 
 1912. South American Archaeology. New York. 
 
 1913. Note on a Gold Beaker from Lambayeque, Peru. Man, 
 
 XIII, No. 37. 
 1913b. The Clan-ancestor in Animal Form as depicted on the 
 Ancient Pottery of the Peruvian Coast. Man, XIII, 
 No. 65. 
 
 1914. Mexican Archaeology. New York. 
 
 1916. Central American and West Indian Archaeology. New 
 York. 
 MacCurdy, George Grant. 
 
 1906. The Armadillo in the Ancient Art of Chiriqui. ICA, 
 Fifteenth Session, II, pp. 147-164. 
 
 191 1. A Study of Chiriquian Antiquities. CAAS, Mem. III. 
 1913. Note on the Archaeology of Chiriqui. AA, XV, pp. 
 
 661-667. 
 Markham, Sir Clements Robert. 
 
 1871. On the Geographical Positions of the Tribes which 
 formed the Empire of the Yncas. JRGS, XLI, pp. 
 281-338. 
 
 1904. The Megalithic Age in Peru. ICA, Fourteenth Ses- 
 sion, II, pp. 521-529- 
 
 1908. A Comparison of the Ancient Peruvian Carvings on 
 
 the Stones of Tiahuanaco and Chavin. ICA, Six- 
 teenth Session, pp. 389-395. 
 
 1912. The Incas of Peru. New York (Second Edition). 
 Mead, Charles W. 
 
 1906. The Six-unit design in Ancient Peruvian Cloth. Boas 
 Memorial Volume, pp. 193-195, New York. 
 
 1909. The Fish in Ancient Peruvian Art. Putnam Anni- 
 
 versary Volume, pp. 127-136. 
 
 1916. Ancient Peruvian Cloths. AMJ, XVI, pp. 389-393. 
 1916b. Conventionalized Figures in Ancient Peruvian Art. 
 
 APAMNH, XII. 
 Means, Philip Ainsworth. 
 
 1917. Culture Sequence in the Andean Area. ICA, Nine- 
 
 teenth Session, pp. 236-252, Washington. 
 1917b. History of the Spanish Conquest of Yucatan and of 
 the ItzaS. Peabody Museum Papers, vol. VII. 
 Cambridge, Mass. 
 
A Survey of Ancient Peruvian Art. 399 
 
 MlDDENDORF, E. W. 
 
 1894-95. Peru Berlin, 3 vols. 
 
 Montesinos, Fernando. 
 
 1840. Memoires historiques sur l'aneien Perou. Edited by 
 
 Ternaux-Compans, Paris. 
 1882. Historia antigua del Peru. Edited by Marcos Jimenez 
 
 de la Espada, Madrid. 
 
 MORLEY, SYLVANUS GrISWOLD. 
 
 1910. Correlation of Maya and Christian Chronology. AJA, 
 
 XIV, pp. 193-204. 
 
 19 1 5. An Introduction to the Study of Maya Hieroglyphs. 
 BAE, Bulletin 57. 
 
 NORDENSKJOLD, BARON ErLAND. 
 
 1902. Pracolumbische Salzgewinnung in Puna de Jujuy. 
 BGA, 1902, pp. 336-341. 
 
 1906. Ethnologischen und archaeologischen Forschungen im 
 Grenzgebiet zwischen Peru und Bolivia. ZE, 
 XXXVIII, pp. 80-99. 
 
 1906b. Arkeologiska Urdersokningar in Perus och Bolivias. 
 Stockholm. 
 Oyarzun, Aureliano. 
 
 1910. Contribution al estudio de la influencia de la civiliza- 
 tion peruana sobre los aborigenes de Chile. ICA, 
 Seventeenth Session, pp. 354-397. 
 Polo Jose Toribio. 
 
 1899. La Piedra de Chavin. BSGL, IX, p. 192 ff. and 262 ff. 
 
 1 901. Indios Uros del Peru y Bolivia. BSGL, X, pp. 445-482. 
 Posnansky, Arthur. 
 
 1910. Tiahuanacu y las razas y monumentos prehistoricos. 
 
 ICA, Seventeenth Session, pp. 267-68. 
 
 191 1. Tiahuanacu y la civilization prehistorica. . . BSGLP, 
 
 IX, pp. 1-53. 
 1911b. Razas y monumentos prehistoricos del altiplano andino. 
 TCCC, XI, pp. 2-142. 
 
 1913. El signo escalonado. . . . Berlin. 
 1913b. Una falsa critica de Max Uhle. Berlin. 
 
 1913c. Die Altertiimer von Tiahuanacu. ZE, XLV, pp. 178- 
 186. 
 
 1914. Una metropoli prehistorica en la America del sud. 
 
 Berlin. 
 
400 Philip Ainsworth Means, 
 
 Putnam, Edward K. 
 
 1914. The Davenport Collection of Nazca and other Peruvian 
 
 Pottery. PDAS, XIII, pp. 17-40. 
 
 OUIROGA, ADAN. 
 
 1 901. La Cruz en America. Buenos Aires. 
 Read, Sir C. Hercules. 
 
 1904. On two Pottery Vessels from the Upper Amazon. 
 Man, IV, No. 32. 
 Reiss, Wilhelm, and Stubel, Alphons. 
 
 1880-87. The Necropolis of Ancon. Berlin. 
 Rivero y Ustariz, Mariano Eduardo de, and von Tschudi, 
 Johann Jakob. 
 
 1851. Antigiiedades Peruanas. Vienna, 2 vols. 
 Rivet, P. ; See Verneau and Rivet. 
 Ross, Denman Waldo. 
 
 1901. Design as a Science. PAAS, XXXVI, pp. 357-374. 
 
 1907. A Theory of Pure Design. Boston. 
 Sancho, Pedro. 
 
 1849. Relation de la conquista del Peru. Mexico. 
 Sarmiento de Gamboa, Pedro. 
 
 1907. History of the Incas. Edited by Sir Clements R. 
 Markham. 
 Sartiges. 
 
 185 1. Voyage . . . dans l'Amerique du Sud. Revue des 
 Deux Mondes. 185 1. 
 Saville, Marshall H. 
 
 1907-10. Antiquities of Manabi. New York, 2 vols. 
 Seler, Eduard. 
 
 1893. Peruanische Altherthiimer. Berlin. 
 Simoens da Silva, A. C. 
 
 1912. Points of Contact of the Prehistoric Civilizations of 
 
 Brazil and Argentina with those of the Pacific Coast 
 Countries. ICA, Eighteenth Session, pp. 302-310. 
 Spinden, Herbert J. 
 
 1913. A Study of Maya Art. Peabody Museum Memoirs, 
 
 VI. Cambridge, Mass. 
 
 1915. Notes on the Archeology of Salvador. AA, (n. s.) 
 
 XVII, pp. 446-491. 
 1916b. Portraiture in Central American Art. Holmes Memo- 
 rial Volume, pp. 434-450. 
 
A Survey of Ancient Peruvian Art. 4 CI 
 
 Squier, E. George. 
 
 1877. Peru .... the Land of the Incas. New York. 
 Stubel, Alphons, and Uhle, Max. 
 
 1892. Die Ruinenstatte von Tiahuanaco. Breslau. 
 Torres Rubio, Diego de. 
 
 1616. Arte de la lengua Aymara. Lima. 
 von Tschudi, Johann Jakob. 
 
 1868. Reisen durch Sudamerika. Leipzig, 5 vols. 
 Tozzer, Alfred M. 
 
 1916. The Domain of the Aztecs. . . . Holmes Memorial 
 Volume, pp. 464-468. 
 Uhle, Max. 
 
 1901. La antigua civilizacion peruana. BSGL, X, p. 93 ff. 
 
 1902. Types of Culture in Peru. AA, (n. s.) IV, pp. 753-759. 
 
 1903. Pachacamac. Philadelphia. 
 
 1908. Uber die Fruhkulturen in der Umgebung von Lima. 
 
 ICA, Sixteenth Session, pp. 347-370. 
 
 1909. Peruvian Throwing-sticks. AA, (n. s.) XI, pp. 624- 
 
 627. 
 
 1910. Tipos de civilizacion en el Peru. BSGL, XXV, p. 
 
 289 ff. 
 
 1910b. Las relaciones prehistoricas entre el Peru y la Argen- 
 tina. ICA, vSeventeenth Session, pp. 509-540. 
 
 1910c. Los origenes de los Incas. ICA, Seventeenth Session, 
 
 PP- 302-353- 
 
 1912. Die Muschelhiigel von Ancon. ICA, Eighteenth Ses- 
 
 sion, pp. 22-45. 
 
 1913. Die Ruinen von Moche. JAP, X, pp. 95-117. 
 
 1913b. Zur Chronologie der alten Culturen von lea. JAP, X, 
 
 PP- 34I-367- 
 
 1914. The Nazca Pottery of Ancient Peru. PDAS, XIII. 
 
 Verneau, R., and Rivet, P. 
 
 1912. Ethnographie ancienne de l'Equateur. Paris. 
 
 VOCABULARIO PoLIGLOTA InCAICO. 
 
 1905. Lima. 
 Wiener, Charles. 
 
 1880. Perou et Bolivie. Paris. 
 Young, Jennie J. 
 
 1879 The Ceramic Art. New York. 
 
LIST OF ILLUSTRATIONS. 
 
 PLATE I. 
 
 Fig. i. A Proto-Chimu vessel showing two hunting-scenes, one 
 modelled and one painted in dark brown on a cream- 
 colored slip. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 2. A Proto-Chimu vessel with a modelled scene represent- 
 ing some ceremony, and an engraved landscape. 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 3. A somewhat grotesque Proto-Chimu vessel of the 
 portrait type. Note the fangs and the feather-like 
 headdress. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 4. A portrait, probably belonging to the Proto-Chimu 
 culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 5. A dipper-shaped Proto-Chimu vessel adorned by a 
 conventionalized starfish or octopus. 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 PLATE II. 
 
 Fig. 1. A Proto-Nasca vessel of the semi-realistic type. Note 
 realistic hands and the modelling of the head. 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 2. A Proto-Nasca vessel with some realism, especially in 
 the spears and spear-thrower. Note the eye-painting, 
 and the type of dress. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
404 Philip Ainsworth Means, 
 
 Fig. 3. A Proto-Nasca bowl showing Centipede-God motif. 
 Note four-digit hands. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 4. A Proto-Nasca vessel with two narrow spouts. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 5. A Proto-Nasca vessel decorated with the Multiple- 
 headed God motif. Note the four digit-hands. 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 6. A semi-realistic modelled vessel, of the Proto-Nasca 
 culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 PLATE III. 
 
 Fig. 1. A Proto-Nasca vessel. Note the headdress, the mouth- 
 mask, the ribbon-and-ring decorations, and the conven- 
 tionalized spear-thrower. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 2. A Proto-Nasca vessel with a variant of the Centipede 
 God motif. Note the four-digit hands and the pro- 
 truding tongue. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 3. A Proto-Nasca vessel. Note the feather-like mouth- 
 mask, the ribbon-and-ring decorations, and the bilateral 
 symmetry of the design. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 PLATE IV. 
 
 Fig. 1. An embroidered Proto-Nasca textile. Note the head- 
 dress, mouth-mask, and the four digit-hands. 
 
List of Illustrations. 4°5 
 
 Courtesy of the Museum of Fine Arts, Boston, Massa- 
 chusetts. 
 
 Fig. 2. A Proto-Xasca vessel decorated with a variant of the 
 Centipede God motif and with a painted human face. 
 Xote the four-digit hands, the protruding tongue and the 
 "tears." 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 3. A Proto-Nasca textile. Note the headdress, the mouth- 
 mask, the four-digit hand, and the elaborately decorated 
 protruding tongue. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 PLATE V. 
 
 A richly embroidered Proto-Nasca textile. The design 
 combines elements from the Centipede God motif with 
 elements from the Multiple-headed God motif. 
 Courtesy of Dr. Denman Waldo Ross and of the 
 Museum of Fine Arts, Boston, Massachusetts. 
 
 PLATE VI. 
 
 A large and beautiful embroidered Proto-Nasca gar- 
 ment. Made entirely of wool. 
 
 Courtesy of Dr. Denman Waldo Ross and of the 
 Museum of Fine Arts, Boston, Massachusetts. 
 
 PLATE VII. 
 
 The Weeping God on the monolithic gateway at Tia- 
 huanaco. 
 
 Photograph by courtesy of the Peabody Museum, Cam- 
 bridge, Massachusetts. 
 
 PLATE VIII. 
 
 Fig. 1. A finely woven garment, probably of the Tiahuanaco II 
 culture. 
 
406 Philip Ainsworth Means, 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 2. A garment from the coast of Peru, probably of the 
 Tiahuanaco II culture. 
 
 Courtesy of Dr. Denman Waldo Ross and of the 
 Museum of Fine Arts, Boston, Massachusetts. 
 
 PLATE IX. 
 
 The Greater Chavin Stone. 
 
 Photograph by courtesy of the Peabody Museum, Cam- 
 bridge, Massachusetts. 
 
 PLATE X. 
 
 Two textiles, either late Tiahuanaco II or Epigonal. 
 Note the results of long-continued conventionalization 
 and elaboration. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 PLATE XI. 
 
 Fig. i. A Red-white-black ware vase. Note the headdress, the 
 modelled face and the painted animal-figure. 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 2. A Red-white-black ware vessel in the form of a cat-like 
 animal. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 3. A Black-ware vessel, Chimu culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 4. A Black-ware flask, Chimu culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 Fig. 5. A Black-ware vessel, Chimu culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
List of Illustrations. 4°7 
 
 Fig. 6. A Black-ware flask, Chimu culture. 
 
 Courtesy of the Peabody Museum, Cambridge, Massa- 
 chusetts. 
 
 PLATE XII. 
 
 Fig. i. A Chimu textile. Note the combination of geometric 
 decoration with greatly conventionalized animal-heads 
 or bird-heads. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 Fig. 2. A very fine pouch, Chimu culture. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 PLATE XIII. 
 
 Fig. i. An Inca or Cuzco aryballus of exquisite shape, deco- 
 rated with quipu motif. 
 
 Yale Collection; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig. 2. An Inca or Cuzco aryballus, decorated with rectilinear 
 meander. 
 
 Yale Collection; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 PLATE XIV. 
 
 Fig. i. An Inca or Cuzco pelike, decorated with lines-and-cross 
 motif. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig. 2. A deep bowl with handles, Inca or Cuzco type. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig. 3. A shallow dish, decorated with saw-tooth motif and 
 with diamond-motif. Note the handle in the form of a 
 human head. Inca or Cuzco type with influence from 
 the art of the coast. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
408 Philip Ainsworth Means. 
 
 Fig. 4. A shallow dish, Inca or Cuzco type. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 PLATE XV. 
 
 Fig. 1. An Inca or Cuzco aryballus. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig. 2. An Inca or Cuzco vessel. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig 3. A vessel with anthropomorphic decorations. Although 
 the specimen was found at Machu Picchu, it may have 
 been carried there from the coast. It is of the Chimu 
 type, not of the Inca or Cuzco type. 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 Fig. 4. A vessel combining Chimu art with Inca or Cuzco art. 
 Compare with Figure 3. 
 
 Yale Collection ; courtesy of the Connecticut Academy 
 of Arts and Sciences. 
 
 PLATE XVI. 
 
 Two Inca or Cuzco type ponchos. Very rich in color, 
 and beautifully woven. 
 
 Courtesy of the American Museum of Natural History, 
 New York City. 
 
 PLATE XVII. 
 
 Two Inca or Cuzco type textiles. 
 
 Courtesy of the American Museum of Natural History, 
 
 New York City. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE I. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE II. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE III. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE IV. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE V. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE VI. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE VII. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE VIII. 
 
Trans. Conn. Aead. Vol. XXI. 
 
 PLATE IX. 
 
Trans. Conn. Aead. Vol. XXI. 
 
 PLATE X. 
 
Trans. Conn. Aead. Vol. XXI. 
 
 PLATE XI. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XII. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XIII. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XIV. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XV. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XVI. 
 
Trans. Conn. Acad. Vol. XXI. 
 
 PLATE XVII. 
 
 
 * + + <. 
 
 
 * * 
 
 • -6- -0- ■*■ H 
 * * ■* « * 
 
 * + «■ -0- * . -0- 
 
 * *■* * * 
 
 Hfr * * * H 
 -!> <h <h <> Hfr 
 + 4> ■!>• 4> -■O- -0- 
 
 •*5*- : 
 
GETTY CENTER LIBRARY 
 
 3 3125 00035 9907 
 
For particulars regarding the previous Transactions of 
 the Connecticut Academy, 
 
 Address the Librarian, 
 
 Andrew Keogh, Yale Station, 
 
 New Haven, Conn., U. S. A. 
 
 CONTENTS OF VOLUME 19 OF THE TRANSACTIONS 
 
 PAGES ^ PRICE 
 
 1-1 10 Fossil Birds in the Marsh Collection of Yale Uni- 
 versity, by R. W. Shufeldt (Feb. 1915) $1.10 
 
 1 1 1-446 Middlemen in English Business, 1660-1760, by 
 
 R. B. Westerfield (Feb. 1915) 3.40 
 
 CONTENTS OF VOLUME 20 OF THE TRANSACTIONS 
 
 PAGES PRICE 
 
 1-131 The Materials for the History of Dor, by George 
 
 Dahl (May, 1915) $1.40 
 
 133-144 New Spiders from New England, XI, by J. H. 
 
 Emerton 15 
 
 145-160 Canadian Spiders, II, by J. H. Emerton 15 
 
 161-240 The Historical Background of Chaucer's, Knight, 
 
 by Professor Albert S. Cook (Feb. 1916) .... 1.00 
 
 241-399 Rural Economy in New England at the Begin- 
 ning of the Nineteenth Century, by Percy W. 
 N Bidwell (April, 1916) 2.00 
 
 CONTENTS OF VOLUME 21 OF THE TRANSACTIONS 
 
 PAGES PRICE 
 
 1-144 The Last Months of Chaucer's Earliest Patron, 
 
 by Professor Albert S. Cook (Dec. 1916) .... $1.60 
 
 145-200 The Relationship of the Tetracoralla to the Hexa- 
 
 coralla, by W. I. Robinson (Feb. 1917) 40 
 
 201-313 The American Species of Marchantia, by Alex- 
 ander W. Evans (March, 1917) 1.20 
 
 315-442 A Survey of Ancient Peruvian Art, by Philip 
 
 Ainsworth Means (May, 1917) 1.50