THE UNIVERSITY OF ILLINOIS LIBRARY FA *\kJ ANGOLA MAP SHOWING ROUTE OF EXPEDITION Scale: 1 inch = 229 miles; indicates route Field Museum of Natural History Founded by Marshall Field, 1893 Publication 329 Anthropological Series Vol. XXI, No. 2 THE OVIMBUNDU OF ANGOLA BY Wilfrid D. Hambly ASSISTANT CURATOR OP AFRICAN ETHNOLOGY Frederick H. Rawson-Field Museum Ethnological Expedition To West Africa, 1929-30 84 Plates in Photogravure and 1 Map Berthold Laufer CURATOR, DEPARTMENT OF ANTHROPOLOGY EDITOR ***** N'V£RS'TY or CHICAGO, U. S. A. 1934 PRINTED IN THE UNITED STATES OF AMERICA BY FIELD MUSEUM PRESS 5' FA Z CONTENTS MM List of Illustrations 93 Preface 103 I. Introduction 105 II. Geographical Factors 108 III. Historical Sources 112 IV. Physical Appearance 128 V. Economic Life 133 Nature Lore 134 Food Supply 140 Collecting and Hunting 140 Fishing 145 Agriculture and Cooking 146 Domestic Animals 152 Trade and Transport 156 Industries 158 Iron-work 158 Wood-carving 161 Domestic Implements 165 Pottery 167 Mats and Baskets 169 Weapons 172 Leather Work 176 Spinning and Weaving 177 VI. Social Life 179 Sexual Relations 179 Courtship 179 Marriage 180 Divorce 181 Pregnancy and Childbirth 183 Naming 188 Terms of Relationship 189 Law and Government 199 Warfare and Slavery 204 Village Organization 206 VII. Education 212 Industrial Training and Division of Labor 212 Standards of Conduct, Manners, and Salutations .... 213 89 90 Contents PAGE Educational Value of Play, Music, and Dancing 216 Initiation 226 VIII. Language 234 Affinities of Umbundu 234 Vocabulary 236 Phonetics 237 Vowels 237 Consonants 238 Tone and Stress 239 The Syllable 240 Grammar 240 The Class System 240 Pronouns 244 Principal Tenses 245 Transcription of Folklore Stories 248 Sign Language 252 Riddles and Proverbs 253 Folklore Stories 255 IX. Religion 262 Supreme Being 262 Survival after Death 262 Religious Beliefs and Conduct 264 Funeral Rites 265 Commoners 265 Medicine-men 270 Kings and Chiefs 271 Hunters 272 Training of Medicine-men 273 Functions of Medicine-men 273 Divination 274 Equipment and Miscellaneous Duties 276 Curing the Sick 278 Rain-making 282 Poison Ordeal 283 Ceremonial Fire 283 Prohibitions and Omens 285 X. Culture Contacts 286 Congo Basin 286 Rhodesia 296 South West Africa 303 Contents 91 PAGB XL Wider Culture Contacts 312 Antiquity of Cultural Traits 312 The Blacksmith's Craft in Africa 313 Bantu Religion and Social System 314 African Puberty Rites 316 Hunting Appliances of Africa 317 African Pottery, Baskets, and Musical Instruments. . 319 Kulturkreis Theory 320 XII. Cultural Processes 327 Analysis of African Cultures 327 Assembling of Traits 331 Cultural Losses 334 Integration of Traits 337 Bibliography 349 Index 356 LIST OF ILLUSTRATIONS Map (facing title-page). Angola. IX. Basket-work Patterns, Ovimbundu, Elende. Cat. Nos. 208944-50, actual size. Fig. 1. Big star (olombun- gululu vinene). Fig. 2. Trunk of epangue tree (ocisila cepangu). Fig. 3. Wave pattern (apuku atito). Fig. 4. Kerchief pattern (cdesu). Fig. 5. Arrow (usongo). Fig. 6. A reclining object (onjandu). Fig. 7. Following one another (imbagu). Fig. 8. Hoe handle (ovipi viatimo). Fig. 9. Spots (atumba). X. Basket-work Patterns, Ovimbundu, Elende. Cat. Nos. 208087, 208879-82, actual size. Fig. 1. Wave pattern (apuku atito). Fig. 2. Pattern of kings (olosoma). Fig. 3. Ribbed pattern (olomati). Fig. 4. Double arrow (usonge wayombo). Fig. 5. Star (olombungululu) . XL Incised and Burned Designs on a Gourd, Vachokue, Kuchi. Cat. No. 208032, dimensions 23 x 23 cm, patterns half actual size. XII. Incised and Burned Decorations on Gourds. Fig. 1. From the chief village of the Vangangella, Ngalangi. Cat. No. 208007, dimensions 26x4 cm. Fig. 2. Ovimbundu, Elende. Cat. No. 208012, dimensions 21 x 8.5 cm. Fig. 3. Ovimbundu, Elende. Cat. No. 208010, dimensions 15 x 10 cm. Fig. 4. Ovimbundu, Elende. Cat. No. 208004, dimensions 31 x 20 cm. Fig. 5. Ovimbundu, Elende. Cat. No. 208013, dimensions 22 x 9.5 cm. Fig. 6. Gourd used to contain beer for a bride and groom, Ovimbundu, Bailundu. Cat. No. 208001, dimensions 39 x 26 cm. XIII. Wooden Combs, Tools, Domestic Implements, and Weapons. Fig. 1. Knife used for cutting branches, Vasele. Cat. No. 208427, dimensions 50 x 5 cm. Fig. 2. Combined ax and adze, Ovimbundu, Elende. Cat. No. 208432, length 73 cm. Fig. 3. Wooden pounder for maize, Ovimbundu, Elende. Cat. No. 208404, length 31 cm. Fig. 4. Small ax used by medicine-man when dancing to drive away evil spirits, Ovimbundu, Bailundu. Cat. No. 208415, 93 94 List of Illustrations length 32.5 cm. Fig. 5. Ceremonial ax, formerly used for beheading slaves at the death of a king, Vasele, near Vila Nova de Selles. Cat. No. 208263, length 42 cm. Fig. 6. Hunter's pouch for ammuni- tion, Ovimbundu, Elende. Cat. No. 209059, dimen- sions 15 x 12 cm. Fig. 7. Three wooden hair-combs, Ovimbundu, Elende, and Vachokue, Cangamba. Cat. Nos. 208462, 208459, 208453, one-fourth actual size. Fig. 8. Iron spear, Ovimbundu, Bailundu. Cat. No. 208244, length 37 cm. Fig. 9. Hoe, Ovim- bundu, Elende. Cat. No. 208429, length 51 cm. Fig. 10. Hoe used by Vangangella, Ngalangi. Cat. No. 208439, length 85 cm. XIV. Pottery and Wooden Utensils. Fig. 1. Earthenware water-jar, Ovimbundu, Elende. Cat. No. 208208, dimensions 42 x 14 cm. Fig. 2. Small cooking pot, Vasele, near Vila Nova de Selles. Cat. No. 208201, dimensions 12 x 10 cm. Fig. 3. Earthenware water- bottle, Vasele, Vila Nova de Selles. Cat. No. 209007, dimensions 24x23 cm. Fig. 4. Earthenware food bowl, Ovimbundu, Elende. Cat. No. 208223, dimen- sions 20 x 9.5 cm. Fig. 5. Cooking pot, Ovimbundu, Elende. Cat. No. 208202, dimensions 16.5 x 13 cm. Fig. 6. Milk vessel, Vakuanyama, Mongua. Cat. No. 208152, dimensions 28 x 21 cm. Fig. 7. Wooden spoon, Vachokue, Kuchi. Cat. No. 208143, dimen- sions 34x9 cm. Fig. 8. Wooden milk jug, Vakuanyama, Mongua. Cat. No. 208243, dimensions 21 x 13 cm. Fig. 9. Wooden spoon and stirrer, Vachokue, Cangamba. Cat. No. 208164, dimensions 60.5 x 9.5 cm. XV. Tobacco-pipes, Snuff Box, Rat Trap, and Basket. Fig. 1. Tobacco-pipe, Vachokue, Ngalangi. Cat. No. 208715, length 74 cm. Fig. 2. Tobacco-pipe, Ovimbundu, Elende. Cat. No. 208709, length 35 cm. Fig. 3. Water-pipe for smoking tobacco and hemp, Vachokue, Cangamba. Cat. No. 208688, length 29 cm. Fig. 4. Carved wooden snuff box, Ovimbundu, Elende. Cat. No. 208708, length 23 cm. Fig. 5. Wicker rat trap, Ovimbundu, Elende. Cat. No. 208039, length 33 cm. Fig. 6. Coiled basket with List of Illustrations 95 inwoven patterns of amber and black, Ovimbundu, Bailundu. Cat. No. 208926, dimensions 37 x 18 cm. XVI. Tools and Weapons. Fig. 1. Spearhead, Vachokue, Munyangi. Cat. No. 208372, dimensions 30.5 x 3.3 cm. Fig. 2. Knife, Vachokue, Saurimo in Lunda. Cat. No. 208375, dimensions 21.5 cm. Fig. 3. Black- smith's hammer, Vangangella, Ngalangi. Cat. No. 208826, length 18 cm. Fig. 4. Saw of native make, Ovimbundu, Elende. Cat. No. 208362, length 71 cm. Fig. 5. Iron-bladed tool for boring wood, Ovimbundu, Elende. Cat. No. 208368, length 57 cm. Fig. 6. Blacksmith's tongs, Ovimbundu, Elende. Cat. No. 208813, length 52 cm. Fig. 7. Blacksmith's hammer, Ovimbundu, Elende. Cat. No. 208823, length 33 cm. Fig. 8. Axhead, Vachokue, Cangamba. Cat. No. 208817, length 26 cm. Fig. 9. Blacksmith's punch, Ovimbundu, Elende. Cat. No. 208816, length 24 cm. Fig. 10. Knife in wooden scabbard, Vakuanyama, Mongua. Cat. No. 208357, length 54 cm. Fig. 11. Brass bracelet made by blacksmith, Elende. Cat. No. 208503, width 5 cm. Fig. 12. Bellows of black- smith, Ngalangi. Cat. No. 208831, dimensions 55 x 34 cm. XVII. Bows and Arrowheads. Fig. 1. The bow and arrow, with details of stringing and feathering, are typical of implements of Ovimbundu and Vachokue tribes. Cat. No. 208663, length 143 cm. The arrowheads were collected from Vachokue hunters from Cangamba to Saurimo in Lunda. Cat. Nos. 208643- 652, 654, 656, length average 80 cm. Fig. 2. Arrow- heads, Ovimbundu, Elende. Cat. Nos. 208623, 615, 618, 616, length of each about 80 cm. Fig. 3. Wooden arrow for shooting birds, Ovimbundu, Elende. Cat. No. 208633, length 88 cm. Figs. 4-6. Socketed iron arrowheads, Vakuanyama, Mongua. Cat. Nos. 208598, 592, 601, lengths 67, 58, 66 cm. Fig. 7. Barbed wooden arrow, Vakipungo, southwest Angola. Cat. No. 208607, length 83 cm. Fig. 8. Flat bow and arrow, Vakuanyama, Mongua. Cat. No. 208659, length 122 cm. Fig. 9. Arrow for shoot- ing lizards, Vasele. Cat. No. 208657, length 87 cm. 96 List of Illustrations XVIII. Hut Furniture and Mask. Fig. 1. Wooden seat with hide top, Ovimbundu, Ngalangi. Cat. No. 208180, dimensions 53 x 28 x 30 cm. Fig. 2. Wooden stool, Ovimbundu, Elende. Cat. No. 208864, dimensions 29x21 cm. Fig. 3. Wooden chair with hide seat, Vachokue, Bailundu. Cat. No. 209006, height 80 cm. Fig. 4. Wooden mask, type used by Vangangella and Ovimbundu, Ngongo, Ngalangi. Cat. No. 208100, dimensions 20 x 18 cm. XIX. Wood-carving, Ovimbundu. Fig. 1. A hawk, Bailundu. Cat. No. 208386, dimensions 37x16 cm. Fig. 2. Crane, Bailundu. Cat. No. 208381, dimensions 27 x 10 cm. Fig. 3. Guinea fowl, burned black and speckled with white paint, Elende. Cat. No. 208382, dimensions 20 x 8 cm. Fig. 4. Lizard, Elende. Cat. No. 208394, dimensions 41x6 cm. Fig. 5. Snake, Elende. Cat. No. 208391, length 24 cm. XX. Staffs of Village Chiefs. Fig. 1. Ovimbundu, Elende. Cat. No. 208778, length 68 cm. Fig. 2. Vachokue, Cangamba. Cat. No. 208775, length 82 cm. Fig. 3. Ovimbundu, Elende. Cat. No. 208757, length 35 cm. Fig. 4. Vachokue, Kuchi. Cat. No. 208754, length 98 cm. Fig. 5. Ovimbundu, Elende. Cat. No. 208772, length 90 cm. Fig. 6. Staff of dead king from hut in which relics of kings are kept, from capital village of Ngalangi. Cat. No. 208748, length 123 cm. Fig. 7. Club of heavy black wood, Ovim- bundu, Elende, workmanship of Vachokue and Lundatype. Cat. No. 208777, length 71 cm. Fig. 8. Staff of office in form of paddle, carried by headman of Lioko, a village of Ngalangi. Cat. No. 208745, dimensions 127 x 16 cm. Fig. 9. Heavy throwing- club for killing small game, Ovimbundu, Elende. Cat. No. 208767, dimensions 40 x 12 cm. Fig. 10. Ornamental club, Ovimbundu, Elende. Cat. No. 208764, length 56 cm. XXI. Carved Human Figures. Fig. 1. Wooden figure nursed in place of dead twin, Ovimbundu and Vangangella, Ngalangi. Cat. No. 208345, length 20 cm. Fig. 2. Wooden figure from divination basket of medicine- man, Ovimbundu, Elende. Cat. No. 208307, length List of Illustrations 97 7 cm. Fig. 3. Wooden figure placed by anvil of blacksmith. Spirit in figure assists blacksmith, Ovimbundu, Elende. Cat. No. 208339, length 36 cm. Fig. 4. Female wooden figure used by Vachokue medicine-man, Cangamba. Body hollow to contain medicine, head detachable. Cat. No. 208355, length 61 cm. Fig. 5. Female wooden figure used by medicine-man for divining correct path on caravan journey, Ovimbundu, Bailundu. Cat. No. 208346, length 27 cm. Fig. 6. Wooden figure used like No. 1, Ovimbundu, Elende. Cat. No. 208326, length 18 cm. XXII. Musical Instruments. Fig. 1. Dumb-bell basket rattle, Vachokue, Cangamba. Cat. No. 208732, length 26 cm. Fig. 2. Wooden flute, Ovimbundu, Elende. Cat. No. 208723, length 29 cm. Fig. 3. Instrument played by rubbing grooves with stick, Ovimbundu, Elende. Cat. No. 208805, dimensions 43 x 21 cm. Fig. 4. Rattles for ankles, made from seed pods, Ovimbundu, Bailundu. Cat. No. 208730, length 21 cm. Fig. 5. Instrument, metal keys on wooden board, Ovimbundu, Elende. Cat. No. 208734, dimensions 18 x 11 cm. Fig. 6. Musical bow, Ovim- bundu, Elende. Cat. No. 208722, length 109 cm. Fig. 7. Strip of rattan. May be bent to form musical bow, Ovimbundu, Elende. Cat. No. 208810, length 62 cm. Fig. 8. Frictional instrument, played like No. 3, Ovimbundu, Bailundu. XXIII. Drawings Representing Life of Ovimbundu, Elende. Made by untrained Ocimbundu youth, Elende, who had lived with Europeans. Fig. 1. Ovimbundu hunters. Fig. 2. Medicine-man of Ovimbundu divining. Fig. 3. Ocimbundu woman making beer. Fig. 4. Relative of deceased interrogating corpse as to cause of death. Chief mourners in group on left. XXIV. Examples of Scarification. Fig. 1. Ocimbundu woman, Ngalangi. Fig. 2. Vachokue woman, Cubango. Fig. 3. Esele woman. Cuts, made to cure pain, were rubbed with charred gourd. Fig. 4. Ocimbundu woman, Elende. Scar made at age of ten years by male operator, charred rubber placed in cuts. Fig. 5. Ocim- bundu woman, Cuma. Scars ornamental and tribal. 98 List of Illustrations XXV. Positions of Hands in Drumming, Ovimbundu, Elende. Front View. XXVI. Positions of Hands in Drumming, Ovimbundu, Elende. Side View. XXVII. Ovimbundu Drummers, Elende. Fig. 1. Long drums. Fig. 2. Flat drum. XXVIII. Ovimbundu, Elende. Fig. 1. Man carrying gourd. Fig. 2. Woman with field produce. XXIX. Agriculture near Ganda. Fig. 1. Tobacco plants on ant hill. Fig. 2. Clearing the bush. XXX. Transportation at Elende. Fig. 1. Portuguese riding an ox. Fig. 2. Ocimbundu carrier. XXXI. Transportation, Ovimbundu, Elende. Fig. 1. Boy carrying pig. Fig. 2. Men bearing sick man in hammock. XXXII. Ritual for Hunters. Fig. 1. Tomb near Luimbale. Fig. 2. Trophies near Elende. XXXIII. Ovimbundu Making Pottery, Elende. Fig. 1. Building pottery vessel. Fig. 2. Molding pot with hands and gourd. XXXIV. Ovimbundu Making Pottery, Elende. Fig. 1. Moisten- ing and smoothing wet pot. Fig. 2. Finished wet pots with incised designs. XXXV. Ovimbundu. Fig. 1. Man using drill, Bailundu. Fig. 2. Girl, Elende. XXXVI. Occupations of Ovimbundu, Elende. Fig. 1. Male cotton spinner. Fig. 2. Woman pounding maize. XXXVII. Blacksmiths at Work, Ovimbundu, Elende. Fig. 1. Working bellows. Fig. 2. Forging axhead. XXXVIII. Blacksmiths at Work, Ovimbundu, Elende. Fig. 1. Pounding on anvil. Fig. 2. Using cutting tool. XXXIX. Ovimbundu, Elende. Fig. 1. Hunter. Fig. 2. Woman making coiled basket. XL. Ovimbundu. Fig. 1. Flute players, Bailundu. Fig. 2. Bark removed for making utensils. Fig. 3. Girl, hair studded with brass nails. List of Illustrations 99 XLI. Ovimbundu Wood-carvers, Elende. Fig. 1. Making human and animal figures. Fig. 2. Carving drum. XLII. Ocimbundu Making Mats, Elende. Fig. 1. Tool for mat-making. Fig. 2. Threading reeds on tool. XLIII. Transportation, Ovimbundu, Elende. Fig. 1. Bridge across swamp and stream. Fig. 2. Carrying chickens. XLIV. Building Construction, Ovimbundu, Elende. Fig. 1. Maize bin. Fig. 2. Framework of house. XLV. Funeral Rites, Ovimbundu, Elende. Fig. 1. Old Ocimbundu questioning corpse and offering food. Fig. 2. Burial place of chief. XLVI. Funeral Rites, Ovimbundu. Fig. 1. Hut where posses- sions of dead chiefs are kept, Elende. Fig. 2. Horns of ox over grave, near Caconda. XLVII. Funeral Rites, Ovimbundu, Caconda. Fig. 1. Baskets and coffin pole on grave. Fig. 2. Hut over grave. XLVIII. Social Life of Ovimbundu. Fig. 1. Men's club house, Bailundu. Fig. 2. Guest house, Elende. XLIX. Physical Types, Ovimbundu, Elende. Fig. 1. Boys, showing tooth mutilation on left. Fig. 2. Girl with field basket. L. Ovimbundu Women and Infants, Elende. LI. Ovimbundu Women and Children, Elende. LII. Ovimbundu Youths, Elende. LIII. Ovimbundu Men, Elende. LIV. Types of Ovimbundu, Elende. Fig. 1. Negro type, teeth mutilated in tribal manner. Fig. 2. Modified Negro type. LV. Pigs Owned by Ovimbundu, Elende, Keltic Breed. LVI. Domestic Animals, Ovimbundu. Fig. 1. Sheep and lamb, Elende. Fig. 2. Dog, Elende. Ears clipped "to make him hear well." Fig. 3. Goat, Bailundu. LVII. Domestic Animals, Ovimbundu, Elende. Fig. 1. Cattle. Fig. 2. Goats. LVIII. Physical Types, Southwest Angola. Fig. 1. Girls of Luvando tribe. Fig. 2. Woman of Gambos wearing omba ornaments. 100 List of Illustrations LIX. Female Types, Southwest Angola. Fig. 1. Luvando girls near Kipungo. Fig. 2. Back view, same types. LX. Female Types, Southwest Angola. Fig. 1. Girls of Vaheneca tribe near Huila. Fig. 2. Woman of Huila. LXI. Female Types, Southwest Angola. Fig. 1. Women of Gambos. Fig. 2. Women of Humbe tribe. LXII. Vaheneca Girl at Huila. Fig. 1. Front view. Fig. 2. Back view. LXIII. Vakuanyama Types, Mongua, South Angola. Fig. 1. Man. Fig. 2. Man wearing omba shells. LXIV. Natives of Huila, Southwest Angola. Fig. 1. Man wearing forehead band and omba shell. Fig. 2. Man wearing beaded collar and omba shell. LXV. Vakuanyama Women, Mongua, South Angola. Fig. 1. Women wearing pleated leather skirts. Figs. 2, 3. Woman wearing head-dress which denotes position as principal wife in polygynous family. LXVI. Storage of Maize, Vakuanyama, South Angola. Fig. 1. Grain basket. Fig. 2. Shelter for basket. LXVII. Fig. 1. Ukuanyama Man Treading Hide for Making Woman's Skirt. Fig. 2. House, Vakuanyama Style, Dom Manuel. LXVIII. Tribes of South Angola. Fig. 1. Vaheneca, right and left; Gambos woman and children, center. Fig. 2. Vakuanyama men, with assagais, clubs, and bows, Mongua. LXIX. Bushman, Cassanga. Fig. 1. Front view. Fig. 2. Side view. LXX. Mussurongo Types, Malange, Northwest Angola. Fig. 1. Girl. Fig. 2. Man. LXXI. Mussurongo Woman, near Malange, Northwest Angola. LXXII. Dugout Canoes, River Kwanza, near Malange, North- west Angola. Fig. 1. Exteriors. Fig. 2. View of stern. LXXIII. Fishing and Hunting. Fig. 1. Fishing spear, Ambrizette, northwest Angola. Fig. 2. Stakes for impaling antelope, Vasele, west-central Angola. List of Illustrations 101 LXXIV. Fig. 1. Hunter's Tomb, Vasele. Fig. 2. Village of Vasele near Vila Nova de Selles. LXXV. Female Types of Vasele, West-central Angola. Fig. 1. Woman wearing nose-pin and cowrie-shell hair-band. Fig. 2. Woman wearing beaded head-band. Note scarification of forehead and cheeks. LXXVI. Men of Vasele, West-central Angola. Fig. 1. Man showing typical mutilation of teeth. Fig. 2. Man with scarification on chest, said to cure pain. LXXVII. Women in Charge of Initiation of Girls into Secret Society, Vanyemba Tribe, Ngongo. Fig. 1. Front view. Fig. 2. Back view. LXXVIII. Costumes Worn in Tribal Initiation Ceremonies for Boys. Fig. 1. Costumes of Vangangella and Ovim- bundu, Ngongo. Fig. 2. Ceremonial dress of mixed tribes at Katoko. LXXIX. Tribal Initiation of Boys, Vachokue, Cangamba. Fig. 1. Novices in enclosure. Fig. 2. Public appear- ance after initiation. LXXX. Tribal Initiation of Boys, Vachokue, Cangamba. Fig. 1. Fiber skirts worn by circumcised novices. Fig. 2. Cages in which boys lie after circumcision. LXXXI. Concluding Ceremonies, Tribal Initiation, Cangamba. Fig. 1. Stilt-walkers, Vachokue. Fig. 2. Medicine- man, Valuchazi. LXXXII. Vachokue, East Angola. Fig. 1. Medicine-man, Can- gamba, performing ceremony to make thief return to village. Fig. 2. Man at Mona Kuimbundu. LXXXIII. Treating Sick Woman, Vachokue, Cangamba. Fig. 1. Medicine-man stroking patient's spine. Fig. 2. Pre- paring for ablution of face. LXXXIV. Vachokue, Mona Kuimbundu, Northeast Angola. Fig. 1. Women pounding maize. Fig. 2. Dwelling. LXXXV. Vachokue Fishing dragging basket, net. at Cangamba. Fig. 1. Women Fig. 2. Man in bark canoe, holding 102 List of Illustrations LXXXVI. Vachokue, East Angola. Fig. 1. Cupping operation, Ngongo, Ngalangi. Fig. 2. Group at Mona Kuim- bundu. LXXXVII. Vachokue Women, East Angola. Fig. 1. Albino, Cangamba. Fig. 2. Group at Ngongo, Ngalangi. LXXXVIII. Vachokue Women, Ngongo, Ngalangi. Fig. 1. Show- ing mutilated teeth. Fig. 2. Showing scarification. LXXXIX. Fig. 1. House Where King Communes with Ancestral Spirits, Ngalangi. Fig. 2. House with Painted Walls near Bailundu. Fig. 3. Group Showing Mixture of Tribes, Ngalangi. XC. East-central Angola. Fig. 1. Woman of Vangangella, Cassanga. Fig. 2. Beehive, Inandongo. XCI. Types of Babunda, Cangamba, Showing Mutilated Teeth. Fig. 1. Woman. Fig. 2. Man. XCII. Fig. 1. Mound of Earth Where Childless Women Are Covered with Mud to Give Fertility, Vangangella, Ngalangi. Fig. 2. Trap for Leopards, Cangamba. PREFACE As leader of the Frederick H. Rawson-Field Museum Ethnological Expedition to West Africa I undertook research in Angola and Nigeria, from February, 1929, to February, 1930. The present publi- cation is concerned with the ethnology of Angola (Portuguese West Africa), where a study of the Ovimbundu and their culture contacts was made. At the outset I must express thanks to the Portuguese govern- ment for permission to carry on this investigation. Owing to the courtesy of the American Vice-Consul, Mr. Arthur F. Tower of Loanda, I received from the Governor General of Angola a letter of introduction which was of inestimable service during my journey in the interior. From Portuguese officials much help was obtained, and sound advice accompanied by practical assistance was always courteously given. In London I was assisted in the most cordial way by Mr. David Boyle, of the Cunard Steamship Lines, and Colonel B. Follett,D.S.O., of the Tanganyika Concessions Company. My base camp in Angola was pitched near the Elende Mission Station, where Dr. Merlin W. Ennis kindly provided safe storage for collections and a room for photography. Dr. Ennis was ever ready to discuss and aid my investigation among the Ovimbundu with whom he has spent thirty years. Mrs. Ennis and Miss Rounds aided my studies of handicraft among women. Dr. Hollenbeck gave valuable notes on the ailments of the Ovimbundu; these have been incorporated with my study of medicine-men. Ngonga, my interpreter and chief informant, was secured by Dr. Ennis to assist the investigation, and this he did with the greatest tact and ability. Ngonga speaks English, Portuguese, and Umbundu fluently. These accomplishments were so combined with a deep regard for the customs of his own people, the Ovimbundu, that he made an ideal interpreter. From the base at Elende three journeys were made into the interior. The first of these led to the far south of Angola, among the Vakuanyama; a second journey took me into the Vasele country of northwest Angola; while the third and longest itinerary led as far east as Cangamba, a center of Vachokue culture, thence northward to Saurimo in Lunda. 103 104 Preface At the end of each of these journeys, which totaled 5,000 miles, I returned to the base camp, developed photographs, shipped collec- tions, and continued my studies among the Ovimbundu. In the interior I received help from Mr. and Mrs. H. C. McDowell of Ngalangi,bothof whomgave considerable aid in investigating initia- tion ceremonies for boys and girls, and in securing three medicine-men and a rain-maker to explain their vocations. At Bailundu Mr. G. M. Childs obtained many valuable objects relating to the medicine-man's work, and the worth of these was greatly enhanced by full descrip- tions, and translations of the Umbundu language in which the explanations were given. Owing to the interest of Professor Edward Sapir, formerly of the University of Chicago, now at Yale, assistance in transcribing phono- graphic records of the Umbundu language was obtained. The phonetic transcriptions were made by Dr. M. H. Watkins and Mr. R. T. Clarke, whose expert aid was greatly appreciated. Records of drum music were transcribed by Dr. G. Herzog of the University of Chicago. Drawings of objects collected have been prepared by Mr. Carl F. Gronemann, Staff Illustrator of Field Museum. This recognition of cooperation would be incomplete without gratefully remembering my servants, who shared the fatigue and hazards of the journey. Abilio Esteves proved to be a thoroughly competent guide and adviser. The servants aided in locating cere- monies, acquiring objects, and dealing with the tribes among whom the expedition passed. Wilfrid Dyson Hambly THE OVIMBUNDU OF ANGOLA I. INTRODUCTION Research work among the Ovimbundu of Angola indicates the presence of numerous cultural traits revealing what are probably- distinct stratifications of culture. These diverse elements have been welded together into a pattern, the examination of which constitutes the present problem. The object of this study is an analysis of these traits with a view to showing the sequence in which they have been received, from whence they came, and the processes which have been responsible for coordinating them so as to form the present social system. My presentation passes from geographical and historical consid- erations to an ethnological approach, locally applied in the first place for complete analysis of the culture of the Ovimbundu and surround- ing peoples. The traits have to be considered singly and in combina- tion. Then follows an examination of some of these traits which are widely distributed outside Angola. We next seek to ascertain the geographical origin of traits, their history, and the psychology of their combinations and assimilation into an aggregate. What elements have been acquired through adoption? And what has been the historical process? Which traits have arisen through independent invention? What factors have been discarded, and why? To what extent have the Ovimbundu utilized opportunities for enriching their culture, and what possibilities have been neglected? What are the classes of evidence which might be expected to assist an inquiry into the growth of Umbundu culture? Field work was of primary importance for obtaining a first hand knowledge of the tribal life of the Ovimbundu as it exists today, and in order to estimate the effects of cultural contacts extensive journeys were made to the north, east, and south of the central territory occu- pied by the Ovimbundu. The results of personal investigation are given in chapters IV-IX, dealing with the economic, social, and religious life of the Ovimbundu. These facts have been kept free from the observations of other investigators whose reports are summarized in chapter III, "Historical Sources." Unfortunately there are no archaeological data which can assist a study of historical processes, for archaeology has not yet been approached in Angola. 105 106 The Ovimbundu Geographical study is valuable in showing that physical factors such as position, topography, soil, and climate have had not only a permissive but a stimulating effect on the development of certain cultural traits. Historical documents dating from the year A.D. 1500 present valuable evidence for tribal movements, the effect of early European contacts, and the existence of certain beliefs and ceremonies. The facts adduced in chapter III are used in chapters X-XII in discussing culture contacts and cultural processes. Inquiry respecting relationships of the Ovimbundu to other Afri- can tribes is aided by a study of the Umbundu language, which is shown to be of pure Bantu structure and vocabulary (chapter VIII). The characteristic features of the Umbundu language are those which form the basic elements of Bantu speech in general. These character- istics are alliterative concord (that is, repetition of the prefix before every word in agreement with the noun); absence of grammatical gender; and a position of the genitive in which the name of the thing possessed comes before the possessor. Umbundu, though structurally assignable to the Bantu group, has its own vocabulary, whose degree of relationship to that of surrounding peoples may to some extent be judged by the vocabularies of F. and W. Jaspert (Die Volkerstamme Mittel Angolas, Frankfort, 1930, pp. 144-150). The fact that the language of the Ovimbundu has become the lingua franca of Angola, still further testifies to the thoroughness of the contacts, which from historical sources are known to have been made in the period 1500-1900. The spelling of proper names leaves a wide margin for individual preference. The Umbundu language requires "M" before "B," yet custom has sanctioned the form Bailundu. Nevertheless I have retained "N" in Ngalangi despite the form Galangi on several maps. Maps show great diversity of spelling, but I have adhered throughout to Kipungo (Quipungo) and Kwanza (Quanza). Among tribal names I prefer Vakuanyama to OvaKwanyama and Vachokue to BaKioko or BaDjokue; in making the choice I have tried to imitate the sounds I heard from the natives themselves. In referring to the papers of E. Torday, who knew the southwest Congo well and spoke several Congo languages fluently, I note that he sometimes prefers to recognize the prefix as in BaMbala and BaYaka, but he also writes Badjokue and Bayaka. J. H. Weeks writes Bangala (BaNgala). C. H. L. Hahn uses the form Ovambo (OvaMbo). Ovimbundu I have preferred throughout; the accent is on the Introduction 107 penultimate syllable as is usual with Bantu words. The alternative form would be OviMbundu. Some writers hyphenate after a prefix. Lu is the plural prefix in the tribal names Luchazi, Luvando, Luena and Luimba. Ocivokue, Ocimbundu and Uluchazi are singulars. The word Umbundu is used adjectivally as well as for the name of the language. Observations relating to the physical appearance of the Ovim- bundu (chapter IV) deal with physique, dress, tooth mutilation, hair- dressing, scarification, and personal ornaments. The object of this chapter is to estimate the results of contacts so far as these affect the traits just mentioned. Of great importance in an inquiry into the growth of Umbundu culture is the inferential testimony to be derived from ethnological study of surrounding tribes. This subject is considered in chapter X, "Culture Contacts," with special reference to the Congo basin, Rhodesia, and South West Africa, concerning which there exists an adequate and reliable literature. In chapter XII, "Cultural Processes," ethnological facts derived from field work and historical sources have been combined with data relating to geography, physique, and language. This has been done in such a way as to present a hypothesis of cultural growth which is consonant with direct evidence and inferential testimony. Finally, personal acquaintance with the daily life of the Ovim- bundu is made the corner stone for behavioristic study. This is intended to explain the operation of social forces and controls in welding the tribal traits whose origin and assembly have been previously discussed. Should my colleagues of the functional school contend that my approach is too static in its historical and anatomical method, I would reply that my research does at least lay a sure ethnological foundation for those psychological and sociological studies which are today rightly regarded as essential for the adjustment of relation- ships between Africans and their European administrators. The correlation of all aspects of tribal life, including culture contacts, cannot be too strongly stressed. The headings chosen for chapters are adopted merely for convenience of presentation. But an endeavor has been made throughout the book, and particularly in the final chapter, to emphasize the coordination of economics, social organiza- tion, education, language, and religion. II. GEOGRAPHICAL FACTORS This chapter is intended to give an outline of the geographical conditions of Angola and to point out the way in which these have influenced the growth of Umbundu culture in all its aspects. A presentation of geographical data is of primary importance here, and the actual relationship between the facts of geography and culture is a matter for gradual evaluation in the following chapters. The extent to which geographic determinism has entered into the cultural growth of the Ovimbundu is particularly well seen in chapter V, "Economic Life," which deals with industries, agriculture, domes- tic animals, fishing, and transport. The area of Angola is a factor of importance, for the greater the extent of any country the more diversified will be the products and the cultural differences. This is particularly true if there are consider- able differences in the elevation, because altitude modifies temperature and affects the distribution of rainfall. Angola has an area of almost 500,000 square miles. The greatest length is a distance of about 900 miles from the river Congo in the north to the region of South West Africa. The greatest breadth is about 700 miles from the Atlantic Ocean on the west to the borders of the Rhodesian plateau in the east. This range over twelve degrees of latitude, from 5° S. to 17° S., in association with differences in altitude, have produced ecological regions which are described in the following paragraphs. The northern part of Angola is ecologically a part of the Congo basin, with conditions of heat and moisture giving rise to a dense tropical flora. Owing to the presence of tsetse fly, the great heat, and the lack of open grasslands, such country is unsuited for the rearing of cattle, while agriculture, including cultivation of maize, is restricted to forest clearings. Manioc, palm trees, peanuts, and sweet potatoes are the chief products serviceable to man. This region was, according to the historical evidence of chapter III, the area in which the Ovimbundu moved before they entered the Benguela High- lands of central Angola. Parts of chapters III and X ("Historical Sources" and "Culture Contacts," respectively) are devoted to an analysis of the cultural factors of the Congo area from the year a.d. 1500 to the present day. The central portion of Angola is the area most important in this research because it is the home of the Ovimbundu, whose cultural 108 Geographical Factors 109 growth is under examination. This central plateau, the Benguela Highlands, rises in places to a height of 6,000 feet, an altitude which reduces the heat of the tropics and so modifies the flora, discouraging some types of vegetation and encouraging others. Hostility of the highlands to palm trees and the banana is appropriately discussed when dealing with cultural losses; while the fostering effect of reduced temperature and the presence of wide expanses of open country on the growth of maize and the keeping of cattle is a cultural gain. The prevailing rains of Angola are from the northeast to the southwest; consequently the high plateau intercepts rain clouds which give an annual fall of sixty inches. This adequate rainfall is another factor which has affected economic life, density of population, com- munal welfare, and powers of expansion. The villages of the Ovimbundu are built on hillsides having a commanding view of the surrounding country. There is in the nature of the land a natural protection from enemies. In addition to this the Benguela Highlands are an admirable base from which expeditions both predatory and commercial might, and actually did set out eastward to the interior of Africa, and southwest to the cattle-keeping country. It is important to note that this central plateau is the watershed for four large river systems; the Kwanza to the northwest; the Cunene to the southwest; the tributaries of the Kasai to the northeast; and the Zambezi and its affluents to the southeast. Fishing, with attend- ant beliefs of a ritualistic kind, is of local importance; so also is the making of canoes. Moreover, the river valleys have marked out a natural means of communication in several directions. In connection with the river system the biological factor of the tsetse fly is important, because the presence of the fly locally dis- courages human habitation and prevents the keeping of cattle. The exact distribution of the fly is imperfectly known, but Glossinia palpalis, the cause of sleeping sickness in human beings, also Glos- sinia morsitans, which carries disease to cattle, are both present along the Kwanza and parts of the other rivers (J. C. B. Statham, Through Angola, p. 294). The western coastal strip is a region of great aridity which has in some years no rainfall whatever, because the northeasterly rains have expended themselves on the high plateau. This region displays vegetation of the semi-desert type; namely, baobab trees, prickly acacias, euphorbias, and aloes. Population in this area is sparse, always nomadic, and in some places non-existent. The coastal strip 110 The Ovimbundu was at times traversed by the Ovimbundu who know of the sea and call it kalunga. Bihean caravans crossed to the coast with slaves, as history shows, but the nature of the coastal strip marks it as a western barrier limiting the expansion of the Ovimbundu in that direction. The coast line itself is of the greatest moment in the consideration of historical factors. From Loanda, Benguela, and Mossamedes on the coast, the Portuguese penetrated the interior, so making contacts with the Ovimbundu to the encouragement of caravan trade. The importance of this European contact will be seen in subsequent chapters. The south and southwest parts of Angola are of particular impor- tance in studying the contacts of the Ovimbundu, but to give here the details of the wealthy cattle-keeping culture of these regions, which were accessible through peaceful proximity, trade, and occasional raiding, would be an anticipation of chapter X, "Culture Contacts." One topographical point is of primary importance; namely, the ease with which the Ovimbundu could descend from their strongholds to the low-lying land of the west and south, whereas the reverse journey is much more difficult for a people unaccustomed to manoeuvring and finding their way among hills. Having described the northern, central, western, and southern areas, there remains only the eastern section to consider; this presents several features of peculiar geographical and ethnological interest. The eastern section of Angola is either slightly undulating or flat, the general characteristic is dryness, and vegetation is somewhat sparse though sufficient to shelter many kinds of antelope. The major population is the Vachokue, a warlike, hunting people, who follow agricultural pursuits but slightly, and do not keep cattle. Examination of the literature describing early exploration indicates the truculent nature of these eastern people with whom caravans of Ovimbundu were in frequent conflict. Umbundu caravans crossed this country when making their way to Rhodesia, culture contacts with which are discussed in chapter X. The results of contact of Ovimbundu and Vachokue tribes, so far as physical miscegenation is concerned, are mentioned in chapter IV, "Physical Appearance"; while the social effects of slavery resulting from hostilities are described in chapter VI, "Social Life." Rhodesia is a cattle-raising country, but the hostility of the Vachokue, the great distance from central Angola to Rhodesia, and the general dryness of the country to be crossed, discourage the idea that the Ovimbundu obtained their cattle from Rhodesia. If it is Geographical Factors 111 argued that the cattle might have been brought along the course of the Zambezi and Kwando there is the objection that there are here several tsetse fly belts. On the contrary, cattle-producing country in the southwest and south of Angola is far more accessible than the Rhodesian plateau. In the south there is the cultural habit of digging wells, especially among the Vakuanyama; but the Vachokue have not developed this trait. The substrata underlying the sand of southern Angola hold water which serves through the dry season, a fact which is advanta- geously employed by the cattle-keepers. The Vachokue lack this well-digging habit, and, even if subsurface water were present, the transient Ovimbundu would have lacked opportunity to dig for it when passing through hostile country. The acquisition of cattle by the Ovimbundu is of great impor- tance, because it is concerned with the grafting of a series of pastoral traits on a culture in no way originally associated with pastoral pursuits. The truth of this will later be made clear by examination of historical and ethnological evidence. Geographical considerations give a picture of the Ovimbundu situated in naturally fortified country from which they had access to four surrounding areas, whose cultural characters agree well with the determinism of topography and climate. The natural advantages of the central highlands, and the results of contact with each of the adjacent areas, are points which will be developed in appropriate sections throughout the ensuing chapters. III. HISTORICAL SOURCES The object of this chapter is the presentation of a summary of literature relating to Angola from a.d. 1500 down to the present day. In this literature there may be accounts of the movements of tribes which will throw some light on the origin of the Ovimbundu, and the date of their migration into the Benguela Highlands. Present- day tradition of the Ovimbundu is unanimous in declaring that the tribe came from the northeast of its present locality, but some historical justification of this belief is desirable. The word Ovim- bundu ("people of the fog") may refer to the heavy morning mists of highland regions, and there is the possibility that the Ovimbundu adopted this name when they settled on the high plateau. If there is historical evidence in favor of the traditional home of the Ovimbundu being the region of the north and northeast of Angola, what was the ethnological background of these areas at the time when the Ovimbundu became detached from a matrix of northern Angolan tribes? In other words, what cultural factors are the Ovimbundu likely to have brought with them into the Benguela Highlands? The ultimate origin of these factors is a point which has not been neg- lected (chapter XI, "Wider Culture Contacts"), but for the moment the primary concern is the historical background of the Ovimbundu, with special reference to tribal movements and cultural traits. As part of this historical inquiry, contacts of the Portuguese and the Ovimbundu are of importance. When and where did the Portu- guese come into contact with the Ovimbundu and with what results to the indigenous culture? Chapters IV-IX deal exclusively with my field work among the Ovimbundu in 1929. Therefore it is pertinent to ask to what extent the ethnological observations of travelers and early explorers will corroborate the information recorded in my own notes. Will there be contradiction or confirmation of personal observations? Possibly the ethnological notes obtained from historical sources will serve, not merely to corroborate my field work, but actually to extend the area of occurrence of important cultural traits which I noted in 1929. The literature describing Angolan history and customs is here presented in the form of a chronological bibliography, which is annotated so as to emphasize points bearing on the purpose of this historical analysis. That there should be a fragmentary presentation of data is an inevitable consequence of the nature of the literature 112 Historical Sources 113 itself. Observations have been made, not in a well-ordered time sequence, but at irregular intervals. Moreover, the writings of the majority of observers have not been undertaken with any specific ethnological purpose in view. Early explorers in particular were prone to intersperse historical and ethnological notes among a mass of descriptive material relating to incidents of travel, animal life, and meteorological observations. At the conclusion of this chapter an effort is made to remedy the disjointed nature of the historical evidence. This object is achieved by summarizing the points which provide an answer to the queries brought forward in the opening paragraphs of this chapter respecting the origin and cultural background of the Ovimbundu. The Portuguese entered the Congo in 1482 under the leadership of Diego Cao (E. G. Ravenstein, The Voyages of Diego Cao, Geog. Journ., 1900, pp. 625-649) and from that time onward Portuguese influence of a political and religious kind was exerted along the course of the Congo. Gradually the Portuguese established them- selves on the coast of Angola. Paolo Diaz founded Loanda in 1576, and about eleven years later built the fort of Benguela. The year 1590 saw the Portuguese making war in the interior of Angola against the Jaggas, a northern tribe among whom Andrew Battell was held in honorable captivity as a leader against the Portu- guese and all natives of northern Angola, who were exploited by the Jaggas. In 1645 another Portuguese punitive expedition penetrated the interior as far as Bailundu, the center from which the strongest caravans of Umbundu traders and slavers set out for central Africa. Caconda in the southwest of Angola was founded in 1682, and a century later the coastal town of Mossamedes became a starting point from which early exploration penetrated the interior in search of the sources of the Cunene River. (For the details of Portuguese pene- tration of Angola see Bibliography: T. E. Bowditch; R. F. Burton; E. G. Ravenstein; T. Lewis.) When the Portuguese landed at the mouth of the Congo at the close of the fifteenth century they came into contact with the kingdom of Congo, ruled with great pomp and ceremony at Ambassa, about 150 miles inland, and identical with the San Salvador of the Portu- guese. The old kingdom of Congo was made up of six strong clans of whose rivalry the Portuguese took advantage to strengthen their own commercial and political position. The slave trade was con- sidered to be as respectable as it was lucrative, and there is no doubt 114 The Ovimbundu that the Church participated actively (T. Lewis, The Old Kingdom of Kongo, Geog. Journ., 1908, pp. 598-600). Political influence of the Portuguese, working often through the agency of Jesuit priests, led to factions within the Congo Empire, and the resulting disturbances caused movements of peoples that affected the whole of northern Angola and the population of the Benguela Highlands. Portuguese penetration of the hinterland of Angola, especially from Loanda to Bih£, was concerned with attempts to subjugate native tribes, the establishment of trading posts, and the encourage- ment of the slave trade (S. Marquardsen, Angola, 1928, pp. 6-10). The value of the Portuguese as allies of Umbundu caravans from Bine" lay in their ability to supply guns and powder to their native henchmen, who gave something more than military service in return. In response to Portuguese demand for slaves and ivory, Umbundu caravans made long journeys into the Congo basin, Rhodesia, south and southwest Angola, and possibly across Africa to lakes Tanganyika and Nyasa. The arms supplied in exchange for ivory and slaves must have helped the Ovimbundu in all their predatory excursions. Thus the Ovimbundu were, in the early centuries of contact with the Portuguese, invaders encouraged in the building up of their tribal life and resources. Never were the Portuguese strong enough completely to subjugate northern Angola. Relationships with natives, especially the Bihean section of the Ovimbundu, were directed toward alliances on a commercial basis. The political result of this was a combination of the Portuguese and the stronger tribes for the exploitation of the weaker. The historical ethnology of the southwest Congo is so complex that the elements are difficult to disentangle. The number of tribes concerned is great, and their movements are not easy to follow; but a gradual extension of people from the Congo in a southwesterly direction through Lunda to the Benguela Highlands (1600-1800) seems to be the summation of all the conflict. I regard the Ovim- bundu as the most southerly branch of these mass movements, during which they received a discipline that enabled them to make their home in central Angola, despite opposition from the Portuguese and earlier arrivals. For these conclusions reliance has been placed on the extensive field work and historical analyses of E.Torday and T. A. Joyce, whose "Notes on the Ethnography of the Bambala" (J.R.A.I., XXXV, pp. 398-426) have proved of particular value in this connection. Historical Sources 115 The kernel of the migratory problem of the Ovimbundu is reached when Torday traces out the history of the Kimbundu, for the Kim- bundu are present-day neighbors of the Ovimbundu, to whom they are closely allied in language and culture (ovi is a Bantu plural prefix, which was perhaps used to express the inclusion of Kimbundu and Babunda under the general name Ovimbundu). Torday's sifting of the historical evidence results in the conclusion that the Kimbundu came from the northeast, fighting their way to the Luando. These Kimbundu divided, with civil war as a conse- quence. One section crossed the river Kwanza, south of which they specialized in agriculture after becoming sedentary. Their sub-chief they called the Kalunga, which is the present-day word used by the Ovimbundu in greeting their chiefs. Traditions of the Ovimbundu point to the northeast as a center from which they spread at least ten generations ago. The rise of the powerful kingdom of Lunda dates from the seventeenth century, and although the details of this concentration of power in northeast Angola are unknown the general effects are understood. There was a great displacement of tribes in a southerly and southwesterly direction. In connection with the suggestion that the Ovimbundu came from the northeast of Angola, there is the necessity for recording the absence of even a fragment of historical evidence or tribal tradition indicating that the Ovim- bundu came from the south or the east of Angola. Cultural affinities of the Ovimbundu with tribes of the southwest Congo in particular, and with western Bantu culture in general, strongly support the foregoing deductions from historical sources (chapter X). There is a probability that tribal disturbances resulting from the rise of the Lunda Empire led to a spread of the Vachokue over eastern Angola. Some of the effects of this contact have already been mentioned in chapter II, where contacts of the Ovimbundu and the Vachokue were discussed. The journey brought me into contact with the Vachokue at Cangamba, Katoko, and Ngalangi, of which Cangamba is the most easterly. Witnessing of initiation ceremonies combined with observation of physique, language, and artifacts, leads me to agree with Torday that Cangamba is probably the ancient center of Vachokue culture. As one proceeds from Cangamba westward this type of culture becomes thinner as the borders of the country occupied by the Ovimbundu are approached. On the border line between the cultures of the Vachokue and the Ovimbundu, notably at Ngongo near Ngalangi, 116 The Ovimbundu there is tribal and cultural miscegenation. Both Umbundu and Vachokue languages are spoken; boys of Ngalangi were seen to wear initiation costumes similar to those worn at Cangamba by the Vachokue, but on proceeding farther westward into territory exclusively occupied by the Ovimbundu, such initiation ceremonies are either absent or attenuated in ritual. Elende, one center of research among the Ovimbundu, repre- sents the purest Umbundu speech and culture observable at the present day. But in chapter IV some physical resemblance of the Ovimbundu to the Vachokue is noted. Warfare commonly resulted in the taking of slaves, a fact which might account for an apparent infusion of Vachokue blood in some of the Ovimbundu. The historical data suggest a northern or northeastern starting point of Umbundu migrations. Therefore an inquiry into the ethnology of these areas will be useful in showing the kind of culture with which the Ovimbundu were in contact before their settlement in the Benguela Highlands. If historical sources disclose the nature of northern Angolan culture from the year a.d. 1500 onward, such evidence can then be considered in relation to cultural traits of the Ovimbundu at the present day. The establishment of strong cultural resemblances between extant Umbundu culture and older cultural patterns of northern Angola, would tend to strengthen the evidence of history respecting the northern origin of the Ovimbundu. The regions dealt with in the following summary of ethnological facts are the Cabinda Enclave to the north of the Congo estuary, the region of San Salvador, the hinterlands of Loanda and Benguela, and the area of northern Angola between Lat. 7° and 9° S. and Long. 13° to 22° E. This covers the whole area with which the Ovimbundu are likely to have been in contact before entering their present home. My survey begins with the observations of Andrew Battell in 1596. His account deals mainly with the northwestern part of Angola, a country bordering on and actually including territory now occupied by the Ovimbundu, who in all probability came in contact with the Jaggas, whose habits of life are described by Battell. Tordayand Joyceidentify the Jaggas with the present-day Bayaka, whose cultural resemblances to the Ovimbundu are examined later. In my opinion, the Jaggas correspond well with the Bihean section of the Ovimbundu, an itinerant and exceedingly warlike people. Historical Sources 117 As early as 1600 there were cattle as far north as Benguela, and the Jaggas regarded these as a most valuable part of their plunder. Palms are a characteristic part of the vegetation of Angola north of the Benguela Highlands. Evidently the Jaggas moved extensively in northern Angola because they cut down palms for making wine (Battell, p. 30). The usual method is to tap the top of a standing tree, but the Jaggas were an itinerant people who did not culti- vate palms. Sprinkling the blood of sacrificed animals on a newly kindled fire I have mentioned in connection with Umbundu rites celebrating the founding of a new village. The Jaggas carried out this ceremony before a raiding expedition, when cows and other animals were sacrificed (Battell, p. 33). Battell mentions the use of red tukula wood for personal decoration. This wood (Pterocarpus tinctorius) is used in many parts of Angola at the present time. Reference to Battell confirms the information given to me in relation to an old iron gong obtained at Ngalangi. He says (p. 20), "The general did strike his gong, which is an instrument of war that soundeth like a bell, and presently made an oration with a loud voice." Battell reported that the Jaggas wore beads of ostrich eggshell. Ravenstein, the editor of this volume in the Hakluyt Series, seems to doubt this statement, saying, "There are no ostriches in Angola, and as to beads made of ostrich eggs I can give no explanation." I noted the presence of captive ostriches in southern Angola as far north as Gambos. The Vakuanyama women greatly value their long necklaces of ostrich-eggshell beads. Necklaces made in the north of Angola are traded to the south and conversely, each kind of necklace having a high value due to remoteness of origin. This instance, like many other points, confirms the reliability of Battell's observations in Angola. Father Jerome Merolla (1682) describes the poison ordeal which was used from the Congo estuary to San Salvador (Churchills' Voyages, II, p. 675). "The aforesaid oath is administered to the supposed traitor by a sort of wizard, who, making a certain composi- tion out of the juices of herbs, serpent's flesh, pulp of fruits, and divers other things, gives it to the supposed delinquent to drink. If guilty (as they tell you) he will immediately fall down in a swoon or trembling to the ground." The marimba was used, and a double iron gong was carried before a chief and struck by an attendant. 118 The Ovimbundu 0. Dapper (Description de l'Afrique, 1732, p. 369) shows a scene in which an ax, such as I obtained from the Vasele country, is being used for beheading a slave. Cavazzi (Istorica descrizione, etc., Bologna, 1687) describes the three kingdoms of Congo, Matamba, and northern Angola. The poison ordeal, the scapegoat, the blacksmith's bellows (pp. 101, 170) are all traits known to the Ovimbundu. The musical instrument made from a large gourd, which has a ridged board attached for rubbing with a stick, is the type I collected. Cavazzi pictures the double iron gong and the long drum held between the knees. He also shows a rain-maker (p. 214) and the sacrifice of two hundred victims at the accession of a king (p. 210). Consideration of the history of maize in Africa is of importance in connection with these early writings, as this grain is the staple agricultural product of the Ovimbundu. I am indebted to Dr. Berthold Laufer for access to his unpublished research on this subject. From the following facts one may assume that, in all probability, the maize culture of the Ovimbundu was derived from the Congo region before their migration into the Benguela Highlands. Father Jerome Merolla remarks that maize was growing in the neighborhood of San Salvador (1683-92). The native names were mampunni and massambuta; from this corn an alcoholic beverage was prepared. According to Cavazzi (Ehrmann, Geschichte der merk- wiirdigsten Reisen, XIII, 1794) maize was not intensively cultivated by the Negroes of lower Guinea, though it thrives well and may be harvested twice or thrice a year. The natives said that the grain was brought by the Portuguese, but they did not esteem it highly, and were accustomed to use it as food for pigs. Bosman (p. 312) records that prior to the arrival of the Portu- guese, Negroes were entirely ignorant of milho ("maize"). The account of Duarte Lopez preserved by Filippo Pigafetta states that the Negroes consider maize the vilest of all grains, so that it is given to swine. This contempt and lack of knowledge of the food value, combined with ignorance of methods of preparation, suggest a recent introduction (Pigafetta, translation by M. Hutchinson, 1881, p. 40). Dapper (Description de l'Afrique, 1732, p. 345) also mentions the cultivation of maize. Battell lived as prisoner in northern Angola about the year 1600, consequently his records of the use of the great Guinea wheat (maize), which the natives call mas-importo, give early evidence for the use of this grain (Hakluyt Soc, 1901, pp. 9, 11, 67). Historical Sources 119 The foregoing facts, when compared with field work among the Ovimbundu, indicate that the old culture of the Congo and northern Angola bears a strong resemblance to Umbundu culture at the present day. A more detailed analysis of this resemblance is made in chapter X. The following notes dealing with exploration in Angola (1800- 1930) are adduced for critical comparison with my own observations among the Ovimbundu. The work of Sir R. F. Burton describing the exploration of Lacerda and other Portuguese pioneers, is more useful for geographical than for ethnological information. Lacerda's journey to Czambe, south of Lake Moero, was performed in 1798. A mention of veneration for the dead and consultation of the deceased on all occasions of war or of good fortune, is made (p. 127). These are important points in the present tribal life of the Ovimbundu. Bowditch (1824) writes from information given to him by Almeida and Saldanha. There are valuable references to the Vacilenge, a people adjacent to the Ovimbundu, who refused to kill their cattle, "rather than do which they will endure famine to extremity." This note was made about Long. 15° E. and Lat. 15° S. (p. 34). The cattle were milked, and cattle-raiding by the Ovimbundu was common at this time. Many present-day Ovimbundu do not milk their cattle, but the Vacilenge still follow their milking custom observed before 1824. Livingstone's journey through northeast and northern Angola in 1853 contains references to the Vachokue who seemed bent on plun- der (I, p. 370). Livingstone noted cotton spinning, which I recorded as an occupation of males of the Ovimbundu at Elende. Livingstone, like other early travelers, encountered caravans of Biheans (Ovim- bundu) bearing elephants' tusks and beeswax, commodities, which, along with slaves, formed the chief merchandise of these caravans from the Benguela Highlands (p. 466). Livingstone notes and sketches (Plate XIII, Fig. 10) the double-handled hoe (p. 442) which is still used. J. J. Monteiro (I, p. 61) saw the poison ordeal administered. The poison itself was prepared from the thick hard bark of a large tree (Erythrophlaeum guineense). The place of observation was Mongue Grande, just south of the Congo estuary, and again Monteiro was present when the poison cup was given to two women at Ambrizette. Of this ordeal there is more to be said when discussing the culture contacts of the Ovimbundu, because the ordeal is a basic factor of 120 The Ovimbundu wide distribution in Africa (C. Wiedemann, 1909) and the ceremony- still survives among the Ovimbundu in modified form. Though an informative writer, Monteiro is sometimes vague; he says (I, p. 278) that circumcision is a universal custom among the blacks of Angola. In view of the extent of Angola, the diversity of tribes, and the fact that Monteiro traveled over only a small part of the country, this information is misleading. Monteiro is the only writer in whose works I have found a reference to the Vasele tribe; probably no part of Angola has been so neglected as the Esele country in the hinterland of Novo Redondo. At the present day the Vasele have a reputation for cannibalism. Monteiro saw human flesh eaten at Cuacra, while the skulls of the victims were placed on adjacent trees. Monteiro states (II, p. 167) that on the death of a king the Mucelis (Vasele) put out all the fires in the kingdom; these were relighted by the succeeding king who used fire produced by rubbing two sticks together. The flat beads of shell called dongos, made from Achatina monetaria (II, p. 168), are made today and traded to the far south of Angola, where their novelty assures them a value far beyond their intrinsic worth. I was fortunate in obtaining an example of the old beheading ax mentioned by Monteiro (II, p. 157) and sketched by Cavazzi (p. 210). Par- boiled and roasted rats were offered to me as food; Monteiro mentions the offering of a roasted rat on a skewer (I, p. 99). The account of Commander V. L. Cameron, who left Zanzibar for his journey across Africa in 1873, mentions several points of anthropological importance in relation to the Angolan section of his journey. Cameron saw a net-covered medicine-man of the Kibokue (Vachokue) wearing a mask and a kilt of grass (p. 384). The function of this man was to frighten devils from the woods. The contest in which boys discharge their arrows at a rolling root was seen by Cameron near Kagnombe (Cangamba?). Skulls of victims killed in war were spiked on poles (p. 399). The diviner was followed by attendants who struck iron gongs, while the diviner himself shook a rattle made of basket-work in the form of a dumb-bell (p. 404). Cameron gives an accurate description of the divination basket and its use without going into details. These I have been able to supply (chapter IX). The explorer met caravans of Biheans, renowned carriers then as they are today. They were usually drunk and abusive; in some instances they attempted to rob the stragglers. The use of caterpillars as food is noted (p. 416) : "A man cut open Historical Sources 121 a large cocoon, extracted the contents, and smacked his lips with great gusto." My field notes mention the use of caterpillars as food. Capello and Ivens (1877-80) remark on the burial places of hunters which are distinguishable by the skulls of antelopes, buffalo, and hippopotamuses, stuck on upright poles, mixed with skulls of oxen killed in honor of the defunct. The writers noted that a heap of stones protected the body. I photographed two types of cairn in the regions of Ganda and Luimbale respectively. Capello and Ivens are not precise in their locality, but I judge it to have been Long. 17° E. and Lat. 13° S., a considerable distance from my own observations (Plates XXXII, Fig. 1; LXXIV, Fig. 1). I photographed the stilt-walkers at the final stages of the initiation ceremonies at Cangamba. These men had no costumes; on the contrary they were almost naked but were covered with white clay. Capello and Ivens (p. 295), saw a stilt-walker with a feathered mask and a netting costume. Such attire I saw on an Uluchazi medicine- man, but not on the stilt-walkers (Plate LXXXI, Figs. 1, 2). I was unable to obtain information about the stilt-walkers, but Capello states that they castigated misdemeanants, punished shame- less women, and accused criminals. Capello and Ivens were 187 miles to the northwest of my area of observation. The Uluchazi medicine-man, who appeared with the stilt-walkers during my visits, was said to make bad magic for women. The men ignored him but women gathered round in a derisive way; they quickly scattered when he pursued them. Serpa Pinto, who made his journey across Angola in 1878, has so many references to customs and objects still extant, that I propose to tabulate his observations because of their value in showing the preservation of indigenous traits in spite of Portuguese contacts. The body of a chief is buried with a covering of oxhide. Many oxen are sacrificed at the death of a king. The heir to the deceased is bound to sacrifice his whole herd in order to regale his people and give peace to the departed (S. Pinto, I, p. 63). The Ovimbundu conform at the present time to similar methods of burial and sacrifice. Near Huambo, Serpa Pinto saw in every village a kind of "temple for conversation." This is the onjango, that I have described and photographed (Plate XLVIII, Fig. 1). It is the house in which all males foregather for the evening meal, which is brought by their women (I, p. 96). The gathering and eating of caterpillars is described. This con- tinues today (I, p. 120). 122 The Ovimbundu Serpa Pinto saw shafts for the working of iron ore in the neighbor- hood of Cubango. The ore was mixed with charcoal and smelted in shallow pits. It is stated that the iron was sometimes tempered with ox-grease and salt. The bellows are of the type made at Elende (I, p. 128). Of the tempering process I have no confirmation. Somewhere near Bih£ Serpa Pinto saw the ceremony of question- ing a corpse which was made to sway to and fro, the people believing all the while that it does so without human intervention. The diviner declared that the soul of a dead person will tell who caused the death (I, p. 130). I observed and photographed this ceremony (Plate XLV, Fig. 1). The ordeal of the poison cup is described; blood-letting, and divination by shaking articles in a basket are also mentioned. Pinto says that in the articles that appear uppermost the diviner reads what his hearers are desirous of learning of the past, present, or future. Sorcery and rain-making are likewise briefly mentioned (I, p. 132). The divination basket, the poison ordeal, and rain-making, are Umbundu cultural traits today. Here Pinto illustrates the existing practice of mounting the skulls of animals killed by a hunter on a pole in the village. There is a further reference to this custom among the Ambuellas (I, p. 333). These instances, combined with those personally noted, give a wide distribution for the practice (I, p. 177). Pinto saw the operation of tooth mutilation among the Luimba. The operation was performed with a knife which was struck by repeated light blows (I, p. 209). Pinto illustrates arrowheads in use in 1879 (I, pp. 277, 346). These are exactly the same as those made and used in 1929 (Plate XVII, Figs. 1-9). Axes used by the Luchazi are sketched with a detail that shows the old forms to persist without alteration (II, p. 36). Water-pipes made from horns and gourds by the Luina of eastern Angola are of the forms now used in that region, also by the Ovim- bundu of the present day (II, pp. 33, 37). For use in checking and supplementing my observations Serpa Pinto's book was found to be of the greatest service, though the precise locality was sometimes difficult to identify. This explorer, in common with others, scattered his ethnological observations among descriptions of the route and botanical, zoological, and other notes. Lux traveled from Loanda due east, following the Kwanza on the northern bank between 9° and 10° S. Lat., and so into Lunda. Historical Sources 123 He draws the crescentic arrowhead (p. 123) still common among the Ovimbundu and Vachokue; the double iron gong (p. 122); and the musical bow, which he calls a viola (p. 121). Sogaur states that iron-working had an advanced technique at Dindo, more than half a century ago. According to Sogaur, the blacksmith was using scrap iron from European sources (II, p. 14). Perhaps the most valuable of Chatelain's books is "Fifty Folk Tales of Angola." The stories are accompanied by translations and ethnological notes, the most important of which are references to the carrying of a corpse on a pole; the building of a cairn of stones over the corpse of a hunter; matrilineal descent and the power of the maternal uncle over the persons of his nieces and nephews (sisters' children). My observations included some details respecting the pawning of a sister's children to redeem the debts of their mother's brother. Chatelain adds that sisters' children are successors to private property and chieftainship (pp. 8-10). My notes agree that property is inherited by children of a deceased man's sister, but my informants said that a new chieftain is normally the eldest son of the principal wife of the dead chief. Chatelain's folklore stories, fifty in number, relate chiefly to animals; this was the only kind of story told to me, but W. C. Bell has recorded a few tales of another type. Marquardsen (1928) devotes only one-fourth of his book to the ethnology of Angola, which he treats in a very general way; there is no section dealing specifically with any particular tribe. The author calls attention to Chapman's observation of rock paintings of South African Bushman type, between Cuma and Luimbale in northwest Angola. I have elsewhere remarked on the occasional occurrence of physical types which show a strain of Bushman blood. Today Bushmen penetrate southern Angola from the Kalahari (Plate LXIX, Figs. 1, 2), but their presence in times past or present does not affect the course of history or the data of ethnology to an appreciable extent. Marquardsen gives some notes of a general kind on Vaheneca, Mahuila and other tribes of southwest Angola. The suggestion is feasible that the Umbundu name Suku, for a supreme being, is the same as the Nzambi of Lunda and the Congo. Marquardsen refers to the importance of the maternal uncle in Umbundu society, to the poison ordeal, and to the burial of an Umbundu chief in oxhide, all of which points were recorded in my observations at Elende. Ferreira Diniz's book is unobtainable, but, judging from the sum- mary given in Anthropos (XX, 1925, pp. 321-331), the information was collected by a questionnaire widely distributed among adminis- 124 The Ovimbundu trative posts. I agree that people between Caconda and Huila represent a racial mixture of the Ovimbundu and the Ovambo. The record speaks of painting-houses for girls at Cabinda, an item which agrees with reports from other sources. The jottings concerning tribes from Cabinda to the far south of Angola are too vague to be used in ethnological work with confidence. There are many points on which the work of A. Schachtzabel should be consulted for the purpose of making comparisons with my own observations. The chief of these are a note and photograph on village construction (p. 130); the musical bow (p. 32); the loom (plate VI) ; transmigration of the soul of a chief into an animal (p. 51) ; and the game of mancala (p. 52). My observations of initiation at Katoko, Ngalangi, and Cangamba agree well with that of Schacht- zabel at Katoko, but I was able to obtain more detail and more numerous photographs of the ritual of initiation. The spinning of cotton (p. 143) is exactly the process so frequently witnessed now at Elende. Iron-working (p. Ill) appears to be comparable with the technique at Elende, but Schachtzabel seems to have missed the ritual. I was not so fortunate as this author in finding the old type of iron-smelting furnace in use. The work of Statham is devoted primarily to hunting and descrip- tions of plant and animal life. The chapter given to a casual descrip- tion of people among whom he passed is not useful as ethnology; but the book is of service in providing a background of natural history. Information on the tsetse fly in Angola (p. 294) assists ethnological work by showing the determining effect of this biological factor. Presence of the fly is prohibitive to cattle-keeping, therefore cattle are not kept along the river courses which are infested with this pest. Alexander Barns made no pretence of writing more than a travel book; nevertheless he publishes photographs of ethnological value, gives useful historical summaries, and deals with economic problems of production and transportation. Tucker's book "Drums in the Darkness," though written to interest the American public in mission work, contains many ethno- logical statements. The chief of these relate to the structure of the compound (p. 37) ; the dress of women (p. 39) ; naming of twins and triplets; and other items which agree well with data from the district in which I worked. There are notes on drum signals (p. 74) and cannibalism (p. 77). Apparently drum signals were in use at Bine" half a century ago. Forty years ago a slave was killed and eaten at the installation ceremonies for a new king, a point that was noted Historical Sources 125 in the works of Battell, Cavazzi, and other early writers. A descrip- tion of the whipping of boys at initiation (p. 99) agrees with my own records, but details of the period and place are not given by Tucker. Questioning a corpse (p. 102) is a ceremony I have witnessed and described, but the instance referred to by Tucker relates to the interrogation of the corpse of a chief with regard to his choice of a successor; there are laws of succession, but these may be waived. The ceremony described in my monograph was conducted to discover the cause of death and not to determine succession. If Tucker has any detailed information with regard to puberty ceremonies for girls his reticence is regrettable (p. 142). Tucker says that suicide among women is common (p. 143). A mention of ocisunji, a feast for spirits at which meat is offered to idols, is interesting (p. 168) but details are lacking, and the use of the word idols is ambiguous. The information respecting use of charms is corroborative of my own observations in two other centers, Elende and Ngalangi. The most recent publication on the ethnology of Angola is that of F. and W. Jaspert of the Stadtisches Volkermuseum, Frankfort, 1930. Their journey was undertaken primarily to make collections for a museum, but linguistic and ethnological information was recorded among several tribes, notably the Kimbundu, Vachokue, Luchazi, Luimba, and Kusongo. The farthest point attained in a southerly direction was just north of Kipungo, and the general line of march was from Benguela to the northeast, into Lunda. There is very little overlapping in the work of the Jasperts and myself. My itinerary took me to the extreme south into the Vakuanyama country, and to the far east of Moxico among the Vachokue. My work is presented as a monograph on the Ovimbundu only, and of their culture I treat exclusively, with some reference to the culture contacts observed on all sides of them. The Jasperts do not give an entire section to any one tribe; but combine their information respecting the tribes in a concurrent way, under such subjects as technology, language, and art. If the sections relating to history and languages are excluded there remain a hundred pages, only one-sixth of which refer to the Ovim- bundu; but there is a difficulty in reading through the book without being confused as to the tribe and exact locality under discussion. The pages devoted to a comparative vocabulary form an important part of the work, and one which I barely touched, though I prepared an outline of Umbundu grammar and made dictaphone records of the 126 The Ovimbundu Umbundu language. The illustrations in photogravure are excellent, but the small-scale map is difficult to follow. As might be expected in a work covering an enormous tract of country, the information is of a very general kind; I have nowhere been able to check in any detail on my own observations. F. and W. Jaspert recognize that Umbundu culture is primarily based on agriculture, maize being the most important crop, but they do not admit the importance of hunting (p. 16). I saw many successful parties of Umbundu hunters, tested their precision with the bow and arrow, and was able to record the ritual connected with the initiation of a professional hunter. There is also a ceremony before the hunter sets out. It would be more accurate to say that, although Umbundu culture is based primarily on agri- culture, hunting still retains some of its earlier importance. The diagrams of houses in different parts of Angola are a useful feature of the work. I made many photographs but did not record details of planning. The masks I obtained are exactly like those pictured, but my observations were carried out a long way to the southeast of the point where the Jasperts touched the Vachokue culture. I do not understand from their description whether the writers saw an initiation ceremony. They state that boys are circumcised and girls are excised when very young, even at the age of five or six years. I observed initiation camps and dances at three centers, Ngalangi, Katoko, and Cangamba; the last named is the main center of Vachokue culture. The male novices varied in age from twelve to sixteen years hence they were older than the novices mentioned by the Jasperts. I was informed that there were no excision operations for girls, though at Ngalangi, Vanyemba girls were secluded in the bush during initiation rites lasting for a month. Bibliographical references to the research of H. Baumann will be found useful for comparative study. His detailed record of initiation among the Vachokue is valuable as a check on my observation, as we worked independently in areas separated by several hundred miles. Consultation of historical sources gives the following answers to questions which were asked as an introduction to this chapter. On the grounds of tribal tradition, historical evidence, and cul- tural affinities, the original home of the Ovimbundu was likely to have been in the southwest Congo. The Ovimbundu undoubtedly possess important cultural traits that have been characteristic of the southern Congo region from the earliest time for which a record exists. Historical Sources 127 In northern Angola the Ovimbundu could not have had other than a warlike existence, which trained them in military tactics and the building up of an aggressive confederacy. In this they were aided by contact with the Portuguese, who supplied guns and powder in exchange for slaves and ivory from the interior. This accumulated wealth further stimulated the building up of Umbundu tribal life. Introduction of maize by the Portuguese gave the Ovimbundu a knowledge of this grain, which later became their staple wealth and food supply. The cultural pattern of the northern Congo was the same in the year 1600 as it is today, and consideration of the ethnology of the Congo region reveals numerous similarities with Umbundu culture of the present time. Writings of explorers in Angola from A.D. 1800 to the present day bring out many points which are in agreement with my field observa- tions. There are no discrepancies which would make me question the validity of information given by my informants. In early records, Battell's observation of cattle in the hinterland of Benguela (1600), is important in showing that at such an early date the Ovimbundu had access to cattle when raiding from their home in the Benguela Highlands. From these fundamentals of geography and history the inquiry turns to a detailed account of my observations among the Ovimbundu and surrounding tribes. IV. PHYSICAL APPEARANCE Among the Ovimbundu there are two main physical types. One of these is of brown skin color and slender build. In this type the calves and thighs are poorly developed, and the chest girth is slender in keeping with the general development. In general the physical characteristics are distinctly different from those of a typical West African Negro as represented by the Kru and the Ibo tribes. The Ovimbundu are Bantu Negroes who possibly result from a crossing of Hamites and true Negroes, a hypothesis which would account for both the light-colored slim type and the shorter, darker, more sturdy type. But E. Torday (Herbert Spencer, Descriptive Sociology of African Races, London, 1930, Preface, p. iii) thinks that differences of physique among Sudanic and Bantu Negroes are sufficiently accounted for by isolations and environmental differences. Torday denies the need for postulating an infusion of Hamitic blood, but his opinion is opposed to that which has found general acceptance. The darker Ovimbundu with more facial hair may represent an infusion of Vachokue blood because there was warfare between the Ovimbundu and the Vachokue of eastern Angola with the result that some of the latter were taken as slaves. Types of the Ovim- bundu are shown (Plates LII-LIV). The Vasele, an Umbundu-speaking people of west-central Angola, form a linguistic and cultural pocket because of their isolation in rugged country. Their physique shows no pronounced differences from that of the Ovimbundu, but the methods of scarification, tooth mutilation, and personal ornament are in distinct contrast with those of the Ovimbundu. The Ovimbundu have only a V-shaped notch in the two upper central incisors, whereas the Vasele chip all their teeth to points (Plate LXXVI, Figs. 1, 2). The Luvando of southwest Angola resemble the Ovimbundu in general physique, though the great difference in hairdressing and personal ornament is likely to give a contrary impression (Plate LIX, Figs. 1, 2). The Vaheneca are particularly well developed (Plate LXII, Figs. 1, 2). Not until Mongua is reached does one find a people who are distinctly different from the Ovimbundu in all respects. The Vakuanyama of Mongua are noticeably tall and slim, much taller and slimmer than the Ovimbundu, while their physiognomy is more refined (Plates LXIII; LXVIII, Fig. 2). The Vakuanyama are a pastoral tribe whereas the Ovimbundu are principally agricultural. 128 Physical Appearance 129 The picture of types gathered at Ngalangi (Plate LXXXIX, Fig. 3) shows, reading from left to right along the back row, then along the front row in the same direction: an Ocimbundu; an Oci- vokue; a man of the Vangangella; an Uluchazi woman; a Lunda man; and two types of the Vangangella. The M'Bunda man and woman photographed at Cangamba display characteristic deformation of the upper central incisors (Plate XCI, Figs. 1, 2). Photographs of a Mussurongo man and two women, taken about a hundred miles east of Malange, show no great difference from the Ovimbundu except in their shorter stature (Plates LXX, Figs. 1, 2; LXXI, Figs. 1, 2). The Bushman (Plate LXIX, Figs. 1, 2) shows a type found wandering in small bands in the south of Angola. These photographs were taken at Cassanga. The dress and personal ornaments of the tribes dwelling to the south of the Ovimbundu (Plates LIX-LXVIII) are entirely different from the clothing and decorative styles observed in the areas occupied by the Ovimbundu. This statement is true with regard to bodily covering, ornaments, tooth mutilation, use of pigment, scarification, and hairdressing. In respect of all these factors, males and females of the Ovimbundu have distinctive patterns which do not appear to have influenced, or to have been influenced by the styles around them. In working southward from Elende I passed through typical Umbundu cultures until the vicinity of Kipungo was reached. At this place the change in physical appearance, ornaments, and hair- dressing was remarkable both for its abruptness and distinctiveness. But the change from agricultural to pastoral pursuits is gradual. A comparison of photographs indicates at once the truth of this statement, which can be further illustrated by a detailed description of the ornaments observed from Kipungo to Mongua. By far the most important of these is the circular omba shell made from the basal part of a gastropod shell of the genus Conns Linn; and allied forms. These are highly prized, not because of any intrinsic worth, but on account of strong sentiment arising from their bequest, which is usually in the female line. I have, however, seen a few males of the Vakuanyama wearing these shells. For the old omba shells monetary offers equal to the earnings of a woman for a period of six months were made, but without success. One woman wavered somewhat, but finally decided that she dared not return to her home without her omba shells. From a Portuguese trader I bought, for a small sum, omba shells showing stages in manufacture. 130 The Ovimbundu The new ornaments had acquired no sentimental value, therefore their price was moderate. Omba shells are to be seen in use from Kipungo southward through Huila, among the Luvando, and among the Vakuanyama of southern Angola. In the places mentioned one may judge the social status of a woman by the number of omba shells she wears. A principal wife is usually well supplied with these ornaments. In this southern journey it was noticeable that there was an increasing use of red pigment which is lavishly employed for smearing every kind of ornament, the body, and the hair. Leather belts and skirts of Vakuanyama women are thickly coated with red pigment which is invariably mixed with grease. The red powder is prepared by desiccating a red wood called tukula by the Vakuanyama, a name which is used through Angola and the southern Congo area. From Kipungo southward through the Vakuanyama country there is a notable absence of decorative wooden hair combs. On the contrary, delicately carved combs are used by both men and women of the Ovimbundu. The Vasele make such combs, but by far the best examples are made and used by the Vachokue tribe of eastern Angola. Here the decorative design usually includes a well-carved human figure at the top of the comb. Among the Vakuanyama, necklaces of disks made from ostrich eggshell are worn by the women only. A woman of importance has a dozen loops of such necklaces, each loop being about 125 cm long. These necklaces are so greatly esteemed that only after much persua- sion can a woman be induced to part with a single link. Perhaps, as is the case with omba shell, there is more than the intrinsic value to be considered. One point is constantly noticed in considering social status and ornament. All the wives of a wealthy man, especially among the Vakuanyama, are made to advertise their husband's posi- tion by the profusion of their ornaments and the quality of their leather skirts and belts. There is, in addition to the necklaces of ostrich-eggshell beads, a highly prized necklace made from small perforated disks of shell having a diameter of about a centimeter. The Vakuanyama women smear a necklace of this kind with grease and tukula powder; the value of a necklace 125 cm long is equivalent to that of an ox. In describing these necklaces there is interest in noting that they are traded to the south of Angola from places six hundred miles to the north; therefore their value is to some extent dependent on rarity and distant origin. Physical Appearance 131 Young unmarried girls of the Luvando tribe wear a large number of leg-bands which extend from the ankles to the knees as a sign that puberty has not been reached. These leg-bands are somewhat roughly twisted from fibrous roots and twigs from which the cortex has been removed. Collars of tough, elastic, cane-like substance are worn by Luvando women, and so numerous are these that the neck is entirely covered. These cane neck-bands are ornamented with burned, incised, geometrical patterns, and, in keeping with other ornaments, are thickly smeared with grease and red powder from tukula wood (Plate LIX, Figs. 1, 2). Women of the Vakipungo and Vakuanyama wear heavy coils of brass or copper wire on their forearms. In southern Angola women wear bracelets of twisted wire which are identical with those worn by Zulu women. Bracelets of beaten trade brass, ornamented with incised geometrical designs, are worn by women of the Ovimbundu, Vakuanyama, Luvando, and Vakipungo tribes. Ovimbundu women now depend largely on trade goods for personal ornament. Ovimbundu men and women are dressed in trade cloth. Unmar- ried girls wear one piece of cloth which hangs from the armpits to the knees. A married woman drapes herself with two pieces of cloth, a skirt hanging from her girdle, and an upper piece so folded as to hold her baby tightly to her back. Men wear a single piece of cloth as a skirt; the upper part of the body is bare (Plates XLIX-LI). From the region of Kipungo to the southern border, clothing is of leather. The Vachokue wear trade cloth or bark cloth. Hairdressing is of many styles. Ovimbundu women braid their hair neatly in strands across their foreheads, and small blue and white trade beads are used to decorate the braids. The hairdressing of the Ovimbundu women is different from that in any other part of Angola. At an early age the hair is trained into two long loops at the back of the head. Then these are covered with black cloth which is bound tightly. The two loops are afterwards studded with brass-headed tacks obtained from a store (Plate XL, Fig. 3). Luvando women in the region of Kipungo dress the hair to form a large triangular projection from the back of the head. Girls of the Vaheneca tribe near Huila mass the hair with clay to form large "cock's combs." Women of Gambos pass each small plait of hair through hollow reeds. Humbe women do not redden or grease their hair which is massed into three high ridges on the top, while at the sides there are hornlike projections. Vakuanyama women dress their hair with grease and tukula powder. A principal wife builds up her 132 The Ovimbundu hair into five high cones. Vachokue women mass their hair into separate balls shaped and held by clay and red coloring matter. I know of no hairdressing for Vachokue men, but ornamental wooden combs are sometimes used. In only one place have I seen a nose pin worn, namely, the Esele country of Vila Nova de Selles. The fashion is out of date, but women of only twenty-five years of age have the septum of the nose bored; evidently the custom has not been obsolete for a long period (Plate LXXV, Fig. 1). The most popular European importations are blue cloth with white spots, metal hair combs, beads, and bright metal crosses bearing a figure of the Virgin Mary with the infant Jesus. This ornament has penetrated to districts far away from Christian missions, and its wide dispersal shows that a newly introduced and attractive ornament from a foreign source may readily be accepted by tribes of different cultures. There are interesting points of psychology in the attitude of the Ovimbundu toward European importations. Only a few patterns of cloth are favored, and there is no sale for any other design. Some designs are thought to be appropriate for young girls while others are favored by older women. The same may be said of colored beads, for whereas both blue and white beads are the usual decoration for young women and immature girls, red beads are worn chiefly by the elderly women. A consideration of personal ornament establishes the general truth that in scarification, hairdressing, tooth mutilation, the use of red powder and grease, the anointing of the hair with palm oil, and the wearing of trade cloth, leather, or bark cloth, there is little exchange of styles. There are diffusions in language and other cultural traits, but decorative elements which have for a long period been regarded as distinctive of tribal life are rigidly preserved. V. ECONOMIC LIFE The words "economic life" are here used with a wide connotation including nature lore, food supply, trade, transport, and industries; all these are combined to form a foundation for every aspect of the social life of the tribe. The truth of this is realized if one pays attention to the rites connected with occupations. For the hunter there is special training, ritual, and a peculiar mode of burial. Final ceremonies in the initia- tion of a young blacksmith are associated with sacred acts such as sacrifice of animals and the sprinkling of the tools with their blood. Even the simple occupation of pounding corn requires that the rock shall be dedicated to this purpose by sprinkling the blood of a chicken on the surface. Similarly, a clay pit has to be consecrated before the raw material may be taken for making pottery. A caravan journey is not merely a commercial undertaking; the accompanying medicine-man carries a wooden figure which he consults with regard to the route (Plate XXI, Fig. 5). Division of labor according to sex is one of the most important principles involved in the economic life, more particularly in occupa- tional groupings, which are strictly observed. Moreover, within any one activity, such as house-building, there are tasks for men only, while other parts of the work are performed exclusively by women or children. The study of industries is of great importance when a comparison of Umbundu and adjacent cultures is being made. This is particu- larly true in the instance of wood-carving, an occupation yielding highly specialized products some of which are connected with religious belief and ritual. All artifacts are valuable as criteria of cultural contacts, though their reliability as evidence of trait diffusion natu- rally varies with their degree of complexity. Study of native indus- tries in relation to European contacts illustrates a cultural process which may result in acceptance, rejection, or ingenious adaptation of new ideas. That a study of the economic life of a tribe is not merely a record- ing of material processes and artifacts, is illustrated by observation of the treatment of cattle. This pastoral pursuit naturally falls under a heading "Domestic Animals," but the ideas associated with cattle, including funeral feasts, use of horns on graves, and the wrapping of the royal corpse in oxhide, lead directly into important matters of belief and ritual. 133 134 The Ovimbundu Nature lore of the Ovimbundu rightly forms an introduction to other aspects of economic life, because there is no better introduction to ethnological research in the field than that of associating with the people in daily occupations connected with the food supply and industries. In this way a field worker realizes that observation on the part of hunters and food gatherers is fundamental, not merely to economic and social life, but to the growth of language and folklore. Observation and experiment have led to the selection of many kinds of timber, each having one or more specific uses. Collection of plants is connected with the making of dyes for baskets, a varnish for pots, and a pharmacopoeia for the medicine-man. Minute observation of the habits of animals, primarily carried out to ensure successful hunting, is clearly reflected in the growth of vocabulary, and likewise in the realism and humor of folk tales and proverbs (chapter VIII). This chapter is concerned chiefly with a presentation of factual material whose psychological and cultural bearing is more fully dealt with in chapters X-XII. Nature Lore The Ovimbundu are keen observers whose knowledge of the natural history of plants and animals is comprehensive. Almost any boy of twelve years of age is able to give the information detailed here. Some of the birds have been identified by Mr. Rudyerd Boulton, Department of Zoology, Field Museum. Ombo. Ostrich. The knowledge of this bird must come from the south of Angola. In traveling south I first saw ostriches at Humbe. These were domesticated birds. Epanda. Wattled Crane (Bugeranus carunculatus). The informant said, "Epanda is a big bird which has long legs and a long neck." Epumumu. This is the Ground Hornbill (Bucorvus cafer) whose black wings are tipped with white. I have observed them from Ganda to Vakuanyama country. They take to flight slowly after hopping heavily for a few paces. Ocamukongo. From the Umbundu word ukongo, a hunter. This bird lives on small buck and rabbits. Etokailo. This is the smaller bustard called by the Boers knorhaan. Ohanga. The Guinea Fowl (Numida meleagris). Onjava. Spur-winged Goose (Plectropterus gambensis). Ongonga. An eagle. The word is used generally for a large bird of prey. Ohokohoko. The Bateleur Eagle (Terathopius ecaudatus). The bird is almost without a tail. This bird is not predatory. It kills snakes but does not take chickens or other small animals. Ociselele. A kite, well known because it preys on chickens. Etalahanga. A hawk which waits in the trees then volplanes on its prey. Okapamba. A small hawk which preys on chickens. Enyamahuti. A hawk. Ocikuamanga. The White-breasted Crow (Corvus albus). Onguali (ua-wa). This is the Red-necked Partridge (Pternistis afer). Economic Life 135 Ekalanga. A species of francolin without red coloring on the legs. Esuvi. This is a bird which comes out at night. It can catch spirits and make them die a second death. Onjimbi. This nocturnal owl (Bubo maculosus) is thought to be very dan- gerous. People who hear the cry of this bird are frightened because the noise is the sound of death. Kacukucuku. The Barn Owl (Tyto alba affinis). Ongongayulombo. An eagle which eats small buck, pigs, and rabbits. Ekuti. Is a Red-eyed Dove (Streptopelia semitorquata) which says "oo— oo— oo" very sharply all day. Onende. A dove (Streptopelia capicola) which makes the same noise as ekuti but on a higher note. Onduva. This bird is of great importance to the Ovimbundu, because the feathers are used for decorating the head of a dead king. A medicine-man sometimes uses them for decoration when he is performing. Ondonga. Is a little bird, which, like onduva, belongs to the royal family of birds. Ukuku. A water bird. Ocisandombunji. This is a bird which feeds on white ants. Epandacokocoko. This is a bird (Geocichla litsipsirupa) whose cry tells the people of a village to make the guest house ready as strangers are coming. Ocilonga. Is a beautiful yellow bird, an oriole (Oriolus monarchus angolensis) . Etua, singular; ovatua, plural. A species of bustard. Etioko. A small bird of the plains (Anthus leucophrys). Okakelekele. This is the Spur-wing Plover (Hoplopterus armatus), a large bird which lives on flat land near rivers. Ocikandi. An unidentified bird. My informant said, "Ocikandi can cry in the same way as any other bird." Okakongonyala. This is a bird somewhat like a pigeon but larger. It has long legs and is able to run well. Undolo. A small spotted bird which always looks in the direction of the sun. Omiapia. This is a swallow. Ocikungumiapia. My informant said, "This bird is like omiapia but larger." Omuipui. "He has a crest on his head and when he sings rain is coming." Esunguaguluve. This bird has a long beak. It lays eggs on the ground. The bird is as big as a man's fist. Omanula. A woodpecker (family Picidae) . Okangongo. Has a white breast spotted with black; the head is blue. Ocinjonjo. This is a little bird having a long beak which is used to probe into flowers. Koseselekete. A little black bird which makes a small fine nest. Ocituku. A river bird. These birds go about in flocks. Ongombo. A small yellow bird. Epilili. A yellow bird rather larger than ongombo. This bird is kept in cages very cleverly made of reeds by small boys. The birds are caught by smear- ing a sticky substance on the branches of trees. Ocikenge. Agapornis roseicollis. A small green parrot. Okalusondonjovo. Is to be seen at the time of onjovo (spring). These birds fly in flocks. A description of bird-calls illustrates a method of reasoning by analogy. This results in a transference of human thoughts and emotions to animals, a process which gives animal fables their strong 136 The Ovimbundu appeal. For example, the pigeon says, Tu kolela oku iva ("We believe in stealing"), and the following are further instances of the same kind. Ocinganja is a bird which has several calls. Early in the year the female says as she looks at the newly hatched brood, "This year I have borne white children." Later the little birds grow black feathers. Then the mother cries, "Each year disappointment." Sometimes the female says to the cock bird, "A stick has stuck in my eye. I wonder whether it will make a growth there." The bird is probably the Black Flycatcher (Melaenornis pammelaina) . Epandacokocoko says, "Where will the guests stay? Where? Where? Where?" This is an unfailing intimation that strangers will visit the village. Omanula. This is another bird which announces the approach of strangers by saying, "Akombe! Akombe! Akombe!" ("Guests! Guests! Guests!") Two birds are supposed to carry on a dialogue. Sankanjuele says, "He who has eaten should leave the rest for the important ones," meaning himself. Ondonga answers, "Do you mean me? The way you scold hurts me to my heart." Ungolombia. The male and female birds ungolombia are about to cross a stream. The female says, "I am wearing four yards of cloth," meaning that she will get wet. The husband says, three times, "If you are wearing four yards of cloth, why do you not cross at the source of the river?" Ombovo says, "When I lay my eggs on the ground, the white ants destroy them." The call continues, "When I lay them up high, kalupamba steals them." Ombovo cries, "Such hardship, goodness gracious me!" The exclama- tion is, "A mat we," literally, "O my mother!" Kacukuku the Barn Owl (Typo alba) and his mate speak together. The female says, "Cimuku, do you eat rats?" He answers, "I do not eat them, they have tails." Ungungu says, "No big animal lays eggs, so the crocodile must be a bird, too." Ocimbamba (night hawk) may be heard on moonlight nights saying, "O lion, here are the people." Ekuti says, "My child is gone to Koputu." Koputu, in Umbundu, means a far off place. Possibly Koputu is a corruption of Oporto. This bird makes a monotonous and continuous cry from sunrise to sunset. Epumumu (hornbills) are large black birds. The female says, "I'm going, I'm going, I'm going to our village." The male replies, "Don't go, don't go, the rain has come; let us plant." Katendipanga. This bird is quiet until September, the month in which rain begins to fall. Then the bird says, "Save! Save! Save!" He means this as a warning to people who eat their corn instead of saving some for seed. The seed should be sown in October when the rains have begun. The derivation of the name of this bird is important. Okutenda, to count; ovipanga, the rows of corn in a field. Kalusundanjovo. The female bird says, "Let's throw away the big drum." The male answers, "When we have thrown it away, what shall we do for dances?" Observations of the weather are of importance in fixing dates for sowing and reaping; such dates are the base of time reckoning. Rainfall is important where cattle are kept, and a knowledge of the stars is of service to hunters and caravan leaders when following unfamiliar routes. The weather is thought to be controlled to some extent by the ocimbanda ("medicine-man") and his performances. The rain- Economic Life 137 maker's dance seen at Ngongo, a village of Ngalangi in east-central Angola, will be described in dealing with the ocimbanda (chapter IX). No man thinks that the course of the sun can be altered, but there seems to be a hope that the setting of the sun may be delayed. A man who is likely to be overtaken by darkness breaks a piece from an ant hill. This fragment is placed in the forked branch of a tree while the supplicant says, "Sun, wait a little while for me." Months are reckoned by observation of the moon. The new moon is osai ("moon") yokaliye ("new"). There is no word for half moon. The phrase for full moon is osai ya tunga ohumba. Ohumba means "basket"; therefore the idea appears to be that the full moon is round like the big basket used for field work. A star is called olumbungvlulu. To describe a shooting star, the word luenda ("it goes") is added. The large bright morning star is tanda. Another bright star is tielele. Three bright stars in a line are the hunter, the dog, and the quarry. The three stars are ukongo ("hunter"), lombua ("the dog"), locinyama ("the animal"). An eclipse of the sun is uteke vutanya. The former word means "night," the latter word means "daylight." Hence the meaning is "night in daylight." Small mammals are captured in cane traps or stunned with blunt wooden arrows. The pursuit of such animals is a pastime for boys who thereby receive their early training in hunting. Rats are used as food after being boiled and roasted. This fact accounts for interest in the following small mammals. Umbili. This is a big black river rat, the largest of the water rats. Ocifelefele. A big gray water rat (Cricetomys?) , not so large as umbili. Kalene. A river rat something like the preceding one in appearance. Kalene can stay in the water for a longer period than that endured by ocifelefele. Epeke. This is a bush rat which is light gray in color. Ekolongonjo. This rat stays in the bush. It builds a nest in an ant hill from which it comes out only at night. The color is gray. Elima (Epomophorus) . This is the name given to the fruit bat. The word elima means "not one thing and not another." Thus elima is the word applied to a mulatto; he is not a Negro, not a white man. Osili (Rhabdomys). This is a large light brown mouse having white stripes on his back. Ongenge. A mouse. Epengue. A black rat. Oeipili. A mouse with a long thin nose, probably a shrew. Osinge. This is a very fat mouse. Nakalongaka. This very small mouse makes a hole around which he piles grass to conceal the entrance. Onjomboloka (Lemniscomys). A mouse with a stripe on his back. Kandoti (Dendromus). A very little mouse something like nakalongaka. 138 The Ovtmbundu Observation of reptiles is a necessary self-protection as there are many poisonous snakes. The flesh of the python is eaten. Snake- skin and lizard-skin are occasionally used; for example, in covering a round, hard fruit in order to make a ball for playing a game. When collecting lizards and snakes I found that each kind had a well-known name, though there was occasional disagreement among the men consulted. Ocivangoko is a lizard (Agama planiceps) about ten inches long whose tail is covered with sharp spines. This reptile has colors of bright blue, red, and orange. Ovangu is a large spinous gray lizard (Agama atricoU). Ekangala (perrhosaurus nigrolineatus) is a brilliantly colored lizard which burrows deeply into the ground. Olutanjila is a long slender green snake which hides in trees to capture birds. Ombandanjila is a long gray-backed snake with a light green belly. The generic name for snake is onyoha. The python is omoma. The chameleon (donatio) seems to be feared, as the men and boys refuse to touch a dead one but always move it with sticks. There is a proverb to the effect that the chameleon though slow always gets there. The refusal to touch even dead reptiles is due to. an exaggerated fear of being bitten. I could find no other reason. Trees are of importance in connection with building houses and wood-carving. The qualities of different timbers are well known to the Ovimbundu. For the main part, the collection of medicinal plants and their uses will be described in connection with the medicine-man. The following are the most common trees which are of economic importance. Usia (pronounced oosha). This tree has an edible fruit the size of a walnut. The kernel is valued as a food. Ombula. This tree provides wood which burns readily. The small skin- covered stools to be found in every hut are often made from this wood. Ukengo. The fruit has a hard rind. Owindo. This tree has a small acid fruit from which a medicine is made for the cure of painful menstruation. Usilosilo. The leaf is compound and palmate. The fruit is black. Usiambiambia. Bears a little red, oval fruit. UsoU. Has a large red fruit. Uaombo. Grows near streams. It has a fruit like that of okulakula, but smaller. Uhuliungu. Has a fruit like the berry of a coffee plant. The fruit is used for making mucilage which is used for capturing small birds. Omanda. When this tree is small, the wood is springy, and is therefore useful for making bows. The wood of the older trees is burned for preparing charcoal which is used in the blacksmith's fire. Onundu. This is an erect tree having no branches on the lower part. The wood is extensively used in building native huts. Economic Life 139 Omue. A large tree having clusters of small white flowers which are visited by bees on account of their content of honey. The tree yields a hard wood from which charcoal is prepared. The bark and leaves yield a pigment which is used for dyeing cloth a yellowish brown. O&ui is valuable because it gives a hard wood used for the corner posts of houses. White ants do not attack this wood, which is therefore useful for making the uprights on which granaries are erected. Okapelangalo. A tree from which planks of hard red wood are obtained. These are used for making doors in village fences. Osasa or ekenge or usamba. These trees have small compound leaves which are very similar. The bark of the latter two is used as rope for binding the uprights and the crosspieces in the framework of native houses. Ekenge and usamba also yield a bark which is beaten into bark cloth in the Ngalangi district. Ociyeko. The bark is used for binding posts, also for the fabrication of bark cloth. Ungolo. The roots of this tree yield a dye for cloth. The leaves are said to have a value for curing sore eyes. The mother of a child afflicted with sore eyes chews the leaves, then spits into the child's eyes. Ongaye. Yields a wood used for making pestles and pounding sticks. Omako ("iron wood"). This tree has a hard wood which is used for the same purpose. Onjunge. Gives a wood used in the making of houses, doors, and beehives. Omone. A large tree which gives planks for building purposes. Uvanje. Yields a useful red timber. Ulondangandu. A tree with very rough bark. The word ongandu means a crocodile. This is the tree which even a crocodile can climb. Onganja. A tree which yields a fruit having a value as a purgative. The antelope is said to be fond of the fruits of this tree. Ocikumbeolemba. Gives a resinous fluid which is used in the preparation of lime for snaring birds. Ulemba. This word is derived from the word ulembo, meaning shade. The ombala (native capital) of Ngalangi is surrounded by such trees. Omia. A tree which produces yellow flowers in September. The fruit is not edible but oil is made from it. Ohuku. A tree having fragrant flowers like those of honeysuckle. The thick bark is used in making mats. Osese. A soft wood which is easily whittled with a knife. The figures from Bailundu (Plate XXI, Fig. 5) are often carved from this wood. Umbolombolo. A soft wood which is not very strong. Umbangalunda. A small tree which produces bright red fruits. These are used by women for the manufacture of bead necklaces. Ocilavi. The wood is used for the heads of arrows for shooting birds. This timber is used in the building of pigpens. The branches have projections which are said to guard (pkulava, to guard) the occupants of the pen. Itata. From the roots of this tree a medicine for pulmonary complaints is made. Ukua. This is the baobab, whose habitat is the dry regions. Some of the trees have enormous girth. They are leafless for a great part of the year. The long fruits make gourds. The seeds in the fruits are bitter. Utuotuo (pronounced ootwdtwd). From the wood the Ovimbundu make wooden platters and spoons used for serving mush from the large cooking pot. Onjiliti. This tree yields a hard red wood which takes a lustrous polish. For this reason the wood is employed for making ornamental sticks and clubs. Upondanjamba. This small tree has roots which girls use for making ankle- bands. 140 The Ovimbundu Okalaluluka. This tree has leaves which are used in treating a skin disease. Uvendanguluve. This small tree, only three feet in height, gives straight twigs which are used in making arrow shafts. Food Supply collecting and hunting Collecting of natural products which serve as food substances is chiefly in the hands of women and children, though an exception has to be made in the instance of honey, which is gathered by men and boys. Boys diligently search for nests, noting their location so as to be able to visit them again when the fledglings are large enough to serve as food. Large numbers of women and children may be seen gathering caterpillars in gourds. The insides of the caterpillars are squeezed into boiling water to make soup. When a cloud of locusts appears, as in 1925, the creatures are gathered. They are sometimes fried, or they may be boiled in water, dried, and preserved with salt in earthenware pots. A number of miscellaneous items of the food supply were men- tioned in connection with names of trees whose fruits are gathered. Boys engaged in food-gathering usually carry small bows and blunt wooden arrows {pcilavi). One type of bird arrow is fixed to the bowstring. The forward end of the arrow is split so that it may contain small stones that are ejected when the string is released. From the wild fig tree mucilage is obtained and this is boiled until it forms a thick paste which is smeared on the boughs of trees. Some of the small birds captured in this way are eaten, others are kept in wicker cages made by children. In all parts of Angola large cylindrical beehives may be seen fixed high in the trees (Plate XC, Fig. 2). Two types of hive have been noted in particular. In the Elende district a hive is made by opposing two half cylinders of wood each about three feet long, so forming a hive which has a diameter of one foot. The ends are covered, with the exception of a small round hole. The whole struc- ture is bound round with grass which is kept in position by lashings of bark. In the neighborhood of Cassanga a difference in the structure of the hives was noticed. This type of hive is made from a cylinder of strong reddish bark, the edges of which are fastened together with stout wooden pegs. The dimensions are the same as for the hive used in Elende, but the hive of bark is uncovered. In the Elende district honey of wild bees is removed from the hives in the months of August and December. One man ascends Economic Life 141 the tree in order to lower the hive with a long rope of bark or plaited fiber, while beneath the tree men are prepared to take the hive, which is opened over a smoky fire. The men wear no protection, consequently they are badly stung. Boys are encouraged to help, and those who run away receive no honey. Honey may be eaten alone or with manioc. No drink is made from honey only, but ochasa is the name given to beer with honey in it. Ovingundu is a drink made from pounded corn which has been soaked in water to which a little honey has been added. The drink is allowed to remain untouched over night; thus it becomes sweet and is mildly intoxicating. Wax is a very important item of trade. In the remote places natives bring to small trading posts balls of wax which are about two pounds in weight. These in former days were a standard of exchange in terms of which other values could be measured. These balls of wax are made into large cakes for foreign export. In the Esele country a fiber strainer is used for cleaning the wax. Honey is sometimes dried in very large baskets which are three feet in diameter and two inches deep. These baskets were not observed among the Ovimbundu, but they are used in the region of Cassanga in southern Angola. The honey of wild bees when eaten in the comb is palatable; it would be more so if one could disregard the presence of numerous dead bees. The bow is the chief weapon of the hunter. The release of the arrow is made with the index and middle fingers (Plate XXXIX, Fig. 1). I have observed this method among the Ovimbundu of the Benguela Highlands, in the region of Kipungo, in the far south among the Vakuanyama, and among the Vasele of the Novo Redondo hinterland. Arrows differ considerably in pattern as the illustrations show (Plate XVII, Figs. 1-9); the Ovimbundu have arrow-points of excellent workmanship. The manufacture of these arrow-points is one of the most skilled occupations of the blacksmith. Each man makes his own shafts and feathers them. The arrows of the Vasele have leaf-shaped iron points; so also have those used by the Vakuanyama. Among the Ovimbundu are specialists who make bows from elastic woods called usia and osambia. The bow itself is ohonji and the arrow is usongo. At the third shot I saw a young Ocimbundu boy split a cane which was placed upright at a distance of thirty- three feet. The throwing club (ohunya) is used for killing small game such as hares. 142 The Ovimbundu The only spear (unga) that I have seen is made entirely of iron. The shaft is covered with the tail of an ox to which the tuft of hair remains attached. This is the spear formerly used in warfare. The distribution is wide. Such spears were purchased from the Ovim- bundu of Elende and Bailundu and also from the Vakuanyama living in the far south of Angola, but I do not think that the Ovim- bundu make these spears, which are probably traded from the south. A hunter is considered exceptionally fortunate if he possesses an old muzzle-loading gun (uta). I have seen only two hunters who owned such a weapon. In one instance the barrel was bound to the stock with hide thongs, while the woodwork was decorated with brass tacks. Powder and fragments of metal are carried in a leather pouch which is attached to a broad, leather waist belt (Plate XIII, Fig. 6). There is certainly a feeling among hunters that the sale of a well-tried weapon will be followed by bad luck. The gun is sometimes fixed to form a trap in such a way that an antelope may tread on the string and so discharge the gun. Some- times a piece of meat is attached to a string which is fastened to the trigger. A heavy beam of wood into which an iron spike is fastened was used in districts where the hippopotamus and elephant were hunted. Such a trap was fixed over a path known to be frequented by these animals. Big game of this type is now rare. In the Esele country the following traps are in use: (1) A deep, narrow, grass-covered pit from the bottom of which sharp stakes project upward; this trap is known as okueve. (2) A simple trap consisting of four long sharp stakes which are fixed in the ground so that they incline toward a gap through which a buck is likely to jump (Plate LXXIII, Fig. 2). (3) The trap (ocisonga) for lions and leopards. This is a heavily built structure provided with a panel-like door which slides down when the entering animal releases a cord fastened to the bait (Plate XCII, Fig. 2). (4) A heavy trap triangular in form. This trap is not an enclosure, but a covering under which the animal has to go in order to reach the bait. In addition to the foregoing examples there is a trap (onjanjo) which is used for snaring antelope. The essential of the device is a loop which is bound to the end of a supple branch lightly fastened to the ground. This, however, did not come under my observation. Long, cone-shaped, cane structures are placed in the grass, which is then fired. Animals disturbed and frightened by the fire rush into the wide end of the trap, then make their way to the narrow end (Plate XV, Fig. 5). One such trap is modified to form a snare with Economic Life 143 a noose which hangs over the entrance. From Ngalangi I obtained a trap formed by suspending a heavy block of wood inside a box. The animal enters a small circular hole, passes under the heavy block and begins to nibble the grain which is strewn on the bottom of the box. Presently he releases a fine string which brings down the block in such a way that he is pinned underneath. I am informed that this trap is used at Elende, but I have not seen it there. Some hunters note the feeding and drinking places of their quarry, which is shot from a hiding place in a tree. Young animals may be run down by a hunter in open chase. Screens are not carried in front of a hunter, but he does sometimes dress in the skins of animals. His disguise is completed by wearing a tuft of the animal's hair on his head. Dogs are used for tiring out young animals in the chase, and also for catching hares. There is a tendency at the present time to improve the breed of native dog by crossing with a large hunting dog from South Africa. In the Esele country I have seen, high on a granite rock, a small cairn of stones which covered the skull of a famous hunting dog. Dogs are used most commonly when the hunt is communal (Plate XXIII, Fig. 1), but the best hunters work alone without the aid of dogs. In the general hunt women and children may take part in driving the game; often a fire is started in the grass, which is very dry in the months of June and July. A hunter who works alone may excite the curiosity of an animal by blowing through a horn of an antelope. Spider's web covers the wide end of the horn. I have observed a general hunt in which thirty men and boys participated, each carrying a bow and arrows. The party was accompanied by many dogs. The antelope which had been killed was carried on a pole slung on the shoulders of two men. There was intense excitement as the troop advanced toward their village, shouting and jumping. In another hunt of this kind muzzle-loading guns were carried. The Ovimbundu do not use nets in hunting, neither do they poison animals. Decoy animals are not used, but in the large wooden trap (ocisonga) a living goat or pig is placed. It is necessary to distinguish between hunting as a general pas- time, in which all males, and even women and children join to a certain extent, and the hunting of animals by a professional hunter. The professional hunter is usually called ukongo (less frequently enyanga) ; but there is no name for the non-professional hunter. A boy who wishes to become a professional hunter has to serve for a time with an ukongo before he himself receives this title. There is 144 The Ovimbundu an initiation feast when the training is ended. At the feast all people of the village may be present but they do not dance; only the profes- sional hunters may do so. The boy who is to be initiated must not speak or move until he "feels the spirit on his head"; then he gives meat to the people. After hunters have captured game for the feast, the blood from these animals is used to smear over the bow, arrows, and spear which have been made for the novice by his tutor. This is analogous to the initiation of the young blacksmith who receives blood-sprinkled tools made by the master blacksmith. There is in connection with the life and death of the professional hunter a certain amount of ritual and precaution. The night before setting out to hunt is a time of dancing and renewal of the imple- ments of the chase, which are kept in a house specially prepared for them. A hunter who is on the eve of departure calls in other profes- sional hunters to share the ceremony, which includes the rubbing of the bows and other implements with palm oil. A libation of beer is poured on the bows, spears, and arrows, but no medicine-man is present. Some of the bows are never used, because they are merely the symbols of the personality and prowess of dead hunters whose names they have taken. The food and cooking pots of a hunter must never be associated with those of ordinary household use. If a hunter is following the tracks of an animal he must not point with his finger as this action will drive the animal away. The correct way to point is by use of the feathered end of an arrow. The hunter must not sleep with his wife the night before setting out in quest of game. I have frequently seen in front of the hut of a hunter a number of skulls of antelope and other animals mounted on poles arranged in circular formation (Plate XXXII, Fig. 2). These skulls seem to be trophies, which are invariably taken away by the hunter if he finds a new home. There is no reason to doubt the statement that these skulls are an offering to the spirits who give good luck in hunting, because such a belief would be in harmony with the general respect for ancestral ghosts, which are thought to influence the affairs of the living. On several occasions, notably near Ganda and in the Cassonge country, I have seen the tombs of hunters. These are large structures built of slabs of granite laid with some symmetry. The rock tomb is invariably placed on the top of a commanding eminence of granite. The skulls of animals which the hunter has killed are piled on the top of the cairn, but I do not know whether these are the trophies Economic Life 145 which are fixed on poles in front of the hunter's home during his lifetime (Plate XXXII, Figs. 1, 2). There are at the funeral of a hunter special observances which will be mentioned under the heading of funeral rites. FISHING In the region of Elende there is fishing with both basket and line; a method of poisoning fish is also practised. Usually a male fishes with a bark line. Women catch fish by the poisoning method, and in addition to this they generally follow the procedure in which baskets are held or weighted in the stream. If the water flows swiftly men may take charge of the fishing operations. At times both men and women fish with nets. Husband and wife may not sleep together the night before fishing, as this is believed to make the male and female fish stay together at the bottom of the river. The fishing line consists of tough green bark which is cut into strips whose length depends on the height of the river's bank. A hole is bored through the body of a grasshopper, a worm, or a grub taken from under the bark of a tree. Through this hole is passed a short stiff piece of grass about half an inch long, to which the line is attached. The fish is caught when the crosspiece of sharp grass becomes fast in its throat. When the fisher throws the line he sings: "0 fish, come and take your good thing. Do not send the little fish to spoil the good thing. Better you come and take the good thing with all your strength." In order to make fish poison the tuberous roots of a plant are taken and soaked in water until a scum rises to the top. The solid part of the poison is not given, because it would sink and the fish which ate it would remain at the bottom of the river. Therefore only the scum of this poisonous infusion is thrown in the water. The stupefied, gasping fish remain at the surface, whereupon they are seized by women who transfer them to gourds or baskets worn around their necks. Usually poison is used only in the dry season when the rivers are shallow. Sometimes there is fishing by means of a weir (olunja) which has an opening in the middle. On the lower side of this gap a basket trap is placed. There is no fishing by torchlight. At the coast, and along the river Kwanza I have seen heavy dugout canoes in use; these were about twenty to thirty feet long and hollowed from single trees (Plate LXXII, Figs. 1, 2). At Ambrizette I noted the use of a fishing 146 The Ovimbundu spear eight feet long, the end of which consisted of ten sharp prongs of palm stem (Plate LXXIII, Fig. 1). Near Cangamba in eastern Angola, fishing in the Kwando River occupied numerous men and women of the mixed tribes in the dis- trict, namely, Vachokue, Luchazi, and Babunda. Men paddled into midstream in small bark canoes from which fishing operations were directed. The fishermen carried small conical string nets, which were attached to stakes in such a way that the openings of the nets faced upstream. Vachokue women, working in pairs, dragged baskets against the current (Plate LXXXV, Figs. 1, 2). AGRICULTURE AND COOKING Osila is the Umbundu word for the granary which stands on wooden supports (Plate XLIV, Fig. 1) ; this osila is for the restricted family, and there is one osila for every house. The Ovimbundu store their corn in bulk, but in the Esele country I noted that the cobs themselves were carefully packed. Each Ocimbundu girl culti- vates a small patch of ground, the produce of which she is at liberty to sell in order to buy brass ornaments, beads, and palm oil. In addition to maize, barley, oats, and wheat, with here and there a little rye, are occasionally grown. There are three colors of beans, red, white, and black. There is no attempt to keep the varieties separate, so they cross-fertilize freely. The Ovimbundu try to cultivate a surplus of beans and maize which they use to pay their taxes to the Portuguese, likewise to sell at the stores of traders. Corn is pounded on the rock which has been used for generations after it has been consecrated by sprinkling the blood of a chicken. Pounding begins as early as five o'clock in the morning, before sun- rise, and from that time to sunset the pounding-rock is in use. The rock is evidently regarded as a meeting place for social intercourse; it is undoubtedly the center of village gossip so far as the women are concerned (Plate XXXVI, Fig. 2). As corn alone is considered a poor food, it is sprinkled on boiling water to which beans are added. Cooking goes on from early morning to sunset over a slow fire. Children sometimes receive as their evening meal a thick plastic cake of mushed corn to which green leaves of a squash are added. There are five kinds of manioc resembling one another in general appearance, but the Ovimbundu distinguish the plants, and reserve for each what they consider to be appropriate preparation. The method varies for sweet and bitter varieties of manioc. Economic Life 147 Olungunga is not a sweet manioc, so is not eaten raw. The roots are placed to soak in a stream for three or four days before they are roasted on a fire, after which they may be eaten with impunity. As an alternative the roots may be dried in the sun; they are then pounded into meal which is scattered into boiling water so that a mush is formed. The Umbundu name for this preparation is iputa viutombo, meaning "mush of manioc." The leaves of olungunga are not soaked in water; on the contrary, they may be cooked as soon as they are gathered, but they must not be eaten when warm. There is no danger in eating them after they have been boiled and have been allowed to become cold. The leaves are served with salt or fat. All the manioc, with the exception of olungunga, is sweet. Kandona has roots which may be eaten uncooked, but both leaves and roots are sometimes cooked in water. Other varieties of manioc known as otetu, elemba, and esela are eaten in the same way as kandona. Manioc is in use all the year, but the greatest quantity is consumed in November and December, a period when the growing corn is not ripe and the storage supplies have dwindled. Sweet potatoes are plentiful all the year with the exception of the months of November and December. They are placed in the pot, without removal of the skin, and boiled for a period of twenty or thirty minutes. They are taken out, peeled, and eaten. Euro- pean potatoes are sometimes peeled and made into a mush. The variety of garden produce naturally depends on proximity to a trading post, a mission, or some other European settlement. In propagating manioc a stem is cut off from the parent plant which is about three feet high ; but probably two or three years pass before the tubers are considered large enough for use. Sweet potatoes are planted in January, but there is very little of this food available in February and March. Toward the end of March or early in April a few potatoes may be ready for consumption. Peanuts (Arachis hypogaea, Leguminosae) are planted in October by women who prepare patches of ground which are drilled with holes one inch deep and eight inches apart. One nut is placed in each hole after the shell has been removed. Above the ground small leaves appear. After the flower-stalk withers it has the peculiarity of elongating and bending down. In this way the young pod is forced underground, and the seeds mature a little way below the surface. Some natives employ irrigation by leading small channels of water from a hillside stream to a garden containing maize and 148 The Ovimbundu bananas, but one could not say that irrigation is generally practised, even when circumstances permit. The papaya and the banana are increasingly cultivated, but they are by no means generally distributed. Here and there I have seen a little sugar-cane cultivated by natives for their own use. Near the main railway natives may be seen selling their products, which include cabbages and tomatoes. Some natives are today planting the guava tree which yields sweet palatable fruits. The fruits olosia are collected from the usia tree. When ripe, the fruit is yellow, round, and about two inches in diameter. These fruits are gathered in September, and the kernels, which are about the size of walnuts, are eaten uncooked. No fruits are collected for storage. Olombula fruits ripen in October, when they are eaten raw. Each wife sends the food which she has cooked in her own kitchen to the onjango ("council house"). After carrying the food to this house of assembly, where the men meet each evening, the women return to their houses to eat alone, or with the young children. When there is a plurality of wives each has her own house and kitchen. Ngonga states that separate kitchens built outside the living houses are becoming more rare. The poor have their kitchens in the living room because they cannot provide separate structures for living and cooking. The first meal is taken between five and six o'clock in the morning, the most usual food being meal sprinkled on boiling water to form a paste which is eaten with sweet potatoes. A mush of beans is eaten at night; generally there is no meal at midday. Three pounds of cooked beans are eaten by a person for one meal. Over the cook- ing pots leaves are placed to keep in the steam, especially when the pot contains sweet potatoes. I have made a meal from the sticky, gluey paste which results from sprinkling meal on boiling water. The chief objection to this food is the unpleasant quantity of grit in the meal, owing to the fact that it is pounded on the rocks. Manioc and sugar-cane are chewed at irregular intervals of the day. The amount of meat consumed by the Ovimbundu is small in comparison with the quantity of vegetable food used. This adoption of diet of a particular kind is largely a matter of habit, and there is no good reason why meat should not form a larger proportion of the food supply. The Ovimbundu do not kill their cattle for food though they will eat the meat of oxen which have died from natural causes. Economic Life 149 The flesh of goats, sheep, and pigs might be more frequently- used if these animals were bred and cared for under some system of animal husbandry. The Ovimbundu are familiar with the preser- vation of meat by smoking and drying it, because flesh of animals killed in the chase is sometimes preserved in this way by hunters, yet the method is not widely and constantly applied in order to assure a regular supply of meat. Failure to utilize the milk of cattle and goats is another example of the neglect of useful commodities. The Vacilenge, who are near neighbors of the Ovimbundu, milk their cows as do the Ovimbundu themselves in some districts, though the practice is by no means general. Even where cows are milked, butter and cheese are not made, though the Ovimbundu know of the process, which is practised by the Vakuanyama of southern Angola. There are three kinds of beer. Ocisangua is a sweet beer which even children may drink because it is not intoxicating. Water is heated slowly in the pot but not boiled; meanwhile meal of Kafir corn or maize is added. Pounded sweet potato is strained in such a way that the liquid part goes into the beer pot; the residual mush is given to pigs. After the beer has cooled out of doors, it is trans- ferred to a large gourd and allowed to stand over night. Next day it is considered a fit drink to consume or to offer to visitors (Plate XXIII, Fig. 3). The making of an intoxicating beer ekundi proceeds as for ocisangua, but instead of adding strained liquid from sweet potatoes a root called ombundi is included in the brew. The large pot con- taining this beer is covered tightly and allowed to stand untouched for twenty-four hours, at the end of which time it is a potent drink. It is important to note that the corn is allowed to sprout in the ground before it is made into the infusion to which the root ombundi is added. Ocimbombo is a strongly intoxicating drink. In order to prepare this brew, corn is soaked for a week; then it is left in the ground for the same length of time until it has germinated. It is then pounded on the rocks and placed in large pots filled with water to which sweet meal of corn is added. Simmering over the fire is continued for two days with constant stirring. This brew differs from ekundi in the longer germination of the corn and the longer period of simmering. On the fourth day after the simmering is completed the drink is said to be ready for consumption. If the people have honey, they add some to the brew on the third day of standing. My inform- 150 The Ovimbundu ant said, "Sometimes a man who has drunk this beer will sleep on the ground all day and say nothing." Salt is a welcome gift in all parts of Angola. Native tribes appre- ciate its culinary value but show no eagerness to barter for the com- modity. At the present time salt is sold in every trader's store, but in earlier times the substance had to be obtained along caravan routes from the coast and was therefore more highly prized than it is today. The Ovimbundu realize the value of salt in the diet of cattle; therefore the animals are occasionally driven to a salt lick in the hills. The Ovimbundu do not use this salt for their own diet, possibly because the salt enjoyed by the cattle is some form of potash and not sodium chloride. The Vachokue extract salt from the leaves of a river plant by burning it to ashes which are soaked and strained. This is a common African method, but I did not hear of it among the Ovimbundu. The probability is that the Ovimbundu have always obtained salt from the coast. In connection with cooking and brewing beer, methods of making fire are of importance. Matches are coming into use among the Ovimbundu, but the necessity for them is not great as the hearth fire is not extinguished. In the center of each hut is a fireplace made of three hearth-stones over which logs of wood are placed with their ends in the fire which is kept alight by pushing the logs forward from time to time. A blaze is made by breaking off bark from the logs, placing it on the center of the fire and blowing. Fire is carried from one place to another by conveying a smoldering log. In the Esele country boys may be seen setting off at dawn to scare birds in the corn field, each carrying fire with him. Usually the children or some other members of the family sleep on mats close to the fire, which requires no attention other than a pushing forward of the logs. The Vasele make fire by the twirling method, and at Ngalangi the same procedure was witnessed. In the twirling method two different kinds of wood are used, soft wood for the base and hard wood for the twirling stick. The twirler used at Ngalangi was a piece of cane into the end of which a piece of hard white wood was secured by binding. Ngonga, my interpreter and informant, thinks that any man of the Ovimbundu could make fire by the twirling method if the necessity arose, but the performance witnessed at Ngalangi led me to doubt the truth of this statement. The operator Economic Life 151 undoubtedly knew the apparatus and the method, but he appeared to have lost the dexterity which is necessary for a rapid moving of the hands from the bottom of the twirler to the top. There was conse- quently a long period of smoldering before the flame appeared. Among the Vachokue a little wooden box of tinder, a piece of quartz, and an iron blade are carried for fire-making. From an Esele man a bag containing tinder and quartz was obtained. The making of fire ceremonially in connection with ritual and sacrifice is described among the functions of the medicine-man, because the sacred and profane uses of fire are quite distinct. Tobacco is a very important item of trade. The cultivation is a domestic industry followed usually by women but to some extent by men. In former days when the Ovimbundu traded extensively in Africa every man had a field of tobacco which he himself cultivated. At the present time each man is likely to have a mound of tobacco plants in the middle of a corn field (Plate XXIX, Fig. 1). Women do sometimes cultivate tobacco near their huts, but in this case the little plantation must be strongly fenced. Goats are numerous, and they eat the leaves of the tobacco plant with avidity. Toward the end of September, when the rains begin, tobacco seed is sown on a patch of ground a yard square to raise seedlings which are planted out in October. A few flowers only are left to produce seed. As a further effort to improve the quality of the tobacco many of the lower leaves are removed. The cutting of leaves intended for use as tobacco is done by men in the months of February and March; women and children assist if the field is large. After the midrib has been removed from each leaf, the leaves are suspended from the roof in a bundle. At the end of five days, when the leaves have turned brown, they are twisted into a long straight roll which is hung in the sun for three days. At intervals the roll is twisted in another direction. This gradually exposes all parts of the leaves, so that the drying is thorough. There are three methods of making up the rolls: ombola is an oval roll; ongalo is the round coil; ocine is the name given to tobacco which has been dried and twisted round a stick. Tobacco-pipes are varied in size and design (Plate XV, Figs. 1-3) ; those for men are larger than those used by women. A mixture of tobacco and hemp is smoked in a water-pipe made from the horn of a cow. Details of structure and ornament are given under "Wood-carving." In order to make snuff a piece of dried tobacco is slowly baked near the fire on the end of a pointed stick which is turned frequently. 152 The Ovimbundu The snuff is pounded and placed in a small wooden box of cylindrical shape. Usually the box is ornamented with incised, burned patterns. The Ovimbundu of Bih6 add ashes of wood to their snuff, so producing a mixture called ulelemo. The Ovimbundu of Elende usually use the snuff without adulteration. Two main species of tobacco plants are grown. Women smoke in all parts of Angola. The Ovimbundu do not chew tobacco, neither have I seen it so used elsewhere in Angola. Boys and girls are not allowed to smoke before the age of thirteen years. In the ombala of the Vangangella near Ngalangi, I asked a girl for her pipe which I desired for my collection. The interpreter took the pipe when she proffered it, explaining that, according to local custom, I had asked for the girl. If I took the proffered pipe from her hand I accepted her. Another social custom associated with tobacco is the passing of the communal pipe from hand to hand in the men's council house. Ngonga, my interpreter, says that he has never seen an Ocimbundu woman smoke hemp, but he has seen a woman of the Vangangella (people to the east of the Ovimbundu) smoking hemp. Hemp (epangue) is cultivated only by the Ovimbundu men who smoke it. Pure hemp is smoked in the water-pipe which is not passed from hand to hand. Only tobacco is used in communal smoking. Smoking of hemp or tobacco consists of a few deep inhalations; there is not usually a prolonged placid smoking. When hemp is placed in the bowl of the water-pipe it is covered with large grains of sand or a piece of tin. This intervening substance prevents the hot coals from coming into contact with the hemp. The object is to secure slow ignition. DOMESTIC ANIMALS The principal domestic animals are cow (onjindi), ox (ongombe, which is also the generic name for cattle), bull (onui), sheep (omeme), goat (ohombo), pig (ongulu), dog (ombua), chicken (osanji) (Plates LV, LVI, LVII). The transport animals, donkey, horse, and mule, are not used by the Ovimbundu of Elende and not to any extent by Ovimbundu of other parts; but in the south of Angola the Vakuanyama have sturdy ponies and well-kept mules. The ox when ridden by Portuguese is provided with a leather saddle which is very comfortable if covered with a blanket. The brass stirrups are broad and massive. Through the septum of the bullock's nostrils there is a short brass rod to the ends of which the reins are attached (Plate XXX, Fig. 1). Economic Life 153 I have seen an Ocimbundu male riding an ox without saddle. A cord was passed through the animal's nose to serve as reins. The Portuguese name for horse is cavalo, a word which the Ovim- bundu use in the form okavalu, though there is an Umbundu word ocingongovala, which means "going with his neck up." The Umbundu words for donkey and mule are ocimbulu and omula, respectively. Generally speaking, the ears of animals are not clipped, neither are cattle branded or otherwise marked to indicate ownership, but sometimes the ears of pigs and goats are cut to aid identification. When asked why the tips of the ears of dogs are mutilated an Ocim- bundu will say that a dog with uncut ears does not hear when called (Plate LVI, Fig. 2). Not many families own large herds of cattle among the Ovim- bundu, and I did not see a big kraal until I was in southwest Angola. Cattle, which are a measure of wealth, are used for paying fines, making funeral feasts, paying debts, and securing wives. The cattle throughout Angola are well-developed, handsome animals (Plate LVI I, Fig. 1). The bull remains with the herd the entire year; there is therefore no particular season for the birth of calves. Usually cows are not milked by the Ovimbundu, consequently these people have no milk, butter, or cheese. The Vakuanyama of the south milk their cows and churn butter in calabashes slung on a pole. The Ovimbundu say that milking the cow makes the calf thin, but where the idea of milking the cow is borrowed from the Portu- guese the Ovimbundu have a procedure which is as follows: The calf is allowed to suck for a few minutes in order to deceive the cow; then the milker begins his work. At intervals of a few minutes the calf is allowed to suck in order to continue the deception. Cattle are killed at the funeral feasts of the rich, and the horns of the slaughtered animals are generally mounted on a pole in the vicinity of the grave (Plate XLVI, Fig. 2). The horns of the cow may be used for making water-pipes in which tobacco is smoked, or they may be employed as magical horns when filled with medicine. The hide is pegged out in the sunlight for one day after it has been scraped; it is then rolled and kept until required. When about to be used, the skin is soaked in water for one day. The hide is used to cover the tops of stools, to make pouches and to manufacture bags for carrying corn. In bygone days each king had a wooden box covered with hide, which contained his powder and metal when he went to war. The cow's tail is used 154 The Ovimbundu as a sheath for covering the iron shafts of assagais, and it is sometimes made into a switch which the rain-maker uses during his performances. Bulls are castrated when two years old. The wound is rubbed with ashes, salt, soot, and palm oil. Bullocks are used for riding, also for pull-carts introduced by the Boers. The herd obtains most of its food by grazing. In the dry season grass withers, with the result that the animals become thin and stall-feeding is necessary. In former days cattle-raiding was a practice of the Ovimbundu, who robbed the Vacilenge. It is certain that the Ovimbundu did not own cattle when they entered Angola. If the general tradition is correct the Ovimbundu came into Angola from a northeasterly direction, from the borders of the Belgian Congo as it is called today; this is not a cattle-raising region. The Ovimbundu of Elende have a joke against the people of Bailundu, because the latter on first seeing a cow offered the animal some food on a wooden platter. This story suggests that the Bailundu people, who are of the Ovimbundu confederacy, did not know the animal and its habits as early as did the Ovimbundu themselves. Although the Ovimbundu do not usually kill their cattle they may do so in the months of June and July, because at this time pasture is withered and food is scarce. Animals which are diseased, aged, or injured, are killed and eaten. The native pig is distinguishable from European breeds by its long thin snout and slender development. This breed is said by the Director of Animal Husbandry, Humpata, to be the Keltic breed (Plate LV). There is, he says, no evidence to prove that the Ovimbundu have at any time domesticated pigs from the wild hogs which are to be found in Angola. There is no family which does not own a pig, and on the whole the pig fares much better than the sheep or the goat. The pig receives water and a daily ration of food, which is usually sweet potatoes and their leaves, together with some corn. Male pigs are castrated at any time between the ages of six months and one year by an operator who is a paid specialist. The fee for castrating a bull is four yards of cloth, but a small gift is considered sufficient reward for performing the operation on a pig. If a pig is thin, the leather is said to be of good quality and therefore suitable for making sheaths for knives. Usually the flesh of the pig is eaten shortly after the animal has been killed, but the meat of the bullock, on the contrary, is sometimes dried over a fire and preserved. But this is not a general practice as the animals are too valuable to be slaughtered. Economic Life 155 Goats are more common than any other animal; there are few, if any, families which do not own one or more goats. These animals are not fed or watered. Goats are able to exist on almost any kind of vegetation; consequently these animals are, almost without excep- tion, well nourished. Goats are not milked. Kids are born at any time of the year, and the young males are castrated. The hides are used for making bags. The goat has the misfortune to be the most desirable sacrificial animal. This is not entirely due to the fact that it is cheap and easily obtainable. The sheep is said to be unsuitable as a sacrifice, because it does not make a noise when killed. The hair of the goat is used for making an ornament named osala, which is worn by medicine-men. Sheep are of the long-tailed Syrian breed. Like the goats, the sheep are not cared for in any way; they find their own pasture and water, and in doing so may wander for a considerable distance, though they always return to the village at sunset. The males are not castrated. Sheep are not so frequently kept as are goats and pigs. The skin is used for making bags. Twin births of calves, kids, or lambs are not regarded with awe; on the contrary, such births are welcome. Almost every man keeps one or more dogs, and I have rarely seen a hut in which there were no dogs. Usually young puppies are near the fire, and on the whole dogs are well treated because they are valuable in hunting. They also give warning of the approach of hyenas, lions, and leopards. I have frequently seen a person run into the road to pick up a dog when an automobile is approaching. A tendency to improve the breed of dogs by crossing the lean native animal with a breed of large dogs from South Africa has been men- tioned. In contrast with the generally considerate treatment of dogs by the Ovimbundu one has to note the very emaciated and diseased condition of dogs in the Esele country. Among the Ovimbundu dogs are regarded as desirable food. Sacrifice of a dog at the inauguration of a blacksmith will presently be described. A medicine-man who is about to perform a ceremony for curing the sick has to make a meal of dog's flesh, but otherwise the flesh of the dog is taboo to him. These points of ceremony, taken in conjunction with the food value of the animal, and its use in hunting, show that the dog is highly esteemed. Poultry are of very mixed breeds. The standards of size and weight are higher than is usual in African chickens, a fact which is perhaps attributable to contact of the Ovimbundu with the Portu- 156 The Ovimbundu guese for a long period . The chicken is highly esteemed as a sacrificial animal, but it could not be said that the flesh is in common use. This failure to develop and utilize to the full, again raises the point of social custom and economic habit. Chickens are cared for, as may be seen in the way they are cooped at night. Sometimes a hen may be seen sitting on her clutch of eggs in a dark corner of a hut, unmo- lested by people and dogs; yet it is certain that there is no concen- tration on the rearing of poultry. Eggs are laid, and chickens are hatched throughout the year, except in the months of November and December when corn supplies are at their lowest ebb. Eggs are boiled, or fried on a fragment of pottery, but they are not sucked. The domestic cat is not raised by the Ovimbundu. Wild cats are common, but there is no evidence of their domestication. Small birds and monkeys are sometimes kept as pets. Trade and Transport Caravan trade, which was at one time an important factor of tribal life, is now confined to short journeys for transporting corn, beans, and beeswax to traders' stores. But in spite of present-day decline of transportation the memory of more prosperous times still exists. Names of distant places survive in the Umbundu language; thus Tanganyika is called Nakandundu, while the name for far eastern Angola is Muacimbundu, the name of a one-time important chief. The Umbundu language is understood in all parts of Angola, far away from typical Umbundu centers of culture. In the old days there were professional leaders of caravans, and a ceremony was conducted before starting. The medicine-man and the village chief were the principal performers in a rite which consisted of bringing from its box the head of a former chief, sewn in oxhide. An animal was sacrificed so that the blood could be used for sprinkling on the chief's head, and on some occasions of this kind the head was sewn up in a new piece of oxhide. Direct appeal was made to the preserved head by the reigning chief, who asked for good fortune on the journey. The medicine-man who accompanied a caravan carried with him a female wooden figure decorated with feathers (Plate XXI, Fig. 5). When a branching of the paths gave rise to doubts concerning the correct way the wooden figure was consulted by the medicine-man. Economic Life 157 At the present time a day's march is twenty-five miles, during which a man carries sixty pounds, while the load for a woman is half that weight. The gait of the Ovimbundu includes a limp at every step, so suggesting that the carrier is lame or tired. This appears to be a method consciously adopted as a protection against fatigue, because all muscles are momentarily relaxed. Loads are carried on the head in a long forked stick to which they are lashed. When the carrier rests, the load is not placed on the ground, but is held upright on the stick (Plate XXX, Fig. 2). Such a method avoids the strain of lifting the load from the ground after each rest pause. Although the Ovimbundu have an exchange of products among themselves by both barter and the use of Portuguese money, there are no large markets, with the exception of those at the coastal towns of Loanda, Lobito, and Benguela. There is absolutely nothing in Angola which can be compared with the great markets in Nigeria. Despite the absence of a system of exchange on a large scale, the Ovimbundu have many terms describing units of measurement. There are native standards of length, area, and capacity, but no measures of weight which are undeniably of Umbundu origin. The unit of length (epaluma) is the distance from the tip of the thumb to the tip of the middle finger when the hand is outstretched. The term for two of these units is apaluma avali. These words are the plural of epaluma; avali means two. These units are used to measure tobacco before it has been coiled. Cloth is measured by stretching the arms to their full extent in line with the shoulders; the distance between the tips of the middle fingers is epeka. The stride for measuring land is elianga. Onjimba is an area about twenty-five feet square. Etemo ("hoe") is an area of land two hundred yards long and thirty feet broad. A large field covers two or three atemo. Measures of capacity are provided by various types of baskets. The large conical basket (ohumba) has an interwoven mark which indicates a measure for corn, meal, and beans. Ocitenge is a coarsely made basket used as a unit of capacity. Uhamba is a basket two feet long and one foot deep. The basket on a rectangular base is also uhamba, but at the present day cans are taking the place of all these old measures. Palm oil is measured in a gourd (ocitau or ombangi) of definite size. This little gourd is also used for measuring a viscous substance from trees. The word ekokoto is used to describe this mucilage. 158 The Ovimbundu Balls of wax and tobacco were, and are now, definite standards of trade. To some extent rubber as a medium of exchange has been used through contact with eastern Angola. The Umbundu word ocilila expresses a weight of about thirty kilos. There is no measure for minutes or hours. I have seen a man of the Luchazi tribe keep account of the number of days taken on a journey by cutting notches on a stick. Ngonga says that the Ovim- bundu reckon by cutting notches, also by knotting a piece of string. Three days would be expressed by the words akumbi atatu ("three suns")- The word day or sun is used in fixing a time. There is no word for week. A month is osai, which is the word for moon. When the corn is ripe the people say, "We are in a new year." Another yearly time mark is the arrival of the first rains, probably in the middle of September. The words oku lima ("to cultivate") yield the word ulima, which designates the period between the beginnings of two rainy seasons. Corn is planted about the time of the first heavy rains, which occur in late September or early October. Naturally, this important occa- sion forms a somewhat uncertain time base; nevertheless it is the one used to express the lapse of years up to five in number. After such a period the estimation of time is unreliable. Industries iron-work Among occupations of primary importance is that of the black- smith (ocivinda). Owing to the increasing importation of hoe blades and other iron goods, together with the facilities for collecting scrap iron, the winning and smelting of iron is increasingly rare. There are probably very few places where the old type of conical clay furnace now exists. Almost any fragments of iron are melted at the forge where a box is kept to hold nails, hoop iron from packing cases, and other fragments resulting from proximity to a European culture. Nothing is mixed with the iron, neither is there any casting in molds. The only process is the forging of red-hot iron. The Ovim- bundu do not draw iron wire, although they know of the process which is practised by the Vachokue. The work of the blacksmith was studied at the village of Njongolo in Elende. The men were for a time reticent respecting the nature of their training and the ceremony of initiation, but the chief of the village helped considerably by persuading the men to speak freely. Economic Life 159 Any boy who wishes to become a blacksmith may be trained for the work. It is not necessary that his father should have been a blacksmith. When the youth begins his training he must be eighteen years of age and physically robust. His first duties are the beating of hot iron on the anvil; he is also required to collect and soften fragments of iron. He works very hard but is never allowed to finish anything; the master has to complete the work. At the end of two years the youth asks the master blacksmith to examine him. What is more important still, the master is asked to make the heavy hammer (onjundo), which is used for beating hot iron on the largest anvil (Plate XXXVIII, Fig. 1). There is no doubt as to the sacredness of this hammer, which is a symbol of the completion of apprenticeship. There is also the idea of the master handing on his skill to the pupil by personally making and presenting the tools. I made repeated efforts before being able to purchase one of these hammers ; finally it was procurable only at a high price. The value is due to ritualistic associations, the large quantity of iron used in the making, and the labor required to weld the head to the shaft. On the day of his inception the boy has to purchase four chickens, two male and two female, one pup, and a goat. The master black- smith makes all the tools for the apprentice, but ritual centers chiefly in the fabrication of the big hammer (onjundo). While the master is making this the boy stands on the small anvil which is close to the ground, between the forge and the large anvil under the tree. When the hammer (onjundo) is made, and while it is still red-hot, the handle is pushed into the belly of the dog. The goat and the four chickens are then killed. All the tools are brought together so that blood from the slaughtered animals may be sprinkled over them. The flesh of these animals is eaten with corn and beans. "The blacksmith calls many people to help him, and they like to eat the food," con- cluded my interpreter. During the entire ceremony, also throughout the feast, the boy stands on the anvil. There he remains until the master says, "You may speak and tell us what name you want." Perhaps the boy says, "I am Ndumbu." The people in the crowd clap hands and make a trilling with their fingers in their open mouths. The boy steps from the anvil; he is a blacksmith. My interpreter continued, "He must work hard and people must pay him. He used to work hard, but the master took the money." There was an ancient belief that a blacksmith owed his skill to the help of the spirit of a person he had killed. Wooden effigies 160 The Ovimbundu of the murdered man were placed near the large anvil (p. 163), or they might be kept in the home of the blacksmith. Such figures are still used (Plate XXI, Fig. 3) but the killing of a victim is not now possible. Blacksmiths are free to marry without restrictions other than those imposed by the classificatory system of relationships. The blacksmith's forge is a thatched house about twelve feet square with low eaves that almost reach the ground. The height from the floor to the point of the dome is fifteen feet. In the middle of the floor is a pit ten inches deep in which there is a quantity of charcoal brought from the charcoal burner's fire about half a mile away. At the sides of the pit are three stone seats for the workers. Two of these seats are occupied by men, each of whom works a pair of bellows. The two-chambered bellows is hewn from a block of wood in such a way as to give two circular air chambers from which wooden tubes lead to the fire. These wooden tubes are continued by clay tubes which project into the fire. Over the two round wooden chambers a piece of hide is stretched and tied over the wood- work. Two slender upright sticks, which the operator works up and down, are attached to the hide. In one corner of the hut there is a heap of charcoal, and in another corner lies the scrap iron. An iron rake with a wooden handle is used for stirring the charcoal in the fire. Plates XVI, XXXVII, and XXXVIII illustrate tools and processes. The principal tools are: (1) A flat stone anvil resting on rocks under a tree. The anvil is at such a height that the striker stands upright. At this anvil the metal is beaten with the heavy hammer onjundo. (2) Onjundo, the most sacred of the tools, is 12.5 cm long. Its value is about that of an ox. (3) There are tongs which can be clamped by a sliding metal ring. The larger tongs are 63 cm long and the smaller ones are 35 cm. (4) The cutter is boat-shaped and triangular in cross section. The back, which is grasped in the hand, is 0.5 cm thick, tapering to a fine cutting edge. (5) An iron holder for an axhead during the heating and ham- mering is 23 cm long. It is octagonal, hollow, and fits like a sheath over the shaft of the axhead. (6) For heating the iron on the small anvil, after it has been roughly pounded to shape with the hammer (onjundo) on the large anvil, there is a smaller hammer (usonjolo) of which there are three varieties. These differ only in size. Economic Life 161 The principal products of the forge are axheads, which can be reversed in the shaft so as to form adzes. Hoe blades, tools for mat- making, brass bracelets, knives, and implements for gouging out the pith of gourds or hollowing out a drum, are also made. Arrowheads are likewise an important manufacture. The blacksmith makes a saw blade, 47 cm in length, from hoop iron. The teeth of the saw are turned alternately to the right and left and the serrated blade is roughly hafted in wood (Plate XVI, Fig. 4). Another product of the forge is an iron tool hafted in wood. The pointed blade, which is round in cross section, tapers to a point used for boring holes in wood, after the tool has been made red-hot. Small axes, many of which are used ceremonially in dances, have remarkably well-fashioned blades decorated with punched designs in the form of geometrical patterns. WOOD-CARVING A glance over the list of trees named and used by the Ovimbundu indicates a complete knowledge of woodcraft which is in the hands of specialists. To name only a few, there is the omanda tree, also the omue, which yield charcoal for the smithy. Ombula wood is used for stools, while the elastic timber from the omanda tree is suitable for bows. The ekenge, usamba, and ociyeko trees supply bark for binding crosspieces to the upright poles when making the frame- work of a house; the same bark is used by Ovimbundu and Vachokue of eastern Angola for making bark cloth. In addition to the use of bark for these purposes it is made into large cylindrical receptacles for maize, while its use for beehives is general throughout Angola. Beehives are often made by professional hunters. A traveler notices the mutilation of trees from which complete cylinders of bark have been removed (Plate XL, Fig. 2); half of the cylinder makes a receptacle for carrying on the shoulder. When a small pig is to be transported, four holes are bored in the bark container. Through these holes the legs of the animal are placed and tied together on the under side (Plate XXXI, Fig. 1). This is a more humane transportation than that of tying the feet of the animal to a pole which is supported on the shoulders of the two men who are carriers. In the neighborhood of Cangamba fishermen make and use canoes of bark about fifteen feet in length (Plate LXXXV, Fig. 2). This illustration shows a man taking his nets into midstream where they will be pegged to the river bed. 162 The Ovimbundu The tools used by the wood-carver are the saw, ax, adze, and knife. The products of this craft may be conveniently divided into the following groups: (1) Figurines of human form (Plate XXI, Figs. 1-6). (2) Animal forms, chiefly snakes, tortoises, birds, and lizards. The dog is sometimes represented (Plates XIX, Figs. 1-5; XLI, Fig. 1). (3) Parts of musical instruments, such as drums and the base- boards of sansas. The latter often have the metal keys mounted on boards decorated with elaborately incised patterns (Plates XLI, Fig. 2; XXII, Fig. 5). (4) Domestic implements and utensils, chief of which are heavy wooden beaters for flattening mud floors of houses, grain pounders, stools, cups, platters, bowls, and a heavy pestle and mortar (Plates XIII, Fig. 3; XVIII, Figs. 1, 2). (5) Carved sticks and clubs; these often show elaborate incised decorations of geometrical patterns. Frequently the head of the stick is carved to represent a human head or a full-length figure. The ornamented stick or club is carried as part of the personal dress and artistic equipment. The throwing club is usually a straight stick with an undecorated knob at the end (Plate XX, Figs. 1-10). (6) Carved wooden posts representing the human form. These are set up at the wayside. One has been obtained from a grave near Bailundu. (7) Tobacco-pipes and snuff boxes (Plate XV, Figs. 2, 4). Figurines of human form require special consideration, because they have claims other than that of aesthetic expression. The figures representing Europeans, or natives using some article of foreign introduction, illustrate the grafting of foreign ideas on older methods of work. The art of the African Negro has of late years been accorded a place of honor in critical circles of Europe and America (P. Guillaume and T. Munro, Primitive Negro Sculpture, New York, 1926). Usually, however, the formal technique of lines, curves, and the general aesthetic effect have been discussed to the exclusion of the ethnological background which determines style and function. By far the best example of carved human figures collected in Angola was the one from Cangamba (Plate XXI, Fig. 4) . This female figure, 60 cm high, is carved from hard, dark, red wood in such a way as to achieve a graceful result by the employment of a few Economic Life 163 straight lines. The legs show the usual flexion of the knees and a shortening which is out of proportion to the body length. The lower limbs of most Angolan figures have these characteristics. The body of this figure is hollow and the head detachable. The incised head- dress is imitative of the coiffure of Vachokue women. Cangamba, a village in eastern Angola, is a confluence of tribal elements, namely, Vambuella, Luchazi, Babunda, and Vachokue, whose physical appearance, hairdressing, tooth mutilation, and tribal marks differ considerably. The man who sold the figurine brought it furtively. He was an Ocimbundu, but the work is of Chokue origin. The figure, I am informed, was filled with medicine, then placed near a patient who was undergoing curative treatment. The figurine (ngeve) of a woman having a number of dark feathers attached to her back has only one use. A caravan setting out for a long journey is accompanied by a medicine-man whose outfit includes such a figurine. Should the caravan leader be in doubt when choosing between two paths, the medicine-man sets up the image at the parting of the ways. He kneels before it and asks questions, then plugs his nostrils in order to make replies in a falsetto voice supposed to come from the figurine. Thus advised respecting the route to follow, the caravan continues its journey (Plate XXI, Fig. 5; p. 156). The wooden figure (Plate XXI, Fig. 3) is of exceptional interest because of its connection with the blacksmith's craft, which is asso- ciated with introductory rites for apprentices. In former days a newly initiated blacksmith was expected to disappear for a period during which he killed a man. On returning to the work of his forge the blacksmith made a wooden figure of which this example, one of five obtained, is typical. The spirit of the murdered man took up its abode in the effigy and in this way helped with the work. The figure, which is of the usual dimensions, namely, 36 cm high, has the greater part of its surface covered with reddened earth (pp. 159-160). Two figurines (Plate XXI, Figs. 1, 6) were obtained from women of the Vachokue tribe in the village of Ngongo, Ngalangi. Here the Vachokue and the Ovimbundu mingle to such an extent that customs are no doubt transferred from one culture to the other. These wooden images are used by childless women, or by women whose infants have died. The woman who sold these figures pressed them to her breasts to show the manner of use. People standing around smiled and nodded their approval. A wooden figure of this kind is substituted for a dead twin. 164 The Ovimbundu A common type of small figurine, some of which are female, others asexual, is represented by an illustration (Plate XXI, Fig. 2). Such little carvings are a normal part of the miscellaneous contents of a divination basket described in chapter IX. These figures have been obtained from Elende, Bailundu, and Caconda, all of which are centers of Umbundu culture. In addition to the figurines described, the collection contains many more whose use may be conjectured. The interrogation of Ovimbundu people indicates clearly that there are specific uses of wooden figures which are consulted by the medicine-man. The fore- going explanations illustrate the nature of the beliefs associated with carved wooden figures, but it is not always possible to elicit a clear account of the specific use of each one. Consideration of this aspect of the wood-carver's art has an important bearing on the culture contacts of the Ovimbundu. In studying this question I have instituted comparisons between these figures from the Ovimbundu of Angola and similar figures from the Kasai area of the Congo region (chapter X). The publications of the Musee Congo Beige provide illustrations for comparison with the figures in Field Museum's collection. The carving of animal forms (Plate XIX, Figs. 1-5) is no more than a means of aesthetic expression resulting from accurate observa- tions of animal life as recorded under "Nature Lore." Inquiry failed to show that figures of animals are, or were at any time, used in rites and ceremonies. The carrying of a carved stick is essential when a chief is visiting, attending a council in his own village, or receiving visitors. The ornamental paddle (Plate XX, Fig. 8) was owned by the chief of the capital village (ombala) of the Vangangella, near Ngalangi. At the death of a chief his staff of office, with his tobacco-pipe and sleeping mat, are placed in a small house where such relics of deceased chiefs are kept permanently (Plate XLVI, Fig. 1). Plate XX, Fig. 6, shows a staff of this kind which was preserved in a sacred house in the capital village of Ngalangi, and on the same plate are drawn short ornamental clubs which French ethnologists call batons de promenade, an appropriate name because of their use as part of the full dress equipment. The Ovimbundu have specialized in the carving of small objects, for, in addition to clubs and staffs, tobacco-pipes and snuff boxes are often elaborately carved (Plate XV, Figs. 2, 4). Economic Life 165 The water-pipe of the Ovimbundu consists of the horn of a cow into the side of which a short hollow pipe stem is introduced ; at the top of the stem is a clay bowl for the reception of tobacco, or a mixture of tobacco and hemp. The wide end of the horn is plugged with clay, while a hole is made at the tip in order to provide a mouth- piece. A gourd water-pipe, similar to that used by the Vachokue, is also found among the Ovimbundu. Cylindrical snuff boxes are ornamented with incised, burned, geometrical patterns. The lid is usually attached to the box by a leather thong. One large snuff box is ornamented with three well- carved female figures. The smoker's equipment is sometimes carried in a leather pouch fastened on a waist belt, or the container may be a hollow cylinder of ivory with a leather cap at each end. To this information respecting the smoking and snuff -taking outfit of the Ovimbundu, some observations on the pipes and snuff boxes of other tribes should be added. When making a journey from Cangamba to Saurimo I seldom met an Ocivokue man who was not carrying a gourd water-pipe for the smoking of tobacco and hemp. Such pipes are usually ornamented with brass nails and are finely bound with thin brass wire (Plate XV, Fig. 3). One long pipe from Ngalangi has a pair of metal tongs attached for taking charcoal from the fire in order to ignite the tobacco (Plate XV, Fig. 1). A pipe with two bowls, closely resembling some Zulu patterns, was smoked by an Ocimbundu woman at Ngalangi. The Vachokue make snuff boxes from a yellow wood, which they ornament by burning portions of the surface in such a way that the yellow color is here and there visible. Both men and women of the Vakipungo and Vakuanyama tribes carry snuff boxes of conical shape on their leather waist belts. At the top and bottom such snuff boxes are neatly bound with brass or copper wire. DOMESTIC IMPLEMENTS At an early age girls become accustomed to the use of the V-shaped pounder which is made from hard heavy wood (Plate XXXVI, Fig. 2). The small end of the shaft is a convenient thickness for grasping in such a way that the knuckles are on the under side of the shaft. This is the very reverse of what appears to be the natural grip. Women use the pounder with an easy, circular swing so that the flat round surface comes into contact with the grain on the rock. Endurance in this work is remarkable, and there is no doubt that 166 The Ovimbundu fatigue is avoided by the method of holding and swinging so that the work of crushing is done by the weight of the implement. Two less usual methods of pulverizing grain are by use of a long pestle, which is worked up and down in a heavy wooden mortar as shown by Plate LXXXIV, Fig. 1, in which Vachokue women are so employed, and the crushing of grain with a cylindrical stone which is rolled on a flat slab. This latter method I judge to be very old, for on a pre-Umbundu site encircled by stone walls I have seen flat slabs of stone and cylindrical rollers. Some of the flat slabs were worn extremely thin in the middle and a few were perforated by the friction. It is noteworthy that there are rocks especially reserved for the pounding of grain with the wooden mallet. Such rocks are to be found close to every village, and the dedication of a new rock for this purpose requires the killing of a chicken whose blood is sprinkled on the rock. Work of this kind is begun before daybreak. From that time to sunset the pounding of the wooden mallets, accompanied by the singsong of the women, marks the progress of the day's work. At intervals the pounded grain is sifted through the hands and spread out on a basket-work tray, which is shaken to separate the fine meal. The unbroken and partly broken grain is replaced on the rock for further pounding. A small brush of grass is used for bringing together the grain which is dispersed by the blows. The sole implement used in agriculture, which is entirely in the hands of women, is the hoe. Of this implement there are several variations according to locality. The Ovimbundu and others of Ngalangi employ the form illustrated in Plate XIII, Fig. 10. The long handles measuring 85 cm no doubt reduce fatigue by minimizing bending. Both hands are used, and, furthermore, the increased length of the handles must give a greater leverage. The smaller hoe used by Ovimbundu women of Elende has short handles only 51 cm (Plate XIII, Fig. 9), while that from the Esele country is distin- guished by a broader blade and still shorter handles only 36 cm in length. Included in the outfit of every Umbundu home is the heavy wooden floor beater. This implement, which is fashioned from a single piece of wood, consists of a narrow handle, round in cross section, and a flat portion for beating the moist, newly made, mud floor of the hut. The total length is about 83 cm. Economic Life 167 Not only the carpenter, but almost every Ocimbundu man possesses an implement which may be readily adapted as ax or adze (Plate XIII, Fig. 2). The wooden shaft, from 50 to 70 cm in length, terminates in a narrow grip at one end, while the other end expands into a large oval knob into which a circular hole is bored. The strong iron blade, one of the main products of the local forge, can be removed. If the round tang of the blade is inserted so that the cutting edge is in the same plane as the shaft, the implement is an ax. On the contrary, insertion so that the cutting edge is at right angles to the shaft converts the implement into an adze. When placed over the shoulder this implement is used for carrying utensils in a fiber bag (Plate XXVIII, Fig. 1). Gourds are of three main kinds, which may be found growing on the ground or resting on the roofs of houses at the ends of the climbing stems that bear them. Ombenge is a gourd which is narrow in the middle. It is often converted into a dipper for ladling liquids. The narrow neck is the handle, while a round hole is cut in the larger part so that the dipper can be filled with liquid. Onganja is round, or perhaps oval in form. There is a size used as a measure of capacity. Onganja can be used as a ladle by fixing it at the end of a stick. Olukuembo has a round body and a narrow hook-shaped neck; like the gourd ombenge, it is used as a ladle. Gourds are elaborately decorated with incised and burned designs (Plate XII, Figs. 1-6). Usually these patterns are geometrical, but human figures and animals are sometimes included. Decorated gourds from Bailundu, used for containing beer at a wedding, are among the best examples of their kind (Plate XII, Fig. 6). The owner of a decorated gourd takes great pains to repair a crack with rattan, which is threaded through holes bored in the edges of the fracture. A large gourd is sometimes carried in a native-made net of vegetable fiber. POTTERY A large and varied collection of pottery has been acquired from the Ovimbundu, the Vachokue, and the Vasele. The shapes and styles are best indicated by the illustrations in Plate XIV, Figs. 1-5. Among the Ovimbundu, only women make pots, which are inferior in workmanship to those of the Vachokue and the Vasele. At Elende, among the Ovimbundu, I was surprised to find a man making pottery. Further inquiry showed that he had learned his craft when young in the Vachokue country of eastern Angola. This man, whose work was not copied at all by Ovimbundu women, was regarded as a 168 The Ovimbundu specialist whose products were in great demand. Instead of making crude patterns with a piece of gourd, as do the Ovimbundu women, he presses a brass bracelet round the rims of the pots. The bracelet is deeply indented with geometrical patterns which appear on the clay (Plate XIV, Fig. 5). This man selects a very fine clay for his work, his products are symmetrical, and a polish is given with a smooth pebble after baking. Like all other potters of Angola, this artisan has no knowledge of the potter's wheel. The vessels made by Vasele women, among whom the trade is confined to female specialists, are unlike the products from any other part of Angola. The chevron design is characteristic, so also is the ornamenting of the pot by laying on strips of clay below the rim. This is applique" work which may consist of only a few bands of clay, or the strips may pass repeatedly round the pot until one-half or two-thirds of the surface has been covered (Plate XIV, Fig. 2). At Elende two Ovimbundu women were observed during the pot- making processes (Plates XXXIII, XXXIV). Woman A built up the pot in a basket lined with wet leaves. Woman B pounded the clay on a stone with the heavy wooden pounder used for pounding corn on the rocks. Between the two women was a gourd of water in which A moistened her fingers. Operator B poured water on the clay which she was pounding. A made a cup of clay five inches in diameter and four inches deep which served as a base on which the pot was built up by the coiling process. This cup was placed in the basket contain- ing moist leaves. The rim of the cup was built in height and breadth by the addition of rolls of clay supplied by woman B. As the rolls of clay were laid on the edges of the pot, the inside was smoothed with a piece of gourd. Meanwhile the outside of the pot was supported with one hand. Shaping of the pot proceeded by applying smaller and smaller rolls of clay as the neck of the pot was approached. Gentle smoothing pressure forced out the greatest breadth of the pot just below the base of the neck. A pause was made to allow a partial drying before the neck was built up. Ornamentation consisted of making deep incisions with a piece of gourd while the pot was still wet. During these processes there was constant wetting of the hands. The pots were sun dried, then fired several at a time by placing them in a kiln of dry grass. Polish was given to a pot while it was still hot by covering the surface with liquid made from a tuberous root, during which process the pot was quickly turned on a stick. Economic Life 169 When the clay was being mixed an old pot was broken and pul- verized so that some of the powder might be added to the new pot. There may be the underlying idea of continuity in the potter's art. The potters said, Sanga yi pita ("Lest it leak"). There may be no purpose other than the imparting of stability to the new clay. I was unable to find any trace of ritual except with reference to the opening of a new clay pit. When a pit is first opened both men and women attend. The head of a chicken is twisted off by a medi- cine-man, then the bird is held over the pit by either a man or a woman. There is no law or ceremony to determine who shall take the first clay from the pit. The art of making pots is in the hands of female specialists so far as the Ovimbundu are concerned. Obser- vation makes clear that women will go for a long distance to obtain clay from the pit which has been opened in a ceremonial manner. On their way to such a pit they pass clay which would serve their purpose well, but they do not use it. Children sometimes amuse themselves by making animals of clay. The pottery of the Ovimbundu includes cooking vessels of many sizes, water containers, and very large pots for brewing beer (olombia vi okukela). MATS AND BASKETS This occupation illustrates division of labor on a sex basis. Baskets are made by women, while mat-making is an occupation for males. As with other trades there is specialization. The majority of women are able to make baskets though the skill of individuals varies. All women who have a knowledge of basketry understand the manufacture of dyes. Only a few men make mats; my informant thought that perhaps one man in ten would have the necessary skill. Such specialization is continued into other occupations; for example, only a few men spin cotton thread, while the majority of people buy pottery from expert female potters. The mat-maker, generally an elderly male, uses two tools, a borer and a needle, both products of the native forge. The borer (utomo) for piercing the reeds, consists of a long thin blade in a wooden grip. The needle (osinja) is threaded with bark fiber and passed through the holes made by the utomo. The bark thread is ombanja (plural olombanja). The sleeping mat (esisa) is made of reeds which are gathered in the early morning by a man who wets and binds his material into bundles, each of which contains reeds of the same length. The length of the reeds varies, of course, with the size of mat he intends to make. 170 The Ovimbundu The name esisa is given to the raw material as well as to the mat. The worker begins by laying out the reeds on the ground, side by side; then the slender tool utomo is passed through the reeds near their ends. This position is made permanent by sewing, and the process is repeated at intervals along the length of the reeds (Plate XLII, Figs. 1, 2). Evinda is a large mat, about 120 cm long and 20 cm wide, while each strip of coarse elephant grass is about 3 cm wide. Ability to make the rush mat (esisa) is fairly common, but skill in making evinda is less usual. The mat evinda has several uses; it may be stretched on the floor or bed as a sleeping mat, or possibly it is rolled so that the ends can be fastened together; so treated it forms a cylinder which may be filled with grain when stood upright on the ground. A number of these mats is sometimes used to form a temporary storage place or shelter. The technique is of the twilled variety in which each weft passes over and under two warps. Ocala is a coarse mat made from long stalks which are called "elephant grass" by the Boers; the agricultural term is "Napier's fodder." The long rods are white and glazed, so forming an artistic contrast to the crossbinding, which is of black bark. The technique is known as check, a term used to describe a structure in which warp and weft pass over and under each other singly. The large mat (ocikanga) which is of soft texture shows neatly woven, diamond-shaped patterns of dark brown grass. The technique of this mat is of the twilled variety in which each weft passes over and then under two warps. String bags, which are used for suspending gourds or hanging them over the shoulder, are made by men only. The root ombundi, men- tioned in connection with brewing beer, is used for making string bags. Two fibers that have been teased out from the root are rolled tightly together by rubbing them between the palm and the thigh. Strong rope is sometimes made by plaiting coarse grass; this occupa- tion is in the hands of males. In basket-making the preparation of dyes is of great importance because colored strands of grass are inwoven to make named geomet- rical patterns (Plates IX; X). The Umbundu expression for dyeing is oku lisa olosovo. Red coloring is produced by taking leaves of a plant named evava and cooking them in water along with the bark of the tree ukondo. Ukondo is the "tooth brush" tree because small pieces of its wood are chewed and used for cleaning the teeth. After the grass has Economic Life 171 simmered in this infusion of evava leaves and ukondo bark, it is buried in a heap of ashes and earth. These processes produce the red coloring. Yellow dye is made from the roots of a wild rhubarb (ocilungtduila) which has sagittate leaves. The roots are pounded and mixed with cold water, after which the grass is placed in the mixture. A pot containing the mixture is put on the fire and the contents are boiled for half an hour; at the end of this time the grass is an amber-yellow in color. In order to dye grass black the leaves of evava, the plant used for making red dye, are mixed with an iron solution obtained from the mud of stagnant pools. If the grass which has been cooked in this mixture is not sufficiently dark it is reboiled in the evava-iron mixture to which the pounded leaves of ungalo are added. Brown coloring is made by mixing the red dye with the yellow. Baskets made by the coiling process are the most common type manufactured by Ovimbundu women (Plate XV, Fig. 6). The large basket ohumba is a woman's field basket in which she carries corn, sweet potatoes, and manioc, along with her hoe and pounder (Plate XXVIII, Fig. 2). Each coil consists of a large number of strands of fine grass which are tightly bound. The coils of the better baskets are wrapped with the grass called osoka; this wrapping fastens each coil to the preceding coil. Coarser baskets have the coils wrapped with strips from the leaf of the screw pine (emanalalo). There is a sewing process in which the coils are bound to one another with the bark olondovi, which is kept damp during the process. I have seen a needle, threaded with bast, used for sewing coils together; the needle was rethreaded every time it passed through a coil. Success in basket-making depends largely on the ability of the worker to keep the coils of uniform thickness. There is constant inspection and plucking out of a strand of grass here and there (Plate XXXIX, Fig. 2). In making basket trays (ongalo), the same coiling process is followed, but the work is kept flat. These trays are used for winnow- ing corn. Wicker-work is used for making conical traps for small game such as hares and rats. Conical fish traps are made of wicker- work. Names of patterns, which are of a simple geometrical kind, are well known to basket-makers. Sometimes a pattern is woven round the basket at a certain height to indicate a generally accepted standard for measurement of capacity. 172 The Ovimbundu Among the Ovimbundu are to be found baskets which they them- selves seldom copy. These examples are from the Vachokue country of eastern Angola. The Vachokue specialize in making winnowing trays and rectangular baskets which are twilled by passing each weft over two or more warps. By this technique, combined with the use of brown and black weft and warp, a great variety of geometrical patterns is produced. Trays of this kind vary in diameter from 50 to 200 cm. The Vachokue also make large strong trays in which honey is exposed to dry. In the neighborhood of Cassanga wicker-work baskets are made. The warps or stakes are rigid while the more flexible wefts bend in and out. At this place there are both Ovimbundu and Vachokue inhabitants. Women of the Vakuanyama in southern Angola make exceptionally neat little conical baskets by lashing coils very tightly. The lashing is done in such a way as to leave the outer surface covered with small chevron-shaped designs. There is an intrusion of European influence which tends to mar the work of the Ovimbundu and other native craftswomen. Soft dyes produced by methods described are sometimes replaced by brightly colored, imported dyes; old ribbons from typewriters are soaked in order to extract coloring. The Ovimbundu are imitating European forms of basketry in a few centers, though the native method of coiling is still used. WEAPONS The bow of the Ovimbundu is made of hard red wood which takes a high polish after use. It is round in cross section and tapers considerably toward the ends. The length is usually 150 cm, not an inconvenient size in view of the fairly open bush through which the hunter has to make his way. The bowstring is made of a thin strip of twisted hide, which is looped over each end of the bow by a slip knot made of two half-hitches. A shoulder at each end of the bow shaft prevents the slip knot from passing down the shaft. In some examples only one end of the bow stave is notched. Usually, after completing the slip knot, the hunter leaves a surplus of bow- string which is wound round the shaft. One bow which is not notched has rattan wound round the stave to prevent the loops from slipping. The bow for shooting bird arrows is small and is used only by boys; the string is of twisted vegetable fiber. Economic Life 173 The foregoing description applies to all bows collected in Elende, an Umbundu center; neither is there any appreciable difference between these bows and those of the Vachokue of the area from Cangamba to Saurimo in Lunda. The bows collected in the Vachokue region are on the average 10 cm longer than those collected at Elende, a structural difference probably connected with the heavier arrowheads used by the Vachokue (Plate XVII, Figs. 1-9). The Vakuanyama of southern Angola make and use a bow which bears very little resemblance to those already described. The arrows, too, are entirely different from those used by the Ovimbundu and Vachokue. The lengths of three bows collected at Mongua, a typical Kuan- yama center, are 111, 123, and 123 cm. The bow stave, which is made from a monocotyledonous wood, is 5 cm broad in the widest part, while the cross section is a flattened ellipse. The bowstring is of twisted leather, looped at each end for slipping over the ends of the bow shaft (Plate XVII, Fig. 8). According to L. S. B. Leakey (A New Classification of the Bow and Arrow in Africa, J.R.A.L, LVI, 1926, pp. 259-294), the bows and arrows of Angola have not been studied. I am inclined to place the bows of the Ovimbundu and Vachokue with Leakey's "knotted string bows" (pp. 266-269). Leakey states that the technique employed in stabilizing flight forms the best basis of classification of arrows, but so far as Angola is concerned, the shapes of arrowheads form a basis of classification according to locality. Ovimbundu, Vachokue, Vakuanyama, and Vasele, have distinctive patterns. Naturally there is borrowing of patterns where Ovimbundu and Vachokue intermingle, but, even so, there are distinguishing signs of Vachokue workmanship. The information respecting bows and arrows of Angola is best presented in paragraph form, and comparison of types is facilitated by refer- ence to Plate XVII, Figs. 1-9. The following arrows are used for killing birds: (1) Elende. Ovimbundu boys use a wooden arrow with a heavy blunt head which is carved into five nodules (Plate XVII, Fig. 3). The head is tanged into a reed shaft and bound there with fine fiber covered with wax. The feathers are whole, bent over, and tied. The number of feathers varies considerably in different examples. (2) Kipungo. The Vakipungo have a wooden-headed arrow with nine well-carved barbs (Plate XVII, Fig. 7). The head is socketed 174 The Ovimbundu fairly tightly into a reed shaft bound at the junction to prevent further splitting of the shaft. There are four whole, untrimmed feathers at the butt; these feathers are bent over and bound. (3) Vakuanyama. Wooden arrows having eleven sets of barbs are in use. The head is bound into a reed shaft with strong gut. Another type of wooden arrow has three sets of barbs with three barbs in each set. There are four whole feathers at the butt. (4) Vasele. The Vasele use an arrow, pointed with fine pieces of sharp bamboo, for shooting lizards (Plate XVII, Fig. 9). The foregoing information (1-4) relates only to wooden arrows which are used by boys when hunting birds. The following para- graphs summarize the structural details of arrows used by men, and the data are arranged to call attention to differences in the patterns of arrows used by several principal tribes of Angola. The Vasele make the arrow shafts of hollow reeds, and into these the narrow, ovate, iron heads are tanged and bound in position with fine bark which is not waxed. Three unsplit feathers are bent and tied to the shaft with fine bark fiber. The nock at the butt of the arrow is rectangular in shape and 0.3 cm deep. No quiver was observed. The Ovimbundu form arrow shafts from hollow reeds, the average length of which is 76 cm. All arrowheads of iron are tanged and the tangs are inserted in the hollow reeds. A binding of fine bark fiber is given, and this is covered with wax. From five to ten feathers are used for each arrow. The feathers may be split or they may be left in the natural state. The trimming of the feathers is roughly executed. The depth of the rectangular nocks is 0.5 cm, and splitting of the shaft is prevented by binding the nock with fine bark. No quiver was observed. The Vachokue of Cangamba, and thence northward to Saurimo, make arrow shafts of hollow reeds whose average length (79 cm) is a little in excess of those made by the Ovimbundu. The shapes of the iron heads are shown in comparison with those of the Ovim- bundu (Plate XVII, Fig. 1). All arrowheads are tanged and bound into the hollow shaft with bark fiber which is then waxed. The feathers are split and bent with a neatness much greater than that shown in examples of Umbundu arrows. The rectangular nocks are 1 cm deep. No quiver was observed. The Vakuanyama make solid wooden arrow shafts having an average length of 65 cm. The arrowheads of iron are socketed. Economic Life 175 Three prevailing shapes of arrowhead are conical, narrow ovate, and V-shaped. The quills are split, bent over, and tied. The usual number of feathers used for each arrow is eight. The workmanship is neat and symmetrical. The rectangular nock is 0.5 cm deep. A quiver of hide is used. In addition to the assagai (Plates XIII, Fig. 8; LXVIII, Fig. 2) commonly used by the Vakuanyama and adopted to a very limited extent by the Ovimbundu, there are spearheads which originate in eastern Angola among the Vachokue, whose blacksmiths are expert. Typical spearheads were collected at Munyangi where Ovim- bundu and Vachokue cultures meet. The leaf-shaped blade and tang have a length of 30 cm and an average breadth of 4 cm across the blade (Plate XVI, Fig. 1). The upper part of the tang, made for insertion into a wooden shaft, is round in cross section while the lower part is square. The blade, which has a keen edge, is well graded from a central, raised midrib to the margin. I did not observe many spears among the Ovimbundu and Vachokue. The former value highly a type of assagai which is owned by every man of the Vakuanyama. At Elende and Bailundu I obtained from Ovimbundu men assagais identical with those collected in the Vakuanyama country. The assagai is made entirely of iron; it has a narrow, leaf-shaped iron blade which is socketed to the iron shaft. The butt is sharply pointed. The shaft passes through a cow's tail on which the tuft of hair remains. Consequently the iron shaft is encased in hide for part of its length. At Elende I obtained from Ovimbundu men two spears said to be a product of the local forge. The flat, narrow, leaf -shaped blades were tanged into very rough wooden shafts. Wax was thickly smeared over the junction. The workmanship, which followed the tanged method of hafting arrowheads, was extremely crude. By far the most distinctive knives made in Angola are those manufactured and used by the Vakuanyama of the extreme south (Plate XVI, Fig. 10). I have never found this knife in use elsewhere in Angola, and similar types are to be found only among the Ovambo to the south of the Vakuanyama. These knives vary in length from 48 cm to 73 cm, with a breadth of 5 cm to 7 cm across the scabbard. The general outline is a well- balanced ellipse. The wood used is hard in texture and dark red in color. On one side the scabbard is left open in such a way as to display the blade, which is long, keen, and tapering. A leather thong attached to the back of the scabbard provides means of attach- 176 The Ovimbundu ment either to the arm or to the belt of the wearer. The crescent- shaped expansion at the tip of the scabbard is sometimes held between the toes while the blade is withdrawn. A small knife obtained from an Ocivokue man near Saurimo in the province of Lunda, northeast Angola, has a black wooden haft neatly bound with fine brass wire (Plate XVI, Fig. 2). The steel blade of Vachokue workmanship is eminently suitable for the purpose for which it is employed, namely, that of carving pipe bowls and snuff boxes, which are sometimes elaborately incised. The knife has a distribution from Saurimo to Cangamba, an area of intermittent Vachokue culture, but I have never observed it in the possession of an Ocimbundu. Men of the Ovimbundu have knives of somewhat poor quality. The roughly made wooden haft is attached to a blade which is protected by a sheath of lizard skin. More frequently than not, an Ocimbundu does not carry a knife; neither does he appear to have borrowed knives or the art of making them from the expert Vakuanyama or Vachokue, who are reluctant to part with their tools and weapons. A knife used by Vasele men in the region of Vila Nova de Selles, in the hinterland of Novo Redondo, resembles one used by the Bangala of the Congo (Plate XIII, Fig. 1). The distribution of the implement, which is used for cutting branches from trees, is local in the Esele country. I have never seen such a knife in use among the Ovimbundu or in any other part of Angola. The preservation of this peculiar form, in common with other specialized traits of Esele culture, is due to isolation of the Vasele among hills difficult of access. LEATHER WORK Leather pouches worn on a broad leather belt are part of the essential equipment of a hunter. The pouch usually contains scrap metal and powder for muzzle-loading guns. This type of firearm is, under present Portuguese regulations, difficult to obtain and still more difficult to furnish with powder, which is forbidden to the native. Consequently such a pouch may contain only a pipe and tobacco. The pouch itself is either square, rectangular, or semi-cylindrical in shape (Plate XIII, Fig. 6). The hair may or may not have been removed from the hide. Some examples show signs of careful work- manship in stamping cross-shaped patterns. Brass-headed nails are used for decorative effect on these pouches among the Ovimbundu of Elende and as far south as Huila. Economic Life 177 In addition to pouches, hide is used by the Ovimbundu for the seats of four-legged wooden stools. Before stretching the hide over the wooden frame the edge of the leather is cut into strips which are interlaced on the under side of the stool (Plate XVIII, Fig. 1). There are few evidences of the hide having been dressed; usually the hair is attached. The Ovimbundu understand the dressing of hides in a crude way. The hide is soaked, after which the inner surface is rubbed with ashes. The hair is scraped away and the hide is tightly pegged out. Hide was formerly used for covering wooden boxes in which a king carried his powder and possessions to war. Scraped hide is used to form the tympanum in tubular wooden drums. In only one part of Angola, notably in the south among the Vakuanyama, is leather used as clothing. Women wear pleated skirts of hide and several broad leather belts, while the men have small leather aprons in front and behind. Men only are the leather workers. The hide is soaked and trampled under foot for many hours in order to soften it and make it pliable; the hair is not removed. Before leather clothing is worn it is thickly greased with a mixture of fat and red powder from tukula wood (Plates LXV, Figs. 1-3; LXVII, Fig. 1). SPINNING AND WEAVING My interpreter Ngonga remembers the use of a heavy upright loom about twenty years ago, but this loom is not used at the present time among the Ovimbundu. The Ovimbundu have been in contact with Europeans for three centuries. This has not been a close and permanent contact, but it has been sufficient to account for the disappearance of a one-time essential art. There are in Elende some very fine raffia palm trees, but no use is made of the leaf fiber for manufacturing clothes. Bark cloth is not made by the Ovimbundu of Elende, but I have seen the Vangangella and Ovimbundu of Ngalangi engaged in stripping, beating, and preparing bark for use as clothing, which is formed of one piece wound round the body. The bark cloth is not dyed or decorated with patterns. Frequently at Elende men are seen engaged in spinning cotton which is used for repairs, and not for fabricating garments. The cotton is obtained from a shrub which has reddish leaves and yellow flowers. The cotton may be seen protruding from the calyx in the months of June and July. The man who is winding holds high in his left hand a slender stick which is wrapped for a length of two feet with fluffy cotton. All work of pressing and twisting the cotton into a thread is done by the thumb and index finger of the right hand. 178 The Ovimbundu In the beginning a thread from the fluffy mass on the stick is fastened to a corncob or to a potato, the weight of which keeps the thread taut (Plate XXXVI, Fig. 1). I have never seen a woman spinning and am informed that in former days the working of the loom was a task for men only. In presenting facts bearing on the economic life of the Ovimbundu attention has been paid to the importance of ritual, specialization in industry, and division of labor on a sex basis. Clearly, the economic life is one of the fundamentals of social structure, while at many points it is a field in which the magician works. Topography and climate have fixed certain conditions which, so far as the Benguela Highlands are concerned, favor agriculture on an extensive scale, also the keeping of cattle. In the Benguela Highlands a moderate temperature, combined with sufficient rainfall, has produced a type of vegetation which provides timbers serviceable to the craftsman, together with bush that affords shelter for game. The bearing of these economic facts on the probable origins of traits, and their assembly to form the tribal life of the Ovimbundu, is reserved for discussion in chapters dealing with culture contacts and cultural processes. In chapters X-XII the factual material presented here under "Economic Life" is correlated with what is known of similar activities in cultures surrounding that of the Ovimbundu. VI. SOCIAL LIFE Sexual Relations courtship There is a certain humor and quaintness of language used by Ngonga in describing sexual relations of the Ovimbundu. So far as possible the interpreter's style of expression has been preserved because of its value in revealing the native attitude. When a man begins to like a girl, he looks at her for several days. Perhaps the boy will talk to the girl who will tell him to go to her father and mother. The boy must ask the parents for a friend- ship, which may last for one or two years. During this time the boy must not do anything to the girl, and if he tries she ought to tell her parents. If a baby were born to them in this time it would be a shameful thing. After the boy has told the parents that he wishes to marry the girl he must find something to give to them. The girl must have agreed to marriage before the boy asks the parents. The first token may be no more than a ball of wax or a present of salt. This token says, "The girl is mine," and no other boy will ask for her because it is understood that she is promised to somebody. In former days there were restrictions against marriage with other tribes, but in eastern Angola at the present day the Ovimbundu occasionally marry with the Vachokue. Soon the boy will start building a house on his father's plot, because he intends to bring the girl to his father's land. The boy must at this stage make a further present to the parents of the girl. In the old days he would take about five kilos of salt; but now he will take a bottle of Portuguese wine, two or three blankets, or a piece of cloth. The people who carry the present must be the brother, father, or a male cousin of the suitor. These relatives will say, "This is the day we have come for our boy's wife." Then the parents will examine the things. Perhaps they will say, "You must bring a better blanket." The father or the girl's mother's brother must call the relatives of the girl to a council (onjango) where the relatives of the boy and girl are gathered. The parents say, "We are taking these things for our daughter; we hope she will be a good girl and not shame us. She is a good girl to us and we hope she will be a good girl in your house." The girl's parents turn to her and say, "We should like to hear that you are hospitable; give food to your husband's relatives when they visit 179 180 The Ovimbundu you." There is no infant betrothal. There is not and never was any compulsion of a girl in marriage, but slave girls were disposed of in marriage by their masters. A girl is not allowed to do anything to show that she loves a boy, for it would be a great shame for her to tell the boy that she loves him. Often a girl who prefers a boy will pretend that she does not like him. The Ovimbundu have a story which states that a man said, "I will bring my cow to the green grass" ; he did so, but the cow would not eat. This expresses the idea that a boy would not like a girl who confessed a preference for him. MARRIAGE The prospective bride chooses one married woman and six unmar- ried girls to accompany her to the house prepared by her husband. Here a feast consisting of a pig and some chickens is provided by the husband for the relatives of both families. For three nights the girl returns to the house of her parents while the boy sleeps at his home. The married woman and six girls sleep at the house prepared by the bridegroom. During these days beer is provided by the boy's parents. The prospective husband is ironically addressed as sandombua. Ndombua means bridegroom, sa is an abbreviation of isia meaning father. The term "father bridegroom" refers to the fact that the youth is a potential husband only; the marriage has not been consummated. On the fourth day the bride brings her supply of domestic utensils. These are the cooking pots (olombia) ; the wooden spoons (ovito) ; the brush of grass for sweeping (olueyo); some meal; also the pounder (upi). For the first month the wife is not allowed to cook in her own home; she cooks food in the home of her husband's parents and sends it to the council house (onjango) where her husband takes his meals with other men according to Umbundu custom. At the end of the first month of married life the mother of the husband invites any three old women who have been happily married to lay the hearth stones in the new home. Each of the old women brings a stone for the hearth. A chicken is killed and its blood is sprinkled on the hearth stones. While the young wife is preparing food at the new hearth, she is helped by the old women. If the girl is stirring with the big wooden spoon, one of the old women places her own hands over those of the girl. There is this kind of guidance in every action. I understand that at the present day virginity in a bride is not so highly valued as in former days. The old custom was an examination Social Life 181 of the girl by her husband, and if she were not a virgin he took a hot stick from the fire and burned a hole through her loin cloth. "The girl began to cry, but she had to take the burned cloth to her mother." In such an instance there does not seem to be an idea of guilt. The husband had accepted something that was damaged, and the payment of a pig by the girl's parents reunited the two young people. In former days, also at present, there are boy and girl companions who sleep together, supposedly without having sexual connections, although they may be seventeen years of age. The girl calls the boy ombaisi, and he calls her by the same name. The Ovimbundu understand something of the physiology of conception; the woman is, however, regarded as only a receiving vessel. "The man puts something into her which grows." Husband and wife do not sleep in the same bed during the wife's menstrual period ; the wife sleeps on a mat at the side of her husband's bed. A woman who is menstruating never cooks food, but women give mutual aid in this matter. A man with more than one wife sleeps either four nights or seven nights with each; the four-night cycle is more usual than the seven-night cycle among the Ovimbundu. Each wife has a separate hut and kitchen. There is no wife lending, but a visitor may be provided with a widow or even with an un- married girl. Then the man would have to pay the woman. Ngonga was able to give information with regard to homosexual- ity. "There are men who want men, and women who want women." Ngonga says he has heard people talk about it, and "they think this very bad." A woman has been known to make an artificial penis for use with another woman. The medicine-man will sometimes dress as a woman. Ngonga, who has seen a man dress as a woman, stated that the man arranged his cloth like that of a woman, put palm oil on his hair, and joined the women to pound corn on the rocks. "The other people laughed and spoke bad words to him. His brother, father, and uncle beat him," but without producing reformation. DIVORCE There are many grounds on which a man may secure a divorce, but it does not follow that divorce is frequent. On the contrary the evidence indicates that the majority of difficulties are overcome by a compromise between the relations of the husband and those of the wife. The main causes of dissatisfaction with a wife are want of ability in cultivating her garden, physical weakness, a habit of thieving from the gardens of other women, incompetence in cooking, 182 The Ovimbundu bad temper, too much talking, some physical defect arising from childbirth, and infidelity. But the husband usually accepts payment from the adulterer, and in that event divorce is not sought. If a mother has no milk, there is a likelihood that her children will die in infancy ; this is a cause for divorce. Barrenness gives great dissatisfaction, but is not necessarily a cause for divorce. Usually the husband marries another girl, while the first wife retains her position as head-wife. With regard to frigidity my informant said that sometimes "a married woman does not want her husband to do those things which husbands like to do, or she may want him very seldom." The husband is so angry that he may go away hunting for a long time. The husband may tell some old people, who talk to the girl. If his wife is a good cook, the husband may keep her and secure another girl as a second wife. Sometimes the husband ties the hands of the girl if she resists him. If a wife is returned to her parents on the grounds of her frigidity her husband does not receive compensatory payment, but he hopes that another man will take the girl and pay him for her. Sterility of the male is sometimes suspected, in which case the husband tries another girl, possibly with his wife's consent. The wife may be allowed to have relations with another man in the hope of producing children. The evidence shows that there is definite experiment to test barrenness of the woman and sterility of the male. In case of impotence (not sterility of the male sperm) the wife leaves her husband. In event of barrenness the woman visits a medicine-man, who gives her a charm consisting of two cowrie shells on a strip of leather which she wears round her neck; a potion also will be given her. If the barren condition is due to malevolence of spirits, the face of the woman is painted with streaks of red and white. Such a woman is said to gain ocitumba. Tumba means "a swell," "a rise." A woman may divorce her husband if he does not treat her well. If he beats her or refuses to give her cloth and palm oil, she will leave him; but she will not leave him if he is merely unfaithful. If a woman is unhappy with her husband, she will tell her people about the trouble. Her father and mother may say, "Go and try again." After a year the woman may still be unhappy, in which case she goes to her parents. The husband visits his wife's parents to ask why she has left him. The parents give reasons and offer to return the token he presented for the girl. The return of the husband's tokens is usually long delayed because the parents of the girl are Social Life 183 hoping that another man will ask for their daughter; this new suitor will have to make payment to the deserted husband. The chief of the village is not consulted unless the return of the husband's presents is long delayed or is in some way unsatisfactory. A woman who claims even remote relationship with the royal family is treated well, because her husband is afraid of the influence which may be used against him. A woman who returns to her parents takes with her the articles she contributed to the home; these are pottery, corn baskets, a wooden pounder, and wooden spoons. If a wife returns to her parents without telling her husband that she intends going, he will beat her if he finds her packing up her utensils. The procedure of divorce contains a very human element. When a man has fully decided that he would like to divorce his wife, he will first of all inform his parents of his intention. The parents may advise their son to try the girl for a longer time. On the contrary, the parents of the man may be mischief-makers. Sometimes the parents will say to the husband, "Do you know that your wife is doing these things? It is better you should send her away." If the man is fond of his wife, he will take her to another place where his parents cannot watch her. When a woman divorces her husband to marry another man, she takes to her new home all children under three years of age. Older children go to the home of her parents. The chief of the village has to witness the final ceremony of divorcing a woman; but the husband and wife, also the wife's father, are the principal people concerned in the divorce ceremony. The husband receives from his wife's father a roll of tobacco and a pig, then he places leaves and palm oil on the back of his wife. He slaps her back saying, "It is finished," after which the woman goes to her father or to another man. When a woman takes the initiative and has declared her intention of divorcing her husband, she returns to her parents as described; but the divorce cannot be completed until the husband has been per- suaded to perform this ceremony of slapping her back and making a ceremonial renunciation in public. PREGNANCY AND CHILDBIRTH When a woman finds that she is pregnant, she makes and drinks an infusion prepared from bark fiber. This is to insure the removal of all stringy matter at delivery. Formerly a husband was not 184 The Ovimbundu supposed to have intercourse with his wife during her pregnancy, but this custom of continence is declining. A pregnant woman in particular must not steal. If she does so, her child will refuse to be born until some one present at the confine- ment goes out and steals something. Eating the flesh of a hare during pregnancy will give the baby a split lip. If the flesh of the owl is eaten, the baby will have large round eyes. The expectant mother must not sit on a mortar, a pestle, or a piece of rock; if she does so, labor will be prolonged. A pregnant woman takes earth from just outside her door. She drinks this in water so that the placenta will be delivered whole. A pregnant woman is not supposed to carry anything in her cloth. If she does so, the child will have a long head. A pregnant woman is a potential corpse. A man ought not to quarrel with his pregnant wife and if she says angry things to him he should not reply. He would not speak to a corpse in anger; on the contrary, he would respect a corpse; he must, therefore, respect a pregnant woman. It is a bad omen to see a pregnant woman up in a tree. A man who sees such a woman is expected to shoot her. If a pregnant woman has scolded her husband and is sorry, she goes out to the fields. There she gets her cloth covered with burrs then returns home. If her husband begins to pick off the burrs there is reconciliation. There are several arrangements whereby a woman who has borne only girls may secure male births, provided she can find a woman who has borne only boys and is anxious to have a girl. The simplest way of reversing the births is for the women to exchange belts. The belt is a string, or possibly a plaited fiber girdle an inch in width which is worn next to the body. From this string depends the woman's lower garment. A second way of reversing the births is for the women to make an exchange of food through a hole in the wall of a hut. The food is handed in on a basket tray. The woman inside the hut receives the food while standing with her back to the hole; she then places the food under the bed. The tray is handed back through the hole by the recipient who still keeps her back to the aperture. There is a third method whereby a woman who has borne boys may be made to bear girls, and vice versa. The woman who has borne only boys gives to the woman who has borne only girls, an arrow, a knife, a bow, and an ax. The articles given in return are a pounding stick, a broom, a tray and a basket. The value Social Life 185 of this symbolism relating to occupations of males and females respectively, is obvious. Many women bear children when they are away in the fields, but help at home is often given. The mothers of the wife and of the husband may not be present at the confinement. No childless woman may be near during the confinement or convalescence. The father may not be present or "the child would be ashamed to be born." Before and during her pregnancy a woman feeds her husband with a prickly plant which is mixed in his food; this makes him faithful to her. The abdomen is bound after delivery in order to keep the uterus in position. There is a very small amount of bleeding after childbirth; in fact the flow of blood is quite finished in two or three days. The nearer the diet comes to that of the white man the greater the bleeding at delivery. A pregnant woman may not beat a drum, or she will bear a drum. A woman who sees the blood from circumcision of a male will not have any children. When a female child is born, the umbilical cord is cut with a hoe to ensure that the female will be a good worker in the field. The cord of a male is cut with an arrow to insure good hunting. A newly born child receives a drink of beer, and a cord is tied round its waist. This is not for support; it is the string from which, much later, the lower garment hangs. Girls do not menstruate until they are fifteen or even seventeen years of age. During the months after first men- struation a girl advances rapidly from childhood to womanhood. The medicines that women take to secure abortions are bitter. A woman will refuse quinine because she thinks it will cause abortion. It is certain that twins are welcome among the Ovimbundu, but there are special observances connected with their birth and death. A twin birth is not thought to imply dual fatherhood. Twins are called Njamba ("elephant") and Hosi ("lion"). Although twins are of opposite sexes each receives one of the two names. Njamba is the first born, and Hosi is the second to be delivered. The medicine- man holds a ceremony to cure the mother of twins. The afterbirth of twins is placed in two gourds and carried outside the village for burial by two midwives. The mother of twins must wear round her neck the horn of an antelope given by the medicine-man. She has to blow this when crossing a river, meeting a crowd of people, or seeing a hawk overhead. People laugh at her and in fun say that she is a pig or a bitch. "The woman says the same kind of words to them." Children may be adopted, but they never really 186 The Ovimbundu belong to the foster parents. When an adopted girl marries, half the presents from the husband are given to the natural parents, and the other half to the foster parents. In a case of triple birth at Ngalangi two infants died at birth; the other succumbed after three months. In another instance a woman bore three children and in the vernacular of my informant "something which was nothing," but all died. In a third case of triple birth a woman had a boy and two girls. The boy died at the age of fifteen years; the girls grew up and bore children. Albinos do not find it easy to obtain wives. I was told that an albino man at Ngalangi could not find a wife until he married a widow who had several children, because unmarried girls of his own age did not want him. One of the charms collected has to be worn by a woman who has triplets. The object is a rattle (olusangu), which the woman shakes when she meets any one. If she gave an ordinary greeting the children would die. The greater part of this information was obtained at Ngalangi by interrogating women who were questioned in Umbundu by Mrs. McDowell. At Elende I asked my interpreter Ngonga to interrogate his wife. According to Ngonga a woman must visit a female prac- titioner in the first month of her pregnancy. The face and the body of the pregnant woman are painted with red, black, and white spots. This means that she will have no difficulty in bearing children, neither will she have any sickness during the nine months of gestation. "When her belly hurts" (i.e., the quickening), the woman goes to the same female ocimbanda, who paints lines of white, red, and black across her breast. The ocimbanda gives the woman a necklace consisting of a strip of leather to which two cowries are attached. Experience proves that such a necklace is difficult to buy. If the woman has a necklace of this kind which was worn by her grandmother, it is a very powerful aid to conception, a relief in painful menstruation, and a means of securing easy delivery and normal gestation. I could not find any trace of an idea relating to the entry of a soul, ancestral or otherwise, into the foetus. Ongandu is the name for a disease of the genitalia; but Ngonga used the word to describe abdominal pain suffered by a woman who has had no children. By this I suppose he means painful menstrua- tion. The curative root which is given is called kayambua. Abortion is never secured by mechanical means. The medicine (ihemba) is made from the root of a plant that is boiled in water and Social Life 187 drunk. The concoction is described in Umbundu as ihemba vioku tundisapo imo ("medicine to take away belly"). When the monthly period begins too early in life, the medicine- man recommends that the young girl should wear the cowrie necklace worn by her mother's mother. A deformed child is destroyed, but not if it has been allowed to survive the first day. If male triplets are born, two stay permanently with the mother. At the age of five years one child goes to the king, whose child he becomes. When a woman has had relationships with more than one man, she will die in childbirth unless the medicine-man is called to cure her. In a Vachokue village of eastern Angola I bought a female wooden figure. It was explained that this would be nursed by a woman, one of whose twins was dead, in order to induce another conception. Moreover, the nursing of the figure prevents the death of the second twin. The Ovimbundu also use these wooden figurines for replacing dead twins (p. 163). Children are suckled for a long time; even those of three years and older come to the breast. There are instances in which milk is present but lacking the nutritive qualities; then the baby is likely to die. The matter is simpler when a mother gives no milk at all because the baby is given to another woman, but not necessarily to a woman who is suckling a child. A baby whose mother has no milk may be given to a woman who has not borne a child for many years, and the sucking of the child quickly induces a milk supply. This is agreed upon by informants at Ngalangi and Elende who have seen a child nursed by a woman with withered breasts. Protection is given to the fontanelle by covering it with a vegetable gum which hardens. Near the chief village of the Vangangella at Ngalangi there is a mound decorated with feathers and painted wooden posts (Plate XCII, Fig. 1). Childless women are placed on the mound which is near a river. They are covered with mud, after which the medicine- man sings songs and administers potions. The women go home and are made to sit on mounds in their kitchens. These mounds, which are made in rows, like earth heaped up after hoeing a trench, may be a symbol of successful agriculture and human fertility. On looking into the subject of blood brotherhood I found that an exchange of blood between two males who swore mutual fidelity was at one time common. At the present day an exchange of blood is 188 The Ovimbundu sometimes made between husband and wife, at night and in secret. People say that those who exchange blood will die at the same time. NAMING In addition to the words chosen to describe twins, there are some points of importance in connection with the naming of children. The father and mother change their names when the first child, male or female, is born, but there is no change of name at the birth of subse- quent children. In a certain family the name of the first child, a girl, was Vitundo. The name of her father, which was Cingandu, was abandoned; he became Savitundo, "the father of Vitundo." The mother's name of Visolela was changed to Navitundo, "the mother of Vitundo." If the first child dies, the parents dislike their names; they there- fore revert to their original names. When another baby is born, the parents again change their names in the way described. A post- humous child is called Lusati. A child born after twins is Kasinda, which means ' 'to push . ' ' Twins are cal led the Lion and the Elephant, or the Elephant and the Hippopotamus. There are no secret names. The names of the dead must not be mentioned ; the deceased is referred to as "the one who has gone." Children may change their own names at the age of about sixteen years, and actually do so if their names are distasteful to them. Ngonga's friend, named Katito ("little"), changed his name to Mukayita (meaning not known). Ngonga's sister, named Ndumbila (meaning not known), changed her name to Cilingohenda, which means "It is a pity." I met a chief near Bailundu who was called Kandimba, meaning "the Little Hare." There may be a change of name during sickness. A man now named Katahali suffered sickness in addition to other misfortunes. His sickness recurred, so he changed his name from Kopiongo to Katahali. The meaning of the former name is not known. The new name, Ka tala ohali, means "He who has seen trouble"; Katahali is an abbreviation. Another instance of change of name, also in the village of Cilembo, was that of a man who changed his name from Lumingu to Kaihemba, which means "the one who lives by medicine" ; because without medicine he would have died. A sick child may receive a bad name, for instance, the name Pig. If one or more children have died a subsequent child receives an ugly name with a bad meaning. There is no totemism, but children may be named after animals. A girl is sometimes named Kambundu ("a little frog"). Other names for females are Esenje ("the rock where corn Social Life 189 is pounded") and Cisengu ("a small bird with a long tail"). A boy- may be Kangwe ("the little leopard"). Names sometimes give an indication of descent. Ngonga's full name is Ngonga Kalei Liahuka, Ngonga ("eagle"), Kalei ("one who works for the king"), Liahuka (the father's surname). Ngonga's sister is Cinyaiiala ("the old basket"). This was the name of her father's father's sister. As the father of Cinyaiiala is Liahuka, the daughter is surnamed Yaliahuka (ya, "of"). The father chooses the names of the three first children whether boys or girls. The mother chooses the name of the fourth child whether male or female. If the child is a boy, the mother probably chooses the name of her brother or of her father's brother. When a first son is born, the father usually gives the name of his father; for example, Ngonga's father's father was Ngonga. If the first baby is a girl the father chooses the name of his sister. Ages are not known, but reckoning of age goes back five years by counting the number of times maize has been sown; maize is planted each October. The period from sowing to sowing is ulima. If a man has a child by a woman who is not a wife or a concubine, the woman keeps the child for eight or ten years. The man must give the mother a cloth in which to carry her illegitimate baby, also oil for her hair. Sometimes the girl will go to the father of her child to be his concubine, but her parents will not let her do so if he has a bad reputation. To bear a child out of wedlock is a disgrace. Terms of Relationship In preparing the following tables Ngonga was the ego, or male speaker, and each term is given in relation to himself, with its reciprocal. The tables give firstly Ngonga's own generation, then his ascendants, and finally some of his descendants. Tables E and F are diagrammatic forms of tables A-D. TABLE A (See Table E III) Ngonga's Own Generation Terms in italics are Umbundu names either for persons or for kindred classes. Reciprocals are placed in brackets; W.S. means, "woman speaking." Ukai wange is my wife; uketu, which means "spouse," is a modern form of address for husbands and wives when speaking to each other (veyange, my husband). Kota or huva is my elder brother (mume wange or manja, younger brother). Kota or huvange is my elder sister. An elder brother speaking to his sister calls her by name, or he uses the term mbuale. When speaking of her he says mukai wange (mume wange or manjange, younger brother). The same terms are used in the same way to apply to my father's brother's son, and my 190 The Ovimbundu father's brother's daughter. Similarly the terms manja or kota are applied to my mother's sister's son. The former is used if this relative is younger than myself; the latter term is employed if the relative is older than myself. Mukai is the term for my mother's sister's daughter (reciprocal, manja, kota, or huva, means mother's sister's son). Manjange is a general name for a mother's sister's child, male or female. Upalume describes my father's sister's son (upalume, mother's brother's son). The word is also applied to my father's sister's daughter. Marriage with a father's sister's daughter is permissible, but it is not favored since the offspring of such a union may be stupid. Upalume also designates my mother's brother's daughter, who, according to custom, is regarded as the most suitable spouse for me. Cepua cange. This term is applied to any child of my mother's brother and to any child of my father's sister (reciprocal, cepua cange). Nawa. The term is used as follows: for my elder brother's wife; (W.S.) my husband's younger brother; my younger brother's wife; (W.S.) my husband's elder brother; my elder sister's husband; my wife's younger brother; my younger sister's husband; my wife's elder brother; my father's brother's son's wife; (W.S.) my husband's father's brother's son; my father's brother's daughter's husband; my wife's father's brother's son; my father's sister's son's wife; (W.S.) my husband's mother's brother's son; my father's sister's daughter's husband; my wife's mother's brother's son; my mother's brother's son's wife; (W.S.) my husband's father's sister's son; my mother's brother's daughter's husband; my wife's father's sister's son; my mother's sister's son's wife; (W.S.) my husband's mother's sister's son; my mother's sister's daughter's husband, my wife's mother's sister's son. This completes the terms of relationship for Ngonga's own generation. TABLE B (See Table E II and IV, also Table F II) First Generation op Ngonga's Ascendants Tate. The term is applied to my father; my father's brother; and my mother's sister's husband (the word omolange, my child, is the Umbundu reciprocal for the English reciprocal terms, son, brother's son, wife's sister's son). Mai. The word is applied to my uterine mother, my mother's sister, and my father's brother's wife (again the Umbundu omolange is the reciprocal for the English reciprocals, son, sister's son, and [W.S.] husband's brother's son). Aphai means my father's sister, and the term is said to designate a "female father." My mother's brother's wife is also aphai (the Umbundu reciprocal for either male or female is ocimumba cange, which is the equivalent of the English reciprocals, brother's son and husband's sister's son). Manu or inanu is my mother's brother (reciprocal is ocimumba cange, which means sister's son). Cikulume. The term is applied to my father's sister's husband (the reciprocal, ocimumba cange, means wife's brother's son). This completes the first generation of Ngonga's ascendants. TABLE C (See Table E I and F I) Second Generation of Ngonga's Ascendants Sekulu yange. The term is applied to my father's father, and literally means my older father (the reciprocal is onekulu yange, meaning son's son). Sekulu. The word designates my mother's mother's brother, and my mother's mother's sister's husband (the reciprocal, onekulu, means sister's daughter's son, and wife's sister's daughter's son). Social Life 191 Kukululu or sekululu. These terms are applicable to my father's father's brother, my father's father's brother's wife; my father's father's sister's husband; and my mother's father (the reciprocal onekulu describes the reciprocals, brother's son's son; husband's brother's son's son; wife's brother's son's son; daughter's son). Maikulu. This designation, which is derived from mai, meaning mother, and kulu, an old person, is applied to my father's mother (reciprocal, onekulu yange, son's son). Maikulu also means my father's father's sister; my mother's mother; my mother's mother's brother's wife; and my mother's mother's sister (the reciprocal is onekulu which equals the English recip- rocals brother's son's son; daughter's son; husband's daughter's child; and sister's daughter's child). This completes the second generation of ascendants. Note: Ngonga's wife uses the names maikulu and kukululu for those relatives of her husband to whom Ngonga himself applies those terms. These rela- tives call Ngonga's wife onekulu. TABLE D (See Table E IV and V, also Table F II, IV and V) Some of Ngonga's Descendants Nunulu or uveli is my first-born son. Other sons are omola, meaning child. Omola ulume means a male child. Omola ukai is a female child. Uveli also means a first daughter (the reciprocal for these terms is tale, meaning father). Omolange means my child. I apply the word, not only to my own children, but to my elder brother's son (reciprocal, late, father's younger brother); to my elder brother's daughter; and to my younger brother's son. Ndatembo. The word is applied to my son's wife; my daughter's husband; my elder brother's son's wife; my elder brother's daughter's husband; and my younger brother's son's wife (the same word ndatembo is used for the reciprocals of these terms; namely, husband's father; wife's father; husband's father's younger brother; wife's father's younger brother; and husband's father's elder brother). The foregoing classificatory system of relationship is not peculiar to the Ovimbundu, but is a cultural trait of many Bantu-speaking and some Sudanic-speaking Negroes. The similarity of the Umbundu system to those of surrounding peoples in Rhodesia and the Congo will be pointed out in chapters dealing with culture contacts. When describing marriage rites reference was made to the fact that a wife goes to live near her husband's relatives; the system is therefore patrilocal. The limited family consists of husband and wife (or wives) with their children. The greatest number of wives observed in the family of a commoner was four, and in the family of a king eleven. A household sometimes contains adopted children, also domestic slaves (pawns) who are working to pay off debts for their maternal uncles. An extended family may include a grand- father and his wife, his sons, their wives and children, and his unmarried daughters, with classes and nomenclatures mentioned in the foregoing tables. The phrase epata lia tate (or aluse) means "family of my father" and includes all relatives on the father's side. The words epata lia 192 The Ovimbundu mat (or oluina) mean "family of my mother." These terms seem to indicate a bilateral rather than a unilateral lineage. Inquiry did not show that there were totems for the mother's or the father's people, or that marriage into any particular local group was com- manded or enjoined. A village is a unit under the administration of a chief (sekulu) and as such is part of a large group of villages forming a kingdom ruled by a king (osoma), but such village units are concerned with government and warfare, not with exogamy. In tracing descent an Ocimbundu gives the names of relatives of both the father and mother, but commoners are unlikely to know the names of their relatives beyond the grandparent class. The children of a king were at one time able to recite many generations of ancestors along both parental lines. But if the king had married a commoner, as he was allowed to do after choosing his first wife from the royal line, the children knew only the genealogy of their father. This is obvious since the commoner mother would not be likely to know her line of descent. The son of a village chief (sekulu) may marry a commoner, but, according to Ngonga, the sekulu, his father, "would have a very good look at her." Ngonga had not heard of the marriage of brothers and sisters, not even in the royal family; the idea of such a union was new and repulsive to him. In Umbundu society there would be no necessity for such a marriage as that of brother and sister or other close relative, because members of a royal family of one part of the country could intermarry with members of a royal family in some area far away. There might, for example, be marriages between the royal families of Bailundu and Ngalangi, which are two principal kingships of the Ovimbundu confederacy. When asked what would be the fate of a man who committed incest with his blood daughter or uterine sister, Ngonga said he would be killed by his brother or by his mother's brother. If he escaped he would have to go far away so that his people could never see him again. If a fine were accepted for incest, the culprit would have to pay his own kin "because he had shamed them." Table A, Ngonga's Own Generation, calls attention to a plurality of terms for brother and sister according to the relative ages of the speaker and the person addressed. Moreover, there is a term for direct address and another which is used when speaking of a brother or sister. Ngonga was questioned with regard to forms of address for his brothers and sisters other than the children of his uterine mother. Social Life 193 Ngonga actually has a brother by his father's first wife, and for this male he uses the same terms as for his uterine brothers. The same terms are used for uterine sisters and sisters begotten by his father through wives other than his uterine mother. Ngonga said, "If people ask you which sister or which brother you can explain it in words." A wife of Ngonga's father, other than Ngonga's uterine mother, is called mai yesepakai; that is, "the mother who is jealous of my mother." Mai means "mother," and the remainder of the term is a derivative from the word esepa, meaning "woman's jealousy." If two women A and B desire to marry the same man and only A is suc- cessful, B calls A sepakai. Under similar circumstances a man would call his successful rival cikuelume cove. There is a distinct word for man's jealousy. In explaining this Ngonga said, "When I see my wife look at another man, I have ukuelume ["man's jealousy"] in my heart." If on the death of Ngonga's father, his mother married again this male would be called by Ngonga tate yesepakai ("the father who is jealous"). A wife calls the children of the family, who are not her own, omala vesepakai; that is, "the children who are jealous of the other children." A mother-in-law taboo operates. Conversation between mother-in-law and son-in-law must always be carried on while the speakers stand back to back. The foregoing note on a mother-in-law taboo has been supple- mented by a letter from Dr. Merlin W. Ennis of Elende, Angola (August, 1931). "Mother-in-law and father-in-law taboos seem to be directed against seeing each other. The persons involved may not see each other. If they meet on the path, one steps aside and turns the back while the other passes on. The one passing by goes through the motions of seeing no one. If it is necessary to converse on some subject, they sit looking in different directions, or one sits out of doors and the other within, around the corner of the door. This holds equally for a man and his son's wife, and for a woman and her daughter's husband. A man may see and talk with his daughter's husband, and a woman may see and speak to her son's wife, but no son-in-law may eat with a father-in-law, likewise daughters-in-law and mothers-in-law may not eat together. Brothers- in-law may not eat together unless they have gone through a certain ceremony; this also holds for sisters-in-law." 194 The Ovimbundu Table A indicates a wide use of the word nawa for "in laws" of the speaker's generation. Thus Ngonga calls the wives and husbands of his brothers and sisters nawa without distinguishing them in any way. The term nawa has a still wider connotation for it includes Ngonga's father's brother's son's wife, and his father's sister's daughter's husband. Ngonga's wife said that she uses the word nawa for all the brothers and sisters of her husband, for all the children of her husband's father's brother, her husband's father's sister's children, her husband's mother's brother's children, and her husband's mother's sister's children. Reciprocally, all these people call Ngonga's wife nawa. Table A introduces the question of cross-cousin marriage, which is the functional form among the Ovimbundu. My informant said that he calls his mother's sister's children brothers and sisters, there- fore he could not marry the girls. Neither are his father's brother's daughters eligible for marriage with him. The table shows the truth of this, for Ngonga calls his father's brother's daughter mukai, mean- ing "sister"; she addresses him as kota or huva ("brother"). Ngonga also addresses his father's brother's son by the term manjange ("younger brother") or huvange ("older brother") according to the relative ages of the speaker and the person addressed. When Ngonga was asked whether he could marry a daughter of his mother's brother, he replied, "I could marry her very well indeed." A marriage of Ngonga with his father's sister's daughter would be permissible but Ngonga said the marriage is not regarded as a good one "because the children will be stupid." Ngonga could marry his mother's brother's daughter's daughter, or his mother's brother's son's daughter. It would also be permissible for Ngonga to marry his father's sister's daughter's daughter, or his father's sister's son's daughter. Ngonga said, "My father's brother is my father, and my mother's sister is my mother." When questioned further my informant said that a marriage with daughters of these relatives would make him ocinyama, which means "an animal." Relatives would say, "You have shamed the family." It will be noted from Table B, First Generation of Ngonga's Ascendants, that Ngonga's father's brother calls him omolange, meaning "my child," for which the reciprocal is tate ("father"). Ngonga's father's brother's wife is mai ("mother"), and she calls him omolange ("my child"). Ngonga distinguishes between his mother's oldest sister {mai yukulu), and his mother's Social Life 195 youngest sister (mat yumalele). Ngonga's father's oldest brother is tate yukulu. Thus far Ngonga was clear about his eligibility for marriage, and in addition to this he was sure that he could not marry a sister of his wife while his wife was alive, but he could marry his wife's sister, after the death of his wife. The wives of a deceased elder brother are divided among his younger brothers, or a man may inherit the wife of his mother's brother. Ngonga always laughed heartily at the idea of marrying a woman much older than himself. He did not know whether he could marry certain older people of the ndatembo and maikulu class but thought the idea amusing because of their age. In discussing law, inheritance, and slavery, the importance of the mother's brother will be indicated, and for this relative there is the term manu (or inanu). The manu calls Ngonga ocimumba cange, but he calls Ngonga's wife ndatembo. Ngonga's mother's brother's wife, also Ngonga's father's sister, are aphai, which my informant interpreted as a "female father." The use of the word nawa has been explained in such a way as to show that Ngonga's wife uses this term for a large class of relatives by marriage who are of the same generation as her husband, while these people reciprocally use the word nawa. Ngonga uses the word nawa in exactly the same way for corresponding relatives of his wife. This term nawa is not to be confused with the term ndatembo whose connotation is indicated by the following examples. The wife of Ngonga applies the term ndatembo to certain of her husband's relations who are not of his generation. Thus she calls Ngonga's mother ndatembo when speaking of her, but she addresses her mother-in-law as max ("mother"), while ndatembo describes the rela- tionship between Ngonga's wife and his father. Moreover, Ngonga's wife uses ndatembo to describe her husband's father's brother's wife, and Ngonga's wife is called ndatembo by her husband's father's sister, her husband's father's sister's husband, and her husband's mother's brother. These are relatives by marriage but they are not of the same generation as her husband. Table C, Second Generation of Ngonga's Ascendants, gives no particular difficulty. Maikulu includes the paternal grandmother, the father's father's sister (great aunt), the maternal grandmother, the mother's mother's brother's wife, and the mother's mother's sister. Therefore maikulu is a class name for the second generation of female ascendants. 196 The Ovimbundu o < s PQ § 2 § H Dh m a i»o J Social Life 197 i I C J! B 1 I i ■ H 1 E W h O h | M EC a z « o 198 The Ovimbundu Sekulu includes the father's father, the mother's mother's brother, and the mother's mother's sister's husband. The terms sekululu and kukululu were used by Ngonga as being synonymous, and therefore they have the same connotation. With regard to the term maikulu Ngonga seemed clear, but in reply to questions concerning the reason for having the terms kukululu and sekululu, he was evidently confused. At last he said, "It does not matter, they are the same people." Ngonga never hesitated in giving the reciprocal for all these terms, which is in every instance onekulu. These terms may have been correlated with different functions which have now become obsolete. The two most important collateral relatives in the parent's generation are the father's sister (aphai) and the mother's brother imanu). The term for father's sister means "female father." As such her relations to ego are similar to those of a father, consequently it is not considered right to marry her daughters. The mother's brother, on the other hand, is without doubt ego's most important relative. The two are linked by a series of reciprocal duties and obligations. Marriage of a mother's brother's daughter, or a mother's brother's widow, is a correct procedure. The value of the more important reciprocal duties is indicated in the following sections on law and government. There was no possibility of devoting more time to the matter of relationship terms, but there is here sufficient to indicate the remarkable persistence of the kinship system in spite of three centuries of contact with the Portuguese and other Europeans. Moreover, the information gleaned by interrogation of Ngonga and his wife gives the main points for comparison with systems of a similar kind that have been given for Ashanti, Uganda, and Rhodesia, by Rattray, Roscoe, and Smith and Dale respectively. These kin- ship systems are the same both in general principles and in con- siderable detail. The foregoing notes and tables were studied by Mr. Zachary Taylor of the University of Chicago, who prepared Tables E and F. Mr. F. Eggan offered useful criticisms. The Arabic numerals on these charts refer to the numbers used to designate terms of relationship on the list on page 199. The ego or speaker is Ngonga, my interpreter, whose direct line of relationships is shown in Table E, while Table F indicates the terms used by Ngonga for his wife's relatives. The levels numbered in Roman numerals on Tables E and F indicate the generation stratification as follows: Social Life 199 I. Grandparents. II. Parents. III. Speaker's generation. IV. Children. V. Grandchildren. Numbers on the left of the sign (=) refer to males, those on the right, to females. The terms of relationship used in Tables E and F are as follows: 1. Ukai wange. My wife (vocative). 1L Ocimumba. Children of spouse's Ngonga s wife calls him veyange, family my husband. 9 tut,.™* ™n~>in ~»™i„~>n„ v™, «- 12. Manu, inanu. Mother's brother brothel man^ar^e- hunger (dir^jt 3^^). 3. Kota, huva. Older brother. 13' Kulv" £n °Jd pe*?°° j,n ^"and- . „,, , _. . , ,« .„ parent s and grandchild s genera- 4. Mouale. Sister (vocative). tjon Mukai wange. Sister (non vocative). g^ yy^ Male d. 5. Nawa. In-laws of speaker s genera- parents. _ -T.10n", T , . ,. 14. Maikulu. Female grandparents. 6. Ndatembo. In-laws of ascending or , c ^ , , „ , .... , ... descending generation. 15- Onekulu. Grandchild of either sex. 7 Tate Father 16. Upalume. Mother's brother's or ft. Mai. Mother. father's sister's children- 9. Aphai. Father's sister (female 17- Ctfculume. Father's sister's hus- father). band- 10. Omolange. Child, used for speaker's Note: Age distinctions are made chiefly family only. in the speaker's own generation . Law and Government In dealing with marriage rites it was stated that a woman had to contribute certain articles to the home; these are the utensils that she uses in her daily work. The baskets, pottery, pounders, and brushes may seem insignificant, but nevertheless there are rules regulating their disposal at death. Her husband will retain some of the articles; the remainder will be shared among the deceased woman's sisters. The disposal of a widow is a matter for discussion among the relatives. The husband makes no bequest to his widow and children, though the children may receive a small gift of a pig or some corn. The property of a man is bequeathed to his mother's brother or to his sister's sons. The eldest brother of the deceased or the maternal uncle of the widow takes her to his house along with her children. If her father is alive he may take her. Ngonga says that each responsible relative says, "It is better that you should take her"; at last some one says, "I will take her." When explaining the system of inheritance Ngonga was clear on two points. In the first place wives and children of the deceased are 200 The Ovimbundu not entitled to inheritance of land, cattle, or anything else belong- ing to the dead man. The greater part of the property would be bequeathed to the eldest brother of the deceased's mother. There appears to be discussion as to the distribution of property, but the maternal uncle is responsible for settling disputes. He him- self has the first claim, and in Ngonga's words, "If there are quarrels about the cattle, land, and other things, the mother's eldest brother will settle them." The mother's brother has rights over his sister's children even to the extent of pawning them to pay his own debts. On the other hand the maternal uncle is responsible for the conduct of his sister's children. He would have to pay fines incurred by thefts they committed, and he provides for his sister's son a wife who is either his daughter or his widow. Ngonga stated that the maternal uncle does not invariably take all the property of a deceased nephew for himself. He may give something to the deceased's mother, the deceased's maternal grand- father, or to a brother of the deceased. Such gifts appear to be the outcome of good will on the part of the dead man's maternal uncle; there is no compulsion. Women never inherit cattle or rights to the use of land. The king is the head of the legal system, though his activities as such are not so great as those of the village headman (sekulu). The olosekulu (there are usually more than one) of a village witness the final act in a divorce ceremony. They used to have charge of trials for theft, murder, adultery, likewise the right of settling argu- ments concerning the ownership of land. It was the sekulu who distributed the land to the extended families when a new village site was opened. The maternal uncles settled the minor divisions among the limited families. There are many kings among the Ovimbundu, but I thought that there was a tendency to confuse the titles of osoma ("king") and sekulu ("headman" or "chief" of a village). The jurisdiction of a king is so well known that any person is able to say under which king he lives. Ngonga said definitely that a man of the Ovimbundu who was under the jurisdiction of a certain king would have to obey the commands of a visiting king, provided they were not in any way disapproved of by the king to whom first allegiance was due. A chief may demand labor for the building of his house or the cultivation of his land. He does not pay for this but usually gives beer to the workers; sometimes he secures the labor and gives nothing Social Life 201 in return. Even a slave used to have rights of appeal to the king if he were starved and beaten without cause. The king was the prime mover in warfare. The whole legal system is at this time directly under Portuguese administration though the kings and chiefs have minor powers. Ngonga says that complainants who are dis- satisfied with the decision of their own chief go to the Fort, meaning of course the Portuguese military post, or office of the Administrador. There is no doubt that in former days the responsibility for crime and debt was transferred to a relative in the absence of the delinquent. If a thief escaped, responsibility would rest with a brother, a father or a son. The mother's brother was often held to be responsible for the misdeeds and debts of his sister's children. The mother's brother has rights that extend to the sale of his sister's children to pay his debts. For this reason Ngonga says, "If I have done wrong and they cannot catch me it is right that he (mother's brother) should pay for me." I supposed the case of Ngonga having stolen a cow; he has been caught but escapes. Ngonga said that the man from whom he had stolen the animal would go to his (Ngonga's) mother's brother in order to name the price required for the cow, or any other possession which had been stolen. If the mother's brother thought the price reasonable he would pay. On the contrary, if the price claimed were too high the mother's brother would appeal, first to the village chief, then if necessary to the king. Ownership of land does not appear to have been a source of dis- pute. Apparently land was distributed by the chief as explained in dealing with village organization. Rights seem to have been well defined and there was always enough land for a stranger. When a man was going away on a trading journey he might lend his land to another who would agree to cultivate it and take the produce for a year. I was unable to satisfy myself that land is now, or was in time past held to be the property of dead ancestors. The right to land depends on its use and allotment by a chief at the time of founding a village. A man who intended to be absent would make a payment so that he might leave his goods at the house of another. The chief would settle any disputes arising from these arrangements. The complainant in a suit is called ombile, the defendant is ovilue. The name for a witness is uvangi. In the old days a false witness had to make recompense to the man about whom he had told a lie; he would probably have a beating as well. A master was in every way responsible for the actions of a slave. The degree of responsi- 202 The Ovimbundu bility of women is mentioned in describing the penalties for adultery and theft. The penalties for offences against the king were undoubtedly more severe than those imposed for the same offences against a com- moner. Ngonga says that the punishments for crimes against a chief or medicine-man were, and are now, the same as the penalties for offences against any other person. It seems, however, that if the chief were a very powerful man, he could claim penalties which would have been appropriate for offences against a king. A thief was punished by beating, in addition to which he would have to pay a fine to the people from whom he had stolen, but he rarely escaped without the beating. The thief would be beaten just as much for stealing one animal as he would for stealing many. If a woman stole from her neighbor's garden, those who caught her would tie her and take her to her husband, who would pay compensation and then beat her. A child caught stealing would be taken to the maternal uncle or to his own father. If a woman stole from relations she would probably have no punishment other than a beating. Her husband would have to compensate the relations. Stealing honey from hives is an offence that is punished according to the general laws relating to theft. If a man was sentenced to a death penalty, or to a beating, there was an official appointed by the king or by the chief to see the sentence carried out. This representative was called ukuenje welombe which means "the servant, or minister, of the court." Ngonga described the penalty for murder saying, "It was a terrible thing they did to the murderer. A tight triangle of wood was fastened on to his neck by a peg and to this a cord was attached so that he was suspended to the roof with his feet barely touching the ground. In some instances his head was placed through a hole in the door of his house while his body was inside the hut. He had very little food or water and people threw things at him. He was kept tied up for a month to see whether he could pay something. Suppose he could not pay, and nobody would pay for him, he was taken outside the village where his head was cut off. If he had many possessions, he said on the first day, 'If you will take this thing off my neck I will pay two oxen and two slaves.' The payment would be made, not to the wife or children of the murdered man, but to the brother of the mother of the murdered man, or to the mother of the murdered man, or to the son of a sister of the victim. If the murderer agreed to make the payment the people kept him tied up until the Social Life 203 fines were in their hands. If he murdered one of the royal family he was killed after he had paid the fine." The punishment of a woman who has committed a murder is the same as that for a man. Ombulungu is the name given to trial by poison ordeal. The medicine-man holds out both his hands, in each of which there is a potato, only one of which is poisoned. The accused man says, "If I have done this thing, this potato will be poison for me. If I have not done this thing, this potato will be food for me." The complainant says, "If this is not the man who poisoned my brother, this potato will be poison for me." They sit in front of the medicine-man, and each must take the potato in the hand opposite to him. The poison makes a man very sick. His mouth swells so that he cannot speak. When a woman was accused of murder her son or her brother would take her place in the poison ordeal. A boy of fifteen years of age would have the same treatment as a man. "I don't think a girl would kill anybody. I never heard about it," said Ngonga. It appears that resort was made to a medicine-man when the accused was thought to have committed murder by poison or magic. If the murderer used weapons, the accuser, who was usually a relative, was entitled to kill him with the kind of weapon used for the crime. The penalties for adultery were the same as those for murder. The woman appears to have escaped punishment, except that "her husband would not have her any more." The male adulterer had to pay the husband; if he could not pay or escape, the husband had the right to kill him. In the old days the price for adultery was two oxen, a pig, and a slave. At the present day a man who has com- mitted adultery and made payment, possibly takes the woman and all children under three years of age. The husband has the right to decide whether the adulterer has the privilege of taking the woman and her children. An adulterer with one of the king's wives was castrated but not killed. He might, instead of mutilation, pay a very high price, while he himself along with his sisters and the sons of his sisters, would become slaves of the king. If a man could not pay, he might be sold as a slave in order to provide money for the fine. Criminal law did not distinguish between responsibility for inten- tional and unintentional offences. It is very difficult to say how much of the old law survives. One feels that underneath the ostensible Portuguese rule there is an active native life that is resisting subjugation. Only two years ago Ngonga paid an ox to prevent one of his female relatives from being pawned. Within the past ten years Ngonga has 204 The Ovimbundu actually paid to redeem his brother and sister who were sold to pay the debts of his maternal uncle. Officialdom is one thing and actual practice is another. Ngonga is right when he says that people appeal to the native law as laid down by the chief or king. If they are not satisfied they pretend a great respect for Portuguese law and therefore go to the Administrador. Warfare and Slavery The Ovimbundu have been for centuries an organized people possessed of a well-developed language, a legal system, and complex social organization. In trade and travel too, the Ovimbundu have been in the vanguard of African tribes. These reasons, in conjunction with the fact of numerical superiority and early contact with the Portuguese, would account for the success of the Ovimbundu in warfare. The defensive federations existing between the ten or twelve main political units, whereby they did not habitually war on each other, helped to assure success. Should the eldest son of a king's principal wife be thought unsuitable as successor, a competent son of the deceased king was elected by a council of village chiefs. There is no evidence that the Ovimbundu poisoned their weapons, neither do they do so today. The Ovimbundu do not use a shield at the present time, neither is it certain that they had shields in former times. A skin-covered powder-box was carried in the king's train when on the warpath. Signaling-drums were about two feet long. They were made from a cylinder of wood over the ends of which skin was stretched. Signaling, which was accomplished by drumming with the palms of the hands, was the task of boys who resided in the king's com- pound, except when they accompanied him to war or on a trading journey. There was no system of smoke signaling. The iron war gong was named ongonge, but the instrument is nowadays tapped by the assistant of a medicine-man during his performance. In former times the gong was struck in the night to give assurance that the enemy was not in sight. In war the oldest son of a king or of a chief went ahead of the war party accompanied by a few men who wore red leaves on their foreheads. Intertribal jealousies, raiding for cattle and slaves, also reprisals for interference with caravan trade, were the chief causes of conflict with surrounding peoples. There was undoubtedly some internal warfare within the confederacy. The Ovimbundu of Elende, Bail- Social Life 205 undu, and Ngalangi are said to have had conflicts. In time past, as at the present day, a king reigned over territory which was extensive but definitely delimited for purposes of administration. Encroachment of one king on the rights of taxation and administra- tion of another led to raids and reprisals. Village chiefs collected taxes in the form of agricultural produce and gave these to the king, who personally visited a village from which payments had not been made. A folklore story begins, "The people had not paid taxes so the king came to the village and told them a parable." A king, if young, accompanied his people on the warpath. There was, however, a permanent leader named kesongo, a derivative from songola ("to lead"). The declaration of war, likewise the tactics, were discussed by a council of olosekulu ("village chiefs") in the ombala ("capital") where the king had, and still has, a royal com- pound. If war had been conducted among sections of the Ovim- bundu, the defeated people had to pay taxes and tribute; moreover, their women and cattle were taken. In event of a successful war against the Vachokue there was plundering of cattle and women, but it was not found practicable to exact periodical payments from the enemy. The subject of warfare is intimately related to that of slavery. There was until very recent times a domestic slavery which followed from the inability of a person to pay his debts. In connection with this reduction of free persons to a condition of slavery there are several points of exceptional interest. The debtor himself is not taken as a slave, neither are his wife or children. The correct proce- dure is a sale of his sister's children; but more frequently the children themselves are taken by the creditor. "The debtor's sister will say nothing because this is the law of the Ovimbundu." If payment of the debt is made later, the children are set free. Usually the word pawning is used by ethnologists to describe this proceeding. Further consideration of Ngonga's payment to his mother's brother clarifies the facts of domestic slavery. "I paid for them, I took their place," said my interpreter. The payment for the return of these two children was two oxen. The girl, who was ten years of age when she was taken to pay the debt, was returned to her people when she was a woman with three children. Ngonga's brother was not actually taken from his home. He was made to pay his mother's brother's debt by working for the creditor. In general, these domestic slaves were not ill treated, though their rights were limited. The position of slaves taken from a hostile 206 The Ovimbundu people, especially if they were from outside the Ovimbundu, was one of absolute and abject servitude. The master had rights of punishment including a death sentence; but Ngonga says that a slave could appeal to the headman of the village if he were starved and cruelly treated. All children of a slave became slaves of the master. As the slave had no property he could not pay a fine; the master would pay the fine then relieve his feelings by beating the slave. Instead of paying the fine the master might sell the slave. Slaves were not branded or marked in any way. Ngonga has seen runaway slaves hunted with dogs. A slave was not allowed to buy his own freedom. A master could dispose of his slave girls in marriage ; for instance, a young girl might be sold to an old man. The condition of the slave is well expressed in Ngonga's own words: "The slave worked hard at everything, then the master said he had done nothing." Slaves used to go to war to fight with their masters against an enemy. A slave might become a blacksmith or a hunter, two very esteemed occupations, but all his work would be for the master. Ngonga says there was no slave market belonging to the Ovimbundu, but every man knew where he could buy a slave. Slaves could act as witnesses in a trial. Slave women were not lent out for prostitu- tion. A slave owner did not have promiscuous intercourse with his slave women, but he chose two or three girls as concubines. A slave girl who was married to a free man would revert to her master along with her children when her husband died. If a master had married his own slave woman she would, at his death, become the property of his eldest brother and her children would go with her. In some instances the slave woman and her children would be given to the son of her master's brother. Village Organization When choosing a site for a new village, a preference is shown for a hillside, though woods or valleys are at times selected. The foot of cliffs is a favorite site. In addition to the shade afforded by the cliff there was in time past the advantage of being hidden from the view of enemies. Further shelter was afforded by the planting of wild fig trees. Sometimes an ombala ("capital") was rendered picturesque by the planting of trees which grow to a great size; such a plan was followed at the ombala of Ngalangi. Caves in the wall of a cliff, likewise rugged hillsides, gave a place of retreat for women and children during an attack. Usually there Social Life 207 are small streams of pure water falling down the cliffs and hillsides. Near Bailundu and Ngalangi, villages still retain their defences which consist of high poles set in a trench. Judging from sites visited at Elende there was a pre-Ovimbundu stone-building culture. One of the sites has such a commanding view over extensive plains and valleys that the position would be almost impregnable. At the present day the line of fortifications is well marked by stone walls three feet high. These are composed of boulders to which the builders had ready access on adjacent hillsides. Large stones were no doubt rolled from the slopes to the small plateau chosen as a building site. This small plain lies midway between the hill crest and the valley. At present this old site, which is enclosed by lichen-covered walls, is overgrown by tall grass and trees attaining twenty feet in height. Photography would be uninstructive unless a preliminary clearance were made — a formidable task owing to the density of vegetation. In the center of the enclosure is a group of transported boulders possibly marking the site of a place of assembly. A search among the long grass reveals stone slabs and cylindrical crushers which were used for grinding grain over a very long period, as may be seen by the wear on the base stone; some thick stones are so worn as to be almost per- forated. Weather-worn stones that were probably used as scrapers, are to be found. Surface potsherds are of the material of which present-day Ovimbundu women make their cooking pots. These sherds mark the places now used by small nomadic bands, hence the surface pottery may have no connection with a pre-Umbundu culture. In the vicinity of this walled stone village are hillside cairns marking the sites of graves. These have been robbed by medicine- men in search of material for their charms. The Ovimbundu have no traditions regarding the site, neither is there legendary or other evidence to show that the Ovimbundu ever made their villages of stone. The raising of a cairn of stones over the body of a hunter is, however, a present-day practice near Ganda and in the Esele country. If a site cannot be systematically worked, it is better left alone until scientific investigation is possible. The preliminary clearance of grass and timber would be a long process. The archaeological material recovered from such a site would be of a uniform and simple kind. There is no evidence of anything beyond an elementary stone age culture. 208 The Ovimbundu At the present time one type of village is surrounded by stout stakes ten or more feet high set in a roughly circular plan. The arrangement of alleyways within the village is complicated without following any symmetrical plan. A visitor sees only a labyrinth of passages between high poles, with here and there a rectangular wooden gate at intervals along the passages. The door itself is made from four stout heavy planks which are roughly hewn and as a rule undec- orated, but occasionally I have seen the panels of a door carved with designs representing the breasts of a woman; or there may have been simple geometrical patterns. The top ends of the door panels are massive spheres of wood bored through with holes. Through the holes in the tops of these panels a pintle is passed in such a way that the door is suspended from above. During the daytime the lower ends of the panels are raised to the level of the lintel. They are there supported on a Y-shaped upright which is erected in front of the doorway. At night the planks are removed from this support and allowed to hang downward in the doorway. There is a sliding cross- piece to push over the lower ends. The same arrangement is some- times used to make a door for a hut. This old type of village doorway, common twenty years ago, is becoming increasingly rare. The house of a king or a chief is larger than that of a commoner. The enclosure which contains a king's house and the other buildings is named elombe, while epandavailo is the word used to describe the entrance to this compound. The elombe is built by men and women who receive no payment. There is a ceremony when the compound is opened. The house of the chief, Kandimba Sanjahulu, near Bailundu, is rectangular, with mud walls raised on a platform of earth and stones. Whitewash has been applied to the outside of the walls and some ornament is given in the form of painted blue crosses. Complete study of structural types and the planning of internal divisions has been made by F. and W. Jaspert, of the Stadtisches Volkermuseum, Frankfort. At a village near Cuma the house of the chief differed from the dwellings of commoners in being somewhat larger. The house had been abandoned, not because the chief died there, but because the chieftainship had been transferred to an adjacent village. The tomb, which will be described in connection with funeral rites, was a few yards from the house. This mausoleum was surrounded by a high wooden palisade, to a stake of which were attached the horns of an ox killed at the funeral feast, while the jawbone lay in the enclosure Social Life 209 (Plate XLV, Fig. 2). The house of bows for holding sacred relics will be described in connection with religion (Plate XLVI, Fig. 1). In former days there used to be a hard mud floor for dancing, centrally placed in the village. In the center of the village is the communal house where all men, and boys over four years of age, gather to eat their food, which is sent by the women. This onjango is the place of assembly for dis- cussion of village affairs. A house of this kind at Bailundu is circular in form with a diameter of seventeen feet (Plate XLVIII, Fig. 1). In the middle of the onjango are stones forming a fireplace. In another such house there were jawbones of oxen. These were suspended from the walls as tokens of the payment of fines. There is no communal house for females. In some villages there is a common kraal for cattle. The king's compound usually contains: (1) The king's house, which invariably has a separate sleeping room. (2) A house and kitchen for each of his wives. (3) A tomb for the burial of kings and their wives. (4) The house of bows for staffs, bows, mats, and tobacco-pipes of dead chiefs. (5) A house of meditation for the king (Plate LXXXIX, Fig. 1). (6) Pens for pigs and chickens. The guest houses (Plate XLVIII, Fig. 2) that I have seen were not in the king's compound. Granaries, which are conspicuous in every village, have been mentioned in connection with food supply (Plate XLIV, Fig. 1). Each house has its own granary. I have noticed garbage pits containing corn husks and refuse from sweet potatoes, but there is as a rule no organized scheme for disposing of refuse ; dogs and pigs are the scavengers. There is no particular place for defecation; people use the bush. Every man has a knowledge of house-building. Women do not actually build, but, along with the children, they pour water into the clay pit and carry clay to the men who are plastering the walls. Ngonga says, "People would be surprised to see a woman building a house. They would call her a he-woman." Men help one another in house-building by giving reciprocal service for which there is a gift of food and beer. A deep rectangular clay pit is made as close to the house as possi- ble, and the children thoroughly enjoy puddling the clay with their 210 The Ovimbundu feet. A trench about eighteen inches deep is dug for the poles (akoso), which form the framework. The wattle work is tied to the uprights with strong strips of red bark before the plastering is begun. The old type of Umbundu house was round, but most of the houses now show a transition to square or rectangular forms (Plate XLIV, Fig. 2). Exterior wall-painting is found only in the northern districts of Angola (Plate LXXXIX, Fig. 2). Between Saurimo and Malange in the north, rectangular, painted houses are frequently seen. Wall- painting is a cultural trait from the Congo region where that form of decoration is common. The method of forming a village site and the right of the chief to allot land have been described. The house of the chief is the first to be erected. There is drinking of beer to celebrate the completion of the village, and the pots of beer have to be stirred with the claws of chickens; these have been killed to provide blood for sprinkling the walls of the new houses. Village chiefs go through the process of stirring and drinking. The medicine-man drinks first, then he offers the cup to the chief. The communal house is built by the united efforts of all men of the village. Ngonga says, "The king never helps with any of the building, but he talks very much." The guest house is also built by communal labor. The king of the ombala of Ngalangi told me that a king always uses the house of his predecessor, which must not be pulled down. No repairs are permitted and the house is used until it is absolutely untenable. The interior of a commoner's house has three hearth stones in the middle of the floor, which is of hard mud beaten down with a heavy wooden implement called ocikandulo. The ceremonial placing of the stones has been described in discussing marriage. There is no chim- ney, consequently the walls and roof are blackened with smoke. The junction of roof and walls provides pockets for the reception of small articles. On the floor near the walls may be several large beer pots. The dark smoky interior is almost intolerable. The sleeping room is sometimes separated from the living room by matting. The bed is made from a cross-work of sticks supported on four Y-shaped posts, one at each corner. On the bed are coarse sleeping mats and possibly a modern blanket. The miscellaneous contents of the hut are gourds, dippers, cooking pots of several sizes, pounders, and baskets. In many of the huts I found cooking going on in the general room, but frequently a sepa- Social Life 211 rate hut is provided as a kitchen. In some dark corner of the hut there is likely to be a sitting hen or a bitch with pups. Goats and pigs intrude from time to time. Lean dogs scavenge round the doorway. Naked little children whitened with dust play on the floor. A woman crouches over the cooking pot stirring the glutinous mass of corn meal. About sunset men are gathering in the council house to which their meals are sent. They push forward the logs into the fire, throw their blankets around them and chatter until the women and children arrive with the evening meal. In the vicinity of the village, boys are wandering about with blunt wooden arrows fitted to their bowstrings, to be ready for the homeward flight of birds. Other boys are driving cattle to the kraal. As darkness falls the long tubular drums are brought out. These are held between the legs and played in compound rhythm (Plate XXVII, Fig. 1). A shuffling dance begins, slowly at first, then with increasing vigor, to be continued far into the night. VII. EDUCATION Industrial Training and Division of Labor Usually there is no formality connected with the industrial and occupational training of Ovimbundu children. The section dealing with play shows that boys and girls learn by spontaneous imitation of their elders. There are, however, exceptions to this general truth. Special rites connected with the training and inauguration of blacksmiths and hunters have been described. The training of medicine-men and women will be dealt with in chapter IX, "Religion." In describing economic life in chapter V, division of labor on a sex basis; also specialization in industry because of peculiar tastes and aptitudes, were noted. There is clearly a twofold division in industrial life. In the first place sex determines occupation, and within this primary grouping there is a secondary grouping depending on personal preference and individual ability. I have found no instance of exclusive hereditary right to any occupation or industry. The net result of the Umbundu system of division of labor is indicated by the following summary: When building houses, men dig the rectangular trench, cut the poles for the framework, lash the crosspieces, cut grass and lay the thatch, then place clay over the wattle walls. Women carry water for mixing the clay and convey the moist clay to the plasterers. Children puddle the clay with their feet. Males are hunters both professional and general. During a communal hunt women and children assist in driving game, which is frightened by the firing of grass. Boys organize games in which they imitate these activities. Men fish with a line and bait, whereas women use baskets and narcotic poison. Men and boys take care of cattle and have charge of dogs. Women and girls give what slight attention is afforded to other domestic animals. Male specialists castrate bulls and goats. During agricultural operations men clear the ground and burn the bush. Women carry on hoe cultivation without assistance from men. Men are the chief carriers in caravans, but women and children may sometimes be seen carrying loads. They march in line with the men but have smaller burdens. Males are employed in blacksmith's work, wood-carving, making weapons and tools, weaving mats, dressing hides, spinning cotton, and formerly in weaving on an upright loom. Men are exclusively the makers of 212 Education 213 musical instruments, and males are the musicians. Men are the only persons engaged in warfare and administration, and the onjango or council house is used exclusively by males. Men follow the occupation of medicine-man, in which there is considerable speciali- zation. Female practitioners deal with pregnancy and women's ailments. In addition to the tasks for women mentioned above, the following are staple occupations: collecting firewood, drawing water, caring for infants, making pottery, weaving baskets, dancing, and singing. Young girls share these activities with older women. The foregoing categories explain division on the grounds of sex. Degrees of specialization are not so easy to formulate, but in general a man follows some one occupation, for example wood-carving. Then within this occupation there is specialization in the making of drums, domestic utensils, or figurines. Almost any woman could make pottery or baskets, but the difference in skill leads naturally to concentration in the hands of expert potters and basket-weavers respectively. These sell their wares to those who either do not make such articles or are inept at the process. Standards of Conduct, Manners, and Salutations The parents of a child, also his maternal uncle, assume respon- sibility for training in the precepts and standards given here. The maternal uncle, who is the mother's oldest brother, is particularly interested because he has to pay fines should his sister's children commit thefts. The evidence indicates that the home and restricted family have formative influences over the conduct of children. Children are beaten if they tell lies, answer old people rudely, or steal food. Ngonga says that his "stealing hand" was once placed in the hot leaves of the cooking pot. If a child steals an egg which is cooking, the hot egg is held between the culprit's hands. One cannot fail to notice the quiet and unobtrusive way in which children sit in the presence of their elders both in the home and in the council house. Children do not speak when their elders are in conversation, unless addressed. A child, likewise an adult, receives a gift with both hands. The implied idea is that reception with one hand is a depreciation of the gift. If a child holds out one hand, the hand is slapped. When receiving, an Ocimbundu says "kuku," literally grandfather or elder. Colloquially the word is used to mean greeting, "I thank you," or "I beg your pardon." 214 The Ovimbundu Lying is strongly disapproved and the liar is called ohembi. The Ovimbundu appreciate hospitality (unu) which also means generosity. A man who is hospitable is said to be ongavi. Greediness, which is disliked, is described by the word oku sapa ("to be greedy"). There are standards of honesty in sales and exchanges. A deceiver in trade is ohembi. With regard to the relation of the sexes in early years it may be said that, in spite of boy and girl friendships, and the communal sleeping of boys and girls at the home of one of the girls, a man expects to marry a virgin, a point which has been dealt with in describing marriage ceremonies. Naturally there is a difficulty in obtaining precise information respecting the details of these nocturnal gatherings of boys and girls. Ngonga said that girls inform their parents concerning the house where the night is to be spent, and there is a point of etiquette requiring that girls must not go to a house where boys are staying together. Standards of conduct already described under courtship are a result of direct teaching by parents. Marriage rules and a classifica- tory system of relationships, with its prohibitive decrees, are taught in the home, in the men's council house, and at initiation ceremonies where such exist. In addition to these sources of instruction there is no doubt an unconscious absorption of ideas and standards. The power of suggestion is always at work through everyday examples. Apart from demands made by tribal custom and direct instruction there are variable personal standards of modesty. Ngonga states that many lascivious stories are told among men, and when the men are drunk, they tell these to women. "Sometimes the women laugh, but the good women do not like to hear these tales." A male commoner when meeting the king bows low, extends his arms, claps his palms and says, "ohosi ["lion"] akuku ["grandfather"]." Only the old people follow the ancient custom of falling on their knees when greeting the king. The Ovimbundu never were in the habit of doing more profound obeisance, but the Vangangella, when greeting one of their kings, rub their chins on the ground and place dust on their chests. Even at the present day a woman or child of the Ovimbundu is expected to kneel when greeting a king, but such an acknowledgment is not now usual in greetings given by a male commoner. The reply of the king to the commoner is "kalunga." This word enters into greetings of all kinds. The literal meaning of the word is "the sea," "king," "god," or "death," but the interpretation Education 215 of the word depends upon the syllable accented and the context, as further explained in the chapter on the Umbundu language. A man or woman of the royal family greets the king with the words na kuku; na means "lord," and kuku is a term used for any old man to whom the speaker intends to show respect. Kuku is also applied to a man who stands in the relationship of grandfather. The king is expected to reply to a greeting given by one of the royal family by placing his right hand on his chest and saying twice, "kalunga." There may then follow from the king a question relating to welfare in general. Possibly the king will inquire the object of his subject's journey. A sekulu ("chief of a village") greets a king in the same manner as does a commoner, and the king replies as he would to a commoner. Commoners greet a sekulu with the words na kalunga ("lord, greeting"). One sekulu greets another of the same rank with the one word kalunga, accompanied by clapping of the palms. Male com- moners clap hands on meeting; this action is accompanied by the word kalunga, from each of them. Two female commoners use the word kalunga as a greeting, but as a rule they do not clap hands. They do, however, clap their palms when greeting a woman of the royal house. Boys and girls must greet their fathers, maternal uncles, grand- fathers, and other old men with either of the terms na kuku or na kalunga, the latter being more usual. Na kuku would be the appro- priate term of respect for any elderly man. Children use the same words (na kuku) when addressing any elderly woman, including a grandmother. A usual greeting of a child to the mother, likewise to the mother's sister, would be kalunga mai ("greetings, mother"). There is no prescribed form of address to the medicine-man. Greetings between a commoner and a medicine-man follow the usage noted for two commoners. Ngonga says, "They will treat him (the medicine-man) like a village chief if he is an old man and good. Perhaps they will say na kalunga as if talking to a chief." In the early morning people pass the greeting oku lipasula. This means "we have been like dead, we are awake." The appropriate early morning greeting of a commoner to the king or to a chief is oku lipasula a kuku. A usual afternoon greeting is oku lanisa. The evening greeting is oku lisuninya. One who desires to pass across a room usually walks in front of other people, but while passing the attitude must not be erect. The thumb and middle finger should be snapped together, while the words 216 The Ovimbundu konyimo oko are spoken. The literal interpretation of these words is "back there." Ngonga says the words actually mean "excuse my back." Spitting in the vicinity of the house of the king or of the chief is not allowed. "If you did that in the old days, you would have to pay something." Some people spit near the hearth in their own houses. A man who spits in the road is expected to cover the spittle. No person would spit in the presence of the chief or of the king. A hand should be placed in front of the mouth by a person who is coughing or sneezing. Shortly after the birth of a boy or girl all who are on good terms with the parents greet them by saying kalunga. The word is repeated four times with clapping of the palms. When words have been imperfectly heard and the listener desires repetition it is customary to say kuku in an interrogative tone. No particular etiquette is observed when eating. Two or three children eat from one platter, helping themselves with their fingers to the mush or sweet potatoes. Rules forbidding the preparation of food by menstruating women, likewise prohibitions relating to stepping over a person, are mentioned in discussing taboos and omens. Educational Value of Play, music, and dancing The educational value of play lies in its imitation of the pursuits of adult life. Though there are specialists in music, dancing, and singing, these exercises are indulged in by everyone. Music, singing, and dancing are more than amusements, they are fundamental as coordinating forces in tribal life at all levels of culture (W. D. Ham- bly, Tribal Dancing and Social Development, London, 1926). Among the Ovimbundu Cimbamba Co Lia is a popular game for girls, who form a circle, join hands, and sing: Part I Omola una, ndo sile vekango, Cimbamba co lia ("That little child was left in the desert, the nighthawk ate it") . At the word Cimbamba they begin to dance, facing from side to side in such a manner that they meet and bow. Some sing, Cimbamba co lia ("The nighthawk ate him"), and others respond, Kalikisi ("'tis the goblins"), as many times as they wish. Finally a return is made to the first words which are sung again. All musical transcriptions have been made by Dr. G. Herzog from my phonographic records. Education 217 1 Ul 1 IZ3 tHrcPr'L1 m a.-s i n r r p Eventually the leopard gets the last child and deposits it on the ground where it immediately begins to imitate the pounding of meal and to sing, Fule, fule, fule, fule, kolohanda ko Luwa ("Pound, pound, pound, pound on the rocks of Luwa"). This ditty, which is sung by women during their daily occupation of pounding maize on the 218 The Ovimbundu rocks, is repeated until all the children have been captured. Lastly, the mother is caught and taken by the leopard to the bushes, where he hides her. -3 sat u u. * ** ** € m 4 1 \tWT% " The leopard goes back and picks up a handful of sand which is an imitation of meal. Each of the children pretends to pick up a handful of sand before going in search of the mother. The leopard J -*»1 c 6Vq. „ -s- . ^ a, . $ ^=ff=fT * £g^ F* ingl fcs fa ^=V -3 - S 5 p«« 53E & ci, T~ Z 1 I. - - J ,1 W I J 30E ruf ft r i »w S 5 g rf3^ ** r r ■ r r K-.5 n3 «4 <•*(*, Cli c.J*^ ^ — * ^S 51 -0 d=J36 CI, f f ■ r f ^v «x 4 4;tr. V I \\ \ f^^ Education 219 says, "Here's your mother." The children look and sing, "She's not there." They throw sand in that direction. The leopard repeatedly leads to places where the mother is not hidden, while each time the same words are repeated and sand is thrown. Finally the leopard leads the children to the place where the mother is hidden. Then the mourning song is changed to a glad dance. The children clap hands and sing, Mai Cisangu weya. Mai Cisangu weya ("Mother Cisangu has come"). After transcribing the music of this song Dr. Herzog reported, "The melodies are rather simple, moving within a restricted range, with a plain rhythm, the same short unassuming melodic fragment being repeated as long as the game may require it, or changed slightly to suit the words. It should be kept in mind that these are children's songs; other songs of the Ovimbundu are probably much more elaborate. The manner of singing songs by a solo and a responding choir is highly characteristic of African singing. Thirds as seen in the transcriptions on page 218 are often used in the music of west Africa. "In the musical notations, S stands for Solo, Ch for the Choir. A as a sign above a note indicates that the tone is sung approximately a quartertone higher than noted. It indicates a short transitional tone of slight rhythmic or melodic significance and of uncertain pitch." There is but one example of children's dolls in Field Museum's collection. This was obtained from a little Ocimbundu girl of Elende. The doll is made from a corncob which is draped in blue cloth of European origin such as women wear. My interpreter said that the dressing of dolls made from corncobs is a general custom at the time of cutting the corn. In view of the frequent occurrence of agricultural rites among Bantu Negroes, the use of dolls led me to inquire into the possibility of there being some kind of fertility cult associated with a corn-mother concept, but my inquiries met with negative results. There is an Umbundu word for game (omapalo; plural, olomapalo) which is used for games in general. The verb oku papala means "to play." As a rule, male adults do not play games, but they have the game of mancala which is called ocela. This is difficult because it involves quick counting. Mancala has a wide distribution as shown by the fact that it occurs in Africa, Syria, the Malay Archipel- ago, South America, and the West Indies. At Dom Manuel in the southeast of Angola I saw an arrangement of holes in the ground which was used for this game. There were four rows of fourteen holes in each row. The counters were nuts from an oil palm. 220 The Ovimbundu In the compound of the king at Ngalangi there was an ocela board consisting of an oval piece of wood on a short base. This board had holes arranged in four rows of seven, twenty-eight holes in all. The king refused to sell the board. He said that it was highly valued and that the gambling stakes were high. Up to the age of sixteen years boys play the game of ocitina. Each side has a number of bulbs from a figwort, which are rolled along between two lines of boys who shoot at them with arrows. The winners are those who have shot the greater number of their opponents' bulbs. The party A rolls bulbs while the opponents B do the shooting. Then the positions are reversed. A game with whipping tops is ongilili which is said to be of Umbundu origin, not a derivative from a Portuguese game. There is no evidence that games are seasonal; any game may be played at any time. Games of mimicry naturally form a diversion, and boys imitate the occupations of their elders. The chief activities of adult males are warfare, hunting, and carrying, all of which used to play a very important part in the communal life up to a few years ago. When boys play at making war they have two sides, the attack and the defense. The attacking party runs about the village taking prisoners, who are tied with bark rope. Girls sometimes play this and other games with the boys, but usually boys and girls play separately. Sometimes strong boys are selected for hunters whose dogs are the little boys. Boys who pretend to be the antelopes or other game go to the tall grass to hide. Toy bows and arrows are made, but sometimes the hunters go through only the movements of shooting. The boys who are pretending to be the game roll over and gasp when shot. The "dead game" has to cling to the pole on which it is borne to the village on the shoulders of the hunters. The little boys go along on all fours barking like dogs. The Ovimbundu were, and still are, renowned carriers whose prowess is imitated in boys' games. Boys make up loads in the correct way; that is, lashed in the fork of two long sticks which can be rested on the ground. These they carry along, singing as they go. There is some wrestling, also stone throwing to test distance and accuracy. Rubber is made into a ball which is bounced rapidly and repeatedly with the open palm. Men and boys sometimes form a ring around which the ball is thrown from one person to another. The Ovimbundu have no game played with string wound around the fingers, and I have been unable to find any evidence for the past or present formation of string figures. I have seen boys making bird Education 221 cages and wooden animals to use as playthings. Girls mold small animals from clay. The bull-roarer is in use as a plaything at Elende. There is a game of hide and seek for boys. A knife is hidden; then a boy who has been absent for a time enters the ring marked off for the game. His proximity to the knife is indicated by playing on a musical bow. There are taps on the instrument which mean that the knife is far away. On the contrary, when the searcher approaches the hidden knife, the boy with the bow plays a distinc- tive note, yelula, yelula, meaning "pick it up, pick it up." Boys readily make a simple apparatus for playing a game of lassoing a hoop. A pliable branch is bent so as to form a circle, then the ends are lashed together with bark strips. The lasso is a piece of rattan or bark having at each end a corncob or a small stick three inches long. One boy bowls the hoop so that it passes in front of, and a few feet away from his opponent, who stands twenty-five feet distant. As the hoop passes in front of him, the boy throws his lasso in such a way that it twines round the hoop and brings it to the ground. There are dances of many kinds, but Ngonga says that he would not know from the steps only what particular purpose the dance served. Several dances have been described, each in its appropriate section. There are no dances specially arranged to celebrate weddings or births. The funeral dance is described along with other ceremonies relating to interment. The medicine-man dances in connection with making rain or curing the sick. In order to say why the dance was being performed, it would be necessary to listen to the words of the songs; these are usually chanted in accompaniment to shuffling movements and the rhythm of drums. Some of the older men and women perform dances and sing songs that are unknown to the younger generation. Dancing is in favor during the months of May and June because there are supplies of maize for making beer. As the dry season advances the maize formerly available for making beer is consumed as food, hence dancing is not so usual. Ngonga says that the old people know a dance which should be performed at new moon "so that there will be no sickness during this moon." Older men dance in commemoration of events during past wars, while women are spectators, and on these occasions there are beer-drinking and the slaughter of an ox. A group of men keeps up a shuffling dance while an old man relates a war story in a singsong voice. The oldest man is the first to cut the meat, after which each man helps himself. 222 The Ovimbundu Almost every evening one may hear the tapping of drums and the songs of dancers, because in addition to dances for a funeral or other special occasion there are ordinary dances of amusement in which lines of men and women advance and retreat, or men and women circle round the drums making arm movements accom- panied by a swaying rhythm. The dance onyaco, performed in June when the corn is ripe, may have an ancient history and special significance as part of an agri- cultural rite. When corn is being stored the people sing, "There is grain in the house, may it never be out." There is no dance in connection with fishing. There is mimicry of animals in the hunting game played by boys. The frog, the leopard, and other animals are imitated in certain games, but I have no evidence of the performance of mimetic dances in relation to any cult for increasing the supply of animal life. The use of the small ball ombunje illustrates the way in which an apparent toy can be used in rites of a religious kind. Ombunje consists of a hard spherical fruit about six centimeters in diameter, in which several hard seeds rattle. The sphere is covered with a layer of cloth over which lizard skin is stretched and sewn. When the people wish to commemorate the death of a king, or when the king is sick, the medicine-man (ocimbanda) says that there must be osaka dancing. A strong man dances for many hours while holding this little ball in his outstretched hand. Other men who are dancing use their fists to hit the muscles of the outstretched arm in an attempt to make the holder drop the ball (ombunje). If he does so another man will promptly take hold of it. The precise nature of the endurance test is unexplained, but there is possibly the idea of giving strength to a sick king by this tension and endurance. My interpreter thought this was so, but could not explain why the dance should be performed to commemorate the death of a king. It would seem natural, however, to transfer the ombunje rite to a commemora- tive festival, if in the first place it was part of the last rites of a dying king. There are among the Ovimbundu specialists in dancing, singing, and the playing of musical instruments. Onjimbi is the word for a singer of merit who starts the choruses. Ocili is a dancer of more than ordinary skill. When a man is required to play a drum or other instrument I have noticed that it is thought necessary to bring a specialist. There is no doubt that drumming requires special aptitude and practice. The man who plays the long drum is usiki, the drum Education 223 itself is onoma. The flat wooden drum without membrane is ocingufu, the player of this drum is usiki wocingufu (Plate XXVII, Fig. 2). The long tubular drums vary in length from 50 to 103 cm. At a funeral, four of these tubular drums, which are usually held between the legs or placed upright, are played in compound rhythm. A performer always warms the skin of the drum at a fire, and sometimes the pitch is altered by sticking a lump of wax on the membrane of a long drum, or on the sides of a wooden drum. Dr. Herzog reports as follows, after transcribing several cylinders of phonographic records of rhythms played on the long drum, and after examining motion pictures which were synchronized with the drumming (Plates XXV, XXVI). "These motion pictures indicate the position of the performer's hands, and suggest that the sharply accented tones are produced by impacts from the phalanges. The higher notes so produced have been marked by notes above the line. A mark placed between notes indicates a very short rest. "No. 1 (dictaphone record 3) represents the drumming of a young player, who is apparently still an apprentice, since his rhythm is very simple. The three forms of his playing given below were used interchangeably, and he shifted from one to the other without stopping. In the second form, the order of the sharp and of the plain beat is reversed. Otherwise, the sharp beat occurs, in the playing of all drummers, on the off beat. The rhythm of No. 2 (dictaphone record 15) is a little more varied because the player was changed. The drummer of No. 3 (dictaphone record 30) was the best performer in the neighborhood, according to Mr. Hambly; and his playing is the most interesting. The drum records consist of a small number of rhythms which are freely interchanged; the two predominating rhythms are given below. The moving pictures were made from this player's performance. "The notations found here do not convey to the reader the bewildering complexity of African rhythm, for this appears only when a performance includes the use of several drums and musical instruments in conjunction with dancing." From Ngalangi a large friction drum 120 cm long and 47 cm in diameter was obtained. This instrument, which was the property of the village, was played only on public occasions. Apparently the drum had been hollowed from a log of wood, one end of which was left open while the other was covered with hide. To the inner side of this hide a long cane was attached. On the side of the drum 224 The Ovimbundu S»a-8 i J*IW» lr' f- " rr ' ' :" ": pp -^ p pit :0 V, a j > a, r p r • p and in the middle was an oval aperture. The performer laid the drum on the ground and took his seat astride it. He dipped his right hand in a gourd of water, then rubbed his wet palm up and down the cane rod, which he could reach through the oval aperture. The sound of the rubbing on the rod was communicated to the membrane. A short friction drum made at Elende measures 21 by 42 cm. One end of the drum is covered with hide kept taut by pegs of wood while the other end is open for the insertion of the player's hand. A cane rod is attached to the inside of the membrane. The assistant of a medicine-man holds this instrument under his arm and plays during divination. The medicine-man sits on a stool shaking his divination basket (Plate XXIII, Fig. 2). Rattles are of three kinds, and of these the more common are a small, long-necked gourd containing hard seeds, and a compound Education 225 rattle made by fixing seed pods on a stick. Rattles of the latter type are sometimes tied to the ankles during dancing and they usually form a part of the medicine-man's outfit (Plate XXII, Fig. 4). From Cangamba a basket rattle of dumb-bell form was obtained; the basket-work receptacles at each end of the connecting rod contain hard seeds (Plate XXII, Fig. 1). At Bailundu three Ovimbundu flute players met a party of visitors whom they accompanied around the native village under the leadership of a chief. The men played the flutes continuously except when the chief was speaking. The wooden flutes varied in length from 20 to 50 cm. The instruments were end-played, and the smallest of them had seven stop-holes. The visitors were con- ducted from the village by this small orchestra which is a permanent guard of honor for the chief (Plate XL, Fig. 1). At Cangamba the Vachokue have an instrument consisting of eight slats of wood which vary in length and thickness. These sounding boards are attached to two parallel cords which are kept tight by two assistants. The performer taps the wooden slats with two sticks terminating in balls of wax. This is the well-known marimba, but the gourds which are usually fastened under the slats of wood were absent in this instance. The instrument ocisanji is played well by only a few men. It consists of a wooden board, often well carved, on which there are from eight to nineteen thin keys of metal. These can be pushed backward and forward in their bridge so as to alter the vibrating length and pitch of the note. The pitch may also be varied by placing small balls of wax on the under side of the keys. When the instrument is held between the palms the player's thumbs are in position for stroking the keys. Sometimes ocisanji is played inside a wide gourd (Plate XXII, Fig. 5). The musical bow is called ombumbumba. The bridge, which is a small stone, or a piece of wood placed at one end of the bow, keeps the string taut. The gourd is tightly fastened to the bowstring with a loop of string. One end of the bow is sometimes placed in the player's mouth. The left hand holds the bow and presses the gourd to the body, then releases it a little from time to time in order to assist tone and resonance. The player holds in his right hand a reed which is tapped lightly on the bowstring, while the thumb and forefinger of the left hand are used occasionally in pressing on the string to alter its vibrating length (Plate XXII, Fig. 6). 226 The Ovimbundu Another musical bow named ekolowa is of simple construction, consisting of a strip of cane from 54 to 70 cm long and 2 cm wide. At each end of the cane is a projection. A string, when tightly stretched between these projections, pulls the cane into the form of a bow. The performer, placing one end of the bow in his mouth, plucks the string with his thumb and forefinger (Plate XXII, Fig. 7). Two rubbing instruments are used by the Ovimbundu. Ogolanda is a large gourd with a slit in the top along which there is a board cut into sixteen notches. The rubbing of a short stick along these notches produces a sound which is greatly amplified by the gourd resonator. The second instrument of this type is in the form of a wooden bow having its thickest part notched for rubbing with a stick (Plate XXII, Figs. 3, 8). Only specialists are skilled in composing songs, and both men and women are composers. The younger people sometimes go to the old people to learn songs which were popular a generation ago. There are no professional itinerant story-tellers. There is a chant for funerals which has been quoted in the appropriate section, but no special wedding songs are used. The Ovimbundu have satirical songs humorously describing individual foibles and peculiarities, and as usual in Negro communities a satirical song is feared by thieves, adulterers, or other offenders. In former days when men were on the warpath they sang, Okaimbo ketu katito eteke tu lisanumbula tu tandako. ("Our village is little today, we attack, we extend.") Another war song is Ocisonde ci likoka ove o kasi vonjila tumdamo. ("Red ant that creeps along, you who are in the way, get out.") When men on the march came to a camping ground occupied by another caravan, they sang as a challenge, Cinene nye? Cinene onjamba kakuli okachama kavela ukuavo. ("What is the largest? There is no animal largest. The largest is the elephant.") During hauling and carrying, men sing, Yende, yende chale, ocimboto lomala vaco. ("Let it go, let it go, the crab, the frog, with its children.") Lifting loads is always accompanied by noises which suggest that someone is injured. Part of the men make deep grunts to which their companions answer with prolonged groans. Initiation Evidence bearing on initiatory rites in Angola shows that the ceremonies are arranged as a process of incorporation into the Education 227 tribes, for everywhere these rites aim at securing ideas of unity, coop- eration, conformity to tribal law, and admission to adult tribal life. The methods used to achieve these aims are seclusion, circum- cision, physical suffering, direct tuition, dancing, hunting, a change of name, and finally a ceremonial return to the tribe with adult status and the right to marry. The following notes give details of ceremonies witnessed at three centers; namely, Katoko, Ngalangi, and Cangamba. For comparison of these rites with others performed in eastern Angola reference should be made to the books of A. Schachtzabel, and to the papers of H. Baumann (III), and F. and W. Jaspert, whose observations were made independently of each other and of my own investigation. The dances of the novices at Klatoko, where there is a mixed population of Ovimbundu, Vangangella, and Vachokue tribes, are part of the final ceremonies following circumcision and seclusion. The social group formed by this collective circumcision, seclusion, and dancing, is called ovinganji, which is the name of the initiatory rites themselves. A boy who has been initiated is not allowed to become friendly with one who has not suffered the ceremony, and all boys who were circumcised at the same time preserve a sense of unity by dancing in a company and moving about together for twelve weeks after their return to the village from which they came. Circumcision, which is prohibited by the Portuguese govern- ment, is still practised secretly in some regions. When circumcision is carried out by primitive methods serious infection may result from the lack of clean instruments, for the knife is, of course, unsterilized. After the operation the patients are subjected to harsh treatment during the period of cure that follows. The circumcised are secluded in a wooded area. Food provided by the parents is placed on the bank of a river near which the initiation camp is built, and after the parents have gone away the boys bring their rations from the river to the camp. Before eating, the boys are obliged to give profuse thanks to the men who are acting as tutors. In some regions there is no food ration for the boys, and the rule is that each boy must subsist on what he is able to catch and collect with his hands. Each boy has a male guardian who takes care of him after the operation. Those who have been operated upon are not permitted to wear clothes, nor are they allowed to warm themselves until quite healed. Any infraction of the rules results in a severe whipping. 228 The Ovimbundu During the period of isolation costumes for the dance are made. These consist of clothing of tightly fitting, coarse netting, masks, and girdles which are for the use of only those boys who have been circumcised. The feasts and dances celebrating the conclusion of initiatory rites are of great importance. Women are not supposed to know that the operation of circumcision is taking place, and they are taught that ovinganji are supernatural beings who have sprung up from the earth; therefore every effort is made to conceal masks and costumes from the sight of women and the uninitiated. No female is allowed to go near the enclosure where novices are confined. A few days after observing the costumes and dances of the newly initiated boys at Katoko I was in the Ngalangi region at the village of Ngongo, about a hundred miles to the north of Katoko. At Ngalangi two boys of the Ovimbundu were questioned with regard to their experiences in the initiation camp. It is certain that initiation ceremonies are held at irregular intervals and not more frequently than once in four years. The name given to the ceremonies for boys in the Ngalangi area is ocinganji (oci, "big"; nganji, "judge," or a masked person). When there is a number of boys who have not been circumcised, these approach the oldest men to ask for a circumcision ceremony. The old men visit the sekulu ("headman") of the village to request that arrangements shall be made. An ocimbanda ("medicine-man"), assisted by other men, prepares a camp in a wooded area. Usually the father of each boy has to arrange that a tutor shall accompany his son to the camp, but sometimes as many as three boys have the same guardian. The guardian receives a small fee, possibly nothing more valuable than a chicken. The camp is made on the side of a stream remote from the village. Each boy takes a chicken to the camp for the purpose of making a special meal, which is given at the name-changing ceremony which follows initiation. The chicken is eaten soon after the boy has been circumcised. The boy changes his own name for a new one which is announced in the village from which he came. While circumcision is in progress the enclosure is surrounded by male drummers and men who shout, so that any noise made by the boys during circumcision will not be heard. The period spent in camp is uncertain, probably from three to six months. The rule is that camp is not disbanded until every boy is healed; which means that the confinement may be greatly Education 229 prolonged if even one boy fails to recover. One of the two youths interrogated said that in his camp there were seventy-eight boys, three of whom died. My other informant said that in his camp there were sixty-eight boys, only one of whom died. The informants agreed that the deaths were due to an epidemic of influenza and not to privations or septic conditions arising from the operation. During the period of seclusion the boys are taught songs and dances used at the ceremony that celebrates the conclusion of their initiation. By privation the boys are taught the value of food and fire. Novices are beaten if they show any disrespect for their guardians, and trifling offences are severely punished. Every boy has to take from the fire a burning stick, which he holds in his hand while running between two lines of men who beat him, and if he drops the stick he has to start his run once more. The boys swear allegiance to one another. A novice thinks that he will die if he gives information to a woman or to a man who has not been initiated. From his mother, each boy receives food in a gourd which he deposits on the bank of a river near the novices' camp. If a boy dies, a hole is bored in the gourd so that when the mother receives this she will not send more food. Each guardian has a stick which is sent to the mother of his pupil when the ceremony of initiation and seclusion is ended. If the boy has died, bark is cut from both ends of the stick before this symbol of death is sent to his parents. When the boys come out from their camp at the conclusion of the rites one man and one woman stand on the bank of the river, and the boys pass under the legs of both the man and the woman. When the boys arrive at the village the whole population comes out to welcome them. There is a feast and beer-drinking bout on the day of return. During seclusion no water is allowed for washing, and the inform- ants laughed as they spoke of the order "wash hands," whereupon the boys merely rubbed their hands together. When the order, "wash dishes," was given the boys rubbed their platters with their hands. On the day of leaving camp the novices bathe in the river. Ablution consists of three dips, after each of which the boy stands on the bank until he is dry. For two months the boys wear strips of bark cloth. During this period all the novices must move together, but there is no objection to their leaving the village provided they do so all in one company. While wearing bark cloth the boys have each day to attend a cere- 230 The Ovimbundu mony at which the older initiates dance while the novices clap their hands. I witnessed the dance of initiated boys at Ngongo, where the costumes resembled those I had seen at Katoko. At night a youth brought for me a set of initiation costumes (Plate LXXVIII, Fig. 1), which resembled those worn during the dance; this he did with great secrecy. While at Ngalangi, the initiation of girls was investigated. After sev- eral days of negotiation with a village headman three female guardians of the girls came from the bush. The illustrations (Plate LXXVII, Figs. 1, 2) show the attitudes of these women and their decoration. There were three male drummers in the orchestra and several women, who sang and clapped their hands. On emerging from the bush the females moved toward the orchestra with their backs toward the players. As there were about two hundred yards to cover, this slow backward movement occupied a long time, since the three women did not take more than a few inches at each step. The dance itself was a slow, shuffling, swaying movement, made while the bodies of the performers remained inclined forward and their heads were bent so that it was almost impossible to see their faces. The women wore no clothing except loin cloths. Each female was plastered from head to foot with alternate bands of red and white clay smeared thickly. These three females are the women who act as instructors for the girls during isolation in the bush where they receive sexual and domestic instruction. The seclusion of girls covers a period of one month. During this time the novices suffer no harsh treatment though an operation is performed, possibly excision of the clitoris, but I am not certain on this point. It was also stated that a corncob is introduced into the vagina. The tribe practising this ceremony for girls was the Vanyemba, living close to the Ovimbundu at Ngongo near Ngalangi in central Angola. In order to corroborate further the two accounts of initiation ceremonies for boys I journeyed to Cangamba in Moxico, eastern Angola. The position of the two first centers visited (Katoko and Ngongo) is on the border line where Ovimbundu and Vachokue mingle, and although there is a mixture of these and other tribes at Cangamba, the culture, language, and physique of the Vachokue predominate. A description of the circumcision ceremonies witnessed at Cangamba is given below for comparison with the accounts resulting from visits to Katoko and Ngongo. Education 231 The compound in which the initiates had been confined for three months was a circular enclosure made of light branches and leafy boughs, at the narrow embrasure of which stood a guardian of the boys, who permitted entrance. In the middle of the large enclosure were seven small cages, each of which was just large enough to allow one boy to lie on his back, and I was informed that the boys lie thus for two weeks after circumcision (Plate LXXX, Fig. 2). The ages of the boys appeared to be from ten to seventeen years (Plate LXXX, Fig. 1), an observation which agrees well with that made at Katoko and Ngongo. The disparity of age among the novices at each center where initiation was observed bears out the statement that initiation ceremonies are held at irregular intervals, but not more frequently than once in four years. During confinement in the large enclosure the novices had made masks and costumes (Plate LXXIX, Fig. 1), and when they were pulling on the coarse netting suits, which fit tightly, I observed that circumcision had been performed thoroughly, evidently some weeks ago, for the wounds were healed. Masks were obtained, and these the instructors of the boys were careful to wrap in bark cloth, at the same time requesting me not to show the objects to women. A few days later these boys returned to their village to dance while wearing the costumes which had been made in the enclosure. At Cangamba the final ceremony lasted twelve hours amid great excitement, including ceaseless drumming and dancing. One boy, who wore a skirt on his fiber costume, lifted this to display an artificial penis of great size (Plate LXXIX, Fig. 2). This organ he grasped in his hand while chasing the women and girls, who ran away screaming and laughing. At this ceremony there appeared stilt-walkers (Plate LXXXI, Fig. 1) and a medicine-man, a member of the Luchazi tribe (Plate LXXXI, Fig. 2). Circumcision among the Bdjok (Vachokue) has been witnessed by C. P. Holdredge and described by Kimball Young (Amer. Anthr., XXIX, pp. 661-669). This ceremony took place in the far north- east of Angola. The man described as the magish of whom women are afraid, but who is known to the initiates, corresponds to the Uluchazi medicine-man seen in the final ceremony at Cangamba. Of this grotesquely dressed person the men and boys took no notice, neither did he pay attention to them. On the contrary he chased women, who ran away screaming. Kimball Young's description of the enclosure where boys are confined, and his account of the 232 The Ovimbundu short fiber skirts worn within this enclosure by the novices, are in agreement with the observations made at Cangamba. There is also a close resemblance between the independent accounts of H. Baumann and F. and W. Jaspert for the Vachokue, also P. A. Delille and E. de Jonghe for the southwest Congo. The question of initiation rites, past and present, among people of pure Umbundu culture is important. The matter is more fully dealt with later in discussing cultural contacts and cultural losses, but there are a few points which should not be omitted here. At Caconda, an Umbundu center of western Angola, Pere R. P. Laagel stated that only two years ago he had visited an initiation camp of Ovimbundu boys in the hope of persuading them to return to his mission school at Caconda. My informant stated that the boys were circumcised, beaten, and confined to an enclosure for several weeks. Those who cried out when flogged received a double portion of blows. Poisoning is the fate of a boy who reveals the secrets of the initiation camp to women or uninitiated boys. This is the most direct testimony I have received concerning initiation rites among unmixed Ovimbundu. From Bailundu and Elende I have obtained slight evidence of initiation rites. At the former place, which, like Elende, is a center of the Ovimbundu tribe, fiber costumes were obtained. The mask from Elende is unlike those used at Cangamba, Katoko, and Ngongo. The fiber suits are, however, no different from costumes used at these places. Initiation ceremonies are not held at Elende today, and beyond Dr. Ennis's statement that circumcision rites have been revived in the past fifteen years of his long residence in the neighbor- hood of Elende, I have no evidence of ritual. The data relating to initiation of boys in Angola suggest loss of the initiatory rites from Umbundu culture, until only a few old masks and costumes, along with attenuated ritual, remain sporadi- cally. On the contrary, in eastern Angola and particularly among the Vachokue, who are farther than the Ovimbundu from coastal influence, initiation ceremonies flourish. At present there is occurring a penetration of initiation rites from the Vachokue area into east- central Angola, which is a region of cultural admixture. The foregoing pages have described certain institutions and social forces which collectively bring individuals into conformity with tribal standards of conduct. The influences at work in this direction may be direct or indirect. Education 233 Among direct educational forces are the home training given by parents and the mutual responsibility of children and their mother's brother, whose reciprocal obligations have been explained. Family life is undeniably an important institution whose power is by no means suppressed by the strength of the village community and tribal organization. Initiation, which may or may not coincide with puberty, is perhaps the most potent direct influence in formation of character and the inculcation of principles tending to tribal solidarity. Formerly warfare and hunting were more important than they are today, consequently initiation rites imposing hardship and emphasiz- ing the importance of concerted action, had a greater social value than they have at present. Formal instruction in industries such as iron-working, or pro- fessional hunting, indicates one aspect of occupational training. In these instances there are apprenticeship and an initiatory rite. All children are, however, subject to an indirect industrial education through the agency of play, while a few, according to desire and aptitude, become experts in some particular occupation. At all times the force of suggestion is at work in the home, in the men's council house, and through everyday observation of the conduct of elders. Probably folklore stories also have an indirect educational value, because some of the fables show the desirability of courage, honesty, and foresight, at the same time deprecating cowardice, stupidity, and low cunning. In chapter IX religious beliefs are discussed, and, in connection with these, standards of conduct are outlined. Although the Ovim- bundu have a belief in a supreme being and creator (Suku), and although they believe in survival after death, there is no evidence to show that conduct and education are influenced by theological beliefs. Suku, who is vaguely conceived, is far away and unin- terested. He issues no commands. Neither is there a belief in punishments and rewards after death. The standards of conduct are based on utilitarian principles which secure harmony and unity. The efficiency of the direct and indirect educational forces of the Ovimbundu is attested by the history of the tribe. The indige- nous system of education, supported by favorable environment, consolidated these people so that they became the dominating power in Angola. Moreover, their cohesion has been such that three centuries of European contacts have failed to eliminate all the basic elements of their tribal life. VIII. LANGUAGE Affinities of Umbundu The Umbundu language is widely understood in Angola owing to the journeys of the Ovimbundu as raiders and traders in large caravans. Umbundu is as important for communication in Angola as Hausa is for Nigeria, or Swahili for the east coast of Africa. Notwithstanding this use of Umbundu as a lingua franca it is necessary to note that tribes surrounding the Ovimbundu have their own distinctive Bantu languages; moreover, the Umbundu language itself has several dialects. For example, the Vasele tribe living in rugged country in the hinterland of Novo Redondo have a dialect of Umbundu (Usele) so distinctive that a competent Ocimbundu interpreter has difficulty in making himself understood, or interpreting what is said. Isola- tion of the Vasele from the Ovimbundu has led to the formation of a cultural pocket in which linguistic developments have shared in a specialization of culture. The differences between the Umbundu of Elende and that of the Ambundu tribes may be noted by comparing this brief outline of Umbundu, recorded at Elende, with the records of Amandus Johnson, whose research relates to the language spoken in the hinterland of Loanda. The present chapter will deal only with notes on vocabulary, phonetics, an outline of grammar, sign language, folklore and proverbs, all of which are briefly treated. In the field phonographic records of the Umbundu language were taken, and for the analysis of tonal values and other aspects of phonetics I am indebted to Dr. M. H. Watkins, a student of Professor E. Sapir, formerly of the University of Chicago, now at Yale. The section on Umbundu grammar conclusively shows Umbundu to be a representative language of the Bantu family of languages. Umbundu possesses all the fundamental characteristics of Bantu speech. The truth of this may be seen by comparing this outline of Umbundu with the analysis of Bantu languages given by A. Werner (Language Families of Africa, London, 1925). Moreover, Carl Meinhof (Grundzuge einer vergleichenden Grammatik der Bantu Sprachen, Berlin, 1906, pp. 112-115, and translation by A. Werner and N. J. von Warmelo, London, 1932) indicates the degree of relationship of Umbundu to other Bantu tongues by giv- ing comparisons of the stems of personal and possessive pronouns, along with other comparative data. 234 Language 235 H. H. Johnston (A Comparative Study of the Bantu and Semi- Bantu Languages, Oxford, 1919, vol. I, p. 350) places Umbundu in his Group X ; that is, the southwestern group of Bantu languages. Johnston's map indicates that throughout Angola there is a gradual transition from the southwestern Bantu group to that of the south- west Congo, which is exactly what might be expected from a con- sideration of the geographical position of Angola. Johnston's grouping is based on root forms and vocabulary. This difficulty of making a comparison of the affinities of Umbundu does not preclude the possibility of recording some preliminary observations, especially with regard to vocabulary. F. and W. Jaspert have prepared a comparative vocabulary of 250 words in the languages of the Ovimbundu, the Luena, the Basongo, the Luchazi, the Luimbi, and the Vachokue tribes. Inspection of the columns indicates a close connection of Songo, Luchazi, and Luimbi; in fact, the vocabularies of the two latter are almost identical. The vocabularies of the Vachokue and the Luena have much in common with those mentioned, but the Umbundu vocabulary has only a minor agreement with the vocab- ularies of these eastern and northeastern tribes. There are, how- ever, some words which are identical in all these languages mentioned by F. and W. Jaspert. Examples of identity, or very close similarity, are found in the words for "bow," "elephant," "to eat," "firestick," "finger," and a few other forms, but the principal deduction from inspection of the columns is the distinctiveness of the Umbundu vocabulary. On the contrary, examination of the vocabularies of the Ovim- bundu and the Vakuanyama, of the south of Angola, appears to justify Johnston's inclusion of Umbundu with southwestern Bantu in so far as similarity of vocabulary is a criterion. The correspond- ence is further borne out by P. H. Brincker's "Lehrbuch des Oshi- kuanyama." In the light of these comparisons the relationship of Umbundu vocabularies is readily seen, for instead of searching for similarities the numerous identities are at once evident. At a glance one sees that the Umbundu and Kuanyama words for "arm," "arrow," "ax," "banana," "beard," "bird," "blood," "bone," "breast," "charcoal," "guinea fowl," "hand," and "hoe," are either identical or closely related. I noted the easy communication of my Ocimbundu interpreter with the Vakuanyama of Mongua. Some of the following chapters show that the Ovimbundu have an undeniable southwest Congo culture; they have, however, taken 236 The Ovimbundu cattle and a few other traits from the Vakuanyama of southern Angola, and there seems to be no doubt that the vocabularies of the Ovimbundu and the Vakuanyama have been influenced by these southern contacts. Yet structure will prove the final arbiter in deciding linguistic relationships, for the Ovimbundu as exten- sive travelers have widely distributed their vocabulary. Caution is necessary in making comparisons of Umbundu with surrounding languages; for, although Kuanyama has received careful attention, and the researches of Meinhof have given a reliable background for Ovambo and Herero speech, the language of the Vachokue of eastern Angola and the speech of the Mussurongo of northern Angola, call for a comprehensive work. Moreover, there yet remains the task of providing a standard Umbundu dictionary and grammar, and pending the collation of this linguistic material there are no means of determining the exact syntactical, phonetic, and other relationships of Umbundu. Vocabulary Apart from a broad question of the linguistic evidence of culture contacts, there are points of interest connected with the recording of vocabularies in a restricted area. Under the heading "Nature Lore" attention was called to an extensive Umbundu vocabulary connected with those activities which underlie the economic structure. The vocabulary showed that the Ovimbundu are keen observers whose activities have resulted in the formation of a large vocabulary arising from trades and occupations. In nature study many fine distinctions are made; thus, there are words to distinguish not only trees and plants of economic importance, but varieties of snakes and lizards which are of no great economic interest. In order to test further mental activity and the acquisition of vocabulary, an Ocimbundu girl of about twelve years of age was asked to name some of the objects in my collection. Without hesitation she made the following list. The object was indicated, then the child gave the name. This information was checked by an adult interpreter who found that the child was correct in every instance, although the objects do not specially appeal to the in- terests of a young girl. The objects named by this girl are as follows: Ocimanda. Wooden dish. Ukinda. Switch made from the tail of Omutopa. Tobacco-pipe made from a an ox. horn. Opatalonya. Leather pouch for belt. Etenga. Pipe for smoking hemp. Upi. Pounding stick. Ocinunga. Brass bracelet. Ombenje. Gourd with a long neck. Language 237 Onganja. Large round gourd. Embungumbungu. Bull-roarer. Onguwa. Net in which gourd is carried. Olosangu (singular, olusangu). Rattles Ohonji. Bow. for tying on the legs when dancing. Usongo. Arrow. Olumbendo. Wooden flute. Olundovi (plural olondori) Bark rope, g^ m oeifiumia0t Wooden hair. Ombulunwumba. Musical bow. comb. Onoma. Drum. Osinja. Long needle used in making OhuHya. Throwing stick. mats Ohanyanga Drill for boring wood 0nemk Wooden j and doye Ongombo. Basket used in divination. _ , -, - . Esanga. Water jar. 0sala- , Head-dress worn by kings, r\ > T t 1 -i warriors, medicine-men and hunters. Ocipupu. Jar for palm oil Small bmsh used b medicine-man. Ocimbangu. Skin-covered box carried _ .„ „ _ , to war by a king. Ocinumariuma. Wooden image of hu- Omusaka (diminutive, okamisaka). man "S111"6- Flour-sifter. Olupunda. Rat trap of cane. Opesi. Tobacco-pipe. Ombntesa. Snuff box. The names for colors are restricted to a few words; but in all other respects the Umbundu vocabulary is extensive. My interpreter said, "When I was learning colors, the women at the school told me many names, but I could see no difference at all." Ngonga contended that there was no difference between the color of a dark blue book and the black box on which it was resting. The following colors have names. Black is tekava. White is yela. Yellow is ondunga. Red is kusuka. Greens and blues are not well distinguished linguistically, but green is anirtamboto. The word tekavisa is used to indicate that a color is not distinctly green or blue. The word yelisa is used to describe gray, and all dark reds. Wumbula is the term which describes a greenish-blue shade. A further study of vocabulary was made by asking Ngonga to speak into the dictaphone. In doing so he gave lists of words including nouns in their singular and plural forms, together with many examples of other parts of speech and illustrations of syntax. These records have been transcribed by Dr. M. H. Watkins. (This section was prepared with use of phonetic symbols that were easily available. For the most recent system, see "Practical Phonetics for Students of African Languages," by D. Westermann and I. C. Ward, London, 1933.) Phonetics vowels The vowels are a, e, i, o, u, and perhaps a. They seem to be some- what lower than the cardinal vowels. There was only one word in which the vowel a was heard, dpdtdl&nd (leather pouch for a belt). This word, however, appears in the field notes as opatalona (opata- 238 The Ovimbundu lonya) and the obscure vowel heard might have been a result of imperfection in the record. These vowels are fairly close to the European vowels, their nearest equivalents in the European languages being approximately as represented below: a, as in German Masse. e, as in French He". i, as in English machine. o, as French eau. u, as in German Buck. 2, as e in German Klasse, and a in about. The nasal vowel q appears in a fairly large number of words, and the nasal g is occasionally heard. Before another vowel, u takes the sound of w: bhohyb (olu-eyo), "broom"; dtumqlq' twayge* (tu-ayge), "my little children." The rising diphthongs (a, je, ju, and jo and the falling diphthong a\ appear in the material, but every combination of two vowels does not make a diphthong. In several instances, when a word terminating in a vowel was followed by another word, the initial letter of which was also a vowel, the final vowel was elided. Final vowels on the low tone tend to vanish; that is, they are only slightly voiced, as in dsaygokaluygi (dsaygd dkaluygi), "he found (a) little hole"; ndav6len6ne~ (ndave"ld enenb), "I am big sick." (I am very sick.) CONSONANTS The table of consonants is as follows: StoDs 1 Voiced I Voiceless Spirantsj™<*f I Voiceless Aflfricatives I Voiced I Voiceless Nasals Lateral Semivowels Bilabial Dentilabial Alveolar Prepalatal Palatal Glottal b d, »d 9 V V t k f s dj, ndj tc h m n I n V w and y The symbols nd and ndj indicate "nasal attack" consonants, i.e., the fully pronounced consonants d, dj, and g are preceded by their homo-organic nasals not completely formed. It is as if one prepared to make the sound of n but before its completion changed to d or dj; likewise the change is made from y to g. Letter c is ch as in church. Language 239 tone and stress Apparently there are three tones in this language, but one cannot be quite sure, since under the somewhat unusual condition of attempting to record his voice on the dictaphone the informant might have distorted the pitch of certain syllables. That three tones can be distinguished when listening to the records, and that these tones tend to follow a definite pattern, for example, in the singular and plural forms of the same word, can be asserted with satisfaction; but there is less assurance in trying to understand all the nuances of these tones and in assigning them, without reservation, a definite place in the phonetics of the language. Nothing short of field work aided by delicate instruments can afford any satisfactory conclusions on this point. The data did not present any clear instance in which tone had semantic value. That is, there were no examples of two or more words, which, otherwise identical, differed only in tone and meaning. The field notes state that the difference between cardinal and ordinal numbers is one of tone only, and this was partially verified from the phonographic records, but the words were not clear enough for transcription. The following words have tones of semantic value, but the phonographic record was not clear. Kalunga, according to tone, can mean "god," "sea," "death," or "greeting." Ombambi can mean a "fever" (low tone) or a "bush buck" (high tone). Onjila means "bird" or "path." Ombundi means "gateway" or the root which is commonly used in brewing beer. Ongongo means "earth" (middle tone), or "hardship" (high tone). In this chapter tones are indicated by placing a grave accent (d) to indicate a low tone, and an acute accent (d) to show the occurrence of a high tone. The middle tone is left unmarked. The tones are not fixed, and they will be shown to shift in context. For example, a high tone on the penultimate syllable tends to be carried along to the last syllable also. This shifting of tone is especially noticeable when a noun is brought into concord with its qualifying adjective or relative, or when a noun is the subject of a verb. The following are instances in which tones do not maintain their original positions as found in isolated words: uti, "tree," utt unkind, "large tree"; blwkyb, "broom," 6lw6y6 luwa, "good broom"; dmunu, "person," omunu utito, "small person." This kind of tonal behavior was clearly indicated in five records of single words, and expressions of two words. But in the transcrip- 240 The Ovimbundu tions of folklore texts tonal behavior will be seen to be less consistent. Discrepancies in the incidence of tones may be due to different speeds at which words are spoken. The placing of tones is perhaps most reliable when transcriptions are made from free, continuous speech, as in the case of fluent reading, or talking in continuous sentences. Many words received a definite stress on the last syllable, and in others the penultimate syllable received a light stress. The majority, however, received slight, if any, stress, hence no definite conclusions could be reached on this subject. In several instances it is obvious that the informant was striving to be clear and distinct in pronouncing each syllable; therefore he possibly sacrificed certain characteristics of his intonation. THE SYLLABLE Every syllable ends in a vowel, consequently a vowel terminates each word, and consonants are pronounced with the following vowel, or with a consonant plus the vowel. The vowel of the penultimate syllable is long when the word stands alone, and in larger sound- groups the vowel is long in the penultimate syllable of the last word, while in the preceding word, or words, the corresponding vowel seems to be only half the length. Vowel length is not indicated in these transcriptions. Grammar the class system The data available were sufficient to establish the following classes of nouns, on the basis of their prefixes in the singular and plural; the formation of the adjectival or relative concords was likewise noted. There is no assurance that this list is exhaustive. Class Prefix Relative Singular Plural Singular Plural 1. omu-, omo- oma- U- va- la. 2. u- u-, oku- a-, ova- ovo- U- u- (ku-) va- a- 3. 4. 5. 6. u- o- olu- otci- ovi-, i- olo- olo-, a- ovi-, i- u- t tci- vi- vi- vi-, a- vi- 7. 7a. 8. e- -, i- oka- a-, ova- ova- otu- li- li- ka- a- a- tu- According to the principle of concord, the noun prefix is the governing element which determines the concordial agreement of parts of speech when these are brought into relationship with the noun. The prefixes also indicate number and, together with the Language 241 concordial agreement as stated above, divide the nouns into several classes or class genders. We need not enter upon a general discussion of concord, which in the Bantu languages is a method of expressing grammatical relations that are of fundamental importance. Concord need not always appear in the form of perfect alliteration, although it frequently does so, for the essential fact is that all nouns of the same class are recognizable as such, and other elements of speech when brought into syntactic agreement with these nouns must carry similar distinctions. Students of Bantu have suggested that the various noun classes probably represent a proto-Bantu classification which formerly betokened a grouping based on common characteristics (A. Werner, Some Bantu Linguistic Problems, Jour. Afr. Soc, XXVIII, 1928-29, pp. 155-165). The nouns listed here under Class 1 may accordingly be recognized as belonging to the so-called personal class, and the prefixes of Class 8 indicate a diminutive class. CLASS 1 Singular Prefix: omu-, omo- Relative: u- Plural Prefix: oma- Relative: va- omunu, person omQlq, child dmanu, people dmfylq, children Illustrations of concordial agreement with these nouns: omunu utito, small person drnQlq" utito, small child dm&nu vdtito, small people dmfylq' vdtito, small children CLASS 1A Singular Prefix: u- Relative: u- Plural Prefix: a-, ova- Relative: va- ulume, man ufeko, girl alume, men afeko, girls ovalume, men dvdfeko, girls ukqi, woman umalehe, lad akqi, women amalehe, lads ovdkqi, women The following forms show the concordial agreement for Class 1A: ulume utito, small man akqi vdtito, small women alume vdtito, small men ovdkqi vdtito, small women ovalume vdtito, small men afeko vene (va-ene), your (pi.) girls ukqi utito, small woman amalehe vavd, their lads 242 The Ovimbundu Singular Prefix: u-, oku- Plural Prefix: ovo- uld, bed ovdld, beds uta, gun ovdta, guns Concordial agreement for uld unenk, large bed ovdld dnSnd, large beds uta utito, small gun Singular Prefix: w- Plural Prefix: ovi-, i- uti, tree oviti, trees upi, handle ovipi, handles Concordial agreement for uti unene, large tree oviti vinenk, large trees Singular Prefix: o- Plural Prefix: olo- dndjo, house oldndjo, houses dmangu, chair dldmangUy chairs Concordial agreement for 6ndj6 yiwa, good house oldndjd viwa, good houses dmaygu yitito, small chair Singular Prefix: olu- Plural Prefix: olo-, a- dluni, fly dl&ni, flies dluhSygo, wild plum dlohtygo, wild plums CLASS 2 Relative: u-, ku- Relative: a- dkulu, leg dvdlu, legs okwdkwd, arm ovdkwd, arms Class 2: otf&A dZifo, small guns dkulu kutito, small leg dvdlu dtito, small legs CLASS 3 Relative: u- Relative: vi- utima, heart ovitima, hearts itima, hearts Class 3: utima utito, small heart ovitima vitito, small hearts CLASS 4 Relative: yi- Relative: vi- dyg&mb&, ox oldygfrmbb, oxen dygulu, pig dldygulu, pigs Class 4: dlomdygu vitito, small chairs dygulu yinSnb, large pig oloygulu vinknh, large pigs CLASS 5 Relative: Zm- Relative: w-f a- olwi (olu-i), stream oldndwi, streams olwiyd, broom oldndwfyd, brooms dlwiyd, brooms Language 243 Concordial agreement for oluni lutito, small fly oldni vitito, small flies olwi lunknk, large stream oldndwi vinkne, large streams Singular Prefix: otci- Plural Prefix: ovi-, i- otcimunu, thief dvimunu, thieves imunu, thieves otcitunu, pit ovitunu, pits Concordial agreement for otcitunu tcinknk, large pit ovitunu vinene, large pits Singular Prefix: e- Plural Prefix: a-, ova- ekd, hand ovdkd, hands ep{d, field ovdpyd, fields ewe, stone ov&we", stones dp&kd, fruit Concordial agreement for ekd litito, small hand dvdkd dtito, small hands Singular Prefix: -, i- Plural Prefix: ova- imbo, belly dv&imbd, bellies Concordial agreement for imb6 linenk, large belly ov&imbd dnene, large bellies isd litito, small eye Class 5: olwSyd luwa, good broom dldndwiyd viwa, good brooms clink, yd dwa, good brooms CLASS 6 Relative: tci- Relative: vi- itunu, pits dtcitd, one hundred ovitd, hundreds dtcipd, skin ovipd, skins Class 6: dtcipd tciwa, good skin ovipd viwa, good skins CLASS 7 Relative: li- Relative: a- dvdpdkd, fruits dpdkd, fruits ekdndu, wrong dvdkdndu, wrongs dkdndu, wrongs hpumu, corncob dpumu, corncobs Class 7: epyd liwa, good field dvdpid dwa, good fields CLASS 7A Relative: li- Relative: a- isd, eye ovdsd, eyes Class 7A: ovdsd dtito, small eyes imbd lidngfrmbe, belly of ox dvdimbd dlongdmbd, bellies of oxen 244 The Ovimbundu CLASS 8 As previously stated, the prefixes of this class signify diminutive forms. In Chewa, one of the languages spoken in Nyasaland Protectorate, these prefixes and the augmentative prefixes when used with the personal nouns convey a somewhat derogatory mean- ing. In Zulu the diminutive and augmentative suffixes, although not determining classes, are likewise of this nature when employed with nouns of the personal class. The records upon which this brief description of Umbundu is based were not clear on this point, but I note that -tito means "small," so that the distinction between dkalume ("little man") and ulume utito, translated in the field notes as "small man," may be of considerable importance. Likewise, otcvmunu ("thief"), Class 6, may be an augmentative-derogatory form in origin, provided that there is an augmentative class and that it coincides with Class 6. Singular Prefix: oka- Relative: ka- Plural Prefix: otu- Relative: tu- okandjo, little house otwalume, little men otundjo, little houses okatcipd, little skin dkalume, little man dtuvipd, little skins Concordial agreement for Class 8: dkamglq kdwa, good little child otum