Hak ee le eae ee = = atten ae ———— ti) HT Tt} bet +t Aree 1; Ht SES lsh so en sesa ss TSS ete teense erie hee eeeens = Peers tree 2 oa pigs: Titeteo eseek Ts The person charging this material is re- sponsible for ‘ts return to the library from which it was withdrawn on oF before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. To renew call Telephone Center, 333-8400 NIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN ST OS Seo L161—O-1096 Net Re a at ANTHROPOMETRY OF THE OVIMBUNDU ANGOLA | BY WILFRID D. HAMBLY CURATOR, AFRICAN ETHNOLOGY MU : aS & HISTORY. > THE LIBRARY OF THE . OCT 121988 UNIVERSITY OF {LLINOIS ANTHROPOLOGICAL SERIES ‘FIELD MUSEUM OF NATURAL HISTORY. j ‘VOLUME: XXV, NUMBER 2 | SEPTEMBER 30, 1938 PUBLICATION’ 426 7 , 7 Bry eye, Ay NAN } iH iy? Wiad pes Oe ye pee Sarbanes th oo ¥ {Y.48 o! a4 7 My =f, “ Sige $ oy mos) ‘he Aah + ‘ Beat ore Ss } ayy es Jee ey ae Roane 4G ; ae . ae tN Sf Ry. AD Sahay me Ths t > ADB AS oa eB Vary SAR: a 5s Sate Ah ey ae 4407S ees he aoeeets Rep a oie con Piet aOR of ‘ Se Ue f Pie e A Fee fae ek Ky Ny & Me at ates} tf . “Fee ‘a ’ See Ss ‘oe A e. ee a ake =) A iy ” A)! 4 ae Coa ie a, 3 ah Bd we Nee \ “ Nat Sei RY BS a my oS hs Pats
  • Fats 7 74. eR ¥ S Riaphit Ne Sees see oN Serte AS we Le D he A es es Lin es ee 1 eh SY tie e Wieg nti i . + ¥ er x ee ag . ka! rf te, ‘7 o ry ¢ fy: ew MY aS ay Zin) Cire, ‘ Eee ‘ BAN: ir WN Me ; aD kee ha y y +s, ks Ax cow PRL Re ! yt nN Pee han is fal hg ‘ i‘ ae ae “ re AP ay A A Ny eet (ftir paces re ‘ eK tt ee Patty’ ” ss 4 * =f See Be os ANTHROPOMETRY OF THE OVIMBUNDU ANGOLA BY WILFRID D. HAMBLY CURATOR, AFRICAN ETHNOLOGY MU gre Se s es EV & HISTORY HE e-------r7) sare HT Rrsrrevirr oh cows Ho gee bE SEES 6008-1 et Sead lee seeesesse fb: FOUNDED BY MAaanDeE. FIELD THE LIBRARY OF THE OCT 12 1938 UNIVERSITY OF ILLINGIS ANTHROPOLOGICAL SERIES FIELD MUSEUM OF NATURAL HISTORY VOLUME XXV, NUMBER 2 SEPTEMBER 30, 1938 PUBLICATION 426 i ll ll, le a ee ee aS ST ee ee ¥i25 CONTENTS PAGE Rae SP DURANT OAT TONNES 600. ia Lente taal wr Sr) por Pres Mehaes Meee ee 25 MURS dre Stra Ce ee SEA ee bee eal ae Sal rca ppm Cue 27 DEE a tS eee Kg gn, a kone tak fe WOR StS Sargent 29 SE RII 5 3 So ce yak ok 3 Eager Pee ee ote 29 Ye Pus MURINE 80 ot rf tog So ee ee LO 29 vn GUE PCLT ONION oo ek oe a Sor an Cong Rec ee a 29 - Nr RIPON tas a, Wie eon is, ee oe pa ip are Casals a 29 bs i PEOMMUINUIEN oc er eV gad tp aidow teed oe Rr eee 30 Foxy PINCUS RAM Fre tc nire ae es eek tks oe, a BR es Ra OS 30 ~ PIG OCES AN POCHMIGUE Fe AE ies eS Se ee ek a alae aha Nog 30 II. OBSERVATIONS AND MEASUREMENTS ON FIFTY-THREE OVIMBUNDU MALES ig CE ABEMES EAP E ee oe fo ao oe, Seek Se ee hg, Be ee Gm I 32 cpaneral (near vacsnn tree. set ee ay See, Cc a ne ee ual eae 32 PRROMPOTRTED GIN TROIOE fot. fe eles: 5 ea Ho ee, a ee 33 PUI fer) obs as eee R Dokl eae brah bce Se eo S ER ned ee ce 33 FIORE en FL eee ie ORG yee eG Mere fein SST AG te hod GE OL ae eal serio 33 PI acre ier ay eee a, al erieen yan A nana gs eRe On SMR. em tee 37 PROB ois ns EET ie a hk et tea de AR a 8, Ue ee 38 YT go oe rete 2 a ate ge ian ie annie eet tr aaa EMR re Neh Ap a me Cig 38 MMU ET MRIG So 6 ee art ch ae ees eb ek ah ee eg fa Alliage NaT aeee 39 do III. COMPARISON OF THE OVIMBUNDU WITH OTHER TRIBES OF ANGOLA oe) CE RMEIE SOR WEE carer cre eae EES 8S ae ales x 51 ~ UR LANTOS ee ee ores Meare. Oe Cera e re til tk, feet im ime oe te 51 as eens, Ten Pe rN se) oa pat ae ay eee 53 SRN SANCASOIOEE ess che eh ot Pele: gett wy we 5 nee egw hoes 53 ESD AE TRS Se Iie are ear Raat Sk ee pe HEPA a RN os CR Wig SAS RPO Ea 54 d IV. COMPARISON OF THE OVIMBUNDU WITH 1 NEGRO TRIBES OUTSIDE ANGOLA iv CP aae SEE ARIES coc ee ak. ee g oie ek ones 56 *: pS OCP ee ee ers ba eer ae © ae aT ge hava a 56 - Statistical Comparison of Ovimbundu, Ashanti, and Dahomeans.. .. 57 Comparative Variability of Ovimbundu Traits and Those of Other WRAL ATIOMID INCITING at lie See A TE ish LO eee 60 % PE Pae aon RUIN 05 fe ol Sh es OD oe! MEO AD om pee LP AS Sh BOTS 61 PANS C4 le Gane Sei eal oe ae muir On ee Pe Sen Pe Sea 62 * SUG ITSO eh ted ge SoG es Me els Dito Wedd Sub & sienna ce) Ge Mein Cee 62 3 LANNE Gan sy paar US ga ae gh aa Ge a's) eo teeiner beng sat Mile RD ee 63 V. SIZE OF FAMILIES AMONG THE OVIMBUNDU CPAMBIRY oso ae Reem 65 URANIUM hag aE ig SS See ww Daioh eh eee 65 O° BiQuers an Seeere GL WNAOFMANGS sk 8 re we ie He 65 i" CR EE ES So a era ee ee ee 68 sy CRIm es Aree Mtoe. Se. a eh el ey sl oe, tel dee 69 23 ~s) ro ™~s 24 CONTENTS VI. SCARIFICATION AND DENTAL MUTILATION . Introduction ae Notes on Scarification . BIBLIOGRAPHY INDEX . 72 pari PA . 72 . 74 . 76 —_—“« eo ee Siete LIST OF ILLUSTRATIONS PLATES I-IX. Front and side views, Ovimbundu males. X-XII. Front views, Ovimbundu males. XIII-XV. Side views, Ovimbundu males. XVI-XVIII. Back views, Ovimbundu males. XIX. Dental deformation. Left: central upper incisors considerably mutilated. Right: central upper incisors slightly mutilated. XX. Dental deformation. Upper left: slight mutilation of central upper incisors; upper right: no mutilation. Lower: mutilation of central upper incisors into chisel form. XXI-XXX. Scarification. Plates described in detail in Chapter VI, “‘Scarifica- COAR AS tion and Dental Mutilation.” TEXT FIGURES Graphs of Frequency Distributions PAGE PEC ree MC BILLING NEISNTS i Oe en lee ees ven Ege a ee ead ac ORT 43 . Sitting height as a percentage of standing height . .......... 44 Chest girths, and chest girths as a percentage ofsitting heights ..... 45 ee OMCLUED MIME PURO WOMENS. 8 a eo eo cc ra Te eee 46 . Bizygomatic and minimum frontal measurements .......... 47 BEG RO 8a ga eI ie nese cs RO kT Ea a we 48 aN ININIRUN (00, 5 eae Se Rae) (es Sigua os kp vay alle muni a SRS 49 Cees RUN ONE DORE ea Sa ct a Wa tak we was 50 25 PREFACE The data presented in this publication were collected in the year 1929 by the Frederick H. Rawson-Field Museum Ethnological Expedition to West Africa. The expedition was concerned primarily with study of material cultures of Angola and the assembly of a representative collection of objects from the Ovimbundu, Vakwan- yama, and Vachokwe tribes. Nevertheless, some time was devoted to a study of social organi- zation, religion, and other subjects that have been discussed in a previous publication (Hambly, 1934). At the end of the expedition- ary period an effort was made to measure a sample of 100 Ovimbundu, at Elende, and to record some data bearing on size of families and mortality. Unfortunately, this object could not be achieved in the time available, but the data recorded for 53 adult males are valuable, owing to the great paucity of information. There are, I believe, no anthropometric measurements of the Ovimbundu available for comparison with my own data, but fortunately we have statistics for four adjacent tribes (Cardoso and Corréa, 1916). The object of this publication is to give a correct impression of the physique of Ovimbundu males by means of anthropometric measurements and photographs, and to compare these data with records for adjacent tribes. It is desirable also to contrast the somatological characters of the Ovimbundu with some Negro tribes of Africa, outside Angola. A few notes on corporal marks and dental deformation of the Ovimbundu are added, and a record of the size of families is given. The data were obtained by questioning subjects respecting the survival of their brothers and sisters, and the survival of their children. The drawings were made by Staff Illustrator Carl F. Gronemann. WILFRID Dyson HAMBLY 27 Apa Ces, ; a m4 neg’ Sa ath ANTHROPOMETRY OF THE OVIMBUNDU ANGOLA I. TECHNIQUE The measurements recorded (Table 1) were based on instructions given in the Report of the British Association for the Advancement of Science (Cunningham, 1909). The instructions are as follows: HEAD MEASUREMENTS Maximum Length—From the most prominent point of the glabella to the most distant point in the middle line on the back of the head, known as the occipital point. Maximum Breadth.—Measured wherever it can be found above the plane of the ear-holes. The callipers should be held in a vertical transverse plane and moved about until the maximum diameter is ascertained. Minimum Frontal Diameter.—From one frontal crest to the other across the narrowest part of the forehead. For recent discussion of these measurements see Tildesley (1938). FACIAL MEASUREMENTS Upper Face Length.—From nasion to the edge of the gum between the two upper central incisor teeth. Total Face Length.—From nasion to the lower edge of the point of the chin. Maximum Interzygomatic Breadth—The maximum diameter between corresponding points on the opposite zygomatic arches. Gonial Breadth.—The diameter between the extreme outer points of the angles of the lower jaw. NASAL MEASUREMENTS Nasal Height.—From nasion to the subnasal point. Nasal Breadth—The greatest diameter, measured without pressure, between the wings of the nose. EAR MEASUREMENTS Greatest Length of the Ear—From the highest to the lowest point of the auricle. 29 380 ANTHROPOMETRY OF THE OVIMBUNDU Greatest Breadth of the Ear.—The maximum diameter at right angles to the length-line, from the ear basis to the hinder border of the auricle. TRUNK MEASUREMENTS Stature.—The subject is to be measured in the erect attitude, with his eyes directed to the horizon, his heels firmly planted, and the balls of the toes in contact with the ground. It is absolutely necessary that the subject should be symmetrically placed so that the mesial plane of the body is in every respect vertical. Height of the Ear-hole-—Measured from the ground. Height of the Acromion.—Measured from the ground. Height Sitting—This is the length of the trunk from the vertex of the head to the lowest points of the ischial tuberosities (height from seat). Chest Circumference.—Ask the subject to raise his arms. Pass the tape horizontally round the chest at the level of the junction of the fourth rib-cartilage with the sternum or breast-bone, then lower the arms and hold the tape tightly. The measurement was made at the end of an expiration of the tidal air. There was no conscious effort on the part of the subject to inflate or deflate his chest. INSTRUMENTS USED The instruments used were those illustrated by Martin (1928, vol. 1). The callipers (Tasterzirkel) with blunt oval ends are shown on page 125 (Fig. 48). The Gleitzirkel are callipers with pointed ends, pictured on page 127 (Fig. 49). The use of the height-measuring instrument is recorded on page 1538 (Fig. 66). For measuring the sitting height a box was used. This was squarely placed on perfectly level ground. The exact height of the seat was recorded and later subtracted from the total sitting height, which was measured from the ground. SS DEFECTS IN TECHNIQUE Table I indicates that measurements from the ground to the acromion were abandoned after 25 subjects had been measured, and the same is true for measurements from the ground to the ear- hole. These records were abandoned partly to save time for more important measurements, and to allow more opportunity for photog- raphy. Moreover, I felt uncertain as to the accuracy of calculating the height of the head in this way. The head-height calculation is made by subtracting from the total sitting height the sitting height TECHNIQUE 31 measured to the middle of the auditory meatus. The slightest deviation from a rigid upright posture during the recording of the two measurements causes a serious error. Yet, despite this objection to the technique, the measurements obtained for head height are what one might reasonably expect, with perhaps the exception of No. 2 (170 mm.) and No. 3 (160 mm.), which seem excessively high. The uncertainties entering into measurement of nasal heights, facial heights, and bigonial widths will be mentioned later when describing the variability of these traits. Even if we discard the incomplete series of measurements for acromial height and head height, we have sufficient data to give a clear impression of head, face, and trunk. Our data for the cephalic index, the nasal index, and the stature enable us to make a compari- son with those three characters as recorded by Cardoso and Corréa (1916) for four tribes adjacent to the Ovimbundu. Cardoso says little about technique, neither does he give individ- ual measurements; he presents averages and ranges. But we have no choice in our selection of comparative anthropometrical data, for so far as I am aware no figures other than those of Cardoso are available for Angolan tribes. II. OBSERVATIONS AND MEASUREMENTS ON FIFTY-THREE OVIMBUNDU MALES (Tables I, II) GENERAL OBSERVATIONS Skin Color.—All were brown, and only four approached a shade that bordered on black. At the lighter end of the scale there were four subjects who might be described as “light brown+.”’ The mode of the color distribution shows 19 individuals described as dark brown. The 46 observations on skin color are tabulated below: Light+-+ Light+ Light Dark Dark+ Dark++ 0 4 8 19 11 4 Reference to the skin color scheme of Martin (1928, vol. 2, p. 206) suggests that my “dark+-+” people are Martin’s grauschwarz division. My “dark+” division is Martin’s schwarzbraun class. The major “dark’’ division of 19 men is Martin’s dunkelbraun class. Our “light brown” and “light+’’ are somewhat like Martin’s reinbraun people, perhaps not quite so light. The color range of the Ovimbundu is distinctly lighter than among the west African Kru or Ibo, or among the Nilotic Dinka; and the general impression is one of medium to dark brown, with the lighter shades predominating among females. Eyes.—All were dark brown, almost black, of tint No. 16, which is the darkest shown in the Augenfarben-Tafel of R. Martin and B. K. Schultz. | Hair.—All subjects had typical Negro hair. The facial hair was sparse even when allowed to develop to full growth. The most hirsute man observed is shown on Plate III. He had a well-devel- oped moustache and chin tuft. There is little sign of hair on the chest, with the exception of the man shown on Plate X, No. 34. Teeth.—Teeth are in good condition. See Plates XIX, XX, and Chapter VI, for notes on mutilation. Scarification.—See Plates XXI-X XX, and Chapter VI, for notes on scarification. Ear Lobes.—One man in three had his ear lobes completely joined to the cheeks. This isa Bushman trait, and since the Bushmen were at one time more numerous in Angola, and farther north than at present, the trait may have been derived from ancient admixture of stocks. 32 OBSERVATIONS AND MEASUREMENTS 38 MEASUREMENTS AND INDICES THE TRUNK Standing Height (Fig. 4).—The average height is 1687.1 mm., which, according to the classification recommended by C. G. Selig- man (1930, pp. 14-15), places the Ovimbundu just within the tall class characterized by a stature of 1680-1720 mm. One individual is of pygmy stature (1406 mm.) but he has no deformity. There are 21 men, nearly one-half the sample, of medium height (1580-1680 mm.). ‘The mode lies at 1630-1722 mm., with 27 individuals in that class. If a height of more than 1720 mm. may be considered as “‘very tall’ (Seligman), then 17 men, that is, about one-third of our sample, are in that class. Generally speaking, the population is “tall” to “very tall.”’ Sitting Height (Figs. 4, 5).—The average is 848.8 mm. The mode is 825-845 mm. The majority of the observations, namely, 36 out of 51, are in the range 825-885 mm. The average index of 50.3 indicates that the sitting height is half the standing height, and the form of the graph indicates that this relationship has only slight variability. Chest Girth (Fig. 6).—The average chest girth is 828.0 mm. and the relationship (index) between chest girth and sitting height is 97.3 on the average. In other words the circumference of the chest is about equal to the sitting height. The majority of the indices (25 out of 47) lie in the range 95-100. THE HEAD Maximum Length (Fig. 7).—The average length is 187.3 mm., and the mode is the range between 182 mm. and 186 mm. Maximum Breadth (Fig. 7).—The average breadth is 136.8 mm., and a frequency distribution of breadths shows the modal value between 132 mm. and 138 mm. Cephalic Index.—The average cephalic index is 73.1, which brings our sample definitely within the dolichocephalic class, with index below 75. The frequency distribution of cephalic indices is of great value in showing that 77.3 per cent of our sample have cephalic indices between 70-75. The Ovimbundu are decidedly long-headed, only 22.6 per cent of the population having an index over 76. Head Height.—For a sample of 25 measurements the average height is 185 mm. é cg 08 Sh 9F 29 OTT OOT O8T OTT Sot OFT S6T $06 L68 OF9T SLLT cI é 99. LZ: OF 2% 99: -S0T- LIT 0ZT: FOL. 8Zl 88.621 O88 FS SPST --OLST lal *syaoyo 03 poutol A[yied saqo] iva ‘iley [eloey jo quowidojaAep wWnIpey[ yep 6G I8 Sh 9S EL GIT 06 SZt OOT SEt THT S8T O6L SH8 SPST PLOT €T ‘di do} puv uryo uo sail -JaAep [jem Ajitey Wey yep 09 OF Th GS ZL PCL OST OZT G6 82T SET PET O€8 8E8 EFST LEOT «aT “doy uo yey A[qveorjou pvey foryqeusoid ‘syeeyo 03 peulof seqo] ey + yep 1¢ Z3& 0G OG 99 SIT SIT L2T BIT 82t. OFT FET OGS8 E28 IST VEST UT yep 9¢ 3& Lb 8F 69 SIT G6 SET SIT OZ OFT 88T Pd 098 s09T SELT +0T “rey [eles asredg = + BIT GG «8% OHH «OOS «G9 «(OTL «(OT «(OT «(OTL «6ST «6GhT 688 «Sah «898 (OLST B8aLT 6 “ulyo pue dy do} uo sey esivdg + 31ep 69 FE HH FG OL EZT TOL OST GOL cat GET S8t 8 S88 SPST S89T 8 *piveq [Tews ‘ayoe} “snow poedo[erep-[[9M é 29 08 8h 6 OL OIT OOT STITT ZOT 92T OFT PET SL8 068 “""' S69T L *syooyo 03 poutol seqo| iva ‘iayjoul [[e} AJA B pey qoofqns ysoye} oy, ++ 3412p LG 9% Th 6H TL OT $6 G2t SOT 6FT OPT S6T OF8 GL8 GOLT 9P8T 9 yep 09 GZ Sh 68 LG ZOT OOT OFT OTT SET rT F8T ST8 088 Z6ST OZLT g "syaeyo 0} paulof seqo] leq yep €9 && PH Sh ZL Z2l OOT SEI OOL EST ShT 861 206 968 O99T O08T +P 99 82 Zh Sh ZO ZIT G6 OT GOT O9T BET LEI 0€8 968 96ST QHLT 448 é +743 1¢ Zé PH OF 19 SOT 06 OZ 86 OLT SET 861 082 OF8 OOLT j +43] 0G &&8 SF 8h G9 EIT 8 €2I OOT ZET S8t FBT 028 218 “°° PEST FT sysvUlagy ites ee ee es As ae as ee) i ae Ae i a a SO Caan Ae y9sue] leq = LT Yypeelq Ivy = oT YUjpReiq [VSeN= Gl 4 4YBIoYy [vlovy [210 = ST [eqUuOl] UNUITUIY, = 6 yjBuUe] peef]{ = 9 gOS 0} UOIWIOLIY = F qysiey [eseN = 77 [eluosig = TT qysloy pvoH = 8 Yyis yseyD = ¢ Wysloy BUIpuLys = B YystIey [elovy Joddy = $7 d1yeMosAzig = OT yypeeiq perey{= Z qYysSIey BUIVIIS = 7 Jequinu [euosleg = [ VIODNY ‘AGN IG ‘AGIUT, NGNNAWIAO AHL JO SHIVIT LINGY ATUHL-ALILY AO SINAWAUNSVAY I WTavi 34 *padojaaep [Jom Ajttey dey [elegy “syooyo 0} paulof seqo| Iva fayouysnowl 4YZI[S "syaeyo 0} peulof saqo| iva ‘iley [Bloey asivds ‘jensn UsYq} Paved s10Ul 9[79] V ‘asivds AJOA Iley [BIB *syooyo 04 peutol saqo, ivq “sHaeyo 0} poutol seqoy| iva ‘iey [eloey oesiedg “1ley [Bloey esredg "Iley [eloey asiedg *pedolaaep [jem Arey dey [eloey “syaoyo 03 poutol ‘payeiojiod saqo, ivq *syooyo 04 poutof Ajjerjied saqo] eg “ey [Vloey asieds ‘syooyo 0} poutol saqo, eq "11ey [Bloey asredg syspUlagy ++ yep ++ y1ep ++ 31ep tae 40109 urs Iv LY 8P 11 panwyywoj—T WIV TL 69 &1 as SIT Sit vor Sit vor STT vit OrT LIT LIT OIT OZT OTT 80t STI Sit e0r vor SIT Itt OTT SIT STT as $6 86 IT 0st vit cel or “uo}BIFNUT 4I00 J, } $28 $ a S6L 918 €18 v08 928 GOL v98 T16 098 ots ¢18 098 698 988 ons 198 SSL ¢18 618 L¥8 0€8 068 vI6 7 *UOTBOYLBIS » 0Z9T ~=s-«6E OS4T 88 Os9T 4218 SE9I 98 9ILIT 8 0Z9T ~=s WHE €89T 8&8 ZI8ST = «a8 Ost 418 9P9T 08 SoLI 462 ait 82 12 O89T 12 o} OL9T 92 e99I 9&2 GoLI = 8 9I9T 8&2 907 22 GZLI 412 cO9T §=s-:«O% re9t—s-« GT. 009T = S&T O99T 4427 GI8t 91 3 I Se Se *painsveul usu Jo Jequinu oy} ey¥oIpUl seseyyueied ul soin3ig t *UOTBOYUBIS » CSOF OTOF 6GLOF EOF EOF 6EO0F SSOF STVOF FEOF ELIF 620F LYOF FITEF 98FF 06LF pee “Aap cys 093 960 99° 987 c6S 908 S89 O2S O6TT LID GeL P09 YI9F E928 ZESL “PIS 9F0F 990F SOF 8POF LIOF PEOF POOF LEOF EEOF POF IZOF SZOF 8E0F 9E0F IS 0F aes “A jo 602 698 869 6¢L STL 81S 928 O99 OOS 188 Lo0e LE 999 FFG ZcEsg 667 "W209 CS OF SZOF 8ZOF SEOF SOF SSOF BS0F S9O0F 8HOF OIF IVOF LIOF PVDF OEVF PIIIF LOF | sueayy 6s¢ T62 Lev 88h aso SPrIt LE S22t BEOL OSET BOET SLIT O'8Z8 B88 EESST TL89T (eo) (gg) (9) (9) (es) (@¢) (g9) (gg) (9) (92) (89) (eg) (6%) (TS) (G2) (ee) é at 9T ST vas &T 6r IT Or 6 8 d 9 g 7 & *syaoyo 03 peutof saqo] ivy yep 1¢ 12 68 OF EL OZT 06 GIT OOT “°° SET S8T O18 828 “°° O09T &$ i 9 08 Iv PS SL OCT 06 S2I SOT © THT LET F28 098 ~~” SBoT cg “syooyo 03 peurof saqoy, ivy yep O¢ 1€ OF Sh V9 LIT OOT Set TOL “°° PET SBT AY vé8 “°"" GPT Ts “syaoyo 0} peurof seqo, ivy yep. 1s 0€ cy LY $9 STI O06 SIT 9G "°° OS SLE — 6h ©" 90ST og “syaoyo 03 peurof seqo, ivy yep 6h O€F FH ZG LL BIT G6 Got ZIT “°° OFT 68T OL8 608 “°° O69T 67 +¥yep gg OF Gh LH 89 S2I G6 OZT GOT “°° OFT 88T O18 ZE8 “°° 8BLT 8P *“syooyo 0} poutof saqo, ivy yep 0¢ LE LH FG BL {V2 G6 OZ OOT “~** OST SLT O88 026 “"** O6LT LY “syooyo 0} paurof saqo, 18q way Sh 9% Iv LY SL OIT 06 OZ ZOT “°° SST PST O06L OV8 “"°" OG9T ov qy31 96 92 Sh OS TL ZIT G6 OZT OOT “°° GST Z8T G08 088 “""° OTLT cv yay yo 18 $8 LH 8G OTT 96 Z2t OTT “*° LET O8T 028 O92 “"** O99T vy ‘dy toddn uo sey pedojeaep-]jam ‘syeeyo 03 poulof seqo] ivq qs 1¢ 2&8 Sh SS 08 9ZE GOT ZI OTT “°° 8h 002 SI8 088 “""" GELT &¥ 14311 ZG 82 Th 0S #9 OTT 06 OTT OOT “°° GET S88T SHB ~"° “*"* geet GV “syooyo 03 poeutof seqo] 1eq +%4lep cc ZE Sh 8h 99 SIT OOT S2t LOT °° S&T O6T G8 OLL “"*" 989T Iv “syaeyo 07 peutol Sseqo, Ive ‘padojeaep [oa Alirey iey [Vloey yep €¢ co Ib FH 2D LOT 06 SIT OOT “°° GET S8Tt OLL FEL ~'°* LOST 40% syapMoy pny Te) ee es Le Se oe Sy |e cee eee eS eee elk ee oe I papnpuoj—T ATaVL 36 OBSERVATIONS AND MEASUREMENTS 87 Cranial Capacity —We have now the data for calculating cranial capacity from a formula. The dimensions of the living head of the Ovimbundu are 187.3 (length), 186.8 (breadth), and 135.0 (height). According to Isserlis (1914) the capacity of the Negro skull is .0003849 X HLB +65 / VN. If we allow 8 mm. reduction in length, 4 mm. reduction in height, and 15 mm. reduction in breadth, to allow for the flesh, our dimen- sions become 179.8 mm., 121.8 mm., and 131.0 mm. for the length, breadth, and height, respectively. Then, according to the formula, the cranial capacity is 1101+8.9 cc., which seems somewhat low. The inter-racial formula of Pearson (1904) is given as .000337 X HLB +406.01. This yields a more likely result of 1370 cc. as the cranial capacity. Hooke’s formula (1926) reads .000366 H’LB+198.9+45.8/VN_ and gives a capacity of 1246+6.3. If we take the figures of cranial capacity assembled by Kitson (1931, Table IV, p. 292), the Teita skulls, 30 in number, are found to have a mean cranial capacity of 1316 cc., and 37 males of Tan- ganyika have a mean skull capacity of 1299.3 cc. Pearson’s formula gives for the Ovimbundu a skull capacity compatible with all the figures quoted by Kitson. According to Pearson’s formula the average cranial capacity of Ovimbundu males is 1370 cc., and the ranges of averages given by Kitson for various Negro samples of east, west, and south Africa is 1299-1422 ce. THE FACE Minimum Frontal Diameter (Fig. 8).—The average is 103.8 mm. The range of this measurement is from 90-118 mm., and 44 out of 53 measurements lie in the interval 99 mm. to 108 mm., so indicating a fairly low variability. Bizygomatic Width (Fig. 8).—The average is 122.8 mm. The range of measurements, namely, 102-140 mm., is somewhat greater than that for the minimum frontal diameter. But 32 cases out of 53 lie in the interval 120-124 mm., and the distribution is therefore well concentrated in the small interval of 4 mm. Bigonial Breadth (Fig. 10).—The average distance between the gonial angles is 97.6 mm., but the measurements are not closely concentrated. Out of 53 measurements 46 lie between 90-100 mm. 38 ANTHROPOMETRY OF THE OVIMBUNDU The extremes of the distribution are 88 mm. and 130 mm., a some- what wide range. Total Facial Height (Fig. 9).—The average is 114.2 mm. The range, 102-126 mm., and the distribution are rather widely scattered. Yet there is a definite modal value of 22 measurements out of 53. Nearly half the distribution is within the small range of 115-118 mm. Upper Facial Height (Fig. 9).—The average is 68.2 mm., and the range is 57-80 mm. The distribution of measurements is not well concentrated. Facial Indices.—The relationship of the upper facial height to the bizygomatic width is 55.9, and for the total facial height com- pared with the bizygomatic the index is 93.7. The jugomandibular index, which shows the proportion of the bigonial width to the bizygomatic, is 79.9. The jugofrontal index, expressing the minimum frontal diameter as a percentage of the bizygomatic, is 85.0. THE NOSE Nasal Height (Fig. 11).—The average is 48.8 mm., and the range is 39-56 mm. The distribution is fairly closely concentrated, since 38 measurements out of 53 lie in the small range of 44-50 mm., that is, within the compass of 6 mm. Nasal Width (Fig. 11).—The average is 42.7 mm., and the range is 33-50 mm. The concentration of the distribution is within the small range of 4 mm. There are 43 out of 53 measurements lying between the values 40-44 mm. Nasal Index.—The average nasal index is 87.9. If we consider 55-70 as leptorrhine values, 71-85 as mesorrhine, and 86-100 as platyrrhine, our average index (87.9) is just within the platyrrhine group. If the index had been two points lower, our sample of Ovim- bundu would have been at the upper limit of the mesorrhine group. When we come to a comparison of our sample with other Negro tribes we shall probably find that the Ovimbundu have relatively narrow noses. With regard to the distribution of 53 nasal indices, 35.8 per cent of our men are mesorrhine, 58.5 per cent are platyrrhine, and only 5.7 per cent are hyperplatyrrhine (index 100+). Measure- ments on the nose show high variability. THE EAR Length.—The average length is 53.9 mm. The distribution is concentrated between 49-57 mm., and within this range of 8 mm. 44 out of the 53 observations lie. OBSERVATIONS AND MEASUREMENTS 39 Breadth.—The average breadth is 29.1 mm. The breadths are, however, more variable than the lengths; the mode lies at 31.0 mm., and 35 out of 53 observations are in the range 28-31 mm. Index.—The average index is 54.0, so indicating that the width of ears is a little more than half their length. The mode of the distribution is close to an index of 50. The distribution is, however, somewhat scattered, the total range of the ear index being from 45-66. Out of 53 indices 37 range from 50-58. VARIABILITY If we ask which traits of the Ovimbundu are most variable, and which seem to be the most constant and well entrenched, the question can be answered by consulting the coefficients of variation for traits and indices (Tables I, II). We are comparing traits of the Ovim- bundu inter se and without reference to the variability of correspond- ing traits in other Negro groups. The probable errors of the coeffi- cients of variation are given in the tables, but here, for practical purposes, we may use only the coefficients. If these are placed in ascending order of magnitude the following arrangement results: COEFFICIENTS OF VARIATION A: FeOee Bresathis. wscc cds oso 3. 27> S182: Chest girth -«..2 22. ae eee 5.56 2. Sittingheight-standing height 14; Bisyeomatic oc. 05.562k ek 5.60 MMIOE i SON ia eee ee Sak 1b: Nasal width... orsscca etek 6.93 8. Cephalic index............. 3.45 16. Total facial height—bizygo- a, FOO POON ae a5 oe ee ee 3.87 matic Index... o46. 123 os 7.07 5. Standing height............ O29: — 175 Bar ene, «eae heeds 7.09 6. Minimum frontal diameter— 18. Upper facial height......... T.18 head breadth index...... * 40-19; Nasal netont: 3-065 sy ccecek 7.29 7. Minimum frontal diameter— 20s BIgOnIal WIGGN::-- «sts sks 8.26 bizygomatic index....... O205 421.) EAT RORGRLD ses eees 8.59 8. Minimum frontal diameter... 5.00 22. Head height............... 8.81 9. Total facial height.......... 5.18 23. Bigonial-bizygomatic index. 9.05 10. Acromion to sole........... Be) sos INMRAL INGCR acces cs ee cle sO) 9.15 11. Chest girth-sitting height 20s: RP MOOR Sos Sas cea oh os 9.40 io)! ily I eC ee 5.39 26. Upper facial height—bizygo- TS. bet hele oc es bs ses 5.44 WAGIC TNGOK: ose cdee sas 9.53 For constancy in measurement head breadth stands first, with a low coefficient of variability, 3.27. Head length is almost as constant, and the cephalic index has also a low variability. The relationship of the sitting height to the standing height appears to be very constant (V=3.34). In the group of coefficients of variation ranging from values 4 to 5 are standing height, minimum frontal diameter, and the indices derived therefrom. Of all facial measurements the minimum frontal diameter is the one most likely to be free from error, since the measurement is made between two well-defined bony crests over 40 ANTHROPOMETRY OF THE OVIMBUNDU TABLE II INDICES FROM MEASUREMENTS OF FIFTY-THREE ADULT MALES OF THE OVIMBUNDU TRIBE, ELENDE, ANGOLA No. 1 2 3 4 5 6 ‘4 8 9 10 1 61.5-. 94.0 °-72:8 62.8. “91.9 61.5 <81.35 76:2 ~°89-6.. 56,0 2 49.4 92.8 66.7 50.8 87.5 75.0 81.7 74.2 95.6 62.7 3 61.3. 92.7 030 47.7% 86.1 18.2) 80.8 16.1". 87,5 60:0 4 49.8 100.7 73.2 64.1 91.7 16.2. 76.2 69:0 97,8 52:4 5 bleed. «92.0 T.2. 40.7 (258 (sk: 78.6) 1720 T10.2> 50.8 6 47.4 96.0 74.9 56.8 96.0 76.0 84.0 71.9 88.7 45.6 7 b2.6 98.8. 72.2: ‘60.9. 300.9 86.9. 38.7 (2-8: O98. .48.4 8 62.0. 96.4 73:0 b3.8 94.6: 71.7 30.8" 77,8 Sl. S876 9 AQ. (87,8. 1726" 52.0. :91:7 :83.8: 97. 172h. 3820: 50.9 10 49:6 -99.4 74.5. 51.9: 85.0 T1.4 °85.0' 380.7 (97:9 57.1 11 50.4 1083.3 72.2 62.0 90.5 92.9 92.9 84.8 100.0 66.1 12 49.7 99.0 71.1 60.0 103.3 108.3 79.2 68.8 74.5 50.0 13 50.5 938.5 77.4 58.4 95.2. 72.0 80.0 70.9 80.3 52.5 14 60.5 100.7 74.3 65.0 90.0 97.5 86.7 78.2 95.7 46.5 15 50.5 100.9 72.5 47.7 84.6 76.9 84.6 78.6 97.8 54.5 16 6054 101.2 68.3: O11. 86.2) 1236- 280.0> Tied 2th bl 7, ba bS.6- COT! .TL.S 62.0. $9.0) 83.8. Sh 14.) 80.7 60.0 18 olv9 10020. 771 62.6" SL. 79.2" 330.8 76.8" 100.0 4950 19 61.8 101.6..%73.7 62.8 88.8: 72.0. -84:0 976.0 «87.5: 48,2 20 SLO) 10864 Tt 2, BOst 98 cdr 1920) Shee eel O0.d 6020 21 60.7 96.6 (8.1) 64.2°°: 36.7 83-3. 83:3 71.9 —93.0° 68.8 22 51.6:.--96.9° 69.8 -68,3:.. 8658. 79.2 38828-7827 -* 89.1 66,3 23 63.3. 98.7. 76:3. 64.6. $0.8. 76.9. 80°58 .76:6> <95.7% °50.0 24 Meas of es 75.4 49.6 85.2 74.1 79.2 75.9 84.0 60.0 25 50.8. 89.3. 41.3 68.3 938.9 °82.6° 86-9: 74:6. “76:0: 49.1 26 68.0: = O92 “U1.0: JOU © 98-8) (80.2 Bind: lace! sebse OO. t 27 51.3 100.9 77.8 64.3 104.3 86.9 86.9 69.4 78.7 58.8 28 Ce en TS GOOCH BOS: VOR OLD A sO, Oece 29 AOS. cis, 10.70 59.2) 93.6: 92.0 “83:2 “TAS 12:7 3|-8622 30 61.9% 98:8 F651 42.5: 114.7 98.1 Obl > 74.6. 58220. 50.0 31 49.7 100.0. 73.4: °b4.2° -91.7 —83.3: 85.0: 72.3 “95.7 68.8 32 60.3". 96.6. 70:5 62.0 -91.2- 80.0 °84.0 7823 “-86.3° 60.0 33 51.3 99.5 70.7 54.4 92.0 74.4 84.0 77.8 86.5 55.8 34 AT2 «98.0: 71.4: 60.9: 04.5: “B18 61.8 (72.0. “98.1. 60.0 35 48.1 98.1 71.9 61.0 100.0 78.0 89.0 78.9 85.4 56.4 36 SO eT «3 Lote 73.7 48.0 838.2 76.0 94.4 84.3 95.5 63.0 37 52.0 96.2 72.7 59.2 94.4 89.6 85.6 75.9 82.7 56.4 38 60:0". «8272 10.7 -56.0.° 99°) 776: 89.6 18.2 018507 54.5 39 49.1 103.8 69.6 56.8 96.0 76.0 80.0 74.1 84.9 -47.0 40 60-:7-—97:.0 “71.8: 63.9 98:0 (78:3 :86-9 75.7 93.2. 46.8 41 45.7 109.7 71.0 62.8 92.0 80.0 85.6 79.2 938.7 68.2 42 5 she? tide evades (168: 60.2) | 94:8: 11756--86.2° 44.8) 232.0. 53.8 om which the flesh is thin. The bizygomatic also gives well-defined eminences; consequently, these measurements and the indices into which they enter are not likely to have a high variability because of difficulties of technique in measurement. On the contrary, the bigonial width (V=8.26) may have a high variability because of the range of measurements resulting from a more or less fleshy jaw. It is difficult to say how accurately we are measuring the width between the bony gonial angles. Inspection of photographs (Plates I-XV) proves, however, that our figures of OBSERVATIONS AND MEASUREMENTS 41 TABLE II—Concluded INDICES FROM MEASUREMENTS OF FIFTY-THREE ADULT MALES OF THE OVIMBUNDU TRIBE, ELENDE, ANGOLA No. 1 2 3 4 5 6 7 8 9 10 43 50.7 92.6 74.0 64.0 100.8 84.0 88.0 74.3 81.8 56.1 44 45.8 107.9 76.1 47.5 90.2 77.9 90.2: 80.3 70.2 57.4 45 - 61.5 91.5 74.2 59.2 938.3 79.2 88.3 74.1 86.0 46.4 46 60.9 94.0 72.38 60.0 96.7 75.0 85.0 76.7 87.2 67.8 47 51.4 90.2 74.3 65.0 101.7 79.2 83.3 76.9 87.0 54.0 48 So.4 + °O7:8 FAG 36.7 101.7 “79.2. 87.5 75.0 89.45.6159 49 47.9 107.5 74.1 61.6 94.4 76.0 89.6 80.0 84.6 61.2 50 na) ht ee 75.6 56.5 100.0 78.3 82.6 73.1 89.4 58.8 51 50.7 94.1 72.4 52.4 95.9 82.0 82.8 75.4 88.9 62.0 52 50.5 96.9 71.6 58.4 100.8 72.0 84.0 74.5 75.9 55.5 53 51.4 98.4 71.3 61.3 100.8 75.6 84.0 75.7 78.0 52.9 1 2 3 4 5 6 7 8 9 10 (51)* (47) (53) (58) = (53) (53) (53) (58) (58) (53) moore 50.3 97.8 73.1 55.9 93.7 79.9 85.0 75.9 87.9 54.0 errors | £0-16 +£0.52 +0.23 +0.49 +0.61 +£0.67 +0.38 40.31 +0.74 +£5.08 of V. | +0.22 +0.37 +0.23 +0.62 +0.46 +0.59 +0.32 +0.29 +0.60 +0.55 Std. 1.68 5.26 2.52 5.33 663 7.23 4.11 3.384 8.04 5.08 dev. | +0.11 +0.36 +0.16 +0.44 +0.43 +0.47 +0.27 +0.22 +0.53 +0.33 *Figures in parentheses indicate the number of men measured. ra 3.34 5.39 3.45 9.53 7.07 9.05 4.83 440 9.15 9.40 $i Sitting height 100 ma BigonialX 100 = Standing height i Bizygomatic (Jugomandibular) ee Chest girthx 100 t= Minimum frontal diameterx 100 Sitting height Bizygomatic (Jugofrontal) ee Head breadthx 100 es Minimum frontal diameterx 100 = Head length - Head breadth P Upper facial height 100 as Breadth of nosex 100 = Bizygomatic cs Length of nose 5 Total facial heightx 100 ‘Be: Width of earX100 =F Bizygomatic $5 Length of ear variability in facial measurements are no doubt genuine records of variability. There are many differences of facial contour, and the conclusion is that constancy of head form is accompanied by fairly high variability of facial measurements. The nasal index is in the highest ranges of variability (V=9+). On the whole the most entrenched features are proved to be head-form, stature, and the proportion of the trunk to the lower limbs. A comparison of coefficients of variation is aided by inspection of Figures 4 toll. These are graphs showing the frequency distribu- 42 ANTHROPOMETRY OF THE OVIMBUNDU tion of some of our measurements and indices. Compactness of a curve, which rests on a small base and rises to a peak without irregularities, is an indication of uniformity in the measurements, while, on the contrary, a straggling graph with a broad base, many irregularities, and a poorly defined modal peak indicates high varia- bility in the measurements. Owing to paucity of data, class intervals are few and the curves show no more than general trends and modal values. Considering that only 53 men were measured, one might suppose that the frequency distributions for all the traits would be somewhat irregular. On the contrary the curves are fairly compact, and those for distributions having a coefficient of variability of 3-4 are particu- larly well defined. We are dealing with a sample from a restricted area in and about the small village of Elende, and, so far as the notes go, only one man married a woman from a distant place. Only three men measured have brothers who were measured, and only four men have cousins (on their father’s side) who were measured. I believe, therefore, that our sample, though perhaps tending to uniformity through local selection in marriage, is not appreciably influenced by measuring men who have a blood relationship. Portuguese records show that the Ovimbundu have inhabited the high plateaus of central Angola during three centuries. The nature of the country tends to isolation, which would be an important factor in producing a uniform type. It is true that the Ovimbundu were caravan people who made long journeys eastward from central Angola; but the slaves, together with the ivory derived from these trading expeditions, were sold to the Portuguese, and in all probability there was little marital miscegenation of the Ovimbundu with slaves from far away. We may, I believe, regard our sample as-fairly representative of the Ovimbundu, who are tall, slim, dolichocephalic, Bantu-speaking Negroes. They are only moderately platyrrhine, and the skin colors range in the values medium brown to dark brown. The anthropometric measurements of the Ovimbundu may now be compared with those taken by Cardoso (1916) among four neigh- boring tribes of eastern Angola. Pere ey a *syyZIey ZurzyIs pus soinqeys JO UOIyNGLysIp Aouenbaly “pF ‘Oly S26 906 S88 S9IB SKE 9ZB SOS G8L GPL SHL GFL 218s COLL OL9/ O98/ e2L4 Q2L9¢ ESI 8ESt Coe, 9¢el OOxs i | | | | | y I | I | S}YBIOYY Busjzys | | | | | Sd1N}BIS T q al| q q i I 4 Sythe ae 9 ane | Auanbas4 it 43 Frequency Sitting height, percent of standing height O } l | | ] | | ! ] 45 46 47 48 49 50 57 S2 55 Fic. 5. Frequency distribution of sitting height as a percentage of standing height. 44 ‘syy Sey BUIQIIS Jo eBezUeoJed B sv SYIIIS Soyo pus ‘syzJ13 ysoyo Jo suoTyNqIIystIp Aouenbalg °9 “OIA LOi #0} 104 86 SE 26 68 98 #8 2B O£L6 906 288 8SE VLE O/B BBL Z9L ELL HL OED ae ER IO Sa I Ra RN ee | la or aN ool aa aloe SION eee a Lang WYySioy Burjps jo aBejusjsod & SB YIIIZ YsaYyA SU}NB YS9y4 @ © £ 6 0b CeO Q = ~ 4 et vad Aduanbos-4 45 *syjZus] pvey pues syypeeiq pvey Jo suoI4{nqii4sIp Aouenbely *) “OI Ole 902 ZOZ BE! HG! OG 9Bi CBs BLt HL} OL/ Oss vel eet cet | S | | | | | | | qT 1 | | | Sy Sua, pray Sypeeiq poy 02 22 Aduanbas-4 46 Od “s]UsWaINSvEW [VJUOI] WNUITUIW PUB dIyBWIOZAZIq JO SUOT{NqII4SIp Aouonbaly °8 “Diy Ort OF' 2Ok BS Het OS! Qi Slt BOs HOF OOF | 1 \ q a ' q I Ay t ' q | T SUIPIM D}BWIOZAZIG SYIPBaig |B}UOI) LUNI Aduanbos+ 47 "syyBray [eloey Jo suorynqiiystp Aouenbaly °6 ‘OT OOf oS: JYSIOY [BINBy [BIO] JYysiay jeioey todd . SS Aduanbo4s4 O02 ce] 48 Frequency Bigonial width ; ; | | L ; | 80 85 90 95 100 105 110 115 120 125 130 Fic. 10. Frequency distribution of bigonial widths. 49 ‘s}YZIley pus syypvelq [VseU Jo suOIyNqIIysIp Aouenbely “TT “S17 82 4 oe te Or cD S}YSIIY [BSBN q SY]PBIIG |BSBN » a Aduanbes4 22 92 50 III. COMPARISON OF THE OVIMBUNDU WITH OTHER TRIBES OF ANGOLA (Tables III-V1I) The only figures available for comparative study are those of Cardoso (1916), whose observations and measurements on males may be classified as follows: Tribe and number measured Habitat Vachokwe (90)........... 8°-13° S. Lat. and 18.5°-22.5° E. Long. Battie (62) ote s os oS as East of Kwanza River, between 11°-12° S. Lat. and 17°-18° E. Long. Pane (161 Sec ce 11°-12.5° S. Lat. and 20°-23° E. Long. Eatchaze (46). ..5.5....... About junction of meridians 19°-21° E. Long. and parallel 12.5° S. Lat. Hambly (1934) gives a map showing the location of these tribes, as does Cardoso (1916, p. 256). The Ovimbundu are centrally situated, and the other tribes mentioned inhabit territory to the northeast, east, and southeast of the Ovimbundu. The number measured is smallest for the Luchaze (46), then for the Ovimbundu, with observations on 53 men. The samples for the Vachokwe (90), the Luena (101) and the Luimbe (82) are more satisfactory. The series is, however, the only one available, so we are obliged either to reject the data, or to accept them with whatever defects there may be. Anthropometric measurements for the Ovim- bundu are given in Table I and the indices for these measurements in Table II. The four tables for comparative study of the data collected for these tribes are Nos. III-VI. STATURES (Tables III, IV) The tallest people are the Luchaze, with an average height of 1704 mm. The Vachokwe are next, with a stature of 1695 mm., only 9 mm. less than the Luchaze. The Ovimbundu rank third, with an average height of 1687 mm., which is almost identical with the average stature for the Luena, namely, 1685 mm. The Luimbe are shortest, with a height of 1671 mm. With the exception of the Luimbe, whose stature is in the highest range of “‘medium”’ height, all the tribes are “‘tall,’’ according to Seligman (1930, p. 12), who follows the system of grading and nomenclature adopted by A. C. Haddon. 51 52 ANTHROPOMETRY OF THE OVIMBUNDU TABLE III COMPARISON OF AVERAGE MEASUREMENTS OF ANGOLAN TRIBES (In millimeters) Cephalic Nasal Tribe Stature index index Ovimbundu’. .....%.. 1687 fever 87.9 Wachokwesast300o 1695 Wek 96.6 EAU DE: cc 3 chee: 1671 75.6 98.6 TAICN AA on oe cmarour 1685 (i a, 97.4 ENUCHaze) Ses oe 1704 f five! 98.5 The frequency distribution of the statures for five tribes is given in Table IV. This distribution has to follow the scheme of class intervals adopted by Cardoso, since he does not give measurements for individuals. If he had done so, we could have made our own class intervals. All the tribes have a very small percentage of men in the class below 1600 mm., the range being 3.0 per cent for the Luena to 6.5 for the Luchaze. The class intervals, 1600-1649 mm., 1650-1699 mm., and 1700 mm. +, emphasize the superior height of the Luchaze, for whom more than half the sample is taller than 1700 mm. The Luimbe, who have the shortest average stature, show, as we might expect, the greatest number of individuals in the classes 1600-1699 mm. (medium height). TABLE IV PERCENTAGE FREQUENCY DISTRIBUTION OF STATURES TRIBES Cass INTERVAL — a — Ovimbundu Vachokwe Luimbe Luena Luchaze Less than 1600......... SiS 3.3 6.1 3.0 6.5 1600-1649......... 22-06 16.7 ARES 23.8 19.6 1650-1699 :......3.0. 2: 34.0 35.6 31.9 29:37 19.6 BQO SES is vee eas 39.6 44.4 36338 Asie 54.3 The testimony of the distribution on the whole bears out the significance of the averages. The order of tallness is as follows: According According to to averages requency distribution Luchaze (tallest) Luchaze Vachokwe Vachokwe Ovimbundu Luena Luena Ovimbundu Luimbe Luimbe The only difference in the grading is that for the Ovimbundu and the Luena. According to averages the former (1687 mm.) are practically the same height as the latter (1685 mm.). There are, however, for the Luena 48.5 per cent in the tall class, but for the Ovimbundu only 39.6 per cent in that class. There are in the ANGOLAN TRIBES COMPARED 53 Ovimbundu sample a few tall men whose exceptional height raises the average, but when classified according to stature frequencies there are not so many tall Ovimbundu as Luena. CEPHALIC INDICES (Tables III, V) In comparative tribal study of this trait we have to recognize distinct differences of head form. Taking first the averages, the Ovimbundu with an average C.I. of 73.1 are decidedly the most dolichocephalic of our samples. Next come the Luimbe and the Luchaze, also dolichocephalic, with indices 75.6 and 75.1, respec- tively. These three tribes constitute a dolichocephalic group. The remaining tribes, Vachokwe and Luena, have cephalic indices that are almost identical, 77.1 and 77.0 respectively; both are in the range of mesaticephaly. The frequency distribution of cephalic indices (Table V) supports the inferences drawn from a study of averages. In the long-headed class (index 69-75) there are 77.3 per cent of the Ovimbundu sample, but less than half that number of the broader-headed Vachokwe and Luena. On the contrary, when we come to consider the tribal dis- tribution in the mesaticephalic class interval 76-80 we find that the number of Ovimbundu in that class is only 22.6 per cent, while nearly three times as many (about 60 per cent) of the Vachokwe and Luena are in that class. Again, in the class of broadest heads (index 81-85) there are no Ovimbundu whatsoever, but there are 10.9 per cent of Luena tribesmen, 8.9 per cent of Vachokwe, and 8.5 per cent of Luimbe. Every tribe except the Ovimbundu has representatives in the round- headed class. TABLE V PERCENTAGE FREQUENCY DISTRIBUTION OF CEPHALIC INDICES TRIBES CLAss INTERVAL cr ‘ Ovimbundu Vachokwe Luimbe Luena Luchaze BURA 10 cit cucu ee 77.3 30.0 50.0 29.7 54.3 90-0 8 Oe 22.6 61.1 41.5 59.4 39.1 ot | a a Rare oF ate pa eae 8.9 8.5 10.9 6.5 NASAL INDICES (Tables III, VI) The list of average nasal indices for the five tribes (Table III) clearly indicates that our sample of Ovimbundu is distinctly more mesorrhine than any other tribal group. The average nasal indices for four tribes have a very small range, namely, from 96.6 for the 54 ANTHROPOMETRY OF THE OVIMBUNDU Vachokwe to 98.6 for the Luimbe. The nasal index for the Luchaze (98.5) is almost identical with that of the Luimbe, and the index for the Luena (97.4) is very close. So we have a homogeneous group of high nasal indices (96.6—98.6) for eastern Angolan tribes, but for the central Ovimbundu an index of only 87.9, about 10 points lower. The frequency distribution (Table VI) brings out still more clearly the comparative narrowness of the nose in the Ovimbundu sample. In the mesorrhine class (N.I. 70-85) there are 35.8 per cent of the Ovimbundu sample, that is, five times as many as the Vachokwe, and seventeen times as many as the Luchaze, which have only 2.2 per cent of their sample in the mesorrhine class. When we turn to the hyperplatyrrhine class interval (index 100+) we find only 5.7 per cent of the Ovimbundu sample present; but 50.0 per cent of the Luimbe, 53.3 per cent of the Luchaze, and similarly large numbers of the Vachokwe and the Luena are in the hyperplatyrrhine class. TABLE VI PERCENTAGE FREQUENCY DISTRIBUTION OF NASAL INDICES OF ANGOLAN TRIBES TRIBES CLass INTERVAL y, A— \ Ovimbundu Vachokwe Luimbe Luena Luchaze Tees than 70.05.2346 S24). ‘ney abet sabi alts Janus LOBED aisle din tex 3" 35.8 6.7 4.9 6.0 2.2 $5-99 9. yo e 58.5 55.5 45.1 51.5 44.4 SOO 2 Sis ata of 0-0 Sa eles 5.7 37.8 50.0 42.5 53.3 CONCLUSION In stature the Ovimbundu are not greatly removed from tribes- men of eastern Angola. But, on the contrary, distinct differences of undoubted significance are seen in the cephalic and nasal indices. The Ovimbundu are decidedly long-headed and mesorrhine, when compared with their eastern neighbors. The explanation of these differences may be geographical, his- torical, or biological. The central highlands, which are from 3,000 to 5,000 feet in elevation, have PEIRCE le lower than those in surrounding country. Therefore the nasal index of the Ovimbundu inhabitants may be lower on this account (A. Thomson and L. H. D. Buxton, 1923). Moreover, isolation, leading to fixation of type, for both nose and head form, may be in some measure responsible for the distinctive characters of the Ovimbundu. From tribal traditions, and from Portuguese sources, we know that the Ovimbundu have built up a strong isolated confederacy, which was at war with surrounding tribes for several centuries. The ANGOLAN TRIBES COMPARED 55 Vachokwe, on the contrary, are known to have migrated from the southwest Congo region. The historical evidence, together with the topographical con- ditions, points to close association of the four tribes—Vachokwe, Luimbe, Luena, and Luchaze—whose mesaticephaly and platyrrhiny are widely distributed traits of the Congo region. We have, there- fore, in our samples: (1) A relatively round-headed, decidedly platyrrhine Congo group. This group consists of four tribes living close together in country offering no barriers to miscegenation. (2) A central, highland group of Ovimbundu, established earlier than the eastern tribes. These people have head form and a nasal index distinct from the eastern group of tribes. IV. COMPARISON OF THE OVIMBUNDU WITH NEGRO TRIBES OUTSIDE ANGOLA WEST AFRICA (Tables VII, VIII) The main samples for comparison are those of 100 west African Negroes measured by Weninger (1927), 100 Hausa measured by Tremearne (1911), and samples given by Herskovits (1937). Wen- inger’s sample, though derived from several tribes, may be taken as representative of Negroes of Sudanic speech in the far west. The Hausa are a linguistic division of Negro affinities, but with some somatic modification. Their language is classed as Hamitic with Sudanic Negro elements. Compared with west African Negroes the Hausa show a refinement of features, including reduction in thickness of lips and width of nose. Tremearne does not describe his technique, but he probably followed that of D. J. Cunningham (1909). Weninger adopted the technique of R. Martin (1928); so also did Herskovits, with certain modifications which he carefully describes (19380, pp. 19-89). I would say that the methods followed make comparisons permissible; but, even if objections are raised, we have no alternative data col- lected by the same observer using absolutely constant technique. In stature, the Ovimbundu (1687 mm.) are almost identical with Weninger’s sample, for which the average height is 1688 mm., a difference of only one millimeter. Tremearne’s Hausa have an average height of 1684 mm., which differs very little from that of the Ovimbundu and Weninger’s west African Negroes. The Ovimbundu are 45 mm. taller than the 48 Ashanti Negroes measured by Rattray (1928, p. 335). Cephalic Index.—The Ovimbundu (index 73.1) are a little more dolichocephalic than Weninger’s west African Negroes, whose index is 74.6, and a trifle higher still is the Hausa index of 75.4. The range for three tribes under comparison is therefore small, namely, 73.1-75.4. But Rattray’s Ashanti have an index of 77.5 and are therefore distinctly more round-headed than the Ovimbundu. Nasal Index.—The Ovimbundu (index 87.9) are definitely less platyrrhine than Weninger’s west African Negroes, whose average N.I. is 92.9. With regard to the frequency distribution of nasal indices, only 5.7 per cent of the Ovimbundu are hyperplatyrrhine with an index of more than 100, but 27 per cent of Weninger’s west 56 AFRICAN NEGRO TRIBES COMPARED 57 African Negro sample is hyperplatyrrhine. The Hausa, also, with an average N.I. of 90.0, are slightly more platyrrhine than the Ovim- bundu. The nasal index of Rattray’s Ashanti is 95. 2, that is, 7.3 points higher than that of the Ovimbundu. STATISTICAL COMPARISONS OF OVIMBUNDU, ASHANTI, AND DAHOMEANS More precision can be given to comparative methods by determin- ing whether the difference between the means is greater than three times the square root of the sum of the squares of the probable errors of the two averages: M:-M: > 3¥ (PEi)?+(PE2)? Only in this case is the difference between the two means significant; otherwise, the difference might have arisen from random sampling. TABLE VII AVERAGE MEASUREMENTS OF THE OVIMBUNDU COMPARED WITH THOSE OF DAHOMEANS, ASHANTI, NIGERIANS, AND HERSKOVITS’ “TOTAL WEST AFRICAN SERIES” Traits oe Ovimbundu Ovimoundu a Pie amg 4 omeans an igerlans est rican series OU esgic ate, Geet ads +1.4 +34.2* +15.9 +11.7 eres sn ee PRney were +e . +4.9 wae et —6. ath —4, Head breadth.......... —9.7* —12.4* as —10.9* Cephalic index......... —2.7* —5.7* ee —4.0* Upper facial height..... +1.3 +3.2* ens +0.9 Total facial height..... —4.6* —4.4* SER. —4.5* Bizygomatic........... —18.9* —16.3* ote es —17.7* Height of nose......... —1.8* —3.9* Spee —2.8* Width of nose......... —1.2* —2.0* —0.8 —1.3* Height of ear.......... —4.6* —3.1* Soe —3.9* Width of ear.......... —4.7* —5.1* pao —4.9* A=difference * averages. *=differences found to be significant. All phe in millimeters. Sign + or — to be read with the word Ovimbundu in front of it Table VII gives average measurements of 53 Ovimbundu com- pared with averages of corresponding traits of 77 Ashanti and of 93 Dahomeans measured by Herskovits (1937). The significant differences are marked with an asterisk. R. A. Fisher (1932, p. 45) recommends the use of the standard error instead of the probable error. Students wishing to follow this system will divide the probable errors by 0.67449 to convert them to standard errors. Ovimbundu and Dahomeans.—Nine of the twelve traits compared show a significant difference. The very small difference of 1.4 mm. in stature indicates that the Ovimbundu are almost exactly the same 58 ANTHROPOMETRY OF THE OVIMBUNDU average height as the Dahomeans. The differences in sitting height and in upper facial height are possibly real differences, though they are not proved to be so by the statistical methods we have used. The head form of the Dahomeans is different from that of the Ovimbundu in three respects, namely, length, breadth, and cephalic index. The heads of the Ovimbundu are shorter, narrower (abso- lutely), and definitely more dolichocephalic than the heads of the Dahomeans. The difference in bizygomatic breadth is significant; so also is the difference in total facial height, the dimensions of the nose, and the size of the ears. The Ovimbundu have narrower, smaller faces than the Dahomeans, their noses are narrower, and their ears are shorter and narrower. Ovimbundu and Ashanti.—Ten out of twelve traits compared are significantly different. The Ovimbundu are very definitely taller than the Ashanti, and probably they have a greater sitting height, though the difference of 12.6 mm. is not demonstrably significant. Neither can we be sure that there is any real difference in head length, but we can be satisfied that the heads of the Ovimbundu are narrower (absolutely) and more dolichocephalic than those of the Ashanti. The Ovimbundu have a C.I. of 78.1, which is 5.7 units less than that of the Ashanti. Herskovits (1937, p. 495) states that the Ashanti shape the skulls of infants to shorten them. This statement seems plausible because the Ashanti are round-headed, but it conflicts with the evidence given by Dingwall (1931, p. 113), whose collated references all men- tion elongation of the Ashanti head by artificial means. The facial heights, both upper and total, are different in the Ovimbundu, who have also a significantly smaller bizygomatic width than the Ashanti. The Ovimbundu have ears that are defi- nitely smaller than those of the Ashanti. In fact, the physiognomy of the two tribes is undoubtedly different. ~ Ovimbundu and West Africans.—Herskovits compounds his data for Dahomeans, Ashanti, and Nigerians into a “total west African series.’ In view of the significant differences, for example, in the head form of the Ashanti and Dahomeans, and because of the small number of groups, the compounding of traits may not be statistically permissible. There are few data for the Nigerians, and the ‘“‘total west African series’ results mainly from the averaging of measure- ments from only two units. The method is rather like averaging AFRICAN NEGRO TRIBES COMPARED a:) the stature of two brothers, one of whom is 72 inches tall, the other 24 inches. The average height of the boys in the family is 48 inches. It is desirable, of course, to have a general statistical basis for African Negroes, prepared from study of many groups whose average measurements are finely graded, as they might be if enough groups were measured. But, owing to paucity of data, perhaps the com- pounding of the few figures we have is permissible. In defense of averaging the data we may say that the three groups—Dahomeans, Ashanti, and Nigerians—are topographically close. Differences between the average measurements of the Ovimbundu and the “total west African series’ are given in the right-hand column of Table VII. In nine out of twelve instances of comparison the traits of the Ovimbundu are shown to be significantly different from corresponding traits of the west Africans. Possibly the Ovim- bundu are taller than the west Africans, and the former may have the greater sitting height and upper facial height, though the plus values for the Ovimbundu are not demonstrably greater. Summary.—In assessing the differences between somatic traits of the Ovimbundu and west Africans our consideration need not be entirely dependent on a mathematical test, though it is true that out of 39 calculations which test the differences of means, 28 are shown to be differences not likely to have arisen from comparison of small samples. In other words, they are probably real somatic differences. But apart from such tests we may glance across the stature columns of Table VII and note that the Ovimbundu have always plus values in stature and sitting height. In every instance of com- parison, the Ovimbundu have the shorter, the narrower, and the more dolichocephalic heads. Upper facial height is always a plus value for the Ovimbundu, but the total facial height is always a minus value. These differences clearly indicate a difference in facial proportions, and this is further attested by the comparatively small bizygomatic width of the Ovim- bundu. We have to admit, however, that measurements of facial heights, owing to poor definition of the points of measurement, are particularly liable to discrepancies when several different persons make the measurements. A glance across the column giving dimensions of noses and ears shows these features to be smallest in the Ovimbundu. Statistical data confirm photographic evidence in showing the Ovimbundu as a Negro type, but with modifications that distinguish them definitely from typical Negroes of far west Africa. 60 ANTHROPOMETRY OF THE OVIMBUNDU COMPARATIVE VARIABILITY OF OVIMBUNDU TRAITS AND THOSE OF OTHER WEST AFRICAN NEGROES Table VIII has been prepared to facilitate comparison of varia- bilities. L. H. D. Buxton’s figures in Rattray’s ‘‘Ashanti’” (1923) give coefficients of variation for traits of a small sample of 48 Ashanti. Herskovits (1937) gives standard deviations for measurements of 77 Ashanti and 98 Dahomeans. From the standard deviations, coefficients of variation have been worked out from the formula: vy 1008 ~ M The standard error of V is V V2N which is converted to the probable error by multiplying by 0.674489. The table should be consulted by glancing across from left to right. Evidently the variabilities of statures for the Ovimbundu, two groups of Ashanti independently measured, and the Dahomeans, are very close, all being within the zone V=3.37 to 4.29. Sitting heights are a little more variable than statures but the coefficients are by no means high. The head form in each tribal group has a low variability, and the variabilities are about the same for the four tribal groups considered. The dimensions of the head, likewise the cephalic indices, seem to be well-stabilized traits. TABLE VIII COEFFICIENTS OF VARIATION AND THEIR PROBABLE ERRORS, OF OVIMBUNDU TRAITS COMPARED WITH THOSE OF ASHANTI AND DAHOMEANS 48 Ashanti 77 Ashanti 93 Dahomeans Traits 53 Ovimbundu (Rattray) (Herskovits) (Herskovits) SSEAGLINO yea Fecee tree biecs sisal 4.29+0.28 3.37+0.23 3.6740.20 3.80+0.19 Sittin’ heights. ances. 5 ok 5.44+0.36 ye: 4.02+0.22 4.14+0.20 fi6ad length... eh Ses 3.87+40.25 2.49+0.17 *3.02+40.16 3.61+0.17 Head breadth............ 3.2740.21 3.04+40.21 3.386+0.11 3.25+0.16 Minimum frontal diameter. 5.00+0.33 3.78+0.26 Paes oie Bizygomatie elo... 5.60+0.387 3.18+0.22 4.06+0.22 3.40+0.17 Bigonial breadth......... 8.26+0.54 5.16+0.36 ciatece pease Upper facial height....... 7.18+0.47 7.47+40.52 6.4440.35 6.21+0.30 Total facial height....... 5.18+0.84 5.42+0.41 4.82+0.26 5.50+40.27 Nasal:heighten): sac. Gas. 7.29+0.48 5.49+0.88 7.62+0.41 7.380+0.36 Nasal breadth........... 6.39+0.45 7.54+0.52 7.16+0.39 7.42+0.36 Har: breadth .%.. kick 8.59-+0.56 bya! 7.87+40.40 6.92+0.34 Wardengtys 2. aces acd: 7.09+0.46 ey 7.03+0.38 7.68+0.38 Cephalic index........... 8.45+0.23 3.18+40.22 3.18+40.17 4.34+0.21 NaSah OS eM ateroh nis 9.15+0.60 9.87+0.69 eae ee Upper facial index........ 9.538+40.62 7.03+0.49 Total facial index. ....... 7.07+0.46 5.07+0.35 AFRICAN NEGRO TRIBES COMPARED 61 The highest variabilities are shown in the nasal index and the upper facial index; for the former the V value is in the 9+ zone. On the whole, the conclusion must be that despite real differences in absolute measurements, the variabilities of the traits for different tribal groups are remarkably close, trait for trait. We are dealing with distinct somatic types and tribal differences. But the groups are equally well stabilized, and in each tribal group head form and stature seem to be the physical features which are most strongly entrenched and least variable. THE CONGO REGION Stature—Figures collated by Montandon (1928, pp. 248, 264, 276) and by Hambly (1937, p. 173) show that most of the Congo tribes are of medium stature, that is, somewhat shorter than the Ovimbundu, who are upper medium to tall. Yet the Ovimbundu (1687 mm.) are shorter than the Bushongo (1747 mm.) and also shorter than the Azande (1701 mm.). Cephalic Index.—The data for head form (Struck, 1922) are more convincing. Negroes of the Congo are definitely in the higher ranges of mesaticephaly with indices from 77 to 80, whereas the Ovimbundu are decidedly dolichocephalic (index 73.1). Nasal Index.—From the same collations of statistics, though these are not as reliable as one might wish, one may make some compari- sons with tribes in the Congo region. Montandon (1928, p. 248) gives 47 averages of nasal indices for different tribes. The averages do not carry equal weight, for they are not based on equal numbers of observations. In some samples there is paucity of data; neither can we be sure of the technique of the many observers whose results are collated by Montandon. Nevertheless, the frequency distribution of the 47 average indices shows that 16 of them have a value from 90-105. That is to say, one-third of the samples from the Congo are definitely platyrrhine and hyperplatyrrhine. About one-third of the Congo samples fall in the class interval 85-90, and to this value the N.I. 87.9 of the Ovimbundu conforms. Montandon’s collation shows 10 of the 47 average indices to lie in the range 80-85, which is somewhat more mesorrhine than the value for the Ovimbundu. Apparently, when the nasal index for the Ovimbundu is compared with a series of 47 average nasal indices for Congo tribes the N.I. of the Ovimbundu occupies an intermediate position. One-third of the Congo samples are more platyrrhine than the Ovimbundu, one- 62 ANTHROPOMETRY OF THE OVIMBUNDU third have about the same degree of platyrrhiny, and one-third are less platyrrhine than the Ovimbundu. We have, however, only one definite and reliable fact emerging from comparison of the Ovimbundu with the Congo tribes; namely, the dolichocephaly of the Ovimbundu and the high mesaticephaly of most Congo tribes. Undoubtedly the Congo region was formerly more widely and more densely populated with Pygmy tribes than at present. Miscegenation of Negroes and Pygmies takes place today, and this process, continued over many generations, probably accounts for the low-medium stature and the comparatively round- headed form of many Negro tribes of the Congo region. SouTH AFRICA Stature—Dr. C. G. Seligman (1930, p. 189), quoting from a manuscript by Turner in the Royal Anthropological Institute, London, states that a miscellaneous group of 4,098 Negro mine laborers of adult age had an average stature of 1670 mm. This is 17 mm. less than the average stature of our sample of Ovimbundu. Stayt (1931, pp. 11, 368-871) shows the average stature of 168 Bavenda males to be 1676 mm., which is 11 mm. shorter than the average for the Ovimbundu. But the 39 Zulu (1696 mm.) and the 23 Batonga (1712 mm.), measured by Cipriani (1930-31), are both appreciably taller than the Ovimbundu. Cephalic Index.—Consulting the same sources, we find that the C.I. for 168 Bavenda is 75.2; that is, two points higher than the head index for the Ovimbundu. The index for the Zulu is 75.4, again just a little higher than that of the Ovimbundu. The 23 Batonga have a C.I. of 72.3. They are just a trifle more dolichocephalic than the Ovimbundu. Yet, on the whole, the C.I. of 73.1 for the Ovimbundu makes a good fit with the indices for the Bavenda, the Batonga, and the Zulu. Nasal Index.—All the average nasal indieés given by Stayt, and by Cipriani, are higher than the index of 87.9 for the Ovimbundu. The Bavenda have an index of 92.3, the Zulu 92.0, and the Batonga 90.9. The Ovimbundu are less platyrrhine than these samples of southeastern Negroes. East AFRICA Stature—Samples of measurements from east Africa are numer- ous (Hambly, 1937, p. 178) but open to the objections previously stated. Roscoe (1911, p. 520) gives the average height of 288 Baganda as 1673:mm. That is 14 mm. shorter than the average for AFRICAN NEGRO TRIBES COMPARED 63 the Ovimbundu. Scanning down a list of statures of 16 tribes (Hambly’s compilation) shows the Ovimbundu to have the highest average. The 101 Wanyamwezi of Leys and Joyce (1913) have an average stature of 1675 mm., almost the same as that of the Baganda, but 12 mm. shorter than the average for the Ovimbundu. Cephalic Index.—The C.I. of the Ovimbundu (73.1) is almost identical with that of 288 Baganda, who have an average index of 73.4. The index for the Ovimbundu agrees well with the samples of average indices for east African Negroes. These samples have a small range of averages, the extremes being 72.6 to 77.6, and the modal value of the averages is about 74.0. Nasal Index.—Roscoe’s 288 Baganda have an average N.I. of 85.4, slightly less platyrrhine than the index of 87.9 for the Ovim- bundu. The 128 Akamba measured by Leys and Joyce (1913) have an index of 86.5, which is between that of the Baganda and the Ovimbundu, but all three indices are in a very small range. Hambly’s “Source Book’ (1937, p. 178) shows that the average nasal indices for tribes in lower east Africa range from 89.4 to 101.1. All the samples, eleven in number, are appreciably more platyrrhine than the Ovimbundu. CONCLUSION In stature the Ovimbundu may be matched by many Negro tribes. The cephalic index of 73.1 is of a dolichocephalic type that is common, except in the central Congo and part of Cameroons, where high mesaticephaly prevails. The Nilotic Negroes are definitely taller than the Ovimbundu, more dolichocephalic, with an index about 70-72, and decidedly more platyrrhine. The Ovimbundu agree well in stature with tribes of eastern Angola but are definitely less platyrrhine and more dolichocephalic. The most interesting feature recorded for the Ovimbundu is the nasal index, which is in close agreement with the indices of adequate samples of the northeastern Bantu, namely, Baganda and Akamba. The series of photographs (Plates I-XX) shows the slim build of the Ovimbundu as compared with typical west African Negroes. In comparative narrowness of nose, reduced width of face, reduced thickness of lips, modified prognathism, and comparatively light skin color the Ovimbundu must be regarded as a modification of the true Negro of the Kru and Ibo type. The general build and physiognomy of the Ovimbundu together with the nasal index suggests a contact, perhaps very remote in time, 64 ANTHROPOMETRY OF THE OVIMBUNDU with that northeast African element that is usually described as Hamitic. But perhaps the modified Negro characters of the Ovim- bundu may be due, not to interbreeding of types, but to isolation under definite environmental conditions. V. SIZE OF FAMILIES AMONG THE OVIMBUNDU (Table IX) DIFFICULTIES OF INQUIRY An effort to discover the age of each speaker, the age at which he was married, and the number of years married was a failure. Ngonga, the interpreter, also Dr. M. W. Ennis and Dr. Hollenbeck, made estimates for a few individuals who were well known to them, but the information obtainable was too uncertain to be used statis- tically. There did seem, however, to be some indication that a youth marries at the age of about twenty-two to twenty-five years. Subjects seemed willing to answer, but some had evidently forgotten the exact number of their brothers and sisters, since some of these had died in very early infancy. Aversion to speaking of the dead may have made our estimated death rate too low. The range of ages may be considered to be from twenty to forty-five years, and this point is of importance when considering the incidence of mortality. BROTHERS AND SISTERS OF THE INFORMANTS To carry out this inquiry satisfactorily the investigator would have to know not only the number of brothers and sisters, living and dead, but many more details of the family structure. Was the family polygynous? If so how many wives constituted the family? How many children of each sex did each wife bear? What was the survival among these children? Was one father responsible for the parentage of all the children enumerated by the speaker? A few inquiries showed that such a questionnaire was far too ambitious, and the only information obtained was that tabulated in Table IX, showing brothers living and dead, sisters surviving and dead, together with totals. These figures I believe to be sufficiently reliable for considera- tion, yet any inferences we draw are only approximations to the truth. For example, a subject of twenty-two years may have a father who is still young enough to beget children by existing wives or by newly acquired spouses. Therefore, by recording the present size of the family we may be underestimating. We must not forget, of course, to add the speaker himself to the number of brothers and sisters he enumerates. The difficulties of inquiry are such that the technique demanded in communities where written records are kept cannot be applied to the Ovimbundu. Brothers.—Fifty informants claim 83 living brothers, so making a total of 133 males, that is, 2.7 living males for each family. For 65 TABLE IX SIZE OF FAMILIES CHILDREN Girls Living | Dead | Living | Dead | Total Boys and girls Boys Living | Dead ASD FI ES AOD SOO BAO MO MOMAS SOM ASSN AOS S OA AE UCSB OFS rt FCS ND Oe Sw SS Fh EPS BIO CLS RON OF ONS. SOCCCOCOMMOOCONNMONOCHOnnAOCOCONCCCCOCOSCSOS SMA HAAANOOANHOSCOSCOCNRNMNOSCOCOMNHROnRANDMHS WOnMNOMMMOOCOMNOCCOCOCONnMooCoConooconNooo Reser HOSOMNA NNO COMOOOCOCONCMOMNHOMMS BROTHERS AND SISTERS Brothers and sisters | Dead | Total iving : L Sisters Brothers Living | Dead | Living | Dead DN Oe DH ANM PAM OAM MAM AH MMO SH He BD HHS = = WDD QV re SH HO HO LOR N OO HAIN He OD OD rH 1 1D 4 OD HILO ANS OPO OADM ROMO ONSMDOONW MAM OOo OW Os ’ MMASANS ANOS OOMONMMAHANANNMHOMHAMMeOONOnr MH HONGC CONNOOCHROCHOHOOCHONAN HO HOM rir ANSON ANAO HOOP HO COONHMOOCHNOOCS MMi! AAMONSCHATMHAOCOCCHONCOCOHMNHANHAANSG OMONS MARRIAGE Probable Years married age at marriage Hehe Oe Oe Oe roe ‘egies Sh 1 2 5 20 3 5 25 ? married unmarried not long married married unmarried = ( Age* No. “Ayjuey sod (peep pue Zuraqy) WeIpIy f'9 20 “sor[TUley OS OY} 10J UeIPTTY OTE JO 1810} B SaATZ sJUBUIOJUI QS ey} SNid (09%) Joquinu sI4y, *(p¥ep pus BUIAT]) S194SIS PUB SIOYIOIG 09Z WIE] SJUBUUIOJUT OG ‘3J2] OY} WOIJ USAES UUINIOO UF APES “Ajrurey Jed sayeul Burlay 1° 10 ‘satires OG ey} 10J SaTwUL BUIAT] EET JO Je}0} B ZuPyeUI os ‘sIaYy}OIq BuUIAT] Eg WIep S}UBULIOJUI /¢ ‘snyL “S}UBULIOJUI JO JequINU 94} SEATS SesayjUOIEd UT BINZYy ey, ij “epun’y Woly ot |b “Y{41q STY e1OJeq pelMss0 syyeep UEeYyA Aj[ePedSe ‘siIeJep JoquioWel OF B]qvUN JUBULIOJUT ‘AouByUl Alive UT GULOS ‘pasvooep S194SIS puB SI94I0I1q 10430 fF *r9MSUB 03 BIQuUy 4 -ayeurxoiddy x Le eV = Seo St... 21 sh, 202 Sa.2 feo te 61 OI Sh Se i eS ee’ "* sedBOAV (OTS) ($8) ($81) 8st 19 L6 93 &¢ SE vy 09 S2I S8I 39 og BO SOR ES er Ee ee s[®70,L (eh) (8%) (8h) (eh) (er) (8%) (€%) | (0S) (0g) (0¢) |](0g) 0 0 0 0 0 0 0 I 0 T 0 0 0 T é é é &¢ I I 0 T 0 0 0 ¢ & va § SG 0 0 é é é eg I 0 T 0 I 0 0 L & Vv 0 0 § Vv é é 0€ TS i I 0 T 0 0 0 I 0 T 0 0 0 T é é Sia og 0 0 0 0 0 0 0 8 0 8 0 v 0 r peyireuun 02 67 “€ a I ae I é 0 s if I 0 I v 0 é é OF 8h 4 I I 0 0 i i! ¥ $ & 0 I T j é é sv LY 4 T & T 3G 0 T 8 0 8 0 v 0 v é é é 9F 0 0 0 0 0 0 0 £ é § é 3G é i peyiewun é tor 0 0 0 0 0 0 0 9 & & & G 0 I pelieurun é vP ¢ 3S g T I I é 4 T T I 0 0 I é é é &P 0 0 0 0 0 0 0 6 9 & g j § T é é vz a4 I 0 T 0 0 0 I $ G & 0 0 G g é é SZ Iv ¥ I & 0 3 IT i 9 v 3S 3 0 j v4 é é Sv OF 9 3 v 3 T 0 § é é é é é é é é é é t68 £ & 0 3 0 I 0 4 3 0 3G 0 0 0 é é og 8& ¢ 0 g 0 v 0 I Ss & SG § 0 0 3% é é SZ Lg 0 0 0 0 0 0 0 L v & I I § 4 pelieurun 86 9& 0 0 0 0 0 0 0 ¢ Sj € 3G T 0 3% é é G3 gE 0 0 0 0 0 0 0 £ é & é 0 é & peliewun 9t | ins (4 0 3S 0 I 0 I 9 I $ i § 0 é é é SZ && Toy, | peed | Suary | peoq | Suary| peoq | Burry | twos, | peed | Surry | peed | SurarT | peed | Surry | popzeu alane s]113 pus shog S11 shog sioqsis pue sisy,O1g: 81948IS si0yjO1g S180 X alqeqorg 423y “ON N@UGTIHD SUMLSIG GNV SUAHLOUG aOVIaUV A SUI'TINV,] 40 AZIS panuyuoyj—XT ATAaV.L 67 68 ANTHROPOMETRY OF THE OVIMBUNDU each of these families there are 1.3 deceased males, so making a total male progeny of 4.0 per family. Since the male issue is 133 living, and 63 dead, the total male progeny is 196, which gives a death rate of 32.1 per cent for males. Sisters.—The group of 50 informants has 52 living sisters, which gives an average of 1.0 living girls per family, and since 62 girls have died there is a death rate of 1.2 girls per family. Out of a total of 114 girls born, 62 died, so giving a death rate of 54.4 per cent as compared with a death rate of 32.1 per cent for boys. More boys than girls are born, according to these family records, and the males survive better. Death rates should, of course, be given in age groups. All we can say is that our informants varied from 20 to about 45 years, and that the deaths recorded are those of individuals from infancy to about 45 years of age. We must remember that our vital statistics are not dealing with mortality in infancy or childhood only. In fact, an informant usually forgets the infantile mortality, but he remembers his contem- poraries, and his figures certainly include those of his sisters who died in childbirth. Probably deaths in childbirth account for the higher mortality among females. Brothers and Sisters——Our informants represent a total family strength of 310 children living and dead, that is, about 6.2 children per family. The word family is used in the restricted sense to mean the father, his wife or wives, and their children. In polygynous families each wife has her own hut in the family compound, and in this hut she resides with her children. There she keeps house and is visited by her husband in a four-night cycle. We have, there- fore, a picture of the father and mother (or mothers) with an average of 6.2 children to the family compound. This is, of course, a high average compared with European standards. Of these 310 children 185 survived and 125 died, which gives a high death rate of 40.3 per cent. Considering that we have not included infants who died very young, and whose numbers are not remembered by the informants, the mortality is probably higher than our data suggest. CHILDREN OF INFORMANTS Examination of the section of Table IX dealing with children of the informants should provide a check on the conclusions derived from that part of the same table which gives statistics for brothers and sisters. SIZE OF FAMILIES 69 We are dealing now with the families of 43 married men, but the total number of children per family will not be strictly comparable with the family strength as revealed by an inquiry about brothers and sisters. Many of our married subjects are in the early twenties, and we cannot forecast the number of their progeny. Some inform- ants have been married only a year or two. On the contrary, when our informants speak of their brothers and sisters, they are giving a fairly complete and reliable idea of family strength, for each informant is of adult age. This explanation perhaps accounts for the fact that the family, judged from the number of brothers and sisters, was 6.2, and only 3.7 as judged from the number of children of the informants. Male Children Surviving.—The informants have a male progeny of 79, of whom 35 have died. The death rate for boys is therefore 44.3 per cent, or somewhat higher than the 32.1 per cent calculated from information respecting births and deaths of brothers. Female Children Surviving.—The informants have a female issue of 79, which is exactly the same as the male progeny. Of the 79 females born, 26 died, so giving a mortality rate of 32.9 per cent. This is appreciably less than the 54.4 per cent death rate calculated from information respecting brothers and sisters. Male and Female Children Surviving—The total number of children begotten by our 43 informants is 158, of whom 61 have died, so giving a death rate of 38.6 per cent. This is a close approxi- mation to the 40.3 per cent death rate arrived at by studying deaths of brothers and sisters. CONCLUSION The only permissible inferences are: (1) That, judged by European standards, the families are large. (2) That, though the birth rate is high, this is balanced by a high death rate. The death rate for the brothers and sisters of the informants was 40.3 per cent of all births, and for children of the informants the general death rate was 38.6. These figures are in close agreement, and since the data are derived from two independent methods of inquiry I think we may be fairly confident that the death rate is about 40 per cent. The difference of —1.7 per cent in the death rate of children as compared with those of the older generation (brothers and sisters columns, Table IX) may be the result of an improvement in medical attention following the erection of government hospitals and missionary medical stations. The 70 ANTHROPOMETRY OF THE OVIMBUNDU difference is so small, however, that it might arise fortuitously from the crude methods of our census. (8) Data relating to the children of our informants indicate that the birth rates for boys and girls are exactly equal. Our 43 informants sired 79 boys and 79 girls. But the inquiry respecting brothers and sisters indicates a definite preponderance of male births. The figures indicate a change in sex ratio at birth, from one generation to the next. (4) Evidence respecting the relative survival of boys and girls is contradictory. The brothers and sisters inquiry shows the death rate of girls as 54.4 per cent, and that for boys as 32.1 per cent. On the contrary, among children born to our informants the death rate for boys is 44.3 per cent, and that for girls is 32.9 per cent. The death rates for both sexes together are, however, very close, namely, 40.3 per cent for the brothers and sisters generation, and 38.6 for the children of our informants. (5) Students wishing to make further study of sex ratios and death rates of males and females at all ages will find a bibliography of sources in Hambly’s compilation (1937, pp. 691-694). L. W. G. Malcolm (1924, pp. 454-478) adduces some evidence to show that masculinity can be higher at birth and in early years among Negroes, for example, the Ibo and the Edo. We have, however, to beware of errors that might arise from a discriminating infanticide that favored the preservation of boys. And when dealing with adults there is a discriminatory emigration to be reckoned with. The subject of census returns is full of pitfalls. P. Ryckmans (1938, pp. 242-258) adduces some evidence to show a preponderance of males at birth in one of the Congo regions, but the number of females is usually in excess of the number of males at birth. Except in one province, the females of the first year sur- vive better than the males. Ryckmans points out that boys of fifteen years of age are more numerous than girls in the provinces studied. But, in the absence of written records and accurate time-keeping, the age of fifteen has to be judged by the investigator. There is a definite tendency to bias the results by classing well-developed girls as adult women even though they may be only fifteen years of age or under. The same error is not so likely to be made when judging the age of boys. The tendency, therefore, is to get too many boys, and too few girls in the fifteen-year age group. The sex ratio data collated for the Ovimbundu are not incom- patible with some of the statistics recorded for other Negro tribes, SIZE OF FAMILIES 71 but the technique and results are all too uncertain to allow of clear and definite census statements. (6) The undoubtedly high death rate of 40 per cent demands further attention. Evidently there is urgent need for more care of mothers during pregnancy and delivery, for increased attention to infant welfare, and for extension of clinical treatment for people of all ages. VI. SCARIFICATION AND DENTAL MUTILATION INTRODUCTION Searification is practiced only to a slight extent among the Ovimbundu, the reasons for the mutilations being therapeutic and esthetic. The following notes give all the information which was available. The data indicate that therapeutic scarification consists of a series of small cuts made over the affected region; no coloring matter is introduced. On the contrary, esthetic scarification, which is adopted chiefly by women, consists of cutting or pricking designs with the point of a knife and rubbing burnt rubber into the cuts. For details of persons scarified and dentally mutilated see Table I. Of the men measured, only one in six had the typical dental mutila- tion of the Ovimbundu; namely, a V-shaped notch between the upper central incisors. Only one man in eleven was scarified, and the marking is not heavy. General observation suggests that mutilation of the teeth is less frequent among females than among males, and that the custom is decadent with both sexes. NOTES ON SCARIFICATION Plate XXI, Left: Subject No. 17.—The marks were made when the subject was ten or twelve years of age. A male operator passed a needle under the skin and cut along the direction of the needle. Plate XXI, Right: Subject No. 24.—A woman made the scarifica- tion when the boy was a child. Plate XXII, Left: Subject No. 40.—The marks were made by a male when the subject was a child. The point of a knife was used and burnt rubber was introduced into the cuts, so giving a dark blue appearance to the cicatrice. Plate XXII, Right: Subject No. 37.—The marks were made by a man when the subject was a child. The appearance of the scar indicates that burnt rubber was introduced into the cuts. Plate XXIII, Left-—The scar, into which burnt rubber was introduced, was made by a male operator when the subject (a male) was a child ten years of age. The scars were made for ornament. Plate XXIII, Right.—Information for right figure is the same as for left figure. The scar shown on the left figure is on the right cheek, the scar shown on the right figure is on the left cheek. The forehead design, which is ornamental, and the abdominal scarification to cure a painful spleen are shown in Plate XXX (right). 72 SCARIFICATION AND DENTAL MUTILATION 73 Plate XXIV.—Two male operators made these designs when the subject (a male) was a boy about thirteen years of age. Each operator made one design. The boy also had his teeth chipped when he was about eight years of age. This subject stated that only males perform tooth mutilation, but either a man or a woman may searify. The subject paid for this treatment by working half a day for each operator. Plate XXV: Subject No. 43.—The marks are ornamental, and they show a distinctly blue tinge under the skin. A man was paid for making the marks when the subject was a child. Plate XX VI: Subject No.32.—The blue cicatrices are ornamental. No further information was given. Plate XX VII, Left.—The subject was a woman who stated that the designs were made by her mother’s brother when she was a child. The cicatrices were faint and dark blue in color. Plate X X VII, Right.—Ornamental scars under the eye of a woman. Burnt rubber had been introduced. Plate XXVIII, Left.—The subject was a woman. The forehead design was made by cutting with a knife, and the other designs by pricking with needles obtained from a white man. The needles were made into a bundle. The operator was her mother’s brother, whom she described as omesali (expert). This woman stated that the marking of girls was at one time very common. The girls preferred different designs, but two or more girls could have the same design if they wished. PlateX X VIII, Right.—This lozenge design was made on a woman. A knife was used for making discontinuous cuts, but no coloring matter was rubbed in. Plate XXIX.—These designs were made when the subject was a young girl. Her mother’s brother was the operator. The marks are of bluish color, and the scar tissue is small. The scar on the left cheek was hardly visible in front view. Plate XXX, Left: Subject No. 13.—The cuts were made during a cupping (bleeding) operation to cure a pain in the chest. Plate XXX, Right.—The forehead design of this male is of the ornamental type, namely, a faint scar of bluish color. The abdominal searification resulted from cupping to cure pain. BIBLIOGRAPHY Abbreviations AnAn Anthropologischer Anzeiger JRAI Journal Royal Anthropological Institute MAG Mitteilungen der Anthropologischen Gesellschaft in Wien ZFE Zeitschrift fiir Ethnologie CarRpboso, F. and Corrfia, A. A. M. 1916. Antropologia angolense, Quiocos, Luimbes, Luenas, e Lutchazes. Arquivo oe phase e de Antropologia. Institut d’Anatomie, vol. 2, pp. 323-356. isbon. 1918. Ibid, vol. 4, pp. 283-321. CIPRIANI, L. 1930-31. Osservazioni antropometriche su indigeni asiatici e africani. Archivo per Antropologie e la Etnologia, vols. 60 and 61. Yemen Arabs, pp. 136- 163; Mozambico, pp. 166-188; Zulu and Batonga, pp. 189-286. CUNNINGHAM, D. J. (and others) 1909. Anthropometric investigation in the British Isles. Report of the Com- mittee. London. DENIKER, J. and LALoy, L. 1890. Les races exotiques a l’exposition universelle de 1889. L’Anthropologie, vol. 1, pp. 257-293. DINGWALL, E. J. 1931. Artificial cranial deformation. London. Ashanti, p. 1138. FISHER, R. A. 1932. Statistical methods for research workers. London. HAMBLY, W. D. 1934. The Ovimbundu of Angola. Field Mus. Nat. Hist., Anthr. Ser., vol. 21, No. 2. Chicago. 1937. Source Book for African Anthropology. Field Mus. Nat. Hist., Anthr. Ser., vol. 26. Chicago. Anthropometry of Negroes, Part I, pp. 168-170, 173, 176. HERSKOVITS, M. J. 1930. The Anthropometry of the American Negro. New York, Columbia University Press. Technique described, pp. 19-39. 1937. Physical types of west African Negroes. Human Biology, vol. 9, pp. 483-497. Hooks, B. G. E. — 1926. A third study of the English skull with special reference to the Farrington Street crania. Biometrika, vol. 18, pp. 1-55. ISSERLIS, L. « 1914. Formulae for the determination of the capacity of Negro skulls from external measurements. Biometrika, vol. 10, pp. 188-192. Kitson, E. ‘ 1931. A study of the Negro skull with special reference to the crania from Kenya Colony. Biometrika, vol. 23, pp. 271-314. Leys, N. M. and Joycn, T. A. 1913. Notes on a series of physical measurements from east Africa. JRAI, vol. 48, pp. 195-267. 74 BIBLIOGRAPHY 75 MALCOLM, L. W. G. 1924. Sex ratios in African people. AA, vol. 26, pp. 454-473. MARTIN, R. 1928. Lehrbuch der Anthropologie. 2nd ed., 3 vols. Jena. MONTANDON, G. 1928. L’ologenése humaine. Paris. Measurements of Negroes, pp. 58, 174, 230-240. PEARSON, K. 1904. On the measurements of internal capacity from cranial circumferences. Biometrika, vol. 3, pp. 366-397. RaTTrRAy, R. S. 1923. Ashanti. Oxford. Table of measurements, p. 335. Roscoe, J. 1911. The Baganda. London. Tables of measurements, pp. 493-522. RYCKMANS, P. 1933. Démographie congolaise. Africa, vol. 6, pp. 242-258. SELIGMAN, C. G. 1930. Races of Africa. London. StayT, H. A. 1931. The Bavenda. London. Tables of measurements, Appendix II, pp. 368-371. STRUCK, B. 1922. boner einer Karte des Kopfindex im mittleren Afrika. ZFE, vol. 54, pp. 51-113. THOMSON, A. and BuxToN, L. H. D. 1923. Man’s nasal index in relation to certain climatic conditions. JRAI, vol. 53, pp. 92-122. TILDESLEY, M. L. 1938. Measurements of head length and breadth. The systems of Broca, Topinard, Monaco agreement, British Association, Martin, compared. Man, No. 103. TREMEARNE, A. J. N. 1911. Tribal marks of Nigeria (and measurements on 100 Hausa). JRAI, vol. 41, pp. 162-171. WENINGER, J. 1927. Eine morphologisch-anthropologisch Studie durchgefuhrt an 100 west- afrikanischen Negern. MAG, Wien. See also R. Routil, AnAn, vol. 8, pp. 270-279, 1932, for a review of Weninger’s data. INDEX Acromion from sole, height of, Ovim- bundu, 30, 34-86, variation of, 39 Ages, of the Ovimbundu who were measured, 65 Akamba, stature compared with that of Ovimbundu, 63 American Negro, anthropometry of, 74 Angola, measurements of tribes, by Car- doso and Corréa, 31; see Ovimbundu Ashanti, measurements of, 75; com- pared with Ovimbundu, 57; Negroes measured, 56 Averages, comparison of significant dif- ferences of, for Ashanti, Dahomeans, Ovimbundu, 57; of measurements of Ovimbundu, 36 Azande, tall Congo Negroes, 61 Baganda, cephalic index, nasal index, and stature of, compared with Ovim- bundu, 638; measurements of, 75 Bavenda, cephalic index, nasal index, and stature of, compared with Ovim- bundu, 62; measurements of, 75 Bibliography, 74, 75 Bigonial, -bizygomatic index, Ovim- bundu, variation of, 39; breadth, coef- ficients of variation for Ovimbundu, Ashanti, and Dahomeans compared, 60; of Ovimbundu, 34-36, 37, varia- tion of, 39 Birth rate, 69 Bizygomatic, coefficients of variation for Ovimbundu, Ashanti, and Daho- means compared, 60; of Ovimbundu, 34-36, compared with Dahomeans, Ashanti, and other W. Africans, 57; variation of, 39; width of, for Ovim- bundu, 37 Bleeding, scars left by, 73 Brachycephaly, see Cephalic indices Brothers, of Ovimbundu men who were measured, 65, Table IX, 66, 67 Ben. ear lobes of, joined to cheeks, Bushongo, tallest of Congo Negroes, 61 Callipers, types used, 30 Census returns, difficulties of, 70 Cephalic indices, 40; of Angolan tribes compared, Table V, 538; coefficients of variation for Ovimbundu, Ashanti, and Dahomeans compared, 60; of 76 Negroes of Congo compared with Ovimbundu, 61; of Ovimbundu, 33, compared with Dahomeans, Ashanti, and other W. Africans, 57, variation of, 39; study of, by Struck, 75 Chest girth, of Ovimbundu, 33, 34-36, variation of, 39; -sitting-height index, 41, Ovimbundu, variation of, 39; technique followed in measuring, 30 Children of Ovimbundu, survival of, 69, Table IX, 66-67 Climate, and nasal index, 54 Coefficients of variation, of Ovimbundu, Table I, 36, 39, for indices of measure- ments, 41 Congo region, Ovimbundu compared with Negroes of, 61, 62; sex ratio of tribes in, 70; Vachokwe migrated from, 55 Cranial capacities, methods of estimat- ing, bibliographical references, 74, 75; various formulae for estimating, 37 Cranial deformation, 58, 74 Cupping, scars left by, 73 Curves, of frequency distribution, 42- 50 Dahomeans, measurements of, pared with Ovimbundu, 57 Death rates, 65-71 Dolichocephaly, see Cephalic indices, Head measurements com- Ear breadth, coefficients of variation of Ovimbundu, Ashanti, and Daho- means compared, 60; Ovimbundu, variation of, 39 Ear index, 40, 41; Ovimbundu, varia- tion of, 39 Ear length, coefficients of variation for Ovimbundu, Ashanti, and Daho- means compared, 60; Ovimbundu, variation of, 39 Ear lobes, of Ovimbundu joined to cheeks, 82; see remarks column, Table I, 34-86 Ear measurements, of Ovimbundu, 38, 89; technique followed, 29 East Africa, cephalic indices, height, nasal indices of Negroes, compared with those of Ovimbundu, 62, 63; physical measurements of Negroes in, 74 Edo, masculinity in, 70 ee a INDEX vf Elende, Ovimbundu measured at, 42 Entrenched traits, see Coefficients of variation, Variability Errors (probable) for coefficients of variation of various tribes, 60; for measurements of Ovimbundu and for indices of these measurements, 41; Table I, 36 Eyes, Ovimbundu, color of, 32 Facial height, of Ovimbundu, total and upper, 34-36, 38 Facial indices, 40 Facial measurements, of Ovimbundu, 37, 38, and W. African Negroes com- pared, 57; technique followed, 29 Families of Ovimbundu, compound and houses of, 68; size of, 65-71; size of, compared with European families, 69 Frequency distribution, and cephalic indices of Angolan tribes, 53; curves of, 42-50; and nasal indices of An- golan tribes, 54; and statures of Angolan tribes, Table IV, 52 Hair, Ovimbundu, color and texture of, 32, on face and body, see remarks column, Table I, 34-36 Hausa tribe, measurements compared with those of Ovimbundu, 56 Head breadth, coefficients of variation of Ovimbundu, Ashanti, and Daho- means compared, 60; of Ovimbundu, 34-36, compared with Dahomeans, Ashanti, and other W. Africans, 57; variation of, 39 Head height, Ovimbundu, 33, 34-36, variation of, 39 Head length, coefficients of variation for Ovimbundu, Ashanti, and Dahomeans compared, 60; of Ovimbundu, 33, 34-36, compared with Dahomeans, Ashanti, and other W. Africans, 57; variation of, 39 Head measurements, of Ovimbundu and W. African Negroes compared, 57; technique followed, 29; see Ce- phalic indices Height, of Angolan tribes compared, Table III, 52; coefficients of variation of Ovimbundu, Ashanti, and Daho- means, compared, 60; Negroes of Congo compared with Ovimbundu, 61; of Ovimbundu, 33, Table I, 34-36, compared with Dahomeans, Ashanti, and other W. Africans, 57; sitting-standing index, 41 Height of ear, of Ovimbundu, compared with Dahomeans, Ashanti, and other W. Africans, 57 Height of face, of Ovimbundu, 38 Height of nose, of Ovimbundu, com- pared with Dahomeans, Ashanti, and other W. Africans, 57 Ibo tribe, masculinity in, 70 Indices, Tables of, 40, 41; see under separate items, Cephalic indices, Nasal index, and similar terms. Individual measurements, Ovimbundu, Table of, 34-36 Instruments used, 30 Ivory, traded by Ovimbundu, 42 Jugofrontal index, Ovimbundu, 41 Jugomandibular index, 41 Keloids, see Scarification Kenya Colony, study of skulls from, 74 Knife, used in scarification, 72, 73 Kwanza River, tribes situated near, 51 Luchaze, habitat, number measured, 51 Luena, habitat, number measured, 51 Luimbe, habitat, number measured, 51 Marriage, Ovimbundu, probable age at, 65 Maternal uncle, scarifies his sisters’ daughters, 73 Medical attention, 69, 71 Mesaticephaly, see Cephalic indices Mine laborers, S. Africa, stature of, 62 Minimum frontal diameter, -bizygo- matic index, of Ovimbundu, varia- tion of, 39; coefficients of variation of Ovimbundu, Ashanti, and Daho- means compared, 60; -head-breadth index, of Ovimbundu, variation of, 39; Ovimbundu, 34-36, 37 Missionaries, medical, 69, 71 Nasal breadth, coefficients of variation for Ovimbundu, Ashanti, and Daho- means compared, 60; of Ovimbundu, 34-36, variation of, 39 Nasal height, coefficients of variation for Ovimbundu, Ashanti, and Daho- means compared, 60; Ovimbundu, 34-36, variation of, 39 Nasal index, of Angolan tribes com- pared, Table VI, 54; and climate, 54; coefficients of variation for Ovim- 78 ANTHROPOMETRY OF THE OVIMBUNDU bundu, Ashanti, and Dahomeans compared, 60; Negroes of Congo com- pared with Ovimbundu, 61; Ovim- bundu, measurements of, 38, varia- tion of, 39, and W. African Negroes compared, 56; theory of Thomson and Buxton, 75 Nasal measurements, lowed, 29 Needle, used in scarification, 72, 73 Nigeria, measurements of Hausa in, 75 technique fol- Ornament, personal, see Scarification Ovimbundu, comparison with other tribes of Angola, 51-55; families, size of, 65-71; further comparison with tribes outside Angola, 56-64; scari- fication and dental mutilation of, 72, 783; study of individual traits, 29-49; traits compared with those of other Negroes, summary of, 63 Pain, cured by scarification, 72, 73 Platyrrhiny, degrees of, Angolan tribes compared, 54; see Nasal index Polygynous family, 65 Portuguese, relations with Ovimbundu, 42, 54 Pygmy tribes of Congo, mixture with Negroes, 62 Rubber, burnt and Pubbed in sears, 72, 73 Searification, 72, 73 Sears, see Scarification er iia 65-71; studied by Malcolm, T \ Sisters, of Ovimbundu men who were measured, 65; Table IX, 66, 67 Sitting-height—chest-girth index, 41 Sitting height, coefficients of variation for Ovimbundu, Ashanti, and Daho- means compared, 60; of Ovimbundu, 33, 34-86, compared with Daho- means, Ashanti, and other W. Afri- ee 57, variation of, 39; technique ol, Sitting-height-standing-height variation of, 39 Skin color, of Ovimbundu, 32 Slaves, sold by Ovimbundu, 42 South Africa, cephalic indices, heights, and nasal indices of Negroes, com- pared with those of Ovimbundu, 62 Standard deviations, of indices for Ovimbundu, 41, of measurements of Ovimbundu, 36 index, Statistical method, bibliography, 74 Statures, of Angolan tribes compared, Table III, 52; see Height Sudanic Negroes, of W. Africa, meas- urements compared with Ovimbundu, Tables, I, individual measurements of Ovimbundu with averages, coeffi- cients of variation, standard devia- tions, 34-36; II, indices of these measurements, 40, 41; III, IV, V, VI, comparing cephalic indices, nasal indices, and statures of Ovimbundu with other tribes of Angola, 52-54; VII, comparing measurements of Ovimbundu with various W. African tribes, 57; VIII, coefficients of varia- tion for measurements of Ovimbundu, compared with Ashanti and Daho- means, 60; IX, size of families of -Ovimbundu, 66-67 Technique, of census returns, 65-70; difficult, 40, 41; particulars of, 29-31; recent discussion of, Tildesley, 75 Teeth, of Ovimbundu, mutilation and condition of, 32; see notes in remarks column, Table I, 34-36 Teita skulls, estimating capacity of, 37 Total facial height, -bizygomatic index, Ovimbundu, variation of, 39; coeffi- cients of variation of Ovimbundu, Ashanti, and Dahomeans compared, 60; Ovimbundu, compared with Da- homeans, Ashanti, and other W. Africans, 57, variation of, 39 Total facial index, coefficients of varia- tion of Ovimbundu, Ashanti, and Dahomeans compared, 60 Trunk measurements, technique fol- lowed, 30; of Ovimbundu, 338; see Chest girth, Sitting height. Upper facial height, -bizygomatic index, Ovimbundu, variation of, 39; coeffi- cients of variation for Ovimbundu, Ashanti, and Dahomeans compared, 60; Ovimbundu, 38, compared with Dahomeans, Ashanti, and other W. Africans, 57, variation of, 39 Upper facial index, coefficients of varia- tion of Ovimbundu, Ashanti, and Dahomeans compared, 60 Vachokwe, habitat, number measured, 51; see Tables III-VI Variability, coefficients of, for measure- ments on Ovimbundu, 36; comparison INDEX 79 of Ovimbundu traits, 39; of Ovim- Width of ear, Ovimbundu compared bundu and several W. African tribes with Dahomeans, Ashanti, and other compared, 57-60 W. Africans, 57 Width of nose, Ovimbundu compared Wanyamwezi, stature compared with with Dahomeans, Ashanti, and other that of Ovimbundu, 63 W. Africans, 57 West Africa, Negro tribes compared Zulu, stature, cephalic, and nasal index with Ovimbundu, 56-61 of, compared with Ovimbundu, 62 THE LIBRARY OF THE OCT 12 1938 UNIVERSITY OF ILLINOIS Field Museum of Natural History Anthropology, Vol. X XV, Plate I No. 2 No. 5 No. 2 No. 5 FRONT AND SIDE VIEWS, OVIMBUNDU MALES Field Museum of Natural History Anthropology, Vol. XXV, Plate II wi» No. 10 No. 12 No. 10 No. 12 FRONT AND SIDE VIEWS, OVIMBUNDU MALES (02 °°N) HIVW NANNAWIOO ‘SMAIA ACIS GNV LNOUd III ld ‘AXX “10A ‘ABojodoryyay £10987 [WINIWN JO MNES PIL (¢% ‘ON) WIVW NGANNAWIOO ‘SMAIA ACIS CNV LNOUA AI *3%Id ‘AXX ‘I0A ‘A80j0do14juy A104stH{ JBInjeN JO wunesnyy PleLy (L2 °ON) AIVW NANNAWIOO ‘SMAIA ACIS CNV LNOUA A id ‘AXX ‘1A ‘AB0jodoiyjuy A30481F] yeINjBN JO wneasny Pen ATVW NGNNAWIOO ‘SMAIA ACIS CNV LNOUA IA ld ‘AXX ‘9A ‘ABojodoiyjuy AJ04SIH{ [BINJeVN JO wNesnypy Pye1T (28 °ON) DIVW NAGNNAWIOO ‘SMAIA ACIS GNV LNOUA IIA *®Id ‘AXX “OA ‘A80jodorqjuy A104S1FJ YeINIVN JO WNasny, PPL WIVW NOGNNGWIOO ‘SMAIA ACIS CNV LNOWA TIIA ®%Iid ‘AXX “OA ‘AB0jodo1yquuy AiO}IFj [BINJVN jo unesnyy PjeLd XI ld ‘AXX ‘IA *AB0j0dorqquy FIVW NGNNAWIOO ‘SMAIA ACIS GNV LNOUA A10981H [BiN{VN JO UNesNW PPh VE “ON 2379T SHIVW NAGNNAWIAO ‘SMAIA LNOU X ld ‘AXX ‘1A ‘ABojodorqquy £10381 [eINIVN JO uMesn PPT 33 “ON $959'T SHTIVW NGNNOAWIAO ‘SMAIA LNOW IX ®¥id ‘AXX ‘lA ‘ABojodoiqjuy A20981F yeINZeN JO wNosny ppd 08 “ON :948ry SHIVW NGNNAWIAO ‘SMOIA LNOU ITX ld ‘AXX ‘OA ‘ABoodoiyqquy Al10jstp{ [eINjeN jo wMesny] Pel SHIVW NGNNAWIAO ‘SMAIA ACIS IITX ¥ld ‘AXX ‘19A ‘ABojodosqquy A104stH jenjen jo umMesny Pl SHIVW NAGNNAWIAO ‘SMAIA ACIS AIX Id ‘AXX ‘19A ‘A80jodorq}0V 410481H [einjeN jo unesnyy Pied SHIVW NGNOAWIAO ‘SMAIA ACIS AX 781d ‘AXX “OA ‘AB0jodorqjuy 410481H [einjeN jo unesny Plt Field Museum of Natural History Anthropology, Vol. XXV, Plate XVI No. 12 No. 10 No. 2 No. 5 BACK VIEWS, OVIMBUNDU MALES See also Plates I and II SHIVW NANNAWIAO ‘SMAIA HOVE IIAX ld ‘AXX “1A ‘A30jodorqjuy £10981 [BINIVN JO WNESNW PIPLA SHIVW NANNAWIAO ‘SMAIA MOVE IIIAX *3¥Id ‘AXX ‘10A ‘AS0jodoiyjuy Ai0xstp{ [eInyeN jo winesnyy Pent NOILVNUOAGC TVINGG XIX Id ‘AXX “19A ‘AB0jodorq3uy £10481] JBINIVN JO wasn PlerT Field Museum of Natural History Anthropology, Vol. XXV, Plate XX DENTAL DEFORMATION Upper right: no mutilation SHIVW NGNNAWIAO ‘NOLLVOIMIUVOS LT "ON IXX ld ‘AXX ‘1A ‘AZ0jodoiyjuy A10481H{ JBanjgeN jo UinesnW PIT SHIVW NGNONAWIAO ‘NOLLVOIAINVOS IIXX ¥id ‘AXX ‘1A ‘AZ0jodoryquy A10981F] JBINIEN JO wMsnyY PPA WIVW NAGNNAWIOO ‘NOILVOIAINVOS IIIXX 31d ‘AXX “1A ‘AS0jodoiqquy A1048ry [eINJeN JO wNESN PPT (28 °ON) HIVW NOAGNNAWIOO ‘NOILVOMIUVOS J i ta IAXX °®Id ‘AXX ‘0A ‘Ad0jodoiqquy A10481f [@IN}JVN JO uNesn PPT N@WOM NAONNAWIAO ‘NOLLVOIAIYVOS — , : 3s a, : ie \ sO a wail IIAXX ®¥id ‘AXX ‘I°A ‘AS0jodoiqjuy £10981H [BANIVN JO Nesn PPL NAWOM NGNNAWIAO ‘NOLLVOIMIYVOS ' ae peli ot” 3a Z J se f Oo“ wae [IIAXX %¥Id ‘AXX ‘OA ‘AB0jodoiqquy A10481H{ [@IN}BN JO unssny PPh NVWOM NONNAWIOO ‘NOLLVOIAINVOS XIXX %¥id ‘AXX ‘IA ‘ABojodoiqjuy A10481f{ YeINjVN JO unasny PPT ST “ON 239T SHIVW NGNNAWIAO ‘NOILVOIMIUVOS ALAA YI . er 73 . “4 Fi Pa eee ORL ly x 7 Pie ae . ELIT ET EES) Se ae c Lot ‘ SON St / FOOTSIE EEULAEEE \ : . Freee EER AN UNS 4444 Tn Mae Wire 1 hs XXX Id ‘AXX ‘OA ‘ABojodorqjuy ALOIS [BINVNY JO Nasn PIPLT 4 e Sak, y Ck ey ra 8 LIPS Gy IMGs ‘ f (OE GES at Pd pl } ft; i rie aie tees Tye NS eas ps gas 4 ark th etait : ae a