DERMATOGLYPHICS I N NEGROES OF WEST AFRICA HBROLD CUMMINS Department of Anatomy, Tulane University FOREWORD The following account forms a supplement to mTilder7s study of “Palm and sole prints of Liberian natives” ( ’13), which was based upon a series of prints collected by Frederick Starr in his expedition of 1912. Through the kindness of Professor Starr,l a duplicate series of these prints has been placed at the disposal of the present writer. Several considerations prompted the restudy of the collection, and it may be mentioned that Professor Wilder encouraged this extension of his original paper. First of all, it is desirable to have, for comparative purposes, a complete analysis of all the dermatoglyphic features accessible in the prints, namely, palmar main lines, palmar patterns, apical finger patterns, and the plantar configurations. Wilder’s account is concerned chiefly with the palmar dermatoglyphs, more especially the main lines, and but incidentally with apical finger patterns and plantar configurations. On account of deficiencies of the prints, Wilder was able to make complete and positive determinations of the main lines in but 101 of the 200 palms in the collection. It was expected that the list of complete formulae could be enlarged by the duplicates, since there is a chance that at least some of these might admit the study of features which were wanting o r illegible in ninety-nine palms of Wilder’s set. A serial number had been attached by Starr to the prints of each man; these numbers are employed by Wilder in his table of palmar ‘ F o r the privilege of the material I am iiidebted not oiily t o Professor Starr, but also t o Dr. Charles Midlo, through whose interest the collection came t o my hands. 9 AMERICAN JOURNAL O F PHYSICAL ANTHROPOLOGY, VOL. XIV, NO. JANUARY-MARCH, 1930 1 10 H h R O L D CUM&fINS formulae and in comments on individual prints, making it possible to identify the duplicates with his records. Wilder 's table 1carries, in addition to the 101 complete and dependable formulae, a number of tentative formulations of the doubtfiil prints. When my own determinations were checked against the formulae in Wilder's table, it was found that some of the questioned interpretations could be verified. extending the reliable determinations to a total of 133 palms. Lastly, an advantage is to be obtained by translating the palmar determinations t o terms of the revised system (Cummins et al., 'as), and with particular guard against errors which have been sliown to exist in palmar interpretation and formulation (Cummins et al., '28). The material is derived from one hundred soldiers, natives of Liberia and Sierra Leone. Since the individuals are all males, the element of possible sexual variation cannot be considered. The tribal composition is a matter of record (Wilder, '13, list on p. 191), though a division of the results according to tribes is not warranted, in view of the small size of the collection. Data regarding the occurrence of racial admixture are lacking. Whether there are any closely related individuals in the series is unknown. The prints are contact impressions, made by the common printers '-ink method. Each individual is represented by prints of tlie palms, finger tips, and soles, though many of the impressions, particularly those of the palms, are incomplete 01- only partially decipherable. Calculations of the incidences of the various features are thus based upon varying numbers of individuals, the total being indicated in each instance. -4 loupe magnifying about 2f diameters was employed in tracing the main lines and in other observations demandiiig magnification. This account does not aim to make exteiisire comparisons with the dermatoglyphic traits of other peoples. Its objectivc is t o make the findings available, in full tabulations, for the benefit of future comparath-e studies in which they may be of service. The literature cited prorides sources for detailed DERMATOGLYPHICS IN NEGKOES 11 comparisons, and only certain seemingly noteworthy characteristics are mentioned. METHODS O F FORMULATION The interpretation and formulation of palmar main lines and patterns follow the methods recently described by Cummins et al. ('29). These methods constitute a revision of Wilder's procedures (Wilder, '16, '22 ; Wilder and Wentworth, '18). I n comparing results obtained by the use of the revised and the original methods, allowance must be made for certain differences in the two systems, both in procedure and the significance of symbols (Cummins, '29). Discrepancies between the incidences of palmar features recorded f o r the present collection in Wilder's paper and in this report are due, however, almost wholly to the addition of thirty-two palms to the series formulated. Galton's ( '92) arch-loop-whorl classification is employed in the formulation of apical patterns. The formulation of plantar configurations follows Wilder's method, of ~vhiclithe most complete account is presented in Wilder and Went w o ~11t. THE PALM ,Vain. limes Of the one hundred prints of right hands, only fift;v-two are complete in every respect. Of the remaining prints, sixteen are unsuited to the tracing of main lines, owing usually to incomplete printing of the distal zone wherein are located the digital triradii from which the lines arise. The remaining thirty-two prints are only partially decipherable, but one, two, or three main lines being traceable in each. Curiously, the prints of left hands are more complete. Eighty prints are formulated in their entirety; thirteen prints lack the formulation of one or more lines, and seven prints are so incomplete that not one main line can be traced. Line A . Following the principles of formulating main lines, there result in this series several varieties of symbols 12 HAROLD CUMMINS iiivolving line A which required special attention in the counting of its terminations. First, the dual formulations 11/3, 11/5', and the like, of which there are several examples, are recorded when the line joins a triradius of a second interdigital pattern, with the consequence that the line is traceable both to a distal termination and to a position on the ulnar-proximal borders of the palm. I n the counting of the terminations (table 1) the distal termination was ignored in every case. The ground for this selection is that the generalized slant of the palmar ridge courses is indicated by the latter element of the symbol, while from the standpoint of a comparative treatment the distal termination is an unimT-4BLE 1 Tervziiiations of liize A , giving the percentile occurrence of each position for the right hand (compiled f r o m eighty palms), for the left hand (cornpaled from ninety palms), and the average of the two percentages POSITION i RIGHT Per cent 3 3 4 5' 5" 35.0 21.2 42.5 1.2 1 &EFT 1 AVEBAGE Per cent Per cent 1.1 57.7 23.3 17.7 0.5 46.3 22.2 30.1 0.6 portant local detail. The single example of the closely related 11-5' was treated in the same manner. Two examples of dual terminations on the ulnar border (4/3 and 4/2) were entered in the count of positions as 3 and 2, respectively, with the belief that the lower termination represents the significant generalized slant of the palmar configuration. When the line enters a hypothenar pattern (giving here several symbols of 3h and 4h) the entrance into the pattern is ignored, the terminations being entered in the count as if the line actually reached the margin. The alternative position in several alternative formulations, such as 4(3), is ignored, with reasonable confidence in the validity of the interpretation given prscedence in the symbol. The remaining terminations of line A are unequivocal. 13 DERMATOGLYPHICS IK NEGROES Perhaps the most striking feature of line A is the want of terminations in position 1 and the represeiitatioii of position 2 by a single instance (which is a left hand). Terminations on the proximal border are thus exceptionally rare. Notwithstanding the extreme reduction of the proximal positions, which are recognized to have a differential incidence in right and left hands, a bimanual distinction is quite evident in the apportionment of termiiiations among the levels of the ulnar border. Position 4, it will be recalled, sigiiifies a termination at an intermediate level in the length of the ulnar margin, while positions 3 and 5 (with its divisions 5” and 5’) embrace terminations estending over the proximal TABLE 2 Terminations of h e B, giving the percenttle occurrence of each positton f o r f h e right hand (comptled f r o m seventy-one p a l m s ) , f o r the left Land (concptlrrl f r o m ninety-one p a l m s ) , and the average of the two percetilnges POSITION 5’ 5w 6 LEFT AVERAGE Per cent Per cent Per cent 14.0 67.6 4.2 34.0 61.7 24.0 64.6 2.1 3.1 RIGHT ~ _ _ _ _ and distal halves, respectively, of this margin. Position 4 is about equally represented in the two hands. In left hands the remaining terminations are concentrated in the lower half of the ulnar margin, while in right hands the terminations are more concentrated, though not so much proportionately, in the distal half. L i m e B. The terminations of line B are assembled in table 2. The proportion of terminations in the distal half of the ulnar margin is rather uncommonly high in both hands. The left hand, as is usual, displays more terminations on the ulnar border than does the left. L i m e C. The positions of line C are shown in table 3. The large proportion of terminations in 5” seems noteworthy. 14 HAROLD CUMMINS Line D. In interpreting line D, a number of dual formulations appear, in these prints invariably represented by the symbol 9/7. Such terminations result from the fusion of the line with an accessory triradius of the fourth interdigital pattern. As pointed out by C u m i n s and Midlo (’as), the more radial position provides the significant indication of TABLE 3 Terminations of line C , giving the perceatile occurrmce of each position for the right Ramd (compiled f r o m f i f t y f o u r p a l m s ) , for the left hand (compiled f v o m eighty-oiie p a l m s ) , and the average of the t w o percentages ~POSITION RIGHT LEFT Per c e n t Per cent 3.7 3.7 33.3 5.5 14.8 38.8 3.7 6.1 53.0 0 X .i” G 7 9 AVERIDE ___Per ceibt 3.7 4.9 43.1 2.7 19.1 26.1 23.4 13.5 ~- TABLE 4 Teimwiatioiis of 1%neD , gzvzng the percentzle occurrence of each posztzon f o r the right lmad (compz.led f r m szxty-two p a l m s ) , f o r t h e left hand (compzled froin ctglity-seven p a l m s ) , and the avo-age o f Ihe t w o percentages - - ___ - -.-POSITION RIGHT ~- LEFT 1 ~ 7 8 9 10 11 - Per c a n t Per c e n t 37.1 4.8 33.8 4.8 19.3 63.5 - 25.7 2.3 3.4 1 ~ ~ ~ AVCR.4GE ____.. Per ceiil 51.3 2.4 31.2 3.5 11.3 ________-. generalized ridge alignment, and in all such cases it is accordingly chosen in counting the terminations for the makeup of table 4. I n the instance of several alternative formulations, such as 11(10), 7(8), and the like, the alternative position is ignored in the count, with the belief that the first formulation is reasonably dependable. The extremely low incidence of positions 11 and 10 is worthy of mention. Positions 7 and 9 are the characteristic DERilIATOGLPPRICS I N NEGROES 15 terminations in this collection, and, as usual, are more common in left hands. Reference should be made also to the grouping of terminations in modal types (table 5). Conaplete main-line formulae. Table 6 presents the occurrences of formulae in accordance with the treatment adopted by Wilder ('22). The abundance of 7.5.5.- formulae is iioteTABLE 5 T h e termanations of line D grouped according t o the and Midlo ( ' 2 6 ) IlLOdd 1clpes of Cwni I I I I ? I S TYPE P e r ceiLt Per 41.9 38.6 19.3 7 I, 11 c'iit I Pcr rent 53.7 34.8 11.3 65..5 31.0 3.4 TABLE 6 Pel centde d t & t b ~ t l ~ ?of~ the mawa-line fonioulae nccot &ng t o t h e y i oiipr?ig suggested by Wzlder ( ' 3 3 ) . Percentages are cnlculated o u the basas o f the ?oiimber of pa1in.s L ~ wkich L the mawt lznes a?-e contplelely foi ~ittilated (fifty t w o r z g 7 ~ t ;eighty l e f t ) _ _- RIGHT - - - I P e r cent 126.96.36.199.6 188.8.131.52.3 184.108.40.206.3 ~ l l o t l i e r s 30.7 -~ ______ LEFT ' 1 ~ e eei,t i 55.0 26.2 2.5 16.2 __ 1 , AVLR4GE 1 - 1 __ Pel ceici 44.8 21.7 9.9 23.4 worthy. 9greater variabilit,y of riglit liands is reflected in a higher iiicideiice of formulae which are unrelated to the t h e e characteristic forms. The main-line formulae may be assorted in accordance with modal types (table 7). With this assortment the 7 and 9 types are shown t o be strikingly in the ascendancy. Bimannal distinctions may be summarized by the relative abundance of the types: in riglit liands 9>7>11, while in left hands 7>9>11. 16 HAROLD CUMMINS Axial triradii Determinations of axial triradii are detailed in table 8. In the absence of varied comparative data, on the basis of the present formulation, it is impossible to state whether any racially characteristic traits are present. I n the Central American Indians ( Cummins, '29), however, there appears to be a greater tendency for distal shifting of these elements (with 58 per cent of the palms bearing a t, and 40 per cent, a t'). TABLE 7 Percentile distribution of the maimline formulae according t o the modal types of Cummins and Midlo. Percentages are calculated on the basis of the number of palms admitting assignment t o t h e modal typcs (sixty-one right; eighty-two l e f t ) ; it should be noted that several palms in which the main lines are incompletely formulated can be so assigned TYPE 7 9 11 RIGHT LEFT AVEBSGE Per cent Per cent Per cent 36.0 44.2 19.6 57.3 39.0 3.6 46.6 41.6 ll.G TABLE 8 Percentile occurrences of axial triradii (compiled f r o m seventy-one right palms and seventy-three l e f t s ) 1 RIGHT ______ t t' tt' t" 1 LEFT I AVERAGE Per cent Per cent Per cent 66.2 30.9 1.4 1.4 70.0 23.2 4.1 2.7 68.1 27.0 2.7 3.0 Patterns The distribution of the varieties of hypothenar configurations is shown in table 9, and the total incidence of true patterns, as contrasted with pa.tternless configurations, is given in table 10. The infrequency of hypothenar patterns (18.2 per cent for all varieties combined) is a conspicuous feature of the collection. A source of possible error in the record of patterns of the L' type should be recognized. Some of the 17 DERMATOGLYPHJCS IN NEGROES palm prints are not adequately rolled over the ulnar margin. and it is possible that an L' having its enclosed end well extended beyond the area printed would not be correctly identified, being then recorded as an A". Careful inspection TABLE 9 Percentile occurrences of the varieties of hypothenar configurations (compiled f r o m ninety-eight right palms and ninety-seven l e f t s ) TYPE L" L' Lr/Ac A"/L" T' A '/Lr A" RIGHT LEFT 1 AVERAGE Per cent Percent j Percent 1.o 12.2 3.0 1.0 12.3 2.0 3.0 1.0 1.0 64.2 G3.9 I, 1i 2.5 2.0 0.5 0.5 64.0 15.8 1.5 1 I __- 0.5 12.2 ~-- TABLE 10 Percentile occurrences of true p a t t e r m and pattern vestiges o f the palm (not including 'multiplications' in the interdigital areas or arches of the hypothenar area). T h e separately li.?ted true hypothenar patterns of table 8 are here combined. T h e numbers of palms f r o m which t h s figures are compiled follow: H y p o t h a a r , ninety-eight right, ninety-seven l e f t ; thenur and first interdigital, one hundred right, ninety-eight l e f t ; second interdigital, s i z t y right, seventynine l e f t ; third interdigital, fifty-two right, seventy-four l e f t ; f o u r t h interdigital, fifty-eight right, eighty-two left ~ PATTERN RIGHT LEFT ___ Per cent Hypothenar Thenar and first interdigital Second interdigital Third interdigital Fourth interdigital Per cen AVERAGE Per cent 18.3 22.4 6.3 18.9 93.9 18.2 9.0 13.3 40.3 84.5 ~ 18.2 15.7 9.8 29.6 89.2 -- of the prints leads to the conclusion that incomplete printing of L' patterns proba.bly is an almost negligible source of error. It appears, therefore, that the scarcity of patterns may be accepted as a definite racial trait, especially since similar low .figures are reported for two other collections of 18 HAROLD C U M M I N S iiegroes (Wilder, '04, a small series of negroes in America; Steggerda, '29, blacks in Jamaica). A patterned configuration of the thenar eminence, representing morphologically the thenar and first interdigital patterns, is fairly frequent. Second interdigital patterns occur in almost 10 per cent of the hands-an unusual frequency which is approximately that reported for Jews (Cummins and Midlo, '27). Third interdigital patterns are relatively infrequent, while patterns of the fourth interdigital area display a remarkably high frequency, unmatched by any race thus far studied and approached only by the Central American Indians, where the average for both hands is 76.6 per cent (Cummins, '29). There are 126 palms (49 rights and 77 lefts) bearing fourth interdigital patterns, as follows: L occurs in 72 palms, 1 in 4, D in 37, d in 1, I,,% in 11, and L/L in 1. APICAL PATTEBNS The finger-print sets are complete f o r only fifty-eight individuals. Calculation of the occurrences of the four basic types of configurations is based upon these alone, the incomplete sets being useless for the purpose 011 account of bimanual and digital peculiarities in the incidence of the types. The percentile distribution of types in the 580 patterns is : whorls, 38.9; ulnar loops, 56.3; radial loops, 1.0; arches, 3.6. While the series is small, it pielcls certain possibly significant indications of a racial trend, which will be of interest in comparison with a tabulation of other racial studies, such as Boiinevie's ('24) table 5. The most outstanding result is the low incidence of radial loops, amounting to only 1 per cent, in contrast to occurrences ranging from 2.7 per cent to 5.8 per cent in a number of peoples assembled for comparison by Bonnevie. The figure of 1 per ceiit represents an occurrence of six radial loops among the 580 apical patterns. Since it is so rare a pattern, the incomplete sets of prints may be scanned to determine whether their addition would appreciably alter the incidence reported for the fifty-eight complete sets. Among the forty-two incomplete sets there 19 DERMATOGLYPHICS IN NEGROES are four radial loops, digit I1 of right hands accounting for three of them, a right digit 117 for the fourth example. Eadial loops, it should be recalled, are practically confined to digits I1 and IV in all peoples ; in this collection 376 of the patterns of these digits are lmowii, out of an expected total of 400, and it seems quite unlikely that the twenty-f our iiidecipherable prints could carry a coilcentration of radial loops sufficient to enlarge appreciably the percentile incidence of this pattern. In table 11 are collected the occurrences of pattern types for each digit separately, of interest in comparison with Boiinevie's tables 6 to 9. TABLE 11 Percrjrtile orcuwcnces of p a . t t w n t y p e s for each digit separately I Arches Radial loops Ulnar loops Irnorls I1 4.7 11.1 31.5 60.7 39.5 49.3 9.6 6.4 7.6 3.1 14.0 53.2 ' 39.7 36.9 - . 111 1.0 80.2 18.7 ~ IV 4.0 I 73.8 23.2 PLANTAR CONFIGURATIONS An analysis of the hallueal aiid three interdigital coiifiguratioiis is carried in table 12. Of particular interest are the large incidences of whorls in the liallucal area and of L T patterns in the third interdigital area (compare Cummiiis and Midlo, '27, table 14). 20 HAROLD CUMMINS TABLE 12 Percentile occurvences of particular pattern types and open fields i n each of the four distal plan.tar areas AREA A N D NUMBER COMPILED Hallucal (197 soles) 1 CONFIGURATION TYPE A B c W 0 First interdigital (191 soles) 0 U n m Second interdigital (188 soles) 0 U n w Third interdigital (189 soles) 0 U n 1 1 OCCURRENCE 25.3 13.7 0.5 55.3 5.0 66.4 0.5 31.9 0.5 28.7 51.0 2.1 18.0 65.0 34.3 0.5 SUMMARY Palmar main lines, palmar patterns, apical patterns, and the configuratioiis of the four distal plantar areas are analyzed in one hundred male natives of Liberia and Sierra Leone. The report forms an extension of Wilder’s study of the same collection (Amer. Anthrop., XV, 1913). The racially significant trends of variation are as follows, the figures being percentile and stating the averages of the occurrences of the respective features in right and left members: proximal terminations of line A (positions 1and 2) are reduced (0.5); line D terminates commonly in positions 7 (51) and 9 (31) ; the most common formula is 7.5.5.- (45), and next, 9.7.5.- (22) ; hypothenar patterns infrequent (18) ; the thenar-first interdigital pattern fairly frequent (16) ; second interdigital pattern fairly frequent (10) ; third interdigital patterns relatively infrequent (30) ; fourth interdigital patterns very common (89) ; among apical patterns radial loops are reduced (1); hallucal patterns have an DERMATOGLYPHICS I N NEGROES 21 abundant representation of whorls (55) ; the third, fibularmost, plantar interdigital area has a high incidence of U pat,terns (34). LITERATURE CITED BONNEVTE, KRISTINE 1924 Studies on papillary patterns of the fingers. Jour. Geneties, XV, 1-111. CUMMINS, HAROLDDermatoglyphics in Indians of southern Mexico and Central America (Sta. Eulalia, Tzeltal, Lacondon and Maya tribes). (In press.) CUMXKINS,HAROLD, AND CHARLES MIDLO 1926 Palmar and plantar epidermal configurations (dermatoglyphics) in European-Americans. Am. J. Phys. Anthrop., 19, 471-502. 1927 Dermatoglyphics in Jews. Am. J. Phys. Anthrop., X, 91-113. CUMNINS, HAROLD, H. H. KEITH,C. MIDLO, R. B. MONTGOMERY, H. H. WILDER, AND I. W. WILDER 1928 Study of error in interpretation and formulation of palmar dermatoglyphics. Am. J. Phys. Anthrop., XI, 501-521. 1929 Revised methods of interpreting and formulating palmar dermatoglyphics. Am. J. Phys. Anthrop., XII, 415-473. GALTON,FRANCIS 1892 Finger prints. Macmillan. KEITH, HAREIETH. 1924 Racial differences in the papillary lines. Am. J. Phys. Anthrop., VII, 165-206. STEGGERDA, INEZ D. 1929 Palmar dermatoglyphics in negro-white crosses. (In “Race crossing in Jamaica. ”) Carnegie Inst. of Wash., Publication 395, 229-252. WILDER,HARRISH. 1904 Racial differences in palm and sole configuration. Am. Anthrop., VI, 244-292. 1913 Racial differences in palm and sole configuration. 11. Palm and sole prints of Liberian natives. Am. Anthrop., XV, 189-207. 1916 Palm and sole studies. Biol. Bull., XXX, 135-172, 211-252. 1922 Racial differences in palm and sole configuration. Palm and sole prints of Japanese and Chinese. Am. J. Phys. Anthrop., V, 143-206. WILDER,H W S H., AND BERT WENWORTH 1918 Personal identification. Badger.