Digitized by the Internet Archive

in 2010 with funding from

University of Toronto

http://www.archive.org/details/anthropologicalrOOthom

ANTHROPOLOGICAL REPORT

ON THE

EDO-SPEAKING PEOPLES OF

NIGERIA.

-^ o

PART I: LAW AND CUSTOM.

BY

NORTHCOTE W. THOMAS, M.A., F.E.A.I., etc.

GOVERNMENT ANTHROPOLOGIST.

LONDON r HARRISON AND SONS.

( 1910.

611863

4-. 7. shs

CONTENTS.

Page.

Section I. — Language and People 5

II. — Religion and Magic 24

III. — Marriage and Birth 47

IV. — Inheritance, Adoption and Property 64

V.— Law 103

VI.— Kinship 112

Appendices : —

(A) Linguistics 123

(B) Genealogies and Kinship 138

(C) Photographic and Phonographic Records 153 Index 161

A 2

SECTION" I.

Language and People.

Area. — The group of peoples with which the present report is concerned occupies the centre and north-east of the central province of Southern Nigeria, and a portion of the southern part of the Kabba province. The area is roughly oval in shape, with its long axis north and south and of a maximum length of some hundred and fifty miles, and a short axis of sixty miles or less. The most southerly point appears to be south-east of Warri, and the most northerly point a few miles north of a line running from I sua to Ibilo.

In this area is one large tribe, the Edo or Bini, of which Edo, as they themselves term it, or Benin City, is the capital; north-east of them lie the Esa or Ishans round Ubiaja; south of them lie the Sobo, with three or more dialects ; and the remainder of the area is taken up with a mass of tribes, some of them limited to a single village, which are sometimes collectively known as Kukuruku. In the south-west of the Kukuruku country we have the Ora country, and further to the east, north of the Esa, are the districts of Uzaitui, Fug a, and Agenigbodi. Over the border, in Northern Nigeria, are the Ibie and Upila districts ; in each of these districts the language appears to be fairly homogeneous ; outside them the linguistic groups are often exceedingly small ; in the north-east angle of the Ifon District, for example, we find Iki, Aroko, Otua, and Okpe, the languages of which are said to be mutually unintelligible, though they are not more than two hours distant from each other.

6

Physical Type. — There is little uniformity of physical type among the Edo-speaking peoples, though it is not difficult to recognise a Sobo or an Edo. The average stature of the men appears to be about five feet five inches. As regards physical strength there is also much variation, if the performances of carriers can be taken as a criterion ; though it must be remembered that a carrier living on strange food and traversing unusual ground cannot approach his normal per- formances. Generally speaking the hill people are superior in physique to the people of the plains, and will carry equal loads in spite of the difference of ground. Although the negro is able to carry loads for long distances, he must go at his own pace ; even without a load his endurance is small when he is called upon to move faster than usual.

As a rule, malnutrition is rare ; but occasionally, as in the Or a country, there is a lack of food owing to sheer laziness.

Tribal marks and fashions in hairdressing are now losing their significance, but the raised scar on the forehead of the Sobo, the crow's foot at the outer angle of the eye in the Ishan, and a few more are yet kept up.

Physical deformities are rarely seen, the main one to catch the eye being umbilical hernia, due to wrong methods at birth ; it is very local in its distribution.

Lunacy is rare ; deaf and dumb people are occasionally met with. Among abnormalities may be mentioned albinism (common in places), red hair (rare), and supernumerary fingers.

Language. — The family of languages whose distribution is here described appears to be distinct from those of the surrounding peoples, which are, on the south the I j o ; on the west the Yoruba, on the north-west and north tribes of the Yoruba stock, on the north-east the Igbira; on the east, but on the other bank of the Niger, are the Igara at Ida ; then come, on the west bank of the Niger, various branches of the I bo, who extend down into the Ijp territory.

Certain words appear to be common to the Edo and other

families of languages, but in some cases their prevalence in both families can be readily explained ; and in any case their number is far too few to afford the presumption of any relation between the languages, whose syntax appears to be entirely different. If there is any resemblance between the Edo family and other West African languages, it is probably to be found with the Ewe tribes of Togoland and the Gold Coast, with the verbal forms of which there are certain obvious resemblances.

Among the words common to Edo andYoruba are such terms as oke (hill), okuta (stone), and the like, and the reason for their appearance in both families of languages is firstly, that the line of kings which formerly ruled the Edo came from the Ife country ; and that, secondly, hills and stones being virtually unknown in the Edo country, terms for them would naturally be adopted from the language of the immigrants.

There are certain resemblances between the Edo and Ibo languages which are more difficult to explain. Parts of the body for example, like the mouth (unu), are known by identical names, and here it is difficult to suggest an explan- ation. Finally, there are a certain number of words, like ekuiye (spoon), which are obviously adopted from a romance language of Europe, and we need have no difficulty on deciding on Portuguese as their source, for it is well known that the Portuguese were in Edo at the beginning of the sixteenth century.

The languages are, as a rule, difficult to learn or under- stand, and the difficulties arise from two sources. In the first place, a sentence or phrase coalesces into a single word with the vowels elided or euphonic letters inserted. It is often difficult to discover from an interpreter what are the words which are the component parts of such a phrase.

In the second place, the Edo languages are, like Yoruba, and certain other West African languages, toned, that is to say, the meaning of monosyllables, dissyllables, and some- times of even longer words, is changed by altering the intona-

8

tion with which they are spoken. To take a simple example, eha means three, while eha means six ; in this case stress and accent coincide. To take a more complicated example, Ide means " I come," while Ide means "I am not coming." Id§ means " I buy," Id§ means "I fall," and Ide, " I tie." When toned words are included in the body of an agglutina- tive word, the difficulty of recognition, which is sometimes considerable in any case, is greatly increased.

In Edo proper there is, as a rule, no plural, but a certain number of words change an initial o into i to form the plural, thus: — oxwo, ixwo, woman; odiQii, elder; idion, elder. The pronoun commonly used for the third person, namely, on a, or o, is inflected in the same way, and has plural in a, or a. Many nouns are in reality the third person singular of a verb, as, for example, om axe (he makes pots), potter ; plural, imaxe. But it would be erroneous to suppose that the plural of non-verbal nouns has been formed by analogy from that of verbal nouns ; for, in a few cases in Edo, we find plural in e, for example, q bo, doctor ; e,bo, doctors; in the second place, in some of the Kukuruku languages, other vowels undergo a change. The vowel in the plural is sometimes a : for example, obo,, hand ; abg, hands, in Ishan.

The personal pronouns are in Edo i, ti, me, or memo for the first person, we for the second person, sometimes found in the form u before gb, w, etc. ; for the third person, ona, o, oi, le, e, etc. In the plural, ma, wa, and ina. Personal pronouns are prefixed to the verb, which is uninflected for a person or tense. The objective pronoun usually follows the verb, but is occasionally inserted in the body of the word. An example of the former is inele from ine, I know, le, it; an example of the latter, inome, I do not marry ; inelomi, I do not marry him.

In connection with this latter example, attention may be called to a peculiarity of some of these languages. There is no negative in inome, but a particle must be inserted to change fcheverb from negative to positive. Thus : iganome means " I am going to marry."

A particle ga inserted after the personal pronoun, is the normal method of forming the future, for the perfect ne may be affixed to the stem of the verb, or fo {i.e., finished) may be used in the same way. Thus: ilue, I do it; igalue, I shall do it ; iluene or iluefo, I have done it.

The negative may be indicated in various ways. The accent may be thrown back as in ide, I come ; ide, I am not coming, or ma may be inserted in the stem of the verb ; for example, imadQ, I do not buy. For the negative imperative ge is used — for example, gemudia, do not stop. The subject pronoun precedes the verb in the interrogative sense, and the question is shown by a particle, by dynamic stress, or by musical intonation. Thus ugbe means you killed him : ugbe, did you (not) kill him ?

Subordinate sentences are comparatively rare. The rela- tive pronoun is ne, and relative sentences are frequent. Subordinate sentences of time are also found, but the ordinary method of expressing a hypothetical clause is to make it co-ordinate with the main sentence. Thus, to express in Edo " If you stay I will stay," one says, wamudia (you stay) idiake (I stay with you).

In connection with the verb mention must be made of the importance of separable verbs, thus: mure,bring,mgrukpare, bring the lamp ; another feature is the piling up of verbs to express what is to us a simple idea. Thus: irhiegade, I bring, is compounded from irie, I take ; ga, go ; de, come.

A certain number of impersonal verbs are found, usually of sensations or emotions, thus: ohamegbime, I am thirsty.

The adjective is uninflected, except for comparison, of which there is only one degree, formed by adding se, past, thus: f)kpalo, big; okpalose, bigger.

There are distinct words in all the Edo languages for numbers from one to ten, twenty, sometimes thirty, and two hundred. The numbers from eleven to twenty are com- pounded in two ways : from twenty upwards they are vicesimal with the exception of thirty, the term for which is sometimes borrowed from the Yoruba.

10

ce 02	Qkpa. era.	a>	2 «■ d .5 .s ©	2 cdV*^	CD jz S 32 ,2 CD ~r ^'■'©r
Kukuruku of Uzaitui.
eg P- ce CT ©	c3 '"S % ^ o Q).h a;	CD p Id cd -^ cc --n! >o • iH CD >iH	■^9 • ce 3 * C3 ,-» C3 . • CC ■-— < ,-=< > r— < ^ CC ^5 CC^=.>0 CDCDCDCDOCJOOCDCDCD p toe be /t£ [tc t tsC p be Tl Tc Tc _ be i|
Sobo of lyede.
•	•	.	.	:•:::: t> : : : :
tJD
Cr.£	_9 CD of r-l CDT	• . i— i • • "o • CD b^^-s S r- cc CD .3! CD	CD • .U x& .S	ikwe iwrovo iwive iwera iwene ixwioli ixwesa, ixw ixwioligbiv ixwiolerba ixwiolene uze . . .
c3 o
ce • Ph ce ^ > CT CD	<3 cc CD	. . CD CD N 5 CD CD S ^ N CD .i-H .r-(	cd : ■g .cd • .3j ■— ._( S3 »r cc	igbe igbokpa igbeva igbesa igbene igbize azenezowig azesazowigc azevasowig azgkpasowi
d o J- <
ce : Cr CD	c3	: '• ce CD °> £ *3 CD CD ^	58 2 P. CD^ -C P ca	2 CD CD ~ m P P P n? ^ cc © CD 2 ce CD ^ cu,cececD.^cDb:^5 CD C CD CD CD rS .S .2 > ^ CD ^^^jd.jd^ ce ce o c«-> tr. to tc tc tc bc-44 M r^ ^ g
6 •73
O O c8 k> c3 *ce	2 <» -	>rH ^3 »i-< ^ '-r -P .,_ CD ..-<	igbe iweva1 iwera iwene ikesugie enelawugie ehalawugie evarawugie gkpadewug ugie
i— i C3	cc	-t IT, Cl l> X C	OHMMTjdOONMaiO

11

Village. — Villages vary in size from the hamlet of three or four houses, and twenty or thirty inhabitants, to places with a thousand or more inhabitants, which are themselves only quarters of a larger town. Especially in the hill country, we find large populations, perhaps as many as ten thousand in some places, in a small area, recognising the authority of one head chief.

Social Organisation. — The organisation of the village varies with the tribe ; in the Edo country there is a tripartite division of the males — senior grade is known as Idign, or elders, next to them comes Igele, and the junior grade is known as the Oloqai.

The Idign are virtually the council of the village, and in pre-European times they dealt with minor law cases, the more important ones being sent to Edo for the judgment of the king.

The Igele, assisted as a rule by some of the junior Idigh, have to undertake such work as roofing houses, and it is from this rank that carriers are usually drawn.

The Ologai carry wood and water, and do some of the work of cleaning the roads. A boy joins this rank as soon as he is strong enough. The promotion from the Igele to the IdiQn is obtained by means of payment.

It is difficult to say what the average proportions of the various grades are, but in a total population of about one thousand I obtained the following figures: IdiQn, one hundred and eighty-one ; Igele, fifty-eight ; Ologai, seventy-five.

The villages of the Edo country were grouped, before the European occupation, some under princes who were known as Ogie (plural Igie), others under chiefs, stewards, or other heads. The Igie were descendants of former kings, whom custom compelled to live away from the capital. Each village, whether presided over by an Ogie or not, was represented in Edo by a chief or king's steward, and communications to the king were normally made through this representative. Tribute in kind was paid to the king, a proportion of which was retained by the representative.

12

In the Ora country, and the greater portion of the Kukuruku country, the adult males of the village who have not attained to chieftainship, are organised in companies (Otu) formed of men of about the same age. An Otu is inaugurated, as a rule, once in three years, and, as a rule the junior Otu has to undertake the cleaning of the roads. Where there is no lack of labour, the senior Otu are dispensed from labour, and the four junior Otu do the work, with one Otu to supervise them. The commencement of a new Otu is celebrated with more or less festivity, and in some places they wear ceremonial dresses and dance round the town in the same wTay as Secret Societies.

The various Otu are known bynames, by collecting which on a large scale it is possible to arrive at some idea of the average age of the oldest men of a village.

As a rule, the oldest Otu in a place of any size appears to be about forty years older than the junior Otu. If they come in once in three years, the oldest men who have not yet become chiefs must be about sixty-five years of age.

In the Ishan country, the organisation comes nearer the Edo tribe. In Irua, they have two Otu, of which the senior is liable for work only when there is much to be done. Above the Otu come the Igele and the Idiqn.

Elsewhere in the Ishan country only IdiQii and Igele are found.

In the Sobo country the social organisation is more rudi- mentary, but there is a division into old men, adults, and small boys. The adult men are often spoken of as a " Club," over which there is a headman.

In the kingdom of Edo there were in pre-European times two kinds of chiefs, a small number of the more important were succeeded by their sons, but, as a rule, a position was sold by the king for the highest figure. A son of the late chief was not necessarily excluded, but he had no special claim to his father's position.

In the Kukuruku country at the present day the position of chief is purchased, and the payment goes to the existing

13

body of chiefs ; in the Ora country the cost appears to be comparatively small ; and in the kingdom of Agbecle, where there is a king, chieftainship appears to be a comparatively minor matter. In Ida, however, and probably in the adjacent parts of Northern Nigeria, the cost of becoming a chief is heavy (about £30), and the system gives rise to grave abuses.

Originally there was considerable advantage attaching to the position of chief, for he could neither be arrested nor attacked in war. At the present day, the only advantage of becoming a chief, if we except a certain amount of prestige attaching to the position, appears to be the share in the fees of the new chiefs, which is squandered in riotous livino-. So far as I could discover, practically the whole of the fees are paid in kind, and the payments are spread over a period of years. Each payment appears to be an excuse for a feast on the part of the recipients, and the only result of the present system is to impoverish the community.

By limiting the number of chiefs a check would be put upon this waste, greatly to the advantage of the people at large ; or, alternatively, it might be possible to enact that a proportion of the fees payable by the new chiefs should be handed over to the Government for the benefit of the com- munity.

In the Sobo country women chiefs are found, and the system by which chiefs are chosen appears to vary consider- ably. In some cases chieftainship is hereditary, in others it depends upon age, and in others upon wealth.

Demography. — It is commonly assumed that in Africa generally women far outnumber the men. I collected statistics from four contiguous villages in the neighbourhood of Edo by the genealogical method. In the nature of things, no exact statistics as to age are available, but, assuming that all the youths above the Ologai are adults, and that all women of the same age or older are married, the proportions of the sexes are almost precisely equal. In point of fact, a certain number of Ologai are married, and it appears there-

14

lore that the males above the age of puberty are more numerous than the females, for there are a certain number of unmarried males, whereas only such females as are widows or too old to marry again, remain unmarried above the age of puberty.

In addition to enumerating the adults, I also recorded the names of all the children. It is necessary to make special enquiry as to the number of children who are dead. The infant mortality appears to be enormous, and children who die soon after birth are so often omitted in the genealogies that the statistics are quite misleading. It may be added that it is also necessary to make special enquiries about deceased wives ; it frequently happens that their names are omitted, and their children may be passed over.

In the villages in question, taking children of all ages, or rather all the descendants of those now living in the villages, I found two hundred and twenty-eight males to one hundred and eighty-three females.

In order to discover the exact meaning of these figures, it is necessary to discover how many of them are of the age of two years or less. At the time of recording genealogies, when my figures were not yet tabulated, I did not realise the necessity for this. Accordingly, it is impossible to say whether a surplus of males runs right through the population up to the age of puberty, or whether more males are born than females, and the male mortality is greater than the female. The question is further complicated by the fact that many of the younger males leave the village to work on concessions. If their names are omitted in the enumeration, the apparent equality between the sexes in adult life, which is inexplicable, if the statistics from descendants are accurate, is readily explained. It must, however, be admitted that the numbers recorded in the genealogies are in any case too small to give a thoroughly reliable result.

In order to test the results thus obtained, I have tabulated the sexes of the descendants of the older generation, that is to say, of the brothers and sisters of the people now living in

15

the village, who are not descendants of people still alive, and, where I could obtain the information, I have included the figures of their descendants. It must, however, be remem- bered (a) that there is a tendency to forget the names and sexes of children who die soon after birth, where the informant is not one of the parents ; and (b) that the numbers of the women in particular are apt to be too small, because, where the information is drawn from the youngest son of a family, the names of the elder sisters may be either unknown to him or forgotten. This was clearly shown by the statistics of the sex of the first-born child ; for, whereas in the present generation, the proportions were seventy-nine males to fifty-five females, including only those oases in which the children were still alive, the proportion in the earlier generation was forty-seven males to nineteen females. The total of the earlier generation was one hundred and three males and sixty females.

I obtained apparently reliable statistics as to the death rate for about one-half the total number of families. The total number of living descendants of people residing in Utekon was one hundred and ninety-six, and the total number of deaths of their descendants, almost all of them in infancy, was one hundred and twenty-nine, giving a death rate of about four hundred per thousand. As some of the living were infants, it may be assumed that not far off five hundred children out of a thousand die.

In the four villages enumerated the average number of living wives of a husband was 1*5, including deceased wives and those who had eloped it is 1*7. Of a grand total of two hundred and forty males, one hundred and fifty-five were married, and no information was obtainable as to ten. More than two-fifths of the married men had two wives, and one- fifth had three wives.

The average number of children per husband, including only those who remain alive, wTas 2*7 ; including all children alive and dead, the average number of children becomes 2 '7 per wife, and 4-5 per husband.

16

I did not record genealogies on a sufficiently large scale elsewhere to give reliable results, but in one of the com- pounds of Fuga I found an excess of females over males, the exact figures being seventy-eight males and ninety-three females.

A fragment taken from another compound showed a small excess of males. I also recorded the number and sexes of the children in the older generation, as in the Edo country, and here, for some reason, the total numbers recorded were considerably greater. Among them, however, were included all the children of women of the present generation who had married out of the compound. The totals thus gained gave two hundred and three males to a hundred and sixty-five females. The proportions obtained by combining these numbers with those already given, suggests that in Fuga the sexes are about equal in number.

If the figures drawn from the Edo villages and from Fuga are even approximately correct they are somewhat difficult to explain. It is a well-established fact that privation and hardship tend to produce an excess of male births. But life in the old days was undoubtedly harder than it is at present, and there can be little doubt, that life in the Edo country is easier than life in the Kukuruku country. Statistics, how- ever, on a much larger scale, are needed in order to give reliable results.

In the Sobo country genealogies were persistently falsified to make the number of children appear as small as possible. I obtained reliable statistics in one village only ; and they showed 3*5 living children per marriage, which was usually monogamous.

Food. — In connection with the genealogies, I also made enquiries as to the amount of property possessed by each man. As might be expected, movable property, other than food products, was comparatively rare in the villages. The statistics of yam production showed that the output varied enormously, according to the industry or otherwise of the labourer. Yams are stacked in ropes (Uga) of twenty- two

17

or twenty-three, ten ropes make an Ekbo, or stack. In some cases an adult man, unmarried, did not produce more than five Uga, while others, under no more favourable conditions, produced five Ekbo. On an average, each adulfc man appears to produce about two Ekbo, or four hundred and fifty yams. The total output for the three villages where the enquiry was made was about a hundred and seventy-nine Ekbo. The adult population was about one hundred and eighty, and there were the same number of children. It must, however, be remembered that in a certain number of cases men were sick and unable to work their farms. Had this not been the case about twenty more Ekbo would probably have been produced.

Yams are the staple food of the Edo proper, but, as these statistics show, they must be largely eked out by other products, such as corn, beans, and koko yams, which are not included in the figures given above.

In the hill country of Otua and Okpe far more corn appears to be produced, and it is common to find large corn- fields quite close to a quarter. I found it, however, impossible to obtain statistics as to the annual output.

The staple food of the Sobo country is cassava. Here again, though from a different reason, it was impossible to give accurate figures ; whereas yams are dug up and stacked as soon as they are ripe, cassava may be, and is, left in the ground for a considerable period. I measured some of the crops which were worked by individual families, from which it appeared that the area for adults was about two hundred square feet. The statistics were collected on a small scale, and it was impossible to say what the total output of cassava for this area would be.

As a rule, the yam stacks of a village in the Edo country are found on the farms themselves, elsewhere they are frequently in the bush, and only chance, or information from an inhabitant of the place, will disclose the locality. As a rule, yams are brought to the village each day in quantities sufficient for consumption.

B

18

Corn is stacked in conical piles with a pole in the centre. These stacks appear to be made, as a rule, on the farms.

In addition to yams, cassava, and corn, banana and plantain are found in most parts. Many different sorts of bean are grown, and in the Sobo and Kukuruku countries, and to a less extent in other parts, a considerable amount of dried fish appears to be eaten. Flesh appears to be but seldom consumed. Eggs and milk are unknown as articles of diet.

The only intoxicating liquor produced over a great part of this area is palm wine, made from the oil palm by (a) a cut below the crown, or (b) by felling the tree ; another kind is produced from raphia vinifera. Where guinea corn is common pito is made from it.

Calendar. — Although an enquiry as to the age of a man in years, or the number of years since a given event, will meet with no answer or a random one, a calendar year is known to all the Edo-speaking people. In the kingdom of Edo, two kinds of years were known, male and female, one of which was probably about a month longer than the other.

The year was divided into months, or, as they are usually termed, moons. In most cases they have no name, in other cases, as in Edo itself, the only names which are known do not stand in any exact relation to the lunar months, but are taken from the ceremonies proper to certain periods of the year. Occasionally, months other than lunar months are known, and at one point in Northern Nigeria, a twenty-day month seems to be used, with two periods of nine months to our year.

The week is everywhere a recognised period of time, and is, properly speaking, four days long, this being the interval between the two markets at any given spot. Occasionally, as in the Ida district, eight-day markets are found, but the names applied to the intervening days clearly show that a four-day week was the primary one.

One of the four days is commonly known as the rest-day

19

and on this day men frequently stop at home, though farm work is not absolutely forbidden. Women, on the other hand, go to market as usual.

Market. — The hour for the market varies in different parts of the country, in the kingdom of Edo and in the Sobo country it begins in the forenoon, and is fullest about mid- day, and the same appears to be the case with the markets in the Kaba province. At Agbede, on the other hand, markets are held in the afternoon, and about three o'clock is the hour at which the market-place is full.

Hausa and Yoruba traders apart, market trading is entirely in the hands of women, and the products which are set out for sale in the market are, as usual in Africa, almost ex clusively food-stuff.

In addition to the market, there is the so-called silent trade. Outside a village may often be seen by the wayside, plantains or other eatables with palm kernels lying near them. These are for sale, and the palm kernels indicate the price in cowries.

With the exception of the gates of Edo itself, there appear to have been no points at which tolls were collected. At the gates of Edo, in pre-European times, the gate-keepers are said to have collected five cowries from each person going to market, or, according to another account, five per cent, of the value of the load.

To the question whether a village with an established market had a right to object to the setting up of another market on the same day in the same neighbourhood I could never get a definite answer, The formalities which attend the setting up of a market are slight, and, on the whole, it appeared that a new market set up under such conditions would probably not attract a sufficient number of people, and that therefore the experiment would probably not be tried.

The market-place is sometimes wholly within the territory of the village, though it is not necessarily anywhere near an inhabited centre. In other places the market-place is neutral

B 2

20

ground, standing midway between two villages, or upon the boundary of their territories. In former times, the seizing of pawns in the market-place or on the way to market was forbidden, and any place which declined to conform to this regulation was unable to carry on a market.

In the present day, holding of the market is largely dependent upon the supply of food-stuff', and, in the time of scarcity, before the new yams come in, it may cease to be held altogether. In the hill country, the markets appear to be unimportant — in fact, there is sometimes a small daily market in the place of the usual four-day one.

Arts and Crafts. — The most important native industry, apart from the production of food, is undoubtedly palm oil making. In each village may be seen the troughs in which the nuts are washed. After cutting them from the tree, the first operation is to boil them. They are then put into the troughs, which are filled up with water, and women trample the nuts until the oil floats on the surface, and the husk is separated from the kernel. The oil is then skimmed off' and put into a pot ; the next operation is to rinse the husks with water a second time, and the oil is again skimmed off. After a third rinsing and skimming the husks are removed from the water and put on one side for use in lighting ; the kernels are put aside and cracked at a later period. The oil run from the husks is boiled and separated, the inferior part being reserved for lighting purposes.

Cotton is grown in some abundance in various parts of the country. It is frequently of good quality, and the thread made from it is strong, but as a rule the native is disposed to grow his cotton mixed up with other crops, such as beans, and the result is that the cotton crop suffers in quantity.

After the cotton fibre is remove/1 from the plant, the first operation is to roll it on a block with an iron bar in order to remove the seed. It is then cleaned with a bow ; to spin the cotton it is wound" round the left arm or on a short stick held in the left hand. The end is made fast to a spindle consisting of a wooden rod passed through a disc made from

21

a broken calabash. Drawing a portion of the cotton out, the woman gives a twist to the spindle, and in a very short time about a yard of thread is twisted. This is wound on the spindle and secured by a hitch, then a new length is made. When the spindle is full the thread may be wound off' on to another spindle, or it may be wound upon a stretcher with movable pegs.

Two kinds of looms are in use, one used by women, which makes a comparatively short but broad piece of cloth ; the other a man's loom, which produces a long and narrow fabric. The man's loom appears to be exceedingly rare among the Edo-speaking tribes, but the woman's loom is common in most places where European stores are not readily accessible.

The woman's loom is upright, and consists of two pillars with crossbars, round which run the woof threads, the shed is formed by a heddle rod wound with cotton for each piece of cloth ; the warp thread is put through with a shuttle, and beaten down with a loom sword. The time taken to make a piece of cloth five feet long seems to be about three days. Coloured threads are employed to produce the patterns, for all of which native names exist. In some cases the dye stuffs are produced locally, from bark, seeds, or roots, in other cases they are purchased from Hausa or Yokuba traders.

As a rule the weaving, apart from the introduction of coloured threads, is quite straightforward, but occasionally an openwork pattern is produced by inserting loose threads and tying up five woof threads at intervals across the whole breadth of the cloth. This done, a warp thread is put in, five new threads are taken up to the right or the left of those taken up in the first instance, and another warp thread is put in. The number of threads between the insertions of the loose threads is forty or more, and when these have all been taken up, by moving the loose threads to the right or left an open chevron is produced. This pattern is repeated at intervals until the cloth is finished. As a rule cloth is

22

made entirely of cotton threads, but in the hill country the inner bark of trees is occasionally used for the woof threads.

The man's loom is on an entirely different system ; heddle frames, worked with foot loops, produce the sheds, and much facilitate the making of the cloth. The warp thread is beaten home with a batten, and the cloth as it is finished is wound upon a revolving bar. The woof threads pass over a cross piece behind the heddles, and are made fast, some distance away, to a stone or other weight, by which the necessary tension is kept up.

Pottery is produced by exceedingly simple processes. The clay is, as a rule, purchased in the market. After being brought home, it is watered and worked with the feet till it is of the proper consistency. Then sausage-like rolls are prepared, and when a pot is to be made the potter seats herself, sometimes upon a piece of wood the size of a door, sometimes on the ground. As a base for the new pot the neck of an old one is taken. Clay is taken in the hand, formed into a lump, and thinned with the fingers until it can be placed in the neck, then a roll of clay is taken and added with the fingers to the base thus prepared ; and the pot is completed by the same method throughout. The neck and lip are formed by methods illustrated in an article published in "Man," July, 1910. After completion the outside of the pot is smoothed, and after being left to dry and decorated with ornamented marks it is baked in the fire, which is usually made of bark, plantain stalks, or like material, and is simply laid over the pot to be fired. The process of firing takes only about half an hour ; the pots appear to be durable, the price varies according to the size, from three pence upwards.

Numerous types of baskets are made in various parts of the country, the largest being some four feet high ; they are used for carrying food products home, or to the market, for storing such things as beans, and for fish traps. Basketwork frames are also used for drying tobacco.

Many kinds of mats are made, some highly decorated with

23

coloured patterns ; they appear to be strong, and able to resist rough usage.

Except in Edo itself there appears to be little brasswork, though bracelets of brass, and in some parts broad brass collars, are not infrequent. Almost everywhere, however, blacksmiths are to be found, and with the rude materials at their disposal they appear to produce good work. At a few points in the hill country, more especially in the south centre of the Kaba province, smelting furnaces are found. Elsewhere iron is acquired from European sources. Hoes appear to be the objects most commonly produced, and the ordinary market price is ninepence. The blacksmith's forge is also much used for the repair of cutlasses and knives. Occasionally other objects, such as lamps and magical instruments, are produced.

24

SECTION II.

Eeligiox and Magic.

Religion. — Eeligion appears to play a considerable part in the lives and thoughts of the Eclo-speaking people. There is no house but has its household shrine and household worship, both of the so-called Ebo and of the ancestors. Each village or each quarter has its seasonal celebrations, and over all is a supreme deity, commonly called Osa or Osalobula. Osa receives no regular sacrifice in many cases, but he is far from being the ordinary type of the otiose creator, remote from mankind and indifferent to them, not only does he figure largely in their folk tales, but his name is constantly upon their lips, and his emblem, a long pole with white cloth on it, is to be seen in nearly every village. He is regarded as the creator of the world, and a myth is told in which Osanowa, or Osa of the house, has an evil counterpart, Osanoha, or Osa of the bush. Osanowa created man : Osanoha created animals. Osanoha also made a sickness house, in which were all diseases. When men and women came near this house, on their way from heaven to earth, rain fell and they entered