THE IGALA KINGDOM. WRITTEN BY JS BOSTON, NARRATED BY HRM LATE KING(ATA) ALIYU OBAJE Redocumented and uploaded by MAKOJI RICHARD son of EJURA OPALUWA A

ENVIRONMENTAL FACTORS

watermen and other people born to the water, somewhat disparagingly as amomonya, children of Igala describe specialist at Onya, which is a far-off town representing the extreme limit of Igala contact with the peoples inhabiting the Niger delta, Ecologically, the Igala live in an area of transition between the high forest conditions of the coastal belt and the drier and more open conditions of the savannah belt. The rainfall, averaging about 50 inches a year, is high enough to support the principal crops of the forest zone, such as yams, cocoyam, maize, pumpkins, and cassava. But conditions also favour the crops on which drier areas rely exclusively, namely millet, guineacorn, benniseed, and various types of beans. The transi-tion is also reflected in the tree crops that play an important part in the Igala diet and in womens' trading. These include the fruits of the oil palm and of moringia, which are both typical forest trees, together with the fruits of Prosopis cashews and Parkia, which reach their southern limit of distribution in Igala. Igala farming, which is carried out principally by the men, is based wherever possible on a combination of yams and maize sup-plemented by the other crops mentioned above. The farmers practise shifting cultivation, planting yams in the best land available and following this up with a rotation of maize and other crops when the initial fertility is lost and it becomes a necessity to shift the yam sequence to a new area, cleared from either secondary bush or from fallow land. The present boundaries of the Igala kingdom are defined by the configuration of the Igala Division, an administrative division of kogi state located around the middle belt geographical landscape Nigeria. These boun-daries do not coincide exactly with the traditional limits of the kingdom, which were slightly more extensive, but the differ-ence is not great, and the modern nucleus includes all but an insignificant minority of the Igala-speaking peoples known to reside in this area. In this study I identified the traditional kingdom with its modern configuration, which is justifiable if it is borne in mind that the Ata formerly exercised a loose form of suzerainty over Igbirra and Kakanda groups respectively. The success of yam cultivation in Igala depends on making skillful use of the dry season watercourses with their plentiful reserves of high forest and well-shaded land.The whole landscape is situated around the confluence where all under are the rule of the Ata igala , he also made appointments to titles, and in some cases exercised direct control, over many Ibo- and Idoma-speaking settlements that lie beyond the present boundary.The total area of the kingdom today is roughly 4,900 square miles, with a total population, in the 1952 census, of 361,119 persons. This gives an average density of seventy-four persons per square mile, which is slightly lower than comparable figures for the Idoma to the east, and much lower than the averages that occur among the neighbouring Ibo in the south and south-east. The low overall density of population, in a fertile and productive area, is a factor of great sociological significance. It can be associated with the system of residence in small, scattered, and impermanent hamlets which tradi-tionally obtained over much of Igala. In the southernmost areas, along the left bank of the Niger, annual flooding of low-lying territory restricts settlement to a few higher ridges, and the traditional pattern here is one of compact, large, and permanent villages. But elsewhere in Igala, and particularly in the central districts to the north and north-east of the capital, the typical pattern is one of small dispersed settlements, loosely grouped together in villages.The mobility of the traditional pattern of settlement in Igala cannot be over-emphasized, although this situation is changing rapidly today with the establishment of a permanent network of improved communications. In the past hamlets were often short-lived, with an average life span of from two to three generations, and in this period of their existence it was common for their composition to alter considerably owing to the abandonment of homesteads and the establishment of new living sites in other hamlets or villages. The question as to whether a householder should move or not is one that fre-quently occurs in oracle consultations, and moves are made as a reaction to political_oppression, to sickness and to mis-fortune of various kinds. The impermanence of settlements is The verb echado, to set up a new dwelling', is an extremely significant term in the Igala political vocabulary. The corresponding noun for an abandoned dwelling is blache. Constantly migration of acient antique diverse forms of art and creativity which influenced the rich cultural heritage of an ethnicity reflected in Igala material culture, which has developed few of the arts that seem to go with long continuity of settlement and prolonged occupation of the same set of dwellings. For instance, the principal altars and shrines used by the Igala are all capable of being uprooted and moved to a new environment. They comprise staffs and masquerades, repre-senting the dead, calabashes and pots to contain medicines llars and mounds which can and fetiches, and simple mud pillars be made anew and rededicated by a series of appropriate offerings if the householder decides to move. The influence of this pattern of dispersal and population movement on social life and on the social and political institu-tions of the Igala_is less_obvious. But it is probable that the characteristic emphases of the social and political system can be explained as part of the system of ecological adaptation. The balance of ties of descent against ties of kinship is related to this factor; similarly the existence of a dual system of administration, combining hereditary and non-hereditary modes of grouping in equal_proportions is ultimately an ex-pression of the underlying pattern of residence and settlement. The igala tend to validate knowledge in general by saying that it comes from the past, and their way of saying that a fact belongs to the widest order of human experience is to say that it was known to the ancestors of long ago, Within the category of knowledge so validated they distinguish ago, abogujo igbili. between narratives that record facts with certainty and, as it were, objectively, and those that embellish the truth by allowing fancy to play upon the sequence of events and accepted connexions between facts. Narratives in the first class are described by the term ita, whilst the second category is termed chiala or ahiaka. The former group includes proverbs, myths, legends, and other forms of inherited objective comment. The latter includes folk tales, fables, and a wide range of popular and imaginative stories. Ita are told for a serious purpose, however comic, their actual content may be. Ahiaka, on the other hand, exist to stimulate the imagination, and facts whose true nature is of cause and effective. An essential feature of all 'ita' is that they try to define the essential character of their subject, subordinating historical development to a statement of the subject's basic relations with the other major features of the world-in-which-it exists.and live. in One typical example ita explains the difference between the bushfowl and domestic fowl, by saying that the former grew tired of being used for for sacrifice, and decided to run away into the bush. In this way the two related species came to lead separate lives. But the bushfowl shows its relationship to the domestic fowl by returning daily, before daybreak to inquire whether fowls are still being used for sacrifice And the domestic cock replies to this by complaining that ma che ololoo. 'they are doing so, much too much. Within the general corpus of ita, the Igala distinguish clan histories as a separate category, ita olopu. They are distinguished in this way partly because they refer to the basic divisions of the social world in which the Igala live, but also because they co-ordinate oral tradition in a chronological as well as a structural sense. In so far as the Igala measure the exact sequence of past time they do so dynastically: cach clan chronicles the proof the past is past by the succession of generations of its own members. And these diverse traditions are co-ordinated and synchronized to some extent by reference to the dynastic history of the royal clan. The Igala identify the history of their nation with the history of the royal clan, and represent both its structure and the traditions of its origins and development by reference to the relationship of the royal clan with other leading descent groups that are representative of the widest divisions of Igala society.

Previous accounts of Igala oral tradition have used the Igala king lists as though they are basically equivalent to the linear time scale employed in European historiography, and they have attempted to translate the legends that record the origins and foundation of the ruling house by the concepts of sequence and historical development that are associated with this type of time scale. The difficulties that arise in this con-nexion, however, constitute major objections to the mode of analysis itself, and in order to show the necessity for a different approach I will summarize briefly some of the ways in which the oral traditions resist any attempt to translate them literally as an objective record of the actual succession of events in the past.

on other historical stories The principal objection to regarding the Igala king lists as an exact dynastic record is that their total time span does not coincide with the much longer span of time which seems reasonable for the history of the kingship points to a connexion between the Benin and Igala kingdoms grounds. There is much evidence, as I will show later, which in the first decades of the sixteenth century, and evidence of a still older cultural connexion with the Yoruba-speaking peoples suggests that kingship at Idah may be contempor-ancous with the period of the late Yoruba and early Benin kingdoms, which fall roughly between the thirteenth and sixteenth centüries. But the fullest known list of Igala kings cannot be made to cover a period of more than a few centuries. Clifford wrote of the immigration which is sup-posed to have established the dynasty: So began the regime of the Atas at. Idah. It is not possible to fix any reliable date for this event, but we shall not be very far wrong in assigning the colonization of the Agatu-Ocheku-Amara area to the early part of the Seventeenth Century, and Ayegba's arrival at Idah towards its close.'

From internal evidence of borrowing in other directions, it seems fairly certain that Clifford based this date on arguments advanced in an unpublished note by a previous administrator at Idah, who calculated the average time span of recent generations and extrapolated this backwards to arrive at a total of 250 years for the whole dynasty. This author wrote:

On the analogy of the past 68 years the rule of the Igala in Idah cannot have extended over a period of more than 250 years. Basing our calculations on the average period from genera-tion to generation at go years (a generous allowance), we get a total of 220 years back to Idoko. Numerous other pedigrees col-lected in various reports in Ankpa and Dekina show a maximum : of six generations taking us back to the earliest Attahs. It would appear therefore that the dynasty was founded about the end of the Seventeenth Century. At the present time, when approximately 30 years have elapsed since this calculation was made we are in a position to assess the average length of the last ten reigns, which span a period of roughly 120 years, from 1834 to 1956 ad. Applying full king list, which the average of 12 years per reign to the full contains-twenty-five-names, a time span of approximately 300 years emerges for the entire dynasty. Allowing for the difference in date between the two calculations, this second assessment is roughly in agreement with Former Attah igala, late Aliyu Obaje's estimate. The comparatively short period of an average reign can prob-ably be explained by the rotating system of succession, in which three other lineages hold the royal office in turn before the cycle is completed and a son succeeds his own father. Igala kings tend to be old, or comparatively mature, when they come to the throne, and the whole system favours relatively short periods in office. Attah Igala late Aliyu obaje, and other writers who have dealt with this problem of chronology, assume that the king list and associated tradi-tions refer only to the last two or three centuries, and that any earlier developments, antedating this period, fall largely out-side the scope of Igala oral tradition. This view, that the Igala retained no formal knowledge of their history before the seventeenth century, is essential to Clifford's argument that kingship was established at Idah by a migration from Wükari, in which a branch of the Jukun royal house was involved. He maintains that there was no earlier dynasty at Idah, but only a system of clan government in which the heads of the in-digenous families, or Igala Mela, participated. In Clifford's thesis Igala history literally begins with the kingship, and with he arrival of the immigrant rulers from Wukari. The early ancestors of the royal house are regarded as shadowy, but nevertheless essentially historical, figures who migrated by a well-defined_route through the north of Igala, stopping for two generations in the vicinity of Amagedde near 'the Benue river, before they finally moved across country to settle at Idah. This analysis, however, gives an entirely false impression on unanimity and agreement in Igala oral tradition, and mis-represents the perspective of Igala legends in order to'evade the analytical difficulties created by the inadequate chrono-logy of the king list and by the diversity that actually exists in the traditions. In a published study of these traditions I have tried to show that there are at least three different attributions of royal origin, existing side by side in the corpus of legends of old There was in those early days no form of central organization, the tribe consisting of a number of moieties each under its own patriarch or petty chieftain; these latter, nine in number, were the primitive fathers of Igala. The comparatively short period of an average reign can prob-ably be explained by the rotating system of succession, in which three other lineages hold the royal office in turn before the cycle is completed and a son succeeds his own father. Igala kings tend to be old, or comparatively mature, when they come to the throne, and the whole system favours relatively short periods in office. Clifford, and other writers who have dealt with this problem of chronology, assume that the king list and associated tradi-tions refer only to the last two or three centuries, and that any carlier developments, antedating this period, fall largely out-side the scope of Igala oral tradition. This view, that the Igala retained no formal knowledge of their history before the seventeenth century, is essential to Clifford's argument that kingship was established at Idah by a migration from Wükari, in which a branch of the Jukun royal house was involved. He maintains that there was no earlier dynasty at Idah, but only a system of clan government in which the heads of the in-digenous families, or Igala Mela, participated. In this thesis Igala history literally begins with the kingship, and with the arrival of the immigrant rulers from Wukari. The early ancestors of the royal house are regarded as shadowy, but nevertheless essentially historical, figures who migrated by a well-defined_route through the north of Igala, stopping for two generations in the vicinity of Amagedde near 'the Benue river, before they finally moved across country to settle at Idah.

This analysis, however, gives an entirely false impression o unanimity and agreement in Igala oral tradition, and mis-represents the perspective of Igala legends in order to'evade the analytical difficulties created by the inadequate chrono-logy of the king list and by the diversity that actually exists in the traditions. In a published study of these traditions I have tried to show that there are at least three different attributions of royal origin, existing side by side in the corpus of legends of

There was in those early days no form of central organization, the tribe consisting of a number of moieties each under its own patriarch or petty chieftain; these latter, nine in number, were the primitive fathers of Igala. Clifford's emphasis on the Jukun connexion overlooks earlier published evidence of Yoruba and Benin influence, and my own research has confirmed the existence of traditions of Benin origin within the ruling house itself. Clifford's bias towards the Jukun attribution is clearly related to the interest aroused by Meck's study of the Jukun kingdom, and to Meek's support of the hypothesis that kingship diffused into Nigeria from Hamitic sources along the valley of the Benue river. It cannot be too strongly emphasized that the igala hold divergent views about the origin of their kingship, and that this divergence occurs within the dynastic framework that Clifford related solely to the period of Jukun contact. In other words the king list and associated traditions represent a total view of Igala history, covering contacts with other peoples and historical developments in the remote past as well as in more recent times. The fact that traditions of Jukun, Yoruba, and Bernin origin can exist in the same corpus of legends shows -that the king list is not bound by the conventions of sequence that occur in a linear time scale. Igala oral tradition clearly takes a synoptic view of the past, selecting important events from different periods and bringing them within a narrower compass of past time. And the fact that these different traditions coexist with one another, as alternatives, is also highly significant. It suggests that the perspective imposed by attempts to translate the material into a strictly historical sequence is largely artificial, and that the oral traditions have a perspective of their own which has not been fully explored. 1 suggest that the early part of the dynastic record, down to and including the period of political foundation represented by the period of Ayegba om Idoko, is largely mythological ingnificance and that the early ancestors are not to be regarded as historical figures in a literal series, but rather as mythical archetypes of structural arrangements that must-have taken a longer period-of-time-to-evolve than is suggested by the idiom of associating events with single reigns. The effect of regarding the king lists as a substitute for written history is to overlook the mythological content of Igala tradition and ultimately to misrepresent the perspective in which the Igala see the past. If the traditions are seen in their true perspective it at once appears that the mythological phase of the past is of the past mount importance, and that historical function is correspond-ingly subordinate to the political function of the myths con-tained in these traditions. The mythical character of the early period of the Igala kingship is reflected in uncertainty about the genealogical status of the first royal ancestors. Genealogically, the apical ancestor of the royal clan is the fifth in line of descent, and there are four Atas, including one woman, before Ayegba om. Idoko, to whom most branches of the royal clan trace their descent. Ayagba's forbears are:

Abutu Eje

Agenapoje

Ebelejonu

Idoko

Idoko immediately precedes Ayegba, and is fixed in this position by the fact that Ayagba's name is invariably given as Ayagba om Idoko, son of Idoko. But the relationship of the other three ancestors is uncertain, in the sense that there are different versions of their connexion, amounting virtually to a complete series of variations on the genealogical positions of these three names. These early ancestors are both structurally and historically protodynastic, they do not belong to the main pedigree of the ruling class and they also represent a different order of time. Ayagba and his father are transitional figures, myth with regard to their achievements and the com pression of historical events, and historical in the sense that they mediate between the period of creation and the later period of development in which events occur within the dimension of historical time.

The protodynastic ancestors tend to be represented in kingship ritual by a single figure or by a single symbolic cult object. In Ojaina, the royal burial ground at Idah, there are only twenty-two royal graves, and the grave of Ebelejonu is symbolically equivalent to the graves of Agenapoje, Abutu Eje, and Idoko, who are not represented by individual graves. There-is, similarly, a single ritual staff, Otutubatu, which re-the protodynastic royal ancestor cult; when sacrifices are made their names are their presence is manifest in the ritual staff, which is placed not called individually as are those of the dynastic kings, but to one side of the other ancestral staffs and first receives offerings. We can relate to this collective ritual role the tendency in oral tradition to regard these early reigns as belonging to the same dimension in time. The protodynastic ancestors are not sharply differentiated from each other, and atis frequently not clear in which of the early reigns the events described are supposed to have occurred. They represent col-lectively the age before Ayagba, and the emphasis on their common function may account for the blurred outlines and indistinct sequence of the mythical epoch.

The transfer of sovereignty ?

The myths that describe the protodynastic epoch place ends in a sequence. that is governed by rules of logical development farther than by objective historical conventions. There are three principal stages which in a sense lead on from one another but nevertheless fit together awkwardly in a his-torical sequetice, because of the omission of intervening stages and also because the historical details given in the legends are specific to each phase and do not always relate exactly to the details of the other phases. The argument proceeds logically rather than historically, and on the whole its development bears out Professor Levi-Strauss' suggestion that the division and symbolism of myths are frequently based on a procedure of modifying ideas that are initially predicated as opposites. It will become clear to what extent this idea is useful in the context of Igala tradition if we analyse the principal myths separately and consider the inter-relationship of their symbolism.

The first phase of Igala protodynastic myth is contained in different legends describing the transfer of sovereignty from the autochthonous population to an immigrant group of royal or noble descent. These legends vary with regard to the tribe or centre to which they attribute the origin of the immigrants. But they coincide in stating that the immigrant founder of the

The staff, Otutubatu and the grave of Ebelejonu are correspondingly more important than the other staffs and graves reserved to single rulers. Otutubatu is cared for by the senior priest, Atębo, and has its own shrine. Igala royal house was himself of royal origin, the section of an elder and senior royal line in another kingdom. The migra-ion to and arrival in Igala may be described in detail, with supporting circumstantial evidence of a geographical char-acter. My account of the migration from Wukari is-a good instance of this, and there is a counterpart, making a different attribution, in the legend of Benin origin that forms part of the inner tradition of the royal house. But this migra-ion can also be represented symbolically, and there is an important/myth of totemic identity between the royal clan and the leopard which avoids the difficulties of historical attribu-tion by attributing the foundation of the royal clan to a union between an anthropomorphic leopard and a woman of the autochthonous group. Mockler-Ferryman gives a version of this myth in which the royal baby was exposed by its mother and then found and fostered by the leopard until it became old enough to re-enter human society. The numerous other references that occur in unpublished and published litera-ture to a 'man who came out of the bush' probably also refer to this myth, although they often graft the purely mythical illusion on to the stem of legends belonging to later phases. Monsell writes, for example:

Ebelejonu who became the first Atta of Idah was grinding corn one day when a handsome youth came out of the bush. She asked him to come and help her, and afterwards took him to her house, bathed and fed him, and ultimately married him became the first Achadu.

The full text of the royal version of the leopard myth, as told to me by one of the elders of the ruling subclan at Idah, is as follows:

The first rulers had a daughter who went daily from the capital to collect firewood in the grove that is now called Ojaina. As she was visiting this grove she met a leopard there who took the form of a young man. He made advances to her and was accepted, so the young girl was going daily to meet-her-husband in the bush He killed game for her and made presents of bushcow and other animals to her parents who began to be curious about the mysterious husband and asked repeatedly to see him. The girl told him of this the leopard promised to appear and fixed a time at which he would show himself to his inlaws. But when the leopard came out of the bush the girl's parents ran away in terror. So the leopard ran and hid himself again in the thick bush at Ojaina and went into the ground there. His wife later delivered a child, Abutu Eje, who was the founder of the royal clan. The Ojaina grave is the spiritual centre of the royal clan and forms the last resting place of all dead Atas.

The theme of these different legends of royal origin is that the foundations of the Igala state were laid by the transfer of sovereignty from a group of clans representing the local population to an immigrant lineage of royal descent. This explanation is only partially, or incompletely, historical in character. As we have seen, the historical aspect is subordinate to the mythical function to the extent that different historical attributions exist as alternatives to one another, and the central idea, of a royal migration, can be expressed in a purely mythological form by using the idiom of ritual unity implicit in the totemic attribution. The function of these legends as a form of myth is to contrast the political implications of royal descent with the principle of non-royal descent. The transfer of sovereignty expresses the royal group's inborn right to rule and fulfils the destiny of the migrant prince created for him by the fact of birth to an older royal line. Other more explicit reasons are added for this transfer of power. In Mockler-Ferryman's version of the leopard myth, when the leopard's foster-child emerged from the bush his first public act was to settle a dispute between two men who were fighting. The legend says

He at once took upon himself the duties of arbiter, rebuking the one and commending the other. So astonished were the people who had during the incident crowded around him that they immediately proclaimed him their King, and refused to permit him to leave the town. This was the first Atta...

Another, unpublished, version of the king's arrival says:

The Okpoto were much struck by the skill of the newcomers in hunting and also by their knowledge of the art of poisoning arrows, in which up to then they were not very expert. Ayegba om Idoko, moreover seemed to be possessed of a very powerful medicine, probably taught him by his father, who was a noted medicine man.

A third explanation says:

The story about this is that a man called Aiyagba or Ajagba, came from, a far country called Apa. He settled in Idah and married an Okpoto woman. He was successful in war and became chief of the whole tribes and country."

These stories embellish the main theme that the immigrant stranger was of royal blood, and if we survey the whole corpus of tradition on this subject it is clear that the principal reason given for the transfer of sovereignty is the personal nobility of the immigrant, reflecting the high status of his own forbears. In other words the first origin myth is concerned wholly with hereditary status, and with the different implications, in terms of political function, between royals and non-royals. This image of a polar relationship between the two classes, in which one is permanently super-ordinate and the other per-manently inferior, reflects a basic division of political func-tions between clans in the Igala political system. I analyse this difference fully in the later sections of of this this study, and need only say here that in this political system rights of political sovereignty, in the widest geographical sense, are vested in large-scale clans of high rank, whilst rights of local sovereignty are vested in small-scale, localized clans who are often re-garded as being the 'landowners' of the areas in which they are settled. The myth represents this basic division of functions and attributes its origin to the introduction of notions of aristocratic rule from other kingdoms with whom the Igala have been in historical contact. The peaceful nature of the transfer of sovereignty is signifi-cant, since it implies a high degree of continuity from the older system and also stresses the complementary nature of the relationship between the two classes of descent group involved. At this stage of the myths the political functions of the in-digenous clans after the transfer of power are not defined pre-cisely. But it is implied that the king co-ordinates their relations without modifying the earlier structure. Furthermore, the myth leaves open the question of relations between the two groups in future generations; it is clear from the structure of the situation that the king's sons inherit a right to rule, but it as also clear that the descendants of the landowning clans inherit something of their forefathers' position as kingmakers.

The Achadu and the kingmakers

Whereas the first phase of these origin myths is not associated with any particular reign in the protodynastic period, the second phase is usually placed in the reign of the female Ata, Ebelejonu. This female figure plays a key role in the legends that describe this phase of development, mediating a marriage alliance with the non-royal elements of the Igala population and reversing the direction of the alliance recorded in the first myth, so that the royal clan, from being a wife-receiving group, reaches the position of giving its own women to the non-royal

descent groups that make up the bulk of the population. A version of this legend told in the royal subclan at Idah says:

When Ebelejonu became Ata she was a yoling girl and had no husband. She was fond of festivals where there was singing and dancing and one day went to a celebration arranged by the Igala Mela clans at Igalogwa, in their own section of the capital. At this festival she saw a slave belonging to the Igala Mela, who was of strikingly handsome appearance. He was a man of Ibo origin who had come to the Igala area for hunting, had lost his way in the bush, and had been found and taken as a domestic slave by

the Igala Mela, in whose territory he was hunting. Ebelejonu-was-so-much-attracted by this person that she had him transferred to her own service, and eventually made him her consort. The Igala Mela and her own followers were jealous of the husband's position and began to abuse him in her hearing. But she retorted that apart from the slave, achai adu, there should none like her own. This description was incorporated no one with power in the title of Achadu which the Ata then bestowed on her console It was also determined that this title would be hereditary in future, and that the Achadu would control all the land lying beyond the

to the east of Idah as Ebelejonu died childless, and the kingship passed to her brother Agenapoje. But the Ata is still described as oy' Achadu, the Achadu's wife, in memory of this early marriage, and this is also why Igala kings have their ears pierced like a woman's,

In all probability the historical events alluded to in this myth are compressed in time, so that developments which may have spanned several generations are assigned to a single reign. The rise of the Achadu's clan to a position of dominance within the federation of kingmakers is described actiologically rather than historically. Similarly, the brief references to contact with the Ibo and to the Achadu's connection with the trans-Anambra region condense, a wide range of historical associations covering a long period of Igala expansion into Ibo territory in this eastern sector of the Igala-ibo border. We can conclude from this synoptic treatment of historical material that is well-known to the Igala in other contexts, that the function of this tradition is to summarize the historical trends for a mythological purpose.

In this phase of the mythical epoch both the elements of the previous stage are present, the immigrant royal/clan being represented by a female ruler, and the indigenous clans being identified explicitly with the Igala Mela. There is also a third element, however, in relation to which these other two ele-ments, previously opposed to one another, symbolically merge and form a unity. In the first myths the principle of opposition is the distinction between royal and non-royal descent. But in this new phase of development descent itself is contrasted.

 In either case sructural change comes about through the realign-ment of lineage genealogies so that they either meet in the person of the hunter, or become coeval with the foundation of the hunter's own lineage. In the case of the Achadu, both pro-cesses occur. There are many lineages in the trans-Anambra area that claim direct descent from Omeppa and are thus full members of the Achttelu's clan. The Igala Mela, on the other hand, have traditions of separate foundation, and claim to be older in point of time. But their genealogies are in fact no greater in depths than that of the Achadu's clan. And in fact also the succession to generations within the Igala Mela genealogies tends to be synchronized with the pattern set by the Achadu's lineages so we can regard the former as being to some extent structurally dependent upon the latter.

The relationship indicated by this myth between the prin-ciples of inherited and achieved status is ultimately one of superordination and subordination. In the legend the Achadu owes his elevation to royal patronage, and similarly in many hunter legends the founder consolidates his achievement by marrying one of the king's daughters and obtaining the grant of a title But this ideal ascendancy of the royal group is modified by the role assigned to their representative in this myth as of Achadu. The king, as 'wife' to the head of this clan, is in some respects in a subordinate position The Achadu's authority over the king is a function of his duties as head of the kingmakers, and is expressed again in the ritual of accens-sion when the Achadu beats and abuses his 'wife', and sends the candidate for the throne to his own senior wife for the cere-mony of piercing the ears. The myth uses the idiom of domestic kinship relations to describe this authority, and in the same way that some of the other kingmakers are described, in the context of accession as the 'father' and 'mother' of the king. I will show later that the division of powers in government between the royal clan and the federation of kingmakers tends to be concentrated on the opposed offices of Ata and Achadu. This concentration and opposition is symbolically expressed often of common the bond between royal clans in the myth by the idiom of a wife-husband relationship. And the associations of these positions counterbalance those that are implicit in the description of the Achadu as a slave and hunter who was ennobled by the favour of the Ata.

Ayegba om Idoko

The third, and concluding phase, of Igala mythology des-cribes the achievements of the apical ancestor of the royal clan, and identifies his contribution to the kingship with the emergence of an autonomous and fully organized státe. Ayagba's reign forms a reference point for most of the office holding clans in Igala, from which they measure their position in the state. These clans sum up and conceptualize their functions in the state system by the mythical relationship that existed between Ayagba and their own founders. Each of these clan traditions makes a separate contribution to the Ayagba myth, and a correspondingly wide survey would be necessary to do justice to its importance in Igala tradition. But some of these secondary sources will be considered later, and it is sufficient here to consider the main outline of Ayagba's reign, about which most of the traditions agree. This divides natur-ally into three major series of events, corresponding to the stages by which the kingdom achieved its independence from alien rule. I give below versions of these legends told by a member of the royal house at Idah: they agree closely with similar legends told in other parts of Igala. The first legend tells the story of Ayagba's accession and of the way in which he provoked a conflict with the Jukun. The Ata had been paying tribute to the Jukun from a remote period. The Jukun came here and made war so that we would give them something. When the Igala grew tired they decided to pay tribute to avoid further war, and went on paying until Ayağba's time. "Idoko was ruling at that time and Ayagba, his son, was being trained by a palace eunuch with the title of Enefola; Enefola looked after Ayagba until he grew up, and taught him to prepare medicines. Ayegba became powerful and made medicines that made him stand out and prowessed furthermore remarkably outstanding successful hunter.

Amana Edime.

A remarkably similar version of this legend was given me by an elder from Akpodo, about 20 miles north of Idah. were successfül. God showed him the way in this. Then Ayagba said that at the end of the year he would go to the Jukun himself. His father approved, being sick at the time, and authorized Ayagba, to to take the tribute on his behalf. Ayagba took dried excrement and put it into nine calabashes of the kind that they use today in festivals, he put it inside them and covered it over. Then he bought tobacco wrapped up in the ancient manner and got that ready. He travelled with these loads, escorted by palace servants, and on arriving in the Jükun country and that he would rest and present the tribute in the morning. So he was shown somewhere to stay. After Ayagba had gone in, the son of the Jukun chief came to look for Ayagba, but found only a young girl there. He greeted her and asked if she was Ayagba's wife, but the girl replied that she was Ayagba's younger sister. The prince said that he found her attractive and would ask Ayagba for her when he came back. This prince then made gifts of food to the girl who said that she needed a special ritual vessel, ane, before she could eat. The prince replied that he would bring his own father's vessel if she would agree to spend the night with him. She consented, so the chiefs son stole the ane after his father had finished eating and brought it to the girl, not knowing that she was Ayagba in another form, That night, when the time came for the girl to go to the prince, he fell asleep, and Ayagba woke his followers telling them to make, haste because he had found what he came for. They ran, and were on the road next day when the Jukun rose and began to call Ayagba When they failed to find him, the Jukun decided to uncover the tribute without waiting any longer. They uncovered the calabashes and pulled out a piece of tobacco which they gave to the king, situng on his box stool. The king smoked it and swooned; his people ran to support him, but he did not revive. So the Jukun looked up and down for Ayagba but did not find him. They sent a messenger to Idah, who returned to say that Ayagba had not returned, but that the old king, Idoko, was dead. The Jukun asked-about-the-succession and were enraged when they heard that Ayagba was the there. They sand that war would reach there first, and that they would destroy the city of Idah.

 Significant comments can be dealt with fully here. It is an Igala conception of royal power that Ayagba inherits a funda-mentally weak position from his ancestors. The notion that stances is a constant theme of kingship tradition in this society the king has to work within the framework of given circumstances and this legend is one of many parables which show that political solutions are achieved by intrigue and cunning rather than by the application of force. It is a typical of this attitude that the symbolic transfer of sovereignty in which the ane vessel stands for the king's rights over the land, is accomplished by a strategy involving the use of medicines. In this context the Jukun overlords represent a whole complex of external forces with which the Igala king has to deal, and possibly also symbolize the opposition of physical force to the qualities of rank inherent in the Igala royal clan.

The second phase of this myth follows on chronologically from the first legend and describes the defeat of the Jukun in their punitive campaign against Ayagba. This legend has be-come well-known in the literature of Igala and is recorded by both Seton and clifford I give below Seton's version, which condenses the essential features and differs only in a few details from Ata ali Obaje's account and from other versions that are told in Igala today.

Inikpi was the daughter of Ayagba Om'Idoko. She is reputed to have been very beautiful and of a noble disposition and her father loved her more than anything else in the world, and she him.

In' Ayagbas reign the Jukuns under Appah attacked Idah, and he were unable to make any headway against them. He asked a Certain learned Nupe, Mallam, what he could do to change the fortune of the war: He replied: 'If you do not wish to lose both your title and your land you must sacrifice the daughter you love so much to the spirits of this place.'

When Ayagba heard this he was overcome with grief, and ap-peared prepared to lose all that he had rather than carry out the sagrifice. She, however, heard what the Mallam had said, and went to her father and begged him to save himself by sacrificing her. She is said to have gone nine times before he consented.

into it with nine slaves and with all her jewels and charms. She called to the people above to throw in the earth. This was quickly A large hole was dug in the market place, and she went down done. She was unmarried at the time of her death....

After her death the Mallam supplied the Attah with some charms, which were thrown into the river Nasallu (near Idah). The Jukuns, who were camped on the far bank, ate the fish taken from the river, and many of them died. The remainder were Scattered by Ayagba and his followers.

This version does not make it clear, as do texts in the original language that both the sacrifice and employment of the Nupe magician were advised by the Ifa oracle. Nor does it reveal the important fact that the oracle specified that the sacrifice should be made to the land. Igala versions of the immolation say that Ayagba offered his daughter to the land. i momanw du nyane

This central incident again reflects the myths' constant preoccupation with the ritual relationship existing between the king and the land. This connexion is shown to be the one dimension in which the king's powers are incomplete, and in which the royal clan cannot fulfil its destiny to rule without external assistance. The sacrifice of Ayagba's daughter shows this incompleteness in an extreme form; the king is in imminent danger of defeat, and his failures are ascribed to the 'spoiling of the land', ane ekpabię. Other versions of this Inikpi legend emphasize the threatened extinction of the dynasty by saying that Ayagba had no male children, and that the sacrifice of his daughter was made to ensure the birth of male heirs to continue the line, Both interprètations attribute this situation to the failure of the king's duties towards the land, and they make the survival of the dynasty conditional upon the restoration of this relationship.

This myth, in other words, restates the theme of the initial phase of the protodynastic period, which is also expressed in-directly in the legend of Ayagba's quest for the ane symbol of the Jukun king. The Ata's powers embrace every aspect of government with the exception of ritual control of the land. The royal clan's control of this aspect of the state is variously represented as being due to a voluntary transfer of sovereignty, the theft of an emblem of sovereignty from the Jukun, and, finally, to a sacrifice in which the Ata gave up his most valued possession. The theme common to these different formulations is that sovereignty over the land exists, as it were, apart from the royal clan, and is not part of the hereditary qualities which are transmitted by descent and form the birthright of each member of the royal group. This conceptual distinction corresponds, as I suggested earlier, to the division in political functions between clans of high and low rank in the Igala, and is the basis of the opposition between royals and non-royals. The full associations of ritual responsibility for the land will become clearer as I describe the detailed working of the political system, and it will then also be clear how these mythical references to the land-cult symbolize a basic aspect of Igala political organization.

The myth of the last phase of Ayagba's reign is the common property of the many clans that trace their foundation to this period and begin their history with an account of their founder's connexion with Ayagba. Perhaps the most important theme that emerges from their different accounts of this epoch is the role of Ayagba as the creator of the system of hereditary titles on which the Igala political organization depends. These titles are usually described as rewards given by Ayagba to the clan founders for their assistance in the Jukun war, for specialized services that have become hereditary within the clan, or, finally, to mark some special kinship connexion with the founder of a particular clan. In many cases the award of a title by Ayagba forms the charter by which immigrant clans justify their arrival and presence in Igala, and by which, at the same time, they demonstrate their assimilation into the state system and into the pattern of Igala culture. A majority of the clans living along the left bank of the Niger fall into this category, and hold titled offices whose duties are concerned with riverain affairs and with contact between the capital and the other nations using this riverain highway. In this instance, the compression of the historical dimension into a single transaction between the king and the clan founder emphasizes the equity of the groups concerned into the Igala system. The historical difference between these with other clans, and the completeness of their assimilation focused upon differences of hereditary political function, and immigrant groups and the indigenous inhabitants of Igala is focused upon differences of hereditary political function, and does not imply the notion of cultural difference or of qualified membership of the political community. In other cases the award of a title redefines an older relationship between the royal clan and an indigenous group, crystallizing functions that are implicit in the earlier traditions and giving them a constitutional form. This is especially true of the kingmakers, who in the earlier mythical phase are identified with the kingmaking role of the autochthonous population, and have authority over the Ata, but do not have precise political functions of the kind associated with the award of a title. Ayegba, after his victory over the Jukun, included the Igala Mela in his distribution of titles: this is recorded in the traditions of the Igala Mela themselves, and is also made ex-plicit by the name of one of the titles, Agbenyo, which is popularly said to mean that the holder forgot, egbenyo, to allocate land to his own clan while he was appointed by Ayegba to fix the boundaries of the Igala Mela territories.

In both these sets of tradition the contractual element of the relationship between the king and his office holders is pronounced. The effect of placing these political connexions in the mythical period of time is to give them permanence and make them unalterable in principle. They become the subject of rituals that are performed annually to renew the ideal con-nexion between the ruling house and subordinate clans. And the rite sets the seal on a perpetual alliance between two descent groups, in which each has its part to play in maintaining a wider structure to which they both belong. One of the functions of historical compression here is to emphasize the contractual nature of the relationship, by balancing the king's inherent superiority directly against his dependence upon the specialist services of others and upon the nexus of kinship ties that form an inevitable corollary of the descent principle. This series of formal contracts that make up the title system also resolve the problem of the king's relationship with the land which was posed by the earlier myths. In awarding titles to his followers Ayagba is said to have defined the boundaries of clan land in all parts of the kingdom, and so to have fixed the relative position of the major clan centres for all time. The significant feature of this distribution of fiefs is that they were awarded principally to non-royal clans, who exercise the rights of local sovereignty that are implicit in the roles of the non-royal clans.

Rights are limited by this distribution to rights of overlord royal population in the earlier myths. Ayagba's sovereign ship, entailing a share of the non-royal clans' perquisites but intermediacy of the non-royal clan heads. Within the context curtailing the king's participation in local government by the means of a hereditary system of government this in turn limits the rights of the royal clan as a whole, and defines their position in relation to the local landowning clans to whom they are superior in rank. These are the basic principles of structure whose working out dominated traditional politics in Igala. I have tried to show that oral tradition presents these principles, provides a model of the political system, by using the names and achievements of different groups of clan founders archetypes. To some extent these archetypes in the manner that Professor Levi Strauss has suggested is typical of the procedure of mythical thought. They form a series of polar types or complementary extremes, which are progressively mediated by the introduction of new terms and new contrasting pairs. and so as are presented.

The logic behind this system of oppositions is probably an historical one and does not obey mechanical laws of philosophy or rules of logical necessity. It probably reflects a type of political evolution in which kingship emerged in a system where discrete clan units were associated with definite tracts of territory over which they held sovereign rights. It would be interesting, in this respect, to compare the Igala system with that of the neighbouring Idoma, where there is no traditional system of kingship. but where the clans are organized on the basis of their local landowning rights. The Igala concept of the land chief, which I describe later, and which underlies the whole concept of royal power, is closely paralleled by the Idoma concept of the land and there are also obvious affinities with the political concepts-of-many other West African peoples who have no system of kingship but nevertheless define clan functions in terms of land ownership

the administrative system by which the Igala are governed at the present time, territorial considerations are paramount, and the division of the kingdom into districts and village areas would form a natural starting point for any analysis of the modern system of government. But it is clear from the form taken by Igala oral tradition that in the past the hereditary aspect of political authority was the major focus of interest, and that the clan system was seen as providing the basic framework of government and administration. This dif-ference is one of emphasis and does not involve a conflict of principle. Igala clans are usually localized, and the rights that they possess in land make up a major portion of their corporate functions. But although the territorial sphere formed a majer dimension of clan consciousness and clan inter-relations, the clan system existed in its own right as a frame-work within which political authority was delegated perma-nently among clans of different orders. And the primacy of this mode of organization over the territorial distribution of power was such that spatial relationships were more fluid in the tradi-tional than in the modern system and the geographical balance of power was correspondingly more dynamic than is possible in an administration, tied to a rigid system of terri-torial divisions. I have tried to follow this bias and bring the main emphases of the traditional system to life by considering the hereditary and the territorial aspects of government in separate sections, and by giving priority to the workings of the chan system, divorced as far as possible from its territorial foundation.

Agnatic links are followed in Igala to determine succession

The village areas of the modern network correspond to the districts, ane, of the traditional system, whilst the modern Districts are closely linked to the traditional spheres of influence of various royal subclans. sub to hereditary political office and to the many different statuses involved in the inheritance of rights over persons and pro-perty. Hereditary rights of this order are vested in patrilineal groups, olopu, whose members exercise jurisdiction over succession and inheritance. These are the groups around which to which much of Igala life is still orientated at the present public and political life revolved in the traditional system, and governed to a considerable extent by the basic fact of clan time. In this society a man's status and expectations are still membership, inherited from the father's side, and in the tradi-tional system clanship was even more important as a determinant of status. Olopu are unilineal descent groups in the sense that their members are related to each other agnatically and share cer-tain rights in common that can only be transmitted through the male line. But they are unilineal groups of clan, rather than of lineage type. Exact genealogical relationships are important over only a limited span in time, usually not more than three or four generations, and beyond this connexions with or through earlier generations tend to be abridged by naming the founder of the group.

When genealogies are collected from clan members it is common to find a basic uncertainty about the actual relation-ship above the third or fourth ascending generation. For in-stance, in a sample of six clans among the kingmakers, the Igala Mela, all the clan heads knew several names of previous office holders, but only three placed the series in exact genea--logical relationships. The others explained that the first three or four men named in each case were patrilineally related, omenekelę, to each other, but that the details were unknown. Further inquiries into one of these imperfect genealogies had the same negative results, producing a shallow genealogy with a very wide span and a mainly putative connexion between the founder and the first-generation. One informant living 20 miles from the capital was in frequent communication with the clan head at Idah, and went to him for advice and assistance on a number of family problems. But their mutual rela-tionship could not be established in detail; it went back to two ancestors in the third ascending generation who were known to be patrilineally related and beyond this it was stated that 'we do not know who their progenitors were', a mene ki bi ma gen.

In tracing agnatic relationship there are a number of conventional formulae which make it unnecessary to establish exact genealogical connexions. The procedure is simplified in the first place by concentrating on sibling relationships, in which the persons named usually belong to the same generation or are brought into line with other persons who do. The discussion then proceeds by steps from one generation to the next, each pair of names being oji ego or equal in generation. When no more names are known, or when it is wished to shorten the discussion the formula for sibling relationship is used and the last persons named are described as omata, 'children of one father'. This term is essentially a category of relationship and is not descriptive as the literal translation implies. It assimilates the relationship between siblings to the connexion between children. Intervening generations can be disregarded, provided the links are agnatic, and two members of any subsequent generations can be regarded as representing ekoji, the children, ekojiomia, of the original founder. The phrase ata ka titi bi ma, 'one father begot them', has the same significance, meaning that the connexion is an agnatic one and that the number of generations involved is irrelevant. If we represent this term diagrammatically it will show more clearly how the step from the founder of a clan to the first generation is often made.

The putative nature of the first level of clan genealogies also appears in the use of title names for clan founders instead of their own personal names. In the majority of clans, the office of clan head is combined with a political title, ofe, having its own name and salutation. On appointment to the title the holder virtually gives up the use of his personal name and is known by the name of his office. As ancestors in the clan genealogy, former titleholders are known by their personal names again, together with any cognomens that may have become associated with them. But the clan founder can also be described, in the genealogies, by his clan title, and in many cases no personal name is given, but only that of the title. For instance, in the imperfect genealogy referred to above the founder's name was given as Aleji. But this is the name of that the clan beads held in succession from the Ata, and ake of the chart itself. It this instance the founder is purely a figurehead, personifying generations of early ancestors whose names have been forgotten. Forgotten is perhaps the wrong term for this process of genealogical adjustment; the shallow depth of the genealogies is related to a rotating system of succession which I describe later on, and it is mainly because they are irrelevant to this structure that the names of early ancestors can be discarded or subsumed under a class name.

Chns have individual names, which are often synonymous with the names of political titles. There are also individual clan sabatations, used in polite address by non-members, and varying according to the sex of the person addressed. In titled clans, salutations for men usually take the form of a diminuive of the title salutation or of the clan name. This is formed by prefixing the term oma, child, to either of these names, as in omata, 'royal clansman', which uses the name of the royal title, or as in omanu, 'Achadu's clansman', which makes use of the Achadu's title greeting. In practice clan salutations are more commonly used towards women than men. In addressing, men salutations can be borrowed loosely from a set of terms that properly denotes seniority within the clan. Thus the term otalu strictly applies to the first-born son of the speaker's father, or, more generally, the eldest surviving member of one's own patrilineal clan segment. Okolo is the correct form of address for the second son or for the next in line to the head of the clan segment. Wodi and doga are the corresponding terms for the third and fourth sons respectively, or for the men in those positions in the order of clan seniority. The first two terms in this set are frequently used by non-members of the clan to acknowledge greater seniority or higher status in the person addressed, and in many contexts it may be more respectful to address a man as doga* than to use his correct clan greeting. In addressing women, there are two terms which denote seniority in relation to the speaker. Iye is used for a woman a generation above one's own, and ouja for a, senior member of one's own age-group. But these terms have kinship associations through being used as familiar terms of address among women in the same local community. And it is often more respectful to address a woman by her clan salutation than by the term based on seniority, the reverse of the pattern that obtains in the salutations used for men. Relative seniority is a major constituent of clan relationships, and in all Igala clans one of the chief functions of clanship is the creation of a pattern of authority based on descent. Seniority is also important in other social contexts, but outside the context of clanship or of clan interrelations it does not have the same jural significance. And in many kinship contexts, as we shall see, the norms of behaviour between blood kin are patterned by contrasting their relationship with the formal relations expressing seniority that characterize the narrower field of agnatic relationships within the clan. The most general distinction of seniority within the clan opposes the elders, abogujo, to the age-groups of younger men, men on abokolobia, and places these in turn above the lowest class of juniors or young children, abimoto. Within this broad classification of distinction are extremely conscious of this distinction and classify the basis of their being oji ego, equal in age, with one another The classification is coordinated roughly throughout Igala by reference to the sequence of generations in the royal house. In 1960, for instance, the Igala regarded the Ata Amaga's genera-tion as being almost extinct, and accorded great respect to anyone who was generally accepted as belonging to this group. Conversely, they used Atabo's reign as a dividing line between groups that might have reached the age of discretion and those who, by definition, were too young to have any under-standing of adult morals and the conduct of public life.

This distinction of seniority between generations vests the authority of the clan in a hierarchy of sibling groups, within which each group is free to manage its own affairs but is nominally subordinate to the senior groups above it. Within the sibling groups, again, relative seniority is distinguished, and an internal hierarchy of authority emerges in any context that brings clansmen of the same generation together. This internal distinction is nominally based on age, and is upheld by the use of kinship terms that denote relative seniority within the same age-group, achogwa for senior sibling and okikili for a junior person. But this criterion is affected by a number of other factors, including wealth, size and status of family, maternal and other cognatic kin ties, lineal closeness to office holders within the clan, personal ability, and so on. The hierarchy that emerges in a given situation is delicately poised in spite of its apparently solid foundations. And it may be the fictional and dynamic nature of this principle that accounts for the tensions and constraint that characterize sibling relationships within the clan. Here a pattern of authority that is outwardly stable and fixed in its order, is in fact competitive and open to change. the most senior member of the uppermost age-group that has the logical outcome of this hierarchical arrangement is that theres authority over the whole clan. The Igala recognize this by making the position into an important clan office, of And they define the pattern of succession to this role by saying that when the ogijo olopu dies the person who is next in enter of seniority should return home to the clan centre, from wherever he may be residing, in order to look after the affairs of the clan, Egba koguja olopu lekuru ene ki charononw a liwunya todu ka a de olopunw

THE OJO CLAN

In the Ojo clan, centred on Igebije, some 25 miles from the capital, one branch of the clan had provided two or three ogijo olopu în succession. But after Okeme's death, his relatives sent for an elder, Idakwoji, living in the Ibaji district about 40 miles away. Idakwoji later quarrelled with Okęme's branch of the clan and gave up the post, so another elder, Odaudu, was invited to return to Igebije from a neighbouring district to become the ogujo olopu in his place. When Odaudu died the office passed to Atoloko, who had moved to Igebije during Odaudu's lifetime. The authority exercised by the ogujo olopu within the clan

"When the senior member of the family dies the next in age should return home to look after the family.

This dispute is described more fully later moral and rinial in character and the office is important not so much administratively as forming a symbol of clan sorveignty, Its holder mediates between the living and the dead members of the clan by wanding next in order of seniority to the ancestors themselves, and his home is an important centre for the ancestral cult carried on within the clan. Arbitration in disputes is often referred to this ritual function, by the ogijo olopu saying for instance that he intervenes 'so that the ancestors will not be angry. The Ifa oracle, which is consulted in all serious misfortunes, often gives its verdicts in the same terms. A case was recorded for instance in which a son who had quarrelled with his father made sacrifices through the ogijo olopu to the clan ancestors who were said by Ifa to be incensed and to be punishing the offender by bringing sickness and other misfortune into his household. The administrative functions of the ogujo olopu are limited by the fact that Igala clans tend to be widely dispersed beyond their original centres. But, more especially, they are limited by the existence of other office in which the external leadership of the clan is vested. In clans that are titled, as the majority are, the roles of external and internal leader are merged in a single office, to which succession is governed by formal rules of rotating succession in bestowing titles, the king, or other patron, uses the formula

Enyini e mudoqujo olopu

Today you become the ogujo olopu."

In other words seniority by age is in this is entirely sub-ordinate to the function of political leadership, which has its own rules of procedure. The duties of the ritual ogujo olopu, or most senior elder, are limited in these cases to deputizing for the titleholder during the succession. When the titleholder dies the ancestral cult symbols okuute are transferred from his house to that of the senior-elder, and the next incumbent formally takes on the duties of ogujo olopu when the okwute arrainded over to him after his appointment. In another case a married woman who became chronically ill, with tuberculosis was sent home to her father when the oracle atributed her illness to find out that she offended her father's ancestral guardians. sickness to ancestral anger and discorvered various ways in which the patient has offended her fathers ancestorial guardians. The Igala also isolate and limit the administrative functions of the clan elder by distinguishing between the office of ogujo olopu and that of onu ane, or 'chief of the land'. The second term describes the political head of the clan in his capacity of heir to the sovereign rights exercised by the clan over the land with which it is associated. In titled clans this function is merged with the political and ritual aspects of clan office to form a single entity. But during, title disputes and the long periods of interregnum that are so characteristic of the Igala title system, the two offices tend to become separated and the demarcation of their respective spheres of jurisdiction and authority is frequently a matter for dispute. I postpone discussion of this complex relationship to the section in which territorial rights are described more fully. One of the basic differences between the position of ogujo lopu and other clan offices is its association with the horizontal plane of clan organization, in which the unity of sibling groups, cuts, across the ane division of the clan into lineages for into subclans the ritual duties of the office, for instance, present relations with the clan ancestors in the light of this unity rather than portraying them as an association of separate lineal groupings. Titled offices on the other hand, incorporating the functions of the land chief, onu ane, tend to emphasize lineal divisions, and vest the hereditary rights of the clan in one lineage or subclan at a time. This distinction will become clearer as we investigate the differences between clans in more detall and at this stage we will illustrate only the principle liniage group unity without supplying the contrast-ing term of solidarity within the lineage. In all matters concerning inheritance and succession, in which, by definition, all clan members have a common interest, the principle of sibling unity is upheld by the rule that seniors take precedence over juniors. This rule is expressed in a traditional formula of inheritance which says that-

achogwa one chene ki a rononw 'the senior sibling is the person to succeed.'

When a person dies in Igala the transfer of statuses that the deceased held is discussed and put into effect by the agnates of the deceased when they assemble for the funeral ceremonies The main stages of the funeral, which may be spread over years, correspond to different stages in the distribution and allocation of the deceased's property. At the first stage, centring allocation of rights, over the dead man's wives and children, on the interment, an interim arrangement is made. A further is made when the family reassembles for the rites that end the widows' mourning. And the final settlement is reached at a further stage of ritual, the Aku ceremony in which the ghost becomes an ancestor definitively.

These funeral assemblies reproduce the sibling organization of the clan in miniature and form one of the most important occasions for corporate action within the clan. The terse suc-cession formula quoted above refers to the pattern of organ-ization that emerges on these occasions and does not directly describe the actual heir or the mode of transferring property. It is related to the norm that claton and feft have a common interest in property matters and a common voice in the determination of inheritance. When the clan, or its local section, meets to discuss inheritance, the sibling grouping takes precedence over division into lineages, and authority over the settlement rests chiefly with the senior siblings, achogwa, of the deceased. The number of those who actually attend varies with the social status of the deceased and the genealogical involvement varies accordingly. At the funeral of a wealthy noble most of the branches of his subclan or clan may be represented, and the inheritance issues are discussed informally over a wide range of clan relationships. At the other extreme a poor commoner living apart from his clarısmen may have only his own and his brothers' children to represent the olopu. But the principle of agnatic right and sibling unity is unaffected by these variations in practice. The clan group that meets to discuss inheritance and perform the funeral rites-for-a-dead-member of the clan is nominally under the authority of the ogijo olopu within the clan. If the deceased is a titleholder this may be symbolized when the funeral presentations are begun, by making the first presentation in the name of the ogijo olopu, But this symbolic authority is rarely put into effect unless a dispute breaks out over the inheritance which is of such dimensions that the funeral procedure has to be suspended. Under normal cir-cumstances, the effective head of the proceedings is the most senior male member of the dead person's sibling group. It is this relative who formally takes responsibility for the whole funeral and adopts the role of ajiegwu, 'He who buries the body.' He presides over the discussion of inheritance questions, and gives the final decision if there is disagreement within the sibling group. In between the formal assemblies the heirs of the deceased consult the ajiegu before they alienate or convert any of the property. Similarly, clansmen who wish to take any of the widows, under the custom by which rights over widows are inherited agnatically, make their request to the ajiegwu who then brings it before the whole sibling group when they assemble at the end of the ritual period of mourning.

Traditionally, the ajieguu should be chosen, where possible, from a collateral lineage of the clan and not from the deceased's own lineage the lineage is the widest group within which self-acquired property and domestic rights over houses, land, and crops are normally inherited. But by the rule of funeral procedure this grouping is made subordinate to the clan as a whole in property matters, and the common interest of clansmen in this sphere is expressed and maintained by giving them authority over the inheritance settlement. More-over, clansmen from collateral lineages are not excluded entirely from inheriting themselves. They can, for instance, inherit rights over widows and children and if the deceased had specially close ties with agnatic relatives outside his own lineage this might be recognized by allowing them to take small items of property, or by allowing them the use of any land or houses that had been granted them in the deceased's lifetime The pattern of co-operation and obedience that emerges in connexion with inheritance is so fundamental to the institution of clanship that the Igala frequently use the principle of responsibility for burial to characterize agnatic relations in general. In acknowledging other groups as clansmen they often say:

awa cholepu katiti a ji eguru ola wa 'We are one clan, we bury each other.' Similarly in title contests, where candidates may try to disqualify their rivals by alleging that they not members, the person so challenged may reply by bringing evidence to show that members of his own lineage have buried or been buried by other lineages whose clan status is beyond question. If the evidence is accepted it counts as a proof that the persons named were agnatically related and that their descendants in the male line are members of one and the same clan.

As the context of this corporate activity shows, the notion of clan solidarity is essentially a ritual one, involving the totality of clan members both living and dead in an indivisible, but essentially spiritual, unity. Perhaps the clearest expression of the ritual identity between clansmen occurs in the agnatic inheritance of ritual prohibitions, elifo, on the use of various animals and plants as food, or on certain kinds of activity that are ritually proscribed. These clan avoidances, elifo olopu, are observed throughout the individual's life, and it is believed that sickness will result if the rules are knowingly disregarded. After death, on the other hand, the prohibition has to be Fitually broken as part of the obligatory funeral rites, icholo, that separate the dead from the living. In the Achadu's clan, for instance, a morsel of the prohibited ouwe fish is added to cooked food of which a portion is then placed on the grave of the dead person. Or, if a certain kind of behaviour is proscribed, such as using three forked sticks to support house. hold vessels, this is done deliberately at fone stage of the funeral to release the dead person from his bonds with the living members of the clan.

These hereditary avoidances form the basis of clan oaths which can be employed by a suspected person who wishes to swear his innocence. The Igala say that if a man swears by the thing which is forbidden to his clan his word is not to be doubted. The oath takes the form of saying 'If I am guilty of this thing may the prohibited thing (naming the food or action) be done to me this year. In other words, the suspect swears by his own clan identity that he will die within the year, if his words are untrue, and implies that his fellow clansmen Sickness caused by disregard of elifo elepu usually takes the form of boils and other chronic skin conditions will assemble to carry out the icholo connected with this prohibition.

These clan avoidances are like totemic prohibitions in that they often concern animals or plants. But there are also many differences between the Igala custom and the use made of animal and vegetable species in the classic system of totemism. The elifo olopu are in many cases not widely known outside the clan, and do not form public symbols of clan identity. By their nature, the prohibitions ought to be observed at all times, and potentially make a symbolic distinction between the members of a clan and those who observe different restrictions. But in practice it is mainly their role as funeral icholo that brings the elifo olopu into prominence in ordinary social life, and in this context they form esoteric symbols of the mystical bond on which clanship is founded.

In general the Igała do not rationalize the different ritual prohibition is observered by clans beyond saying that they belong to funeral ritual, enw icholo, or that they are an ancient custom, enw egwu. This absence of logical myths and legends in connexion with the elifo olopu is especially typical of the avoidances connected with inanimate objects or with miscellaneous activities such as whistling at night or cracking palm kernels outside the house. Where the prohibitions refer to animals, explanatory traditions are more common. Typically, this class of prohibition is explained by stories that attribute the hereditary relationship between a clan and its totem to the animal's intervention at a moment of danger in the life of the clan's founder. For example, the Ofojo clan at Achigili relates that in the distant past its members, who were then living together in one village, were asleep at night when an edikura monkey disturbed the rest of one of the elders by chattering in a tree above his house. Each time the man tried to get back to sleep the monkey became more insistent, and behaved in such an unusual way that the elder at last understood that it was giving warning of impending attack, danger, raiding, Rasty, from another village. So the elder roused the, village and led his clan to safety In some cases the prohibited activities are carried out by thasieratles, and to perform them in any other context would be to betray a secret of masquerading. For example, whistling at night is an activity associated with the Amed masks which are performed by certain clans. dispersed the footprints they left on the sandy paths, so that clansmen to safety, whilst a group of the edituu monkeys the enemy would not know which direction they had taken This type of legend rationalizes the fact of clan dispersal in Igala by postulating the break up of an earlier, gnore con centrated type of patrilocal settlement through a war or series of wars. In upholding the myth that each clan originally had a common founder and a common birth place the legend transfers some of the loyalty owed to these geographical symbols to the special relationship with a particular animal species, which at once recalls the mythical unity of the clan in remote time, and the factors that led to the clan's dispersal. One other element stressed by this type of legend is the clan's solidarity in its external relationships with other descent groups. The prohibitions on inanimate objects and miscellaneous activities tend to be symbols of the internal unity of the clan and of the ritual bonds between members exclusive of their connexions with members of other clans. But the prohibitions on creatures tend to refer to the clan's participation in a wider scheme of political relations. For instance in respect of certain of their individual eliſo olopu the royal clan is opposed as a class to the non-royal clans and to the Achadu's clan. The royal clan taboos the flesh of dogs and has a positive affinity with the leopard. The Igala Mela clans, on the other hand, taboo ewolo or civet cat, and in this respect are typical of those indigenous clans throughout the kingdom who per-form the Egwu Afia play, which uses a costume incorporating the pelt of ewolo. The Achadu's clan, finally observes an abso-lute prohibition on the ouwe fish. These three classes of totemism are widely known in Igala and to some extent. serve to symbolize the identity and individuality of the clan groups concerned within a system of political co-operation. Although Igala clans are basically similar in construction and in function, there are many differences between individual clans, and diversity is a characteristic of the clan system as a matters of burial custom, masquerading, and other features partly differences of detail, in of ritual symbolism. But they also extend to clan structure in whole. The differences are

This story occurs also in Idoma country, and among some the groups in the Nsukka area. that there are marked differences of scale within the clan system, and variations in the order of internal segmentation. This structural variation is associated with rank and political ascendancy, the largest and most complex clans being those of aristocratic descent whilst the smallest and simplest descent groups tend to be low in the social scale and insignificant politically.

Some idea of the range of variation in the structure of Igala clans can be gained by considering two clans placed at oppo-site poles of the system. At one extreme the royal clan divides genealogically into at least five major subclans, which in turn subdivide individually into several maximal lineages. The sub-clans are centred in different parts of the kingdom, and the clan as a whole is so numerous and so widely dispersed that it has representatives in practically every village throughout the kingdom. It is difficult to calculate the total membership of this group, but a rough estimate can be made of the Idah sub-clan as totalling at least 35,000 members. In 1963 the Ocholi Descendants Union, comprising members of one of the four lineages that make up the ruling subclan, had a paid-up membership of 8,907 persons, mostly adults. The member-ship included some uterine kin in addition to agnates, but this accession must be balanced against the fact that not all full members of the lineage belong to the Union. The four line-ages of the ruling subclan based on Idah are roughly equal în size, and it would therefore be reasonable to estimate the total membership of this subclan at four times the total of any of its lineages. The relationship between the numerical size of the Idah subclan and the other royal subclans located in other parts of the kingdom is even more difficult to estimate accurately. But we know that some of these groups are not far below the ruling house in rank and political importance, and that their role and pattern of dispersal closely parallels the position of the ruling

Strictly speaking, the Igala term olopu describes the patrilineal nucleus of a clan and the term aju the descendants in either line of a common ancestor. In practice the second term is often used by Igala to mean a patrilineal clan with its membership increased by including uterine kin, omonobule.

On a conservative estu wate the combined membership of the provincial subclans at least equals that of the ruling house at Idah. So we can conclude that the total membership of the royal clan as a whole is in the order of 70,000 persons, which is equal to about 24 per cent. of the Igala-speaking population.

At the other extreme of clan structure are clans of low rank whose political functions are limited to controlling the district within which they are centred. A typical example of this type of clan is the Ojo clan centred on the village of Igebije in Gwolawo District. As the genealogy in figure 3 on page 34 shows, the clan divides into five major lineages and has no intermediate stage of subclan organization. Its total adult membership in 1963 was not more than twenty persons, or a maximum of forty if we include closely associated uterine kin.

The genealogical depth of clans in Igala seems to vary with their numerical strength and political importance. The sub-clans of the royal descent group, for instance, vary in depth between seven and nine generations. In the Ojo clan, and a group of similar clans in the surrounding area, the variation is between four and seven generations.

This discrepancy in size and scale between clans of high and low rank in Igala is a fact of great cultural and political significance. Apart from their overall political functions, the aristocratic clans perform a unifying and centralizing role through their great size and wide dispersal. In other words political control is a function of clan size and spread as well as of judicial administrative and military arrangements.

Culturally also, these clans are regarded as the natural guardians and arbiters of tradition, and this may be connected with the fact that their corporate activities form a complete demonstration of the rules that are basic to Igala social structure. Offices held within the royal clan, for instance, repre-sent the widest possible extension of the principle of seniority, and also demonstrate-in-full-the-relationship between this principle and that of lineal segmentation. Cases of inheritance and succession in the royal clan form an important class of legal precedent for this reason and are remembered and discussed long after similar cases concerning commoners have been forgotten. But above all, by their unity and political

ascendancy, the aristocratic clans provide the clan system itself with a framework within which clans are oriented to the centre by hereditary alliances and by ties of kinship and marriage. The overall dominance of the royal clan within the clan system is balanced politically by grouping other, lesser clans together in opposed classes. I demonstrate this by discussing the grouping of clans within the capital and the division of functions between them. I then go on to show the repetition of this pattern in the provincial centres of government, leaving the purely local aspect of clan relations to be described in the following chapter.

The royal clan.

From some points of view the various groups claiming royal descent in Igala make up a federation of clans rather than a single unilineal entity. The royal subclans are localized in different parts of the kingdom and in the past were virtually autonomous within their own spheres of political influence. Relations between them were often hostile, and even today the branches are so jealous of their own prerogatives that they tend to regard each other as rivals in a struggle for political power. But in spite of the degree of separation between subclans, and of a recurrent absence of cooperation, members of the clan have a strong sense of royal identity, based on common descent and this sense of belonging to a class set apart from other, non-royal clans, compensates for the centrifugal tendencies inherent in the political functions of the clan.

The family tree of the Igala royal clan, like Igala clan genealogies in general is essentially an aggregate of individual lines rather than a balanced and symmetrical structure of the type that emerges in a lineage system. For convenience, since it provides a useful way of summarizing genealogical information about the different subclans, I have shown the pedigree of the royal clan in figure 4-on-page-46-as-a-comprehensive genealogy. But it should be emphasized that this synthesis is to a large extent an artificial origin and has no exact parallel in Igala concepts of clan structure. There is no occasion or context on which the pedigree is set out, recited, or otherwise preserved and outside the interest of most of the members of the clan. as a whole, and the total structure is beyond the knowledge To some extent the genealogy of the royal clan

Within particular subclans, have to be submitted to the Ata at ized through the fact that all royal titles, though hereditary Idah for appointment when an office falls vacant. In investing contestants the king, together with gating the claim of rival concerned his own titled siblings and the elders of the subclan concerned, investigate the genealogy of the contestants, tracing their relationship to the founder of the subclan and from this founder to the point at which the subclan diverges from the body of the parent clan. These investigations, on which the award of the title depends, define for a time at least the relationship of the subclan to the senior house, at Idali, and it is by assembling each subclan's version of its title ancestry that a comprehensive picture of the clan's structure can be obtained.

But this process of genealogical co-ordination has many limitations, and does not keep the parts of the system in balance as the processes of investigation and recording do in a lineage system. Royal title disputes involve a dynastic relationship between two subclans only, and exclude members of the clan who belong to neither the ruling house nor to the subclan in which the vacancy occurs. Moreover the settlement reached in one generation is not regarded as binding upon the next, because, through the system of rotating title succession, a different lineage holds office each time the title falls vacant and the incoming lineage does not feel bound to uphold the version of a pedigree laid down for its predecessor in office. A similar lack of continuity occurs on the king's side also, if the king happens to have changed between the time when the útle was last awarded and the time of the next vacancy. In this event the two lineages concerned, from the ruling subclan "On the other, work out their relationship afresh. on the one hand and the subclan in which the vacancy exists

their apical andestor, and are in this sense structurally equi Most branches of the royal clan take Ayagba om Idoko as Members of the clan are sometimes described collectively as valent, as Subclans, with the ruling subclan based on Idah. Amoma 'Ayagba, the children of Ayagba, and royal individuals who wish to stress their genealogical identity may do so by Closer saying Ayegba bi wa, 'Ayegba begot us'. The clan pedigree is however, typical of Igala clan pedigrees in general in being so broad laterally in relation to its generation depth that the notion of common descent from Ayegba is clearly a fiction, a convenient term for the notion of common ancestry acquaintance with the pedigrees of individual lingages shows less uniformity in the structure than Igala suppose; both at Idah and in the provinces there are branches of the clan that claim descent from ancestors earlier than Ayegba, and pressume equity with all the branches that stem from Ayegba himself.

The most important case of dissent from the notion that Ayegba is the true apical ancestor of the royal clan occurs among the northern subclans in Igala, who tend to regard the Ankpa branch of the clan rather than the Idah one as their immediate superior. This argument turns on the status of Atiele, from whom the founders of these subclans are descended, and who in the commoner version of the clan pedigree is said to be a son of Ayegba. The Ankpa branch of the clan now accepts this version and gives genealogies that correspond with the one set out in figure 4 on page 46. But in 1935, administrative officers collecting historical information at Ankpa were told that:

Atiele, a younger brother of the Ata Ayegba came to the Ankpa area trading. The people of Ojja who were settled there gave him a wife called Anagba who was the mother of Oguchę Ekwo. She was a relation of a local chief called Agbaji. Oguchệ Ekwo got so strong eventually that Agbaji gave up his title and later Oguche Ekwo went to Idah and was given beads by the Ata." This version which places Atięle and Ayegba in the same generation is supported in the administrative records by the traditions of an offshoot of the Oguchekwo lineage, based at Okenyi. It also occurs in my own notes from the Ankpa area, The best known example of this is the subclan owning the Ochai Ata title. Different kings have broken the rule of succession to this title by bestowing it on their own collateral relatives, but the title strictly belongs independent royal subclan, forming a separate branch of the royal house, and is clearly, a major alternative in Ankpa tradition to the version that regards Atiele as a son of Ayagba. Which of these two accounts is dominant in any period probably depends upon the state of political relationships between Idah and Ankpa These typical branches of the royal clan have formed the focus of political ppposition between the central and the north-eastern (districts of the kingdom, and Ankpa has tended to regard the Schority of the Idah branch as being spiritual rather than political in character, At the turn of the century the Ankpa group rejected the Atas authority in political matters by refusing to accept the king's choice in the succession to headship of the subclan. This led to lighting and to eventual intervention by the British Government on the king's side of the quarrel. Since that period administrative changes have tended to incorporate the northern districts more closely within the framework of the kingdom. And the Ankpa subclan has finally accepted a subordinate position, as its own genealogical records fiow testify. But the claim to equality or near equality with the Idah branch still exists in the notion that Aticle was a junior brother of Ayagba om Idoko, and not his son..

At Idah itself there are three royal descent groups that stand genealogically outside the framework of the main ruling subclan. Each of these groups has its own hereditary title, whose holder is confirmed by the Ata, and the difference in genealogical status between these lines and the main subclan reflects a degree of structural separation between their own offices and the offices that are hereditary in the Ata's subclan. The first two of these subclans, taking the title of Ochai Ata and Achenya Ata respectively, trace descent from Idoko, theAnkpa District Notebook, 1903. An expedition sent from Lokoja to uphold Oguchay's appointment by the Ata (see figure 7. p. 5o) was ambushed at Enabo on the return journey and suffered heavy casualties. A punitive expedition, which was strongly resisted, followed and a military garrison was established at Ankpa to pacify the area. A company of the West African Frontier Force was stationed in Ankpa from 1904 tο 1933.

In 1968 Ankpa led movement supported by ninė northern. Igala Districts which aimed to establish a separate Native Authority at Ankpa, so instead of daould be to Ananda's claims to that tax payments would be made to Ankpa and administere from political equality with Idah. administered locally. An alternative to the collective terra amom'Ayagba for the members of the royal clan exists in the name amomala which classifies them by reference to the oval title rather than by reference to the founder. But this term is ambiguous since although in one sense it means any member of the royal clan, in another sense it describes the titled members of the royal clan and their descendants over two generations. We can conclude that there is no single term for membership of the royal clan that is without political overtones; the concept of royal descent by its very nature implies distinction of seniority and relative status, and it is only in relation to other groups of clans that these divisions disappear. The nobility of the royal clan is a basic fact of igala social life, maintained by a number of economic and political privileges, and expressed in these and various other customary forms of respect and deference. Basically, however, the qualities that are thought to justify this high rank are spiritual attributes, hereditary within the clan and descended in an un-broken line from its remotest ancestors. They are symbolized to the members of the clan and to the Igala as a whole by the totemic affinity that exists between the amomata and the leopard, omataina, whose name combines the singular form of the clan name with the adjective ina meaning sacred or holy.In one sense the term ina means 'the greatest' or 'supreme'. But it also suggests the notion of a spiritual or sacred counterpart of a secular reality. It occurs in the name ojaina for the royal burial ground (oja = group, or community), in Omaina, the term for the afterbirth, and in ofuina, discharge occurring in acute forms of dysentery. Within the roval clan, the leopard affinity is explained by a myth that traces ultimate royal ancestry to an anthropomorphic leopard. This a myth is not generally known in detail outside the royal clan, but the sacred leopard affinity is common knowledge, and knowledge of these ritual associations appears in the fact that the leopards a quasi-sacred animal to all Igala. In hunting it is gready feared for its ritual as well as its physical qualities. Its name, must not be pronosticed in the hunting field, as this act is sufficient to contaminate the hunting medicines, and ata communal hunt surprises a leopard whilst beating the bush, the hunting has to be abandoned for the day. If a leopards killed its death is mourned in the same way as a member of the royal clan, and this obligation does not end until a member of the royal clan has been sent from Idah to collect the pelt for the Ata and supervise the ritual disposal of the leopard's body.

Again, when members of the royal clan are buried their funeral ceremonies include two rites, icholo, that express the idea of leopard affinity. These form perhaps the commonest expression of the idea, through which everyone is reminded of the spiritual identity of the clan, and involve, as we shall see, a direct contrast with other forms of clan totemism which are expressed the same context of burial.

In the first stage of funeral ceremonies, in which the corpse is wrapped in white cloths, okpe, brought by mourning kins-men and friends, before the body of an Omata can be removed to the grave the shroud has to be marked down both sides of the body with spots. This is done with three pigments, white, red, and dark blue, or black. The marks are said to represent the spots of the royal leopard. As the marking is

These rites are referred to as icholos, "funeral rites', and include the act of wrapping the pelt in white funeral cloths, skpe, in the manner of a normal burial

The pigments are made from chalk, camwood and indigo, or ashes respectively. A further instance of this piecoccurs in an administrative report from Dekina where the chief Onu Ajobi, is head of a royal subclan. The report, written in 1950, says: 'It is believed that the souls of departed Onus Ajobi can enter into the bodies of leopards. (In this connexion it is inter-esting to note that before burial the bodies of deceased Onus Ajobi are marked with brown and white clay (in imitation of leopards' spots.)

Carried out, by the agnatic kin of the dead person, the women and other bystanders sing:

E ma deken e da ama

an ama yen). 'If you don't turn into a leopard then turn into later, in a series associated with the second mortuary ceremony, aku, a bier, abaihi is displayed in the same manner as the shrouded corpse before the actual interment and when this is about to be carried to the grave a rite is performed that the snatching of the body by a leopard symbolic hereditary distinctive item of royal ritual that is in use amongst the whole clan is a protective medicine called Ejebilo This is to be found in most ancestral shrines within the clan in the form of a small calabash filled with a black substance, stoppered, and often tied round with white cloth (okpe) strips. It has the function of an ode oji (odoji), 'medicine of the head', that the family head applies to the foreheads of new-born babies, placing them as it were, under the protection of the ancestors from whom the medicine was inherited. All clans use protective medicine (Odoji) for their children, but the variety, Ejebi, is specific to the royal family. Ayagba om Idoko is said to have appointed the Agaidoko clan, a prominent family in Idah of Igbirra extraction, to make it för himself and his successors. One intelligence report describes the medicine at Ankpa as being used especially for partürition, in cases of difficult childbirth, and the name could be inter-preted literally to mean 'agree to bear children' or 'facilitating birth'. But it also has the wider functions attributed to pro-tective medicines. As my informant said, 'It helps the royal clan in their farming, and to begét children, it brings luck their way and closes the road to witchcraft which might harm them and their families.' for a reference to chiefs who wore a piece-of-leopard's-skin sewn onto their gown or cap.

Burial rites of the Ata

For this rite women are driven into the houses and away from the site of the funeral, and a leather bag or similar container is beaten to represent the noise made by a leopard which is supposed to snatch the body from the bier. This rite is in fact a kind of masquerading. is supposed The ritual qualities of royalty have negative as well as positive aspects, and there are conditions of pollution which have to be avoided. It is forbidden, elifo, for an omata to eat dog's flesh, and not only the flesh is prohibited, but any kind of association between the concept of a dog and the concept of royal food. So that if an omata is eating and the common term for dog, abia, happens to be uttered, the food should be thrown away. Negative food prohibitions are expressed in the context of death in the same way as the positive totemic association. When the titled heads of royal lineages die, one of the funeral rites, icholo, consists of offering a cooked meal containing dog meat Indeed ancestors of each sub, the period of the group's foundation, since throughout the clan are remembered. When a new titleholder takes up office royal group only the last nine title the okuute of the earliest member of his lineage is discarded, custom is possibly with the result that only the last two or three ancestors of each lineage are represented in the cult. This custom related to the degree of separation between the different branches of the royal descent group.

Although the various royal subclans are ritually auto-nomous in the field of ancestral ritual, the Idah subclan occupies a special position in relation to the roval dead. Its for clan is framework of royal ritual which stresses the uniqueness of the Ata's position as direct descendant of Ayagba om Idoko, and heir to the insignia of office that are believed to have been instituted by Ayagba. In addition to these insignia, there is a physical association between the early ancestors of the royal dynasty and the capital which makes it ritually impossible for the functions of the ruling house to be transferred to any other centre in Igala. At Idah in addition to maintaining the graves of the last nine kings, the royal festivals also maintain the grave of Ebelejonu, representing the earliest or protodynastic ancestors, and the grave of Ayagba om Idoko, standing for the emergence of the royal clan as a political entity. And when the festival of royal ancestors is held, it must be initiated at these graves in the capital, and the first stage be completed by the king, before the royal subclans in the provinces can perform their own ancestral rites.

The ritual, a social and political relationship of the dif-ferent branches of the royal ancestors to one another and to the senior branch at Idah, can best be summarized by saying that the Ata is in one capacity the ogujo olopu, or ritual elder, of the entire clan. It is in this capacity that he confers titles within the entire royal clan, the holders of these royal titles forming, as it were, a ceremonial sibling group which reproduces the unity of the clan in the first generation of it Around 1950 a serious attempt was made by government to transfer the headquarters of the kingdom from Idah to a more central site, at Ochaja. But this was so strongly opposed by the Igala that the idea had to be abandoned, and work on the new site was discontinued. The crises of life also contain an element capable of polluting the qualities inherent in royalty. Childbirth and new-born children are supposed to be avoided by an omata; a corpse has the same ritual significance and should not be seen for this reason. Contact with menstruating women, and the act of sexual intercourse are also contaminating and may have to be ritually cleansed. These last two forms of pollution have a wider incidence than the royal clan, and I shall try to show later which elements of royal ritual are exclusive and which are of common occurrence.

One other minor taboo associated with royalty prohibits the use of ugbakolo (Ficus capensis), a tree whose leaves are a common ingredient of native medicines for coughs and catarrh. Amomata may not make use of its timber for firewood, nor employ the leaves, bark, or fruits in their medicines. Again, there is a funeral icholo connected with this fruit which breaks the prohibition and so emphasizes the separation of the newly dead from the living. Whereas these ritual prohibitions affect the whole clan and recur in royal burial ceremonies throughout the kingdom, other fields of ritual emphasize the clan's division into auto-nomous subclans. The cult of royal ancestors, for instance, is fragmented amongst the component branches of the royal clan. Each subclan has its own burial ground, similar to the Ojaina grove at Idah, where the kings are buried, and its own Set of ancestral staffs, okwutę. There is no emphasis in the ancestral ritual of the clan on genealogical continuity from It is also reputed to be an ingredient of medicines used in sorcery.

ided ancestors of each sub the period of the group's foundation, since throughout the clan are remembered. When a new titleholder takes up office royal group only the last nine title the okuute of the earliest member of his lineage is discarded, custom is possibly with the result that only the last two or three ancestors of each lineage are represented in the cult. This custo related to the degree of separation between the different branches of the royal descent group.

Although the various royal subclans are ritually auto-nomous in the field of ancestral ritual, the Idah subclan occupies a special position in relation to the roval dead. Its for clan is framework of royal ritual which stresses the uniqueness of the Ata's position as direct descendant of Ayagba om Idoko, and heir to the insignia of office that are believed to have been instituted by Ayagba. In addition to these insignia, there is a physical association between the early ancestors of the royal dynasty and the capital which makes it ritually impossible the functions of the ruling house to be transferred to any other centre in Igala. At Idah in addition to maintaining the graves of the last nine kings, the royal festivals also maintain the grave of Ebelejonu, representing the earliest or protodynastic ancestors, and the grave of Ayagba om Idoko, standing for the emergence of the royal clan as a political entity. And when the festival of royal ancestors is held, it must be initiated at these graves in the capital, and the first stage be completed by the king, before the royal subclans in the provinces can perform their own ancestral rites.

The ritual, a social and political relationship of the different branches of the royal ancestors to one another and to the senior branch at Idah, can best be summarized by saying that the Ata is in one capacity the ogujo olopu, or ritual elder, of the entire clan. It is in this capacity that he confers titles within the entire royal clan, the holders of these royal titles forming, as it were, a ceremonial sibling group which reproduces the unity of the clan in the first generation of its Around 1950 a serious attempt was made by government to transfer the headquarters of the kingdom from Idah to a more central site, at Ochaja. But this was so strongly opposed by the Igala that the idea had to be abandoned, and work on the new site was discontinued. Foundation. When a titled member of the clan dies the king has to be informed by sending some of the title-beads cut from the dead person's wrist, together with a hat, and in return the Ata gives permission for the burial to proceed by sending a burial cloth which becomes the first of the many cloths pre-sented at the funeral. Similarly in inheritance disputes, or in any controversy that affects royal clansmen, the king can inter-vene as ogujo, olopu and give a ruling which is absolutely final, short of rebellion.

In principle, members of the royal clan are under the king's personal jurisdiction for serious offences, and cannot be punished by any other court than the king in council. Amomata who were condemned to death, as they might be for murder, kidnapping, or raising war against the king, were executed by the Ochai Ata, the titled head of one of the lineages in the royal subclan at Idah, instead of by the igainya or by the royal eunuchs, who executed other felons. The place and method of execution were also different; it took place at the Ochai Ata's court (okęte Ochai Ata) and was carried out by a form of strangu-lation. It is doubtful, however, whether this judicial pro-cedure was followed as a matter of course where the outlying subclans were concerned. The members of these groups were under the jurisdiction of their own subclan head for most misdemeanours, and there are few concrete instances of criminals being taken to Idah from the provinces for punish-ment because they were of royal blood. The rule was more effective in the case of members of the ruling house, but for the kingdom at large it formed part of the mythology of royalty, and of the ideal and fundamental difference between the main descent classes.

It will be obvious from this account of the royal clan that its corporate functions as a descent group are extremely rare. Execution by cutting the throat and beheading took place either at

erane, the land shrine, or at ere Inikpi, the shrine of the Ata's daughter, Inikpi Kidnappers and incorrigible thieves were sometimes executed by impaling, edunyokpa, and this sentence was usually carried out near one of the main approaches to the capital. Offences against the king, such as adultery with a royal wife, were punished by clubbing and throwing the offender from the

summit of the high cliffs, overhanging the Ochęchę stream. The offender was prostrated and suffocated by a stick held across the throat. stricted, and that the bond between royal clansmen is mainly, a ritual and moral one. The assumption of common descent provides this bond of unity, but the clan, as a whole has no precise genealogical structure. The widest unit within which relations between clansmen are controlled by the descent model is the subclan, whose branches are inter-related as parts of a total system. The clan itself is co-ordinated geo-graphically rather than genealogically, and although there are institutional expressions of clan unity these tend to be corn-centrated in the ruling house to the exclusion of the outlying branches. A major part of the contacts between the senior sub-clan and the other branches of the royal group are purely political in character, concerned with the spatial relationships of the different centres, and recognition of the descent tie is limited mainly to the annual festivals and to certain other formal occasions such as succession to vacant offices. The divisions of the royal clan that I term subclans are political groupings of lineages rather than genealogical units in the strict sense of the term. If Atiçle is regarded as a son of Ayagba for instance, there is no structural difference between the lineages centred at Idah and those centred at Ankpa which would justify separating them on genealogical grounds alone. The royal subclans can be regarded as descent groups only in the sense that they each monopolize the inheritance of a par-ticular title; it is their descent from the titleholder rather than descent from the individual founder that provides the key to their unity. The members of each subclan share exclusive rights to a particular title or group of titles and it is this joint prerogative that gives the subclan the status of a hereditary corporation in its dealing with other groupings of the same order. The major titles taken by the different royal subclans are shown in the table on the following page.

In all the royal subclans there exists in addition to the major title held by the leader of the group, a number of minor titles that are inherited within-the-same-rotating framework as the central title. The distribution and inter-relations of these secondary titles are most significant for the structure of the Subclan, whose internal segmentation can only be understood in relation to the inheritance of the whole complex of titles belonging to the subclan. To show how this system operates. Title

Salutation

Subclan centre

Ata

Agabaidu

Idah

Onu Ankpa

Doga

Ankpa (Ankpa District)

Achema

Doga

Dekina (Dekina District)

Onu Imani

Oji, or, Duga

Imani (Imani District)

Onu Ojoku

Agbo, or, Doga

Okaba (Ojoku District)

Obaje Atabaka

Doga

Okpo (Olamaboro District)

Ojibo Akpoto

Doga

Iga (Olamaboro District)

Ochai Ogugu

Doga

Ogugu (Ogugu District)

Onu Aloga

Doga

Ikka (Enjemma District)

Onu Okenyi

Doga

Okenyi (Ankpa District)

TABLE 1. TITLES HELD BY ROYAL SUBCLANS

and to analyse its structural implications, I will describe the inheritance of titles within the ruling subclan at Idah and con-sider in detail the distribution of these offices within one of the component lineages of the subclan, the Aju Ocholi group. It was not possible for me to spend long enough in the other subclan centres to work out their own patterns of title inherit-ance in similar detail. But the evidence available, which I sum-marize at the end of this section, suggests that the principles are the same and that a closely similar structural pattern exists within these groups.

The ruling house at Idah divides into four maximal lineages¹ that are equal to each other in the succession although they are not of the same status genealogically. As figure 5 on page 57 shows, two of the lineages have emerged from the bifurcation of the group founded by Akumabi and were at one stage of their existence major lineages within this maximal lineage. The Igala rationalize this change from a rotation of three branches to one based on four lineages by saying that Amacho, Aku-mabi's eldest son, died before his installation as Ata could be completed. The succession therefore passed to his junior brother, Itodo Aduga, and since then has alternated between the descendants of these two men and the members of the dynasty itself

In central Igala the term olopu describes the segments of a clan as well as the entire descent group. Branches of this unilineal group can be dis-tinguished loosely as 'hands',owo, and in some eastern districts major and maximal lineages are referred to as atakpa, "houses', or ofegbili, "hearthstones'.

56 THE HEREDITARY BASIS OF CENTRAL GOVERNMENT

older maximal lineages, founded by Akogu and Ocholi. The lineages are referred to descriptively as either olopu Itodo Aduga, olopu Amacho, and so on, or alternatively as aju, descendants, of these men.

Ayagba om Idoko'

Akumabi

Akogu

Ocholi

Amachos

Itodo Aduga 1 Amocheję"

Ogala

Idoko Adegbę

Onuchę

Ekalaga

Aku Odiba¹2

Okoliko

Ocheję Onokpa¹ 1 Obaje19

Amaga¹¹ Atabo

Oboni16 1 Ame 20

Oguche Akpa

Alii

FIGURE 5. THE RULING LINEAGES

(omitting protodynastic kings, and two later rulers who left no male issue)

Politically and socially the four branches of the ruling house combine on many occasions in pairs, following the lines of the original genealogical split between Akumabi's descendants and the members of the other two lineages. The first pair, comprising the Itodo Aduga and Amacho lineages has the collective name of Ajaku, an abbreviation of aju Akumabi. The second pair, comprising Aju Ocholi and Aj'akogu has no special name, but its members often describe the close con-nexion of the two lineages by the kinship term omaiye, which denotes children who have both parents in common. The complementary term for their relationship with the Ajaku moiety is omora, or omata, implying that in this direction they are related through the father only. These terms are, of course, used-figuratively-and-do not imply an actual kin rela-tionship; there is no sense in which the two lineages could be described as forming a corporate kin group on the basis of matrifiliation.

This opposition between the two major divisions of the ruling house is a basic fact of Igala political structure and

57 THE IGALA KINGDOM

appears to have been in existence from a remote period. As Clifford first pointed out, the traditional layout of the palace area at Idah reflects this dichotomy. The Itodo Aduga and Amacho lineages build their palace in the same half of the enclosed area known as olegbo Ata, whilst the other two lineages, during their own tenure of office, occupy the other half of the compound. The square tower, odage that forms a major landmark in the royal quarter serves as a boundary marker between these two sectors.

Politically, the Ajaku branch of the ruling house tends to act in concert against the other two lineages, who regard each other as natural allies in struggles over the succession or the allocation of political power. At different times in the past the allies on either side are said to have been involved in inter-necine conflicts, with aju Itodo Aduga fighting aju Amacho and aju Ocholi fighting aju Akogu. But these hostilities were limited in scale and importante, and arose mainly from local quarrels over slave-raiding and pillaging in villages that were under the protection of the allied lineage. In disputes over the succession, which turned on at least one occasion into a major war, the Akogu and Ocholi lineages traditionally fought on the same side and assisted each other against the combined strength of the Ajaku federation.

Outside the capital, the lineages of the ruling house are so widely dispersed through the metropolitan area and beyond, and their pattern of distribution changes so much from one generation to the next, that it is difficult to delimit their res-pective spheres of influence geographically. The Igala simplify this problem by associating each of the lineages with a centre where it is in the majority and where one or more past kings have lived before coming to the throne.

In 1907, for instance, Capt. Byng-Hall wrote that:

"The first Atta of Idah after Aiyagba setfled at Akwacha.

The succession now runs:

1. Chief of Arapa.

2. Chief of Akwacha.

3. Chief of Onupu.

4. Chief of Alo.

The villages named in this list coincide with names that are often quoted by the Igala today in describing the geographical distribution of the ruling lineages. If for the term 'chief of in Byng-Hall's notes we substitute the phrase 'head of the lineage associated with his list is directly comparable with one that frequently recurs in my own field notes, as follows:

Lineage

Lineage centre

District (in modern system)

Itodo Aduga

Ayangba

Okura

Amacho

Onukpo

Dekina

Akogu

Arapa

Biraidu

Akpacha

Ife

Ocholi

Igaliuwo

Alo

Itobe

TABLE 2. THE GEOGRAPHICAL LOCATION OF LINEAGES IN THE RULING SUBCLAN

The actual pattern of distribution of the royal lineages is infinitely more complex than this description suggests, and there are today no moral, ritual, or political forms of leader-ship which make these villages predominant within the settle-ments of the lineages concerned. The association between the lineages and these hypothetical lineage centres is an historical one; they represent the final positions taken up by the differ-ent lineages at the end of a bitter war fought in the last cen-tury, and probably developed as the fortified sites from which each lineage conducted its own campaigns at that time.

In the early part of the nineteenth century the Ata Ekalaga was mitardered at the end of an exceptionally long reign by members of the next lineage in the line of succession. His supporters, belonging to the Akogu lineage, were expelled Amocheję came to the throne around the year 1835, and it is probable that Ekalaga's death occurred in this same decade. The Igala explain the assassination of Ekalaga by saying that his reign was excessively long and blocked the Ajaku faction's succession to the kingship and other offices.

 In the capital, the Ajaku group pressed the other two lineages northwards, and consolidated its hold on the ridge of high land that runs north from Idah and debouches in the central area at Ayangba. Mean-while the Ocholi lineage fell back on the river Niger, and took up a defensive position at Alo, about 8 miles from the left bank near Itobe. Their allies, the Akogu lineage, bore the main brunt of the fighting and were driven still further northwards into the hilly regions of what is now the Ife District, and also to the north-east of the Ocholi positions, in the Abocho area.

When peace was established the lineages dispersed from these centres, and in the following reigns continued the pattern of founding widely separated small hamlets which is more typical of the pattern of royal settlement than the pattern of concentration for defensive purposes. In Amaga's reign, when the Akogu group returned to power there were further hostilities when the king took reprisals against Ayangba for the part that its people had played in harrying the Akogu settlements throughout the metropolitan area after Ekalaga's murder. This reign re-established Akogu and Ocholi settlers in the vicinity of Idah. For instance, some of Amaga's own sons settled at Gwolawo, lying between the capital and Ayangba, and their homesteads attracted further groups of settlers from their own lineage and from the Ocholi lineage. Similarly, the area between Idah and the Anambra river came within the coalition's sphere of influence at this time, with the Ocholi lineage settling at Ogbogbo and Okoainwili and the Akogu group settling in the intervening villages of Oforachi, Achokpa, and Ikeſi. Since Amaga's reign, which ended in 1900, the need for defensive alliances within the ruling house has disappeared, and the settlements of the royal lineages have tended to legends attribute Ekalaga's longevity to powerful medicine and say that it was to overcome the power of these medicines that his enemies devised a ritual manner of putting him to death by suffocation. become so intermingled with one another that it is difficult to speak of spheres of influence in other than an historical sense. Broadly speaking, the Ocholi lineage tends to be settled along or near the left bank of the Niger, from the confluence down-stream to the southernmost reaches of the kingdom, in the Ibaji district. Akogu settlements are often found on a small scale in close proximity to the Ocholi hamlets, but the main centres of this lineage are situated along the northern margins of the metropolitan area. Ajaku settlements are most numerous on the main route north from the capital and tend to form a barrier between the outlying, and oldest, Akogu hamlets and the approaches to the capital. In the immediate vicinity of Idah homesteads of the different lineages often coexist in the same village, but the local landowning groups are traditionally oriented towards a particular branch of the ruling house. The fluidity that characterizes the pattern of royal settle-ment outside the capital is related to the mobility of royal settlement in the capital itself. Formerly, whenever the Ata died a physical transfer of population took place in connexion with the arrival of the new king. Most of the great offices of state, which are hereditary within the royal subclan, fell vacant with the king's death, and the previous holders were expected to vacate the capital to make way for the new king's own nominees. These offices were filled mainly by members of the king's own lineage, so the royal succession involved a transfer of power from one lineage to another, and was not confined to the king and his entourage. It is clear from informant's statements and from early administrative records that the exodus from the capital that took place on the king's death was not confined to the office holders in person, but affected most of the members of the dead king's lineage who were resident at the time in the capital. In 1901 Partridge commented on this dispersal and wrote:

The-district-and-town-(of-Idah)-are-always-occupied-by-the-reigning family, while the other three branches, not being per-mitted to live there, retreat into the interior. At the death of an Ata. Hence the discontent created by Ekalaga's long reign within the next lineage in line. Atta, then, a sort of 'double twilight' takes place the late reigning family with all their chiefs, followers, and slaves, have to leave the homes in which many of them have been born and brought up, and immigrate to towns in the 'bush' which they know only by name. At the same time the new reigning family 'come in' and their people settle in the compounds occupied by their forefathers four reigns ago.

There is no direct earlier reference to the exchange of lineages in the capital, but we can perhaps associate with this custom the marked fluctuations in the population of Idah that attracted the attention of the early explorers in the nineteenth century. With the advent of colonial rule royal patronage was so drastically curtailed that the exchange of lineages was modified, and each group began to build permanent houses in the capital. But administrative officers found it, difficult to eradicate the idea that office holders in the royal lineage changed with the appointment of a new king, and as late as 1929, when the Ata Atabo died, the District Officer found it necessary to recommend to the Resident that 'the custom of making a clean sweep of titled officials-be no longer followed'.

Within each of the royal lineages descent is traced by relatively exact stages through the male line from the group's founder, and much more importance is attached to genealogical links than in the wider groupings of subclan and clan. The sense of genealogical identity amongst members of the group is correspondingly stronger in the lineage than in these wider groupings. It is inconceivable to the Igala, for instance, that members of the same lineage should fight on opposite sides in. any armed conflict. And if homicide occurred within the lineage it would have grave ritual as well as social consequences." In the royal clan, also, the scale of its The offender would be handed over to the local district head who would send the case to the Ata' or the provincial chief for trial. But in addition offerings would be made locally to the ancestors and to the land, to 'restore the land', enuychi ane. An act of homicide committed by a member of the same subclan, from a different lineage, would have similar consequences. Igala oral traditions say that sacrifices were made to the land at the end of the royal war that followed the murder of the Ata Ekalaga. organization is so great that lineages are the widest units with an effective interest in property relations; the functions of burial and settlement of inheritance issues are performed by members of the same lineage rather than by the subclan or clan. Collateral branches of the lineage are opposed in this case in the way that collateral lineages are opposed within smaller clans. This ideal of unity and identity amongst members of the same lineage is also expressed in several aspects of marriage regulations. The royal lineages are exogamous groups, and marriage between members is condemned as ebita. Sexual intercourse between agnates is less strongly condemned than a permanent liaison, but the distinction is not an easy one to make in Igala. Ebita marriages involve no marriage payments and there is no formal transfer of rights from the wife's group to the husband's when this kind of liaison is formed. But paternity is determined by the mother's statement, and if, as often happens, she names a lover, the child can be claimed by the progenitor's clan on payment of a fine to the legal husband's group. Sexual intercourse, therefore, potentially creates a lien on children in the same way as marriage, and in a limited sense it is equivalent to marriage because it determines the primary clan affiliation of the woman's children. And the rule of exogamy covers this case in addition to the possibility of marriage contracted in the normal way. The same principle of lineage unity in relation to marriage underlies the Igala custom of inheriting rights over widows in the agnatic line. When a man dies his wives become oyogwu, 'wives of the estate' and the rights formerly held by the husband over their services are transferred to his agnatic kin. A reciprocal extension of this principle is the custom of sororal polygyny, edunola, which favours marriages with classificatory sisters. The actual incidence of this type of marriage is low in Igala, amounting in my experience to three or four marriages in several hundred cases. But it is much favoured in principle for a first or senior wife to promote a second marriage for her husband within her own lineage.

The royal lineages in the ruling house subdivide genealogically into major segments that have some individual corporate functions. And these segments divide further into minor or daily audiences held at the palace they are similarly privileged. The amomata sit to the right and left of the Ata, in close proximity, whilst the rest of the audience face the king, at a greater distance. When the Ata enters the audience chamber, the councillors rise and screen the throne momentarily by standing in a close circle around it with arms raised, so that the king can sit and arrange his robes unobserved. In daily life the high rank of the royal councillors was symbolized and maintained by a number of privileges. Traditionally, only they, the king, and the Achadu, had the right to wear a type of bright red cloth, ododo, to build rectangular houses, unyi aje, or to ride horses in the ceremonial processions that took place during the annual festivals. Wives were frequently given to them without payment, and they claimed the nominal right to appropriate the wife or daughter of any commoner in marriage by the act of sending a messenger to place the councillor's bead insignia over the woman's head. In the provinces the councillors were treated with great deference, as the Ata's highest representatives, and whenever they went on tour, it became the duty of each village head that they visited to collect materials for their food and drink and for gifts from each section of the village. Councillors out of office still enjoyed great prestige and became centres of politi-cal influence wherever they settled in the provinces.

These are the basic divisions of palace organization, and-also-represent the major categories into whięli shares of wine, kola, and other royal gifts were divided at feasts. When perquisites are being shared the senior member of each class receives the group's share and then determines the allocation to individual members on the basis of relative seniority. The high rank of the councillors was shared by their sons, and to a lesser extent by their grandsons, and in terms of rank, it is the whole group of the councillors and their children to-gether with the king's immediate descendants that is described by the term amomata. They had great influence in local politics. and, as I show later, their kinship network always constituted an avenue of communication with the central government, alternative to the formal chain of command based on terri-torial divisions. The daughters of councillors also share the rank of their father, and are known popularly as amom'ukpaihi 'children of power', because they frequently do not contract regular marriages, but stay with their patrikin and bring up their children as full members of their father's group. Their own children are in turn privileged and often mediate between local groups in the provinces and the central government..

Most, although not all of the titleholders in the king's council were formally responsible for collecting tribute, enw'ire, from a particular village or district that was tradi-tionally linked with each title. These areas of jurisdiction are shown in the third column of the list of councillors' titles. The councillor acted as intermediary, ohięgba, between the locality concerned and the central government in judicial, political, and other matters, and derived a regular income from this connexion in the form of a share of the tribute paid by the area under his jurisdiction. In considering the significance of this administrativecontrol it is important to remember that the councillors' closest and most influential connexions were with the localities dominated by their own major lineages. It is vital to distinguish the formal administrative network of control, whose limitations are discussed later, from the nexus of hereditary ties that formed the real basis of royal power. I discuss these hereditary aspects of royal authority later in more detail.

The titled councillors were also responsible for mobilizing the royal army in time of war, and on major expeditions accompanied the king into the field at the head of their own contingents. Descent and kinship ties were as important in this connexion as in the case of councillors' formal suzerainty over a particular locality. Councillors drew most of their support from the area in which their own major lineage was dispersed, and the political obligations of the outlying districts to assist the king in time of war were expressed to a considerable extent in terms of the kinship bonds created by marriage between the royal lineage segments and local communities. Within the council there are six titles that are regarded as senior to the others and take précedence over them. They are:

Amanata Ochai Ata

Odokina Ata Odoma Ata (Odomata),

Omakoji Ata Ohięmogbolo

One token of their seniority is that they formerly had traditional titled compounds, okete, at fixed points in the capital, whereas the other councillors were free to build on any site. The Amanata lived in the quarter known as Ukwaja, close to the Igala Mela sector; the Ochai Ata lived among some of the most important non-royal chiefs; the Odokina Ata and Omakoji Ata were close to the palace, among the settlement of royal eunuchs, and the Odorna Ata lived slightly further away at Idoma, between the palace and the Achadu's compound. The Ohiemogbolo's compound lay nearer the cliff, behind the residence of one of the titled eunuchs.

But seniority within the royal council was also a matter of the king's favour, and any member of the group admitted by the Ata to his circle of confidants enjoyed a status as high if not higher, than that of the nominally senior members. The corporate functions, of the council were more clearly defined than the duties of individual members, and there was so little specialization of offices that the different councillors' positions were virtually interchangeable. The councillors did not spend all their time in the capital, but usually also had connexions with a rural area where they might reside for months at a time. With the exception therefore of the two or three members who were especially favoured by the king, the composition of the council changed from day to day. There was an absolute duty to attend at the annual festivals, or at the king's pleasure, but constant attendance-at-court was expected only of the king's-favourites, and the other members seem to have been free to spend much of their time elsewhere.

Their distribution coincides with the principal entrances into the royal quarter. The history of the king's council under modern conditions illustrates the lack of any clear internal structure and the difficulty of assigning specific functions to the titled offices. In the first two decades little use seems to have been made of the traditional machinery of government, and Igala was administered directly by the divisional officers, working through professional District Heads, who were mainly non-Igala. The earliest reference to the composition of the Ata's council occurs around 1919, and says:

The big council consisted of the Attah's brothers, nine in number:

Amanata

Odomata

Umakoji Ata

Omalogba

Inalogu

Oshomogbolo

Oshomakube

Egbena

Ekpa

No attempt seems to have been made by the colonial government to systematize the composition of the council until 1929, when Clifford, as District Officer, was commissioned to reorganize the administration on traditional lines.

Writing of the council, Clifford said later:

The form of government exercised by the Atas was simple and effective. At the centre was the Ata himself as Priest/King with the Achadu as chief executive-then the hereditary officers of state (kinsmen of the reigning Ata) and a Council composed of the nine Igala chieftains...

The reorganized council included five royal titleholders and three Igala Mela members, who were all assigned departmental duties. Elsewhere Clifford described the council on

Igala Divisional Report, 1931.

• The duties were:

Achadu

Vice-President

Ochai Ata

Justice and Titles

Odonata

District Affairs and Tax

Amajitata

Departmental

Etemaihi

A co-opted member

Egena Ata

Supernumerary member, for missions to province

Odokina Ata

Supernumerary member, for missions to province Supernumerary member, for missions to province

Abaigbo

The Egena Ata and Odokina Ata were later stationed permanently in the provinces, as Resident Councillors at Ankpa and Dekina, whilst the other councillors without specific duties were employed as Touring Councillors. the provinces, the Onu Ogugu and the Onu Dekina.

It is clear from the balance of power in the traditional council and from the failure of attempts to develop the council into a system of specialized offices that the traditional titles were relatively undifferentiated. In so far as the council was hierarchical in its organization it reproduced the pattern of seniority that characterizes sibling groups in Igala. The council itself was in a sense a model sibling group, and is so regarded by the Igala themselves. But the question of seniority within the council, apart from a purely ceremonial ranking, turned on the king's favour so that the roles of senior and junior sibling which the Igala assign to the different titles are in the last resort nominal positions, and not predetermined. But it is also a feature of this system that the difference in status between the various titles was not great. The titled offices were the key links in the complex chains of descent and kin-ship ties that oriented outlying villages towards the centre. And the essential duty of a titled omata was to act as spokes-man for his own major lineage and for the wide kin group that centred upon this segment of the reigning group. The titled councillors commanded and mobilized the support of the villages in which their own agnatic kin were settled and with which this lineage group had hereditary connexions from the past. So that in the last resort any analysis of the workings of the traditional council is essentially concerned with the dis-tribution of power and offices between the various major segments of the maximal royal lineage. In appointing councillors from his own lineage the Ata chooses a fairly high proportion of his own siblings, nοι more distantly related than through a common grandfather or great-grandfather. Collateral branches are usually allocated around half the total number of offices available, so that the senior branch, of those in the direct line of succession, receives a larger share than any other lineage segment. To illustrate this point, figure 6 on pp. 72-3 gives an outline genealogy of the Ocholi lineage, showing succession to titles. The record is not complete, because knowledge of the succession was scanty, and the facts had to be pieced together from widely scattered sources. But this reconstruction does show that the senior branch has a share in most of the titles, and shows also that offices could be transferred from one line to another or from the line in which they were usually hereditary to the senior branch of the lineage. Table 4 summarizes the number of turns taken by each of the five major lineage segments in respect of

the different titles. Although they are not hereditary in an exclusive sense, titles nevertheless tend to be regarded as the hereditary property of major segments, or major lineages. Allegiance to the title-holder, and his descendants, creates an important bond of unity within the lineage segment, and gives a focus to cor-porate activity which would otherwise be lacking. The title holder mediates between the segment members and the central government, and his house at Idah serves as an unofficial méeting place for the group. In Ocholi lineage, for instance, the late Odomata, Agono Okenya, was the undisputed head of the lineage segment founded by Ohięmini Gadagu; its members were sometimes described, by the rest of the lineage, as abo Odomata, the Odomata's people. At the annual royal festivals, the descendants of Ohiemini Gadagu were reunited dispersed between Itobe and Idah. The lineage founded by Ogala has associations with Shintaku and Adenekpa, near the Niger-Benue confluence. But its members are also settled in the metropolitan area and the Ibaji District. The other major lineages are similarly widely scattered, and can only be located spatially by referring to the homes of the reigning or the most recent titleholder from that particular group.

Within the major lineage the settlements of the titleholders give rise in time to smaller, localized groupings that can appropriately be termed localized lineages. I give examples of this type of grouping in discussing local communities and need only say here that in structure they are shallow agnatic groups structured around the descendants of a titleholder. Under modern conditions these units frequently have a small section based in the capital in addition to the parent body situated in an outlying district. Leadership within them tends to be inherited by the titleholder's children in order of seniority and then by their own eldest children in turn. But the groups are so shortlived that they tend to disintegrate before the third generation is reached; the sons of the founder inherit something of his high political status but this impetus is lost by the next generation unless a new political office is achieved.

Localized lineages do not form part of a total system of segmentation in the same way as major lineages within the maximal lineages or as maximal lineages within the subclan. They are as it were the nuclei into which the major lineages contract in the period when the maximal lineage is out of office, and from which the corporate organization of the major lineage can grow again when the whole lineage resumes office. But relations between localized lineages are competitive in the long term to the extent that different groups may be favoured when a new king is appointed from the maximal lineage to which they belong. For instance, within the Ohiemini major lineage in the Ocholi group the whole major segment was in Oboni's reign oriented towards the Oguma group which held the title of Ohiemogbolo. But when the Ocholi lineage ne came to the throne, in Ame's reign, a rival minor lineage, led by Okenyi Aga's son at Ofenya, 'took over a new title, Odo-nata, and became the leaders of the Ohiemini segment. The next localized lineage centred on Oguma was given no title in thiy reign, but still retains the prestige of having supplied two holders of the Ohiemogbolo title in previous reigns. There are other groups of similar scale within this major lineage, but the struggle for leadership of the segment has become focused on the relations of these two titled groups, and the other units have not attained the same degree of corporate organization or corporate status. They would be better described as ez tended families rather than localized lineages, since their uniry is domestic and not political in character, But, potentially, they are capable of taking on the same functions, Le, of achieving the same degree of cohesion if their leaders manage to achieve royal favour at any time in the future.

The royal subclans centred in outlying districts of the king dom divide in the same way as the Idah group into maximal lineages that are the basic units of a rotating tile system These maximal lineages are exogamous groups and show the same kind of unity in relation to marriage as the equivalent segments of the ruling subclan at Idah. They are likewise associated with different centres, so that the genealogical rela tionship between the different segments can be represented by a spatial model. But as in the metropolitan area this associa tion with different centres is mainly an historical one, and the maximal lineages are are in fact widely dispersed. In the following table the segmentation of the different subclans is summarized by showing the number of maximal lineages into which each group divides. The last two columns show the names of the lineages and their respective centres.

Subclan (Title-in brackets)

No. of lineages

Lineage names

Lineage centres

Ankpa-(Onu Ankpa)

Onoja Ikoja Aba Ojogobi

Abacy Acerane Ofugo Ojogobi

Dekina (Achema Ajobil

4

2

Akomata Ogado

Subelan (Title in brackets)

No. of linéages.

Imani (Onu Imani)

Ojoku (Onu Ojoku)

8 (in two groups of four)

Okpo (Obaje Atabaka)

Iga (Ojibo Akpoto)

4

Ogugu (Ochogugu)

2

Ikka (Onu Aloga)

Okenyi (Onu Okenyi)

4

Lineage names

Ohimu Okorikoto Onyuke Adede

4

Ekçle Anawo Oko

Enyikwole Ekelę Onegelegu Eje Agbane Oboni

Obaje-odo Odemu Ochiguma Onojaka Ote

Omoko Ede Abodiga Ake

Onoja Obekpa

Ame-Alebe Idoko Okpano Idagba Ogilijaja

4

Ujakpabana Ogbogiri Okpe Oduma

In some cases maximal lineages that participate in the title succession have been assimilated into the subclan from else-where, and are not members of the descent group in the strict

Lineage centres

Agariga

Ekekpe

Abo (Imaniabo)

Ogenç Ago

Bagele

5

Ogwowala

Okaba

Onupi

Utalu

Ojoko Ojeje

Inyologu

Abodo

Agbaduma

Okpe

Ocheku

Etutękpe

Okpo

Ogenę

Ocheba Agala

Ocheba

Omagaba Okegbi

genealogical sense. For instance, four out of the five lineages making up the Okpo subclan trace their descent frofit Oguche Ekwo, who also founded the Ankpa subclan. But the fifth lineage in the succession, named after its founder, Obaje Ote, traces its descent to the Ajaku branch this instance the assimilated lineage belongs belongs ultimately to the same clan, but in other cases of assimilation the immigrants are of alien clan origin. For instance one of the four lineages in the Ikka subclan is of Ibo origin and claims that its founder was rescued from slavery by the the founder of one of the other lineages. In another case, in the Imani group, the Onyuka Jineage was founded by a daughter of the first chief, and its members are technically uterine kin, omonobule, and not agnates. They nevertheless take an equal share in the rotation, and qualify politically for full membership of the descent group. It is in cases of this kind that the fiction of common descent is used by classifying members as children of the title rather than as children of a particular person. The category amomonu 'children of the chief', assumes a different kind of unity from that obtaining within the maximal lineage, and adapts the descent model to refer to a kind of positional succession.

Maximal lineages can also arise in these outlying subclans from fission of an existing group, in the same way that the Ajaku branch of the ruling subclan at Idah has bifurcated. As the following genealogy of the Ankpa subclan shows the Tkoja and and Ojogobi lineages are genealogically major rather than maximal lineages and both belong to one lineage. Adanawo which is equivalent in order of segmentation to the other two maximal lineages, Onoja and Aba. But the four groups have become equivalent politically, and the two divisions of the Anawo lineage rank as maximal lineages in the title succes-sion. There is moreover within the same segment sa third linete. Ekelę Onugbaję, that claims equal rights in the suc-cession, and if this claim is allowed in future it will expand the existing rotation into a system of five lineages.

On being appointed, the heads of the provincial royal sub-clans select councillors from the members of their own maxi-mal lineage and appoint them to a range of minor titles that are hereditary within the subclan concerned. This system of governance was limited by the segmentary nature of the clan organization, which opposed units of similar scale to one another at dif ferent levels on a descending scale. Within the clan as a whole the king's political functions were paralleled by those of the other subclan heads who were to a considerable extent in-dependent of his authority. Similarly, within each subclan the division into maximal lineages tended to restrict the effec tive administrative authority of the subclan head to the area controlled by his own lineage. At each level of segmentation the units were defined by contraposition to units of similar - scale and functions, their mutual relations were competitive and often hostile, and the units merged only in relation to the pattern of division existing in the order above them. The Igala descent system is, however, unlike a segmentary lineage system in the emphasis that it places on relative seniority within the system. In each order of segmentation the units that emerge are 'ranked in a pattern of seniority. At the highest level of division this pattern is fixed so that the royal house of amajata is permanently in the ascendant. But in the lower orders of segmentation the principle of seniority is applied on a rotating basis or on the basis of age so that the locus of authority shifts in each generation, and the units.take it in turn to exercise rights that are vested ultimately in the whole group.

The balance of power between the major segments of the royal clan is maintained ultimately by the groups that control the succession to their principal corporate offices. The king exercises this control over appointments to the headship of the provincial royal subclans, but his own office is subject in this respect to the authority of the kingmakers at Idah, who control the election of a new king and umpire the rotation of authority amongst the four maximal lineages of the ruling house. In the royal provinces there are non-royal clans whose. functions are analogous to those of the kingmakers at Idah, and who hold a similar range of titles. But their authority over the provincial chiefs is ritual and moral in character rather than political, since succession to these provincial offices is determined within the hierarchical structure of authority run-ning through the royal clan as a whole. To illustrate the king-makers' functions in the political system follow the same pattern as before of considering the metropolitan area in detail, followed by a brief comparison of this system with the provincial pattern. Igala oral tradition explains the role of the kingmakers in the-state-system-by postulating a transfer of political sovereignty from the original landowners at Idah to the immigrant founders of the royal clan. As we saw earlier, this transfer created a contractual relationship between the one group of clans and the royal descent group. But the modification of this primeval arrangement is acknowledged by the assimila-tion of another immigrant group which consolidated the indi-genous landowning clans under the authority of its own head, the Achadu. The two stages of this myth summarize the powers that are hereditary in the federation of kingmaking clans, and validate the dominance of the Achadu's descent group within the federation.

The Igala Mela clans are nominally nine in number, and, in principle also, the descendants of the autochthonous popula-tion of Idah. In practice neither the group's composition nor its membership are as rigidly exclusive as these ideals imply. At least eleven clans lay claim to the status of belonging to this group. And again, the rule of indigenous origin is sufficiently flexible to allow one clan with traditions of immigrant origin to be regarded by all Igala as a full member of the federation. One of the reasons for this discrepancy between the norm and the reality is that the term, 'Igala Mela', refers to spatial as well as to structural arrangements. It describes a sector of the capital lying between the inner and the outer defences of the old city, occupied by a cluster of homesteads belonging to the Igala Mela clan heads. Any titleholder whose family has lived for more than one generation on this low ridge of land Iying between Igalogwa and Ukwaja tends to be classed by the people of Idah as belonging to the Igala Mela. The Aleji clan, for instance, which was originally of Igbirra stock, has become fully assimilated through long residence. The Ochijenu clan, on the other hand, claims to have been physically established there for several generations, and is regarded by members of other clan groups as Igala Mela, but is not yet fully accepted by the Igala Mela themselves. Finally, there is the example of a clan, taking the title of Achadu Kikilį Ukwaja, which has moved from the quarter and is in the process of losing its Igala Mela affiliation, although it is still regarded as a full member in connexion with the kingmakers' ceremonial duties. A-list of the eleven clans claiming Igala Mela status, is given in the table on pp. 84-5. This table also summarizes the salutations that are used in greeting the men and women of each clan, and shows, in the last column, the elifo olopu that has to be observed by all members of the clan in each case. With the addition of the Achadu's own clan, which is shown separately at the foot of the table, the clans listed comprise the complete federation of kingmakers.

The clans in the Igala Mela are individually much smaller in scale than the royal house at Idah to which they are opposed politically. This is true of their genealogical span and depth, as can be seen from the following summary of the number of generations and maximal lineages, in each of the clans.

Clan

Nunber of maximal lineages

Generations

Protodynastic ancestors

Etemaihi

5

4

Onubiogbo

2

4

33--

5

Okwęję

2

6

4 (?)

Unana

Agbenyo

6

Obajadaka

2

4

5

Achadu Kikilį Ukwaja

4

6

4

Ochijenu

3

9

2

Onęde

3

5

Clan tille

Title

Women

Eja4ęmaihi

Takida

Onubiogbo

Aguba

Clan salutations Men

Ohiowa

Onu Igala

Ugodo

Onu Igala

Ritual prohibitions

Flesh of monitor lizard, abaji

Flesh of crocodile, ortye Id Clan title

Clan salutations

Title

Men

Women

Ritual prohibitions

Onęde

Ojogobi

Onu Igala

Onu Igala

Use of ugbakolo or ikpokpo wood for firewood

Aleji

Ebije

Ahiebu

Anyebu

Forbidden for a needle, role, to be brought in, contact with food, or named when a member of the clan is eating

Okweje

Ede

Ahiebu

Anyebu

Flesh of the ground-squirrel, okweję

Unana

Onu Igala

Ahiebu

Anychu

Flesh of palm civet, ewolo

Agbenyo

Onu Igala

Oyowo

Oyowo

Forbidden to use palm leaf fans, utowo ekpe. Also forbidden to use three-forked sticks 25 pot supports

Achanyuwo

Ede

Ahiebu

Anyebu

Flesh of palm civet, ewolo, and wood of ugbakolo tree

Obajadaka

Doga

Qyo

Oyo

Forbidden to touch food with an unhafted knife, ukpakelle obe

Ochijenu

Ude

Oohi

Onyete

Forbidden to use palm fronds, ak-paga, for fire-wood

Achadu

Agukili

Ohiukwa

Aina

Kikili Ukwaja

Aleaf vegetable, oro dudu, and flesh of ouwe fish

Achadu

Anu

Omanu,

Ugodo

The ouwe fish

The corresponding numbers for the ruling subclan are four maximal lineages, seven generations, and at least four proto-dynastic ancestors.

The difference in scale between this group and the royal house is also reflected in the fact that the clan as a whole is exogamous in each case, instead of the maximal lineages. Finally, there is an associated difference in the numerical size of the Igala Mela clans which individually contain no more than a few hundred adults each. One can conclude that the Igala Mela clans, in size and scale, are more closely compar-able with the maximal or major lineages of the ruling house than with the whole subclan. But they are distinct descent groups in each case, with the same corporate functions and tokens of identity as the royal clan itself. They have no tradi-tions of common origin in the genealogical sense, but form a closely united federation whose members are bound together by a strong sense of historical unity and common political opposition to the reigning royal subclan. Within the federation of kingmakers' clans there is a marked discrepancy of size and scale between the Igala Mela clans and the Achadu's own descent group, which is closely comparable with the ruling house. In the same way that the royal descent group, in virtue of its size and dispersal co-ordinates the cor-porate activities of small localized clans throughout the king-dom into a unified system, so the Achadu's clan performs a co-ordinating role within the much more restricted field of kingmakers' activities. The Achadu's dominance and leader-ship of the kingmakers is a function of his clan's preponder-ance within this group of clans, which ultimately provides the Achadu with a greater command of human and physical re-sources than the heads of the Igala Mela descent groups.

The Achadu's title is hereditary, as figure 8 shows, within a group of four lineages who trace their descent from a com-mon male ancestor. Omeppa. As in the case of the Idah and Ankpa subclans in the roval descent group. two of the lineages have been created by a process of segmentation within a unit of the same order as the other two major branches. They now form maximal lineages with the same functions as the older branches in the rotation, and can inter-marry with one another or with either of the older branches. All four lineages are widely dispersed in the modern districts of Adoru, Igala Mela, and Igalogwa. But they have individual historical associations with different villages, and their settlements are to some extent concentrated around the following centres.

Lineage

Centres

Onogu

Ojuwo (Adoru District)

Abutu Ejibo

Alaba (Gwolawo District) Adoru (Adoru District)

Atiko

Alloma (Alloma; Igalogwa District) Okula (Igalogwa District)

Akpadodo (Igalogwa District)

Inedu Otubi

Igoti (Igalogwa District) Gbamaka (Igalogwa District) Agojeju (Igalogwa District) Mamerębo (Igalogwa District)

Omeppa

Atokwo

Inedu Otubi

Ame Aranejo

Onogu

Obera Ajuma

Atiko

Abutu Ejibo

Ogohi

Odo

Ogala Omere