Community, Clan, and Village Systems and Chieftainship

Communities, Clans, and Villages.

The native populations of the Mafulu area are scattered about in small groups or clusters of villages or hamlets; and, as each cluster of villages is for many purposes a composite and connected whole, I propose to call such a cluster a “community.” Friendships, based on proximity and frequent intercourse and intermarriage, doubtless arise between neighbouring communities, but otherwise there does not appear to be any idea in the minds of the people of any general relationship or common interest between these various communities of the area. Each community regards the members of every other community within the area as outsiders, just as much so as are, say, the Ambo people to the north and the Kuni people to the west. If a community, or group of communities together, were the subject of an attack from either Ambo or Kuni natives, each of these being people whose language is different—as regards the Kuni utterly different—from that of the Mafulu, there would apparently be no thought of other Mafulu-speaking communities, as such, coming to assist in repelling the attack. Hence in dealing with the question of inter-village relationship, I have to fix my mind mainly upon the community and its constituent parts.

Concerning the situation as between one community and another, as they regard themselves as quite distinct and unrelated, the only question which seems to arise is that of the ownership of, and rights over, the intervening bush and other land. The boundaries between what is regarded as the preserve of one community, within which its members may hunt and fish, clear for garden purposes, cut timber, and collect fruit, and that of an adjoining community are perfectly well known. The longitudinal boundaries along the valleys are almost always the rivers and streams, which form good boundary marks; but those across the hills and ridges from stream to stream are, I was told, equally defined in the minds of the natives, though no artificial boundary marks are visible. These boundaries are mutually respected, and trouble and fighting over boundary and trespass questions are, I was told, practically unknown, the people in this respect differing from those of Mekeo.

A community comprises several villages, the number of which may vary from, say, two to eight. But the relationship between all the villages is not identical. There is a clan system, and there is generally more than one clan in a community. Often there are three or more of such clans. Each clan, however, has its own villages, or sometimes one village only, within the community, and two clans are never found represented in any one village,1 or any one clan spread over two or more communities.

Fig. 3 is a diagrammatic illustration of a typical Mafulu community, the circles representing villages of one clan, the squares those of another clan, and the triangle being the sole village of a third clan.

I have said that the entire community is for many purposes a composite whole. In many matters they act together as a community. This is especially so as regards the big feast, which I shall describe hereafter. It is so also to a large extent in some other ceremonies and in the organisation of hunting and fishing parties and sometimes in fighting. And the community as a whole has its boundaries, within which are the general community rights of hunting, fishing, etc., as above stated.

But the relationship between a group of villages of any one clan within the community is of a much closer and more intimate character than is that of the community as a whole. These villages of one clan have a common amidi or chief, a common emone or clubhouse, and a practice of mutual support and help in fighting for redress of injury to one or more of the individual members; and there is a special social relationship between their members, and in particular clan exogamy prevails with them, marriages between people of the same clan, even though in different villages, being reprobated almost as much as are marriages between people of the same village.

The Mafulu word for village is emi, but there are no words signifying the idea of a community of villages and that of a group of villages belonging to the same clan within that community. As regards the latter there is the word imbele, but this word is used to express the intimate social relationship existing between the members of a clan, and not to express the idea of an actual group of villages. Communities and villages have geographical names. The name adopted for a community will probably be the name of some adjoining river or ridge. That adopted for a village will probably be the name of the exact crest or spot on which it is placed, the minuteness of the geographical nomenclature here being remarkable. Clan-groups of villages, forming part of a community, have, as such, no geographical names, but a member of one such group will distinguish himself from those of another group by saying that he is a man of———, giving the name of the chief of the clan occupying the group.

I was assured that, when there are two or more villages of a clan with a common chief and emone, they have originally been one village which has split up, an event which undoubtedly does in fact take place; while on the other hand the several villages of a clan, presumably the outcome of a previous splitting-up of a single village, will sometimes amalgamate together into one village, which thus becomes the only village of the clan. But two villages of different clans could never amalgamate in this way. The following are examples of these village changes:—

Near to the Mafulu Mission station is a community called Sivu, which includes seven villages occupied by three clans, as follows2:—

1. Voitele	Belonging to a clan whose chief, Jaria, lives at Amalala, where the clan emone is.
2. Amalala
3. Kodo-Malabe
4. Motaligo
5. Malala	Belonging to a clan whose chief, Gito-iola, lived at Malala, where the clan emone is. (He has recently retired in favour of his eldest son, Anum’ Iva, who is the present chief, and also lives there.)
6. Gelva
7. Seluku	Being the only village of a clan whose chief, Baiva, has recently died. His eldest son, who has succeeded him, is an infant. There is no regency.

Also near the Mission station is a community called Alo, which includes four villages occupied by two clans, as follows:—

1. Asida	Belonging to a clan whose chief, Amo-Kau, lives at Asida, where the emone is.
2. Kotsi
3. Ingomaunda
4. Uvande	Being the only village of a clan whose chief is Iu-Baibe.

Referring to these villages, in the year 1899 the clan now occupying the four villages Voitele, Amalala, Kodo-Malabe and Motaligo had only a single village, Kaidiabe, the clan’s chief being the above-mentioned Jaria. Then there was a Government punitive expedition, following the attack of the natives upon Monseigneur de Boismenu (the present Bishop of the Mission of the Sacred Heart in British New Guinea) and his friends, who were making their first exploration of the district, in which expedition a number of natives, including the brother of the chief, were killed. After that the village was abandoned, and the three villages of Voitele, Amalala and Motaligo arose in its place. Subsequently after a big feast, which was held at Amalala in the year 1909, that village put out an offshoot, which is the present village of Kodo-Malabe. Also in the year 1909 the village of Uvande was represented by seven villages, all belonging to one clan under the chieftainship of Iu-Baibe, the names of which were Ipolo, Olona, Isisibei, Valamenga, Amada, Angasabe and Amambu; but after the feast above mentioned the people of that clan all abandoned their villages, and joined together in forming the present village of Uvande.

The chief, that is the true chief, of a clan has his house in one of the villages of the clan, and if, as sometimes occurs, he has houses in two or more of these villages, there is one village in which is what is regarded as his usual residence, and this is the village in which is the emone of the clan.

As regards the relative predominance of the various clans of a community and their respective chiefs in matters affecting the whole community (e.g., the arranging and holding of a big feast), there is no rule or system. The predominance will probably, unless there be a great disparity in the actual size or importance of the clans, and perhaps even to a certain extent notwithstanding such a disparity, fall to the clan whose chief by his superior ability or courage or force of character, or perhaps capacity for palavering, has succeeded in securing for himself a predominating influence in the community.

The word imbele and certain other words are used to designate the closeness or otherwise of the connection between individuals. Imbele signifies the close connection which exists between members of one clan, and a man will say of another member of his clan that he is his imbele. The word bilage signifies a community connection, which is recognised as being not so close as a clan connection; and a man will say of another, who is outside his own clan, but is a member of his own community, that he is his bilage. The expression a-gata signifies absence of any connection, and a man will refer to a member of another community, Mafulu, Kuni, Ambo, or anything else (there is no distinction between these in the use of the term) as being a-gata, thereby meaning that he is an outsider.

This brings me to the question of the use by me of the term “clan” to designate the intimate association above referred to. To begin with, there is a considerable difference between the situation produced by the clan system, if it may be regarded as such, of Mafulu and that of, say, Mekeo, where one finds several clans occupying one village, and where members of one clan may be scattered over several more or less distant villages; though this latter difference might perhaps arise in part from natural geographical causes, the flat lowlands of the Mekeo people being highly favourable to inter-village communication over their whole areas, and to the holding of their recognised and numerous markets, whilst it may almost be assumed that such intercommunication would be more restricted, at all events in days gone by, among the Mafulu inhabitants of the mountains.

Then again in Mafulu there are no clan badges, nor are there any realistic or conventional representations of, or designs which can to my mind be possibly regarded as representing, or having had their origin in the representation of, animals, birds, fishes, plants, or anything else. As regards this, however, it may be mentioned that the Mafulu people are very primitive and undeveloped, and have not in their art any designs which could readily partake of this imitative character, their artistic efforts never producing curves, and indeed not going beyond geometric designs composed of straight lines, rectangular and zig-zag patterns and spots.

Also I was unable to discover the faintest trace of any idea which might be regarded as being totemistic, or having a totemistic origin. In particular, although enquiry was made from ten independent and trustworthy native sources, I could not find a trace of any system of general clan taboo against the killing or the eating of any animal, bird, fish, or plant. It is true that there are various temporary food taboos associated with special conditions and events, and that there are certain things the eating of which is regarded as permanently taboo to certain individuals; but the former of these restrictions are general and are not associated with particular clans or communities, and the latter restrictions relate separately to the individuals only, and apparently are based in each case on the fact that the food has been found to disagree with him; though whether the restriction is the result of mere common sense based upon individual experience, or has in it an element of superstition as to something which may be harmful to the individual concerned, is a point upon which I could not get satisfactory explanation.

Again, still dealing with the question of totemism, I may say that the community and village names (as already stated, there are no clan names) do not appear to be referable to any possible totemistic objects. There is no specific ancestor worship, in connection with which I could endeavour to trace out an association between that ancestor and a totemistic object, and there is no special reverence paid to any animal or vegetable, except certain trees and creepers, the fear of which is associated with spirits and ghosts generally, and not with ghosts of individual persons, and except as regards omen superstitions concerning flying foxes and fireflies, which are general and universal among all these people, and except as regards the possible imitative character of the Mafulu dancing, which, if existent, is probably also universal.

Moreover, I was told that now, at any rate, the people regard their imbele or clan relationship as a social one, as well as one of actual blood, a statement which is illustrated by the fact that, if a member of one clan leaves his village to reside permanently in a village of another clan, he will regard the members of the latter clan, and will himself be regarded by them, as being imbele, although he does not part with the continuing imbele connection between himself and the other members of his original clan.

On the other hand the association between members of a clan is exceedingly close, so much so that a serious injury done by an outsider to one member of a clan (e.g., his murder, or the case of his wife eloping with a stranger and her family refusing to compensate him for the price which he had paid for her on marriage) is taken up by the entire clan, who will join the injured individual in full force to inflict retribution; and, as already stated, the members of a clan share in one common chief and one common emone, intermarriage between them is regarded as wrong, and apparently each group of villages occupied by a single clan has in origin been a single village, and may well have a common descent. I think, therefore, that I am justified in regarding these internal sections of a community as clans.

Chiefs, Sub-Chiefs and Notables and Their Emone

At the head of each clan is the amidi, or chief of the clan. He is, and is recognised as being, the only true chief.

He is the most important personage of his clan, and is treated with the respect due to his office; but, though he takes a leading part in all matters affecting the clan, he is not a person with any administrative or judicial functions, and he has no power of punishment or control over the members of the clan. In public ceremonial matters of importance, however, he has functions which rest primarily upon him alone, and he does, in fact, always perform these functions in his own village; and on the occasion of a big feast (as to which see below), he does so in whatever village of the clan that feast may be held.

The chief lives in one of the villages of the clan, but may have houses in other villages of that clan also. In the village in which he mainly resides is his emone or club-house, which is the only true emone of the clan; and for the upkeep and repair of this he is responsible. This is the ceremonial emone in his own village, and is always the one used in connection with the ceremony of a big feast in any village of the clan; and, if the feast be held in a village other than that in which is his then existing emone, another one is built in that village in lieu of his former one in the other village.

There is not in connection with these chiefs and their ceremonies any distinctive difference in importance between the right and the left as regards the positions occupied by them on the emone platform or the structure of the emone, such as is found among the Roro people.

Next in rank to the chief, and at the head of each village of the clan, there is a sub-chief, or em’ u babe, this term meaning “father of the village.” He is not regarded as a true chief, but he is entitled, and it is his duty, to perform in his own village all the functions of the chief, except those connected with the big feast. He and the similar sub-chiefs of the other villages of the clan are the persons who take the prominent part in supporting the chief in any ceremonial function concerning the whole clan in which the latter may be engaged, and in particular at the big feast. The em’ u babe is usually a relative of the chief, and at all events is an important personage. He also has in his own village his emone, which is the principal emone of that village, and is used for all ceremonial functions in that village except the big feast, but it is not regarded as being a true emone. The chief holds in his own village of residence both his office of amidi and that of em’ u babe, there being no other person holding the latter office in that village.

Next in rank to the sub-chiefs come a number of ake baibe, which means “great men.” These are the leading people—the aristocracy—of the clan. There are no distinctive social grades of rank among them. Their number is often very large in proportion to the total number of male inhabitants of a village; indeed sometimes almost every member of a village will claim to belong to this class. These people are in no sense office-bearers, and have no special duties to perform, though on a ceremonial occasion they are entitled to have their importance borne in mind. Each of them also is entitled to have an emone (here again not a true emone) in his village, but in fact their numbers often make this practically impossible, and you rarely see more than two or three emone in one village.

The above are all the chiefs and notables of the clan. There is no such thing as a war chief.

Aristocracy in its various forms is not a condition to which a man attains on getting older—it is attained by inheritance.

The office of the chief is hereditary in the male line by strict rules of descent and primogeniture. On the death of a chief his office descends to his eldest son, or if that son has died leaving children, it descends to the eldest son of that son, and so on for subsequent generations. Failing the eldest son or male issue in the male line of the eldest son, the office devolves upon the late chiefs second son or his male issue in the male line. And so on for other sons and their issue. Failing such male issue the office passes to a collateral relation of the late chief on his father’s side (e.g., the late chief’s next eldest brother or that brother’s son, or the late chief’s second brother or that brother’s son), the ascertainment of the devolution being based upon a general principle of nearest male relationship in the male line and primogeniture.3

The chief holds his office for life, but he may in his lifetime resign it in favour of the person entitled to succeed him, and this in fact often occurs. He cannot, however, on the appointment of his successor still continue in office himself, so as to create a joint chieftainship, as is done in Mekeo. He, as chief, is subject to no special taboo, and there is no qualification for office, other, of course, than hereditary right; but no chief can perform the functions of his office, or build for himself an emone, until he has married. There is no ceremony on the chiefs accession to office on the death of his predecessor; but there is a ceremony (to be described hereafter) on a chief’s abdication in favour of his successor. Cases have, I was told, occurred in which a man has in one way or another forced himself into the position of chief, though not qualified by descent, and has thus become a chief, from whom subsequent chieftainship descent has been traced, but I could learn nothing of the circumstances under which this had occurred. Also it has happened that, when a chief has been weak, and has not asserted his position, a sub-chief has more or less usurped his power and influence, without actually upsetting his chieftainship or supplanting him in his performance of ceremonial duties.

If the chief on acquiring office by inheritance is a child, or not qualified to act (e.g., unmarried), he is nevertheless chief; but some person will usually act as his guardian, and perform his functions for him until he has qualified. This person will probably be one of the young chief’s eldest male paternal relations (e.g., the eldest living brother of the last previous chief), and will presumably be a person of consequence; but he will not necessarily be one of the sub-chiefs.

All the above observations concerning the hereditary nature of a chief’s office and subsequently explained matters apply also to the case of a sub-chief, except that there is no ceremony on his resigning office in favour of his successor, and that the usurpation of the office of a sub-chief, of the occurrence of which I found no record, would perhaps be more difficult of accomplishment. In the event of a village throwing off an offshoot village, or itself splitting up into two villages, the then existing sub-chief of the original village would continue his office in it or, in case of a division, in one of the villages resulting from the split, and the other village would have for its sub-chief some one of the ake-baibe of the original village, probably the one who was most active in organising the split. On the other hand, if several villages united into one, one only of their sub-chiefs could be sub-chief of the village arising from the amalgamation, and the others would sink to the rank of ake-baibe.

The observations concerning the hereditary nature of a chiefs rank also apply to the ake-baibe. I have no information concerning them on the other points; but these are not so important as regards these people, who have no official position and have no duties to perform.

There are, as will be seen hereafter, a number of persons who are employed from time to time to perform various acts and functions of a ceremonious or superstitious character, notably the man who has the important duty of killing pigs at feasts; but these men are not by virtue of their offices or functions either chiefs or sub-chiefs, or even notables or important personages. It is in each case a matter of the specific personal power which the man is believed to possess. Any of them might happen to be an important personage, and the pig-killer, whose office is a prominent one, would probably be one; though in his case muscular strength would, I understand, be an important element of qualification.4

Villages, Emone, Houses and Modes of Inter-Village Communication

Villages and Their Emone and Houses.

The Mafulu villages are generally situated on narrow plateaux or ridges, sloping down on each side; but the plateaux are not usually so narrow, nor the slopes so steep, as are those of the Kuni district, and the villages themselves are not generally so narrow, as the contour of the country does not involve these conditions to the same extent. Also the Mafulu villages are on the lower ridges only, and not on the high mountains; but the actual elevations above sea-level of these lower ridges are, I think, generally higher than those of the top ridges of the Kuni. Plate 54 shows the position and surroundings of the village of Salube (community of Auga), and is a good representative example, except that the plate does not show any open grassland.

The villages are, or were, protected with stockades and with pits outside the stockades, and sometimes with platforms on trees near the stockade boundaries, from which platforms the inhabitants can shoot and hurl stones upon an enemy climbing up the slope. The stockade is made of timber, is about 15 to 25 feet high, and is generally constructed in three or more parallel rows or lines, each of the lines having openings, but the openings never being opposite to one another. These protections have now, however, been largely, though not entirely, discontinued.1 It is, or was, also the practice, when expecting an attack, to put into the ground in the approaches to the village calthrop-like arrow-headed objects, with their points projecting upwards.

The average size of the villages is small compared with that of the large villages of Mekeo, some of them having only six or eight houses, though many villages have thirty houses, and some of them have fifty or sixty or more. The houses and emone are much smaller than those of Mekeo, and much ruder and simpler in construction and they have no carving or other decoration. There are no communal houses.

The houses are ranged in two parallel rows along the side of the ridge, with an open village space between them, the space being considerably longer than it is broad, and more or less irregular in shape. The houses are generally built with their door-openings facing inwards towards the village enclosure.

At one end of the village, and facing down the open space, is the chief’s or sub-chief’s emone. These are, like the Roro marea and the Mekeo ufu, used, not only in connection with ceremonies, but also as living houses for men, especially unmarried men, and for the accommodation of visitors to the village. There are probably also in the village the emone of one or more of the notables before mentioned, of which one will be at the other end of the village and any others will be among the houses at the side of, and facing into, the village enclosure. There are not often more than three emone, true or otherwise, in one village.

You of course do not find the surrounding palm groves of Mekeo and the coast; nor do you generally see the waste space behind the houses, or the ring of garden plots outside the waste space, the position of the village on its ridge being usually hardly adapted to the latter. You may, however, often find garden plots very near to the village. Each family has its own house, and, except as regards the emone and their use, there are no separate houses for men or women, or for any class of them.

The Mafulu emone is an oblong building, erected on piles of very varying height, the interior floor being anything from 3 to 15 feet above the ground. In size also it varies very much, but generally it is internally about 12 to 15 feet long from front to back, and about 8 to 12 feet in width. The roof, which is thatched with long, rather broad leaves, is constructed on the ridge and gable principle, with the gable ends facing the front and the back, and the roof sloping on both sides in convex curves from the ridge downwards. Remarkable and specially distinctive features of the building are the thatched roof appendages projecting from the tops of the two gable ends (front and back), the forms of which appendages are somewhat like a hood or the convex fan-shaped semicircular roof of an apse, and in construction are sometimes made as rounded overhanging continuations of the upper part of the roof, and sometimes as independent additions, not continuous with, and not forming parts of, the actual roof. In front of the building, but not at the back, is a platform at a level about a foot below that of the inner floor, extending the whole length of the front of the building, and projecting forwards to a distance of from 2 to 5 feet. The approach from the ground to this platform in the case of a high-built emone is a rudely constructed ladder, but when the building is only low and near the ground it is generally merely a rough sloping piece of tree trunk, or even only a stump. The two gable ends are enclosed with walls made of horizontal tree branches, two or three of which are, at both the front and rear ends of the building, discontinued for a short distance in the centre, so as to leave openings. These openings are, say, 2 feet or more above the level of the front outside platform, and 1 foot or more above that of the inside floor, and are usually very small; so that, in entering or leaving the building, you have to step up to, or even climb, and wriggle yourself through the opening, and then step down on the other side. Inside the building you find the centre of the floor space occupied by a longitudinal fireplace, about 2 feet broad, extending from front to back of the building; and the floors on each side of this fireplace slope upwards somewhat from the visible level of the fire-place towards the sides of the building. The fireplace part of the interior is, in fact, dropped to a level below that of the adjoining floors, so as to form a long trough, which is filled up with soil upon which the fire can burn; and it is the visible top level of this soil covering which is practically flush with the inside lower level of the adjacent upward-sloping floors. Some distance below the roof there is usually an open ceiling of reeds, used for the purpose of storing and drying fruits and other things, and especially, as will be seen hereafter, for drying fruit required in the preparation for the big feast.

Fig. 4 is a diagram of the front of an emone, disclosing the internal plan of the floor and fireplace, for which purpose the front hood of the roof and the front platform are omitted from the plan, and of the horizontal front timbers the third up from the bottom is shown at the ends only, the middle part being omitted, and small portions of the timbers immediately above them are omitted. The words in parentheses appearing in the explanatory notes to the figure are the Mafulu names for the various parts of the building.

Explanatory Notes to Fig. 4.

(a) Main posts, one at the front of the building, one in the middle, and one at the back (apopo).

(b) Posts supporting roof, a line of them running along each side (tedele).

(c) Posts supporting outer edge of flooring, a line of them on each side (emuje or aje).

(d) Post supporting inner edge of flooring and hearth, a line of them on each side (foj’ ul’ emuje).

(e) Lower ridge pole (tanguve).

(f) Main downward-sloping roof work, strongly made, going all the way back, only four or five of them on each side (loko-loko).

(g) Upper ridge pole (tope).

(h) Main horizontal roof work, resting on f (gegebe).

(i) Upper downward-sloping roof work, not so thick as f resting on h, going all the way back at intervals of about 1 foot (engala).

(k) Upper horizontal roof work, not so thick as h resting on i (gegebe)

(l) Thatch made of leaves (asase).

Note.—The roof (excluding the hood) projects forward and overhangs a little beyond the post a, so as to overhang the greater part, but not the whole, of the platform; the hood (not shown in this figure) is really intended to shelter the platform.

(m) Pole supporting roof (karia).

(n) Pole supporting outer edge of floor (karia).

(o) Pole supporting inner edge of floor and enclosing hearth (jakusube).

(p) Floor, composed of transverse woodwork (koimame) with thin light longitudinal lath work on top of it (ondovo).

(q) Pole above inner edge of floor and edging hearth, not so thick as o (bubuje).

(r) Floor of fireplace, upon which soil is put (foj’ ul maovo).

(s) Pieces of wood supported by c and d, going right across building and over floor of fireplace, but under its earth, all the way back (kooije).

(t) Wall timbers below top of door-opening, at front and back (kautape).

Note.—t(1)goes right across under door-opening, but the middle portion of it is omitted from the diagram, and the lower edges of timbers t (2) are partly broken off, so as to show floor and fireplace.

(u) Wall timbers above top of door-opening (dibindi).

Note.—t and u together-the whole wall-are called bou.

(v) Uprights bracing together t and u (Mafulu name unknown).

(w) Ceiling made with reeds and used for storing and drying fruit, etc. It may occupy the whole length of the building and the whole width of it, or part only of either or both of these (avale).

(x) Space filled up with soil and used as hearth (foje).

(y) Door-opening, one at back also (akomimbe).

Fig. 5 is a diagram of a transverse section across the centre of an emone, showing the internal construction. The explanatory note only deals with portions not explained in those to Fig. 4.

Explanatory Note to Fig. 5.

Post a is the main central support of the building corresponding with post a in Fig. 4. Posts b b are central side supports to the roof. Poles c and d are attached to posts a b b, and help to strengthen the fabric. These poles are also used for hanging up sleeping hammocks, the other extremities of which are hung to the loko-loko of the roof (Fig. 4, f). The name for post a is dudu, but this word is often used to express the whole structure a b b c d.

I have endeavoured in the diagrammatic sketch—Fig. 6—to illustrate the apse-like projection of the roof of an emone and the platform arrangements. I have in this sketch denuded the apse roof of its thatch, showing it in skeleton only; and I have shaded all timber work behind the platform, in order more clearly to define the latter.

Explanatory Notes to Fig. 6.

(a) Front end of thatch (asase) of main roof.

(b c d) Front apse-shaped roof (siafele), the thatch having been removed to show its internal construction.

(b c, b e, b d) Downward-sloping roof work (engala).

(f f, c d] Horizontal roof work (gegebe), carried round in curves.

Note.—Sometimes the apse-shaped roof is constructed as a continuation of the main roof of the building, in which case the gegebe of the former are a continuation of those of the latter. Sometimes the apse roof is a separate appendage, not connected with the main roof, and in that case the gegebe of the former are separate from those of the latter, and are fixed at their extremities to the loko-loko of the main roof.

(g) Posts supporting the platform (purum’-ul’ emuge).

(h) Horizontal platform supports resting at one end on g and at the other end fixed to either the tedele or the emuje.