Chapter XXIII: Appendix: A (1)

THE TECHNICAL CULTURE OF THE ANDAMAN ISLANDERS

In this appendix I shall give a brief account of the technical culture of the Andaman Islanders, with a few comparative notes on the technology of the Semang of the Malay Peninsula and the Negritos of the Philippine Islands. The Andamanese, the Semang and the Philippine Negritos are so similar in physical characteristics that it is reasonable to suppose that they are descended from a single stock. It is on the basis of this hypothesis that they are all spoken of as belonging to one race, the Negrito race. It is therefore of some interest to compare the culture of these three different peoples to see if we can determine what was the culture of their ancestors.

In such hypothetical reconstructions of the past it is necessary to proceed with extreme caution, as there is no means of controlling results. The method I have adopted is to compare first of all the different types of technological products or activities found in different parts of the Andamans in order to determine as far as possible what was the technical culture of the ancestors of the Andamanese when they first reached the islands, and what changes have taken place since the islands were occupied. It is only this primitive or generalised Andamanese culture that can be compared with that of the Semang or the Philippine Negritos.

From the point of view of technical culture the natives of the Andamans must be separated into two main divisions, which will be spoken of as the Great Andaman Division and the Little Andaman Division respectively [226]. The most plausible explanation of the differences of culture and language between these two divisions has been mentioned already. We must assume that when the islands were first peopled, or at some later time, the inhabitants of the Little Andaman became isolated from those of the Great Andaman. The language and the technical culture of each of the two groups underwent a number of changes during the many centuries that followed. At a much later date, after the differences between the two divisions had been developed, and probably not many centuries ago, a party or several parties of natives must have made their way from the Little Andaman as far as Rutland Island. Here they came in conflict with the natives of the Great Andaman Division, and in this way arose the antagonism between the Jarawa (the immigrants from the Little Andaman) and the other natives of the South Andaman (who formed in 1858 the Aka-Bea tribe), which has lasted down to the present day. We shall find that the technical culture of the Jarawa has been only very slightly influenced by contact with the natives of the Great Andaman Division, and therefore differs very little from that of the Little Andaman at the present day.

Primitive Andaman Culture
|
_____________/ \____________
/ / \
________/|\________ / \
/ | \ / \
N. Sentinel Little Jarawa Southern Northern
Andaman Group Group
\____________ ____________/ \_____ _____/
\/ \/
Little Andaman Division Great Andaman
Division

I have provisionally included the natives of the North Sentinel Island in the Little Andaman Division. The ground for so doing is that the form of bow in use in the North Sentinel Island is similar to that of the Little Andaman and unlike that of the Great Andaman. Unfortunately, almost nothing is known about the technology of the North Sentinel, and nothing whatever about the language. It is possible that the natives of this outlying islet have been isolated for many centuries from both the Little Andaman and the Great Andaman, and further information about them might show that their technology is different in important respects from that of the Little Andaman.

Within the Great Andaman Division there are a number of differences in technology between the tribes of the North Andaman and those of the South Andaman and Middle Andaman.

In order to render the exposition and argument that follows more easily understood the supposed relations of the different types of culture are shown in the form of a diagram or tree. The justification for this arrangement will appear as we proceed.

There is only very scanty information available about the technical culture of the Negritos of the Malay Peninsula, whom we may speak of as the Semang. There are differences of technology between the Semang of different parts, and a careful study of these differences would serve to throw much light on changes that have been introduced since the Semang were isolated from the rest of their race. There is no doubt that the Semang have adopted many elements of their present culture from their neighbours the Sakai and others. In some instances it is possible to trace this external influence, but in others it is doubtful whether we are dealing with a primitive Semang form or with a form adopted from their neighbours [227].

The same thing must be said about the Philippine Negritos. Our information is not sufficient to enable us to discuss the local differences, nor to determine what elements of the culture have been introduced by contact with other races [228].

HABITATIONS.

The huts of the Andamanese are best understood by considering first of all the simplest and most temporary structures. A man away from the main camp at night (on a hunting expedition) erects for himself a simple shelter of leaves. Such hunting shelters vary considerably according to circumstances. In the rainy season they are built much more substantially than in the dry season. Sometimes shelter is found between the buttresses of large trees, a few leaves being added. There is, however, one type of hunting shelter that is usual. Two poles are cut and erected perpendicularly in the ground so that they stand about four feet apart and about four feet high. To the top of these is tied, with cane or creeper, a horizontal pole. A few poles or sticks of sufficient length are placed so as to lean against the horizontal pole at an angle of about 45°, the ends resting on the ground. On these are placed any leaves that can be obtained, preferably the leaves of canes and other palms. The shelter thus consists of a single rectangular roof, one end of which rests on the ground, while the other rests on a horizontal bar attached to the top of two perpendicular supports. The shelter is built facing to leeward.

The usual family hut of the Andamans is built on exactly the same principles as a hunter’s shelter, but, being intended for occupation for some weeks or even months, is more carefully built. For a small hut, to be occupied by one family, four posts are erected, two at the back being from two to four feet high above the ground, while two at the front are from five to seven feet high. Two horizontal poles are attached, one to the top of the front posts and the other to the back posts, with strips of cane. If poles of a convenient size and forked at the top are available these may be used for the posts of the hut, the horizontal poles being supported in the forks, but a native would not trouble to search for such timbers, being satisfied with an unforked post. A few slender timbers, preferably of mangrove wood (Bruguiera), are placed on the two horizontal poles and bound to them with cane. These rafters, as they may be called, project for a foot or two above or beyond the higher horizontal, and similarly project a foot beyond the lower one so as almost to reach the ground.

In the better kind of hut a mat is made of palm leaves, and this mat is placed on the rafters and tied to them with strips of cane. To make a mat a number of strips of bamboo or cane of sufficient length are taken and placed on the ground parallel to each other. Leaves of a species of cane are collected and each leaf is divided into two parts down the middle. These half-leaves are then attached to the strips of bamboo or cane, by means of strips of the outside of canes, the technique being wrapped-twined work. The half-leaves are attached so that the leaflets, which are attached to the leaf-stem at an angle, incline alternately to one side and the other. The photograph reproduced in Plate VII shows a hut of the kind here described. Mats in course of making are shown lying on the ground in Plates VI and VII.

A quicker, but less efficient way of thatching the hut is to take the half-leaves such as are used for making a mat and fasten them in bundles of five or six directly to the rafters.

Huts of this type, each occupied by a single family, are built by the natives of the Great Andaman Division in the form of villages. The Jarawa and the natives of the Little Andaman build similar huts in their hunting camps, occupied during the fine weather.

Some huts of this type are provided with a floor raised above the ground. Such a floor is erected on short posts, and may be made of bamboos or of planks or pieces of broken-up canoes. A floor of this kind, raised a foot or so above the ground, is shown in Plate VII. Huts are sometimes to be seen with a floor raised as much as three feet above the ground.

The simple Andaman hut as above described is entirely open at the front and on each side. In an exposed situation screens of palm-leaves may be erected at the side. If still more shelter is required, a hut may be built with two roofs. Such a hut requires six posts, two taller ones six or seven feet high, and four shorter ones, two on each side. For such a hut two mats are made, and are so attached that one mat projects above the other. No attempt is made to fasten the two mats together at the top, but on the contrary a space of several inches is left between them to allow the smoke of the fire to escape, rain being excluded by the overhang of one of the mats. Huts of this type may be seen in Plate VI.

Each of the huts hitherto described is occupied by a single family. In order to understand the communal huts it is necessary to consider the arrangement of small huts in a camp or village. In the tribes of the Great Andaman there are two main types of such arrangement. The first type is that of the hunting camp, which is occupied for a few nights only. In this all the huts are placed facing in the same direction (to leeward) and in a line with one another. The second type is that of a village to be occupied for some weeks or months. In this the huts are arranged round an open space, all facing inwards, as described earlier in this book [229]. All encampments in the Great Andaman tend to conform to one or other of these types, but variations are introduced according to the nature of the site occupied. Thus in a hunting camp the site may not permit of the erection of the huts in one line. A village is, as a rule, only put up at a spot that has been used from time immemorial, where there is an open space of sufficient size, but if, for any reason, a site is selected where there is not room to arrange all the huts around the dancing ground, the arrangement of the village may be irregular.

The hunting camps of the Jarawa are sometimes arranged on the same principle as those of the tribes of the Great Andaman Division, i.e., all facing in one direction and as nearly in one line side by side as the site will allow.

The natives of the Little Andaman erect hunting camps in the fine weather. In the only one that I have seen the huts were arranged irregularly so as to make the best use of the available space.

A few words must be said on the sites chosen for encampments. It must be remembered that the islands are entirely covered with forest. The natives will not, if they can avoid it, put their camp under high trees, for fear of the danger of falling branches in a storm. At the same time they prefer a situation where there is an open space surrounded with forest so that they are sheltered from the wind. The coast-dwellers always camp immediately within the jungle on the shore of the sea or of a creek. The forest-dwellers usually choose a position on a hill or ridge, and this is particularly the case with the Jarawa. The camp must be close to a supply of fresh water. In the tribes of the Great Andaman Division no precautions are taken against a possible attack by enemies, but the Jarawa do take precautions, clearing the trees around their camps so that they have a good view of the approaches, and even, apparently, placing look-out stations at the tops of the paths [230].

Amongst the coast-dwelling tribes there are sites that have been used for encampments for many centuries. At these spots there are found heaps of refuse that have accumulated year by year. These kitchen-middens, as they are sometimes called, consist of the shells of molluscs, bones of animals, stones that have been used for cooking, fragments of pottery, and loam produced from decayed wood and other refuse.

The two types of camp arrangement which are seen in the village and the hunting camp are exhibited in two different types of communal hut. One of these, corresponding to the hunting camp, may be termed the long shelter. It is apparently only used in the North Sentinel Island. A hut of this type was seen by Mr Gilbert Rogers in 1903. It was rectangular, 40 feet long and 12 feet wide. The roof was supported on three rows of small posts ranging in height from 3 feet at the back to 6 feet at the front of the hut. The roof projected about 2 feet in either direction beyond the posts and was about 2 feet from the ground at the back and 7 feet above the ground at the front. There were twelve places for fires, six in front and six at the back of the hut, and near each, on the right-hand side, was a platform supported on four sticks, of the usual Andamanese type, for keeping food. There were two rows of sleeping places which were separated by small poles, making rectangles on the ground about 5 feet by 4 feet, each of which was probably occupied by a man and his wife and small children [231].

The relation of this type of communal hut, in which all the members of one local group are brought together under a single roof of one slope, to the ordinary family hut of the Andamans, and the arrangement of the hunting camp in a line, is obvious.

To the arrangement of huts in a village around a central open space corresponds the second type of communal hut, which may be called the round hut. Communal huts of this type were formerly built by the natives of the Great Andaman Division, but have fallen into disuse in recent times, owing to the natives having become much more migratory in their habits. Huts of the same type are built at the present day by the natives of the Little Andaman and by the Jarawa.

In its typical form this kind of hut is built by erecting two circles of posts, a smaller circle of tall posts, and a wider circle of shorter posts. The tops of these posts are connected by horizontal and sloping timbers, which make the framework of the roof. The roof is made of a number of mats of palm-leaves, which are laid on the rafters and tied to them with strips of cane. The mats are made in exactly the same way as the smaller mats used for the small huts and already described. They are sometimes rectangular in shape, though occasionally an attempt is made to make them narrower at the top and broader at the bottom. They are arranged on the roof so as to overlap one another and thus make the hut rain-proof. They are not joined in the centre, but a small space is left for the smoke of the fires to escape, and the rain is prevented from entering by letting one or two of the mats overhang the others at the top.

In the round huts of the Jarawa and the Little Andaman there is no centre-post, and according to the statements made to me by the natives of the Great Andaman they did not use a centre-post for their huts. In the description attached to a photograph in the British Museum Mr Portman speaks of the centre-post of a communal hut, which is shown in the photograph still standing, although the hut had been pulled down. It would therefore seem that in the Great Andaman the natives did sometimes erect a centre-post for their round huts. The typical round hut, however, has no centre-post.

It is clear that the round hut has been developed from the village. If all the small huts of a village be drawn together so as to touch each other, and if the mats of thatching be lengthened so as to meet and overlap in the middle, we have a round hut in its typical form. The evidence that this is so is afforded by the thatching, consisting of separate mats, often rectangular in shape (like the mats used for family huts), placed so as to overlap one another. This crude way of thatching could hardly have originated in any other way. Further evidence is afforded, as we shall see, by the internal arrangement of the hut.

Although the hut is here called a round hut, it must not be supposed that the shape is always regularly circular. It may be somewhat oval, and in any case is rarely very regularly constructed. In general, however, the shape approaches more or less nearly to a circle.

Huts of this kind vary in size according to the number of families occupying them. The height in the middle may be as much as 30 feet and the diameter may be 60 feet. The smallest I have seen was a Jarawa hut on Rutland Island, which was only nine feet high and 15 feet maximum diameter. In exposed situations the mats of thatching reach as far as the ground, but huts are sometimes built in sheltered situations with a space of a foot or two left between the ends of the thatch and the ground. A low doorway is provided on one side.

Within the hut there is a central space that is the common part of the hut and corresponds to the dancing ground of the village. In the wet season the communal fire is situated in this open space, and here the communal cooking is performed. In Jarawa huts the roof of the central part of the hut is hung with trophies of the chase consisting of pigs’ skulls bound with cane. In former times the natives of the Great Andaman Division hung similar trophies in their round huts. Around the central space are the spaces allotted to the different families, these being marked off by means of short lengths of wood laid on the ground.

It is thus clear that the basis of Andamanese architecture is the use of a single rectangular roof giving a shelter open at the front. This is the usual form of the hunter’s shelter and of the family hut in the village. For additional shelter two such roofs may be used, but no attempt is made to join them, one being made to overlap the other. There are two customary modes of arranging huts, either side by side facing in the same direction or round an open space facing inwards. Where, instead of separate roofs for each family, we have a united roof, these two arrangements of the camp give rise to two different types of communal hut, the long hut and the round hut.

In a village each hut is occupied by one family. In the communal hut (of either type) each family has a special portion of the hut marked off for its special use. Whether in a village or in a communal hut each family has its own small fire, at which the family meals are prepared. At one side of this fire is erected a small platform about a foot above the ground, supported on either three or four upright sticks. This platform is used for storing food. The natives of the Little Andaman erect low bamboo platforms to serve as beds, arranged round the communal hut, each family having its own. In the Great Andaman the natives, as a rule, make a bed of leaves on the ground and lay a sleeping mat on the top of this. In damp situations, however, they sometimes, as already mentioned, make a floor to the hut, raised a foot or two above the ground, and sleep on that. The Jarawa have a habit of sleeping in the wood-ashes of their fires in their cold weather hunting camps.

Turning now to the Semang, we find some differences in respect to their habitations. Those of the Semang who have not been influenced to a great extent by their neighbours and have not settled down to agricultural pursuits, never camp in the same spot for more than a few days, and have therefore no need to build anything except temporary shelters [232].

The Semang often erect their shelters in trees, well above the surface of the ground. This is a feature which distinguishes them from the Andamanese. It seems probable that these tree-shelters have been adopted by the Semang as a protection against wild beasts [233]. As there are no dangerous beasts in the Andamans, the extra labour involved in building a shelter in the branches of a tree instead of on the ground would serve no useful purpose. The difference in this respect between the Semang and the Andamanese is therefore due to a difference in the circumstances in which they live.

The typical form of Semang shelter, occupied by one family, is erected by planting three or four stout sticks or poles in the ground in a row at an angle of about 60° or 75° and lashing palm-leaves across these. The screen or roof thus formed is further supported, if necessary, with one or two poles used as props in front [234]. These shelters are similar to the Andaman shelters in having a single sloping rectangular roof, but differ from them in being supported, not by upright posts, but in an altogether less adequate manner. However, the Semang shelters are apparently very easy to erect, and as they are only occupied for a night or two there is no inducement to the natives to make them more substantial.

The Semang sometimes make a shelter by planting a number of palm-leaves in the ground in a semicircle so that the overhanging ends meet in the centre [235].

As the Semang are constantly moving from place to place, they have little use for a communal hut of substantial build. One communal shelter has been described, which contained eleven sleeping-places arranged in two long rows. The upright timbers of the shelter consisted of young saplings planted in two opposite rows, across them being lashed the leaves of a palm. There were, besides, two central posts or pillars, each about a third of the distance from either end of the shelter, and a dozen poles placed as props or wind-braces in various positions and at various angles, in order to strengthen the structure and keep it from being blown over in a high wind. The two slopes of the roof were not united over a ridge-pole, but a longitudinal aperture was left between them for about two-thirds of the entire length of the roof, and through the gap thus caused the greater part of the smoke from the many fireplaces issued. All round the walls were ranged a number of bamboo sleeping-platforms, five to six feet in length by about three feet in breadth. The owner of each sleeping-platform or family unit possessed a separate fire or hearth [236].

We have only scanty information about the huts of the Philippine Negritos. In Zambales (Luzon) a certain number of the Negritos have adopted a settled mode of life and depend on agriculture for some part of their subsistence. The most advanced of these have adopted the form of hut common amongst their neighbours. The less settled Negritos of Zambales erect huts which are almost exactly the same as the family huts of the Andamanese. Two short upright posts are erected for the back of the hut, and two taller ones for the front, and on these four posts a rectangular roof of one slope is erected. A bamboo floor or platform is erected a foot or so above the ground, just as in some Andaman huts [237]. In the Zambales huts the upright posts are forked and the horizontal poles are supported in the fork.

At Casiguran the Negritos erect palm-leaf shelters similar to those of the Semang. A few poles are thrust into the ground at an angle and in a row and palm-leaves are attached to these, the screen being further supported with props [238].

A comparison of the three branches of the Negrito race in the light of present information shows that the usual form of habitation amongst them is a sloping roof or screen of palm-leaves. One form of this, the simplest to construct, but only suitable as a temporary shelter, is in common use amongst the Semang and is found amongst the Negritos of Casiguran. The other form, more permanent but requiring more labour to erect, is in common use in the Andamans and amongst the Negritos of Zambales. Of communal huts we have no evidence in the Philippines. The communal shelter of the Semang consists of two screens leaning towards one another. The two types of communal hut of the Andamans are both derived from the family hut.

HUNTING, FISHING, ETC.

The Andaman Islanders depend for their subsistence entirely upon the natural productions of the forest and the sea. They make no attempt whatever to cultivate the soil. Until the introduction of dogs in 1858 they had no domestic animals. Young pigs are occasionally kept in captivity till they are grown, but they are killed for food and are not bred in captivity. Thus the Andamanese provide themselves with food by three different forms of activity: (1) collecting such things as roots and fruits and honey, (2) fishing in the sea and in the creeks, (3) hunting the wild animals of the forest.

For hunting the Andamanese rely entirely on the bow and arrow. Since they have had dogs they occasionally make hunting spears, but they did not do so in former times. They make no use whatever of any method of trapping game or birds. For fishing they also make use of the bow and arrow, wading out on to the reefs and shooting the fish, and in this they are very skilful. Crustaceans, such as crabs and crayfish, are captured in the same way. In the North Andaman a sort of short fish spear was formerly in use as an occasional substitute for the bow and arrow. In all parts of the islands small nets are used by the women for catching small fish and prawns. In the Great Andaman large nets were formerly used for capturing turtle, dugong and large fish near the shore. At the present time the natives of the Great Andaman Division make use of harpoons with which they capture turtle, dugong and large fish from their canoes. Harpoons are not used in the Little Andaman. The Andamanese are also aware of methods of poisoning or stupefying fish in pools by means of certain plants that they crush and place in the water, but I have never seen them use this method of fishing, although they say that they formerly did so. They have no fish hooks and no fish traps. At the present time a few of the natives have learnt to take fish with hook and line, but they are unable to make hooks for themselves, and have to obtain them from the Settlement at Port Blair.

In collecting roots a digging stick is used, and a hooked pole is used for gathering fruit, but they have no other special implements in use in collecting natural productions, and have no need of any. The adze is used for obtaining molluscs and for cutting honey-combs from hollow trees.

It is thus clear that by far the most important utensil of the Andamanese is the bow and arrow. We may say that they are essentially a bow and arrow people. This is even more true of the natives of the Little Andaman Division than of those of the Great Andaman Division.

It may be noted here that the Andamanese have no weapons that are used only for fighting. They fight with their chief hunting weapon, the bow and arrow. Nor have they any special defensive weapons, the shield being unknown.

The Semang in their natural condition depend for their subsistence on collecting roots and fruits from the forest, on catching fish in the streams, and on hunting animals. Their mode of subsistence is thus essentially the same as that of the Andamanese. One difference is that they have not the sea from which to draw supplies, and another is that the forests in which they live afford a much larger variety of game. A number of the Semang now practice a little rude agriculture which they have undoubtedly adopted in imitation of their neighbours of other races [239].

The principal weapon of the Semang, as of the Andamanese, is the bow and arrow. In hunting they also use spears [240], thus showing a difference from the Andamanese. Some of the Semang make use of the blow-pipe with poisoned darts, but it is practically certain that they have adopted the use of this weapon from their neighbours the Sakai [241]. They also make use to some extent of traps with which to capture jungle animals and birds. The wilder Semang living in the mountains have little opportunity of obtaining fish. Those of them that dwell near rivers use fish-spears and harpoons for catching large fish, and a small basket-work scoop for catching small fry [242]. They also fish with rod and line, the hooks being, as a rule, roughly manufactured from bits of brass or other wire [243]. The Semang have no special fighting weapons either offensive or defensive.

In the Philippines some of the Negritos practice a little rude agriculture [244]. It is practically certain that they have only adopted this mode of subsistence through contact with agricultural peoples of other races. They originally depended entirely upon collecting, fishing and hunting, and even those who now grow a few scanty crops devote a large part of their energies to hunting and collecting the natural products of the jungles [245]. The chief weapon of the Negritos of the Philippines, as of the Andamanese and the Semang, is the bow and arrow. They use the bow and arrow for shooting fish, having special fish-arrows [246]. It seems doubtful if they use spears, unless they have adopted them from their neighbours. In hunting deer the Negritos of Zambales use large nets like fish nets. They are acquainted with the use of traps for game but they seem to prefer to depend on the bow and arrow [247]. In the larger streams of Zambales they make fishing weirs of bamboo, after the manner of the Christianised natives of the same part [248].

As the most important weapon of the Andamanese, and indeed of the Negritos in general, is the bow, we may consider this first. Different kinds of bow are in use in different parts of the Andamans, but by a careful comparison of them it is possible to show how they are all derived from one original pattern.

The first kind of bow to be described is that in use in the Little Andaman. These bows are all made of a reddish-brown wood (possibly Mimusops littoralis). They are cut with an adze from a straight piece of wood, and are planed but not polished. The length varies within fairly wide limits. Six specimens selected as typical have lengths of 131, 150, 159·5, 163, 168 and 188 centimetres, giving an average of about 160 centimetres (= 63 inches). In section the bow is markedly convex on the one side and slightly convex on the other. The two figures (Fig. 1) show the section at the middle and at a point 7 cm. from the end of a typical specimen. The shape in section varies a little from one example to another, and the dimensions of breadth and thickness also vary. At the broadest point, which is in the middle, the average breadth of six bows is 3·2 cm., the broadest being 3·7 cm., and the narrowest 2·3 cm. The average thickness in the middle is 1·8 cm., the actual figures ranging from 2·1 to 1·3 cm. From the middle the bow tapers slightly towards each end. At a distance of 7 cm. from the end of the bow the average breadth is 1·8 cm., and the average thickness 1·2 cm.

The flatter side is the inside of the bow. Referring to the figures, the side marked A is that which faces a man as he holds the bow ready to shoot (called here the inside). C is thus the right-hand side and D the left-hand. By breadth is meant the distance from C to D, and by thickness the distance from A to B.

At each end of the bow there is a shoulder, as shown in Fig. 2. The length from the shoulder to the end of the bow, i.e. the length of the point on which the string is looped, is about 10 to 12 mm. Both ends are the same. For a few centimetres below the shoulder the bow is served over with string or fibre. In a carefully finished bow the serving is usually done with ornamental string, i.e. with string round which is twisted the dried yellow skin of the Dendrobium. In other examples plain string is used, or a strip of twisted Ficus fibre, or even nowadays a twisted strip of cotton cloth.

Bows are never ornamented in the Little Andaman either with paint or with incised patterns.

The bow-string of the Little Andaman is made of strips of the bark of the Ficus laccifera. The number and width of the strips used depend on the size of the bow. In a small bow now in the Cambridge Museum the string is made of a single strip of bark about 1 cm. in width. This strip is simply twisted, the twist being that of a right-hand or male screw. When two strips are used they are not twisted around one another in the way that a two-ply rope is made, but are laid flat together and twisted together, so that when the string is finished only one of the strips is visible, the other being inside. In a stout string for a large bow three or four strips may be twisted together in this way. The bow-string is not a rope, but a twisted strand (Fig. 3.)

At one end of the string a loop is made, as shown in Fig. 4. The end is doubled over to make a loop of the right size, a round turn is made over the standing part (A) and the end (B) is twisted in with the standing part by untwisting the latter, laying the end in, and twisting up again. If this splicing, as it may perhaps be loosely called, be not sufficiently secure, it is served over or stopped with finer fibre of the same kind. This loop is of sufficient size to slip down over the shoulders. At the other end the string is attached to the peg either with a knot, or else by means of a small loop (just large enough to go over the peg, but not large enough to slip over the shoulder) made in exactly the same way as the loop already described. When the bow is to be strung the larger loop is slipped over the peg at the top end of the bow and is pushed down over the shoulder. The other end (with the smaller loop) is then slipped on the peg at the lower end, resting on the shoulder. The lower end is placed on the ground, while the top end is held in the hand. The man places his foot against the middle of the bow and draws the top towards him until he is able to slip the top loop of the string up over the shoulder so that it catches the peg or tip. The bow is then ready for use.

Toy bows are made for small boys of exactly the same general pattern as the large bows. A toy bow of this kind, now in the Cambridge Museum, is 107 cm. long and 18 mm. broad in the middle.

The next type of bow to be considered is that used by the natives of the North Sentinel Island. I have only been able to see one specimen of this type, which is in the British Museum. It is made of a different kind of wood from that used in the Little Andaman. The length is 155·5 cm., and the breadth at the middle is 4·3 cm. The section in the middle, which is shown in Fig. 5, is slightly different from that of the average Little Andaman bow, but it has the same feature of greater convexity on the outside and less convexity on the inside, and it lies just within the range of variation of the Little Andaman type. The ends of the bow are shaped in the same way as those of the Little Andaman bow. The breadth at the shoulder, however, is 2·5 cm., which is greater than the corresponding measurement of the Little Andaman bow. The bow is not ornamented either with a painted or incised pattern. The string is missing. There is no binding at the ends below the shoulders, but this has possibly been present and come off, as the specimen is one that has been thrown away by its owner owing to the wood having split. So far as we can tell from this single specimen the bow of the North Sentinel differs very little from that of the Little Andaman.

We now come to the bows of the Jarawa of the South Andaman. The Jarawa of Rutland Island, of whom there are now very few, but of whom there were a larger number twenty or thirty years ago, make bows exactly like those of the Little Andaman, and apparently do not make any other kind. The Jarawa to the north of Port Blair, who have been driven northwards by the spread of the Penal Settlement, also make bows of this type, which it is not possible to distinguish from Little Andaman bows. These northern Jarawa, however, also make bows of a different kind. These will be spoken of as belonging to the “modified Jarawa type.” They are larger than Little Andaman bows, having an average length of about 185 cm., with a breadth of about 5 cm. The section, throughout the greater part of the length, is either plano-convex, or, more frequently, concavo-convex. The section of a typical example is shown in Fig. 6. At the middle of the bow, where it is held in the hand, there is a slight thickening produced by a protuberance on the inside, i.e. on the flat or concave side. In a certain number of specimens the bow, instead of being straight, is slightly recurved outwards. Finally, the wood from which these bows are cut is not the same as that used in the Little Andaman.

The Little Andaman bow, the North Sentinel bow and the Jarawa bow are all varieties of one type. The Little Andaman form is probably nearest to the original of the type, and I shall show later how the modifications found in the modified Jarawa type came to be adopted.

We now come to bows of a different type, of which there are two varieties, one used in the South and Middle Andaman, and the other used in the North Andaman.

The bow of the South Andaman tribes is not cut from a straight piece of wood, but is cut from a tree that has bent in the course of its growth into a suitable curve. A tree has to be found that will provide a piece of wood of the required shape. From this the bow is shaped with an adze, and is finished by planing with a boar’s tusk.

Bows of this kind vary in length between 180 and 210 cm., the most usual length being between 190 and 195 cm. At the upper end the bow is brought to a point approximately circular in section. From this point it broadens out until, at a distance of about 50 or 55 cm., or between one-quarter and one-third of the length of the bow, it reaches its maximum breadth, which is, on the average, about 5·5 cm. The section of the bow at this point is convex on the outer side, while on the inner side it may be flat or slightly concave, or even in rare instances slightly convex. In many specimens there is a very slightly raised keel running down the middle of the inside of the bow. The thickness of the bow at the point mentioned is usually 1·5 to 1·75 cm., and there is little variation in this respect in different specimens. (See Fig. 8.)

At the middle the bow decreases in breadth and increases in thickness to form a handle. At the handle the usual section may be described as pear-shaped, the greatest diameter being the thickness (from inside to outside) and not the breadth.

From the handle towards the lower end the bow again increases in breadth, so that the lower portion is about the same breadth and thickness as the upper portion. At the lower end it tapers to a point, circular in section, but the point is blunter than at the upper end.

Thus the whole bow consists of a leaf-shaped upper portion or blade which is straight (i.e. neither curved inwards nor outwards), a waist or handle, and a lower blade that is curved backwards or outwards at about its middle, this being the position of the bend in the wood from which it is cut. A bow of this type is shown in Plate V.

Near each end the bow is served over with string for a distance of 3 or 4 cm., leaving a bare point at the upper end of about 7 cm. long, and a point at the lower end of about 1·5 cm. (Fig. 7.)

A bow-string is made from the fibre of the Anadendron. A number of strands of the fibre are taken and are waxed with black bees’-wax. Four of these strands are taken, three of them are placed together and the fourth is wound spirally round them. When the end of the active strand (the one being twisted round the others) is neared, a new strand is taken and laid in. The twisting is continued for a few turns and the newly inserted one is then taken, the end of the first active strand being laid in and wound over in its turn. The process continues in this way, new strands being added until a cord of sufficient length has been made. This is again waxed over on the outside.

At one end of the cord a knot is tied. At the other end a loop or eye is formed. To make this eye, when the cord is of sufficient length, the end of it is bent over to form a loop of about 1 cm. or a little more in diameter. This loop is then served over with thread made of Anadendron fibre. The serving is continued over the neck of the loop for about 1·5 cm. This gives an eye with the appearance shown in Fig. 9. A loose strand of fibre is left at the neck of the loop. This is wound spirally over the cord, as described before, new strands being added one after another until the cord has been treated in this way for about 35 cm. from the eye. It is then stopped by serving it over for about 2 cm. with Anadendron thread. It is clear from this description that the cord is somewhat thicker for about 35 cm. from the end with the eye than it is in the rest of its length.

To string the bow the knotted end of the bow-string is fastened round the top end of the bow with a slip knot, so that it rests on the top of the string serving. The bow is then turned upside down and the top end (now temporarily at the bottom) is fixed in the ground or against a stone, so that it will not slip. The other end of the bow is taken by the left hand, while the cord is held in the right, the right foot is placed against the handle or middle of the bow and the bow is bent, the end held by the hand being drawn towards the operator until he is able to slip over it the eye or loop at the end of the string.

After the bow has been strung the upper portion, which before was straight, is now curved inwards, and the bow therefore appears as S-shaped when seen from the side. When a man starts out hunting or fishing he strings his bow and tests it, and it remains strung till he returns, when he unstrings it and places it in his hut.

The advantage of having a knot at one end of the string seems to be that should the string be stretched by use it can be tightened by altering the position of the knot.

At the point where the nock of the arrow is placed when the bow is drawn, it is usual to serve the string over with thread of Anadendron fibre.

The peculiar features of the South Andaman bow depend on the fact that the bow takes advantage of the greater toughness and elasticity of wood that has been compressed in the course of its growth. When the bow is drawn the strain does not fall evenly, but, by reason of the shape of the bow, is concentrated on one portion, namely the lower portion of the bow where it is curved outwards. This is easily seen when a bow is strongly drawn, for from the S-shape that it has before, it becomes very nearly true arc-shaped when fully drawn. The lower portion of the bow works as though hinged, and thus the strain is largely borne by the curved portion of the bow. Now this portion is cut from the concave side of a tree that has been bent while growing, and consequently the fibres of the wood are here stronger, tougher and more elastic. The result is that for a given amount of energy spent in drawing the bow a greater force of propulsion is given to the arrow than with a bow of the Little Andaman type.

The breadth of the bow is necessitated by its shape, for if it were narrow the string would slip round on to the outside of the bow. The narrowing at the handle is necessary for holding the bow. The adoption of tapered ends instead of shoulders is a definite improvement as it makes the bow less liable to split at the ends.

The bows of the South Andaman group of tribes are always decorated with incised patterns. The conventional pattern is shown in Fig. 10. One line of such pattern runs down each edge of both the inside and the outside of the bow, and on the inside a similar line of pattern runs down the middle. When bows are newly made they are often also decorated with designs in red paint and white clay, particularly if they are intended as gifts. These painted designs soon wear off and are not renewed.

In these tribes bows are sometimes made of a size so large as to be almost useless for hunting. One such bow, now in the Cambridge Museum, is 220 cm. long and with a maximum breadth of 10 cm. Such bows are very carefully made and decorated and are intended as gifts. A man generally makes such a bow with the deliberate intention of giving it to some person whom he wishes to please. The bow that I have was specially made to give to me in this way. A man who possessed such a bow would not dream of using it in hunting, but he might use it in a shooting match, in order to show his skill.

In the South Andaman tribes toy bows are made for boys of somewhat the same shape as the ordinary hunting bow. An example of such a bow, now in the Cambridge Museum, is 121 cm. long. It is cut from a bent piece of wood in such a way that the lower portion is curved outwards. The section in the middle is plano-convex, very nearly the half of a circle, the breadth being 26 mm. and the thickness 13 mm. It is broadest in the middle, and tapers towards each end. When strung it assumes the typical S-shape of the South Andaman bow. It is served over with thread at one end and with a strip of cotton cloth at the other, leaving two points for the string. The latter is of the usual South Andaman type, but of smaller dimensions.

We must turn now to the bow used by the four tribes of the North Andaman, which is of a somewhat different pattern from that just described. It is, in the first place, shorter, lighter and more slenderly made. Of ten typical specimens the shortest is 153 cm. long, and the longest is 182 cm. The usual length is about 160 to 165 cm. In its broadest part the North Andaman bow is broader than the South Andaman bow, the breadth varying from 6·5 to 7·5 cm. in different specimens.

Although the North Andaman bow is, as a rule, cut from a curved piece of wood, it may, on occasion, be cut from a piece that is practically straight.

At the upper end of the bow there is a long point. In a specimen that is in every way typical, at about 5 cm. from the point the section is circular and the diameter is about 5 mm.; at about 30 cm. from the end the section is slightly flattened or oval, and the maximum diameter (the breadth from side to side) is about 1·5 cm. From this the bow broadens fairly rapidly, until at a distance of about 60 cm. from the end it is 7 cm. broad. The section at a point 60 cm. from the end is shown in Fig. 11, where it may be seen to be convex on the outside and only very slightly convex on the inside. At about 75 cm. from the end the bow narrows in breadth to form a handle, at the same time increasing slightly in thickness. The handle is approximately circular in section in the middle, which is about 95 cm. from the upper end, and about 80 cm. from the lower end, the diameter being about 2·2 cm. Below the handle the bow broadens out once more into a lower blade which in shape and section is similar to the upper blade. At a distance of about 30 cm. from the lower end the bow once more narrows off to a point approximately circular in section. The lower point of the bow is not so long or so tapering as the upper point.

The whole bow thus consists of two blade-shaped portions tapering to a point at each end, and with a waist or handle between them. The upper blade is straight, i.e. is not curved either outwards or inwards. The lower blade is curved outwards (like that of the South Andaman bow) in nearly every newly made bow and in every bow that has been in use.

The upper part of the bow is served over with string for about 1·5 cm. (at a distance of 15·5 cm. from the end in the bow that has been described), and the lower end is similarly served (at a distance of 6 cm. from the end). The general shape of the bow as seen from inside is shown in Fig. 12.