Volume 49 1940 > Volume 49, No. 194 > The native pottery of the Fiji islands, by R. R. C. Maclachlan, p 243-271
THE NATIVE POTTERY OF THE FIJI ISLANDS
THE material utilized in the preparation of this paper consists of the collections of Fijian pottery in the chief museums of New Zealand.
Acknowledgments are due to the Director of the Dominion Museum for photographs and information; to Mr. G. S. Stevenson of the Auckland Institute and Museum; to Mr. R. S. Duff of the Canterbury Museum for the preparation of the three excellent drawings used for figs. 45, 45a and 49; and to the Director of the Otago Museum for permission to figure examples.
Numerous references have been obtained from books in the Otago University Library and the Hocken Library, but the chief source of information has been the very fine collection of publications relating to Fiji in the Reference Department of the Dunedin Public Library.
The accompanying map shows the distribution of pottery-manufacturing centres together with the chief trading routes for ceramics in the Fiji group.
Each of the pots figured has been subjected to a series of measurements, together with a test for glaze-texture, as suggested by Benjamin March. 1
According to Roth, pottery is made to-day “on the Ra coast; in the Namena district 2 of Tailevu Province; in the - 244 district of Vutia in Rewa Province; in the district of Nasingatoka and Malolo in Nanddonga Province, and in the Province of Tholo West, all on Viti Levu island; on Yanutha island in Serua Province, in the Province of Lomaiviti; and at Nalotu 3 village in the Yawe district on Kandavu” (12, p. 217).
Laura Thompson obtained sherds from the following sites: Kambara, Wangava, Fulaga and Mothe; while a pot was procured from Levuka, Lakemba (15, p. 109).
Schurig states that pottery is made at the present time on Malaki island, Koro, Kondaevo, and Ono, while Nasilai and Rewa river are given as ancient sites of manufacture (13, map 2).
Throughout Fiji native pottery is fast disappearing, being replaced by iron and cheap china from the local stores (12, p. 217), but a number of good accounts by witnesses during last century and earlier throw light on various aspects of the process of potmaking.
C. F. Gordon-Cumming (6, p. 125) states “… the town of Na Sava, which is peopled by the former inhabitants of the island of Malaki, from which they were driven out by the whites as an act of vengeance for the murder of a white man… That, at least is one version of the story. Malaki… is a pleasant spot, grassy and wooded, but now left desolate. To its people is attributed the honour of having been the first in these isles to invent pottery, an art which is here carried to a perfection far surpassing anything found in other groups of the Pacific…
“… At So So and Rewa the women just beat out a flat piece of clay on their hand, and then gradually mould it into a cup-like form, with the help of a smooth stone held inside and a wooden spatula with which to beat the outer surface. When their modelling is finished the pieces are left to dry in a house for six to eight days, and are then taken to a quiet sheltered nook between the sea and a great rock. Here a pile of light wood and small sticks is built, and on this the pots are laid. Dry grass is lightly piled over- 245 - 246
them, and small twigs over all. This pile is set on fire and kept burning for about half an hour. Then when still hot, the cooking pots are well rubbed with an infusion of tire—i.e., mangrove bark—which is a dark red dye, and gives the pots both colour and a slight glaze. Ornamental pots and those for water are kept in the house from four to eight days. They are first baked with a light grass fire, afterwards with wood and while still hot glazed with the nadakua resin… (6, p. 128).
“There are slight variations in the process in different parts of the group, as on the north of Vanua Levu, where all the pottery we procured was unglazed. Several of the finest pieces I have seen came from Na Sava…” (6, p. 128-9).
“The pottery is made entirely by hand, nothing in the nature of a wheel being known. The clay having been mixed with a fine sand is rolled into long sausages, and these are coiled one above the other, into a hollow circle, thus forming the base of a round pot. Having partly moulded this into shape, the potter takes a smooth round stone in her left hand and holds it inside the clay, while with the other hand she beats the exterior with a flat piece of wood like a spoon, and constantly moistens the clay. Fresh sausages are constantly built up round the top and gradually narrowed till there remains only room to insert one finger (if for a water pot) or the food (if for a cooking pot); these are in like manner beaten to a smooth surface both inside and out. The rim of the vessel must now be fashioned, and then comes a final wetting and smoothing of the whole, and probably a very elaborate geometrical pattern is, last of all, marked with a small sharp stick. Sometimes a pattern is laid on in raised work, almost like clusters of grapes. The work must be done ere the day wanes, as towards sunset the clay falls and will not mould obediently to the potter's hand… ”(6, pp. 129-130).
The above accounts clearly show the difference between the two predominating techniques employed throughout Oceania, namely the coil and patting methods.
Wilkes (21, p. 348) gives a good account of potmaking in Fiji, but does not specify the locality in which he witnessed the process. He says “They dig the clay and - 247 carry it in baskets to the village, where they knead and temper it with sand to a proper degree of tenacity. Their tools are very simple, namely: a flat mallet (tala); a small flat round stone (vatu); and a circular cushion made of coco-nut leaves.
“A lump of tempered clay is first taken, which is fashioned somewhat into the shape of the vessel the workman desires to form; the stone being introduced into the inside, the mallet or spatula then being used on the outside with the left hand. The different parts are fashioned or made separately and afterwards joined. The joints are very nearly closed, so much so as to escape detection. The strokes with the mallet are exceedingly hard at first, but as the vessel approaches the intended shape they become more gentle, and the finishing is given by smooth pressing. Many of the vessels are extremely graceful in shape and must require a very true eye to form the various parts so as to fit. The figures or tracings that are seen upon them are executed by young girls with the fibres of a coco-nut leaf, The pots are baked before an open fire after which the glazing, or rather, varnish is put on, consisting of the resin of a species of pine (resembling the kauri pine of New Zealand) called makandi, mixed with a decoction of the Mangrove bark.”
Smythe, speaking of Rewa, says that it is “… one of the few places in the Group where pottery is made, and as we were desirous of seeing the native process, Mr. Waterhouse took us one day … up the river to Vutia, the town of the fishermen of Rewa. We were so fortunate as to find on our arrival a woman in the act of modelling a large pot. Her only tools were a flat piece of wood and a smooth round stone to hold against the vessel inside. Bending from the waist over the vessel, which was placed on the ground she beat it into shape with these simple implements. The clay which was yellowish in colour and very plastic is found in the neighbourhood. After being allowed to dry in the sun for a few days, the pots are slightly burned in a fire of brushwood; and finally they are glazed by being rubbed over, while still hot, with kauri gum.” (15, p. 38-9).
It is interesting to note that Cumming, writing some twenty years later, says (6, p. 189): “The town of Rewa - 248 consists of a cluster of villages, inhabited by various divisions of the tribe, all subjected to a central power…
“… the fisher town stands quite apart from the homes of the agricultural population and intermarriage is equally rare. Thither we wandered in search of the curious pottery made by the very low cast women of the fisher tribe.”
Williams gives a brief account of potmaking in Fiji, but does not state where he witnessed the process. He says:“… In the manufacture of their pottery, the Fijians employ red and blue clays tempered with sand: their apparatus consists of a ring like cushion, four flat mallets, (tata), and a round flat stone; and yet the pots are often made with as true an outline as if they had been turned with a wheel. Lines and figures are traced on the vessels while yet moist; and after drying a few days, a number of them are piaced together, covered with very light fuel, such as reeds, nut leaves, grass, etc.; this is set on fire and by the time it is burnt out, the pots are baked. While yet hot such as are to be glazed are rubbed over with resin of a species of pine.” (22, pp. 70-1).
The technique outlined in the above account is very similar to that described by Roth in his Potmaking in Fiji (12, p. 217). He records the manufacture of pottery at the village of Naivuvuni, Ra province, Viti Levu island. The two women who produced this pottery “… belonged to a group of people (Yavusa)… whose pots have a reputation for being well made. It is said locally that they are the only people whose pots are worth having.” (12, p. 217). Roth states: “The essential constituent is a plastic clay called nggele kuro (clay for pottery) and in colour it is a dingy shade of dark purple. It is obtained from certain points of land on Malaki, an island off the coast; at Navoranga near Naivuvuni, at Ndelaivaimboro near Navunitongoloa; and at the head of Vaindrika. It is forbidden to smite or strike the hole from where clay is taken for pottery making purposes.
“Nowadays a husband and wife will go together to the place where the clay is to be found. The wife will choose the clay and they will both gather up what is sufficient for their needs, and take it off to their village. Similarly they go out to collect the sand nuke loaloa (black sand), which - 249 is obtained from the sea-coast nearby, and is dark grey in colour.” (12, pp. 218-9).
The clay is first kneaded between the fingers to find any extraneous matter, which is thrown away. Next a large board hollowed longitudinally is taken and sprinkled with water and then with sand. On the top of this is placed some of the clay broken up into small lumps. The potter works the clay with the heel of one foot, balancing herself with the aid of a sapling, using it as a prop. From time to time she bends down and folds the clay over upon itself, adding a little water at intervals.
A lump of the prepared clay is shaped with the hands to form a thick pencil, which is laid upon a circular stand made from the leaves and skin of the banana tree, rolled up to form a flat ring on which the pot will remain until ready for firing. The ends of the clay pencil are coiled round until they meet to form a ring of clay. Three further rings are added, each successively upon the top of the previous one. These will constitute the body of the pot, the uppermost will become the base of the pot, the first ring will form part of the neck. The potter next walks slowly round the vessel in an anti-clockwise direction, pressing the rings together by running the fingers vertically upwards (12, p. 219-20).
The clay is next beaten out to the desired shape, a process which is a matter of hours, the operator constantly walking round in an anti-clockwise direction. Water is frequently applied to the work by dipping the blade of the tool into a basin of water and rubbing the surface of the stone which is held as a counterpoise inside the pot with the blade. Water is occasionally sprinkled upon the pot itself. This prevents the tools from sticking to the clay and preserves the plastic qualities of the latter (12, pp. 221-2).
As the patting process continues the upper part is gradually closed until no more than a small aperture the thickness of a finger remains, and finally this is sealed up so no trace of its existence remains.
The pot is left to dry in the wind but out of the direct rays of the sun, the centre of the vessel being covered by a piece of cloth. The pot is left in this state until the exposed portion is strong. The centre of the pot has further treatment to undergo and must remain plastic. The pot is left - 250 in this state until the exposed portion is no longer pulpy, and is then inverted and work resumed. Roth continues: “It is noteworthy that a considerable amount of energy is expended on beating out the clay at certain stages, and yet there is a corresponding high degree of accuracy and regularity of the strokes conveyed by the tools to the pot. What seems rather an ordinary series of operations is in fact a work for experts, and one cannot help admiring the dexterity which they display in the execution of their task.” (12, p. 222).
The neck and mouth of the pot are formed next. The beating gradually narrows the neck, but if there is not sufficient clay in the pot itself to form a satisfactory one another coil is added and pressed to the clay of the pot. The vessel is now placed in the shade for about an hour to enable the rim of the aperture to harden sufficiently to permit the addition of the neck and lip. During this period the body of the pot is smoothed over with the hands (12, pp. 222-3). Roth proceeds: “In the formation of the mouth of the pot and the neck which supports it, a first pencil of clay is laid round the edge of the aperture… and the thick rim so formed is then turned or rolled back outwards upon itself from the inside until the diameter of the aperture is the length of the operator's finger span… Then a second pencil is laid round on top of the first and pressed into union… so as to become one with the pot.” (12, p. 223).
The flat surface of the sau-tool is then used to pat down the edge, next the same tool with its appropriate stone is used to form the neck. To do this “the operator walks round the pot three times in forming this neck; the first two turns are occupied in gently beating up the clay, using the angular surface of the tool; on the third round she uses much vigour to finish off the neck. Another pencil is now placed round the neck to strengthen the joint and smoothed down neatly with the fingers… ” (12, p. 223).
Roth also gives a good account of the pottery of the Mbua Province, Vanua Levu island (12, pp. 229-232), where the art of potmaking is now extinct, although once a flourishing industry. The women who made it for Roth's observation had not produced any before, although they had - 251 all, when children, seen their womenfolk do so. The process recorded from Mbua differs somewhat from that of Ra, the chief divergences being as follows:
Roth says: “Pottery was used in barter in former times. It is still bartered to-day, but to a very limited extent. One pot—an elongated form of kuro—which I saw at Koro village… had been exchanged for mats with some people of Tholo West. Koro is at least 20 miles from the nearest village in Tholo West (12, p. 229).
“Pots made on the north coast of Mbua were formerly transported up into Mathuata Province by canoe and exchanged for bark-cloth and whale's teeth, food and mats. Exchanges were also made with the people inhabiting the inland area of Mbua” (12, p. 232).
Pottery is transported for trade purposes either by means of a shoulder-stick, with a pot attached at each end, or in a large basket supported by a rope passing under the armpits and over the front of the shoulders. These two methods are employed by men and women respectively (12, p. 228).- 252
Laura Thompson states that: “the Levuka women manufacture pots in Levuka for trade in the southern islands. They also accompanied their men, who were professional sailors and traders to such islands as Kambara and Oneate, where suitable clay was to be found, and made pots on the islands to supply the local demand (18, p. 111).
“On Kambara, according to old informants, they lived in the village for which they were making pots… The Kambara women traded mats for Levuka pots. Occasionally the Kambara men obtained pots in Levuka (18, pp. 111-112).
“Pots were imported into southern Lau from the island of Matuku. They were made in the village of Nggalikarua (18, p. 112).
“Pots from Kandavu were also traded in southern Lau. On Kambara they were exchanged for barkcloth… wooden kava-bowls… perfumed oil… and digging sticks. There was no standard of value, but the women say one large Kandavu pot was equivalent to about one wooden Kava-bowl or to one sheet of bark-cloth… (about three by three metres) and one container of coco-nut oil (18, p. 112).
“Pots from fabrication centres in Tailevu on Viti Levu island also found their way to Lau” (18, p. 112).
Williams (22, p. 93) states that: “The commercial transactions of the Fijians, though dating far back, have been on a small scale, consisting of a barter trade, which is chiefly in the hands of the Levuka, Mbutoni, and Malaki people, who regard the sea as their home, and are known as ‘the inhabitants of the water’ Although wanderers they have settlements on Lakemba, Somosomo, Great Fiji and other places. They exchange pottery for masi, mats and yams. On one island the men fish and the women make pots for barter with the people of the main. Their mode of exchange is very irregular. The islanders send to inform those on the mainland that they will meet them, on such and such a day at the trading place—a square near the coast paved for the purpose. The people of the continent bring yams, taro, bread, etc., to exchange for fish…
“… The inhabitants of the Friendly islands still depend upon Fiji for their canoes, spears, sail-mats, pottery…” (22, p. 94).- i - ii - iii - iv - v - vi - vii - viii - ix - x - xi
- xii Page is blank- 253
The question now arises: Do the potters constitute a separate caste? Early investigators almost without exception maintained that this was the case. Wilkes, writing in 1838 says: “The potters also constitute a separate cast, of which the women only exercise the art and do no other work” (21, p. 348).
A similar view is expressed by Jenkins who states: “… These pots are manufactured by women who follow this employment only” (8, p. 350).
Stonehewer Cooper, speaking of the Fijian potters, writes: “The manufacturers of… crockeryware and other articles of… household necessities are chiefly confined to the women, and are also tribal specialities; hence certain districts are noted for particular kinds of pottery…” (17, p. 99).
“… Pottery work was entirely confined to the women of sailors… and fishermen' families” (17, p. 119).
Deane, however, includes pottery among the articles of personal property “… which a member of a clan has made or produced in his spare time… These possessions may be alienated, and are often bartered away at what is called the “Solévu.” On such occasions one town will arrange with another for an exchange of goods. The people of one community will bring for example, pigs or mats, which they will exchange for pottery or salt, made by the inhabitants of the other town…” (4, pp. 118-119).
Thomson, however, attributes the localization of industries in Fiji to the natural products of the country. “No tribe, however wide its territory was entirely self-supporting. Salt came only from the salt pans in the mangrove swamps, cooking pots from the clay pits in the outlying islands;… the carving of bowls and the building of canoes were the craft of the carpenter clans and no other. The comfort, if not the existence, of a tribe depended upon barter…” (19, p. 280). This took the form of the Solevu, or formal presentation of property by one clan to another. (19, p. 280).
Following the establishment of a European government in Fiji the Solevu was for a time encouraged by the authorities as it was considered that should it be abolished the manufacture of mats, pottery, salt, and other articles - 254 would fall into disuse, and that the material comfort of the people would be adversely affected (19, pp. 286-7).
It would appear from the above accounts that potmaking in Fiji is a tribal speciality, but as to whether the potters constitute a group of specialists within these tribes is a question rendered more complex by the indiscriminate use of the terms tribe, clan, cast, group, by some of the older writers. That a rigid group of specialists does not exist at the present day is shown in the statement by Roth that it is not considered degrading for women of the highest rank to manufacture pots; indeed, women of chiefly rank are expected to produce an article superior to that of ordinary women (12, p. 232).
In respect of form the native pottery of Fiji may be divided into the following classes:
CLASS A.—Simple open bowls.
Fig. 1, a rather crude glazed open bowl figured by Edge-Partington (5, pl. 114, vol. 1). This is a typical example of the fundamental shape of all clay receptacles. No particulars are given as to colour, locality, or measurements.
Fig. 2 figured by Edge-Partington (5, pl. 114, vol. 1), is very similar to the above. No particulars are given. The pot itself would appear to be of a somewhat deeper and less regular form than that shown in fig. 3 below.- 255
Fig. 3 (Auckland Museum) is almost hemi-spherical in shape and is red-brown in colour. In this example the aperture is somewhat constricted.
Fig. 4, a cone shaped thick-walled pot, is figured by Edge-Partington (5, pl. 56, vol. 1), who gives no further particulars. This shape, though rare in general, is present in the Solomon islands, as exemplified by fig. 51, in the Otago Museum. For a pot of this type to remain upright it must be thrust either into soft ground or placed in a stand, hence the cone-pot was one of the chief products of the early Egyptian potters (11, pl. 5, 44d).
CLASS B.—Bowls with an in-curving rim or shoulder.
A simple example of this group is fig. 5, a small yellow-brown pot in the Otago Museum. The walls of the vessel are very thin, with a delicate in-curving rim, 2 cm. in width.
Fig. 6, a bowl in the Auckland Museum, the walls of which closely resemble in outline figs. 2-3, but it has a narrow in-curving rim which appears to be of about the same width as that of fig. 5. The glaze, however, is much deeper, and the pot is reddish-brown in colour.
Fig. 7 a yellow-brown pot in the Otago Museum. Here the rim is broader—3.5 cm.—and the vessel of much thicker pottery. Unlike the two preceding examples of this class, the rim rises sharply from the sides of the bowl, making a sharp angle with the latter. At the same time the true spherical shape of the bowl is disappearing, since a distinct flat base is present.
CLASS C.—Vessels in which an out-curving lip has been added to the in-curving shoulders.
Fig. 8 (Auckland Museum), a typical example of this class in a transition stage. The shoulders, which rise steeply from the bowl, are surmounted by a narrow out-curving rim, the whole being very similar to a class of pot from the New Hebrides (J.P.S., vol. 48, p. 34, figs. 34-44). It will be noticed that the point of junction between base and shoulder is marked by decoration in both the Fijian and New Hebridean examples.
Fig. 9 (Dominion Museum), closely resembles fig. 8. In this case, however, the junction of shoulders and bowl, although very distinct, is undecorated. The out-curving - 256 lip is much more pronounced, and, like the entire pot, of more delicate material than fig. 8.
Fig. 10 (Otago Museum), a small pot similar to the above, and black in colour. Fig. 9 and fig. 10 both strongly resemble in shape a fragment from Santo (New Hebrides) in the Canterbury Museum (J.P.S., vol. 48, p. 50, fig. 46).
Fig. 11 (Canterbury Museum), shows a much larger pot of the same type. The division between shoulders and bowl is clearly indicated, but the out-curving rim is, compared with fig. 10, much reduced. The pot is vandyke-brown in colour and slightly glazed. An interesting feature of the outer surface is a series of minute ripples producing an effect similar to that of some sherds from the neolithic site of Quang-binh (Anam), excavated by l'Ecole Francaise d'Extreme-Orient. These sherds, which are much weathered, appear to be combed in a vertical direction (3, pl. 51, 1-2, 9).
Fig. 12, a pot in the Dominion Museum in which the bowl and shoulder blend imperceptibly in an almost perfect curve. The comparatively wide brim has a slight concave depression, and comes away from the shoulders at an acute angle. An interesting feature of the brim is that four oblong apertures are present to facilitate suspension.
Fig. 13 (Auckland Museum), shows another example of this class. Here, however, the bowl and shoulders make an extremely wide angle and are surmounted by an almost vertical rim.
CLASS D.—Vessels in which the out-curving lip has been carried up, so as to roof completely over the bowl or base. It is to this class that by far the greatest number of Fijian pots belong.
Fig. 14 (Dominion Museum), one of the simplest of this class for here the shoulders terminate in a plain aperture—i.e., no rim is present—and the point of junction between bowl and shoulder is clearly indicated, but the former is elaborated by the presence of a base. Another feature is the presence of the spout or air-vent, an essential feature of this class, when the aperture is small.
Fig. 15 (Dominion Museum), a more elaborate vessel of the same type as fig. 14, but differing from it by having no base or air-vent, a feature not required when the aperture - 257 is large. The shoulders are composed of four distinct tiers each sloping up at a sharp angle.
Fig. 16 (Otago Museum), shows a dull-brown vessel closely resembling fig. 15. In this example, however, a base is present, and the four tiers forming the shoulders are built up at an angle of 45°, while the fourth, which is vertical, forms the neck, for which there is an earthenware stopper. The shoulders are further elaborated by the presence of eight handles.
Fig. 17 (Dominion Museum), has four large handles. The shoulders are not formed into definite tiers, but slope up to form a vertical lip. The handles arise a short distance above the junction of bowl and shoulders, this point being emphasized in each quadrant so formed by the presence of three blunt projections, which may be related to the turtle-motif described below.
In fig. 18 (Auckland Museum), the handles are much reduced and are more decorative than utilitarian. Two slight projections above the junction of bowl and shoulders suggest much-modified tiers.
In fig. 19 (Dominion Museum), vestigal handles are present, represented by four slight double ridges on the shoulders. In each of the resulting quadrants a projection is present on the point of junction of base and shoulders, as in fig. 17.
Fig. 20 (Dominion Museum), closely resembles the preceding figure. The vestigal handles, indicated by a double row of serrations, extend three-quarters of the way up the shoulders from the point of junction of the latter with the bowl.
Fig. 21 (Otago Museum), has lost the handles. The shoulders rises sharply from the bowl and terminate in a fairly massive lip. Traces of much-reduced tiers are evident in the two ridges immediately above the junction of bowl and shoulders.
Fig. 22 (Dominion Museum), almost identical in form with the preceding figure, but has a more pronounced bowl, while a delicate neck is present below the lip.
Fig. 23 (Otago Museum), very similar to figs. 21-22, excepting that the central aperture has a permanent earthenware - 258 ware stopper, but this is counterbalanced by the presence of a large aperture to one side of the neck. It will also be noted that in this example the shoulders assume a slight curve, thus forming an inverted bowl.
Fig. 24 (Auckland Museum), has the curve of the shoulders more pronounced, and the point of fusion of the latter with the bowl is much less conspicuous, although a small spout is present.
Fig. 25 (Dominion Museum), here the bowl and shoulders blend imperceptibly, producing a gourd-like form. The only aperture is at the neck, which is emphasized by the pronounced brim.
Fig. 26 (Auckland Museum), has the body an almost perfect sphere. As in fig. 25 the only aperture is the neck. This vessel is almost identical with one from the Lau islands figured by Laura Thompson (18, p. 113). The pot is of a rich-brown colour, with a dull glaze on the upper surface, but brighter and of a lighter tone on the lower.
Fig. 26a, a vessel very similar to the above, except for the fact that the neck and lip are undecorated. This pot is figured by Edge-Partington (5, vol. 1, 12, pl. 113) who gives no information beyond stating that it is an “Earthenware Cookery vessel.”
Fig. 27 (Dominion Museum), perfectly spherical in shape, a type which appears to have arisen as a result of the gourd-shape mentioned above becoming slightly flattened. The sole aperture is the neck, which is somewhat reduced.
Fig. 28 (Dominion Museum), a good example of a subdivision of this class, namely a group in which the neck itself has become vestigial, due to the elaboration of the handles. As in the case of fig. 17 the handles are a characteristic feature, and the aperture of the neck would appear to be reduced as a result of their convergence upon it. In fig. 28 the four handles have met immediately over the spot where the neck-aperture would be, and only a vestigial opening is left at the point of fusion of the four handles. An aperture at the base of one of these acts as the neck, while a spout is present at the point of fusion of bowl and shoulder.
Fig. 29 (Otago Museum), a small bowl of the same class. The aperture of the neck is completely lost, a vestigial - 259 the four vestigial handles are present, indicated, as in fig. 19, opening is present at the point of junction of the handles—which are reduced to three in number, while a small opening between two of these serves as a neck.
Fig. 30 (Dominion Museum), a small vessel in which the handles are reduced to three and the vestigial neck aperture is absent from their point of junction. A large aperture which functions as a neck is present to one side of the handles while to the right a small spout is visible at the junction of bowl and shoulders.
Fig. 31 (Otago Museum), has the handles reduced to a single loop. As in the above, no neck-aperture is present beneath the handle, but the usual opening is present at the side of the latter, while a small spout is present on the right. Three modified tiers just above the point of fusion of bowl and shoulders should be noted.
Fig. 32 (Otago Museum), a very similar pot. The loop-handle shows a tendency to further modification. The usual aperture and spout are present, and three conventional tiers.
Fig. 33 (Otago Museum), a small vessel in which the loop-handle has become so much reduced that it can serve only as an attachment for a suspending-cord. The usual spout and aperture are present. An interesting point is that the four vestigial handles are present, indicated, as in fig. 19, by four double protuberances—the whole being very similar to a pot figured by Smythe (13, p. 89).
Fig. 34 (Dominion Museum), a good example of this class of vessel. These pots are produced by completely roofing over the bowl, the only openings being a small aperture through the ‘carapace’ and a spout, usually situated at the point of fusion of bowl and shoulders— (the roof). The head and ‘flippers’ are added after the completion of the body of the pot.
Fig. 35 (Auckland Museum), an interesting feature is the presence of four ‘flippers’ at the anterior end—two being formed by the projections of the ‘carapace’ and two by the protuberances from the bowl. The spout is situated at the posterior end of the carapace.- 260
Fig. 36 (Otago Museum), differs from the above pots in that it is an almost perfect sphere.
In respect of form the ‘units’ which compose this class of pot would appear to be very similar to the vessels shown in figs. 32-33.
Fig. 37 (Otago Museum), the units are placed in communication by a small passage running through the support between each unit immediately below the handle. The aperture for filling is visible on the left of the figure, while the unit on the right has a small spout.
Fig. 38, after Edge-Partington (5, pl. 114, vol. 1) shows an interesting variation from fig. 37. Two small spouts are present at the ends of the units at the point of fusion of roof and bowl, being put into communication by a passage in the connecting-bar. The sole aperture for filling is at the summit of the hollow handle, which connects with both units.
Fig. 39 (Auckland Museum), an almost identical vessel: this pot is of rich greenish-brown colour with a high glaze.
Fig. 40 (Otago Museum), a quadruple vessel. The aperture for filling is shown in the right hand unit, which is connected to the other by a small passage through the sides. An outlet is provided by a small spout in the left hand unit. The handle is solid, and is attached to the two central units by two slender processes.
CLASS G.—‘Fruit-Cluster’ pots.
This class is closely related to the preceding group, since each globe is actually a small bowl completely roofed over, the point of fusion between bowl and roof being usually clearly indicated.
Fig. 41 (Otago Museum), a typical triple-cluster. A line of serrations in high relief emphasizes the point of junction between roof and bowl, while a large aperture is present in the globe on the left, which communicates with the remaining two by a small passage in the thick central connection. The sole outlet is the small spout in the globe on the right of the figure. An interesting feature, which provides the most direct evidence as to the derivation of this type, is the large vestigial aperture present at the apex - 261 of the handle. On comparison with figs. 38-39 it is apparent that all that is required to transform the present example into a ‘unit’ pot of the double type, is the removal of the third globe together with a slight compression of the remaining two, which would then be connected by a hollow handle with an aperture at the apex.
Fig. 41a, after Edge-Partington (5, pl. 114, vol. 1), is of a more complicated type than the above two. Here the aperture at the summit of the handle appears to be the only opening to and from the globes, of which there are five, connected by the hollow handle. It does not appear from the drawing that the globes are definitely connected with one another other than by the handle, but if this is the case, the pot could not have been of any practical use, although a small spout is present to the right of the figure.
CLASS H.—Open bowls with a base.
Fig. 42, after Edge-Partington (5, pl. 112, vol. 1), a kava-bowl in the von Huegel collection, Cambridge. From the shape it would appear that this class of vessel is built up in an inverted position. The colour and glaze are not indicated.
Fig. 43 shows a curious type of dish which is figured by Edge-Partington (5, pl. 56, vol. 1). The triple-bowls are mounted upon a comparatively massive stand, but beyond stating the width—twelve inches and a half—Edge-Partington gives no further information.
Fig. 44, a cauldron like pot also from Edge-Partington (5, 2, pl. 113, vol. 1). The legs appear to be somewhat crudely fused with the bowl, and the out-curving lip is clearly a secondary addition. Unfortunately, beyond stating that the vessel is an earthenware kava-bowl, no further particulars are given.
Fig. 45, a kava cup of yellowish-green earthenware in the Canterbury Museum. This vessel consists of a bowl twelve inches in diameter mounted upon three legs, the lower extremities of which are fused in a circular base, the whole being very similar to a wooden one also in the Canterbury Museum.- 262
Figs. 46-47a, two pottery baskets in the Otago Museum. The handles are somewhat crudely attached, while the sides of the bowls are thrown into scallops for a third of their depth, and a small base is present. Both the pots are a warm-brown in colour, and have a high glaze.
Figs. 47-48, an interesting pot recently acquired by the Otago Museum. The body of the vessel is similar to that of fig. 13, but in this example the whole is surmounted by a carefully-modelled human head, with details in bold relief. The hair alone is rendered in a conventional style, terminating in a decided ridge a short distance above the forehead. The ears are of interest in that they have been attached with their axis in a vertical direction, while that of the head is at an angle of 45°. The pot itself is undecorated, save for a curious projection at the base of the neck and an incised design consisting of six lines running round the neck at the level of the above-mentioned projection. From the former several vertical groups of five to seven incisions are given off in a downward direction, the whole being very similar to a sling for the pot, as shown in the figure. Two apertures are present in the head—a small one representing the mouth, and a larger one at the crown of the head—both of which at the time of receipt were stopped with corks. The upper part of the vessel is of a light red colour, which blends with a dull green from about the middle downwards.
Fig. 49, a three-tiered pot in the Canterbury Museum; an interesting feature is the manner in which the handles are reduced and converge round the neck, thus affording an intermediate stage between figs. 17-18 and 19.
The decoration upon Fijian pottery is almost entirely confined to the upper part of the vessel—i.e., shoulders, neck, and lip—a tendency which is also evident in the pottery of the New Hebrides. (J.P.S., vol. 48, p. 40.)
Figs. 18, 21-23, 32, and 41, show an interesting ‘rickrack’ design formed by a series of inverted Vs. This type of decoration appears to be a derivative from basket-work. Lehmann records this design on various articles from the - 263 Carolines (9, Tafel 4); New Caledonia (9, Tafel 5); North and South Solomon islands (9, Tafel 7); and the Admiralty islands (9, Tafel 8). The addition of Fiji to this list shows that the motive is one of the most widely-spread of Oceanic decorative elements.
A motive which appears most consistently on Fijian pots is a series of protuberances which usually mark the point of fusion between bowl and shoulders, as in figs. 8, 15-16, 18, 21-24, 28-30, 49, 52. In some examples this is accompanied by a series of incisions, as in figs. 16, 22, 35, 38, while in fig. 46 protuberances adorn the edge of both bowl and handle, but in fig. 47a they are confined to the latter.
All the ‘unit’-pots figured here show this latter type of decoration as in figs. 37-40. In figs. 38, 40 the protuberances form a row up the centre of the upper surface of each unit, while in fig. 39 a double row is present, which continues along the angles of the handle.
Closely allied to the protuberance as a decorative element is the knob or stud, which is very prevalent in Fijian pottery. Fig. 20 shows a simple example; here four large flattened knobs are present, one in each quadrant of the shoulders.
Fig. 30 shows a pot in which the knobs form a series of vertical rows upon the shoulders, interspaced by a similar group of serrations.
Fig. 29 is a pot in which the stud motif is more complex than in fig. 30. The shoulders have a series of double and treble vertical rows of knobs, interspaced by a group of vertical serrations.
Fig. 28 shows a very interesting design. Here numerous knobs form a scattered design upon the upper part of the shoulders, while on the lower a triangular design is formed by a series of serrations, each of which has a knob in the centre. This design appears to be closely related to that which appears on a sherd from the New Hebrides, now in the Canterbury Museum, where a knob is present along with a rectilinear group of ripples. (J.P.S., vol. 48, p. 50, fig. 46.)
Fig. 2 shows the knob-motif expressed in a slightly diflerent form. Here the numerous protuberances are confined to the upper part of the bowl, and the desired effect produced by leaving blank a number of circular spaces.- 264
Fig. 3 illustrates a very interesting design in which the knob plays a conspicuous part. Here the lower surface of the vessel is covered by an irregular mass of knobs, while a single row is present just below the lip, each two of which are interspaced by a Y shaped projection. This element is recorded by Lehmann as occurring on various articles in New Caledonia (9, Tafel 5), and the Philippine islands (9, Tafel 15).
Fig. 39 shows a modification of the ‘coco-nut-leaf’ design (12, p. 231), which in this example consists of rows of V-shaped serrations on the upper surface of the pot. Lehmann records this motive on various articles from the Carolines (9, Tafel 4), Berlin Harbour (territory of New Guinea) (9, Tafel 10), and the Torres Straits (9, Tafel 11), and North New Britain (9, Tafel 7). A similar design appears on some sherds from Mohenjo-Daro (10, pl. 91, fig. 28).
Fig. 50 shows a fragment of the ‘carapace’ of a turtle-pot in the Otago Museum. This design is identical with that on a sherd from Nguna, New Hebrides (J.P.S., vol. 48, p. 44, fig. 24), while it is very similar to the decoration on the shoulders of the pots shown in figs. 27, 31, 52, though the latter may be a modified basket-work design. Lehmann records this type of decoration in wood and basket-work from the Marquesas (9, Tafel 1), New Caledonia (9, Tafel 5), South Solomons (9, Tafel 6), and the Diri of Central Australia (9, Tafel 12), as well as Ceylon and South India (9, Tafel 20).
Figs. 19, 21, 23, 32, have on their shoulders a serrated design which appears to represent a type of sling, similar to one figured by Bossert (2, pl. 24, fig. 13), from the Empress Augusta river (territory of New Guinea).
Fig. 24 shows a matting-design on the shoulders. This type of design is similar to that of a confinement-mat of traditional pattern from Nauru (2, p. 353, fig. 4).
Fig. 45 has on the stand a ripple-design of the same type as mentioned above, while the legs are serrated in a manner which gives the appearance of crude fluting. The design on the outer surface of the bowl is of a simple rectilineal type, but the chief interest with respect to decoration lies in the lip, which shows the ‘coco-nut-leaf’ motif - 265 to perfection (fig. 45a). The design is divided into four unequal quadrants by groups of eight to eleven transverse notches, being similar in this respect to the lip of fig. 12.
Fig. 49 shows a very unusual type of decoration, consisting of double rows of wavy lines making opposite angles with one another, and continuing unbroken round each tier. This design is recorded by Lehmann on various articles from the South Solomon islands (9, Tafel 6), the Empress Augusta river (territory of New Guinea) (9, Tafel 10), the Ainu of East Asia (9, Tafel 21). Bossert figures a reed-house in the Matthew group (2, p. 335, fig. 6) in which this design is worked along the front, between the eaves. A vase of painted pottery from Mohenjo-Daro shows a similar design, which in this example forms a band round the centre of the vessel (10, pl. 90, fig. 23, pl. 87, fig. 4). This motif is discussed in detail by Stolpe (16, pp. 14-57) in dealing with the degeneration of the human figure as a decorative element, while a further consideration is given by Dr. H. D. Skinner in his Morioris of the Chatham Islands (14, p. 67, figs. 4-5, p. 68, figs. 6-7).
Since the pottery of Fiji is largely confined to storage-vessels (chiefly for liquids) its efficiency is scarcely impaired by any shape it may assume; in fact, a pot in which the aperture is reduced is decidedly more suitable for fluids than an open-mouthed jar. This may have been a contributing factor in the development of classes D-G above. Gordon-Cumming suggests that Fijian pottery types had their origin in the clay nests of certain spiders inhabiting these islands (6, p. 250). Schurig, however, states: “. . it is improbable that the shapes of Fijian pottery are direct imitations from prototypes in nature: bunches of fruit, flowers, canoes, etc.… Primitive man no more fashions things direct after nature than the child does; his inspirations come direct from his own imagination and from settled traditions… But even the natural objects which he fashions out of his head are not done because of the aesthetic satisfaction which he derives from them, but because they bear some relation to the vessel which he wants to make. This relation may be of a technical nature, but it may also - 266 be magical. It is possible that the composite vessels recall the coco-nut shells which are covered over with a netting and bound together in pairs for use as water bottles. And turtle shells were first perhaps used as containers; then later on the whole creature was fashioned in clay …” (13, p. 138 on). It is evident, however, that the modern pottery of Fiji is very similar to the more ancient examples from the New Hebrides, so much so as to suggest that both arose from the same source, but for a long period have been following an independent course, and acquiring individual characteristics, yet both Fijian and New Hebrides pottery may be regarded together as a single division of Melanesian ceramics. Ethridge, however, describing some ancient sherds of Mallicollan pottery (5a) states that “the glazed and artistically decorated ware of the Fiji group offers no resemblance …” (5a, p. 199). This appears to be true in so far as the ‘fish scale’ pattern is concerned (5a, p. 198), but if the most elaborate of the New Hebrides sherds (5a, pl. 35, fig. 3) be compared with figs. 16, 17, 19-21, 23, 24, 26, 31, 33-34, 36 and 53, it will be noted that all the elements of the design in that piece are also present to a varying degree in the latter.
Up to the present time the only discovery of native pottery in Polynesia has been that of the Bishop Museum expedition in Tonga, where altogether 1,577 sherds were collected. This was a highly porous, lightly fired and fragile ware. McKern, in the Bishop Museum Bulletin dealing with the expedition states that “Tongan pottery seems to resemble that of Fiji in many essential traits, but to differ from it in many details. The similarity between the cooking ware is much more pronounced than between finer, decorated ware …” but it is evident that “Tongan pottery is certainly Fijian in origin and basic type.” 4
As mentioned above, two methods of construction are employed in Fiji, namely, the coil- and patting-methods, while the coil-technique alone is present in the New Hebrides. In this respect it is interesting to note that as far as localization has been possible with the examples figured here, the pieces exhibiting features analogus to New Hebrides charac- - 267 teristics are from districts employing the coil technique (13, map 1).
In respect of elements of decoration, affinities are distributed from the Indus Valley in the west to the Marquesas islands in the east, with an area of greater concentration centreing in the Fiji islands. The area to the east of the Fijis affords fewer parallels than that to the west, a line of evidence which may indicate the route by which the art of ceramics has entered Oceania, a definite answer to which can only be given after a survey of the region centreing round New Guinea and Malaysia, concerning which the writer intends to publish in the near future.
Since the completion of this paper Leo Austen's work relating to the Megalithic Structures of the Trobriands has appeared in Oceania (1) in which he gives a description of the funerary pots of that locality. The decoration upon those which he figures shows no similarity with any on Fijian pottery, but appears to have marked affinities with New Guinea and the Solomon islands. The Trobriand pots show no similarity with respect to shape when compared with the Fijian examples, but one important point is, that the decoration is confined to the shoulders, and the junction of the latter with the bowl is indicated by either a row of serrations or scratches (1, figs. 1-8), as in the case with the New Hebridean and Fijian pottery.- 268
KEY TO MAP
1 This test is based upon the reflection of a light-source. In measuring the gloss of a piece of pottery a pocket-torch is held so as to make an angle of about 55° with the glazed surface, and almost in contact with the bulb of the torch. The reflection is then noted and compared with the standards given in Pl. 1 of March's Standards of Pottery Description, University of Michigan Press, 1934.
2 In which district there is a village called Nggelekuro=clay-for-pottery.
4 “Archaeology of Tonga,” by W. C. McKern, B.P. Bishop Museum Bulletin 60, pp. 119 and 122.