Arts and Crafts of the Cook Islands
The original house type was evidently rectangular in ground plan with median ridge posts supporting the ridgepole directly. For small houses, a median ridge post at each end was sufficient but for a longer house, additional median ridge posts were needed to prevent the ridgepole from sagging in the middle. This form of architecture is widespread (fig. 261, a).
The thatch material of pandanus sheets for better class houses and coconut leaf for the rougher houses was probably old. The sliding door which reached New Zealand was probably brought in with the first settlers.
The use of sennit in lashing house framework together with neat lashing patterns was old, but the use of ornamental patterns on purlins and rafters (figs. 17, 18) was evidently confined to Mangaia and may be regarded as a local development.
Four-legged wooden seats made in one piece were present in only the Cook, Society, Austral, and western Tuamotuan islands. They must have originated in the second period and diffused from the Society Islands, where wooden pillows were made. The simplest form is from Tahiti (fig. 260, b), where the seat is flat or slightly curved. The four legs are straight, round in section but with a vertical edge on the outer side, and without feet. This form without feet must follow an old Tahitian type, as a specimen in the Royal Swedish Museum, Stockholm (fig. 260, c) was collected by Sparrman of Cook's second expedition. As this expedition did not call at the Cook and Austral Islands, there is no doubt that the seat belongs to Tahiti.
A second type—with a well-curved seat, curved legs with heart-shaped feet, with the edge on the legs and feet directed toward the middle transverse line—is known to have been made in Atiu in recent times (fig. 260, a). This is a local variation of the Tahitian type. The long seats of Aitutaki (fig. 19, b) and the single pedestal seat of Rarotonga (fig. 19, c) are also local insular variations. Although two seats of the second type in the British Museum (C.C.St. 870 and 1905/1-20/4) have been attributed to Tahiti, and another in the Royal Scottish Museum, Edinburgh (?) to the Austral Islands, the technical details prove that all three seats belong to Atiu.
The seat of a third type of stool has an even greater curve than that of the Atiu type. The two ends are deeply concave and the middle part of the seat at the ends is raised in an even curve above the general plane of the seat. Neither of these features is present in the two types described above. The legs have feet which are circular, evenly projected from the legs, and with a flat upper surface that meets the legs at right angles. The under surface of the feet is convex. The legs are curved but the concave curves facing the ends of the seat come straight down to the feet and do not slant back as in the Atiu type. In section, the legs are somewhat heart shaped with the edge facing toward the median transverse line as in the Atiu type. One of the seats was figured by Ellis (25, p. 181) and attributed by him to Tahiti. However, four seats of the same type are carved along the sides and ends of the seat with carving that is typical of the Austral Islands. The four carved seats are located page 422as follows: British Museum (L.M.S., 471); Oakland Public Museum, California; Leipzig Museum (Po. 388); and the Peabody Museum, Cambridge (48,450). Another seat, partly carved, is in the British Museum (Tah. 3), and one without carving is in the Copenhagen Museum. The technique, which is so different from that of the authenticated Tahitian seats of type one (fig. 260, b, c), and the presence of carving lead me to attribute this type to the Austral Islands (fig. 260, d).
The fourth type, from western Tuamotu, follows the Tahitian flat seat with legs with an outer vertical edge and no feet. The legs are higher and near the ends of the seat and the lower ends are sometimes flared. A wooden rung is usually lashed between the end legs with sennit (fig. 260, e).
Figure 260.—Types of small four-legged seats. a, type one, Cook Islands (British Mus., 1905/1-20/4): 1, curved seat with slight concavity at ends, not carved; 2, leg heart shaped with edge (3) toward middle line; 4, foot heart shaped with point toward middle. b, type two, Tahiti (Bishop Mus., C4190): 1, seat with little curve, straight ends; 2, straight legs, round in section but with edge (3) to outer side; no feet. c, type two, Tahiti (Royal Swedish Mus.): 1, seat with greater curve than b, concave ends; 2, legs round in section but with vertical edge (3) to outside; no feet. d, type three, Austral Islands (British Mus., Tah. 3): 1, seat with great curve, concave ends; raised plane in middle at ends (5), and carved edge; 2, legs heart shaped, with edge (3) toward middle; 4, round foot. e, distinctive Tuamotuan seat (Bishop Mus., 5973): 1, flat seat with straight ends; 2, long legs rounded in section set close to ends of seat; 4, no feet but lower ends flared to roughly rectangular shape; 5, rung between legs.
For Samoa, I describe (73, fig. 295) a three-legged seat said to have been., used by fowlers. The legs were separate and lashed to the seat. Brother Fred page 423Henry of Leone, Tutuila, informed me through correspondence that wooden seats were used in Western Samoa during the installation of a high chief. Available information indicates that the seat was a length of coconut tree trunk with the lower end scooped out to form three or four legs in the same technique that was used to make the wooden anvils (malaise) used in beating coconut husk to clean fibers for making sennit (73, pl. 18, B, 4). The seats were thus elongated anvils and had no affinity in technique with the four-legged stools described above. Four-legged stools are confined to the central area of Polynesia, where they probably spread from the Society Islands to the Cook and Austral Islands, and to western Tuamotu. Tahitians retained the simplest form, whereas Cook and Austral Islanders elaborated on an early design by making a deeper curve to the seats and adding feet to the legs. The Aitutaki people on the other hand made a longer seat but retained the Tahitian form of legs without feet. Aitutakians and the Austral Islanders added their own form of carving to the edges of the seats.
Figure 261.—Types of house ridgepole support, a, median ridgepost, Cook Islands. b, kingpost with tie beams supported by two lateral posts, Samoa. c, kingpost with tie beam supported by wall plate, post-missionary. 1, median ridgepost; 2, ridgepole; 3, wall posts; 4, wall plate; 5, principal rafters; 6, lateral supporting posts; 7, plate supporting tie beam; 8, tie beam; 9, kingpost; 10, plates resting on ends of tie beam and supporting rafters; 11, median plate supporting kingpost.
Note: the second ridgepole, purlins, and thatch rafters are omitted. The Samoan rafters are curved and kept in position by collar beams which are also omitted.
In Mangaia, a house was made with tie beams (te'a) stretched across the wall plates and with king posts (pou te'a) erected on the middle of the beams page 424to support the ridgepole (fig. 261, c). At first glance this form of framework might be classed with the king-post type of western Polynesia but there is an important structural difference involved. In the Mangaian house, the tie beam rests on the wall plate which is supported by wall posts whereas in the western houses, as I have said, the wall posts and posts supporting the tie beam are distinct from each other and set at different distances from the middle line of the house. In addition, the western tie beam is supported solely by its own two posts and does not rest on the wall plate which is at a lower level.
The terms te'a for tie beam and pou te'a for king post are Tahitian words in which the hamza represents the dropped k. If the form of house were old Mangaian, the terms used in the Mangaian dialect would have been teka and, pou teka. The Mangaian form of king-post house described for the Society Islands by Handy (41) has mortices and tenons that are not practicable without steel tools. I conclude, therefore, that the European form of king-post house was introduced into the Society Islands by European missionaries in their own dwelling houses, schools, and churches and that it was later introduced into the Cook Islands by missionary teachers who went there from the Society Islands. They also carried with them the Tahitian building terms. An addition with a rounded end (termed po'o ta'a) is attached to the ends of some of the rectangular houses. A number of radiating rafters are attached above to the top of the gable end and below to a curved end plate. This technique is used in the fare pote'e of Tahiti (41, pp. 10, 11) in which both ends are rounded, whereas in the Cook Islands they are added to only one end for a kitchen. I believe that this addition was also borrowed from Tahiti in post-missionary times and that the Rarotongan term of po'o ta'a is a version of the Tahitian (fare) potee which should be pote'e.
Another addition is a square lean-to, in which parallel rafters extend from the end rafters of the gable end to rest on a straight end plate supported on posts. This technique resembles that of the fare taupe'e of Tahiti (41, p. 17) in which, however, the end plate has a convex bend upward. Again, the Tahitian fare taupe'e has the lean-to addition on each end, whereas the Cook Islands lean-to is attached to but one end and used as a kitchen. It is probable that this technique was also borrowed from Tahiti in post-missionary times.