The Coming of the Maori
The walled houses were originally designed to give greater height so that people could move about in the erect position. This was accomplished by making side walls consisting of wall posts which supported a horizontal wall plate at their upper ends. The extra height of the walls had to be added to the length of the two median ridgeposts which supported the ridgepole. The lower ends of the rafters, instead of resting on the ground, were supported by the elevated wall plates and thus the extra height was secured.
In central Polynesia, the rectangular walled house was the type in general use. The upper ends of the rafters (oka) were crossed in pairs over the primary ridgepole to provide a crotch for a second ridgepole rendered necessary by the technique of pinning on the ridge thatch. The lower ends of the rafters projected beyond the wall plates to form eaves. Horizontal poles or battens (purlins) were spaced on the rafters and lashed to them with sennit braid to steady the roof framework (Fig. page 11710d, e). The lashings around the rafters were made in patterns to add decoration to utility.
No detailed description of the technique of the framework of the common Maori houses, including cooking houses, is available to me and my memory cannot supply details which were never noted in spite of frequent contact with them. However, it may be assumed that the early settlers brought the knowledge of walled houses with them from whatever part of Polynesia they came. This assumption is supported by the terms used for some of the principal elements in the framework of the houses. In both New Zealand and the Cook Islands, the name for the ridgepole is tahuhu (Cook Is., ta'u'u) and for the ridgeposts pou tahuhu (Cook Is., pou ta'u'u). The Cook Islands' dialect resembles that of the Taranaki and Whanganui tribes in using the glottal stop for the h sound. The wall posts in Maori are poupou and in Rarotongan, pou-turuturu. The principal rafters are termed oka throughout Polynesia and the use of the term oka for the slabs which represent rafters in roofed store pits indicates that oka was also the term for rafters in common houses until the elaborate development in superior houses led to the rafters being termed heke. The term kaho was used in both areas for the wooden elements to which the thatch was tied. The surviving terms indicate that the common walled houses of New Zealand were based on the central Polynesian pattern.
The thatch material used in New Zealand led to marked differences from the technique now prevailing in central Polynesia. There, the thatch consisted of split coconut leaves plaited in lengths of a few feet or of pandanus leaves formed into sheets by doubling the butt ends over a rod and pinning the doubled parts together with dry coconut leaflet midribs.
The foundation for the thatch was formed by placing a second light ridgepole in the crotch formed by the upper crossings of the main rafters, laying long rods termed Kaho (Rarotonga, ka'o; Tahiti, 'aho) on the second ridgepole and the purlins, and tying an eaves-batten to the lower ends of the kaho which projected beyond the wall plate to form eaves. The kaho rods were spaced a few inches apart and, as they ran parallel with the principal rafters and were for the purpose of attaching the thatch, I have termed them "thatch rafters" (Fig. 10e). The thatch sheets were laid horizontally on the thatch rafters with their stiff edges above and they were lashed (ato) in two places with sennit passing around the stiff element of the upper edge and the appropriate thatch rafter beneath. Successive sheets were overlapped from the eaves to the ridgepole and sections one sheet wide were added from below upwards until the roof was covered (Fig. 11a). Special thatch sheets were doubled over the ridge and kept in position by long wooden pins passed through between the two ridgepoles (Fig. 11b).page 118
Fig. 11. Roof thatch
a, thatch sheets, Cook Is.; b, ridge sheets, Cook Is.; c, thatch bundles, Maori; d, ridge thatch, Maori.
The Hawaiians used a similar technique with pili grass and, though pandanus grew abundantly, thatch sheets were not made with its leaves as in central Polynesia. The horizontal rods to which bunches of pili grass were attached in horizontal rows were termed 'aho, the equivalent of the Maori kaho. It is evident that kaho in its dialectal forms was applied to the wooden rods to which thatch was tied, whether their direction was obliquely upwards like rafters or horizontal like purlins. I once thought that the use of thatch purlins in New Zealand was due to the absence of pandanus, but the use of the same method in Hawaii where bunches of pandanus or ti leaves were sometimes used with the pili technique, dispels that theory. In Rapa, the thatching also consists of bunches of pandanus leaves tied to horizontal rods. The use of thatch purlins in the marginal areas of Hawaii, Rapa, and New Zealand indicates that the method is old and may have been present in central Polynesia before dispersal took place. However the use of thatch rafters with thatch sheets is so widely spread throughout the Pacific that both methods must have been known at an early period.
The framework of the side walls consisted of spaced wall posts with a wall plate lashed to their upper ends. The wall plate supported the lower ends of the main rafters. In central Polynesia, the wall spaces were usually filled with vertical stakes which allowed air to circulate through the interstices. In Samoa, the walls were formed by plaited coconut sheets strung together like Venetian blinds. In Hawaii and Rapa, the walls were matched in the same way as the roof. In New Zealand, the walls were thickly thatched to keep out the cold air.