Ou-Kaapse meubels en stinkhoutidentifikasie: moontlikede met kanoniese veranderlike-analise en bistippings.

Die identifikasie van houtsoorte waarvan Ou-Kaapse meubels vervaardig is, is 'n wesenlike probleem vir kultuurhistorici, meubelhandelaars, versamelaars en museumkundiges. Houtidentifikasie word meestal net op uitwendige voorkoms en seide op houtanatomiese kenmerke gebaseer. Hierdie artikel bied 'n kultuurhistoriese perspektief op die vraagstuk, waarna die omvang van suidelike Afrika se inheemse houtsoorte met die gepaardgaande probleme met hout-identifikasie onder die loep geneem word. In besonder word probleme met die uitkenning van verskillende houtspesies wat aan dieselfde genus behoort, uitgewys. Ten slotte word aangetoon hoe statistiese metodes soos kanoniese veranderlike-anaIise en meegaande bistippings, uitgevoer op kwantitatiewe houtanatomiese data, aangewend kan word om op 'n objektiewe wyse spesies soos stinkhout en imbuia van die genus Ocotea te identifiseer. Onsekerhede verbonde aan sodanige identifikasies word ook gekwantifiseer. Old Cape furniture and the identification of Stinkwood : possibilities with canonical variate analysis and biplots The identification of woods used in manufacturing Old Cape furniture is a fundamental problem for cultural historians, furniture dealers, collectors and museologists. Usually the identification of these woods is based only on the external appearance of the wood, and rarely on wood anatomical properties. This article presents a cultural historical perspective on the problem, followed by a discussion of the extent of southern Africa's indigenous woody species and the associated wood identification problems it poses. Specifically, the problems associated with the identification of species belonging to the same genus are highlighted. The potential of objective wood identification based upon quantitative wood anatomical data together with statistical techniques such as canonical variate analysis and accompanying biplots, to distinguish between stinkwood and imbuia of the genus Ocotea, is illustrated. Uncertainties associated with such identifications are also quantified.