A preliminary study was made of the chemistry of pyrogallol- and catecholtannins, their general properties and methods of extraction and determination. Three methods of estimation - Lowenthal, powdered hide and spectrophotometry - were compared using two control solutions, four samples of wood and one of bark. Using the empirical powdered hide method, tannins of both types were estimated in wood and bark of various tropical species (some separately and some as a mixture), Moroccan oaks (Quercus suber and Q. ilex), and European oak 9Q. petraea). Further tests were made on the wood and bark of the two mangrove species, Rhizophora mangle and R. racemosa, by subjecting them to successive extraction with a range of solvents. None of the woods tested had as much as the 10% of tannins considered necessary in economic sources. The bark of the two mangroves, of Eucalyptus urophylla and of Prosopis africana had tannin contents over 10% and the latter two species had very favorable tannin/non-tannin ratios. All the tropical species, with the probable exception of E. urophylla, had only catecholtannins. Only the oaks and E. urophylla bark gave positive results when tested for gallotannins.