hurricanes or when heavy N.W. swells or seas and the salt spray completely kills off all the leaves and vegetation which do not recover for many months. At such times certain species of birds most suffer almost if not 100 per cent mortality. The last serious hurricane or rather sea from a hurricane further north to completely destroy all green vegetation etc. of these islands in the N.W. was in 1936 (Jan.) and normality was not reestablished till 1937. During my stay there (May '38) most species were established and reasonably plentiful possibly as plentiful as food supplies warranted. In fact considering the comparative scarcity of bird life in the eastern ti tree strip on the mainland one would consider these islands in the west carried a population above capacity. Water borne seed trees and so on and coconuts which are almost entirely absent in the ti tree strip of the east are probably responsible for the denser bird life of these islands. It is very noticeable on these islands that Zosterops, which does not visit coconuts on the mainland to any great extent to feed, spends much of its time of the islands closely searching the hessian like fibres of the coconut plam heads. Under this fibre countless insects find shelter and doubtless Zosterops and Lichmera and Pachycephala also, all of which I have seen feeding thus, have long since discovered this fact. Also the eastern mainland coastal strip is subject to salt burning at all seasons of the year and much more frequently, whereas these northwest islands are only subject to such during the hurricane season and that at wider intervals and are more or less protected during the nine months of the prevailing wind (S.E. and N.E. and E). Such N and N W to S W winds as do occur during this season are generally of short duration and the former very light. Hurricane winds or