39. Australia may be contributed but if so probably via New Caledonia. The most likely case is that different areas have all contributed and probably Caledonian species are more numerous than others. Any treatment of the subject of bird distribution by hurricanes would necessarily be a great deal of surmise as so many factors would have to be taken into consideration which could so easily alter the completion of the whole subject. (The greatest factor is of course the irregular behaviour of such storms themselves that being the only really regular thing about them). Actually the distance between land masses either from the east (Lifu) or south (N.C.) is not too great for direct flight migration to have taken place. This is especially true as regards coming from Lifu, especially as birds would be flying down wind on a calm or normal day of prevailing wind of steady direction. Also they are close enough for this to be more or less true for an average hurricane which in such a distance would hardly be likely to curve enough to affect the general more or less straight course between the islands and thus deflect birds to mill the island altogether. The same might be said to be true as regards the S E corner of the New Caledonia mainland as from there the wind curve of a circular storm would tend towards the Loyalties. It is probable the Loyalties have received birds from both north (the N.H.) and south (N.C.) but until comparisons of specimens can be made the amount each has contributed cannot be gauged. Though I have no definite grounds for saying so I believe the actual case is that most species or close species common to all 3 areas have come with hurricanes from the New Hebrides to New Caledonia, and developed their individual charac- teristics in their new localities. Some species of course having