Inrecent discussions on the geographical distribution of animals and plants the subject has been approached from a novel point of view. An attempt has been made to investigate certain problems of distribution by the statistical or bio‐ metrical method; and it is thought by some that this method may throw new light, not only on the phenomena of distribution, but upon the general principles of biology, of which the facts of distribution must be in some degree the ex‐ pression*. There appear to be certain purely statistical aspects of the problems raised, which have so far been ignored; and since some of these considerations seem to be of a fundamental character, it is perhaps worth while to place them upon record. I advance these considerations with diffidence, because the points which I am about to make are in no way obscure; they are on the contrary very simple; and the only excuse for discussing them is that their fundamental im‐ portance for the statistics of distribution appears to have been overl