IN view of the recent correspondence in NATURE1 bearing on the question of reproductive periodicity in organisms, I should like to direct attention to some observations made by Miss S. M. Marshall and myself in Australia in 1928–29, which appear to indicate a type of such periodicity not previously recorded. In shallow pools on the Low Isles coral reef, one of the common corals is the branched species Pocillopora bulbosa. On some occasions, collections of branches of this coral will give off large numbers of planulae at other times, large collections of branches will give off no planulae at all. During the period August 1928 to July 1929, about eighty collections of branches were made, representing hundreds of colonies. When the results of these collections were plotted, it was found that the production of planulae was discontinuous, periods of production alternating with periods of no production. In the months December to April the productive periods occurred at about the time of new moon ; in July and August at about the time of full moon ; and in May and June a transition period was demonstrable from new moon to full.