Bivariate generalizations of Neyman's Type A distribution

Three generalizations of the Neyman Type A distribution (Neyman, 1939) to the bivariate case are introduced, each of which is such that each of its marginal distributions is univariate Neyman A. The models from which they arise seem appropriate in certain ecological situations. Method of moments estimators of the parameters are derived, and an example is given of an attempt to fit the three distributions to ecological data. The problems of maximum-likelihood estimation and the efficiency of moments are not discussed here.