The classes of distributions, such as the normal, gamma, Pareto and the like, which are known by their names and have some attractive properties, form a very small part of the space of all distributions, or even of continuous distributions. A greater variety of distributions is obtained by transformations of these common (named) distributions, and it can be further extended by other operations, such as convolutions. But none of them offer as much flexibility as mixtures (Everitt and Hand, 1981; Titterington, Smith and Makov, 1985).