[Random sample planning in morphometry]

This report tries to explain principles and presuppositions of planning sample sizes. The aim is estimation of the optimal, that means minimally necessary sample size, justifying an investigation ethically and economically. The problems are complicated because of the necessity of nested sampling or samples within samples. We describe the most important presuppositions using an example of mean value estimation. These are: establishing of demands for exactness, of error probability, and knowledge of variance of the parameters. It is explained in detail by means of binomially distributed variables as we find them in the point counting method.