A search for pattern and form

Seed science is a microcosm of biology which deals with a wide range of hie rarchical levels of organization - from molecular biology to population eco logy. information at every level accumulates at an alarming rate. Making se nse of it all calls for generalizations. Those which enable predictions to be made for circumstances different from those in which the observations we re first made are particularly useful. Many scientists approach these probl ems by searching, consciously or unconsciously, for patterns or forms. Patt erns for this purpose are defined here as discontinuities with some measure of repetition, whereas forms are thought of as continuous shapes in two or more dimensions; both can be visualized as spatial structures (most people have difficulty in conceptualising more than three or, at most, four dimen sions). If one considers the whole range of hierarchical organization in bi ology, from molecules to communities, the recognition of the patterns and s tructures of processes at one hierarchical level of organization may assist in understanding the processes at adjacent levels, but such knowledge may not always be helpful in explaining processes at more remote hierarchical l evels where different rules and mechanisms may predominate. Arbitrary curve fitting sometimes has a role in the recognition of the form of processes, but it is usually better to try and discover relationships where the coeffi cients have some biological meaning. Not only can a search for patterns and forms of processes help in the interpretation of data and the development of new ideas, but it can also sometimes help in the design of experiments. This personal view deals with some examples taken from seed research in the hierarchical levels in which I have been involved - usually somewhere betw een molecules and ecology.