The role of time and size in ontogenetic allometry: I. Review.

Several interconnected issues are a part of most studies of ontogenetic allometry, the relationship between size and shape during growth. One issue is the choice of a model, either linear or one of a series of nonlinear models that have been proposed in the literature. The independent variable, against which growth is assessed, can be a measure of time, e.g., age or a measure of size, e.g., weight. These categories of independent variables have become confused with Medawar's (1945) classification of allometric studies as empirical or deductive. Medawar's distinction may be less useful than Tukey's (1980) between exploratory and confirmatory studies. The "correct" choice for some of these options will be a function of the data set being analyzed. However, an understanding of the implications of these issues is necessary to make the correct choices.