Petal inception and the problem of pattern detection.

Abstract This comparative study is based on seven cases of petal inception which differ mainly in their relative position with respect to the stamen primordia (Fig. 2). By means of our ordinary botanical language the seven cases may be classified into mutually exclusive patterns. However, this kind of classification (homologization) leads to contradictory and ambiguous conclusions. Therefore a conceptual model is proposed which overcomes these inadequacies (Fig. 3). In this model the concepts “primordium I and II” are used instead of “petal” and “stamen” primordium. It is shown that in all seven cases after the formation of primordium I, a primordium II is initiated in different locations. The differences between any two of the seven cases can be understood quantitatively as the distance between the sites of primordium II inception. In all cases the outer primordium, whether it is primordium I or II, develops into the petal. Some of the conclusions are the following. 1. (i) The model converts qualitative differences into quantitative ones. In this way the problem of which cases are essentially similax and which ones are fundamentally different is eliminated. In this sense the model is an example of a (semi-) quantitative homology concept. If “patterns” should be distinguished, the notion of “prototype” must be eliminated from the definition of “pattern”, so that the difference between patterns would be a quantitative one only. 2. (ii) Although it is known that a certain primordium will develop into a petal or stamen it cannot be called a “petal” or “stamen” primordium in comparative morphology. 3. (iii) Pseudoquestions and pseudoproblems are eliminated by the model. (iv) Gross distortions of similarities are avoided.