Morphology provided the bulk of evidence for evolution at the times of Darwin. Its roots lie in predarwinian concepts of typical similarities, organizing diversity of organisms and organs over representation by types. While morphology was subsequently redirected towards reconstructions of evolutionary history, one concept of the type sees a reflection of typical similarities with regards to construction (Bauplan) in typical similarities with regards to development (Werdeplan). The type itself is constructed for a particular group of organisms, or organs, so that the members of the group can be derived by the least number of natural steps. Despite this uniquely clear concept of Naef (1935), subordination of morphology under phylogeny and a general decrease in interest in questions not phrased in the terminologe of genetics and molecular biologe sidelined such questions. It became clear over the recent past that morphology can deliver more than a platform for comparing character states and classifying them as homologues or homoplasies. Wagner (1989) relates in his biological homology concept character identity to the organization of a character within an organism. Repeated notions of latent characters (e.g. Wake, 1999) point to a level of character organization transcending or underlying (dependent on the viewpoint) observed states. I propose to outline a biological type concept starting from Naef' s (1935) type concept, as it is explicitly aimed at organizing natural diversity so as to reflect norrns of development. This comes at an expense - phylogenetic reconstruction becomes less easy within such a framework. Molecular methods, becoming increasingly popular, more than fill such a conceivable void. My case study are paired fish fins. Their phylogenetic derivatives, tetrapod limbs are a successful example of how a typological approach can work (Shubin and Alberch, 1986). However, they are an elegant case, masterfully elaborated, therefore avoiding the conceptual difficulties with a typological approach aimed at reflecting underlying variational tendencies, which provide a test for homology in a biologically meaningful way (Wagner, 1999). In other cases, signals might often hide in noise so that even the most gifted can't detect them. To cue in the search for signal is a task for theoretical work. Paired fish fins have additional degrees of variation as they are mostly not elongated, like tetrapod limbs, but semicircular, allowing for more ways of a particular skeletal element to differ from its position in another group. I want to organize patterns of paired fin skeletons within monophyletic groups of fish over types according to Naef's (1935) criteria and then to organize these types likewise into a general type. As phylogenetic reconstruction is not the goal, more degrees of freedom exist to construct the simplest type, from which all observed patterns can be derived by the least number of natural transformational steps, steps which are modelled after e.g. a developmental sequence. Aided by a geometric and statistic analysis of transformations, I aim to solve conceptual and methodologival difficulties of the biological type concept by working it out on a concrete example.