For Aristotle, living beings are complex composites of matter and form, where form is to be understood functionally, and not merely as shape, as a specific combination of soul-capacities that characterizes the kind of living being in question. It is a commonplace in Aristotelian scholarship that both these forms of living beings and the enmattered animal species to which they give rise are ‘fixed’. Forms are ‘fixed’ in the sense that they – without being eternal themselves – are replicated eternally through sexual reproduction from father to offspring: the offspring receives a ‘potential for form’ that is formally (but not numerically) identical to the form its father possesses in actuality. Since this potential for form is for the most part – i.e., if nothing greater impedes – realized in the same way, and since the cosmos is eternal in time, Aristotle believes that there is a continuous generation – and corruption – of organisms that are one in form and that instantiate the same range of animal species. Within these confines, there is no room for a transformation of species.
In seeming contradiction with this fixity of species, however, Aristotle’s biological works often stress the flexibility of nature during embryogenesis and the later development of animals. Aristotle typically describes the process of the actualization of a potential for form into an actual animal in terms of ‘formal natures’ ‘making’ or ‘building’ the animal. This formal nature instantiates the animal’s efficient, formal, and final cause (which is identical to its soul), and following the craft-model already employed by Plato, Aristotle personifies this internal principle as acting for the sake of something, while following a certain logos or ‘guideline’ for building. Given the unchangeability of animal species, one would expect these ‘guidelines’, or ‘definitions of the animal’s substantial being,’ to be rather strict and precise, but instead Aristotle suggests that they are underdetermined: rather than being specified by the animal’s logos, Aristotle claims that it is somehow ‘up to’ formal natures to determine which parts to produce for the performance of a certain function, how many of those parts it should make, and where in the animal body it should place them. Evidently, even though the starting point (i.e., the potential for form) and end point (i.e., a fully developed animal of a given species) of animal generation are ‘fixed’ and ‘permanent’ features of the world, the individual actions of the formal natures through which this process is carried out are not similarly predetermined.
In a paper I will be presenting at SUNY Brockport this week, I argue that Aristotle’s tendency to think of the activities of formal natures in the production of animals as if performed by tinkering craftsmen shows that for him, the structure and functionality animals have is as much a result of the teleology embedded in their forms as it is of the goal-directed actions performed by their formal nature in realizing those forms. This indicates that at least at the level of embryogenesis, Aristotle does allow for something like a transformation of species and that his view of development is ‘epigenetic’ in the true sense of the term. That is, the ‘guidelines’ for building with which nature works do not predetermine in advance all the changes that need to be made, but rather leave room for nature to make adjustments to the animal design ‘on the fly,’ as the development proceeds. For his theory of natural teleology this means that the underlying physiology must be rather complex and flexible: the potentials for form that guide the realization of those species are dynamic, rather than blind, automated self-regulating principles, as they are shown to be capable of responding to emerging problems and possibilities.
 For the idea that Aristotle’s understanding of natural kinds is not as such ‘anti-evolutionary’, see already Balme (1972), 97-98; Granger (1987), 110-116; Henry (2006), 451-455; and Lennox (2001a), 160-162.