Developmental anatomical data are insufficient to discuss plausible intermediates between an ancestral, scaled, reptilian skin and appendage-bearing, avian skin. We also review adult tissue replacement and ubiquitous mechanisms underlying skin morphogenesis. Combining developmental data sensu lato with consideration of necessary biological roles permits evaluation of major form/function trends in skin evolution. New data on feathers reveal retention of the sauropsid synapomorphy of vertical alteration of α- and β-keratogenesis. By identifying roles that were obligatorily maintained throughout evolution, we demonstrate constraints on hypothetical skin morphologies in preavian taxa. We analyze feather origins as a problem of emergence of complex form via modulations of morphogenesis. While existing data do not permit presentation of sequential, hypothetical, intermediates culminating in a plumage, the analysis: (1) implies that a protofeather and its follicle are most easily derived from isolated, flattened, elongate, reptilian scales; (2) explains diversification of feather morphs from a contour-like “basic” feather and the similarity between feather and hair follicles; and thus (3) reveals several developmental constraints on structures proposed as antecedent to avian feathers, whether hypothetical constructs or palaeontological interpretations. Although these conclusions do not depend on any previous scenario, they are consistent with Regal's (1975) model and the limited, fossil evidence, especially that of the “basal archosaur” Longisquama.