750 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS also contain the necessary information for determining whether whiskers or pelage hairs will develop (4). The adult whisker dermal papilla also has a profound effect on adult epidermal cells, locally promoting the development of an increased and highly mitotically active germinal population, including suprabasal mitoses, associated with a columnar basal layer. This seems to be the prelude to many epithelial developmental situations (e.g. Wessells (45)). Indeed the papilla may then go on to induce the formation of hair follicles, regardless of whether the epidermis contained epidermal follicular elements or had no previous history of being follicular epidermis. None of these follicles was typical of the site of origin of the epidermis or, indeed, of the implant site. Rather they more closely resembled small whisker follicles, although lacking the usual whisker dermal features. Two further observations merit consideration. Not all papilla/epidermal associations formed functioning hair follicles. Even where they were formed development proceeded slowly. Straile (46) believes that the neuro- vascular contribution to hair follicles, as inferred from histological studies of the tylotrich follicle, may be vitally involved in hair growth. It is there- fore conceivable that the associated development, or not, of this ill-under- stood component was a limiting factor in the present induction studies. It may be relevant that while in vitro cultures of embryonic skin develop follicles and grow a short length of hair, continued hair growth has not been achieved (47). Of great interest was the influence of whisker dermal papilla on oral epithelium. Available evidence has suggested that the specific pattern of keratinization of this epithelium, like feather epidermis, is innately deter- mined so that it is genetically incapable of being modulated by contrary dermal influence. However, instances of attempts, some more or less success- ful, at follicle formation in the presence of papilla cells are highly indicative that provided the dermal influence is strong enough oral epithelium is also capable of modulatory behaviour. Just as intriguing, and reinforcing the above interpretation, was the development of stellate reticula, histologically highly reminiscent of the enamel organ of developing teeth. Here again an unblocked but previously repressed genetic potential seems to have re- vealed by some stimulatory effect, presumably non-specific, of the papilla cells. The derreal papilla, the hair cycle and the hair The continuing presence of the dermal papilla adjacent to the epidermal

DERMAL PAPILLA AND THE DEVELOPMENT AND GROWTH OF HAIR 751 component is essential for hair growth. On removal of the whisker papilla from the follicle, hair growth ceased and was only renewed when a papilla spontaneously regenerated or, failing this, when one was implanted. Induction of hair growth where the bulbar region was removed and a dermal papilla implanted has several implications. It confirmed the deduction of Montagna and Chase (48), from the results of exposing scalp to X-rays, that the source of the epidermal germ for the next hair growth need not be con- sidered in terms of a specialized population of cells surviving from the previous matrix. Extension of the follicle associated with the initiation and growth of hair, reminiscent of proanagen in pelage follicles, is highly suggestive that the dermal papilla is primarily involved in the naturally occurring initiation and development of proanagen. Furthermore, regard- less of the obvious anatomical differences which initially exist, the morpho- genetic changes and presumed interactions which occur at proanagen appear, in essence, to be identical to those which occur in ontogeny following establishment of the follicle plug. The fact that the adult dermal papilla can induce follicle development gives added weight to this supposition. It is also not unreasonable to suspect papilla complicity in the termin- ation of anagen, perhaps even exercising some control over the orderly regression of the follicle during catagen (2). Evidence is flimsy on this point. While the matrix disappeared, hair growth ceased and the hair was lost from the follicle after removal of the dermal papilla, it could equally be argued that these sequelae were a wounding response. Nevertheless simple cessation of cell division in the matrix, following cancer drug therapy, did not of itself precipitate the follicle into catagen (49). While systemic factors have been shown to influence the growth of hair, their effect seems to be one of modification of an intrinsically determined behaviour acquired by each follicle during its development (50-53). Current results strongly suggest that it is the dermal component of hair follicles which carries this behavioural information. However explanations for the cyclic nature of hair growth include the suggestion that inhibitors accum- ulate somehow and somewhere in the epidermal component and, on attain- ment of a threshold level, cause cessation of mitosis in the matrix (29, 54). A leaching out or loss of inhibitor would then allow growth to be resumed. While such factors may exist and be of importance the stimulatory influence of the dermal papilla should certainly not be forgotten. Indeed, even apart from morphological changes, the cyclic activities which occur within the papilla itself, including the known ones of presence of acid mucopolysaccharides and alkaline phosphatases during anagen and