Around the turn of the century, tonofibrils and contractile myofibrils were observed within the same cells. These findings have been largely forgotten. To clarify the topical relations of these proteins in epithelial cells, duplicate sections of methacarn-fixed human and canine tissues were treated with the tannic acid-phosphomolybdic acid (TP)-Levanol Fast Cyanine 5RN reaction for myosins and the PAP technic for prekeratin, respectively. In bronchi, lingual and sweat glands, liver and pancreas, myosin was confined to the terminal bar-terminal web system, including pericanalicular layers. Prekeratin occurred throughout the epithelium of bronchi and ducts; secretory cells showed little or no reaction. Observations on myosin in kidney confirmed data by Harper et al. (1970). The PAP technic colored transitional epithelium and collecting tubules intensely; convoluted tubules did not react. Staining of segments of Henle's loops varied from case to case. Both reactions colored thymic epithelial cells. In myoid cells of Hassall's corpuscles myosin was gradually replaced by prekeratin and keratin. Basal cells of epididymis reacted strongly with the PAP technic, but did not contain myosin. Prekeratin is apparently identical with epidermin, whose composition and structure were well known in the 1950's. Epidermin undergoes chemical changes as cells move from the stratum basale to the stratum corneum. According to DAKO, the antibodies used in this study were prepared with prekeratin extracted from stratum corneum. Data in the literature and observations in this investigation indicate that some samples of antibodies do not react with all tonofilaments.(ABSTRACT TRUNCATED AT 250 WORDS)