For the past 50 years there has been considerable debate over the sub-structure of the fully differentiated (oxidised) trichocyte keratin intermediate filament. Depending on the staining and preparative procedures employed, IF observed in transverse section in the transmission electron microscope have varied in appearance between that of a "ring" and a "ring-core" structure, corresponding to the so-called (8 + 0) and (7 +1) proto-filament arrangements. In a new analysis of the fine structure of the 1 run equatorial region of the X-ray diffraction pattern of quill we show that the observed pattern is consistent with the (8 + 0) model and we are also able to assign values to the various parameters. In contrast, we show that the observed X-ray pattern is inconsistent with a (7 +1) arrangement. Furthermore, in the (7 +1) model steric hindrance would be encountered between the core protofilament and those constituting the ring. The appearance of a central "core" in transverse TEM sections, previously attributed to a central protofilament, is explained in terms of portions of the apolar, disulfide-bonded head and/or tail domains of the trichocyte keratin IF molecules, including the conserved H subdomains, lying along the axis of the IF, thereby decreasing the efficacy of the reducing agents used prior to staining. The H1 subdomain, previously shown to be important in the assembly of epidermal IF molecules at the two-to four-molecule level, is likely to have a similar role for the trichocyte keratins and may form part of a central scaffold on which the molecules assemble into fully functional IF.