In previous papers on the micellar organization of the nerve fibre, attention has been directed almost entirely to the lipoid-containing sheaths. Polarized light and X-ray diffraction studies have emphasized the similarity of the architecture of these structures over a wide range of animal forms, from the highly myelinated nerves of vertebrates to the so-called non-medullated fibres of invertebrates (Schmitt, Bear and Clark 1935; see preceding paper for other references). To date little more than a beginning has been made with analysis of the ultrastructure of the axoplasm. This has been due largely to technical difficulties arising from the small size and weak birefringence of the axons so far studied, from the unfavourable position of the axis cylinder within more or less complex sheaths, and from the extreme sensitivity of this colloid to the action of reagents. Valentin (1861), Klebs (1865), v. Ebner (1882), and Friedlaender (1889), found no birefringence in the fresh axis cylinder, although Valentin mentioned that coagulation produced slight positivity. Ambronn (1890) described a positive birefringence common to all types of nerves and refers this to a “ground substance”. Though Apathy (1890) in early works concluded that neurofibrils are optically isotropic, in later studies (1897) he considered that the fibrils were uniaxially positive with respect to their length and demonstrated dichroism of the neurofibrils in gold preparations. Finally, Göthlin (1913) showed beyond doubt that the axis cylinder is weakly uniaxially positive in life,particularly at the regions of nodal constriction (vertebrate axons).