Investigations of the micellar structure of fibre substances have given rise to two theories. The older theory (Meyer and Mark 1930; Mark 1932; Siefriz 1934; Meyer 1930; and Nageli 1928) considers the micelles as separate crystallites, between which lie the intermicellar spaces. The micelles consist of “Hauptvalenzketten” bound together along their length by homeopolar bonds and in the transverse direction by van der Waals’ forces, the intermicellar binding being also attributed to van der Waals’ forces. The original model suggested in work published by K. H. Meyer (1930), for cellulose, depicts the micelles arranged like bricks in a wall (fig. 1), and doubtless this is the simplest explanation of the X-ray results. But it is difficult to understand how such an arrangement can give a micellar structure its peculiar mechanical properties, and further how it is possible, when both inter- and intramicellar cohesion are attributed to the same type of force, to cause by swelling experiments an enlargement of the intermicellar spaces, while the “Hauptvalenzketten” remain unaffected. An alternative theory has been put forward by O. Gerngross, K. Herrmann and W. Abitz (1930), W. T. Astbury (1933), A. Frey-Wyssling (1936) and E. Guth and S. Rogowin (1936). These authors suppose that a given “Hauptvalenzkette” is not confined to a single crystalline region but may stretch through more such regions. In general, the arrangement of the neighbouring chains will be truly lattice-like, but a chain may lie at too great a distance from its neighbours or not lie exactly parallel to them, so that the structure as a whole will show statistically distributed spaces. In fig. 2 ordered crystalline regions may be distinguished (drawn in thick line), but their significance is physically different from that of the crystallites of the Meyer model. They are not self-contained units; the whole system is linked together due to the “Hauptvalenzketten” extending beyond a single micelle. Astbury considers that in a substance of high molecular weight of a type capable of swelling that part which produces the X-ray spectrum is the concentration centre of a complicated network of thread-like molecules. He draws an analogy between micellar structure and the secondary structure of Zwicky. He suggests that it is possible that micellar systems, which are characterized by a mixture of perfection and imperfection, are the counterpart in compounds of high molecular weight of the well-known mosaic structure of the more familiar crystals. Frey-Wyssling is of the opinion that the micelles, growing together, enclose lens-shaped spaces running parallel to the fibre axis. Between these intermicellar spaces are small rod-shaped regions of undistorted lattice, which are the so-called micelles of the earlier work (fig. 3). In this figure, which gives a pictorial representation of Frey’s theory, the statistically distributed hollow spaces are shown black; some of these are enclosed in undistorted crystalline regions. A lamellar structure consisting of superimposed monomolecular layers suggested by O. L. Sponsler and W. H. Dore (1930) has been shown to be untenable from the work on double refraction by Baas-Becking and Galliher (1931).