SHORT COMMUNICATIONSBabinet's principle of optical reversibility. From this principle the scattering depends only on the presence of discontinuities, in this case of electron density, irrespective of the sign (high or low) of the densities. In polycrystalline solids it is not possible, therefore, to decide whether the parameters calculated must be attributed to the solid particles or to the interstices between them. Guinier & Fournet (1955) have pointed this out already for X-ray scattering at small angles. As a result, the question remained undecided for a long time, whether in cellulose the microcrystallites (so called 'elementary fibrils') or the voids (or other materials between them) are associated with the parameters calculated. In the case of cellulose this question has been raised by Statton (1956) and Caulfield (1971). Statton maintained that the voids (interstices) rather than the microcrystallites are the units in question and Caulfield even added that, because of the inherent ambiguity of interpretation, the method of scattering at small angles may have rather 'confused than solved the issue of the size of the elementary particles in cellulose'.One way to solve this problem is by impregnating the intercrystalline regions with substances of another (higher) electron density and studying the effect on the scattering. The author applied intercrystalline swelling and by studying the effects on the radial distribution of the scattering elements arrived at the conclusion that the microcrystallites are the units to which the parameters calculated belong. This method can only be applied, however, when interparticle interference occurs in the unswollen sample.The observations with high-resolution microscopy mentioned above directly confirm the conclusions from the swelling experiments and radial distribution, that the parameters calculated must be attributed to the microcrystallites and that the voids have other diameters. From the above analysis it appears that the assumptions made concerning the uniformity and orientation of the particles and the particle sizes determined with the X-ray scattering method are fully confirmed by electron microscopy, and that the latter complements the X-ray method in the identification of the scattering entities in the case of a polycrystalline solid. Cell Biol. 29, 181-197. HEYN, A. N. J. (1969 Appl. Cryst. (1975), 8, 342] claim that the nucleation of stable forms of certain substances may preclude the subsequent preparation of previously known metastable crystalline forms 'often even in laboratories many miles away'. However, the literature on these substances does not indicate that any authentic metastable forms have ceased to be obtainable, and we see no reason to believe that the effects of seeding are other than local and temporary.
References