iron oxide is possible, particularly in the hydrogen clay Padegaon-B which has a comparatively large percentage of iron, and would account for the observation that both the Si02/AlaOs and the SiOz/RzOr ratios of this hydrogen clay differ materially from the corresponding ratios calculated for either montmorillonite or beidellite.* In the derivation of this equation it has been assumed that (i) Lamb's equation for viscous flow holds good, (ii) "end effect" is negligible, and (iii) the velocity of diffusion can be neglected.