THE transformation of crystalline native cellulose (cellulose-I) into the hydrate modification (cellulose-II) by the action of caustic soda solutions is known to proceed by the formation of intermediate crystalline soda-cellulose complexes1. In a previous publication2 from this Laboratory, it was reported that the complete native-hydrate transformation is impossible in raw jute, the presence of appreciable quantities of cellulose-I being always apparent from the X-ray photographs. In ramie the conditions for a partial transformation of this kind are much more critical, and appear to be such that only part of the cellulose has reacted with the alkali. In order to determine whether the effect in jute can also be explained in this way, we have studied the ‘soda-cellulose’ obtained by treating raw jute without tension in caustic soda solutions of various concentrations, and find that in general the native modification is absent for concentrations greater than about 10 per cent (w/w), but that it reappears when the alkali is removed by washing in water. In jute, therefore, incomplete transformation to cellulose-II is not due to the failure of the alkali to penetrate into the cellulose crystallites, but to at least partial reversibility of the native → soda-cellulose reaction.