The interconversion of sugars in nature has for long been a source of interest and speculation. Many attempts were made to postulate a mechanism for the smooth and facile transformations but until 1933 the theories were unsupported by any experimental evidence. The first direct conversion of one aldose to another by means of changes involving a Walden inversion was achieved by Irvine and Hynd who obtained either d-glucose or d-mannose from d-glucosamine, according to the experimental procedure. It was obvious that in one case Walden inversion took place on C₂, but an interpretation of reaction was complicated by the removal of the nitrogen atom. Later, by the action of ammonia on methylglucoside-2-bromohydrin, Fischer et al. obtained “methylepiglucosamine” which was later shewn to be 2-amino-methylaltroside. The research which is now to be described was undertaken with a view to extending to the pentose series the application of the general principles which had been worked out for the hexoses. Although the research was incomplete, in the sense that no interconversion was achieved, many interesting facts emerged in the course of the work and some important derivatives of arabinose were prepared for the first time.