“…Certain sugar epoxides may be conformationally stable, even though they contain no 4,6-O-benzylidene group or 1 ,6-anhydro ring. Thus methyl 2,3-anhydro-4,6-ditosyl-a-~-alloside probably exists almost entirely in that conformation in which the bulky tosyloxy groups are equatorial and, if this is so, the attack of lithium aluminum hydride a t Cz (22,23) is another example of diaxial opening: H Anhydrides of monocyclic pyranoses, whose conformations are not stabilized, will normally exist as an equilibrium mixture of two conformations, corresponding to the two half-chair conformations of a six-membered ring. Hence, even if ring opening is always diaxial, two products are possible, as shown in figure 6 for 2,3-anhydrides, and each of these products can exist in either or both of two conformations.…”