Two different crystalline models of cellulose have received attention from crystallographers: the "straight chain" model of Meyer and Misch and the "bent chain" model of P. H. Hermans. We have examined these models and have given a mathematical interpretation of each in terms of the conformational angles [Formula: see text] and ψ and the glycosidic angle τ. For a given geometry of the glucose residue, the "straight chain" corresponds to a unique value of τ; for a value of τ greater than the unique value, two "bent chains" are possible, one allowing an intramolecular hydrogen bond (O-3 … O-5′) and the other not. It is suggested that the former only be referred to as the Hermans conformation.