The experiments of Rivlin and Saunders and Gent and Rivlin are consistent with W having the form (33) in which C is a constant and f is a function of I2−3 only, which decreases monotonically as I2−3 increases. With this interpretation of the experiments, the term f (I2−3) represents the departure from the predictions of the kinetic theory. This conclusion is reinforced by the finding of Gent and Rivlin that W1 exhibits no hysteresis, while all of the hysteresis (i.e., departure from ideal reversibility) exhibited by the rubber resides in W2 (i.e., in the term f (I2−3)). This suggests that the model for a vulcanizate adopted in the kinetic theory must be modified by the introduction in the undeformed rubber network of crosslinks which are broken when the rubber is deformed and are re-formed, nearly reversibly, when the rubber returns to its undeformed configuration. However, so far no satisfactory mathematical development of this idea has been carried out; the construction of a convincing theory which accounts for the departure of W from the neo-Hookean form remains an unresolved problem.