DIFFICULTIES encountered by Stoner1,2 and Guggenheim3–5 in their attempts to apply thermodynamics to systems which contain magnetizable matter have made these writers aware of important defects in the usual development of the notion of energy in the theory of magnetism and in the analogous theory of electrostatics. The difficulties occur at the outset, when it is desired to write a general equation to express the first law of thermodynamics; such an equation must allow for density change in the material substance of the system and must include a term to allow for variation of magnetic energy. It is then essential to express this latter variation in a form which is valid when the material bodies in the system are deformable. Stoner and Guggenheim have not done this, and, besides, their discussion excludes systems in which hysteresis can occur. In the remarks which follow I refer to the analogous electrostatic problem.