A criterion of equilibrium. Reversible processes

As we have seen, the only changes which can take place in an adiabatic system are those in which the entropy either increases or remains constant. The same applies, of course, if the system is completely isolated, that is, if it does no work, as well as being within an adiabatic enclosure, so that the internal energy is constant. Thus whenever a system could change from a state of lower to one of higher entropy, dthin an enclosure of constant volume and energy, it is possible, in a thermodynamic sense, for this change to occur. 

The trend towards equilibrium is shown diagrammatically in Fig. 9. It does not follow, of course, that when an increase in entropy is possible, the change will necessarily take place at an observable speed. This is shown by the example of a hydrogen-oxygen mixture and many other thermodynamically metastoble systems which may persist for centuries without visible change. Thermodynamics has nothing to say about the rates of processes. 

In 1 10 a path between two states A and of a system was defined as being reversible if the cycle A J5 - A could be completed without 

The reversible path must be one for which the internal forces of the system differ only infinitesimally from the external forces, and for which all heat transfers take place over temperature differences which are only infinitesimal. That this is the case may be seen by considering two examples.f 

Thermal equilibrium. If the otdy process which a body undergoes is the addition or removal of heat the effect of this addition or removal is to cause a definite change of state (since in this case dq dU) and thereby a definite change of entropy. The equation 

---