A theory for the viscosity and structure of glass as a function of temperature and composition had been advanced, proposing that at a few critical temperatures silicate glasses undergo structural transformations. Between these transformation temperatures, the activation energy and lattice coefficient are constant. A series of special glasses were melted and viscosity was measured to test this theory. The results indicate that the viscosity‐temperature behavior of glass does not conform to the previously proposed model. The activation energy appears to vary continuously over the entire temperature range studied. However, a critical temperature does exist at which the temperature coefficient of the activation energy changes. This discovery represents a more subtle change than that previously envisaged. Copyright © 1979, Wiley Blackwell. All rights reserved