The hysteresis which normally appears in the isotherms of the silica gel-water system has been attributed by Patrick to the presence of permanent gases in the system. Only one case has been found in which a reversible isotherm has been recorded in the silica gel-water system. For alcohol and benzene only one case of hysteresis has been reported. These results seem to be independent of the presence or absence of air or other gases.The sorption of water, benzene and ethyl alcohol has been examined using the sorption balance. A hysteresis loop appears for water only. This hysteresis may not be eliminated by special methods of evacuation and must be considered as a real effect. The isotherms of alcohol and benzene, on the other hand, are completely reversible. It has been shown that the dimensions of the hysteresis occurring in the water system may be affected by the manner of addition of vapor to the apparatus. Only when the vapor pressures remain reasonably constant during sorption are the dimensions of the effect evident. If very large pressure changes take place the hysteresis may disappear.A comparison of the isotherms for water showing hysteresis, and those of the sulphur dioxide system (upon which the original suggestion as to the cause of hysteresis was based) show that there is not necessarily any relation between the two.