It is demonstrated that fullerenes, prepared via the standard method (an arc between graphite electrodes in a partial pressure of helium), on heating to high temperatures release ^4He and ^3He. The amount corresponds to one ^4He for every 880,000 fullerene molecules. The ^3He/^4He isotopic ratio is that of tank helium rather than that of atmospheric helium. These results convincingly show that the helium is inside and that there is no exchange with the atmosphere. The amount found corresponds with a prediction from a simple model based on the expected volume of the cavity. In addition, the temperature dependence for the release of helium implies a barrier about 80 kilocalories per mole. This is much lower than the barrier expected from theory for helium passing through one of the rings in the intact structure. A mechanism involving reversibly breaking one or more bonds to temporarily open a "window" in the cage is proposed. A predicted consequence of this mechanism is the incorporation of other gases while the "window" is open. This was demonstrated through the incorporation of ^3He and neon by heating fullerene in their presence.