The mean obliquity of Mars may have increased over geologic time due to climatic changes associated with obliquity oscillations; this mechanism is dubbed ''climate friction.'' Given here is the mathematical theory of climate friction. Also, a 10 m.y. numerical integration of the equations for a hypothetically large amount of climate friction is performed, for the cases of the obliquity oscillations being (1) a single sinusoid and (2) a sum of three sinusoids. Theory and numerics agree in both cases within about 12% on the size of the secular increase in obliquity, with most of the discrepancy coming from the theory giving obliquity amplitudes ~10% too large. Further, one possible mechanism of climate friction is investigated here: ''postglacial rebound'' on Mars. The idea is that giant polar caps form when the obliquity is low. These caps slowly squeeze out an equatorial bulge. When the obliquity is high, the caps disappear, but the bulge takes some time to collapse, due to mantle viscosity. Thus the equatorial bulge oscillates but is out of phase with the obliquity oscillations. Like tidal friction, this causes a secular increase in the average obliquity. Using the preferred climate model of Fran¿ois et al. (1990), the total secular increase in the obliquity over the age of the solar system can amount to about 10¿ if Mars' mantle has an Earthlike effective viscosity of 3¿1021 Pa s (3¿1022 P). The secular change is negligible for effective viscosities which differ by more than a factor of 10 from this value. Last, how much climate friction tipped the Earth is presently unknown.