In 1900, Otto Lehmann observed that the texture of a cholesteric droplet heated from below can rotate continuously (Ann. Phys. (Leipzig), 2 (1900) 649). This observation (which has never been reproduced, to our knowledge) was explained in 1968 by Leslie (Proc. R. Soc. London, Ser. A, 307 (1968) 359) from symmetry arguments accounting for the chirality of the material. In 1982, Éber and Jánossy showed experimentally that a similar thermomechanical effect also exists in a compensated cholesteric (in which the helix is completely unwound). This result was immediately questioned by Pleiner and Brand who claimed that only the symmetry of the phase (and not that of the molecule) determines the structure of the macroscopic constitutive equations (Mol. Cryst. Liq. Cryst. Lett., 5 (1987) 61). According to them, the Lehmann effect should necessarily vanish at the compensation temperature. In order to understand the correct interpretation, we conducted very carefully the experiment in two complementary geometries. Our results agree with those of Éber and Jánossy, confirming the predominance of microscopic symmetries over macroscopic ones.