Abstract

We study a model proposed by Liboff (Found. Phys. Lett. 10:89, 1997) to violate the second law of thermodynamics. Discs are moving without friction in three connected channels inclined by π/3 with respect to each other. Based on the geometry considerations, it was argued that eventually all the discs end up in the middle channel regardless of their initial positions. This would mean a decrease of the entropy of the system and violation of the second law. We argue that no such anomalous behavior occurs in the system