NATURE of the phenomena. It is at this point that thermodynamical relations become important, since the mechanism of the processes can then be neglected ; but with the obvious drawback that these relations throw very little light on the mechanism, and are therefore more satisfying to the physicist than to the chemist.