Focuses on generalizing the notion of variation in a set of numbers to the notion of variation in a set of probability distributions. The work collects known ways of comparing stochastic matrices, and then generalizes these, and establishes the relations of implication or equivalence among some.