Bertrand Russell has found the paradox that bears his own name in the spring of 1901 and offered a version of the so-called “simple” theory of types as measures against it in an appendix to The Principles of Mathematics (1903). This theory was devised to deal with the class-version of that paradox. But he formulated it also in terms of “predicates” and the type theory has no effect to this formulation. In this paper, I shall show that Russell offered measures also against the “predicate”-version of that paradox in the Principles and it is very interesting in a sense that it enables to avoid the paradox without forbidding self-predications of predicates in general.