If we accept the traditional definition of knowledge according to which S knows that P provided that (i) P is true, (ii) S accepts P and (iii) S is justified in believing that P, the Gettier's counter-examples will prompt us to look for a fourth condition to complete the analyses of knowledge. Before we determine whether such a fourth condition is, indeed, needed and, if so, how it is to be formulated, we may do well to examine with some care the situations which, according to Gettier, are counter to the traditional definition of knowledge. There are two reasons why this course of action is to be recommended. First, it may enable us to expose the presuppositions, other than those involved in the traditional definition of knowledge itself, which Gettier appeals to in order to construct his counter-examples. Second, this may provide us with an insight into exactly what it is that is supposed to be wrong with that definition.