Our notation and terminology basically follows that found in [2], except with regard to notation for unions and intersections; in a few instances we cite other references for special terms.The following two propositions are established by a fairly straightforward moveable-markers technique; either proof is only a minor variation on the other.
\Jr, and (iii) (a-r) -y» is immune f or all i.It was shown by Yates, in [5] (in answer to a question of Dekker and Myhill), that there are basic retracing functions, some of them retracing unique infinite sets, which do not retrace any infinite recur* sive set. In each of Yates' examples, all of the sets retraced by such functions have nonimmune complements. The above propositions demonstrate the existence of examples in which an infinite set a is retraced by a basic function and a has immune complement. In any example of this latter type, the function in question must retrace a unique infinite set, which, of course, cannot be recursive.We remark that all of the sets y* obtained by us in proving Propositions A and B are, owing to the nature of the proofs, hyperimmune (for the notion of hyper immunity, see, e.g., [5]). This is closely related to the following general assertion : 1 We are indebted to Paul Young for a conversation which took place in August, 1963. At that time he made a suggestion which has proved to be susceptible of elaboration into proofs of Propositions A and B.523