IMO 1972 LL MON23

Origin: MON

Problem

Does there exist a 2n-digit number a2na2n−1 . . . a1 (for an arbitrary n) for which the following equality holds: a2n . . . a1 = (an . . . a1)2?