The similarity between prime numbers and irreducible polynomials has been a dominant theme in the development of number theory and algebraic geometry. There are certain conjectures indicating that the connection goes well beyond analogy. For example, there is a famous conjecture of Buniakowski formulated in 1854 (see Lang [3, p. 323]), independently reformulated by Schinzel, to the effect that any irreducible polynomial f (x) in Z [x] such that the set of values f (Z + ) has no common divisor larger than 1 represents prime numbers infinitely often. In this instance, the theme is to produce prime numbers from irreducible polynomials. This conjecture is still one of the major unsolved problems in number theory when the degree of f is greater than one. When f is linear, the conjecture is true, of course, and follows from Dirichlet's theorem on primes in arithmetic progressions.It is not difficult to see that the converse of the Buniakowski conjecture is true; namely, if a polynomial represents prime numbers infinitely often, then it is an irreducible polynomial. To see this, let us try to factorof positive degree. The fact that f (x) takes prime values infinitely often implies that either g(x) or h(x) takes the value ±1 infinitely often. This is a contradiction, for a polynomial of positive degree can take a fixed value only finitely often.There is a stronger converse to Buniakowski's conjecture that is easily derived (see . We will give a proof of this fact that is conceptually simpler than the one in [1], as well as study the analogue of this question for function fields over finite fields. More precisely, let F q denote the finite field of q elements, where q is a prime power. Fix a polynomialThe analog of Cohn's theorem to be proved in what follows is thatis irreducible in F q [t, x]. The proof in the function field case is much simpler and is motivated by the following elementary result, which can be viewed as somewhat of a strong converse to the conjecture of Buniakowski.