Since degf>degg, when x is sufficiently large we have ∣g(x)/f(x)∣<1. That is, there exists a real number R such that for all ∣x∣>R, we have ∣g(x)/f(x)∣<1. For all such x and all primes p, we have
∣pf(x)+g(x)∣≥∣f(x)∣(p−∣f(x)∣∣g(x)∣)>0.
Therefore all real roots of the polynomial pf+g lie in [−R,R].
Let f(x)=anxn+an−1xn−1+⋯+a0 and g(x)=bmxm+bm−1xm−1+⋯+b0, where n>m, an=0 and bm=0. Replacing f(x) and g(x) by an−1f(x/an) and am−1g(x/am) respectively, we reduce the problem to the case an=1, in which the leading coefficient of pf+g is p. If r=u/v, (u,v)=1 and v>0, is a rational root of pf+g, then v is either 1 or p.
Suppose there are infinitely many instances with v=1. If v=1 then ∣u∣≤R, so there are only finitely many integers u. Hence there exist distinct primes p and q such that we get the same value of u. Then the polynomials pf+g and qf+g share a common root, from which we obtain f(u)=g(u)=0. So in this case f and g have a common integer root.
Suppose there are infinitely many instances with v=p. Comparing the powers of p in the denominators of pf(u/p) and g(u/p), we take m=n−1, so that pf(u/p)+g(u/p)=0 simplifies to
(un+an−1pun−1+⋯+a0pn)+(bn−1un−1+bn−2pun−2+⋯+b0pn−1)=0.
This shows that un+bn−1un−1 is divisible by p, and since (u,p)=1, we get u+bn−1=pk, where k is some integer. Also, all roots of pf+g lie in [−R,R], so
p∣pk−bn−1∣=p∣u∣<R,∣k∣<R+p∣bn−1∣<R+∣bn−1∣.
Hence there are only finitely many possible values of the integer k. So there exists an integer k such that for infinitely many primes p, ppk−bn−1=k−pbn−1 is a root of pf+g. For these primes, we have
f(k−pbn−1)+p1g(k−pbn−1)=0.
Hence
f(k−bn−1x)+xg(k−bn−1x)=0(1)
has infinitely many solutions of the form x=1/p. Since the left-hand side of equation (1) is a polynomial, equation (1) holds for all real numbers x. Substituting x=0 into equation (1), we obtain f(k)=0. Thus the integer k is a root of f.