The following question was asked: if $f$ is a germ of an analytic function, regular in $z$ and in $u$ in Weierstrass sense, is there a polynomial in $(z,u)$ with analytic coefficients, that divides $f$? The answer is negative, a counterexample is given.