I have a larger problem involving divergences and curls, but the correct answer requires ∇°B (divergence of B) = 0. I understand the proof of this in Griffiths, but the definition of divergence in cylindrical coordinates is: After using the product rule to split the first term, we get the divergence of B is B_rho / rho + 0 + 0 + 0, or simply Div(B)=B_rho / rho; however, this clearly contradicts what we know about the divergence of B being zero. Can someone please clarify this for me, I've been stuck at it for hours. Thanks. P.s. First post!