1see it can be proved using simple arguments..First of all..cylindrical symmetry needn't neccessarily imply that potential is same over a whole complete cylindrical surface... it may be of a kind like that of a p-orbital..(*p-orbital has cylindr. symmtry..*)....so acc to the formula of potential..which is kp cosθ/r2 at a dist r frm d center of d dipole..a circle could be obtained having the same potential... that's y a CYLINDRICAL SYMMETRY!