Sometimes you want to find the electric field of a loong cylinder carrying for instance a uniform volume charge. Then for easy calculations you approximate the cylinder with an infinitely long one and then put a gaussian surface over the entire thing. Symmetry then dictates that the field point in the same direction as the area elements of the cylinder and then you can find E. However, the ends of the cylinder are always accounted for in a way that I don't like. Because my solutions manual simply says that E is perpendicular to the area elements here. Why is that? Surely I can see that for an infinite cylinder at each point will have a field pointing radially out, but as soon as you consider what's happening at the end you are more or less actually assuming that the cylinder is finite. Can someone explain this ambiguity and why the procedure works even though the steps are pointless (in my opinion)?