The model of a double auroral arc presented by Gouml;ran Marklund and his colleagues in their recent article appears crucially to disregard the fact that electrostatic equipotentials are closed surfaces. Figure 1 of the article depicts equipotentials as open-ended lines, and the text reinforces this concept by referring to acceleration by a U-shaped potential structure. The equipotentials of the model are, therefore, not closed even in two dimensions. If the equipotentials were to be closed, it would be obvious immediately that electrons, which in the model experience them as static, will-contrary to the claim made-gain no energy from traversing them. Work done on entering, and leaving the potential wells will necessarily exactly cancel.