The calculation of the inviscid, incompressible flow past non-lifting three-dimensional bodies has been greatly facilitated by A. M. O. Smith and his collaborators, using distributions of elementary singularities over the surface of the body. There is considerable current interest in the extension of this work to lifting bodies, involving the representation of the real viscous wake by a vortex sheet of zero thickness in the inviscid model of the flow. In view of this interest, it is appropriate to consider the way in which such a vortex sheet behaves at the trailing edge of a wing with non-zero trailing-edge angle. Is the sheet tangential to the upper surface of the wing, to the lower surface, or to neither?