“…Assuming potential flow, in which the velocity potential φ has been made dimensionless with U h, x and z with h and t with h/U , and in the absence of vortex sheets, we have to solve where η may well be multi-valued, with The initial conditions are appropriate to the jet collision problem under discussion. Hence, over this timescale, we have an 'unsteady Helmholtz flow' for which, as discussed in King & Needham (1994) even the small time solution is very complicated. Nonetheless, we anticipate that this solution will describe the growth of a vertical column of liquid centred near the initial impact point and that, as t → ∞ with F = ∞, the flow will tend to a steady Helmholtz flow in a frame moving with the spout.…”