In the original Hele Shaw problem n = 2 and ,U = 6, the Dirac measure at the origin. A couple of, essentially equivalent, weak formulations of this problem have been given in [4, 8, 91 (and in [3] f or a variant of the above problem) in the case n = 2. It has been proved in [4, 81 that a classical solution (this notion been made precise in two different ways in [4, 81) also is a weak solution, and in [4, 8, 91 (and [3]) that a weak solution exists and is unique for the time interval [0, ~0) for arbitrary given initial domains Do. It has also been proved [4, 81 that, at least under some extra hypotheses which probably are fulfilled for almost all f > 0 in nonpathological cases, aD, consists of analytic curves when t > 0 if {D,: t 2 0) is a weak solution. All the above results except the last one generalize without much trouble to arbitrary n B 2 (as noted also in [9, Section 31). Some variant of the last result probably also holds for n > 2, with “analytic curve” replaced by “real analytic hypersurface”. No results seem to have been published on the existence or uniqueness of weak solutions for time intervals of the kind [-T, 0] (T> 0) with Do = D given, i.e. backward solutions. The above mentioned results on the analyticity of dD, show (or indicate) that a backward solution