I've seen something like it where the asymptotes are some other two intersecting lines ax + by = c in the plane. (It's probably the result of applying a linear mapping to the curve (e^x-1)(e^y)-1 = 1). It was used in a relatively recent proof that zeta(2) = π^2/6, based on (I think) integrating to get the area bounded by such curves and straight lines. Sorry I don't have the specifics. —Dan ----- Right. You can get the numerical factor by finding the area under the curve. Nobody has seen this curve before? -----