This is a laboratory zonal-jet study using a rotating water tank. The bottom topography has a tent-shaped radial cross section designed to generate two critical latitudes, i.e. two positions where βe, the radial gradient of the potential vorticity (PV), changes sign. This configuration is motivated by observations indicating Jupiter and Saturn have not only multiple zonal jets, but multiple stable critical latitudes. It is known that “supersonic” critical latitudes (with respect to Rossby waves) are stable, whereas “subsonic” critical latitudes are posited to be unstable. Because Rossby waves are uni-directional, “supersonic” critical latitudes come in two varieties: Rossby Mach number M_R > 1 and M_R < 0, where the latter holds when the waves are directed downstream. Experiments focus on: i) how do zonal jets emerge from localized forcing in a system with alternating PV gradients? and ii) what differences are there between the evolution of various types of critical latitudes? The water is forced by mass injection along one radius. Laboratory altimetry provides accurate, unobtrusive records of the circulations that reveal the emergence of counter-propagating β-plumes (Rossby-wave envelopes), which expand into tank-encircling zonal jets. The tank’s negative βe annulus is characterized by M_R ~ 1, which is the condition surmised for Jupiter and Saturn. The weaker critical latitude (in terms of jumps in the PV-gradient) adjusts its position by ~4% of the tank radius and maintains M_R ~ 1. In contrast, the stronger one vacillates while maintaining |M_R| << 1, and may be relevant to steep oceanic seamounts.