Numerical experiments are carried out by using the Yamagata-Flierl equation for eddies of a scale intermediate between the Rossby deformation radius and the beta-scale. These experiments are performed for two types of zonal flows, one has a linear shear and the other is sinusoidal velocity profile along meridian. Through these experiments, we tried to investigate the motion and the life time of the intermediate scale eddies in the cases of linear shear zonal flow, and attempted to compare the results with the development of the South Tropical Disturbance (STrD) observed in 1979-81 on Jupiter for the cases of sinusoidal velocity profile.From the experiments for the case of linear shear flow, we found the following facts.(1) If the eddy and zonal flow have the same sign of vorticity, the eddy can survive stably. If the eddy and zonal flow have the opposite sign of vorticity, the eddy disappears or is deformed remarkably. (2) In the case of uniform zonal flow, a cyclonic eddy can survive stably in a high speed westward flow, whereas an anticyclonic eddy can survive in a low speed westward flow. When the westward zonal flow has the non-dispersive wave velocity, both cyclonic and anticyclonic eddies can survive stably. (3) In a low speed westward flow, a cyclonic eddy shifts poleward and an anticyclonic one equatorward. In a high speed westward flow, a cyclonic eddy shifts equatorward and an anticyclonic one poleward. (4) Anticyclonic eddies move westward with higher speed than that of non-dispersive wave. On the other hand, many of cyclonic eddies move westward with lower speed. In one of our experiments for the sinusoidal velocity profile, a cyclonic eddy, whose radius is comparable to the latitudinal width of the westward jet, can go poleward through this jet, and is elongated in northeast and southwest direction in the northern hemisphere before it disappears. The STrD seems to be this kind of cyclonic eddy, because the STrD also had cyclonic vorticity and developed poleward, and because the STrD also was elongated in the same direction as our results. The life time and longitudinal motion of this eddy is not consistent with those of the STrD, but through collisions with small vortices the STrD may get a long life time.