READERS' FORUM 1023and cases 1-3 in Ref. 2. One can speculate that for the case of supersonic flow over a blunt fin, 4 at least in the region ahead of the blunt fin, the flow is strongly dominated by the geometry and other inviscid mechanisms. The strength of the primary vortex was closely predicted; hence so were the main features. For the case of supersonic flow over a swept corner, 2 the mechanism for generating the primary vortex was probably not correctly simulated; hence so were not the strength of the vortex and its associated'features.