Simple analytical solutions are presented for topographic waves in elliptical basins whose isobaths form a set of confocal ellipses. It is shown that the previous solution presented for this problem is incomplete. The complete solutions no longer agree so closely with observations in Swiss lakes and the Great Lakes. The class of profiles considered is extended to include flat-bottomed basins, allowing a more realistic estimate of the velocity distribution in such basins and suggesting a slightly closer agreement of predicted period with observations.