Kelvin's discovery of Taylor columns

In 1868 Kelvin observed that a sphere moving on the axis of uniformly rotating water takes with it a column of liquid as if this were a rigid mass. This observation anticipates by thirty years Hough's analytical demonstration that steady slow motion in a rotating fluid is two-dimensional, and by fifty years Taylor's perception and experimental demonstration of the consequences thereof. This note discusses Kelvin's unpublished observation in the context of his lifelong interest on rotatory motion and later investigations on the subject by Hough, Proudman, Taylor, and others.