Ashworth and Davies (1977) have recently discussed space-time geodesics and spatial geodesics on a rotating disk. By hypothesis, timelike geodesics in space-time represent possible paths of free particles, null geodesics in spacetime represent possible paths of photons, and spatial geodesics are defined as paths along which the distance between any two fixed points is a minimum. A special feature of their work is that the geodesic solutions are expressed in terms of two different coordinate systems Sl(r l , 01, zl , tl) and $2(r2, 02, z2, t2), each of which is associated with the rotating disk. According to Ashworth and Davies, the two solutions 01 = 01(rl) and 02 = 02(r2), for each type of geodesic considered, are fully equivalent, the distinction between them being related to the different types of measurement which observers on the disk would be expected to make. However, it will be shown in this note that 01, unlike 02, does not have the status of a coordinate in the normal sense (one of four variables which serve to label events in space-time) and that the solutions quoted are therefore not equivalent. We argue nevertheless that 01 does possess some significance, i.e., it is a plane polar coordinate in a particular type of representation (on a Euclidean plane) of geodesic paths on the disk. Consider a disk which rotates with constant angular velocity a, with respect to an inertial frame, and assume space-time is flat. Using the notation of Ashworth and Davies (except that we suppress the z coordinate throughout), let (r, 0, t) and (r2, 02, t2) be coordinates in the inertial and disk frames, respectively, related by the Galilean transformation