Some readers of the January issue of the Journal may have noticed rather strongly discrepant statements on the subject of self-contained detection devices in the several papers on collision avoidance. For example, on page 43 Gaudillère suggests either constant bearing or constant closing speed as an indicator of impending collision between aircraft; and on pages 82–87 Whitfield and Cade suggest that ships may determine transverse velocity (and hence may identify a collision course) by using a differentiated doppler radar return. But on pages 25–26 the author of this note states that the derivatives of range and azimuth required for this use are in fact too small for successful physical measurement.