Sime et al. (1984, 1985) have argued that two common characteristics of looplike coronal mass ejections are inconsistent with interpretation of such loops as shock waves moving through the corona. These are the cessation of lateral motions of the loop sides to form nearly radial, stationary bright columns or ''legs'' in the late stages of many mass ejections and the pushing aside or bending of pre-existing coronal features well outside the loop sides. Both are contrary to the notion of a propagating wave with a shock at its front and in disagreement with quantitative models of shock propagation in the corona. We report here Solar Maximum Mission (SMM) coronagraph observations of a looplike coronal mass ejection that occurred on July 6, 1980, and that was atypical in the two characteristics mentioned above. The sides of the loop moved laterally while they were visible and did not stop to form stationary ''legs.'' A group of raylike features initially outside the mass ejection loop were not pushed aside or bent until the laterally moving loop sides passed over them. We thus propose the July 6, 1980, mass ejection as a candidate for identification as a shock wave. Quantitative analysis of the motion of the bright loop is also consistent with such an identification. The loop top moves outward with an apparent radial speed of ~1000 km s-1, faster than all but one other mass ejection observed by SMM in 1980 and probably faster than the Alfv¿n speed in the corona. The loop sides move laterally at a speed of ~800 km s-1 and show no evidence for a significant deceleration. The excess mass in the bright loop is estimated at ≤2¿1014 g, comparable to the expected error in its measurement and well below the value found to be typical of coronal mass ejections. ¿ American Geophysical Union 1987