My interest in an acoustic measure of global ocean warming came tip suddenly in April 1988. I had given a lecture invited by, the Office of Naval Research (ONR) on the subject of a 1960 experiment (Shockley et al., 1982) in which 300 lbs. of dynamite were detonated in the sound channel off Perth, Australia. The explosion was clearly recorded by Bermuda hydrophones halfway around the earth. The authors had interpreted the result as a transmission along the unimpeded oceanic great-circle route between Perth and Bermuda. Munk, O’Reilly and Reid (1988) asked the question whether the neglect of lateral refraction and of the earth’s ellipticity was justified. Accordingly, they constructed a set of refracted geodesics radiating out of the shot site. Surprisingly, the north-south temperature gradient associated with the Antarctic Circumpolar Current deflects the rays to the left south so severely that they never make it to Bermuda: the island lies in the geometric shadow. It is not clear what scattering process was responsible for putting enough energy into the shadow zone for the explosions to be heard at Bermuda.