In working back to the set of bright flashes in the 3:34-3:37 UT period, a discontinuity showed up. The BRIGHT flash times did not appear at the right time. For example, assuming a 23.4 second period from bright flash to bright flash, and working back from a known bright flash at 3:42:02.7, then approximately 3:36:58.5 UT should be a bright flash. But the bright flash occurs at 3:36:46.8. The difference is 11.7 seconds which is half the full period of 23.4 seconds. My conclusion is that sometime after 3:37, the Bright flash is shifted by 1/2 period (11.7 sec) Note that at 3:37, the satellite was hard to view and in the 3:38:30 to 3:42 period, the flashes are nearly equal intensity. Can someone confirm this conclusion? Ron Lee