I have tried to refine the eclipse ephemeris of DV UMa using the following available mid-eclipse times. HJD-2400000 source ----------------------------------------- 46854.7485 Howell et al. (MNRAS 233, 79) 46854.8339 " 46854.9195 " 46855.6944 " 46855.7798 " 46855.8660 " 46855.9517 " 46856.4646 " 46856.5503 " 46856.6370 " 46857.4963 " 46857.5815 " 49775.9696 Nogami et al. (vsnet-alert 56) 49776.0554 " 49776.1410 " 49776.2269 " 49779.0601 " 49779.1457 " 49779.2317 " 50079.289 Kato (unpublished) The best fit ephemeris is Min.HJD = 2446854.7487 + 0.085852614 E (1), which expresses all the observed times within an error of 0.002 day. The second best (viable) solution is Min.HJD = 2446854.7483 + 0.085877839 E (2) Using the equation (1), predicted times of eclipses on Apr. 9 and 10 would become: YYMMDD hh:mm:ss (UT Geo) 970409 00:36:14 970409 02:39:53 970409 04:43:31 970409 06:47:09 970409 08:50:47 970409 10:54:26 970409 12:58:04 970409 15:01:42 970409 17:05:20 970409 19:08:59 970409 21:12:37 970409 23:16:15 970410 01:19:53 970410 03:23:32 970410 05:27:10 970410 07:30:48 970410 09:34:26 970410 11:38:05 970410 13:41:43 970410 15:45:21 970410 17:48:59 970410 19:52:38 970410 21:56:16 The equation (2) predicts eclipses 1h 18m later. Hope this would help, and someone could confirm or reject this prediction. Regards, Taichi Kato