Considering rotating rockets, the normal behaviour indicates an exponential rise of the rotation period with time. But in our archive (PPAS) we have a lot of accelerations. (see the web pages: experimental results of flash period measurements). Most of the time there is also a change in mean motion, and accordingly in drag values on the same time. (see again web pages of the BWGS: flash 91, 92-8B accelerated). Thanks to Ted Molczan and Mike McCants(the kilroy archive), I've been collecting all element-files since May 92. Now I can start to compare them with the accelations in our archive. Questions: - Has somebody already done this before? Are there any 'correlations', or is the correlation very fuzzy (as it seems to me on a first look). - Do you know where to get older elements (before may 92)? In our archive (the PPAS) we have a lot of old, good observed accelerations. Maybe some (official?) organization can make them accesible through the net? greetings, Kurt Jonckheere