Roger, it may have been the distnt triggered quakes. I thought I was seeing a pattern of two or more M6+ quakes occurring on the same day separated by 1000s of km.

I don't feel like looking through the archives...was a couple of years ago. If you were to re-do this, it would be M6+ quake, then other M6+ quakes that occur >1000 km away (for example). I suppose it could be a plot of time to next M6+ quake in hours. If there were 100 non-aftershock M6+ quakes in the world then you might expect an average of 88 hours to the next quake of there is no relation, and some normal distribution around 88 hours(I think there are more, the actual number has been posted on this page and is probably hiding on the USGS-NEIS site). I may have my statistical logic completely wrong.

You might also exclude aftershocks as the initial quake from which you search for following quakes. For example, there would be so many from Japan (or Chile, or Sumatra) that they would overlap each other. Also, the Sumatra rupture was 1400 km-long so if you look for quakes >1000 km away on that one, you would be including aftershocks.

Chris

Chris