Hey Don,

Just a suggestion. August also has 31 days in it, February has 28 usually, so there is going to be a bias.... What about running the numbers based on average number of quakes per day? Might be better at avoiding the number of days in the month probem. Better yet would be an average number of quakes per day through the entire year, then look for groupings within the year ....

Just rambling,

Michael