As it happens, we are in the midst of doing the study of that very subject, and although I don't have formal data to present, our preliminary findings are that in the course of looking at the past five years' worth of long weekends, looking at the price before and the price after, and looking at price changes in the weeks immediately preceding and following those long weekends, the difference is negligible. It is close to zero. In other words, there are no objective data that bear out the myth that pump prices rise before long weekends.