Galileo reported that if two balls with different densities are dropped simultaneously, the less dense ball will initially fall faster, but later lag behind the more dense ball. The authors observed this same behavior for a lead ball and a billiard ball of different size. Numerical calculations show that due to the complex dependence of the drag coefficient on the velocity and size of a ball, it is possible under certain coditions to drop two balls simultaneously and find that one initially lags behind but then passes the other. However, the specific behavior observed by the authors, and probably Galileo also, was most likely due to a systematic error in the dropping apparatus.