I think he's eluding to what I said. As the water rises, it follows the low contours and surface area increases, so the quantity of water necessary to raise the first inch is less than the qty of water for the subsequent inches.