One hundred students, 25 each in grades 2, 4, 6, and 8, were individually interviewed. A cylindrical beaker with five equidistant levels marked on it was presented first, and a pear-shaped separation funnel was also presented. The funnel had a "faucet" at the bottom that allowed water to start and stop running into the beaker. The interviewer asked the child to mark the water levels on the funnel when the water had run to the first level in the beaker, to the second level, to the third level, and so on. The levels on the funnel thus varied a great deal depending on its diameter. The interviewer then asked, "Did the water take the same amount of time to go down from this level to the next in the funnel as it did to go up from this level to the next in the beaker?" It was found that correct judgments about durations were not made before eighth grade (the criterion being 75% of the eighth graders giving correct answers). This grade level was considerably later than the age of 7 or 8 (second grade) that Piaget reported. (Contains 3 figures and 1 table.)