Now let us take the steel out and examine it. The act of tapping it on the anvil in order to knock off the surplus water gives us a hint that our test piece has undergone some sort of a change. For now it rings with a bell-like clearness and gives the hammer with which we strike it a quick snapping rebound which in itself indicates great hardness. Next, we test the piece with a hardened steel file with which we could easily have made a deep ridge before we attempted the heating operation and to our surprise the file has as little effect as if it had been made of wood. And to our surprise on closer examination, we actually find that our test piece has scratched the file—surely it must be very hard. We are convinced that some marked change must have taken place. What can it be? Why it must be that due to the rapid cooling in the pail of ice water we brought the temperature of the test piece down below the critical range =before= the abnormal condition at which it existed while at and above the critical range had found =time= to change back to its former condition. And we remember that if one of these allotropic changes is going to take place at all, nature says it =must= do so while the steel is within the critical range and therefore having forced the steel through that critical range which separates one allotropic condition from another, before it had found =time= to effect its desired change, we managed to entrap the abnormal condition so that we could see it and feel it and get familiar with it at room temperature. If we so desire we can now make other hardness tests on our piece of steel at our leisure. For these scientists have invented several machines. One of the most common is called the scleroscope in which a hardened steel ball is allowed to drop from a given height on to the piece of steel to be tested. Then the rebound of the ball is carefully noted. The higher the rebound, the harder the piece. That is natural isn’t it? We know that if the ball were allowed to drop on butter, it wouldn’t rebound at all, because the butter is so soft. A piece of wood would possibly record a very tiny rebound, while a piece of hardened tool steel would effect a very material action of the scleroscope ball, thus indicating extreme hardness.