Continually lower comes the temperature of the piece—1500°, 1400°, 1375°, 1350°, 1325°, 1300°, 1275° F.,—and lo! the piece jumps, and from this all the way down to cold it responds to the attraction of our magnet. Just to make sure that we are not “seeing things,” we again start our furnace, and, as the steel heats, we test it with the magnet. So far there is no doubt about its being magnetic! At 900° F., the pieces begins to show dark red, at 1000°, 1100°, 1150°, 1200°, 1250°, stronger and stronger red. At all of these temperatures the steel is attracted. So it is at 1275° and at 1300° F. But just as we are thinking that we must have been mistaken before, we find that again the steel is suddenly “dead” to the pull of the magnet! And at what temperature? The pyrometer indicates 1320° F. But was not this the same or very nearly the same reading at which the pyrometer needle paused on the way up, and do you not remember that it was only a little below 1250° F. that it paused on the way down, and the disagreement of the two temperatures we ascribed to “lag”? No, we made no mistake. Steel loses all of its magnetic properties at the “critical range” and has none above it. Dilatation and Conductivity Certain other great changes, too, occur here. We know that most materials expand uniformly upon heating and contract as they cool. Steel is no exception, but at the critical range on heating it becomes fickle and for a short space contracts instead of continuing its uniform expansion. Conversely, during cooling, it ceases its uniform contraction and suddenly dilates or expands for a short period when it reaches the critical range, after which aberration it again resumes its old habit of uniform contraction as the temperature falls. Just so with its electrical conductivity. At the critical range the electrical conductivity suddenly decreases abnormally as the piece gets hotter and as abnormally increases as the steel cools through the critical range on the return trip. There are certain other happenings at or near this particular temperature but we will not consider them here. Manifestly all of this has a deep meaning. Recalescence Indicates the Hardening Point You remember that we said the divergence between the going up and the coming down pauses of the pyrometer needle was due to tardiness or lag? Among humans habitual tardiness is not considered a desirable trait, but it is undoubtedly through this very lag or tardiness that steel becomes so serviceable to us. [Illustration: HEATING AND COOLING CURVES OF STEEL WITH .9 PER CENT OF CARBON ]