If the crack be as shown in [No. 2], the middle of the blade has been hotter than the corners: snipping and comparing the grains will tell the story.
If the crack be more nearly a straight line, as shown in [number 3], the chances are that there is a seam there and the steel is at fault.
How to Tell a Seam from a Water-crack.—A seam is caused by a gas-bubble in the ingot which has not been closed up by hammering or rolling; it always runs in the direction of the work; in bars it is parallel to the axis.
The walls of a seam are always more or less smooth, the surfaces having been rubbed together under heavy pressure during hammering or rolling, and they are black usually, being coated with oxide.
The walls of a water-crack are never smooth, they are rough and gritty, and they may have any of the temper colors caused by the action of water and heat.
There need never be any question as to which is which.
If a long tool cracks down the middle, it may be from too much heat, from seams, or from a lap.
A lap is caused by careless working under a hammer, or by bad draughts in the rolls, folding part of the steel over on itself. Laps, like seams, run parallel to the axis of a bar, and usually in very straight lines.
Any long piece of steel may be split in hardening by too much heat. In making the experiment of heating a piece continuously from scintillating, or creamy color, down to black, to show the differences of grain due to the different heats, the sample almost invariably splits down the middle as far as the strong, refined grain, or nearly that far.
As stated before, a round bar will be almost certain to split if it be heated up to medium lemon, although a square bar may endure the same heat without cracking.