putty, and, unlike the steam hammer, there is no _jarring_ on the
material, and it is manipulated with the same ease as a small hammer
by hydraulics.

The tensile strength of steel used for shafts having increased from 24
to 30 tons, and in some cases 31 tons, considering that this was 2
tons above that specified, and that we were approaching what may be
termed _hard steel_, I proposed to the makers to test this material
beyond the usual tests, viz., tensile, extension, and cold bending
test. The latter, I considered, was much too easy for this fine
material, as a piece of fair iron will bend cold to a radius of 1½
times its diameter or thickness, without fracture; and I proposed a
test more resembling the fatigue that a crank shaft has sometimes to
stand, and more worthy of this material; and in the event of its
standing this successfully, I would pass the material of 30 or 31 tons
tensile strength. Specimens of steel used in the shafts were cut off
different parts--crank pins and main bearings--(the shafts being built
shafts) and roughly planed to 1½ inches square, and about 12 inches
long. They were laid on the block as shown, and a cast iron block,
fitted with a hammer head ½ ton weight, let suddenly fall 12 inches,
the block striking the bar with a blow of about 4 tons. The steel bar
was then turned upside down, and the blow repeated, reversing the
piece every time until fracture was observed, and the bar ultimately
broken. The results were that this steel stood 58 blows before showing
signs of fracture, and was only broken after 77 blows. It is
noticeable how many blows it stood after fracture. A bar of good
wrought iron, undressed, of same dimensions, was tried, and broke the
first blow. A bar cut from a piece of iron to form a large chain,
afterward forged down and only filed to same dimensions, broke at 25
blows. I was well satisfied with the results, and considered this