ay at sea. The introduction of the triple expansion engine, with
the three cranks, gave better balance to the shaft, and the forces
acting in the path of the crank pin, being better divided, caused more
regular motion on the shaft, and so to the propeller. This is
specially noticeable in screw steamers, and is taken advantage of by
placing the cabins further aft, nearer the propeller, the stern having
but little vibration; the dull and heavy surging sound, due to unequal
motions of the shaft in the two-crank engines, is exchanged for a more
regular sound of less extent, and the power formerly wasted in
vibrating the stern is utilized in propelling the vessel. In spite of
all these improvements I have mentioned, there remains the serious
question of defects in the material, due to variety of quality and the
extreme care that has to be exercised in all the stages during
construction of crank or other shafts built of iron. Many shafts have
given out at sea and been condemned, through no other cause than
_original defects_ in their construction and material.

The process of welding and forging a crank shaft of large diameter now
is to make it up of so many small _pieces_, the _best shafts_ being
made of what is termed scrap, representing thousands of small pieces
of selected iron, such as cuttings of old iron boiler plates,
cuttings off forgings, old bolts, horseshoes, angle iron, etc., all
welded together, forged into billets, reheated, and rolled into bars.
It is then cut into lengths, piled, and formed into slabs of suitable
size for welding up into the shafts. No doubt this method is
preferable to the old method of "fagoting," so called, as the iron
bars were placed side by side, resembling a bundle of fagots of about
18 or 20 inches square.

The result was that while the outside bars would be welded, the inside
would be improperly welded, or, the hammer being weak, the blow would
be insufficient to secure the proper weld, and it was no uncommon
thing for a shaft to break and expose the internal bars, showing them
to be quite separate, or only partially united. This danger has been
much lessened in late years by careful selection of the materials,
improved methods of cleaning the scrap, better furnaces, the use of
the most suitable fuels, and more powerful steam hammers. Still, with
all this care, I think I may say there is not a shaft without flaws or
defects, more or less, and when these flaws are situated in line of