ng, pounding; bumps, blows, collisions,
oscillations, running aground--stresses that crumpled steel or
scissored the rivets in two.

It was hard to foresee the critical stress that should mean life or
death to the ship and its people. Some went humbly forth and came home
with rich cargo; some steamed out in pride and never came back; some
limped in from the sea racked and ruined; some ran stupidly ashore in
fogs; some fought indomitably through incredible tempests. Some died
dramatic deaths on cliffs where tidal waves hammered them to shreds;
some turned turtle at their docks and went down in the mud. Some led
long and honorable lives, and others, beginning with glory,
degenerated into cattle-ships or coastal tramps.

People were but ships and bound for as many destinations and
destinies. Their fates depended as much and yet as little on their
pilots and engineers, their engines and their frames. The test of the
ship and of the person was the daily drudgery and the unforeseen
emergency.

Davidge believed in preliminary tests of people and boats. Before
he hired a man or trusted a partner he inquired into his past
performances. He had been unable to insist on investigation in the
recent mad scramble for labor due to the sudden withdrawal into the
national army of nearly every male between twenty-one and thirty-one
and of hundreds of thousands of volunteers of other ages.

He had given his heart to Marie Louise Webling, of whom he knew little
except that she would not tell him much. And on her dubious voucher he
had taken Jake Nuddle into his employ. Now he had to accept them as he
had to accept steel, taking it as it came and being glad to get any at
all.

Hitherto he had insisted on preliminary proofs. He wanted no steel in
a ship's hull or in any part of her that had not behaved well in the
shop tests, in the various machines that put the metal under bending
stress, cross-breaking, hammering, drifting, shearing, elongation,
contraction, compression, deflection, tension, and torsion stresses.
The best of the steels had their elastic limits; there was none that
did not finally snap.

Once this point was found, the individual metal was placed according
to its quality, the responsibility imposed on it being only a tenth of
its proved capacity. That ought to have been enough of a margin of
safety. Yet it did not prevent disasters.

People could not always be put to such shop tests beforehand. A
reference or two, a