the middle part of the boat--just where the
large wheels come in the engraving--two of the lines are longer than the
others, so that a man might use them as stirrups, and thus be enabled to
clamber into the boat even without assistance. The rudder descends
considerably below the keel--to give it more power--and has to be raised
when the boat is being launched.

The shaded parts of the diagrams show the position and form of the
air-cases which prevent a lifeboat from sinking. The white oblong space
in Figure 2 is the free space available for crew and passengers. In
Figure 3 is seen the depth to which the air-chambers descend, and the
height to which the bow and stern-chambers rise.

It is to these large air-chambers in bow and stern, coupled with great
sheer--or rise fore and aft--of gunwale, and a very heavy keel, that the
boat owes its self-righting power. The two air-chambers are rounded on
the top. Now, it is obvious that if you were to take a model of such a
boat, turn it upside down on a table, and try to make it rest on its two
_rounded_ air-chambers, you would encounter as much difficulty as did
the friends of Columbus when they sought to make an egg stand on its
end. The boat would infallibly fall to one side or the other. In the
water the tendency is precisely the same, and that tendency is increased
by the heavy iron keel, which drags the boat violently round to its
right position.

The self-righting principle was discovered--at all events for the first
time exhibited--at the end of last century, by the Reverend James
Bremner, of Orkney. He first suggested in the year 1792 that an
ordinary boat might be made self-righting by placing two watertight
casks in the head and sternsheets of it, and fastening three
hundredweight of iron to the keel. Afterwards he tried the experiment
at Leith, and with such success that in 1810 the Society of Arts voted
him a silver medal and twenty guineas. But nothing further was done
until half a century later, when twenty out of twenty-four pilots lost
their lives by the upsetting of the non-self-righting Shields lifeboat.

Then (1850) the late Duke of Northumberland offered a prize of 100
guineas for the best lifeboat that could be produced. No fewer than 280
models and drawings were sent in, and the plans, specifications, and
descriptions of these formed five folio manuscript volumes! The various
models were in the shape of pontoons, catamarans or rafts, north-co