the stacks are at the boilers' fire-door end. The steam lines came
off the crown of the boilers and probably passed through the ends of the
wheelbox to the engine; a trunk for the steam lines would undoubtedly
have been necessary.

[Illustration: Figure 15.--SKETCH OF 130-GUN SHIP proposed by Patrick
Miller to King Gustav III of Sweden in 1790. In Statens Sjoehistoriska
Museum, Stockholm.]

[Illustration: Figure 16.--PATRICK MILLER'S manually propelled
(paddle-wheel) catamaran ship _Experiment_, built at Leith, Scotland,
1786. Scale drawing in Statens Sjoehistoriska Museum, Stockholm.]

The engine is shown to have had counterbalanced side levers, one on
each side, and a single flywheel on the outboard side. The cylinder is
over the condenser or "cistern," connected by the steam line and valve
box on the side. The cylinder crosshead is shown in the inboard profile
to have reached the underside of the beams of the upper deck. The
crosshead was connected by two connecting rods to the side levers. These
levers operated the paddle wheel by connecting rods to cranks on the
paddle-wheel shaft. There is another pair of connecting rods from the
side levers to the crosshead of the air pump. All connecting rods are on
one arm of the side levers, the other end having only a counterbalance
weight beyond the fulcrum bearing. The flywheel has a shaft fitted with
two gears, and is driven through idler gears from gears on the
paddle-wheel shaft; it turns at about twice the speed of the paddle
wheel. No other pumps or fittings are shown in the engine hull, although
manual pumps were probably fitted to fill and empty the boilers. Piping
is not shown.

[Illustration: Figure 17.--PAINTING OF THE _Experiment_ in the Statens
Sjoehistoriska Museum, Stockholm.]

The four rudders, toggled in pairs, are shown in both the lines and
inboard drawings, but the shape is different in the two plans. Operation
must have been by a tiller under the gun-deck beams. The outer end of
the tiller may have been pivoted on the toggle bar and the inboard end
fitted, as previously described, with steering cable or chain tackles.
This seems to be the only practical interpretation of the evidence.

Reconstructing the Plans

In the model it was necessary to reconstruct the deck arrangements
without enough contemporary description. The outboard appearance and
hull form, rig, and arrangement of armament require no reconstruction,
for all that is of importance