eight per horsepower of feather-weight motors appeared to range from
2 1/4 pounds up to 7 pounds per brake horsepower, some actual figures
being as follows:

Antoinette........ 5 lbs.
Fiat.............. 3 lbs.
Gnome....... Under 3 lbs.
Metallurgic....... 8 lbs.
Renault........... 7 lbs.
Wright.............6 lbs.

Automobile engines, on the other hand, commonly weighed 12 pounds to 13
pounds per brake horsepower.

For short flights fuel economy was of less importance than a saving
in the weight of the engine. For long flights, however, the case
was different. Thus, if the gasolene consumption was 1/2 pound per
horsepower hour, and the engine weighed 3 pounds per brake horsepower,
the fuel needed for a six-hour flight would weigh as much as the engine,
but for half an hour's flight its weight would be unimportant.

Best Means of Propulsion.

The best method of propulsion was by the screw, which acting in air
was subject to much the same conditions as obtained in marine work.
Its efficiency depended on its diameter and pitch and on its position,
whether in front of or behind the body propelled. From this theory
of dynamic support, Mr. Lanchester proceeded, the efficiency of each
element of a screw propeller could be represented by curves such as were
given in his first lecture before the society, and from these curves the
over-all efficiency of any proposed propeller could be computed, by mere
inspection, with a fair degree of accuracy. These curves showed that the
tips of long-bladed propellers were inefficient, as was also the portion
of the blade near the root. In actual marine practice the blade from
boss to tip was commonly of such a length that the over-all efficiency
was 95 per cent of that of the most efficient element of it.

Advocates Propellers in Rear.

From these curves the diameter and appropriate pitch of a screw could
be calculated, and the number of revolutions was then fixed. Thus, for a
speed of 80 feet per second the pitch might come out as 8 feet, in which
case the revolutions would be 600 per minute, which might, however, be
too low for the motor. It was then necessary either to gear down the
propeller, as was done in the Wright machine, or, if it was decided to
drive it direct, to sacrifice some of the efficiency of the propeller.
An analogous case arose in the application of the steam turbine to the
propulsion of cargo boats, a problem