THE STEAM TURBINE—PARSONS. 107 
The application of the turbine to the propulsion of vessels involved 
some interesting problems. The most important was, how slow could 
a turbine be made to rotate consistently with the maintenance of its 
efficiency in steam consumption, and at the same time be of moderate 
weight and cost ? 
In the same problem naturally arose the question of how fast could 
a screw propeller be made to revolve when propelling a vessel of a 
given size and at a given speed—in other words, when delivering a 
given propulsive horsepower at a given speed? The first question as 
to designing a low-speed turbine was solved in 1894 to 1896, by the 
aid of the accumulation of accurate data from experiments on land 
turbines; and the modification arrived at in the turbine has been 
chiefly directed to the splitting of it up into two or three or more tur- 
bines in series on the steam, and each working a separate shaft. This 
splitting up of the turbine results in a twofold advantage. It makes 
the turbine (which otherwise would be very long) much shorter, and 
because of being shorter finer clearances and less loss by leakage re- 
sults, and the whole engine is hghtened. A secondary gain, resulting 
from the division of the power over several separate shafts, arises 
from the fact that smaller propellers may be used, making higher 
speeds of rotation admissible, which again acts in lightening and im- 
proving the economy of the turbines. 
The second question, that of the propeller, was much more difficult. 
It was not simply the problem of designing a screw with a moderate 
slp ratio and a moderate loss by skin-friction of the blades in the 
water, but it was complicated by cavitation, or the hollowing out of - 
the water and the production of vacuous cavities caused by the force 
of the blades tearing through the water, a phenomenon first noticed 
by Sir John Thornycroft and Mr. Sidney Barnaby in 1893, and by 
them named cavitation. This apparatus shows the phenomenon. 
| A small tank was shown, with a model of the screw of a cross- 
channel boat or of an Atlantic turbine liner. It was pointed out 
that it was very difficult to make the screw cavitate, because it was 
especially designed not to cavitate; it was, however, made to do so’ 
in the tank by removing the atmospheric pressure from the surface of 
the water above the propeller by the air-pump. The removal of the 
atmospheric pressure, which helped to keep the water solid, enabled 
cavitation to be induced at a much lower speed of revolution. In the 
tank there was a head of about 14 inches of water above the topmost 
blades. If the tank had not been exhausted there would have been a 
head equivalent to 32 feet, plus 14 inches, plus capillary forces, tending 
to keep the water solid. Therefore, instead of 1,500 revolutions (the 
speed of the propeller when serious cavitation was induced), a speed 
of at least 20,000 revolutions would have been required (because forces 
