a6 EXPERIMENTS ON THE FROUDE. 
The thrust-deduction is computed by the usual method 
T—R 
t= a (8) 
in which T and R are the thrust and the resistance, both in pounds. 
Finally the hull-efficiency is found by the equation 
I-—? 
sane (9) 
in which w is the wake and ¢ is the thrust-deduction. 
An investigation of the meaning of the results of our experiments 
should be preceded by a discussion of the degree of precision to be attributed — 
to the several elements that enter into the formation of the preceding table. 
The speeds, revolutions and the apparent slips are determined with all 
desired precision; the only uncertain element is the personal equation of 
the observer in giving the signal at the passing of the ranges at the beginning 
and end of the course; even with allowance for this element the error may 
be estimated at a fraction of 1 per cent. 
As has already been stated a variation of two pounds direct thrust of 
the propeller shaft could be definitely observed and measured by the thrust 
mechanism. At good speeds (from six to seven knots), for which the thrust 
is 370 to 150 pounds, the error in the determination of thrust was small; 
from one-half of 1 per cent. to 14 per cent. At the lower speeds (four knots 
and less) the error inthe thrust is considerable, amounting in some cases 
to 4 per cent. = 
We were advised by our confreres of the Department of Electrical 
Engineering that the rating of our motor in the laboratory had a precision 
of one-half of 1 per cent. and that we might expect a like degree of accuracy 
in our measurements of power input to the motor. By using stand- 
ardized shunts this precision was maintained for all except low powers. 
During the season of 1910 we relied for determination of friction on a 
brake which was bolted abaft the propeller when the boat was drawn up. 
For various reasons the installation and running of this brake was trouble- 
some and we were not entirely satisfied with the results. Nevertheless, a 
careful consideration of the results of the tests and a comparison with tests 
by a good brake installed in the boat in 1911 convinces us that our error 
is probably not greater than 1 per cent. in the determination of powers for 
high speeds; for low speeds the error is somewhat greater. At full power 
the frictional loss amounts to about 10 per cent. which is not very different 
from the friction commonly assigned to asteamengine. The propeller shaft 
