EXPERIMENTS ON THE FROUDE. 89 
the boat and gives signals at the beginning and end of a run; the assistant 
below in the house takes the electrical reading for the measurement of power; 
during the first season a voltmeter and an ammeter were used for this 
purpose, but during the past summer a wattmeter was substituted and was 
found somewhat more satisfactory. The assistant also makes necessary 
adjustments of machinery and apparatus before and after a run. 
The experiments consist in making series of progressive speed trials 
over a measured course an eighth of a knot long; or in some cases a course 
one-quarter of a knot is used. ‘This course is parallel to the sea-wall on 
the Cambridge side of the Basin where the water is partially protected from 
the prevailing winds. The conditions for such trials are as near ideal as can 
be found in an unenclosed space. ‘There is sufficient depth of water (22 feet), 
which is unaffected by tidal or other currents, except surface currents due 
to the wind. Our first year’s work showed an unexpectedly large influence 
of the wind even on fair summer days, and accordingly the observers began 
work early in the morning this present year; from four to eight o’clock in the 
morning were favorite hours. At this time also there was little interference 
from other boats. 
For progressive speed trials a continuous recording device is most 
desirable and for our purpose with only two observers it was indispensable. 
Our device is carried by a stand of convenient height, with rubber feet to 
deaden vibrations from the hull, set by the side of the propeller shaft, under 
the eye of the assistant observer who sees that all the pens are working 
properly and uses rubber stamps to indicate the beginnings and ends of 
runs and other information. A strip of paper six inches wide is drawn at 
any desired speed by a little electric motor, over guiding rolls, and across a 
horizontal metallic plate. Above this plate and bearing on it are a number 
of fountain pens with multi-colored inks; some are actuated by electro- 
magnets and others are connected with the mechanism which weighs the 
thrust of the propeller. One pen, which is controlled by a clock, marks half 
seconds, and another controlled by the principal observer indicates the 
beginnings and ends of the runs. Another pen records the revolutions of 
the propeller shaft. For convenience in counting every fifteenth time 
impulse and every tenth revolution is omitted. 
The thrust of the propeller is weighed directly and a continuous record 
is kept by the recording-device, described in Appendix A. For this purpose 
the thrust-block is given a fore-and-aft motion of about three-eighths of 
an inch on a bed plate. The block bears against the short vertical arm of a 
bent-lever and lifts a long horizontal arm with a sliding weight which 
measures the major part of the thrust. The remnant of the thrust is 
