EXPERIMENTS ON THE FROUDE. 109 
APPENDIX A. 
THRUST MECHANISM. 
The thrust mechanism which is an essential feature of our apparatus 
and which has some special features is represented by Figs. 27, 28 and 29, 
Plates 58-59. In order to allow the propeller shaft to be shifted fore-and-aft 
it was threaded for its entire length, save for the part that might be in the 
shaft-log. On the shaft, as shown by Fig. 27, Plate 58, there was a com- 
position sleeve which was locked by a nut both forward and aft, when the 
shaft was set in any desired position. ‘There was one thrust collar at the 
middle of the sleeve with adjustable steel races for the balls that carried the 
thrust to the thrust block. Near the ends of the sleeve were cylinderical 
races with balls to carry the weight of the shaft. 
The bent-lever for weighing the thrust is shown by this figure and by 
Figs. 28 and 29, Plates 58-59. The long horizontal arm (shown broken) is 
40 inches long; the short horizontal arm carries a counter-weight for use 
in making adjustments. The short vertical rod is forked and bears directly 
against the thrust-block knife edges; at the bearing surface the short arm 
has hard steel plates to guard against indentation of the knife-edges. The 
shell of the thrust block is split diametrically for convenience in construction 
and assembling. It carries two lugs on each side as shown by dotted lines 
in Fig. 27, Plate 58, and by full lines in Fig. 28, Plate 58. These lugs are 
split and there are four hard steel knife edges inserted with their edges in 
the horizontal diametrical plane. The bent-lever has a ball-bearing pivot 
as shown by Fig. 29, Plate 59. The entire thrust-block rests on a bed-plate 
and has a fore-and-aft motion of three-eighths of an inch; this is more than 
necessary so that troublesome adjustments may be avoided. When the 
boat was maneuvering to come on the course the block could be wedged 
fast, and when on the course it was cast loose. 
As already described, the major part of the thrust was counterbalanced 
by a large sliding weight on the long arm of the bent-lever. To allow for 
and measure minor fluctuations the extreme end of the bent-lever bore 
against an indicator spring that was secured to the frame of the boat. The 
deflections of this spring were carried by a light wire over multiplying and 
guiding pulleys to a pen that traversed across the tape of the recording 
