247 
This formed the boat. A crank-handle was attached to a ver-. 
tical shaft opposite the seat. The lower end of the shaft was 
about 7 inches below the surface of the water. This end was 
so contrived that boards of various shapes and sizes to imitate 
a fish’s tail might be attached at pleasure so as to lie hori- 
zontally just under the surface of the water. A number of 
deal boards of various lengths, taper and shape were tried suc- 
cessively, but with not much success. Finally a large steel saw 
about 8 feet long was introduced. This was so far successful 
that had the saw been broader, and stiffer at the fixed end, it 
was clear that fair results might be anticipated. As it was, with 
very slight exertion, although the boat was very heavy, being 
nearly water-logged by water having found its way into the 
pipes, the boat moved forward at the rate of about two miles 
per hour. The action of the saw seemed to be precisely that 
of the tail of a fish, lashing from side to side and driving the 
water astern. When the crank-handle was turned to one side 
the saw was bent, and in trying to recover itself produced a 
pressure upon the water partly to one side and partly astern: 
it was the reaction of this latter portion which drove the boat 
forward. Any other form of boat would have done equally 
well, the upright shaft taking the place of the rudder. A pliant 
steel rod covered with India rubber or gutta-percha in the form 
of a tail might be even more successful. Within the last few 
months it has been reported in the newspapers that boats pro- 
pelled in this way have been tried successfully on some of the 
American canals. 
If such a mode of propulsion could be introduced with 
success, the machinery for driving it might be very simple 
