U5 



ratio of breadth to the diameter of the propeller, from which 

 a maximum result will be found. I tried a propeller having 

 eight arms, the angles of which were at 56°, and the propor- 

 tion of the breadth to the diameter about Jth, which 

 answered remarkably well. I thought that probably another 

 with four arms at the same angle, and of the same length, 

 but double the breadth, would answer as well; but I found 

 that the speed of the boat was thereby greatly diminished, 

 and the distance performed not so great. I have also found 

 that there is a decided loss by too many arms in the propellers : 

 the third propeller mentioned in the last paragraph, having 

 six arms, propelled the boats 150 links in 1' 43"; another 

 of the same diameter and angle, having sixteen arms, required 

 2' 10" to accomplish the same distance. From this it would 

 appear that in addition to the particles of water which are 

 put in motion by the actual contact of the propeller, there is 

 an extra quantity moved by the attraction or cohesion of the 

 particles of water to each other, and that therefore the dis- 

 tance of the propellers from each other should be such as to 

 take full advantage of this principle ; and that if they be too 

 much crowded in the wheel, they prevent the proper escape 

 of the water, and tend to retard the velocity of the vessel. 



The success attending these experiments in propelling 

 upon water, induced me to turn my attention to the same 

 principle to propel air. I have now in my possession a model 

 which clearly illustrates the great capability of a wheel of 

 this construction. The model is twelve inches in diameter 

 and two inches broad ; it has eight vanes, the periphery being 

 at an angle of 45 degrees ; it is mounted upon a frame at 

 one end of a covered opening, a little exceeding the diameter 

 of the wheel, and about eighteen inches long, and is put into 

 motion by a wheel and pinion, by which a great velocity can 

 be given to it. In the other end of the opening, I fix one of 

 the wheels I have made for measuring the velocity of currents 



