8 On a New Foi'm of 



acting against the water, and will so continue to the end of 

 the stroke. However rapid the movement of the propeller, 

 a certain time is required for the floats to get into this 

 position, and therefore, especially if the floats have a large 

 surface, there will be a slip, amounting probably in extreme 

 cases to one foot, which it will be, perhaps, impossible to pre- 

 vent ; at the end of the stroke, when the action of the pro- 

 peller in the water ceases, the floats will be rapidly feathered 

 by the forward movement of the ship). In the act of feather- 

 ing there will be a certain resistance and loss of power from 

 the sudden lateral displacement of the water, caused by the 

 rotation of the floats upon their axes. 



The projecting neck of each float will be thrown forward, 

 but, as it is very narrow, and may be made still more so than 

 shown in the model, this portion of the float will offer no 

 resistance of any consequence. 



The only loss of power, therefore, will arise from the back- 

 ward and lateral movement of the body of the float. 



The backward movement of the float will rapidly take 

 place at the end of the stroke, and before the commencement 

 of the return stroke, the fixed extremity of the float being 

 carried forward by the motion of the ship, while the free 

 extremity is left behind. 



The float therefore will turn on its centre, the water being 

 displaced partly forward and partly backward. A certain 

 loss of power must result from this displacement which it 

 would be difficult to estimate, but from the rapidity of the 

 movement, and the extremity of the float being left un- 

 restrained in its motion, it is difficult to see how the resist- 

 ance could be great. 



During the lateral movement of the float the return stroke 

 of the piston will begin to operate on it, and the act of 

 feathering will be completed in a gradual manner during its 

 rapid forward movement. The extent therefore of the 

 lateral movement of the float will not correctly express the 

 amount of displacement or the actual loss of power, the float 

 being drawn obliquely rather than forced laterally into its 

 position parallel with the shaft. Thus the actual lateral 

 displacement of the water will be practically very much 

 modified and diminished. 



In theory, therefore, the loss of power in feathering the 

 floats would appear to be very small, and in a length of stroke 

 of ten feet could scarcely exceed one foot, and supposing one 

 tenth part of the stroke of the propeller to be ineffectual for 



