26i 



capability thus given of making any additional head of water 

 occurring at intervals, by freshes or any other cause, avail- 

 able, by letting the water on the wheel at higher and higher 

 levels. 



The first course of experiments is devoted to the deter- 

 mination of the comparative value of two water-wheels, the 

 one whose diameter is equal to the whole fall ; the other to 

 the head and fall, or to the total descent. By the head, the 

 author always means the efficient head, or that due to the 

 actual velocity of efflux at the sluice or shuttle, as determined 

 by Smeaton's experimental method, — this was equal to six 

 inches in all cases. 



The apparatus employed in these researches consisted of 

 two accurately made models of these wheels, with curved 

 buckets, made of tin plate, the arms, &c. of brass, and the 

 axes of cast iron, working on brass. Special contrivances 

 were adopted to measure the weight of water passed through 

 each wheel in each experiment, which was in every case 

 1000 lbs. avoirdupoise ; and others, to preserve the head of 

 water quite constant, — to determine the number of revolu- 

 tions made per minute, and thence the speed of the wheels. 

 One wheel was 25.5 inches diameter, the other 33 inches 

 diameter. The value of the labouring force was determined 

 directly by the elevation of known weights to a recorded 

 height by a silken cord over a pulley ; the altitude was read 

 off, on a fixed rule placed vertically against a lofty chimney. 

 The relative power of the wheels was determined by the 

 speed of rotation of a regulating fly or vane. 



All the principal results given in the ten tables are the 

 average o^five good experiments. From the accurate work- 

 manship and large size of these models, the peculiar con- 

 trivances for ensuring accuracy of observation, and the care 

 taken in the experiments, the author reposes considerable 

 confidence in his results as practical data. 



The velocity, in reference to maximum effect, is determined, 



