Mr. J. Taylor's Description of a Rain-gauge. 407 



and measured ; and fig. 2. the arrangement of the wheels and 

 hands, which indicate the quantity that has passed through 

 the instrument. 



The whole is inclosed in a japanned tin case, of which the 

 upper part is the receiving funnel, from which the water drips 

 into a smaller funnel and pipe, which conducts it to a water- 

 wheel A, having three buckets. One of these buckets is al- 

 ways detained in a position to receive the flow of water by the 

 bent lever and weight B, which presses against two of the 

 three pins ccc, and keeps the wheel still, until a quantity of 

 water has entered the bucket sufficient to raise the weight, 

 when it instantly escapes; and the next bucket is presented in 

 the same position, to receive another charge, being detained 

 by the next pin coming in contact with the lever. A slight 

 spring d is attached, which takes the pin as it rises to retard 

 the velocity of the wheel, and to prevent its overshooting its 

 proper position. For the same reason the lever is slightly 

 notched where the pins rub against it in their passage. 



The instrument is adjusted to the proper measure of water 

 by the weight on the bent arm, which screws backward and 

 forward for this purpose, and admits of its being regulated to 

 great accuracy. 



It is evident that such an instrument, if the principle can 

 be correctly applied, would measure any flow of water which 

 might not be too large for being conveniently submitted to an 

 apparatus of this kind. 



Fig. 2.' shows the arrangement of the wheels and graduated 

 circles for denoting the quantity of rain. The large funnel 

 has an area of 72 square inches, and the water-wheel is ad- 

 justed to revolve once with 7*2 cubic inches, or what is equal 

 to 1-lOth of an inch in depth, on the area of the funnel. Thus 

 each bucket will upset with 2"'l- cubic inches; but as the wheel 

 may not be exactly in equilibrium, it is better to adjust it to 

 one complete levolution, by suffering 7*2 cubic inches to drip 

 in and correct it by the counterpoise. 



On the axis ol" the water-wheel is a pinion of 8 teeth {e) 

 driving a wheel of 80 teeth {/), which thus revolves once for 

 ten revolutions of the water-wheel, and by a hand on the arbor 

 indicates one inch by completing the circle, and the divisions 

 into tenths as marked on the circle; but as each bucket carries 

 the hand but one-third of a division, the observations nuiy be 

 made to l-30th of an inch in depth. 



To extend the register of the (juantity to a long period, tlie 

 other wheels are added ; and on the arbor of the wheel f is 

 a j)iiiion of 8 teeth {g), driving a wheel of 80 teeth {h), on the 

 arbor (jf which is a pinion of 20 teeth (/), driving the wheel of 

 80 teeth (/), on the arbor of which is another hand pointing 



to 



