lo Sept., 19 io.] Pump and Water Measurement. 569 



1. A double pulley fastened on top of rim of tank with a large wooden 

 float in the tank and a heavy — on account of wind action — indicator 

 running down the outside. When full, the indicator will be 

 at the bottom. 



2. Unless the flume is running very fast, a system of 

 float measurement should be suitable. Float a piece of 

 wood along a length of, say, 100 feet. Take number of 

 seconds and thus find feet travelled per minute. Reduce 

 this by 20 per cent. This, multipHed by the square 

 feet in section, will give cubic feet per minute. Each cubic foot equals 

 6^ gallons. For instance, suppose a triangular flume 12 inches wide at 

 surface of water and 6 inches deep, the area is | x | = \ ; and suppose 

 that float travels 100 feet in 55 seconds, then 55:6o::ioo = 109 feet 

 per minute, deduct one-fifth or 20 per cent., the velocity is 87 feet per 

 minute, and this multiplied by the area, ^ sq. foot, gives, say, 22 cubic 

 feet or 137 gallons per minute. 



For weir measurement, the area of the notch must be not more than 

 one-sixth, say, of approach channel and, the depth passing over the notch, 

 must be measured to still water. The discharge in cubic feet per minute 

 equals length of notch in feet by the square root of the cube of the depth 

 in feet by 200, or in symbols Q = 200 Iht For instance, notch is 

 18 inches or ih feet long and depth flowing over is 3 inches or \ foot 

 then — 



Q = 200 X i^ X g = 375 cubic feet per minute. 



3. The soap used in packing will not repel kerosene for any time. 

 Get asbestos packing, but see that it is somewhat larger than space between 

 spindle and gland. It should be well rubbed with powdered graphite and 

 packed into space tightly a little at a time. Coiling it on the spindle 

 and trying then to put all in at once will not give ^. ^ _^ 

 satisfaction. 'l l'>'~^ ' i' ~~? 



4. There are several methods of which the most f q 1 ^ 

 practicable, in your case, is the Prony brake, which ^V_v^ ^ 

 consists of a square frame with shaped brake blocks - -' 

 surrounding the circular shaft or pulley with a long lever arm. The 

 brake is screwed up by the bolts till the engine is exerting its full power, 

 a weight on the end of the lever keeping the brake from revolving round 

 the wheel. Sand poured into a box is a good weight as it can be added 

 gradually. The formulra is: — 



W = weight in pounds at end of arm. 



d = length of arm measured from centre of wheel in feet. 



R = revolutions per minute. 



X.TTP ,T> , , V W X r 14- X 2d X R 



rJUf (Brake-horsepower; = — 



33,000 



Gradually screw up brake, adding sand as found necessary. We 

 cannot, however, recommend you to take on the testing yourself, unless 

 there are special reasons. A considerable amount of experience is required 

 to provide against unforeseen events, which may affect results seriously. 



^^^t%^^ 



