184 HOT-WATER BOILERS AND PIPES. 



that may sometimes be generated on the surface of the water in 

 the flow-pipe, would find egress, unless the supply-pipe were 

 bent in the shape of an CQ to prevent it, which is a very good 

 plan ; and, as a small lead pipe of about l inch bore is suffi- 

 cient to supply a boiler of considerable size, the pipe can easily 

 be bent in any shape to answer the purpose. 



3. Impediments to circulation, fyc. The power which pro- 

 duces the circulation of the water in the pipes is the specific 

 gravities of the two bodies in the return and flow-pipes ; whether 

 this force acts on a pipe 100 feet in length, or on one only 5 

 feet in length, the result is precisely similar. 



Now it is evident that if this unequal pressure is the vis viva, 

 or motive power, which sets in motion the whole quantity of 

 water in the apparatus, in order to ascertain the exact amount 

 of this force, it is only necessary that we know the specific 

 gravities of the two columns of water, and the difference will, of 

 course, be the effective pressure, or motive power. This can be 

 accurately determined when the respective temperatures of the 

 water in the boiler and in the descending or returning pipe are 

 known. 



As this difference of temperature rarely exceeds a very few 

 degrees in ordinary cases, the difference of the weight of the two 

 columns must be very small. But, probably, the very trifling 

 difference that exists between them, or, in other words, the 

 extreme smallness of the motive power, is very imperfectly com- 

 prehended, and will, perhaps, be regarded with some surprise, 

 when its amount is shown by exact computation. 



In order to ascertain, without a long and troublesome calcula- 

 tion, what is the amount of motive power for any particular 

 apparatus, the following table has been constructed. An appara- 

 tus is assumed to be at work, having the temperature in the 

 descending pipe 170, and the difference of pressure upon the 

 return-pipe is calculated, supposing the water in the boiler to 

 exceed this temperature, by from one to twenty degrees. This 

 latter amount will exceed the difference that usually occurs in 

 practice. 



By referring to the annexed table, it will be found that when 



