Transactions of the Horticultural Society. 179 



A will also decrease in consequence of its expansion ; but as soon as the 

 column (c (V) of fluid above the centre of the upper pipe is of a greater 

 weight than the column/e above that centre, motion will commence along 

 the upper pipe from a to b, and the change this motion produces in the 

 equilibrium of the fluid will cause a corresponding motion in the lower pipe 

 from B to A ; and, in short pipes, the motion will obviously continue till 

 the temperature be nearly the same in both vessels ; or if the water be made 

 to boil in a, it may also be boiling hot in b, because ebullition in a will assist 

 the motion. 



" 5. The causes which tend to retard the motion of water in the pipes 

 are, first, the contraction of the moving fluid at the orifice of the pipes ; 

 secondly, the friction of the fluid in the pipes, which sets the limit to the 

 distance to which the pipes can be extended to produce the proper quantity 

 of useful effect : but it is remarkable, that the higher the temperature of 

 the moving fluid the less its friction ; thirdly, the motion is retarded by the 

 cooling of the fluid in its progress along the pipes, such cooling having a 

 tendency to produce a double current ; and, fourthly, by bends and changes 

 of form. 



" 6. It will be evident to any person of philosophical research, however, 

 that in considering water the only liquid capable of being employed, we 

 should be losing sight of one of the greatest advantages resulting from the 

 knowledge of natural phenomena; for all liquids expand by heat, and 

 hence in all of them its partial application would produce motion under 

 proper circumstances ; while the boiling points of different liquids are at 

 such different temperatures that we may vary the ultimate temperature of 

 the heating surface from 100° to 600", that of water being 212°. This mode 

 of considering the subject opens a new source for speculation and for im- 

 provement, which it will be desu-able to consider more in detail after ana- 

 lysing the laws of the motion of liquids by heat. 



" 7. A general investigation, embracing all the circumstances concerned 

 in the motion, would be extremely intricate, and hence I shall not attempt 

 to include more than those which are of sufficient importance to have an 

 influence on the results requiring attention in practice ; and, for a like 

 reason, I adopt the most simple formula of hydraulics tliat applies to the 

 case : — 



Put /=the sum of the lengths of the pipes in feet. 



/i=the depth of the liquid in the boiler in feet below the centre of the 

 upper pipe. 



e=the expansion due to the mean difference of temperature at the 

 extremes of the apparatus. 



/=the friction of the liquid against the surface at the mean temper- 

 ature for 1 ft. in length and 1 in. in diameter, 



</=the diameter of the pipe in inches; and 



i'=the velocity in feet per second. 



" The friction of a pipe is as its surface, and the square of the velocitjf 



N 2 



