of Processes in the Fine and Useful Arts. 347 



vessels, and would level the water in them, if the connecting 

 tube D were stopped. The whole, therefore, must be in per- 

 fect equilibrium ; and it will remain so as long as there is an 

 equal temperature in the two vessels. 



The fluid being thus at rest, unstop E, and if heat be 

 applied to A, the water in this vessel will expand, and the very- 

 small quantity of fluid in the end of the tube in A, just above 

 and below the surface, becomes specifically lighter than that 

 in the other end in B ; this (almost imperceptibly at first) de- 

 stroys the equilibrium ; the tube commences to act as a siphon ; 

 a small quantity of warm fluid is drawn higher in the tube, 

 and cold water descends from the other end, which causes the 

 water in B to flow into A through the connecting tube D ; 

 this further destroys the balance ; and the circulation will nolv 

 go on with a sort of rapidity that could hardly be anticipated ; 

 particularly if the descending part of the Thermosiphon be 

 kept as cold as possible. That is, it will be in proportion to 

 the quantity of heat abstracted or given out. 



As a considerable quantity of air is given out by water 

 when it is first heated, some of this air lodges in the upper 

 part of the Thermosiphon, where it expands, and often stops 

 the circulation. It will be necessary, therefore, to re-fill it in 

 the way already described, viz. to stop F and F^, unstop G, 

 and fill the Thermosiphon with the sort of fluid already used : 

 also, when the height of G approaches fifteen or twenty feet 

 above the level of the water in the vessels, and the water in 

 the boiler is at 1 80° or 200°, steam may collect, or be formed, 

 in the higher part, and expand, which will, after some time, 

 prevent the circulation. This may partially be remedied by 

 pouring a small quantity of oil into the cock G when nearly full, 

 so that it may cover the water in the tube with a thin film. 

 This film will always swim on the surface of the water, and, 

 in some degree, prevent its being converted into an elastic va- 

 pour. These unavoidable imperfections are, however, very 

 trifling, when it is considered that the operation of filling is so 

 much simplified by the use of the cocks F F' and G. In fact, 

 this operation requires only the most common attention. It 

 is done in less time than one minute ; and this might not be 



