HOT WATBH, 



313 



siphon, by taking a piece of lead pipe, say of half an inch 

 bore, and four or five feet long, bending it like a siphon, 

 but one leg a good deal more bent than the other, in order 

 to give the descending water time and space for giving out 

 its heat ; and then, filling this tube with water, and placing 

 one hand on each end to retain it full, immerse the extre 

 mities in a pot of water over a fire, as represented in the 

 annexed diagram. Supposing the water of j,. 34 

 a uniform temperature in both legs of the 

 siphon, no circulation would take place ; but 

 supposing it to cool sooner in the long leg 

 than in the short leg 5, then the equilibri 

 um would be destroyed, and the water in 

 the long leg a would descend, and draw up 

 water through the short leg b . and this cir 

 culation would continue as long as the water 

 c was maintained at a temperature above 

 that of the surrounding atmosphere,&quot; 



Mr. Rowley s adaptation of the siphon is one of the 

 simplest and most efficient that has been proposed. In 

 Fig. 35, a c e are the two legs of a siphon, through the 

 upper of which the heated water ascends, and by the lower 

 descends. Immediately over the descending bend, a pipe 

 connected with an air-pump is inserted, in order to fill the 

 pipes, or remove the air which collects in the superior 

 iiinb. Instead of the air-pump, a funnel with air-tight 



Fig. 35. 



