for Preventing Water- Pipes Bursting during Frost. 241 



I therefore procured a globular glass vessel of these dimensions, 

 accurately fitted with a cylinder and piston, and filled it with water. 

 On exposing it to freeze, however, the water expanded equally on all 

 sides of the vessel and burst it, instead of elevating the piston. A thin 

 sheet brass vessel I employed for the same purpose succeeded no 

 better. But, after a number of similar experiments, on vessels of 

 different shapes, which I need not detail, I ultimately found that a 

 true and accurately bored cylinder was the only vessel in which water 

 could bo frozen under pressure. This it was of great importance to 

 ascertain ; for had it occurred that ice expanded uniformly in a 

 lateral direction it is plain that no vessel of moderate strength could 

 have resisted the pressure. From these experiments I found that 

 in a simple tul)e of any convenient dimension, fitted with a piston, it 

 would bo, without bursting, elevated by the freezing of the water it 

 ^•ontained, and with a force far beyond what was absolutely necessary. 

 This could be applied to shut a cock, or operate in any other manner 

 found to be most appropriate. Such is the principle of the apparatus 

 represented. 



The only apparent objection that presented itself next, was, that 

 the water contained in the pipes being subjected to the same degree 

 of cold as that portion confined in the small tube, would freeze as 

 soon, or at all events so nearly at the same time that there would not 

 be space intervening of sufficient duration to allow of the water being 

 shut off. This objection, however, is easily disponed of by the ap- 

 plication of a tube possessed of superior powers of conducting heat 

 to that of the lead pipes. Copper at once suggests itself as the most 

 appropriate, and for two reasons : \st. Copper is to lead, as a 

 conductor of heat, as 18 to 89, or, as nearly as possible, 5 to 1 ; and, 

 2(i, Its tenacity or cohesive power is exactly ten times that of lead* 

 These, it will be seen, are valuable properties for the purpose, since 

 it not only guarantees a quicker conduction of heat, or, in other 

 words, a quicker freezing of the water it contained, but a power of 

 resisting ten times the pressure. But there is no occasion to employ 

 a copper tube of the same dimensions as the lead pipe. One of half, 

 or even a fourth of the size, I have found to answer quite as well. 

 This, again, gives a still greater advantage, since the conduction of 

 heat holds a direct ratio to the mass of the body. In short, it is 

 very apparent that a properly-constructed copper tube, containing a 

 small portion of water, will be the first to freeze, and so shut off the 

 water. 



The practical application deducible from these principles, consists 

 of a very simple apparatus. A copper tube of any convenient dimen- 

 sions fitted with a piston filled with water, and firmly attached to 

 the lead pipe, would, from its superior conducting power, be the 

 first to freeze, and by the vertical expansion of the ice to elevate a 

 piston and shut a stop-cock, while the water in the lead pipes was 

 still fluid. On similar principles it would be the first to become 



VOL. LI I. NO. CIV. — APRIL 1852. , Q 



