143 On the Freezing of Water in leaden Pipes, 



the metal itself, which prevents them from bursting at all 

 times. Thus they can only allow of expansion from con- 

 tained frozen water a limited number of times ; and old 

 pipes, and such as may have been frequently repaired by 

 soldering, must be expected to give way in every rigorous 

 season. 



The pipe being new when first laid down, the low tem- 

 perature of the succeeding winter occasions a congelation 

 of the water throughout the whole length of the tube ; an 

 expansion every way ensues by this process of icefaction ; 

 the ductility of the metal allows of an enlargement of its 

 diameter throughout its whole length ; its thickness is pro- 

 portionally diminished ; a milder temperature ensues, and 

 a thaw subsequently takes place in the tube without occa- 

 sioning damage. A second, third, and perhaps more win- 

 ters thus pass on ; the water each time frozen in the pipe, 

 each time enlarging its diameter ajid diminishing its thick- 

 ness : at last, the diminution of this latter renders it no 

 longer capable of withstanding the force of internal disten- 

 sion, and the succeeding frost occasions a rupture in that 

 part of the tube where the resistance may prove the least. 



If the. tube had been constructed of a more refractory 

 metal, as iron, &c, the first expansion would have burst 

 the same; while it is obvious that the durability of a leaden 

 pipe must be very precarious after the first occasion for re- 

 pairing a rupture therein : for, in easting a pipe, if the cool- 

 ing were slowly and uniformly conducted (a precaution not 

 Bt all attended to in their manufacture), its texture and duc- 

 tility would be the same throughout, and we should have 

 good reason to conclude that the diminution in thickness 

 throughout the pipe, by all the previous expansions of the 

 freezing water, would be alike ; and the necessity of repair- 

 ing a rupture in one part would be a certain indication of the 

 extreme thinness of the whole, and its incapability of un- 

 dergoing any further distens'ion. But as the tube is never 

 -uniformly cooled, but cast in small lengths of two to three 

 feet, each of which is allowed to cool as speedily as the 

 temperature of the surrounding air will admit, and after- 

 wards joined together to form one pipe of any required 

 length, there may be certaiu portions the ductility of which 

 •mav better allow of progressive distension than others ; and 

 I add, that the mode of cooling a pipe after casting the 

 fame, were it attended to, would have considerable influ- 

 ence on its ductility ; instances of which we have in every 

 substance capable of fusion (whether metallic or not), 

 whereby the regular annealing of the same, or cooling them 

 4 in 



