that do not burst with Frost. 247 



was not the case when it was put in the freezing-mixture. 

 The most noticeable thing, however, was the fact that all the 

 unburst pipes had become good water-hammers, and this 

 showed that leakage could not have occurred. The screws 

 were removed, the pipes filled, again screwed up and refrozen, 

 and this was repeated till all were burst. The round lead and 

 the thin elliptical composition pipe burst at the third freezing, 

 and the elliptical lead pipe at the sixth. Judging from the 

 fact that it required three freezings to burst the round pipe, 

 one might be led to suppose that a round pipe would last 

 equally long under ordinary conditions, which is certainly not 

 the case. The reason is that, under ordinary conditions, a 

 greater length freezes at a time and more slowly; and a slight 

 inequality arising, the expansion from a greater volume of 

 water is concentrated on the weaker places, which therefore 

 give way during the second, if not the first, freezing of the 

 water. 



It might be thought that, as the outer layers of ice are 

 below the freezing-point when the pipe is being cooled, they 

 would not act as a plastic body and accommodate themselves to 

 the changing form of the pipe; but no doubt can remain as to 

 their behaviour in this respect when the pipe is cooled in air; 

 for in the freezing-mixture, where the rate of cooling must be 

 much more rapid, such accommodation takes place perfectly, 

 even thin composition pipes changing their form and becoming- 

 round. The apparent plasticity of the ice may depend on 

 fracture and regelation; for if the outer layers are below the 

 freezing-point, and a bursting-pressure is brought to bear on 

 the compound pipe formed of lead and ice, it might yield, the 

 lead bending and the ice cracking, and so allowing the water 

 to penetrate the cracks and freeze in them. Whether this 

 action takes place or not does not much matter : the result, as 

 in the somewhat different case of glacier-motion, is much the 

 same. 



If the pipe is made of such a form that it will not become 

 round till it has been completely frozen, say, three times, it 

 will take a great many frosts to burst it, as those parts that 

 do not freeze easily will protect the more exposed portions ; 

 so absolute security may be relied upon till ordinary round 

 pipes have burst once or twice ; and then the now nearly round 

 ones may be squeezed back to their original form. The choice, 

 then, is between two evils ; either burst pipes, with the usual 

 damage and cost of repair, or the trouble of inspection every 

 second or third time that the neighbours find that the " thaw " 

 has burst theirs. 



"We thought it possible that iron might be sufficiently elastic 

 X2 



