HALIFAX WATER WORKS. — JdHNSTON. 105 



case of the 2T-incli, a leak (level(ii)e(l (Inrinii,' the winter follow- 

 ing the laying of the pipe, and on digging down to the main a 

 joint was discovered to have drawn out about ^ <d" a;i inch. 

 This was canlked with cold lead and gave no trouble until the 

 following winter, when it agai.i shuwed signs of leaking, and on 

 investigation the joint was found to have drawn another ^ inch. 

 The blocking of the pipe^ on each side had apparently not 

 settled out of ].)lace, being laid om the to]) of the ledge rock. It 

 was thought that this drawing apart of the joint might have 

 been due to the contraction of the pi])es. Assuming the differ- 

 ence of temperature to have been 30 degrees, which is a fair 

 estimate between the tem^xM-aturc of the ])ipes when laid ami 

 when the leak developed, the contraction to opea the joint ;{ 

 inch would have to take [ilace through 324 feet of pipe. If this 

 -took place on each side of the defective joint there would be a 

 strain on the joint of over !('» tons, the pipes weighing a ton 

 and a quarter to the 12-foot length, if the leakage was from this 

 cause, it should close up again in the sunuuer when the tem- 

 perature of the pipe rose. I'nfortunately, the | inch which is 

 said the pipes separated in the second winter was not measured 

 accurately, but was estimated by the foreman and may have 

 been overstated; but assuming it to be correct, the writer can- 

 not advance any theory for the increase in the opening from 

 this cause, as there would not be any more difference in the 

 temperature than the amount given above. It is possible that 

 the jorat may not have been driven home, and as at this ])oint 

 there-, was only about four pounds pressure when testing, the 

 oxide paint used may have prevented the leak showing whea 

 the pipe was tested on being biid, and a settlement may have 

 occurred in one or two lengths of jiipe distant Ironi the leak 

 dragging the ])i]>e apart at the weakest point. 



It will be seen from the descri])tion of the method of laying, 

 that the process of lowering and blocking is exactly the same as 

 for ))lain ]ii])e, excejit xh" ramiiiing hoin(\ A\liich lake^ but \'erv 

 little mf)re time than the extra care required in centering the 

 pi])c for a. Ic^ad joint and then the joint is complete; whereas, 



