78 THE SEVERN TUNNEL. 



or by pipes to let it through the brickwork. It can readily 

 be understood then that in building in the successive rings, 

 that put in first against the rock would bo the roughest and 

 most pervious to the water, the next put in would be better, 

 and the third better still ; the conditions under which each 

 successive ring is put in being constantly more favourable 

 the further it is removed from the water. The natural 

 result of this would be that the rings, as a rule, were least 

 leaky towards the top of the invert, and most leaky towards 

 the rock bottom. 



In the case alluded to, after the work liad been put in 

 two or three months, when the cement was supposed to be 

 set, the pipes were corked. The water pressure came on 

 the brickwork and had reached a pressure of about GO lbs., 

 when in certain places the top ring of the invert was blown 

 up. After the top ring was gone the second was blown, and 

 after that the tliird was rising, when the contractor opened 

 the pipes again and took off the water pressure. 



The explanation appears to be simple. The top ring in 

 these places was the tightest or least leaky of all the six 

 rings put in, and therefore the water which came with com- 

 parative ease through the lower five rings was checked by 

 the top ring, and nearly the whole of the water pressure 

 came upon that one ring, which was only 4| inches thick — 

 a pressure possibly of nearly two tons on the square inch — - 

 and this was more than that one ring could bear ; the cement 

 joints yielded each a little (for no crushed brick was found), 

 and the ring was blown up. The same process was repeated 

 at the second ring, and afterwards the third was following 

 suit when the pressure was taken off by uncorking the 

 pipes. 



This only occurred in certain parts of the invert ; in the 

 other places, no doubt one of the deeper rings might have 



