1898.] on Bringing Water to Birmingham from Wales. 687 



Ijensation water for another 100 days. In this way a drought of 

 180 days is provided for, the water for supply during the 100 days 

 coming from the Pen-gareg and Craig-goch reservoirs, higher up the 

 Elan. They hold together 3330 million gallons, and are therefore 

 fully competent to ensure this. 



The water darkly shaded on the diagram above the submerged 

 dam and below 782, cannot of course be counted as effective storage, 

 as it cannot be drawn down without leaving the aqueduct inlet high 

 and dry, but it will of course be in no sense stagnant, because the 

 quantity going to Birmingham must always be running through it. 

 When the second and following instalments are required for supply, 

 the reservoirs on the Chierwen will have to be made in succession 

 as required, and the addition of the water obtainable from them 

 will enable the 40 feet " slice " between 822 and 782, which they 

 will always be rei:)leting, to maintain the increased delivery by way 

 of the aqueduct and the com25ensati()n as before, leaving the 2565 

 millions below 782 for the last 100 days of the drought. In order 

 to delay as long as possible the making of the Claerwen reservoirs, 

 a tuimel 1^ mile long is to be driven from the Dol-y-mynach 

 reservoir on that river to above the submerged dam, so that its 

 natural unstored waters can be used for supply, the respective levels 

 at each end admitting of this being done comfortably. 



In this country Imudreds of impounding reservoirs have been 

 constructed for the storage of water for canal purposes and for town 

 sup2}ly, and a very large majority of these have banks of earth sup- 

 porting an internal wall of puddled clay, which forms the watertight 

 part of the barrier. 



There are still only very few stone dams of any great size in 

 England, although many are to be found on the Continent of Europe. 

 The Ehan and ('laerwen valleys were, however, peculiarly adapted 

 for such structures, the dam sites being all on rock practically to the 

 surface, and plenty of stone for building at no great distance, the 

 material for earth banks being, on the other hand, deficient. 



It may be interestino to show a cross-section of one of these stone 

 dams, and on Diagram No. 6 you have the Caban Coch which we are 

 now building, and alongside it that of the Bouzey dam, near Epical, 

 in France, which failed about three years ago with very disastrous con- 

 sequences. I invite you to compare these two profiles, and note the 

 relative thickness of the walls at the same depth below the water sur- 

 face, which, of course, determines the pressure. In this dam (Bouzey) 

 the line of stress, instead of falling within the middle third of the 

 profile, as it ought to do, was very much nearer the down-stream face at 

 the point of failure ; the weight of the structure was under 130 lbs. 

 per cube foot, and neither the stone nor the mortar of which it was 

 built was of good quality. The failure was no doubt due to the fact 

 that when the reservoir was full the water face of the wall at the 

 point of fracture, owing to the improjjer form of cross-section, was sub- 

 jected to a tensile strain which the material was not competent to boar. 



