[ 263 ] 



rancan paiTagc, in which the boats unload 

 in Cajllc-Jield. The contrivance and de- 

 fign of thefc doors are admirable, but as 

 many of them will occur in this tour of 

 the navigation, it will not be improper to 

 explain the conftrudion of them here. 



In the fketch, Plate IV, Fig. 4, A re- 

 prefents one fide of the channel of the 

 canal, being walled ; B, is the floor of it j 

 C, C, are two doors, fixed in the pofition 

 they appear in, and turning on hinges at 

 bottom, d, dj d, d. The doors have a free- 

 dom of rifing, but cannot fall lower ; now 

 it is obvious from this, that in cafe the 

 bank A, breaks, and the water ruflies out 

 of the breach, the decreafe of prefTure on 

 the doors C, C, will raife them up at once 

 to e, e, (where they will be fixed againft 

 projeding irons made for that purpofe,) 

 as the water at F, F, will then natu- 

 rally force them up. The confequence 

 of which is, the lofing no more than the 

 water contained between the doors i and 

 if the bank fhould break at F, yet the 

 quantity of water loft would be but trifling. 

 I have thrown the doors near one another, 

 for your fully underflanding the principle 

 upon which they are defigned ; but in the 

 S 4 canal 



