364 
" was attributed to error and defective workmanship, but it 
" has since been the cause of solid content and admiration." 
No such undulation occurs in the main rib of SaUsbury, 
but the circumstance has been referred to as a matter of 
interest in connection with the subject before us. 
The main rib of the vaulting is shown in plate 3, No. 8, 
the lines ABC, converging to the point D, are the subsi- 
diary ribs forming the groins, and dying into the pendant D, 
which is attached to the wall E E, (this is the clerestory wall 
E, on the section over the columns F, which separate the 
nave from the aisles.) It is against this wall E that the thrust 
of the vaulting is directed, and concentrated on the point D ; 
this thrust would necessarily drive out the wall, were it not 
carried forward by the flying buttress G over the aisle roofs, 
and made to die into the wall buttress H. 
The principle of the vaulting, and the use of the flying 
buttress, may be rendered more familiar by a very homely 
illustration. Let us presume that the several converging 
lines ABC are the ribs of an umbrella, the point D the 
point at which the snick or pin is introduced. Now so long 
as this pin remains fixed, the umbrella is expanded ; the 
moment it is forced in by the hand, the umbrella collapses. 
I have simply introduced this illustration as a popular one, 
but must admit that the thrust of the vaults is in a different 
direction, which can only be withstood by corresponding re- 
silience, and this is aflforded by the flying and wall buttresses. 
In order more clearly to understand the efiective character 
of the flying buttress, as a medium of communicating the 
thrust from the vaulting to the wall buttress, let us refer 
to the annexed diagram. 
