354 
label mouldings occupy the place of these bands) ; these 
bands lap over each other, so that the united or crossing 
band at the level of the springing is somewhat increased 
beyond the width of a single band. The width of these 
bands being taken out of the diameter of the enclosing circle, 
prevents the possibility of the pointed arches being other- 
wise than acute : let me remark that only two pointed 
arches are included under the enclosing semicircle. This 
semicircle is shown by a darker tint. 
The next example is from Norwich Cathedral, and shows 
still more forcibly the correctness of my position. Here 
the width of two bands (nearly) is taken from the enclosing 
semicircle, so that the pointed arches are more acute than 
in the preceding example. (Plate 1, No. 4.) 
The last example is from the Church of Saint John, 
Devizes. Here three pointed arches are included under 
the enclosing semicircle, leading to greater intricacy in the 
interlacing, and rendering the pointed arches extremely 
acute. (Plate 1, No. 5.) The properties of circles being 
alike, whatever their diameter, an useful key is here given 
to all the varieties of multiple windows having the same 
springing line. 
I am not aware that these arcades have ever been viewed 
in a similar light to that in which they are now presented. 
They have been studied synthetically, but not analytically, 
(i. e. by considering the various forms obtained under one 
enclosing semicircle ) ; and the important inferences which 
I shall hereafter draw from the peculiarities appertaining 
to these arcades, will, I humbly hope, induce others to 
direct their attention to the subject. It is clearly the duty 
of every architect (now that the study of Gothic Architecture 
has happily revived) to think for himself, to investigate for 
himself ; to respect the opinions of his predecessors and his 
elders ; but not to rest satisfied with an acquaintance with 
