architectural effect aimed at were greater, we find two or more such
orders as the above, with, perhaps, columns supporting the outer ones;
the receding orders, in either case, doing away with undue obstruction
of light or view; the sill always well sloped, to throw off the water,
and having usually a string-course below, to prevent it from running
down and discolouring the walls. In all this, strict regard to practical
reason and utility is manifest; every step is argued out on the basis of
construction and requirement, and every decoration is founded on, and
results from, the conclusions come to on these practical grounds.
In domestic architecture, if a window were beyond the width of a single
casement, a small pillar was often interposed, and the inner order of
the window was divided into two arches, while the outer one, if there
were any, was in one, the casements or shutters falling into rebates in
the back of the column, by which a window of double width, which would
not otherwise be conveniently attainable was produced. In window-like
openings in which glazing was not needed--as in triforiums, cloisters,
and screens--this system was used for beauty where not demanded for the
same reasons as in windows, and the subdivisions were often increased to
three or four under one comprising arch.
[Illustration: Fig. 144.--St. Trophimus, Arles. Cloisters, north side.]
In other instances of the same kind, where light arcading was needed, as
in cloisters, and the wall was too thick to rest upon a single capital,
two small columns were placed one behind another, or a sort of bar or
double corbel placed on the capital of a single pillar to support the
springer of the arch, for the sake of avoiding the use of thick piers,
which were not needed for strength, and would obstruct view and light;
and all these practical contrivances were made elements of beauty and
varied effect.
[Illustration: Fig. 145. Priory Church, Bridlington. Part of remains of
Cloisters.]
Another legitimate exercise of reason on the part of the Romanesque
builders, was the rejection of the fixed rules of proportion observed by
the ancients between the diameter and height of their columns. These
rules were good in their place, but they had been worked out for a
totally different system; and we know that the ancients themselves were
anything but as slavish in their adherence to them as their modern
imitators. In a purely arcuated system, however, it became clear that
such rules were out of place and inconsistent with reason.
Circumstances, in a majority of cases, prescribed the height of a
column, from reasons wholly irrespective of the question of its load. It
followed, then, that the diameter must be regulated rather by the load
than the height, so that every variety of proportion became admissible.
Take, as an example, the crypt under the choir of York minster. Its
height being prescribed by circumstances, and the portion of it required
for the vaulting being fixed by the width of the arched bays, it
followed that the height of the columns was also rigorously defined; but
some of these columns had to carry those of the church above, and with
them the whole superstructure, while others had no load but the vaulting
of the crypt and the floor of the church. Surely, then, the simplest
exercise of reason dictated that their diameters should vary with their
load, irrespective of their height. The system of clustering columns
both helped to moderate the extremes of such variation in proportion,
and, at the same time, introduced still wider liberty; for, though a
pier destined to carry a vast load might be subdivided, and its apparent
proportions thus lightened, the individual shafts of which it was
composed, not having each its own proper load, might be viewed as
decorative only, and be made exceedingly thin for their height. The use
of such thin shafts did not, however, originate in the Middle Ages.
Canina shows in his work on _Domestic Architecture Decorated with
Ornaments of a Light Form_, that it was frequent among the ancients,
though not often adopted by modern Classic architects. Even for really
constructive pillars it is admissible where the material is of
remarkable strength, as in the case of metal columns, and, in a less
degree, with those of marble or granite where the load is very small;
but it is especially so where the columns are of a decorative rather
than a functional character, in which case it is not only lawful, but
correct, to show this by making them of slender proportions. The
liberty, however, which I here defend, must, as all other liberty, be
kept within reasonable bounds, and must be regulated by a correct eye
and sound judgment.
Another sound exercise, as I think, of reason and liberty, which was
universal among the Romanesque and Byzantine architects, was the
departure from the rule of the ancients that all capitals and other
recurring objects of a like nature should be worked to one and the same
pattern. It may be that the unity of a colonnade, united by a single and
unbroken entablature, demanded this. I am not finding fault with it in
Grecian or Roman architecture; but where the capitals are separated by
arches, or did not form a continuous range at all, the effect would be
most painfully monotonous if the sculptured capitals were all alike, as
if cast in a mould by the hundred. We accordingly find it established as
a universal law that, though moulded or other mechanically-formed
capitals might, if you please, be alike, no such slavery should be
imposed upon the sculptor; but that he should have the fullest scope,
within the reasonable limits suggested by the requirements and the
general balance and harmony of mass and outline, for the freest exercise
of his own imagination.
Now, though these and other developments of the Romanesque period were
founded on a thoroughly practical and logical course of reasoning, it by
no means follows that a perfected form of arcuated architecture had yet
been arrived at, any more than that the decorative system had been
brought into a thoroughly refined or artistic form.
Towards the middle of the twelfth century the efforts of the architects
were redoubled towards the attainment of these two objects; and the
advancement made, both in correcting defects in construction and
refining the decorative system, were most strenuously followed up, and
all improvements made were founded strictly on reason. The great
constructive difficulty met with arose from the powerful outward
pressure of the round arch when of great span or carrying any great
load, and especially so when used in situations where it was difficult
to give it any very massive abutment.
The cases of failure from this cause were most frequent; so much so,
that besides the numerous instances recorded of buildings wholly or in
part falling from the failure of the arches, we find among the buildings
still remaining abundant evidences of the insufficiency of the round
arches for their load, and of the abutments to resist their pressure. In
ordinary architecture we cannot, as in bridges, viaducts, etc., give our
arches an unlimited abutment proportioned to the pressure, whatever it
may be; we are limited in our means of doing this by innumerable causes:
thus, in a central tower, if the arms of the cross have aisles, the
natural abutments of the tower arches are reduced to the frail aid of a
continuous arcade upon detached pillars; and even if there are no
aisles, the abutting walls are perforated with windows. The abutments,
again, of a chancel arch are perforated either by arches or windows,
while the gable over the arch loads it heavily at its weakest point.
The abutment of an arch, again, has often to impinge upon a pier at half
its height, as in the case of a nave arcade abutting upon the detached
piers of a central tower. In all such situations the undue pressure of
the round arch was found to be most prejudicial. Still more strongly was
it felt where the nave was spanned by stone vaulting. The Romans had got
over this, as in the Baths of Diocletian, by breaking the continuity of
the aisles by vast abutting walls across them. But in a church this was
impracticable. Its uses demanded continuity of aisle and moderation in
the size of the pillars. Failures often occurred from these adverse
causes, and the ingenuity of the architects was naturally directed to
obviating the defect.
I have, in a previous lecture, described the series of tentative
experiments, all of them dictated by constructive and practical
requirements, by which it was attempted to avoid these difficulties. I
will not weary you by recapitulating them. The two obvious desiderata
were an arch of less pressure and an abutment of greater resistance; and
these were the two objects aimed at in most of the succeeding
developments. The first demand was met by the pointed arch; the second
by the systematised use of the buttress, whether of the solid or arched
description. It was perfectly well known that the outward thrust of an
arch diminished as its height increased; that the resisting power of an
abutment depended mainly on its extension in the direction of the
pressure; and that where sufficient extension of abutment could not be
obtained without inconvenience or dissight, the deficiency might be
compensated by loading it from above: and by arguing on these three
facts the constructive characteristics which distinguish Gothic from
Romanesque, or the pointed-arched from the round-arched Gothic, were
logically worked out.
The strictly mathematical mode of increasing the height of an arch
would, I suppose, be by using a semi-ellipse, its major axis being
vertical. The form is, however, most unpleasing to the eye and
troublesome in execution, from its constant variation of curvature, so
that by far the most natural and practical means of effecting the object
is the adoption of an arch of two centres, or what is commonly called
the “pointed arch.” We accordingly find, as I have shown by ample
evidence in a previous lecture, that this form was in the first instance
used just in those situations in which a reduction of outward pressure
or an increased power of bearing weight were of the greatest importance.
I have shown that this form was not adopted at first as a matter of
taste, of fashion, or of fancy; nor even, as has been suggested by a
highly talented writer, as a means of meeting the difficulties arising
from the varied heights of the arches of vaulting, but simply from
_structural_ and _mechanical_ necessity. It matters not whether the form
was new or old, whether it occurred to them without external suggestion,
or whether they saw it in the East, in their own intersecting arcades,
or in the first proposition of Euclid. It was not the seeing of it in
any such manner which caused its introduction, but the simple fact that
they had arrived in the course of their constructive development at a
practical problem of vital importance, which _absolutely demanded_ the
pointed arch for its solution.
The first situations in which the pointed arch was substituted for the
semicircle are the wide spanning arches of vaulting and the arches
carrying central towers and gables. We next find it in the wide arches
of nave arcades; and it is not, as a general rule, till it became
customary in those positions where it was demanded for practical
reasons, that it began to be used as a matter of taste in other
positions.
Having secured the first object--an arch of reduced pressure--the
second, viz., the abutment of an increased resistance, was attained by
the systematic development of the buttress--a feature very much
neglected by the Romanesque builders; and, as the vaulting of a lofty
nave could not be directly supported by the ordinary buttress, the
arched or flying buttress was introduced, spanning the aisles and
conveying the pressure to the buttresses beyond. That this was
introduced for utility only, and not from taste, is proved by the
attempts in early instances to conceal it; so that we may with certainty
conclude that all these beautiful features of Gothic architecture
originated not from taste or caprice, but from reasoning upon practical
and urgently pressing constructional requirements, and that the beauties
to which they gave rise proceeded from the application to them of the
great principle of Gothic architecture, the decoration of constructive
or useful features.
Let us, however, suppose for a moment that our building is not vaulted,
but has timber roofs; there still remains an advantage in the use of the
pointed arch. If it has, for instance, a central tower, the demand for
an arch of reduced thrust is still greater than if the church had been
vaulted, for the arms of the cross, from their reduced weight, are less
effective as abutments.
The chancel arch, again, demands height, and the more so if it be wide,
as in our own day is necessary. The nave arcades are better pointed than
round, as are any others carrying any considerable weight. Buttresses
remain necessary at the ends of the arcades, and are desirable as a
steadiment to the outer walls, particularly where roofs without a direct
tie are made use of, and are further useful as permitting the
introduction of larger windows than might be safe without them. In all
cases, indeed, where roofs or floors are so constructed as to
concentrate pressure upon points, it is clear that buttresses are
desirable; and when the efficient size cannot be given them without
inconvenience or dissight, it is equally clear that the deficiency may
be readily compensated by loading them with lofty pinnacles. It is wrong
to use buttresses without any object but appearance, but there are
numbers of cases where they are of great advantage, besides those in
which we know them to be indispensable. If so many of our arched and
vaulted buildings in these days were not mere pretences in lath and
plaster, we should have more practical experience of the need of the
buttress and of the pointed arch. I was once told by the English
Commissioner in Scinde that the European engineers had difficulty in
making the native builders there believe that any but a pointed arch
will stand.
Let us now inquire as briefly as may be into the _rationale_ of _ribbed_
vaulting as distinguished from the _arris_ vaulting of the Roman and
earlier Romanesque builders.
A groined vault does not of absolute necessity demand the use of ribs
any more than the plain waggon-head vault. Even the latter was from an
early period frequently divided into compartments or bays by transverse
ribs, which were useful as a means of giving it rigidity; but in
_groined_ vaulting these were of nearly constant use, both for the same
reason, and because the vault, being reduced at its springing to so
narrow a footing, required this additional strength. The arrises,
however, or diagonal lines of intersection, were always left without
ribs.
Why, then, was the custom changed? For two important reasons. The first
was this: that the intersection forms naturally a feeble line, both from
the difficulty, particularly with the rough materials usually employed,
of making its construction sound; from its forming an arch of greatly
increased width without corresponding increase of height: and from its
reduction at the springing level to a pin’s point.
The second was of a more intricate nature, and requires to be explained
more in detail. When the two intersecting vaults of a groin are similar
and equal in their section, or when the section of one is the
mathematical resultant of that of the other, the line of intersection
falls in a _plane_. When vaulting, however, became general, all sorts of
irregularly-formed spaces would have to be so covered, and would present
problems of considerable difficulty, in which it would be impossible in
all cases that the vaulting surfaces should be portions of cylinders or
regular cylindroids, and in which the intersecting lines could not,
without much twisting of the surfaces, be brought to fall into planes.
The introduction of the diagonal rib met both of these difficulties. It
strengthened the weak angle and gave it a substantial footing; and it at
the same time gave to the lines of intersection a certain degree of
independence of the vaulting surfaces; so that, instead of the surfaces
governing the intersection, they were thenceforth governed by the ribs,
and the latter could be made to fall into planes, and to avoid unsightly
forms even in vaulting spaces of the most irregular and abnormal forms.
The substitution of the rib for the arris worked as great a revolution
in the principles of vaulted construction as did the pointed arch
itself. Nothing in the way of vaulting was now impracticable or
unsightly; the architect was absolutely master of his work, and could do
what he liked with it. The facilities it offers are quite marvellous in
the eyes of the modern practical man when once they are opened to them.
I have myself found one of the most practical men I ever met with, who
had for years taken the leading management of the business of the
greatest builder of our day, though hitherto uninitiated in Gothic
construction, almost in ecstasies at finding a difficult problem in
vaulting he had been puzzled over for days and making models of in vain,
solved in an instant by seeing the absolute liberty of action exercised
in a similar case in Westminster Abbey. The old builders themselves
perfectly luxuriated in their newly-discovered liberty: not only could
they vault spaces of any conceivable plan, every dimension of it
varying, and the difficulties increased by the necessity of pushing up
windows in its sides in all kinds of difficult positions, but they could
make the result so pleasing and apparently so straightforward and
natural, that not one observer out of a thousand ever finds out that
there was any difficulty to be got over at all. Sometimes, indeed, we
find them rejoicing so much in their freedom as to set themselves
needless puzzles for the very luxury of solving them. There is a most
remarkable instance of this in the crypt under Glasgow Cathedral, where
the pillars which support the floor have been placed in a variety of
intricate positions for no reason, apparently, but to produce curious
perplexities in the vaulting and create strange problems, for the mere
pleasure to be derived from their solution and the beauty of the puzzle
when solved.[59]
It has been argued that the Gothic vault is less refined than some of
the previous forms, because less strictly mathematical; that a refined
system of construction should in all cases possess an exact mathematical
solution, though the builder may, when once master of the true theory,
depart from it in execution; that the work, in short, though irregular
in execution, should be perfect and mathematically accurate in its
theoretical type.
I agree with this doctrine in the main; but I hold that the Gothic vault
complies with its conditions.
The square groined vault, with semicircular arches, is perfect in its
theory, and gives elliptical arches for its arris lines. The same, if
vaulted with the pointed arch, is equally true in theory, for the
diagonal ribs may be pointed arches, formed each of portions of two
ellipses. The oblong vault, again, is perfect if the wide arch is a
semicircle, the narrow one a vertical semi-ellipse, and the arrises
horizontal semi-ellipses of the same height; but the ancients generally
chose to stilt the narrow arch instead of using the vertical ellipse,
and by doing so threw the diagonal arris out of the plane and out of
shape; but the theoretical form remained, nevertheless, perfect. In like
manner, if the same figure be vaulted across its widest span by a
pointed vault, and if the narrow vault have a pointed arch composed of
two portions of ellipses, and the intersections be of the same figure as
resulting geometrically from the intersection of the two vaults, the
theoretical form is perfect. Now, if in either case the architect thinks
the elliptical pointed arches inferior in beauty to those composed of
parts of circles, and by using ribs finds himself enabled to throw the
error resulting from the substitution of the latter form into the
vaulted surfaces where it will be invisible, surely he is only using
that discretionary power of introducing irregularities upon a perfect
theory which is claimed as his right; and this is exactly what the
Gothic architects introduced.
The fact is that, besides its unpleasing form, especially when the major
axis is vertical, the use of the ellipse entails such an annoying series
of difficulties as greatly to increase the trouble and consequent cost
of execution. The constant change of curvature, the troublesome methods
of striking it, and of finding the true lines of the arch-joints, not
to mention the mathematical fact that the same joint line is never true
both for the extrados and intrados, and that, if the rib-mould remains
unchanged in depth, the extrados and intrados cannot be both true
ellipses at all; all these furnish quite sufficient practical reasons
for its rejection in cases where not only is there no necessity but an
abstract mathematical idea to be satisfied by its use, but the beauty of
the work is greatly improved by dispensing with it.
Though the pointed arch was introduced from purely constructive reasons,
there was another of a more æsthetical nature, which rendered its
adoption more general when once introduced. It was a double one; not
only did the general tendency towards lofty proportions render it
necessary to make use of an arch more in harmony with the general
feeling of the architecture, but the rejection of a fixed code of
proportions for pillars and other parts demanded for the arch an equal
power of varying its own proportions. The semicircular arch is absolute
and invariable, and though the use of smaller segments would meet the
case in one direction, there were no means of proportioning it to
features of _increasing_ height. This was attempted both in Romanesque
and Byzantine works by the expedient of _stilting_, but this is, after
all, more a semblance than a reality. As in cases already cited, the
mathematical solution of the problem is the ellipse; but only imagine
anything so unpleasing as a series of elliptical arches placed the
length-way upwards! Good taste would not suffer it. But the pointed arch
at once met the difficulty. To illustrate my meaning, I will beg you to
take an internal bay of a Norman cathedral (Fig. 146), and to suppose
yourselves to have to increase its height throughout in the ratio of
one-third (Fig. 147).
[Illustration: Fig. 146.]
[Illustration: Fig. 147.]
You first, after setting out your widths as in the original, increase
the whole height and that of each storey by one-third; you then increase
the pillars and the jambs of the triforium and clerestory windows in the
same proportion: this brings you to a stand, for the arches, being
semicircles, are invariable. Either you must leave them unaltered and
throw all the extra height into the wall above them, or you must stilt
them each to the extent of one-third of their height unless you can make
use of an elastic arch which will change its proportion at pleasure.
The ellipse occurs and meets the case, but it offends your eye. At
length, however, the pointed arch suggests itself, and gets rid of the
whole difficulty. So similar are a Romanesque and an Early Pointed bay
in all other respects, that the change of proportion which I have
described seems at once to effect the whole change in style.
Had the constructional motive alone existed, the pointed form would have
been confined to arches of considerable span; but the demand for a
variable arch adding æsthetic to the constructional claim, caused its
speedy adoption in positions where strength alone would not have
demanded it, though the semicircle, the plain segment, and the segmental
pointed arch, were, at all subsequent periods of the style, used side by
side with the true pointed form.