is seen on the capital shown in Fig. 197, the eggs of which are amongst
the largest known. They are pointed and beautifully cut, and the
shell is sharply defined. A slightly earlier but equally perfect example
of the enrichment has already appeared in Fig. 13.
With slight degradation of form it has been retained, as possibly
the most characteristic of all Classic enrichments, throughout the
Roman and Renaissance times. A quite common Roman variation
has appeared in Fig. 42, the eggs being somewhat widely spaced and
the tongue being replaced by a dart, or barbed arrow head. A much
more crude Romanesque variant has appeared in Fig. 45, the eggs having
there more the appearance of elongated balls and being without the
straight top which they ought to possess in accordance with true
Classic precedent. But this was revived during the Renaissance period,
most of the examples of which date can hardly be distinguished from
their Classic prototypes, though one is given in Fig. 198, carved upon
one of the projecting corbel beams of a house in Halberstadt, North
Germany, which shows that further variations were possible. It was,
however, quite a common thing both in Roman and Renaissance work
to over- load the eggs and the darts (or tongues) alike with foliage.
The leaf and tongue may possibly be a mere variant of the egg
and tongue, suited for carving upon the cyma reversa moulding instead
of upon an ovolo, and having a cyma or double-curved outline. The
^ > 3 O
■» t ■> y
,» > »»
^1 > »
' ^ » ' , > , ' ■" ,
■J > > ' ' >
Fig. 215 — Portion of Huniplucy de Boliun's Monu-
ment, Hereford Cathedral (from a cast in the
Fig. 219 — Ornament on IStli Century Tomb,
Fig. ^217 — Ball Flower (from a cast in the Architectural
Facing jxtge 102
C C I ..<«', ,.
Fic. ■-'20 -Kiirich ncnts of I'ducls in Arch Soffit, St. Giiistina, I'adiia.
(l'"roiii ■■ ( )riii!iu';il il Dctiils of the Itahan Renaissance."')
Facing page 103.
THE EVOLUTION OF ARCHITECTURAL ORNAMENT 103
example illustrated in Fig. 199 is again taken from the later Temple
of Diana at Ephesus. It will be noticed that the tongue is precisely
the same as that in the enrichment of which this may perhaps be a
derivative, while the so-called leaf has a border just such as the egg
has, though its own modelling is essentially different and there is a
strong centre line to it. This ornament proved to be capable of a great
deal of variation. Fig. 200 shows it as it appears on one of the exhibits
in the Grseco-Roman room in the British Museum, retaining to a
considerable extent the refined precision of outline of Greek work,
but with a triple leaf introduced in the centre of the leaf pattern and
with the point differently formed from that which is generally accepted
as being typical. The purely Roman example in Fig. 201 rounds off
the outline, in conformity with the usual Roman inclination to adopt
segments of circles instead of delicate hand drawn curves or portions
of conic sections, while the tongue has been changed into something
more nearly approaching a leaf in shape.
When the Renaissance came, the leaf and tongue was revived in
almost precisely the same form as previously, sometimes more like the
Greek, at other times more like the Roman, but frequently, as was the
case with the egg enrichment, overladen with foliage. One form
was, however, introduced which is never found in the Greek work,
though it has been discovered in the Roman, and that is a series of
leaves without any intervening tongues or darts, as is shown in the
small piece of leaf enrichment from the mantelpiece of Langley Park,
Kent, of quite late Renaissance date, illustrated in Fig. 202, very nearly
The dentil is another enrichment of which anx>iugin can be found
without any extraordinary stretch of imagination. Examination of
the Lycian Tomb shown in Fig. 203, or of that which has already
been illustrated in-'F'ig. 130, would indicate that the timber construction
which was being copied necessitated the appearance of a number of
rafter "ends below the cornice. Similarly, purlin ends appear within
the verge of the arched roof. It may not necessarily follow that the
denti) was directly derived from these timbers, but that it originated
104 THE EVOLUTION OF ARCHITECTURAL ORNAMENT
in some such timber construction is fairly obvious from the position
in which it was mostly found in Greek work, as indicated in Fig. 204,
where it appears beneath the cornice on the entablature of the Mauso-
leum at Halicamassus, as this is now re-erected in the British Museum.
This exactly coincides with the position which the rafter ends occupy
in the Lycian Tombs. The effect, however, is that of a series of square
blocks of stone, and in many subsequent uses these occur as the tiniest
of enrichments, cut out either in woodwork or masonry and employed
almost invariably to break up the density of a shadow ; in this manner
sustaining the tradition of their original employment underneath an
overhanging cornice. It will be noticed that, in contradistinction to
this, the function of the egg and the leaf enrichments was to break
up the light, where it impinged upon curved surfaces, in a similar
The idea of breaking up a shadow by regularly spaced blocks
was employed again in Romanesque times, but it was then often effected
by meai^ of circular blocks and not square ones. The biUet ornament
shown in Fig. 205 was that which was employed, and although the
name was given to it at a much later date there is every probability
that it had a timber origin. There is no connection that can be
traced between this and the Classic dentil, but the effect achieved
is much the same. It occurs here (in Fig. 205) beneath the hood
moulding of the windows round the Templars' Church at Laon. But
it is rare to find it upon the continent of Europe. It is much more
common in English Norman work, as is shown in Fig. 206, similarly
placed beneath the hood moulding of the nave arcade of St. Mary's
Church, Dover, a church whose western arch leading from the nave
to the tower, which appears in the photograph, is of pre-Norman date
and depressed in shape, forming a slight horse-shoe, while the jambs
are not quite vertical.
It will be noticed that the effect, both of the billet and the dentil,
when small, is not entirely unlike that of the bead and reel, of which
there are many variations in Classic and Renaissance times, though it
will suffice to indicate here that two different Greek forms are to be
> , ' ~>
Fig. 221 — Diaper on Rood Screen, Southwell Minster
Fig. 21S — Bauk of Reredos, Peterborough Cathedral
(photographed by Mr. T. R. Somerford, A.R.I.B.A.
Fig. 222— Leaf Enrichniout, Vostil.ule to Henry VII. 's
Fig. 223 — Lower Part of Bay Window, Henry VII. 's
[Facing page 104
•-y>T'T - i "" '*^'» -^-
-J^KJt. — -^ I
Fig. 226 — Ptolemaic Palm Capital.
I<"IG. 224 — Hathor-headed Capital.
V-: \ '
Fig. 225 — Ptolcmaio I,otu.s Capital.
Fig. 227 — Capital at Kiitrancc to Priiiciiial Cliaiiiber
in the Catacombs at Alexandria.
[Facinu page 105.
THE EVOLUTION OF ARCHITECTURAL ORNAMENT 105
seen on the small rolls beneath the echinus with its large egg and
tongue enrichment, both of the Archaic and later Ionic columns at
Ephesus, shown in Figs. 196 and 197.
When discussing Fig, 181, reference was made to the appearance
of a square-nail head enrichment within the Greek key pattern thereof.
Such a nail head has rarely been found in Classic work, but it was
quite a common enrichment in the continental Romanesque and the
English Norman, having apparent origin in a crude attempt to decorate
a surface without much effort by persons who were not possessed of
elaborate tools. It is seen, for instance, in the corbel string over the
choir stalls of St. Nicholas at Blois, shown in Fig. 207, alternating with
simple hatched sinkings of the surface, these being the two different
methods of obtaining texture employed thereon. Some much more
elaborate examples are to be found in the jamb of the south doorway
of Bourges Cathedral (Fig. 208), where the nail heads appear in several
places, occupying, in fact, the whole surface of one of the shafts.
This nail head is nothing other than a straight outlined four-
branched star, but its capability of developing into an ornament
is very great. This seems to have been recognised earliest upon the
continent. One of the first examples of its evolution into what one
may call an incipient dogtooth occurs on the arches of a 12th century
house in the old town of Dol, in Brittany, which is shown in Fig. 209.
Here the original nail head has been replaced by a four-leafed flower,
still rudely carved and varying considerably throughout the range.
It will be noticed that a ball centre has been given to it. This idea,
crossing over to England, developed into the Norman dogtooth,
shown in Fig. 210. Neither of these forms is common, but they may
well be considered to be steps in the evolution of the true dogtooth
enrichment, though the last example is even more definitely a flower
than is that at Dol ; the centre ball is more pronounced, and has a small
knob upon it, while other little ball knobs appear along each of the petals.
The true dogtooth ornament, as shown in Fig. 211, appeared
simultaneously in France and England though in France only where
English influence was predominant. This particular example is another
io6 THE EVOLUTION OF ARCHITECTURAL ORNAMENT
taken from the Templar's Church at Laon ; the dogteeth are cut in
a hollow between two rolls, here, as in all cases, so placed that they
could be easily carved out of a rectangular block of stone, in this
instance that from which the vault rib was fashioned ; and also,
again, as in most other cases, appearing as the enrichment of a hollow
and intended to break up the extreme depth of its shadow.
It is more rarely employed beneath a hood moulding in the position
commonly occupied by the Norman billet or nail head, but it is found
there sometimes, as can be seen from the photograph of a Romanesque
archway at Lou vain in Belgium (Fig. 212).
The earlier examples are all fairly shallow, but as time went on
the hollows became deeper, and by the middle of the 13th century
the dogtooth was often scarcely discernible in their heavy shadow.
The effect produced consequently, one would think, scarcely jus-
iiied the large amount of labour expended. That so much trouble
should be taken to produce so slight a result as is indicated in the
deep shadow in Fig. 213, may well be wondered at, if it were not
known that the great Gothic carvers of that date spared no pains to
obtain perfection in their work. The example is taken from the
Chapter House of Lincohi Cathedral, but there is a large amount of
such work in England.
To what extent the diaper work of this same 13th century was
derived from the dogtooth it is impossible actually to say, but certainly
there is a close resemblance between the one and the other — except,
of course, that the diaper is a shallow ornament worked upon a surface.
Fig. 214, however, is so closely allied in its detail to Figs. 209 and 210
that it is impossible not to imagine that some connection must have
existed, and it is by no means exceptional. It is one of many
different forms of diaper ornament to be found in Westminster Abbey.
There is a star-shaped four-leafed flower with a ball at the junction
of the petals, which appears to have originally itself represented a small
flower, while there are other flowers, probably primroses, introduced
between the great petals, following the general tendency of the 13th
century carvers to represent spring foliage.
THE I EVOLUTION OF ARCHITECTURAL ORNAMENT 107
Subsequently this star-shaped arrangement of a flower becomes
quite common ; it is found in the latter part of the 14th and during
the 15th century, introduced as an ornament in a shallow hollow and
varying in outline according to date. Two different forms appear
in the portion of Humphrey de Bohun's Monument in Hereford
Cathedral, of which a sketch is given in Fig. 215 ; the leaves are now
of a perfectly natural type, but the centre of one represents a flower
bud, while the centre of the other is itself a small four-leafed flower,
remarkably like the dogtooth in its general suggestion. Later again
it occasionally took some such entirely conventionalised form as is
shown in Fig. 216, which is a late example from Yatton Church in
Somersetshire. These last examples belong to the 15th century, or
possibly even later.
During the intervening 14th century an ornament appeared in
the West of England to which the name of " ballflower " has been
given. It was considered by all the older writers upon Gothic archi-
tecture to be the distinctive ornament of the " Decorated " period, but
as a matter of fact it is only found in a few counties, where it was
used largely, occupying the same position in the hollow mouldings of
that period and district which was filled by the dogtooth of the 13th
century all over England, But as the hollows were not so deep the
effect was a different one. These balls, occurring in constant suc-
cession all round the tracery as well as in the true mouldings of a win-
dow, give it much the appearance of knotted lacework. A detail of
one is shown in Fig. 217. An outer ball, like a seed pod, is slightly
opened, displaying another one within it. A range of them occurs
at the back of the reredos in Peterborough Cathedral, shown in Fig.
218, in the hollow cornice, though not so close together as in more
typical examples, while the form is somewhat altered, probably owing
to the fact that the work was executed in one of the eastern counties
of England, where it is quite a rarity, and not in the west, where it
is common. The outer petals, which are always three, are more
pronounced, and there is no sign of the inner ball.
Occasional examples are found in the 15th century of a com-
io8 THE EVOLUTION OF ARCHITECTURAL ORNAMENT
bination of these motives. Fig. 219 shows a four-leafed patera in the
cornice of a 15th century tomb in Westminster Abbey. The general
idea is that of the leaves in De Bohun's tomb (Fig. 215), but the centre
is a ball, as if the ball flower and the four-leafed flower Vv^ere combined.
Other similar examples are by no means uncommon.
Paterae like these are not entirely confined to Gothic work. It
is very rare indeed that one can trace Gothic influence in Renaissance
ornament, but perhaps it is only necessary to introduce here an illus-
tration of the paterae in the arch soffit of Sta. Giustina at Padua
(Fig. 220) for the similarity of motive to that of much of the late
Gothic ornament to become at once apparent, with the advantage of
indicating how large an amount of variation is possible of quite a
simple original idea.
Diaper work is not always arranged on a diagonal or other regular
scheme, at any rate so far as the pattern itself is concerned, though
it is always in square or diamond-shaped blocks. Occasionally isolated
leaves occupy the diapers, as in some of those in the rood screen of
Southwell Minster ; an example is given in Fig. 221. The relief is
not great, and of course the form of the leaf is that which is indicative
of the period, which in this case is that of the 14th century. Similarly,
when leaves were used as minor enrichments in the 15th century,
they also partook of the character of the time. An illustration of this
will be found in Fig. 222, the leaf being here as unquestionably autumnal
as is that shown in Fig. 221 the open leaf of summer. It is one of
a series of leaf enrichments in the hollow cornice moulding of the
vestibule to Henry VII. 's Chapel, Westminster. A photograph of the
exterior of a portion of the same building is given in Fig. 223, mainly
with the object of illustrating a minor enrichment which became
common during the 15th century — that known as the crenelle. It
appears invariably upon the top of a ornice as a kind of cresting
thereto, and has every appearance of having been derived from the
embattlements of castle walls with their alternating embrasures. This
is probably the case, and if so it is perhaps the most prominent example
which exists of what was originally a wholly essential feature in a
THE EVOLUTION OF ARCHITECTURAL ORNAMENT 109
building intended for defensive purposes, being afterwards employed
in a purely decorative manner on a much smaller .scale. Incidently,
this same photograph shows various smaller enrichments in the quartre-
foils of the panels. The portcullis, indicating Westminster, can be
seen, and so can the fleur-de-lis, the sign that the building was regal
and that England claimed the sovereignty of France. There is a
third enrichment, which consists of a double five-leafed flower — the
well known Tudor rose, formed by placing the roses of York and
Lancaster one above the other. There is some doubt whether the
roses represented are wild dogroses or primroses ; at any rate, they
are five petalled flowers of some amount of similarity to those of which
a sketch has just been given in Fig. 214.
Some Alexandrian Capitals— An Appendix
; • 1 S
Fig. 228 — The Khartoum Monument,
Fig. 229 — Pompey's Pillar, Alexandria.
[ Facing page 112.
- -: .4^-^-iU-l
I V "■
Fig. 230 — A Doubly Rupportcrt Respond Capital.
Ji'iG. 231 — A crude Roman Capital.
Fig. 232— a Doric Capital.
i''ia. 233 — A Greek lonio Capital.
Fig. 234 — A Roinano-Tiyzantine loiiie Capital.
Fig. 235 — A Hyzantinc Basket Capital
Far! I) ft /'".'/'' l':!.
Some Alexandrian Capitals — An Appendix,
On account of its geographical position and fine natural harbours,
easily made more safe by artificial means, Alexandria has necessarily
been the most important seaport of Egypt for many centuries, while
it was, for a considerable period, one of the very greatest cities of the
civilised world. Founded by Alexander the Great in 332 B. C, it
became the centre of Greek intellectual life during the third and
second centuries B. C, precisely at the date when the fourth great
building period of Egypt, the Ptolemaic, was at its zenith. Later on
it became a most important Roman port, and then the headquarters
of one of the great branches of the Christian Church, to be eventually
destroyed by the Arabs in 641 A. D., having risen afresh after almost
as complete destruction in 116 A. D., in 215 A. D. (when the inhabitants
were massacred by Caracalla), and in 616 A. D., when it was taken by
the Persians. Throughout this long period of a thousand years, its
merchants traded with all parts of the Mediterranean by sea, and
simultaneously with Egypt and Nubia by way of the Nile, and with
Syria and Assyria by means of caravans.
Thus Alexandria has been the natural meeting place of many
influences. Yet there is little known of it, and the archselogist and
the enquirer into the evolution of architectural forms have troubled
little about what its ruins might have to tell. In truth, investigation
has been unusually difficult, owing to the occupation of the site by a
modern city, and to the gradual sinking of the land below the ancient
water level, while the one scientific attempt at excavating, made by
Mr, D. G. Hogarth in 1893, was remarkably barren in its results.
Since then a considerable number of fragments of buildings, mainly
capitals of columns, have been discovered. Most of these are stored
114 THE EVOLUTION OF ARCHITECTURAL ORNAMENT
in the Museum, but with little to indicate either their probable date
or even whereabouts they were discovered. For all this, they are
worthy of study, and, as the study proceeds, various points of interest
disclose themselves, and at least one somewhat startling conclusion
forces itself upon the investigator.
Near as is Alexandria to Cairo and Ghizeh, there are no indications
of buildings having existed there when the Pyramids were building,
nor even during the subsequent building periods of the Theban dynas-
ties, if we except the one Hathor-headed capital (Fig. 224) carved in
basalt, of which examples of Theban date have been found at Deir-
et-Bahari. That this capital should itself be Theban is, however,
exceedingly doubtful, for all other known examples, except those at
Deir-et-Bahari, are Ptolemaic ; that is, they belong to the Egyptian
style which was prevalent during the period when Alexandria
was the great centre of Greek culture, from B. C. 332 onwards. The
face exhibits all the typical characteristics both of Egyptian work
(impassiveness) and of sculpture in a hard material, being devoid of
detail and treated in broad flat surfaces.
Fragments of unquestionably Ptolemaic capitals, such as those
shown in Figs. 225 and 226, occur more frequently in the Museum.
In every case they are carved in the local limestone or coral rock,
and so typical are they of Egypt as to indicate with some conclusiveness
that Alexandria was simultaneously a Greek and an Egyptian city,
with two highly developed civilizations co-existing side by side, like
the European and the Arabian at the present time. The ornamentation
of one of these capitals is based upon the lotus, and that of the other
on the palm, the origin being obvious in each case ; but they are merely
typical examples selected from a considerable number, and no two are
precisely alike, though all are small. The probability is that they
were decorative rather than constructional.
On the other hand, the capital shown in Fig. 227 (or, rather, the
pair of capitals, for there are two of them) appears to be unique and con-
structional. It is rock cut ; that is, the column it belongs to has been
left standing when the solid rock around it has been hewn away, and
THE EVOLUTION OF ARCHITECTURAL ORNAMENT 115
serves as a support to the roof of coral rock at the underground entrance,
itself excavated, to the sacraficial chamber of the catacombs, dis-
covered only a few years ago. The column is Ptolemaic Egyptian in
type in all respects save one ; but the exception is important, for it
consists in the introduction of voluted leaves and caulicolse in the upper
part of the capital. Taken in conjunction with the occurrence else-
where in this small catacomb chamber of the Greek form of the egg-
and-tongue enrichment, and of purely Egyptian palm capitals, there
is indicated a tendency to combine Greek and Egyptian ornaments in
a manner which is at least unusual.
This would not be so obvious if there were not ample indications
elsewhere that the voluted tendrils and leaves of Corinthian capitals
were no rarities at Alexandria. Examples of the Greek Corinthian
form of capital are extremely few, and it is even generally supposed
that the only true examples are those of the Tholos at Epidauros and
of the Choragic Monument of Lysicrates at Athens, with the possible
addition of the much destroyed capital, having an elliptical plan, which
was brought from Ephesus and is now in the British Museum, (Fig. 37).
Yet there are a large number at Alexandria, mostly in the Museum
and recently discovered, though the largest, carved in granite, now
surmounts the Khartoum monument (Fig. 228), erected a few years
since to commemorate the retaking of Khartoum, from miscellaneous
fragments pieced together. All, too, are of practically the same design,
having a double row of alternately tall and short acanthus leaves
springing from the necking and rising about half-way up the bell of
the capital, which is exposed above except for the occurrence of broad
voluted leaves at the angles and tendril volutes, forming caulicolae,
So distinctive is the type, and so beautiful, that it may well be