found under the old Excise Office, Broad Street. I have an original
drawing of it by Fairholt, dated March 1, 1854 (Fig. 90). The best
authorities are two large original coloured drawings, one by Archer in
the British Museum and the other at the Society of Antiquaries. The
central panel had a white ground and black border; the Nymph had reddish
flesh and a light greenish scarf; the Panther seems to have been a
grey-buff spotted black. There was much black and white in the pattern
work, and some of the fillings were of black and white triangles.
[Illustration:
FIG. 91.
]
It was described at the time of finding as having formed a square of 28
ft.; it was diagonally about north and south and 15 ft. below the
surface. The central subject was “Ariadne or a Bacchante reclining on a
panther.” In _V.C.H._ it was said to be “Europa on the Bull,” but the
drawings agree with the former description. The composition is very
similar to the Bacchus, and doubtless a wine cup was held by the
Bacchante also. Notice that vases appear elsewhere in the design. The
panel was about 2½ ft. square. This fine floor was taken to the Crystal
Palace, where it seems to have disappeared. From its size and subject we
may suppose that it was the floor of the central dining-hall of some big
house. The drawing and balanced design of the central group is
wonderfully skilful as space filling. Fig. 91 is based on original
drawings of the floor at the Society of Antiquaries and the British
Museum and a sketch in the Wollaston Collection at South Kensington.
This mosaic should be compared with a floor found at Bignor, which is
very similar in its details, and probably, I think, by the same artist.
There the centre is occupied by Jove’s eagle and Ganymede, the cupbearer
to the gods.
_Vase-Panel Mosaic._—In his account of discoveries at Bucklersbury,
Price describes a floor found in St. Mildred’s Court which must have
been one of the finer kind. “A square enclosed a circle containing a
vase in brown, red and white with the addition of bright green glass.
Around the vase there appeared portions of a tree with foliage; also an
object resembling an archway with embattled figures and other objects,
the meaning of which is difficult to describe without an illustration.
Around the whole were two simple bands of black tesseræ separating the
circle from an elaborate scroll of foliage and flowers, analogous to
that on one of the pavements at Bignor. At each corner was a flower
showing eight petals of varied colours. From the centre of each sprang
two branches, which united in a leaf in form like that within the
scroll. The entire design is bordered by the guilloche in seven
intertwining bands of black, red, brown and white tesseræ. A drawing of
this interesting floor was in the possession of Mr. G. Plucknett.” The
central panel must have been a formal landscape—a large wine krater
backed by a tree and an arcade with figures on the parapet. In another
place Price names it again amongst mosaics which had glass tesseræ;
probably the tree was of green glass. This pavement also doubtless
occupied a dining-hall. In an earlier account (_London and Middlesex
Archæol. Soc. Proceed._ iii.) Price says: “When perfect it was of some
extent, resembling those discovered at East India House and the Excise
Office. In the centre was a vase similar to those at the Excise Office,
and around it a scroll of foliage beautifully arranged. The fragments
were packed in cases and sent to the workshops of Messrs. Cubitt.”
_An Orpheus Mosaic (?)._—Roach Smith reported the existence, below
Paternoster Row, of what must have been an exceptionally fine pavement,
which was broken up before any proper record of it could be made. This
“superb pavement extended at least 40 ft.; towards the centre were
compartments in which in variegated colours were birds and beasts
surrounded by a rich guilloche border.” The wording suggests a square
room, and the two former examples show that large square rooms existed
in London. In the villa at Woodchester the chief central room was nearly
50 ft. square; the pavement had “a central circular compartment; within
the border was a wide circular band containing representations of
animals, inside was a smaller band containing birds; on the southern
side was a figure of Orpheus.” The description of the London mosaic
suggests that it, too, had for subject Orpheus charming the beasts. It
was found about 1840 at a depth of 12 ft. In 1843 part of a mosaic
floor, “with birds and beasts within a guilloche border,” was found at a
depth of 12½ ft. below the offices of the Religious Tract Society at the
corner of Cannon Row (_V.C.H._). Is it not probable that this was
another part of the pavement described by Roach Smith?
_Inscribed Floor._—A mosaic pavement found in Pudding Lane as lately as
1886, and bearing an important inscription, was destroyed before any
sufficient record of it was made. A printed version of the lettering was
given in the _Archæological Journal_ of the same year by Dr. Haverfield,
with some comments. (Also see _S.A. Proceedings_, xiv. 6, and _V.C.H._)
In the collections of the Society of Antiquaries I find a sketch of it
by Henry Hodge, a careful draughtsman of the time. This drawing is said
to have been made “from a sketch by I. W. Jolly and fragments,” so that
its strict accuracy is questionable. It appears that it was complete on
the right but imperfect on the left-hand side. On the right some parts
of the pattern covering the rest of the floor and a border are shown and
some dimensions are given. It looks as if the panel was about 5 ft.
across and was the centre of a strip 7 ft. or 8 ft. wide. The letters
were about 3 in. high, black on a white ground; the last four seem to
have been D. S. P. D.—_de sua pecunia dedit_—and this would imply that
the mosaic belonged to a temple. The destruction of these mosaics is a
sad witness to the nineteenth-century type of intelligence. Of all of
them only the fragments of the Bacchus pavement are now known to exist.
I should like to find out what became of the Bacchante pavement sent to
the Crystal Palace, and whether the vase mosaic is still in packing
cases at Messrs. Cubitt’s. I wonder, too, what became of Mr. G.
Plucknett’s drawing, and wish I could get tidings of it. The great
pavement in Paternoster Row seems to have been destroyed without even a
drawing being made; while the sketch taken by Mr. Jolly of the inscribed
floor has, so far as I know, been burnt. And this was the high age of
university education!
_Bucklersbury._—The most perfect of the existing mosaics is the complete
and restored pavement with an apsidal end found in Bucklersbury. A good
account of it while yet in its place is given in _The Builder_ (1869):
“It lies fresh and bright as when it was first put down.... It is to be
hoped that some pains will be taken to trace the remaining walls of the
building to which this speaking pavement belongs.” Here, again, although
the apartment was not large and the ornamental mosaic was more than a
central panel, there was a broad border of the coarse tesseræ. Besides
having been a saving, the contrast of the plain red with the variegated
central area seems to have been liked. The interlacing squares of this
pavement resemble those of the Excise Office floor, and its central rose
is like a panel in the same floor. An angle-filling is similar to a
quarter of the central pattern filling the centre of the small India
Office pavement, which, again, had interlacing squares. A single
cross-like pattern filling a panel in the British Museum is again like
that of the India House mosaic. Many such references could be carried
much further, not only in regard to London pavements, but including the
country ones also. I reach the conclusion that they are for the most
part nearly of the same date, and that many were by the same artists.
_Fenchurch Street._—A fragment of what must have been a fine floor was
found in 1859 and is now in the British Museum. It is part of a panel
which contained a vase and two birds. An illustration given in Price’s
_Bucklersbury_ shows that there was a margin of coarse tesseræ beyond,
and that the panel must have been one of a series making up a handsome
border. A fragment of a floor with a wide border divided into panels has
lately been found at Colchester. Roach Smith described the former as
“what would seem to have been an extensive pavement,” and he calls the
bird a peacock. A good coloured drawing, in the Archer collection, of
the fragment shows the bird’s neck a bright blue; the blue tesseræ were
of glass. Fig. 92 is from Price, but I have dotted in on the top
right-hand corner the line of a more modern building from Roach Smith’s
illustration. This is one example of many cases in which more recent
walls have been carried up from the Roman level and square with a Roman
building. (A in fig., and compare Fig. 90.)
[Illustration:
FIG. 92.
]
[Illustration:
FIG. 93.
]
[Illustration:
FIG. 94.
]
[Illustration:
FIG. 95.
]
_Birchin Lane, etc._—In 1785 a small piece was discovered here of “a
fine tesselated pavement of very small bricks and stones; of this, only
one corner appeared, which is composed of black, green, and white stones
and brick, forming a beautiful border.” Another account says that “the
tesseræ measured about one-quarter of an inch and were of various
colours.” I am particular about this, for the bright colours were
doubtless of glass. I find a contemporary drawing of this fragment in
the Guildhall Library, from which it appears that there was a fair blue
besides the colours mentioned. (Fig. 93; compare Fig. 92 and a border
illustrated by Mr. Ward.) Outside it were big red “brick” tesseræ. There
is in the Guildhall Museum a fragment of another mosaic found in Birchin
Lane. It is part of a star-shaped all-over pattern of a well-known type
(the Barton Cirencester, etc.). Fig. 94 A shows the fragment, and Fig.
95 is a diagram of the complete pattern. Another piece at the Guildhall
has a sea-monster of small scale but most skilful execution. The place
of finding is not noted, but it is probably a fragment discovered in
Birchin Lane in 1857, described in _V.C.H._ as part of a pavement
“representing a sea-horse.” Two other small pieces in the same museum
are very similar in colour and quality, and may have come from the same
source. One of these seems to have belonged to a pavement of square
panels of knot-work framed in scroll bands (Fig. 94, B), or it may have
been part of a panelled border similar to Fig. 92. Morsels of painted
plaster were also found in Birchin Lane, where there must have been a
good house.
[Illustration:
FIG. 96.
]
A fragment of mosaic at the London Museum comes from another all-over
star-pattern similar to that at the Guildhall, but this piece was next
to the outer border of the pavement. This fragment is of particularly
beautiful colouring—quite a purple floor. I give a sketch of the
fragment in Fig. 96; it must have come next the border of a pattern like
Fig. 95.
_Threadneedle Street._—Several pieces of London mosaic are shown in the
Roman corridor at the British Museum, but not very effectively. Two are
exhibited as given by Mr. E. Moxhay, but it is not added that they were
found in Threadneedle Street in 1841. One is part of a passage and the
other is a square from the centre of a room. (See illustrations in Roach
Smith’s _Roman London_, from which Fig. 97 is taken.) Another piece
found at East India House, Leadenhall Street, is not set up rightly. The
pattern is of two interlacing squares; the margin should not be parallel
to either of these, but it should touch two of the points of the star
form. (Fig. 98. See Sir W. Tite’s illustration in _Archæologia_; compare
also the Bucklersbury pavement at the Guildhall.) This floor came from
the same level as the Bacchus mosaic and not far away from its position;
probably the small chamber to which it belonged was part of the building
which contained the large square hall of the Bacchus mosaic.
[Illustration:
FRAGMENT OF ROMAN TESSELLATED PAVEMENT DISCOVERED AT THE DEPTH OF 14
FEET UNDER THE FRENCH PROTESTANT CHURCH IN THREADNEEDLE STREET.
APRIL 1841.
FIG. 97.
]
_The Bank._—A fourth piece in the Museum is a square panel from
Lothbury. Allen describes it as “An ornamental centre, measuring 4 ft.
each way, of an apartment 11 ft. square; beyond this were tiles of an
inch square extending to the sides of the room.” It is another example
of the plan of having a comparatively small central panel liberally
framed in much plain red work. The device in the centre is a cruciform
pattern. I can hardly think that from, say, 250 A.D. it would not have
been recognised as a cross indeed. Compare the small cruciform centres
of two squares of mosaic exhibited close by.
[Illustration:
FIG. 98.
]
The floor mosaics at the British Museum are dispersed in two galleries
and a staircase, and even so each one is badly presented. Fragments of
the Bacchus floor are shown without any key-plan of the whole. Of five
on the north wall of the Roman gallery, the place where only one was
found is told. The interesting little Orpheus mosaic discovered at
Withington is shown by three single fragments, although an excellent
restored engraving was published in _Archæologia_ when it was found. I
wish space could be found for setting them in their due relation and
completing the composition in outline. The surface requires careful
cleaning and some repolishing. The floor from Thruxton on the north
staircase has lost its centre since it was engraved. The engraving
itself is shown in the gallery a hundred yards away, without any
reference from one to the other. In this case, I think, the centre
should be painted in on the plaster filling of the original.
These mosaics must have been drawn out on the levelled beds prepared to
receive them by the master artist and filled in by him and his
assistants. The preparation for such a floor is made clear in the
description of a London mosaic found in 1785: “This pavement, as well as
most of the rest, was laid in three distinct beds; the lowest very
coarse, about 3 in. thick, and mixed with large pebbles; the second of
fine mortar, very hard and reddish in colour, from having been mixed
with powdered brick; this was about 1 in. in thickness, and upon it the
bricks [tesseræ] were embedded in fine white cement” (_Archæol._, vol.
viii.). The Bacchus pavement described before “was bedded on a layer of
brickdust and lime of about an inch.” Powdered brick (tile) and lime
made a strong cement which would finish perfectly smoothly and provide
an inviting surface to draw and work upon.
Several mosaics while not quite plain were simpler in design and perhaps
coarser in execution than those already described. A star-shaped
fragment found in Bishopsgate Street, illustrated by Roach Smith, was of
black and white tesseræ. It was probably the central panel of a floor,
as Roach Smith said. A mosaic found at Lincoln had a similar star-shaped
panel at the centre. About 1840 a tessellated pavement was found in
Bishopsgate-Within “of black and white tesseræ in squares and diamonds”
(_V.C.H._). In Bush Lane “a pavement of white tesseræ” is recorded. On
the site of the Guildhall “irregular cubes of dark-grey slate and white
marble” were found (_Journal B.A.A._ xix.).
Another pavement, found in Lombard Street in 1785, was “composed of
pieces of black and white stone one-third of an inch square, probably
deposited in regular order” (_Archæol._ viii.). These black and white
mosaics were doubtless like the counter-changed patterns found at
Wroxeter, Silchester, etc. At the latter the Christian church had a
square space for the altar paved in this way. Several years ago I drew a
fragment of an identical design at Lincoln. This was probably a
fourth-century fashion. The others may be a little earlier generally,
but they overlapped into the Christian period.
Many floors have been found in London which were wholly of coarse
tesseræ of red, or of a few simple colours accidentally distributed. One
of these is described as of irregular tesseræ about 2 in. by 1½ in.,
mostly red, but some black and white. A room 17½ ft. by 14 ft., in
Leadenhall Street, had coarse tesseræ red, black and white, 1¼ in.
square, and a similar floor “of red bricks about an inch square with a
few black ones and white stones” was found in Lombard Street. Some
floors found at Silchester had circular and polygonal tiles used with
mosaic cubes filling up the interspaces. At Bath, if I recollect aright,
there are fragments of pleasant floors in which larger irregular pieces
of marble are set here and there in a floor mainly of large red tesseræ.
At Silchester a polishing tool is said to have been found, being a lump
of marble with an iron socket for the attachment of a handle
(Middleton’s _Rome_).
A general comparison of the British mosaics brings out the resemblances
between the members of certain groups. The similarity of the Cirencester
and Woodchester Orpheus pavements has already been mentioned. The London
floor found at the Excise Office was very like the mosaics at Bignor in
both the patterns and figure work. The same pavements resemble others
found at Silchester, and also the Cirencester and Woodchester mosaics. A
pavement found at Stonesfield, near Woodstock, had a wreath of foliage
springing from a head similar to that of Woodchester. It is obvious that
elaborate works in isolated villas cannot have been home-made, and it is
likely that this group at least was the work of craftsmen established in
some central city. No centre is so likely as Londinium, a wealthy town,
the most conveniently placed for the importation of materials. We think
of these works as “decadent,” but really there was a new life in them.
The centre of origin of the later type seems to have been Alexandria,
and similar works to our own are found in Asia Minor, North Africa and
Gaul. The use of glass in these mosaics is likely to have been an
Alexandrian innovation. Price gives a list of five London mosaics in
which glass was used, and I may add the fragment at Birchin Lane
described above. Glass was also used at Cirencester and Woodchester: the
purple tesseræ in the fine border of the latter in the British Museum
must be glass.
[Illustration:
FIG. 99.
]
[Illustration:
FIG. 100.
]
Taking into consideration the great similarity of mosaics found in the
East—those from Halicarnassus, for instance, now in the British
Museum—to those found in the West, the character of the patterns, the
mystical nature of some of the figure designs, and the swift ability of
the workmanship, I am drawn to the conclusion that the craftsmen are
likely to have been Greeks. Some confirmation of this is to be found in
the fact noticed by Wright, that the Greek H sometimes appears for E in
the few mosaic inscriptions which exist. Mosaics must, I think, have
been works of the prosperous Constantinian age. The floor at Frampton
had a XP monogram on it (Fig. 99); the Orpheus pavement at Horkstow,
accepted as Christian by Cabrol, had crosses (Fig. 99); and a second one
at Winterton has a red cross by one of the animals; the pavement at
Thruxton, in the British Museum, has crosses set in the border in what
seems to be a significant way (Fig. 100). The other details on this
figure also have a Christian look; the top one is from Bignor, the
bottom one from Frampton. Fig. 101 is from the Orpheus mosaic at
Withington. If the Orpheus pavement at Frampton was Christian, the
others are likely to have been so too. At least, they symbolise the
Harmony of the Universe; they are not “mythological.” These pavements
are evidence of the cosmopolitan nature of Romano-British culture.
[Illustration:
FIG. 101.
]
[Illustration:
FIG. 102.
]
[Illustration:
FIG. 103.
]
Any idea of thought in decoration is difficult for us to apprehend. The
records of the pavements which have been found in Britain deserve study
from this point of view. The whole art of the time witnesses not only to
the professional skill of artists, but to the thoughts and desires of
the provincial Romans—and natives too, doubtless—who demanded such
works. They speak of a time when the old beliefs had been for a large
part allegorised and fitted into a sort of poetic cosmogony; the designs
often dealt with the order of Nature. Many interesting details are to be
found in these mosaics; Fig. 102 is a sundial which appears with a
celestial sphere on the pavement at Bramdean. The fragment of
inscription (Fig. 103) is from Thruxton. Large square mosaics which seem
to have been the floors of central halls have been mentioned. In two
cases, such floors found in Britain had sunk water basins at their
centres. At Woodchester four columns were placed about the central
space, and there was doubtless an opening in the roof above. Such a
central hall would have been an _Atrium_, and this helps to explain the
planning of Roman houses in Britain.
------------------------------------------------------------------------
+CHAPTER VIII+
WALL PAINTINGS AND MARBLE LININGS
BY putting together, in our imagination, the mosaic floors, the
fragments of wall paintings, and the marble linings, we can gain a
fairly certain knowledge of what the finer Roman interiors in Londinium
were like, and we may further add to the impression by remembering the
many precious objects in silver, bronze, pottery and glass, which are in
our museums. Broken remnants of wall paintings have been found in large
quantities, and pieces are preserved at the British, the Guildhall, and
the London Museums, also at the Society of Antiquaries. Some account of
several of them was given by Roach Smith in his _Illustrations of Roman
London_, from which Fig. 104 is reduced. The fragment (5, Fig. 104), now
with the others in the British Museum, is part of a pilaster-like strip
about 8 in. wide, of foliage springing symmetrically on each side of a
central vertical stem; it is on a dark ground, and marginal lines divide
it off from a red space which covered the main surfaces of the wall.
This “pilaster” was doubtless one of several. The morsels (6 and 7, Fig.
104) evidently belonged together; the one-sided nature of the design
suggests that it was next the angle of a room; and the loop in the upper
part of 7 looks like the end of a festoon; 9 is somewhat similar; and
the others may all have belonged to “pilaster” strips.
[Illustration:
FIG. 104.
]
The method of dividing up the wall space with strips of plain colour or
with “pilasters” was very general. A simpler scheme was to have marginal
borders only, and these were frequently of considerable width, made up
of many bands and lines of colour. Dadoes were very general, sometimes
only a plain band of colour or a horizontal bar running into the
margins; at other times they were fully decorated: two examples lately
illustrated in _Archæologia_, from Caerwent and Silchester, are really
fine work. The latter had a row of “panels,” alternating square and
round, set with leaves and ears of corn, on a red ground between dark
top and bottom bands.
_Stripes and Margins._—A piece of wall of considerable height was found
at Bignor, having a quadrant skirting at the bottom, a plain dark band
as a low dado, and the space above divided into panels. At Cirencester a
fragment was found which showed a band of fair yellow, edged with
margins of white separating spaces of a cool grey-green. At the Society
of Antiquaries is a piece of plaster showing fine red and green spaces,
divided by a white band and a black line—very simple, but beautiful
colour (Fig. 105).
[Illustration:
FIG. 105.
]