The Architectural Antiquities Of Western India

00 ir

u< OU_158650>[g

THE ARCHITECTURAL ANTIQUITIES
OF WESTERN INDIA

THE ARCHITECTURAL

ANTIQUITIES
OF WESTERN INDIA

BY

HENRY COUSENS, M.R.A.S.

LATE OF THE ARCHAEOLOGICAL SURVEY OF INDIA

THE INDIA SOCIETY

3, VICTORIA STREET, LONDON, S.W. i
1926

PRINTED IN GREAT KKITATN

CONTENTS

EARLY ARCHITECTURAL REMAINS -

STUPAS OR CINERARY MOUNDS -

EARLY STRUCTURAL TEMPLES -

LATER CHALUKYAN TEMPLES - - - - - - ' 2 7

TEMPLES IN NORTH KANARA - - - - - - 34

THK SOLANKI TEMPLES OF GUJARAT AND KATHIAWAD - - 35

TEMPLES IN THE DAK HAN - - - - - -5*

MUHAMMADAN BUILDINGS AIIMADA13AD AND GUJARAT - - 58

ARCHITECTURAL REMAINS AT 13IJAPUR - - - - ?">

MUHAMMADAN BUILDINGS IN S1ND - - - - -82

LIST OF ILLUSTRATIONS

P % late.

1. The Great Cave at Karli

2. Interior of the Great Cave at Karli

Cave at Nasik showing Wooden Construction

3. Chaitya Cave at Ajanta

4. Front of Vihara Cave at Ajanta
Interior of the Tin Tal Cave at Elura

5. Exterior of the Tin Tal Cave at Elura

6. The Great Monolithic Temple of Kailasa at Elura

7. Nagaraja and Nagadevi at Ajanta
Colossal Buddha at the Kanheri Caves

8. Pillar in the Indra Sabha Cave at Elura
Pillar in the Jagannatha Sabha Cave at Elura

9. Capital of Broken Column at Ajanta
Structural Buddhist Chaitya at Ter

10. The Thul Mir Rukan in Sind

n. Stupa Mound at Mirpur-Khas in Sind, before excavation
Stupa at Mirpur-Khas, after excavation

12. Terra-cottas and Moulded Bricks from Stupas in Sind

13. More Terra-cottas and Moulded Bricks from Sfttpas in Sind

14. Head of a Buddha from the Mirpur-Khas S/upa
Image from the Mirpur-Khas Stupa

15. Old Temple at Gop in Kathiawad

16. Old Temple at Bilesvara

Old Temple of Surya at Sutrapada

17. The Temple of Meguti at Aihole
Two Styles of Towers at Pattadakal

18. Interior of Lad Khan's Temple at Aihole
Interior of the Temple of Virupaksha at Pattadakal

19. The Durga Temple at Aihole

20. The Temple of Virupaksha at Pattadakal

21. The Temple of Papanatha at Pattadakal

22. The Temple of Kasivisvesvara at Lakkundi

23. Doorway Mouldings of the Temple of Kasivisvesvara

viii LIST OF ILLUSTRATIONS

Plate.

24. Sculpture on the Walls of a Temple at Halabid, Maisur

*$. The Temple of Mahadeva at Ittagi

26. Shrine Walls of the Temple at Ittagi

Corner Slab of Ceiling of Hall of Temple at Ittagi

27. Hoysala Sculpture from Balagamve
Interior of Temple at Annavatti

28. Sculpture at Base of Tower of Temple at Balagamve
Window in a Temple at Narsapur

29. Bracket Figure from a Temple at Kuruvatti
Image from Narayanapur, in Haidarabad Territory

30. The Temple of Dodda Basappa at Dambal
The Walls of the Temple at Dambal

31. Pillar in the Temple of Sarasvati at Gadag
Another Pillar in the Temple of Sarasvati at Gadag

32. Gateway of the Town of DaMioi, in Gujarat
Pillars of the Rudramala at Siddhapur

33 The Temple of Somanatha at Somanatha-Pattan, in Kathiawad

34. The Temple of Navalakha at Ghumli
The Temple of Navalakha at Sejakpur

35. Inter ioi of Vimala Sha's Temple at Abu

30. Ceiling in Vimala Sha's Temple at Abu

Central Pendant of Ceiling of Tejahpala's Temple at Abu

37. Ceiling Panel from Temple at Abu
Another Ceiling Panel from Temple at Abu

38. Temple of the Sun at Mudhera : closed Hall

39. Temple of the Sun at Mudhera : open Hall

40. Interior of the Temple of the Sun at Mudhera

41. Kirttistambha at Vadnagar

42. Southern Peak of Satrunjaya Hill
Torana from Temple at Abu

43. The Temple of Ambarnatha

44. The Temple of Govindesvara at Sinnar

45. Architrave in Temple of Aesvara at Sinnar

46. Temple at Balsane

LIST OF ILLUSTRATIONS

Plate.

47. Central Portion of Facade of Jami Masjiil at Ahmadubad
Mosque in Mirzapur Quarter at Ahmadabad

48. Window in Sidi Sayyid's Mosque at Ahmadabad
Ruins of Haram at Sarkhej

49. Queens' Tombs at Ahmadabad

Ceiling Panel from the Jami Masjid at Chrmpanir

50. The Jami Masjid at Bijapur
Interior of the Jami Masjid at Bijapur

51. The Ibrahim Rauza at Bijapur
Window in the Ibrahim Rauza at Bijapur

52. Walls of the Tomb of the Ibrahim Rauza

53. The Mehtar-i-Mahal at Bijapur

54. The Gol Gumbaz at Bijapur

55. Coloured Tiles in the Walls or the Jami Masjid at Thatha

56. Surface Carving on a Mihrab at Thatha

57. The Tomb of Jam Ninou at Thatha

Diagram showing Relative Sizes and Heights of St. I'aul'.s . 1 the Gol
Gumbaz at Bijapur

BIBLIOGRAPHY

FOR complete information on the subjects dealt with in this mojjo-
graph consult the following works :

Report on the First Season's Operations in the Belgarn and Kalaclgi Districts.

(Burgess : 1874.)

Report on the Antiquities of Kathiawad ami Kachh. (Burgess : 1876.)
Report on the Antiquities of the Bidar and Aurangahad Districts. (Burgess :

1878-)

Report on the Buddhist Cave Temples and their Inscriptions. (Burgess : i88.v)
Report on the Klura Cave Temples and the Brahmanical and Jain Caves in

Western India. (Burgess : 1883.)

The Antiquities of the Town of Dabhoi in Gujarat. (Burgess and Cousens : 1888.)
On the Muhammadan Architecture of Bharoch, Camhay, Dholka, Champanir and

Mahmuclabad in Gujarat. (Burgess : 1896).

The Muhammadan Architecture of Ahmadahad. (Burgess : Part I., 1900.)
The Muhammadan Architecture of Ahmadahad. (Burgess: Part II., 1905.)
The Architectural Antiquities of Northern Gujarat. (Burgess and Cousens : 1903.)
Portfolio of Illustrations of Sind Tiles. (Cousens : 1906.)

Bijapur and its Architectural Remains, with Historical Outline. (Cousens :iQi6.)
The Antiquities of Sind, with Historical Outline. (Cousens : in the Press.)
The Chalukyan Architecture of the Kanarcse Districts. (Cousens : in the Press.)
Somanatha and Other Medieval Temples in Kathiawad. (Cousens : in the Press.)
The Medieval Temples of the Dakhan. (Con- ens : in the Press.)

THE

ARCHITECTURAL ANTIQUITIES
OF WESTERN INDIA

EARLY ARCHITECTURAL REMAINS

ARCHITECTURAL remains in Western India, so far as at
present known, begin with the Buddhist period. There are
found, in small numbers, rude stone monuments, such as
dolmens and cromlechs, of uncertain age; but, though they may be
considered buildings of sorts, they cannot be classed as architecture.
Relics of very great antiquity have, recently, been brought to light
in Sind, but it is doubtful whether anything in the shape of archi-
tecture, of the same period, will be found in connection with them.
Apart from Sind, which is a most promising field for the spade and
pick, and Gujarat, the rest of Western India gives but little hope of
buried remains of any extent, since the rock comes so near the surface
that there is seldom earth enough upon it to bury anything but small
and recent objects. Much of what was left upon the surface from
early times has been removed by the people as suitable material for
building their village homes; and, in this respect, old brick buildings
of all ages, when once abandoned, owing to dilapidation or desecration,
have suffered badly, the old bricks being more easily handled than
stone. Railways, roads, and canals have taken their toll of both brick
and stone as ballast, road metal, and building material. The hand of
the vandal has been very heavy.

In the following account, Western India will include, in addition
to the Bombay Presidency with Sind, districts immediately contiguous,
since territorial boundaries do not coincide with those of the various
styles of architecture that will be noticed. Thus it will be necessary
to overstep those boundaries into the Baroda State, the Nizam's

2 THE ARCHITECTURAL ANTIQUITIES

dominions, Maisur, and the Madras Presidency to pick up some oyt-
lying examples of the styles described.

Scattered throughout the length of Western India, from the
northern frontiers of Sind to the River Savitri, dropping suddenly
in one great cascade, 830 feet, into the Konkan, and separating the
Bombay Presidency from Maisur, is found as great and varied a
collection of ancient monuments as are to be met with in any other
part of India of equal extent; indeed, nine-tenths of the rock-cut
temples of the Buddhists are to be found within this area, Buddhist,
Jain, Hindu, Muhammadan, with the later Portuguese, Dutch,
Armenian, and English remains, help to make up a goodly assort-
ment. Stupas, cave and structural temples, mosques, tombs, palaces,
forts, churches, convents, and graveyards are all represented, often
quite near one another, succeeding each other from one end of the
country to the other. Just as varied as these relics of the mason's craft
are the peoples, languages, and religious beliefs which gave rise to them.

Apart from Muhammadan and European, there are practically no
remains of civil architecture to be found. Wells and tanks there are,
but these partake more of a religious character, it being a particularly
meritorious action to construct them as ensuring a happier state in the
hereafter. Hinduism was Hinduism, all over the country, and such
was the case with other religions, so that all people respected the
temples of their gods to whatever state they paid allegiance; they were
spared during invasion and raids, and, generally, the persons of their
priests were sacred. Not so with civil buildings, such as palaces, which
an enemy showed no hesitation in demolishing. Hence the construc-
tion of buildings of that nature was a bad speculation, and they were
run up cheaply, though, perhaps, of gaudy materials. They have all
gone the way of those who built them, but we get some idea of what
they were like from sculpture and painting, such as may be seen in the
sculptures at the Sanchi tope and the frescoes at the Ajanta caves.
Once a temple was desecrated it was abandoned, however costly a
structure it may originally have been; it then fell a prey to the vandal
for the sake of its material, or it was put to the basest of uses, from
a cattle shelter to a village latrine. Many of the earliest temples.

OF WESTERN INDIA 3

following those of wooden construction, were of brick firstly, brick
with wooden door-frames, pillars, and beams, and then brick with stone
door-frames, pillars, and beams. Of military architecture there f
very little beyond the ruins of an occasional fort or town defences, until
Muhammadan times.

Commencing, then, with Buddhist remains, in the absence of
anything of Hindu origin claiming priority, the cave-temples stand
first in Western India. Perhaps the predilection for cutting temples
in the rock is to be found in the great desire for lasting merit, which
the solid rock offers over a structural edifice; the merit gained by the
act would last as long as the work. These rock-temples, and the
monasteries and nunneries which generally accompanied them, are to
be found, mostly, in the lonely defiles and ravines of the Western
Ghats, but, in some cases, in or near ancient passes and communications
leading up from the seaboard to the Dakhan above. Some, again, are
found farther inland, cut in the sides of isolated hills and ranges.
They were, thus, secluded from the busy centres of population, where
the monks could, free from all distractions, concentrate their minds
upon their studies and worship amidst the silence of the hills and
forests. In these sequestered retreats they fashioned, in the bowels of
the mountain, columned temples and halls, finished, to the minutest
detail and ornament, with the utmost care. With wild beasts prowling
around, they chanted their evening services, and then barred and
bolted themselves in for the night. It must not be imagined that the
cave-cutters chose natural caverns to facilitate their work, and enlarged
and fashioned them to their requirements; a natural cavern means
rotten rock, where fragments may drop at any time, and so make
living within them very dangerous. They rather selected cliffs where
the rock was solid and free from cracks and fissures, and the commence-
ment of a cave, subsequently abandoned on account of hidden flaws
showing themselves, may still be seen. The driving in of the shafts
was all carried out by manual labour, no such thing as blasting helped
them, nor would it have been allowed, for it would have loosened the
rock overhead and have made it unsafe. From markings upon the
surface in unfinished excavations it appears that a tool in the shape of

4 THE ARCHITECTURAL ANTIQUITIES

a heavy pickaxe was used for roughing out, and chisels for the finish-
ing. The rock, in most cases, has been the Dakhan trap, no easy
material to work in. Many of the caves, like the great cathedral at
Karli (Plate i), were built by subscription. We learn this from
inscriptions, where we are told that such and such a man provided a
pillar; in other words, as much money as would suffice to cut and
finish one.

A close examination of these rock-temples shows that they follow
wooden prototypes, and this may be seen in the Nasik caves, where
not only are the beams and joists faithfully copied, but the wooden
pins in the ends are not even omitted (Plate 2). In the earliest chaityas,
which are vaulted chapels, wooden ribs exist to the present day, which
were evidently added in imitation of the arched bamboos that supported
the original thatch, and such primitive erections are still put up by the
Todas of the Nilgiris. When the cave-cutters gained confidence in the
stability of the rock, they ceased to use the wooden ribs and simply
imitated them in the stone in the later examples. Since these temples
were cut in the rock they have, with the exception of Kailasa at Elura,
no exteriors beyond the fa$ade; and that one does not help us to recon-
struct those of the earlier caves, as that is a comparative late excavation
whose exterior is just a copy of a Dravidian temple still existing. There
is, however, an early chatty a, built in brick, at Ter, whose exterior gives
us some idea of what the earlier ones were like, and is unmistakably
derived from a thatched original (Plate 9). What the exterior of the
flat-roofed temples or monasteries was like we cannot very well judge;
they were, possibly, mud roofs on wooden beams and joists.

The great chaitya, or cathedral cave, at Karli, between Bombay and
Poona, on the old line of communication between the Konkan and the
Dakhan, by way of the Bhor Ghat, is the finest of its class (Plates i
and 2). The outside has suffered badly by the fall of much of the
rock face, which has carried away the greater part of the facade and
court before the entrance. This might, in great measure, have been
spared had the excavators properly drained the slope of the hill im-
mediately above the facade. The inner fa$ade is fairly well preserved,
and is typical of most others of this class. It is a wall of rock closing

OF WESTERN INDIA 5

the front of the cave, and is pierced, below, with three doorways giving
access to the nave and two side aisles respectively, and with a great
horse-shoe archway above, which admits a flood of light and air into
the interior. The light through this arch illumines the dagoba, the
object of worship, which stands under the apse at the end of this
basilica. In this bright lighting of the dagoba the chaitya is the
complete antithesis of the Hindu temple, where the object of worship
is engulfed in the gloom of an inner shrine, and where a lamp is always
necessary to make it visible. Wood occurs, again, in the ornamental
lattice in the top of the archway. Immediately below this, and above
the doorways, is the music gallery, which must have been reached by
a ladder, since no other means of access exist. Passing within, we find
ourselves in a great cathedral-like nave, 45 feet high, with vaulted
roof, supported upon two rows of columns, which continue round, and
meet behind, the dagoba, and these are encircled, again, by narrow
side aisles also meeting behind them. Owing to the columns being
set very close together the side aisles are very dark. In the apsidal
end of the nave, the ribs, in the roof, radiate from a centre above the
dagoba. The length of the cave is 124 feet, and width, 45 feet 6 inches.
The earliest chaityas are very plain, as a rule, but this one is freely
adorned with figure sculpture and a kind of basket-work ornament very
common in caves of this age. Above the columns, and under the
architrave, are groups of persons upon elephants, all well carved.
These face inwards to the central nave; but, facing into the side aisles,
are men upon horses, and the latter seem relegated to this position, in
the dark, since the horses were failures compared with the elephants.
The pot bases and capitals of the columns are very characteristic of
this period about the beginning of the Christian era. The dagoba is
a representation of a stupa or burial mound; in this case, the stupas
built over the Buddha's ashes. It was the first, and only object of
worship, in the early caves : subsequently the image of the Buddha was
introduced. Surmounting the dagoba is the Tee, over which, again, is a
wooden umbrella, which, constructed of boards, dips on two sides only
and not all around. This is carved on the underside, but the smoke
and dirt that have accumulated upon it prevent it being seen from below.

6 THE ARCHITECTURAL ANTIQUITIES

In the top of the Tee, under the umbrella, is a deep square hole, which',
when we opened it, probably for the first time since the cave was
atandoned by the priests, was found to be full of ancient sawdust in
which, very likely, some relic had been deposited, and secured by the
close fitting stone lid. The outside of the facade had, originally, a
number of large inscriptions, some of which have been cut into, at
a later time, when pairs of erotic figures were sculptured. In the
north and south ends of the forecourt are sculptured life-sized
elephants, standing forward in half relief from the walls, and, above
them, are other pairs of figures, but which are of better proportions
than those below, and may be original. Just in advance of the outer
facade screen, which has mostly fallen away, is a tall pillar surmounted
by a lion, on the shaft of which are inscriptions. Although these give
no date, the style of the letters, and the architecture of the cave, would
put the chaitya at about the beginning of the first century before Christ.

Other chaityas of note are found at Bhaja and Bedsa, in the range
of hills opposite Karli, Elura, Ajanta, Nasik, Junnar, Kondani, and
Pitalkhora. They vary in details from the plainest at Junnar to the
ornate ones at Ajanta (Plate 3). The outer fa9ades of many of these,
like that at Karli, have been destroyed. The Visvakarma cave at Elura
is about the most perfect.

The flat-roofed caves were the residences of the monks, and their
refectories. Around the central hall are little cells, each with a stone
bench, against one wall, to sleep upon. The cells, in some cases, had
doors fitted to them, and, as very little air could enter the cave by the
main door, and, perhaps, a window, or two, one can imagine what the
interior must have been like when these were shut up for the night.
Some of the later viharas, for they go by this name, had, in addition
to the sleeping cells, a larger one off the back wall, which served as
a shrine for a colossal Buddha a private chapel, in fact. In addition
to the caves, rock-cut cisterns were always provided, which were sup-
plied by springs and by water running down the hill face and guided
to them by surface channels. In a refectory cave at Kanheri, near
Bombay, there are long, low stone tables running down the length of
the room, along each side of which the monks sat on the floor.

OF WESTERN INDIA 7

There are some interesting caves in a range of hills just outside
the town of Aurangabad not the famous Elura caves, which are much
farther away. They are Buddhist, in three groups, about three quarters
of a mile apart. The largest cave, in the first group, is very much like
the larger viharas at Ajanta, with the same type of pillars (Plate 4).
On entering the shrine, in which is a colossal Buddha, and after getting
accustomed to the gloom, groups of life-sized figures, to the right and
left, kneeling out upon the floor, loom up out of the darkness and are
rather startling; it requires a second look to know that they are not
really living beings. There are thirteen figures altogether. They
are remarkable for their head-dresses, which are full wigs, in horizontal
and vertical curls, and their features are very Egyptian in outline. Such
wigged images are found in old temples of the sixth and seventh
centuries, and one was found in the stupa which was excavated at
Mirpur-Khas in Sind (Plate 14).

In the second group of caves is one which, judging from the images
upon it, was possibly a nunnery. Nearly all the images, and those the
principal ones, are of females, whose head-dresses surpass those of the
worshippers, in the cave in the first group, in gorgeousness. Those of
the chief ladies remind one much of the bearskins of the Guards inter-
woven with strings of jewels and flowers.

Another piece of sculpture in this last cave is worthy of notice, as
it is rather rare, and this is a good example. It represents Avalokites-
vara or Padmapani, of colossal proportions, standing full to the front,
with eight little groups, four up each side of him. This represents the
Buddhist litany. The top group shows two persons kneeling, and
beseeching him to save them from fire which rises in a mass of flames
behind them. The second asks for protection from the sword; the next,
from captivity; the lowest, from shipwreck. The top, on his left, desire
to be protected from lions; the next, from serpents; the next, from
wild elephants; while the lowest shows disease and death in the shape of
an old hag.

There are remains of painting upon the ceiling of one of these caves,
where the Greek fret or key pattern is seen.

"STUPAS" OR CINERARY MOUNDS

S were, originally, great mounds of earth raised over the
ashes of a chief or religious leader, and surrounded by wooden
rails to protect them. Later they were built in brick or stone with
an earthen core, a stone railing taking the place of a wooden one.
Later on, again, the hemispherical mound was raised upon a dwarf
cylindrical drum, leaving a passage round the dome for circumambula-
tion in connection with funeral rites. Such is the general shape of the
early stupas built over the relics or ashes of the Buddha and important
priests; but, by degrees, the height of the basement increased until the
shape as seen in the stupa known as Thul Mir Rukan was reached
(Plate 10). This is, more or less, the shape of the rock-cut dagobas,
or imitation stupas, found in the cave temples. Decoration was lavishly
applied to the exterior, on both stone and brick, the ornamental details
being worked out in the brick as delicately as in the stone. The
stupas that have, up to the present, been found or uncovered in Western
India are of brick. The great Boria stupa, discovered and excavated
in Kathiwad, had a solid burnt brick core, but the find of a few
sculptured marble slabs showed that some stone work existed, prob-
ably as railing and crowning umbrella. In the Bombay Presidency
proper we have discovered three stupas, or rather the remains of
them namely, that in the jungles near Junagadh, already referred
to; one near the village of Supara, 33 miles north of Bombay,
supposed to be the Ophir of King Solomon's time; and one at
Kolhapur, in the Southern Maratha country. None of these gives
us any clue as to their original shape, so ruined are they. In Sind,
however, from a study of that of Mir Rukan and the remains of the
one unearthed at Mirpur-Khas (Plate n), it is possible to recon-
struct such as were built in that province. From the top of a high
square basement rose a cylindrical tower, finishing off in a dome, on

8

ARCHITECTURAL ANTIQUITIES 9

the top of which was, probably, a Tee and umbrella in stone or wood.
The sides of the tower had a slight slant, as may be seen in the photo-
graph of Thul Mir Rukan. The square basement of that stupa J|is
been destroyed, but the broken brickwork of the lower part of the
tower shows where it surrounded it. The core of these stupas was
generally built of sun-dried bricks, while the exterior casing was of
burnt brick, so well made that much of it, especially the decorative
parts, might be termed terra-cotta.

The general scheme of decoration consisted of rows of flat pilasters,
in tiers one above the other in the towers, standing upon a deep
moulded base, with niches between the pilasters containing images of
the Buddha. Highly decorative string courses further adorned the
walls. Much of the ornament, especially the quasi-Corinthian capitals
of the pilasters, has a classic look, and some details are almost line
for line copies. Around these stupas there appear to have been other
religious buildings, judging from the number of foundations found
around that of Mirpur-Khas. These were, no doubt, monasteries for
the monks, who lived upon the offerings of the faithful. Amongst the
ruins have been found portions of small votive stupas , and quantities
of clay tablets such as Chach found the samani making at the monastery
near Brahmanabad when he visited him.

Buried within the cinerary stupas were the ashes and other relics,
together with a miscellaneous collection of offerings, a small burnt
brick chamber receiving them. In the stupa which was unearthed at
Mirpur-Khas, in Sind, was found a roughly dressed block of stone, in
the flat top of which was a cup-shaped depression in which was a small
rock crystal bottle, within which, again, was a small silver case about
the size round of a lead pencil, within which was still another case, of
gold, containing a small particle about the size of a pin's head. That
this was the relic there seems to be no doubt, since the silver case was
wrapped round with gold leaf that had not been disturbed, and was
fresh and untarnished. In the crystal bottle, beneath, but not in the
gold and silver cases, was a quantity about an egg-spoonful of ash,
apparently bone ash. Another block of stone, with a corresponding
cup-shaped hollow, but shallower, capped the first stone, and the space

io THE ARCHITECTURAL ANTIQUITIES

around the crystal bottle was filled with sand, in which were a quantity
of seed-pearls, coral and crystal beads, a copper ring, a small gold ring,
aiVi some copper coins.*

In the stupa at Supara was found a more elaborate stone receptacle,
which had been carefully moulded and turned with a proper lid, and,
within this, the relic was found enclosed within several small reliquaries,
one within the other, surrounded by eight bronze images. The stupa,
of brick, was almost entirely destroyed.! In the great Boria stupa y
in the jungles near Junagadh, in Kathiawad, the burnt brick seemed
to have been used throughout the building, and the relics were found
embedded in this at a considerable distance above ground level. As
at Mirpur-Khas, the little caskets containing them were in a cup-shaped
hollow in the top of a roughly dressed block of stone. The caskets
were of stone, copper, silver, and gold, one within the other. The
relic was a small chip as big as one's small finger nail and about an
eighth of an inch thick.J

Other stupas that have been found in Sind are that near Tando
Muhammad Khan, one at Depar Ghangro, near Brahmanabad, and one
at Mohen-jo-dhado, north of Sehwan. These link up with others in
the Panjab which join those in the Kasmir valleys and neighbourhood.

Besides the cinerary stupas^ others were built to commemorate
places where the Buddha is supposed to have halted when in Sind. The
Chinese pilgrim, Hiuen Tsiang, in the seventh century, wrote : " When
Tathagata was in the world, he frequently passed through this country,
therefore Asoka-raja has founded several tens of stupas in places where
the sacred traces of his presence were found. Upagupta, the great
Arhat, sojourned in this kingdom, explaining the law and convincing
and guiding men. The places where he stopped and the traces he
left are all commemorated by the building of sangharamas and the
erection of stupas. These buildings are seen everywhere. 35 He also

* These are all described in full in my volume upon the " Antiquities of Sind,"
now in the Press.

t Described in the "Journal Bombay Branch R.A.S.," XV, 273.

t A further account of this stupa will be found in the "Journal of the Asiatic
Society of Bengal," LX, Part I, No. 2, 1891 (Misc.).

OF WESTERN INDIA 11

tells us that, in his day, there were several hundred sangharamas in Sind,
occupied by about 10,000 priests, and that they studied the Little
Vehicle. The thul Mir Rukan was possibly one of these commemorf-
tive stupaS) no relics having been found when it was examined many
years ago; but I am not satisfied on this point.

EARLY STRUCTURAL TEMPLES

THE earliest structural temple in Western India is, as far as
we know, without doubt, the Buddhist brick chaitya at Ter,
in H.E.H. the Nizam's dominions, and about 30 miles
east of Barsi (Plate 9). The late Dr. Fleet identified this place with
the long-sought site of the ancient city of Tagara of the early Greek
geographers, and an examination of the place, made subsequently,
shows that an ancient city did occupy this position. Heavy brick
foundations are seen, from which the villagers still dig out the bricks.
It was, evidently, a place of some importance, first with the Buddhists,
and then with the Hindus and Jains, and the place has its traditions
and legends, though the latter do not help us in the identification.
Most important, among the remains, is the old brick chaitya, standing
in the village and facing east, which has been appropriated by the
Vaishnavas for the worship of one of Vishnu's avataras, or incarna-
tions, Trivikrama. The image now occupies the position of the dagoba,
portions of which were found lying about. The building consists of
the vaulted chaitya with a flat-roofed hall before it. The former
measures 31 feet long by 33 feet high. It is just a plain copy, in
brick, of the rock-cut chaitya, whose waggon-vaulted roof rises to a
ridge on the outside, and is completed with an apsidal end. The
facade, above the hall roof, is a rough counterpart of that of the
rock-cut Visvakarma cave at Elura, excepting, perhaps, the little
niche holding a Hindu image, which is, probably, a later addition.
There was, in all likelihood, a small window here to let light into the
interior. Heavy mouldings, around the base of the walls and the
eaves, with slender pilasters between them, are the only decoration on
the outside walls, over which was a coating of plaster. There are no
pillars within the chaitya, its small size not requiring them. The
lotus ornament upon some of the sculptured stone fragments, is very

12

ARCHITECTURAL ANTIQUITIES 13

similar to what is found upon the Sanchi stupa and that which stood
at Amravati, near Bezvada.

There were also found at Ter several old brick temples in the
early Dravidian style such as are found in stone at Pattadakal and
neighbourhood, with all their decorative details reproduced in moulded
brick. They are of the same style and age as the old ones at Kukkanur,
in the Nizam's dominions, and may not be later than the seventh or
eighth centuries, but perhaps older. The doors, beams, and ceilings
are of wood, richly carved, stone being used nowhere. In this respect
they differ from the old brick temples at Sirpur, in the Central
Provinces, of the seventh and eighth centuries, where stone has been
used for these parts. It was not because stone was unprocurable, for
the country is all trap rock, in which cave temples have been cut at
Dharasena (now Osmanabad), 12 miles south-west of Ter. The
woodwork here is reminiscent of similar old work found upon very
old temples in the Chamba Valley in the Himalayas. They are dedi-
cated, as they always have been, to the older Vaishnava worship.
Outside the village, on the south ind west, are great mounds from
which the bricks are taken. The original city overflowed to the
opposite side of the river on the east, where huge mounds of debris
still exist. No exploration has yet been undertaken at Ter.

For other very early temples we must turn to Kathiawad and the
Bijapur district of the Bombay Presidency. Along the southern shore
of Kathiawad are found a few old shrines which bear a striking resem-
blance to those very early ones in Kasmir. The best and oldest
example is that at Gop, in the Barda Hills (Plate 15). Of this
particular temple, which is considered the oldest in Kathiawad, little
remains the interior walls and roof of the cella. It is so remarkably
Kasmirian in style that there can be little doubt of its place of origin.
The stepped-out pyramidal roof, with its prominent window-like arched
niches, and the trefoil arches around its basement, are marked features
of the northern style. There are, however, considerable modifications
of these in the next older temples found along this coast, due, no
doubt, to the descendants of the builders of the Gop temple, settled
in Kathiawad, losing touch with those of the north, and introducing

i 4 THE ARCHITECTURAL ANTIQUITIES

modified and new features of their own invention. This is seen in
the temples at Kadvar, Bilesvara, Sutrapada (Plate 16), and a few other
jJaces along the coast line, which, without the Gop example, might
have suggested an indigenous origin. The little arched niches, upon
the tower, with, sometimes, little heads in them, are but miniature
imitations of the great arched facades of the cave chatty as of the
Buddhists, such as is seen in the photograph on Plate 3. In the
Buddhist caves they are also found as a small ornament, placed at
short intervals along the larger heavy mouldings, and are frequently
shown with heads in them. They also occur in the old brick temples
at Ter, and are continued, more or less modified, in the medieval
Solanki and Chalukyan shrines, as will be seen farther on.

Passing to the southern end of the Presidency, old temples of
about the same age that is, between the sixth and eighth centuries
are found in many respects strangely like these old temples in
Kathiawad; but these, instead of travelling down from the north, had
their origin in the old Dravidian or Pallava buildings of the south of
India. They are found chiefly in the villages of Aihole and Pattadakal,
in the Bijapur district, and the immediate neighbourhood. The
Kathiawad temples have, in their sikharas or towers, the same stepped-
out pyramidal arrangement, heavy mouldings, quarter round in section,
and the same comparatively plain walls, decorated, when decorated at
all, with shallow pilasters at intervals, so very unlike the much cut up
and highly decorated surfaces of the later Solanki period. The plans,
too, in their simple designs, are much alike; indeed, the older Dravidian
or early Chalukyan, since this style was followed by the early Chalukyan
builders, corresponds to the northern examples, not only in the contrast
of styles as compared with the later work, but even in their dates. The
Chalukyas, like their kinsmen the Solankis, were great temple builders;
but, as the latter did not come into power until the end of the tenth
century, the former had several hundred years' start of them. The
oldest dated building of the early Chalukyas that we know of is the
temple of Meguti at Aihole a purely Dravidian one (Plate 17).
Here there is a definite date to start from, for an inscription upon
a large slab, built into its eastern wall, tells us it was built by the

OF WESTERN INDIA 15

Western Chalukyan King Pulikesin II. in the Saka year 556 (A.D.
634-5). It gives no information of the origin of the Chalukyas, but
they are said to have come from Ayodhya in Northern India. They
lived under the guardianship of the Seven Mothers, or Saptamatri, and
their favourite deity was Kartikeya, the god of war, while the god
Narayana presented them with the boar standard. The early Chalukyas
followed the Vaishnava cult, but, later on, they were equally attached
to the Saiva. The country, over which they eventually held sway,
embraced the southern portion of the Bombay Presidency with part
of the Nizam's dominions adjoining, the northern part of Maisur, and
portions of the Konkan. The extent of their possessions varied con-
siderably at different periods, according to their fortunes of war with
their neighbours.

Unfortunately, this temple of Meguti is not complete, having
suffered badly at the hands of Time and the vandal. It is possible
that it was never completed, the lower stage of the tower being left
as we find it. The interior is very dark, due, in some measure, to
the sides of the hall having been walled up, and dwarf rubble walls
added to its roof, with pot tiles through them for shooting through,
at some late period when the temple was used as a place of defence.
Windows in Hindu temples are purely ornamental, the interstices in
the trellis and arabesque designs admit but a very little amount of
light, and only make the gloom between them all the more dense;
and one has to be careful not to stumble over a dethroned deity or
drop into hollows from which the paving stones have been taken.
The windows in Meguti are no exception to the rule. Peering through
this semi-darkness, where the oil lamp has long ceased to burn, is
seen a colossal Jin seated upon his throne, too large to have been
brought in through the doorway. It was probably placed in position
before the walls were raised. There is nothing to show which
tirthankara it is intended to represent. Lying in the passage around
the shrine is a great slab bearing an image of a devi, which seems to
represent Ambadevi or Ambaji, a favourite goddess with the Jains,
whose two children are held by attendants, one on either side. The
walls of the shrine run up through the roof and form the sides of the

1 6 THE ARCHITECTURAL ANTIQUITIES

first story of the tower, which frequently contains an upper shrine in-
Jain temples.

The outer walls are very plain, being relieved by alternate square
projections and recesses. It was intended to embellish the exterior
with sculpture, for the faces of some of the panels have been left
rough for this purpose, and intermediate ones are sunk to receive
loose images, some of which were found lying about. A band of
small figures in panels, with some arabesque, runs round the plinth.

Just below the brow of the hill upon which Meguti stands, and
not far from it, is a curious two-storied temple which is partly
structural and partly excavated in the rock. It consists of two long
verandahs, one above the other, with a frontage of four heavy square
pillars and two pilasters. Off the verandah of the upper story are
a long room and three shrines, cut in the rock, and off the lower is
the beginning of a shrine. In relief, upon the centre of the upper
verandah ceiling, in front of the shrine door, is a figure of a small,
seated, clothed (Swetambara) Jina, with a triple umbrella above him.

Older, perhaps, than Meguti is the temple of Lad Khan in the
middle of the village (Plate 18). There is an inscription upon the
front wall, in characters of the eighth or ninth century, recording a
grant, but, as it has no reference to the temple, the latter was only
a convenient and permanent place to record it. There is no temple
that impresses one so much with its cave-like character. Its general
massiveness, its unnecessarily heavy columns, the simplicity of its
construction, and its plan and details have much more in common
with cave architecture than later temples have and with cave archi-
tecture not of the latest. The wooden forms, from which cave
architecture sprang, are seen throughout. The pillars are the most
characteristic feature, being remarkable for their great massiveness,
and are more suited to support the heavy rock roof in a cave than
the lighter one of a structural temple. The rock cutters, when they
built this, had not then learnt much concerning the relative strength
of materials, and so transferred their cave-cutting proportions to this
building. A comparison may be made between this and the interior
of the Tin Tal cave at Elura (Plate 4). The walls are not walls in

OF WESTERN INDIA 17

the ordinary sense of stone masonry, being composed of posts, at
intervals, joined up by stone screens containing lattice windows, just
as might be expected in wooden framing. Sometimes these lattio?
windows, in early buildings, are carved in imitation of reticulated
bamboo work. The flat roof, and its want of elevation, are cave-like
characteristics; but, more than anything else, the massive pillars, with
their roll bracket-capitals, evince a simpler and more dignified style
than many of those decorated ones in Cave III. at Badami, a few miles
away, which was excavated about A.D. 578 by Mangalesvara, the pre-
decessor of Pulikesi II., and have certainly an air of greater age than
that cave.

The very unusual position of the shrine, which is placed within
the great hall and against the back wall, has a very primitive look.
At first sight, it would seem that the building had been intended as
a simple hall or matha in which, by an afterthought, a shrine was
clumsily inserted to convert it into a temple. But this was not so,
as is clearly shown by the fact that, in the similar temples of the
Kont-Gudi group in the village, the beam from pillar to pillar, before
the shrine, has been placed on a higher level in the original construc-
tion, in order to admit of the lofty doorway being seen to its full
height, and lion brackets project beneath the beams, one on each
side, to further decorate the entrance to the shrine. The panel of the
wall of the hall, at the back, is blank, whereas those on either side of
it are perforated with windows. In the cave-temples we find shrines
within the main hall, notably at the Dumar Lena at Elura and at
Elephanta, and in structural temples in the north, such as that of Biles-
vara, in Kathiawad. The temple of Lad Khan was Vaishnava, but, at
present, it contains the linga and Siva's bull. On the block, above
the shrine door, which is the safest indication of the original dedication
of the temple, is Garuda, the vehicle of Vishnu.

Though the decorative details on this temple are spare, they are
vigorous and expressive; they are suited to their positions, and are not
so crowded and meaningless as they are in many later buildings. The
windows around the building are, as in Meguti, of perforated slabs
in various patterns; but the most decorated part of the temple is the

1 8 THE ARCHITECTURAL ANTIQUITIES

front porch, the pillars of which have life-sized images upon them in
bold relief. The water-pot ornament, which is seen here, is a very
''favourite device in the decorative details of many of these old temples,
and it occurs much elaborated with foliage in the caves. It is found,
again, but conventionalised, in the later medieval temples of North
Gujarat as part of the design of many of the pillars.

Another temple at this village, remarkable in its great likeness to
the chatty a caves, is that called the Durga temple, not on account of
its dedication to the goddess of that name, durga also meaning a
" fort," and it being within its precincts. Its plan is practically the same
as the Buddhist cathedral caves. As in these and the old chaitya at
Ter, the end of the temple is round or apsidal, and it has a central
nave and side aisles separated by rows of pillars. But, in the place
where the dagoba would be, is a shrine for an image. It was not so
easy, here, to imitate the vaulted roof of the chaitya as it was at Ter
in brick, so the nearest approach to it is the lofty central roof with
a lower sloping roof of slabs over the aisles. Above the shrine,
however, rises a sikhara, or tower, which the Buddhists do not appear
to have had. This tower, in what has been called the northern style,
is similar to those of about the same age in Orissa. Unfortunately,
the upper part has fallen, but a more complete example is that of the
temple known as Huchchimalli-Gudi outside the village.* Surround-
ing the temple, outside, is a pillared verandah, which broadens out
in front to form an open pillared hall, or porch. The pillars are not
quite so massive as those in Lad Khan, but still preserve the same
simplicity of design, being square blocks, without bases, surmounted
by very plain bracket-capitals. Those in the front porch are adorned
with pairs of human figures in full relief, and some are further
enriched with bands and medallions of arabesque.

The images in the niches around the walls, outside, are an impartial
mixture of Saiva and Vaishnava deities; for in those early days there
was not that sharp division between the two cults that crept in in
later times, especially with the rise of the Lingaits. There is no doubt

* These sikharas are described in Fergusson's "Indian and Eastern Architecture,"
Vol. II, pp. 96 and 98.

OF WESTERN INDIA 19

tBat the temple was originally Vaishnava, like that of Lad Khan, for,
over the shrine door, Garuda, the vehicle of Vishnu, presides. In
each hand he grasps the tail of a Naga which continues down each
side to the upturned body. The doorway is much after the style of
those of the viharas at Ajanta. The central niche, at the back of the
temple, which generally gives a clue to the deity originally occupying
the shrine, has been removed; and the absence of this image from other
old temples would lead one to believe that it was intentional, the
removal having been due to some other sect who had subsequently
appropriated the temple to their own form of worship. It is quite
possible that the Durga temple was dedicated to the Sun-god, Surya.

It is interesting to find, not only here but also at Pattadakal, a
few miles away, examples of the northern type of tower side by side
with others of the Dravidian, making this district a kind of meeting-
ground of the two. It is from the blending of these two that the
later Chalukyan style was evolved, though it eventually became a
separate local style of its own (Plate 17).

The temple of Huchchimalli, in the fields a short distance away,
is possibly older than the Durga one, which might be considered
contemporary with Cave III. at Badami. Its general style and heavier
cyclopean masonry give it a more ancient appearance, and it is
certainly older than Meguti. Here, again, the shrine is contained
within the hall, being placed towards the east end so that a passage,
or pradakshina path, is left around it. As already mentioned, this
temple has a spire in the northern style, but its details are more archaic-
looking than in that of the Durga temple. It is interesting to compare
it with the temple of Parasuramesvara at Bhuvanesvara, in Orissa,*
when it will be seen how strikingly alike they are. Fergusson, in his
first edition of his " History of Indian and Eastern Architecture," was
inclined to place the tower of the latter temple, with its cella, at about
A.D. 450, the mandapa, or hall, a little later. In the revised edition,
Dr. Burgess has brought the date down to the seventh or eighth
century, but this is, probably, too late an estimate. The outline of

* Fergusson's <4 History of Indian and Eastern Architecture," Vol. II, p. 96.

20 THE ARCHITECTURAL ANTIQUITIES

the tower has an older look than the Aihole one, though, in its detaifs,
the latter is simpler and heavier looking. But, in judging the age of
'these temples, local peculiarities have to be taken into account.

The interior of the temple is perfectly plain, excepting the shrine
doorway, which follows the style of some of the cave doorways. It
will be noticed that there are no pilasters in the walls opposite the
pillars as found in later temples. The interior of the shrine has been
totally wrecked, most likely by treasure hunters, and the paving of
the floor is in great disorder. Unlike the last two temples we have
considered, which face east, this one faces the west. Meguti, being
a Jain temple, and not bound to any particular direction, faces the
north. In the porch ceiling is a representation of Kartikeya, the god
of war, from which it is evident that the original dedication was Saiva.
Upon the front of the temple is an old Kanarese inscription which
records a grant of oil to the priest of the temple by the king, Vijaya-
ditya, in A.D. 708.

It is necessary to pass briefly over other temples in the village, of
which there are many, and all of more or less interest. The temple
known as the Kont-Gudi, which is of the same style as Lad Khan,
and, like it and most of the other temples in the village, in disuse
and neglected, is another Vaishnava shrine which faces the west. In
the hall, the shrine stands against the back wall and is more clumsily
placed than in Lad Khan. The central ceiling has an ashtadikpala
slab such as is frequently found in later temples. It is divided into
nine compartments, in the centre of which is Brahma, while the other
eight contain the eight regents of the points of the compass. Garuda
presides above the shrine doorway. Beside this temple there are
others, one being still embedded within the walls of the adjoining
houses and used as a dwelling.

Another interesting temple is in the fields to the south of the
village. This is in the same heavy massive style, but, in this case,
it is an advance upon the others in that it has a separate shrine in
continuation of the hall, with an almost complete tower of the northern
type. It has also image niches round the outside of the shrine walls.
In the central aisle, above the pillars, are some vigorous and well-carved

OF WESTERN INDIA 21

bands of dwarf-like figures and arabesque, while spanning the
architrave are great ceiling slabs, or rather were, for three, upon which
are representations of various deities, had been carried away many
years ago. Upon the dedicatory block we again find Garuda. The
shrine, which has been badly upset, has part of the seat of the image
within it. Upon the front of the tower is a large slab bearing the
tandava, or dance, of Siva, but this, and a similar one upon Huch-
chimalli-Gudi, both loose, were probably placed there at a later time.
Upon the south side of the temple is an inscription, in letters of the
seventh or eighth century, which seem to record the name of an
architect. It runs thus: "Hail! There has not been, and there
shall not be, in Jambudvipa, any wise man, proficient in (the art of
building) houses and temples equal to Narasobba." This may be the
name of the man who had charge of the reconversion to Saiva use.

The method of roofing the halls of these temples is worthy of
notice. The nave is covered over with great flat slabs lying across
from one architrave to the other, and the side aisles are covered
by similar slabs which slope downwards to the outer walls. As the
stones are not so accurately dressed as to make them watertight,
the joints are covered by long narrow stones, grooved on their under-
sides, and fitting into channels, cut down each side of the joint, in
the slabs. The one thus lies over and fits into the lower, much like
the half-pot tiles in ordinary native roofing.

Pattadakal, or, as it was known in ancient times, Pattada-Kisuvolal,
is now an insignificant little village, tucked away in rather an out-of-
the-way corner of the Badami district; but it must have been at one
time a great centre of religious activity, and, with Aihole and Badami,
an important spot in the kingdom of the early Chalukyas. When
approaching the village, one notices the many old temple towers which
rise above the roofs of the mud-walled houses (Plate 17). There are
old shrines, both in the Dravidian and northern styles, standing side
by side. Among the former are the three largest and most interesting,
since their inscriptions leave us in no doubt as to their origin.

It will be sufficient to describe two or three. The first and most
important, and one that appears to have been in uninterrupted use all

22 THE ARCHITECTURAL ANTIQUITIES

down through the ages by the Saiva community is that of Virupaksha
(Plate 20). While others have lost their original names, this one has
^retained that of the deity to whicb it was at first dedicated. Though
much later than those at Aihole, it has much of their cyclopean style
of masonry. It is, however, a far larger building, and is a great
deal more imposing. An inscription tells us that it was built for
Lokamahadevi, the Queen-Consort of King Vikramaditya II., a
descendant of Pulikesi, who erected Meguti at Aihole, in commemora-
tion of his having thrice conquered the Pallavas of Kanchi (Con-
jeveram). He reigned in the middle of the eighth century. In the
temple of Rajasimhesvara, at Conjeveram, he left an inscription, which
tells that he did not confiscate the property of the temple, but returned
it to the god. Again, the Vakkaleri copper-plate grant of his son
Kirtivarman tells us that, after Vikramaditya's conquest of Kanchi,
he made gifts to the Rajasimhesvara temple, and was so impressed
with the images and the other sculpture that he saw there, that *he
had them overlaid with gold. There is so much in common between
that temple and Virupaksha that there can be no doubt that he brought
away architects and masons to build another like it at his own capital.
And this seems to be confirmed in one of the inscriptions at Pattadakal
which speaks of the builder of Lokesvara (Virupaksha) as the most
eminent sutradhari (architect) of the southern country. Upon the
temple of Papanatha, a short distance to the south of Virupaksha, is
an inscription in praise of a certain Chattara-Revadi-Ovajja, who is
said to have "made the southern country" that is, was the builder
of the temples of the southern country, and hence, by inference, the
builder of this one. He belonged to the same guild, the Sar-
vasiddhiacharyas, as the builder of Virupaksha. The monolithic
temple of Kailasa, at Elura, already mentioned, is, to all intents and
purposes, a copy of this temple, although cut out of the solid rock, not
as a cave, but as a complete structural temple. In the reign of
Vikramaditya's son, Dantidurga, of the Dakhan, led a Rashtrakuta
army against the Chalukyas and took possession of the bulk of their
territories. At this time Dantidurga was cutting caves at Elura, so
that it is more than probable that he was so impressed with Virupaksha

OF WESTERN INDIA 23

fhat he had it copied in the rock by his cave-cutters. The line of
small elephants around the base of Virupaksha was represented in
Kailasa by almost life-sized ones, since it was necessary to raise th^
temple somewhat out of the pit in which it is cut. It is the most
northern example of Dravidian work (Plate 6).

The temple of Virupaksha consists of the sanctum containing
the great linga, or phallus, the spacious pillared hall with its porches,
the Nandi pavilion before it, sheltering the bull of Siva, and the
eastern and western courtyard gateways, linked up by the engirdling
walls. Over all, it measures 250 feet, of which the main building
occupies a length of 120 feet. Within the hall are eighteen heavy
square pillars supporting the roof, being arranged in four rows from
east to west, the two central rows having five pillars in each. The
pillars are all of one pattern, only the sculptures upon them differing;
and the very deep architraves above them divide the whole ceiling
into parallel aisles running east and west. This arrangement is
different to that usually followed in later temples, where the pillars
are arranged around a central square, the compartments of the ceiling
corresponding. Like those in Lad Khan's temple at Aihole, these
pillars have no bases; a roll bracket capital spreads the bearing under
the beam above. Upon the pillars are broad bands of bas-reliefs
representing scenes from the Mahabharata and Ramayana.

The exterior of the temple is a mass of heavy mouldings and
sculpture. The former run round the high basement and are very
vigorous and effective. They also follow the eaves of the main walls
as well as those of the stories of the tower. The sculptures, unlike
the stiff conventional forms of later Chalukyan work, are natural and
forcible, and they need no labels to assist in interpreting them. The
beautiful perforated scroll-patterned windows form one of the finest
features of the building.

We are informed by the inscriptions that the junior queen of
Vikramaditya, Trailokyamahadevi, a sister of Lokamahadevi, also built
a temple to the north of Lokesvara, which is easily identified with that
now known as Mallikarjuna, whose courtyard touches that of Viru-
paksha at its north-west corner. The inscription which gives us this

24 THE ARCHITECTURAL ANTIQUITIES

information also says that Virupaksha or Lokesvara was built to tht
south of the temple of Vijayesvara, which was erected by Vijayaditya-
gatyasraya (A.D. 696-733). This temple, too, exists, though neglected,
in the position described, under the name of Sangamesvara. It is a
very plain, simple, and massive building, but its hall has been badly
damaged. An inscription upon a large slab, which has been placed
inside the temple for safety, records a grant to the temple of the god
Vijayesvara. The name occurs several times in short inscriptions upon
the pillars.

The temple of Trailokyesvara, or Mallikarjuna, standing between
the other two, is of the same general plan and design as that of Viru-
paksha; the interiors, save for a slight increase in the width of the
latter, being exactly the same, even to the two little shrines, one on
either side of the entrance to the main shrine. It has not, however,
been completed, some of the shadowy forms of the sculpture, in
places, only just emerging from the rough. A comparison of the
towers of the two temples, Trailokyesvara and Lokesvara, shows that
they are alike in all respects save for the crowning member which, in
the former, is round instead of square. There is no doubt that the
square form of Virupaksha is more in keeping with the rest of the
tower. At Kailasa this member is octagonal. Garuda presides over
the shrine door, and thus indicates that this temple was dedicated to
Vaishnava worship.

The temple of Papanatha, in the south of the village, strikes quite
a new note (Plate 21). Its tower, which is fairly complete, is in the
northern style. Like the others, it faces east. It was built originally
for Vaishnava use, and was possibly dedicated to Surya-Narayana, or
the Sun-god; for, though Papanatha is a name of Siva, and was possibly
given to it by the Lingaits in later times, Garuda, the vehicle of
Vishnu, is found above the shrine doorway and that of the inner hall.
In the niche, at the back of the shrine, in the pradakshina, or circum-
ambulatory passage round it, is an image of Surya. Instead of an
image, a buckle-like ornament occupies the dedicatory block over the
outer or main doorway. The temple is in disuse as such, but is used
as a byre, cattle being tethered within, and along its walls without.

OF WESTERN INDIA 25

This is probably an older temple than that of Virupaksha, though
the pillars within, with their bases and cushion capitals, might suggest
a later date. They are, however, clumsy, and are not so well designed
as those of a somewhat similar pattern in the caves of Ajanta. Though,
at first sight, it looks very unlike Virupaksha, it will be found that
the difference is almost confined to the tower. Its plan is but a
modification of its neighbour; in fact, it is a nearer approach to that
of Kailasanatha at Conjeveram than to Virupaksha. The outer walls
are practically the same as those of the latter, a variation in the arrange-
ment of the sculptured panels alone being made. These temples, with
northern spires, are not a class by themselves apart, architecturally, from
those bearing Dravidian spires, save for these differences.

In the front porch are two massive pillars, without bases, upon
which are sculptures after the fashion of those in the porch of the
Durga temple at Aihole. One of them represents a warrior in his
war chariot, drawing his bow against his enemies who crowd around
him. Some of the many sculptures, upon the outside walls, show
scenes from the Ramayana, many of which are labelled in characters
which may be as old as the sixth century. The beams between the
ceiling panels of the larger hall for there are two halls are most
exquisitely carved and decorated in hanging arabesque, it being some
of the richest work of its kind in Western India. A continuous frieze
of little figures, bearing up the loops of a running festoon, such as is
frequently seen in early Buddhist ornament, trails along the top of the
walls of the hall.

As at Aihole, the village here is full of old temples, each one of
which has some feature peculiar to itself, either in design or sculpture.
In a Jain shrine, west of the village, are some very fine makaras y or
conventional monsters, with florid tails of arabesque, and life-sized
elephants flanking the shrine doorways.

In this corner of the Bijapur district are many other interesting
remains, notably at Badami and Mahakutesvara. Badami was an
early capital of the Chalukyan dynasty, when it was known under the
name of Vatapi, the old form of its present name; indeed, it is safe
to assume that it was in the hands of the Pallavas of the south before

26 THE ARCHITECTURAL ANTIQUITIES

it came into those of the Chalukyas. There is an inscription on one
of the rocks here which mentions Vatapi and speaks of the Pallava as
the foremost of kings. This inscription, judging by the script, is
possibly earlier than that in Cave III., which is dated A.D. 578. The
town occupies a very picturesque position between the horns of the
rugged hills, which embrace it, with the pretty lake behind. The
crags and boulders, covered with wild cactus and jungle the ever-
green tints contrasting with and enhancing the ruddy hues of the red
sandstone leave but a few boulder-strewn passages up through the
gorges and rents, between the detached masses, to the ruined forts above.
Perched upon an overhanging rock, on the north of the village, is the
temple of Malegitti, or the female garland-maker. It is a small but
compact temple, complete in all its parts, but now in disuse. It is
in the same heavy style as those we have been considering. The name,
as given in the inscription Malegitti-Sivalaya indicates clearly its
original dedication. There is an inscription on one of the pillars of
the porch of the time of the Vijayanagar king, Sadasivaraya, dated
in A.D. 1543, which records the construction of a bastion, which is
probably the one just above the temple. A similar inscription, of the
same date, in Cave III., records the construction of a bastion on the
southern fort, above the cave. Upon the hill, above the temple, is
another in ruins, while, upon the farther margin of the lake, is the
picturesque group of Bhutanatha of somewhat later date, making a
very attractive picture with its reflections among the reeds of the lake.
The temple and group of the Mahakutesvara shrines lie in a secluded
glen, with its trickling water and screw pines, between Pattadakal
and Badami, and are reached by a rough path across an intervening
tableland.

LATER CHALUKYAN TEMPLES

BETWEEN the earlier and the later Chalukyan temples there
appear to be no transitional examples to bridge over the gap,
excepting, perhaps, Kallesvara at Kukkanur, and the old Jain
temple at Lakkundi in the Dharwar district. About the middle of the
eighth century the Rashtrakutas of the north, whom we have already
seen cutting rock-temples at Elura, swept down upon the Chalukyas,
took possession of their dominions, and held them until about A.D. 973,
when a descendant of a side branch of the Chalukyan family regained
possession of their ancestral lands. This may account for the absence
of buildings during the period the country was in the hands of
strangers. Not only in style is there a great difference between the
earlier and the later work, but the material used has changed from the
rougher grained sandstone to the more compact, tractable, and finer
grained black stone known as chloritic schist, which dresses down to
a much finer surface, and has enabled the sculptors to produce so much
of that beautiful, delicate, lace-like tracery which characterises the later
work, and which it would have been difficult to produce in the coarser
material. With it the circular shafts of the pillars have been brought
to a very high state of polish (Plate 27).

Lakkundi, the ancient Lokkigundi, now rather a deserted-looking
village in the Dharwar district, full of the ruins of old temples
enclosed within the old fort walls, was once a place of considerable
importance. We learn that in A.D. 1191, the Hoysala king, Ballala II.,
made Lokkigundi his headquarters after finally extinguishing the
Chalukyan family and annexing their territory to his own in the
south.

The old Jain temple in the west end of the village is, probably,
the oldest temple here. We have far fewer temple inscriptions to
help us in dating these later temples, though the country is full of
inscribed tablets recording miscellaneous grants. A grant to a temple

27

28 THE ARCHITECTURAL ANTIQUITIES

may not give us any clue to the date of its construction, yet we can
gather from it the simple fact that the temple is not later than the
inscription. This Jain temple is not far removed from the temple of
Kallesvara at Kukkanur, which we may take as an intermediate example
between the earlier and the later style. In both the size of the
masonry has diminished, and we no longer find the heavy cyclopean
blocks as used in the former. They are yet sufficiently heavy enough
to be piled up without any cementing material, and, practically, no
through or bond stones. In fact, all through the period of the older
and medieval stone temples, here and elsewhere, no mortar or cementing
material was used. The masons depended upon flat horizontal beds
with the weight of the superstructure to retain all in place. When
their buildings did give way, through unequal settlement of the
foundations, or cracking of beams or supports, the masonry generally
rolled down like a house of cards. Their foundations were poor, and
of hardly any depth; many buildings were raised upon a layer of*
great undressed boulders spread upon the surface of the ground, and
this not for want of funds, since they lavished an abundance of ex-
pensive decoration upon the superstructure (see Plate 22). Over
most of the Dakhan the soil is firm, and the rock rises to or very
near the surface. In Northern Gujarat, where the soil is very sandy,
most of the more important stone temples rested upon a substantial
brick foundation.

The dedicatory block, above the doorway, which has frequently
been alluded to, is a small, square, projecting panel on the middle of
the lintel, upon which is generally carved an image of the deity to
whom the temple is dedicated, or of some member of his train; and,
thus, should the image in the shrine be missing, it is not difficult to
say what deity occupied it. An apparent exception to this rule is the
image of Gaja-Lakshmi, or Lakshmi with her elephants, frequently
found upon the outer doorways of Jain temples in the Chalukyan
districts, but never upon the shrine doors. This, in the past, has led
to many mistakes, for she is found upon the outer doors of nearly
all Brahmanical temples in this district as well. Because she was
seen in old photographs over Jain doorways, it was taken for granted

OF WESTERN INDIA 29

that all temples having her in the same position were Jain; but the
great majority of temples in the Chalukyan country are Brahmanical.
Lakshmi, here, was a favourite with both communities. Around *he
shrine walls of a temple are, usually, three principal images, one on
each side and one at the back. The one at the back gives another
clue to the original dedication of a temple, for it has some immediate
connection with the image in the shrine. Jain shrines have, as a rule,
less figure sculpture upon them, and what there is is a repetition of
the Jina which is seated in the shrine. A Jain image always presides
upon the dedicatory block over the shrine door of a Jain temple.

In Hindu temples the object of worship within the shrines of
the Saiva or Lingait cults is the linga, or phallus, of Siva, except in
temples to goddesses, which have an image of the appropriate deity.
In Vaishnava temples there is installed an image of Vishnu or of some
avatara or incarnation of him, as well as such allied deities as Surya-
Narayana. In Jain shrines is always found an image of one of their
twenty-four jinas, or tirthankaras (teachers), selected by the donor of
the particular temple as his special favourite. These are either clothed
or nude, according as the donor belonged to the swetambara (white
robed) or digambara (nude, sky-clad) sect, and are found either sitting
or standing with the arms hanging by the side in the latter case. The
digambara images are nude to the waist, and in this particular they
differ from Buddhist statues, which are shown with but one arm and
shoulder bare, or fully clothed.

Passing on to the more elaborately decorated temples of the tenth
to the thirteenth centuries, we may take that of Kasivisvesvara at
Lakkundi as a fair example. Plate 22 shows the south side of the
same. In this will be seen the storied arrangement of the tower, still
discernible, though much masked by the superabundance of ornament;
and it will also be seen how the vertical bands of the northern tower
are simulated by the arrangement of the trefoil niches, ranged, in line,
one above the other. As a purely decorative detail, the miniature
sikhara, or tower, of the northern type is often introduced into niches
around the temple walls. The rich lace-like work, for which these
temples are noted, is seen in the mouldings around the doorways, much

30 THE ARCHITECTURAL ANTIQUITIES

of it being so thoroughly undercut as to have the appearance of the
finest fretwork, standing away from the stone, and held to it by but
a lew tiny struts beneath (Plate 23). The village is full of these old
shrines, which are, now, nearly all deserted or used for other purposes.

Even more elaborate is much of the carved work in the old
temple of Mahadeva at Ittagi, 6 miles away across the border
(Plate 25). This is a much larger building than that of Kasivisves-
vara, it having what the other does not possess, a large open pillared
hall. An inscription gives its date as Saka 1034 (A.D. 1112), and
styles the temple a devalana-chakravarti, a very emperor among
temples, a title it, no doubt, fully deserved, as being probably the
finest temple in the Kanarese districts after that of Halabid in Maisur
(Plate 24).

The tower, as it now stands, rises in three tiers, or stories, with
a few of the lower courses of the masonry of the fourth which carried
the finial. These stages are quite distinct, and are not so cut up and
masked by decorative details as in Kasivisvesvara and others of the
same class. The little cusped niches, which adorn the middle of each
story, rising one above the other, are very handsome; and the delicate
work is further enhanced by the rich dark shadows in the niches. Some
eighteen to twenty-four inches of the basement is still covered up with
the silt of ages, the lowest mouldings being hidden.

The triangular corner slabs of the central ceiling of the outer
pillared hall are marvels of fretted stone (Plate 26). The great slabs
have been worked into a rich heavy mass of hanging arabesque foliage
and makaras, which emerge from the jaws of a kirttimukha y or
grotesque face. The spirited convolutions of the design, with their
circling excrescences and bewildering whorls, form, altogether, as rich
a piece of work as will be found amongst these temples. What adds
to the effect is the remarkably deep undercutting, so that the whole
of this petrified mass of foliage hangs from the ceiling by a forest of
little struts, hidden in the dark shadows behind it, and connecting it
with the main body of the slab. Similar deep carving may be seen
upon one of the Hoysala temples at Balagamve, just over the Maisur
border (Plates 27 and 28).

OF WESTERN INDIA 31

The great open hall, at the east end, was originally supported upon
sixty-eight pillars, of a totally different style from those of the earlier
temples. By this time the masons had learnt something about t4ie
strength of materials, and that a much lighter pillar was sufficient.
Twenty-six of these are large ones standing upon the floor and forming
the main support of the roof; the rest, which were dwarf pillars, stood
upon the surrounding bench, or scdile, and carried the sloping eaves.
The larger columns are of different patterns, but are arranged
symmetrically with regard to their designs. The four central ones
are of the most complicated design, and are similar to those in the
porch of the temple of Dodda Basappa at Dambal. There are also
round and square shafted pillars as in Kasivisvesvara at Lakkundi, the
round sections, as usual in this type of pillar, having been turned in
the lathe. It must not be supposed that they were turned in a hori-
zontal position, for that would have been impossible. They were
turned in a vertical position, and the fixed tool was brought up against
the roughly dressed shaft as it revolved. On many pillars the fine
grooves of the sharp-pointed tool are left, but, in the more finished
examples, the surface has been polished almost as smoothly as glass.
This may be seen in the interior view of the temple at Annavatti on
Plate 27.

Other highly decorated temples, of about the same period, are
found in many places within the old Chalukyan boundaries, and there
is hardly a village that has not some remains, either as the ruins of
an ancient shrine or inscribed slabs. The more notable remains are
found at Gadag, Haveri, Hangal, Bankapur, Kuruvatti, across the river,
Dambal, Chaudadampur, Niralgi, Harahalli, Galagnatha, Harihara,
Rattihalli, Balagamve, across the border in Maisur, Unkal, Degamve,
and Belgaum. The little temple of Sarasvati, the goddess of learning,
at Gadag has some elaborately designed pillars, and there are no others
that are equal to these for the crowded abundance of minute detail
which covers their surfaces (Plate 31). It consists, chiefly, of repeti-
tions of miniature shrines, tiny pilasters, panels containing Lilliputian
deities and attendants, rampant lions, and a host of other detail. The
little images are adorned with necklaces, bracelets, anklets, and a

32 THE ARCHITECTURAL ANTIQUITIES

profusion of other jewellery, each bead and jewel being fashioned
with the most careful and delicate touch. Other pillars in this shrine
arc decorated with a beautiful diaper pattern around their shafts, of
which there is just sufficient to make us wish for more. At Kuruvatti
some of the flying figure brackets are still to be seen (Plate 29). These
existed in most of the temples, but nearly all have been removed or
destroyed. The evidences of the religious zeal of the iconoclasts of
Islam are everywhere apparent among these shrines. At Dambal we
come across a temple built upon a totally different plan from those
generally met with at least, so far as the exterior is concerned
(Plate 30). The star-shaped plan, occasionally indulged in, was
obtained by revolving a square about its centre that is, the inter-
section of its diagonals. The angles of the square will thus travel
round upon a circle, and, if the angles are stopped at points equi-
distant from one another, they will form the periphery of the plan
of the proposed building, all the angles of which are right angles. The'
hall follows the same plan, though larger, the two giving an outline
of a serrated 8. But, less frequently, temples have been found where
the angles were not all right angles. A circular plan is not found,
the nearest approach being a circular corridor, containing images, sur-
rounding an ordinarily planned temple at the Marble Rocks, near
Jabalpur. The Hangal temple is noted for its very large domical
ceiling in the main hall, which rises, like some of those in North
Gujarat, in concentric circles of cusped mouldings, and then, at the
apex, falling again in a great rosette or pendant.

There is no part of Western India so full of inscribed tablets
and memorial stones as the Karanese districts that is, the country of
the Chalukyas. There is scarcely a village that does not possess one
or more; possessions that the people do not, in the least, appreciate,
where they do not lend themselves to some base use as a culvert slab
or a washing stone. Very few of these are now found standing upright
in their original positions; for the most part they lie about half buried
in the mud of fields; some are set up outside the village where the
kine can have a last rub before being turned in at night. Many have
perished. It is mainly from these lithic grants and deeds, often

OF WESTERN INDIA 33

demanding infinite patience in their decipherment, that the knowledge
we have of the dynasties who, with their feudatory chiefs, ruled over
these districts has been gleaned. These records are divided into two
classes those that record grants of land, money, or periodical pay-
ments in kind to temples, communities, or private persons; and those
that are memorials of deceased persons. The former have little
sculpture upon them beyond the sun and moon, indicating that the
grant is to last as long as those two luminaries, and the linga, or other
symbol, showing the religious sect to which the donor belonged, the
remainder of the stone being occupied with laudatory accounts of the
donor and his ancestors, the grant coming in as an afterthought at
the end. The memorial stones are more elaborately sculptured in
bands or panels, showing the manner of the hero's death, his ascent to
the heaven of his god, and the symbols of his faith.

Other remains of those times are columns, which were, as a rule,
set up before temples. There is one bearing a very important inscrip-
tion at Mahakutesvara. A curious one stands in the village of Bala-
gamve which has on the top a human statue with two birds 5 heads,
known as ganda-bherunda, the great enemy of elephants, set up here
to scare them away from the village crops. It is over 30 feet high,
and was erected in A.D. 1047. Before Jain temples they held images,
such as the chaumukha, and before Brahmanical shrines either images
or a brazier for a light. Such columns were even copied in the mono-
lithic temple of Kailasa at Elura, upon the top of which is the sacred
tnsula^ an emblem of Siva.

TEMPLES IN NORTH KANAKA

ADNG the seaboard, below the Ghats, in North Kanara, are
found some stone temples of curious construction, but very
much later than the Chalukyan, just described, having very
little in common with that style. Although this district was within
the old Chalukyan limits, comparatively little temple building in stone
was carried out at that period. The method of building, above the
Ghats, where the rainfall was very much less than in the Konkan, was
not so suitable there, where it is excessive, for such buildings, without
mortar joints, would be but sieves. It was left to the masons, under
the Vijayanagar kings, to devise a method of roofing that would meet
requirements. Similar buildings are also found in South Kanarar at
Mudabidri.

The most notable features in these temples, then, are the plain,
sloping roofs of flat, overlapping slabs, and the peculiar arrangement
of stone screens which enclose the sides. There is a great likeness
between these and similar ones, built in wood for the most part, in
Nepal, which has led some writers to wonder whether there could have
been any connection between the two. Perhaps the only connection is
that the same conditions produced the same result. These roofs may
be seen repeated in the thatched cottages in Bhatkal, where such temples
occur, even to the double story; and the like is seen in the Jami Masjid,
a modern construction in wood, which is, perhaps, more like the
Nepalese buildings than the stone temples. The barred screens, which
enclose and protect the sides, have their prototype in the large coarse
screens hung around the verandahs of the houses in the village, after the
fashion of Venetian blinds, where they are made of battens or slats
from the split stems of the supari palm, strung together with rope.
They are very reminiscent of the Buddhist rail. There is nothing of
much note within these temples the sculpture is very coarse and the
pillars are squat and ugly.

34

THE SOLANKI TEMPLES OF GUJARAT
AND KATHIAWAD

K TURNING to Kathiawad and Northern Gujarat, we find a
large class of medieval temples which are closely akin to those
of the Chalukyan districts. After the fall of Valabhinagar, the
ancient capital of the Valabhi kings, a new power arose in North
Gujarat with Anhillavada-Pattan as its centre. The founder of this
new dynasty was Wan Raj, of the Chavada family, whose members
are reputed by some to have been Sun-worshippers. During the reign
of one of his successors, bonds of matrimony joined up the families
of Anhillavada and Kalyana, and a Solanki (Chalukya) became King of
Anhillavada in the tenth century under the title of Mul Raj. With
this chief's accession began the building of those temples, the ruins of
which are now scattered about the country, and which reached its
greatest height during the reign of Siddha Raj and Kumarapala during
the eleventh and twelfth centuries. Indeed, Siddha Raj has been
credited with so many buildings that, at the present day, if a villager
be asked who the builder of any old ruin was, he promptly replies,
"Siddha Raj." The descendants of the old salats still exist in this
part of the country and follow their calling on the traditions of their
forefathers, but in such modified style as less wealth and less zeal have
developed. They still possess some of their old Sanskrit books, which
they profess to follow, but it is doubtful whether many of them can
understand them intelligently, since they are written in slokas, or verse,
which are often obscure in their meaning. It was these men who were
pressed into the service of the Muhammadans, when they gained the
ascendancy in Gujarat, and who built the mosques and tombs of the
conquerors; and this is the reason why we find these buildings so
Hindu in feeling, the only limitation imposed being the rigid exclu-
sion of figures upon their walls.

35

36 THE ARCHITECTURAL ANTIQUITIES

The largest and most elaborate temple of which we find remains
in Gujarat was, no doubt, the Rudramala, at Siddhapur, the city called
after Siddha Raj; of which just sufficient remains, together with the
help of more fortunate examples, to enable us to reconstruct it with
tolerable exactness. Siddhapur is upon the Sarasvati River, 64 miles
north of Ahmadabad, and it was originally called Sristhala, or Sristha-
laka. Mulraj, towards the end of his reign, about the middle of the
tenth century, wandering about the country in search of ease for a
troubled conscience, settled down at this holy place and commenced
to build a temple to his gods, but he does not appear to have com-
pleted it. It was reserved for the great Siddharaj, two centuries later,
to take in hand the unfinished and now ruined fane, and, in its place,
to raise such a temple as had not been seen in the land before the
famous Rudramala (the garland of Rudra or of the eleven Rudras),
or Rudramahalaya (the abode of Rudra or Siva), the scanty, but
colossal, ruins of which now remain embedded amongst the houses *of
the town, near the river bank (Plate 32). These remains consist of
five columns of the front or eastern porch, with their beams above
them; four columns in two stories of the northern porch; four greater
columns, also in two stories, which stood on the west side of the hall
and before the shrine; and one kirttistambha, or arch of fame, the
only one of a pair, in the courtyard at the north-east corner of the
temple. In addition to these are fragments of some small subsidiary
shrines at the back of the temple, which have been converted into a
Muhammadan masjid.

It was with the help of a salat from Pattan that the writer was
able to reconstruct the plan of this great building from the few
portions left and an intimate knowledge of the style. It was, no
doubt, the largest, or, at least, the second largest, in Gujarat, measuring
about 145 feet by 103 feet, rising in, at least, three stories. Another
temple, which stood at Vadnagar, 20 miles to the south-east, may
have been still larger, judging from the size of two kirttistambhas now
left, which are larger 35 feet 6 inches high than that at this temple
(Plate 41). Such magnificent piles were not likely to escape the
attention of the Muhammadans in their first onslaughts upon Hinduism ,

OF WESTERN INDIA 37

in this province; and in proportion to their grandeur was the complete
destruction that overwhelmed them. Between the front, or east, porch
and the flight of steps, or ghat, leading down to the river, stood the
Nandi pavilion, and, perhaps, before this again, a third kirttistambha.

These kirttistambhas were favourite additions to temples in this
part of the country; they seem to serve no definite purpose other than
ornament, but they serve that purpose well. There are, as a rule,
three in connection with a temple, one flanking each of the forward
corners of the building, and one in front of the entrance, facing the
river, as in the present case, or the flight of steps leading down into
an ornamental tank, as in the case of the temple of Surya at Mudhera.
They are not, however, equally graceful in outline, those at Vadnagar
being finer in every respect than the one at the Rudramala. It will
be seen that they are simply composed of two columns, of a larger
size, as we find in the halls of the temples, supporting a deep orna-
mental entablature and pediment. The squareness between the vertical
columns and the straight soffit of the cross beam is broken by the
light cusped torana, or flying arch, thrown across between them. This
torana, again, as will be seen in the Abu temples, was also a pet device,
whose convolutions are like some beautiful creeper running wild
between pillar and pillar in cusps and curves. Where a kirttistambha
is found singly beside an ornamental tank, as at Kapadvanj, it is,
probably, the only remnant of an old temple which has disappeared.
Among Chalukyan temples they are also found, but so much smaller
and more insignificant that they hardly catch the attention. In the
case of a number of these they served a special purpose, and were
erected in the courtyards of temples where, in olden times, a great
pair of scales was slung to the cross beam, on those rare occasions
when a king, in fulfilment of a vow, had himself weighed against gold,
which was distributed in charity.

A most imposing ruin is that of the temple of Surya at Mudhera,
a village about 28 miles south by west from Siddhapur (Plates 38
and 39). Around the great temple of the Sun, which stands upon
high ground and looks eastwards, over the village, to the rising sun,
the ground shows signs, in extensive brick foundations, of a former

38 THE ARCHITECTURAL ANTIQUITIES

town of much more consequence than the present little mud-walled
village. The temple, which is complete in all its parts, save for the
tower and the roofs of the halls, which have disappeared, consists of
the shrine and great closed hall, forming, together, one block, and
the detached open pillared hall in advance of these. In continuation,
eastwards, of the last is the ornamental Surya kunda, or tank, with
the two pillars of a small kirttistambha between them, at the head of
the steps leading down into the tank. Colonel Monier Williams, when
Surveyor-General, visited this temple at the beginning of last century,
and wrote of it : " There is one of the finest specimens of ancient
Hindu architecture at Mundera I ever saw. It is a pagoda very similar
in structure to those of the present day, but ornamented so profusely
that it is very evident the founder was determined to make it the
most finished piece of work that it was possible for the compass of
human art to effect. . . . All the upper part of it is supported on
pillars, which are of an order the most elegant, and enriched with
carved work of exquisite beauty, and which would be considered in
this refined age as the conception of a correct taste, and the execution
of a masterly hand. ... I do not recollect observing in any building
that I have seen in India such marks of the sheer effects of time as
many of the stones about this pagoda and tank display. We spent
some time every day in inspecting the place; but such is the variety
of its beauties that it would have taken a much longer time than we
had to spare to have discerned them all, or have gained a faint idea
of the general design."

The plan of Kumarapala's temple of Somanatha at Somanatha-
Pattan is practically the same as this one, except that in the former there
are no signs above ground of an outer pillared hall or tank. It is
likely that, that temple being dedicated to Siva, these adjuncts of the
Mudhera temple were replaced by the Nandi pavilion. The floor of
the shrine has been destroyed blown up by gunpowder, it is said,
by the Muhammadans. Beneath the original floor is an underground
shrine where, beneath the fallen debris from the tower and upper
floor, was found the seat of the Sun god with his seven horses upon
the front of it, but there was no trace of the image; that may be found

OF WESTERN INDIA 39

buried, upside down, in the pavement in front of some mosque. But
Surya is displayed in all the prominent niches around the walls, and
up and down the sides of the doorways. The shrine doorway has been
badly damaged.

The pillars are of the same pattern as are found in the temple of
Vimala Sha at Abu, with the graceful fretted toranas arching over
the span from pillar to pillar, and softening the otherwise rigid effect
of the straight square beam (Plate 40). The outside of the wall is
a mass of delicate detail, all carried out in the red sandstone so
universally used in this province. It is very doubtful whether marble
would have had any better appearance, except for sheltered interiors
such as at Abu; for where it has been used for exteriors, as in the old
Jain temple at Sarotra in the north-west of the Pahlanpur State, it has
mostly weathered to a dirty-looking, smoky black which does not show
the details to advantage.

The open hall is a magnificent pile of pillared splendour, even as
it stands in ruins; and with these old temples it is questionable whether
they looked better fresh from the builders' hands than they do now
with all the mellowing effects of the ages that have passed over them,
and when all the rawness of new masonry has vanished.

Anhillavada-Pattan, the old capital, has little to show in the way
of architectural remains, the reason for which is that, being a large
town, old material has been used up in later work, and we find many
an old sculpture or pillar in the walls of the town. In the country
around there are numbers of old ruins of interest which overflow into
Kachh and the borders of Sind. Pattan became a quarry from which
the kings of Ahmadabad and others carted away the material for
building their own palaces, mosques, and reservoirs. This, in the
old buildings, they got ready to hand, and, consequently, the only
remains now of the Solankis and Waghelas are but fragments. The
present town does not occupy the site of the old city, but is to the
east of it. Outside the western gate, stretching far across the fields,
are mounds, fallen walls, and other indications of the position of old
Pattan. On this side is the ruin of the great Sahasara-linga talav, or
tank, around which have gathered old legends of the time of Siddha-

40 THE ARCHITECTURAL ANTIQUITIES

raja. Most of the stone of its embankment has been taken away for
building purposes. Near this are the remains of the Rani's van), or
well, and, judging from the portions that remain, it must have been
second to none in Gujarat in size and splendour. Beyond this, on an
island in the river, is Sheikh Farid's tomb, partly a converted temple;
and, lying in the courtyard of the same, is a huge detached ceiling
panel of magnificent winding foliage. This great scroll, executed,
unfortunately, in friable sandstone, occupied the space in the ceiling
over the east porch, which has fallen into the river, but it has been
salvaged and placed in its present position. It is, certainly, the finest
piece of scroll work yet found in Gujarat. A copy, on a much smaller
scale, and not so delicately chiselled, is seen in one of the marble
ceilings at Abu. In a Jain temple in the town is the reputed image
of Wanraj, the founder of Pattan. There is some very fine work in
marble in many of the Jaina temples, but it is not, as a rule, very old.
In the adjoining province of Kathiawad, and also in Kachh, afe
many more of these temples, built by the same men, under the same
rulers and their tributaries, or local chiefs in the south with whom
the Solankis were frequently at war. One of the most notable of
these is the well-known and famous shrine of Somanatha at Somanatha-
Pattan (Plate 33). The present ruin is a rebuilding by Kumarapala
of that which was attacked by Mahmud of Ghazni in A.D. 1025. The
temple was then desecrated, but was not destroyed; its destruction was
carried out by the lieutenant whom he left behind him to govern the
country. Bhimadeva of Anhillavada, who had been hovering about
on the heels of Mahmud, appears to have retaken the town on his
departure and to have rebuilt the shrine of Somanatha; and it was in
consequence of it being again sacked, or of having fallen into ruin,
that Kumarapala undertook its reconstruction. Subsequently, the
building has been attacked on more than one occasion by the Muham-
madans, and has been converted into a masjid^ when two little minars
were set up above the main entrance and a new dome was put over
the main hall. It is now quite deserted, and near it lie scores of
old images from its walls and collected from various old buildings
in the town.

OF WESTERN INDIA 41

The plan of the temple probably followed that of the Rudramala
at Siddhapur, and was of nearly the same dimensions, being in length
some 140 feet over all. Though these two buildings were remank-
able for their size, compared with others of the same class, they were
by no means imposing when compared with sacred edifices in Europe.
St. Paul's Cathedral could contain, comfortably, three such buildings
within its walls. That there were larger temples than these in other
parts of India may be gathered from the size of the shrine of a colossal
temple all that remains of it at Bhojapur in the Bhopal territory,
where the shrine door measures about 30 feet in height by 1 5 broad
or about three times the measurements of that of Somanatha.

The temple, which faces the cast, consisted, when entire, of a large
closed hall, with three entrances, each protected by a deep and lofty
porch, and the shrine, which stood upon the west side of the hall,
having a broad circumambulatory passage around it, between the
inner and outer walls. This passage was lighted by a large balconied
window in each of its three sides, and these formed a very pleasant
feature in the general appearance of the building. It is quite possible
that, like the temple of Surya at Mudhera, this had an open pillared hall
in advance of the present main entrance, but there is no sign of it, now,
above ground. Of the original pillars and pilasters in the hall, which
are much like some of those in the temple of Tejahpala at Mount Abu
(A.D. 1232), only a few now remain, and these are in a very bad state
of decay owing to the corroding effect of the moist sea air and spray.
The interior of the shrine has shared in the general wreckage, but
retains most of its domical ceiling. The shrine door has been removed
and has been replaced by a roughly built one.

The sculpture, upon the exterior of the temple, has been so
effectually effaced by the despoiler that it is almost impossible to
identify the few images that remain. The general workmanship shows
it to be rather later than that of the best period that is, the eleventh
century, when the temples of Surya at Mudhera, Ambernatha near
Kalyan, the Rudramala at Siddhapur, and Vimala Sha on Mount Abu
were raised.

The well-known story in connection with the sack of Somanatha,

42 THE ARCHITECTURAL ANTIQUITIES

of Mahmud carrying away the gate or doors, is a myth, but may have
some foundation in fact. It is almost certain that, if the doors had
bien overlaid with silver embossed work, as was a common practice,
he stripped the silver off and appropriated it. The door brought back
from Ghazni, with so much ostentation, and which is now at Agra,
is an original door of Mahmud's tomb, the workmanship and propor-
tions of which are not Hindu. Learning by bitter experience, at the
hands of the Muhammadans, the Hindus, in their last rebuilding of
the temple, in the town, constructed a subterranean shrine in which
the linga was placed for safety, a hidden passage leading to it; another
linga was placed in the shrine above for ordinary use.

The ruins of another fine temple of this class are found at Ghumli,
in the Barda Hills, the seat of the Jethwas from the tenth to the
fourteenth centuries (Plate 34). Though not so large as Somanatha,
being less than two-thirds the length and breadth, it supplies some of
the missing parts of that temple. The tower has mostly fallen, but
there is sufficient left in the miniature sikharas, in the base of the
tower, to show what the whole tower was like, both here and at
Somanatha. A large, open, two-storied hall takes the place of the
closed hall of Somanatha. This temple, known as Navalakha, may,
possibly, be somewhat older than the other.

Of the same class, a very ornamental temple stands, in ruins, at
Sejakpur, about 6 miles to the south-east of Dolia Railway Station
(Plate 34). It, too, goes by the name of Navalakha, or nine lakhs
a favourite name for temples as well as palaces. It is not known what
the "nine lakhs" refer to; it may possibly be intended to convey the
idea that the building cost that amount of the current coin to build
it, or that the number of images upon it were so many that they were
past counting. It became but a high-sounding name for such build-
ings, like the thousand-pillared hails of Southern India, which never
contained anything like that number.

Though the temple is in a very ruinous condition, the crispness of
the carving upon it is almost as fresh and sharp as when it left the
sculptor's hand; and, being in a reddish-yellow sandstone of uniform
tint, the appearance of the work, under a brilliant sun, is very pleasing.

OF WESTERN INDIA 43

As usual, the plan of the building is designed with a number of
projections and recesses, and these, again, are further nicked out into
smaller ones in the basement mouldings, which is in striking contmst
with the simple plans of the very early temples. Unlike the star-
shaped plan of the Chalukyan district, this is worked out upon fixed
squares. By dividing a square into 121 smaller ones, 1 1 on each side,
and discarding 10 squares in each corner, by a zigzag line across the
corner, the approximate shape of the periphery of the shrine or hall
is obtained. The domical ceiling of the hall is constructed, not with
regular radiating voussoirs, but with horizontal rings, with the indi-
vidual stones lying upon flat beds, each ring being corbelled inwards
above that below, until they meet near the crown, where the curve
reverses in a central hanging rosette or pendant. The weight of the
masonry of the roof, packed around the dome, holds the whole together
without any cementing material. They are thus similar in construction
to those in the Chalukyan temples of the same age. Two very fine
examples are found in the temples of Vimala Sha and Tejahpala at
Mount Abu. The pillars are of the same type as those at Mudhera, the
dwarf ones around the margin having the conventional water-pot and
foliage decoration so common throughout Gujarat and adjacent districts
of that period. The filigree fretwork, covering the surface of the
tower, is as rich as the best found elsewhere. The whole temple is
built upon a high, solid, brick foundation.

Gujarat, Rajputana, and Kathiawad have been plentifully supplied
with wells, many of which, especially the step-wells, are not much
behind the temples in architectural pretensions. The descent to the
water-level is by a series of flights of steps with landings at intervals,
and a superstructure of pillars and architraves repeated in stories up
to the ground-level. As the steps descend, these stories increase in
number over the various landings. At the far end from the entrance
is the deep, circular well, with an arrangement at the top, on the
ground-level, for hoisting water for irrigation purposes. The advan-
tage of the step-well over the ordinary kind is that the water can
always be reached as it sinks from step to step : the disadvantage is
that, being so accessible, it is easily fouled. There is a fine example

44 THE ARCHITECTURAL ANTIQUITIES

at Ahmadabad, known as Dada Hariris, and an even finer one at
Adalaj, north of that place, and Wadhwan possesses two good ones.

^ Much attention was paid to the architectural appearance of city
walls and gates. Most old towns of any consequence were provided
with these, properly built in stone, in contrast to the mud walls of
ordinary villages. Among the most entire, still existing, are those of
Dabhoi, near Baroda (Plate 32). The walls are built of great blocks
of sandstone with heavy bastions at intervals, in which will frequently
be seen odd stones and sculptures from older ruined temples. Of
the gates, by far the finest is the Eastern, or Hira, Gate (Gate of
Diamonds). It was covered with handsome carvings representing
groups of warriors, animals, birds, and serpents, some of which
remain; but the gate has been, for the most part, rebuilt in Muham-
madan style. Upon this gate is a representation of a camel, an animal
not found in earlier sculptures.

From this account of the old architectural monuments of Western
India we cannot omit mention of the Jain temples on Mount Satrun-
jaya in the south-east corner of Kathiawad, though, with the excep-
tion of one or two, the buildings cannot be classed as ancient or even
medieval. But Satrunjaya is a place apart from all others, its position
being unique. Rising gently from the plain to twin summits, 2,000
feet above sea-level, linked together by a saddle, or shallow valley,
its top is crowded with hundreds of temples, all walled in, in separate
enclosures, with a total absence of dwellings of any kind (Plate 42).
The shrine of Adinatha dominates the southern summit, while a great
chaumukha temple surmounts the northern. This is one of the five
great tirthas of the Jains, the others being Samet Sikhara, or Mount
Parsvanatha, in Bihar, Mount Arbuda, or Abu, in Sirohi, Girnar in
Kathiawad, and Chandragiri in the Himalayas. This hill is sacred to
Adinatha, or Adishvara Bhagavan, the first of the twenty-four
tirthankaras, or great teachers, who were born into the world, at
different periods, to reform it, after it had fallen from grace.
Although great antiquity is claimed for many of the shrines, more
especially for that of Adishvara, the rebuilding and repairing of the
older ones have left but little, if anything, of the original work in

OF WESTERN INDIA 45

evidence anywhere. Inscriptions and mutilated images and simhasanas,
or image seats, built in, in fragments, in modern cells and shrines, do
not go back beyond the twelfth century. Among the oldest temples
may be reckoned the Panch Pandava, and, perhaps, that ascribed to
Kumarapala in the Vimalavasi tuk, or enclosure. The rest are later
erections, and the building of shrines still goes on apace, since the
building of a temple or the setting up of an image ensures to the donor
a percentage of the merit that accrues to those who worship, merit
being the royal road to salvation, or nirvana.

Forbes, in his " Ras Mala," describing Mount Satrunjaya, writes :
" There is hardly a city in India, through its length and breadth, from
the river of Sindh to the sacred Ganges, from Himala's diadem of ice
peaks, to the throne of his virgin daughter, Roodra's destined bride,
that has not supplied at one time or other contributions of wealth
to the edifices which crown the hill of Paleetana; street after street,
and square after square, extend these shrines of the Jain faith, with
their stately enclosures, half palace, half fortress, raised, in marble
magnificence upon the lonely and majestic mountain, and like the
mansions of another world, far removed in upper air from the ordinary
tread of mortals. In the dark recesses of each temple one image or
more of Adeenath, of Ujeet, or of some other of the tirthankaras is
seated, whose alabaster features, wearing an expression of listless repose,
are rendered dimly visible by the faint light shed from silver lamps;
incense perfumes the air, and barefooted with noiseless tread, upon
the polished floors, the female votaries, glittering in scarlet and gold,
move round and round in circles, chanting forth their monotonous,
but not unmelodious, hymns. Shutroonjye indeed might fitly repre-
sent one of the fancied hills of Eastern romance, the inhabitants of
which have been instantaneously changed into marble, but which fay
hands are ever employed upon, burning perfumes, and keeping all
clean and brilliant, while fay voices haunt the air in these voluptuous
praises of the Devs." Satrunjaya has a mahatmya, or legend, sup-
posed to have been composed by Danesvara at Valabhi, by command
of Siladitya, King of Surashtra, extending to about 8,700 lines in
Sanskrit verse, and containing 108 names for the hill.

4 6 THE ARCHITECTURAL ANTIQUITIES

The Jain shrines upon Mount Abu, an isolated hill within the
southern confines of Rajputana, are better known to tourists than
th*se upon Satrunjaya, being more accessible. But, here, only a small
portion of the top of the hill is occupied by them, the rest being
partly occupied by bungalows and barracks, perched among the crags
and boulders, as a hot weather resort. The plateau extends some
14 miles from north to south, while its breadth varies from 2 to 4 miles.
A hundred years ago, Colonel Tod was the first European to tread this
delightful resort, 4,000 feet above sea-level, and to thread his way
through the tangled lanes, hedged in with wild dog-rose, to look upon
the lovely pink blossom of its wild peach-trees, and to gaze upon the
fantastic forms of the Toad and the Nun Rock the black-robed sister
who is ever on her knees.

About a mile north of the civil station lie the Dilvada temples,
which take their name from the adjacent village, and the village from
the temples Devalvada, or Temple-hamlet. Another 7 miles north is
Guru Sikhara, the highest point, rising to 5,653 feet, and, about 4 miles
north-east of the station, is the old fort of the Paramara chiefs of Abu,
Achilgarh, beneath whose shattered battlements is the famous shrine
and sacred tank of Achalesvara. The Dilvada group consists of four
principal temples, each, with its subsidiary shrines and corridors,
contained within its own enclosed quadrangle, as is the usual arrange-
ment with the Jains. The principal temples are those of Vimala Sha
and Tejahpala, whose marble halls were added about A.D. 1032 and
1232 respectively (Plates 35-37 and 42). They are constructed, almost
entirely, of white marble, quarried in the plains to the north, and hauled
up miles of rugged hillside by some means that we are now altogether
ignorant of. Age and the weather have exercised their mellowing
influence upon the stone under cover, giving it a soft, creamy tint which
is very pleasing.

The amount of beautiful ornamental detail spread over these
temples in the minutely carved decoration of ceilings, pillars, door-
ways, panels, and niches is simply marvellous; the crisp, thin, trans-
lucent, shell-like treatment of the marble surpasses anything seen
elsewhere, and some of the designs are veritable dreams of beauty. The

OF WESTERN INDIA 47

work is so delicate that ordinary chiselling would have been disastrous.
It is said that much of it was produced by scraping the marble away,
and that the masons were paid by the amount of marble dust $o
removed. Amongst all this display of the sculptors' skill, the two
great domical ceilings of the halls of Vimala Sha and Tejahpala attract
most notice, and receive most praise. That of Vimala Sha is carried
out in the same style as those in other temples of the same period,
that in Tejahpala's is even more elaborate, the great central pendant
being a magnificent piece of work. How the weight of this hanging
mass of fretted marble is supported is a mystery. In the older temple,
that of Vimala Sha, the pillars of the hall are of the same type as
those in the halls of the temple of Surya at Mudhera and that of
Ambarnatha, and are all of a uniform pattern; but in the temple of
Tejahpala the architects had lost the early vigour of well established
forms, and had allowed themselves a freer fancy in rather nondescript
and whimsical designs. The general plan of this temple has been
copied from the older one.

Though the detail work of these temples is exquisite, and the
designs are not only novel in their treatment but extremely pleasing
to the eye, yet the setting of the work is at fault. The general out-
lines of the buildings are not in the best proportions. Such large
domes are set too squat; they require greater height. The corridor
ceilings, especially, are much too low, and the unnecessarily heavy,
massive beams, which separate them, are the antithesis of the fairy
lightness of the sculpture and gossamer tracery. The latter is literally
boxed up in deep compartments. As it is but a few feet above one's
head, it is hardly possible to see more than one bay at a time, and
only by straining the neck to view it. Had half the depth of the
beams been buried in the masonry of the roof, and the bald squareness
of their soffits been chamfered, the effect would have been improved,
and the panels and beams would have merged into one general ceiling.

In Jain temples there is a great sameness in the images, especially
of the tirthankaras. They are all of one pattern and one stereotyped
cast of features, and it is only by their symbols, or chihna, shown
beneath them, that we know the one from the other. Parsvanatha,

48 THE ARCHITECTURAL ANTIQUITIES

alone, differs from the rest in that he always has the serpent's multi-
cephalous hood above him. The long ranges of corridor cell-shrines,
winch surround the court of the temple, form galleries of sculpture
where the statues are all replicas of one another. Goggle eyes peer
out at one from every cell, in a stony vacant stare, from images that
ever sit in a calm and listless repose. Badly proportioned, straight-
limbed, and muscleless, they are but poor attempts to delineate the
human form. All Hindu and other Jain statuary, of this period, is,
as a rule, more or less characterised by these faults, and the figures
are often found in the most strained and unnatural attitudes. The
difference between male and female is indicated, chiefly, by the magni-
fied hips and breasts of the latter, the dressing of the hair occasionally
helping to distinguish them. This is the more remarkable since the
sculptors reduced their drapery to a minimum, apparently with the
express purpose of displaying the limbs *he better.

These two temples were each, originally, a small black stone shrine,
to which the marble halls and corridors were added, and were, no
doubt, in existence long before Vimala Sha's time. The present
image of Rishabhadeva, in the main shrine of Vimala Sha, is not the
original one, but is the second, or, perhaps, the third, in succession.
The earlier image or images were smashed by the Muhammadans, and
the curly-haired head of a colossal Jina, in black stone, was found
lying in an underground room under the south corridor. Going the
round of the corridor cells, it is found that nearly all the projecting
mouldings of the image seats have been broken away in the middle
by the falling of the images when thrown down. Many, in these
cells, are not the original ones, which were destroyed; the old inscrip-
tions upon the vseats record the setting up of different tirthankaras
from those now occupying them, as may be seen by their chihnas.
There is a story of one of the sultans of Mandu having made the
people eat their images by converting them into lime, the pan-supari^
in India, being always eaten with lime.

Jain temples are, generally, well supplied with inscriptions, and
these show that, in Vimala Sha's temple, the building of the corridor
does not appear to be quite coeval with the building of the great hall.

OF WESTERN INDIA 49

In the cells the earliest date is 1063, and there is only one of that date.
The next one is of 1 144, one of 1 145, five of 1 146, two of 1 150, six
of 1 156, one of 1 170, and one of 1 1 8 1. Soon after this last date there
would appear to have been special activity in the setting up of corridor
images, for there are twenty-five inscriptions dated in 1189, seve n i
which are over doorways. North Gujarat and Rajputana were overrun
by Mahmud of Ghazni, the idol-breaker, in his march upon Somanatha,
'via Ajmir and Anhillavada-Pattan; it is more than likely that his
troopers discovered the shrines of Abu, and wreaked their vengeance
upon them. So we find Vimala Sha, of Pattan, in 10/52, restoring and
improving the shrines of his fathers. Later on, again, North Gujarat
was invaded by the Muhammadans, still as keen as ever upon their
favourite pastime of idol smashing, under Qutb-ud-din and Muham-
mad Ghori in 1194 and 1195, and no images were set up during
those troublous times. In 1216 and 1217 some grants were made to
the temple, and two inscribed stones were set up to record it; but it
was not until more peaceful times returned, and the tide of Muham-
madan invasion had again receded, that, in 1232, the rich banker
brothers, Tejahpala and Vastupala of Anhillavada, turned their attention
to the desecrated shrines, and, copying Vimala Sha's work to a great
extent, added the beautiful marble halls and corridors to the old shrine
of Neminatha. Other temples of less pretentious workmanship were
added in later times to form the present group. Earthquakes, such as
occurred in 1825, 1849 or 1850, and 1875, have, no doubt, with
previous ones not recorded, been responsible for a great deal of damage,
especially to beams and pillars. An inscription upon an iron pillar,
at the temple of Achilesvara, records the utter rout of a Muhammadan
attacking party by a swarm of bees that chased them down the hill.

Another Jain tirtha is Taranga, about 26 miles to the east-north-
east of Siddhapur, which is a celebrated place of pilgrimage. It is
situated in a wild and picturesque spot high up in the hills. After
a long and weary ascent, on the western side of the range, through deep,
sloping sand, where for two steps gained one is lost, and up a lovely
glen shaded with great trees, down which ripples a stream of clear, cool
water, the pathway leads to a great gateway at the crest of the hill.

4

50 THE ARCHITECTURAL ANTIQUITIES

The way then winds, by a very gentle descent, into a great natural
amphitheatre in the hills, in the arena of which, surrounded by an
''extensive courtyard, stands the temple of Ajitnatha, built by Kumara-
pala (i 143-1 174). On several of the higher peaks around are perched
little canopies, like little watch-towers. The temple is one of the
largest Jain temples in Gujarat. On one of the doorposts an inscrip-
tion records the visit of Akbar and the names of those who entertained
him.

TEMPLES IN THE DAKHAN

BETWEEN the Solanki temples in the north and the Chalukyan
in the south are the so-called " Hemadpanti " temples which,
though not so numerous or handsome, yet cover quite as much,
if not more, ground. There is really no definite boundary, nor is the
class itself very different from its northern or southern neighbours,
with whom they are contemporaneous. They were mostly erected
during the reigns of the Yadava chiefs and local governors under them
in that part of the country between the Narmada and Krishna Rivers.
The Rashtrakutas, who preceded the Yadavas, busied themselves more
in cave cutting than in temple building, and we possess nothing in the
structural line that we can definitely ascribe to their period of rule.
Indeed, these " Hemadpanti " temples have no tradition and very few
inscriptions to connect them with any particular reign. Hemadpant,
or Ilemadri, was the famous minister of the Yadava king, Ramchan-
dradeva, about the second half of the thirteenth century, who, in his
leisure, found time to compose several literary works and to build
many temples, which tradition numbers at some three hundred.
Having been in a position to command the requisite funds it is
probable that, like the brothers Tejahpala and Vastupala of Gujarat,
he caused numbers of temples to be built, and started a revival in
building which was followed by others. The earlier or more
decorated temples in the Dakhan, however, were built long before
his time, but the villagers, in their ignorance, ascribe all old temples
to him. The true " Hemadpanti " temples are characterised by heavi-
ness, inclining to clumsiness, with severely plain exteriors. Very few
temples of this class exhibited image sculptures on the outside walls,
which would have invited the attention of Muhammadan iconoclasts,
who were at this time making their presence felt in other parts of
the country.

The earlier buildings are now in a very shaky condition, and in

5*

52 THE ARCHITECTURAL ANTIQUITIES

a more or less dilapidated state, chiefly due to the material used in
their construction the amygdaloidal trap of the country, quarried, as
-a rule, upon the spot. Though a hard, tough stone, it is full of flaws
and minute cracks, which render it very unsuitable for such parts as
beams supporting heavy masses of masonry. Inadequate foundations
helped to bring about disaster. For reasons of economy or lightness
brick has now and then been used for the towers and superstructure
above the cornice, which, when plastered and moulded, and discoloured
by age, can hardly be distinguished from the stonework below. The
stone walls, which were of variable thickness, and far heavier than
would be built at present, were run up in two shells an outer and
an inner the space between being filled up with loose boulders or
dry rubble. There is a conspicuous absence of " through " or bonding
stones or clamps, and no cement or mortar was used. Thus, in many
cases, where the outer shell has fallen away, the inner has remained
standing intact and supporting the superstructure.

The plans of these temples are either rectangular or star-shaped
in their general outline, and revel in a great multiplicity of angles,
which are carried up through the walls and the towers. These, again,
are cut up by numerous horizontal mouldings into a bewildering
mass of projections and recesses, with their sparkling lights and deep
shadows. The star-shaped plan is not so common as the others.

The temple of Ambarnatha, though not in the Dakhan, strictly
speaking, is, from its solitary position, more conveniently taken with
the Dakhan temple group. The little village of Ambarnatha is
situated about 4 miles south-east of Kalyan, the junction of the
north-eastern and south-eastern branches of the railway near Bombay.
Beside a small stream, to the east of the village, stands the venerable
pile of the ruined temple of Ambarnatha (Plate 43). It is built in
black stone, and is now in disuse, save for the attention of a solitary
pujari who, once or twice daily, strews flowers upon the deserted
linga. Upon a beam over the north door, inside, is an inscription
which tells us that in Saka 982 (A.D. 1060) this temple was built
(possibly upon the site of an older one) during the reign of the
Silahara chief Mummuni, or Mamvani. It is thus of the same age as

OF WESTERN INDIA 53

the temple of the Sun at Mudhera and that of Vimala Sha at Abu,
and was just fresh from the builders 5 hands when William of Normandy
set foot upon British soil. It measures 89 feet by 73^ feet. It is
closer, in style, to those of the same age in Gujarat than to those in
the Dakhan above.

The building faces the west, and is made up of the sanctum and
mandapa, or hall, the latter being provided with three entrances, each
with its own porch. The central ceiling of the hall is supported by
four pillars, and these, with six others in the three porches, are all
the free standing ones found in the temple. These pillars, which are
alike, are similar in style to those in Vimala Sha's temple at Abu.
The general plan is peculiar, being, apparently, made up of two squares
set diagonally to one another, touching corner to corner the smaller
being the shrine and the larger the hall; but, in reality, it is formed
of two squares touching side to side, whose sides have been whittled
down to narrow central panels by the deep recessing of the corners
and by a line of angles running straight between the diminished sides.

The floor of the shrine is sunk below the outside ground-level,
and some eight feet lower than that of the hall. It is approached by
a flight of steps, space for which is obtained by sacrificing nearly the
whole depth of the usual antechamber, thus bringing the shrine door
considerably forward. There seems to be little doubt that the original
shrine was on about the same level as the hall, but that the fall of
the top of the tower crushed down through the floor and destroyed it.
The temple of Surya at Mudhera, as we have seen, was also provided
with an upper and a lower shrine, the upper one having been blown
up by the Muhammadans. We have also noticed the sunk shrines
of Pattan-Somanatha and Ahmadabad. The Hnga, in this lower shrine,
is a projection, through the floor, of the natural rock, and, probably,
for this reason was considered of more account than that which
occupied the shrine above. A rock linga^ like this, is found in the
temple of Omkara at Mandhata, in the Narmada, in the Central
Provinces.

The ceiling panels, within the hall, are very richly decorated in
the prevalent style of the best work of the eleventh century. The

54 THE ARCHITECTURAL ANTIQUITIES *

central dome, owing to its small size, the black stone, and the badly
lighted interior, does not show to advantage. In these black stone
temples, with closed halls, the interior decorations are hardly seen
until one's eyes get accustomed to the gloom.

A noticeable feature in the basement mouldings, on the outside,
is the torus or half round moulding as found in very old temples.
It will hardly be found again in later work, where it changes to a
knife-edged moulding, somewhat wedge-shaped in section, as in the
temples at Balsane in Khandesh.

The Mummuni of the inscription was one of the Silahara maha-
mandalesvaras of the Northern Konkan, whose capital was at Puri,
which was, possibly, situated on Salsette Island. These Silaharas were
tributary to the Rashtrakutas and the later Chalukyas of Kalyana in
the Dakhan. Mummuni was thus, in all probability, a feudatory of
the Chalukya king Somesvara I.

The finest group of temples in the Dakhan, but of later date than
the last, is that of Gondesvara at Sinnar, 20 miles to the south of
Nasik (Plate 44). The town, in old records, goes under the names of
Sindinagara, Seunapura, and Sindinera. In A.D. 1025, Bhillama III.,
of the Yadava family, was ruling at his capital of Sindinagara as a
feudatory of the Chalukyan king Jayasimha II. It is also recorded,
in a grant of 1069, that Seunachandra I. founded the town of Seuna-
pura. According to tradition, however, Sinnar was founded by a
Gavali (Yadava) chief, Rao Shinguni, about seven or eight hundred
years ago, and his son, Rao Govind, is supposed to have built the
great temple of Gondesvara, or Govindesvara. It is possible it may
have been named after Govindaraja, one of the Yadava princes, who
ruled about the beginning of the twelfth century.

This temple is the largest and most complete in the Dakhan. It
is a panchayatana group that is, it consists of the central temple
with four other small ones arranged as satellites around it, thus making
the five that the name implies. The main temple, containing the
linga, is dedicated to Siva, whereas the smaller ones probably con-
tained images of Vishnu, Ganapati, Surya, and Parvati, the consort
of Siva. Before the main entrance of the temple stands the Nandi

OF WESTERN INDIA 55

pavilion, m which is Siva's sacred bull, and, around the whole, was
a wall, enclosing a great courtyard, with two entrance gates. It is
all built of the amygdaloidal trap, most likely from the quarry, no^
filled with water, on the east side of the group. The masonry has,
therefore, not weathered well, the disintegrating action of the wind
and rain having left the exposed portions of the finer work rather
honeycombed.

The plan and general arrangements are much the same as in the
temple of Ambarnatha, the length over all being 78 feet. The style
and disposition of the mouldings of the basement and walls are
similar, except that the latter has a greater display of images, which,
to a great extent, are here replaced by lozenge-shaped ornament. The
figure sculpture is chiefly confined to small images around the porches,
and these are very poorly carved; but there is a particularly fine piece
of carving in the gargoyle through which the waste water from the
shrine passes out on the north side. It is in the shape of the con-
ventional makara, which is seen in the fine architrave over the shrine
entrance in the little temple of Aesvara, to the north-west of the town.
The tower and the roof of the hall are practically the same as in
Ambarnatha, but more complete. The crowning member, however,
as seen in the photograph, is not original; it is the villagers' idea of
restoration a Muhammadan dome on a Hindu temple.

The little temple of Aesvara is in the Chalukyan style, and so
does not belong to the class of work found generally in the Dakhan,
and it is the most northerly example of that class. There are some
details that appear to have been borrowed from northern temples,
such as the kichaka, or small squat bracket figure under, and support-
ing, the beams at the capitals of the pillars, which is not found in
Chalukyan work. A very favourite Chalukyan ceiling is found here,
representing the ashtadikpalas, or guardians of the eight points of the
compass. Above the shrine door, which is fairly well decorated, is a
line of images representing the saptamatris, or seven mothers, another
sculpture peculiar to Chalukyan work, and on the dedicatory block
above the doorway is found the Chalukyan Gaja-Lakshmi, instead of
Ganapati, which is usual above Dakhan doorways.

56 THE ARCHITECTURAL ANTIQUITIES

A most superb piece of carving overarches the entrance to the
antechamber, which, in three bands, rises from the jaws of two makaras,
/)ne on either side, with the most wonderful tails of flowing arabesque,
which turn up over their backs, and forward, on to the arched bands.
The central semicircular panel is occupied by a representation of Siva
dancing the tandava (Plate 45).

The village of Balsane, in Khandesh, possesses a group of nine old
temples of about the same date and style as that of Ambarnatha, but
they are in ruins and uncared for save by the Archaeological Depart-
ment. In a field, just above the bank of the stream, are three. The
largest one, which has three shrines around a common hall, is most
elaborately and profusely sculptured, but has suffered badly from the
weather (Plate 45). Parts of the sikharas, or towers, of two of the
shrines remain, but that of the third has fallen. Triple-shrined temples
are not uncommon, and are found, also, among Chalukyan and Solanki
temples, but they do not always contain the same deities. What may
be called the main shrine, that facing the west and the entrance to
the temple, is dedicated to Siva, and contains his linga and bull. The
other two shrines are now empty, but, judging from the images on
the outside, the south one contained a Vaishnava image. The north
one has two images of goddesses, so it is uncertain what the presiding
deity was. The outer walls are more richly encrusted with ornamental
detail even than those of Ambarnatha; in fact, we have here the style
in its fullest development, crystallised into its richest details, and
sparkling with light and shade from basement to summit.

On the east of the village is a temple with a single shrine and
a hall with three porches, facing east. Like the last, it is in ruins, and
the shrine has mostly disappeared; it can be seen that it was of the
star-shaped plan, but the corners have sides at different angles with one
another. Unlike the temple at Dambal, which has both shrine and
hall on a star-shaped plan, it is confined in this to the shrine only.

With regard to the orientation of temples, it may be stated that
Saiva shrines only face east or west, Vaishnava ones may also face the
north, especially those to Krishna. Temples to Ganapati face the
south, those to goddesses the north. Jain shrines are found in

OF WESTERN INDIA 57

all directions. Muhammadan mosques face the east that is, the
worshipper, with his face towards the mihrab, is facing Makka.

In the group to the east of the village of Balsane is one of *
very different plan. Around the sides of the large closed hall is
a series of ten small shrines in addition to the main shrine at the
back. It is evidently a Vaishnava temple, the small shrines having con-
tained images of the ten avataras, or incarnations, of Vishnu. In the
main shrine, which probably held an image of Vishnu, is now placed
the image of a horseman, which may possibly be that of the Kalki
avatara, taken from one of the side shrines, and which escaped the
profane hands of the idol-breakers.

Other old shrines are found scattered about through the Dakhan,
both of the class we have just been considering as well as the later
" Hemadpanti," especially at Pedgaon and Kokamthan in the Ahmad-
nagar district, Malsiras and Velapur in the Sholapur district, Anjaneri,
Devalana, and Jhodga in the Nasik district, and at Vaghli and the
deserted city of Patan in Khandesh.

MUHAMMADAN BUILDINGS

AHMADABAD AND GUJARAT

FOLLOWING close upon the period of the later medieval Hindu
remains come the mosques, tombs, and palaces of the Muham-
madans. The advent of the uncompromising followers of the
Prophet checked the further development of temple building for a
time, and what little was carried out, with very few exceptions, was
in keeping with the chastened mood of the Hindu builders. Money
spent upon raising such piles as those we have been considering would
have been a risky investment, and the salats and builders of the temples
found more certain employment in raising the mosques and tombs,
palaces and public works of their new masters. On temples con-
structed under the new conditions the builders were careful to show
as few images upon the outside as possible, so as not to excite the
puritanical feelings of the idol-breakers. In addition to the secret
underground shrines, in which they might hide their images, we find
the Jains, at Satrunjaya, building miniature idgahs, or Muhammadan
places for prayer, upon the roofs of their temples, knowing that the
Muhammadans will not willingly destroy a mosque, or idgah, when
once built, and the temple could not be demolished without the idgah
being involved.

When the Muhammadans began to settle down in the districts
they had raided, they made it an almost universal practice to build
their first mosques out of the material of the demolished shrines,
and often upon their very sites. In many cases they were content
to adapt, with a few alterations and the mutilation of images, the
temples as they stood, as was the case with the temple of Somanatha
in Kathiawad, and the temple in the fort at Bankapur in the Dharwar
district. In the latter, ignoring the shrine, they built a wall across
the west side of the hall and converted it into a masjid. This was

53

ARCHITECTURAL ANTIQUITIES 59

the quickest and readiest way of meeting their immediate needs; and
they followed this up with more pretentious buildings, such as the
Jami Masjid, at Somanatha-Pattan, constructed wholly, or largely, out
of temple materials, but on their own plans. The Hindus, unlike
the Muhammadans, did not hold their own temple material too sacred
to be used again for any other purpose. The Muhammadan, on the
other hand, looks upon every stone of his mosque, even when the
latter is totally ruined, as having been consecrated to God's use, and it
is a desecration to use it in any other way. When several great slabs,
bearing long Sanskrit inscriptions, were discovered built into the wall,
with their faces inwards, in the mosque known as Raja Bhoja's school
at Dhar, in Central India, though they allowed them to be replaced
by others, they absolutely refused to let the old slabs leave the mosque,
and so they have been framed and left there.

The principal centres of Muhammadan building activity in
Western India, during the fifteenth, sixteenth, and seventeenth cen-
turies, were Ahmadabad and its surrounding districts, Champanir,
Broach, Cambay, Bijapur, and Sind; and there are many isolated
notable buildings in other places. Those of Ahmadabad, North
Gujarat, and Kathiawad were mostly erected during the period when
the Ahmad Shahi family ruled this part of the country, after its first
king had thrown off his allegiance to Delhi, and before Akbar again re-
conquered it some hundred and sixty years. Ahmadabad is said to
have been founded by Ahmad Shah, the grandson of Muzaffar Shah,
who had asserted his independence in A.D. 1412. During his reign
were raised some of the finest buildings in the city his own private
mosque in the Bhadr, or citadel, the Jami Masjid, the people's mosque,
Rani Sipri's mosque and tomb, Haibat Khan's mosque, Sayyid Alam's
mosque, the Tin Darwaza, or triple gateway, which spans the principal
street of the city, Sidi Sayyid's mosque, which has two noted windows,
most superb specimens of the sculptor's art, and his own tomb, with
that of his queens. In the suburbs was erected Malik Alam's mosque,
about 2 miles to the south of the city.

Muhammad Shah, who succeeded him, in 1443, appears to have
done very little during his short reign; Qutb Shah, however, added

60 THE ARCHITECTURAL ANTIQUITIES

to the embellishment of the city by building his own mosque in the
Mirzapur quarter. During his time were completed the mosque and
tomb of Sheikh Ahmad Khuttu Ganj Bakhsh at Sarkhej, the Kankaria
tank, Dharya Khan's tomb, and the Batwa and Usmanpur mosques.

Mahmud Bigarah, following, in 1459, added many more buildings
to the city during his long reign. He was the most famous of the
kings of Ahmadabad, and it is said of him that " his personal strength,
courage, and military skill are as conspicuous as his religious bigotry
and his stern but far-sighted statesmanship. His love for architecture
is attested by the cities of Mustafabad and Mahmudabad, near Kheda,
as also by the numerous and elaborate additions which his nobles,
following his taste, made to Ahmadabad and its environs." During
his reign were built the mosque and tomb of Sayyid Usman, on the
west bank of the Sabarmati, Miyan Khan Chisti's mosque, Muhafiz
Khan's mosque, Achyut Bibi's mosque and tomb, Dastur Khan's
mosque, Dada Harir's well, and the Shah Alam group, 3 miles to the
south of the city.

Five other rulers succeeded Mahmud Bigarah until 1572, when
Akbar reduced the province, and left a governor to carry on its affairs.
The Marathas made their appearance about the beginning of the
eighteenth century, and carried their plundering raids into Gujarat.
The viceroys, much enfeebled by the absence of aid from either Delhi
or surrounding chiefs, could not cope with these new enemies, and,
"in A.D. 1755 the Muhammadan power in Gujarat was finally ex-
tinguished, and the Marathas, though divided among themselves,
took their place. They burnt and plundered the property of friend
and foe with almost equal energy and spared neither mosque nor temple
which it suited them to destroy. Innumerable are the architectural
monuments which have thus suffered from their wantonness and
malice, and the interesting ruins of which the very materials have
been carried away for building purposes. Ahmadabad, in particular,
felt so heavily the effects of their internal feuds and grinding rule,
combined with the natural decay of the Muhammadan population,
that its suburbs almost disappeared, large quarters within the walls
became desolate, many splendid buildings were destroyed. . . . Fortu-

OF WESTERN INDIA 61

nately the tide of anarchy was arrested ere it had wrought even more
disastrous results by the supervention of the British power."

The earliest mosque in the city is that of Ahmad Shah in tfc
Bhadr (circa A.D. 1414). The pillars, throughout the mosque, have
been rifled from one or more Hindu temples, and on one of them is
a Devanagari inscription dated Samvat 1307 (A.D. 1251). The fafade
of the mosque, with its archways, is purely in the Muhammadan style,
which is very common among the earliest buildings; but some have
no such fafade, the pillared interior not being enclosed. Hindu
temple pillars are not very lofty, their height generally being governed
by the length of the stone blocks which it is possible to get from the
quarries for the shafts, from the capital to the base, this part being
always a single stone, and not built of two or more. Hence, the
Muhammadans, finding them too low for their more lofty ideas, super-
imposed one upon the other in order to get the height necessary, and so
we find them used in this mosque. The minars, at either end of the
building, with staircases winding up through them to the various
galleries encircling their shafts, are, of course, purely Muhammadan
features, and were used as elevated positions from which the muezzins
called the faithful to prayer. In this, as in many of the northern
mosques, there is a raised and screened enclosure in the north end
which was intended for women worshippers; and, to make it quite
private, a separate entrance leads into it from the outside of the
mosque. In the mosque of the Kaabah at Makka is a similar en-
closure for the same purpose. Some mosques, such as the Jami
Masjid at Mandu, in Central India, have an enclosure at both ends,
in which case one is probably a dummy built to preserve the symmetry
of the design of the building. The domical ceilings are richly carved
and have, probably, like the pillars, been taken wholly from despoiled
Hindu temples, any images upon them having been mutilated before
being used in the mosque.

The essentials of a mosque are the mihrab, or direction niche,
towards which the worshippers turn in prayer, so that they may be
facing Makka; a pulpit from which the leader addresses the congrega-
tion; a covered apartment for shelter, and a small tank, or reservoir,

62 THE ARCHITECTURAL ANTIQUITIES

of water, or other means by which they can wash their feet before
entering the body of the mosque. Other desiderata are minarets from
\^hich to call to prayer, a courtyard to enclose the mosque and separate
it from outside dust and dirt, and certain sanitary arrangements, usually
in one corner of the courtyard. The plan of a mosque is very simple.
The building is a square or oblong hall, the roof of which is closed
in with one or more domes supported by pillars or arches, and the
wall round the three or four sides. In the back, or west wall, is one
or more arched recesses, or mihrabs. When the front is enclosed by
a wall, the latter is pierced by one or more arched openings, according
to the size of the mosque, and the mihrabs in the back wall are
generally opposite these openings. When minarets are added, they
are either placed one on each side of the great central arch, or one at
each extreme end of the facade. The dome, over the central bay of the
mosque, is, as a rule, a very large one, while the others, if more than
one, are smaller and lower. Visitors leave their shoes either outside
the entrance to the courtyard or just inside the doorway. It is the
neglect to do this that causes friction between European visitors and
the mosque custodians. They may wear their hats all through the
building, but not their shoes, which are likely to carry in some defile-
ment upon them from the streets. The Hindu's objection to shoes in
their temples is that they are of leather from the sacred cow. In many
cases they object to a camera, even when used by a Hindu, because
of its leather bellows and gelatine coated plates.

The largest and most important mosque in the city is the Jami,
or public, mosque, the mosque of the assembly, or the Juma (Friday),
as it is sometimes called, Friday being the Muhammadan Sabbath,
when the people present themselves there for worship (Plate 47).
Most of the other mosques in the city have been built by rulers or
private individuals for the use of themselves and their families, and are
often an accessory to the tomb of some member of the family. But
many of these have, long ago, passed out of private hands into those of
the local communities and are now, to all intents and purposes, public.

Fergusson wrote of this mosque : " Though not remarkable for its
size, it is one of the most beautiful mosques in the East." With its

OF WESTERN INDIA 63

courtyard, it covers 382 feet by 228 feet, the prayer chamber being
210 feet by 95 feet. Within it are 252 pillars and 76 pilasters, sup-
porting 1 5 domes symmetrically arranged, the centre 3 being somewhat
larger, and considerably higher than the others. The very tall,
graceful minarets, which flanked the great central arch of the facade,
once famous as the " shaking minarets of Ahmadabad," were thrown
down by the earthquake of 1819. In connection with these, it may
be of interest to repeat a note from the writer's Progress Report
for 1904-5: "The following extract from Grindlay's 'Scenery and
Costumes and Architecture of Western India ' refers to them : c But the
most remarkable circumstance attached to this building is the vibration
which is produced in the minarets, or towers, rising from the centre
of the building, by a slight exertion of force at the arch in the upper
gallery. Many theories have been suggested to account for this, but
they all fail in affording a satisfactory explanation of this architectural
phenomenon; which is still further involved in doubt by the circum-
stance of one minaret partaking of the motion of the other, although
there is no perceptible agitation of the part connecting the two on
the roof of the building. 5 Colonel Monier-Williams, then Surveyor-
General, gives, in his journal, the results of his personal observations
on the subject: <3ist May, 1809. We found on examination to-day
that the minarets of the Jumma Musjid shook just as much, or even
more, than any of the others, and that one communicated the motion
to the other fully to as great a degree as those of the Bee-bee Sahib's.
Indeed, we tried the experiment upon every perfect pair of stone
minarets within and about the town, to-day, and the effect was just
the same with them all. As the motion that one of them receives
from the shaking of the other might be supposed to be communicated
to the whole intermediate building, I lay down on the terrace roof,
equidistant between the two minarets, while people were above shaking
them; but I was not sensible of the smallest motion or agitation what-
ever in the building under me.'

"When at Ahmadabad, on the 3ist May last, nearly a hundred
years after Colonel Monier-Williams made his experiment, my atten-
tion was called to the fact that Siddi Basir's minarets, near the railway

64 THE ARCHITECTURAL ANTIQUITIES

station, possessed the same peculiarity. I ascended to the topmost
balcony of one while two or three men ascended the other. Putting
their hands together against their minaret they began to throw their
weight upon it. For a few seconds I felt nothing on mine; but, as
I was about to put the thing down to imagination, it began to swing
with a soft, noiseless, and resilient motion, which gradually increased
until I was, perforce, obliged to call out to them to stop it." All that
remains of Sidi Basir's mosque are these two minarets and the arch
masonry connecting them below.

The Jami Masjid was built by Ahmad Shah in 1423, as we are
informed by an inscription above the central mihrab. Save for the
pavement, which is constructed of a coarse kind of white marble, the
whole building is of fine sandstone. The domes are of the usual
Hindu style, richly carved in concentric circles. The forward central
dome, just within the great arched entrance, rises to a height of 44 feet
from the pavement, or floor, in three stories; flanking this, on either
side, the next domes rise to 37 feet in two stories, while the domes
in the ends of the building rise to 26 feet in a single story. The
second and third stories admit light and air into the building between
their pillars and through the interstices of the perforated screens. A
little dome in the entrance porch to the zanana gallery has the unusual
design of a spiral; it is met with, once or twice, in the Dilvada temples
on Mount Abu. The pillars, though cut purposely for the mosque,
have followed, in their general design, the stilted Hindu pillars in
Ahmad Shah's mosque in the Bhadr; but their shafts, being unen-
cumbered with the many mouldings and image niches upon the Hindu
ones, are far more graceful. A purely Hindu feature, the torana, or
flying arch, has been introduced, with much success, between the
central pair of pillars in the entrance. A black marble slab seen in
the pavement, in front of the entrance, and said to be the inverted
part of a Jain image of Parsvanatha, reminds one of the similar use
to which Mahmud of Ghazni put a fragment of the linga which he
carried away from Somanatha-Pattan.

A door in the east colonnade of the courtyard leads into the court
of the tomb of Ahmad Shah. This is a great domed mausoleum, in the

OF WESTERN INDIA 65

central hall of which are three richly carved tombstones, and at each
corner of this hall is a smaller one, while the intermediate spaces are
open pillared halls.

Beyond this, again, across a street, are the tombs of the queens,
which, standing in the open air, are surrounded by a shallow pillared
corridor (Plate 49). The tombstones fill the enclosed space and are
elaborately carved in white and black marble with the remains of some
fine inlay work of mother-of-pearl.

Sidi Sayyid's mosque, at the north-east corner of the Bhadr, is
noted for the two magnificent perforated windows which it possesses
(Plate 48). The actual perforated work measures TO feet wide by
6 feet high, and is cut in fine sandstone, but is deserving of a better
material. The windows are pointed lunettes, and, though the general
appearance of both are alike, the designs differ. In the one, it is of
a single tree whose pliant stem and many sinuous branches twine and
intertwine among themselves. Rising up the middle is a palm whose
stem is embraced by that of the other tree. In the other window
there are three trees and four palms which go through similar con-
volutions. The windows are made up of four rows of thin slabs, set
edge to edge, one on the top of the other. In this mosque we find
the minarets placed at the extreme ends of the fa?ade.

A mosque of distinction is that known as the Queen's, or Rani
Rupwanti's, in the Mirzapur quarter (Plate 47). It is a very sub-
stantial building, heavily loaded with decorative detail, the bases of
the fallen minarets, on the front of the mosque, being filled with
exquisite designs in stone tracery. It measures 103 feet by 46 feet,
and has three domes standing upon twelve pillars each. The building
and tomb belonging to it are now in the hands of the butcher com-
munity.

Rani Sipari's, or Rani Asni's, mosque and tomb (A.D. 1574), near
the Astodia gate, is " the most exquisite gem at Ahmadabad, both in
plan and detail. It is without arches (except a small one over a side
doorway) and every part is such as only a Hindu queen could order,
and only Hindu artists could carve." It is one of the smallest in
the city, measuring 48 feet by 19^ feet. The balcony windows, in the

5

66 THE ARCHITECTURAL ANTIQUITIES

end, are elaborately wrought, and the mihrabs are of marble and carved
with much care, though not so richly as in some of the other mosques.
r As it is so small, it would have been impossible to provide it with
minarets containing staircases, so very slender solid ones were
substituted, more after the prevailing style of those at Bijapur.

Muhafiz Khan's mosque, in the north of the city, though not a
large building, is a well-proportioned one, and has its minarets intact.
It was built, as its inscription tells us, by Jamal-ud-din Muhafiz Khan,
governor of the city under Sultan Mahmud Bigarah, in A.D. 1492. It
is distinguished for the exquisite character of its details; the galleries
around the minarets, and the brackets which support them, display
great richness of decoration. The carving is very Hindu in character,
the under-cutting of the arabesque being so complete as almost to
separate it from the block on which it is wrought. The mihrabs are
about the finest in Ahmadabad, and are minutely and elaborately
carved. The facade of the building has the minarets at the extreme
ends, and has three small arched entrances instead of one large one.

There are many other first-class buildings, both in and around
the city, but space will not permit us to take up more than these few
specimens. Among the civil buildings may be mentioned the palace
of Azam Khan, used as public offices, on the east side of the Bhadr.
Not far from it, and spanning the principal street of the city, is the
Tin Darwaza, or triple gateway, built by Ahmad Shah. The Moti
Shahi Bagh, a palace built in the reign of Shah Jahan, and now the
residence of the Commissioner, is a fine imposing building, built upon
the high banks of the Sabarmati, about 3 miles north of the city.
The Kankaria tank, south-east of the Rayapur gate of the city, is an
ornamental expanse of water of no mean pretensions, entirely sur-
rounded by tiers of cut-stone steps, and with an island and pavilion in
the centre.

About 3 miles to the south of the city is the tomb and mosque
of Shah Alam, which, with their subsidiary buildings, form quite an
important group. This tomb is notable for the amount of perforated
marble and brass screen work. The interior of the dome is inlaid
with mother-of-pearl, and painted work adds to the general em-

OF WESTERN INDIA 67

bellishment. The minarets of the mosque are two of the loftiest in
or around the city, and rather dwarf the main body of the building,
which has an open fa?ade of seven arches and six pillars. Though it (
may seem daring to say so, in the face of public opinion generally,
I consider these minarets of Ahmadabad finer than those bald erections
standing around the Taj Mahal at Agra; they are, probably, the finest
in India.

At the village of Sarkhej, 5 miles to the south-west of Ahmadabad,
is an important cluster of Muhammadan buildings of the fifteenth
century, among which is the tomb of Sheikh Ahmad Khattu Ganj
Bakhsh of Anhillavada-Pattan, which was built between 1445 and
1451. The tomb-chamber is enclosed by screens of perforated brass in
beautiful patterns, and the exterior walls are filled with windows of
perforated stone tracery. To light the interior of the dome, which,
otherwise, would have been in perpetual gloom, the very unusual
expedient has been resorted to of four tracery windows in the dome
itself. Near by is the mosque, carried out entirely in the pillar and
lintel style. Fergusson says : " This mosque is the perfection of
elegant simplicity, and is an improvement on the plan of the Jami
Masjid. Except the Moti Masjid at Agra, there is no mosque in
India more remarkable for simple elegance than this."

Notwithstanding the elaborate work lavished upon these buildings,
and the great variety of detail, they begin, to some extent, to weary
the eye by their sameness of design. A mosque, with its rigid require-
ments to which the building must conform, does not lend itself well
to any originality of design, so that we get a few stock plans which, in
the main, are repeated over and over again. The design is too
geometrically balanced about its centre a complete plan to which
nothing can be added and from which nothing can be taken away.
There is no chance of stumbling upon some unexpected adjunct or
pleasant variation : this is because they were built once for all, and did
not grow gradually like the great cathedrals of Christendom.

Another centre of old Muhammadan buildings is Champanir, or,
properly, Mahmudabad, near that old Hindu site. It lies 30 miles
north-east of Baroda, and about 72 miles to the south-east of Ahmad-

68 THE ARCHITECTURAL ANTIQUITIES

abad. The city was founded by Mahmud Bigarah of Ahmadabad, in
A.D. 1484; but, beyond the mosques and tombs, there is little sign of
its having been a city of any account. It began as a fortified camp
when Mahmud, with his army, took up his position here until he
reduced the old fort of Pavagadh, whose beetling cliffs and defiant
towers overhang and dominate the place. The old Hindu town of
Champanir, whose ruins lie at the foot of the fort, is said to have been
founded by Champa, a merchant and companion of Wan Raj of
Anhillapur-Pattan, and was also known as Champakadurga and Champa-
kamera, which names, probably, referred more to the fort above.
Ruins of small Hindu temples are found upon the site as well as on
the fort. Mahmud's siege lasted, on and off, some twelve years before
the fort fell into his hands, and even then it fell only by treachery.
Therefore, in order to show the enemy that it was not his intention to
abandon it, he started to build in sight of the fort a new city, which
he called after himself; and this is not the only town of that name that
he founded during his warlike excursions abroad. Scattered over the
site, and shut in by the jungle that has overgrown the site, are a number
of very fine mosques and tombs, all in a ruinous condition, save for the
little attention the Archaeological Department has been able to give
them : the bat and the owl now hold their midnight court there.

The principal building is the Jami Masjid, which is one of the finest
mosques in Gujarat. Here the minarets, which are fairly complete,
rise from the centre of the fafade, one on either side of the great
arched entrance, as in the earlier examples in Ahmadabad. The mosque,
within, measures 1 69 feet by 8 1 feet. The roof contains eleven large
domes, the central one rising through three stories to a height of
57 feet above the floor. On the first and second floors a carved
balcony runs round the octagon, under the dome, which is decorated
by deep carved ribs, of which there are sixteen. There is some very
beautifully carved work in the ceilings (Plate 49).

Among the many remains at Mahmudabad is a beautifully decor-
ated small tomb, the dome of which has fallen. "The pilasters at
the corners and jambs of the doorways are carved in patterns of the
richest floral designs. Except the two famous windows in Sidi

OF WESTERN INDIA 69

Sayyid's mosque at Ahmadabad there is hardly anything elsewhere to
match these twelve pillars in richness and variety of decoration." What
adds to the beauty of the design is the plain flat surfaces, two upon
each of its four walls, which serve as an admirable foil to the rich work
around. Another mosque of note is the Nagina Masjid, which is upon
the same general plan as that of the Jami Masjid. It has some fine
carving, in panels, on the bases of the minarets. Out before it stands
a pillared tomb, graceful but much ruined.

Mahmudabad-Champanir became the favourite residence of Mah-
mud Bigarah, and continued to be so for the rest of his days; and it is
said to have been the political capital of Gujarat until the death of
Bahadur Shah, in 1536.

Another Mahmudabad, or, as it is more generally spelt, Mehmud-
abad, founded by Mahmud Bigarah, lies about 17 miles south-east
from Ahmadabad. He is said to have formed, beside it, a large
deer park, at each corner of which he built a pleasure house with
gilded walls and roof. About a mile and a half to the east of the town
is the tomb of Mubarak Sayyid, one of Mahmud Bigarah's ministers,
erected in 1484. It is a finely proportioned, massive building, though
simple in its design. Of this building Fergusson writes: "The most
beautiful, however, of these provincial examples is the tomb at
Mahmudabad, of its class one of the most beautiful in India. . . .
Though small it is only 94 feet square, exclusive of the porch
there is a simplicity about its plan, a solidity and balance of parts in
the design, which is not always found in these tombs, and has rarely,
if ever, been surpassed in any tomb in India. The details, too, are
all elegant and appropriate, so that it only wants somewhat increased
dimensions to rank among the very first of its class. Its constructive
arrangements, too, are so perfect that no alterations in them would
have been required, if the scale had been very much increased."

Other centres of Muhammadan work in Gujarat are Cambay,
Dholka, and Broach, at all of which places are buildings of consider-
able note.

ARCHITECTURAL REMAINS AT BIJAPUR

UPON the breaking up of the Bahmani Kingdom of the Dakhan,
towards the end of the fifteenth century, the local governors
of the provinces threw off their allegiance to their suzerain
and set up petty kingdoms for themselves. One of these was Yusuf
Khan, who was in charge of the Bijapur division, and who assumed
independent power under the title of Yusuf Adil Shah, in 1489. He
is said to have been a son of Murad, Sultan of Turkey, who, on the
death of the latter, was smuggled out of the country by his mother,
to escape the fate of younger sons on the death of a sultan. After
many adventures he arrived at Bidar and took service under the
Bahmani king, where he rose to high posts in the State, and was finally
appointed to the governorship of Bijapur. Hence we find all the
principal buildings in that city surmounted with the Crescent, the
Turkish national symbol. He had been succeeded by eight rulers of
the Adil Shahi dynasty, when Sikandar Adil Shah surrendered the city
to Aurangzib, in 1686, and it became a district under Delhi. During
its two hundred years of independence, which seems to have been made
up mostly of troublous times wars without and insurrections within
the city was enclosed with lofty masonry walls, some 7 miles in circum-
ference, and was adorned with many beautiful buildings mosques
and tombs, palaces and reservoirs. In the middle of the city is the
walled citadel, or arkilla, within which were the king's palaces, courts,
and assembly hall, the ruins of several of these buildings having been
converted into dwellings and offices for the local authorities when
Bijapur was made the headquarters of the Collectorate in supersession
of Kaladgi. The walls of the Gagan, or assembly hall, still stand gaunt
and bare with its great arch, 60 feet 9 inches in width, spanning the
front, but the roof has gone. It was on the greensward, before this
building, that tournaments, trials of strength, and military displays
took place before the king and his nobles assembled in the Darbar

70

ARCHITECTURAL ANTIQUITIES 7*

hall. Thus it was necessary that the facade arch should be of sufficient
span to offer no obstruction to the view. Behind this is the Sat Manjli,
or "Seven-Storied Palace," with a neat little water pavilion before it^
Other buildings, which were converted, are the Adaulat Mahal, the
Anand Mahal, the Arash Mahal all now dwellings and the Chini
Mahal housing the general offices. The little Makka Masjid, built for
the ladies of the court, is enclosed, as one would expect, by a lofty
wall. It has no mimbar, or pulpit, since no man could enter to address
them. The earliest mosque in Bijapur, that of Karim-ud-Din, built of
materials from a desecrated Hindu temple, stands not far from the
south-eastern gateway of the citadel.

Amongst those buildings which stand out most conspicuously in
the city and its environs are the Jami Masjid, the earliest and most
dignified building, the tomb and mosque of Ibrahim II., the most
lavishly decorated, and the great Gol Gumbaz, or tomb of Sultan
Muhammad, a contrast to the rest by its massiveness and simple
exterior. The Jami Masjid is, as its name implies, the principal
public mosque in the city, and, in it, we have the style, which is so
peculiar to Bijapur, in its full development, in its purest and best
form. It is evidently not the work of local builders, for there could
have been no indigenous craftsmen, at that time, capable of attempting
anything approaching it. Not only its style, but its architect and
builders must have been imported; and their descendants, no doubt,
raised the subsequent buildings in the city. No later building is equal
to this for its perfect proportions or sublimity of effect.

The domes of Bijapur were built for external display : internally,
they are lost in their own gloom, running up very high compared
with their diameter, and, generally, having no clerestory lights to
show them to advantage, some rising from the general roof-level,
almost as hollow towers. These are unlike the domes of the Ahmad-
abad buildings, which are raised upon disengaged pillars, as a rule,
with plenty of light and air entering them from open galleries around
the springing levels. The obvious corrective for this is double doming,
the inner, or lower, dome forming a ceiling sufficiently lighted from
below. But this device is met with nowhere save in the Ibrahim

72 THE ARCHITECTURAL ANTIQUITIES

Rauza tomb, which, in this case, produces a second story entered by
stairs through the thickness of the walls. But, as this is a flat ceiling,
goved around the margins, it can hardly be called a dome.

The arches are, mostly, two-centered, the curves being carried up
from the springing to a point whence they are continued as tangents
to the crown. This is the prevailing form of arch, but others are in
use, such as those in the tomb of Ali Adil Shah II, , which are purely
Gothic, the ogee arch, in one case the segmental, and an almost flat
arch. They are often ornamented with richly moulded cusps, as in
Ali Shahid Pir's mosque, where they look particularly well.

A prominent feature in the Bijapur buildings are the graceful
minarets that flank the mosque facades, and, sometimes, rise above
the tombs. These are purely ornamental, being too slender to contain
a spiral staircase within them, as is the case with the Ahmadabad ones.
The great clumsy brick and plaster minarets standing before the
Makka Masjid have stairs winding up them, but they are parts of
some very early mosque which has disappeared.

Except in the case of the two converted temples in the citadel
and the guardroom at the gateway, there is no other example of the
pillar and lintel style which is so common in Gujarat, and these are
not finished specimens, but rough and ready conversions to serve the
purpose of the first invaders, whose prolonged stay in the place was
problematical. The mosque of Malik Karim-ud-Din is wholly made
up of pillars, beams, cornices, and other portions from old, dismantled
Hindu shrines, the porch being a part of a temple in situ the hall,
or mandapa, with its pillars and niches, but wanting its roof. The
shrine, which was on the west of this, was cleared away to give through
access to the mosque courtyard beyond it. Across the whole west
side of the courtyard is the mosque, or prayer chamber, made up of
pillars, of all patterns and heights, brought to one uniform level with
odd blocks; and, over these, are laid cross beams and slabs forming the
roof. Like the Ahmadabad mosques, the central portion of the roof
of this one is raised by pillars placed one upon the other, thus admitting
light and air into the body of the building. An inscription within
tells us that Malik Karim-ud-Din erected the upper part of the mosque

OF WESTERN INDIA 73

in A.D. 1320, and a builder, Revoya, of Salhaodage, carried out the
work. This Karim-ud-Din was the son of Malik Kafur, the general
of Ala-ud-Din, who conducted several successful campaigns against th^
Hindu kingdoms of the south.

The great Jami Masjid, with its courtyard, embraced the largest
area of any building in the city about 91,000 square feet, which, with
a later extension of the courtyard on the east, amounts to 1 16,000 feet
in all (Plate 50). The main building, the mosque proper, stretches
across the west side of the courtyard, and the latter is closed in, for
the most part, with side arcades or corridors. Plate 50 gives a good
idea of the interior of this noble building. The great dome, over the
central bay, follows more, in its outline, the shape of western domes,
being practically segmental. Most of those in Bijapur are bulbous,
being curved in at the springing, a very good example of which,
though not extreme, being that over the tomb of Ain-ul-Mulk, to the
east of the town, and another is that of Khawas Khan, one of the " Two
Sisters."

As may be seen in the photograph, the interior of the mosque, save
the decorated mihrab, or prayer niche, is severely plain. There is an
air of quiet simplicity about it which adds much to the impressive
solemnity of the place. The whole front and recess of the mihrab is
covered with rich gilding, whereon are representations of tombs and
minarets, censers and chains, niches with books in them, and vases
with flowers, while the whole is interspersed with bands and medal-
lions bearing inscriptions. From these we learn that Malik Yaqut
completed the mosque, or, in other words, put the finishing touch to
it in these decorations, under the orders of Sultan Muhammad Adil
Shah in A.H. 1045 ( A - D - I ^3^)- The mosque was commenced, how-
ever, by Ali Adil Shah I. (A.D. 1557-1580), and was, no doubt,
practically completed by him. Sultan Muhammad was very partial
to painted decoration, and he adorned the walls of the Asar Mahal,
where the sacred relic of the Prophet is enshrined, and the little water
pavilion at Kumatgi, 10 miles to the east of the city, with mural
paintings, in which were introduced human figures and groups in
defiance of the religious prohibition against it. He also gilded and

74 THE ARCHITECTURAL ANTIQUITIES

decorated the walls of the Sat Manjli, the palace of his favourite
mistress, Rhumba. The fanatic Aurangzib, when he captured the
$ity, had all these mutilated wherever the human form was portrayed.
The floor of the Jami Masjid was divided by him into over 2,250
spaces, or pews as it were, by thin black lines upon the plastered and
polished surface, each ample for a worshipper to kneel in when at
prayer.

The Ibrahim Rauza, or tomb and mosque of King Ibrahim II.,
stands in the open, a short distance to the west of the city, beyond the
Makka gate. Upon a high platform, within a great square enclosure,
which was a royal garden at one time, are two imposing buildings
facing one another, with a reservoir or fountain between them
(Plates 51 and 52). The building standing upon the east side of
the platform is the king's mausoleum, in which are also interred his
queen, Taj Sultana, and four other members of his family, namely,
his mother, two sons, and daughter Zorah, the men's tombs being
distinguished by a long arched ridge-stone upon the top of the
tombstone. The sepulchral chamber, which is 39 feet 10 inches
square, contains the six tombs in a row, from east to west, each lying
north and south. In the middle of each of the four sides of the
room is a doorway, with richly carved wooden doors, whose orna-
mental iron bosses were once gilded, while extracts from the Quran,
in raised Persian letters, fill the panels. On either side of each of
these doors is a large fanlight-shaped window, they being beautiful
specimens of perforated stone work. Each window is filled with inter-
laced Arabic letters, forming extracts from the Quran, the letters
being so interwoven as to be difficult of decipherment by any but an
expert. The lights are the perforated blank spaces dispersed between
the letters. These remind one of the perforated windows in Sidi
Sayyid's mosque at Ahmadabad, already described, the pattern of the
latter being in tree forms instead of letters.

These windows, with their very small interstices, admit, with the
doorways, when open, but a very subdued light into the chamber,
just enough to reveal a most remarkable flat stone ceiling. Most of the
buildings in Bijapur have something peculiar to themselves, either in

OF WESTERN INDIA 75

construction or decoration. This ceiling appears to have been the
chef d'ceuvre of the architect who designed this building. It is simply
a hanging ceiling, with another room above it, under the great dom*.
The whole span is the breadth of the room, namely, 39 feet 10 inches,
of which a margin 7 feet 7 inches broad, all around, curves upwards and
inwards to a perfectly flat square in the centre, 24 feet square. Upon
closely examining this, it is found to be composed of slabs of stone,
like an inverted pavement, with, apparently, no support. There are,
certainly, two deep ribs, running across each way, but these are purely
decorative and are made up of separate stones, which do not, in any
way, support the slabs in the nine bays into which they divide the
ceiling. This has been a daring piece of work, carried out in defiance
of all ordinary rules and regulations governing the construction of
buildings. But the architect not only foresaw exactly what he wanted,
and knew how to accomplish it, but he had the thorough confidence
in his materials, in this case, chiefly, superb mortar, which caused him
no hesitation. 7'hat it has stood to the present day, three hundred
years, is sufficient answer to any carping critic. The buildings have
deteriorated considerably externally, but this has been due, in great
part, not to the want of skill on the part of the builders, but to the
stone used. When the Dakhan was divided up into small States, as it
was then, each jealous of the other, and more often at war than other-
wise, suitable building material could only be sought for within the
narrow limits of the State; the amygdaloidal trap of the Dakhan,
found in this area, was far from being the best that could be desired,
with its many flaws and cracks, some so fine as hardly to be noticed
when freshly quarried, but which in time worked mischief. Hence
the deplorable condition of the brackets and slabs of some of the fine
cornices, notably that of the great Gol Gumbaz, to repair which our
Public Works Department had to go further afield in search of better
material.

The whole secret of the durability of the masonry of those days
lay in the great strength and tenacity of the mortar, and this those
builders knew how to make. Perhaps, too, there were fewer con-
tractors or middlemen. This, then, is the secret of this ceiling : it is,

76 THE ARCHITECTURAL ANTIQUITIES

in fact, a solid concrete ceiling to which the stone slabs are merely
stuck by its own adhesiveness. Partly destroyed domes, constructed
i* this manner, may be seen with their lining of stone slabs still
adhering. There was, of course, the danger of individual stones
falling out, but this was provided against by rabbetting the edges, and,
in some cases, fastening adjacent stones with iron clamps. The ceilings
of the corridors are constructed in the same way.

Around the sepulchral chamber, outside, runs a mezzanine gallery
supported, outwardly, upon pillars which are very Hindu-like in style.
The exterior walls of the chamber, between and above the windows
and doors, are elaborately decorated with shallow surface tracery in
stone of arabesque and beautifully interlaced verses from the Quran.
The effect had been further enhanced by colour and gilding, which
has weathered badly (Plate 52). An inscription on one of the doorways
says:

"Heaven stood astonished at the elevation of this building, and it might be said,
when its head rose from the earth, that another heaven was erected. The garden of
Paradise has borrowed its beauty from this garden, and every column here is graceful
as the Cyprus tree in the garden of purity. An angel from heaven announced the
date of the structure by saying, * This building, which makes the heart glad, is the
memorial of Taj Sultana/"

The last sentence gives the date A.H. 1036 (A.D. 1626). From the
inscription it would appear that the mausoleum was built as Taj
Sultana's tomb, but Ibrahim, dying before her, was buried in it.

Both the tomb and the mosque opposite are noted for their deep
rich cornices and graceful minarets, which, as we have already seen,
are purely decorative. The amount of labour expended upon these
features has been unstinted. Under the cornice of the mosque may be
seen, hanging, the remnants of heavy chains with pendants, each of
which has been carved out of a single block of stone. Some beautiful
specimens of these may be seen at the tomb of a saint at Rauza, near
Aurangabad, and also on the Kala Masjid at Lakshmesvara, in the
Dharwar district, where they are still made. The perforated parapets,
round the skyline of the buildings, look, from a distance, like a fringe
of petrified lace, and the grouping of the miniature minars, round the
bases of the corner minarets, is very pleasing. An inscription, near

OF WESTERN INDIA 77

the south door of the tomb, tells us that Malik Sandal, Bijapur's most
noted architect, expended 150,000 huns on the building. Carried out,
on the same lines of construction and decoration, are several othefr
buildings in the city, amongst them being the beautiful little mosque
of Jahan Begam, with its well-proportioned cornice, rich facade, and
graceful minarets, said to have been built by Ibrahim II. in honour
of his daughter, whose name it bears; and the Mehtar-i-Mahal with
its elegant balconied windows supported upon delicately carved stone
brackets which look more like fine wood-work (Plate 53).

In marked contrast to the buildings just noticed is the great Gol
Gumbaz at the other end of the town. Ibrahim II. had carried his
decorative style to its utmost limits, and had thus left to his son,
Muhammad, no chance of surpassing him in the same line. Since the
Ibrahim Rauza represented the last word in decorated effort and
graceful effeminacy in stone, he resolved to strike out in a more virile
and broad-shouldered style, which would, by mere mass, overwhelm
the work of his father. How far he succeeded may be gathered from
a reference to the photograph of his own mausoleum (Plate 54). No
doubt there was great rivalry in the building of the royal tombs, and
Ali II., his successor, started his own tomb upon so ambitious a scale
that he died before he had done more than raise the arcade of Gothic
arches which were to have surrounded the death chamber. Sultan
Muhammad certainly succeeded in raising the most conspicuous mass
of masonry in the Dakhan : indeed, one of the greatest in the world.

The general contour of the building follows that of a great cube
surmounted by a huge hemispherical dome, with an octagonal tower
running up each of its four corners, crowned by a smaller dome.
Between these towers the only prominent feature upon the face of the
building is the great, deep, overhanging stone cornice, the rest of the
masonry, except for the door and window-frames, being rubble and
plaster. The comparatively small doorways and windows, dwarfed by
the immense expanse of wall surface around them, by no means assert
themselves, and, from a short distance, are hardly noticeable. The
monotony of the plastered masonry of the four sides is relieved by
the corner towers, which are riddled from base to summit with arched

7 8 THE ARCHITECTURAL ANTIQUITIES

openings, pigeon-hole fashion seven in each of the seven stories into
which the towers are divided.

* The interior of the building is very impressive. Upon entering
this vast chamber, which is 135 feet 5 inches square, rising to the
apex of the dome, 178 feet overhead, the visitor's ears are assailed,
on all sides, by uncanny sounds and echoes which, rebounding from
wall to wall, roll round the great dome, whence they are returned
in a multiplicity of weird reverberations. Through the gloom,
109 feet above the floor, a gallery may be seen, looking more like
a cornice, running round the base of the dome. In reality, the gallery
hangs out 1 1 feet from the walls, and is reached by long, spiral stairs,
in each corner of the building, which, also, give access, as they ascend,
to each story of the towers. The uninterrupted floor space is the
greatest area of any single apartment known among ancient buildings,
measuring 18,337 square feet. The interior diameter of the dome
is 124 feet 6 inches, while its exterior measurement is 144 feet. The
total height, outside, above the platform upon which it stands, is
198 feet 6 inches, exclusive of the pole at the top, which originally
carried the metal finial, and would add another 8 feet to the total.

Upon a high platform, in the centre of the chamber, are the
tombstones, but not the graves, of the grandson of Muhammad; his
younger wife, Arus Bibi; the Sultan himself; his favourite mistress,
Rhumba; his daughter; and his older wife, in this order from east to
west. The real graves, where the bodies lie, are, as is usual, in the
vault immediately below these, the entrance to which is by a stair
under the western entrance. Over Muhammad's tomb is erected a
wooden canopy.

The method, used so successfully at Bijapur, of working up to
the dome from the sides of the square room below, by means of
pendentives, or squinches, in the corners, is here seen at its best.
Theoretically, there is no limit to the size of the building that could
be raised or covered on this principle; but, with the material the
Bijapur builders had to deal with, it is very doubtful whether they
could have exceeded the dimensions reached in this case without great
risk of accident. The building being planted upon virgin rock, they

OF WESTERN INDIA 79

had no trouble with the foundations. The dome, in spite of its great
size, was probably constructed, with the exception, perhaps, of a small
portion at the crown, without centering support. It is TO feet thick*
at the springing, and appears to have been built of solid brick and
mortar in flat rings of masonry, not radiating, as with voussoirs, each
ring being corbelled inwards to the curve and locking itself. Con-
sidering the size of the brick in comparison with the mass, the dome
is a rigid brick concrete shell which might have stood upon a few
separate supports, like the shell of an egg, without breaking away,
provided the concrete was perfectly homogeneous throughout.

A remarkable feature in this building is its whispering gallery,
already alluded to. The sounds that assail one on entering the
chamber below are much intensified upon stepping into the gallery,
by a passage through the dome, when the footfall of a single person
is enough to awaken the echoes of the tread of a company. Strange
weird sounds and mocking whispers emanate from the walls around.
Loud laughter is answered by a score of fiends hidden behind the
plaster. The slightest whisper is heard across from one side to the
other, the ticking of a watch being quite audible, while a single loud
clap is echoed over ten times distinctly.

Over the south doorway, within, is a large displayed inscription,
boldly cut in three compartments. Each of these sections is a com-
plete sentence in itself, and each, on computing the numerical value
of the Persian letters, gives the date A.H. 1067 (A.D. 1656) as that of
Muhammad's death. The sentences read :

The end of Muhammad has become laudable.
Muhammad Sultan whose abode is in Paradise.
The abode of peace became Muhammad Shah.

The portion added to the back, or north side, of the mausoleum
is said to have been intended to afford a resting-place for Jahan Begum,
the wife of Muhammad Shah, but it was never finished. She evidently
refused to be so treated, and is supposed to have been buried at
a spot some distance to the east of the city, where a mausoleum was
commenced upon the same plan and scale as the Col Gumbaz, with
the same arrangement of corner towers. But the dome, in this case,

8o THE ARCHITECTURAL ANTIQUITIES

would have been much smaller, as it would have been carried by four
inner walls enclosing the sepulchral chamber. It is quite possible
,tfiat this was intended as the last resting-place of the mother of
Sultan Muhammad.

Upon the west side of the Gol Gumbaz is its mosque, which is
a fine building in itself, and worthy of notice, but it has been com-
pletely dwarfed by the great tomb beside it. On the south side is the
Nagarkhana, or gallery for musicians, with the entrance gateway to
the enclosure beneath it.

Not content with the capital of his fathers, Ibrahim II. was
possessed with the mad idea of constructing another, to be called
Nauraspur, upon a site but 4 miles off to the west. He summoned
masons and 20,000 workmen from all quarters, and put the task
under the superintendence of Nawab Shavaz Khan. His nobles and
ministers, as well as the more wealthy merchants, were pressed into this
work, and were prevailed upon to add their quota to the buildingsr
rising upon the new site. What now remains of this ambitious under-
taking are the ruins of the outer wall, enclosing more than half
of the site, which, if completed, would have been as large again as
Bijapur, or some 9 miles in circumference; the ruins of the Nauras,
or Sangat Mahal, the Nari Mahal, and the Tagani Mahal, as well
as mosques and tombs and other buildings of sorts are embedded
in cactus and jungle. The Sangat Mahal was a duplicate, on a rather
smaller scale, of the Gagan Mahal in Bijapur. Trouble with Nizam
Shah of Ahmadnagar brought that State's troops down upon the un-
finished works, which were completely wrecked, and all further con-
struction was abandoned.

Bijapur depended for its water supply chiefly upon two sources,
both without the city walls, which, when the city was invested by an
enemy, was cut off; but provision to some extent was provided against
this contingency by the number of reservoirs, or tanks, and wells,
which, once filled, served the town for some time. The Begam
Talao, to the south of the town, was one of these sources, from which
place it was brought in through earthen pipes, embedded in concrete,
with relieving towers at frequent intervals. The water, entering these

OF WESTERN INDIA 81

at a low level, was drawn off again at a high level, so as to leave
its sediment behind, which was cleaned out at times, steps leading up
the sides to give access, the top of the tower being open. This
aqueduct led to distributing towers and tanks within the walls, whence
smaller pipes, some of copper, connected up with the principal buildings
and gardens. The other source was from Torweh, the site of Nau-
raspur, 4 miles out to the west. The water was brought in from here
through a subterranean tunnel, which was lined with masonry on one
side, the other side being the natural murum wall. As the strata,
through which it is cut, slope athwart the direction of the tunnel,
the subsoil water trickling from the one side was caught, while the
masonry on the other side prevented leakage. What was thus caught
formed a valuable addition to the flow from the head source. Among
the principal reservoirs in the city are the Taj Bauri and the Chand
Bauri, both important works.

There are many notable mosques, scattered about, singly, in different
parts of the country, one of the most ornate being the Kala Masjid
at Lakshmesvara, in the Dharwar district. It is in the same style of
work as the Ibrahim Rauza at Bijapur. Its stone chains, hanging
from the minarets and cornice, which are now sadly damaged, together
with its beautiful tracery in perforated stone, make it one of the
most elegant little mosques in Western India. The gateway to the
courtyard is almost as grand as the mosque itself. These stone chains
are still made at this town as souvenirs, many being intricately
fashioned with double links, all loose and independent, and hollow
fretted pendants with one or more loose, small, stone balls caged
within them. Each chain and pendant is cut from a single stone. The
town is full of the remains of old decorated Hindu and Jain temples.

MUHAMMADAN BUILDINGS IN SIND

OF Hindu remains in Sind, little is to be found, even in ruins,
owing to the havoc wrought by the Arab conquerors. That
such buildings did exist is plain from the great temple at Deval,
which they destroyed, and the fragments built into the tomb of Jam
Nindo at Thatha (Plate 57). A very interesting discovery was made, a
few years ago, in a field near Mirpur-Khas, of a half-life-size bronze
image of Brahma. When the Muhammadans began to settle in the
land as rulers, they started building their tombs and mosques, but
they displayed poor taste in their architectural endeavours. They
sought to make up for their shortcomings in that respect by lavish
decoration, occasionally in surface carving in stone, but more frequently
in the application of coloured tile work, introduced from Persia, with
which they literally covered their buildings (Plate 55). In the earlier
tile-work there are some very beautiful designs, and, in many cases,
good taste has been shown in the harmonious grouping of the few
colours they used two blues, cream, and white. Later on, when
they added other colours, such as the various shades of green, brown,
yellow, and even a dirty red, their work became garish, and would
supply suitable patterns for cheap linoleum. The tombs of the
Talpurs, at Haidarabad, are good examples of this.

Sind being pre-eminently a brick country, most of the buildings,
particularly those covered with enamelled tiles, were built in that
material, with stone foundations to save the brickwork from destruc-
tion by the kalar, or saltpetre, rising in the walls and pulverising
them. With the exception of the tomb of Isa Khan, at Thatha, these
mausolea are rather heavy and clumsy looking in outline, and, to
some extent, are only saved by their elegant finials, and, in a few
cases, by the very pleasing innovation and effective finish to the
dome, the ornamental lantern, a very rare feature in India. These
tombs are, as a rule, great cubical blocks of masonry surmounted

82

ARCHITECTURAL ANTIQUITIES 83

with heavy hemispherical domes. Save for the decorated facades, the
walls are often bald plastered areas, which are relieved only by rows
of shallow panels. There is but one door to most of them, and tha
,too small in proportion to the heavy mass of walling above and around
it; and there is seldom a porch, or any advanced shelter, to give a
sense of dignity to the entrance. Yet, notwithstanding these defects,
the mere mass makes them, to some extent, imposing; and, in their
more or less neglected state, mellowed by age, they look better than
they did when newly built. Of a more pleasing type are the few
stone pillared buildings, decorated with surface fretwork after the
manner of Fathipur-Sikri, the best example of which is the mausoleum
of Isa Khan, already noticed (Plate 56). Marble is but sparingly used
in Sind, and then only upon grave slabs and railings within the
buildings.

INDEX OF PLACE-NAMES

(Roman figures refer to the Plates.)

A. j Brahmanabad, 10

ABU, 37, 39, 40, 41, 43, 44,1 Broach > 6 9
49, 53, 64, XXXV, XXXVI,
XXXVII, XLII

Adalaj, 44
Agra, 67

Ahmadabad, 44, 53, 58-69, 74,
XLVII, XLVIII, XLIX

Aihole, 14, 20, 21, 22, 23, 25,
XVII, XVIII, XIX

Ajanta, 2, 6, 7, 19, III, IV, VII,
IX

Ambarnatha, 52, XLIII

Amravati, 13

Anhillavada-Pattan, 35, 39

Anjaneri, 57

Annavatti, 31, XXVII

Aurangabad, 7

Ayodhya, 15

B.

Badami, 17, 19, 21, 25, 26
Balagamve, 30, 31, 33, XXVII,

XXVIII

Balsane, 54, 56, XLVI
Bankapur, 31, 58
Bedsa, 6
Belgaum, 31
Bezvada, 13
Bhadr, The, 61, 64, 65
Bhaja, 6
Bhatkal, 34
Bhojapur, 41
Bhuvanesvara, 19
Bijapur, 71, 8 1, L-LIV, LVII
Bilesvara, 17, XVI
Boria stupa, 8

c.

Cambay, 69
Chamba Valley, 13

Champanir, 67, 68, XLIX
Chandragiri, 44
Chaudadampur, 31
Conjeveram, 22, 25

D.

Dabhoi, 44, XXXII

Dambal, 31, 32, 56, XXX

Degamve, 31

Depar Ghangro, 10

Deval, 82

Devalada, 46, XXXV-XXXVII,

XLII

Devalana, 57
Dhalka, 69
Dhar, 59
Dharasena, 13

E.

Elephanta, 17

Elura, 4, 6, 12, 16, 17, 22, 33,
IV, V, VI, VIII

F.

Fathipur-Sikri, 83

G.

Gadag, 31, XXXI

j Galagnatha, ^ i

I Ghumli, 42, XXXIV

84

INDEX

Girnar, "42
Gop, I3 , XV

H.

Haidarabad, 82
Halabid, XXIV
Hangal, 31, 32
Harahalli, 3 1
Harihara, 31
Haveri, 31

I.
Ittagi, 30, XXV, XXVI

J-

Jabalpur, 32

Jhodga, 57
Junagadh, 8, 10
Junnar, 6

K.

Kadvar, 14
Kanheri, 6, VII
Kapadvanj, 37
Karli, 4, I, II
Kolhapur, 8
Kondani, 6
Kukkanur, 13, 27, 28
Kumalgi, 73
Kuruvatti, 31, 32, XXIX

L.

Lakkundi, 27, 29, 31,

XXIII
Lakshmesvara, 76, 8 1

M.

Mahakutesvara, 25, 33
Mahmudabad, 60, 67^ 68,
Mandhata, 53

XXII 5

69

Mandu, 61

Mirpur-Khas, 7, 8, 9, 10, 82, XI,

XIV

Mohen-jo-dhado, 10
Mudabidri, 34
Mudhera, 37, 41, 43, 53,

XXXVIII, XXXIX, XL
Mustafabad, 60

N.

Narsapur, XXVIII
Nasik, 4, 6, II
Nauraspur, 80
Niralgi, 31

o.

Osmanabad, 13

P.

Parsvanatha, Mount, 44
Pattadakal, 13, 14, 19, 2 1, 22, 26,

XVII, XVIII, XX, XXI
Pattan (see also Somanatha Pattan),

35>39)40, 57
Pedgaon, 57
Pitalkhora, 6

R.

Rattihalli, 31
Rauza, 76

S.

Sanchi, 2, 13
Sarkhej, 67, XLVIII
Sarotra, 39

Satrimjaya, 44, 58, XLII
Sehwan, TO

Sejakpur, 42, XXXIV
Siddhapur, 36, 37, 41, XXXII
Sinnar, 54, XLIV, XLV
Sirpur, 13

86 INDEX

Somanatha Pattan, 38, 40, 41, 53,

58, 59, XXXIII
Supara, 8, 10
Sutrapada, 14, XVI

T.

Tando, 10

Taranga, 49

Ter, 4, 12, 13, 14, 18, IX

Thatha, 82, LV-LVII

Thul Mir Rukan, 8, 9, u,X

Unkal, 31

U.

V.

Vadnagar, 36, 37, XLI
Vaghli, 57
Velapur, 57

W.

Wadhwan, 44

PKINTilJ I'.V HILLING AND hONS. LTD., (.tlll.DKJUl) AM) tSiIEK

PJLATK 1

THE <,REAT CAVE AT KARL,!

PJ,ATK 5>

INTERIOR OK THE t,REAL CA\ 'V AT KARLl

CAVE AI NASIK SHOWINC; \VOODEN CONSTRUCTION

PLATE

C H A I I N A CAVE A 1 \ f A N I A

PJ.ATK 4

1-KoNl

/'///,//? 1 CAVE AT AJANTA

INIEKIOR OF THE TIN '1 AL CAVE AT EI.VRA

PLATK r,

PLATE (j

Ct,

"^J
1
W

^

o
1

w

o
^

c

UJ

X

PLATE 7

W

a:

^

UJ

X

Q

Q

o
o

Y*

^

<

PLATE 8

oi

a

X
X

PLATK 9

Q
a3

PLATE 10

IHF. THUI. MIR Kl'KAN IN SI NO

PLATE li

STL'T.f MOUNl.) AT M I RIM' K- K H AS IN stNl), HM-ORK K\CA\ATION

i-v^.f *" ;|H^

*y$^"i^$|?!*

., '..-..-^^-'^^f "i2^H 4
v^f^vj;,!*' .1' iWs'c^iKr^.' L 1 r *f"

/ A I MIKIM'R-KHAs, Ab'lKK KXCAV A I 1 < >N

PLATE

1M,ATK

u
p

Q

LJ

O

PLATE

X
ct

PLATE 15

0,

O

<

UJ
Q-

o

O

PLATK 16

AI

OLD TEMPLE OF SURVA AT bU I RAPADA

PLATE 17

1 H>< 1F-MIM1- 01- Mhi.I Tl A I AIHOI.F

<**- -^^. ^ *"* ^ i^V'-^w^

C^fcS!&*tf&^^? :^^^^^fc;t*:i

1'WO M M.KS OF 10\VU<b A r

PLATE IS

U!

pi
'^j
&

&
yu
h

yi

< W

S2

PLATK 520

PLATE J>

ATE y

Y.

**j

u:

Ou
P 1

UJ

IM.ATK 'J:

MOI ; I IMN<;S

AKA

PI,ATK i

pi

'*J

s.

tx-

PLATF. ',>->

PLATE f>(i

-3

X

w *
r ^>

PLATE <rr

FROM BA1./V(,AM\F

INTERIOR 0! 'I I-M PI I- A I ANNAVAI

PLATE 2

at

ex
f
/j

t

Q
X,

u: *^

A CC

PQ :-

<

U4
S

L_

OU

\~>
L>
J*

PLATE 5><)

JJRACKKT Hdl'RI- I ROM A
AT KA Rl'V ATI 1

IPMPLK 1MA(;K 1'ROM NA

HA1I3 \RABAD

AN APT R, LN

PJ.A'l K 150

I UK 'I KM PI. F Oh r>(>mx\ BASAI'PA A1 n \MRAI.

THI WALLS OF THK TFMPIF. AT DAMPAL

PLATK J31

i-U

a.

13 <

a.

X
X

_j
-I

HI

PLATK

C*
'.->
0-

X

Q
Q

PLATE M

o

is

<

^

^r

PLATE ,'31-

HF TKMPLF 01- N \\ATAKII A A I C.Hl'MII

THE 'I EMP1.E OF NAVA1AKHA AT StJAKPTR

PLATE :$->

O
5

INC; IN XIMAIA SHA

CKNTRAI. PENOANI OF CKII.INC. OK 'I F I AHPAI.A^ 'IfcMIM.K A. AHL

PI.ATK ^7

I MI INc; PA NT I- FROM TFMPI.F AT ABI T

ANOTHFR CEILINU PANEL FROM TEMPLh AT

PLATE JJS

PLATK !5J)

w

Cu
O

X
U,

SI
w

f'"j ft " '

l ,:-' .

PLATE 40

INThRloR 01- '1 UK I KM Pl.t oh 'I H K SL'N A'l Ml'UHI-RA

PLATE 11

A //v'YT/S /' i.MHH I A I \ADNAC.AR

PLATE 4^2

bOl'lHI-RN PFAU OK SA'l Rl ' N J A\ A HUT.

hROM 1 KM 1*1.1-. A I AI51 T

PLATK 1!J

r*T4s^^ " ^''^V'V 'ij.^'"**dft

I III 1 hMI'l I- OK XMI'.YKNA I II \

PLATE 44

Q
x:

I

H

PLATE

PI.ATK Hi

PLATE 17

II-NIKAI. I'ORlloN OK KAl'AIJfc. Ob JAM! MASjm AI .\HM\n.\H.\t)

PLATK

\\IMJO\V IN smt SA^MlA MOSQl'F AT \HMAU\HAD

RUINS OK Il.IR.1M AT SARKHEJ

PLATE 49

KKNs' 'I OMBS AT A H M A I )A ISAIJ

W& ' &$&

is%ik ; l^

CKlJINt. P \NI-L tROM THP: JAMI MASJIL) A 1

E 50

THE | A MI MASJII) AT IUJAIMJR

OF THE MASJIO AT WIJAPUR

PLATE 51

1 lib IBRAHIM RAC'/A A I HlJAPl'R

WINDOW IN THE IBRAHIM RAUZA AT BIJAPL'R

PLATE W

I

PLATE

THE MKHTAR-t-MAHAI. AT BIJA1>UK

PLATK 54

at
L>

(X

<

S3

O
O

w

I

PLATE 55

.*iS

COI.OURtlJ 'III.F.S IN THK WALLS OF THF JAMI MA\|1I> A I THATHA

PLATE 56

sl'RFACF LARVINC. ON A MIIIR.W AT THATHA

PJ.ATK 57

5.
oc S
Q ^

t: c

O y
X '^

1 ^^

^ o^

o

CQ

C

U4

X