A History of Fine Art in India and Ceylon, from the Earliest Times to the Present Day c11: Chapter V. Sculpture of the Kushān Period, Other than the Gandhāra School

A History of Fine Art in India and Ceylon, from the Earliest Times to the Present Day c11

Language: English

c11.s1 Section I. General Observations — 10 parts
c11.s2 Section II. Mathurā and Sārnāth — 72 parts
c11.s3 Section III. Amarāvatī — 28 parts
c11.s4 Section IV. Dānavulapāḍ — 6 parts

Chapter V. Sculpture of the Kushān Period, Other than the Gandhāra School

Section I. General Observations

c11.s1.1

In Chapter IV our attention was confined to the prolific Hellenistic school of sculpture which flourished during the first three centuries of the Christian era in the regions of the north-western frontier constituting the home provinces of the Kushān empire, and known by the name of Gandhāra.¹ In this chapter we propose to study certain schools of sculpture, partially contemporary with the Gandhāra school, and to some extent related to it, but distinct in style, and located in interior India.

c11.s1.2

¹ The name is written *Kushana* in the Kharoshthī script, which ordinarily does not mark long vowels; but the long *ā* is justified by Chinese transcriptions and the legends of certain Sassanian coins.

c11.s1.3

As I have occasion to observe more than once, the chronology of the Kushān dynasty is still unsettled, and decisive proof is lacking for any one of the many rival theories on the subject. Six sovereigns of the dynasty are of importance for the history of India and of Indian art. The first two are most conveniently cited as Kadphises I and II. The next four kings, Kanishka, Vāsishka, Huvishka, and Vāsudeva I, certainly reigned in that order for a century in round numbers.² As a working hypothesis I revert to Professor Oldenberg's old theory, and assume that Kanishka came to the throne in A. D. 78. Thus the first and second centuries after Christ are approximately filled by the rule of the leading kings of the dynasty, the name of which may be given to the period up to about A. D. 300, because the next principal dynasty, the Gupta, did not begin until A. D. 320. The sculpture at Amarāvatī, although in a southern locality, is closely related to that of the north, and may be conveniently treated under a classification based on the reigns of a northern dynasty. From the far south we have no certain sculpture of the period, and there is little, if any, in Ceylon which can be assigned to it. I have doubtfully inserted the Jain images from the Cuddapah District, Madras, in this chapter.

c11.s1.4

² Inscribed pillar of Mahārāja Shāhi Vāsishka dated in 24th year (Mathurā Museum, Q 13, *Catal.*, p. 189). In India the reigns of Kanishka, Vāsishka, and Huvishka overlap. Probably Vāsishka reigned in India only, and Huvishka succeeded to the whole empire on Kanishka's death about A.D. 123.

c11.s1.5

The art of the sculptors of the Kushān period, as thus defined, may be sufficiently illustrated by study of the remains in three notable localities, namely Mathurā (Muttra) on the Jumna, Sārnāth, near Benares, and Amarāvatī on the bank of the Krishna (Kistna) river, in the Guntūr District, Madras. Of course, contemporary sculptures occur elsewhere, but it is not necessary for the purposes of this history to consider the remains in minor localities.

c11.s1.6

In the early centuries of the Christian era Mathurā on the Jumna (27° 30′ N., 77° 41′ E.), a city of immemorial antiquity, and prosperous to this day in spite of many disasters, was sacred in the eyes of the adherents of all the three indigenous Indian religions—Jainism, Buddhism, and Brahmanical Hinduism. The abundant supply of excellent red sandstone at Rūpbās and other quarries in the neighbourhood favoured the development of an active school of sculptors, whose workshops supplied all parts of Northern India with idols, much as Jaipur does now. The craftsmen, of course, were prepared to supply whatever was wanted by their patrons of any religion. The character of the local stone is so distinct that the products of the Mathurā studios are easily recognized wherever they may be found. Wealthy worshippers did not hesitate to undertake the cost of transporting heavy, even colossal, statues for hundreds of miles. For instance, unmistakable Mathurā images of large size occur at Sārnāth, some four hundred miles distant.

c11.s1.7

Geographically Mathurā occupies a central position intermediate between Gandhāra to the north-west, Amarāvatī to the south-east, and Sārnāth to the east. It is therefore not surprising that the local school of art should display intermediate characters, linking it on the one hand with the Hellenistic art of Gandhāra, and on the other with the more purely Indian schools of the interior. At one time I believed the Hellenizing sculptures of Mathurā to be earlier than those of Gandhāra, but that view has been proved to be erroneous. The Mathurā sculptors continued to turn out creditable work during the Gupta period, as will appear in the next chapter.

c11.s1.8

Sārnāth, like Mathurā, was holy ground to the Jains as well as the Buddhists. The richly adorned buildings of both religions, crowded with sculpture, were involved in common ruin by the violence of the fierce hosts of Islam at the close of the twelfth century. The Brahmanical Hindus lavished their devotion on the neighbouring city of Benares, and shared the misfortunes of their rivals. The sculptors of Sārnāth ordinarily used the excellent pale sandstone from the quarries of Chanār (Chunar) in the Mirzāpur District, which had supplied the blocks for Asoka's pillars. But, as already observed, wealthy donors sometimes preferred to import red sandstone images from Mathurā. During the last few years much progress has been made in unearthing the buried treasures of Sārnāth, but much more remains to be found.

c11.s1.9

Far away to the east of south the ruins of a vanished Buddhist *stūpa* at Amarāvatī (16° 35′ N., 80° 24′ E.) have furnished a multitude of marble sculptures in relief, certainly assignable to the Kushān age, and of such excellence that competent critics have held them to mark the culminating point of Indian art. Their style connects them in certain respects with both Mathurā and Gandhāra. While Indian elements predominate, traces of Hellenistic or Western Asiatic influence may be recognized without difficulty.

c11.s1.10

I now proceed to discuss characteristic examples of the three chief localities.

Section II. Mathurā and Sārnāth

c11.s2.1

The most distinctly Hellenistic sculpture from Mathurā is the mutilated group, 2 feet 5 inches high, known as 'Herakles (Hercules) and the Nemean lion', discovered by Cunningham serving a lowly purpose as the side of a cattle-trough, and now in the Indian Museum, Calcutta (Fig. 79).

c11.s2.2

The hero grasps the beast with his left arm, and presumably threatened it with a club in the missing right hand. He is nude, except for a skin hung behind his back, and fastened by the paws round his neck. He is fairly well modelled in a way that suggests Greek reminiscences, but the lion, so far as can be seen, is a poor, feeble creature.

c11.s2.3

[Figure: FIG. 79. Herakles and the Nemean lion, from Mathurā. (Photo. 842, I. M. List.)]

c11.s2.4

The only other Indian work of art resembling this group is a corroded bronze or copper statuette, 2¼ feet high, treating of the same subject, which was discovered in a mound at Quetta in Balūchistan.

c11.s2.5

The motive is of great antiquity, going back to Assyrian art, which represented Gistubar, the 'Assyrian Hercules', clubbing and strangling a lion in the same way. India, however, probably borrowed the idea from some Hellenistic work of Asia Minor. The Mathurā group was believed by Sir A. Cunningham, whose skilled judgement on the point may be accepted, to date from the time of either Kanishka or Huvishka. But his notion that the sculpture might have been executed on behalf of a hypothetical Greek colony at Mathurā may be dismissed as fanciful, although at present it is not possible to explain the meaning of the group from an Indian point of view. The popular Buddhism of Mathurā indulged in some queer manifestations, and most likely the composition should be classed as Buddhist. The Indian version of the story called by the Greeks the struggle of Herakles with the Nemean lion may turn up somewhere.¹

c11.s2.6

¹ The Mathurā group is M. 17 in Indian Museum, Calcutta: Anderson, *Catal.*, Part I, p. 190; Cunningham, *A. S. Rep.*, vol. xvii, p. 139, Pl. XXX. For the 'Assyrian Hercules' see Bonomi, *Nineveh and its Palaces*, 2nd ed., p. 136, Fig. 36; Maspéro, *Ancient Egypt and Assyria*, transl. Morton (London, 1892), p. 302, Fig. 152. The Quetta statuette is described in *J. A. S. B.*, Part I, vol. lvi, p. 163, Pl. X; and vol. lviii (1889), p. 141. It seems to be more ancient than the Mathurā group, and the treatment differs.

c11.s2.7

Certain groups and statuettes from Mathurā or the neighbourhood, all dealing with strong drink and intoxication, which may be classed together as 'Bacchanalian', have excited much interest and discussion, in spite of which their interpretation is still far from clear. The supposed Greek character of the composition first discovered was much exaggerated by the early commentators, and some of the connected sculptures have nothing Hellenistic about them. All of them, like the Gandhāra figures, demand explanation as representations of Indian, not Greek subjects, even though certain details may be foreign.

c11.s2.9

The block discovered in 1836 by Colonel Stacy, somewhere at Mathurā, and now marked M. 1 in the Indian Museum, Calcutta, was at first supposed to represent Silenus, and so became known as the ‘Stacy Silenus’. But everybody now acknowledges that the subject is Indian, although the sculptor was influenced by the Silenus model. The stone is 3 feet 8 inches high, 3 feet broad, and 1 foot 4 inches thick, with a circular basin on the top 16 inches in diameter and 8 in depth, seemingly intended to serve as the socket for a column. Both this block and its replica, to be described presently, were carved on back as well as front, and were evidently designed to be viewed from both directions. Apparently they were the bases of columns, which may have stood at an entrance, or entrances, the proper position for Yakshas. But the difference of dimensions suggests that the two blocks may have belonged to distinct buildings.

c11.s2.10

The front group (Plate XXXI) comprises four persons in two pairs, each consisting of a man and woman standing under an *asoka* tree in flower. The stout man on the right has his left arm round the waist of his female companion, who holds his right hand in hers, thus giving him the support rendered necessary by his intoxicated condition, due to the liquor, pots of which stand on the ground. The couple on the left stand facing, in attitudes apparently indifferent, but their countenances have been destroyed, so that their expression is lost. Traces of chaplets may be discerned on the heads of all.

c11.s2.11

The reeling man wears nothing except a pair of short bathing-drawers, and a scarf or cloak (? chlamys) hanging behind his back and fastened round his neck by a knot. The slighter and perfectly sober man on the left is decently dressed in long drawers extending to his ankles, and a close-fitting tunic reaching below his knees. Both of the women are clad in a short tunic coming down a little below the waist, and possibly also in a long skirt. Each holds a piece of loose drapery across her legs. The woman on the left has it thrown over her left arm in the fashion adopted by some of the Gandhāra Bodhisattvas. Both women are adorned with heavy Indian anklets, armlets, and collars.

c11.s2.12

The reverse group, much mutilated, comprises five figures, of whom the principal is a fat elderly man sitting on a stone seat with his left leg tucked up, and so drunk that he has to be supported on his left side by a man and a boy, and on his right by a woman dressed like the females in the front group. The drunkard does not wear drawers like the merry fellow in that composition, but has a waistcloth loosely fastened. In style both reliefs are similar, the modelling being life-like, and the action clearly expressed.

c11.s2.13

The companion block of nearly the same dimensions, but somewhat larger, was discovered many years later by the late Mr. F. S. Growse at Pālī Khera, a suburb of modern Mathurā included within the limits of the ancient city. The reverse group, exhibiting the effects of deep potations, being almost identical with the reverse of the Stacy block, need not be further described. The front group, however, differs from its companion. Five figures under an *asoka* tree again appear. The principal is a fat man, seemingly nude, seated on a low stool made of stones laid in courses, with his left leg tucked up. The stone seat apparently is intended as a symbol of Mount Kailās, the abode of Kuvera. He is drinking from a noggin, apparently of wood, which a male attendant is ready to replenish. The proceedings are watched by another man, a woman, and a small boy (Fig. 80).

c11.s2.14

Two other Bacchanalian groups, found among the sculptures in the Mathurā Museum by Dr. Vogel and described by him, throw welcome light upon the date and meaning of the earlier discoveries described above. One of these groups, 1 foot 2 inches high (Fig. 81), represents a corpulent, coarse-looking man, apparently nude,

c11.s2.15

[Figure: FIG. 80. Pālī Khera block, front group. (Photo. 227, I. M. *List*: Mathurā Museum, C. 2, *Catal.*, p. 83, Pl. XIII.)]

c11.s2.16

squatted, and holding in his right hand a cup, which a female attendant is about to fill from a jar. His left hand grasps a long object, presumed to be a money-bag. This last attribute and the physique of the obese drinker permit of little doubt that the personage represented is Kuvera, the god of riches, whose podgy form has become familiar from the many images collected of late years in connexion with Buddhist monasteries from the Panjāb to Ceylon. Kuvera (also called Vaisravana and Jambhala) was king of the Yaksha demi-gods or sprites, and forms of his effigy are closely related to certain images from Gandhāra, such as Fig. 1 in Pl. XI of *The Gandhāra Sculptures*. Dr. Vogel probably is right in associating all the Bacchanalian sculptures of Mathurā with Yaksha worship.¹ But the exact meaning of the Stacy and Pālī Khera reliefs is far from clear. Whatever it may be, those sculptures have considerable merits as works of art.

c11.s2.17

¹ ‘Études de Sculpture bouddhique’ (*Bull. de l’École française d’Extrême Or.*, t. viii (1908), Nos. 3, 4, Fig. 2); *Ann. Rep. A. S., India*, 1906–7, ‘The Mathurā School of Sculpture,’ pp. 137–60. The second group in the Mathurā Museum differs little from the one figured.

c11.s2.18

Mr. Growse published a Mathurā statuette in his collection, 2 feet 8 inches in height, representing a pot-bellied youth with both hands raised, the right holding a bunch of grapes and the left grasping a goblet of calabash shape (Fig. 82).

c11.s2.19

The same scholar also published a mutilated statue, 3 feet 1 inch high, lying at Kūkargrāma in the Saadābād *pargana* of the Mathurā District (Fig. 83)—a singularly graceful figure of a Nāga youth with a canopy of seven cobra heads, holding his right

c11.s2.20

FIG. 81. Bacchanalian image, ? Kuvera, from Huvishka's monastery, Jamālpur mound. (A. S. photo., No. 33 : Mathurā Museum, C. 5, *Catal.*, p. 85.)

c11.s2.21

hand above his head, while his left grasps a cup similar in shape to that seen on the Pālī Khera block, but apparently without the curved handle. A garland of wild flowers is twined round his body, and he wears a high head-dress of ancient pattern. The pose of the youth is essentially the same as that of the female in the Woman and Tree motive, the garland taking the place of the stem of the tree. The worship of the Nāgas, the spirits of the waters, was much favoured by the ancient inhabitants of the Mathurā region.

c11.s2.22

The drinking Nāga is related to another fine life-size statue of a Nāga water-sprite from Chhargaon, near Mathurā, now in the Mathurā Museum, the approximate date of which is fixed by an inscription on the back, recorded in the fortieth year during the reign of Huvishka. According to the chronology provisionally adopted in this work, the statue (Fig. 84), which is 5 feet high, may be ascribed to the year A. D. 117 or 118. The modelling is good. The arrangement of the waistcloth in a twisted roll is found in other cases. The broken left hand probably held a cup.¹

c11.s2.23

¹ Vogel in *Prog. Rep. A. S., N. Circle*, 1907–8, p. 38; *J. R. A. S.*, 1910, p. 1313 note.

c11.s2.24

These so-called 'Bacchanalian' sculptures of Mathurā cannot be at all understood if considered by themselves. They evidently belong to a large class of Buddhist works of art, represented by the 'scènes bacchiques' of Gandhāra, which fill two plates of M. Foucher's book,² several reliefs on railing pillars at Mathurā, the 'Indian

c11.s2.25

² Foucher, *L'Art gréco-bouddhique du Gandhāra*, Figs. 127–33 *b*.

c11.s2.26

FIG. 82. Bacchanalian statuette. (*J. A. S. B.*, Part I, vol. xliv, Pl. XIV, 2 : Mathurā Museum, C. 6, *Catal.*, p. 87.)

c11.s2.27

FIG. 83. Bacchanalian Nāga, from Kūkargrāma. (*J. A. S. B.*, Part I, vol. xliv, Pl. XIV, 1 : Mathurā Museum, C. 15, *Catal.*, p. 90.)

c11.s2.28

Bacchus' of the Tank silver dish (*post*, Chap. X, Sec. 5), and the festive scenes depicted in the Aurangabad and Bāgh Caves. All such works appear to be expressive, as Mr. Growse suggested, of a little understood sensual form of popular Buddhism, not indicated by literature until a time seemingly much later than the second century. But when the true history of Indian Buddhism comes to be written it must be based on the evidence of the sculptures and pictures as much as on the books. M. Roller's question, addressed to Christian ecclesiastical archaeologists with reference to the art of the Catacombs, may be repeated to Indianists : 'La pierre ne servirait-elle pas à contrôler le manuscrit ?'³

c11.s2.29

³ *Les Catacombes de Rome* (1881), Preface, p. ii.

c11.s2.30

The excavations at Mathurā have yielded numerous specimens of pillars of stone railings associated with *stūpas*, both Jain and Buddhist. Most of the Buddhist ones were, I think, found on the site of Huvishka's monastery in the Old Jail or Jamālpur mound, now entirely removed. The Jain specimens came from the Kankālī mound, which included the remains of an early *stūpa* and two temples belonging to the Jains. The pillars have high-relief statuettes, usually of. females, on the front, and other

c11.s2.31

FIG. 84. Nāga statue, with inscription of Huvishka's reign, from Chhargaon. (A. S. photo., No. 47: Mathurā Museum, C. 13, *Catal.*, p. 88.)

c11.s2.32

FIG. 85. Yakshī on dwarf; Mathurā Museum. (A. S. photo., No. 47, 1908–9.)

c11.s2.33

panelled scenes, one above the other, or floral patterns on the back. The style of art is much the same, whether the work was intended for Jain or Buddhist use.

c11.s2.34

The rather immodest females adorning many of the pillars were supposed by Cunningham to be dancing-girls, an opinion certainly erroneous. They appear rather, as argued by Dr. Vogel, to belong to the Yakshī class, like the somewhat similar figures of the Bharhut rail (*ante*, p. 74). Some of the figures are really meant to be naked, but in others the apparent nudity is merely an artistic convention. The female [Figure: FIG. 86. Two Yakshīs (?); Indian Museum.

c11.s2.35

[Figure: FIG. 87. Female, half-back view; Mathurā Museum. (A. S. photo., No. 47; 1908–9.)]

c11.s2.36

[Figure: FIG. 88. Female with right arm bent; Mathurā Museum. (Photo., No. 47, A. S.)]

c11.s2.37

[Figure: FIG. 89. Female with right leg bent; Mathurā Museum. (Photo., No. 47, A. S.)]

c11.s2.38

is frequently posed in what I call the Woman and Tree arrangement. A few examples may be given. Many more have been published by Cunningham and other authors.

c11.s2.39

Fig. 85 represents a variant of the common Woman and Tree motive. The female stands on a prostrate dwarf, a male Yaksha. The pose, as in many other cases, is easy and graceful. Here the lady is draped.

c11.s2.40

A sculpture in Calcutta shows two females together, unmistakably nude, but for the bead girdle (Fig. 86).

c11.s2.41

A pillar in the Mathurā Museum (Fig. 87) presents a half-back view of a female with singular want of success. The difficulties of the pose seem to have been too much for the sculptor.

c11.s2.42

FIG. 90. Female and child; Mathurā Museum. (Photo., No. 47, A. S.)

c11.s2.43

FIG. 91. A soldier; Mathurā Museum. (Photo., No. 47, A. S.)

c11.s2.44

FIG. 92. Lion and rider; Indian Museum. (Photo. 844, I. M. *List*)

c11.s2.45

The unusual attitudes shown in Figs. 88 and 89 are treated much more skilfully. A child is introduced into Fig. 90.

c11.s2.46

The male figure, seemingly of a soldier, in Fig. 91 is quite exceptional and effectively designed.

c11.s2.47

A well-executed sculpture in the Indian Museum (Fig. 92) represents a youth riding a conventional lion, and may be of earlier date.

c11.s2.48

Dr. Vogel describes a mutilated statue (height 3 feet 10 inches or 1 m. 17) of a male deity standing with his left hand resting on his hip (Mathurā Museum, E. 12, *Catal.*, p. 108), which evidently had three heads, of which that on the proper right has been lost. The style indicates that the image belongs to the Kushān period. ‘It is of interest as the only polycephalic image which can be attributed to that epoch.’ Another sculpture in the Mathurā Museum (J. 7, *Catal.*, p. 143, Pl. XXII), 2 feet 7½ inches high, described by the same author, seems to date from the time of Kanishka. The subject of the alto-rilievo, which is on the front of a railing pillar, is ‘a male figure of Faun-like appearance with elaborate turban, necklace of beads, and other ornaments. He is standing under a mango-tree in blossom, with his right hand raised to his lips, and with his left placed against his thigh.’ This is a very artistic and well-modelled composition.

c11.s2.49

FIG. 93. A Bodhisattva, from the Katra; Mathurā Museum. (A. S. photo., No. 37 of 1908–9: Mathurā Museum, A. 1, *Catal.*, p. 47, Pl. VII.)

c11.s2.50

FIG. 94. Bodhisattva from Mathurā. (Photo. 843, *I. M. List*.)

c11.s2.51

A seated Bodhisattva (Fig. 93) in the Mathurā Museum, bearing a dedicatory inscription, ‘for the welfare and happiness of all beings,’ incised in script not later than the first or second century after Christ, is of special interest as exhibiting at that early date the saint seated in the traditional *yogī* attitude, which became general subsequently, and with his right shoulder bare. But for the inscription the principal figure might well be supposed to be of much later date. His drapery is excessively formal in its folds. The attendant figures are more ancient in appearance and well modelled. The simple foliage decoration is in good taste, and the flying spirits above are fairly successful. ‘Palaeographical evidence,’ Dr. Vogel observes, ‘points to the sculpture belonging to the early Kushāṇa period. It is, with the Buddha image of Anyor (No. A 2), the oldest image of Sākyamuni of which the date can approximately be fixed by an inscription, and must be one of the first Buddhist images made in Mathurā. On account of its artistic merit and excellent preservation also, this Bodhisattva is one of the most remarkable sculptures preserved in the Museum.' It is called a Bodhisattva in the inscription, and must represent Gautama Sākyamuni before he became a Buddha, because the tree shown over the image is the *pīpal* (*Ficus religiosa*), the tree of Gautama.

c11.s2.52

A Bodhisattva in the Indian Museum, Calcutta (Fig. 94), is a good sample of the more usual Mathurā type of the Buddhist saint, akin to the ‘royal personages’ of Gandhāra art, and very different in spirit from the seated image. It is almost a duplicate of an image found at Sārnāth.¹ No attempt is made to idealize the human form, which is treated naturalistically and without exaggeration.

c11.s2.53

¹ *Ann. Rep. A. S., India*, 1904–5, Pl. XXVI *c*.

c11.s2.54

A ‘tablet of homage’ (Plate XXXII), with a relief sculpture of a Jain *stūpa* (2 feet 4 inches high, 1 foot 9¾ in. wide), now in the Mathurā Museum, was found embedded in a wall near the Holī gate, but is said to have come from a field near the village of Maholī. It was dedicated by a certain courtesan named Lonasobhikā to the Arahat Vardhamāna or Mahāvīra, and gives a good picture of an ancient Jain *stūpa*, which was constructed and decorated on exactly the same lines as the Buddhist edifices of a similar kind. In this case the building depicted stood on a high plinth, and was approached by nine steps, leading to a *toraṇa* gateway of the Sānchī type, with a garland hanging from it. The *stūpa* was surrounded by a plain railing, and two similar railings were carried round the drum. The posturing females are unmistakably nude. The side columns are modified Persepolitan of poor design. The script of the inscription is of about the first century B.C.²

c11.s2.55

² Some of the details in *Jain Stūpa*, p. 61, based on Mr. Mukharji’s information, are erroneous.

c11.s2.56

FIG. 95. Nude female on Jain railing pillar, Mathurā. (*Jain Stūpa*, Pl. LXII, front.)

c11.s2.57

One example of the nude female, or Yakshī, on the pillar of an actual Jain railing, may be given (Fig. 95). She holds a weapon, which may be a broadsword of the old Indian kind.

c11.s2.58

A draped bracket figure for the side of a *toraṇa* (Fig. 96) may be compared with the similar, but superior, figure at Sānchī (*ante*, Fig. 46).

c11.s2.59

The cleverly drawn and executed bas-reliefs on the medallions of railing bars, probably Jain, representing animals, real or mythical, and other objects, may be com- Plate XXXII. Tablet with relief sculpture of a Jain *stūpa*. (Mathurā Museum, Q. 2, *Catal.*, p. 184, Pl. V. A. S. photo., No. 35 of 1908–9 = Pl. CIII of *Jain Stūpa*.)

c11.s2.61

pared with the more or less similar designs at Bharhut, Sānchī, and Bodh-Gayā (*ante*, Chap. III). Four specimens will suffice (Figs. 97–100).

c11.s2.62

Other decorative motives are illustrated and discussed in Chapter XI.

c11.s2.63

More than forty years ago Cunningham formed the opinion that it would not be advisable to undertake any further excavations at Sārnāth.¹ But happily the opinion

c11.s2.64

¹ *A. S. Rep.* i. 129.

c11.s2.65

FIG. 96. Draped bracket figure. (*Jain Stūpa*, Pl. XXXV, front.)

c11.s2.66

FIG. 97. Fish-tailed elephant. (*Jain Stūpa*, Pl. LXXIII, 1.)

c11.s2.67

expressed by so high an authority has not deterred Mr. Oertel and Mr. Marshall from fruitful explorations, which have proved the superficial character of the earlier researches, and have revealed an astonishing wealth of sculptures extending from the age of Asoka to the twelfth century. During that long period of about fourteen centuries, Sārnāth, the scene of the first preaching of the Law, continued to be regarded by Buddhists as one of the four most sacred spots in the world, and to be visited by myriads of pilgrims. Pious kings and monks lavished their wealth and devotion on the adornment of the numerous buildings which sprang up on the holy

c11.s2.68

[Figure: FIG. 98. Bull. (*Jain Stūpa*, Pl. LXXV, 3, front.)]

c11.s2.69

[Figure: FIG. 99. Shell. (*Jain Stūpa*, Pl. LXXI, 7.)]

c11.s2.70

[Figure: FIG. 100. Modified vine-leaf. (*Jain Stūpa*, Pl. LXXI, 3.)]

c11.s2.71

ground, and from age to age renewed or replaced the monuments fallen to decay. When the armies of the idol-hating Muslims sacked Benares in the closing years of the twelfth century, the rich and splendid establishments at Sārnāth were ruthlessly burnt and reduced to irretrievable ruin. But in these latter days the pick of the patient excavator has brought to light the remains of the ancient monasteries, temples, and statuary in such abundance that the history of Indian sculpture from Asoka to the Muhammadan conquest might be illustrated with fair completeness from the finds at Sārnāth alone.

c11.s2.72

The discoveries include many works of art dating from the Kushān period, all executed in sandstone, generally of the Chanār kind, but often of the red variety imported from Mathurā and wrought in the style favoured by the sculptors of that city. Several statues of Bodhisattvas, executed in the round on a large scale, are almost identical with the Mathurā specimen reproduced above (Fig. 94), and one of these is dated in the third year of the reign of Kanishka, which may be regarded provisionally as equivalent to A. D. 80. The Kushān age of such works is thus definitely determined. Halos, when present, are plain, not highly decorated as in the Gupta period.

c11.s2.73

A finely executed bas-relief, which once decorated a doorway and exhibits artistic lotus and vine patterns, besides a picture of an elephant worshipping a *stūpa*, is quite in the Mathurā style, and may be assigned with some confidence to the first century of the Christian era.¹ The style of the Sārnāth works is so closely related to that of Mathurā that illustrations may be dispensed with.

c11.s2.74

¹ For these sculptures see 'Excavations at Sārnāth,' *Ann. Rep. A. S., India*, pp. 59-104, with numerous illustrations, especially Pl. XXVI *a*, and XXVII, and text illustration, Fig. 12.

Section III. Amarāvatī

c11.s3.1

The sculptures from the *stūpa* of Amarāvatī and its surrounding railing or screen of marble may claim the distinction of being the best known specimens of early Indian art. No visitor to the British Museum, however indifferent to Indian curiosities, can help seeing the spoils of the *stūpa* and railing displayed on the walls of the grand staircase, and everybody at all interested in Indian antiquities is more or less familiar with Fergusson's *Tree and Serpent Worship* and other works treating of the sculptures. Objects so comparatively hackneyed and familiar do not call for prolonged discussion or extensive illustration in this volume.

c11.s3.2

The small town of Amarāvatī (16° 35′ N., 80° 24′ E.) on the south bank of the Krishnā (Kistna) river, in the Guntūr District, Madras, represents a more important ancient city called Dharanikota, a place of considerable note from at least 200 B. C. A richly decorated *stūpa*, known to have been in good repair and still venerated in the twelfth century, continued to exist to the south of the town up to the close of the eighteenth or the beginning of the nineteenth century, when it was utterly destroyed by a greedy local landholder, eager to obtain cheap building material and convinced that marble slabs, plain or carved, formed excellent food for a lime-kiln. About a century ago Colonel Mackenzie visited the place and had drawings made of numerous slabs, now no longer in existence. Various archaeological explorers have salved remnants of the sculptures, which are now mostly housed in either the British Museum or the Central Museum, Madras. Our knowledge of the extraordinary richness of the decoration of the *stūpa* and its railing is derived from the poor remnants thus rescued and Colonel Mackenzie's drawings, which have been published fully by Mr. Fergusson and Dr. Burgess.

c11.s3.4

The *stūpa* in its earliest form was of high antiquity, dating, as inscriptions prove, from about 200 b.c., and some fragments of sculpture perhaps as old still survive. But the great mass of the sculpture is much later, and belongs to the Kushān period. The authority of the Kushān kings, however, did not extend as far south as Amarāvatī, which was then within the dominions of the powerful Āndhra dynasty of the Deccan. By the help of two inscriptions mentioning Āndhra kings, the testimony of Tāranāth the historian of Buddhism, and other evidence of various kinds, the construction of the great railing may be assigned to the half-century between 150 and 200 after Christ. The highly ornate slabs which cased the *stūpa* itself may be a little later. We are almost certainly safe in saying that all the sculptures of the railing and casing fall within the hundred years between a.d. 150 and 250. Until quite recently everybody believed that there used to be two railings, and all the printed descriptions give details of an 'outer' and an 'inner' railing. But Dr. Burgess now states that he and everybody else were mistaken, the fact being that no more than one railing, the so-called 'outer' one, ever existed. The slabs supposed to have belonged to an 'inner' railing really formed a casing applied to the body of the *stūpa*.¹

c11.s3.5

¹ Fergusson, *Hist. of Ind. and E. Archit.*, 2nd ed. (1910), vol. i, p. 119.

c11.s3.6

The railing, by far the most magnificent known example of such structures, was 192 feet in diameter, about 600 in circumference, and stood 13 or 14 feet high above the pavement. It was constructed of upright slabs connected by three cross-bars between each pair of uprights, which stood upon a plinth and supported a coping about 2 feet 9 inches in height. On the outer face each upright was adorned with a full disk in the centre and a half-disk at top and bottom, minor sculptures filling the interspaces (Fig. 102). Similar but ever-varying disks decorated the cross-bars, and the coping was ornamented with a long wavy flower-roll carried by men, numerous figures being inserted in the open spaces (Fig. 103). The plinth exhibited a frieze of animals and boys, often in comic or ludicrous attitudes. The decorations on the inner face were even more elaborate; the coping presenting a continued series of bas-reliefs, and the central disks being filled with delicate sculptures, treating every topic of Buddhist legend. Thus every part of the structure, with a surface of about 16,800 square feet (600 × 2 × 14), was covered with sculptured reliefs.

c11.s3.7

The slabs forming the casing of the lower part of the *stūpa*, 162½ feet in diameter, were carved more richly even than the inner face of the railing, if that be possible. Apparently there were twelve in each quadrant, the principal object depicted on each slab being a highly decorated *stūpa* with its railing, the rest of the surface being covered with an infinite variety of sculptures. Study of Plate XXXIII, reproducing the best preserved of such slabs, will dispense with the necessity for detailed description, and at the same time give a good notion of what the appearance of the Amarāvatī *stūpa* must have been in the days of its glory. When fresh and perfect the structure must have produced an effect unrivalled in the world. However much severe taste may condemn the characteristic Indian lavishness of decoration which scorned to leave an inch of plain surface, the vast expanse of sculpture in white marble gleaming in the brilliant sunshine of India by day, or the light of myriads of lamps fixed on the surface of the dome by night, cannot have failed to exhibit a scene of unequalled splendour.

c11.s3.8

While abstaining from minute description of Plate XXXIII, which serves as a synopsis of the sculptures generally, I may invite the attention of the reader to a few points. In the relief picture the sculptured decoration is carried high up the dome, but the extant slabs seem to have been attached only to the lower part of the Amarāvatī

c11.s3.9

FIG. 101. Slab with representation of a *stūpa*, from votive *stūpa*, Amarāvatī. (A. S., Madras, photo.)

c11.s3.10

*stūpa*. It is possible that higher bands of decoration may have existed and been wholly destroyed. The railing in the relief has four cross-bars, and not only three as in the real monument. The 'moonstone' at the entrance agrees in form, though not in design, with the Ceylonese examples. The lions and some of the architectural forms are survivals of the Assyrio-Persian patterns of the Asokan age. The meaning of the five *stelae* or pilasters on the face of the *stūpa* is not known. The worshippers in the central scene adoring the chair occupied only by an object which may be a turban, might have appeared in a Sānchī or Bharhut relief, where images of Buddha are unknown; but here, at the top of the picture, we also find Buddha seated in the conventional *yogī* attitude. The flying figures are not quite successful, having too much the appearance of resting upon solid support. The frieze at the top of the slab contains nearly fifty figures, and the general effect, like that of nearly all the reliefs, is marred by excessive elaboration. But the skill of the artist in design and drawing, and his technical powers of execution, are beyond dispute.

c11.s3.11

FIG. 102. Basal medallion on pillar of rail, with plinth. (I. O., photo.)

c11.s3.12

FIG. 103. Undulating roll motive on coping of rail, Amarāvatī. (Photo. 774, I. O. *List*.)

c11.s3.13

Buddhists consider the multiplication of *stūpas* in any material to be a work of high merit, and accordingly the ancient Indian worshippers were in the habit of erecting in the precincts of sacred spots multitudes of minor *stūpas*, ranging in size from considerable masonry buildings to tiny models. The drum of such a votive *stūpa*, eleven feet in diameter, was disclosed at Amarāvatī by recent supplementary excavations, and a panel of its decoration is reproduced in Fig. 101. It will be observed that sundry details differ from those on the larger slab from the great *stūpa* and that the carving is slightly less crowded. The standing man and woman at the base are natural and well executed. The lions, as usual at Amarāvatī, are stiffly and conventionally designed. The whole composition, regarded as a display of skilled craftsmanship rather than as fine art, deserves high praise.¹

c11.s3.14

¹ *Ann. Rep. A. S., India*, 1905–6, Pl. LI, LII.

c11.s3.15

FIG. 104. Pilaster, Amarāvatī. (Photo. of Madras A. S.)

c11.s3.16

The infinite variety of the patterns used in the medallions and bars may be realized by study either of actual examples or of the relief pictures. Fig. 102 is an excellent and well-preserved example of a charming decorative design based on the lotus-flower motive. The beauty and delicacy of the floral devices in the border and plinth deserve special notice and admiration. They will repay minute examination with a magnifying glass.

c11.s3.17

Fig. 103 gives a characteristic specimen of the Amarāvatī treatment of the wavy garland or roll motive. The Hellenistic nude Erotes have developed into full-grown Indian men in waistcloths, and the imbricated Roman roll of Gandhāra is replaced by a much thicker roll of tinsel covered with elaborate patterns.

c11.s3.18

The pilaster shown in Fig. 104 is another of the recent discoveries. It exhibits the old worship of symbols combined with the seated image of Buddha, as in the slab already described. Buddha, who has both shoulders covered, is seated in the *yogī* posture, which, as we have seen (*ante*, Fig. 93), was already adopted at Mathurā at about the same date, probably not later than the middle of the second century after Christ. The four-petalled flower often occurs in Gandhāran art, and in Khotan, and is still a traditional pattern in Western Tibet. The horse's head is good. I cannot identify the legend.

c11.s3.19

The treatment of floral and animal decorative motives has been illustrated above by photographs on a small scale. Three specimens may be added from Mr. Rea's drawings on a larger scale, which have not been published except in his book (Figs. 105-7).

c11.s3.20

A few separate images have been found at Amarāvatī. Two large marble statues, 6 feet 4 inches in height, discovered during the recent supplemental excavations, are illustrated in Fig. 108. The opaque drapery is treated in a formalized style, quite different from the smooth transparent robes of the Gupta period, to be discussed in the next chapter, but to a certain extent resembling Gandhāra work, and

c11.s3.21

[Figure: FIG. 105. Lotus forms. (Rea, *S. Ind. Buddhist Antiq.*, Pl. XLII, 2.)]

c11.s3.22

[Figure: FIG. 106. Lotus and *makara*. (Ibid., Pl. XLIII, 1.)]

c11.s3.23

[Figure: FIG. 107. A pond. (Ibid., Pl. XLVII, 1.)]

c11.s3.24

the paintings on the columns in Cave X, Ajaṇṭā. These images may date from the third or fourth century.

c11.s3.25

Fergusson's opinion that the sculptures of the Amarāvatī school mark 'the culmination of the art of sculpture in India', which was generally accepted until recently by English writers, including myself, does not now command such ready assent. I will not presume to say which work marks the 'culminating point', but it is certainly safe to affirm that the pre-eminence claimed for the Amarāvatī reliefs may be effectively challenged by compositions of later date, at least in some respects. All critics, however, can agree with Mr. Havell that the marbles of Amarāvatī offer 'delightful studies of animal life, combined with extremely beautiful conventionalized ornament', and that 'the most varied and difficult movements of the human figure are drawn and modelled with great freedom and skill'. The obvious overcrowding of the compositions unfortunately is a defect common in Indian art. Historically, the sculptures are interesting as an academic development of the style of Sāñchī and Bharhut, with a stronger tinge of Hellenistic influence, perhaps coming by sea from Alexandria and through the ports rather than from Gandhāra. Considering the geographical and political separation of the Kushān and Āndhra empires, I think the presumption is that the sculptors of Amarāvatī had not direct knowledge of the Gandhāra school, although it is possible that they may have had it. Hiuen Tsang, the Chinese pilgrim, in the seventh century, did not really describe the *stūpa* as being 'ornamented with all the magnificence of the palaces of Bactria (Tahia)', as Fergusson and Burgess suppose him to have done. A slight slip of the pen in the Chinese text used by Julien introduced the word mistranslated as 'Bactria'. The pilgrim really praised two monasteries in the Deccan as 'having all the artistic elegance of a great mansion and all the beauty of natural scenery'. The assumption made by Dr. Burgess and other authors that the account of *two monasteries* given by Hiuen Tsang should be applied to the *stūpa* of Amarāvatī is far from being established. Thus disappears the basis for Fergusson's argument that the school of Amarāvatī should be considered the offspring of the marriage of the art of the North—that is to say, Bactria as represented by Gandhāra—with that of interior India as represented by Sāñchī and Bharhut.¹ Mr. Havell may be right in believing that origin-

c11.s3.26

¹ Fergusson, *Hist. Ind. and E. Arch.*, reprint of 1899, p. 103; new ed. by Burgess (1910), vol. i, p. 123. 'Instead of the *ta-hsia*, a "great mansion", here, the B text, used by Julien, has *ta hsia*, which is a Chinese name for the country called Bactria. But this is evidently a slip of the pen, and the proper reading is that of the other texts which means a “great mansion”' (Watters, *On Yuan Chwang's Travels in India* (1905), vol. ii, p. 218). This material correction and Mr. Watters's comments on the current 'identification' of the pilgrim's *monasteries* with the Amarāvatī *stūpa* have been overlooked in the revision of Fergusson's book. 'It is hard,' Mr. Watters observes, 'to understand how any one could propose to identify a large monastery among hills and streams, and having spacious chambers and great corridors, with a building which is only a remarkable tope situated on a plain.' The error concurred in by Julien, Fergusson, and Dr. Burgess will not readily disappear from books on Indian art and antiquities.

c11.s3.27

[Figure: FIG. 108. Marble Buddhas. (Photo., A. S., Madras; *see Ann. Rep. A. S., India*, Pl. LI, 3.)]

c11.s3.28

ally the effect of the Amarāvatī marbles was heightened by colour, and in holding that technically they should be regarded as 'painted rilievos' rather than as true sculpture. But whether they were painted or not, they must have formed, when perfect, one of the most splendid exhibitions of artistic skill known in the history of the world.

c11.s3.29

FIG. 109. Sculptured and inscribed pedestal, 2½ feet high, in front of Jain shrine at Dānavulapāḍ, Cuddapah District. (Photo. 535, A. S.)

Section IV. Dānavulapāḍ

c11.s4.1

I am disposed, although with hesitation, to refer to the Kushān period, about the second century after Christ, certain Jain sculptures excavated a few years ago at Dānavulapāḍ in the Cuddapah District, Madras, and associated with large bricks (1'~9″ × 9″ × 4″) of the dimensions of those found in the ruined *stūpas* of the Krishnā (Kistna) District. The images (Figs. 109–10) are carefully modelled, and their style is not inconsistent with the date assumed, but they may be a century or two later. The inscription on the circular base is not dated. The material is a fine white limestone or marble.¹

c11.s4.2

¹ Rea, *Ann. Rep. A. S., India*, 1905-6, pp. 120-7, with illustrations.

c11.s4.3

[Figure: FIG. 110. Jain Tīrthaṅkara and Yakshī; near Dānavulapāḍ, Cuddapah District. (Photo. 537, A. S.)]

c11.s4.4

*Mathurā and Sārnāth*:—ANDERSON—*Catal. Archaeol. Coll. I. M., Calcutta* (1883): CUNNINGHAM—*Archaeol. S. Rep.*, vols. i, iii, xi, xvii, xx: GROWSE—*Mathurā, a District Memoir*, 3rd ed. (Allahabad, 1883): MARSHALL and OERTEL—*Ann. Rep. A. S., India*, 1903-4 and 1904-5; *J. R. A. S.*, 1907, pp. 995-1000: SMITH, V. A.—*The Jain Stūpa and other Antiquities of Mathurā* (Allahabad, 1901): VOGEL—'The Mathurā School of Sculpture' (*Ann. Rep. A. S., India*, 1906-7, pp. 137-60, in progress).

c11.s4.5

*Amarāvatī*:—BURGESS—*The Buddhist Stūpas of Amarāvatī and Jaggayapeta* (1887, being vol. vi of *New Imp. S. of A. S. Rep., India*): FERGUSSON—*Tree and Serpent Worship*, 2nd ed. (1873); *Hist. of Ind. and E. Archit.*, 2nd ed. by BURGESS (1910): REA—*Ann. Rep. A. S., India*, 1905-6, pp. 116-19; *South Indian Buddhist Antiquities* (1894, being vol. xv of *New Imp. S. of A. S. Rep., India*).

c11.s4.6

Other references in footnotes.