Pramāṇa · Search · About · API

History of Indian and Indonesian Art c8.s3: Gupta Period 320—600 A. D.

History of Indian and Indonesian Art c8.s3

Language
English

Gupta Period 320—600 A. D.

c8.s3.1

A rājā of Pātaliputra, who assumed the name of Candragupta I and extended his dominions as far as Allahābād (Prayāg), established the Gupta era, 319—20, to commemorate his coronation. Samudragupta extended the kingdom to the Satlaj, and made conquests in Southern India. Candragupta II, the legendary Vikramāditya, annexed Mālwā and Ujjain, and dispossessed the Saka rulers of Surāṣṭra, known as the Western Satraps; he removed the capital to Ayodhyā. The White Hūns invaded northern India in the reign of Kumāragupta I, and in the time of Skandagupta, about 480, broke up the empire. In about 528 the Hūns, under Mihiragula, were defeated by Balāditya, a later Gupta, allied to a rājā of Mālwā, and those of the Hūns who were not permanently settled in Rājputāna retired to Kaśmīr. As a culture period, and for the purposes of this book, the Gupta period is taken as covering the years 320—600 A. D.

c8.s3.2

The outstanding characteristic of the art of India at this time is its classical quality. In the Kuṣāna period the cult image is still a new and important conception, and there we find, quite naturally, magnificent primitives, or "clumsy and unwieldy figures", according to our choice of terms. In the Gupta period the image has taken its place in architecture; becoming necessary, it loses its importance, and enters into the general decorative scheme, and in this integration acquires delicacy and repose. At the same time technique is perfected, and used as a language without conscious effort, it becomes the medium of conscious and explicit statement of spiritual conceptions; this is equally true of sculpture, painting, and the dance. With a new beauty of definition it establishes the classical phase of Indian art, at once serene and energetic, spiritual and voluptuous. The formulae of Indian and of which the influence was felt, not only throughout India and Ceylon, but far beyond the confines of India proper, surviving to the present day.

c8.s3.3

The period is often described as one of the revival of Brāhmaṇism and of Sanskrit learning and literature. But actually there is no evidence of any preceding lack of continuity in the development of Brāhmaṇical culture. The *kāvya* style is already foreshadowed in the *Rāmāyaṇa* and fairly well developed in the second century A. D. Certainly there had never existed a "Buddhist India" that was not as much and at the same time and in the same areas a Hindū India. In any case, an age of heightened aesthetic consciousness, of final redactions of the Epics and *purāṇas*, and of codifications and systematisation in the arts¹ must have been preceded by centuries, not of inactivity, but of intense and creative activity. The period is thus one of culmination, of florescence, rather than of renaissance. No more than a passing allusion can be made here to the close parallels that exist at this time between the development of art and literature: the same abundance pervades the Sanskrit *kāvya* literature, the Ajaṇṭā paintings and the decoration of the Gupta reliefs.

c8.s3.4

¹ An important piece of evidence given by Hsüen Tsang proves what might in any case have been inferred, the existence of *Śilpa-śāstras* in the late Gupta period: he mentions five *Vidyās* or *Śāstras*, of which the second is the *Śilpasthānavidyā* (Beal, I, p. 78).

c8.s3.5

The rich decorative resources of Gupta art are to be understood in terms of its inheritance, indigenous, Early Asiatic, Persian and Hellenistic. The Gupta style is unified and national. Plastically, the style is derived from that of Mathurā in the Kuṣāna period, by refinement and definition, tendencies destined still later, in the natural course of events, to imply attenuation. Meanwhile Gupta sculpture, though less ponderous than the ancient types, is still distinguished by its volume; its energy proceeds from within the form, and is static rather than kinetic, a condition that is reversed only in the mediaeval period. In all these respects Gupta art marks the zenith in a perfectly normal cycle of artistic evolution. In India, as elsewhere, we find a succession of primitive, classical, romantic, rococo, and finally mechanical forms; the evolution is continuous, and often, especially in the earlier periods, rapid; and wherever our knowledge is adequate, Indian works, like those of other countries, can be closely dated on stylistic evidence alone.

c8.s3.6

The school of Gandhāra, in the earlier part of the Gupta period, continues to flourish in the North West, though in more or less Indianised forms. The remains at Jauliāñ and Mohṛā Morādu (Taxila) afford a good illustration of its character. The former consist of a main and smaller *stūpas*, chapels, and a monastery; the latter with an assembly hall, refectory, kitchen, store-room, bathroom and latrine, indicating a comparatively luxurious development, and that the monks no longer depended upon the begging bowl for all their food. The sculptures date from a little before or after 400 A. D. There is no evidence that any appreciable production in stone took place after the third century; almost all the Jauliāñ sculpture is executed in clay or stucco, once coloured and gilt. The style is still fairly vigorous, freer, indeed, and more animated than that of the earlier work in stone; it is at once less refined, less well-considered, and less academic. It is profoundly Indianised; but it cannot be equated in any aesthetic sense with the central productions of Gupta art, and in comparison with these is essentially provincial¹.

c8.s3.7

¹ Marshall, 6, 7.

c8.s3.8

The Bhallaṛ *stūpa*, of third or fourth century date, at Taxila, with an unusually high drum, stands on a rectangular basement approached by one flight of steps, illustrating the simple form from which the many-terraced types of Kaśmīr and Java and Burma were later evolved².

c8.s3.9

² Marshall, 5, pl. XXVIII.

c8.s3.10

Farther to the North West, at Chārsada and other old sites near Puṣkalāvatī, Gandhāran stucco and clay figures have been found, similar to those of Jauliāñ, but of finer and more pleasing quality, though likewise dating about 400³.

c8.s3.11

³ Marshall and Vogel.

c8.s3.12

Other stūpas of early Gupta date are found in the Sind valley, and of these the Mīrpur Khās example is the most important; it is a brick structure standing on a square basement, and chiefly remarkable for the existence of three small chapels or cellas within the mass of the basement on the western side, affording the only Indian instance of a type of structure combining *stūpa* and chapels in a way later on to be greatly elaborated in Burma. In the central chapel there is a true brick arch⁴. The decoration consists of carved bricks, like those of Jamāl-gaṛhī, Bikanīr, and other early Gupta sites, both Buddhist and Hindū. There are also terracotta Buddhas, with Gandhāran affinities, and the figure of a donor, still preserving its original colours, the flesh wheat-coloured, the hair or wig black, the waist-cloth red⁵. Most of the Gandhāra sites seem to have been wrecked by the White Huns under Mihiragula in the latter part of the fifth century, and this practically ended the activity of the school. The original influence, nevertheless continues to be apparent in the architecture and sculpture of Kaśmīr, and that of a few related monuments, such as that at Malot (fig. 274)¹, dating from the time of Kāśmīrī domination in the Pañjāb.

c8.s3.13

⁴ It may be remarked here that many isolated occurrences of a true vaulted arch are found in Indian architecture of pre-Muḥammadan date. E. g., Piprāwā, Peppé and Smith; Pāṭaliputra, Maurya arch stone, A. S. I., A. R., 1921—22, pl. XXXVI; Bhītārgāon, Cunningham, 4, vol. XI; Nālandā, A. S. I., E. Circle, 1916—17, p. 45; Bodhgayā, Cunningham, 3, pp. 85, 86, and 4, vol. XI, pp. 42, 43, and Mitra, pp. 104 ff.; Konch, Peppé, J. A. S. B., XXXV, pt. 1, p. 54; Kāfir Koṭ, A. S. I., A. R., 1920—21, pt. 1, p. 7; Kaśmīr, Sahni, 3, p. 73; Kiyul, Cunningham, 4, vol. III, p. 157; Burma, Fergusson, 2, p. 352; and at Poḷonnāruva, in Ceylon. All these represent true voussoirs, not merely the pointed arch form, which also occurs in monolithic and corbelled construction. As a rule in the Indian arches the bricks are placed sideways so that the thin edges are in contact. For pre-Muḥammadan buttresses see A. S. I., A. R., 1922—23, p. 118; for domes, Fergusson, 2, vol. 1, pp. 312—319. See also page 12 and fig. 146.

c8.s3.14

⁵ Cousens, 5, pl. XXXVIII; 8, pl. 14.

c8.s3.15

¹ Burgess, 8, pls. 237—241; A. S. I., A. R., 1918—19, p. 5, and 1920—21, pl. 11.

c8.s3.16

The Buddha figure in the early Gupta period is fully evolved, and this classical type is the main source of all later forms both in and beyond the Indian boundaries. The only example of the old Kuṣāna type with shaven head is the Māṅkuwār image, dated 448/9 A. D. (fig. 162). This figure at the same time exhibits a peculiarity rather common in the Gupta period, that of webbed fingers². Apart from this exceptional figure, the Gupta type is characterised by its refinement, by a clear delineation and definition of the features, by curly hair, absence of *ūrṇā*, greater variety of *mudrās*, elaborately decorated nimbus, the robe covering one or both shoulders and extremely diaphanous, clearly revealing the figure; and by a lotus or lion pedestal, usually with figures of donors. Scarcely any trace of Hellenistic plasticity is apparent.

c8.s3.17

² Other examples, B (b) 103 and 181 at Sārnāth, B 10 Lucknow Museum (from Mathurā), and reliefs at Cave XIX, Ajaṇṭā.

c8.s3.18

The leading variations are exemplified in the fine fifth century image by the colossal standing image from Mathurā (fig. 158), the beautiful but less vigorous seated figure B (b) 181 at Sārnāth (fig. 161) and others at the same site, the Sultāṅgañj copper image of over life-size in Birmingham (fig. 160), and the figures in relief at Ajaṇṭā, Cave XIX (fig. 154), and those of Kārli, Kaṇheri (fig. 164) and other western caves.

c8.s3.19

All of these are executed in local material, at Sārnāth, for example, in Chunār sandstone; it is obvious that by this time local ateliers existed at every sacred site. But that Mathurā still maintained a high reputation is illustrated by the existence of Buddha images in Sikrī sandstone, e. g. at Kasiā (colossal Parinirvāṇa image made by Dinna of Mathurā and seen by Hsüan Tsang)³, at Bodhgayā⁴, Prayāg (Māṅkuwār, mentioned above) and Sāñcī.

c8.s3.20

³ For Kasiā (= Kuśinagara, site of the Parinirvāṇa) see Sastri, H., 1; Vogel, 5, and 13, and in A. S. I., A. R., 1906—07, pp. 49 ff.; and Sahni, 4.

c8.s3.21

⁴ Cunningham, 3.

c8.s3.22

In view of the wide distribution of Mathurā images in the second, third, fourth and fifth centuries, it is easy to understand the evident derivation of the Gupta from the Mathurā type, and the fact that, as Smith remarks apropos of the Sārnāth figure B (b) 181, the Gupta Buddha is "absolutely independent of the Gandhāra school"⁵. As Marshall too observes, "Hellenistic art never took a real and lasting hold upon India”¹. In fact “le buddha de Mathura, ce prototype d’inspiration et de facture indiennes et peut-être même l’authentique ancêtre de toutes les images du Bienheureux, ne s’est pas éclipsé au contat de l’art Gandharien et ... a survécu à la vogue classique sans avoir subi d’altération essentielle”².

c8.s3.23

⁵ Smith, 2, p. 170.

c8.s3.24

¹ Marshall, 8, p. 649.

c8.s3.25

² Goloubew, in B. É. F. E. O., 1923, p. 451.

c8.s3.26

Thus the famous theory of the Greek origin of the Buddha image, propounded by Foucher, and since adopted by many scholars, proves to lack all solid foundation, and falls to the ground, and with it the implied Greek inspiration of other Indian images, Brāhmaṇical and Jaina. The fact that a Hellenistic element, plastic and iconographic, of some kind, enters into and is absorbed by Indian art, remains. Opinions may differ as to its extent and significance; its importance is slight, and perhaps rather historical than aesthetic.

c8.s3.27

Gupta architecture may be discussed under heads as follows: (1) *stūpas*; (2) excavated *caitya*-halls and *vihāras*; (3) structural *caitya*-halls and apsidal Hindū temples; (4) flat-roofed temples; (5) *śikhara* shrines, and exceptional types such as those at Gop, Bodhgayā, and the Maṇiyār Maṭha; (6) palace and domestic architecture and the theatre.

c8.s3.28

Only two structural *stūpas* of Gupta date survive outside the Gandhāra area in anything like a fair state of preservation, both of the cylindrical type, the globular dome of the monolithic *caityas* being, no doubt, difficult to realise constructively. The first is the well-known Dhamekh *stūpa* at Sārnāth, probably of sixth century date. The structure consists of a circular stone drum, resting on the ground level without the usual rectangular basement; above this drum rises a cylindrical mass of brickwork to a total height of 128 feet. Halfway up the base are four niches which must have held Buddha images; immediately below these niches is a broad course of exquisitely carved elaborate ornament, geometrical and floral, in the manner of the painted ceilings at Ajaṇṭā. The other *stūpa* is the later of the two Jarāsandha-kā-Baiṭhak at Rājagṛha, a tower-like erection, rising from a substantial basement, and dateable about 500.

c8.s3.29

The caves afford numerous examples of monolithic forms; here, e. g., at Cave XIX, Ajaṇṭā, there is usually a high cylindrical drum, decorated with standing or seated Buddha figures between pilasters crowned by a *makara*-arch, richly ornamented, and supporting a globular dome (*aṇḍa*) with the usual pavilion (*harmikā*) and range of umbrellas (*chatravali*). This form is directly derived from that of votive *stūpas* of the Kuṣāna period such as N 1 in the Mathurā Museum³. Outside Cave XIX, at Ajaṇṭā, on the right hand side, there is a relief apparently representing a pavilion with a globular dome and umbrellas; this is not really a domed pavilion but the elevation of a solid *stūpa* like that within the hall¹.

c8.s3.30

³ Vogel, 13, pl. IV.

c8.s3.31

¹ A structural stone domed pavilion is unmistakeably represented at Amarāvatī, see fig. 146.

c8.s3.32

There exist many “caves” of the Gupta period. At Ajaṇṭā Caves, XVI and XVII are *vihāras* dating about 500 A. D., Cave XIX a *caitya*-hall dateable about 550; all of these contain paintings, referred to below.

c8.s3.33

The two *vihāras*, XVI and XVII are pillared halls with the usual cells and the addition of shrines in the back wall containing seated Buddhas in *pralambapāda āsana*, “European fashion”, which now appears for the first time. The beauty and variety of the pillars in these *vihāras* is remarkable, the types in the two caves differing, and no two of any type being exactly alike. In Cave XVI, vertically or spirally fluted pillars are characteristic, with rounded bracket capitals, sometimes with horizontal ribs like the early Pallava brackets of the South. In Cave XVII the pillars are square above and below, the centre is fluted, and the brackets are provided with squatting figures of *gaṇas* supporting the horizontal cross-beams, and this placed back to the roof and face downwards; this type of *gaṇa* capital becomes almost universal in mediaeval architecture.

c8.s3.34

The *caitya*-hall, XIX, retains the plan of the early types, but with extensive changes in the façade and a great development of (Mahāyāna) sculpture. The façade (fig. 154) is a further development of the Nāsik type, but in place of the railing, which at Nāsik extends across the whole width of the wall from side to side, separating the doorway from the window above it, there is a double roll cornice decorated with *caitya*-windows framing heads, a form most likely of Āndhra origin, but already common in early Gupta work. Above these cornices, the frame of the great window stands out in relief against a many storied screen of architectural reliefs; below it is the flat-roofed entrance porch supported by four pillars, and very shallow. The aisle pillars within are richly ornamented fluted columns with pot and foliage capitals, and massive, decorated, rounded brackets, supporting an elaborate frieze of niches with Buddha figures. The *stūpa* is of the type already described, with a range of three heavy umbrellas, far removed from wooden forms. Outside, right and left of the façade and on the walls of the excavated court in front of the cave are many more Buddha figures in relief; the type is full-fleshed, but gracefully *hanché*, and the drapery is treated with the greatest possible simplicity, closely moulding the body. It is these types, or those of Sārnāth, which are as nearly as can be indicated, analogues of the pre-Khmer Indianesque Buddhas of Romlok, while, as remarked elsewhere, the Kaṇheri sculptured reliefs are no less closely related to the Stoclet Avalokiteśvara. In this connection it may be remarked that the Vākāṭaka kings to whom Caves XVI and XVII are due, to some extent successors of the Āndhras in the Dekkhan, controlled the Telugu country almost to the mouths of the Godāverī, and by this route the Gupta tradition found easy access to the East¹.

c8.s3.35

¹ For Ajaṇṭā see Fergusson, 2, vol. 1; Fergusson and Burgess; and Burgess, 4.

c8.s3.36

Closely related to those of Ajaṇṭā are the *vihāra* and *caitya* caves at Bāgh, which are likewise painted, and date about 500².

c8.s3.37

² For Bagh see Haldar; Luard; Dey. A full publication by the India Society is announced.

c8.s3.38

At Elūrā, the Viśvakarmā *caitya*-hall is internally like Cave XIX at Ajaṇṭā; externally, it is remarkable for its unique façade, of which the lower storey is a verandah with pot and foliage capitals and the upper (fig. 155) contains a divided window flanked by two niches in which are standing figures of Buddha. The superstructure of these niches is two-storied, with angle *āmalakas*, and is topped by a *kīrttimukha*. The excavation is Gupta or early Cāḷukyan, dating about 600³.

c8.s3.39

³ Fergusson, 2, vol. 1, p. 159; Burgess, 8, pl. 275.

c8.s3.40

Of caves in Kāṭhiāwāḍ, the most interesting and beautiful is the two storeyed pillared hall in the Uparkoṭ at Junagaṛh. The varied fluted columns, capitals with elaborate figure groups like Ajaṇṭā paintings, cornices with *caitya*-window niches like those at Gop, the acanthus ornament of the pillar bases like the Bhumara lintel, all point to a late Gupta date. The excavation includes a bath and lacks the ordinary cells of a monastery; it would almost seem that it may have been the underground summer chamber of a palace. It is in any case one of the most elegant of all works of the Gupta period, and fully the equal of the little temple, no. 17 at Sāñcī⁴. The important group of cave and structural temples at Udayagiri, Bhopāl, mostly Brāhmaṇical, is nearly related to the same Sāñcī type and that of the Tigowā series⁵.

c8.s3.41

⁴ Burgess, 1.

c8.s3.42

⁵ Cunningham, 4, vol. IX.

c8.s3.43

Several structural apsidal temples, planned like the *caitya*-caves, have survived, including one Brāhmaṇical example. At Ter (= Tagara), Sholāpur District, the structural brick *caitya*-hall, of fourth century or perhaps earlier date, seems once to have enshrined a *stūpa*, and only subsequently to have been converted to Vaiṣṇava usage. Characteristic external features are the barrel-roof, rounded at the rear end, and terminating above the entrance in a gable-end of *caitya*-window form, enclosing an architectural relief; roll mouldings; and walls decorated with simple pilasters. The *maṇḍapam* is perhaps a little later in date⁶. At Chezārla, in the Kistna District, the Kapoteśvara temple (fig. 147) is similarly a structural *caitya*-hall, originally Buddhist and later converted to Hindū usage. Here the gable end is decorated with reliefs including both architectural forms and figures; the roll mouldings are more developed, but the wall is plain. Near this temple is a curious little rectangular cella recalling Indianesque types of Hanchei; and a number of small monolithic votive shrines, with domed roofs decorated with single *caitya*-arches, like the rock-cut Pallava shrines at Bhairavakoṇḍa in the Guntūr District, and the Arjuna Ratha at Māmallapuram¹.

c8.s3.44

⁶ Cousens, 1.

c8.s3.45

¹ Burgess, 2, vol. I, p. 126; Longhurst, 3, pl. XIII; A. S. I., Southern Circle, A. R., 1917—18; Diez, p. 29.

c8.s3.46

The Brāhmaṇical Durgā temple at Aihoḷe (fig. 152) is probably of sixth century date, and rather early Cāḷukya than Gupta, but is connected with the types now described. Entirely of stone, it follows the plan of the apsidal *caitya*-halls, but the roof is flat and constructed of stone slabs, a northern *śikhara* rises above the *garbha-gṛha*, and there is a verandah, roofed with sloping slabs, supported by massive square columns with heavy brackets. The whole stands on a high basement of several horizontal courses, of which one is fluted, another decorated with *caitya*-arches, and another with reliefs².

c8.s3.47

² Cousens, 4.

c8.s3.48

Small, flat-roofed shrines consisting of a cella with almost plain walls, generally with a shallow verandah, and often surrounded by a pillared hall, and without any kind of *śikhara* are typical of the early Gupta period. The beautiful little shrine at Sāñcī, temple 17 (fig. 151), is a good example. Here the verandah pillars exhibit a typical development; the capitals are square and very massive, with addorsed animals now separated by a tree; this form is found also at Tigowā, Erān, Gaṛhwā, and Udayagiri. It is characteristic, too, that the line of the verandah architrave is carried round the wall of the otherwise plain cella as a string course³.

c8.s3.49

³ Cunningham, 4, vol. X; Marshall, 5.

c8.s3.50

At Tigowā, C. P., there is a flat-roofed Hindū shrine of identical design; the roof slabs are fitted together by overlapping grooves, as in the case of many of the flat-roofed temples at Aihoḷe. In the case of the Patainī Devī temple near Uchahara the roof consists of a single slab⁴. Other and simpler flat-roofed shrines are illustrated in the Goṇḍ temples of the Lalitpur District⁵.

c8.s3.51

⁴ Cunningham, 4, vol. IX.

c8.s3.52

⁵ Mukerji.

c8.s3.53

Two extremely interesting flat-roofed temples have been found at Bhumara⁶ in Nagoḍh State and at Nāchnā-Kuṭharā⁷ in Ajaigaṛh, Bundelkhaṇḍ. A description of the former will suffice for both. The Bhumara Śiva temple consists of a masonry cella (*garbha-gṛha*) with a flat slab-roof and a carved doorway having representations of river-goddesses on the jambs and a fine bust of Śiva, with flying figures, on the lintel. The cella contained a Śiva-liṅgam of the type of the still finer example existing at Khoh in the same State. Around the *garbha-gṛha* are the scattered remains of a larger chamber which surrounded it, providing a roofed *pradakṣiṇā patha*, and of a *maṇḍapam* attached to and preceding this enclosure. These remains consist of a *great* variety of columns which are not monolithic, of richly carved lintels that supported the roofing slabs, of *caitya*-window niches from the cornice, of parts of the doorway, and of carved slabs which decorated the lower part of the outer wall like a deep wainscot. Some of the *gaṇa* figures have *rākṣasa* faces on their bellies. There is a tendency to unrestrained development of arabesque¹.

c8.s3.54

⁶ Baneryi, 3.

c8.s3.55

⁷ A. S. I., Western Circle, 1919, pl. XV—XVII and pp. 53, 60.

c8.s3.56

¹ For Bhumara, see Banerji, 3. The *rākṣasa* faces are found also in the Durgā temple, Aihoḷe, in Cave III at Aurangābād, and at Prambanam in Java. The motif seems to have originated in Gandhāra (Spooner, I, fig. 3).

c8.s3.57

The most interesting flat-roofed temple in the Dekkhan is the Lāḍ Khān at Aihoḷe (fig. 148). This temple, dating about 450 A. D. is very low and flat, its walls consisting of stone slabs set between heavy square pilasters with bracket capitals; roll-mouldings decorated with small well-spaced *caitya*-arches are characteristic of the roof. On the pillars of the porch are figures of the river-goddesses, which are most characteristic of Gupta work and persists into the mediaeval period, extending also to Java. On the roof is a small square cella of slab construction, with a porch, forming an independent shrine of the Sun. The walls have central projecting niches with reliefs. The windows are stone slabs, perforated in a variety of beautiful designs. Three other temples at Aihoḷe have either never possessed, or did not originally possess a *śikhara*, that of Kont Guḍi having been added as late as the tenth or eleventh century².

c8.s3.58

² For the Lāḍ Khān and other Aihoḷe temples see Cousens, 4. The resemblance of the slab cella at Hanchei to the roof shrine of the Lāḍ Khān will be remarked in respect of the construction, with Bhumara as a very possible analogy for the surrounding building, which may have been of wood at Hanchei.

c8.s3.59

There are other low, massive, flat-roofed, cave-like temples, not unlike the Lāḍ Khān, but provided with simple Nāgara *śikharas* above the cella; these shrines, originally Vaiṣṇava, have been later converted to Śaiva usage, and it is just possible that their *śikharas*, together with that of the Durgā temple, are later additions. The best examples of this group are the Hucchīmallīguḍi (fig. 153), which is not at all unlike the well-known Paraśurāmeśvara temple at Bhuvaneśvara (fig. 216), but much more severe, and with only two courses between successive angle-*āmalakas*; and the temple in Field 270³.

c8.s3.60

³ Cousens, 4; Fergusson, 2.

c8.s3.61

Of these pre- and early Cāḷukyan temples at Aihoḷe, some (Durgā, Lāḍ Khān, Hucchīmallīguḍi, and Meguti) have shrines detached from the back wall, as at Bhumara and Gop, providing for *pradakṣiṇā* with the roofed area; others (Kont Guḍi, etc. follow the mediaeval plan, in which the cella is connected with the back wall, so that *pradakṣiṇā* is only possible outside, in the open air. Where this plan is followed in a cave, of course, circumambulation is altogether precluded. Both in caves and structural temples the two plans appear side by side during several centuries; the older arrangement, for example, persists at Elephanta, but it is doubtful if any later instance could be cited.

c8.s3.62

The northern *śikhara*, as we have already seen in the case of several temples where it is an accessory rather than an essential, begins to appear in the late Gupta period. In more characteristic examples in the Ganges valley the *śikhara* and cella together form a tower, which may be provided with a porch, but forms the main part of the temple. These early towers are built up of elements similar in design to the cella itself, and with straight or nearly straight edges, and are thus nearer to the type of the Dieng Caṇḍi Bhīma (fig. 346) than to the fully developed curvilinear form under which the northern *śikhara* is most familiar. What may be regarded as a prototype of the early towers in which the reduplication of the main structure is still quite apparent, may be studied in a Kuṣāna railing pillar, J 24 in the Mathurā Museum (fig. 69)¹, and still better in the "Bodhgayā plaque" (fig. 62).

c8.s3.63

¹ Vogel, 13, pl. III.

c8.s3.64

The Bhītārgāon brick temple is a good example of the kind of tower referred to. The plan is square, with doubly recessed corners, double cornices, and a recessed frieze of carved brick. Above the double cornice rises the pyramidal roof with tiers of *caitya*-niches in horizontal courses. The walls are decorated with terracotta panels of Brāhmaṇical subjects. The general effect is not far removed from that of the early towers in Campā².

c8.s3.65

² Burgess, 8, pls. 303, 304; Cunningham, 4, vol. XI, pls. XIV—XVII; Vogel, 8.

c8.s3.66

Other brick and stone towers of similar character but more developed are found at Sona Tapan and Chinpur near Bānkurā, and several others at Manbhum and Dalmi, all in Bengal. In the case of the brick Kevaleśī shrine at Pūjārī Pālī, Bilāspur District, the tower is provided with angle-*āmalakas* on each storey³.

c8.s3.67

³ Some of these illustrated in Burgess, 8, pls. 288—290, 298, 300.

c8.s3.68

The well-known Gupta Daśāvatāra temple at Deogaṛh, near Lalitpur, dating about 600, is of stone, with plain walls, except that on three sides there are recessed sculptured panels, representing the Gajendramokṣa episode, Viṣṇu-anantaśayin, and a scene betwen two ascetics, and on the fourth a sculptured entrance with river-goddesses on the door jambs. The basement was decorated with fine panels representing *Rāmāyaṇa* scenes, an almost unique instance of an arrangement quite common in Java. The tower was of several stories, with *caitya*-arches and angle-*āmalakas*⁴.

c8.s3.69

⁴ Burgess, 8, pls. 248, 252; Mukherji.

c8.s3.70

The great Buddhist temple (fig. 210), known to archaeologists as the Mahābodhi, was most likely originally designated “Gandhakuṭī of the Vajrāsan”¹; as it now stands it is a restoration (1880—1881) of the Burmese restorations of 1105 and 1298, and still earlier mediaeval renovations and restorations. It consists of a high straight-edged pyramidal tower of nine storeys, with an angle *āmalaka* at each stage, surmounted by a *htī* with a fluted, bulbous, *āmalaka*-like lower member; this tower in its lower part, over the entrance, has tall narrow lancet opening, admitting light to the sanctum, and a part of the construction clearly shown in photographs taken before the last restoration consists of true arches. There is a porch on the east side, later than the main part of the shrine; and the whole stands on a single high *pradakṣiṇā* terrace. On the western edge of this terrace the *Bodhi*-tree was still growing until its decay and fall in 1876.

c8.s3.71

¹ The words Vajrāsana “adamantine seat” and Bodhi-maṇḍa “place of enlightenment” are both used by Hsüan Tsang to designate the seat occupied by the Bodhisattva on the occasion of the Great Enlightenment (Mahāsambodhi). The term Gandhakuṭī used to designate a Buddhist temple is derived from the name of a cell occupied by the Buddha in his lifetime. For descriptions of the temple, see Cunningham, 3, and Bloch, 2; the former seems the better account.

c8.s3.72

This temple was certainly standing when Hsüan Tsang visited Bodhgayā in the seventh century; he describes it rather minutely and gives its dimensions practically as they now are, quoting the height exactly and the width approximately². Fa Hsien states that there existed a temple at each of the four sites at which the Four Great Events of the Buddha’s life had taken place; proving that some temple existed here in the fifth century. Other considerations make it probable that the present temple, substantially in its present form, but of course without the later porch, was erected in the second century A. D., at any rate not later than the very beginning of the Gupta period. These reasons include (1) the presence of a coin of Huviṣka amongst the relics deposited at the foot of the interior Vajrāsan (2) a coping inscription in Kuṣāna or very early Gupta characters, referring to the “Great Gandhakuṭī temple (*pāsāda*) of the Vajrāsana” (3) a Kuṣāna inscription on the edge of the outer Vajrāsana placed against the back wall of the basement, on the ground level. Further, the “Bodhgayā plaque” (fig. 62) found at the Kumrāhār site, Patna (Pāṭaliputra), and the Kuṣāna relief reproduced in fig. 69 both show that temples of this kind might very well have been built as early as the second century A.D. In all probability then the new temple was built to enshrine a Buddha image, at the time when images were coming into general use; it was built, of course, as Hsüan Tsang expressly states that it was built, on the original site, following the usual rule in such cases. The building of a roofed temple, however, involved the removal of the Bodhi-druma to its modern position on the edge of the terrace at the back of the temple; there could have been no objection to this, so long as the Bodhi-maṇḍa was kept in its original place, where, indeed, Hsüan Tsang saw it. And in fact, Cunningham discovered behind the mediaeval grey sandstone Vajrāsana in the cella, another plastered throne, and behind this a polished sandstone slab resting on four pilasters exactly as represented in the Bhārhut relief (fig. 41) and undoubtedly of Aśokan age.

c8.s3.73

² Beal, 1, vol. II, p. 118.

c8.s3.74

The one other Vajrāsana referred to above as the “outer Vajrāsana”, found by Cunningham when the late mediaeval buttress of the back wall was removed, is large and beautifully decorated¹, on its upper surface with a simple geometrical design of circles and squares, on its sides with *haṁsas* and palmettes like those of some of the Aśokan capitals.

c8.s3.75

¹ Cunningham, 3, pl. XIII. It must be of early Śuṅga, if not Maurya date.

c8.s3.76

The famous centre of Buddhist learning at Nālandā, South Bihār, was founded by Narasiṁha Balāditya (467—473). Hsüan Tsang describes the great brick temple over three hundred feet in height, erected by this king, as resembling the tower at Bodhgayā, and says that it was exquisitely decorated and magnificently furnished. Nothing survives but the massive basement²; some of the niches on this basement representing fully developed curvilinear Nāgara *śikharas* may be later additions. Nothing at Nālandā, the most famous of mediaeval monasteries and centres of learning, antedates the fifth century, or postdates the twelfth.

c8.s3.77

² Burgess, 8, pls. 227, 228.

c8.s3.78

The temple at Gop in Kāṭhiāwāḍ (fig. 191) is more or less unique, but evidently connected in some way with the Kāśmīrī school of architecture. The square tower which is now its conspicuous feature was once surrounded by a flat-roofed hall providing for *pradakṣiṇā* under cover, concealing half its height; it is surmounted by a double pent-roof of the Kāśmīrī type, and decorated with large *caitya*-niches containing figures of deities. The basement of the outer structure, decorated with a *gaṇa* frieze, still remains. The shrine is Brāhmaṇical, and dates about the end of the sixth century, and is thus early mediaeval rather than Gupta properly so-called³.

c8.s3.79

³ Burgess, 2, and 8, pl. 266.

c8.s3.80

The hollow circular building at Rājagṛha, known as the Maṇiyār Maṭha is quite unique; traditionally known as a treasury, it is just possible that it represents a colossal *liṅgam* like those at Faṭehpur, near Bārāmūla, Kaśmīr, and Tiruparakuṉṟam near Madras. All that remains is the circular basement, with a small portion of the superstructure. All round the base are niches, separated by pilasters, and containing stucco images of fine and sensitive workmanship representing a *liṅgam*, Bāṇāsura, a six-armed dancing Śiva, and many Nāgas and Nāginīs (fig. 176). A date between 250 and 500 A. D. has been suggested, the fifth century seeming most likely⁴.

c8.s3.81

⁴ Marshall, 1.

c8.s3.82

We have already had occasion to refer to Nāgara and Drāviḍa *śikharas*. Both are towers rising above the *garbha-gṛha* of a temple, the chief difference being that the Nāgara type comes to have a curvilinear form and forms a real spire, while the Drāviḍa type retains its original terraced formation, with ranges of cells at each level. Much discussion has been devoted to the question of the origin of the Nāgara curvilinear spire, which has variously been derived from the *stūpa*, the simple domed cell, and the bamboo processional car¹. For the most part these theories represent deductions drawn from appearances presented by the fully developed form, not taking into account what may be called the primitives of the type. The original view propounded by Fergusson² I believe to be the correct one. This is that the Nāgara spire, however elaborately developed, really represents a piling up of many superimposed storeys or roofs, much compressed. The key to this origin is the *āmalaka*; properly the crowning element of a tower, its appearance at the angles of successive courses shows that each of these corresponds in nature to a roof. Thus the Nāgara and Drāviḍa towers both originate in the same way, but in the case of the former the storeys are so compressed and multiplied that at last the vertical effect completely dominates that of its horizontal components, while in the latter the storeyed principle and horizontal lines are never lost sight of. In the later northern towers, indeed, the suppression of the horizontal elements in many late examples is carried so far as to produce a smooth-surfaced pyramid with continuous outlines unbroken by any angle-*āmalaka*. It may be remarked that the northern tower developes convex curves, while in the southern *gopuras* the ultimate outlines are concave.

c8.s3.83

¹ The theories are summarised in Chanda, 2, with references.

c8.s3.84

² Fergusson, 2, vol. II, p. 119. This view is shared by M. Parmentier (6).

c8.s3.85

In both cases the aspiring aspect of the mediaeval towers contrasts most markedly with the static character of the early low flat-roofed temples. Just in the same way in Burma and Siam the *stūpa*, originally a hemispherical dome with one umbrella and clearly differentiated division of parts, develops into soaring types like those of the Shwe Dagon at Rangoon, with a continuous convex curve from base to pinnacle. The change from horizontal and domed to vertical and pointed forms is the most conspicuous tendency represented in Indian architecture, and must reflect an emotional qualification taking place in religious psychology not unlike that which distinguishes Gothic from Romanesque. A parallel tendency in India in narrative art has been traced by Foucher, contrasting the reserve of the earlier *Jātaka* scenes with the emotional emphasis already so marked at Ajaṇṭā³. The same development can be followed in the literature, and no doubt, if we knew enough about it, could be recognized in music and dancing.

c8.s3.86

³ Foucher, 4.

c8.s3.87

Indian palace architecture, with rare exceptions, mainly in Rājputāna of late date, has always been one of wooden construction, and for this reason no very ancient examples have survived. But palace architecture is very well illustrated in the sculptures of Amarāvatī and in the paintings of Ajaṇṭā, and from these it is evident that a palace consisted essentially of connected groups of one or two-storeyed pillared halls with flat or pointed roofs, the wooden pillars and capitals, cornices, &c., being elaborately decorated with painting and carving. It is in fact just this kind of palace architecture that survives in Burma (Mandalay), Siam (Bangkok), Cambodia (Phnom Peñ), Java (Yogyakarta and Surakarta) and in Japan(Kyoto); from these sources a very fair idea of the planning and appearance of much older Indian palaces can be gathered¹.

c8.s3.88

¹ For palaces in Burma, see Ko, 3; in Cambodia, Groslier, 3.

c8.s3.89

The classical palace was always provided with a picture gallery (*citra-śāla*) and a concert-hall or theatre (*saṁgītā-śāla*, or *nāṭya-maṇḍapa*)². The former, of course, was a hall specially decorated with frescoes, such as we find it described in the *Uttara-Rāma Carita*. The latter was an open pillared hall, with a stage raised somewhat above the level of the ground, and visible to the spectators from three sides, the "head of the stage" on the fourth side being a decorated partition shutting off the green-room. There was no curtain separating the stage from the audience, but two curtained doors led from the green-room to the stage, just as in a modern Chinese theatre, and it is with reference to these doors that we have the common stage direction "Enter with a toss of the curtain". Another constructional feature that survives in Far East is the low railing that runs round the edge of the stage platform. The outer walls were solidly built of brick, and "like a mountain cave" i. e. an excavated *vihāra*, without angles or projection, to the end that the voices of the performers and the low notes of the *kutapa* might be adequately heard.

c8.s3.90

² Bharata, *Nāṭya-śāstra*, Ch. 2 (ed. Grosset, Paris, 1898). Cf. *Mahāvaṁsa*, Ch. LXXIII, v. 82.

c8.s3.91

The general characteristics of Gupta sculpture have already been referred to. In the following paragraphs some of the more important examples are listed.

c8.s3.92

(1) Buddhas at Mathurā or of Mathurā origin include a magnificent standing Buddha from the Jamālpur (Jail) mound (fig. 158), A 5 in the Mathurā Museum, and a similar figure in the Indian Museum, Calcutta, both of the fifth century³; another with webbed fingers, from the Kaṭrā mound, B 10 in the Lucknow Museum, dated equivalent to 549/50 A. D.⁴; colossal reclining Buddha of the Parinirvāṇa shrine at Kasiā (Kuśinagara), with fifth century inscription mentioning the donor, the Abbot Haribala and the sculptor, Dinna of Mathurā⁵; seated Buddha with shaven head (the only Gupta example) and webbed fingers (fig. 162), from Māṅkuwār near Allāhābād dated equivalent to 448/49 A. D.¹; seated inscribed Buddha from Bodhgayā, dated equivalent to 383 A. D.²; seated Buddha at Sāñcī³. The two colossal Nāgas, in Mathurā stone, at Sāñcī may also be mentioned⁴.

c8.s3.93

³ Vogel, 13, p. 49, and pl. IX; Smith, 2, fig. 117; A. S. I., A. R., 1922—23, pl. XXXIX.

c8.s3.94

⁴ A. S. I., A. R., 1911—12, p. 132.

c8.s3.95

⁵ Vogel, 5; Cunningham, 4, vol. XVIII, p. 55, and vol. XXII, p. 16.

c8.s3.96

¹ Smith, 2, p. 173 and fig. 119; Bloch in J. A. S. B., LXVI, p. I, p. 283.

c8.s3.97

² Cunningham, 3, pl. XXV; A. S. I., A. R., 1922—23, pl. XXXVIII a, and p. 169.

c8.s3.98

³ Marshall, 4, 5 and 10 (pls. 1, 2).

c8.s3.99

⁴ Marshall, 10.

c8.s3.100

(2) Other Buddhist sculptures in stone include the wellknown seated Buddha from Sārnāth (fig. 161)⁵; other Sārnāth Buddhas and Bodhisattvas⁶; the Sārnāth lintel, with representations of Jambhala and *Jātaka* scenes⁷; Buddha figures in relief at Ajaṇṭā, cave XIX and Nāgarāja group at the same site⁸; Buddha figures of the façades at Kārlī, Kaṇheri, &c.⁹; Avalokiteśvara litany groups at Kaṇheri fig. 164), Ajaṇṭā Cave IV¹⁰, and Auraṅgābād; Bodhisattva torso from Sāñcī in the Victoria and Albert Museum, London, early Gupta or perhaps late Kuṣāna¹¹.

c8.s3.101

⁵ Sahni and Vogel, pl. X; Smith, 2, pl. XXXVIII; A. S. I., A. R., 1904—05, p. 81.

c8.s3.102

⁶ Sahni and Vogel, pl. XIII b, XX, etc.; Marshall and Konow; Vogel, 2; Hargreaves, 2, pl. LXIII.

c8.s3.103

⁷ Sahni and Vogel, pls. XXV—XXIX; Marshall and Konow.

c8.s3.104

⁸ Burgess, 8, pl. 200; Coomaraswamy, 7, pl. 72.

c8.s3.105

⁹ Burgess, 8, pls. 168, 212.

c8.s3.106

¹⁰ Burgess, 8, pl. 185.

c8.s3.107

¹¹ Cunningham, 4, vol. X, pl. XXI; India Society, 1; Smith, 2, p. 64 (misdated).

c8.s3.108

(3) Buddhist sculpture in metal: the most remarkable figure is the colossal (copper) image (fig. 160) from Sulṭāṅgañj, Bhagalpur District, Bengal, now in the Museum and Art Gallery, Birmingham, date ca. 400 A. D.¹². Other important examples include the richly decorated, copper and silver inlaid, brass figure (fig. 163) from Faṭehpur, Kāṅgṛā¹³; the Boston bronze Buddha, said to have been found in Burma (fig. 159)¹⁴; the rather clumsy statuettes from the Bāndā District, Bengal¹⁵; and the fragments from Bezwāḍā¹⁶; small gold Buddha in the British Museum¹⁷.

c8.s3.109

¹² Smith, 2, p. 171 and fig. 118; Rūpam, no. 21. A magnificent figure, seven and a half feet high and weighing over a ton.

c8.s3.110

¹³ Vogel, 4. The *ajourée* pedestal is closely related to one found at Suvarṇapura in Siam, J. S. S., vol. XIX, pl. XV.

c8.s3.111

¹⁴ Coomaraswamy, 9, 2, p. 61 and pl. XXI.

c8.s3.112

¹⁵ Smith and Hoey.

c8.s3.113

¹⁶ Sewell, R., 2. Some perhaps later: inscriptions of tenth (?) century.

c8.s3.114

¹⁷ Smith, 2, pl. LXXIV.

c8.s3.115

(4) Brāhmaṇical, &c.: colossal Varāha Avatār relief at Udayagiri, Bhopāl, about 400 A.D. (fig. 174)¹⁸; Paurāṇik and epic panels of the Gupta temple, Deogaṛh, early Gupta (fig. 167)¹; Umā-Maheśvara group from Kosām (Kauśāmbī) near Allāhābād, dated equivalent to 458/59 A. D., now in the Indian Museum, Calcutta²; Nativity of Mahāvīra or Kṛṣṇa from Paṭhārī, in the Museum at Gwāliar (fig. 178)³; slab with flying Gandharvas and Apsaras from Sondani, in the Gwāliar Museum (fig. 173); pillars from Chandimau with scenes from the Kīratārjuniya of the *Mahābhārata*, in the Lucknow Museum⁴; river goddess from Besnagar, in the Boston Museum (fig. 177)⁵; *toraṇa* pillars at Maṇḍor, Jodhpur State, with Kṛṣṇa Līlā scenes (fig. 166)⁶; pillars and architrave from Gaṛhwā, in the Lucknow Museum⁷; Narasiṁha from Besnagar, in the Gwāliar Museum (fig. 170); stucco reliefs of the Maṇiyār Maṭha, Rājagṛha (fig. 176)⁸; Kārttikeya belonging to the Bhārata Kalā Pariṣad, Benares (fig. 175)⁹; the Bhumara and Khoh *liṅgams* and Gaṇeśa¹⁰; sculptures of the Bādāmī caves and early temples at Aihoḷe¹¹; sculptures of the Rāmeśvara cave, Elūrā, especially the verandah pillars (fig. 190); small bronze of Brahmā (fig. 168) from Mīrpur Khās, in the Karāchi Museum; upper part of a bronze Śiva in the Victoria and Albert Museum, London; bronze-coated iron plummet from the River Surma, Bengal, in the British Museum (fig. 169)¹²: sacrificial pillars (*yūpa*) of Viṣṇuvardhana at Bijayagaṛh, 371 A. D.¹³.

c8.s3.116

¹⁸ Cunningham, 4, vol. x, pl. XVIII; Coomaraswamy, 7, pl. 99; Burgess, 8, pls. 216, 217.

c8.s3.117

¹ Burgess, 8, pls. 250—252; Smith, 2, pls. XXIV, XXV.

c8.s3.118

² Fleet, *Gupta inscriptions*, p. 256; Banerji, 4, pl. LXX b.

c8.s3.119

³ Smith, 2, pl. XXVI. Probably rather later in date.

c8.s3.120

⁴ Banerji, 2.

c8.s3.121

⁵ Smith, 2, fig. 112; Vogel, 18.

c8.s3.122

⁶ Bhandarkar, 2; Marshall and Sahni.

c8.s3.123

⁷ Smith, 2, figs. 114, 115; Burgess, 8, pls. 242, 243; Cunningham, 4, vol. X, pls. VI, VII. For another lintel with *Mahābhārata* scenes see Banerji, 1, pl. LIV. For *toraṇa* pillars like those of Gaṛhwā, found at Bilsaṛ, Cunningham, 4, vol. XI, pl. 13 and pls. VI, VII.

c8.s3.124

⁸ Smith, 2, fig. 113; Marshall, 1.

c8.s3.125

⁹ Rūpam, no. 21, 1925, p. 41.

c8.s3.126

¹⁰ Banerji, 3. Another fine *mukha-liṅgam* at Ataria Khera, Nagoḍh State.

c8.s3.127

¹¹ Burgess, 8, pls. 267—274; Cousens, 4 (the four massive roof slabs, of which three are from the temple in Field 270, Aihoḷe, reproduced in Cousens, 4, pl. LXXVI, and dating about 600 are in the Prince of Wales Museum, Bombay; the fine roof slab (fig. 165) of Cousens, 4, fig. 6, may be still *in situ*.

c8.s3.128

¹² Coomaraswamy, 4, fig. 100. In the same volume, p. 77, there is reproduced a rubbing of a handsome Gupta seal in copper; the text should be understood as “Seal of the Warden of the Frontier of Śrīvadra”.

c8.s3.129

¹³ A. S. I., A. R., 1902—03, p. 207.

c8.s3.130

A colossal Hanuman from Pārkham, D 27 in the Mathurā Museum appears from the style and fine modelling of the torso to be of Gupta age¹⁴. Four colossal images and groups at Rūp Bās, Bharatpur State include an image of Baladeva with cobra hoods, over twenty seven feet in height, his wife Ṭhākur Rānī, a Nārāyaṇa with Lakṣmī over nine feet in height, and a group supposed to represent Nārāyaṇa standing on the head of Yudhṣṭhira, who is surrounded by the Five Pāṇḍavas. No information is available as to the style or date of these evidently important sculptures¹.

c8.s3.131

¹⁴ Vogel, 13, p. 100. An earlier example is illustrated in A. S. I., A. R., 1923—24, pl. XXXV, K.

c8.s3.132

¹ Cunningham, 4, vol. VI, p. 21, gives only a list of the images and their dimensions.

c8.s3.133

(5) Terracottas, mostly Brāhmaṇical: panels of Brāhmaṇical subjects, decorating the brick temple at Bhītārgāon²; *Rāmāyaṇa* subjects, Saheṭh-Maheṭh (Śrāvastī³); large image of Hārītī, and Buddha figures, at Kasiā (Kuśinagara)⁴; Mīrpur Khās, Buddhas and donor⁵; seals and small terracottas from Basāṛh (Vaiśālī)⁶; seals and small terracottas from Bhīṭā⁷; figures from Kurukṣetra, Delhi⁸; Bikanīr (more likely late Kuṣāna)⁹; carved and moulded bricks at Bilsaṛ¹⁰.

c8.s3.134

² Vogel, 8; Cunningham, 4, vol. XI, pls. XIV—XVII.

c8.s3.135

³ Vogel, 7.

c8.s3.136

⁴ Sastri, 1; Vogel, 5. In the Lucknow Museum.

c8.s3.137

⁵ Cousens, 5.

c8.s3.138

⁶ Bloch, 1.

c8.s3.139

⁷ Marshall, 3.

c8.s3.140

⁸ Cunningham, 4, vol. XIV, pl. XXVII.

c8.s3.141

⁹ A. S. I., A. R., 1917—18, pt. 1, pl. XIII; and see page 69.

c8.s3.142

¹⁰ Cunningham, 4, vol. XI.

c8.s3.143

Indian literature of all kinds and at all periods, at any rate after the Maurya, makes incidental references to painting. It may be taken for granted that from a very early period, not only were sculptures and architectural details covered with thin plaster and coloured, but that the flat walls of temples and palaces were decorated within and without with pictures or with painted “wreaths and creepers”. In the Epics we often hear of painted halls or chambers (*citra-śāla*) in palaces. A whole scene of Bhavabhūti's *Uttara-Rāma-Carita*, dating from the close of the Gupta period, is laid in such a gallery, where Rāma and Sītā are represented as viewing newly executed paintings of scenes from their own life, which awaken in Sītā a longing to revisit the forests, creating in her a “latent impression” (*bhāvana*)¹¹. The *Viṣṇudharmottaram* distinguishes the kinds of painting appropriate to temples, palaces and private houses; and applies the theory of *rasa* to painting. Paintings are there classified as *satya*, *vaiṇika*, *nāgara* and *miśra*, which I am inclined to render as true, lyrical, secular and mixed, mainly with reference to their themes¹². The same text devotes considerable space to the question of foreshortening as applied to the features and limbs; and lays great stress on adherence to canonical proportions. The necessity of giving expression to the movement of life (*cetanā*) is emphasized; he understands painting who can represent the dead without life movement, the sleeping possessed of it. Finally it is said, with good reason inasmuch as both are occupied with the exact expression of emotion, that without a knowledge of dancing (*nṛtya-śāstra*) it is hardly possible to understand the true skill of painting¹.

c8.s3.144

¹¹ Belvalkar, p. 26 (= Act 1, v. 39). For painting representing events in the life of a still living king, cf. *Mahāvaṁsa*, Ch. LXXIII, v. 83.

c8.s3.145

¹² *Satya* seems to mean here “true to life, realistic“, perhaps with reference to portraiture. *Vaiṇika* suggests pictures of musical modes (cf. p. 129, note 1). *Nāgara* perhaps = erotic; *nāgarika* (see p. 88) might be translated “man about town“.

c8.s3.146

¹ *Vinatu nṛtyaśāstreṇa citrasūtram sudurvidam*, *Viṣṇudharmottaram*, III, II, 3. The *Viṣṇudharmottaram* (see translation, Kramrisch) is a mediaeval composition apparently embodying older, and probably Gupta materials. A later mediaeval text by Śrī Kumāra, the *Śilparatna*, Ch. 64, deals in a similar fashion with painting; translated by Coomaraswamy, 13. Keyserling has remarked of Indian dancing and religious images the "identity of the spirit in both appearances".

c8.s3.147

Painting appears in all lists of the sixty-four *kalās*, the fine arts or accomplishments². Portrait painting, usually from memory, and on wooden panels, is a device constantly employed in classical Sanskrit plays³. The *Kāmasūtra* of Vātsyāyana, a work essentially of the Gupta period, mentions the drawing panel, paints and brushes as parts of the ordinary furniture of a gentleman's (*nāgarika*) chamber and taken in its context this throws some light on the meaning of the term *nāgara* as used to define a kind of painting. It is quite evident that, in the Gupta period at least, painting was not exclusively an ecclesiastical, but also a secular art, practised by amateurs as well as by professional members of gilds; it was a social acomplishment, at least among princes and ladies of the court, and in the "fast set"⁴.

First 147 of 171 passages. Read on

Cites (79)