Pramāṇa · Search · About · API

The Hindu Temple v1.c8.s7: (B) The Superposition of Shapes Along the Vertical Axis

The Hindu Temple v1.c8.s7

Language
English

(B) The Superposition of Shapes Along the Vertical Axis

v1.c8.s7.1

The images which are given concrete form in the building of the temple have been seen and worded in revelation (śruti) and the sacred tradition (smṛti).

v1.c8.s7.2

The Hindu temple is a synthesis of many symbols. By their superposition, repetition, proliferation and amalgamation, its total meaning is formed ever anew. In the vertical direction, the superposition of forms leads towards the culminating point. The solid socle or base (Adhiṣṭhāna, etc.) functions as the altar on which the offering is made in the shape of the temple. Its main part is the walled in, dark space of the Garbhagṛha. From the vertical walls of the Garbhagṛha, the original dolmen shape, rises the superstructure and above it, the finial. The high superstructure is not derived from any particular roof shape; these, in certain types are embodied in its form. It is not a roof with increased height; it is a form of sacred architecture, complete in itself and is placed above the walls of the Garbhagṛha⁸⁹.

v1.c8.s7.3

⁸⁹ Generally, the pyramid of the South Indian temples is super-added to its straight walls. The survival of the dolmen type in flat roofed temples of the mediaeval period—such as those near Jhansi (central India), at Candpur, Dudahi and Ladhaura (ASI, U. P. Photographs, 1937-38, Nos. 6785, 6758, 6763) has been pointed out already. The Megalithic nature of these flat roofed temples is shown by the roof where it consists of one monolithic slab. In certain flat roofed temples, moreover, the front is composed of large, vertical stone slabs (Siva temple in Kuttikondabilam, Guntur; ASI, Madras Photographs 1936-7, No. 328).

v1.c8.s7.4

The height of the storeys (bhūmi) of the pyramidal superstructure diminishes, on certain temples, in an arithmetical progression; each successive storey is ¼ or also ⅕ less than the lower. In this progression however is not included the ground floor (samsthāna) of the temple. In order to moderate the abrupt superposition of the pyramid with its miniature storeys on the relatively high wall of the ‘cubical’ Garbhagṛha, this wall frequently appears divided externally in two storeys, each complete with its base, pillars, capitals and roof cornice. This architectural consideration belongs to the Cola age (Temples at Tanjore, 1000 A.D.; Gangaikonda Colapuram about 1025 A.D.); and while later temples show an increase in the number of simulated storeys on the walls of the Garbhagṛha (Temple at Tiruvarur, about 1600 A.D.), Pallava temples are free from this aesthetic deceit (Shore temple at Mamallapuram, Kailāsanātha temple at Kāñcipuram (650-700 A.D. approximately). See Plates LI-LIII; LVII-LIX in P. Brown, ‘Indian Architecture’ and Part VI.

v1.c8.s7.5

The superposition of the pyramidal, storeyed form on the vertical walls of the Garbhagṛha, though generally is not necessarily always observed. The pyramidal structure of the Vaikuṇṭha Perumal Temple (ib. Pl. LIV) rises directly from its socle. Here too, the storeys are not simulated, miniature replicas; they house a Garbhagṛha on each floor. In the majority of the temples in South India, including the Kanarese country, a complete structural pyramid of this type, appears raised on the Samsthāna, the groundfloor or one storeyed temple. This development appears already completed in the seventh century rock cut temple models which an ingenious king (Narasimhavarman) was pleased to have cut out of the rocky boulders near the shore of Mamallapuram.

v1.c8.s7.6

The diminution of the height of the Bhūmis of the superstructure of a South Indian temple is carried out according to more than one consideration, such as the number of these ‘storeys’, etc. (see Part VII, Chap. 5).

v1.c8.s7.7

In Northern India, the diminution of the Bhūmis of a curvilinear superstructure such as the Sikhara of an Orissan temple does not form a series; if, for example, there are ten Bhūmis, of Above the superstructure and no longer part of the body of the temple, is the finial or Stūpikā; it is however proportionately related to the body. All these shapes are piled one upon the next, along the vertical axis. It connects the central point of the floor of the Garbhagṛha with the high point of the finial.

v1.c8.s7.8

On this vertical axis are threaded the levels of the building, its floors (bhūmi) and profiles, their projections and recesses. Expansion proceeds from the central point of the Garbhagṛha, in the horizontal, in all the directions of space; this spread with its proliferation and particularisation is gathered up towards the apex; the broad mass with its many forms is reduced to a point, beyond its total form.

v1.c8.s7.9

The piled up altar, the base, the dolmen-cell, the Garbhagṛha, and the superstructure are the architectural constituents of the whole image of the temple; it rises like a mountain. Its mass diminishes while it is drawn along the vertical to a high point, straight above the centre in the dark small space of the interior. The image of the mountain and the cave is known in nature, and is given form by the architect. This total aspect of the temple is not in continuation of an extant architectural type like its parts, the altar, the dolmen, the Tabernacle and the other constituents of its superstructure. It preserves however as they do, the memory of a cult, the cult of caverns; and this corresponds moreover to the immediate realisation of the cavity of the heart. These are immediate symbols in nature and in man; to this day the Himālayas are full of natural caves, small or large, and sacred.

v1.c8.s7.10

The cave temples are an elaboration of these primeval sites. No architectural forms however were evolved there; on the contrary, forms of structural buildings were adjusted to the exigencies of the rock, pillars for instance increasing in

v1.c8.s7.11

which the lowermost has 5 units, the height of the following is: 4 7/16 ; 4 ; 3 15/16 ; 3¾ ; 3⅝ ; 3 9/16 ; 3 7/16 ; 3 5/12 and 2. ('Canons of Orissan Architecture', op. cit. p. 111). No diminution however in a geometrical progression can be seen, as stated by M. M. Ganguli, 'Orissa and Her Remains', p. 128, on the accompanying Plate II.

v1.c8.s7.12

Apart from the main types of the superstructure and their several components (see Part VI) efforts are documented in the centuries around the beginning of the present era, of superimposing and combining buildings along the vertical axis.

v1.c8.s7.13

An ancient type of temple consisting of two super-added buildings is shown on the reverse of Audumbara Copper Coins (Fig. 16, Pl. I of J. N. Banerjea, 'The Development of Hindu Iconography'). The lower part of the structure has a sloped roof, above it is placed a smaller structure which also has a sloped roof. This type of a Śiva temple, of the second to first centuries B.C. in the Panjab, in the superposition of structures, resembled Bengal temple types (Gauriya) of the present day, in Bengal and Orissa. Their roofs are sloped and curvilinear. Buildings of these and similar types preceded, and contributed to, the formation of the superstructure of the temple (Part VI).

v1.c8.s7.14

Another ancient form of the temple is represented in Barhut, about 100 B.C. and in Mathurā in the second century A.D. (Coomaraswamy, 'HIIA', Figs. 42 and 70). In these reliefs a small, domed shrine is encased in another structure which is roofed by a series of superposed 'slabs' of increasing width, the topmost forming a flat and crenellated roof. This method of superposition and encasement though it is neither represented by any of the later temples nor mentioned in Vāstu-śāstra, corresponds to the vision of the Throne of Supreme Blessedness on the Brahmānanda mountain (Shri Svami Hariharanand Sarasvati, 'Viṣṇu, The All-Pervading Principle', JISOA, XII. p. 154).

v1.c8.s7.15

Re the various combinations contemporary with the main types of the superstructures, which were not destined to become leading types, see, for example, p. 169, note 94.

v1.c8.s7.16

width so as to support the weight of the hill which in part rested on them. Layana, place of rest, is the name for rock cut temples. They have no Śreṇi, which means no superstructure with its cluster groups of similar shapes; they are without buttresses (niryūhaka), while a circumambulatory (bhrama) and windows (gavākṣa) should be carved in the rock ('Samarāṅgaṇasūtradhāra', LIX. 236-237)⁹⁰ in imitation of structural temples.

v1.c8.s7.17

⁹⁰ The Layana is equipped moreover with stairs, a gateway (pratolī), roll cornice (vitaṅka= kapota-pālikā) on the façade, and doors. It is raised on a socle (vedi) and has a portico (prāgriva). This description refers only to cave temples such as those in Bādami. The cave temples, in the earlier examples (3rd century B.C.—6th century A.D.) are interiors only, having a façade; to these types were added (Mamallapuram, Elura, Kalugumalai), complete replicas of structural temples hewn out of the rock in their exterior, and excavated within (Kailāsanātha Temple and Indrasabhā in Elura). Although the last named temples, are the one Hindu, the other Jain, the majority of the rock cut temples and sacred abodes are Buddhist. Out of a total of 1200 rock cut temples 900 are Buddhist, 200 are Jain and 100 are Hindu. Some of the sanctuaries in Elura (Daśavatāra, etc.) and the Śiva temple in Elephanta are, though posterior to the sixth century, interior excavations only with a façade. Re. these, and later excavations in Northern India—at Dhamnar (Rajputana) and Masrur (Kangra) see P. Brown, 'Indian Architecture', (Buddhist and Hindu), Chapters V, VI, XII and XI.

v1.c8.s7.18

Symbols such as the vertical axis or pillar along which the varied forms are threaded on different levels or the cave in the mountain, and architectural forms such as the convergence of ascending lines which connect the perimeter of the building with the end of its vertical axis, or the various shapes of the superstructure, these and other images and forms constitute the symbolical and concrete structure of the temple. The temple under the name of mountain resembling it by its peaked form, is always the One Mountain, an image of manifestation in its hierarchy along the central axis of being. This axis passes through all the strata of existence, and shows them linked to the highest point, at different levels⁹¹. From the highest point the line passes in the centre and pierces the ground in the middle of the Garbhagṛha where the Liṅga or image is. From the perimeter of the Prāsāda towards its highest point rises the bulk of the building, a vesture of the central axis; in its folds and throughout its extent, it is an exposition of the total meaning of the temple in the particular application to each single spot.

v1.c8.s7.19

⁹¹ René Guénon, in 'Le Symbolisme de la Croix' and other works has made clear the meaning of these and other symbols; A. K. Coomaraswamy has explained them in their application in art. P. Mus has interpreted the Barabuḍur in the aspect of these perennial symbols.

v1.c8.s7.20

The names of the three first temple types recorded in the early texts are those of the Mountain; Meru, Mandara, Kailāsa; another type among the twenty temples

v1.c8.s7.21

The square, dark, small Garbhagṛha is not transferred from the cave temple to the structural temple. The flat roofed 'Gupta' temple is not derived from Brāhmaṇical excavated sanctuaries contemporary with it (Udayagiri in Bhopal, etc.), nor from earlier excavated cells with a flat ceiling; the early rock cut sanctuaries have domed or vaulted interiors, whereas cells and halls in the rock cut monasteries have a straight ceiling. There is no scope for a flat roof in rock cut temples, the Caitya halls prior to the Gupta age; it belongs to the flat roofed porch only of the sanctuary proper. The flat roof of the rock cut Kailāsanātha Temple in Elura of the eighth century is in imitation of a structural temple of that age.

v1.c8.s7.22

Any shape can be cut into the rock; no structural form is born there.

v1.c8.s7.23

In the 'terrace temples' at Ahicchatra, etc., the axis or shaft cuts across the terraces; the shrine, in continuation of the shaft, rises from the highest terrace.

v1.c8.s7.24

described in the ‘Bṛhat Saṃhitā’ is called Guharāja, King of caves (LV. 17)⁹²; it is equally telling by its name as the four last named types, in chapter LV of the ‘Bṛhat Saṃhitā’, the Round, the Square, the Octagonal and the Sixteen sided one are by their form, which should be dark inside, "so that light from outside will not enter these Prāsādas" (‘Bṛ. Saṃh.’ LV. 25; 28, with comm.).

v1.c8.s7.25

Guharāja, King of caves, is a name as suggestive as it is unique among the ever-increasing types of temples enumerated and described in the texts⁹³. The name however occurs also as that of actual temples, such as the ‘Kuraja (Guharāja) Bir’ Temple⁹⁴. Kuhara, or cave, is a synonym of Śālā, or room, in the ‘Bhaviṣya Purāṇa’, where the type of temple, called Meru, is described as having many Kuharas (ch. CXXX. 27)⁹⁵.

v1.c8.s7.26

The language of the texts connects the mountain and the cave while describing works of architecture, or forms of nature and also the residences of different classes of gods, Devas and Dānavas, Pannāgas, Yakṣas, Rākṣasas, Guhyas, Gandharvas, Vidyādharas, Siddhas, Kinnaras and Apsarās who live in Indra’s grove on mount Sitānta full of rock and cave-houses⁹⁶. The ‘Vāyu Purāṇa’ tells of the various kinds of residences of the gods on the different mountains⁹⁷. It is a topography of the mountains where the gods reside and of their habitations on an Olympus with many peaks⁹⁸; it is not a description of temples built by man on mountain tops for the gods to dwell in.

v1.c8.s7.27

The caves are ancient residences of the gods.. It is there too, and not only on the banks of rivers that they love to dwell; their presence there is felt so strongly that cave and god are one; "on the Viśākha mountain there is a great dwelling belonging to Guha, the Secret one (Kārttikeya), the god who is very fond of living in caves" (guhā; ‘Vāyu Purāṇa’, XXXIX. 55). To these natural habitations of

v1.c8.s7.28

⁹² ‘Bhaviṣya Purāṇa’, CXXX. 32 which derives from the Br. S., or from a source common to both, substitutes Gṛharāja for Guharāja, see Part VII, 2nd chart (cf. also S.S. Ch. XLIX. 16; 160). ⁹³ The ‘Samarāṅgaṇasūtradhāra’, LIX. 193-197, describes the temple called Guhādhāra ; its name however is derived from the division of the door-frame (dvārabheda) into several compartments (guhā) as it is also described elsewhere. It should resemble the temple Simha. This possibly includes a half forgotten identification of the temple type called Siṃhāsya (‘Matsyapurāṇa’, CCLXIX. 28) and the Guharāja of the ‘Bṛhat Saṃhitā’ (see Part VII, second chart). "Viśvakarman has described 3000 types of temples" says the ‘Bhaviṣya Purāṇa’ CXXX. 36. ⁹⁴ ASIAR, 1915-16, Pt. II, p. 17. It is situated between Deogarh and Candpur in central India. This temple, of about the 9th century, similar in this respect to the Pārvatī temple at Nachna Kuthara has on the flat roof of its Garbhagṛha, another though smaller cubical sanctuary—and on top of this a curvilinear Śikhara. In principle, the two temples types, the one with several storeys and the other with a curvilinear Śikhara are placed here one above the other. ⁹⁵ The Ananda Pagoda in Pagan, Burma, being a ‘Ku’ or cave has such Kuharas, ‘caves’ or halls, in the 4 directions, radiating from a massive centre. The Burmese name for structural brick temples is ‘Ku’; cf. note 104. ⁹⁶ ‘Vāyu Purāṇa’, Ch. XXXIX. 55-57 (śaila-gṛha ; guhā-gṛha). ⁹⁷ Ib., śl. 1. This does not mean, as P. K. Acharya, ‘Indian Architecture’, p. 21, opines that "the ‘Vāyu Purāṇa’ maintains its unique position by dealing with the construction of various temples built on mountain tops". ⁹⁸ Ib., śl. 57 ; the large residence (bhavanam) of Kuvera and also ‘Harmya-prāsādas’ are described on mount Piśācaka.

v1.c8.s7.29

the gods have to be added, as places of worship, retreat and congregation, other natural caves and also those cut into the rock for similar purposes. The Ājīvikas, a Jain sect, and the Buddhists were the first to do so; they were non-orthodox. Within Brāhmaṇism, the substitution of excavated caves for natural ones took time to evolve. There, as elsewhere, the sacredness of the particular site was to begin with, sufficient in itself; the Tīrtha, in this case was specially marked by its being a cave. The heterodox sects preceded the Hindus by many centuries in their interference with, and transformation of nature. They had already achieved magnificent results in such large, apsidal vaulted ‘churches’ as the cave temple at Karli and in rock cut monasteries like those at Nasik or Ajaṇṭā (Nos. 8, 12, 13), when about 400 A.D. Brāhmaṇical worship cut its entry into the rock. It has the shape of a small, flat roofed Garbhagṛha. To it is added, as the contemporary Gupta temples, a structural porch or maṇḍapam (Udayagiri, Bhopal, C. I.).⁹⁹

v1.c8.s7.30

⁹⁹ This earliest Brāhmaṇical rock-cut temple is amongst the earliest fully preserved Brāhmaṇical shrines. It dates from the reign of Candragupta II, 382-401 A.D. (Cunningham, ASR, Vol. X, p. 41). The dated Ājīvika caves in the Barabar Hills, Bihar, were excavated in the reign of Aśoka, in the 3rd century B.C.

v1.c8.s7.31

No apsidal temple was cut into the rock for Viṣṇu, Śiva or any Brāhmaṇical form of divinity¹⁰⁰. The rock cut Hindu Garbhagṛha is an equivalent of the structural flat roofed stone temple¹⁰¹; yet it is also preceded by the small, single rock cut cell and also by those which surround in numbers Buddhist monastic halls.¹⁰² It is however significant that a relatively large cell, at the centre of the far end of the hall and serving as the main sanctuary, appears in the Buddhist excavations¹⁰³ only at an age when the flat roofed Garbhagṛha had been set up in stone Prāsādas and cut into the rock.

v1.c8.s7.32

¹⁰⁰ The apsidal plan was however adjusted to the use of Brāhmaṇical worship in structural temples, such as the Kapoteśvara temple at Chezarla, the Durga temple at Aihole, etc., and according to the ‘Samarāṅgaṇasūtradhāra’ (Ch. XLIX) must have been widely used. Cf. also SS. XLIX. 103-4; LII. 17, the Hastijātīya’ type based on the square plan.

v1.c8.s7.33

¹⁰¹ In the Lad Khan, Kont-gudi and 2 other small temples at Aihole, the Garbhagṛha on the other hand is built against the back wall of the pillared hall (ASIAR, 1907, p. 201 f.) ; this position would correspond to a Garbhagṛha cut deep into the rock and preceded by its maṇḍapa.

v1.c8.s7.34

¹⁰² Garbha, Pāli ‘gabbha’, is also the name of rock-cut cells of the Buddhist monks connected with their large monastic halls. Such an establishment is called a ‘navagabha Maṇḍapa’ (Karli cave inscr., ‘Ep. Ind.’, XI, p. 119), if there are nine ‘gabha’ or cells ; or also ‘paca-gabha maṇḍapa’ and ‘sattagabha maṇḍapa’, when their number is only 5 or 7 (Junnar cave inscriptions, pp. 131, 136). Any small room is finally called ‘gabbha’, in Pāli texts and denotes various kinds of chambers or rooms (‘Cullavagga’, VI. 3. 3.) which may be square or rectangular, etc. The use of terms such as ‘pāsāda’, or its equivalent ‘vimāna’; gabbha, and also ‘guhā’, cave (‘Cullavagga’, VI. 1. 2) in civic architecture is also current in the Epics. Dhātugarbha (dagaba) is the stūpa as receptacle or womb of the relics (dhātu) of the Buddha. The Buddhist rock-cut cells again had their equivalent in structural cells; these in stone buildings, such as are preserved in Gandhāra were also used as shrines for a Buddhist image or a stūpa (Takht-i-Bahai). In this particular instance however they are not square, but rectangular, etc., and they have not flat roofs.

v1.c8.s7.35

¹⁰³ In Vihāra No. 5 for instance in Ajaṇṭā, or No. 3 at Aurangabad.

v1.c8.s7.36

In the temple at Udayagiri, for the first time the walled in stone quadrangle as it were entered the mountain; one stone-form, that of the dolmen, was put back into the primeval stone, the living rock. There the Garbhagṛha retained its flat roof by adapting a natural ledge of rock.

v1.c8.s7.37

In the quest for secrecy, the enclosure of the shed in Vedic rites, or also of the dolmen for purposes of the Hindu temple, was one way of attaining it. Another way less widely and only comparatively later resorted to in architectural form by the Hindus, led to cutting into the interior of the mountain, the living rock. The final solution is the Garbhagṛha within the Prāsāda with its superstructure like a mountain.

v1.c8.s7.38

The type of temple called Guharāja, has the shape of a cave (guhā) according to Utpala (Comm., Bṛ. S. LV. 25). Its height, 32 cubits, follows the general rule, of being twice the width of the Prāsāda, and implies a superstructure as high as the walls of the Garbhagṛha (Part VII. First chart). This King of caves, Guharāja, whatever its actual shape was, shares part of its name with Burmese brick built temples. In Burma, brick built temples with inner spaces are simply called 'Ku' or cave.¹⁰⁴ One of the temples at Pagan bears the name Shwe Ku, Golden cave. The Burmese Glass Palace chronicle tells about the erection of the Ānanda temple of Pagan, how King Kyanzittha requested eight Arhats to produce by their concentrated thought an image of the cave Nandamūla in the Gandhamādana mountain. This they did and the King built a large Ku=Guhā, a 'cave', or temple in the likeness of the cave Nandamūla and called it Nanda.¹⁰⁵ The name of the cave, which properly is the Garbhagṛha, appears here as that of the whole Prāsāda.

v1.c8.s7.39

¹⁰⁴ R. Heine-Geldern, 'Weltbild und Bauform in Südostasien', 'Wiener Beiträge zur Kunst und Kultur Asiens', Vol. IV, p. 63. The central space in the Ānanda temple, Pagan, is a brick mass; the 'caves' are in the four directions.

v1.c8.s7.40

¹⁰⁵ Ib., p. 15. The Gandhamādana mountain, acc. to the 'Viṣṇupurāṇa' II. Ch. II, 17, lies to the south of Meru. Cf. note 78.

v1.c8.s7.41

Cave and Mountain, in the architecture of Greater India are names for the total temple, Ku (Guhā) in Burma, Giri (mountain) in Cambodia and Meru, in Bali.¹⁰⁶ In India itself, and originally, they denote the interior respectively¹⁰⁷ and the high exterior shape of the Prāsāda. The interior with its cave darkness corresponds, to the deity known 'ab intra'. The exterior with its mountain slopes along the superstructure and the perpendicular walls of the Prāsāda displays to the light of day, the seed which has taken root, and sprouted.

v1.c8.s7.42

¹⁰⁶ L. Finot, 'Sur quelques traditions Indo-Chinoises', 'Bull. de la Commission Archéologique de L'Indochine', 1911, p. 20; J. C. Von Eerde, 'Hindu-Javaansche en Balische Eer edienst', 'Tijdschrift voor Indische Taal, Land en Volkenkunde', LXV. 15-16.

v1.c8.s7.43

¹⁰⁷ The 'water in the cave' is in the Garbhagṛha the water with which Liṅga or image are laved in the daily rites. It passes from the image to a drain (praṇāla) on the floor which traverses the middle of the north wall of the Garbhagṛha, and leaves through a spout carved in the likeness of a Makara, etc. The water in which the Liṅga or image has been bathed is sanctified and therefore is made to flow to the north. The Ganges too, is most sacred where its course turns northwards. The northern direction implies an upward course, back towards the origin—high up in the mountains and higher still, in the celestial region.

v1.c8.s7.44

Yet another secret place which also became integrated into the temple, is the place of the Omphalos, in the womb of the earth and below its surface. The cave, under-ground, the crypt,¹⁰⁸ is the main Garbhagṛha of several preserved temples.

v1.c8.s7.45

¹⁰⁸ Guhā (cave) and 'gupta', secret, both from the same root, appear also as verbal equivalents of the 'crypt'.

v1.c8.s7.46

Tamas, darkness, is the descending tendency, it is the quality proper of the underground crypt. Above it, the Prāsāda arises, ascends in height according to the Sattva-guṇa, and expands its perimeter as far as Rajas, requires it.¹⁰⁹ Tamas, darkness, is the causal body, the ‘kāraṇa rūpa’. As it was in the beginning when out of primordial darkness evolved all things that be, so also from the deep, central darkness of the Garbhagṛha the meaning of the temple shines forth on its walls and reaches the high point of the finial. Thus in certain temples there are two Garbhagṛhas; above the crypt-Garbhagṛha is the upper sanctuary, accessible or visible to all. The secret chamber of the Sūrya temple at Modhera, Gujerat, built in 1026-1027 A.D., is sunk to eleven and a half feet below the level, and is underneath the floor of the Garbhagṛha of the temple.¹¹⁰ At Aundh, the principal Liṅga is in the crypt below; in the upper Garbhagṛha is another Liṅga; steps lead down into the crypt from an opening in the floor of the upper shrine.¹¹¹ The present-day temple of Somanātha, Pattan, Kathiawar, also has a lower shrine. It surrounds the Somanātha Liṅga, symbol of the self existent Omphalos. A ‘duplicate’ for every day worship is in the upper shrine.¹¹² In the Jambukeśvara temple, near Trichinopoly, in South India, the Garbhagṛha below the level of the ground enshrines a Svayambhū Liṅga standing in water. The great sanctity of the non-man-made Liṅga, the hidden darkness of the not only innermost, but also of the lowermost, Garbhagṛha, are proper to the Guhā, the secret chamber, around the omphalos, the navel and centre of the Earth and of Being; of this the Ādhāraśilā is one symbol and its place is the same.¹¹³ Above it is the Garbhagṛha for the daily rites, and above these graded levels of secrecy and sanctity is the superstructure. The finial above it shines golden, high up, straight above the omphalos, or centre of the Garbhagṛha, the womb and cave in the mountain. Or else no floor separates the lower and the upper chamber, they are one; only the sunk level is preserved. The one and only Garbhagṛha is often much lower in level than the hall, the Maṇḍapam by which it is approached; stairs lead down to it, to a depth of seven or eight feet, or less;¹¹⁴ the actual extent into depth finally is immaterial as long as the descent is marked by the level of the floor being lower than that of the threshold, be it even by one step only. In Orissa, the name for the Garbhagṛha is Gambhirā, the deep lying.

v1.c8.s7.47

¹⁰⁹ The three Guṇas, Sattva, Rajas and Tamas, active in every form of manifestation, have in the form of the temple a comprehensive visual symbol.

v1.c8.s7.48

¹¹⁰ J. Burgess—H. Cousens, ‘Architectural Antiquities of N. Gujerat’, ASWI, Vol. IX, p. 73.

v1.c8.s7.49

¹¹¹ ib., p. 75. About the temple at Aundha, Hyderabad, Deccan, where the floor of the Garbhagṛha is considerably sunk below that of the maṇḍapam and the Liṅga is not seen from the “hall doorway”, see H. Cousens, ‘Mediaeval Temples of the Dakhan’, ASI. IS. Vol. XLVIII, p. 78.

v1.c8.s7.50

¹¹² Cousens, ‘Somanātha’, etc. ASI. IS. XLV. p. 28.

v1.c8.s7.51

¹¹³ The underground situation of the place of greatest sanctity has been explained as a protective measure from the Mohammedans, as all the temples where it is so placed, were built when the Muslims had entered India. Apart from the fact, that the vast majority of preserved Hindu temples dates from these centuries, (10th to 12th), the presence of the Svayambhū Liṅga, the natural Omphalos, disproves the assumption.

v1.c8.s7.52

¹¹⁴ In the temple of Amaranātha (Ambarnath), Thana District, Bombay, 1060 A.D., the floor of the shrine is sunk below the outside ground level and about 7' or 8' lower than that of the hall. Stairs descend to it. The possibility of an original shrine on the same level as that of the hall (Cousens, op. cit., p. 13) does not detract from the fact of the sunk level, as it now is, of the Garbhagṛha. Other temples in the Deccan, in Gujerat, Rajputana, the Central Provinces and Orissa are sufficiently widely distributed to be valid examples of a practised form of worship and architecture. In the Deccan, the Nāgeśvara temple at Karjah (Ahmednagar ; l.c., p. 58) has a shrine of which the floor is 6' below the floor of the hall approached by a flight

v1.c8.s7.53

The different levels below ground and above producing two sanctuaries are destined for various rites; they are not restricted to a definite position. The dividing line may be the ground surface; it need not however be only there but is applied in the vertical direction so that one sanctuary is above the other. Made independent of a definite level, the principle is that of vertically superposed sanctuaries. This is known from Gupta and Early Cālukyan temples, in northern India, the Deccan and in South India. It is as if the sanctuary from below the ground with its omphalos in the shape of Liṅga or image had arisen to the higher levels. Sanctuary upon sanctuary, they are superposed in several storeys, particularly and consistent with the total symbolism of the respective temples, in South India. There, the special application of this principle is to those temples of Viṣṇu where in seven superposed storeys, the lowermost cell enshrines the standing (sthāna), the next higher one the seated (āsana), the one on the third floor the recumbent (śayana) image of Viṣṇu, as in the Vaikuṇṭha Perumal Temple at Kāñcīpuram, and in the yet higher storeys the images of Brahmā, Mahāviṣṇu, Sadāviṣṇu and the four armed Nārāyaṇa. "Like a hollow cane of bamboo (venurandhravat) are the cells placed one above the other in the vertical axis of the Prāsāda" ('Vaikhānasāgama', VI).¹¹⁵

v1.c8.s7.54

¹¹⁵ Cf. the construction of the Liṅgarāja Prāsāda, Bhuvanesvar, and others, note 49.

v1.c8.s7.55

The ascent of the cave along the vertical axis of the Prāsāda is by a twofold process, which is one in nature; by coalescence, and reduplication or repetition, the crypt arises on to higher and higher levels. The vertical axis of the Prāsāda always passes through its centre.¹¹⁶ By its ascent from underground, the crypt of steps leading down from inside the shrine doorway. The ante-chamber is 2' lower than the hall floor, a graded descent to the origin and centre, just as inversely, the superstructures with their finials, of several halls mark the graded ascent, level upon level, in many 'bhūmis' towards the final point on high (Pl. I). As above, so below, with the corresponding changes of direction, form and accessibility. The sunken shrine of the Śiva temple of Lonad (Kalyan) is three feet below that of the Maṇḍapa (ibid., p. 21).

v1.c8.s7.56

¹¹⁶ The plinth or also the terrace on which the temple stands do not necessarily take part in this vertical extension. Some of the highest Prāsādas, especially in Orissa, as those of the Liṅgarāja Temple, Bhuvanesvar, (middle 11th century), have no socle. The Prāsāda rises straight from the ground, although lesser temples, but not the very least ones rise from a plinth (M. M. Ganguly, 'Orissa and her Remains', p. 107).

v1.c8.s7.57

The shrine of the Śiva temple at Rajur (Buldana) is still deeper than that of the temple of Amaranātha, and similar to it is the temple at Chandol (Burgess-Cousens, op. cit., p. 15). In Rajputana, in the temple at Visalpur, near Deoli, of the year 1174 A.D. the Garbhagṛha is three and a half feet lower than the Maṇḍapam (Cunningham, ASI. Report VI). The Jyotirliṅga temple of Oṅkāra at Mandhata (Narmada ; Cousens, 'Mediaeval Temples of the Dakhan', p. 13) and the temple of Boram Deo (near Chapri, Chattisgarh; ASI. vol. XVIII. p. 34) show also by the difference of floor levels of Garbhagṛha and Maṇḍapa, the secret (gupta) nature of their crypt and sanctuary, the Garbhagṛha.

v1.c8.s7.58

In many of the temples in Bhuvanesvar, Orissa, dating from the ninth to the twelfth century (Mārkaṇḍeyeśvar; Bhāskareśvar, Pāpanāsinī, etc.) the floor of the Garbhagṛha is several feet below that of the 'Jagamohan' or Maṇḍapam.

v1.c8.s7.59

The so-called "Bhāskareśvar" temple (M. M. Ganguly, l.c.) is called Megheśvar by the local people and vice versa. The above remark refers to either of these shrines.

v1.c8.s7.60

comes to take the place of the cave in the mountain, which is the image of the complete Prāsāda with its superstructure.

v1.c8.s7.61

In some shrines (Kandarīya Temple, Khajuraho, Fig. i; Pt. VI) the Garbhagṛha far from being lower is even higher in level, than the Maṇḍapam. It is raised into the height of the Śikhara, itself partaking in its ascent.

v1.c8.s7.62

In the terrace temples (p. 149) a hollow central shaft below the high sanctuary is extended vertically across the terraces.¹¹⁷

v1.c8.s7.63

¹¹⁷ This and all other information about Ahicchatra has been received by the courtesy of Dr. R. E. Mortimer Wheeler, Director General of Archaeology in India, through Mr. A. Ghosh, Superintendent of Archaeology, Excavations Branch, who writes: "Of the two most prominent temples at Ahicchatra, one was more or less completely explored, while the other and bigger one had to be abandoned before it could be fully understood. There are some other temples. . . . Both the temples underwent several repairs and restorations, resulting in horizontal and vertical increases in their dimensions. In all their stages, however, they are square in plan with projections on the west for flights of steps.

v1.c8.s7.64

The superposition of cavities, in the interior of the temple, took place along the vertical axis; below ground and above, on various levels. In principle, it inheres in the temple with its high superstructure. The buildings however, but for the type of the Vaikuṇṭha Perumal temple, do not show the vertical series of their internal cavities.

v1.c8.s7.65

The underground crypt is secret; and in the vast majority of temples there is but one Garbhagṛha, closed on top by a flat ceiling or shallow dome which seals, as it were, at the same time, the interior of the superstructure (Fig. i, Part VI).

v1.c8.s7.66

The Garbhagṛha, the Cave in the Mountain, lies below its highest point. Along this axis, on any level of the temple, there is, in principle, this secret centre. Even though it is inaccessible from within, its position is marked by the superposition of surrounding storeys (bhūmi) on the monumental body of the Prāsāda (Fig. h, Part VI).

v1.c8.s7.67

The sum of all the possible cavities, one above the other, in the centre of the temple, is "like a hollow reed". It traverses, in principle, though not structurally, the temple from the apex to its foundation. On top of the superstructure, the "hollow reed" or shaft is seen to emerge from the body of the temple. It exceeds it by a short span and is crowned by a 'dome' (Figs. f-h, Part VI) or by the Āmalaka (Fig. i; and Pl. I) of the temple.

v1.c8.s7.68

The socle (pīṭha) where present, is a solid substructure, the upper surface of which forms the floor of the building. The crypt may be within the socle or base or extend below the outside ground level. The Pīṭha, where present, is piled on top of the completely filled in, solidly, or by cells, originally excavated site of the entire building. This is the general practice and the texts corroborate it.

v1.c8.s7.69

The fully exposed temple has three storeys in its last three stages. The first and earliest stage being buried very deep below the later superstructures was imperfectly explored.

v1.c8.s7.70

The plan of each single storey is square. There is no shrine on each storey, the only one being on the top. Each terrace leaves a sufficient space between the central part and the parapet for serving as ambulatory.

v1.c8.s7.71

The axis of the temples consists of a hollow central shaft, filled with debris, on the top of which the sanctuary was erected. In neither case has the central shaft been exposed to the lowest depth. In one case it was dug down to 12 feet from the top.

v1.c8.s7.72

The earliest stage of the temples evidently belongs to the Gupta period, as one of them was founded on a level yielding typical pottery of the Kuṣāṇa period. . . . They continued in their last stages till the end of the tenth or eleventh century".

Linked texts (29)