Undavalli cave temple

The great rock-cut temple at Undavalli is the largest of a group of early cave temples on both banks of the Krishna at Vijayawada (then Bezwada).¹² Its date and its makers are unknown, because none of its inscriptions records its origin.¹² It is described here because one historian gave it to the Vishnukundins.³

The temple

A. H. Longhurst described the temple in 1924.²⁴⁵ In his account it has a ground floor and three upper storeys cut out of the natural rock.²⁴⁵ Its front is about 90 feet long and about 50 feet high.²⁴⁵ The ground floor is an unfinished pillared hall.²⁴⁵

The first storey seems to have been designed as a triple shrine, probably for the Hindu Trinity.²⁴⁵ But most of its sculptures show Vishnu and his legends.²⁴⁵ The second storey is a large pillared hall with a small empty shrine at one end.²⁴⁵ At the other end a sanctuary holds the huge image of Vishnu lying on the serpent couch (Anantashayana).²⁴⁵ The top storey is unfinished.²⁴⁵

Longhurst noted Shaiva (Shiva) figures as well as Vaishnava (Vishnu) ones in the second storey.⁵ He concluded that Undavalli was a stronghold of the cult of the linga (stone emblem of Shiva) in the 7th century, “in spite of the presence of Vaishnava figures in the big temple”.⁵ Four smaller unfinished rock-cut shrines and a group of miniature memorial shrines are cut in the hill nearby.⁶

Inscriptions

According to Longhurst, the earliest inscription is an incomplete Telugu record of about the 9th century.² It is on the cornice of the ground-floor front and is later than the monument.² Another record shows the temple “in good order and well patronised by the Reddi chiefs of Kondavidu” in the 14th century.² A record from South Arcot district says that the Vijayanagara king Krishnaraya visited the temple and gave it gifts in the 16th century.² These records call the temple Anantasayigudi or Anantasyanagudi.²

Lions and a vase

On the base of two pillars Jouveau-Dubreuil found carvings of lions, and at the foot of another pillar a vase.³ He compared the lions with the lion on the seals of the Vishnukundin grants.⁷ In particular he quoted H. Krishna Sastri’s description of the Ramatirtham seal.⁷ It shows “an advancing lion or tiger with its left fore-paw raised”, its mouth wide open and its tail looped over its back.⁷

He also noted that coins with a lion on one side and a vase on the other are found on the Godavari and the Krishna.³⁷ These coins had been given to the Pallavas, but he was convinced that they belonged to the Vishnukundins.³⁷

Longhurst described the same carvings.⁴⁵ In his account, the bases of two front pillars on the second storey carry small lions with one fore-paw raised.⁴⁵ Other pillar bases carry a vase ornament.⁴⁵ He found both “similar in style to those appearing on certain coins which have been attributed to the Pallavas”.⁴⁵

Vishnukundin or Pallava?

Jouveau-Dubreuil argued in The Pallavas (1917) that the caves of Undavalli were the work of the Vishnukundins.³⁸ He reasoned that the dynasty ruled on the Krishna and Godavari before the Chalukya conquest.³⁸ He also reasoned that its crest was the lion of the carvings.³⁸ He thought the Vishnukundins had taken the idea of cutting caves from their Vakataka relatives, whose inscriptions are found in the caves at Ajanta.³⁸ He named the caves at Vijayawada, Mogalrajapuram, Undavalli and Sitanagaram as their work.³⁸

He went further.⁸ He suggested that the Pallava Mahendravarman I copied these caves when he cut his own temples round Kanchipuram.⁸ He believed Mahendravarman to be a Vishnukundin king’s grandson.⁸ In 1920 he repeated the attribution, citing the lion and vase on the pillars at Undavalli and Mogalrajapuram.⁹

Longhurst, writing in 1924, allowed that the Pallavas might have learned rock-cutting from the Vishnukundins, but saw no proof.¹⁰ Until more was known about the Vishnukundins and their monuments, he preferred to treat the Vijayawada group as Pallava monuments of the early 7th century.¹⁰ He based this on architectural grounds.¹⁰ He found the plan and style of the Undavalli storeys typical of the time of Mahendravarman I.⁴⁶

K. R. Subramanian, in 1932, reported Dubreuil’s attribution.¹¹ He noted that the Madras Archaeological Department considered the caves Pallava.¹¹ The epigraphical report for 1909, on the other hand, had placed them in the early centuries CE.¹¹

Footnotes

1. Longhurst, Pallava Architecture, Part I (1924), p. 22 (archive.org↗ (opens in a new tab)). ↩ ↩²

2. Longhurst, Pallava Architecture, Part I (1924), p. 27 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶

3. G. Jouveau-Dubreuil, The Pallavas (1917), p. 33 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

4. Longhurst, Pallava Architecture, Part I (1924), p. 28 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴

5. Longhurst, Pallava Architecture, Part I (1924), p. 29 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵

6. Longhurst, Pallava Architecture, Part I (1924), p. 30 (archive.org↗ (opens in a new tab)). ↩ ↩²

7. Jouveau-Dubreuil, The Pallavas (1917), p. 34 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵

8. Jouveau-Dubreuil, The Pallavas (1917), p. 35 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

9. G. Jouveau-Dubreuil, Ancient History of the Deccan (1920), p. 92 (archive.org↗ (opens in a new tab)). ↩

10. A. H. Longhurst, Pallava Architecture, Part I (Memoirs of the Archaeological Survey of India No. 17, 1924), p. 5 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³

11. K. R. Subramanian, Buddhist Remains in Andhra and the History of Andhra between 225 and 610 A.D. (1932), p. 118 and note 3 (archive.org↗ (opens in a new tab)). ↩ ↩² ↩³