No. 18. PĀLU MÄKICCÄVA ROCK-INSCRIPTION OF GAJA PALU ĀLU MÄKICCÂVA is the present name of an abandoned tank in the jungle, about 120 yards off the high road to Trincomalee, at a distance of sixteen miles from Anuradhapura. The inscription is on a rock at the vana or 'spill-water' of this tank. Dr. Müller gives an account of it, with a transcript, an English translation, and a lithographic plate in his Ancient Inscriptions in Ceylon1; but Mr. Bell's examination of it resulted in the publication of a revised text and translation in his Seventh Progress Report for 1891. The present article is based on a good estampage and an eye-copy, with which he has since supplied me. It will be seen from the accompanying facsimile (Plate 27) that, as Mr. Bell has already said, the inscription is in a well-preserved state. It is engraved in six lines, covering an area of 3 by 8 feet of the surface of the rock. The letters, with the exception of the fifth, sixth, and seventh in the first line, are quite clear. Their size varies from 4 to 7 in. in height, and 2 to 5 in. in breadth. They belong to the Southern Brāhmi alphabet of the latter part of the second century A.D., exhibiting, as they do, nearly all the graphic peculiarities of the script of the Nāsik, Kuḍā, and Junnar inscriptions of about the same period. There is besides a complete agreement in the admixture of types, old and young, in the same record. Those most noticeable in the present one are:-(a) the angular E ja and the archaistic' round ɛ ja in ll. 4 and 2; (6) the cerebral & in vaḍamanaka of 1. 4, which differs but little from the shape of the cerebral in 1. 5, appears to belong to the alphabet of the Northern Kṣatrapas 2; (c) ti with the fully developed loop as on some Andhra coins, and in Nāsik, No. 174, ti with the lower end of the vertical stroke ending in a curve to the left, which is most likely the precursor of the looped form: these two, both in the second line, are the only variations from the usual type of ta h with the broad and more or less round side-limb; and (d) the forms of ba with the left side of the quadrangle slanting outward either concavely as in (1. 1) and (1. 5), or with an angular bend as in □ (1. 3) and (1. 5). 1 See No. 10, pp. 28, 74, and 110. • See, for example, the form of the aksara da in Soḍāsa's record at Mathura (Ep. Ind. vol. ii. p. 199, No. 2, Plate). For examples, see Rapson's Catalogue of the Coins of the Andhra Dynasty in the British Museum. .If we now carefully compare these various types in the Pālu Mäkiccäva inscription under discussion with those in the rock records of Perumäiyaṇkuļam and Maharatmale, dealt with above at pp. 58-74, we shall be constrained on palaeographic grounds to regard the Perumäiyankulam record as the oldest of the three. That at Pālu Mäkiccäva comes next, as the majority of its letters is in type distinctly older than the more ornamental forms with longer curves noticeable in the script of the Maharatmale rock. The contents too of these three inscriptions give us the same age, for the first, i.e. the Perumäiyaṇkuļam, record belongs to the reign of king Vahaba (P. Vasabha, 130-174 A.D.), the Pālu Mäkiccäva inscription to that of his grandson king Gamiņi Abaya (P. Gajabahuka Gāmanī Abhaya, 177-199 A.D.), and the Maharatmale one to the time of king Naka (P. Mahallaka Nāga, 199-205 A.D.). The next point to be considered is the four-inch square symbol engraved just below the akṣaras va and ṇa of the word kahāvaṇa at the end of the fourth line. It is subdivided into four equal squares, and resembles at first sight the conventional representation of the railed enclosure of the sacred Bōdhi tree, so often found side by side with other religious or auspicious symbols on ancient Buddhist coins, seals, and inscriptions1. But whether this is an instance of the railed enclosure alone without the tree being used as a Buddhist emblem or not we cannot decide until other examples are brought to light. There is also no ground to suppose that it is a Ceylon variation of the rounded figure for 90 occurring on Kṣatrapa coins of the second and third centuries A.D.2 In view, however, of its isolated position immediately under the name of the coin, Mr. Bell's suggestion that 'it may denote the standard by which the value or weight of the kārṣāpana was reckoned, i.e. by padas or quarters,' seems quite plausible. The symbol may further signify that the coin in question was the silver kārṣāpaṇa, which had its half and its quarter size, pādika or pāda. As regards the words in the text, the following remarks may be offered :- 'See, for instance, the Sohgaura plate (J. R. A. S., 1907, p. 510). * See Rapson's Indian Coins, pp. 144 ff. A. S. C. Seventh Progress Report, p. 44. • See Cunningham's Coins of Ancient India, p. 47; Rapson's Indian Coins, pp. clxxxiii ff., and Rhys Davids' Ancient Coins and Measures of Ceylon, p. 13. • See above, p. 60. though it occurs once or twice in the Habarana inscription 1. Possibly the later Sinhalese form Vähäp with modified a-sound is a derivative of it. Manumaraka (1. 2) is also spelt marumakane. See above, p. 63. Puti and maharaji (11. 2−3) have the Magadhi nominative in i3. Upala-vibajaka (1. 3) is probably identical with Upala-bijaka of Tammännekanda record. Upala may be either the Skt. upala, a stone or rock, or a derivative of Skt. utpala, P. uppala, the blue-lotus, often used as a personal name. Vibajaka is Skt. vibhājaka. The form paca (1. 4) for Skt. and Pkt. pañca, Sinh. pasa, paha, occurs also in Indian inscriptions of about the second century A. D. Kahāvana is Skt. kārṣāpaṇa, Pkt. kāhāpaṇa, kāhāvana, P. kahapana, Sinh. kahavana. Hence paca-sahasi-kahāvaṇa (1. 4) is P. pañca-sahassž-kahāpaṇā". Dariya and kanavaya are derivatives of gerunds in ya from the causatives of Skt. dhy and khan. Their equivalents in Pāli would be dhariya and khanāpiya, the Sinhalese causative sign -va being derived from Skt. -paya (P. -paya, -pē, Pkt. -ve). Tubaraba is no doubt a Sinhalese form of Thūpārāma as suggested by Mudaliyar Gunasekara 10. Buka11-sagahaṭaya (1. 5) is P. bhikkhu-sanghassa atthaya. The inscription contains nothing historically important. It confirms the Ruvanväli Dāgaba 12 and Maharatmale 3 records as regards king Gamiņi Abaya being the son of king Tisa and grandson of king Vahaba. It further states that he spent 5,000 kārṣāpaṇas on the excavation of the Vaḍamanaka Tank in the Upala district, which he then granted to the Buddhist priesthood at See A. S. C. Seventh Progress Report, p. 47, and A. I. C. No. 55. Cf. also Upaladonika-vavi in A. I. C. No. II. 6 As Uppalo, the father of Ummāda Phussadēva (Mv. xxiii. 82). See, for example, Nāsik, Nos. 1 and 5 (Ep. Ind. vol. viii. pp. 59 and 73), and Bharhut, No. 144 (Z. D. M. G. xl. p. 75). 7 Cf. P. dasa-sahassi-lõkadhātu. I • Prakrit dialects hardly make any distinction in the use of the suffixes tvā and ya to form gerunds. They are added indiscriminately to simple roots as well as to those with prepositional prefixes (Pischel, op. cit. § 581). For some examples of gerunds in ya from simple roots, see Whitney's Skt. Gram. § 990 a. • Cf. also the Tamil causative in -pi or -vi. 10 A. S. C. Seventh Progress Report, p. 44, note 8. Cf. ohu Mahabō vànda dā vaňdanaṭa Tumbarup yannahu (Dharmapradīpikā, p. 324). "For the forms bika and buka, see above, p. 73. For examples of the change of original i to u and vice versa, see Pischel's Gram. der Pkt. Spr. §§ 117 and 124, and Geiger's Litt. u. Spr. der Sinh. § 12 (2). 12 A. I. C. No. 5. 13 See above, p. 62. though it occurs once or twice in the Habarana inscription1. Possibly the later Sinhalese form Vähäp with modified a-sound is a derivative of it. Manumaraka (1. 2) is also spelt marumakane. See above, p. 63. Puti and maharaji (11. 2–3) have the Magadhi nominative in ¿ 3. Upala-vibajaka (1. 3) is probably identical with Upala-bijaka of Tammännekanda record'. Upala may be either the Skt. upala, a stone or rock, or a derivative of Skt. utpala, P. uppala, the blue-lotus, often used as a personal name. Vibajaka is Skt. vibhājaka. The form paca (1.4) for Skt. and Pkt. pañca, Sinh. pasa, paha, occurs also in Indian inscriptions of about the second century A. D. Kahāvana is Skt. kārṣāpaṇa, Pkt. kāhāpaṇa, kāhāvaṇa, P. kahāpaṇa, Sinh. kahavana. Hence paca-sahasi-kahāvaṇa (l. 4) is P. pañca-sahassi-kahāpaṇā". Dariya and kanavaya are derivatives of gerunds in ya from the causatives of Skt. dhy and khan. Their equivalents in Pāli would be dhāriya and khanāpiya, the Sinhalese causative sign -va being derived from Skt. -paya (P. -paya, -pē, Pkt. -ve). Tubaraba is no doubt a Sinhalese form of Thūpārāma as suggested by Mudaliyar Guņasēkara 10. Buka 11-sagahaṭaya (1. 5) is P. bhikkhu-sanghassa atthaya. The inscription contains nothing historically important. It confirms the Ruvanväli Dāgaba 12 and Maharatmale 3 records as regards king Gamiņi Abaya being the son of king Tisa and grandson of king Vahaba. It further states that he spent 5,000 kārṣāpaṇas on the excavation of the Vaḍamanaka Tank in the Upala district, which he then granted to the Buddhist priesthood at 1 A. I. C. No. 61, 1. 10. 2 Cf. also Tisaya for Tisa (op. cit. No. 2). * Compare, however, vähäp in E. S. p. 85. • See Pischel's Gram. der Pkt. Sprachen, p. 249. * See A. S. C. Seventh Progress Report, p. 47, and A. I. C. No. 55. Cf. also |