नाट्यशास्त्रम् (प्रथम भाग - अभिनवभारती एवं हिन्दी टीका सहित)

.XVIII. 58 ] RULES ON THE USE OF LANGUAGES 333

livelihood by [collecting] wood and leaves should be assigned Śābarī¹ as well as the speech of forest-dwellers.

54. For those who live in places where elephants, horses, goats, sheep, camels or cows are kept [in large numbers] Ābhīrī¹ or Śābarī² has been prescribed, and for forest-dwellers and the like, Drāviḍī³ [is the language].

55. Oḍrī is to be assigned to diggers of subterranean passages, prison-warders, grooms for horses¹ ; and Heroes and others like them while in difficulty are also to use Māgadhi for self-protection.

Distinguishing features of various local dialects

56. ¹To the regions [of India] that lie between the Ganges and the sea, should be applied a dialect abounding in e²

57. To the regions that lie between the Vindhyas and the sea should be assigned a language abounding in na¹ ( or ta).

58. Regions like Surāṣṭra and Avanti lying on the north of the Vetravatī one should assign a language abounding ca¹.

---

² See 54 note 3.
54 (C.52b-54a; B.XVII.56). ¹ Ābhīrī dialect is not available in any extant drama. ² See 53 note 1.
³ Drāviḍī dialect is not available in any extant drama. It is possible that it was not a pure Dravidian speech (See 44 above). Possibly a Middle Indo-Aryan dialect in which Dravidian phonetic and lexical influence predominated was meant by this. Its habitat was in all likelihood some region of North India. Cf. Nitti-Dolci, p. 120-122.
55 (C-54b-55a; B.XVII.57). ¹ Emend oḍhrikārasarakṣasāṃ (ḍha in B) into oḍrī kārāśvarakṣatāṃ (C. oṣṭrikānāñ ca rakṣatāṃ and B. sandhikārāśvarakṣātāṃ); for Oḍri Pkt. see 48 note 3. and Nitti-Dolci, pp. 120 f.n. 4 and 122.
56 (C.55b-56a; B.XVII.59). ¹ B. again reads 44 after 55.
² This "e" is perhaps termination of the nominative singular the a-bases in AMg.
57 (C.53; B.XVII.60). ¹ This relates to a dialect of Pkt. which changes na always into ṇa. Though according to some grammarians Pkt. is always to change na into ṇa, it seems that such was not strictly the case with all its dialects. For example in the so-called Jain Pkt. (AMg. of Hemacandra) has initial n and intervocal ṇṇ.
58 (C.56b-57a; B XVII.61). ¹ It seems that at the time of the