Introduction] IRREGULAR GRAMMATICAL FORMS [xxvii gular contraction after the dropping of Visarga is somewhat frequent, especially in the case of pratīkas ¹. In two instances a is elided after the dual e: drsyete 'lpaḥ (i. 8) and dre'numatch (iv. 88). Sandhi is further applied to a Pragrhya in vaidadasvy rysi (v. 62; also 89). The irregularities in regard to nouns, pronouns, and verbs are not inconsiderable. The weak base is used in the feminine participle arohati (vii. 130). The base suläsa is once used instead of sulas (vi. 34). Adjectives derived from compounds, especially from Devata-dvandvas, are often irregularly formed: visvabhaisaj ya (viii. 50), trisāmvatsarika (v. 97); aindra-soma (vi. 27), soma-pausna (iv. 91), and others as in the Sarvanukramani 2. The nom. plur. of patni appears as patnayah (ii. 8, 9, 12; iii. 92) and patnih (viii. 128) beside the regular patnyah. Prathamasyām (beside dvitiyasyām) occurs once (viii. 95). The form caturbhih is frequently used instead of catasrbhiḥ, sometimes even in direct association with a feminine noun 3. Enam appears to be once used for enad (i. 73). Mantra is once used as a neuter (viii. 129). The neuter haras appears once (v. 19) as a masculine. In the verb, unaugmented forms occur several times: silat (iv. 113), barhayat (iv. 69), bodhayat (iv. 115), samayal (iv. 53). Anusasati once stands for the 3rd sing. present (vii. 37); the present base of landh is twice badhya (v. 134; vi. 14); apahanat stands for apahan (iv. 114), mocire for mumucire (vi. 15), asusvapat for asusupat (vi. 13). The form mathyati has a passive sense (iii. 62). Compound gerunds sometimes take -tvā: namaskrtva (i. 1), samgrhiträ (ii. 47), a-nitrā (v. 18); while -ya is once used in the simple gerund säntvya (iv. 3). With regard to syntax, the nominative is a few times used for the accusa- tive: hetavah for hetun (i. 28); pañcasat for pañcāśatum (vi. 54); bhrataraḥ for bhrätyn (iv. 32); and the accusative with iti instead of the nominative occurs several times. The order of the words is frequently disregarded, especially in the case of iti5. This irregularity goes altogether too far in two or three cases 6, making the passage either unintelligible without external evidence, or positively misleading. Vocabulary. With regard to vocabulary, the vowel of a few words appears shortened under the influence of the metre: sikată (vii. 79), asih (v. 170), Nubhaka (vii. 128). To the same cause are due the peculiar formations uruvasinī (ii. 59) = urvasi, and varisukapa (ii. 69) = vārṣākapa. Brhataspati is meant (ii. 3) for an etymological explanation of brhaspati. 1 Jātavedätha, ii. 30; tisroktah, vi. 136; vidvām- sarsigatām, vii. 112; samiddhapryah, iv. 62; dandeti, i. 50; ileti, iii. 138; tivrendravayubhyam, iii. 94; payasvatyuttara, vii. 10. There are besides a few irregular contractions of pratikas of a different kind: vitatädiḥ (vitatau), i. 57; raksohanadi, vii. 142. 2 See my edition, p. x, top. 3 See Index of Words, under catur. 4 Probably also in ii. 132. 5 v. 46; vi. 91; vii. 12, 46, 11I. See i. 115; ii. 13ª; iii. 92.
