________________
201
is concerned deponentia do not exist in Pkt. It should be noted that Alsdorf (1937: 61-8) after studying those verbs in Apa. earlier identified as deponentia, reached a similar conclusion regarding that language.
An obvious case is tīrae. This is not, as Weber (Abh., p. 64) assumes, a deponens but a regular passive of the verb tara-, the usage of which is comparable to that of Skt śakyate. gasijjihii in 804, likewise, is a regular passive. Gathā 804 (only found in S) reads as follows:
vārijjartT navako- mā te amvupisão
mui tti ma putti aṁgane suvasu caído tti muhaṁ gasijjihii.
If someone tries to stop you, saying '(it is) full moon', you should (obey and not sleep in the courtyard, little girl, lest your face will be swallowed by Rāhu) who may think the (real) moon (only) a piśāca in the water (arvupisão).
A different case is pijjai in 678. Gathā 678 (only found in w and Y ; and in G and ) reads as follows:
pijjai kannańjalihim duddha jalas amilia
janar avamilia pi tujjha salāvam să vālā rāahaṁsi vva.
As far as the first half is concerned the passive pijjai is in perfect order. An anomaly arises only in connection with the second half. One may therefore consider the possibility that the Gathā is a patch-work verse assembled, with little regard for syntax, from two independently existing half-verses. Note in this connection the awkward repetition of the word (Saṁ)miliań. The fact that Hemacandra in IV 10 includes the passive stem pijja- among the Adeśas for pā- is no independent support for Weber's (and the traditional) interpretation for pijja- in this Gathā. The inclusion of pijja- in Hemacandra's sutra may in fact rest upon a misunderstanding of an instance like this one. Another instance of the 'deponens' pijja- quoted by Pischel (in his translation of Hem., p. 132), namely pijjanti in Śakuntala (ed. Pischel, p. 29, line 5) is in my opinion a regular passive.
