INTRODUCTION xlix Malli.’s observant eye detects that the mas. ending of māsara and ācāma should dispel all such misapprehension: द्वन्द्वान्तश्रूयमाणपुंस्त्वानि (p. 584). Great care is taken by Malli. where mispronunciation of sibilant sounds is generally observed.1 Regarding Sunāsīra (Indra), Malli. says the initial sibilant is a palatal and not the dental sibilant s; he quotes as authority शुनासीरेणेति हरिप्रबोधयमकात् तालव्यादिः (p. 32 ) and points out that the word is spelt as sunāsīrah also, quoting Ascaryamanjari: शक्र इव नाकविख्यातसुनासीरः इत्याश्चर्यमञ्जरीश्लेषाद् दन्त्यः (p. 31). Though Malli. does not as a rule give the derivation of words, he does so when a particular sound is irregular, doubtful or novel, e.g. in angaņam (courtyard ) the cerebralization of na is irregular as pointed out by Subhūti: अङ्गत्यस्मिन्निति अङ्गणम् । पृषोदरादित्वाण्णत्वमिति सुभूतिटीका । अङ्गनमिति च कतिचन कथयन्ति । (p. 203) Similarly, in kavātah ( door), the middle sound can also be a labial as per the Rabhasakośa: पवर्गमध्य इति रभसकोशकारेणाभ्यधायि; कपाटः कपाटीति (p. 206) .. Amara gives carmakaşā as a synonym for saptalā (soap tree); Malli. notes that the sibilant may be sa, as the Subhūtitikā sanctions it : चर्मकसेति दन्त्योपधः सुभूतिटीकायां व्युत्पादितः । (p. 303) In this context, he gives a whole verse describing its merits and the place where it is commonly found. Other examples are dłşad (stone) which may be drśad on the authority of Aruna- datta and others : 1 शशः षषो मा भूत् (Mahābhāsya, I. 1. 1). D
