Praudhamanorama p177

p177

On "vibhāṣā ṅiśyoḥ" (Pā. Sū. 6.4.136): śi is the one enjoined by "napuṃsakācca", but not the śi of jas and śas, because in its presence the bha-designation is impossible. Because by 'bhasya' only a non-sarvanāmasthāna beginning with y or a vowel is implied, he says this: 'asarvanāmasthāna...'.
On "ata eva": because of the usage 'doṣaṇī'. Although the Bhāṣya can be well explained even when a single remainder is made of an aggregate dvandva, nevertheless it is logical that kakud is a separate inflected word. Otherwise, the singular of kakud and the dual of the two arms could not be obtained; this is the purport.
"Raising his right arm, the night-roamer rushed at him", thus Raghuvaṃśa. There, dakṣiṇaṃ doḥ ('the right arm') is connected as the object with udyamya ('raising').
On "doṣaṃ tasya": "Taking up the arm of that bow, which held such an arm, while grasping its string / finding fault with that bow, which incurred such a fault, while appreciating its merit." Such is the usage of Śrīharṣa. There, the relation is: taking up (gṛhṇan) the arm (doṣam / hastam) and fault (dūṣaṇam) of the bow which had (bhajataḥ) it, and grasping (gṛhṇan) the bowstring (guṇam / maurvīm) and excellence (atiśayam).