The final a and i (both long and short) of a bha aṅga (stem) are elided before a taddhita affix.,
Thus दाक्षी, प्लाक्षी, सखी from दाक्षि, प्लाक्षि and सखि with the feminine affix ई 4.1.65 The above forms could also have been evolved by the simple rules of sandhi, as दाक्षि + ई = दाक्षी ॥ It might be asked, where was then the necessity of eliding इ? There would arise difficulty in the forms like अति सखि ॥ From the word सखी formed by ङीष् under 4.1.62, we form the compound सखीमतिक्रान्तः = अतिसखि, the final becoming short by 1.2.48 (the samasanta affix टच् is not added here as required by 5.4.91, that rule applying to the masculine form सखि and not to the feminine सखी). Now if in forming सखि + ई = सखी, had we not elided the letter इ of खि, then the long ई being the single substitute for both इ + ई, would be considered as like the final इ of सखि by 6.1.85. That being so, the word अतिसखि would be considered a non-घि word by (I .4. 7), for सखि has been specifically excluded form घि class. Not being a घि, we cannot have the form अतिसखेः in the Ablative and Genitive singular. Hence the necessity of eliding इ before the feminine ई, so that the ekadesa rule 6.1.85, should not apply.
The short इ is elided before Taddhita affixes:- दुलि — दौलेयः, वलि — वालेयः, अत्रि — आत्रेयः ॥ अ is elided before ई as :- कुमार — कुमारी, गौर — गौरी, शार्ङ्गरव — शार्ङ्गरवी ॥ अ and आ are elided before the Taddhita :- दक्ष - दाक्षि, प्लक्ष — प्लाक्षिः, चूड — चौडिः, बलाका — बालाकिः, सुमित्रा — सौमित्रिः ॥
Vart:- Prohibition must be stated in the case of ई (शी), when it is the substitute of औ 7.1.18, being the case-affix of the nominative, and accusative dual in the neuter, and of certain feminine in long आ ॥ Before this ई, the preceding vowel, इ and अ are not elided. The stem before this ई is also bha by (I .4.18) in the case of Neuter nouns. Thus काण्ड + ई = काण्डे; कुड्ये ॥ सौर्य + ई = सौर्ये (the य would have been elided by 6.4.149. It is owing to this Vartika, that we have inserted the word 'feminine' in the translation of the sutra. This may also be done by reading the anuvritti of श्यां 6.4.136, and न 6.4.137 into this sutra.
Vart:- इयङ् and उवङ् are superseded, when they would apply simultaneously with this lopa. Thus इयङ् and उवङ् have unimpeded scope in श्री or भ्रू + औ or अस् (जस्), as श्रियौ, श्रियः, भ्रुवौ, भ्रुवः ॥ Lopa has unimpeded scope in कमण्डलू — कामण्डलेयः (with ढञ् 4.1.135), and भाद्रवाहेयः ॥ But in वत्सप्री + ढञ् 4.1.135, both the rules 6.4.77 and 6.4.148, present themselves. The former is superseded, and we have वात्सप्रेयः (वत्सं प्रीणाति = वत्सप्रीः, तस्यापत्यं); so also लैखाभ्रेयः with the ढक् affix, this word belonging to Subhradi class 4.1.123.,
