अनुदात्तस्य च यत्र+उदात्तलोपः

The word udatta is understood here. Thus कुमार꣡ + ई॒॒ = कुमा॒॒री꣡ ॥ The word कुमार has acute on the last, when the unaccented (anudatta) ङीप् is added to it, the अ꣡ is elided 6.1.148, the anudatta ई॒॒ becomes udatta. So also प॒॒थिन्꣡ + श॒॒स् = प॒॒थः꣡ 7.1.88, प॒॒थिन् + ए॒॒ = प॒॒थे, प॒॒थिन् + आ॒॒ = प॒॒था꣡ ॥ The पथिन् has acute on the last. So also कुमुद + ड्मतुप् = कुमुद्वा꣡न् 4.2.87, न॒॒ड्व꣡त्, वे॒॒तस्वत् ॥ The words कुमुद &c, are end-acute, and the affix मत् (वत्) is anudatta 3.1.4.

Why 'an anudatta vowel gets &c'. Observe प्रा॒॒सङ्गः + य॑त् = प्रासङ्ग्यः ॥ (प्रासङ्गं वहति 4.4.76). Here though प्रसङ्ग꣡ is end-acute 6.2.144 formed by घञ् with _gati)) it is followed by य॑ which is svarita 6.1.185, this svarita causes the lopa of अ꣡ udatta: but it does not itself become udatta. No, this is not a proper counter-example. For by the general rule of accent 6.1.158, when one syllable is ordained to be udatta or svarita, all the other syllables of that word become anudatta. Therefore when य॑ is taught to be svarita, all the other syllables (like ङ्ग꣡) will become anudatta. So there is no lopa of udatta when य॑ comes. The word अनुदात्त is used to indicate that the initial anudatta becomes udatta. Had anudatta not been used, the sutra would have been यत्रोदात्तलोपश्च, which would mean यस्मिन् प्रत्यये परत; उदात्तालुप्यते तस्यान्त उदात्तो भवति, for the anuvritti of अन्त would then be read from the last. There would then be anomaly in the case of मा हि धुक्षाताम्, and माहि धुक्षाथाम् ॥ From the root दुह we have formed धुक्षाताम् and धुक्षाथाम् in the Aorist Atmanepada, dual. दुह् + क्स꣡ + आताम् = दुह् + स् 7.3.72 + आताम् ॥ Here आताम् causes the elision of the udatta अ of स ॥ Therefore the final of आता꣡म् would be udatta, which is not desired. The augment अ is not added because of मा, had it been added, the अ꣡ would have been udatta 6.4.71. Had हि not been given, then also the whole of धुक्षाताम् would have become anudatta in माधुक्षाताम् because of तिङ्ङतिङः 8.1.28. See 8.1.34 (हि च). Why have we used the word यत्र in the sutra? If it was not employed, then the subsequent anudatta may be such which would not have caused the elision of the previous udatta; such udatta being elided by some other operator, and still such anudatta would have become udatta. Thus in भार्गवः, भार्गवौ, भृगवः ॥ Here भृगवः is plural of भार्गवः ॥ The word भार्गव is end-acute and this acute (अण्) was elided in the plural, before the affix जस् was added. The elision here is not caused by जस्, but is a subject of जस् (not निमित्त but विषय). Therefore जस् does not become udatta, which would have been the case, had यत्र not been used. प्राक् सुबुत्पत्ते गोत्रप्रत्ययस्य लुक् ॥ The gotra affix was elided prior to the adding of the case-ending. Why do we say \when an udatta is elided\? Observe बिद + अञ् 4.1.104 = ब꣡द ॥ The feminine of this will be बैद꣡ + ङीन् 4.1.73 = बैदी꣡ ॥ Here ई causes the elision of anudatta_ अ of द, and hence remains unchanged. So also आ꣡ र्वी ॥,