adḍ is substituted for the Nominative 1st-Case and Accusative -2nd case singular endings after the Pronouns ḍatara etc. (i.e. the aṅga (stems) formed with the affixes 1. ḍatara and 2. ḍatama, and the aṅga (stems) 2. itara 4. anya and 5. anyatara,
Karika
अपृक्तश्चेदमो दोषो निवृत्ते डतरादिषु ।
अद्डित्त्वाड्डतरादीनां न लोपो नापि दीर्घता ॥
These are the five pronouns, which in the list of Sarvanamans are read together (See 1.1.27):- डतर, डतम, इतर, अन्य and अन्यतर ॥ Thus कतर + अद्ड् = कतरत् (the अ of katara elided by ड्): as कतरत् तिष्ठन्ति, कतरत् पश्य ॥ So also कतमत्, इतरत्, अन्यतरत् and अन्यत् ॥ Why after these five only ? Observe नेमं तिष्ठन्ति, नेमं पश्य ॥ Why do we make the affix have an indicatory ड्? In order to prevent the lengthening of the vowel in the nominative singular : as कतर + अत् = कतरात् by 6.1.102. In the case of the accusative, अत् being the substitute of अम् will be sthanivat, and give us कतरत् by 6.1.107 even without ड् ॥ Why not make the affix merely त् and not अत्ड्; it would give कतरत् &c without the application of any rule of Sandhi? The simple त् would not give us the Vocative हे कतरत्, The final would have been elided in the Vocative, as being an aprikta. See however 6.1.69.
Karika :-If in the sutra 6.1.69, there is the anuvritti of the word 'aprkta', then there is fault with regard to अम् (i. e. the vocative of members in अम् will not be elided, हे कुण्डम् will be the form required and not हे कुण्ड); if however, the anuvritti of aprkta ceases, there will be anomaly with regard to pronouns कतरत् &c (i.e. we shall not have the form हे कतरत् but हे कतर) Therefore, by reading the affix अत् with an indicatory ड् i. e reading it as अद्ड्, we get out of this dilemma, and so there is not the elision of अत् in कतरत् &c in the vocative; nor is there lengthening of the vowel (कतरात्, which would have been the form had there been no ड्).,