The affix aṇ comes in the sense of descendant after prātipadika (Nominal Stem) denoting the name of a ṛṣi or the name of a person of the family of 1. andhaka 2. vṛṣṇi 3. kuru,
This debars the affix इञ् ॥ The Rishis are like Vasishtha, Visvamitra &c. Thus वासिष्ठः, वैश्वामित्रः ॥ So also श्वाफल्कः, रान्धसः being persons belonging to the family of Andhaka; and वासुदेवः, आनुरुद्धः being persons belonging to the family of Vrishni, and नकुलः, साहदेवः being persons belonging to the family of Kuru.
The word अत्रि is the name of a Rishi, in forming its Patronymic, the present sutra enjoins अण्; but (IV. I. 122) enjoins ढक्; thus there being a conflict, we apply the maxim of interpretation contained in (1. 4. 2). So that the Patronymic or अत्रि will be formed not by अण् but by ढक् ॥ Thus आत्रेयः ॥ Similarly उग्रसेन a person of Andhaka family, forms its patronymic औग्रसेन्यः 4.1.152 by ण्य; and not by अण्, so also वैष्वकसेन्य 'son of Vaishvaksena' a person of Vrishni, family; so also भैमसेन्यः 'son of Bhimasena' a person of Kuru family.
Note.- The words or sabdas are eternal; how is it that rules applying to them should be made dependant upon their occurring in particular families or not, when families themselves are non-eternal? In other word, how a rule affecting a permanent object is conditioned by an impermanent adjunct? To this, some reply that Panini has by mere coincidence (and not as cause and effect) classified certain definite number of words as belonging to certain families, and then the rule made applicable to them. Moreover a permanent object is sometimes designated by an impermanent object : as the permanent time is designated by the name of Saka. Others say that the families of Andhaka &c., are also permanent, and the words 'Nakula' 'Sahadeva' &c., found therein are, of course, permanent.,
