________________ xlix was perhaps used at the end of Pathita-ganas, in Paninian system, is also considered to be a pointer to the Akrti nature of a Gana. In the sutras of the grammars of Sakatayana and Hemacandra, the Akrti-ganas are denoted by the plural form of the initial word of the Gana 38, while the Pathita, or Paripurna, Gana is characterised by the singular. There are, however, sutras where this formal distinction is not found in sakatayana, as in the case of Gahadi-gana, which though mentioned in plural is still treated as Akrti-gana139. As against this, Pathita is a name given to the Gana in which the author has supplied the complete and exhaustive list of words, e. g. Sarvadi, Svaradi, Kathadi, etc. Since it was not possible to add any new words to such lists, they were also called Nitya-ganas. Like the Akrti-ganas, the Pathita-ganas also are referred to in the sutras by their first word of their respective lists coupled with the word 'adi' or 'prabhrti or by its plural form. But, although Buddhisagara too has enumerated both these types of Ganas, it was not possible for him to adhere to this method of distingushing these two types by the use of the words 'di' or 'prabhrti' as also the plural, since he had to adjust his sutras in the structure of suitable metrical form, and many a times the exegencies of metre compelled him to use the Gananame in singular, dropping the words 'adi' or 'prabhrti' as also the plural, * while mentioning it in the concerned Sutra. However, he took the opportunity to make good the deficiency in his auto-commentary. Thus, for instance, the Svaradi-gana and the Svagatadigana, both of which are Aksti-ganas, are mentioned in singular in PGBV 2.2.62 as "Svara-svagatadeh', while the Taulvalyadi-gana which is a Pathita-gana is mentioned in plural, but without the use of the word 'di' in the PGBV 2.2.3.11 as 'In ataulvalibhyah'. 9.7. Vyakarana in the last analysis means, at least in the case of regularly formed words in Sanskrit, their reduction to verbal roots, and no system of Sanskrit grammar, however, crude, could go on long without some stock of well-listed roots. It would be natural to suppose that Panini and the predecessors he has mentioned by name, had their own Dhatupathas 140). The Paninian Dhatupatha is the oldest one that has come down to us, and the pattern once fixed thereof is mostly followed by his successors. Here every root is given along with its meaning, these two forming one sentence, which is usually called a Dhatusutra. Generally only one meaning is assigned to a root and is given in the form of an action noun in the locative case; sometimes, however, it is given in the form of a Bahuvrihi compound with. "artha' as the second member, e. g. "gatyartha', sabdartha', etc., this later method being preferred when there are many roots against one meaning. The
