The Sanskrit consonants, as single sounds, remained in princi ple unchanged, except that all sibilants : S, s, s were confused un der one form, s. On the other hand consonants in contact with each other were liable to change, a stop being assimilated to a following stop (e.g., satta, duddhatit from Skt. saptd, dugdhcim) and a continuant to a following or preceding stop (e.g., akko, aggi from Skt. arkdl agni1J). If the assimilated consonant was a sibi lant, the resultant group was aspirated (e.g., atthi, vaccho, acchi from Skt. dsti, vats4, ciksi). Lastly, all final consonants had dis appeared (e.g., vijju from Skt.
Its grammatical forms are clearly derived from Vedic Sanskrit. But the process of simplification and normalization has proceeded much further. In the declension of the noun the dual has disappeared, leaving only two numbers, singular and plural. Among the cases the dative has almost lost its separate existence, its place being taken by the genitive; and in general, chiefly through the action of sound-change, the number of cases with separate terminations has been greatly reduced. The declen sion in the singular of a feminine stem in -a illustrates this: Sanskrit Pali Sanskrit Pali nom. kanya kannd abl. kanyayah kafiriaya
acc. kanyam kafifiaM gen. kanyhyah
inst. kanyayd kannaya loc. kanyayam kafinaya dat. kanyayai kannaya voc. lednye kaiiiie.
The changes in the verbal system are still greater. The athematic stems of Vedic (in which the termination is added directly to the root) have been mostly replaced by thematic stems, in which the insertion of the vowel a preserves the individ uality of both root and termination (e.g., ledhi, lihdnti have
been replaced by leh-a-ti, leh--&-nti). The middle voice is in process of disappearance. Among the moods, traces are still found in the oldest stratum of the subjunctive, but like the im perative and optative it is confined to the present stem. Of the tenses, the perfect has ceased to play any part in regular conju gation, while the imperfect and aorist have combined into one tense. The use of participial phrases in place of finite verbal forms is increasing.
In general the numeral system is that of Sanskrit, alterations in form being due to sound-change, rather than to the introduction of any new principle.
Gender is grammatical and is divided into three categories : masculine, feminine and neuter—following the sys tem of Sanskrit. But there is already some disturbance of the system. The neuter of the numerals for "two" and "three" is used sometimes with masculine and feminine nouns. A growing confusion between masculine and neuter stems presages the loss in most of the modern languages of all distinction between these genders.
The vocabulary, though based primarily on that of Vedic Sanskrit, shows changes and developments of meaning and the admission of new words. Some words of the modern languages make their first recorded appearance in Pali, e.g., kaddhati, "drags" (replacing Skt. kcirsati) appears to-day in Panjabi kaddhe, Hindi karhe, etc. Later the growing influence of literary Sanskrit caused the inclusion of numerous new words of that language disguised under a Pali form.