________________
48
THE INDIAN ANTIQUARY
[FEBRUARY, 1913.
But even taking Paņiņi at the usual Hindu valuation, there are many difficulties in utilizing his sitras in an investigation of Sanskrit pronunciation. His last sátra is "aa" (VIII., iv., 68) and is nsually interpreted to mean that though in the body of the stras vowels have been described to be open (rivrita), short a is not open, but close (sazvita). This information can be utilized only it we know for certain how short a was pronounced by Paņiņi. This letter is pronounced in South India like the u of' but when accented and like the shortened form of the e in her 'wben unaccented. In Northern India when it is unaccented it loses all individuality and practically vanishes. In Bengal and Orissa, the accented a approximates to o. In which of these ways did Pâņiņi intend the sarrita a to be made? This is a question difficult to answer. And then there is the further question, whether these different pronunciations of a are far off reminisconces of the fact that Sanskrit a represents Indo-Germanic a, e, and o. Again in modern Hindt we certainly hear short e and short o. Whence come these sounds?
It is fairly well-known that the Hindus are divided into two great groups, the five Gaucas and the five Drávidas. These groups are distinguished from each other, firstly by the fact that the Brahmans of the former group eat fish and the flesh of " five five-nailed "animals, and those of the latter do not, and secondly by the fact that the Drâvidas pronounce and as sh and y, and the Gaudas in many cases pronounce them kh and j. Thus when they begin words or syllables, there are invariably kh and ; ; jama, jamuná, kha!, púkhar, y in the middle of n syllable is y as in syát; sh when it is the first part of a conjunct consonant is sometimes attempted to be pronouncel, and then it approximates to s, thus shashti becomes khasti. 1, the nasal of ch-series is pronounced alike throughout India, when it preceded ch or j, but when it succeeds j as in the words yajia or juana, it is pronounced differently in different parts of India. The Tamil has in his own tongue a distinct 1 sound, occurring by itself in words, e. 9., fiúyiru but it cannot be easily pronounced after ji bo he pronounces these words as yagha, gñana. The North Indian makes the first word jagya and the second gyána; the Maratha makes the former yadnya.
As regards sibilants, there are four sounds, the English 8, the Tamil é, the English sh, and the Indian sh sounds, all made by the friction of air passing between the palate, beginning from behind the teeth and gradually receding to the mid palate. There is no difficulty with regard to the first of these sounds. The second is the sound made in South India and the third in North India when reading T. Seeing that Pâniņi was a Sindhi, it is probable that he followed the modern North Indian practice. South Indians claim that their pronunciation of this letter is the proper one, but there is no shadow of evidence to prove this, though when « South Indian speaks Sanskrit, the car can much more readily detect the difference between and . But this is perhaps due to the fact that to the South Indian, Sanskrit is absolutely a foreign language, his mother tongue belonging to the Dravidian family and he is therefore plus royaliste que le roi. With regard to the last of these sounds, too, there is a difficulty. The Drávida makes the sound by doubling the tongue, and contacting the blade with the middle of the palate. The Gauda makes a kh of it, Where the South Indian reads tushara, the Gauda reads tukhára. The Gauda and not the Drâvida has spoken Sanskritic languages continuously from the beginning of the historic age in India, and hence his pronunciation must be regarded as the genuine Sanskrit pronunciation and the Drávida one but a modification of it by a foreign tribe attempting to acquire it. The main language of Afghanistan is Pashto in its S. W. parts and Pakhto in the N. E. Here we have over again the Dravida-Ganda difference. The S.W. sh may be due to the proximity of a Dravidian language, the Brahûs. It is to be noted that Herodotus speaks of them Paktues and the Rig Veda refers to them as Pakthas. Apparently Pakhto was the ancient form and Pashto a recent one. This fact rendere it probable that was kh in Sanskrit till the Drávidas made it into sh. This view will react on the discussion of certain problems of linguistic science. Collitz derires ksheti from a root kshei and kshayati and kshináti, both from a root ghehee. But it is a disputed question whether the Indo-Germanic had a sh sound. If, as with the Gaudas, Sanskrit is really kh and Iesh is really kilch and if y developed from Iudo-Germanic k onght to be pronounced sh, the above disputed question ought to be rediscussed in the light of this. As an example of a mis