________________
OCTOBER, 1919]
MISCELLANEA
(see Burnell, South Indian Pal., p. 51, n. 1; Caldwell, p. 10); the cosmographer of Ravenna. records the name as Dimirica. Now, as Dr. Caldwell has justly observed, the transmutation of dr-into d- is Prakritic (cf. also the Pali Damilo in the Mahavamso). How great is the probability that a parallel transformation has occurred in Tamil itself! In any case, it is ineonceivable that, when the word Dravida made its appearance in Sanskrit, it was not a transliteration of an authentic indigenous word. Whereas it is impossible to suggest any previously existing Sanskrit model on which an indigenous word more closely resembling the surviving Tamil could have been moulded into Dravida.
If we now come to consider the chronology of the processes considered above, we may first state that the simplification of the old grouped consonants must have occurred at about the same date in Tamil as in Indo-Aryan languages; at least, if the testimony of the geographers' records authenticate local usage and not forms belonging exclusively to the Indo-Aryan dialects which borrowed Dravidian place names.
195
As for the surding of sonant consonants, we have seen it proved for the initial by the modern pronunciation tamil, and confirmed for intervocalic consonants by the testimony of Kumarila, and, above all, generally by the absence of sonant symbols in the alpbabet of a language which possesses voiced sounds now and which no doubt had them in prehistoric times also. So we may conclude that this loss of sonority must be sought for somewhere between the beginning of the Christian era and the time of Kumarila.
At what date, then, did the surds thus obtained again become sonants, as they now are, between vowels? We do not know. But we may infer that the change is comparatively recent. The Nangal, written about A.D. 1200, (see Barnett, Cat. of Tamil Books in the British Museum, preface, p. III) still inculcates (III, 20) that in the transcription of Sanskrit words, the first letter of each varga represents the three following letters (for example the letter k does duty for k, kh, g and gh, not only without distinguishing sonants from surds, but also without indicating any difference of sound due to the place of the letter in a word). Besides, the existence of doubled consonants is expressly recognized (II, 55), but without the faintest allusion to any difference in articulation. Finally, the doubling of the initial letter of the second members of compound-words (IV, 15ff.), although it may seem to indicate a difference in pronunciation between the initial and the intervocalic consonant, is by no means conclusive, even on that point. Its occurrence may depend on various conditions, among them the nature of the preceding sound (compare Ko-pparakesarivraman as opposed to madirai-konda in the inscription of Nandivarman the Pallava, VIIIth century; see Hultzsch, South Indian Inscriptions, II, p. 370); and it is easy to conceive a stage in the history of the language, (whatever be the future alterations), when the initial consonants may have been uttered with a special stress; this would not imply that the intervocalic consonants were necessarily weaker. So we may admit that in A.D. 1200 there is not yet any clear trace to be found of the change in question.
MISCELLANEA.
PALESIMUNDU.
The Periplus applies the name 'Palasimundu' to the island which was called by the ancients' Tapro. bane' (Ceylon). Pliny knows the name (VI, 24). Ptolemy too notes that the ancient name of the
island was Simundu. According to Lassen the word Palasimundu' is the Sanskrit Påli-stmanta, "abode of the law of piety"; i.e., the Dharma of Gautama Buddha. (See Schoff's Periplus, p. 249 1 This view though ingenious is far from satisfactory.
•
