between short u and shorter u is that the former is rounded and latter is unrounded. T. P. Meenakshisundaram considers this shorter u as “a mere vowel release’ [1965 : 67]. Pope's remark that it is pronounced very lightly means the same thing. That may be the reason for his transcription of the word ma: tu without the final vowel. The second part of the statement that it is often turned into i by the common people may represent a change of -u > i when preceded by -i. This feature is found in the colloquial speech even today. “u : like u in rule [ru : 1 ) or oo in school [sku : 1 }” Ip. 6] i.e. high back long vowel. “o and o : like o in opinion [o pinijan] and opium soupfam] respectively” (p. 6]. Here also, we are not able to get any clue for the qualities of these vowels from the modern phonetic transcriptions of the evamples given as in the case of e and e :. 1.4. DIPHTHONGS equates with â'i or ài “ai as in aisle [ail] and ai in [meid]= â * or ai” (p. 6]. The transcriptions give two values [ail and [ei], which have already been noted by Beschi [1848 : 13]. But the significance of Pope's equation of the initial vowel in ai with short and long a is not clear. “ai seldom has its fu' sound, except in the first syllable of a word. In other cases, it is often sounded e : y, only differing from e : in the slight y sound in which it ends. In the middle of a word, it is often sounded almost like a” (p. 12]. Therefore. there are three values for ai: ai in the initial syllable of word, a medially and e : y finally. These cannot be taken as allophones of ai. “au is often written av or avu, thus, maunam “silence' is written mavunam. This letter does not really belong to the language and was introduced, apparently under Sanskrit influenees’ auvai'a certain poetess' -avvai [p. 13]. So, au need not be considered as a scparate sound. 273
