page 132 note 1 Cf. Pillai S. Anavaratavinayakam, The Sanskritic Element in the Vocabularies of the Dravidian Languages, pp. 27–48, Madras, 1923 (Dravidic Studies III).
page 132 note 2 2 The borrowing could be the other way round.
page 133 note 1 In the case of this word it is very likely that Indo–aryan has borrowed from Dravidian. Tamil and Malayalam hardly ever substitute –r–, a peculiarly Dravidian sound, for Skt. –1–, and Sīhala–, Simhala– could easily have been adapted by popular etymology from an original Dravidian form *dram. It is to be presumed that the Dravidians of South India had a name for Ceylon before the coming of the Indo–aryans. The following words need to be considered in this connection: Ta. tjram toddy, Ta. Ma. īravar the caste of toddy–drawers in Tinnevelly, Travancore, and Malabar, Ka. īdi toddy, īdiga a man of the toddy–drawer′s caste, Te. īdiga the toddy–drawer caste, indravadu a member of that caste, idiga id. The iravar are said to have migrated from Ceylon in the remote past and to derive their name from iram Ceylon. This tradition may be only a deduction from their name, since it is impossible to separate Ta. Ma. irava– from Te. īdiga, etc., and the latter caste has no tradition of Ceylonese origin. The words are most simply derived from Ta. īram. toddy; possibly īram Ceylon was named from īram toddy, after the palm–trees with which it is well stocked and the toddy produced from them. An earlier form of the word would be *ciram (whence Sīhala–), the c– being lost later as in the many examples quoted below. It should be noted that Kanarese toddy–drawers are Telugu immigrants, and the Kanarese words loan–words from Telugu.
page 134 note 1 BSO., VIII (1936), p. 703.
page 135 note 1 Linguistic History of Certain Dravidian Words (J.D.L., Calcutta, 1929), p. 7 n.; JORM., iv (1930), p. 171; JA., vol. 61, p. 29.
page 135 note 2 S Dravidian, Amer Journal of Philology, vol. 40 (1919), pp. 76–84.
page 136 note 1 Cēra Kings of the śangam Period, p. 18, London, 1937.
page 139 note 1 Ta. utai garment according to this etymology represents *cutai, and with the usual changes would have appeared in some areas of Dravidian as *sode. This could be the origin of Skt. śāta (ka–), etc. Since Skt. has no short –o– it has to choose some other sound to represent it in Dravidian words. Here it is –6–, more commonly –ā–. Skt. śāfa is then borrowed into Dravidian: Ta. ātai, etc.
page 141 note 1 These Dravidian words are reminiscent of Skt. RV., etc., sīra– n. plough. J. Bloch (BSOS., VIII, 414) draws attention to this, but decides against a derivation from the Dravidian of the Skt. word in favour of an IE. etymology (*sēi to sow). In favour of a Dravidian origin is the frequent use of yuj– to join with sīrā– (yunakta sīra RV., x, 101, 3, saiyogam sīram AV., vhi, 9, 16, etc.), since Dravidian ēr, sēr has just the meaning and etymology which suits these contexts. The fact that Skt. lāngala has been borrowed from Dravidian (BSOAS., XI, 131 and 603) also favours this view. The main difficulty is vowel (Skt. ī Dr. ē), as Bloch says. Such a change does indeed occur in Brahui (khīsun red: Kur. khēso), but there is no reason to believe that this is ancient.
page 143 note 1 This is one of the few examples where the Northern Dravidian languages (Kurukh, Malto; Brahui) preserve an original guttural in contradistinction to the Central and Southern languages (with affricate or sibilant). The discussion of this phenomenon must be reserved for a future occasion. This special origin of c–/s– in a certain number of words does not seem materially to affect their treatment in the South Dravidian languages.
page 146 note 1 The change of s– to h– in Kuvi and dialects of Gondi (Kuvi hāru salt, etc.; Go. (Maria dial.) hovar id.), and the occasional dropping of this h– (Go. Mar. ovar according to Lind) is a modern change of the same kind but independent of the earlier S. Dr. change.
page 146 note 2 Geiger W., Grammar of the Sinhalese Language, pp. 82 ff.
page 146 note 3 Dravidian Studies III, BSOAS., Vol. XI (1943), pp. 122 ff.