¹ The following books and articles are referred to by the authors' names.

Leonard Bloomfield, Language; New York, 1933.

Rhys Carpenter, Folk Tale, Fiction and Saga in the Homeric Epics; Berkeley, Calif., 1946 (Sather Classical Lectures, No. 20).

J. G. Frazer, Taboo and the Perils of the Soul; London, 1911 (The Golden Bough³, Part II).

Louis H. Gray, Foundations of Language; New York, 1939.

A. I. Hallowell, Bear Ceremonialism in the Northern Hemisphere, American Anthropologist N.S.28.1–175 (1926).

Otto Keller, Thiere des classischen Alterthums in culturgeschichtlicher Beziehung; Innsbruck, 1887 (referred to as Keller 1).

Id., Die antike Tierwelt, 1er Band; Leipzig, 1909 (referred to as Keller 2).

A. Meillet, Quelques hypothèses sur les interdictions de vocabulaire dans les langues indo-européennes, Linguistique historique et linguistique générale² 281–91; Paris, 1926 (Collection linguistique publiée par la Société de Linguistique de Paris, VIII; the article was first published privately in 1906).

Holger Pedersen, Linguistic Science in the Nineteenth Century; Cambridge, Mass., 1931 (translated by J. W. Spargo).

G. Richter, Manual of Coorg; Mangalore, 1870.

O. Schrader, Reallexikon der indogermanischen Altertumskunde², 1er Band; Berlin, 1917–23.

² If significant, significant of what? The bear, in one species or another, is or was found over the whole of Eurasia. Does Pedersen mean that the bear's population density was greater in that part of IE territory inhabited by Germanic and Balto-Slavic speakers at the time when the linguistic displacement occurred? This may be so. But it is probably of more significance that, as we shall argue, these speakers lived in an area that was part of, or bordering on, the ‘circumpolar’ territory and thus open to the diffusion of culture traits that were current in that territory. Can we, in fact, presume that the whole of that territory is, or was, distinguished by a high density of the bear population? Culture traits, we are told, are relatively independent of traits of the physical surroundings, short of an absolute nonexistence of a physical trait necessary for the existence of some cultural trait.

³ Bloomfield (400) says shortly: ‘a ritual or hunters’ tabu'. It is in sincere gratitude for much instruction and inspiration that this attempt at elucidation is offered to him.

⁴ I am indebted to Professor Archer Taylor, who called my attention to these modern German and English vocabularies.

⁵ Journal of the Asiatic Society of Bengal 72.3.51. Since this was written, I have received Man in India 27.1 (1947), which contains (44–56) an article by Asutosh Bhattacharyya, The tiger-cult and its literature in Lower Bengal. This article gives more detail, none of which however affects my conclusions.

⁶ For the sources, partly in books, partly in my field-notes, see Lang. 21.184, note 1 (1945).

⁷ Other examples of Kota -ǰ representing -li or -lli of other Dravidian langauges are to be found in my article, A Dravidian etymology of the Sanskrit proper name Nala, University of California Publications in Classical Philology 12.259, note 27. To these may be added: Kota ku ǰ ‘daily wages paid in money’ : Tamil, Kannaḍa, Telugu ku li ‘wages, pay’; Kota neyǰ ‘sticks with which fire is made by twirling’ : Toda nisy; cf. the verbs Tamil ñeli-k-, ñeli-nt- ‘to make fire with firesticks’, Toda nis-, nis-θ-.

⁸ On Toda s in relation to Proto-Dravidian l, see my preliminary account in the article cited in fn. 7 above.

⁹ This was pointed out by L. V. Ramaswami Aiyar, Quarterly Journal of the Mythic Society 22.459.

¹⁰ In work with a Telugu speaker from Rajahmundry at the Linguistic Institute of 1947, I found that his word for ‘jackal’ is nakka. I do not know its history. Is it from *narka < *narr-ka? I can only instance such parallels as ceppu ‘shoe’ < *cerpu : Tamil ceruppu, nippu ‘fire’ < *nirpu : Tamil neruppu, guḍḍi ‘blindness’ : Tamil kuruṭu, emmu(ka) ‘bone’ < *elmu : Tamil elumpu. None of these has i in the second syllable, but the word for ‘rat’ shows i of the other languages replaced by u in Telugu on addition of a -ka suffix: Tamil eli, Kannada ili, Kota eyǰ, Toda isy, Coorg eli, Tulu eli, Telugu eluka, Kolami elka.

¹¹ It would be interesting to know whether the Holeyas (poleyë), the very low-ranking servant caste, who also speak the Coorg language, have the same usage. Nothing, unfortunately, is known in detail about their speech.

¹² For weddings, see my article Kinship and marriage among the Coorgs, Journal of the Royal Asiatic Society of Bengal 4.1.123–47, especially 131–3 (1938). One curious detail, irrelevant here, involved in the killing of a tiger is that after a ‘tiger-wedding’ the Coorg man may grow mutton-chop whiskers.