^{page 69 note 1} ´ ´zil, L'ideologie tripartie des Indo-Européens (Collection Latomus, Vol. XXXI), Bruxelles (Berchem), 1958.Google Scholar Where no other indication is given, references to pages in the present article are to the pages of this book.

^{page 69 note 2} This summary of a summary is given merely for convenience of reference. In the interests of‘fair play’ (p. 104, §dot; 5) it must be said that the summarization, though carried out con-acientiously (and, I hope, ‘loyalement’), is no more than it claims to be. The quotation of pusages out of context, and without detailed discussions of evidence, may lead to distortions rf perspective. The reader is therefore warned of possible bias, and is referred to Professor Dumézil's own writings.

^{page 69 note 3} Since the English word ‘function’ would be more than a little misleading, it has seemed advisable to retain the French term for this technical usage.

^{page 70 note 1} For convenience, the roman numerals I, II, III will be used as labels to indicate the three ‘fonctions’, or, where appropriate, any correlating hierarchization.

^{page 71 note 1} The correlation, however, is not in every case equally clear. It would require more ingenuity than I find myself capable of to correlate the ‘fonctions’ with the Indian ‘aims of existence’, dharma artha kāma, quoted on pp. 24 and 97 in the unusual order dharma kāma rtha, and translated as ‘la loi morale, la passion, l'intérět économique’. This still seems to me to be I, III, III—‘fécondité’ is given pride of place in the characterization of III (p. 19). The point can hardly be simply ‘the lascivious soldiery’, nor, specifically, the notorious libertinism of Indra, since, although Indra himself is II, his adultery is held to be a sin against III (p. 30); and we would not in any case wish to believe that sinful behaviour could really be a purusārtha. Perhaps this is one of the cases where we have to admit ‘quelque flottement dans la représentation ou définition des trois fonctions ’(p. 32).

^{page 72 note 1} In an earlier book, L'héritage indo-européen à Rome, 1949Google Scholar, the condemnation of the argument from ‘la nature des choses’ appears in less forthright terms, and it is even admitted (p. 239) that ‘la tripartition consciente et explicite de la société ou de la partie directrice de la société en pr^tres, guerriers et agriculteurs n'est pas propre au monde indo-européen’. But the final conclusion is the same (circ;ibid., pp. 240–1): ‘Trés précisément, si Ton considere la portion d'humanité déjà vaste que connaissaient les anciens, on constate qu'un système triparti conscient et explicite ne se rencontrait que dans des civilisations où une puissante composante indoeuropèenne est incontestable’.

^{page 75 note 1} See p. 106, § 19, for bibliographical references. I have not, however, been able so far to trace a Hebrew counterpart of the ‘dieu borgne’ (Ódinn, Horatius Codes). One is naturally tempted to think of the blinded Samson, but this would probably be a mistake: see Dumézil, , Déesses laiines et mythes védiques, p. 97, n. 2. In this note we are promised a study of the ‘dieu sonverain aveugle’, and it will be interesting to see whether this study will throw light on the ‘vakur qualificatrice, fonctionnelle’ of the blindness of Samson. It is important to emphasize the fact that in these studies the mere absence of a particular element is never significant, since we must naturally always make allowance for loss and change with the passage of time.Google Scholar

^{page 77 note 1} The Indian ‘MÄnnerbund’ is attached to Indra (II), and the Germanic to (Ódinn (I, type Varuna) (p. 95, § 10); and in Rome, as we have seen, we find a second set attached to Mucius Scaeuola (I, type Mitra).

^{page 80 note 1} In L'héritage indo-européen á Rome, p. 240Google Scholar, Professor Dumézil actually noted that all three ‘fonctions’ could be attributed to David: ‘Le petit berger David (III) tue le champion philistinsur la ligne de bataille (II) et bientôt il sera 1'oint du Seigneur (I) ’. It is difficult to see therelevance of this example to the general proposition which it purports to illustrate: ‘Ce quidomine 1'idéologie (sc. dans la Bible), c'est bien plutôt le sentiment de l'omnivalence—moyennantla volonté divine—de chaque être et de l'équivalence de tous’; since there are after all not manypersons in the Biblical narrative who could exemplify a similar omnivalence in respect of allthree ‘fonctions’. What is surprising is that David is held to be disqualified on this score froman interpretatio trifunctionalis, although Romulus, who was also shepherd, warrior, and king Bulletin of the School of Oriens accepted without difficulty. Professor Dumézil has supplied his own comment on this type ofprocedure: ‘Nous voici de nouveau en pleine philologie du bon plaisir’ (Déesses latines et mythes védiques, p. 21).Google Scholar

^{page 82 note 1} It is naturally tempting to see in the double representation of II a reflexion of Indra and Vāyu.

^{page 83 note 1} L'héritage indo-européen à Rome, p. 1 UMritage indo-europeen a Rome, p. 240.Google Scholar

^{page 84 note 1} 'L'héritage indo-européen à Rome, pp. 35–6.Google Scholar