________________ REVIEWS 299 To conclude this review I venture to submit the following remarks which are meant at the same time as a tribute to Mr. Norman's fine scholarship and as a small contribution to the study of this important text. The first pada of 9 (=885a) has only seven syllables: svagatam napagatam. CPD proposes to add a svarabhakti vowel (see s.v. apagata). Mr. Norman admits svarabhakti .vowels in dvara and tvam (p. lxiv), but does not state his reasons for rejecting the same solution in this case. Verse 16 speaks of a bhaddo ajanno nangalavattani sikhi. The commentary explains that ajanna can refer to a bull, a horse or an elephant. In this verse the commentary takes ajanna to be a bull. However, Mr. Norman remarks that sikhin does not apply to a bull. He proposes to solve this difficulty by assuming that the thoroughbred is a horse and that narigala means here 'tail' or is a mistake for nangula. However, according to the Medinikosa sikhin can also refer to a balivarda. Perhaps, in this case, sikha means the tuft on the belly of a bull. Mr. Norman always translates ajanna by 'thoroughbred'. I suppose that in 173 and 659 he takes it as referring to a bull. In both verses Mr. Norman translates dhura by 'load'; 173 vahate dhuram 'draws its burden'; 659 dhure yutto dhurassaho 'yoked to a load, enduring a load'. According to PED dhura is used figuratively in the meaning of 'burden, load, charge, office, responsability'. This may be true or not but, in any case, it seems preferable to translate dhura in these two verses by 'yoke' (cf. also 359 viriyadhuraniggahito where dhura is used figuratively: 'restrained by the yoke of energy'). In note 22 Mr. Norman suggests taking jhayam as a namul-gerund. He quotes other examples of namul-gerunds in Pali. One of these: jiva-gaham ca nam aggahesi can better be taken as an example of a cognate accusative (cf. J. S. Speyer, Sanskrit Syntax, $ 44; L. Renou, Grammaire sanscrite, p. 289). To these gerunds can be added alumpakaram which has to be read in Dhp-a II.55,22: ubhohi hatthehi alumpakaram gutham khadi. Norman's edition has alumpakaragutham, but one of his manuscripts (C) has the reading alumpakaragutham.CPD quotes this passage under alumpakarakam but states wrongly that manuscript C reads alumpakarakam gutham and that Norman's edition has alumpakaragutham. CPD does not refer to BHS alopakaram and alopakarakam both taken as gerunds by Edgerton (BHS Grammar 22.5, 35.3, 35.5; Dictionary ss.w. alopa and -karakam). Cf. also Pali sannidhikarakam, BHS samnidhikaram (Mahavastu 1.343.18). In verse 55 a new interpretation for asandim is proposed by Mr. Norman who takes it as the 1st sg. aorist of a form a-sad- showing a nasal infix. A Sanskrit version of this verse is to be found in the fragments of the Sthaviragatha published by Heinz Bechert (op.cit., p. 263). Instead of asandim this version has asannam which probably resulted from a misunderstanding of asandim. Mr. Norman remarks that in the first line of 55 a finite verb is missing. However, the second line is a cliche which occurs often in the Theragatha (24, 66, 107, 108, 220, 224, 286, 562, 639, 886, 903). In two verses (117, 349) the first pada has tisso vijja ajjhagamim for tisso vijja anuppatta. If one replaces in 55c anuppatta by ajjhagamim, a finite verb would not be required in the first line. The Sanskrit version of the second line is slightly different: tisro vidya maya (praptah krtam buddhasya sasanam). This shows that alternative readings existed in the second line. In 104 lahuko vata me kayo phuttho ca pitisukhena vipulena, Alsdorf suggests reading piti m.c. and deleting ca, in order to restore the Arya metre. However, instead of phuttho the original reading must have been phuto 'suffused', cf. 383 pitiya phutasariro. Therefore ca has to be maintained. In 305 sugatavara is rendered by the best of the well-farers'. The commentary gives 4 CPD quotes variant readings from Norman's edition, using the same signs. This is likely to create confusion. For instance, according to the system of CPD S indicates a Siamese text whereas Norman's S refers to a Sinhalese print.
