^{page 459 note 1} ZDMG, XXXVIII, 1884, 610.Google Scholar

^{page 459 note 2} Hermann Oldenberg, Die Hymnen des Ṛigveda, I, Metrische und textgeschichtliche Prolegomena, 1888. See pp. 58 ff.

^{page 459 note 3} Arnold, E.V., Vedic metre in its historical development, 1905.Google Scholar See sections 42–8.

^{page 459 note 4} Arnold, p. 13. The stanza is Rv. iv.50.4.

^{page 460 note 1} See p. 461 for this note.

^{page 461 note 1} 600 trimeters; being Rv. i.85, U54, vii.61, vii.63; Rv. ii.26, ii.33, ii.35, iii.59, iii.62, iv.50, iv.51; Rv. i.32, vii.82, x.129; Rv. v.83, vii.49, x.15, x.168. The references have been arranged in four groups to indicate the periods to which they belong on Arnold's view: strophic, normal, cretic, and popular. No ‘archaic’ hymns are represented in the samples. The number of pādas in each group is: strophic 124, normal 244, cretic 100, popular 152.

^{page 461 note 2} About two per cent find no place in the grid. See the next paragraph for these.

^{page 461 note 3} Unfortunately dependent on an arbitrary serial order, particularly of the openings. Order of frequency was the leading principle: but other considerations modified this, in the arrangement of the openings.

^{page 462 note 1} Nos. 6 and 7 are of the type which associates an ‘iambic break’ with what Arnold calls a ‘secondary caesura’ after the eighth syllable; chiefly found in the Vasiṡṭha hymns (Rv. vii). Word-end at the eighth syllable is (Arnold, p. 180) ‘in all other parts found in about 35 per cent of the verses … but in the Vasistha hymns in no less than 57 per cent’. The association of break E₅ with this ‘secondary caesura’ is more than twice as common in the Vasiṡṭha hymns as elsewhere in the Rv. (p. 181): and Arnold calls verses of this type ‘Vāsiṡḥhī verses’. Thus No. 6 is a Vāsiṡḍhī verse. The fact that both Nos. 6 and 7 have short eighth syllable connects with the fact that both have word-ending at this syllable: for, as Arnold states (p. 185), a short eighth is twice as common when that syllable is final as when it is initial or medial. (Short eighth final vowels are more numerous by half than short eighth closed syllables, as appears from Arnold's list of instances of short eighth in his section 220, pp. 202–3: this is important in view of the divergence of the saṃhitā and the pada text in their treatment of the length of final vowels.)

^{page 463 note 1} Arnold gives the name Bhārgavī to a pattern normally hendecasyllabic because it has the jagatī ending instead of the triṡṭubh ending characteristic of decasyllables; and further differentiated from the decasyllables by having caesura either at the fourth or at the fifth syllable. It is used notably in Rv. x.77 and 78, ascribed to Syūmaraśmi Bhārgava.

^{page 463 note 2, 3} Arnold's theory of the rhythm of decasyllables is that they are triṡḍubhs with a gap or ‘rest’ replacing a syllable either before or after the caesura. The rest therefore fills a ‘place’ or position in the triṡḍubhs pattern, and he always includes it in his count of syllables, as if it were an actual syllable. He marks the ‘rest’ by a point. Since the ‘rest’, in the present example, follows the caesura, his scansion would be and the actual fifth syllable is for him the sixth.

^{page 464 note 1} I use the term syncopation in the sense usually given to it by metrists, that of syncope or the disappearance of a syllable between two other syllables. When Arnold calls ‘trochaic’ and other openings ‘syncopated’ he uses the term in the musical sense of displacement of beat or ‘ictus’.

^{page 464 note 2} This case is in fact recorded in the ‘grid’ (constructed several years ago) as the sole example of 5F₄ (reading, with Macdonell, náinam áṃho'ś;noty ántito ná dūrt).

^{page 464 note 3} Arnold records 27 cases of prolonged cadence (pp. 208–9 = section 224) and 35 cases of prolonged opening (pp. 215–16 = sections 229 and 230). He also notes, as ‘requiring to be removed by emendation’, four apparent cases of hypersyllabic breaks (section 219, p. 202).

^{page 465 note 1} The fifty examples found in Mbh., Sabhā, retain much of the freedom of the Ṛgvedic pattern, but they avoid the short fifth syllable and slightly prefer the heavier D-type centre. The distinctive character of the Sabha triṡḍubhs was pointed out by Edgerton, Franklin, JAOS, LIX, 1939, 159–74.Google Scholar

^{page 465 note 2} See his chapter xiv In his example (another verse of the same poem) the fifth syllable happens to be long anceps, and he scans the line calling it The alcaic hendecasyllable is related to the sapphic by according to Dale, A.M., ‘The metrical units of Greek lyric verse. III’, Classical Quarterly, NS, I, 3–4, 1951, 127–8Google Scholar: that is, the final anceps of the sapphic becomes the initial anceps of the alcaic.

^{page 465 note 3} Nor is there any catholic virtue in Taktgleichheit, divisibility into equal ‘feet’. It is one method of introducing measure into patterns, the method of mātrā-vṛtti, but not the only method.

^{page 465 note 4} For Rv. i.120.1–9 see Prātiśākhya xvi.20, 40, 41, 49, and 44. Arnold regards the text as corrupt, and rewrites these verses in a more even form (pp. 232–3).

^{page 466 note 1} More than one in every six pādas in the puruṡa-sūkta, Rv. x.90, is a ‘glyconic’.

^{page 466 note 2} See the table of triṡḍubhs patterns, Nos. 5A (six cases) and 6A (three cases). Two of the cases occur in one stanza, Rv. vii.63.3b and c: eṡā me deváḥ SavitF181; ccKhanda yáḥ samānáṃ ná, pramināti dhma. Hephaestion xiv calls the sapphic and the Pindaric hendecasyllable alike ‘epichoriambic’: a sufficiently appropriate name for the A-type triṡḍubhs (with choriambic centre).

^{page 466 note 3} See the table, No. 11A. Hephaestion describes it as an epichoriambic trimeter with an antispast as first metron (in precise correspondence with the analysis taken as the basis of the above table. The table records three occurrences, or a percentage of one half; e.g. Rv. iv.50.4b mahó jyótiṡaḥ paramé víoman().

^{page 466 note 4} See in the grid 8E, 11E, and 12E:

Rv. iv.50.3b.

Rv. i.88.6b. prátiṡṭobhati vāgháto ná vnī

Rv. vii, the refrain found in many Vasiṡṭha hymns, e.g. vii.61.7d:

yūyáṃ pāta suastíbhiḥ sádā naḥ

^{page 466 note 5} Kühnau, E., Die Trishtubh-Jagati Familie, Göttingen, 1886.Google Scholar

^{page 467 note 1} The undetermined or indifferent syllable should not be called syllaba anceps, because it obeys no laws, whereas anceps is used under restrictions. See Dale, , ‘The metrical units of Greek lyric verse. I’, CQ, XLIV, 3–4, 1950, 143.Google Scholar There is (perhaps) an example of metres with indifferent syllables in the Kanarese types described in chapter vii of Jayakīrti's Chandonuśāsana, printed in Velankar's, H.D. Jayadāman, Bombay, 1949.Google Scholar See also Professor Velankar's article, ‘Jayakīrti and ancient Kannada metres’, JBBRAS, XXI, 1945.Google Scholar

^{page 467 note 2} Statistics are not available in Arnold, since he accepts the view that the first syllable is indifferent. See his table of varieties of trimeter rhythm (p. 188).

^{page 468 note 1} ‘The long-short sequence is called samānī. The short-long sequence is called pramāṇī. A pattern is different.’ I suspect that Kālidāsa's closing puṡpitāgrā stanza in Raghuvaṃśa vi has a śleṡa based on a memory of Pingala's vitānam anyat:

pramudita-vara-pakṡam ekatas tat

kṡitipati-maṇḍalam anyato vitānam

uṡasi sara iva praphulla-padmaṃ

kumuda-vana-pratipanna-nidram āsīt.

The secondary meaning is that the opening runs in unison with the rest of the pāda, whereas being a pattern, runs differently.

^{page 469 note 1} I follow Miss Dale's analysis. See her concise statement at p. 142 in the Classical Quarterly, XLIV, 3–4, 1950 Google Scholar (section 2 in the article cited above).