This page has not been fully proofread.

34
 
INTRODUCTION
 
This version is well-determined, differing considerably from the
rest of the archetype, except Y₁. Most remarkable are the omissions of
many otherwise accepted slokas, such as eke satpuruṣāḥ [221] and brahmã
yena [285]. The major differences of order from Y T G M are seen from
the chart to be transposition of yadāsīd ajñānam [6] from Ś 59 to V
101, and another of a whole group of stanzas between V 8 and V 35. Most
MSS derive from Sātārā district, probably even BM 256, and IO Mar.
A. 3. 47, the last being a misch-codex.
 
X¹= Rājāpūr 2; size 8-1/4″ × 4″; fol. 111, l. 6-7, let. 25. Dated A. D.
1755 July 15. The first seven stanzis of N being lost on a missing folio,
the first six of these were collated from the defective ISM Rajavade 51,
which also lacks the 7th. Colophon:
 
शके १६७७ वर्तमाने युवनामसंवत्सरे आषाढशुद्धसप्तमी भौम्यवासरे हस्तानक्षत्रे त्रिप्रहरे भर्तृहरीन्थग्र
लेखन समाप्त श्रीकृष्णार्पणमस्तु । सन्न ११६५ छ ६ माहे शौवाल श्रीराम जय राम जय राम श्रीरामनाम ॥
 
Sātārā
 
X2- ISM Barve 244, from Limb, Satara Dist. Size 8" × 5--1/4"; fol.
77; 1, 18-20, let. 20-23. Written crosswise, leather bound, modern, with some
extra ślokas and many blank leaves. Colophon: zfà užzkatuliandì gufacadei
वैराग्यशतकं संपूर्ण । श्रीसांब सदाशिवार्पणमस्तु । श्रीरस्तु । छ ।
 
Both of these give the Sanskrit text followed by Marathi samavṛtta.
The complete Sanskrit text is given without translation in another MS of the
ISM Barve collection, while RASB 576 is the text of the Niti alone.
 
Version Y TG M: This, charted under Y, is the general version.
current in the Dravidian peninsula, though plenty of devanagari copies have
penetrated for into N territory such as the Nepal MS, supported by Jodhpur
6 of A. D. 1823. Also, HU 2133, copied at Jaipur in A. D. 1705 whose Y type
with division into paddhatis is unmistakable, though Ś numbers are rather badly
upset after $ 19. Bikaner 3277 is another Y MS from the extreme north of
Bhartrhari territory; dated samvat 1900, and complete though corrupt; it is
supported in its order by Hain. 1836 and Bik 1027, which is a misch-codex with
Y type N and a V of N; also by Bik 3285, which has the V alone with northern
interpolations towards and after the end. Finally, Bik 3286 with the same
Niti and [uncompleted] S to 67 is in Mahārāṣṭrian hand, as well as older,
showing that the type was imported. These two minor versions and HU 2133
have not been collated.
 
A remarkable example of its influence is to be seen in Punjab 2885
which, particularly in Niti, is a W type [heavily emended in its readings] forced
into decades by simply omitting the numbers beyond ten in each set. This
rigid subdivision into decades for N and V, with scores for Ś [still further
subdivided into irregular portions] is the main feature of this major version.
Nevertheless, the complete uniformity thereby attempted has not been achieved,
and there is a surprising number of minor versions. Also not collated
are Sringeri 309, an aberrant Nandinagari palm-leaf which gives unusually
many N concordances, apart from strong differences in sloka content; and the
10 Telugu printed edition of 1810 or 1848 (no title page!] which is fully
attested in slokas as well as its queer readings by Mysore KB 340, a Kanarese
MS on European factory paper. The eighth paddhati in Niti is generally
disturbed, perhaps because of an excess over the 100 of ślokas at hand when