This page has not been fully proofread.

32
 
INTRODUCTION
 
S: The Southern and Western Recension
 
This is shorter than N, and more logically arranged by paddhati
divisions that attempt to group together stanzas of similar content.
Individual readings are generally much smoother, and fit Papinian grammati-
cal rules.
 
Archetype Y.
 
This consists essentially of a single western [Mahārāṣṭrian ] Version W
which has as its feature a deficient Vairagya of 86 [properly 85] ślokas,
though plenty of inflated copies exist which are to be discounted for lack
of uniformity, repetitions of previous stanzas and, in general, absence of
the commentary [though this is sometime variously written in]. Even
with an N of 110 stanzas, the total is 296. Moreover, there is every reason
to suspect the last eight of N as later additions, for the final karmapaddhati
ends at bhimam vanam = N 102; the rest belong to no paddhati at all, and
are almost always labelled "pratyantare slokāḥ", stanzas from another version.
The last two of Ś may also be added later. The original version must,
therefore, have been one of the shortest, hence oldest, preserved. Though
it is valuable for testimonial purposes as the sloka number and order are
well-determined, the actual readings are almost useless, for most copies are
corrupt or heavily emended. The commentary is so sketchy that when it
exists at all it is difficult to distinguish a divergent commentary reading
from gloss in important cases; the most plausible interpretation of this
divergence is that the original readings agreed better with N, but the text
having been modified by more olegant versious of S, these readings now
survive only in the commentary.
 
The commentary itself is anonymous in the several dozen W MSS I
have been albe to examine personally, though it has been ascribed to
Maheśvara on the colophon found in a single fresh Bengali copy of the
Nīti, as described bp Rajendralal Mitra, Notices IX. 2837. I suggest that
this is incorrect, possibly by confusion with our Y₁ commentator. In the
first place, the center of the distribution is unquestionably Mahārāṣṭrian,
most of the numerous Bhartrhari MSS found in Mahärāṣṭra belonging
to this version. Further, wherever we can trace the provenance of W MSS
from other parts of the country, they show Maharāṣṭrian associations as for
example the Benares collection, made by Maharaştrian pundits, as also the
Ujjain and SVP lot; Baroda 787 was, according to its colophon, copied at
Devagadh-Baria in Gujarat for one Narayana Kaceśvara Ekabote; another
at Chindwārā in C. P. comes from the collection of Prabhakar Śāstrī Rode.
Incidentally, PU 516, whose commentary Poleman misinterprets as being
by Bhartṛhari himself, is merely the ordinary W Srigara. Finally, even the
mistakes are characteristically from Marathi, as 177° sa tu bhavatu daridrī,
and 325ª vistīrņe sarvasve, which occur even in our Telugu W₁ and Śrügeri
9/15, which is a devanagari paper MS copied by some Gujarati scribe. 1SM
Kaļamkar195, Punjab 2885, and RASB 9510 are misch-codices, preponderantly
 
W.
 
All the MSS collated have both text and commentary, being:
 
W₁= Anandāśrama 6384; size 10-1/2" x 5"; fol. 63, numbered separately