This page has not been fully proofread.

INTRODUCTION
 
58
 
no. 6 the Kavyasamgraha] not one is in Bengali, the nearest being
5592, an Oriya palm-leaf of the Vairagya. This, like my Yr, is of peninsular
[YTGM] type, being divided into paddhatis; it follows stanzas from
the Santiśataka, ascribed by a colophon to one Simhunadeva, though eleven
(of 24) stanzas belonging to that section also survive and have in common
with Bhartrhari our stanzas nos. 190, 218, 301. The remaining are devanagari
paper MSS, of which 7747 is a poor E type probably from Rajputana like
our Az and perhaps 7743, 7804. Surprisingly many show Mahārāṣṭrian
influence. Nos. 9356 and 9937 [ both of N] belong actually to the W version
and 576 [N] is still more clearly of Maharastrian provenance, for it is
of the version X, localized in Satara. Benares 57-4 is a modern Bengali
copy of the fragment of the devanagari Niti preservod in Benares 60-10,
and the version is again from Räjputana, represented perhaps by Bikaner
3280. IO 14196 gives e śrngäraśataka in modern Bengali, and is a late
copy of a W original, while IO 1151 is a V of northorn type, with some
features common to my version C, but generally mixed; copy in Bengali
on modern European paper. Three MSS. one only of each śataka, were
reported at the Vangiya Sahitya Parishat, and these upon examination
turn out to be of type generally curient in Rajasthan. One of my conclusions
upon studying MS evidence available to me was that Bhartrhari roached
Bengal late, often through Maharāṣṭrians, from some base at Benares. This
was confirmed unexpectedly by the researches of P. K. Gode, who showed
[Ind. Culture XII, 1945, 47-56] that in the 17th century there existed
settlements of Maharāṣṭrian Brahmins in Bengal, even East Bengal.
Moreover, these were based upon Benares, and had originally migrated
for deeper study of tarka and nyaya works. In any case, the contrast is
undeniable when we compare the absence of complete satakatraya MSS
with the innumerable copies of the Säntisataka found in Bengali script.
 
Rajendralal Mitra's Notices mention only three Bhartṛhari MSS.
No. 706 fearfully misdescribes the Padyasamyraha containing our Y₁ as
a work on smrti rites. No. 2837 notices a fresh Bengali MS with
commentary of "Maheśvara", but actually the N from our version W; notice
3325, unless Mitra made another of his colossal mistakes, reports a
commentary by Sri Minanatha, which happens to be the name of a
Nathapanthiya pontiff.
 
This negative evidence is again to be supplemented by reference to
anthologies. The oldest published was the so-called Kavindravacana-
samuccaya [F. W. Thomas, Bib. Ind. Calcutta and London 1912], definitely
an eastern anthology, perhaps of the 12th century; it does not cito Bhartrhari
by name, and contains very little that is definitely Bhartrhari's. I submit
that this is the same as the RKB., an eastern anthology of the 11th century.
The fragment published by Thomas must have been about a third of a
later copy of a Ngor type codex. For this dating, and for finding the
citation Bhartrhareh (from atrocious photo-duplicates) on fol. 36%, I am
indebted to Prof. V. V. Gokhale. Now the remarkable fact is that RKB.
citations follow the stanzas, while of the five stanzas (succeeding the name)
which end the (penultimate) section, three are 6, 302, 218 of this edition