This page has not been fully proofread.

INTRODUCTION
 
drop out of Group [. Besides culotlamsita-[1], the only possible mangalacaraṇa
for Bhartrhari, we have only one käki nihsprhah 185 which places faith in
Siva. Neither is particularly heavy Vedanta, and such casual mention shows
merely that there was a peg upon which Vedantie garments could be hung for
Bhartrhari by later redactors. On the other hand, the mere existence of Saiva
stanzas in versions copied and commented upon by Jains show that the
Vedantic accretions come early enough to be well established. Changes like
sajjanānām or punyabhājām for brāhmaṇānām ju 66ª, jano for jino in 224",
arhati for sūlini in 43%, and an occasional addition like stanza 219 is all that
can be assigned to Jain influence, which does not tamper with the Saiva
 
verses,
 
*****
 
63
 
The evidence for a single defective copy being the source of half
our manuscripts rests upon the extraordinary phenomenon of two different
second halves for the stanza asarāḥ [83]. Which reading it has depends
upon whether it is followed by the stanza bhavanto vedānta-[287*] or
not; if not, the second half is that which normally goes with 287*. This
is immediately explained as a case of haplography, for both second halves
begin with tathapy. I maintain that this could not have happened in more
than one manuscript, and even then only at that remote epoch when copies
of Bhartrhari were rare, with very few interested in reading or comparing
the version with others-if any others then existed. One should note that
Krsnasastri Mahabala's NSP edition omits 83 altogether, BU 114/7 has
both 83 and 287 with the same ending; Mysore 582 has, (like Bik 3277,
almost unique among YTGM MSS) both of these in the A succession,
though it otherwise follows M1.5. Occasional MSS like Bikaner 3280, PU
496, HU 2145 have the proper succession of these two stanzas, but with
the second in order cdab. Finally, the haplography has actually begun in
HU 196 but is corrected on the margin. A similar double ending for 234
cannot be explained; but the stanza is not of Group I.
 

 
There is secondary confirmation for this grouping in that the
clearly northern H version gives on occasion southern readings while the
southern W agrees with extreme northern MSS against others of its
 
recension.
 
3. 2. Omissions and inclusions. Just how and when the satakatraya
idea started is hard to determine, but it must have been fairly early.
Recension S is more logically arranged than N, whence it should be the later;
at best we may see in N a slight tendency to group two or three similar
stanzas, as for example 68, 59, 60 on kinghood, 109, 229*; 110 on meretrices.
On the other hand, S is decidedly shorter than N, which should place it
earlier. Therefore, we have to adopt the explanation that S was rearranged
from an N which continued to expand and develop. Both make up their
three centuries in different ways, though with a substantial common nucleus.
It is our main critical task to see how far this common and therefore
presumably original portion could be restored.
 
Every scribe who has left us a complete MS of the satakatraya
manifests his anxiety to include every possible stanza which he be-
lieved to Bhartṛhari's. This would be attested by the numerous verses