This page has not been fully proofread.

66
 
state of doubt indicated by an asterisk after the sloka number in the
apparatus; stanzas omitted in three or more complete MSS, whether
constituting a version or not, are also struck out of the group. C, D, I are
treated as single MSS.
 
INTRODUCTION
 
The total of exactly 200 stanzas in group I is accidental. Of these, the
first seven do not belong to the same sataka in all archetypes, hence are
labelled 'unplaced' and separated from the rest. Cudottamsita-, which is placed
in all three centuries by turn is taken as the general benedictory stanza, while
the rest are given in alphabetical order. Possibly, stanza 82 might also have
been included in these. The remainder of group I is subdivided into the
traditional N-S-V classes, each being provisionally in the order of version A
though the general N order shows some marked differences on occasion.
 
Group II: This is printed in the same type as group I, and contains
stanzas found in more than a single well-determined version, but not included
in I. The criterion is generous, because of the critical conclusions in section
3.3 about 83 and 287. After all, each of the versions began from a single MS
at some time in the remote past. That they have become rooted attests their
strength, but not necessarily their absolute reliability. Why does J alone
omit 341? Why are 285, 350, 221, and a few others of the sort omitted only
in X? To such questions, I can find no answer. But the age and contents of
these versions being incontestable, there is no option but to include them in II.
On the other hand, the group is undoubtedly much too broad, inasmuch as
the authentic stanzas cannot possibly include all of those we find here; so I
have drawn a minor dividing line of my own by reporting some stanzas as
"omitted in" and others as "found in", whichever was shorter. Clearly, the
latter subgroup is much less likely to be authentic, though again the boundary
is not sharp. A few stauzas that might, without doing any great violence to
the canon, have been included in Group I have been starred.
 
Group III: This naturally contains everything in satakatraya MSS
which cannot be ascribed to the first two groups, and which is also not clearly
indicated by the scribes as an interpolation from some other sources. In
many cases, as for example in HU 2145, it is impossible to be certain that the
scribe wasn't just exercising his memory and his pen, without much regard to
Bhartrhari. But it is impossible to neglect any of the additions indicated as
extras because one or the other of them generally turns up elsewhere in the
body of a satakatraya MS. This process is also ancient enough not to be
ignored. Thus, bhavad bhavan-[627] has forced me to include the extras
in X2 though X is well-determined version; cetohara [499] meant the
inclusion of the extras in J3. There are plenty of stray verse jottings that I
have left out because they were clearly not meant by the scribe as part of
the satakas. Not every scribe is so careful as that of Bikaner 3287 who labels
his extras Prästāvikävali, with an emphatic word on the margin na tu
Bhartrharakṛtiḥ, though several of the stanzas do occur in the Bhartṛhari text.
Bikaner 3276 gives many scattered marginal extras of which one [not
taken here] is labelled Siladevasüreh, whence the inference is that the rest
are to be taken as Bhartrhari's and indeed most of them are found in
other Bhartrhari MSS; Bikaner 3279 gives a couple of extraordinary stanzas