This page has not been fully proofread.

6
 
EDITOR'S PREFACE
 
in error not only because of such confusion, but because he forgot that the F4
order was V-N-S against his and Carey's S-N-V. Of course, not all the F
stanzas have been charted, and ahau vā hāre và has been kicked into his supple-
ment, though found in all his MS sources, simply because the exact hundred
had to be preserved: "a nobis quidem expulsa ne numerum centenarium trans-
grederetur liber". The notes are ostentatious enough, usually missing all diffi-
cult points; even there, the parallel between bhavanti namrás and Sa'adi's
"nehad śākh pur mevaḥ sar bar zamin" has not been recognized.
 
Two at least of Bohlen's successors refused to be outdone. The Bhartṛ-
hari version in Benfey's Chrestomathie is a simple copy of his predecessors, hence
adds nothing to the textual information. The learned editor maintains in his
preface that though duskha is the correct form, he would not insist upon it in
his text, seeing that duḥkha had been sanctified by usage; nevertheless, he has
been unable to resist the temptation of euriching the Sanskrit language, and
the form does appear in parts of the Chrestomathie, which thus takes a step fur-
ther away from common sense as well as the MS evidence. Decidedly worse is
Haeberlin's edition in the Kavyasamyraha anthology (Calcutta, 1847). The
elitor has not condescended to explain his method, if indeed he had one. As
nearly as can be ascertained by comparison, the Bhartyhari portion is an exe-
crable reprint of Carey's e lition, with some verses broken in two, each half being
numbered separately, some missing altogether; there are gaps in the numbering,
and quarters end in the wrong place. It might be possible to accuse the editor
of plagiarism, but never of any such taint as knowledge of proof-reading, or of
the Sanskrit languago. The Kāryasamgraha has been improved in successiv
editions by competent scholars like Jīvānanda Vidyāsāgara, so that the later
Calcutta editions are at least readable, though there is no evidence of any new
 
MS basis.
 
Indian editions. Our earliest printed editions, like the Telugu of 1840
or 1848 [no title page! 10 2. L. 33] could have been used as local versions, being
extensions of the MS tradition. Unfortunately, they are singularly rare, for it
was found more profitable to copy with emendations from other printed editions
rather than go back to any codex. Among the earliest is a lithographed edition
published at Benares in 1860 [Divākara Press; 10 H 12a and 13] and re-issued
immediately. This was by Harilal of Gaya, resident in Benares, and seems,
according to the colophon, to have been completed in 1842. The solitary Bhar-
trhari MS at Alwar [probably the same as no. 940, p. 39, Peterson's 1892 Cata-
logue] is a defective copy of this text and commentary, dated samvat 1900.
The edition takes its text for N entirely from version W, as well as most of the
commentary which he calls the Subodhini; the S is generally northern, but Sam-
bhusvayambhu- has been omitted because of displacement in W, from which
again the commentary is borrowed with additions; the V is definitely of the N
recension, with arbitrary changes. Harilal's edition was copied without acknow-
ledgement in some of the Bombay lithographs, unless Harilal had himself copied
some such carlier edition which I have missed. Naturally, there is no ques-
tion of acknowledgement on the part of those who have continued the bor-
rowing. Among these, the Venkateśvar Steam Press edition at Bombay in 1884
adds a Hindi translation. While the Bombay lithographed editions such as IO
279. 3. B. 10 [1868] and 10. C. 5 [Jagadiśvara Press, 1882] generally favour W