This page has not been fully proofread.

INTRODUCTION
 
75
 
in doubt; we have three separate virtues, of which one develops according
to the environment; the difficulty arises from the extreme compression
and is characteristic of the Bhartyliari nucleus, for the poet generally takes
syntax and grammar in his stride The southern reading is so obviously
a paraphrase emendation that it is hardly necessary to state that the N
variant could never have arisen from it. On the other hand, there are
cases where the S reading must have been the original, the most notable
being sakayuvati- in 257%. The simile describes piper-betel leaves, of which
the common variety is a dark green, which agrees with the otherwise
senseless northern reading śwlayuvati. But it is not generally known that
there exists a rare superior variety, beloved of the connoisseur, grown only
in a few places [as in one near Benares] which has a pale golden colour
that coincides exactly with the highly prized pale golden complexion of the
now forgotten Saka maidens,
 
The reading that best shows the need for applying the method to con-
stitute our text letter by letter is perhaps param devatū in 70°. The general N
text gives param daivatam as against parā devata in S, so that it would have
been necessary to invent the reading had it not been found in Ao.1 It,
a few Jain anthologies, and one MS used by K. T. Telang. No grammatical
difficulty remains if para is taken adverbially in the sense of eva, though even
the A commentator neglects to do this.
 
Decidedly un-Pāṇinian is rūgādhikkyatamoṣṭha- in 1198. There is a
minor difficulty the samdhi -tama + ostha which should normally have become
tamauṣṭha, but is optional (Vänt. Pän. 6. 1. 94) in the present case without
the diphthong. However, this would naturally cause an inclination to take
the preceding ending as participial, -tam in place of the superlative -tama, which
really constitutes the major difficulty. The noun has so inferior a position to
the verb in Sanskrit that adding -tama even when an adjective is not indicated
is permissible, as in stritara, strītamā, in our dictionaries, or kumăritarā,
brāhmaṇitamā in the Käsikä on Pap. 1. 1. 22. We have dvijamukhyatamaḥ
in Mṛcchakatika I. 3, suhṛttama in Mudrārākṣasa I, aśvatama in Pañcatantra
'V. 10.
The real objection is to having two taddhita affixes, for adhikya
is already so developed. The meaning not being in doubt, and the emenda-
tions so palpable, I prefer to take this as another of the poet's short cuts,
rather than a vedic survival.
 
Other difficulties are easier to resolve. In 152, kendhitair narair
tends to form a unit "hungry people", i. e. beggars, which gives rise to
the variants, particularly niranna-; as a matter of fact, kudhitair qualifies
śiśukair, and the narair is emphatically in the sense of Cato's viros alienos,
that one's wife should not be seen in so desperate a plight by other men.
The variants for manorathoparicita in 196 derive from the tendency of
paricita, a strong compound, to form a unit; the correct resolution, as indicated
by our hyphen, is manoratha+upari+cita-. It 98 sumanasaḥ has clearly
been mistaken for an incorrect nominative plural "flowers", which should
have been sumanāmsi; the original meaning, however, seems rather to have
been the genitive singular of sumanas taken to mean "the man of good
mind", to go with madam in d, for we already have kusumāni in b; of course,