This page has been fully proofread once and needs a second look.

इति (तथा) सौदामिनी[^23] लसति किं जलदे प्रयाते[^24]
 
अस्तंगते[^25] शशिनि तिष्ठति चन्द्रिका किम् । (पूर्वपीठि-५३)
 
इति च [^25a]चौर पञ्चाशति बिल्हणः ।
 
राजन् विभान्ति भवतश्चरितानि [^256]तानि
इन्दोर्द्युतिं दधति यानि रसातलेऽन्तः
धीदोर्बले अतितते उचितानुवृत्ती
आतन्वती विजयसंपदमेत्य भातः ॥
 
इति काव्यप्रकाशे (सप्तमोल्लासे उद्धृतम्) ।
 
हे [^26a]रोहिणि त्वमसि शीलवतीषु धन्या
एनं निवारय पतिं सखि दुविनीतम् ।
 
[^23]. This is the 53rd verse in the same pūrvapīṭhikā, where the printed edn.
reads bhavati, instead of lasati found in the MSS. of the S.K.U.
[^24]. This verse is also quoted for the same illustration of the violation of
sandhi, which is considered compulsory.
[^25]. The printed edn. here also as in the previous verse (See Note 21)
reads : prayāte hy astaṃgate, inserting hi in the middle. It appears to be a
device adopted to avoid the difficulty of sandhi. It is not clear what was the
original reading, if it was by Bilhaṇa.
[^25a]. Our author does not use the word Caurapañcāśikā. He uses the word
Caurapañcaśat, and sometimes Caurapañcaśatīya.
[^26]. This verse is cited in the Kāvyaprakāśa as an illustration for the two
types of visandhi-doṣa, of which one is
(I) the aicchikasandhyabhāva based on the dictum vākye tu sā (saṃhita)
vivakṣām apekṣate; and the other
(II) pragṛhyādihetukasandhyabhāva, depending upon the rules Īdūdeddvi-
vacanaṃ pragṛhyam (1.1.11) and Plutapragṛhyā aci nityam (6.1.125).
 
In the first case the non-application of a sandhi even once (sakṛd api)
between words in each half of a verse is considered a fault, according to the
poetic convention though grammatically it is not wrong.
 
In the second, adopting sandhyabhāva more than once (asakṛt) simply
because it is permitted by express grammatical rules, is also counted to be a
fault. Both these points were first noted by Daṇḍin thus :
na saṃhitāṃ vivakṣāmīty asandhānaṃ padeṣu yat,
tad visandhīti vikhyātaṃ, na pragṛhyādihetukam. (Kāvyādarśa III.159)
In IV.27 of his Kāvyālaṃkāra Bhāmaha gives an example of the second
type alone, without defining the doṣa, and he does not illustrate the first type.
Perhaps aicchikasandhyabhāva was not considered to be a doṣa by him. Later
writers followed Daṇḍin. Mammaṭa also says:
saṃhitāṃ na karomīti svecchayā sakṛd api doṣaḥ.
pragṛhyādihetukatve tv asakṛt.
(p. 333, Jhalakīkar edn., 1965)
[^26a]. This verse (he rohiṇi etc.) is also quoted in the Durghaṭavṛtti under
2.4.34, and this was probably the source for Bhaṭṭoji.