बहुव्रीहौ सक्थ्य्-अक्ष्णोः स्वाङ्गात् षच्

Thus दीर्घे सक्थि यस्य = दीर्घसक्थः, so also कल्याणाक्षः, लोहिताक्षः, विशालाक्षः ॥ The word सक्ष्टपक्ष्णोः is genitive dual, while it ought to be ablative. Why do we say \in a Bahuvrihi compound\? Observe परमसक्थिः, परमाक्षिः ॥ Why do we say \after sakthi and akshi?\ Observe दीर्घजानुः, सुबाहुः ॥ Why do we say \when denoting a part of the body?\ Observe दीर्घसक्थि शकटं, स्थूलक्षिरिक्षुः ॥

The new affix षच् is employed instead of टच् for the sake of the accent Thus चक्रसक्थी स्त्री, दीर्घसक्थी स्त्री ॥ By sutra 4.1.41 the feminine of words ending in an affix having an indicatory ष् is formed by ङीष् and not ङीप् ॥ Had the affix here been टच्, then the feminine would have been formed by ङीप् 4.1.15, which is anudatta 3.1.4. Now by rule 6.2.198 the second term 'saktha' has optionally udatta accent, on this final in a Bahuvrihi. When therefore, it is not oxytone, then by adding ङीप् the final of दीर्घसक्थी would have been anudatta, so that the feminine in one case would be oxytone, in the other not, which would be anomalous. But when ङीष् is added, it is always udatta, which is desired: and removes the anomaly. The anuvritti of Bahuvrihi extends up to the end of the chapter.,