Devarâja, in his commentary on gnâ, explains it: “Gamer dhâtor dhâpṛ́vasyajyatibhyo naḥ (U. S. III. 6) iti bahulakân napratyayo bhavati ṭilopaś ca; ṭap. Gatyarthâ buddhyarthâḥ jânanti karmeti gnâḥ. Yadvâ gacchati yajñeshu; abhí yajñám gṛṇîhi no gnâvaḥ (patnîvaḥ) Rv. I. 15, 3. Chandâṃsi vai gnâ iti brâhmaṇam iti Mâdhavaḥ. Asmấ îd u gnấś cid (Rv. I. 61, 8) ity api; gâyatryâdyâ devapatnya iti sa eva. Tasmâc chandasâm gâyatryâdînâm vâgrûpatvâd gnâvyapadeśaḥ.”
In his remarks on Nigh. III. 29, it is quite clear that Devarâja takes gnâḥ as a nom. plur., not as a nom. sing. He says: Menâ gnâ iti stríṇâm; ubhâv api śabdau vyâkhyâtau vânnâmasu. Mânayanti hi tâḥ patiśvaśuramátulâdayaḥ, pûjyâ bhûshayitavyâś ceti smaraṇât. Gacchanty enâḥ patayo patyârthinaḥ. The passage quoted in the Nirukta III. 29, gnâs tvâkṛntann apaso ’tanvata vayitryo ’vayan, is taken from the Tâṇḍyabrâhmaṇa I. 8, 9. “O dress! the women cut thee out, the workers stretched thee out, the weavers wove thee.”
Thus every support which the Nighaṇṭu or the Nirukta was supposed to give to the form gnâḥ as a nom. sing. vanishes. And if it is said s.v. gnâspati, that in this compound gnâḥ might be taken as a nom. sing., and that the Pada-text separates gnâḥ-patiḥ, it has been overlooked that the separation in Rv. II. 38, 10, is a mere misprint. See Prâtiśâkhya, 738. The compound gnâspatiḥ has been correctly explained as standing for gnâyâspatiḥ, and the same old genitive is also found in jâspatiḥ and jâspatyam. See also Vâjasan. Prâtiśâkhya, IV. 39. It is important to observe that the metre requires us to pronounce gnâspati either as gnăāspătĭḥ or as gănāspătĭḥ.
There is, as far as I know, no passage where gnâḥ in the Veda can be taken as a nom. sing., and it should be observed that gnâḥ as nom. plur. is almost always disyllabic in the Rig-veda, excepting the tenth Maṇḍala; that the acc. sing. (V. 43, 6) is, however, disyllabic, but the acc. plur. monosyllabic (I. 22, 10). In V. 43, 13, we must either read gn̆āḥ or ōshădhī̆ḥ.
The beginning of V.43.13c is also given as “ghnā vasāna oṣadhīr”. Printed text (breve over n, combined breve and macron over i) may be intended to read “gnăâḥ” or “gănâḥ” and “oṣadhĭîḥ”.
[31.] Sthâ, svâbhiprâyabodhanânukûlasthiti, to reveal by gestures, a meaning not found in our dictionaries. Wilson renders it wrongly by to stay with, which would govern the instrumental. Śap, cursing, means to use curses in order to convey some meaning or intention to another person.
[32.] Wilson’s Sanskrit Grammar, p. 390.
[33.] In verbs compounded with prepositions the accent is on the penultimate: e.g., samídhe, atikráme, etc.
[34.] See M. M.’s Translation of the Rig-Veda, I. p. 34.
[35.] Morris, Historic Outlines of English Accidence, p. 52.