v1.0
Publishing history:
v1.0: 19/03/26
geodha m. 
In Lewis, feminine (?occasionally masculine: cf. (OS 1843–82) Geodha Bèug (i.e. [An] Geodha Beag ‘the small cove’) NB515663).
Cf. [kjɔ·ɔ] (Holmer 1938, 174: Islay), /ˈco?o/ (Grant 1987 I, 178: Islay), [ɡ̊’ɛ̯ɔ.ɔ] (Wentworth 2003a, s.v. cove: Gairloch), /gʲɔ.ə/ (AFB˄); occasionally spelt geòdha, e.g. Grant (ibid.) and Coates (1990, 179: St Kilda). Alternative (disyllabic) spellings include geogha (Faclan bhon t-Sluagh˄) and geotha (ibid.; Holmer 1938, 174; also Craigie 1894, 162, but who states incorrectly that th here represents [h]).
Sometimes spelt geò, e.g. MacBain (1896; 1911), Dwelly (1911), Henderson (1910, 199: geō: North Highlands), Sommerfelt (1952a, 231: Canna) and Coates (1990, 179: St Kilda).
Cf. (Beul na Geodha) [ˌbjɑʟ nə ˈɡˊɔː] ‘the opening (of the ravine)’ approx. NB109446 (Cox 2002a, 183.533). In unstressed position, for example as generic in name-initial position in qualified place-names, the word is generally pronounced [ˈɡ̊ʲɔ] or [ˈɡ̊ʲɔˑ], e.g. SG Geodha a’ Ghàrraidh [ˌgˊɔ. ˈɣaːRi] ‘the dyke of the ravine’ (Oftedal 1972, 121).
Extended senses include ‘hollow in a rock’ (MacEachen 1842), ‘small cave (Harris); sea cliff (North Uist)’ (Faclan bhon t-Sluagh˄, s.v. geotha). Holmer (1938, 174, s.v. geotha) writes ‘“gullet”, “inlet”, common in place-names’ for Islay, but we should probably read ‘“gully” ...’, cf. Grant’s (1987 I, 178, s.v. geòdha: Islay) ‘ “narrow inlet of sea with high rocky sides”, “ravine”, “gully” ’.
MacBain (1896; 1911); Mackay (1897, 94: from Ice. gja [sic]); Watson (1906, 365); Henderson (1910, 199); MacLennan (1925); Christiansen (1938, 4, 16); MacPherson (1945, 36); Sommerfelt (1952a, 231); Fraser (1978, 7: gja [sic]; 1979, 23; 1984, 38); Coates (1988, 21); and McDonald (2009, 357–58).
Holmer (1938, 174); Oftedal (1972, 121; 1980, 173); Cox (1991, 492; 2002a, 183.533); and McDonald (2009, 357–58).
The development of ON /aː/ > /ɔː/ is attested in writing in Old Norse and Old Icelandic texts from 1200 (Haugen 1976, 206); see also Gordon 1981, 267 §8, and Iverson 1973, 10 §6.6 b.
An exception is SG Geodha na Caillich [lit. ‘the creek of the old woman’] NB39525152, in Lewis (ECC, 319, 322–23).
Alternative forms of SG geodha include geothadh (Shaw 1780; MacFarlane 1815; Dwelly 1911: geòthadh, see geòdha; Faclan bhon t-Sluagh˄: ‘gully’, Coll), perhaps in origin a genitive form, 
Unless final -adh here is a back-formation in dialects where historical -adh has become [ə] (e.g. Mid-Argyll and Lorne, cf. SGDS Items 330–331 Donnchadh, Points 45–50, 58–64) and would therefore represent [ə].
LASID (≈IV, 209, Item B1071 ‘harbour’) lists SG (Arran) ‘[oˌo] (or [oːb], pl. [leg. diminutive] [oːbɑ̆n])’, to which Ó Baoill (2012) comments ‘?not geodha’, although LASID’s [oˌo] is possibly a typesetting error for *[oˌob], cf. (B1153 ‘... in the harbour’) (Arran) anns an òb [ənsə ˈnoːb] (s.v. òb).
There is a degree of confusion among some early lexicographers. The homonymous EG (eDIL˄) 2gáeth f. ‘wind’ and 3gáeth (generally) m. ‘the sea; a stream, estuary’ are represented in the following entries in Lhuyd’s (1707) Irish dictionary: ‘geoth for gaoth “wind”’ (citing ‘Libs. Pl.’ 
‘Pl.’ = Plunkett’s (1662) Latin-Irish dictionary, but ‘Libs.’ is unexplained: neither ‘Libs.’ nor ‘Libs. Pl.’ seems to be employed elsewhere in Lhuyd’s volume.
Contrast O’Clery 1643: gaoth .i. fairrge ‘sea’; O’Brien 1768: gaoth ‘the sea’; Ó Dónaill 1977: 2gaoth ‘inlet of sea, estuary’.
Noted and derived by Craigie (1894, 162) from ON gjá.
In a separate entry, O’Reilly also lists gaoth ‘the sea’. Cf. Armstrong 1825: †gaoth ‘dart; sea; theft’, which combines three of O’Reilly’s entries; so also Dwelly 1911, although AFB˄ lists gaoth ‘shallow (esp. estuarine) stream; marsh’ separately, cf. gaoth ‘marsh’ (Faclan bhon t-Sluagh˄: Skye) and gao ‘inlet of sea, marsh by sea’ (ibid.: Strathglass); for gaoth ‘marsh’ in Scottish place-names, see Watson 1926, 492–93. Lhuyd’s (Appendix) plural form gaotha ‘streams left at low water’ is cited in HSS 1828, where it is described as feminine. For the question of the etymology of EG gáeth ‘the sea etc.’, see Tempan forthcoming.
For the variation between hiatus ([ˈɡ̊ʲɔ-ɔ]) and a long vowel ([ˈɡ̊ʲɔː]) in SG geodha ~ geòdh, cf. SG (Arran) òb [oˌob] ~ [oːb] (fn 11; see also Watson 1999, 350–52: 51). 
For similar variation in place-names in Lewis, see Cox 2022, 127–28.
Pace McDonald (2009, 357–58 
Noting that Taylor (1968, 126) and Cox (1989, 6; 1994, 31) give examples of such loan-names.
Goodrich-Freer (1897, 67) suggests that ON geo [sic] probably yields SG geòb ‘a partial opening as of a door or mouth’, but SG geòb ‘wry mouth; little mouth; gaping mouth; creek’ (Dwelly 1911), while apparently confused to some extent with geodha, may go back to Scots gaup [gɑ(:)p], [gǫ:p] ‘to gape; one who gapes’, although MacBain (1911) suggests a derivation from Eng. gape.
Derivatives: SG geodhach, with the adjectival suffix -ach (e.g. AFB˄).