Kaczyńska ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Reflections on a Cretan Gloss Denoting a Salt Source Elwira Kaczyńska WO MODERN EDITORS of the late-ancient lexicon by Hesychius of Alexandria print with the so-called crux philologorum the Cretan gloss †ἀλλάθαρον· ἁλµυρόν. Κρῆτες (α 3110) “†allátharon – something salty, [so called by] the Cretans.”1 The crux generally indicates that the lemma is corrupt or doubtful. A different view is expressed by R. A. Brown, who thinks that the Cretan word ἀλλάθαρον is correct and represents a purely Greek formation: “ἁλµυρόν is to be derived ultimately from the stem ἁλ- = salt. There is no reason to suppose that the Cretan word is not also from the same stem (psilosis of an initial vowel is not uncommon in Cretan forms).”2 He refers directly to M. Lejeune, who registers a number of Cretan forms with psilosis.3 Unfortunately, Brown explained the prior part of the supposed compound as a purely Greek element and did not give any 1 I 110 Latte; I. C. Cunningham (ed.), Hesychii Alexandrini Lexicon I (Berlin 2018) 146. See also DGE II 155. Earlier editors or authors published the gloss in question without a crux, e.g. M. Schmidt, Hesychii Alexandrini Lexicon I (Jena 1858) 126 (α 3112); G. M. Kleemann, De universa Creticae dialecti indole (Halle 1872) 22–23; LSJ9 68. 2 R. A. Brown, Evidence for Pre-Greek Speech on Crete from Greek Alphabetic Sources (Amsterdam 1985) 27. 3 M. Lejeune, Phonétique historique du mycénien et du grec ancien (Paris 1972) 281– 282. The psilotic forms are regularly attested in the Lesbian poetry, cf. Aeol. ἄλµυρος adj. ‘salty, brine (of the sea)’, e.g. ἄλµυρον ἐστυφέλιξε πόντον (Alc. 334.2); θάλασσαν ἐπ’ ἀλµύραν (Sapph. 96.10), cf. C. Neri, Saffo, testimonianze e frammenti (Berlin 2021) 737. T ELWIRA KACZYŃSKA 149 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 etymology for the latter. A. Th. Vasilakis also treats the Hesychian gloss in question as a credible item of the Ancient Cretan vocabulary. He quotes it with no crux4 and does not suggest any etymology for the lemma ἀλλάθαρον. H. B. Voretzsch explains the Cretan word as an alleged com- pound derived from *ἁλµᾱ-ταρον by a progressive assimilation of two consonants -λµ- > -λλ-; cf. Gk.5 ἅλµη, Dor. ἅλµᾱ f. ‘sea- water, brine’, ταριχεύω vb. ‘preserve food by salting’.6 G. M. Kleemann does not accept Voretzsch’s suggestion that “λλ ex λµ assimilatione factum es[t],” and adds: “Neque infitiandum est, magnum in Creticis vocabulis campum assimilationem habere. Licet tamen de hac voce dubitare.”7 Other researchers think that the etymology of the Cretan word ἀλλάθαρον is unclear.8 1. A new explanation for the Hesychian lemma I would like to suggest a new etymology of the Cretan gloss †ἀλλάθαρον· ἁλµυρόν. Κρῆτες. Generally, both the lemma †ἀλλάθαρον and the explanatory word ἁλµυρός adj. ‘salty, briny’, also ‘brackish’, metaphorically ‘bitter, distasteful, piquant, unpalatable’, seem to represent two analogous compound formations containing the well-known Greek noun ἅλς (gen. sg. ἁλός) m. ‘salt’, f. ‘sea’.9 The neuter form of the adjective ἁλµυρός 4 A. Th. Vasilakis, Το κρητικό λεξιλόγιο (Heraklion 1998) 35. 5 Abbreviations used: Arm.: Armenian; Cz.: Czech; Hitt.: Hittite; Hom.: Homeric (epic) dialect; IE: Indo-European; Mod. Gk.: Modern Greek; OChSl.: Old Church Slavic; PGk.: Proto-Greek; Srb.-ChSl.: Serbian variant of Church Slavic. 6 H. B. Voretzsch, De inscriptione Cretensi qua continetur Lyttiorum et Boloentiorum foedus (Halle 1862) 12. 7 Kleemann, De universa Creticae dialecti indole 22–23. 8 Cf. DGE 155 s.v. †ἀλλάθαρον: “Etim[ología] desc[onocida].” 9 J. Pokorny, Indogermanisches etymologisches Wörterbuch (Bern 1959) 878–879; Hj. Frisk, Griechisches etymologisches Wörterbuch I (Heidelberg 1960) 78–79; P. Chantraine, Dictionnaire étymologique de la langue grecque I (Paris 1968) 65; R. 150 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 is commonly used in the popular phrase ἁλµυρὸν ὕδωρ ‘salt (sea-)water’ or θαλάσσης ἁλµυρὸν ὕδωρ ‘salt water of the sea’, which appears as early as in the Homeric epics (Od. 4.511, 5.100, 9.227, 12.236, 12.240; Strattis fr.14 K.; Arist. Mete. 353b15; Nonnus Dion. 31.57; cf. also Thuc. 4.26.5 ὕδατι ἁλµυρῷ).10 It is very probable that the Cretan lemma †ἀλλάθαρον refers to ‘salt water’ too. There is no reason to reject the traditional hypothesis that the Cretan gloss †ἀλλάθαρον represents a primitive compound, whose first member is a reflex of the old Indo-European term for salt (IE *sal-s, hence Lat. sal m. ‘salt’; Arm. ał ‘salt’; Gk. ἅλς m. ‘salt’, f. ‘sea’). In fact, the Cretan term †ἀλλάθαρον could be regularly derived from *ἁλλάθαρον [hal'latharon] as a result of a distant regressive dissimilation of two aspirated consonants (Grassmann’s Law).11 For the dissimilatory loss of aspiration in two non-adjacent syllables, consider the following examples: 1.1. Gk. Ion. ἀδελφεός, Att. ἀδελφός m. ‘brother’ < PGk. *ἁ­δελφειός < IE *sm-̥gue̯lbhei̯ós adj. ‘(descending) from the same womb’; cf. Ved. ságarbhya- adj. ‘born from the same womb’, sagarbhā- f. ‘pregnant woman’.12 See also Gk. δελφύς f. ‘womb’, Ved. gárbha- m. ‘womb’. 1.2. Gk. Hom. ἄλοχος f. ‘a bed-fellow, spouse, wife; co- habitant; concubine’ < PGk. *ἅλοχος < IE *sm-̥loghos ‘having the same bed; lying together in a bed; the action of lying together’; cf. OChSl. sǫlogъ m. ‘bedmate, husband, lover’, Srb.-ChSl. Beekes, Etymological Dictionary of Greek (Leiden 2010) 74–75; R.-G. Curcă, “Terminology of Salt in Ancient Greek,” Studia Antiqua et Archaeologica 21 (2015) 221–233. 10 DGE 166; LSJ9 72; M. Glezos, ϓδωρ – Αύρα – Νερό (Athens 2001) 333; F. Montanari, The Brill Dictionary of Ancient Greek (Leiden 2018) 96; J. Diggle, The Cambridge Greek Lexicon I (Cambridge 2021) 64. 11 See N. E. Collinge, The Laws of Indo-European (Amsterdam 1985) 47–61; B. Vine, “Grassmann’s Law,” Encyclopedia of Ancient Greek Language and Linguistics II (Leiden 2014) 26–27. 12 Pokorny, Indogermanisches etymologisches Wörterbuch 473; Chantraine, Dictionnaire étymologique 18–19; Beekes, Etymological Dictionary 20. ELWIRA KACZYŃSKA 151 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 сулогъ m./f. ‘bedmate; σύγκοιτος’, Cz. souloh m. ‘intercourse’. See also Russ. суложь f. ‘bedmate, wife’.13 1.3. Gk. Hom. ἄσταχυς m. ‘grain growing in the field; (mowed) cereal together with the stem and ear’ < PGk. *ἅ-σταχυς ‘having the same corn-ear’; cf. Gk. στάχυς m. ‘corn-ear’. Generally, I agree with Voretzsch’s opinion that the Cretan geminate -λλ- must represent a primitive cluster of consonants. However, the derivation of -λλ- from *-λµ- hardly seems pos- sible. It should be emphasized that a different gloss of Cretan origin, attested in the Hesychian lexicon, points to the original cluster *-λµ-: 1.4. αὕµα· ἅλµη, ὑπὸ Κρητῶν (Hesych. α 8324)14 “haúma – sea- water, [a word used] by the Cretans.” The Cretan lemma αὕµα (< Dor. Gk. ἅλµᾱ; cf. Hom. Gk. ἅλµη f. ‘saline water, sea; salt sediment; salt solution, brine’ ← Gk. ἅλς m. ‘salt’, f. ‘sea’ < IE *sal-s ‘salt’), registered by Hesychius, documents a velar pronunciation of the liquid consonant [l], which is perfectly attested in the central area of Crete from the fifth century B.C.15 In other words, the cluster *­λµ- appears frequently in the ancient Cretan dialect. The suggested pro- 13 Pokorny, Indogermanisches etymologisches Wörterbuch 658–659; J. P. Mallory and D. Q. Adams, The Oxford Introduction to Proto-Indo-European and the Proto- Indo-European World (Oxford 2008) 209; Beekes, Etymological Dictionary 74, 852. 14 281 Latte; V. Pisani, Manuale storico della lingua greca (Brescia 1973) 111; M. Bile, Le dialecte crétois ancien (Paris 1988) 120; 380 Cunningham. 15 This process consists in the velarization of the Greek phoneme [l] at the end of a closed syllable and its regular transformation into the Cretan phoneme [u̯] (cf. ‘dark’ l in English). See Kleemann, De universa Creticae dialecti indole 27; A. N. Skias, Περί της κρητικής διαλέκτου (Athens 1891) 75; A. Thumb and E. Kieckers, Handbuch der griechischen Dialekte I (Heidelberg 1932) 156; Bile, Le dialecte crétois ancien 120, and “Cretan,” Encyclopedia of Ancient Greek Language II 396–398, at 396; M. Morani, Introduzione alla linguistica greca. Il greco tra le lingue indeuropee (Alessandria 1999) 82; A. Bartoněk, Dialekty klasické řečtiny (Brno 2009) 191, and Chréstomatie starořeckých nářečních nápisů (Brno 2011) 142; J. Méndez-Dosuna, “The Doric Dialects,” in A.-F. Christidis (ed.), A History of Ancient Greek: From the Beginnings to Late Antiquity I (Cambridge 2015) 444– 459, at 455. 152 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 gressive assimilation *-λµ- to -λλ- is completely unknown in the island of Crete. There is only one example of the regressive assimilation of *-λµ- to -µµ-, attested by the proper name Σαλµώνιον/Σαµµώνιον/Σαµώνιον, a promontory in the eastern part of Crete.16 In my opinion, the spelling -λλ- in the Hesychian gloss in question could reflect the original cluster -λϝ- [lu̯]. In fact, the digamma was preserved for a long time in Crete. If this suggestion is correct, we can connect the latter member of the Cretan dialectism ἀλλάθαρον with the Hittite noun u ̯attaru- n. ‘well, spring, source’.17 A. Kloekhorst rejects the traditional derivation of Hitt. u ̯attaru- n. ‘well, spring, source’ from Hitt. u ̯ātar n. ‘water’ (< IE *wódr̥ n. ‘water’), correctly stressing that this explanation “is impossible in view of the geminate spelling -tt- in u ̯attaru-, which points to an etymological *t, which contrasts with the etymological *d in u ̯ātar.”18 His position is confirmed by Hittite lexical data, e.g. nom.-acc. sg. u ̯a-at-ta-ru (KUB 31.143a+; VBoT 124 III 21, KUB 8.41 III 14 fr.), u ̯a-at-ru (KBo 40.34, 5), gen. sg. u ̯a-at-ru-aš (KBo 8.41 III 3), u ̯a-at-ta-ru-aš (KUB 31.143a + VBoT 124 II 11, Bo 4767, 4), [u ̯a-at-]ta-ru-a-aš (KBo 25.112 III 8), allative sg. u ̯a-at-tar- u ̯a (KBo 3.7 IV 12, KUB 17.6 IV 9), dat.-loc. sg. u ̯a-at-tar-ú-i (KBo 24.12 obv. 6), also TÚL-i (KUB 12.66 IV 15), nom.-acc. pl. [u ̯a- 16 Voretzsch, De inscriptione Cretensi 12; Bile, Le dialecte crétois ancien 120 n.204. 17 D. M. Weeks, Hittite Vocabulary: An Anatolian Appendix to Buck’s Dictionary of Selected Synonyms in the Principal Indo-European Languages (Los Angeles 1985) 13 (‘spring, well’); J. Friedrich, Kurzgefaßtes hethitisches Wörterbuch. Kurzgefaßte kriti- sche Sammlung der Deutungen hethitischer Wörter (Heidelberg 1991) 250 (‘Brunnen, Quelle’); J. Tischler, Hethitisches Handwörterbuch (Innsbruck 2001) 199 (‘Was- serloch, Quelle, Brunnen’); A. Kloekhorst, Etymological Dictionary of the Hittite Inherited Lexicon (Leiden 2008) 989 (‘well, source’). 18 Kloekhorst, Etymological Dictionary 989. It should be added that earlier scholars (e.g. J. J. S. Weitenberg, Die hethitischen u-Stämme [Amsterdam 1984] 195; Weeks, Hittite Vocabulary 13–14) derived the Hittite term u̯attaru- n. (u- stem) ‘well, spring, source’ from Hitt. u̯ātar n. ‘water’. As noted, such an etymology is impossible for phonological reasons. ELWIRA KACZYŃSKA 153 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 at-ta]r-u ̯a (KUB 19.37 III 54).19 Finally, Kloekhorst reconstructs *uot(H)-ru- as a primitive archetype of this Hittite word.20 It is worth emphasizing that the suggested cluster *tH may create an aspirated voiceless dental stop [th], which should be rendered as θ [th] in Ancient Greek. Moreover, it is possible that the Anatolian term for ‘well, spring, source’, reflected in Hitt. u ̯attaru-, is to be found in the following gloss: 1.5. Βαθάλη· κρήνη. Ἀµερίας (Hesych. β 47)21 “Bathálē – a (name of) spring/source, [according to] Amerias.” K. Latte and I. C. Cunningham, the two modern editors of the Hesychian lexicon, follow Benseler, who argued that Βαθάλη represents not a purely Greek appellative but a proper name (of foreign origin) referring to a spring (“Bezeichnung einer Quelle”).22 The Pre-Greek hydronym, registered by two Greek lexicographers, Amerias and Hesychius, can be related to Hittite u ̯attaru n. ‘well, spring, source’. The Greek grapheme β (originally [b], later [ƀ] and [v])23 may render the digamma [u]̯, pronounced later as [v].24 On the other hand, the Greek phoneme θ [th] agrees with an Anatolian phoneme [th] or [tH], as suggested by Kloekhorst. The liquid consonant, written as λ [l] in the gloss offered by Amerias, may replace the Hittite liquid 19 KBo = Keilschrifttexte aus Boghazköy (Osnabrück 1923–); KUB = Keil- schrifturkunden aus Boghazköy (Berlin 1921‒1990); VBoT = A. Goetze, Verstreute Boghazköi-Texte (Marburg a.d. Lahn 1930). 20 Kloekhorst, Etymological Dictionary 989 s.v. u̯at(ta)ru-, concludes: “So, although I know no IE cognates, I would mechanically reconstruct this word as *uot(H)-ru-.” 21 306 Latte, 413 Cunningham. 22 G. E. Benseler, Wörterbuch der griechischen Eigennamen3 (Braunschweig 1911) 191. 23 Bartoněk, Dialekty klasické řečtiny 142. 24 Cf. J. Brause, Lautlehre der kretischen Dialekte (Halle 1908) 38–56. 154 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 phoneme [r].25 Following the consensus that a reflex of IE *sal- ‘salt’ appears in the former part the Cretan gloss †ἀλ-λάθαρον, I propose that the latter member -λάθαρον is identical with the above- mentioned Anatolian noun for ‘well, spring, source’, which many Hittite texts register as u ̯attaru and was presumably pro- nounced [ua̯tharu]. It is difficult to decide whether the Cretan term should be treated as a hybrid (Greek-Anatolian) formation, as it ostensibly appears. The Anatolian term for ‘salt’ is in fact unknown.26 Theoretically, one can propose that the Cretan lexical item was an early adaptation of the Anatolian (Hittite or Luwian) *sal- u ̯atharu- ‘salt source’.27 Some researchers believe, however, that 25 The λ/ρ variation seems to be one of the features of the Pre-Greek substrate. Cf. E. J. Furnée, Die wichtigsten konsonantischen Erscheinungen des Vor- griechischen (The Hague 1972) 387; R. Beekes, Pre-Greek. Phonology, Morphology, Lexicon (Leiden 2014) 21. 26 Weeks, Hittite Vocabulary 84. The meaning ‘salt’ in the preserved Hittite texts is rendered by the Sumerian ideogram MUN, as well as by the Akkadian term ṬĀBTU. Cf. H. A. Hoffner, Jr., “An English-Hittite Glossary,” RHA 25 (1967) 7–99, at 77; Friedrich, Kurzgefaßtes hethitisches Wörterbuch 286. It is worth adding that C. Watkins suggests that ḪUR.SAG Šaliwana/i- means something like ‘Mount Saltrock’: “Luvo-Hittite ‹‹lapan(a)-,” in D. Disterheft et al. (eds.), Studies in Honor of Jaan Puhvel I (Washington 1997) 29–35, at 34). Thus, he derives the first component of the Hittite oronym, šali-, from IE *sal- ‘salt’. Reflexes of the Indo-European word for ‘salt’ in the onomastic landscape are discussed by A. Falileyev, “‘Salty’ Geographical Names: A Fresh Look,” in M. Alexianu et al. (eds.), Archaeology and Anthropology of Salt: A Diachronic Approach (Oxford 2011) 209–214. 27 The Hittite or Anatolian terms for ‘salt source’ or ‘salty water’ are unknown. It should be emphasized, however, that Strabo 12.3.12 mentions salty sources or saltworks (Gk. ἁλαί) located near Halys, the longest river in Asia Minor (Cappadocia). He believes that the hydronym Halys (Gk. Ἅλυς, also Ἄλυς and Ἄλις, now Kızılırmak or Acı Su) is motivated by these saltworks (cf. D. W. Roller, The Geography of Strabo [Cambridge 2014] 524). Therefore the river in question is also known by the Turkish name Acı Su (‘bitter water’). Moreover, the Ancient Greek hydronym Halys (literally ‘Salt River’) derives ELWIRA KACZYŃSKA 155 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 the linguistic substrate of Crete, the Aegean islands, and some parts of continental Greece in the second millennium B.C. was of Anatolian origin.28 It was also assumed by some scholars that “most Greek place names in -νθος are likely to be legacies from one (or more) languages close to the Anatolian branch which was spoken in certain regions of Greece before the Mycenaeans settled there.”29 The supposed Hellenic adaptation of the Anatolian lexeme *sal-u ̯atharu- (u-stem) n. ‘salt source’ could be connected with the replacement of an original u-stem by a highly productive o-stem. The final result of prehistoric Cretan- Anatolian language contacts would be attested by ἀλλάθαρον. The Hesychian form of this Cretan term appears to be correct and the gloss should be printed without a crux. 2. Can the Hesychian gloss refer to a Cretan source of salt water? There are three abundant saltwater springs on the island of Crete. All of them are called Armirós or Almirós (Mod. Gk. Αρµυρός, earlier Αλµυρός ← Gk. ἁλµυρός adj. ‘salty’).30 2.1. Mod. Gk. Αλµυρός/Αρµυρός, salt springs and marshes located ca. 300 meters west of the village of Georgioupolis (in the Chania region). They are attested implicitly in the river name τοῦ Ἁλµυροῦ τὸ ποτάµι (see §2.4), first mentioned in the so- called Chrysóboullo (1182, a counterfeit created in the 13th cent.).31 via Anatolian (or via Phrygian) from IE *salús adj. ‘salty/salzig’; cf. D. J. Georgacas, “From the River Systems in Anatolia: The Names of the Longest River,” Names 12 (1964) 197–214, at 212; J. Tischler, Kleinasiatische Hydronymie. Semantische und morphologische Analyse der griechischen Gewässernamen (Wiesbaden 1977) 60. 28 S. Rose, “Greek and Anatolian Languages,” Encyclopedia of Ancient Greek Language II 27–31, at 28. 29 Y. Duhoux, “Pre-Greek languages: Indirect Evidence,” in A History of Ancient Greek I (Cambridge 2015) 223–228, at 228. 30 O. Rackham and J. Moody, The Making of the Cretan Landscape (Man- chester 1996) 42. 31 G. A. Sifakas, “Το χρυσόβουλλον Αλεξίου Βʹ Κοµνηνού και τα δώδεκα 156 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 The salt source is explicitly attested in modern times, e.g. Αλµυρός – [πηγή] στη Γεωργιούπολη Αποκόρωνα (1996);32 Αλµυρός or Αρµυρός (my autopsy; June 2003). The Cretan proper name derives from the adjective αλµυρός, also αρµυρός ‘salt/y/, briny, saline’.33 The alternative form αρµυρός documents a delaterali- zation typical of some idioms of Modern Greek, including the Cretan dialect.34 2.2. Mod. Gk. Αλµυρός/Αρµυρός, also Αλµυρίδα/Αρµυρίδα, an abundant source of saltwater northwest of the village of Gazi (Heraklion region), from which a small river called Armirós/ Almirós flows (see §2.5). A small lake, the Kolímba Armiroú, forms near the source. The attestation for the salt spring begins in the 16th century: Armirò (1590);35 Armiris (KIK 1941);36 Αλµυρός αρχοντόπουλα (1182 µ. Χ.),” CretChron 2 (1948) 129‒140. See also Th. E. Detorakis, History of Crete (Heraklion 1994) 135‒136. 32 Rackham and Moody, The Making of the Cretan Landscape 42. 33 D. N. Stavropoulos, Oxford Greek-English Learner’s Dictionary (Oxford 1992) 34. 34 B. Newton, The Generative Interpretation of Dialect. A Study of Modern Greek Phonology (Cambridge 1972) 112–113. 35 M. Arakadaki, “Σταµάτη Φασιδώνη (Λιµενάρχη Χανίων): ‘Περιγραφή των ακτών της Κρήτης’ (1590 ci.),” in Ch. Charalampakis (ed.), Τα κρητικά τοπωνύµια: Διήµερο Επιστηµονικό Συνέδριο I (Rethymnon 2000) 109–158, at 128, 149. 36 Historical maps and archive sources: Bas 1618 = D. Calabi, F. Basilicata, Regno di Candia: atlante corografico di Francesco Basilicata 1618 (Marsilio 1993); Bosch 1651 = M. Boschini, Il Regno tutto di Candia (Venice 1651); Cor 1681 = P. M. Coronelli, Isola. Regno di Candia (Venice 1681); EC-Road 2003 = Ανατολική Κρήτη / Eastern Crete, Road Editions, Scale 1:100.000 (Athens 2003); KIK 1941 = Karte der Insel Kreta, Maßstab 1:266.666, Sondergabe I.1941; Merc 1590 = G. Mercator, Candia cum insulis aliquot circa Graeciam (Duisburg 1590); Ort 1584 = A. Ortelius, Candia Insula. Archipelagi insularum aliquot descrip. (Antwerp 1584); Rossi 1622 = D. di Rossi, Descriptio Crete, nunc Candia Moderna (Venice 1622); Sans 1658 = N. Sanson, Isle et Royaume de Candie (Paris 1658 [map]); TLK 1953–1957: Τοπωνυµικό λεξικό της Κρήτης, Historical Museum of Crete. ELWIRA KACZYŃSKA 157 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 (Αρµυρός) – µεγάλη καρστική πηγή (1965);37 Αλµυρός – [πηγή] στο Γάζι Τεµένους [sic!] (1996/2004);38 Αλµυρίδα (EC-Road 2003). The oldest record, Armiris (1941), represents an alternative form taken from katharevousa (= Mod. Gk. Αρµυρίδα). It derives from the Cretan dialectal word αρµυρίδα f. ‘saltwater source’ (< Gk. ἁλµυρίς, gen. sg. ἁλµυρίδος f. ‘salty moisture; salty foam; salti- ness; salty terrain’).39 2.3. Mod. Gk. Αλµυρός/Αρµυρός, salt springs and marshes located ca. 2 km south of the town of Agios Nikolaos (Lasithi region). The marshes called Almirós form one of the most im- portant hydrobiotopes in Greece. The salt marshes are clearly registered as Armiro in the 17th century (Bas 1618; Bosch 1651; Cor 1681) as well as in modern times, e.g. Almyro or Armyro – a copious source of brackish water (1865);40 Almyros (1869);41 ὁ Ἁλµυρός (1890);42 ὁ Ἁρµυρός (1891);43 ὁ Ἁλµυρός (1902);44 ὁ Ἁλµυρός (1984);45 Αλµυρός (2006).46 There are also rivers called Almirós or Armirós on the island. Two of them are connected with the above-mentioned salt springs. 37 E. K. Platakis, “Ο Αλµυρός του Ηρακλείου. Ειδήσεις και πληροφορίες,” CretChron 19 (1965) 55–104, at 55. 38 Rackham and Moody, The Making of the Cretan Landscape 42. 39 Montanari, The Brill Dictionary of Ancient Greek 96. 40 T. A. B. Spratt, Travels and Researches in Crete I (London 1865) 145. 41 V. Raulin, Description physique de l’île de Crète (Paris 1869) 379. 42 N. Stavrakis, Στατιστική του πληθυσµού της Κρήτης µετά διαφόρων γεωγραφικών, ιστορικών, αρχαιολογικών, εκκλησιαστικών κτλ. ειδήσεων περί της νήσου (Athens 1890) 51. 43 E. G. Generalis, Επίτοµος γεωγραφία της νήσου Κρήτης (Athens 1891) 11. 44 E. G. Generalis, Επίτοµος γεωγραφία της νήσου Κρήτης3 (Athens 1902) 10. 45 P. Faure, “Hydronymes crétois,” Κρητολογία 10 (1984) 30–61, at 45 = Recherches de toponymie crétoise. Opera selecta (Amsterdam 1989) 303–337, at 318. 46 N. Andriotis, Πληθυσµός και οικισµοί της ανατολικής Κρήτης (16ος‒19ος αι.) (Heraklion 2006) 30. 158 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 2.4. Mod. Gk. Αλµυρός/Αρµυρός or Ποταµός, earlier του Αλµυρού το ποτάµι, a river only one km long, flowing from the spring by the same name (see §2.1) about 300 m west of Georgioupolis (Chania region). The river is attested for the first time in 12th or 13th cent.: τοῦ Ἁλµυροῦ τὸ ποτάµι (1182 [13th cent.]). Historical sources or maps registered it by means of a Latin calque, e.g. Salsum flumen (1415);47 Salsa fluvius (Rossi 1622). The Cretan river frequently appears as Armiro (= Mod. Gk. Αρµυρός) in Venetian and post-Venetian times, e.g. Armiro (1577/8);48 Armirò (1590);49 Armiro (Bas 1618); Armiro Fiume (Bas 1630); Armiro F. (Bosch 1651); Armiro fl. (Sans 1658). The river is navigable all the way from the salt springs to the mouth. 2.5. Mod. Gk. Αρµυρός/Αλµυρός, a small river, ca. two kilometers long that flows from an abundant saltwater spring called Almirós Armirós or Almirída Armirída (see §2.2) to the Cretan Sea near the mouth of the river Gazanós (Heraklion region). This river appears frequently in maps starting in the 16th cent.: Armiro flu. (Ort 1584); Armiro fl. (Merc 1590). It is mentioned as Armiro or Armirò by Francesco Barozzi (1577/8), Stamatis Fasi- donis (1590), and Francesco Basilicata (1630).50 2.6. Mod. Gk. Αλµυρός/Αλυµρός/Αλυµβρός, a periodic stream flowing through the gorge of the same name, located 1.5 km northwest of the town of Ierapetra (Lasithi region). It crosses some suburbs of Ierapetra. Attestation: Αλµυρό (στον) – ξεροπόταµος (TLK 1953‒1957: Ιεράπετρα A/815[638].49); ὁ Ἁλυµ(β)ρός (1984)51; Αλµυρός (my autopsy; July 2003). Two al- 47 M.-A. van Spitael, Cristoforo Buondelmonti, Descriptio insule Crete et liber insularum, cap. XI: Creta (Heraklion 1981) 100, 246. 48 S. Kaklamanis, Francesco Barozzi, Descrittione dell’isola di Creta (Περιγραφή της Κρήτης) (1577/8) (Heraklion 2004) 269. 49 Arakadaki, Τα κρητικά τοπωνύµια 125, 146. 50 Kaklamanis, Francesco Barozzi 268–269; Arakadaki, Τα κρητικά τοπω- νύµια 128, 149; Francesco Basilicata, Relazione 1630, in S. G. Spanakis (ed.), Μνηµεία της κρητικής ιστορίας V (Heraklion 1969) 35. 51 Faure, Κρητολογία 10 (1984) 49 = Recherches de toponymie crétoise 322. ELWIRA KACZYŃSKA 159 ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 ternative forms Ἁλυµρός [Alimrós]/Ἁλυµβρός [Alimbrós], noted by P. Faure, probably arose from local metathesis. The gorge seems to take its name from the river. Hesychius of Alexandria thus seems to have recorded a Pre- Greek name of one of three Cretan sources of abundant salt water. If so, the corresponding Hesychian gloss should be re- stored as Ἀλλάθαρον· ἁλµυρόν. Κρῆτες. Conclusions The above reflections on one of the Cretan glosses recorded by Hesychius of Alexandria lead to the following conclusions: 3.1. The Cretan gloss †ἀλλάθαρον· ἁλµυρόν. Κρῆτες (Hesych. α 3110) cannot be treated as corrupt or doubtful. Both the lemma (ἀλλάθαρον) and the explanatory word (ἁλµυρόν) contain the same lexical morpheme *ἁλ- [hal] ‘salt(y)’ (cf. especially Gk. ἅλς, gen. sg. ἁλός m. ‘salt’, f. ‘sea’), which evidently derives from the inherited Indo-European term for ‘salt’ (IE *sal-). 3.2. There is no obvious psilosis in the Cretan word ἀλλάθαρον. The spiritus lenis can readily be explained as the outcome of a distant regressive dissimilation: [ø] – [th] < [h] – [th]. As shown above (§1.1–1.3), the dissimilatory loss of aspiration in two non-adjacent syllables is attested in several instances. 3.3. The Cretan word ἀλλάθαρον (correctly glossed as ἁλµυρόν ‘something salty’) likely represents an Anatolian loan, whose original archetype is *sal-u ̯atharu- n. ‘salt source’; cf. Hitt. u ̯attaru- (u-stem) n. ‘well, spring, source’. Alternatively, the Cretan lexeme may be interpreted as a Greek-Anatolian hybrid formation. 3.4. The relation between this Hesychian gloss and three sources of abundant salt water in Crete, commonly called Armiros or Almiros, is unclear. We cannot dismiss the possibility that the first word of the lemma should be capitalized, Ἀλλάθαρον· ἁλµυρόν. Κρῆτες “Allátharon – a salt source, [so called by] the Cretans.” This would parallel the Hesychian gloss of 160 DENOTING A SALT SOURCE ————— Greek, Roman, and Byzantine Studies 62 (2022) 148–160 foreign origin Βαθάλη· κρήνη. Ἀµερίας, “Bathálē – a (name of) source, [according to] Amerias” (§1.5). March, 2022 University of Lodz, Faculty of Philology Department of Classical Philology elwira.kaczynska@uni.lodz.pl ORCID 0000-0002-4545-1927