Appendix V. On the Doric Dialect.
1. The ancient grammarians divided the Greek language into four distinct branches—the Doric, Ionic, Attic, and Æolic; the latter including all dialects not comprised under the other three heads, because only one branch of it, the Lesbian, was the written language of one species of poetry: and yet this latter division must unquestionably have contained different species less connected with each other than with some branches of the other three dialects. It is, however, pretty well agreed that the several Æolic dialects together contained more remains of the primitive Grecian or (if we will so call it) Pelasgic language, than either the Doric, Ionic, or Attic; and that at the same time many forms of the latter were preserved with great fidelity in the Latin tongue; partly because the life of the Italian husbandmen bore a nearer resemblance to that of the ancient Greeks than that of the later Greeks themselves, and because neither their literature, nor any fastidious sense of euphony and rhythm, induced them to soften and refine their language. But of the more polished dialects, that of Homer, though differing in many points, yet in others doubtless closely resembled the original language, which must once have been spoken from Thessaly to Peloponnesus, and was variously metamorphosed in the Doric, Ionic, and Attic dialects. Thus, for example, the genitive case of the second declension, in the ancient form, was ΟΙΟ, which was preserved in the Thessalian dialect,[1907] perhaps also in the Bœotian,[1908] and in Latin I or EI is also perceivable; [pg 418] whilst in the Doric Ω and the Attic ΟΥ this vowel was entirely lost. The nominative of masculines of the first declension in Α belongs to the Latin, Homeric, Dryopian, Thessalian, Bœotian, Macedonian, and Elean dialects. In the Doric it was probably of rare occurrence, and more accidental.[1909] The Æolic dialect, which was spoken in Bœotia, likewise contains remarkable traces of an ancient Pelasgic language, and has striking coincidences with the Latin: thus in the ancient Bœotian inscriptions the dative of the first declension ends in ΑΕ. Gradually, however, it departed from this language, as the diphthongs ΑΙ and ΟΙ, which anciently were written ΑΕ and ΟΕ, were changed into Η and Υ: and thus almost all the vowels and diphthongs received a new form. On the other hand, we must be cautious of supposing the Latin to be the ancient form, in cases where a transmutation of letters has already taken place. The following is a remarkable example to this effect. ΟΠΩ, from whence “the eye,” ὄππα in the Æolic dialect,[1910] ὄφθος in the Elean,[1911] ὄπτιλος in the Spartan. In other dialect, ὄκκος, hence ὄκταλλος in the Bœotian, in the Latin oculus, where Π and Κ bear the same relation to each other as in the words πέτυρες (Æolic) quatuor, πέμπτος, quintus, ποῖ, quo, πόθι, alicubi. Moreover the Latin has a very large number of words derived from the Campanian and Doric Greeks, which must be distinguished from the primitive Greek dialect.
2. These remarks are merely premised in order to point out the authorities upon which all investigations into the form of the most ancient language of the Greeks should be founded. We have already intimated our dissent from those who, in opposition to Pausanias,[1912] suppose the Doric to have been the native dialect of Peloponnesus, not only disallowing the claim of the Dorians to its introduction, but even denying that they were the first to adopt it. This supposition would leave us without any means of explaining how the dialect of the Dorians of Peloponnesus agreed [pg 419] in so many peculiar idioms with that of their fellow-countrymen in Crete, the close and general connexion between the two being of an earlier date than the Doric invasion of Peloponnesus. The ancient Peloponnesian dialect was certainly that language which may be recognized in the Latin and in Homer, many of the peculiarities of which occur indeed, but many of the most essential are not found, in the Doric dialect. This latter dialect was, however, very widely diffused over that peninsula by the preponderance of the Dorians, being not merely adopted by the Helots (who even at Naupactus spoke Doric), the Orneatæ,[1913] the Laconian Periœci, and the Attic inhabitants of Colonides;[1914] but even by the independent Arcadians, who, according to Strabo, used indeed the Æolic dialect, but were generally supposed to adopt the Doric (δωρίζειν), as also did Philopœmen.[1915] Unfortunately we have little information respecting the dialect of the Arcadians, our chief guide being the names of their towns, in which several Dorisms occur; as, for instance, Καφυαὶ (from Κηφεὺς), Νᾶσοι, Ἀνεμῶσα (ἀνεμόεσσα), and some anomalous forms, such as Λαδοκέα for Λαοδικέα, Θελποῦσα for Τιλφοῦσσα, Dor. Τιλφῶσσα, Κραρεῶτις, a tribe of Tegea, for Κλαρεῶτις.[1916] The Eleans, on the other hand, spoke nearly pure Doric; which is shown indeed by their use of the digamma,[1917] by their broad accent, and the Ω in the genitive case; but chiefly by the frequent use of Ρ, which, besides the ΤΟΙΡ, ΤΙΡ in the well-known treaty of the Eleans,[1918] is also proved by the Elean forms δίκαρ (for δίκας or δικαστὴς), οὗτορ, ἵππορ and similar forms, whence the Eleans were called βαρβαρόφωνοι.[1919] Moreover, the Apollo Θέρμιος of the Eleans was the same as Apollo Θέσμιος, in Attic Greek.[1920] Eretria was founded by Eleans in conjunction with other Greeks, whence the frequent use of the Ρ in that town;[1921] and from this city the neighbouring [pg 420] Chalcideans also adopted it;[1922] whilst among the Carystians another peculiarity of the Spartan Elean dialect prevailed, in the change of Θ into Σ.[1923] The Eretrians, however, received from the Eleans another peculiarity of the pure Doric, viz. the use of the aspirate in the place of Σ; and imparted it to the Oropians, their neighbours, and sometimes their subjects, on the other side of the strait.[1924] Thus it is evident that the dialect of the Eleans was very similar, nay, almost akin, to the Spartan. Now it is very improbable that this strict observance of the Doric dialect should have been learnt by mere intercourse, since on no side were they in immediate contact with Dorians. It is much more probable that the Ætolians, who conquered Elis, used, from their vicinity to the Dorians, the same dialect: that they spoke Doric in later times, is proved by the testimony of ancient authors and monuments extant;[1925] and the same was also the language of the inhabitants of the ancient Epirus Proper.[1926] It seems, therefore, that this dialect was formed in the northern and mountainous districts of Greece, particularly in the vicinity of mount Pindus, from whence the Dorians brought it in their migration to the more southern parts of the country, where they were in consequence commonly regarded as the race with whom it first originated.
3. To determine with any degree of precision how much climate and the nature of the soil contributed to the formation of this dialect, would be a matter of extreme difficulty; although the comparison of the corresponding dialects of different languages with the various localities in which each was formed may lead to several interesting observations. There can be no doubt that a mountain life is favourable to the formation of the pure, broad, and long vowels, such as Α and Ω; as also that a residence in the lowlands and on [pg 421] the coast produces rather modifications of the long vowels[1927] and short syllables. It should, however, be borne in mind, that the influence of these causes upon language was in full operation at one period only, when the organs generally evinced greater pliancy in adapting themselves to the various peculiarities of situation. In later times, Doric was spoken in maritime towns, as low German is now in mountains and highlands. We must likewise remember, that not only the country, but also the people, bore a distinct national character, the influence of which upon their language must have been full as great as of the former. The hypothesis that the ancient dialects were determined more by internal than external influence, more by the nature of the men than the influence of place, is confirmed by a remarkable passage of Jamblichus,[1928] who had probably derived this sentiment from the schools of the early Pythagoreans: he pronounces the Doric dialect to be the most ancient and best, comparing it, on account of the sounding vowels with which it abounded, to the enharmonic style of music, as he does the Ionic and Æolic dialects to the chromatic style. The only meaning of this remark can be, that the long vowels Α and Ω were pronounced in as clear and marked a manner (particularly when, as was often the case, they were circumflexed) as a bar separated by a double bar in the tetrachord strung to the enharmonic pitch, so much used for music of the Doric style.[1929] Otherwise a manly character is always attributed to the Doric dialect:[1930] its fitness for solemn occasions and simple expression is shown by the literary remains which have come down to us.
4. It cannot be expected that we should here enter into a minute examination of all the peculiarities of the Doric dialect: the following brief remarks will, it is hoped, be received as an attempt rather to set forth the most remarkable features of the spoken language, than to explain the niceties [pg 422] of the polished style used in writing and poetry. The frequent use of Α prevailed indeed partially in the ancient dialect, and in most cases the use of Η originated in the Ionic, which in this respect bore nearly the same relation to the ancient Greek as the English language does to the German.[1931] The broad pronunciation (πλατειασμὸς) of the Dorians frequently, however, exceeded that of the ancient language, as may be seen from the Latin. Thus φαγὸς, fagus—φάμα, fama—μᾶλον, malum—ἀρχᾶς, terras (genit.) κᾶρυξ, (caduceus), and the like, are clearly the genuine ancient forms. On the other hand, the change from Α to Η in the temporal augment existed in the most ancient Greek, as is evident from ago, ēgi, ἦγον, capio, cēpi, &c. The Doric dialect, however, here also used Α in the place of Η. I am not aware whether another change very nearly coinciding with the latter has ever been noticed, viz. the frequent use of the short Α for Η, especially in the enclitics, as κᾶ (which however is long) for κε or ἂν, a form common to all the Dorians, and in the same manner γα for γε,[1932] κα for the correlative τε in τόκα, πόκα, ὅκα in Sophron, Theocritus, and others, to which corresponds θα in πρόσθα, ἐξύπισθα (Alcman), ἔμπροσθα, ἄνωθα.[1933] The same change is also observable in ἅτερος for ἕτερος, τράφω for τρέπω,[1934] Αρταμις[1935] for Αρτεμις, τάως, παραιτέρω, in the Cretan dialect,[1936] τάμνω in the Heraclean Tables and elsewhere, σκιαρὸς, φρασὶν, in Pindar; and innumerable examples of a similar kind. Η, either as a contraction of ΕΕ, or a lengthening of Ε, occurs in many instances in the place of ΕΙ in the other dialects (the reverse took place among the Bœotians), as in ποίη, πλήων, μήων,[1937] ὄρηος, Λύκηος (Alcman), κοσμῆν, κατοικῆν (Theocritus, and the Byzantine Decree in Demosthenes[1938]), δήρας for δείρας in [pg 423] the treaty of the Latians in Crete,[1939] χῆρες in Cretan, and also used by Alcman, κῆνος or τῆνος in Alcman and others; πεπόνθης, ἀπολώλη Theocritus and the Heraclean tables: and thus in contractions from ΑΕΙ, Η has frequently preponderated over Α, as in the pure Doric form ὁρῆν,[1940] ἡ καρδία παδῆ Sophron;[1941] although it must also be allowed that the diphthong ΑΕ was contracted into Η, as in ὅρη, &c. ἦραι for ἆραι,[1942] and ἐνίκη for ἐνίκαε in a Laconian inscription in Leake's Morea, vol. III. Inscript. n. 71.:[1943] to which instances we should probably add the following cases of crasis, κἠν, κἠπὶ, κἠκ. The reverse of this, which we find in the words πει in Sophron,[1944] and ὅπει in a Corcyrean inscription[1945] for πῆ and ὄπη, is a remarkable variety. The Dorians, consistently with their love for the pure and long Α, were equally partial to the Ω. This letter frequently forms the original sound, as in the accusative case Ἀργείως, Argivos; and hence the abbreviated form θεὸς for θεὼς in Cretan and Coan[1946] inscriptions, and in Theocritus, was probably formed by an elision of the characteristic vowel, as δεσποτᾶς in the first declension. We frequently also find use made of the vowel Ω as a prolongation of Ο, instead of the common form ΟΥ, produced by the elision of consonants: thus in the form of the participle [pg 424] feminine in ωσα, used in Crete and Peloponnesus, and also in the Heraclean Tables, whilst the softer form in οισα, where οι was also derived from οντ (as in the third person plural ναίοισιν, and in the masculine participle τύψαις), was perhaps peculiar to Sicily. Ο also, when followed by Ε, overpowers the latter letter, and is changed into Ω, as for instance in Κοιλῶσσα (a mountain near Phlius), λωτρὸν, ὑπνῶν for ὑπνόεν, Laconian forms in Aristophanes, παμῶχος, and similar words in the Heraclean Tables; though whether this is the case when the Ε precedes the Ο is doubtful, for in εὐορκῶσι and similar forms in Cretan inscriptions, it is ΕΩ, not ΕΟ, which is contracted into Ω. In this case ΕΟ is generally contracted into ΕΥ, or it is changed into ΙΟ, as ΕΩ into ΙΩ; thus μογίομες, λυχνοφορίοντες in the Lysistrata of Aristophanes (according to the old reading), ἐπαινιῶ, ὀμιώμεθα ib., ανιοχιων for ἡνιοχέων in the Laconian inscription in Leake, No. 71. with which compare ἐμμενιῶ in the oath of the Latians, πραξίομεν in the decree of the Istionians, and παμωχιῶ in the Heraclean Tables.[1947] In the above cases there is no reason for assuming any other changes, than from ΕΟ into ΙΟ and ΕΩ into ΙΩ, as the Dorians appear to have been very unwilling to tolerate Ε with Ο; the short Ι, however, before the lengthened Ο must have been particularly suited to their ears. The long Α in Ἀλκμὰν, Ἀτρείδα, Ἀγησίλας, πρᾶτος was without doubt a thick sound between Α and Ο, for which there was no distinct character. The Spartan dialect frequently has ΟΥ for Υ (which change regularly occurs in the Bœotian dialect), as δίφουρα for γέφυρα (Hesychius in v.), φούἱξ for φυσίγξ (Valck. ad Adoniaz, p. 276.), μουσίδδω for μυθίζω (ibid. p. 279.), φούαξιρ (vol. I. p. 384. note f.), μοῦκορ for μυχὸς (Koen p. 343.), καμπούληρ, a species of olive-tree (in Hesychius), derived, I believe, from κάμπτων ὕλην, κάρουα for κάρυα (Hesych. in v.); οὐδραίνει, περικαθαίρει according to Hesych. for ὑδραίνει, τούνη for σὺ (Hesych.), ἀπεσσούα for ἀπεσύν in the letter of Hippocrates (compare Coray ad Plut. Alcib. 28.). ΟΙ for Υ is only found in Ποίθιοι, according to Photius.
5. The consonants in the Doric dialect were in some cases so brought together as to give the words a roughness which was avoided in other dialects, and consequently it possessed more of that ancient fulness of consonants which was preserved with greater fidelity in the Latin language than in the Greek; partly from the neglect of that law, which was so constantly observed by all the dialects of the Greek, that every word should end either with a vowel or semi-vowel. The Doric has at least the ancient form of the participle τιθὲνς (Lat. ns, in ancient Gothic ants), which is quoted as a Cretan and Argive form;[1948] and the preposition ἐνς for in with the accusative (into), which in other dialects was changed into εἰς; but in the Doric it became, by the omission of the final Σ, ἐν in the sense of into, as in Crete and in Pindar,[1949] although Cretan inscriptions of considerable antiquity have εἰς, which appears to have been the usual Laconian form. Thus also the Cretans and Argives formed the future in σπένσω, merely throwing out δ, as a τ is properly omitted in τιθένς.[1950] The Rhegians adopted the same usage from the Messenians.[1951] It is clear that the organs of the ancient Doric race were better fitted for this rough pronunciation than the more delicate ones of the other Greeks, who even changed the Roman Hortensius into Ὁρτήσιος. The same remark may be applied to the word μάκαρς in Alcman (fragm. 66.), and some similar forms.
Another more striking characteristic of the Doric dialect is the aversion to Σ, the σὰν κίβδαλον; hence the Doric lyric poets, Lasus and others, wrote poems without that letter; a practice in direct contradiction with the partiality shown by the Ionians for that sound. To this principle may be traced various other peculiarities: first, the interchange of Σ and Τ, which, however, is on the whole merely a relic of the original dialect, as in the adjectives ἐνιαύτιος and πλούτιος,[1952] [pg 426] in τὺ or τοὺ, tu, in τέσσαρες, quatuor, in the third persons δίδωτι, φατὶ, which still retain this form in Sanscrit (while in the Latin and German languages T is always the last letter of this third person). Also in the name of Neptune the Doric was doubtless the original form, having the same root as πόντος, ποταμός; the original form was Ποτίδας (in Epicharmus and Sophron[1953]), and the Megarian in Aristophanes says Ποτείδας; so also the Corinthians; and hence their colony Ποτειδαία,[1954] Ποτειδᾶν (from Ποτειδάων) was the Spartan and the Rhodian form.[1955] It is singular that in some cases the Dorians also used Σ for Τ, as σᾶτες for τῆτες,[1956] corresponding to which we find σάμερον in Pindar, Theocritus, and the Tarentine dialect (a word, according to Hesychius, synonymous with νῦν); the σὰ for τὰ of the Megarians, and this latter for τίνα is the same change.[1957] It was this aversion to Σ, noticed above, which led the Spartans in the double consonants ΣΤ, ΣΚ, ΣΠ, to reject the Σ and double the other consonant; hence the Laconian forms κτίτταρ for κτίστης, ἐττὰν for ἐς τὰν, ἀμπίτταρ for ἀμφιστὰς,[1958] ἀκκὸρ for ἀσκὸς.[1959] Valckenær lays down the following rule: “literam Σ Lacones in sequentem consonantem non liquidam mutant;” and of this change he finds traces in the Tarentine dialect, to which we may add, that Hecate, according to Hesychius, was there called ἄφραττος, i.e. ἄφραστος. The most interesting example of this change in the Spartan dialect is the form ἄττασι for ἀνάστηθι (derived from ΑΝΤΤΑΣΙ), in which word more than three Laconisms are discernible. With this point is immediately connected the change of Ζ, i.e. ΣΔ into ΔΔ, for instance in verbs in ζω, Laconice—δδω, many instances of which occur in the Lysistrata and Acharneans of Aristophanes. There is no evidence of the same change occurring in verbs whose [pg 427] characteristic is Γ; although the Dorians were induced by analogy and a partiality to the letter Ξ to introduce the termination ξω, where the characteristic letter was not Γ but Δ, which is evident by the formation of the substantive καθίππαξις (as should be read in Hesychius for καθίπταξις), δεικηλίκτας, &c.[1960] Even in the Laconian dialect, however, the soft sound of ΣΔ is used instead of ΔΔ, as ἀγίσδεο, μελισδόμενος, τράπεσδα in Alcman, and in the pretended apophthegm of Lycurgus, ἂν πτωχοὶ μένητε καὶ μὴ μέσδω (i.e. μείζω) ἅτερος θατέρω ἐράῃ κτῆμεν.[1961] It would however be erroneous to suppose, with regard to the mode in which this transition was effected, that the sound of Ζ, when already formed, passed into ΔΔ or ΣΔ. The ancient dialect appears to have had a separate Δ, pronounced with a peculiar compression of the mouth; the Dorians in several cases, agreeing with the Ionians, added the Σ, and formed either Ζ, where the sounds were more combined, or ΣΔ. In other cases the Dorians merely gave additional force to the Δ. With the Æolians there was scarce any distinction between the harsh and the common Δ, as in Δεὺς for Ζεὺς, δυγὸς for ζυγὸς &c.; in the same manner Ζεὺς in the Latin became Deus, ῥίζα radix, ὄζω odor,[1962] and hence the long Ζ was wanting in that language; but the peculiarity of the original sounds of this consonant is evident from the circumstance that the Latins substituted for it I; for example in jugum from ζυγὸς, major from μείζων, &c.; in like manner the Æolic dialect interchanged δια and ζα, καρζὰ, καρδία.[1963] The change of the last letters of verbs ending in -σσω into -ζω in the Tarentine dialect, instead of ττω like the other Dorians, as ἀνάζω for ἀνάσσω, is quite peculiar to that town.[1964]
6. Another mode of avoiding the sound of Σ was to omit it altogether. This suppression was made at an early date in the third person plural, which consequently retained a nearer resemblance to the original form in the Doric than in the Ionico-Attic dialect, in which the preservation of Σ soon caused the ΝΤ to be dropped. Examples of this, as πεινῶντι, ἀποδίδωντι, κεχάναντι, αἰνέοντι (bhavanti, in Sanscrit, corresponding to the ancient high German ant; the Bœotians wrote -ωνθι, -ανθι) are found in all the Doric inscriptions; yet Alcman uses the termination -ουσι as well as the ancient form. Sometimes this elision of σ lengthened the preceding vowel, as in Πηρεφονεία Lacon. for Περσεφονεία, according to Hesychius, with which we may compare πῆριξ for πέρδιξ in the Cretan dialect (ibid.); also πρειγεύτας, πρείγιστος, πρειγηία in Cretan inscriptions for πρεσβεύτης, &c.; the Argives also used Γ for Β in πέργεις. (See Hesychius.) Concerning the omission of Σ before Φ, e.g., φὶν for σφὶν, in the Laconic dialect, see Koen p. 254.; the Syracusans changed the place of the Σ, and converted ΣΦΙΝ into ΦΣΙΝ, i.e. ψίν. This aversion to Σ also appeared in the substitution of the aspirate for this consonant, in which change the pure Doric dialect is directly at variance with the Latin, in which the aspirate was often replaced by Σ, for example, ἅλς, sal, ἡμι, semi, ὕλϜη, sylva,[1965] &c. The Laconians, on the other hand, used μῶἁ, instead of μῶσα, and on the same principle μωἱκὰ, music, as also in the participles κλεῶἁ, ἐκλιπῶἁ, &c, to which we may add ὅρμαὁν for ὅρμησον, as in Aristophanes; also ποιῆἁς, πᾶἁ, βίὡρ for ἵσως,[1966] βουὅα for βουσόα;[1967] the same usage also prevailed among the Argives, as we learn from Dercyllus, among the Eretrians, who borrowed it from the Eleans, and also among the Pamphylians; with whom several Argive and Rhodian peculiarities of dialect appear to have been preserved.[1968] Lastly, with this aversion to Σ is [pg 429] connected the rhotacismus, which we have already observed in the Spartan and Elean dialect, and of which the interpreters of the decree against Timotheus,[1969] particularly Casaubon, have collected many examples. Of these I will only cite ἐπιγελαστὴρ, the mocker; καλλίαρ, an ape (Hesych. in vv. comp. Boeckh Exp. Pind. Pyth. II. p. 251); κιλλακτὴρ, an ass-driver (Pollux VII. 13. 56.); σάριρ, a palm-branch (Hesych.); τίρ, τίς, (ib. and in the Elean Rhetra), παλαιὸρ (Aristoph. Lys. 988.), σιὸρ θεὸς, πὸρ ποῦς, νέκυρ νέκυς, βόμβυρ a kind of flute (Hesych. in vv.). Whether in the oblique cases Σ could always be changed into Ρ is uncertain, since, besides the Elean Rhetra, no genuine monument, and only a few and obscure glosses, afford any information on the point. However, ἀμ᾽ ἀρκᾶρ for ἀπ᾽ ἀρχᾶς (according to Koen's conjecture ad Gregor. p. 283.) is an instance, as also the Cretan τέορ for σοῦ (Hesych.), where the pronoun is declined, as ἐμοῦς, ἐμέος, ἐμεῦς in Epicharmus.[1970] We may observe that generally the Latin is in this respect very different from the pure Doric; though it resembles it in some words. Thus the Laconian ἀκτὴρ is the Latin actor, and in gubernator we see the Doric form κυβερνατὴρ, and so in other instances.[1971]
7. Notwithstanding this fuga sibili—this aversion to the Σ—to which almost all the changes mentioned in the last two sections may be traced—yet the Doric dialects always retained in the first person plural the final Σ from the ancient language (as is proved by the Latin -mus);[1972] and Laconians, Megarians, and Doric Sicilians said ἥκομες, ἀπορέομες, &c. It does not appear that in the Doric dialect any original consonant passed into Σ, except Θ; and this change probably arose from a desire to soften the harsh sound of the aspirate. Instances of this Laconism in Alcman (Ἀσᾶναι, ἔσηκε, σάλλεν, σαλασσομέδοισαν), in the Lysistrata (ἤνσε, [pg 430] ἔλση, σιγεῖν, μουσίδδειν, &c.), and the grammarians (e.g. σὶνκασεύδει, κασαίρηὁν, for καθαίρησον, according to Koen, κασαρεύειν, according to Valckenær) are well known, and particularly σεῖος ἀνήρ; comp. Valckenær, p. 277, sqq. who has treated this point with great ability. Also in Hesychius, συμβουαδεῖ, ὑπερμαχεῖ (for συμβοηθεῖ) we should probably write συμβουασεῖ (otherwise Hemsterhuis), and κασελατίσαι, καθίσαι, ibid. is from ἕλλα, ἕλα, κάθεδρα, sella; whence ἑλατίζειν, καθελατίζειν, sedere facio. In this respect the colonists of Sparta at Tarentum did not follow the idiom of their mother city; as they said θυλακίζειν, not συλακίζειν, to beg:[1973] the Rhodians also retained the original Θ in ἐρυθίβη (Strabo XIII. p. 613. Eustath. ad Il. α. 34.): in Cretan this change only occurs in σεῖναι for θεῖναι in Hesychius, and in σίος in the treaty of the Olontians: for Corinth may be cited Σίσυφος for Θεόσοφος, according to Phavorinus, p. 403. Dindorf; for Sicyon perhaps σειρὸν, θέριστρον, Hesych. and also στίαι for θριαὶ, Schol. Apoll. R.h. II. 1172. That the Eleans were acquainted with this variety has been shown above.