Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Roth.Mixed_Aorists_in_Homeric_Greek.1990.
One class of “mixed aorists” consists of the imperatives ἄξετε and ἄξεσθε, οἶσε οἴσετε and οἰσέτω, ὄψεσθε (the singular imperative ὄψεο· ἰδέ often cited as from Hesychius is actually a conjecture of Cobet’s, which Latte does not accept in his edition  ); the secondary tense form ἄξοντο; and the infinitives ἀξέμεν and ἀξέμεναι, οἰσέμεν and οἰσέμεναι.  These all have stems which are normally restricted to the future. The aorist of ἄγω is ἤγαγον, not ἦξα, except in the mouth of a barbarian (Timotheus Persians 165). The verbs” bear” and” see” are inherited from Indo-European as suppletive (cf. Latin fero, (te)tuli , latus ): οἴσω and ὄψομαι regularly provide their futures. A root with durative aspect, like those of φέρω and ὁράω, was evidently unsuited to the future as well as to the aorist.  The stems οἰσ- and ὀψ-, like ἀξ-, are rarely and weakly attested as sigmatic aorists. In Herodotus l.l57.3, ἀνοῖσαι is a conjecture for ἀνῶσαι in the manuscripts. Pindar has ἐπόψατο (fr. 78-9), presumably an artificiality based on the epic forms. The subjunctive ἐπιόψωνται (Plato Laws 947c) comes from a different verb, and perhaps directly from I 167 ἐπιόψομαι. The evangelist Luke (13.28) uses a subjunctive ὄψεσθε (p75 and many mss, others ὄψεσθε or ἴδητε). None of these has any value as evidence for the Homeric and earlier stages of the language. In Homer, the imperatives οἴσετε, οἷσε, and ἄξετε are clearly associated with futures, as at Γ 103-5 (οἴσετε … οἶσομεν … ἄξετε) or χ 101-6 (οἴσω … δώσω … οἶσε). So we must connect these Homeric forms with the future system in Greek, and not with Sanskrit thematic imperatives of the sigma tic aorist (nesa etc.). 
In the case of οἴσετε we might expect to shed some light on the problem by looking into the origin of the root. Unfortunately, however, the etymology of οἴσω is unknown. Meillet calls it a racine obscure;  Frisk, ohne Etymologie.  Although it has been compared with Sanskrit véti, chercher à atteindre, this comparison is open to the objection that neither the Arcadian inscription (IG 5.2.3, see below, chapter two) nor the epic shows any sign of a digamma.  Prellwitz suggests * oit-so from the root * i – “go”, comparing Sanskrit étah “hurrying” and Latin utor “I use” (Old Latin infinitive oitier).  He also suggests that οἶτος “fate” would be to οἴσω as Latin fors , fortuna is to ferre. Semantically, he interprets utor as “I advance myself (my interests)” (ich fördere mich) and οἴσω as “I shall advance (something)” (ich werde fördern). Ribezzo considers οἴσω a causative or factitive formation, again derived from * i – “go”; that is, οἴσω should mean “I cause (something) to go.”  He compares Sanskrit ésyati “vuol fare andare, mette in movimento,” identical with the Atharva-Veda future of eti “go.” Magnien expresses no opinion on the etymology, but lists οἴσομαι as the basic form (in accordance with the tendency for futures to be of middle voice) and translates it je me porterai, j’irai as if he thought it might be the future of εἶμι.  The only Homeric passage he cites for this meaning is Θ 400: … οὐ γὰρ καλὰ συνοισόμεθα πτόλεμόνδε. Cf. also [Hesiod] Scutum 358: … εἰ δὴ νῶι συνοισόμεθα πτολεμίξειν, and, with διοίσομαι, H. Merc. 255: ἐπεὶ τάχα νῶι διοισόμεθ’ οὐ κατὰ κόσμον. From this intransitive middle he derives the transitive active οἴσω je porterai and in turn the transitive middle οἴσομαι je porterai pour moi. This scheme is theoretically attractive but is based on no real evidence. The meaning of “going” belongs to the compounds συμφέρομαι and διαφέρομαι, not particularly to the future οἴσομαι. It is next to impossible to prove a relation with a root containing only the one sonant i. Furthermore, the existence of οἰστός and οἰστέος (not to mention the medical term οἰσοφάγος) suggests that the root of οἴσω is ois– (or oi + dental) rather than oi-.  That is, the s does not reflect the suffix * -se/o- , although it may have helped οἴσω become established as a future tense. Originally, οἴσω would have been a present whose meaning had some affinity with the future, like εἶμι. There are other words in Greek with a base ois -: οἶμα from * ois-ma “rush” (Π 752, Φ 252) and οἶστρος “gadfly” — literally, an instrument which causes something to rush. The nouns οἴμη and οἶμος more likely belong to εἶμι, like οἶτος. As Frisk says, οἶμα and οἶστρος, together with Avestan aesma and Latin ira (*eisa ), imply an Indo-European verb * ois- “to set in violent motion.  One way of setting something in motion (though not in violent motion) is to fetch it; thus there might be a semantic connection between the hypothetical * ois- and οἴσω as a present with the meaning “I fetch,” which then could have an imperative “fetch.” As will be seen, “fetch” is something like what οἶσε means in Homer. But the connections are too tenuous to be convincing, and οἴσω is clearly a future in the Homeric state of the language. Let us see what we can learn from the Homeric evidence.
To discover the original function of the imperatives and related forms in the Greek epic language, one must analyze their functions in the oldest attested instances. To pick out the oldest instances, one must demonstrate that some of the usages are derivable from others. The secondary types can then be set aside, and the investigation be restricted to the basic types.
The first type that can be eliminated is the indicative. Only one example occurs Θ 545: έκ πόλιος δ’ἄξοντο βόας καὶ ἴφια μῆλα. If this form were considered by itself, one might call it an “imperfect of the future” and compare it to the conditional tenses in Sanskrit and Old Irish, as Magnien does.  But it is more economical to suppose, with Debrunner, that the line is derived from Θ 505:  έκ πόλιος δ’ἄξεσθε βόας καὶ ἴφια μῆλα. As a general rule, one can predict an imperfect third person plural in -οντο corresponding to a given second person plural imperative in -εσθε. This can be simply expressed by an analogical diagram:
|ἄγεσθε||→||ἄγοντο (unaugmented imperfect)|
Furthermore, the middle imperative itself occurs only once, and may well be secondary to the active imperative ἄξετε;
This does not affect the independent status of ὄψεσθε, which has no corresponding active form.
Similarly, the infinitives ἀξέμεν(αι) and οἰσέμεν(αι) can be eliminated from consideration. It should be noted that, where they are used as future infinitives (Ψ 668, Ψ 221), the forms do not give any difficulty. The problem arises when they are treated as present (or aorist) infinitives. They nearly always appear in indirect commands. The obvious explanation for this restriction is that they are derived from imperatives, by the transformation which turns direct commands into indirect.  This again can be expressed diagrammatically:
x= οἰσέμεν κελεύει
Compare Γ 103, with the imperative: οἴσετε ἄρν’, ἕτερον λευκόν, ἑτέρην δὲ μέλαιναν and Γ 119f, with the infinitive: … ἠδ’ ἄρνε κέλευεν//οἰσέμεναι. In just this way the expression εὐφημεῖν κελεύειν was derived from the imperative εὐφήμει (εὐφημεῖτε) and Latin iubeo saluere from salue ( saluete ).  The athematic infinitive ending -μεν(αι) on a thematic stem, though not unusual in Homer, nevertheless supports the conclusion that the infinitives are a relatively late development.
Two passages show a freer use of these infinitives. In Z 53: δώσειν ῷ θεράποντι καταξέμεν· ἀλλ”Αγαμέμνων the infinitive is both a reflection of a command and an adaptation of a formula like that in E 26: δῶκεν ἑταίροισιν κατάγειν κοίλας ἐπὶ νῆας. As for Ω 662f. – … τηλόθι δ’ὔλη//ἀξέμεν ἐξ ὄρεος – this also is an extension from the indirect commands. The expression ἀξέμεν ὔλην (Ψ 111, cf. Ψ 50 and Ω 778) undergoes a transformation embedding it in the nominal sentence ὔλη τηλόθι. So all the infinitives can eventually be derived from the imperatives; if these can be explained, the rest will follow.
If the imperatives occurred in repeated formulae, we would have some evidence for their antiquity in the epic language. We might be able to establish a hierarchy of formulae and thus simplify the data still further. As it turns out, this is not possible. There is no real formulaic pattern in Homer. The impression may arise, however, that οἶσε (οἴσετε) could be an old word in ritual contexts. In Homer the objects of οἶσε, οἴσετε, and ἄξετε are usually offerings, water, firewood, or fire. The usage of Callimachus and Theocritus has some similarity to the Homeric. A person is asked to fetch something for ritual purposes. What is desired may be fire: Ο 718 οἴσετε πῦρ, ἅμα δ’αὐτοὶ ἀολλέες ὄρνυτ’ ἀὕτήν χ 481 οἶσε θέειον, γρηῦ, κακῶν ἄκος, οἶσε δέ μοι πῦρ Theoc. 24.48 οἴσετε πῦρ ὅτι θᾶσσον ἀπ’ ἐσχαρεῶνος ἑλόντες or water: υ 153f. … ταὶ δὲ μεθ’ ὔδωρ ἔρχεσθε κρήνηνδε, καὶ οἴσετε θᾶσσον ἰοῦσαι γ429 ἔδρας τε ξύλα τ’ἀμφὶ καὶ ἀγλαὸν οἰσέμεν ὔδωρ Call. H. 5.47f. σάμερον αἱ δῶλαι τὰς κάλπιδας ἢ ‘ς Φυσάδειαν ἢ ἐς ‘Αμυμώναν οἴσετε τὰν Δαναῶ or sacrificial animals: Γ 103 οἴσετε ἄρν’, ἕτερον λευκόν, ἑτέρην δὲ μέλαιναν cf. Γ 120 or ritual implements: Call. H. 5.l5ff. μὴ μύρα λωτροχόοι τᾷ Παλλάδι μηδ’ἀλαβάστρως… οἴσετε μηδὲ κάτοπτρον· ἀεὶ καλὸν ὄμμα τὸ τήνας Call. H. 5.31 οἴσετε καὶ κτένα οἱ παγχρύσεον, ὡς ἀπὸ χαίταν … In Callimachus Hymn 6 (to Demeter), οἶσε is used in prayer to the goddess: 6.136 φέρβε βόας, φέρε μᾶλα, φέρε στάχυν, οἶσε θερισμόν. The religious usage in Callimachus and Theocritus may well have been derived from Homer, but that is not the only possibility. The Alexandrian poets had access to traditional poetic and religious material, so here their evidence may be independent of Homer. Perhaps οἷσε belonged to some Doric tradition; Callimachus Hymns 5 and 6 and Theocritus Idyll 24 are apparently in Doric dialect, and Pindar uses the infinitive οἴσειν in a present sense (Pythian 4.102). Another piece of evidence is the inscription from the temple of Athena Alea at Tegea in Arcadia (IG 5.2.3):  εικ επι δομα πυρ εποισε δυοδεκο δαρχμας οφλεν … “If anyone brings fire to the temple, let him owe twelve drachmas …” Here εποισε is a subjunctive — one can hardly say a subjunctive of the nonexistent sigmatic aorist. Someone might like to connect the inscription with the poetic passages, and to posit an old ritual formula οἴσετε πῦρ. It may be objected, however, that the inscription is not concerned with a ritual use of fire but with the avoidance of accidental conflagration, such as occurred in 395 B.C. (Pausanias 8.45.4).  One of the Homeric passages also deals with conflagration, namely 0 718: οἴσετε πῦρ, ἅμα δ’αὐτοὶ ἀολλέες ὄρνυτ’ ἀὕτήν· where Hector is urging his men to set fire to the Greek ships. The subjunctive εποισε, then, probably does not reflect a phrase for the ritual use of fire, but simply the confusion of future indicative and aorist subjunctive. This confusion is widespread in Homer; consider, for example, B 229ff: ἧ ἔτι καὶ χρυσοῦ ἐπιδεύεαι, ὅν κέ τις οἴσει Τρώων ἱπποδάμων ἐξ ‘Ιλίου υἷος ἄποινα, ὅν κεν ἐγὼ δήσας ἀγάγω ἢ ἄλλος ‘Αχαιῶν … So the inscription does not shed light on the Homeric imperatives. Furthermore, the poetic imperatives themselves do not show sufficient uniformity to justify positing a ritual usage of any great antiquity. It will be necessary to look for their origin within the epic tradition. P. Wahrmann reached this same conclusion, but she did not offer satisfactory motivation for the development in the epic language. 
Leumann, following Debrunner’s opinion of ἄξετε explains οἴσετε, ἄξετε, and ὄψεσθε as imperatival futures.  In some instances it would be possible to translate οἴσετε “you shall bring” (e. g. Γ 103), ὄψεσθε “you shall see” (e.g. Ω 704). The objection to this view is that the imperatival future does not seem to be a Homeric usage. Schwyzer can cite no clear examples from Homer of an imperatival future in the second person; the nearest thing is the future used in the sense “(afterwards) you may do…” (Z 70, Ω 717).  If the “mixed aorists” are the only Homeric examples of the imperatival future, we had better look for some other explanation.
To determine what place these imperatives have in the systems of the verbs to which they belong, we may compare their functions with those of the regular forms of the verbs. For ἄγω an important fact at once emerges. For the most part, ἄγε and ἄγετε do not function as the imperatives of ἄγω. Rarely does either the singular or the plural occur alone as an imperative meaning “bring.” Usually they occur in combination with another imperative or the equivalent, with a very much weakened meaning. They have been degraded to the status of mere particles. An example typical of many is B 72, with a short-vowel aorist subjunctive: ἀλλ’ ἄγετ’, αἴ κέν πως θωρήξομεν υἷας ‘Αχαιῶν. Notice that no serious metrical difficulty was experienced with ἄγετε, contrary to what Wahrmann implies.  For the imperative of ἄγω in its primary function, a more strongly marked form has been introduced, namely ἄξετε. The original imperative ἄγε (ἄγετε) survives in the secondary function as an adverb, where its connection with the verb ἄγω can be forgotten.  Thus there is a good reason for the spread of ἄξετε, but its origin has not yet been explained.
The particular usage of these imperatives must be examined more closely. When ἄξετε is used, it is addressed to someone nearby, telling him to go and get something from another place and bring it back: Γ 105 ἄξετε δὲ Πριάμοιο βίην, ὄφρ’ ὅρκια τάμνη (Menelaus addresses the Trojans on the field; Priam is in the city.) Ω 778 ἄξετε νῦν, Τρῶες, ξύλα ἄστυδε … (Priam and the Trojans are in the city; the wood is outside.) In contrast, when rarely ἄγε is used as a real imperative, it involves leading someone who is present away to another place: κ 266 μή μ’ἄγε κεῖσ’ ἀέκοντα, διοτρεφές … (Eurylochus does not want to go with Odysseus to Circe’s house.) o 199 μή με παρὲξ ἄγε νῆα, διοτρεφές… (Telemachus does not want to go home with Peisistratus.) Likewise, οἴσετε and οἷσε are used of going to get something and bringing it back, in opposition to φέρετε and φέρε, which tell someone to bring an object which he already has near him: P 718 νεκρὸν ἀείραντες φέρετ’ ἐκ πόνου … (All those involved are out on the field together.) μ 9f. δὴ τότ’ ἐγὼν ἑτάρους προίειν ἐς δώματα Κίρκης οἰσέμεναι νεκρὸν ‘Ελπήνορα τεθνηῶτα. (Odysseus and his companions were on the shore; the body was in Circe’s house.) φ 369 ἄττα, πρόσω φέρε τόξα· τάχ’ οὐκ εὖ πᾶσι πιθήσεις (Eumaeus has the bow with him.) χ 106 οἶσε θέων… (τεύχεα) (Telemachus has to go inside to get the weapons.) Similarly, ὄψεσθε directs the hearers to go somewhere and look at something: Ω 704 ὄψεσθε, Τρῶες καί Τρῳάδες, Ἕκτορ’ ἰόντες (The Trojans are to come from their homes to look at Hector’s body.) Θ 313 ἀλλ’ ὄψεσθ’, ἵνα τώ γε καθεύδετον ἐν φιλότητι (The gods are to come from their homes to look at Aphrodite and Ares.) In sum, all these “mixed aorist” imperatives involve a sense of going in order to do something.
Sometimes, indeed, the imperatives are accompanied by a participle “going.” This is an important clue. Clearly, there is a relation between Ω 704 ὄψεσθε … ἰόντες and (for example) Ξ 200: εἶμι γὰρ ὀψομένη πολυφόρβου πείρατα γαίης, and between υ 154: ἔρχεσθε κρήνηνδε, καί οἴσετε θᾶσσον ἰοῦσαι and N 167f: βῆ δ’ἰέναι παρά τε κλισίας καὶ νῆας ‘Αχαιῶν οἰσόμενος δόρυ μακρόν, ὅ οἰ κλισίηφι λέλειπτο. Consider also θ 254ff: Δημοδόκῳ δέ τις αἶψα κιὼν φόριγγα λίγειαν οἰσέτω, ἥ που κεῖται ἐν ἡμετέροισι δόμοισιν. ὣς ἔφατ’ ‘Αλκίνοος θεοείκελος, ὦρτο δὲ κῆρυξ οἴσων φόρμιγγα γλαφυρὴν δόμου ἐκ βασιλῆος. The same relation exists between η 188: νῦν μὲν δαισάμενοι κατακείετε οἴκαδ’ ἰόντες and A 606: οἱ μὲν κακκείοντες ἔβαν οἷκόν δε ἔδαστος.  The verb κατακείω was properly called a desiderative in the Homeric state of the language, whereas the future had lost its original desiderative character in its finite forms. The future participle (like the infinitive) continued to have a desiderative sense, and was frequently used to express intent, often with a verb of going; e.g. A 12f: … ὁ γὰρ ἧλθε θοὰς ἐπὶ νῆας ‘Αχαιῶν λυσόμενός τε θύγατρα φέρων τ’ἀπερείσι’ ἄποινα. Here the future participle obviously expresses an intent rather than a future fact: the ransoming did not take place on this occasion. The construction of a verb of going with a future participle is common in Homeric and later Greek, both prose and verse. Another construction involving “to go” and some other verb seems to be specifically poetic: the combination of an imperative with a participle “going.” Homeric examples are reasonably frequent; to select three: Γ 406 ἧσο παρ’ αὐτὸν ἰοῦσα, θεῶν δ’ἀπόεικε κελεύθου Z 490 ἀλλ’ εἰς οἶκον ἰοῦσα τὰ σ’αὐτῆς ἔργα κόμιζε I 42lf. ἀλλ’ ὑμεῖς μὲν ἰόντες ἀριστήεσσιν ‘Αχαιῶν ἀγγελίην ἀπόφασθε – τὸ γὰρ γέρας ἐστὶ γερόντων. Compare also Sophocles Antigone 768: δράτω, φρονείτω μεῖζον ἢ κατ’ ἄνδρ’ ἰών and Oedipus at Colonus 1393: καὶ ταῦτ᾿ ἀκούσας στεῖχε, κἀξάγγελλ’ ἰών … The natural (unmarked) way of saying such things is with two imperatives; for example: Γ 432 ἀλλ’ ἴθι νῦν προκάλεσσαι ἀρηίφιλον Μενέλαον Ψ646 ἀλλ’ ἴθι καὶ σὸν ἑταῖρον ἀέθλοισι κτερέιζε. In opposition to this ordinary kind of expression, the type ἧσο ἰοῦσα is stylistically marked. People did not talk that way. Basically, however, it means the same as if one said ἴθι (καὶ) ἧσο. An optional transformation subordinates the verb of going by making it a participle. In fact, the phenomenon is a kind of enallage of verb endings comparable to the well-known poetic device of changing an adjective ending (the “transferred epithet”). With a desiderative verb, the situation is a little different. The unmarked construction consists of the finite verb of going plus the desiderative participle (κακκείοντες ἔβαν, like ἧλθε λυσόμενος). The desire to lie down is not really parallel to the act of going, but rather indicates a state of the subject who goes; so the natural construction involves subordination rather than parallelism. One might compare the constructions with λανθάνω. From a Greek point of view, λανθάνειν expresses a circumstance attending the accomplishment of another verbal idea, so that the natural construction would be the participle λαθὼν with a finite verb (e.g. Thucydides 1.65 ἔκπλουν ποιεῖται λαθὼν τὴν φυλακήν). The common reversal of this (e.g. Thucydides 2.2 ἔλαθον ἐσελθόντες), which is an inherited construction, seems less natural to Greek. Accordingly, the ordinary imperative of κακκείοντες ἔβαν should be something like *ἴτε κατακείοντες, as ἴτω θύσων is of εἷσι θύσων (Plato Laws 909d). Instead, we find κατακείετε … ἰόντες, with the verbs reversed (cf. ἔλαθον ἐσελθόντες). Evidently a stylistic transformation, a kind of enallage like that which produces ἧσο … ἰοῦσα, operates on the imperative “go” plus desiderative participle to produce a desiderative imperative plus a participle “going.” The remarkable fact is that this same enallage is applied to “go” plus a future participle, even though future imperatives do not normally exist in Greek. Thus, instead of *ἴτε ὀψόμενοι, Homer says ὄψεσθε … ἰόντες. The basic meaning remains the same, so that the imperative ὄψεσθε keeps the desiderative sense characteristic of the future participle but not of the future indicative. This “future imperative” still does not exist in the verbal system of the language; such existence as it has in the epic poetry depends on the sense that it is derived from the type εἷμι ὀψόμενος. In much the same way a “transferred epithet” construction is understood with reference to the normal construction. If one can say Θήρωνος ‘Ωλυμπιονίκαν ὕμνον (Pindar 01.3.3), knowing that it will be understood as Θήρωνος ‘Ωλυμπιονίκᾶ ὕμνον, it does not follow that ‘Ωλυμπιονίκας ὕμνος is an expression possible in the language.
Something more can be said about the conditions influencing the creation of these “future imperatives.” The regular imperatives of the verb “to go” ἴθι and ἴτε, like ἄγε and ἄγετε, tend to have very little meaning of their own when combined, as they often are, with other imperatives. There would have been some motivation to create a stronger kind of imperative, as there was to replace ἄγετε with ἄξετε. In particular there would have been an inclination to place in the emphatic initial position of the verse some word which carried more meaning than (ἀλλ’)ἴτε. The transformation that produces ὄψεσθε … ἰόντες has the effect of putting the marked grammatical category (the imperative) and the marked lexical item (the looking, as opposed to the going) into the same word and that in the most strongly marked position in the verse. Generally speaking, doubly (or multiply) marked structures tend to replace the simpler structures which get weakened by frequent use. 
Furthermore, if ἴτε had little lexical value, at least it still had the syntactic function of expressing the imperative. When it is transformed to a participle, it loses even that function. In οἴσετε … ἰοῦσαι (υ 154), οἴσετε carries all the meaning that matters. Consequently, the participle can be omitted; οἴσετε alone is derived from the type οἴσετε … ἰοῦσαι. The participle is omitted in many of the instances of οἴσετε (οἷσε) and all of those going with ἄξετε. Yet in each case the notion of going is preserved, because it is only the derivation from the compound structure which makes the use of the simple imperative possible.  It may indeed be that ἄξετε was formed by analogy with οἴσετε rather than independently from a transformation of *ἴτε ἄξοντες. The verbs φέρω and ἄγω are commonly associated, as in the idiom ἄγειν καὶ φέρειν; and ἄξετε in Γ 105 follows closely upon οἴσετε (103). But that makes no essential difference. The explanation of the structure is the same whether ἄξετε belonged to it from the beginning or was a later addition.
Once the forms have entered the epic tradition, they can survive in a limited way even if the sense of derivation is lost. Antimachus of Colophon uses οἰσότων in a passage reminiscent of Θ 254ff. (fr. 19 Wyss): ἄλλοι δὲ κρητῆρα παράργυρον ἠδὲ δέπαστρα οἰσόντων χρύσεια, τὰ τ’ἐν μεγάροισιν ἐμοῖσι κείαται. Pindar uses οἴσειν as present with no apparent special meaning (Pythian 4.102): … φαμὶ διδασκαλίαν Χίρωνος οἴσειν … And Aristophanes puts οἶσε into the mouth of Lamachus as parody of the grand style (Acharnians 1099; cf. also Ach. 1101, 1122, and Frogs 482): ἄλας θυμίτας οἶσε παῖ καὶ κρόμμυα. For these poets οἶσε is opposed to φέρε not because it has any different meaning but simply because it helps to characterize the style as that of high poetry. It is meant to be recognized as a Homeric word; it is no longer part of a living poetic language.
[ back ] 1. Cobet 365; Hesychius (Latte) gl. 2095.
[ back ] 2. The substance of this chapter appeared in Glotta 48 (1970) 155ff.
[ back ] 3. Meillet, Festschrift Kretschmer 141.
[ back ] 4. Wackernagel, Kleine Schriften 809f.
[ back ] 5. Festschrift Kretschmer 141.
[ back ] 6. Frisk 2.370.
[ back ] 7. Boisacq 694; Vollgraff, BCH 70 (1946) 623ff.
[ back ] 8. Prellwitz 326; Pokorny 295.
[ back ] 9. RendNap n.s. 21 (1907) 229f.
[ back ] 10. Magnien 1.102ff.
[ back ] 11. Meillet, Festschrift Kretschmer 141.
[ back ] 12. Frisk 2.369.
[ back ] 13. Magnien 2.lf.
[ back ] 14. IF 40 (1922) I11f.
[ back ] 15. Wahrmann, Festschrift Kretschmer 311.
[ back ] 16. Benveniste 278, 284.
[ back ] 17. Sokolowski no. 67, pp. 135ff.
[ back ] 18. Danielsson 54f.
[ back ] 19. Festschrift Kretschmer 312.
[ back ] 20. Leumann, K1eine Schriften 240.
[ back ] 21. Schwyzer 2.291.
[ back ] 22. Festschrift Kretschmer 312.
[ back ] 23. Kuryłowicz, Esquisses 1inguistiques 79.
[ back ] 24. Kuryłowicz, Esquisses 1inguistiques 79.
[ back ] 25. Kuryłowicz, Esquisses 1inguistiques 70.
[ back ] 26. Kuryłowicz, Esquisses linguistiques 75.