Considerable advances have already been made in elucidating the text of the Latin Duenos inscription (6th century BCE). Nevertheless, many difficulties remain, and the ongoing discussion has given rise to several misconceptions that ought to be cleared away. First of all, it is not a votive inscription, in spite of what is sometimes asserted. The fact that the Duenos vase may have belonged to a votive deposit does not imply that the vase and its inscription (incised before baking) were intended for religious purposes. The inscription is not of dedicatory nature. It has been suggested that UIRCO represents some kind of divinity (Proserpina, Athena, or Fortūna), but there is no positive evidence for such an interpretation. It needs to be emphasized that the words DEIUOS “gods” and MITAT (also attested in the Tibur Pedestal inscription) do not allow the conclusion that the vase was initially intended to be a votive offering. The presence of DEIUOS (accusative plural) is to be explained by the mention of an oath (IOUESAT “iūrat”). The verb MITAT (“gives” uel sim. rather than “sends”) is better taken as referring to the aristocratic gift-exchange system. As for the function of the vessel, it has been interpreted as a receptacle for cosmetics. Alternatively, this small piece of elegant design may have been purely decorative and not intended for concrete use, like the Poggio Sommavilla fiaschetta (ST Um 2, around 600 BCE). Several linguistic hypotheses have to be rejected. The alleged infinitives OITESIAI (E. Sievers), OITESI (R. Godel), and OISI (E. Tichy) are excluded on morphological grounds. Since the long final /ī/ of amārī continues *ei, the sequence of six letters PACARI cannot be the ancestor of the Latin passive infinitive pācārī. The sequence UOIS cannot be the source of uīs “you want” because the [o] vocalism could not be accounted for. Since the Latin conjunction nisi is a shortened form of nĭsī (which in turn may derive from nīsī < *neisei with a shortening similar to quăsi < quam sī), a comparison between NOISI and nisi is impossible. (It would of course be unjustified to correct the word to **NOISEI.) Moreover, the /o/ grade in the negation would be unexpected. The putative Umbrian parallel nosue (TIg VIb 54) that has been adduced as supporting the vocalism of the first part of NOISI is not probative since nosue may be traced back to *nauswai < *n(e) au swai < *ne ayu(d) swai or *ne (h)ayu(d) swai, cf. the Latin ne-que haud (Plaut. Men. 371, Ter. Andr. 205; haud < *ĝʰ/gʰi/e h₂éy-u-d, cf. Vedic hí or gha) and the Greek οὐ < *oyu < *h₂óy-u(d). Therefore, the interpretation of NOISI as “nisi” has to be abandoned. It is argued that the second line has to be segmented as follows: AST / TED / NOISI / OPET / OITES / IAI / PACA / RIUOIS. The analysis of OITES [oite(n)ss] as a participle ancestor of the Latin ūtēns (R. Thurneysen, R. Meringer, A. Götze) seems unassailable from a phonological point of view; regarding the omission of the nasal, cf. COSMIS < *konsm° < *kom-sm° (not **ko-smi-s) and also Lat. QETIOS [kʷeŋtios] (cf. Quinctius, not Cottius), perhaps PLACIOM (cf. Plancius), Etr. araz (Rome, S. Omobono, 580-540, cf. aranθ, Arruns). As expected, OITES governs the ablative case (RIUOIS). The fossilized adverbial locative IAI may be equated with the first part of the Umbrian iepi < *iyāi-kʷid (containing the same ending as the Oscan svai), hence a complex conditional system NEI..., AST..., IAI... “if... not..., and/but..., then...” (ASTED < ast + tēd similar to ASTU < ast + tū in the protocols of the Arval Brothers). The verb OPET (< *h₁op-éye-ti, cf. optāre) is construed with the ablative (TED... OPET “(if) she wants from you”). Although it may at first seem counter-intuitive, OPET takes a genitive object according to a construction represented in Early Latin with verbs of desiring (studeat tuī, domī cupiō). Since the girl (still a uirgō) has to be satisfied with the discharge of some fluids (RIUOIS, a classic case of a double entendre), it seems fruitful to turn our attention to the possibility that the word NOISI itself refers to a liquid. It is tempting to establish a connection between NOISI and the root of the Vedic nédati “to overflow” (cf. the Khotanese gganīh- “to moisten”), which is well attested in European hydronymy (Nida in Poland, Niedà in Lithuania, La Nied in Lorraine, Nidd in Yorkshire, cf. also the Lusitanian Langa-nida-eigui). Regular sound laws suggest for NOISI the proto-form *noid-to-s > /noissos/ (as tussis < *tud-ti-s, cf. tundō), with a genitive /noissī/ spelled NOISI. The formation *noid-to-s is structurally comparable to *ĝʰor-to-s (Latin hortus, Oscan húrz, Greek χόρτος, Old Irish gort). A *C(C)oC(C)-to-s formation is precisely attested for the semantically related root *pleu- (Greek πλοῦτος: “flood” > “large quantity” > “wealth”). The use of the partitive genitive is further facilitated when such a meaning is assumed for NOISI (cf. impleto aquae purae). The Duenos inscription thus contains a facetious advertisement with erotic overtones in which RIUOIS (separated from its governing participle OITES to create a punchline) is co-referential with NOISI.