One special subgroup of vocalic nouns are those that end in the vowel e, or “e-nouns”. Frequently these have inflectional suffixes identical to those of other vocalic nouns, but they differ significantly in plural forms. In particular, they do not (normally) use the plural suffix -r, but instead form their plurals by changing their final e to i. The details of these forms and their historical developments are discussed in the entries on plural nouns and for the individual noun cases.

Origins of e-nouns: There are two sets of primitive nouns that would produce e-nouns: those that primitively end in ē (and ĕ if this vowel survived into modern Quenya) and those that primitively end in ĭ, because short final [i], [u] became [e], [o] in the phonetic development of the Elvish languages. For purposes of discussion, let us label these subgroups as ē-nouns and ĭ-nouns. The second subgroup differs in its singular forms and in compounds, where the final ĭ is usually preserved. Thus for the nouns lasse “leaf” and súre (súri-) “wind”:

• lasselanta “autumn, (lit.) leaf-fall” (LotR/1107) from ancient ✶lassē + ✶lantā.
• súriquesse “wind feather, a species of grass” (TAI/197) from ancient ✶sūrĭ + ✶kwessē.
• Instrumental singular lassenen, locative singular lassesse (Plotz).
• Instrumental singular súrinen (LotR/377), locative singular súrisse (TAI/197).

The plural forms of ē-nouns and ĭ-nouns would be the same in most noun cases, but would differ in the adverbial cases (allative, ablative, locative) where the plural suffixes -nnar, -llon, -ssen were appended directly to the noun stem:

• Plural lassi and instrumental plural lassínen, but locative plural lassessen (Plotz).
• Plural *súri and instrumental plural *súrínen, but locative plural *súrissen.

Although the plural forms of ĭ-nouns are not attested, the values given above seem the most likely. No special inflectional rules are needed for ĭ-noun, provided you use the correct noun stem (such as súri-) where appropriate.

Conceptual Development: Tolkien discussed the distinct inflections for ē-nouns and ĭ-nouns in his early 1930s document, the Declension of Nouns. They mostly follow the patterns described above, except that in this conceptual period, the plural adverbial case suffixes were added to the plural rather than stem, so they were identical in the two subgroups. For lasse “leaf” and sire (siri-) “river”:

• Plural lassi, instrumental plural lassínen, locative plural lassissen (PE21/6).
• Plural siri, instrumental plural sirínen, locative plural sirissen (PE21/10).

However, the ĭ-nouns were further subdivided into two subgroups based on whether the final ĭ was preceded by a single consonant (siri-) or a cluster (lindi-). Where there was a cluster, they largely behaved as described above for declensions from the 1950s and 60s: singular suffixes were added to the stem. Where preceded by a singular consonant, they sometimes behaved like the ı̯ǝ nouns described in the entry on vocalic nouns, with consonantal inflections intruding for some noun cases and the final ĭ becoming asyllabic ı̯ [y]:

• Accusative sirya vs. normal vocalic sire.
• Genitive siryo vs. normal vocalic sirio.
• Dative siryen vs. normal vocalic sirer.

But Tolkien indicated these were archaic variants, replaced by the forms of ē-nouns, with the ancient ĭ vowel preserved only in the singular adverbial cases: sirinta, sirillo, sirisse (the allative suffix was -nta in this conceptual period).

Conceivably, similar inflections may exist in Late Quenya, so that the genitive singular of súre might be *súryo rather than *súrio. I consider that unlikely, however, and would use súrio in my own Neo-Quenya writing.