16.1These are the basic numbers: 1: minë, 2: atta, 3: neldë, 4: canta, 5: lempë, 6: enquë, 7: otso, 8: tolto, 9: nertë, 10: cainen, 11: minquë. For the number 12, only the stem RÁSAT is given in Tolkien's papers, but there cannot be much doubt that this would yield *rasta in Quenya (Primitive Quendian *rasatâ ). 100 may be *haran - in LotR:1142, we find a reference to "the last [year] of a century (haranyë)". It is not clear from the wording if a haranyë is a century or just the last year of it, but the form of the word itself suggests the latter: If *haran is 100, *haranya would be "hundredth" and haranyë a nominalized form *"hundredth one". The words for higher numbers, such as 1000, have not been published. Numerals from other Elvish languages may not be of much help in reconstructing Quenya words; see WJ:423.
16.2 In LotR:1141 we are told that "the Eldar preferred to reckon in sixes and twelves as far as possible". They evidently often (always, according to WJ:423) used a duodecimal system in which the basic number is twelve rather than ten. In such a system, a number like 11 does not mean "(1 x 10) + (1 x 1)", hence eleven, but "(1 x 12) + (1 x 1)", hence thirteen. The Elvish yén or "long year" of 144 solar years would in a sense be their equivalent of a "century", as 144 is the first three-digit number in a duodecimal system, like our 100.
16.3 In Tengwar, the alphabet native to Quenya, numbers are written "backwards": "The numerals were written according to a positional system like the Arabic, beginning at the left with the lowest number and rising to the highest on the right" (Letters:423). Hence, 61 does not mean sixty-one, but sixteen - or, if the duodecimal system is used, eighteen: 6 + 1 x 12. (SD:129, 135) This must reflect Elvish pronunciation and suggests that for a number like 16 one would say *enquë cainen "six (and) ten" rather than *cainen enquë (or, if duodecimal counting is used, *canta rasta "four [and] twelve" rather than *rasta canta). Greater numbers behave in the same way: In SD:129 we find "6341" meaning 1436, so the Eldarin pronunciation was evidently the equivalent of "six thirty four hundred and thousand" instead of our "(one) thousand four hundred and thirty-six".
16.4 In Sindarin, the various multipla of ten (twenty, thirty, forty etc.) seem to be compounds of words (or stems) meaning "two, three, four" etc. and the word for "ten"; our attested example is nelchaenen "thirtieth" (SD:129), clearly a compound based on the stems NEL "three" and KAYAN, KAYAR "ten". Transferring this principle to Quenya, for which language we lack examples, we may construct forms like these: *attacainen 20, *nelcainen 30, *cancainen 40, *lempecainen 50, *enquecainen 60, *otsocainen 70, *toltocainen 80, *nertecainen 90. (In the case of 30 and 40, I have used the short prefixes nel- and can- mentioned in the Etymologies [NEL, KÁNAT] instead of the long forms neldë and canta; cf. Sindarin nelchaenen. This does not mean that the long forms *neldecainen and *cantacainen are necessarily banned as "wrong".) If we accept *haran as the word for 100, we may similarly construct *attaharan (or *tataran for *tatharan) 200, *neldeharan (or *nelaran for *nelharan) 300, *cantaharan (or *canaran for *canharan) 400, *lempeharan 500, *enqueharan 600, *otsoharan 700, *toltoharan 800, *nerteharan 900. Using this system, we can at least count to 999 in the decimal system (*nertë nertecainen nerteharan).
16.5 The duodecimal system uses multipla of 12 (*rasta) instead of 10: *attarasta 24, nelderasta (or *nellasta for *nelrasta) 36, *cantarasta (or *carrasta for *canrasta) 48, *lemperasta 60, *enquerasta 72, *otsorasta 84, toltorasta 96, nerterasta 108, *cainerrasta (for *cainenrasta) 120, minquerasta 132. The next series, corresponding to the hundreds of the decimal system, would be the multipla of 144, namely 288, 432, 576, 720, 864, 1008, 1152, 1296, 1440, and 1584. But since we do not know the single word for 144, we cannot construct this series. Hence we can only count to 143 in the duodecimal system (*minquë minquerasta, sc. 11 + 132). It has been suggested that hosta might be the Quenya word for 144; its Sindarin cognate host has this meaning (LR:364), but Quenya hosta is simply glossed "large number". If we accept hosta = *"144", we get *attahosta/*tatosta 288, *neldehosta/*nelosta 432, *cantahosta/*canosta 576, *lempehosta 720, *enquehosta 864, *otsohosta 1008, *toltohosta 1152, *nertehosta 1296, *cainenosta 1440, *minquehosta 1584. This would allow us to count to 1727 in the duodecimal system (minquë minquerasta minquehosta, 11 + 11 x 12 + 11 x 144).
16.6 The ordinals evidently have the ending -ya, but only the words for "first, second, third" are attested: minya (Silm:434), tatya, and nelya (originally the three clans of the Elves were simply called Minyar, Tatyar, Nelyar "Firsts, Seconds, Thirds"; cf. also Maedhros' Quenya name Nelyafinwë "Finwë third", PM:352). "Fourth" might be *cantya or perhaps rather *canya, following the pattern of nelya instead of neldya (neldë "three"), indicating that the shortest form of the stem is used to form ordinals (cf. the prefixes nel- and can- mentioned under NEL, KÁNAT - but *nelya instead of *neldya may also be due to the loss of d before y in Noldorin Quenya). "Tenth" may be *cainenya (and hence *attacainenya "twentieth" etc). "Hundredth" is evidently *haranya, the basis of the attested noun haranyë. However, it is not clear how the ending -ya will be added to lempë, enquë, otso, tolto, nertë, minquë. One could use reduced forms of the stems and construct something like *lemya (or *lepya?) "fifth", *entya (for *encya) "sixth", *otya "seventh", *toltya, "eighth", *nertya "ninth", *mintya (for *mincya) "eleventh". Alternatively another adjectival ending, such as -a, might be employed to produce forms like *lempëa, *enquëa, *otsa (*otsëa?), *tolta (*toltëa?), *nertëa, *minquëa. We do not know.
16.7 Besides minya, two other words for "first" are known. Inga is glossed "first" but seems to mean *"foremost, most prominent" rather than simple numerical "first" (the stem it is derived from, ING, is defined as "first, foremost"). There is also yesta, isolated from yestarë *"Firstday". (LotR:1142. In the Etymologies, under the stem ESE, ESET, we find esta instead, but Tolkien marked this stem with a query because it yielded words that clased with another stem, and the word yestarë suggests that he eventually decided to prefix a y.)
16.8 Besides tatya, we also have neuna for "second". Its more basic meaning seem to be *"following"; the stem from with it is derived (NDEW) is glossed "follow, come behind". In most contexts, tatya is to be preferred, but neuna may be used when "second" strongly connotes "following".