HW is a voiceless w , as in English white (in northern pronunciation).
Rule 2
The secondary accents were placed (subject to other considerations) on alternate syllables in either direction from the main stress. The secondaries that preceded the main stress (e.g. in oxytone or paroxytone words) were the strongest.
Rule 4
Strong secondary accent fell on initial or long syllable separated by one or more shorts from stress.
The secondary accent falling on the initial syllable, or any non-initial long syllable, that was separated (by an intervening stressless long or one or more stressless short) from the main was strong.
In the very frequent cases of words ending in two short syllables, […] they received a light stress that could be used metrically.
Exception 1
¶ In Q. when this was added with suppression of aorist-present vowel the expression was very imperious or urgent: tulā́, come at once! oryā́, get up.
In this passage in what Tolkien explicitly calls Quenya, both length and final stress are denoted.
Exception 2
In quadrisyllables and more the stress (chief) can never go farther back than three syllables, except in one case, that of quadrisyllables where the two internal syllables consist of 2 short vowels in hiatus . This is also the only case where a short final vowel bears a strong secondary stress. Areanor is accented Áreanòr , but Areanóre is accented Àreanóre .
This is contradicted by an early draft on Eldarin accent:
But before the development of the Parmaquesta the main stress became fixed. It then fell on the penultimate syllable, when that was long ; where it was short , the main stress fell on the ante -penult . It never fell further back, nor did it fall on final syllables other than monosyllabic words.
But stating that it never falls outside of the main rule would mean a rejection of Exception 1, of which we know the imperative must be stressed finally.
Exception 3
Orome , Óromèa , Óromèo Tampio , Támpiòa
Exception 4
In the earlier stages of Quenya (Lindarin) it appears to have been normally retracted to the initial syllable, though certain prefixes such as al , wa > o , etc. were often not stressed.
Exception 5
roquen is < *roko-kwen with Quenya syncope, *roko being an older simpler form of the stem, found in some compounds and compound names, though the normal form of the independent word ‘horse’ had the fortified form rokko . These compounds being old were accented as unitary words and the main stress came on the syllable preceding -quen […]
Exception 6
But in normal long words, especially those made with derivative suffixes etc., the secondary accent was attracted by a long syllable (not adjacent to stress), and never fell on a short final syllable. A strong secondary accent was borne by a long syllable separated from the main (a) by one syllable, long or short: types màndălḗ , kártastā̀ ; or (b) by two syllables: type kárpalimā̀ . A weak secondary was borne by the furthest of a series of two or three short syllables that preceded the main , types àlakánda , kằlŭmălánda , kằlŭmălṓ .
Note that a long syllable adjacent to the stress did not attract the secondary, so èrĕmăloitḗ .
Note that this paragraph talks about CE stress before the Quenya stress shift. However, we assume that the logic of secondary stress placement stays the same:
The secondary stresses were usually reorganized and governed by the main stress; but on same principles as those described above for the archaic period — except that a secondary accent (often[?] with low tone) could stand on an initial syllable immediately before a long main stress, as in cṑmallo above.
Rule 3 & Exception 7
The initial syllable usually retained some degree of stress. In long words, especially recognized compounds, it was, though lower in tone, often equal in force to the main stress: as in óromárdi , fálmalínnar , etc. It was weaker when immediately preceding the main stress, as in Àndū́ne , ṑmáryo , Tìntálle , Rṑméllo ; and in such cases, if it was short it became unstressed, as in avā́niër .