Although this poem was not published until the early 1950's, Clark Ashton Smith (CAS) discusses it in letters to his mentor George Sterling from 1913. His initial description of the poem, from a letter to Sterling, is worth quoting*:
I enclose a rather fantastic experiment, in which I've tried the irregular repetition of lines and phrases, and a desultory rhyming of words with themselves. I don't remember seeing it done this way before -- only in an arbitrary stanza, as in Poe.
Sterling responded with some brief criticism in a letter of his own**:
The poem is very weird, poetical and impressive. It's only fault is the "too much Poe." His exact spirit, in fact.
"Luna Aeternalis" does indeed make much more use of literary devices such as alliteration than is typical for CAS' verse, and of course such techniques are strongly associated with the poetry of Edgar Allan Poe (of whom CAS was a great admirer). It seems correct for CAS to label this poem as an experiment, for while it does indeed echo the voice of Poe, it seems not to contain so much of the voice of CAS himself. I like to think he toyed with this approach early in his career, quickly recognized its derivative quality, and returned to creating work that was truly his own.
*See letter #64 in The Shadow of the Unattained: The Letters of George Sterling and Clark Ashton Smith (SU) published by Hippocampus Press.
**See letter #66 in SU.