Raymond Roussel does seem like the kind of writer who would write acrostics.  His works are steeped in wordplay.  The procédé is based on homonyms; “Parmi les noirs” throws in a rebus and a cryptogram; there’s a sonnet with a hidden message in La Poussière de Soleils; and so on.  Acrostics seem inevitable.

And a couple of them have been found.  The word “caca,” for example, is spelled out in “L’Inconsolable” (in the seventh section).

Cette phrase s’étale en or mat: “A remettre
Après la valse du quatrième tableau
Commençant par ce vers: ‘La brise ride l’eau…’
A Mademoiselle Ève, au théâtre des Bouffes.”

This phrase is laid out in matte gold: “To return,
After the waltz in the fourth tableau
Beginning with this verse: ‘The breeze ruffles the water…’,
To Miss Eve, at the variety theater.”

This is certainly a seductive idea, given Roussel’s penchant for scatology: Pangloss’s sojourn in the manure in L’Allée aux lucioles, or many instances in Nouvelles Impressions d’Afrique.  But they’re common letters, and I suspect it’s merely accidental, like the acrostic “shit” that the Baconian William Stone Booth found in The Merchant of Venice.

Some readers also see an acrostic reference to Grasset d’Orcet in “La Meule.”  This is even more tempting: d’Orcet’s “language of the birds” is based on puns; d’Orcet is cited by Fulcanelli; Roussel was friendly with Julien Champagne and Eugène Canseliet, both of whom claimed to know Fulcanelli, and both of whom are suspected of being the man behind the pseudonym.  We could even, if we liked, read an allusion to the “language of the birds” in the strange bird ballet that Alfred describes in Les Noces.  But, when I open “La Meule,” I find this:

D’un piquet planté dans l’établi comme un mât;
On y trouve, servant de réclame at d’appât,
Cet avertissement provocant: “JE REPASSE.”

Of a stake planted in the bench like a mast;
One finds there, serving as advertisement and enticement,
This provocative warning: “I REPASS.”

This is not really an acrostic; and spells “d’Oncet,” not “d’Orcet.”

Acrostics do appear in Roussel’s work, but as a subject, not as a technique.  In the last chapter of Locus Solus, the cock Mopsus coughs out acrostics in blood onto a sheet of ivory.  In Comment j’ai écrit certains de mes livres, Roussel explains how the procédé produced the Mopsus episode, but not why the cock spits acrostics.  But, given the context, it’s likely that that detail came directly or indirectly from the same method.  And it may be telling that Roussel does not give the verses.

There’s a particularly intriguing mention of acrostics in Les Noces (Au Bois de Vincennes I, line 1121).

“Oui, dit Claude, un poète, un faiseur d’acrostiches
Ne pourra pas chanter bien des cheveux postiches,
Pas plus qu’un râtelier complet ou des appas
Suspects… Il vaut beaucoup mieux suivre pas à pas
Sa destinée; un jeune auteur à qui la rime
Et facile n’a plus de souffle pour la frime,
Comment faire sur les cheveux tout un sonnet
Quand on a vu la dame ôter comme un bonnet
Sa perruque de blonde? On a tort quand on lutte
Contre l’âge.”

“Yes,” says Claude, “a poet, a maker of acrostics,
Will not be able to sing well about artificial hair,
No more than about a set of dentures, or suspect
Charms… It is far better to follow, step by step,
One’s destiny; a young author for whom rhyme
Comes easily has no breath to waste on sham;
How could one write a whole sonnet about hair
When one has seen the lady remove, like a bonnet,
Her blonde wig? One is wrong when one fights
Against age.”

This is odd for several reasons.  It’s surprising to see Roussel object so to artificiality, when his work is devoted explicitly to the artificial and the imaginary, and to obsessive descriptions, not of things, but of representations of things.  Most of Les Noces itself is taken up with a minute account of a play seen by its supposed protagonists.  The dismissal of blonde wigs is particularly curious, given that Impressions d’Afrique was to offer three transvestites in blonde wigs (Carmichaël, Talou, and Yaour).

Roussel’s unfinished poems reveal that he first established the rhymes, then filled in the rest of the lines: he worked from the ends of his verses, not the beginnings.  And the procédé is developed from that technique; Roussel notes: “Ce procédé, en somme, est parent de la rime.  Dans les deux cas il y a création imprévue due à des combinaisons phoniques.”  “This method, in short, is related to rhyme.  In both cases there is unexpected creation due to phonic combinations.”  Sound is the generator; and the reference to acrostics above is due to the fact that he had a rhyme for it.