The first act is a play. A good play. Poetic, strange, compelling in the way that the best symbolist theater compels. Everyone who has read it agrees on this point. The second act is something else entirely. Everyone agrees on this too, but none of them can articulate what they mean. Some of them can't articulate anything after that.

Catalogued under designation THE KING IN YELLOW, this text represents the highest-priority memetic threat in active files. It is not a grimoire in the traditional sense. No rituals, no summoning circles, no instructions for communion with entities beyond the veil. It is a play. Two acts. The danger is not in what it teaches. The danger is in what it does to the reader, silently, irreversibly, between one page and the next.

The Text Itself

Two acts. The first introduces Carcosa. A city on the shore of the Lake of Hali, beneath twin suns that set in impossible colors. Characters wear the Pallid Mask. They speak of the Yellow Sign. The theatrical tradition is recognizably symbolist. Maeterlinck by way of nightmare, Ibsen if Ibsen had glimpsed something through the curtain he could not un-see. The dialogue is beautiful. The staging notes are precise. Reviewers who stopped at Act I called it a masterpiece of avant-garde theater, a work of staggering originality that deserved wider production.

Reviewers who finished Act II did not write reviews.

The first act functions as literature. It earns its audience through craft. This is not incidental. This is the mechanism. The play is engineered, if a pre-human intelligence can be said to engineer, to ensure the reader reaches the second act willingly, eagerly, having been given every reason to trust the text.

The Second Act

Not entirely clear, even to those who have read it. Language shifts. The vocabulary remains English, or French, or Latin, depending on the edition, but the syntax begins to operate on principles that are not linguistic. Sentences produce meaning that exceeds their component words. Narrative perspective collapses. The text addresses the reader directly, and does so with a specificity that a printed document should not possess.

Readers report imagery that is personally relevant in ways that defy explanation. A woman in Paris described her childhood bedroom in detail the text could not have contained. A professor at the University of Vienna encountered a passage describing the precise arrangement of objects on his desk. The desk he was sitting at while reading. The consensus among the B̵̈͡P͔̊͝K͔̍M͒͗̌ꫀI̯͊̔Ç̷̒R̵̡͌ẌE̊ documented second-act exposures is unanimous: the second act is not fiction. It is not theater. It is an address from Hastur to you, individually and specifically. As though the text had been waiting for you to open it.

The line between reading a play and being read by one is the line the second act erases.

Authorship

Unknown. No manuscript exists. No playwright has ever been identified, and no theatrical tradition claims it. The earliest documented edition appeared in 1895, but references in marginalia of older texts suggest the play, or something ancestral to it, has existed in some form for considerably longer. Provenance trails terminate in contradictions. Printers who produced editions deny knowledge of how the typeset pages arrived at their shops. Booksellers recall stocking copies they have no record of ordering.

The play may be a translation of something non-human. The original language, if language is even the correct term, has never been identified. Whoever wrote it may not have had a choice in the matter. Whoever translated it almost certainly did not.

Containment Status

Periodically suppressed. Periodically resurfaces. The pattern is old enough to suggest that containment is not a strategy the text permits. Copies confiscated in one decade appear in estate sales the next. Editions burned in public ceremonies are found, intact, in the private libraries of the officials who ordered the burning. The text resists containment the way water resists being held. Not through force, but through the fundamental nature of what it is.

The digital age has introduced monitoring concerns that previous generations of archivists did not face. Partial transcriptions circulate on forums whose operators are unaware of what they host. Full-text scans have been intercepted, but the interception rate is not one hundred percent. It will never be one hundred percent. The King in Yellow does not want to be contained. The King in Yellow has never been contained. Every suppression effort in recorded history has functioned, in retrospect, as distribution.