“Surama of Atlantis” and “The Automatic Electric Executioner” (1953) by H. P. Lovecraft & Adolphe Danziger de Castro

To say that “The Surama of Atlantis” and “The Automatic Electric Executioner” are discoveries is a bit of a misstep: they were never really lost. After de Castro’s death in 1959 his papers made their way to the Hebrew Union College-Jewish Institute of Religion, who in turn donated them to the American Jewish Archives in 1988 as Manuscript Collection No. 348. The excellent inventory of the de Castro collection by Chris Powell in 1996 notes both the existence of the texts and their relation to Lovecraft’s revisions. J.-M. Rajala noted in the 2011 Lovecraft Annual:

2 linear feet of de Castro’s papers, including unspecified manuscripts, are in the American Jewish Archives of the Hebrew Union College of Cincinnati (The Jacob Marcus Rader Center, MS-348), and I wonder if these have been carefully examined by anyone. (56)

Powell had, writing in “The Revised Adolphe de Castro” in Lovecraft Studies #36:

He minimally revised “The Electric Executioner” and retitled it “The Automatic Electric Executioner”. He also revised “The Last Test”, creating “The Surama of Atlantis”. He made minimal revisions to most of the story but made more substantial changes at the beginning to describe the origin of the shadowy character, Surama, and to the final outcome of the story. (24)

Powell also noted that:

“Surama of Atlantis” is planned to be included in The Nyarlathotep Cycle being edited by Robert M. Price for upcoming release by Chaosium. (24n14)

However, “Surama of Atlantis” and the other texts were not published in The Nyarlathotep Cycle or anywhere else, though Price thanks Chris Powell in his introduction. There is likely a story there, but the end is that “Surama of Atlantis” has remained in obscurity to the present day.

The original texts by Adolphe de Castro which Lovecraft worked from are those published within In the Confessional and the Following (1893). As is characteristic of Lovecraft, he completely rewrote both “The Automatic Executioner” and “A Sacrifice to Science.” Presumably this would also have been the case with the third revision, though no text of this revision is known to survive.

This work would initially have been done by hand; though no manuscript copies survive, and were later typed by someone (de Castro for “The Last Test,” Lovecraft for “The Electric Executioner”) for submission to Farnsworth Wright at Weird Tales. These final typescripts are also non-extant, so it is not clear what editorial changes, if any, that Wright made when they were published in the pages of Weird Tales. Martin Andersson has pointed out that there are also minor differences between the Weird Tales texts and the version of the stories published in Something About Cats and Others (1949, Arkham House); these differences are not reflected in the AJA texts, so we can be reasonably certain de Castro did not reference the Arkham House text.

For the three new typescripts, the variations from the Weird Tales text include variant titles, spellings (and misspellings), typographical errors, and changes in phrase and formatting that range from slight to rewriting entire paragraphs. The most substantial differences are with “Surama of Atlantis,” which is about 500 words longer, adding a relatively substantial beginning scene and slightly expanded ending, among other changes.

It is difficult to say who is responsible for the differences between the Weird Tales texts and the AJA typescripts; part of the differences (typos, dropped and repeated words, etc.) can be put down to typist error, but not the insertion or substitution of phrases and entire passages. As these do not appear in the Weird Tales texts, they are either survivals from a previous draft (Lovecraft was known to do multiple drafts of stories), or were added in afterwards (almost certainly by de Castro). The possibility of both cannot be ruled out; that is de Castro may have re-typed “Surama of Atlantis” from an older draft of “Clarendon’s Last Test” by Lovecraft and made his own alterations on top of that. The very-unLovecraft-like passage that ends “????” is almost certainly from de Castro.

The passengers on the Satsu Maru cascaded down the gang-plank, glad to be once more on American soil. There was a slight pause in the flow and then a tall thin man, wearing a gray ulster and a fedora hat that shaded his bespectackled eyes, short nose and bearded chin appeared.

Following him was a pretty young woman, dressed in gray with a large straw hat, the brim held down over the ears by a wide blue ribbon. Her left hand held the chain of a gold mesh bag, while her right clutched the collar around the neck of a magnificent St. Bernard dog.

Closely following them was an individual, tall beyond the ordinary, garbed in a long black cape that hung on his shoulders, and covered his entire body, what was seen of his face when the wind lifted the wide brim of his soft large slouch hat, was shocking; it indicated that the head had no hair. His eyes, like glinting black obsidian, were set so deep in the sockets that they seemed black pools in a cavernous skull. In fact, a closer view strengthened the assumption that it was a skull. there was no nose other than a depression and there were no lips over the large yellow teeth.

A moment he stood still, gazing at the sunlit wharf, at the people, and the large Hotel bus which was stationed a short distance from the gangplank.

As the black-clad skeleton halted, it irritated the bespectackled gentleman who turned and said, “What makes you so slow, Surama?”

The individual called Surama, grinned horribly and said, “Coming, doctor.”
—Adolphe Danziger de Castro, “Surama of Atlantis,” Jewish American Archives (MS-348)

What parts of “Surama of Atlantis” were written by Adolphe de Castro, and why? What parts might be strange survivals from an earlier Lovecraft draft? Is there any way to tell? Objectively, no. Without access to Lovecraft’s original manuscript, it is impossible to say definitively one way or another. Yet we can say a few things.

Lovecraft did not make any fuss over substantial errors when “The Last Test” was published, so we can assume the text in Weird Tales is predominantly as he wrote it in the final draft. Also, given that the first title Lovecraft mentions in his letters is “Clarendon’s Last Test,” it is apparent that “Clarendon” was not a change made by de Castro to the manuscript sent to Weird Tales. In “Surama of Atlantis,” the doctor’s last name is “Schuyler” rather than Clarendon. It is also notable that the name “Schuyler” does appear in the Weird Tales script, as Alfred Schuyler Clarendon and Frances Schuyler Clarendon both attest. This oddity could be the result if Schuyler was the original name, and that Clarendon was then added later—but this presumes two drafts, an early draft and a final one. In any case, it is notable that at no point does de Castro revert to the names in “A Sacrifice for Science” (i.e. Clinton for Clarendon/Schuyler, et al.)

The character of Surama evolved from the character of Mort in “A Sacrifice to Science,” but the Atlantean background was pure Lovecraft—“De Castro wanted it excluded at first—but as it turned out, that was the one thing which Wright singled out to mention in describing the tale!” (Dawnward Spire, Lonely Hill 165) Much of the variant material in “Surama of Atlantis” focuses on Surama, from the title and the opening scene to the bizarre ending. If we do accept the idea that there was a previous draft, and that “Surama of Atlantis” retains several features from it, the most obvious are the ones that describe turtle-like attributes to Surama. In “The Last Test” only a single such descriptor exists:

Unlike the ideal subordinate, he seemed despite his impassive features to spend no effort in concealing such emotions as he possessed. Instead, he carried about an insidious atmosphere of irony or amusement, accompanied at certain moments by a deep, guttural chuckle like that of a giant turtle which has just torn to pieces some furry animal and is ambling away toward the sea.

In “Surama of Atlantis,” however, numerous references of turtle-like characteristics are applied to Surama. The idea of Surama as a kind of monstrous turtle-man works within the logic of the story except for one part: the double-aftermath.

In the afternoon the leisurely firemen overhauled the ruins and found two skeletons—or rather one human skeleton with the skull intact with the frame; of the other only the skull—a very human skull, but with osseous outlines disturbingly suggesting a saurian of some sort. The skull reminded people of Surama, and only well-cut clothing could have made a body as indicated by the skeleton look like a man.

An added horror to the situation was a big hole found under the stout fence back of the destroyed building, and Pat McMonigall, the street car watchman, returning from his beat, assured neighbors that he had seen “a turtle as big as a house” ambling down the hill to the bay, a statement that was not take quite seriously, although Pat McMonigall was a rather abstemious chap. Did Surama go back to Atlantis????
—Adolphe Danziger de Castro, “Surama of Atlantis,” Jewish American Archives (MS-348)

It is completely incongruous for Surama to make his way back down to the sea and to leave behind a skull and skeleton for paleontologists to muddle over, at least not unless Surama was wearing the skull like a hat and peering out through the eye-sockets. If this is a survival from an earlier draft, it isn’t clear how this discrepancy might be resolved—either de Castro wrote the entire ending himself (the unLovecraftian line “Did Surama go back to Atlantis????” is almost certainly his; it reads like an annotation accidentally copied onto the line), or possibly de Castro borrowed verbiage from the Weird Tales version and grafted it onto the earlier draft to give a double-aftermath where there was only one before.

It is also worth noting that in the Weird Tales version, it is specified “that’s all that can reach him, James, unless you can catch him asleep and drive a stake through his heart.” A very vampire-like touch which is at odds with Surama as a kind of turtle, but also much simpler and perhaps more reasonable in keeping with his corpse-like appearance. This is changed in “Surama of Atlantis,” with the suggestion that no ordinary weapon can pierce his shell.

A less evident but more serious problem in “Surama of Atlantis” is the question of narration: who is the narrator? Lovecraft gives no specific identity, the story is related anonymously. Yet the beginning of “Surama of Atlantis” specifies that the narrator is a reporter who was injured by Surama and has not long to live, while the ending states:

Dalton probably gave Dr. Jackson an inkling of the truth, and that good soul had not many secrets from his son, who is the writer of these lines.” (ibid.)

It is a very unLovecraftian mistake; but is it the case of a botch in trying to merge an earlier draft with the later Weird Tales text, or de Castro failing to notice the discrepancy as he added his own additions to the story? Notably, there is a reporter who is injured (though not seriously) by Surama in the course of the story, who plays a key role in events; however, if Lovecraft provided such an opening, would he not have also included a suitable closing? It seems an odd plot hook to leave hanging.

It is important to note several places where the Weird Tales text has expanded on the same text in “Surama of Atlantis.” The appointment of Schuyler/Clarendon to San Quentin is given much more space in “The Last Test,” which foreshadows the governor’s political struggles and losing his appointment power. The benefits of Schuyler/Clarendon’s appointment are noted at greater length, as is the emotional argument when Dalton asks for Georgina’s hand in marriage. Instructing Dalton to blot out the Greek passages but send Miller the notebooks makes much more narrative sense than just telling Dalton to blot out and burn everything, as happens in “Surama of Atlantis.” These are the kind of changes which reinforce the narrative as a whole—and they are the kind of changes that one would expect to see between earlier and successive drafts.

It is the omissions as much as anything which suggests that de Castro was not working directly from a copy of Weird Tales. There does not seem to be any narrative reason to have not copied these sections as-is, since they have little overall impact on the other changes in the story. Yet this cannot be taken proof positive of an earlier draft; it could simply be that de Castro made all the changes on his own. It is notable that of the major changes, two—the extended opening and ending—provide an identity for the narrator of the story. The burrowing-turtle aftermath leaves the story open for a hypothetical sequel.

The rest of the changes are minor, and a couple are mysterious. “The Last Test” refers to the Royal Hotel, which burned down in 1906; “Surama of Atlantis” instead refers to the Phelan Building—which also burned down in 1906, but was then rebuilt in 1908 and still survives today; it’s not clear what benefit one has over the other, since both were still standing in the 1890s when the story takes place. The focus at on “microbio death” is a bit weirder:

“Don’t look so shaken up, old fellow! A veteran politician-fighter like you must have seen plenty of unmaskings before. I tell you, I never had even the start of a fever cure. But my studies had taken me into some queer places, and it was just my damned luck to listen to the stories of some still queerer people.
—Adolphe Danziger de Castro & H. P. Lovecraft, “The Last Test”

“Don’t look so shaken up, old fellow! A veteran politician-fighter like you must have seen plenty of unmaskings before. I tell you, I never had even the start of a fever cure. But my studies had taken me into some queer places, and it was just my damned luck to listen to the stories of some still queerer people who preach microbio death.
—Adolphe Danziger de Castro, “Surama of Atlantis,” Jewish American Archives (MS-348)

It seems to be an effort to provide some reason for Schuyler’s mania with watching people die from the black fever, but the phrase makes so little sense in context it’s hard to see what the author was getting at. If that phrase was a survival from a previous Lovecraft draft, dropping it for the simpler obsession on the power of life and death over a patient seems much cleaner.

The searchers had found the place only because of the chanting and the final cry. It had been close to five that morning, and after an all-night encampment the party had begun to pack up for its empty-handed return to the mines. Then somebody had heard faint rhythms in the distance, and knew that one of the noxious old native rituals was being howled from some lonely spot up the slope of the corpse-shaped mountain. They heard the same old names—Mictlanteuctli, Tonatiuh-Metzli, Cthulhutl, Ya-R’lyeh, and all the rest—but the queer thing was that some English words were mixed with them. Real white man’s English, and no greaser patter. Guided by the sound, they had hastened up the weed-entangled mountainside toward it, when after a spell of quiet the shriek had burst upon them. It was a terrible thing—a worse thing than any of them had ever heard before. There seemed to be some smoke, too, and a morbid acrid smell.

Then they stumbled on the cave, its entrance screened by scrub mesquites, but now emitting clouds of fetid smoke. It was lighted within, the horrible altars and grotesque images revealed flickeringly by candles which must have been changed less than a half-hour before; and on the gravelly floor lay the horror that made all the crowd reel backward. It was Feldon, head burned to a crisp by some odd device he had slipped over it—a kind of wire cage connected with a rather shaken-up battery which had evidently fallen to the floor from a nearby altar-pot. When the men saw it they exchanged glances, thinking of the “automatic electric executioner” Feldon had always boasted of inventing—the thing which everyone had rejected, but had tried to steal and copy. The papers were safe in Feldon’s open portmanteau which stood close by, and an hour later the column of searchers started back for No. 3 with a grisly burden on an improvised stretcher.
—Adolphe Danziger de Castro, “The Automatic Electric Executioner,” Jewish American Archives (MS-348)

Unlike “Surama of Atlantis,” the changes between the three texts of “The Automatic Electric Executioner”/“The Electric Executioner” are much more minor, although strangely a bit more complicated since there are three texts to work with, and in addition to the small but substantial changes, both of the typescripts in the de Castro Archive contain numerous typos and errors of spelling, as well as idiosyncratic formatting differences.

The 1930 text of “The Electric Executioner” in Weird Tales may be assumed to be the oldest of the texts; “The Automatic Electric Executioner” text is bound in a manuscript dated 1953, and may be assumed to be the newest. However, the undated, unbound typescript of “The Electric Executioner” does not sit neatly between the two; textually, the undated text and the Weird Tales text follow each other more closely than “The Automatic Electric Executioner” and the Weird Tales text, but the undated typescript contains several small but notable additions and rephrasing not in either of the other texts.

Most likely, this means that both “The Automatic Electric Executioner” and the unbound text of “The Electric Executioner” represent two different branches of transcription, both copied from the same source (either a copy of Weird Tales, or the typescript received from Lovecraft) and copied and altered at different times, without reference to one another. This would explain the differences and similarities between the three texts without requiring any hypothetical earlier drafts. (Why de Castro would type out “The Automatic Electric Executioner” fresh without making reference to the undated typescript, which has differences from the published version, is another questionbut not one with any ready answers.)

Of the substantial differences, they are few: the title of “The Automatic Electric Executioner” is a combination of “The Automatic Executioner” and “The Electric Executioner”; the story is set in 1899 in the standalone typescript of “The Electric Executioner” and 1889 in the others; the discovery of the body scene in “The Electric Executioner” is a bit longer, and there are some minor geographical differences.

S. T. Joshi in his annotations for this story notes that the San Mateo Mountains are actually in New Mexico; and this is apparently an error on Lovecraft’s part. De Castro’s “The Automatic Executioner” has the protagonist go to Mexico City, and from there towards Orizaba which is in the Sierra Madre Oriental range, although it is never named. “The Electric Executioner” standalone typescript has the nameless narrator headed both to Guadalajara and via Guadalajara to Mexico City—it isn’t clear why the change was made, but was obviously a bit of geographic confusion. In one text the narrator goes to the The Fonda Nacional (“National Inn”) and in the other to the Hotel Interbide; both were hotels in Mexico City. Why the change from one to the other is also unclear.

Something perhaps notable is that all three texts retain the odd racism against Mexicans expressed by Arthur Feldon in the story, which reflects something of Lovecraft’s own prejudices and understanding regarding Mexicans and Native Americans—a combination of racial and class prejudice. From De Castro’s other writings, he either agreed with these generally or at least appears to have felt no need to alter them, as the key phrases (“I hate greasers but I like Mexicans!” etc.) remain intact in every textual variation. Treatment of Mexican characters in “A Sacrifice to Science” (right down to using the slur “greaser”) would seem to suggest no major disagreement between de Castro and Lovecraft on the matter.

Aside from noting how difficult some of the Nahuatl and quasi-Nahuatl names appear to have been for de Castro to type, the most interesting part about the variations on “The Electric Executioner” is simply their existence. There isn’t any evidence that the undated typescript came from Lovecraft’s typewriter (at least, the number of typos would seem to argue against it, given Lovecraft’s punctiliousness), and the variations between the texts are comparatively minor. While it is not impossible that Lovecraft was responsible for some of the bits that don’t appear in the Weird Tales text, the changes are so small and affect so little of the story, compared to “Surama of Atlantis” that like as not the average reader would miss them on a read-through unless specifically pointed out.

As a point of hardcore Lovecraftian scholarship and nerdism, “Surama of Atlantis” and “The Automatic Electric Executioner” provide interesting insight on how changes to Lovecraft stories can not work, and perhaps reflect on the difficult process of drafting and revision. That these stories have gone unpublished is perhaps not surprising; the audience for a variorum of such texts is small, and the rights would presumably remain with de Castro’s estate. But that they exist at all should interest and thrill Lovecraft fans: who knows what else may yet remain, in some dusty archive, or in an amateur journal not yet thoroughly picked-over?


Bobby Derie is the author of Weird Talers: Essays on Robert E. Howard & Others (2019) and Sex and the Cthulhu Mythos (2014).

Leave a comment