Alaska’s Lady in Blue: How Baranof Castle Became Haunted

Joe Nickell

Baronof Castle, Sitka, Alaska, ca. 1893. The castle burned to the ground on March 17, 1894.

Among the earliest recorded ghost stories in Alaska is the tragic tale of the lovelorn bride of Baranof Castle. I encountered the promontory once topped by that historic site—and later learned of its captivating legend—when I arrived at Sitka in 2006 on a Center for Inquiry cruise.1

Baranof Castle

The great rock outcrop known as Castle Hill, now officially the Baranof Castle State Historical Site, provides a commanding view of Sitka Harbor and the surrounding city. Historically, families of a native Tlingit clan lived there, but after the Russian-American Company led by Alexander Baranov (or Baranof) sought to establish a trading post, war followed. Finally, the outnumbered Tlingits ceded the promontory to the colonizing Russians for their seat of government (“Castle Hill” 2018).

The Russians destroyed the Tlingit houses there and built a succession of frame buildings, then the two-story brick “castle” or Governor’s House in 1836. It had a cupola that enabled it to also become a lighthouse (see Figure 1). In 1867, Russian Alaska was acquired by the United States, and the transfer ceremony was held at the site. The castle was occupied by U.S. Army commanders for a decade, then remained an administrative center until it burned in 1894. Alaska statehood came in 1959, and—unofficially and in secret—the first forty-nine-star American flag was raised on the site (“Castle Hill” 2018).

It was there, in the days of the old castle, that the ghost story began. Accurate details are hard to come by—and that is a story in itself.

The Legend

As early as the 1880s, there began to appear versions (or as folklorists say, “variants”) of what we now recognize as the Blue Lady legend—although we must not be startled to learn that, initially, she was a ghost of a different color. The story is told in one of a series of travel letters that appeared in newspapers in 1883–1884 and were collected into a book two years later. Its author was a noted woman of her day, Eliza Ruhamah Scidmore.

Scidmore (1856–1928) was an American writer, photographer, and inveterate traveler who became the first woman to serve on the board of the National Geographic Society. In 1883, she spent some valuable time investigating the ghost of Baranof Castle. Scidmore’s account begins by observing shrewdly that the “sad story” of the “beautiful Russian” lady “is closely modeled on that of the Bride of Lammermoor”—that is to say, on Sir Walter Scott’s 1819 novel of that title. (The novel involves a young couple and another’s plot to take advantage of the man’s absence to trick the woman into an arranged marriage, with tragic consequences.2)

Scidmore (1885, 160) writes of the Castle ghost:

At Easter time she wanders with sorrowful mien from room to room, and leaves a faint perfume as of wild roses where she passes. Innumerable young officers from the men-of-war have nerved up their spirits and gone to spend a solitary night in the castle, but none have yet held authentic converse with the beautiful spirit, and learned the true story of unresting sorrow. By tradition, the lady in black was the daughter of one of the old governors. On her wedding night she disappeared from the ball-room in the midst of the festivities, and after long search was found dead in one of the small drawing-rooms. Being forced to marry against her will, one belief was that she voluntarily took poison, while another version ascribes the deed to an unhappy lover, while, altogether, the tale of this Lucia of the norwest isles gives just the touch of sentimental interest to this castle of the Russian governors.

Not only was the ghost originally the Lady in Black, but at the same time she was only a governor’s daughter—although, as we shall see, she evolves in retellings into a Russian “princess.”

Scidmore’s account also contains another illuminating piece of information. At the time of her writing (1883), the castle had been long abandoned and neglected. Indeed, she writes, it “has been stripped, despoiled, and defaced … and the place is little better than ruin.” I think this is very important as it suggests how the place began to be “haunted.” Many times I have traced what I call “the abandonment-creates-haunted-place phenomenon” (Nickell 2017). A site without any apparent ghost becomes abandoned and run down until it looks like a dark and spooky place. Soon there are whispers of ghosts, and people are dared to visit and even spend the night.

This is apparently just what happened with Baranof Castle. Scidmore goes on to state that the local Russian residents “cannot identify this ghost with any member of the governors’ families, and say that the whole thing has been concocted within a few years to keep sailors and marauders away at the night, and to entertain the occasional tourist.”

The Lady in Blue

Other sources have recounted the story of the castle ghost, including one “in the early eighties” by Frederick Schwatka, reprinted in Alaska News, December 24, 1896. I first came upon the tale in a book Strange Stories of Alaska and the Yukon (Ferrell 1996, 7–71), and it quoted a passage from John Arctander’s The Lady in Blue (1911).

This is the fullest, book-length elaboration of the legend, and it purportedly derives from an “old yellow manuscript” of a late Army chaplain named Cramer. However, while it appears to give the tale an earlier history, the “old manuscript” is a well-known literary device (like Poe’s “Ms Found in a Bottle” or the recounting of a legend in Conan Doyle’s The Hound of the Baskervilles), which allows the author to create a story within a story. I was suspicious that it was fictional, and in examining the questioned text (that of the purported old manuscript) in comparison with that of author Arctander, I found such similarities as to persuade me that Arctander wrote it all.

Specifically, I assessed the similarity of the “two” texts by first conducting a standard readability-scale analysis of each. This measures the length of words and sentences to provide an indication of the education level of the intended audience and, by inference, of its author (Bovée and Thill 1989). For accuracy, I chose two sample passages from Arctander and two from the purported chaplain’s manuscript, and I averaged the results.3

The indicated education levels, for Arctander 14.5 and for the chaplain 15, were remarkably close and show that both were consistent with the same level of fifteenth grade—that is, college junior, and therefore common authorship. (Naturally, Arctander would have been writing for a popular rather than scholarly audience.)

As it happens, both Arctander and his “manuscript” writer produced uncommonly long sentences—the former as long as 127 words and the latter one of 92. In each, the longest sentences were a single paragraph.

There are other similarities between the two writings. The purported manuscript sometimes uses an outdated British form of double punctuation—a colon, comma, or other mark immediately followed by a dash. This is said to have lingered in British legal documents (“English Language and Usage” 2018). While the usage might seem to support the antiquity of the manuscript, the telling fact is that Arctander (1849–1920) was born in Sweden and apparently learned English at a university in Norway. He became an American attorney of some note.

Arctander (1911, 13) knows the high quality of the writing in the old manuscript—after all, he wrote it!—and he has fun with this by having the purported chaplain’s son-in-law say, “The old gentleman must have been as good a writer in his younger days as he was a preacher when I first met him … .” Indeed, Arctander really acknowledges the pretense of the whole thing by labeling the book, in its subtitle, “A Sitka Romance”—a romance being “any long, fictitious tale of heroes and extraordinary or mysterious events” (Morris 1970). Never mind if others might mistake his book for history rather than fiction.


Some have tried to make the ghost a real person—attributing her to this or that earlier Russian administration. Like Arctander, writers have promoted her to “princess” and suggested that she was “cruelly separated from her true love and forced into a loveless marriage”; however, “Apart from that, storytellers don’t seem to agree on much, including whether the ghost loiters about in a black, blue, or white dress” (“Ghost” 2015).

Numerous such story variants—e.g., the lady took poison, stabbed herself, or was murdered (Scidmore 1885; “Ghost” 2015)—reveal the narratives to be folklore and/or fakelore at work. We learned from Scidmore (1883) that the tale grew out of the period when Baranof Castle was abandoned, and so it became a spooky place where young soldiers gathered the courage to spend a night. She said this began “after the troops left,” which was 1877 (Andrews 1922, 24). Therefore, we can place the probable time of the ghost story’s origin to 1877–1883. (It might have been earlier, but there is no prior version in print, so I think Scidmore’s statement—that local residents say it was concocted “within a few years” of abandonment—is credible.)

While the tale seemed historical, recall that locals could not “identify this ghost with any member of the governors’ families” (Scidmore 1885, 160). Another source adds skeptically, “One might also imagine that the presence of a Russian princess in Alaska might have been commented on by more historians over the years” (“Ghost” 2015). So, what appeared historical only resulted from the narrative having been set back in time (as Arctander and others did) when it was, in fact, simply made up.



I am grateful once again to CFI Libraries Director Tim Binga for his expert help in obtaining published materials.


  1. The cruise, featuring our “Planetary Ethics” conference, snaked from Seattle along the coastal reaches of British Columbia and southern Alaska. I spoke on “Mysterious Entities of the Pacific Northwest” and, as opportunity allowed when we pulled into ports, did some on-site investigating (Nickell 2007a; Nickell 2007b).
  2. Scott’s novel has a more complex plot, however, and the bride’s name is “Lucy,” not “Lucia.”
  3. For Arctander, I chose a 119-word segment, pp. 9–10, “Southwestern … archipelago”; and another of 112 words, p. 7, “I had … else.” For the “Manuscript,” I selected passages of 106 words, pp. 24–25, “Reaching … hallway”; and 110 words, pp. 22–23, “How long … skirt.” Arctander’s level was 14.5 (the average of 15 and 14), and the chaplain’s manuscript was 15 (the average of 17 and 13). Note that even within a given person’s writing, there can be significant variation, which is why I did the analysis twice and produced an average.


  • Andrews, C.L. 1922. The Story of Sitka. Seattle: Lowman & Hanford Co.
  • Arctander, John. W. 1911. The Lady in Blue: A Sitka Romance. Seattle: Lowman & Hanford Co.
  • Bovée, Courtland L., and John V. Thill. 1989. Business Communication Today. New York: Random House, 125–127.
  • Castle Hill (Sitka, Alaska). 2018. Available online at (Sitka,_Alaska); accessed November 9, 2018.
  • English Language and Usage. 2018. Available online at https:/; accessed November 15, 2018.
  • Ferrell, Ed. 1996. Strange Stories of Alaska and the Yukon. Kenmore, WA: Epicenter Press.
  • The Ghost of Baranoff [sic] Castle: Love’s Labor Lost. 2015. Available online at;the-ghost-of-alaska-baranoff-castle-loves-labor-lost/; accessed November 9, 2018.
  • Morris, William, ed. The American Heritage Dictionary of the English Language. 1970. New York: American Heritage and Houghton Mifflin.
  • Nickell, Joe. 2007a. Mysterious entities of the Pacific Northwest, part I. Skeptical Inquirer 31(1) (January/Febraury): 20–22, 60.
  • ———. 2007b. Mysterious entities of the Pacific Northwest, part II. Skeptical Inquirer 31(2) (March/April): 14–17.
  • ———. 2017. Haunted Buffalo asylum. Available online at; accessed November 2018.
  • Schwatka, Frederick. 1896. An account of the Baranof Castle ghost; reprinted in Alaska News (Juneau), December 24, 1896.
  • Scidmore, Eliza Ruhamah. 1883. From a series of letters published in St. Louis Globe-Democrat summers of 1883–84. Later republished (Scidmore 1885).
  • ———. 1885. Alaska: Its Southern Coast and the Sitkan Archipelago. Boston: D. Lothrop & Co.

Joe Nickell

Joe Nickell, Ph.D., is Senior Research Fellow of the Committee for Skeptical Inquiry (CSI) and “Investigative Files” Columnist for Skeptical Inquirer. A former stage magician, private investigator, and teacher, he is author of numerous books, including Inquest on the Shroud of Turin (1998), Pen, Ink and Evidence (2003), Unsolved History (2005) and Adventures in Paranormal Investigation (2007). He has appeared in many television documentaries and has been profiled in The New Yorker and on NBC’s Today Show. His personal website is at

Among the earliest recorded ghost stories in Alaska is the tragic tale of the lovelorn bride of Baranof Castle. I encountered the promontory once topped by that historic site—and later learned of its captivating legend—when I arrived at Sitka in 2006 on a Center for Inquiry cruise.1 Baranof Castle The great rock outcrop known as …

This article is available to subscribers only.
Subscribe now or log in to read this article.