The Clara Nevada: Gold, Greed, Murder and Alaska's Inside Passage

by Steven C. Levi

  The lighthouse on Eldred Rock was built in 1903 and is located between Juneau and Haines on the Lynn Canal. It is appropriately located because Eldred Rock, where the Clara Nevada went down, is the only spot in the Lynn Canal where the rocks break the surface of the water midchannel. It is, in essence, the top of a mountain that rises 1,500 feet from the depths of Lynn Canal. Courtesy of author’s personal collection.

  Headlines of the sinking of the Clara Nevada. Courtesy of author’s personal collection.

  This is the Candle Creek Roadhouse, date unknown. In the early days of Alaska, roadhouses were established a day’s walk from each other. The food was good enough to keep you alive, and the liquor was usually illegal. Courtesy of the University of Washington Nowell Collection 2246.

  Dog “sleds” were not just used during the winter. Here a dog “sled” is used as a water wagon in Nome, 1900. Courtesy of the Anchorage Museum of History and Art B87.7.60.

  Everyone tried to “make a buck” on both the Klondike and Alaska gold rushes. Here are some wily entrepreneurs advertising in the Seattle paper. Courtesy of the Museum of History and Industry in Seattle LB 124087.

  Seaolin was supposed to have been named “Sea Lion,” but the painter had allegedly visited a few saloons on his way to work. Original source unknown, this photo appeared as a reprint in the Alaska Sportsman, May 1966.

  An authentic four-man Alaskan gold mining operation. Note the large steamer and the pipe that directs the steam into the ground. Because the ground was frozen, it had to be thawed before the gold could be extracted. The earth that was thawed was put in a spoil pile, which you can see on the left-hand side of the photograph. Then the spoil pile was washed one pan at a time in water from the creek, which can be seen on the far left side of the photograph. Courtesy of the University of Washington Nowell Collection 143.

  Main Street in Nome in 1900. Note the Dexter on the left hand side of the photograph. That was Wyatt Earp’s saloon. Courtesy of Jim Green.

  Bessie’s Roadhouse in Nome, presumably with Bessie in the wagon on the far left. Women were more than welcomed during the Alaska gold rush. That’s because there were so few of them, about 10 percent of the population. Note the “Ladies Entrance” at the back right of the photograph. Courtesy of the University of Washington Nowell Collection 6650.

  All kinds of cargo were transported to Dawson. Here, during the summer of 1898, a mule train takes crates of turkeys over the White Pass out of Skagway. Courtesy of the University of Washington 8291.

  Cold weather did not stall business during the Alaska gold rush. Courtesy of author’s personal collection.

  Express dog teams from Iditarod in Seward. These sleds were for cargo, not mail. Note the length of the dogsled at the far left. Courtesy of the Alaska Museum of History and Art B81.19.2.

  Here are the United States Army troops who kept order on the Alaska side of the border during the heady days of Klondike strike. This picture was taken in Skagway on July 4, 1899. Please note that these are all buffalo soldiers, black men, and it appears that the officers are black as well. Courtesy of the Alaska State Library PCA-75-144.

  Clara Nevada on her only voyage. Courtesy of the Jefferson County Historical Society, Port Townsend, Washington.

  The Washington-Alaska Bank in Fairbanks. It was the creation of E.T. Barnette, the man in the white hat in the center of the photograph. In 1911, Barnette absconded with all of the money in the bank. He eventually went to trial but was found not guilty. He lived the life of a lord in Mexico for a decade and then moved to Los Angeles, where he died under mysterious circumstances in 1933. Courtesy of the University of Alaska Fairbanks 63-46-147N.

  Lighthouse duty has never been easy. Duty on them could last years, and the men who maintained these structures could go for months without seeing another soul. Eldred Rock is on a very small island, so small you can walk its perimeter is less than five minutes. Sometimes the waves of Lynn Canal are so high that you cannot walk on the island at all. This lighthouse was established in 1903 and was fully operational until the modern era. Today it is a visual reminder of where Alaska’s ghost ship, the Clara Nevada, went down along with 165 people and 100,000 ounces of gold. Courtesy of the United States Coast Guard.

  Generally speaking, there were two types of paddle-wheelers, those with a single paddle wheel at the back, called stern-wheelers, and side-wheelers such as this one. The Eliza Anderson is famous for “the stranger who came aboard.” Just as the ship was about to sink off the coast of Kodiak, a giant of a man broke into the wheelhouse and saved the ship. After the ship was safe in a protected cove, the man disappeared from the ship. Half a century later, a historian discovered documents that indicated the stranger was actually a stowaway, not a ghost. Courtesy of the Alaska State Library DeArmond Collection 134-252-4.

  Alden Blethen, editor of the Seattle Times. Courtesy of the Seattle Post-Intelligencer.

  An authentic Alaska gold rush saloon. This one was in Koyukuk about 1900. Note the pinups of the day on the wall behind the bar. Courtesy of the Alaska Museum of History and Art Wyman Collection 201.

  Photograph of the Alaska Commercial Company store in Arctic Village. Note the mastodon tusks, common discoveries in Alaska where men were digging up thousands of cubic feet of earth every year. Courtesy of the Alaska State Library C.L. Andrew Collection 45-277.

  An authentic Alaska gold rush dance hall, not a roadhouse. This dance hall was in Arctic Circle, so-named because the boomtown founders thought the town was on the Arctic Circle. (It wasn’t.) There are two women in the front, and if you look carefully, you will see a light bulb on the moose above the front door. Also note the boots the men are wearing—this was deep mud country. Courtesy of the Anchorage Museum of History and Art Wyman Collection #202.

  Map of the Lynn Canal and headline of the Clara Nevada missing. Seattle Times, February 16, 1898.

  Teamsters in the days where “teamsters” meant men who drove horses. This photograph was taken in Nome in 1907. Courtesy of the University of Washington Nowell Collection 5885.

  Not all dogsleds were pulled by dogs. Here is a “sheep” dogsled. Contemporary accounts stated that the sheep could pull as much weight as the dogs and were superior on the trail because they could forage for their own food. This meant the musher did not have to carry feed for the animals. Courtesy of the University of Alaska Fairbanks Arne Wahte Collection 64-34-19.

  The loading of the steamships was functional but not elegant. Courtesy of the Anchorage Museum of History and Art Messer Collection.

  This is a shot of incoming prospectors during the Sushanna Stampede in Cordova in 1913. From this photograph, it is clear that landing facilities had improved since the Klondike strike. Of interest because you can compare this photo to others, it is easy to see that these are stampeders coming in. Note their clean boots. Courtesy of the Anchorage Museum of History and Art 377.5.22.

  An authentic Klondike gold rush cache. Men would transport their one thousand pounds of supplies one hundred pounds at a time. They would carry that one hundred pounds for two or three miles and put it in a pile. Then they would return for another load. After the one thousand pounds had been moved the two or three miles, the process would start over again. The word cache, which means to store something until you returned for it, is one of the few Alaskan words to make it into modern usage. Another is hooch, meaning a low quality of liquor that originally came from the Native American village of Hoochenoo located just outside of Juneau. Courtesy of the Anchorage Museum of History and Art Messer Collection.

  This is a look at the legendary ice staircase of the Chilkoot Pass outside of Dyea. At the bottom of the staircase was Alaska; at the top, Canada. This is a significant historical photograph. Most historical texts on the Klondike gold rush state that the stampeders were constantly on the move up the staircase. It was either up or out. If you could not keep up with the pace of men lock-stepping up the staircase, you stepped off the staircase and slid to the bottom. This is supposedly the
origin of the term “getting cold feet.” However, note that there are at least three areas where men are resting alongside the staircase. Also note the half canoe on the back of the man in the foreground. Everything that had to go up the staircase was reduced to its smallest component, in this case half of a canoe. Courtesy of the author’s personal collection.

  Chapter 5

  The Search for Answers

  “History,” I often tell students in my history class, “is not a re-creation of the past. It’s an assessment of the past based on documents provided by people in archives and museums who will answer your letters.” In piecing together the saga of the Clara Nevada, I learned just how true my statement was.

  It did not take me long to realize that something was amiss with the accepted version of the wreck. Just a cursory examination of the facts suggested that something far more sinister than just a shipping disaster. There were too many loose ends.

  Here was a ship with $165,000 in 1898 dollars, reputedly in raw gold, which sinks in twenty-four feet of water on one of the few reefs between Skagway and Juneau where the bottom usually plunges to depths of at least six hundred feet. That the gold was never recovered seemed odd. What happened to the gold?

  In my quest to answer this question, I began to search the historical record for clues to a disaster that was almost one hundred years old. It meant going through all newspapers, archives, memoirs, diaries and letters of the era to discover what documents still remained. What I thought would be a relatively quick examination ended up taking more than three years, and even then I was unsure I had seen every relevant document still in existence. Such is the anxiety of the historian, the fearful belief that as soon as you complete your research and write your findings, someone will find a single document that refutes everything that you have concluded to be true.

  I quickly discovered that the most reliable and unreliable sources at my disposal were the same: the press. Reading the papers still in existence, I found plenty of information, but much of it was contradictory. But that was the smallest of my headaches. It was nothing in comparison to the problems I had in tracing individuals. Names were spelled differently in different sources. Minor facts, such as the geographic location of evidence, were not in sync.

  Take, for instance, the case of the discovery of the only body, that of George Foster Beck. I found three newspaper articles that stated the body had been found in three different areas, which stretched along a forty-mile front. Complicating matters, each of the three articles included critical information the other two did not. This meant that no paper was completely right, but then again, no paper was totally in error either.

  (I also found an article in the Morning Oregonian a month after the sinking that two more bodies had been found, and that one was believed to be that of Captain Lewis. The two corpses were buried without their clothing being checked. No other source verified this story.)

  Then there was the difficulty in actually reading the papers. Some of the newspapers I needed were in Alaska, but they did not constitute the bulk of my research. That came from the Post-Intelligencer and Seattle Times. As both of these were Seattle papers, I had to get the microfilm through interlibrary loan. When I started my examination, to my horror, I discovered that many of the pages were illegible, faint or smeared. This was as much a sin of printing as of microfilming. Some issues of the Seattle Times had dates of the previous day, and occasionally a paper was printed with no date. Tracking from secondary sources and the handful of indexes also proved to be extremely difficult, and often I found the cited March 23 item on March 10 or 15. Sometimes I did not find it at all.

  Worse yet, 1898 was a long time ago, and many local newspapers for that era no longer exist. Often I found that critical days were missing; even more agonizing, sometimes it was months. I was particularly distressed to find that the microfilm of Juneau’s Alaska Miner, a critical paper for the study of the Clara Nevada, had a heart-stopping, gut-wrenching gap from January 22 to March 12 and then another gap until April 9. The March 12 edition, incidentally, contained an article on the discovery of the body of George Foster Beck as well as an editorial on the Clara Nevada. Another Juneau paper, the Searchlight, had only a single issue still in existence in the critical time frame for the Clara Nevada: February 12. The third local paper I could have used was the Daily Alaska Dispatch, whose surviving issues started well after the Clara Nevada went down. If those specific papers exist, I have yet to find them.

  With the Seattle papers, I was luckier. But then again, there were no indexes. Critical articles had to be discovered by reading the papers. This meant well over two hundred hours in front of a microfilm screen, scanning and hopefully finding every article and filler that existed.

  This work is the culmination of that search through the dusty niches of historical documentation in the Pacific Northwest. It is a story based as much on luck as on knowing what I was doing. Unlike many historical works, this book is not based on a single body of documents with peripheral material added for spice. In fact, this entire work is one of peripheral material drawn from a wide variety of sources. I refer to this as scholarship by accident because how successful the writer is depends entirely on how clever he or she is at finding those documents that still exist.

  My initial task was to search the historical record for any clues. I started with the United States census of 1900 for Alaska, Oregon and Washington, under the assumption that any survivor would most likely be a crew member and would still be in one of those three areas. A passenger who survived would have gone on to his final destination while a crew member would simply have signed on to another ship and would probably have been listed in some census in the Pacific Northwest. A search was made of the newspapers and every name identified with the Clara Nevada was listed.

  However, using the 1900 census proved to be more difficult than spending an hour or so with an index of names. The census returns from Oregon and Washington, the most likely home ports for many of the mariners, were not indexed. Finding an individual meant I had to go through the returns of each county in each state and comparing each name to the list of forty-five that I had. This tedious task was made even worse by the fact that for some individuals I had no first names, and for others, different newspapers had spelled the last names differently. Further complicating the search, some of the individuals listed had first names that were probably nicknames, like “Bat” and “Peanuts.”

  Then there was the problem of reading the handwriting of the census takers. This was made infinitely more difficult by the fact that I was reading microfilm rather than the original documents. All of the data had been entered by hand, and each section of the census offered me the chance to interpret a new bit of penmanship. Page by page, the handwriting could vary. A poorly written “o” could be read as an “a” or an “e.” Sometimes the writing was so light, it could not be read at all. Other times it was so dark that the ink had come through the paper and obscured the name on the next page.

  Eventually I abandoned the search of the Oregon and Washington census data and concentrated on Alaska. Here I had an advantage: an index. But it was not a conventional index. Census data in those days was sequenced on the basis of sound and recorded in a “Soundex.” Names were coded on the basis of how they were pronounced, not how they were spelled. This coding system, unfortunately, was difficult to interpret. As an example, take the name Hildebrand. A researcher would go to the “H” section of the Soundex because the word started with an “H.” Since vowels had no assigned number, the “i” would be ignored. The next letter, “l,” had an assigned number “4.” The next consonant, “d,” had the assigned number “3.” The next letter, “e,” would be skipped because it was a vowel and the final letter in the coding, “b,” had a designation of “1.” Thus, for the purposes of the 1900 census, “Hildebrand” would be found in an index as “H431.”

  Once the names of all passengers and crew had been assigned a Soundex number a search of the Alaskan cen
sus data was conducted. The problem now was to identify the exact name with one that had been listed in the newspaper. A close match would not do; it had to be an exact match.

  When I finished, I had six possible matches for the crew in Alaska. Some, however, were quite speculative. Was Bat Hurley, the quartermaster on the Clara Nevada, for instance, “William Hurley,” a waiter on the City of Topeka? There were, after all, two men on the Clara Nevada that were listed as quartermaster. One of those listings could have been a mistake.

  Four “George Rogers” showed up on the census, one of them a seaman on the steamer Alpha of Victoria B.C. and the second as a waiter on the steamer Tess. The two others were a miner and a passenger. Making a match of George Rogers might be a moot point; however, some sources state that Rogers may not have been onboard when the Clara Nevada left Skagway.

  Then there was Frank O’Donnell, the steward on the Clara Nevada. In this case, there was a close match. There was only one whose occupation was listed as seaman. He was on the steam whaler Jeanette of San Francisco. But there is no way to definitively state that “O’Donnell” with two “ls” is rare enough to assert that these two men were the same individual.

  While “O’Donnell” might be unusual, other common names were not. In the case of Ed Kelly, pilot, and (no known first name) Kelly, waiter, of the Clara Nevada, there was no trouble finding similarly named individuals. Four references to “Charles Kelly,” two for “Edward Kelly” and one for an unnamed Kelly exist in the record. Three of them lived in Skagway, the last port of call of the doomed steamer. In my opinion, none of these could be called a close match.