The Man from the Train

by Bill James

  The Christmas family was well liked—and white, besides—so an intense investigation began immediately. Jeremy and Martha had had ten children, six of whom had passed on before the attack, three of whom were left after the attack. Bloodhounds were brought up from Marianna, Florida, but it had rained hard before the hounds reached the scene, and they were unable to find a trail. A reward fund of $2,500 was raised by donations. A private detective named W. C. Franklin rushed down from Chicago to Cottonwood.

  This particular private detective was a real lulu.

  Dothan, Ala., March 18.—Following the arrest yesterday of Will Christmas, a son, and Walter Holland, a son-in-law, in connection with the triple murder of the Christmas family, new and sensational developments were brought to light today through the medium of a detective, his assistant, a ventriloquist, a superstitious negro, and the negro’s mule.

  The detective, who has been at work on the case, declared the whole mystery is solved.

  Hokay; well, I don’t know why anyone would have any trouble with that story, but let me tell you the whole story, and you can be the judge. As I have said before, private detectives in the nineteenth century were essentially spies; the basic job description of a nineteenth-century private detective was to infiltrate criminal communities and sell out his confidants. Since there was a reward fund for the solution of the Christmas family murders, Detective Franklin decided to see what he could do. To infiltrate the criminal population of Cottonwood, he hid out in the woods near the small town, pretending to be an escaped murderer from Georgia. Since the escaped criminal couldn’t go into town to get his meals, he hired a black man to bring him his meals, out in the woods, and the black man had a mule. After a week or so the wily detective began to sense that the black man knew something about the case, but wouldn’t tell him what it was. So the detective snuck back north, where he knew this guy who could throw his voice. He brought the ventriloquist back with him to his hideout in the woods; not making any of this up. The ventriloquist hid behind a tree or a rock or a bush or something, and threw his voice to the mule, convincing the humble pizza delivery guy that his mule was talking to him; the black man is described in the newspapers as “superstitious,” although I don’t know that believing that your mule is questioning you about an unsolved murder is exactly a superstition. Also, he wasn’t really delivering pizzas; I just said that to help a modern reader relate to the role.

  Anyway, the black gentleman explained to his mule that the Christmas family had been murdered by Will Christmas, an adult son, acting in concert with a son-in-law. The detective and the ventriloquist overheard this explanation, so they went to the sheriff and explained to him what had happened. Since the information came from an unimpeachable source—that mule had never lied to anybody—the sheriff then arrested Will Christmas and Walter Holland, and also Mr. Holland’s wife, who was Will Christmas’s sister, and charged them with the murders.

  We assume that these three were released; we assume this because we can find no record of their being prosecuted. In 1906 police would arrest you for murder based on no evidence at all; it wasn’t like it is now. In this case, several people had been arrested in connection with the murders in the first two days after the crime, but they had all been released. Perhaps the mule refused to testify against them.

  The story of the mule and the ventriloquist was reported coast to coast at the time, with no skepticism detectable between the lines. This pathetic effort to claim the reward fund was, as far as we know, the closest that Houston County ever came to solving the murders. And we could tell you that they are a part of the series, but there is a problem with that, too, as you will discover in the next chapter.

  * * *

  If this series of crimes was a one-man epidemic, then the epicenter of the epidemic was Marianna, Florida. At that time Marianna was a town of about a thousand people. In 1903 the Kelly or Caffey family was murdered eight miles west of Marianna. Cottonwood, Alabama, is twenty-four miles north of Marianna, and what may be The Man from the Train’s worst atrocity would be committed three months later near Milton, Florida, which is a hundred miles due west of Marianna, just north of Pensacola. Jacksonville is two hundred miles east of Marianna.

  * * *

  Several crimes in this series were committed within five miles of a state line; in this case it is about three miles from the Florida state line to Cottonwood, and about fifteen miles to the Georgia state line. This may not mean anything, but Ted Bundy, as a college student, wrote a paper about jurisdictional issues in investigating related crimes. Bundy realized that if two crimes were investigated by different police agencies, the police most often would never connect the dots. The Man from the Train may have had a similar thought, in a less sophisticated fashion; that is, he may have realized that once he crossed a state line, different authorities would be involved, and they would probably not make any connection between the crimes. He probably was riding north from Marianna on the Chattahoochee and Gulf Railroad, and he may well have thought, “OK, the first little town I come to after I get into Alabama, that’s where I’m going to get off.” The first little town he came to would have been Cottonwood.

  Of the last five murders that I have told you about, at least three and possibly four were committed in what could be described as the first little town you would come to after you crossed the state line.

  A mule interrogates a witness in the Cottonwood murders. (Logansport [Indiana] Pharos-Tribune, April 7, 1906)

  CHAPTER XXIX

  Murder in the Cold

  The attempt at rape by Snelgrove in the absence of Mr. Stetka caused much talk, and the Stetkas felt the position keenly, and said they were going back to Austria where they formerly belonged.

  —St. John Daily Sun

  Sydney, Nova Scotia, is about six hundred miles due east of Shirley, Maine, where we had an earlier incident. If you are a real ’merican and not from Maine, you will probably think that that’s impossible; if you go hundreds of miles east of Maine you would be in the middle of the ocean. But actually you would not; you would be in Nova Scotia, on Cape Breton Island, which is not exactly an island; it is a finger of land detached from the mainland by a river or two and long, long stretches of empty roads. Surrounding Sydney and running underneath much of the island is a large underground seam of high-sulphur coal, the Sydney Coal Field. Eleven miles northwest of Sydney is a coal mining area, and in the mining area is (or was) a small town called Dominion. Dominion was incorporated in 1906, the year of our tragedy, and was dissolved in 1995. Prior to its incorporation it was also called Old Bridgeport.

  The Stetka family lived not in Old Bridgeport but south of it; they were one of a few families living scattered among the coal mines. There were a cluster of five or six houses right next to the Sydney and Louisburg Railroad. Anton Stetka had a wife and two children; unfortunately the newspapers did not choose to share their names with us. There was a boy, aged four, and another child, aged two, who we are guessing was a girl because the newspapers mentioned that the other one was a boy.

  The Stetkas had been in Dominion for several years and were respectable, hardworking people living in a tiny five-room one-story house. The house was sixteen feet by twenty feet. As a frame of reference, a living room area rug commonly is seven by ten feet. The Stetkas were purchasing the house from the mining company. It may be presumed that almost everything the Stetkas bought was purchased at the mining company store, which was the only store in the area. In March 1925, this area would become a flash point in a notorious union dispute; local clergy would describe “children clothed in flour sacks and dying of starvation from the infamous ‘four cent meal.’ ”

  One of the other miners who lived in this desperate settlement was named Snelgrove. In January 1906, weeks before the murders, Snelgrove attempted to rape Mrs. Stetka while her husband was away. Snelgrove was arrested and jailed in Sydney, but the incident ruined the relationships among the neighbors, some of whom felt that Mrs. Stetka was tresp
assing conventions of community trust in prosecuting Snelgrove. Snelgrove’s mother lived just sixty feet from the Stetkas. It was horribly awkward, and after the rape attempt, the Stetkas had begun planning to return to Europe.

  In fact, on the day before the murders Mrs. Stetka had met with the president of the mining company, Henry Mitchell, to ask about getting their money back out of the house. Mitchell said he would pay back their investment in the house, but he couldn’t do it right now because he had something else to take care of. Come back tomorrow.

  Lamps were seen burning in the house as late as 10:30 p.m. on February 15, perhaps because they were packing for their impending departure, and perhaps Mr. Stetka may have worked until late in the evening; the miners often worked twelve-hour shifts. He picked up his pay late in the day.

  A little after 1:00 a.m. on February 16, Engine 661 was pulling a string of empty coal cars into Dominion when the engineer, Alexander McKinnon, saw that a house was on fire. McKinnon “sounded the alarm,” presumably giving several blasts of the train whistle, and brought the train to a halt. McKinnon and his fireman rushed to the house. The Stetkas’ house was fully engulfed in flames, which could not be extinguished until the walls collapsed, around three in the morning. According to the St. John Daily Sun, February 17, 1906, “Between the places where the two beds had been placed was the remains of a sewing machine and close by it lay an axe. The back part of the man’s skull was separated from the head and the woman’s skull was fractured and also those of the children. . . . Near by lives a partially demented man. The matter thus far is merely one of conjecture. An inquest has begun.”

  Another report, not entirely consistent with the first, notes that a miner named McAulay saw the fire at 1:30 a.m., “returning from his shift at Reserve as he approached the house.” McAulay observed that the porch to the house was on fire and ran to a neighbor’s house to ask for help. The neighbor reported that she had heard a dog barking. About this time the train crew arrived, and the group of them ran to the house and pounded on the door, which was locked or jammed shut. A dog was barking ferociously inside. McKinnon covered his face with his jacket and smashed the door. As it broke open the dog rushed out, nearly knocking McKinnon flat. Due to the heat and flames, he was unable to enter the house.

  The coroner’s inquest lasted for eight days, and thirty witnesses were called. Several of the witnesses were recalled for further testimony. According to the Manitoba Free Press, February 24, “the enquiry was very careful and exhaustive and exploded the accident and suicide theories completely, but has not located the murderers.” The coroner concluded that there was no evidence to connect the deaths of the Stetkas to the upcoming prosecution of Snelgrove.

  There were three front-row explanations for the murders. First, it was suspected, as is always true in these cases, that Mr. Stetka had murdered his family and himself. But since Stetka’s head had literally been knocked off of his shoulders, this could not be true.

  The second explanation was that the crime was somehow connected to the upcoming trial. But the Stetkas had made plans to leave the area and return to Europe before the trial; Mrs. Stetka had decided not to testify. Setting that aside, Snelgrove was in jail, and it is unclear who else might have acted on his behalf. There are no good candidates.

  The third obvious explanation, to which speculation turned when the first two failed, was that the crime had been a robbery. In this connection it was mentioned that Mrs. Stetka had gone to the mine president, Mr. Mitchell, to get back the money that they had put into the house—and then noticed that she had failed to get the money. Mr. Stetka had picked up his last fortnight’s pay just the previous day, and it was noted that no coins were found in the ashes of the house, nor was Mr. Stetka’s watch. On this basis, some concluded that the crime must have been a robbery.

  “Robbery” is a theory that fills the void for motive when a crime seems otherwise without explanation. Halifax, Nova Scotia, was the nearest city large enough to have a police force. The chief of police in Halifax, Nicholas Power, came to Dominion after the coroner’s inquest. “I am here at the request of the attorney general,” Power reported; who the attorney general was and what he was attorney general of was assumed background knowledge, not mentioned in the St. John Daily Sun of February 28. We will assume that he was attorney general of Nova Scotia. Anyway, Mr. Power’s report concluded (1) that Mr. Power was a great detective who had solved other cases that seemed beyond solution, (2) that he had done a thorough investigation of the crime, and (3) that there was no murder here; it was just an accidental fire.

  This appears to have been a one-man investigation; a detective as great as Nicholas Power, chief of police in Halifax, Nova Scotia, does not require the assistance of subordinates. The investigation by Mr. Power tied a bow around the case, reassuring the citizens of Nova Scotia that they could go on about their business and quit worrying about an axe murderer being on the loose. But since this very thorough one-man investigation was done after the coroner’s inquest was reported to the crown, and since the results of this investigation were in the newspapers three days after the coroner’s inquest was completed, we might perhaps wonder how thorough the investigation really was.

  The murders of the Stetkas have eight characteristics suggesting that they may have been committed by The Man from the Train:

  1. An entire family was murdered in a single event, without warning, and obviously not by a member of the family.

  2. The murders were committed, it appears, within twenty-five yards of a railroad line.

  3. The murders were committed with an axe.

  4. The axe was left in the room.

  5. The front door was left locked or jammed shut.

  6. The house was set on fire as the murderer left.

  7. The family was murdered between midnight and 1:00 a.m., the time period most favored by this criminal.

  8. The crime was committed in a semirural settlement too small to have a police force, as many of the other crimes were.

  Simple as these eight circumstances may seem, there should be zero “random matches” to this bundle of facts in a normal year, in all of North America. We might also note

  • That it does appear from the words used that the murders were committed with the blunt side of the axe,

  • That this was a mining area, as were the locations of a few other crimes; it is likely that the criminal worked primarily as a lumberjack and as a miner, and

  • That Mr. Power reported that the children’s faces were covered with their blanket—a signature behavior of The Man from the Train.

  But at the same time, there are four conditions here that would make us wary of declaring this crime to be a part of the series:

  1. Nova Scotia in February is wretchedly cold. The Man from the Train hated cold weather. Other than his breakthrough crime in 1898, we know of no other crime in the series that was committed in cold weather.

  2. There is the possibility of another motive, either robbery or the other crime story in which Mrs. Stetka was a victim. Of course, side stories like that crop up in almost every case and alternative explanations are proffered in every case, but this side story is more credible than is, say, the pig-faced man in Paola or the alleged business rivalry with Frank Jones in Villisca.

  3. The presence of the dog at the murder scene could be taken to indicate the possibility that the dog was familiar with the murderer. The dog was reported to have been barking loudly, but it is unclear when. It appears that the dog started barking after the murders had been committed and the house set on fire.

  4. The first three contra-indicators are weak and wouldn’t cause us to question whether the crime should be considered a part of the series. The only real problem is this: this crime occurred just eight days after the murders of the Christmas family in Cottonwood, Alabama, and just over two thousand miles away.

  If we ask the question “Would it have been possible for our culprit to have traveled that distance in that amoun
t of time, in 1906, by hopping trains?” the answer is yes, it was clearly possible. It is possible that he might have made that trip in as few as four days, conceivably three, certainly in six. It is also entirely possible that The Man from the Train committed the murders in Nova Scotia but not the ones in Cottonwood.

  But one of the things that normally identifies crimes that are a part of our sequence is that they have simple and obvious time-and-distance connections. If one crime is committed in Florida, the next tends to be in Georgia or Alabama; if one is committed in Colorado, the next is committed in Kansas, and most often after an interval of a month to six months.

  We find occasional spatial gaps, such as the 1911 jump from Texas to Oregon, of course, and there are other cases in which one crime follows quickly after another. Obviously we should not expect a criminal’s patterns to be rigid. My point is, though, that if we argue in other cases that the sequence of crimes is knit together by time and place creating a logical line of march, we obviously have to note that this is very much not true with regard to these two crimes, or to these three crimes, if we include the next one, which will take place back in northern Florida.

  It is possible that The Man from the Train may have worked in the mines in this area in warm weather at some earlier point; he may even have known the Stetka family and may have nurtured a dislike of them for some reason, and he may have returned to this area as he was fleeing northward after committing the crime in Alabama the previous week, deciding to commit a quick atrocity in the far north, and then go back south.

  Also we have to clear this up: the word lumberjack was not used a hundred years ago the way it is now. People who chopped down trees for lumber at that time were generically called woodcutters, and the word lumberjack was used only to indicate certain tightly knit communities of woodcutters who were ethnically distinct; in other words, it applied more to the community than to the job. Most woodcutters in that era were not called lumberjacks, but that is the term we use now, so that is the term I have generally used and will continue to use.