Sax Rohmer, Harry Houdini & Fire-Tongue

Revised: December 23, 2001

In Sax Rohmer's biography, Master of Villainy, an entire chapter, "Houdini to the Rescue," is devoted to the tale of how Harry Houdini helped Sax Rohmer "finish the story he had already given up as impossible" (134). Much of the chapter is based on Rohmer's own first-hand account as given in one of his "Pipe Dreams" -- articles in which he reminisced about his work and his life. They were published in the Empire News, Manchester, England. (Some were later reprinted in The Rohmer Review.) The March 27, 1938 installment was titled "Were Houdini's Feats Supernatural" and in it Rohmer expresses his great admiration for Harry Houdini, explaining how they became friends and relating  numerous anecdotes about some of the experiences they shared over the years.

                                        Harry Houdini

As the title indicates, Rohmer wanted to address the question of Houdini actually using the supernatural. Rohmer notes that after Houdini's death, Sir Arthur Conan Doyle and others firmly believed that Houdini "employed supernatural powers in some of his illusions." Doyle went on to condemn Houdini for discrediting the supernatural feats claimed by others when he was secretly using the supernatural in his "act."  Defending his friend, Rohmer observed that "whilst I discredit entirely the supernatural theory held by Sir Arthur , I am by no means convinced that the means employed were not super-normal." Rohmer maintained that Houdini was simply a superior athlete who had employed unique training methods. Their disagreement actually caused a rift in the friendship of Rohmer and Doyle.

                      Sax Rohmer

Perhaps the most interesting story related by Rohmer in his Houdini "Pipe Dream" is that of Houdini's role in the writing of Fire-Tongue -- the later basis for the "Houdini to the Rescue" chapter in Master of Villainy. While at home on Bruton Street in Mayfair, Rohmer tells his readers that he began work on "a mystery story" commisioned by the American magazine, Collier's. Rohmer then offered a description of how he worked to illustrate how he approached the Collier's assignment:

     "Up to this time I had held a theory that the best way to write a mystery story of the 'Who did it?' type was for the author to be as mystified as the reader. In other words, my method was this: I would stage a murder under conditions which seemed to preclude any human agency.  Every possible outlet for the murderer would be securely sealed. It would be, in short, the perfect crime.
     "At this stage I would introduce my investigator and I would approach the problem from the point of view of this character: I mean that I would examine any clue which I myself had left and, working on from that point, would seek the solution. Obviously, if the author does not know the criminal, the reader has small chance of discovering him!"

Rohmer used this method to write the opening chapters of Fire-Tongue:    Having met Paul Harley briefly in India, Sir Charles Abingdon calls on him for help. Sir Charles is clearly afraid for his life but offers scant information. Instead, he invites Harley to dine with him and promises to explain the basis of his fear during the meal. He dies right after the soup. In the third chapter, Sir Charles' physician and friend, Dr. McMurdoch, concludes, "My certificate will be 'syncope due to unusual excitement'; and I shall stand by it," Harley's job is to prove it wasn't a heart attack, but a murder.

     "At the outset I caused a celebrated medical consultant, retired, to be murdered at his own dinner table under circumstances which pointed to death from natural causes. By every means at my command during the early stages of the story I eliminated any possibility of human agency. The result was dramatic and very puzzling. Then, having introduced my investigator, by name Paul Harley, I assumed his duties and looked about for clues pointing to a criminal.
     "I could find none! It is unlikely that I shall forget the night when the fact dawned upon me that the mystery I had created was not susceptible of solution. I had defeated myself. I had smoked several ounces of tobacco -- and had drunk more than half a bottle of Jamaica run -- when at the hour of 4 A.M., I ceremoniously tore up the typescript which lay upon my desk, and went to bed."

The next day he began work on an entirely different story but used the same protagonist, Paul Harley. Over the course of  a month, he completed the 80,000 word manuscript for Bat Wing. At this point he was confident that all was well; Collier's wanted a "mystery story" and he had a complete story in hand. Rohmer sent a cable to his agent in New York to inform him that Bat Wing was complete. The agent's reply was completely unexpected:

     "Unknown to me he had taken a copy of the opening chapters, totalling some 30,000 words, with him and had delivered them to the New York magazine. And the magazine had commenced publication! My immediate presence in New York was demanded. One installment had already appeared; a second was about to follow. Only three were in hand.
     "Under these dreadful conditions I sailed for America on a fast ship, preoccupied the whole way over with my notes, reviewing the problem from every conceivable angle, starting from the beginning and working up to the point where Paul Harley took over the case.
     "When I reached New York no solution of the mystery had presented itself."

The editors at Collier's did everything they could think of to help Rohmer write. Joseph Coll had again been asked to illustrate the story and Rohmer found Coll's original drawings decorating his hotel room. And despite Prohibition, the room had been supplied with a small bar.   A dictaphone and a typewriter were also in place. Having read and re-read the opening, Rohmer tells us the problem:

      "One discovery I had made in this maddening problem: the alteration of a single line of dialogue in Chapter Three would have given me a clue... But Chapter Three had already been published!"

It was at this point that his friend Houdini appeared. Insisting that Rohmer couldn't stay in the hotel room for days on end, Houdini forced Rohmer to leave the hotel to relax and get away from it all. He introduced him to people he knew, had parties for him, and even had him for Thanksgiving dinner, an occassion on which Rohmer tells us he wore "a tweed suit over pyjamas, for he would take no refusal." Meanwhile, Rohmer was no closer to a solution--until, he tells us, Houdini came "to the rescue."

     "I suppose in fact that I was fairly near the cracking point when Houdini offered me the solution of the mystery! Unannounced, he appeared one night when I was pacing the floor on the verge of desperation. The door opened, and he was there!
     "He carried a copy of the magazine in which that installment including chapter Three had appeared. I had not told him of the piece of dialogue which, had it never been printed, would have enabled me to save the situation; in the circumstances I had thought that to do so would be useless. But he had read every line of the story, approaching it as he would have approached a problem of escape from a locked box. Now he opened the pages and pointed to a sentence which he had underlined. He had found it for himself!
     "'The character who said that has been dropped out,' he remarked. 'Bring him back, and have Paul Harley tax him on that statement. Think of a reason why he lied -- make him change the words... and you're saved!'
     "It was true! By means of this simple device -- or it seemed simple when Houdini pointed it out -- of forcing one character to admit that he had lied, my difficulties vanished like smoke!"

As told by Rohmer in "Pipe Dreams" and retold in Master of Villainy, this is a wonderful story. But who was the character who had been "dropped out" and what was "the single line of dialogue in Chapter Three"? Chapter Two concludes with the death of Sir Charles:

"A stifled shriek sounded from the doorway, and in tottered Mrs. Howett, the old housekeeper, with other servants peering over her shoulder into that warmly lighted dining room where Sir Charles Abingdon lay huddled in his own chair--dead."

Chapter Three is comprised entirely of Paul Harley's coversion with Dr. McMurdoch, the family physician who concluded, "My certificate will be 'syncope due to unusual excitement'; and I shall stand by it." He later appears in Chapters Six, Seven and Eight and is then "dropped out." A close reading of Chapter Three reveals no line of dialogue that prevents a solution to the murder or makes the murder impossible and Dr. McMurdoch is never brought back to admit a lie.

If the line of dialogue is not there, what of  Rohmer's claim that there were no clues in the first two chapters? There are, in fact, a number of clues. In Chapter Two, Harley goes to Dr. McMurdoch's home for dinner. Upon arriving, he is told that the doctor has been called to see a patient, "Mr. Chester Wilson on the other side of the square."  In his absence, Benson, the butler, admits Harley to the dining room. While Harley is waiting, Mrs. Howett, the housekeeper, enters the dining room and, encouraged by Harley, goes on at length about the "new parlourmaid."

"Four hours off has that girl had to-day, although she was out on Wednesday. Then she has the impudence to allow someone to ring her up here at the house; and finally I discover her upsetting the table after Benson had laid it and after I had rearranged it."  

Russian playwright Anton Chekov's first rule of drama was that if an author included a gun over the mantelpiece in Act 1, he had better be sure that it was used by the end of Act 3. If Rohmer uses two pages of dialogue to tell us the parlourmaid was doing something with the things on the table, there is a strong likelihood this will be of use or significance later in the story. A second clue is revealed when, Sir Charles returns and tells Harley that no one from the Wilson household called.

"My dear sir," cried Sir Charles, and the expression in his eyes grew almost wild, "no one in Wilson's house knew anything about the matter!"

"What! It was a ruse?"

"Palpably a ruse to get me away from home."

And finally, just pages before Sir Charles death, we are again directed to take note of the parlourmaid.

"Benson attended at table, assisted by a dark-faced and very surly-looking maid, in whom Harley thought he recognized the housekeeper's bte noire."

But how was it done? In the sentence immediately preceding the onset of symptoms, we are told of the last thing Sir Charles did.

"He raised his serviette to his lips and almost immediately resumed: I was about to tell you, Mr. Harley, about my daughter's----"
      He paused and cleared his throat, then hastily pouring out a glass of water, he drank a sip or two and Paul Harley noticed that his hand was shaking nervously."

Sir Charles is dead within paragraphs, having uttered his cryptic last words: "Fire-Tongue . . . Nicole Brinn." To be sure, Rohmer masterfully throws in much misdirection, but one is hard put to believe that following this murder scene he would believe the mystery he had created "was not susceptible of solution."

Why then would Rohmer have claimed "Then, having introduced my investigator, by name Paul Harley, I assumed his duties and looked about for clues pointing to a criminal. I could find none"? The answer may be the same as why he gave various accounts of how he got his name or how he created the character of Fu Manchu. The "Pipe Dreams" were casual non-fiction, embellished to make a good story. In this case, the entire piece was about Houdini, now dead ten years, and Rohmer may simply have embellished the story. Houdini was doubtless there, for the parties, the theater trips, and Thanksgiving, but it is not likely at all that he that he found a line of dialogue that solved the problem as described.

Seven years before the "Pipe Dreams" article was published, Rohmer gave an interview to Carl Warton of the Boston Sunday Herald. In an article titled "Houdini Saved the Day for Sax Rohmer" (March 8, 1961), Warton relates a totally different version of the story.  Once again, Rohmer's technique is explained: 

     "I used to think that the proper way to write a mystery story was to create a situation as perplexing to the writer as to the reader. It was my idea that I should not know the answer to the puzzle when I started but should work my way to the solution just as the reader would.
     "One experience, however, was sufficient to convince me that the whole plot, its unfoldment and solution, must be completely pre-determined. Since then I have always prepared a scenario for all my stories."

John Harwood, who first reported this article in "Houdini To The Rescue" in The Rohmer Review #4 (March 1970), notes that "Rohmer was behind in the story because he had started it with no idea of how it was going to conclude." Rohmer had already revealed that the parlourmaid switched the serviettes while Sir Charles was off to Chester Wilson's home following the false call by the villain, Ormz Khn. In this version of the Houdini tale, Rohmer had all the plot elements but could not figure out how to end the story. Harwood relates this alternate version of events.

"The affair had reached the point where there were only four more installments to be published and Rohmer was no nearer to the solution than when he had first started writing. Suddenly Houdini jumped to his feet and cried out that he had the answer to the whole problem. When Rohmer asked  what it was, Houdini explained that if the writer used the flashback method he could include all the facts required to conclude the adventure."

This version is more credible than the "Pipe Dreams" version. The story does, indeed, conclude with a five chapter statement to the police in which Nicole Brinn relates what happened in India seven years earlier. But it is not unexpected that he would do so in one form or another. Brinn has spent the entire story maintaining his silence. As early as Chapter Four, "Introducing Mr. Nicole Brinn," he tells Harley something happened to him seven years earlier in India but he can't talk about it. He later tells him "There isn't any one I would rather confide in," but "I must ask you again to be patient. Give me time to think--to make plans."

It is simply inevitable that Brinn eventually tell someone what happened in India. Once the reader knows of the mysterious and frightening events in India, their anticipated revelation is a perfect device to keep the attention of readers of  a serial. The novel ends with Brinn finally relating what happened. That Brinn makes his final statement to the police rather than Harley is also quite logical as it allows the reader to see that he has turned himself in and justified Detective Inspecor Wessex's trust in him. This is hardly an unusual literary device, and it is not likely that Rohmer needed Houdini to come up with it.

John Harwood also noted that Rohmer had given a third version of the writing of Fire-Tongue to another reporter, H. Allen Smith:

"Although Smith had misplaced his notes of the interview, which had taken place in the Thirties, he recalled that Rohmer had said something to the effect that 'I used to start off without an inkling of who committed the crime. But no more. I'll never follow that bloody procedure againe'" (Chicago Tribune Book World, April 21, 1968, 6).

So what really happened? It's not likely we will ever know, but it is safe to assume that Rohmer dropped writing Fire-Tongue in favor of Bat Wing. There is evidence that his idea for Bat Wing was clear by virtue of his claim to have written it in a month. It is also quite likely that his American agent, oblivious to the switch to Bat Wing, did, indeed, sell the story to Collier's, and that Rohmer rushed to New York and worked feverishly to complete it. As he tells us, "The printers began to catch up on me. At one point I reached a stage where I was turning out copy page by page from my apartment to a team of messengers connecting with the printing press!"

In the midst of all this pressure, his friend, Harry Houdini who was then living in New York made every effort to get Rohmer to go out, to meet people and to relax. They very likely discussed the story, but given Rohmer's talent and previous successes, it is not likely that Houdini made any great literary contribution. This is not to say his real contribution, helping Rohmer stay sane in the midst of the daily pressure to produce copy, was not real and significant. It is clear that Rohmer was deeply appreciative. Perhaps less clear is that over time that appreciation was expressed with a little embellishment.

By 1950, Rohmer was telling the editors of Blue Book Magazine that "Houdini credited Rohmer for having developed one or two of his tricks, and Rohmer gave Houdini the credit for having presented him with a solution to two or three of his own mystery stories when he couldn't find the murderer or a rational explanation for the crime himself."

At the end of his "Pipe Dream," Sax Rohmer says of Harry Houdini:

"He was one of the greatest showmen the stage ever known. And he was my friend." 

That much we can be sure of.

Go to The Page of Fu Manchu

Copyright 2000 Lawrence Knapp. All rights reserved.

scarab.jpg (2690 bytes)