By Joyce Kilmer.
(1886-1918)

DR. WATSON has found a friend at last. Of course, there is always the great Sherlock Holmes, but Mr. Holmes's friendship was not purely disinterested. He liked to have the gentle and astonished Doctor with him as audience and errand boy. But seldom did he give him so much as a word of praise.

      But now there appears an enthusiastic defender of Dr. Watson; a man who says that the Doctor, or some one closely resembling him in simplicity and appreciativeness, is necessary in good detective fiction; and that not the great detective himself, not the actual creator of the book, but Dr. Watson himself is the ideal person to tell the story. And the man who says this is Burton Egbert Stevenson, author of "The Mystery of the Buhl Cabinet," "That Affair at Elizabeth," and many another exciting mystery tale.

      "Of course Dr. Watson is necessary," he said to me the other day. "It seems to me that the foil to the detective plays a very important part in fiction of this sort. He represents the general public; he is mystified, enlightened, surprised, as the general public is mystified, enlightened and surprised.

Burton E. Stevenson

      "And it seems to me that Dr. Watson — that the foil — should be the narrator. You see, the story should be told in the first person. If it is told in the third person it is evident that the omniscient narrator has information which he is concealing from the reader; therefore, he is not playing fair with the reader, as the writer of this sort of work should do. If the great detective himself tells the story there can be no surprise. He must reveal bis deductions and conclusions as he goes along; he will not be surprised, and the reader will not be surprised — that is, unless he holds back a part of his information, thereby not playing fair with the reader.

      "And so the only way in which the story can be told without reservations from the reader is for the foil to tell it, as in the Sherlock Holmes stories. Then the foil, being also the narrator, keeps even with the reader in his acquisition of information, and the climactic surprise comes naturally to them both."

      "Then you believe," I said, "that it is unfair for the writer of a detective story to base his climax on an idea not hitherto introduced into the story?"

      "I think that that is a fault," said Mr. Stevenson. "The writer of a detective story, or of a mystery story (for the sort of story that I have in mind need not have a detective for one of its characters) must above all things play the game with his reader. He must put all his cards on the table; he must not keep one up his sleeve and then pull it out and slap it down at the end of bis book.

      "He must not, in other words, astound his reader by an unexpected dénouement, but he must astound the reader by giving an unsuspected twist to the dénouement which he does suspect.

      "Even in Gaboriau's stories you sometimes find the fault to which I refer. In the first part of the book he will elaborately create a mystery; he will do it very skillfully and the reader will be intensely interested. Then for the unraveling of this mystery he will refer back to the grandfathers and grandmothers of his hero and heroine; he will bring up some old feud or family tragedy that dates back to the days of the Empire.

      "This certainly is not playing fair with the reader. If that feud or family tragedy was necessary to the story the reader should have known about it at the beginning of the book. All the cards should have been on the table.

      "Nevertheless Gaboriau is still the best of the writers of mystery tales. I think that the greatest moment in all mystery fiction is the clock instance in 'The Case of Orcival.'

      "You remember the circumstances. There was a room in terrible disorder, and among the wreckage on the floor was a clock which had stopped with the hands pointing to 3. Or was it to 5? At any rate, the fact that the clock had stopped at that hour seemed to indicate that the crime had been committed at that hour.

      "But the great detective was not satisfied with this conclusion. So he started the clock going — and it struck 11! In my opinion that is the very greatest moment in all mystery fiction.

      "You see," said Mr. Stevenson, "a mystery story is like a piece of mathematics. Writing mystery stories is an exact science. And the construction at the plot ls the writers most important problem.

      "The reason for Archibald Clavering Gunter's success was that he was really a genius at plotmaking. He could not write, but he could make wonderful plots. His book 'That Frenchman' has a terrific plot, but the style of it is absolutely puerile. Gunter had great plots, but he wrote in about the manner of Horatio Alger, Jr."

      I asked Mr. Stevenson what he considered to be the great test of a detective story.

      "The great test of a detective story," he answered, "is merely whether or not the current of the story is strong enough to carry the reader along and make him surprised and pleased at the end. I have no patience with the reader who tries to work out the solution of the plot ahead of the author. That is not what he is supposed to do; he should be swept along with the current of the story.

      "The trouble with many stories is that the climax is merely a surprise, a bit of guesswork. It should be logically inevitable. The greatest detective story is the story the conclusion of which is reached with absolute logic."

      There may be people so unfortunately lofty of forehead as not to read detective stories, and therefore to be ignorant of Mr. Burton E. Stevenson's successes in this field. But if they read. poetry, they must know of Mr. Stevenson's distinguished achievements as an anthologist — particularly of that fat and admirable red volume called "The Home Book of English Verse." It may be that his occupations of compiling collections of poetry and writing mystery stories suggested to Mr. Stevenson the ideas which he advanced concerning the connection of the mystery story and the poem. And it may be that his two chief interests furnish a proof pf the theory which he propounded.

      "Any real detective story," he said, "is in the beginning as much a matter of inspiration as is any poem. If Edgar Allan Poe's 'To Helen' is an inspired work, then so is 'The Murders in the Rue Morgue.'

      "Every poem, every great poem, starts with inspiration, and its development is conditioned by the limitations of language, by the use of words that rhyme. So every great detective story starts with inspiration, and its development is conditioned by logic. And, as I said before, the greatest detective story is the one of which the logic is strictest.

      "Of course this is not true of other forms of fiction. In the novel of character, for example, logic plays only a small part. But the detective story is a highly artificial thing. It really is a piece of sleight of hand.

      "Certainly there are many instances of the close connection which obtains between poetry and mystery stories. The instance that comes first to mind is that of Edgar Allan Poe. But there is Conan Doyle; he has written many really fine poems. Sir Arthur Quiller-Couch, an admirable poet and critic of poetry, wrote in 'Wandering Heath' and 'The Pavilion on the Links' two of the best of modern mystery stories. And there is Browning's 'The Ring and the Book' — a splendid murder mystery done into verse.

      "The greatest writers have not disdained to play the game of mystery story writing. There is Henry James's 'The Turn of the Screw.' There is Thackeray's 'The Notch of the Axe' — one of the best of thrillers.

      "Wilkie Collins was master of mystery — 'The Moonstone' is a classic of its sort. And Charles Dickens — well, of course, he wrote a number of detective stories, but his greatest mystery story — one of the greatest of all mystery stories, because its mystery is still unsolved — is 'Edwin Drood.'

      "I went all over the ground of that story when I was in Rochester. I thought that if I got a clearer idea of the locality I might be able to find the solution which Dickens intended.

      "But Dickens had modified the locality so greatly that my study of the Cathedral Close and its neighborhood threw no light on the mystery. I am inclined to agree with Andrew La'n g in his belief that Dickens had got tangled up in his web of complications, and that he himself didn't know how the story was going to end.

      "You see, the story was appearing as a serial before he had finished writing it. So Dickens couldn't do what most writers of detective stories find necessary of the story and change it to suit the end.

      "I think there is truth in the theory that anxiety over 'Edwin Drood' hastened Dickens's death. You know, he got up from working on that novel, walked to the dinner table, and dropped dead.

      "One of the biggest of modern detective stories," said Mr. Stevenson, "is Moffat's 'Through the Wall.' 'The Man Without a Head,' by Tyler De Saix is good, but there is too much machinery in it. The central character in it is a sculptor who makes images of the heads of the people whom he designs to kill, and then masks himself to resemble them so that he can walk away unharmed after killing them.

      "In the best detective stories there Is seldom any elaborate machinery introduced. This business of introducing a death-dealing ball at the end of a cord of having a special sort of house, a secret staircase, an ingenious hiding place, and that sort of thing is wrong. In the modern detective story there should be a modern house in a modern city — all the setting should be commonplace. In the Sherlock Holmes play Professor Moriarty's. underground office with its door opened by a lever at his desk is highly artificial.

      "The best of the Sherlock Holmes stories is 'Silver Blaze' — it is the most consistently logical of his stories. 'The Speckled Band' is interesting chiefly because the idea is so outré. 'The Red-Headed Men's League' is one of the most effective of these stories because the start is so far from the finish. But the end of that story is rather weak."

      Mr. Stevenson is at work now on the solution of a mystery. A French writer of detective stories proposed this mystery as an impossible plot, as a situation which no novelist could explain. Here it is, as Mr. Stevenson outlined it to me:

      "Imagine," he said, "a little French village, resting in the peace of a Summer afternoon. An automobile speeds in and comes to a sudden stop. A woman leans out of the window of a nearby house, tears a sheet of red paper across twice, scatters the four pieces to the winds, and goes back into the house. The automobile leaves the village. Two days later there is a wedding in the village. But as the bride approaches the altar she drops dead. In her hand she holds one of the pieces of red paper. There are three other sudden and mysterious deaths, and in every case one of the pieces of red paper is found gripped in the fingers of the corpse. The French writer said that this mystery could not be solved, but I think it can."

      Mr. Stevenson is inclined to believe that Poe's ability as a writer of detective stories has been overrated.

      "Poe only wrote three detective stories," he said, "and one of these is a failure. He tried to work out the mystery of the death of Mary Rogers — Mary Roget, he called her. But he failed. There is a legend to the effect that a British officer came to Poe and confessed that he had committed the murder in the way in which Poe described it in his story, but this legend has never been verified. Poe spent days going over the scene of the crime, but he failed to bring the story to a satisfactory conclusion.

      "And I have never been entirely convinced by the psychology of Poe's 'The Purloined Letter.' You know the point of the story is that a thing for which people are seeking is best concealed when it is put in a place where it might naturally be expected to be. But the letter that escaped notice because it was placed in the letter rack was first crumpled and torn half across. Surely this was a mistake on Poe's part; a letter thus marked would attract attention to itself."

      Mr. Stevenson does not believe that there is much. resemblance between detective work in fiction and detective work in real life.

      "Actual detection of crime," he said, "is simply a matter of investigation. If a piece of jewelry is lost, people go around to all the pawnshops looking for it. If a man has been wounded and then disappears, people go around to all the doctors and ask them questions, and to all the taxicab drivers and other people who might have seen the wounded man. It is all done on a wholesale scale, and patience is the chief quality displayed. This process is, of course, too cumbersome to use in a story."

      As every magazine editor will testify, detective stories are being written by the thousand, and very many of them fail to find their way into print. I asked Mr. Stevenson how he thought ambitious writers of this sort of story could best prepare themselves.

      "The best training for a writer of detective storms," he said, "is newspaper work. Of course, that is the best training for a writer of any sort, if he gets out of newspaper work in time. He gains facility, he learns not to take himself too seriously, he is brought into contact with all kinds of people. he goes to all kinds of places, and he gains the power of writing in any circumstances. And particularly valuable to the writer of detective stories is the opportunity which newspaper work affords of getting a view of phases of human nature which most people do not suspect.

      Here is a list of "Don'ts" for the guidance of all who desire to have people breathlessly follow the adventures of their lynx-eyed, sleuth. Mr. Stevenson is a successful writer of detective stories, as well as a shrewd critic of them, so this is eminently practical:

      "Don't invent any intricate machinery to confuse the reader's mind.

      "Don't make utterly innocent people act as if they were guilty. Innocent people do not act as if they were guilty.

      "Don't base your plot on an unfinished sentence or an incomplete exclamation. The world has little patience with stories which hinge upon the fact that the heroine refused to take advantage of the opportunity to relieve the hero's mind by telling him that she was there at 3 instead of at 2, or that it was her sister who wore the green dress, or something of the sort.

      "Don't have your hero assume the sins of the villain.

      "Don't have your detectives possess some personal information which no one else possesses. A detective must have the same data and the same sources of information as every one else. This is a fundamental requisite for good detective fiction. One admirable thing about the Sherlock Holmes stories is the way in which Holmes always gets his information from the newspapers. They get into the railway carriage, you know, and Holmes asks Watson to read him the newspaper account of the case.

      "Don't let your story have merely an impossible solution. It must have the one inevitable solution. Don't, for the sake of the effect, allow your story to end with an absurdity."