to belong to school-examinations and text-books, and to reduce the matter to the level of the famous examination-paper which, in reference to Wordsworth's "O cuckoo, shall I call thee bird, or but a wandering voice?" instructed the student to "state alternative preferred, with reasons for your choice." In a sense, classification is always arbitrary and belittling; yet to the novelist's mind such distinctions represent organic realities. It does not much matter under what heading a school-girl is taught to class "Vanity Fair"; but from the creator's point of view classification is the choice of a manner and an angle of vision, and it mattered greatly that Thackeray knew just how he meant to envisage his subject, which might have been dealt with merely as the tale of an adventuress, or merely as the romance of an honest couple, or merely as an historical novel, and is all of these, and how much more besides-is, indeed, all that its title promises.

The very fact that so many subjects contain the elements of two or three different types of novel makes it one of the novelist's first cares to decide which method he means to use. Balzac, for instance, gives us in "Le Père Goriot" and in "Eugénie Grandet" two different ways of dealing with subjects that contain, after all, much the same elements; in the one, englobing his tragic father in a vast social panorama, in the other projecting his miser (who should have given the tale its name) in huge Molièresque relief against the narrow background of a sleepy provincial town and three or four carefullysubordinated characters.

There is another kind of hybrid novel, but in which the manner rather than the matter may be so characterized; the novel written almost entirely in dialogue, after the style, say, of "Gyp's" successful tales. It is open to discussion whether any particular class of subjects calls for this treatment. Henry James thought so, and the oddly-contrived "Awkward Age" was a convinced attempt on his part to write "a little thing in the manner of Gyp"-a resemblance which few readers would have perceived had he not pointed it out. Strangely enough, he was persuaded that certain subjects not falling

into the stage-categories require nevertheless to be chattered rather than narrated; and, more strangely still, that "The Awkward Age," that delicate and subtle case, all half-lights and shades, all innuendoes, gradations and transitions, was typically made for such treatment.

His hyper-sensitiveness to any comment on his own work made it difficult to discuss the question with him; but his greatest admirers will probably feel that "The Awkward Age" lost more than it gained by being powdered into dialogue, and that, had it been treated as a novel instead of a kind of hybrid play, the obligation of "straight" narrative might have compelled him to face and elucidate the central problem instead of suffering it to lose itself in a tangle of talk. At any rate, such an instance will probably not do much to convince either novelists or their readers of the advantage of the "talked" novel. As a matter of fact, the mode of presentation to the reader, that central difficulty of the whole affair, must always be determined by the nature of the subject; and the subject which instantly calls for dialogue seems as instantly to range itself among those demanding for their full setting-forth the special artifices of the theatre.

The immense superiority of the novel for any subject in which "situation" is not paramount is just that freedom, that ease in passing from one form of presentation to another, and that possibility of explaining and elucidating by the way, which the narrative permits. Convention is the first necessity of all art; but there seems no reason for adding the shackles of another form to those imposed by one's own. Narrative, with all its suppleness and variety, its range from great orchestral effects to the frail vibration of a single string, should furnish the substance of the novel; dialogue, that precious adjunct, should never be more than an adjunct, and one to be used as skilfully and sparingly as the drop of condiment which flavours a whole dish.

The use of dialogue in fiction seems to be one of the few things about which a fairly definite rule may be laid down. It should be reserved for the culminating moments, and regarded as the spray into which the great wave of narrative breaks

in curving toward the watcher on the shore. This lifting and scattering of the wave, the coruscation of the spray, even the mere material sight of the page broken into short uneven paragraphs, all help to reinforce the contrast between such climaxes and the smooth effaced gliding of the narrative intervals; and the contrast enhances that sense of the passage of time for the producing of which the writer has to depend on his intervening narration. Thus the sparing use of dialogue not only serves to emphasize the crises of the tale but to give it as a whole a greater effect of continuous develop

ment.

Another argument against the substitution of dialogue for narrative is the wastefulness and round-aboutness of the method. The greater effect of animation, of presentness, produced by its excessive use will not help the reader through more than half the book, whatever its subject; after that he will perceive that he is to be made to pay before the end for his too facile passage through the earlier chapters. The reason is inherent in the method. When, in real life, two or more people are talking together, all that is understood between them is left out of their talk; but when the novelist uses conversation as a means not only of accentuating but of carrying on his tale, his characters have to tell each other many things that each already knows the other knows, and to avoid the resulting shock of improbability their dialogue must be so diluted with touches of realistic commonplace that, as in the least good of Trollope's tales, it rambles on for page after page before the reader, resignedly marking time, arrives, bewildered and weary, at a point to which one paragraph of narrative could have carried him.

III

IN writing of the short story I may have seemed to dwell too much on the need of considering every detail in its plan and development; yet the short story is an improvisation, the temporary shelter of a flitting fancy, compared to the foursquare and deeply-founded monument which the novel ought to be.

It is not only that the scale is different;

it is because of the reasons for its being so. If the typical short story be the foreshortening of a dramatic climax connecting two or more lives, the typical novel usually deals with the gradual unfolding of a succession of events divided by intervals of time, and in which many people, in addition to the principal characters, play more or less subordinate parts. No need now to take in sail and clear the decks; the novelist must carry as much canvas and as many passengers as his subject requires and his seamanship permits.

Still, the novel-theme is distinguished from that suited to the short story less by the number of characters than by the space required to mark the lapse of time or permit the minute analysis of successive states of feeling. The latter distinction holds good even when the states of feeling are all contained in one bosom, and crowded into a short period, as they are in "The Kreutzer Sonata." No one would think of classing "The Kreutzer Sonata," or "Ivan Ilyitch," or "Adolphe," among short stories; and such instances prove the difficulty of laying down a hard-andfast distinction between the forms. The final difference lies deeper. A novel may be all about one person, and about no more than a few hours in that person's life, and yet not be reducible to the limits of a short story without losing all significance and interest. It depends on the character of the subject chosen.

Since the novel-about-one-person has been touched on, it may be well, before going farther, to devote a short parenthesis to its autobiographical or "subjective' variety. In the study of novel-technique one might almost set aside the few masterpieces in this class, such as the "Princesse de Clèves," "Adolphe" and "Dominique," as not novels at all, any more than Musset's "Confessions d'un Enfant du Siècle" is a novel. They are, in fact, all fragments of autobiography by writers of genius; and the autobiographical gift does not seem very closely related to that of fiction. In the case of the authors mentioned, none but Madame de La Fayette ever published another novel, and her other attempts were without interest. In all the arts abundance seems to be one of the surest signs of vocation. It exists on

the lowest scale, and, in the art of fiction, belongs as much to the producer of "railway novels" as to Balzac, Thackeray or Tolstoy; yet it almost always marks the great creative artist. Whatever a man has it in him to do really well he usually keeps on doing with persistency.

There is another sign which sets apart the born novelist from the authors of selfconfessions in novel-form; that is, the absence of the objective faculty in the latter. The subjective writer lacks the power of getting far enough away from his story to view it as a whole and relate it to its setting; his minor characters remain the mere satellites of the principal personage (himself), and disappear when not lit up by their central luminary.

Such books are sometimes masterpieces; but if by the term "art of fiction" be understood the creation of imaginary characters and the invention of their imaginary experiences and there seems no more convenient definition-then the autobiographical tale is not strictly a novel, since no objectively creative effort has gone to its making.

It does not follow that born novelists never write autobiographical novels. Instances to the contrary will occur to every one, and none more obvious than that of "The Kreutzer Sonata." There is a gulf between such a book and "Adolphe." Tolstoy's tale, though almost avowedly the study of his own tortured soul, is as objective as Othello. The magic transposition has taken place; in reading the story we do not feel ourselves to be in a resuscitated real world (a sort of Tussaud Museum of wax figures with actual clothes on), but in that other world which is the image of life transposed in the brain of the artist, a world wherein the creative breath has made all things new. If one happened to begin one's acquaintance with Tolstoy by reading "The Kreutzer Sonata" one would not need to be told that it was the creation of a brain working objectively, a brain which had produced, or was likely to produce, other novels of a wholly different kind; whereas of such books as "Dominique" or "Adolphe," were one to light on them as unpreparedly, one would say: "This is not the invention of a novelist but the self-analysis of a man of genius."

There is one famous book which might be described as the link between the real novel and the autobiography in noveldisguise. This is Goethe's "Werther." Here a youth of genius, as yet unpractised in the art of fiction, has related, under the thinnest of concealments, the story of his own unhappy love. The tale is intensely subjective. The hero is never once seen from the outside, the minor figures are hardly drawn out of the limbo of the unrealized; yet how instantly the difference between "Werther" and 'Adolphe" declares itself! The latter tale is completely self-contained; it never suggests in the writer the power or the desire to project a race of imaginary characters. "Werther" does. Every page thrills with the dawning gift of creation. The lover has not been too much absorbed in his own anguish to turn its light on things external to him. The young Goethe who has noted Charlotte's way of cutting the bread-and-butter for her little brothers and sisters, and set down the bourgeois humours and the sylvan charm of the ball in the forest, is already a novelist.

IV

THE question of form-already defined as the order, in time and importance, in which the incidents of the narrative are grouped-is, for obvious reasons, harder to deal with in the novel than the short story, and most difficult in the novel of manners, with its more crowded stage, and its interweaving of individual and social analysis.

The English novelists of the early nineteenth century were still farther enslaved by the purely artificial necessity of the double plot. Two parallel series of adventures, in which two separate groups of people were concerned, sometimes with hardly a link between the two, and always without any deep organic connection, were served up in alternating sections. Throughout the novels of Dickens, George Eliot, Trollope and the majority of their contemporaries, this tedious and senseless convention persists, checking the progress of each series of events and distracting the reader's attention. The artificial trick of keeping two stories going like a juggler's ball is entirely different from the