There is, perhaps, no other manuscript piece which includes such a variety of effects being striven for by the author, but there are many poems for which the variants show, one by one, the same processes of composition. Thus, in the "Poème de la femme," there are nine variants of the suppressed stanza, of which eight employ the rhymes tordue, mordue, Clésinger, singer (67). Of La Mansarde " there are five variants, four of which present the line while the fifth gives "Par Montmartre et ses trois moulins," and the following isolated lines are noted in the margin of the folio: "Montmartre et ses moulins à vent" "Montmartre avec ses trois moulins." The same piece gives three versions of a line with the rhymeword silhouette (68). Within the poems also there is at times verbal fluctuation which is apparently the result of an attempt to find the poetic, in the adaption of words and ideas to a definite verse-ending. In "Colère," a signed, autograph copy differs from the printed version: Diverse alternates are proposed for several of the lines of a vari ant of "Sainte Casilda": "En errant (dans) à Burgos, (par) dans l'église déserte Mon regard fut frappé d'un tableau saisissant A la place des seins deux ronds couleur de sang l'on voit chaque fibre et "D'où {chappe un rubis par } } chaque veine ouverte La sainte, l'œil nageant, et la joue enflammée, Sourit comme une amante, Semble, aux bras de l'amant, une vierge pâmée” (70). The same process of composition is employed in “ Josué arrêtant le soleil," and here a further indication of Gautier's method is given by the notation of phrases, of ideas to be incorporated, etc., on the back of the folio: "la bataille les grands chars ravageaient la melée et les lourds / éléphants / aux trompes / déroulées" (71). It would seem, then, from the evidence of the manuscripts, that poetical composition was not without its difficulties for Gautier. He lacked an instinctive feeling for rhythm; his hesitations in metre, in the use of strange rhyme-schemes, in the adaptation of words to a desired musical effect, are notable. In certain cases, where the rhyme-words were found, he was limited by them, and fluctuated in his choice of the remainder of the line largely through preoccupation with sonorities. It might be said, indeed, that his process of composition often partook of the nature of his amusement: the bouts-rimés, with the additional task of transforming his verbal facility into poetry. " Je lèche et me pourlèche Marine et le Monde est méchant, qui ne vont pas encore à ma fantaisie" (72). Bergerat and Feydeau, and later critics after them, have emphasized the point that their master's manuscripts are a monument to his facility in poetry as well as in prose; that they are specimens of a most beautiful handwriting, unmarred by corrections. It is true that some such autograph copies, frequently signed by Gautier, exist among his manuscripts, but they do not constitute their whole body. There are, on the contrary, numerous folios covered with variants-some for the same pieces which also exist in pristine autographs, others which offer no inter mediate step between the early brouillons and the printed version-where the piece is not given in its entirety, where the partial manuscript is much corrected and reworked, where the original words of the author are at times wholly effaced by the numerous superpositions which are necessary before an acceptable form is arrived at (73). It is interesting, in this connection, to compare the appearance of the manuscripts of Gautier's prose compositions with those of his poetry. The notable facility which he had in writing prose may be expected to manifest itself in the original versions, and this is the case: here the manuscripts—the brouillons, for autograph copies of a corrected piece are lacking except for certain fragments of the earliest short-stories (later Gautier's prose went direct from the writing-table to the printer, to fill his weekly feuilletons)-present a very different aspect from the workingcopies of his poems. Certain manuscripts of the Voyages may be taken as examples. Of the Voyage en Espagne, which was written in the actual course of Gautier's travels, various chapters exist in autograph. The corrections are very few in number when compared to those of the poems written at the same period. They may be divided into three classes, (1) changes which show verbal preoccupation: grammatical corrections or improvements; corrections of phrases from a conversational to a more literary tone (". . vraiment un très bel effet . ." becomes "... un effet riche et majestueux ..", etc.); corrections which work for more exactitude and clarity of expression; (2) changes in accentuation, in the euphonic composition of the sentence or paragraph; and (3) fluctuations in expression when the idea to be expressed is an abstract one of which the author does not, himself, seem very sure (cf. p. 210, where the reflections on progress, on the use of clothing, etc., seem to entail a hesitation in expression which is linked to the philosophical Aufgabe). The manuscript of Ce qu'on peut voir en six jours is similar, although here the changes in vocabulary are fewer and of less importance than the changes in order of words and the corrections for euphony (74). In the Vacances du lundi there are still fewer verse. changes, and they all work in the same direction as those of earlier travels; in general they appear to have been almost instantaneous (to judge from the sequence of the sentences, from the grammatical evidence of agreement of adjectives, etc.) One paragraph of this manuscript, however, has undergone a great amount of correction. It is found at the beginning of the chapter on Mont Cervin (75), and deals with the construction of Latin Technical description of poetic form seems to offer, to Gautier, difficulties comparable to those found in the composition of corresponding verses. The manuscripts extant of the author's short stories and novels are less great in total volume, but their testimony also confirms the processes of composition noted from the Voyages, so that, in Spirite (76), Gautier has attained to a sureness of touch equivalent to that of the Vacances du lundi. The Critique dramatique (77), of which examples extend from 1837 to 1872, demonstrates the same verbal facility: here there are very few corrections, above all in the later years, and such as are found seem of little importance and consist of a very small number of verbal improvements, changes in order, etc. There is nothing which is not present more abundantly in other literary composition-in the Voyages, for example, where the author's preoccupation with style was greater (as for an original work) and where the actual writing was, in consequence, more carefully done. In the Critique dramatique, indeed, as in the yearly Salons, Gautier was working on material which was practically identical from year to year, where the end was simple: a narration or a description, with little necessity for comment or criticism, or for original developments. Here habit could strengthen enormously the author's initial facility, and the result of his skill and ease is visible in the manuscripts of these critical studies which, in their first draught, show almost absolute conformity to the printed version. An inquiry into Gautier's process of literary composition is not complete without some indication of the relation between his prose and verse compositions. In his notice on "Marine (Jettature), fragment d'un poème inédit," Lovenjoul writes: "C'était du reste l'habitude de Théophile Gautier, s'il faut en croire MM. de Goncourt et leur intéressant journal de commencer en vers les ouvrages qu'il finissait par écrire en prose. Pour notre part, nous connaissons deux exemples de ce fait: celui-ci, et ses vers sur la Péri. Ce fragment seul fut écrit, et l'auteur renonça ensuite à son projet. C'est une des pièces auxquelles il a le plus travaillé, et nous tenons à citer ici comme preuve de notre affirmation, toutes les variantes inédites que nous en avons recueillies . . ." (78). Gautier, according to the report of the Goncourts, had his first inspiration to write in verse, and in the end he wrote in prose; according to the variants cited by Lovenjoul for the pieces of which the versified originals remain, the labour involved in this beginning of composition was great. The facility of his prose seems here to have had a decisive influence on the type of composition in which the author ended. In some cases, however, poems were written on subjects which had already appeared in the prose writings of Gautier; there seems to have been no distinct line between the subjects which the author considered fitting for prose reporting or narration, and those which he might submit to poetical treatment. Within the two versions, nevertheless, there is possible a difference in method of development, and some of the poems are more than versifications of what Gautier had already written in prose. It is true that the prose version of "Rondalla," which appears in the dramatic feuilleton of April 19, 1843, is very similar to the stanzas which were incorporated later in Militona, but with "Magdalena" and "Le Thermodon " the process of composition is different. In the first piece, the prose version is a representation of the Magdalen of Ruben's Descent from the Cross, with a detailed description of the picture and the notation of the critic's very strong emotional reaction before it. The poem, which was published two years later, gives at the beginning some indication of the pictorial source, but is principally occupied with the development of the theme of Christ and the Magdalen, entering into quite another field of Gautier's reflections from that expressed in the account of the picture and in the Toison d'or inspired by it (79). The development of "Le Thermodon," which remains closer to the prose original, is more easily compared to it, and at the same time is equally demonstra |