But there could be no question that he did attain very interesting and artistic effects, and there happened to be at the time a reaction against what was called "slovenliness" and a demand for careful preparation and planned effect in prose-writing. Even so, however, it was not at once that Stevenson took to fiction. He began with essays, literary and miscellaneous, and with personal accounts of travel: and certain critical friends of his strongly urged him to continue in this way. During the years 1878 and 1879, in a short-lived periodical called _London_, which came to be edited by his friend the late Mr. Henley and had a very small staff, he issued certain _New Arabian Nights_ which caught the attention of one or two of his fellow-contributors very strongly, and made them certain that a new power in fiction-writing had arisen. It did not, however, at first much attract the public: and it was the kind of thing which never attracts publishers until the public forces their hands. For a time he had to wait, and to take what opportunity he could get of periodical publication, "boy's book"-writing, and the like. In fact _Treasure Island_ (1883), with which he at last made his mark, is to this day classed as a boy's book by some people who are miserable if they cannot classify. It certainly deals with pirates, and pieces of eight, and adventures by land and sea; but the manner of dealing--the style and narrative and the delineation of the chief character, the engaging villain John Silver--is about as little puerile as anything that can be imagined. From that time Stevenson's reputation was assured. Ill health, a somewhat restless disposition, and an early death prevented him from accomplishing any great bulk of work: and the merit of what he did varied. Latterly he took to a teasing process of collaboration, which his sincerest admirers could have willingly spared. But his last completed book, _Catriona_ (1893), seemed to some judges of at least considerable experience the best thing he had yet done, especially in one all-important respect--that he here conquered either an unwillingness to attempt or an inability to achieve the portraiture of feminine character, which his books had previously displayed. The general opinion, too, was that the unfinished _Weir of Hermiston_ (1897), which he left a fragment at his death, was the best and strongest thing he had done, while it showed in particular a distinct relinquishment, for something freer and more spontaneous, of the effective but also rather affected and decidedly laboured style in which he had hitherto written. For us, however, his style is of less importance than the fact that he applied it almost wholly to the carrying out of that rejuvenescence of romance of which we have been speaking, and which may be taken, as anybody pleases, either for a mere alternative to the domestic novel or as a definite revolt against it. It was speedily taken up by writers mostly still living, and