preserve the continuity of its historic associations, so far as might be consistent with the public interest, and with what was due to progress in knowledge, was one of the first duties of those responsible for a new edition; and just as the Ninth Edition carried forward, with notable additions or substitutions, work contributed to the Eighth and earlier editions, so it provided matter for utilization in the Eleventh, which in its turn had to accommodate the new knowledge of a later generation.
In considering the treatment, however, of the mass of material thus handed down, the editor of the Eleventh Edition had an entirely new situation to deal with. It is necessary here to explain why it is A new departure.that the Eleventh Edition is much more than a revision--is, indeed, a new edifice as compared with the structure of the Ninth Edition. In the whole architecture of the latter there was a serious flaw, due to no want of ability in editors or contributors, but to the conditions imposed upon them in the system of publication.
The economic and mechanical obstacles to the production of a great encyclopædia otherwise than in a series of volumes separately issued at intervals during a number of years were formerly considered The old system of production.prohibitive. Thus the Ninth Edition, the first volume of which was published in 1875 and the twenty-fifth in 1889, was incomplete for some sixteen years after its real inception. Not only does such a long interval between the start and the finish involve the possibility of a change in editorial direction and conception such as happened in 1881 when Spencer Baynes was compelled by ill-health to hand over the reins to Robertson Smith; but even if the same editorial policy remained to dominate the work, the continual progress of time was constantly changing the conditions under which it was exercised. With such a system of publication an encyclopædia can have no proper unity of conception or uniformity of treatment. It cannot be planned from the beginning so as to present at its completion a satisfactory synoptic view of any department of knowledge. The historical record is restricted by the accident of the dates at which the separate volumes are published, in such a way that the facts induded in one volume may contradict those in another. Individual volumes, the contents of which are arbitrarily determined by the alphabetical order of headings, may indeed be abreast of the learning and accomplishments of their day, but Defect of division under different dates.each time a later volume appears the circumstances have altered, and there is every chance that some integral portion of what had previously been published may be stultified. Those who were responsible for the execution of the Ninth Edition of the Encyclopædia Britannica did their best under an impossible system. They made it a collection of detached monographs of the highest authority and value. In their day the demand of a modern public for "up-to-date-ness" had not come into existence, and it seemed perfectly reasonable in 1879 to bring the article on the history of England no further than the accession of Queen Victoria. But it was not their failure to appreciate the importance of dealing with the latest events in history that made so much of the Ninth Edition useless in preparing its successor. When only this was in question, later history could be added. It was the fact that, owing to its system of publication, its arrangement was not encyclopædic, and that in preparing an edition which for the first time had the advantage of being systematic in the distribution of its material, there was no way of adapting to its needs what had been written originally on a faulty principle.
Until the year 1902, when, within nine months, nine supplementary volumes of text were issued by The Times, no publisher had cared or dared to attempt to produce at one time the whole of any Novelty of the method now employed.work of similar magnitude. It was the regular practice to issue volume by volume. On this system the public has been furnished with the Oxford New English Dictionary (still incomplete in 1910, though work had begun in the early 'sixties and the first volume appeared in 1888) and with the Dictionary of National Biography, while the French La Grande Encyclopédie, which took even longer than the Ninth Edition of the Encyclopædia Britannica to complete, was coming out in its thirty-one volumes between 1885 and 1902. But the proof obtained in 1902 of the practicability of simultaneous production in the case of the supplementary volumes which