Of course different countries will require different divisions, e. g., Ecclesiastical history, Mythology, Religion, Theology will not often be required for the same {97} country. And often it will be expedient to combine those divisions in which there are very few titles into one more general; thus Botany, Herpetology, Ichthyology, Zoölogy, would join to give Natural history a respectable size, and Geology, Mineralogy, Palæontology, Physical geography would combine, or in very small countries all these would go together under Description. Under some countries other divisions will be required; in the list are given only those in actual use; but the arrangement is elastic and admits of new divisions whenever they are needed. In regard to a few (such as Epitaphs, Fables, Names, Proverbs) there is room for doubt whether they ought to be under countries; whether the subject cohesion is not much stronger than the national cohesion. Many others are not usually put here (as Numismatics, Philosophy, Religion, Science, Theology, Zoölogy). The former usage was to put under the country only its history, travels in it, and the general descriptive works; and books that treated of the Art, Architecture, Ballads, Botany, Drama, etc., of that land were put with the general works on Art, Architecture, etc. But the tendency of the dictionary catalogue is towards national classification; that is, in separating what relates to the parts of a subject, as is required by its specific principle, it necessarily brings together all that relates to a country in every aspect, as it would what relates to any other individual.

It may be asked (1) why the parts of Natural history are here separated and the parts of Language and Literature not; and (2) why we do not divide still more (following out the dictionary plan fully), so as to have divisions like Liliaceæ, Cows, Horses. As to (2), in a library catalogue of a million volumes it would no doubt be best to adopt rigidly this specific mode of entry for the larger countries; for a catalogue of one or two hundred thousand, arrangement in classes is as well suited to quick reference and avoids the loss of room occasioned by numerous headings. With few books minute division has a very incomplete appearance, specialties occurring only here and there, and most of the titles being those of general works. This may be compared to the division of a library into alcoves. One of from 10,000 to 20,000 volumes has an alcove for Natural History; from 20,000 to 50,000 it has alcoves for Botany and for Zoölogy; from 50,000 to 100,000 it has alcoves for Birds, Fishes, Insects, Mammals, Reptiles, but it must be either very large or very special before it allows to smaller divisions of Zoölogy separate apartments. On an expansive system it is easy to make new alcoves as they are wanted; a similar multiplication by fission is possible in the successively enlarging editions of a printed catalogue. A card catalogue, designed for continuous growth, should have more thorough division than can be put into print, because it must look into the future, while the printed catalogue has no future.

As to (1) I can only say that the divisions of Language seem to me too intimately connected to be dispersed in catalogues of the present size, but that those of Literature have a more substantive existence and ought to be separated sooner. A double subdivision, however, ought to be avoided. Under Language there should be only one alphabet. It is better to arrange

than

Any subdivision of the groups under countries has been strongly opposed as being troublesome to make, useless, and even confusing, or as being an unlawful mixture of classed and dictionary cataloguing. But suppose you have four or five hundred {98} titles under France. History. Will you break them up into groups with such headings as House of Bourbon, Revolution, Empire, Restoration, etc., with references and other devices for those works which treat of several periods, all of which it must be confessed is a little formidable at first glance, or will you leave them in one undivided mass, so that he who wants to find the history of the last half of the 15th century must read through the 500 titles, perhaps, to find even one and certainly to find all? You would divide of course. It is true that grouping may mislead. The inquirer must still be careful to look in several places. The history of France during the ascendency of the House of Valois is to be found not merely under that heading but in the comprehensive histories of the country. The inquirer may be a little less likely to think of this because the titles of these two groups are separated from the many other titles which have nothing to do specially or generally with the House of Valois, but if he does think of it he is greatly assisted by such segregation.

K. Etc.

259. In a supplement, catalogue the whole of a continued set, not merely the volumes received since the first catalogue.