in the relative positions assigned to them. The libraries should not stereotype shifting views in Science.
Classification in libraries appears to resemble in many points classification in the Natural Sciences. In each, the growth of knowledge has made classification necessary. In each, it was artificial for a time, the alphabetical author-catalogue resembling the artificial systems in Science. In each, the need of better methods has led to the search after a truer basis of classification, which has given us the natural systems in the Natural Sciences and the arrangement by subjects in libraries. In each, there is much to be done in reaching on to a higher goal; and in each it is unlikely that a perfect system will ever be attained. In each, evolution is the key to the line of progress. Librarians have their own special difficulties to meet; but you have one great advantage in not having to make allowance for the many missing links, as we must in Botany and in Zoology. A most serious difficulty common to both is the necessity in actual use of attempting to express in linear arrangement the very complex relationships that exist between the several groups.
Dewey's "Decimal Classification and Relative Index" is doubtless well known to you all. Its value is attested by the number of libraries in which it, or some modification of it, is used; and its use continues to spread, until there is reason to anticipate its predominance, at least in countries where English is spoken. Its simplicity and power of sub-division render it very attractive, and its excellent index is an immense advantage. But just because of these good qualities, and of the position that it has in consequence secured, it may become a danger to true progress. As with the Linnean system in Botany, its adherents may become so satisfied with it as to refuse to improve upon it, or to try to advance to any better one. Its relative excellence makes it desirable to examine in how far it supplies our needs, and to subject it to criticism from this point of view. My remarks will apply to the 4th edition, published in 1891.
A full analysis of the treatment of "Natural Science" in it, including certain departments that fall under this also, though placed for convenience under other heads, would occupy much time. I will confine myself to a far more limited task, viz., the consideration of a few points as examples, taken from the divisions with which I am most familiar.
The first fact that strikes one is the absence of gaps to provide for future extensions of knowledge. Yet experience warrants the
