and practical at the time of its introduction, but it grew more and more cumbrous year by year, and at last broke down under its own weight. There are features in the Museum catalogue of the highest value to men of letters, such, for example, as the grouping of academical transactions and periodical literature under the headings "Academies" and "Periodical Publications," but which it would be inexpedient to reproduce in the catalogues of libraries not largely resorted to by students and scholars, or where such books did not constitute a considerable proportion of the collection. It is therefore difficult to lay down many precepts universally applicable for library management which are not sufficiently discoverable by the light of nature. Beyond this, however, there is a wide debatable region which may be explored with advantage, it being always borne in mind that the question is not so much that of the abstract fitness of principles, as of their applicability to the needs of individual libraries. Some few principles may be taken as axiomatic. It may be regarded as established, for instance, that the alphabetical arrangement of entries in a catalogue is the best that can be adopted; and no less so that the alphabetical catalogue gains enormously in value by the addition of a good index of subjects. If, however, advancing a step further, we inquire into the best form of subject-index, we find ourselves involved in controversy, which may continue long, inasmuch
