the administration of social and political affairs. The Branch Associations should make strenuous efforts to interest public men in the library, and to induce them wherever possible to become members of the branch, if not members of the Library Association also. At the present moment, as I pointed out at Perth, the number of influential public men in the Library Association is decreasing, to the weakening of the work and of the influence of the Association in many directions. That we do not get these men as we ought to do is a reflection upon us. The men are there, they can be got. Look at what these men have done for education. And so, if we can show them that the library movement is a live one, that there is something to be done with it, they will work for us, they will interest themselves in us, they will take us up, as they have taken education up. Only we shall have to show them that we know what we want; that we all want, speaking broadly, the same thing; and that we are capable of some self-sacrifice in order to attain it. I don't want to be ethical, but I am profoundly convinced of the fact that just as in the end individuals get pretty well what they deserve to get, no more and no less, so movements, and bodies of men, get what they deserve. If we as librarians deserve better treatment in the future than we have received in the past, we shall get it. Everything depends on us; we need never doubt our cause.

As most of you are aware, I have recently addressed a meeting of the Northern Counties Branch of the Library Association on the question of the formation of a separate branch in the West Riding of Yorkshire. That branch if formed, as I hope it will be, will give us three efficient and satisfactory branches in the North. It will then be for the librarians in the Midlands and in the South to follow the lead of the North and organize branches in these districts. I propose to supply your Secretary with printed copies of my remarks at Harrogate, and to ask him to send a copy to each one of you. In that address you will find more with regard to the organization of branches than I have time to repeat

now.

There is a further point upon which I am not in a posi

tion to enlarge now, but which I desire to submit for your consideration. I suggested in my address at Perth that in view of the work before us, and of the need for making the work which we are asking non-professional people to share more interesting to them, that the Association should seek combination with other bodies having kindred aims, in order that we may conserve and unite our resources. The difficulties in securing the amalgamation of separate bodies, each with its own constitution, and its own tradition, is very great, but I should not suggest amalgamation but rather federation. The societies which were in my mind were in particular, The Museums Association, The Bibliographical Society, and your own, The Library Assistants' Association. It should not be difficult to draw up a scheme of federation under which each of these societies would preserve its constitution intact and its autonomy, while it would pool its rooms, its office organization, and perhaps its periodical. A certain number of meetings of the whole federation would be held each year, which would give an opportunity for the members of these kindred Associations to meet each other, and would tend inevitably in the future towards a real amalgamation; but that could be left to come about naturally, and as it were by the efflux of time itself. A federation on some such lines as these would be of enormous value in focussing effort. It may be that some formal proposals of this kind will come in due course before you, in the meantime you can turn over the idea in your minds and look at it from every point of view.

I said awhile back that I was not pessimistic as to the future of the library movement and of the library profession. So far from being pessimistic, I am definitely, decidedly, determinedly, optimistic. I believe, nay I am certain, that if we put our shoulders together, if we drop for the next few years anyhow all our little domestic squabbles and suspicions one of another, if we make up our minds that we will each fill as well as we know how our own niche in a great hierarchy of effort, that we shall within the next few years render the position of the public library and of the public librarian secure during the next half-century.

PRINCIPLES OF BOOK CLASSIFICATION.

BY E. WYNDHAM HULME.

Chapter IV. On the Definition of Class Headings-
Practical Tests.

WE

WE are now in possession of two distinct tests for estimating the worth of any series of headings in a given classification. These tests are: (1) Does the series provide exact classification for every subject, or combination of subjects, represented by a specific literature in book form? (2) Can such registration be effected at the cost of single entry for each work registered or a fraction over ?

To bring home the practical importance of these tests, we must assume that we are dealing with literature on a large scale, for the benefit of classification has a direct relation to the size of the collection classified. Thus let it be granted that fifty works have been published on (A) Incandescent Lighting with Gas and Oil, and 500 (B) Gas Manufacture, Distribution and Supply-the relative strength of the two classes being one to ten. If no separate provision has been made for (A), information on (A) will have to be sought in (B), and on every occasion that this is required ten entries have to be examined before a single work of the required nature is found. But if only one work on (A) has been published the proportion is one to five hundred, and in such cases the probability is that the search will be abandoned before the entry is discovered or that the entry will be overlooked. Hence, provided the subject of a work is sufficiently distinctive, separate provision must be made in the HeadingsList immediately upon the appearance of the first monograph. According to Mr. Dewey new topics, when they arise, are

always closely related to some existing headings; but probably the reverse of this statement is nearer to the truth. For where discovery has opened a new avenue of thought or action, the literature of such subject cannot be accurately classed under any existing heading. Thus to class Artificial (nitro-cellulose) Silk with Natural Silk, Radiography with Photography, or Metallography with Articles made of Metals, is mere dumping.

The case, however, is entirely different when the problem is one of the mode of defining a series of classes combining common subject in varying proportions. Here the warrant for definition is purely quantitative. The appeal is not to reason, but to facts revealed by the literary census. Thus assuming that investigation has shown a strong literary warrant for the following series :

A Chemical analysis,

AI Commercial or technical analysis,

A 2 Food and Drug analysis,

we should unhesitatingly reject the alternative series :

B Chemical analysis,

BI Analysis of Drugs and Medicines,

B 2 Analysis of Foods and Drinks,

on the ground that the admittedly extensive literature of A1 and A2 cannot be exactly or economically registered under B-B 2.

The principles of definition and of exact shelf classification are therefore identical.

Separate provision must be made for every distinct book class immediately upon the appearance of the first monograph; while in the grading of classes, definitions must be based upon a purely quantitative literary warrant.

We now submit an illustration on a larger scale of headings prepared on these principles.