larger size better. As a rule, use the smallest size for the outside of books. For numbering shelves, alcoves, etc., use, of course, the larger sizes.

Paper and color. The maker and the Supply Department have given time and money to repeated experiments to get the best. Be satisfied with their results, and save your own time, money, and patience by not attempting odd notions in color, paper, size, style of type, etc. The standard, as kept on hand, is probably better than any you would devise after all your trouble, and it will be large extra expense to all concerned to fuss with these odd things made to order.

Making to order. If, however, you must have something different, send exact descriptions of what you want in writing.

They can be made at head-quarters much better and cheaper than you or any one else can get them made. You will, of course, understand that 50 numbers made to order will cost you more than 500 numbers taken from a lot of several millions made by machinery. Better leave color, paper, etc., to the maker's judgment. Confine your description to what is necessary, and make that plain. Figures and letters combined for the new classification which Mr. Cutter is getting out will be provided. If you are so unfortunate as to get up some other new requirement,tell just what you must have, and estimates of the cost will be sent you. Spoiling numbers. By trying hard enough, you can sometimes spoil most of your stock of numbers. I quote from a letter of the maker the following suggestions:

"Number-labels in sheets should be kept flat, well wrapped, and under some pressure. They will then be secure from the effects of very dry or very damp weather.

"Numbered labels separately, i. e., cut up, are best kept in series of little envelopes with open flaps [this reference is to the regular numbercases supplied by the co-operation committee], from which any number wanted can be readily taken; when not in use, press the little series of pockets (or envelopes) closely together, and thus exclude dryness, dust, and dampness.

"In short, Mr. D., hay-fever weather is as bad for numbers as for clergymen. When 'the amount of moisture in the air is large,' look out for polygamous tendencies among the gummed labels. They become 'sealed' to a degree that outmormons the Mormons."

Making them stick. The trouble with all numbers is their coming off. With the Van Everen numbers this is because they are not properly put

The best plan we know for putting on the numbers is to pick them up with a pair of delicate forceps, such as chemists use to pick up their minute weights. The Supply Department can send these in brass at 30c.; nickel plate, 50c. With them one can pick up numbers twice as fast as with the fingers, as the points always catch on the paper, whether lying flat or in a corner. Keep a cup of water on the table, and immerse the number quickly in it, shake off the extra moisture that will cling in a drop, apply it at once to the exact position where it is wanted, as can easily be done with the forceps. Press it firmly down with a bit of blotting-paper, which absorbs the extra moisture. Moistening numbers on a sponge or with the tongue is apt to remove some of the gum, and sometimes all of the appetite.

The covers of some books need attention to secure firm sticking, specially black roan, which has an oily, polished surface. This is best guarded against by roughening the surface where the number is to go with a knife-edge or a bit of sand-paper. The glossy surface of some cloth-bound books can be made good by removing the glossing with moisture before applying the number. Mr. Van Everen writes me on this point, commenting on the above:

"The labels stick well on paper or leather covers, and on most cloth covers. However, if the book where the tag is to stick is moistened with a weak solution of tannin and water, it removes the glossing and insures the sticking. This need be done only in the case of new books. I have never had any trouble with labels put on as you have suggested; but the tannin tea has worked so well that I mention it."

Ordering. This gives us frequent trouble. People scold because they cannot buy one or five or ten of a single number. The cost of each is so minute that it would be like selling single pins or needles instead of papers of them. Few understand that the expense of hunting out these little orders costs more than to give the whole sheet, and the maker would prefer to send the larger quantity at the same price. The difficulty will be removed by throwing away any numbers not wanted. There is, however, a better way. The numbers are put up in different forms: 100 on a sheet, all alike; 100 on a sheet, no two alike, running 1-99, 100-199, etc., up to 10,000. They are also furnished in envelopes of 50 or more numbers, cut apart, and all alike. If you wish only one or five or ten of certain numbers, buy so many sheets including those numbers. You will have then just what you want, and the extra numbers can be put into the envelopes of their own denomination-i.e., buy sheets and tear out the numbers wanted for any of which you are sure you don't want 50. You thus do your own work of sorting numbers, and cannot complain at

the price charged for it, and very likely will never notice that it is any labor at all. It cannot be made less, and the gain is so great over any other system that you will not be dissatisfied.

Compound Numbers.-People often order numbers made for them in the form of fractions, which is indeed the most common way of applying them to books. This is expensive and very troublesome, as it requires so vast an assortment. It is the Chinese method of a type to each word instead of each letter. Make your fractions by pasting on two numbers, one just above the other, and all is simple and cheap. In an emergency, any number can be made by cutting apart or putting together other numbers. This latter is a poor plan as a rule, for if one-half the number were poorly put on it might come off, and the remaining part would make quite a different number. This trouble does not apply to the fractions, because the absence of a denominator would show at once that something was missing, and the inside of the book would be consulted.

Keeping in order. Above all things, don't let your box of numbers get in confusion. Keep them sorted carefully. In tearing out special numbers from sheets, as soon as the sheet gets honeycombed with vacancies and begins to show signs of going to pieces, tear it up and distribute the numbers to their proper envelopes. The greatest danger of ready-printed number-labels is "getting mixed."

Ungummed. Mr. Cutter expresses a preference for his labels ungummed. He applies them with a mucilage of gum arabic and a little gum tragacanth. Mr. Cutter also uses fingers instead of forceps in "picking up." I found it clumsy, slow, difficult to place the label accurately where wanted on the back of the book, and the label's numbers were constantly sticking to the hand or fingers. Forceps are less necessary with ungummed numbers. MELVIL DUI.

LIBRARIES IN THE CENSUS.

DR. HENRY RANDALL WAITE, special agent, census office, Pelham Manor, Westchester Co., New York, has charge of the statistics of libraries for the present census. He is using every effort to make the information as full and reliable as possible, and has prepared his schedules of queries in consultation with leading members and officers of the A. L. A. A most valuable body of information will thus be collected and printed by the government, and librarians will reap all the benefits, at no cost except co-operation in giving the desired information.

We urge every reader of the JOURNAL to do

his utmost to make these returns complete and

accurate.

In spite of this protest to the contrary, there are some so short-sighted that they will either forget these inquiries entirely, or else fill the blanks in haste and with little regard to either completness or accuracy. Librarians who have not the facts asked for already made up where they can copy the figures, should make it their first duty to get at these facts for their own information. We can hardly overestimate the practical value of all these facts when tabulated and diagramed as they will be, and we feel safe in pledging to Dr. Waite the heartiest co-operation of every member of the A. L. A., and of every reader of the LIBRARY JOURNAL. M. D.

BLANK-BOOK INDEXES.

A LIBRARY is or ought to be a mass of alphabetical indexing. At the best much must be in mss.— supplements to catalog, to Poole's index, contents of volumes, of essays, and miscellany, names of borrowers, etc., etc.

For quick reference, cards are vastly inferior to a book, beside the danger of misplacement or loss. The leaves of a book can be turned through the hand, and the eye runs down a narrow column so rapidly that nothing can take the place of a book index. I am speaking as one who believes in and uses cards very largely. Our English friends, who don't like cards, would be so much the stronger in expressing the need of a good blank-book index to fill out for these various uses. Many libraries use the common indexes, put everything beginning with S in a helter-skelter series on the pages headed S; but this is abomination, as all who have tried it will testify. To scatter along, leaving Many blank pages or spaces, is little better. readers look at the first page, and not finding what they seek, conclude that they have seen all the entries under that letter, as nothing indicates that more are to be found on turning the leaf. Another improvement (?) is the Vowel Index, on the plan of Dr. John Todd's famous Index Rerum. It ought to be a penal offense for any librarian ever to use so forced and unnatural a plan. It classifies the entries under A, E, I, O, U, Y, i. e., under the first vowel that occurs after the initial letter. This puts Staples before Seaman, and outrages and confuses all one's instincts of alphabeting. If such a book is to be used at all, it would serve a much better purpose to put on the page Sa, all entries having the second letter a, b, c, or d; under page Se, all second letters e, f, g, or h; under Si, all having i, j, k, l, m, or n; under So, o, p, q, r, s, t; under Su, u, v, w, x, y, z. This does fairly well,

giving 4, 4, 6, 6, and 6 letters in the groups. One objection to adopting this division is that enough people have used the pernicious vowel system to be misled by appearances.

Another and better index is made by some of the stationers, and has second letters, both vowels and consonants, printed on the margin of the leaves. This enables one to classify more closely, and greatly expedites consultation. This style of index is made for the ledgers of commercial houses having a vast number of accounts, but it is more nearly than anything else what the librarian wants for making mss. book catalogs, supplements to Poole, or other alphabetical indexes. I have recently made a tour of the leading stationers in Boston, examining every style of index to be had. No librarian wants any, except of the kind last mentioned above, and of that kind his choice lies between two-Burr's and Graves's-the first recently brought out, the other some three years old in its present improved form. These cost no more than similar books without their great improvements, and the prices, running as low as $1.50, make them available for small work in poor libraries. I find nothing else that compares with either of these in excellence.

Without prejudice in favor of either, I reach the following conclusions, after an entire evening's careful comparison. The advantages common to both are: They discard the absurd vowel system, and arrange like any alphabetical catalog; they do not give mechanically so much space to each letter, but carefully adjust space to the combinations of letters that actually occur, omitting all that do not occur at all. Great pains have been taken to make these assignments of space accurate and practical. Burr's has cloth carefully pasted on the backs of the letters and parts exposed to the most wear. Graves's has the same protection on the face, the letter being printed on the cloth, which takes the direct wear of the fingers in using. I am not clear which is best, and should like testimony from experience. Each has an extension on the back leaf, which greatly aids quick consultation, and is a device that might well be added to our catalogs and Poole's index. This is a strip of very stout manilla (in fact, a light board), projecting from the edge of the last leaf about 5 mm., or 4 inch. In turning the leaves, the fingers under this can much more readily manage the leaves than when under the cover or under the leaves, without this stiff extension. The Graves's extension folds over out of sight when not in use, but it is hardly an improvement. In both, three rulings for initials keep them in column, and aid in finding any name. The Pages column answers for book numbers, and

the wide column, Residences, takes short titles nicely. Both books are well made, of good paper, and strongly bound, and are "printed" indexes, i. e., have the initial letters printed on the left of the pages, as well as on the outside margins.

Burr's is superior in having the notches all cut on curves, thus making tearing less easy than in the common angles, where tears always begin. The printed letters on the pages are of a lighter face, and match better with ms. than the overheavy letter of the Graves's, which is too black and staring. The binding and mechanical work of the Burr's is a little finer and better than the Graves's, and it has in the Library Index three pages of interesting general matter on indexingunfortunately, however, devoted to showing its importance rather than to giving practical hints. In comparing prices (which I have not yet done), the number of pages and lines should be noted in each. If one gives the names that a book will hold as equal to the total number of lines, and the other allows for the multitude of blank spaces that occur in every case, thus indicating the number that will fill up certain parts of the book and render a new index desirable, comparisons will be misleading.

The Graves I find best in the greater compactness of the thumb-notches, and because the second letters are in all cases above the first on the page, and within reach of the forefinger without moving the thumb from its place on the first letter. In the Burr, the notches are made larger, and are scattered up and down the whole length of the page. In practice, I find that I can open to fifty different references in the Graves much quicker than in the Burr, with no trouble from the letters being too compact in the Graves, i. e., the finger does not open at the wrong place, or cover two notches at once. The Graves wastes much less space on the page because of this compactness. Dots before the second letter show if it is on the left page, thus saving time. For emergencies there are extra pages at the back, indexed.

The main feature of difference is, however, that Graves prints the first two letters all down the page, while Burr only gives them at the top once, leaving the rest to be written out. Fine penmen might prefer Burr, but it takes more time and is less legible. The printed letters guard against entries in the wrong place, being a constant warning, and, to my mind, the Graves index would be quicker to write up, and quicker to consult, the two most important features.

I think an index could be made for library use, combining the merits of both these, with rulings specially adapted to library wants. This should

also contain as a preface some brief practical hints on making and using indexes. Their importance seems almost an impertinent topic to a librarian, though it is admirable as an advertisement in the book-store. I hope to get one of these makers to get out such a Librarian's Blank-book Index, combining all the excellencies to be found for our special work.

Both Burr and Graves make large indexes divided down to three letters, and adapted for as many as 200,000 entries. For many library uses I believe these printed indexes are to be very widely used, and those of the profession who seek practical labor-savers should test the new candidates. MELVIL DUI.

NEW CHARGING SYSTEM.

At the recent re-opening of the library, July 6, 1880, the following system was put in operation, which, as will be seen, is a modification of that described by Mr. Cutter (LIBRARY JOURNAL, Dec., 1879, v. 4, P. 445–46).

At the end of each book, held in place by a slit in the paper covering (or by a "pocket" if the book be not covered), are two slips (white and manilla), of uniform size with the standard catalogue card; and at the top of each is entered the booknumber. The runner, when taking a book from the shelf to bring it to the delivery-desk, writes the applicant's registration number on both slips. The delivery-clerk, carefully comparing these numbers, dates the white slip and the borrower's memorandum, places the latter in the book and issues it, and places the two slips in their respective racks, whence they will be taken, for filing, at the close of the day. The file of white slips (bookfile) is arranged in the order of the book-numbers, and in a separate numerical order for each one of the 14 days. The file of manilla slips (borrowerfile) is made up of all the issues of the 14 days in one common order, which is that of the registration numbers. [Note.-As the heading of both white and manilla slips is the book-number, the arrangement by registration-number follows always the one last entered on the manilla slip.] When a book is returned, the slips are taken from their files and replaced in the book, and the book replaced on the shelf. The whole official record of the book is, therefore, so entered as not to be exposed to loss by careless handling, and satisfactorily answers the following questions:

(1) Is a given book in or out? (2) If out, who has it? (3) When did he take it? (4) When is it to be sent for, as over-due? (5) Has the book never been out? (6) How many times (and at what dates) has the book been out? (7) Has a

given person a book charged to him? (8) How many books (total) were issued on a given day? (9) How many in each class? (10) How many books are now out, charged to borrowers? (11) How many are at the bindery, etc.? (12) How many persons now have books charged to them? (13) Are these the applicants who registered earliest or latest? (14) Has a person had a given book before? (15) How many times has the book been consulted at the library?

[NOTE. The above is an exact reprint of a lithogram copy received from Mr. Foster. It is evidently intended for use in answering inquiries about his new system. Such use of the lithogram is worthy of commendation. Anything that one is likely to want to explain, sooner or later, to a number of different people, can be written in this way, and a copy sent as occasion calls. The lithogram is already much used for obtaining information. It is certainly philanthropic for one to take the same pains to give new ideas that others take to get them.-ED.]

SLIP-INDICATOR.

JUDGE CHAMBERLAIN has introduced in the Bates Hall of the Boston Public Library a device to obviate the annoying long delays before a reader learns whether a book called for is in or out. In the old system a reader has sometimes waited twenty or thirty minutes only to learn that the book was not in. This difficulty inheres in all great libraries where there are, at certain hours, crowds of readers, and where runners are compelled to go long distances to look for books. By the new plan every book issued has a duplicate slip, postalcard size, filled with its numbers. These are arranged like a card catalog, with projecting guidecards, bearing alcove and range numbers. This check-box contains, also, slips for all books lost or in the bindery, or withdrawn for any reason from regular places on the shelves. A large wooden block, brick size and shape, pushed against the end of the column of slips, keeps them upright.

The call-slips, as handed in to the desk, go first to this new check-box. A glance shows whether the book is off the shelves. If so, the slip goes back at once to the reader, thus saving his time and the legs of the runner. One girl does all the extra work, and at leisure erases the pencil numbers from the fine, heavy cards used for slips, which are thus used over and over like a slate. The idea of this is to afford a large card, of good quality, so as to handle easily. By using a very soft pencil the marks erase almost as easily as chalk.

It is thought that one-third the time is saved by this device, which I have named the slip-indicator.

Though original with Judge Chamberlain, it has also been used in many other libraries, and devised independently by one other, at least (Amherst College). Usually the box has done duty as a charging-box as well as an indicator. I have never known it used but with success.

In Amherst, I reduced the size of the slip to 5×5 cm. (2X2 in.), and used cheap manilla paper, costing only 12c. per 1000. We could thus afford to dispense with erasures. The reduced size saved three-fourths of the space, and answered every purpose. Finally, I devised the sloping check-box, described in v. 3, pp. 230 and 370, and v. 4, p. 14, of the JOURNAL. The little separate boxes enabled attendants to distribute, consult, or remove the slips in half the time required before. Instead of finding the place in a long row, and separating out the slip wanted, it was found alone, or with a few others standing loosely in its own box. In fact, it was common to read the numbers without touching the slip at all, and we saved by as much as sight is quicker than clumsy fingers. Often the box was empty, and the clerk saw at a glance that the book must be in, as none were out from the five shelves represented by each box. By constant use, one attains a marvelous speed in consulting such a box, and the time saved to runners is a very important item. It becomes doubly valuable in small libraries, where no special runners are employed, and the librarian gets all books himself. At Boston, whenever the runners are at leisure (every day more or less), they verify the box by taking a handful of slips, and going through the shelves to see that all books are out as marked.

We verified at Amherst only monthly, and found few mistakes. I cannot too strongly recommend this check-box and slip-indicator to libraries where books are scattered or on different floors. In small collections, where the books are all within three or four steps of the desk, it is of course unimportant for this use, though serving others almost as important.

MELVIL DUi.

USELESS WORDS.

I RECENTLY had occasion to visit a library where all the notices had about a half more words than necessary. They were less easily read, more crowded, and in all ways undesirable because of this verbiage. Words, words wherever they could be grammatically put in-c. g., insted of “List of Periodicals," as heading for that list, a great staring THE is prefixt-and so on. In notices, rules, etc., don't use an unnecessary word. fill out a rounded sentence. Be clear, but, above all things, be brief. Don't, in the name of common

Don't

sense, preface a great, staring NOTICE to everything. It is meaningless, and hides rather than brings into prominence. A row of placards headed alike "NOTICE" are seldom read. Choose the significant word, and make that large and black. If the notice is "No Smoking," bring out those words so they catch every eye-and omit the senseless NOTICE. Intelligent people don't go about studying notices, but if their eyes catch a word bearing on something they are about to do, they will read and heed. A man may carry his umbrella through an art gallery with 50 NOTICES not to carry umbrellas, but 2 or 3 UMBRELLAS would catch the eye of every man carrying one. In short, in rules, notices, etc., bring out with black face or caps the significant word, and convey the sense clearly in the fewest possible words.

COMMUNICATIONS.

M. D.

THE FUTURE OF THE A. L. A. NEBRASKA STATE LIBRARY, LINCOLN, Nov. 29, 1880. DEAR SIR: By all means, "go ahead" with work of the "A. L. A.," and as long as the present incumbent of the Nebraska State Library is in office, put him down for $25 a year in aid of the work. If there are 400 members of the Association, it seems to me the question of salary is easily solved by making annual dues $5, and fixing the salary at $1,500, certainly not less than $1,200. Very truly yours,

GUY A. BROWN.

Letters like the above are full of encouragement. Mr. Brown suggests an admirable lady for the A. L. A. work. On the other hand, a gentleman of long library experience, a graduate of Harvard, and with several years' practical experience in the book trade and in special cataloging, has offered to undertake the A. L. A. work, and open as well the Clearing-house for Duplicates if that is decided upon, at only $1,200 a year guaranteed salary.

Mr. Brown's plan of raising annual dues would meet our wants, or, better still, the legacy or large gift which the A. L. A. so richly merits, and for which some of us have not only been hoping, but quietly working. The Washington Conference, with its other good work, ought to settle the questions raised in the last JOURNAL under "Past, Present, and Future of the A. L. A." I suggest for the new office the title of Treasurer. This office is vacant, and in fact the worker at the center must be Corresponding Secretary and Treasurer, and might better be called so. M. D.