College and Research Libraries By J O H N M I N T O D A W S O N Xerography in Card Reproduction Mr. Dawson is assistant director, Univer- sity of Chicago Library. TH E R E P R O D U C T I O N of catalog cards has been a persistently annoying problem since libraries began using card catalogs. A t first, cards were laboriously written in li- brary hand; then, with the introduction of the typewriter, as many cards as were re- quired were typed individually, a practice still employed by some libraries. Cards have been printed by libraries, usually by the use of the Multigraph, though sometimes on a job press. T h e Library of Congress, of course, continues to print its cards for its own catalogs and for distribution to other libraries; but most libraries which printed their cards found the cost of setting type and distributing it too expensive. Even when the type was not distributed after printing but was sold as metal, still the cost of typesetting and of buying new type was exorbitant for the limited number of copies run off. T h e next step after the introduction of the typewriter was the adaptation of various duplicating processes to the reproduction of catalog cards. Experiments with gelatin reproduction processes were generally un- successful ; the analine dyes used in the ink were of an unpleasant color, and, more serious, they soon faded; the images were fuzzy, and results were uneven and un- certain. T h e use of stencils proved much more successful, and many libraries now rely on this process with improved stencils, inks, and duplicating machines. A variation of the stencil method is that employed in some public libraries where stencils normally used for addressing envelopes are used.1 H o w - ever, the limited space available for text on these stencils restricts the use of this method largely to fiction, or to other materials re- quiring only brief cataloging, so that it is not feasible for research libraries. M a n y libraries have, in recent years, turned to the special offset presses which were designed mainly for office use and which are simple enough to be operated— after a brief training period—by unskilled personnel. These office offset presses use paper mats on which copy is typed or printed, or photographed (on photosensi- tive mats), and employ a permanent ink, similar to the ink used in regular printing, which produces clear images. One difficulty encountered in using this method is that the ink, having an oil base, is not readily absorbed by the smooth surface of the 100 per cent rag card stock, which sometimes causes smearing or the offsetting of the image from one card onto the back of an- other card during the long period before the ink has completely dried. However, using a quick drying ink or a combination of ink and a drying compound with the ink appears to eliminate this difficulty. T h e Library of Congress prints offset mats when printing its cards so that when the original stock is exhausted additional copies may be inexpensively reproduced by offset print- ing. One of the most promising means for card reproduction which has recently been developed is the combination of xerography2 1 B o w e r s , M . E-., " C a t a l o g i n g with S t e n c i l s , " Library Journal, 65:462-63, J u n e 1 , 1940. 2 X e r o g r a p h y , or, to use the trade name, X e r o X , w a s developed by the Battelle Memorial Institute and the Haloid Corporation. A discussion of xerography in li- braries and a list of articles on it may be found in a JANUARY, 1954 57 and offset printing. Xerography is an elec- trical and mechanical rather than a chemical method of reproduction, as the various pho- tographic methods are. Instead of a photo- graphic negative, xerography employs an aluminum plate covered with an extremely thin coating of selenium, a "photoconduc- tive" material. T h e peculiar property of selenium, which is also the one that makes xerography possible, is that it conducts elec- tricity only when exposed to light; in the dark, it is a nonconductor. In xerography, the selenium covered alu- minum plate is sprayed with an electrostatic charge. In a camera unit, the plate is ex- posed to the material to be copied. As light is reflected from the white surfaces of the copy onto the charged selenium, the selen- ium becomes a conductive material and dis- charges the electrostatic charge into the aluminum part of the plate; but the dark surfaces, i.e., the writing, drawing, or typ- ing, do not reflect the light, so that there is a pattern of dark on the selenium plate, and the selenium remains a nonconductor and retains its electrostatic charge. There is, then, in effect, an invisible pattern of electrostatically charged selenium which is a mirror image of the material to be copied. T h e plate is kept from the light and a dry developing powder is cascaded over it. Par- ticles of the powder are attracted to and held by the electrostatic charge. N o w the plate may be exposed to light, and a mirror image of the original copy can be seen on the plate. Next, a sheet of paper or a paper offset mat is placed on the plate, and is smoothed down so that the paper is in contact with the plate. T h e paper is then sprayed with an electrostatic charge which attracts the powder from the surface of the plate. W h e n the paper is removed from the plate, a true pamphlet edited by Hodgson, J a m e s G., The Use of Xerography in Libraries. Fort Collins, Colorado, Colo- rado A . & M . College L i b r a r y , 1 9 5 2 . A new edition was issued in 1 9 5 3 . image of the material copied can be seen. A t this point, the powder is held to the paper only by the electrostatic charge, so that, if desired, unwanted parts of the copy can be wiped off. T h e paper is then in- serted into an oven which fuses the powder onto the paper in a permanent reproduction of the original copy. A sheet of copy or a strip of cards can be reproduced in from three to four minutes. If the powder has been transferred to a sheet of paper, only one copy will have been produced. If, however, the powder has been transferred to an offset mat, as many copies as are required may be obtained when the mat is run on the offset press. It is this combination of xerography and offset print- ing which is of interest in card reproduc- tion. U S E A T C H I C A G O For many years, the catalogers at the University of Chicago Library have typed their copy as they catalog material, thus providing the typists with clear and legible copy from which to type stencils or offset mats. A study of the copy slips (or cards) showed that approximately 95 per cent of these cards were free from typographic er- rors and could be reproduced without alter- ation or correction.3 T h a t meant that 95 per cent of the copy produced by the cata- logers could be reproduced by the xerog- raphy-offset process. Since typists were dif- ficult to obtain, expensive to train, and had a high rate of turnover, this method seemed to hold promise for relieving an unfortunate situation. It seemed, too, that this might well be a faster and more economical method than having the copy typed on mats and printed by offset, since it eliminated the need for proofreading and correcting the typist's work. W e therefore began experimenting with 3 T h e catalogers are not considered to be typists and no effort is or wrll be made to urge them to produce perfect copy. W h e r e copy is not suitable f o r reproduc- tion, it is sent to a typist to copy on an offset master. SO CO. LEGE AND RESEARCH LIBRARIES xerography-offset for reproducing our cata- log cards. In studying the production of cards by offset, we had concluded that re- producing one set of cards at a time was not efficient, and that, with the equipment we had, the most efficient way was to print four sets at a time.4 T o do this, we had typed copy for four cards on half an offset mat cut vertically, and reproduced them'on strips of card stock cut according to our specifications, 30 cm. by 12.5 cm., with four holes drilled at the appropriate places. A relatively simple homemade jig on a guillotine paper knife made it possible to cut these strips accurately into four cards 7.5 cm. by 12.5 cm., the standard size of catalog cards. It was obvious that it would be equally advantageous to reproduce four sets of cards simultaneously by xerography- offset. T h e X e r o X camera unit is equipped with somewhat primitive registration guides, to- tally inadequate for the precision required in copying catalog cards. A simple guide made of two pieces of plastic forming a right angle was devised and affixed to the camera unit. A piece of stiff binders board was cut with true corners to fit on the camera unit, and two strips of masking tape with adhesive on both sides were pressed onto this board, so placed that when cards are placed on these strips, accurate registration of the images on the xerographic plates is possible when the board is placed within the guides affixed to the camera. This process of reproducing catalog cards requires, then, these steps: 1. Sorting the cards produced by the cata- logers to remove those which require re- typing, and, at the same time, sorting these cards by the number of copies required. 2. Placing the cards on the registration board and placing this board on the camera unit. 4 W e have a Multilith model 75. Some libraries with a Multilith model 1 2 5 0 or similar equipment reproduce eight sets of cards at a time. 3. Making an offset mat by xerography as described above. The lines between the cards and the holes in the cards cast shadows which would be reproduced if they were not removed. They can be removed at either of two different stages in the process: from the plates before the image is transferred to the mat, or from the mat before the image is fused. Cotton swabs have been found most satisfactory for this purpose. 4. Printing by offset the required number of strips of catalog cards. 5. Cutting the strips into four parts. (The strips for more than one set of cards can, of course, be cut at the same time.) After determining that the xerography- offset method of card reproduction was tech- nically satisfactory, the questions which re- quired answering were: H o w much does it cost? D o we save time? If so, how much? Records were kept of the time required for the operations involved in both processes over a period of time, and the cost of labor and materials divided by the number of units (the number of sets of cards) pro- duced. In estimating the cost of rental of the X e r o X equipment (it is available only on a rental basis), the monthly cost was divided by the average number of sets of cards reproduced per month. W e had already determined that the cost of producing ten copies of a card by typing and offset printing was 23.3 cents, and re- quired about 7.8 minutes of labor, exclusive of sorting by number of copies required. O u r study of the cost of the xerography- offset method indicated that the cost for ten copies of a single card was 20.8 cents, and that 2.7 minutes of labor (again exclud- ing the sorting by number of copies re- quired) were required. These costs were determined on the basis of our written cards only, and the cost differential was not great. T h e labor differential is substantial, and the relatively high cost of the xerog- raphy-offset process, in relation to the labor involved, is due to the fixed rental charge JANUARY, 1954 59 which must be apportioned to the number of sets reproduced. If we reproduce L C cards5 from our proofsheet file and our old depository catalog, the cost of the reproduc- tion of our written cards decreases from 20.8 cents to approximately 16.8 cents for ten copies of a card. T h e cost of reproducing L C cards is the same as the cost of reproducing our own written cards, or 16.8 cents per set of ten copies. This is considerably less than the cost of ordering a set of ten L C cards, which, if ordered by number, would cost 6 cents for the first card and 3.5 cents for each additional card, a total of 37.5 cents. T h i s 37.5 cents cost does not of course, include the cost of ordering the cards. Since the proofsheet or depository card must be found in either case (since we order by card number), this cost is the same whether we reproduce L C cards or order them. T h e actual cost of placing the order and receiv- ing and sorting the cards when they are re- turned costs us approximately 1.2 cents per set of cards. T h e cost of ordering L C cards, then, excluding the determination of the card number, is 38.7 cents, in contrast to our cost of 16.8 cents (again excluding the locating of the proofsheet or depository c a r d ) . There is another economy: when the L C proofsheet goes to the cataloger the call number is added and corrections made, so that, when the proofsheet is reproduced, all cards have call number and corrections on them; whereas the call number and the corrections must be added to each card when the set is obtained from the Library of Congress. It must be noted that in the cost figures given above, not all factors are accounted 5 " . . . anyone is f r e e to copy L C catalog cards, adapt them, use them, or even sell them. There is no copy- right on them and L C retains no literary r i g h t s . " LC Information Bulletin, V , 12-.N0.22, p.12, J u n e 1 , 1 9 5 3 . for, and these costs in no instance reflect the full cost of sets of cards. In no instance is overhead, either direct or indirect, in- cluded, nor is the depreciation of equipment included. These elements are common to each of the processes and procedures and have not been considered. In determining the cost of the X e r o X refttal per set of cards, the monthly rental fee was divided by the normal number of sets of cards reproduced each month. A c - tually, the rental should be divided by the number of exposures made each month, whether these exposures are for card repro- duction or for the reproduction of any other type of material. This, of course, will tend to reduce the cost of card reproduction by xerography-offset since the share of the rental allotted to card reproduction will be less. T h i s method of card reproduction is not recommended without reservation. In the University of Chicago Library, the fact that the catalogers type their copy cards is a fac- tor in the success of our operation. If type- writers had to be acquired and catalogers trained to type their copy, it is doubtful that the method would be advantageous for some time. Secondly, the rental fee for X e r o X equipment is substantial (our equip- ment rents for $60 per month, but newer models and more flexible camera units have higher rental fees), and a large volume of work is required to reduce the unit cost to reasonable rates for libraries. T h e xerography-offset method has been in regular use at the University of Chicago Library since October, 1952. It is a normal part of our operations, and has enabled us to eliminate a backlog of cards awaiting re- production and to speed up the reproduction of cards in spite of an acute shortage of typ- ists. SO CO. LEGE AND RESEARCH LIBRARIES