John Willis Clark

RECAPITULATION. INVENTION OF THE STALL-SYSTEM. LIBRARY OF CORPUS CHRISTI COLLEGE, OXFORD, TAKEN AS A TYPE. SYSTEM OF CHAINING IN HEREFORD CATHEDRAL. LIBRARIES OF MERTON COLLEGE, OXFORD, AND CLARE COLLEGE, CAMBRIDGE. THE STALL-SYSTEM COPIED AT WESTMINSTER ABBEY, WELLS, AND DURHAM CATHEDRALS. THIS SYSTEM POSSIBLY MONASTIC. LIBRARIES AT CANTERBURY, DOVER PRIORY, CLAIRVAUX.

If the evidence brought forward in the last chapter be accepted, the Library which a Monastery or College built in the fifteenth century was a long narrow room lighted by rows of equidistant windows. Occasionally, if neighbouring buildings allowed, there was a window at the end of the room also. The fittings were lecterns of wood. On these the books were laid, each volume being fastened by a chain to a bar usually placed over the desk, but occasionally, in all probability, in front of it or beneath it. The readers sat on benches immoveably fixed opposite to each window. It is obvious that reading was convenient enough so long as the students were few, but if they were numerous and the books chained too closely together much annoyance must have been caused. When the University of Oxford petitioned Humphrey Duke of Gloucester in 1444 to help them to build a new library, they specially dwelt upon the obstacles to study arising from the overcrowded condition of the old room. "Should any student," they said, "be poring over a single volume, as often happens, he keeps three or four others away on account of the books being chained so closely together[332]."

Further, the lectern-system was so wasteful in the matter of space, that, as books accumulated, some other piece of furniture had to be devised to contain them. The desk could not be dispensed with so long as books were chained; and it therefore occurred to an ingenious carpenter that the required conditions would be fulfilled if the two halves of the desk were separated, not by a few inches, but by a considerable interval, or broad shelf, with one or more shelves fixed above it. Thus a case was arrived at containing four shelves at least, two to each side of the case, which could be made as long as the width of the library permitted. I propose to call this system "the stall-system," from the word staulum (sometimes written stalla, stallus, or stallum), which is frequently applied to a case for books in a medieval library.

There are at least five fine examples of this system at Oxford—none, I am sorry to say, at Cambridge. There was a set at Clare College, supplied to the old Library about 1627, but they have since been altered by the removal of the desks. Those at Oxford are at Corpus Christi College (1517), S. John's College (1596), Sir Thomas Bodley's library (1598), Merton College (1623), Jesus College (1677-79), Magdalen College (of uncertain date).

As a type of this system I shall take the library of Corpus Christi College, founded in 1516 by Richard Fox, Bishop of Winchester. The library was ready for the fittings by the end of March in the following year, as we learn from a building account preserved by Hearne:

8 Henry VIII. This boke made from the xvth day off March unto the xxxti day off the same Moneth [30 March, 1517].

Md. couenauntyd and agreid wyth Comell Clerke, for the makyng off the dextis in the liberary, to the summe off xvi, after the maner and forme as they be in Magdaleyn college, except the popie heedes off the seites, thes to be workmanly wrowght and clenly, and he to have all maner off stooff foond hym, and to have for the makyng off on dexte xs the sum off the hole viii. li.[333]

The arrangement and appearance of these most interesting cases will be understood from the general view (fig. 70) and from the elevation (fig. 71), but I shall proceed to describe them with some minuteness.

The library occupies the first floor of the south side of the quadrangle opposite to the entrance. It is 79 feet 6 inches long, by 21 feet broad, with ten equidistant windows, about 3 feet 6 inches apart, on each side. At the west end there is an inner library, occupying the angle between the south and west sides of the quadrangle. On each side there are nine bookcases, each 8 ft. 6 in. high, 2 ft. wide, and 7 ft. 6 in. long, divided by partitions into three compartments.

I have carefully studied these cases on several occasions, and it seems to me that the only alterations introduced since the original construction are: (1) the addition of about two feet to the upper portion of the case in order to provide additional shelf-room; (2) a slight change in the arrangements of the desk for the reader; and (3) the addition of the catalogue frame, which by its style is evidently Jacobean, to the end next the central alley. Originally each case had two shelves only, one on the level of the desk (Fig. 71, G, H), and the second about half-way between it and the original top of the case (ibid. E, F). Before chaining fell into disuse the cases were heightened so as to provide an additional shelf (ibid. C, D). At present the number has been further increased by the addition of a fourth shelf above the desk (ibid. A, B), and two below it (ibid. I, K, L, M). The desks have been altered by a change in the position of the bracket, and by the suppression of the slit through which the chains usually passed, as I shall explain below.

The system of chaining used for the lectern-system required modification and extension to suit this new arrangement of shelves. At Corpus Christi College most of the iron-work remains (fig. 70); but it is necessary to go elsewhere to find chained books actually in use. Of such chaining I know no better example than the Chapter Library in Hereford Cathedral, from a study of which I will describe the system, and shew that it is the same as that employed at Corpus Christi College and elsewhere.

The Chapter Library at Hereford was originally over the west cloister, and there is evidence that it was being fitted up in 1394, when Walter de Rammesbury, B.D., gave £10 for the desks[334]. The original building has long since been destroyed, and the books were transferred from one place to another until the present beautiful structure was built on the old site in 1897.

Throughout these changes some very ancient bookcases have been preserved. They have been taken to pieces and altered several times, but are probably, in the main, those put up in 1394. Above all, one of them possesses, in thorough working order, the system of chaining, parts of which are to be met with on the cases at Oxford which we have been considering. Of the accompanying illustrations the first (fig. 72) gives a general view of the most complete case, that which now contains the manuscripts, and the second (fig. 73) shews one compartment of the same case with the books, chains, desk etc. This case is 9 ft. 8 in. long, 2 ft. 2 in. wide and 8 ft. high, exclusive of the cornice. The material is unplaned oak, very rough; the ends are 2-3 in. thick, made of three planks fastened together with strong wooden pegs. The desk has been a good deal altered, and is now inconveniently low, but, as the books were chained, it is evident that there must always have been desks on each case, and moreover the hook which held them up is still to be seen in several places. The frames to contain the catalogue, which closely resemble those at Oxford, are known to have been added in the 17th century by Thomas Thornton, D.D., Canon Residentiary.

As the books were to stand upright on a shelf, not to lie on their sides on a desk, it was necessary to attach the chain in a different manner. A narrow strip of flat brass was passed round the left-hand board (fig. 74) and riveted to it, in such a manner as to leave a loop in front of the edge of the board, wide enough to admit an iron ring, an inch and a quarter in diameter, to which one end of the chain was fastened. The book is placed on the shelf with the fore-edge turned outwards, and the other end of the chain is fastened to a second ring, rather larger than the former, which plays along an iron bar (fig. 75). For the two upper shelves these bars, which are ½ in. in diameter, are supported in front of the shelf, at such a distance from it as to allow of easy play for the rings (fig. 73). Each bar extends only from partition to partition, so that three bars are needed for each shelf. For the lowest shelf there are also three bars, set two inches behind the edge of the shelf, so as to keep the rings and chains out of the way of the desk. The bars for the upper shelves rest in iron sockets screwed to the woodwork at the juncture of the horizontal shelves with the vertical divisions and ends respectively. The socket fixed to the end of the bookcase which was intended to stand against the wall is closed by an iron plate (fig. 76), so that the bar cannot pass beyond it. At the opposite end, that which would usually face the alley between the two rows of bookcases, the bars are secured by lock and key in the following manner. A piece of flat iron is nailed to the end of the bookcase, just above the level of the uppermost shelf (fig. 77). Attached to this by a hinge is a hasp, or band of iron, two inches wide, and rather longer than the interval between the two shelves. Opposite to each shelf this iron band expands into a semicircular plate, to which a cap is riveted for the reception of the head of the socket in which the bar rests (fig. 77); and just below the middle shelf it drops into a lock and is secured by a key (fig. 73). A second hasp, similarly constructed, secures the lowest of the three bars; but, as that bar is behind, and not in front of, the shelf to which it belongs, the arrangements described above are reversed. One lock and key serves for the ironwork belonging to the three shelves.

The chains are made of links of hammered iron as shewn in the sketch (fig. 78) which represents a piece of one of the actual size. There is usually a swivel in the centre, probably to prevent twisting. They vary somewhat in length, and in the length of the links, according to the shelf on which the books to which they belong are ranged, it being obviously necessary to provide for the convenient placing of a book on the desk when a reader wished to consult it. The most usual dimensions are 3 ft. 4 in., 3 ft. 6 in., 4 ft. 3 in.

The removal of any of the volumes, or the addition of a new one, must have been a tedious and inconvenient operation. The bar would have to be withdrawn, and all the rings set free. Moreover, if this change had to be effected in one of the compartments remote from the end of the case which carried the lock, the bar belonging to each of the other compartments would have to be withdrawn before the required volume could be reached.

If the views (figs. 70, 71) of the book-cases at Corpus Christi College, Oxford, be attentively examined, it will be seen that the ironwork exactly resembles that at Hereford. We find similar sockets to contain the bars at the junction of the horizontal shelves and vertical uprights, and a similar system of iron hasps to prevent the bars from being withdrawn.

The desk for the reader would of course vary according to individual taste. As a general rule it was attached to the ends of the case by strong hinges, so that it could be turned up and got out of the way when any alteration in the ironwork had to be carried out. Iron hooks to hold it up were not unfrequently provided. One of these, from the Bodleian Library, is here figured (fig. 79). It was also usual to provide a slit in this desk, about 2 in. wide, as close as possible to the shelf, for the chain attached to the book in use to pass through. This is well shewn in the view of a single book-case in Merton College, Oxford (fig. 83).

I will next describe the library of Merton College, Oxford. There is still considerable doubt respecting the date of some of the bookcases, but the appearance of the library is so venerable, so unlike any similar room with which I am acquainted, that it must always command admiration, and deserve study[335].

The library occupies the whole of the first floor of the south side of "Mob Quadrangle" and the greater part of the same floor of the west side (fig. 80). It is entered through a doorway in the south-western angle of the court, whence a staircase leads up to the vestibule (fig. 81). This room is separated from the two divisions of the library by lofty oak screens, elaborately carved and ornamented in the style of the early renaissance.

The two rooms into which the library is divided have a uniform width of 20 ft. 6 in. The west room, called by tradition Old Library, is 38 ft. 6 in. long (A, B, fig. 81); the south room, or New Library, is 56 ft. 6 in. long (C, D, fig. 81).

The west room is lighted by seven equidistant lancet windows in each of the west and east walls, and by two dormer windows of peculiar design on the side of the roof next to the court. The south room is similarly lighted by ten lancets in each of the north and south walls, and on the side next to the court by two dormer windows like those in the west room. This room moreover has an open space at the east end, about 10 ft. long, lighted by a window of two lights in each of the north and south walls respectively, and by an oriel of five lights in the east wall. In both rooms there is a waggon-roof of five cants boarded, and divided into panels by molded ribs with little bosses at the intersections (fig. 82).

The blank wall at the north end of the west room is panelled with oak of an elaborate and beautiful design for a height of about 12 ft. (fig. 82). The space above this is decorated with panels of plaster-work. The large square central panel contains the arms of the college; the circular panel to the west those of John Whitgift (Archbishop of Canterbury 1583-1604); and the similar panel to the east those of Sir Henry Savile (Warden 1585-1621).

The east end of the south room is similarly treated, but the oak panelling is less elaborate. In the plaster-work above it the arms of the college are flanked on the north by those of George Abbot (Archbishop of Canterbury 1611-1633) and on the south by those of Sir Nathaniel Brent (Warden 1621-1651).

Both rooms are floored with rough oak planking. On this are laid four sleepers, each about 5 in. square, parallel with the side-walls. The two central sleepers have their outside edge roughly chamfered. Into these the bookcases and the seats are morticed. The central alley, 5 ft. wide, is in both rooms paved with encaustic tiles.

In the west room there are twelve complete cases and four half-cases; in the south room there are twenty complete cases and two half-cases (fig. 81); in both rooms arranged in the usual manner with respect to the walls and windows.

In order to present as vivid an idea as possible of these beautiful cases, I reproduce here a photograph of a single compartment from the west library, with a seated reader at work (fig. 83). The case is made to look rather higher than it really is, but this distortion can be easily corrected by comparing the height of the standard with that of the seated figure.

In the west room each case (figs. 82, 83) is 7 ft. 5 in. long, 1 ft. 5 in. wide and 6 ft. high from the top of the sleeper to the top of the cornice. The material is oak. The ends are nearly 2 in. thick, and next the wall are shaped roughly with an adze. Each case is separated into two divisions by a central partition; and originally there was a desk 1 ft. 3 in. wide on each side of the case. These desks were immoveable, and nailed to rough brackets. There were two shelves only to each case: one just above the level of the desk, and a second about half-way between it and the cornice (fig. 84).

The system of ironwork by which the books were secured can be easily recovered by studying the scars on the ends of the cases next the central alley. At the lower end of the standard, two feet from the ground, was an iron bar which carried the chains of all the books which stood on the shelf just above the level of the desk, without reference to the side from which they were to be consulted. This bar was secured by a separate hasp and lock. The bars for the upper shelf, one on each side of the case, were obviously secured by a system similar to that described above at Hereford and Corpus Christi College. The whole system has been indicated on the elevation (fig. 84), which should be compared with the reproduction of one of the cases in the west room (fig. 83). Originally no books stood below the desk. The comfort of readers was considered by the insertion of a bar of wood to rest the feet on, between the seat and the bookcase (fig. 84).

In the south room the cases are on the same general plan as in the west room; but the system of chaining appears to have been slightly different, and to have approximated more closely to what I may call the Hereford type.

In both rooms each case has a picturesque enrichment at the end of the standard above the cornice, and a small oblong frame just below it to contain the general title of the books within the case. The west room is devoted to LIBRI ARTIUM, with the exception of the three cases and the half-case at the north end of the east side, which are marked CODICES MSS. These are protected by latticed doors of wood. In the south room the cases on the south side are all lettered L. THEOLOGIAE; on the north side the first three are lettered L. MEDICINAE; the next L. MEDIC. IURISPP. and the last five L. IURIS PRVDENTIÆ. In this room the last cases at the east end on each side have latticed doors like those on the corresponding cases in the west room.

The building of this library is recorded in four separate account-rolls extending from the beginning of the first year of Richard II. to the third year of the same king, that is from 1377 to 1379. From these documents it appears that the building cost £462. 1s. 11½d.

From this first construction to the beginning of the sixteenth century—a space of 125 years—the accounts furnish us with no information; but, from what we learn afterwards, it would appear that the internal walls were unplastered, that the roof-timbers were unprotected, and that the only light was admitted through the narrow lancet windows.

In 1502-3 the panel-work (celatura) on the roof of the west library was put up at a cost of £27. 6s. 0d. The account contains also a charge for painting the bosses (nodi) at the intersection of the moldings that separate the panels. Mr Henderson points out that these ornaments prove the existing ceiling to be that put up in 1503; for among them are the Tudor Rose, the dolphin of Fitzjames (Warden 1483-1507). and the Royal Arms used from Henry IV. to Elizabeth, but altered by James I.

After this another long interval occurs during which no work done to the library is recorded; but in 1623 the south room was taken in hand, and the changes introduced into it were so extensive that it is referred to in the accounts as New Library (Nova Bibliotheca) a name which it still retains.

In the first place the room at the east end (fig. 81) was thrown into it, and the oriel window constructed, together with the two large dormers on the side next the court (fig. 80). These works, by which light was so largely increased, prove how gloomy the library must have been before they were undertaken. Next, after important repairs to the walls and floor, and the construction of the decorative plaster-work at the east end, the old bookcases were sold, and Benet the joiner supplied twenty new cases and one half-case. The only old case remaining is, by tradition, the half-case against the screen on the north side as one enters from the vestibule.

It is therefore certain that the cases and seats in the south room date from 1623. It is unfortunately equally certain that we know nothing about the date of those in the west room; and we are therefore unable to say whether the cases in the south room were copied from them in 1623, or whether the reverse process took place at some unknown date. If we adopt the pleasing theory that in the west room we have very early cases, constructed possibly when the library was built, we must still admit that these relics of a remote past have been altered at some subsequent period, so as to be brought into conformity with the cases in the south room; for the cornices and the frames for the titles are precisely similar in the two rooms.

The difference between the two sets of cases in the method of chaining, to which attention has been already drawn, may bear on the question of date. As time went on chaining would be modified in the direction of simplicity; and to replace a single central bar by two lateral ones is a step towards this, for under such conditions the addition or removal of a book would entail less displacement. Further, it must be recognised that these cases, whether extremely ancient or comparatively modern, differ in many particulars from those to be met with elsewhere. They are lighter, narrower and more elegant. Again, when the ground-plan of the library is considered (fig. 81) it will be seen that their ends occupy nearly the whole space between a pair of windows. In other examples of the stall-system this is not the case.

The only explanation I have to offer for the whole difficulty is the following. The library was constructed for the lectern-system, with wall-spaces not more than 2 ft. wide, and was so fitted up. When books had become numerous the western library was taken in hand, and the lecterns altered into stalls, the single central bar being retained. At the same time, in all probability, the dormers were inserted. It is remarkable that these changes should not be recorded in the accounts, but possibly they were carried out as the result of a special benefaction[336]. In 1623 the stalls which had been placed in the west room, having been found convenient, were copied for the south room.

I will in the next place briefly notice the distinctive points of the other examples of the stall-system in Oxford.

At S. John Baptist's College the library was built in 1596, and we may presume was fitted up soon afterwards, as Wood records numerous donations of books in the years immediately succeeding, and the appointment of a keeper to take charge of them in 1603[337]. This library, on the first floor of the south side of the second quadrangle, is 112 feet long by 26 feet wide, with eight windows of two lights in each wall. The bookcases, of which there are eight on each side between the windows, with a half-case against the west wall, are rather larger than those at Corpus Christi College, being 10 feet high, and 2 feet 6 inches wide. They have a classical cornice and terminal pediment. The titles of the subjects are painted at the tops of the stalls as at Merton College. A few traces of chaining are still to be detected. The desks have not been altered. Each is in two divisions, as at Corpus, separated by a central bracket, and it has the slit to admit the chains. The long iron hinges are evidently original. The seats resemble those at Corpus.

The bookcases at Trinity College, set up in 1618, and those at Jesus College, made probably in 1679, call for no special remark.

Between 1598 and 1600 Sir Thomas Bodley refitted the library over the Divinity School. This noble room is 86 feet long by 32 feet wide. These dimensions contrast forcibly with those of the long narrow rooms to which we have been accustomed; and it is probably on account of the great width that the 10 windows on each side have two lights apiece. At right angles to these walls, which face north and south, there are nine bookcases on a side with a half-case at each end. Here again we find so close a resemblance to the cases at Corpus Christi College, that a particular description is unnecessary. It should be noted, however, that, as at S. John's College, they had been made of a greater height (8 feet 4 inches) in the first instance, so as to accommodate two shelves above that on the level of the desk. These shelves are proved to be original by the existence, at the juncture of the shelves with the upright divisions, of the plates of iron which originally carried the sockets for the bar. The rest of the ironwork has been removed, and it is difficult to detect traces of its former existence, because modern shelves have been set against the ends of the cases. The hole for the lowest bar, however, remains in the same relative position as at Corpus Christi College; and, as the ironwork for supporting the bars is identical with what still remains there, it seems safe to conclude that no new principle was introduced. The desks are modern, but the large and ornamental brackets which support them are original, and the iron hooks (fig. 79) still remain by which they were prevented from falling when turned up. The position of these hooks shews that each desk was 19 inches broad. There were originally seats between each pair of cases, as may be seen in Loggan's view of the interior of the library, where their ends are distinctly shewn.

A special feature of this room is the beautiful open roof, practically that which Sir Thomas Bodley put up in 1599. The principals and tie-beams are ornamented with arabesques, while the flat surface between them is divided into square compartments on which are painted the arms of the University. On the bosses that intervene between these compartments are the arms of Bodley himself.

I mentioned at the beginning of this chapter that the stall-system had been represented in the library at Clare College, Cambridge. The old library was a long narrow room over the old chapel, and we know on the authority of William Cole[338] that it was "fitted up with wainscote Classes on both sides." These "classes" had been put up shortly before 1627, when the Duke of Buckingham, then Chancellor of the University, was taken to see them. When this library was pulled down in 1763 they were removed to the new library which had been fitted up 20 years previously, and ranged round the room in front of the modern shelves. They are splendid specimens of carpentry-work, and bear so close a resemblance to the cases in the library of S. John's College, that it may be assumed that they were copied from them[339]. When the removal took place they were a good deal altered, and a few years ago some fragments which had not been utilised were found in a lumber-closet. One of the standards (fig. 85), with its brackets, shews that the cases were once fitted with desks, the removal of which was ingeniously concealed by the insertion of slips of wood in the style of the older work[340]. I have not been able to discover any traces of chaining, but as there are a number of seats in the library, very like those in Corpus Christi College, Oxford, it is more than probable that chains were once employed.

The stall-system was not only popular in Oxford itself, but was adopted as a standard for bookcases, and reproduced elsewhere.

The first example I will cite is at Westminster Abbey[341], where part of the dorter was fitted up as a library during the years 1623 and 1624 by John Williams, Bishop of Lincoln and afterwards Archbishop of York, who was dean of Westminster from 1620 to his death in 1650. In the flowery rhetoric of his biographer Bishop Hacket:

This library—which has not been materially altered since 1625—occupies the north end of what was once the dorter. It is 60 feet long, by 33 feet 4 inches broad. There are twelve bookcases—evidently the "desks" recorded by Williams' biographer. Each is 10 feet 10 inches long, 2 feet broad, and 8 feet 3 inches high, divided by plain uprights into three compartments. There are three shelves, below which is a desk for the reader, resting on brackets, and provided with the usual slit for the chains to pass through. These desks are hinged. The cases are quite plain, with the exception of a molded cornice; above which, on the end of each, is some scroll-work. There is also a small frame to contain the catalogue. It is probable that there were originally seats for readers between each pair of cases. I cannot discover any certain evidence of chaining, and yet "chains" are distinctly enumerated among the dean's benefactions. There are faint scars at the intersection of some of the shelves and uprights which may be screwholes—but I cannot feel certain on the point.

I have already given the plan of the cathedral library at Wells (fig. 42). After the Restoration this building was re-fitted during the episcopate of Robert Creighton (Bishop 1670-1672), with the help of donations from the celebrated Dr Richard Busby, and Dr Ralph Bathurst, who was dean from 1670 to 1704. It is important to remember that Bathurst was also master of Trinity College, Oxford, an office which he retained until his death. As he is described in the MS. List of Benefactors preserved in the library as having taken a foremost part in fitting it up (in Bibliothec hac instaurand e???d???t??), the selection of the bookcases may with much probability be ascribed to him. His own college has still bookcases which once must have been excellent specimens of the stall-system.

There are eight bookcases at Wells, of plain unpainted deal, projecting from the west wall between the windows (fig. 42). They are 8 ft. 6 in. long, 8 ft. 1 in. high and 3 ft. broad. Seven of them have desks on both sides, but the last—that placed against the partition at the south end, which screens off a small room for a study—has a desk on one side only. There is no shelf below the desk, but two above it. They are fitted with the usual apparatus for chaining. Between each pair of bookcases, in front of the window, is a seat for the reader. These cases resemble so closely those at Corpus Christi College, Oxford, that the source from which they were derived cannot be doubtful.

Was this library ever chained? A Walton's Polyglot, 1657, had evidently been prepared for chaining, and in a novel fashion, the plate to carry the chain being attached to the left-hand board close to the back of the volume (fig. 86)—so that it was evidently set on the shelf in the ordinary way, and not with the fore-edge turned to the spectator, as is usual in chained libraries. But with this exception I could not discover indications of the attachment of a plate on any of the volumes. If I am right in concluding that the books in this library were never chained, the cases are a curious instance of the maintenance of fashion. Dean Bathurst ordered a bookcase, and it was supplied to him with all its fittings complete, whether they were to be used or not.

My last example is from Durham Cathedral, where John Sudbury, dean from 1661 to 1684, fitted up the ancient Frater as a library. The room is about 115 feet long by 30 feet wide, with nine windows in each side-wall. Their sills are ten feet from the ground.

The cases (fig. 87) are evidently the work of a carpenter who was thoroughly conversant with the stall-system. They had originally two shelves only above the desk, the entablature, now visible on the ends only, being carried along the sides. The shelf below the desk is also modern. These cases are ten feet apart, and between each pair, instead of a reader's seat, is a dwarf bookcase terminating in a desk. Attached to it on each side is a seat conveniently placed for a reader to use the desk on the side of the principal case.

I have shewn that the stall-system made its appearance at Oxford early in the sixteenth century, but I have not been able to discover who introduced it. My own impression is that it was monastic in its origin; and I can prove that it fits at least two monastic libraries exactly. This theory will also explain the prevalence of such cases at Oxford, and their almost total absence from Cambridge, where monastic influence was never exercised to the same extent.

I will begin with Canterbury, where, as I mentioned above[343], the library was over the Prior's Chapel. The construction of this chapel is described as follows by Professor Willis:

Above this chapel Archbishop Chichele built the library which Prior Sellyng fitted up. It stood east and west, and of course must have been of the same size as the chapel beneath it, namely, according to Professor Willis, 62 feet long on the north side, 59 feet long on the south side, and 22 feet broad. The door was probably at the south-west corner, at the head of a staircase which originally led only to the chapel beneath it.

From these measurements I have constructed a plan of the room (fig. 88), and of the bookcases which I am about to describe. The windows are of course imaginary, but, I submit, justified by the uniform practice of medieval libraries.

I am able to reconstruct this library because I have had the good fortune to come across a very curious document[345] which gives sufficient data for the purpose. It is contained in a MS. volume, now the property of the Dean and Chapter of Canterbury, composed of several quires of paper stitched into a parchment cover. They once belonged to, and were probably written by, Brother William Ingram, who was custos martirii in 1503; and in June 1511 was promoted to the office of Pitancer. The accounts and memoranda in the book are of a very miscellaneous character. The part which concerns the library consists of a note of the books which were repaired in 1508. This is headed:

The writer goes round the room, beginning at the west end. He proceeds along the north side, and returns along the south side, to the point whence he started, enumerating on his way the bookcases and their shelves, the volumes removed, and, occasionally, a note of the repairs required. For my present purpose I will content myself with his account of a single bookcase, the first on the list. The writer begins thus: "From the upper shelf on the east side in the first seat (de superiori textu[347] ex orienti parte in prima (sic) sedile)." Three volumes are enumerated. "From the lower shelf (de inferiori textu)," two volumes. "From the upper shelf on the other side of the same seat (de superiori textu ex altera parte eiusdem sedilis)," seven volumes. "From the lower shelf (de inferiori textu)," five volumes. In this way eight seats, i.e. bookcases, are gone through on this side of the room. The writer next turns his attention to the south side, and goes through eight more seats, beginning with: "From the east side of the upper shelf on the south side (de textu superiori ex parte australi incipiendo. In parte orientali)." The examination was evidently thorough, and, as the same number of seats is enumerated for each side of the room, we may, I think, safely conclude that all were examined, and that the whole number in the library was sixteen.

The passages I have quoted shew that each of these bookcases had an upper and lower shelf on each side, on which the books stood, so as to be conveniently consulted by readers on each side; the books were chained; and, in consequence, there must have been a desk, presumably below the shelves on each side; and a seat for the reader. I have embodied these requirements in the accompanying sketch or diagram (fig. 89), which indicates a bookcase of the same type as those at Corpus Christi College, Oxford. If we may suppose that each of these cases was two feet wide and eight feet long like those at Merton College, we can accommodate eight cases on each side of the room (fig. 88), with the same interval between each pair as at that college.

Let us now consider whether the library as thus arranged would have had sufficient shelf-room. Each bookcase being 8 feet long would contain 32 feet of shelving, and the 16 cases a total of 512 feet. The catalogue made in the time of Prior Henry of Eastry (1285-1331) enumerates 1850 volumes[348]. If we allow two feet and a half for every ten of these we shall require 462½ feet; or in other words we can arrange the whole collection in 14 stalls, leaving 2 over for the additions which must have been made in the interval between the middle of the 14th century and the date of Brother Ingram's researches.

If the sketch here given of the probable aspect of the library at Christ Church, Canterbury, be compared with the view of the library at Merton College, Oxford (fig. 82), a fairly correct idea of a great conventual library will be obtained. A very slight effort of imagination is needed to make the necessary changes in the shelves, and to replace academic students by Benedictine monks. Then, if we conceive the shelves to be loaded with manuscripts, many of which were written in the early days of the English Church, we shall be able to realise the feelings of Leland on entering the library at Glastonbury:

I propose in the next place to print a translation of the Introduction to the catalogue[350] of the Benedictine Priory of S. Martin at Dover, which was a cell to Canterbury made in 1389 by John Whytfeld. This catalogue does not indicate the stall-system; in fact I am at a loss to define the precise system which it does indicate. I print it in this place on account of its internal interest, and the evidence which it affords of the care taken in the last quarter of the fourteenth century to make books easily accessible to scholars.

The meaning of the word "distinction" is the principal difficulty in the way of understanding the above description. I thought at first that it denoted merely difference of subject, and that gradus, as in the catalogue of Queens' College, Cambridge, was a side of a lectern. But the statement that the grades are numbered "from the bottom and proceed upwards" can hardly be reconciled with any arrangement of lecterns. Distinctio probably denotes a bookcase or press, divided into 7 grades, and probably placed against the wall, the word gradus here meaning a flat shelf, instead of one set at an angle as in former instances. If this explanation be correct we have here a very early instance of shelves in such a position.

My second example of a monastic library fitted up according to the stall-system is the library at Clairvaux. As I have already printed a full description of it[353], I need not do more in this place than translate the passage referring to the fittings:

As there were so many as 48 bookcases, that is, 24 on each side, the bookcases were evidently spaced without reference to the lateral windows, which were probably raised high above the floor.

The catalogue, from which I have already quoted the verses commemorating the building of the library, contains much useful information respecting the arrangement of the books. The verses are succeeded by the following introductory note:

The most important passage in the above note may be thus translated:

The catalogue as well as the description makes it perfectly clear that each desk, that is to say, each bookcase, had four shelves; and further, as the authors of the Voyage LittÉraire (1708) mention chains[355], it may be concluded that there were desks, and seats for readers, between each pair of bookcases. If we place two shelves on each side of the case we get a piece of furniture precisely similar to that in use at Canterbury.