CHAPTER II.
ROLLS, BOOKS AND BOOKBINDINGS.

Rolls of papyrus and vellum—Quaternions—The sewing of books—Headbands—The rounding and backing of books—Mediæval books—Irish cumdachs—Byzantine bindings—Oriental books—Modern methods of sewing and binding.

The length of ancient rolls of vellum has often puzzled bibliophiles. Rolls of narrow breadth are found 16 or 17 feet or more in length. A learned scholar on being asked how he accounted for this extraordinary length was quite puzzled, never having realised that the roll was in one piece. The neck of the giraffe seemed the only possible solution.

The writer however consulted a clever leather worker and gave him a skin measuring about 3 feet by 2, suggesting certain ways of cutting it. He produced eventually, by wetting, pulling and pinning, a beautiful roll of nearly 4 inches in breadth and 16 feet 9 inches in length.

Mediæval leather workers were no doubt more skilled and practised in this particular art than anyone now is, and the experiment showed that there is really nothing out of the way in the very long rolls which at first sight seem so surprising.

Rolls were written upon in three ways. In the oldest rolls the usual way was to write lines across the breadth of the roll, which was held upright before the reader, and unrolled from the top downwards.

A rare form of writing upon rolls is that found in the prayers written on the strips which are rolled up inside Buddhist prayer wheels. Such prayers, however, are never read, but are counted as being read through on each revolution of the wheel. Each line of manuscript runs along the entire length of the roll, which is unrolled sideways. The prayer wheels vary immensely in size, the best known being the little hand ones chiefly used in Thibet; they are variously ornamented.

The form of writing in rolls that is of most interest so far as books is concerned is a modification of the Thibetan form. Instead of each line running the whole length of the roll, a space limit is now fixed, and the lines of writing follow under each other, so that the page form is at once apparent. This form, a late one, can be seen in the case of the Jewish scrolls of the law. The roll is unrolled sideways, and the rollers at each end are often very handsomely decorated.

But writing of this last kind on rolls has suggested another arrangement in which the reading is more easy, and the re-rolling of the roll itself avoided.

It will be seen that a blank space is left between each of the written “pages.” Now if the vellum, bark, or paper be folded across these vacant spaces, one after the other, backwards and forwards, like accordion pleating, we shall find that we get a form of book well known in the East and also among primitive nations.

Curious examples of such converted rolls can be seen in most museums, and they are generally kept flat by means of two boards front and back, but not otherwise fastened.

The Chinese and Japanese have taken this particular form of evolution from the roll to the book a step further, and by help of the ancient device of “stabbing” the flattened roll along one of its sides, they produce a form called an “Orihon,” easy to consult, strong, and the blank back of the roll so hidden up that its existence is frequently not realised. But if some of the leaves of an Orihon are cut, its real structure becomes at once evident, and a book will be produced with letterpress and blank paper alternately in pairs. A similar kind of alternation shows now and then in the case of MS. rolls that have been cut up, but they are oftener arranged letterpress and blank alternately.

Without realising it we still preserve this blank and letterpress sequence, found in the converted roll, in official and legal manuscripts as well as in those intended for the printer.

When the printing press took the place of the scribe, the blank leaves had no further raison d’être, so they dropped out for good.

Following the rule that the forms of binding have always followed tolerably closely the forms of the manuscript they have covered, we find that rolls were kept in cylindrical boxes, called “scrinia.” Each roll was usually provided with a little tag, so that if there were several of them in one box they could easily be distinguished. The same sort of tags are used to-day in the case of rolled maps kept on shelves.

Until a late period the term parchment must be understood to mean vellum. Now we call sheep skin “parchment” and calf skin “vellum,” and they are prepared in the same way with lime, so that not being tanned they are not strictly “leather.” The finest vellum is prepared from the very youngest and smallest calves, and it is the most beautiful and suitable material for writing or printing upon that has ever yet been found. The surface is singularly even and offers little or no resistance to a pen, so that every sort of handwriting, square or round, is put upon it with equal ease. Vellum has one fault alone, particularly when bound in book form as distinct from a roll, and this is that the edges are apt to cockle, and by so doing they not only make the pages look ugly, but they also admit the dust. Often and often magnificent vellum books, especially at the top, show large vandykes of dust-stained spaces due to this cockling, and all such books should be provided with a close fitting cardboard cap, to be kept upon them whenever they are not in use.

Parchment and vellum were both well-known to the ancients, but their value as materials for writing upon does not seem to have been fully acknowledged until the second century B.C., until which period papyrus had held undisputed pre-eminence for that purpose. At that time for some reason the supply of papyrus from Egypt ran short, and Eumenes II., King of Pergamum, successfully introduced parchment in its stead. Parchment is so called because it was first produced at Pergamum.

Until about the fifth century A.D., vellum MSS. were in the roll form, but then came a change to the book form as we now know it. This change was probably due to the fact that stabbed binding, the only sort then used, was not suited to vellum. The few papyri that exist in book form were stabbed, that is to say, the rectangular pages were kept in position by a binding cord laced through holes pierced sideways right through the entire thickness of the back of the book. The marks of these holes can often be seen along the inner margins of ancient papyri, and they also show in many instances of rebound copies of our early English printed books. To-day plenty of examples of this form of sewing can be seen in the Chinese and Japanese books, “Orihons,” which are really links between the roll and the book form. It is also largely used for thin books of little value, and a modification of it can be seen in numbers of magazines, books of advertisements, and the like, which are kept together by abominable little clamps of wire attached on the same principle. When such books have to be properly bound the little clamps have to be carefully removed, and it is generally found that they have made an indelible stain of rust on the paper, even if they have not also torn it considerably.

In mediæval libraries or monasteries when a book was to be made, vellum leaves were cut into the required size and folded once across the middle. The folded leaves were then fitted inside each other in groups of four (quaternions) or five (quinternions), or whatever other number seemed good to the bookmaker. The leaves were then marked in some slight way so that their order might not be lost, and sent to the scribe to be written upon.

The books count as folios because each original sheet only forms two pages; the fact of their being arranged in groups is accidental, and does not theoretically alter the size of the book. For this the original skin of vellum would have to be folded with certain further divisions, and in the case of early manuscripts this was never done.

Vellum shows a different surface on the flesh side to that on the hair side, and the scribe usually made his rule marks with a blunt style to guide his writing on the hair side, so that on this side the ruled lines are slightly indented, whereas on the flesh side they show as little ridges.

This point is apparently trivial, but if, as seems likely, both Greek and Latin scribes were really very particular and consistent in the way they alternated the two sides of the vellum, then the matter becomes one of much critical importance. Indeed, it has already been of the utmost value in deciding questions as to whether new pages had been added or not to an old book.

When the scribe had made his rulings he then doubled up his sheets once, and arranged them as he desired with regard to the hair or flesh sides. When a section of four sheets were fitted into each other, so that when pressed together they made a solid gathering, such a gathering is called a “Quaternion,” from which term we derive our word “Quire.” This is a normal state of things, but it is obvious that abnormal arrangements might easily be made, from the insertion of a single leaf to that of an entirely additional section.

Now the question arose of how best to fasten the quaternions together, not only as to themselves but also as to the other quaternions, which together formed the entire book.

As to the quaternion itself, it must have been evident at once that a stitch of thread or fibre run from the innermost fold right through to the outermost would hold the leaves firmly together. It is likely enough that this was done separately at first, and then the binder would have looked at a small heap of such gatherings wondering how best to keep them together, and it would soon occur to any constructive mind to knot the loose ends of the threads together, or else to supply a supplementary cord or cords laid at right angles to the back of the sections on which the projecting ends of the threads might be tied or sewn.

This, in fact, was done, and very shortly the best way of fastening the sections on to such cords or threads was hit upon—a method indeed that we have never bettered, and which can still be found in the work of many of our best modern binders.

At first this fastening together of the sections of a book was no doubt done uncomfortably and roughly by hand, but it soon became evident that some simple device in the form of a skeleton frame might be contrived which would render the operations of sewing and binding much easier. Not only easier to execute but also giving a more regular and workmanlike result.

The earliest known sewing frames were the same as are used now. There are two strong columns of wood fixed on a broad platform, with a slot between their bases. From capital to capital extends a bar, and the strips of leather, vellum, or hempen cords which are to form the bands of the book are looped upon it, and are kept taut by means of metal keys attached to the other end, which lock into the slot at the bottom. The bands can be quite easily adjusted to any space the worker desires.

The book, ready for sewing and in proper order of sections, is laid near the worker’s hand, and he, or she, takes it up by sections, one at a time. The section to be worked upon is laid downwards on the little platform, with its back close against the bands, and the worker’s left hand keeps the section open in the middle, while with the right hand a thread is drawn through the back fold, from the inside to the outside, round the band, and then back from the outside to the inside, and so on until all the bands have been caught round. Then to end up, the thread is passed through near the extreme end of the fold and knotted, forming what is called a “kettle stitch,” and from this point the whole operation is repeated, backwards, with the next section. It sounds complicated, but is not so really, and several sections could have been sewn together in the time that it has taken me to describe the sewing of one of them.

Now comes an interesting point, and that is the exact way in which the threads are passed over the bands. We will first see that if the thread is drawn through the back fold of the section, round a hand and back again through another hole, that there is a weak construction, inasmuch as the thread will have a strong tendency to cut through the paper at a and b, because there is in each case a side strain. It therefore seemed necessary that the thread should make its return journey by the same hole through which it emerged, but if simply done, this did not mend matters much, as the strain still tended to pull this hole open wider. A line of exit and entrance without any side pull was needed, and this was found by means of a very clever device. A broad band of leather or vellum was cut through lengthways, leaving a solid piece at each end.

Now when the sewer came to this band he ran the thread straight through the slot, then brought it round the entire band, under it, between it and the back of the book, and down again, back through the slot, and in at the same point from which it emerged. The result of this is that there is no strain at all on the needle hole through which the thread passes, as the pull is quite straight both coming and going. This is undoubtedly the best form of sewing a book on bands, and a book so treated is said to be sewn on “double bands.” But our recent work is not always true, although from the outside it appears correct, because in numbers of cases such double bands are simply glued on the outside of the back, the real sewing, of a very inferior kind, possibly even done by machinery, being hidden underneath the leather. Few great binders except Charles Lewis have ever used sham bands.

Flexible sewn books can be had now if wanted, but the sewing on the bands is not quite the same as the fine mediæval double bands I have described. It is, however, practically nearly as good, and the bands themselves can be made smaller.

The modern method is to bring the needle and thread through the back of the section as usual, then give it one turn onwards over the band and back again through the same hole. It will be seen that this is a thoroughly sound principle, and brings no strain upon the back of the section.

The ends of the bands of limp vellum books have always been treated in the same way; they are drawn straight through the vellum at the joint and then back again and fastened inside by means of the end paper.

The manner of drawing the bands of a “limp” vellum bound book through the limp vellum cover is of much interest, and it survives in many instances where boards are used.

In principle it is the same process as is described further on with regard to boarded books, but in the case of the limp vellum bindings the ends of the bands are normally visible for a short length, but in the case of boarded books they are always covered up with the exception of some vellum-covered boarded books in which the limp vellum peculiarity is preserved.

Another small point of interest about the old limp vellum bindings is that the head and tail bands are made of the same sort of thin vellum strip as the main bands and carefully drawn diagonally through the vellum at the corners and fastened inside. Dealt with in this way the head band becomes of real structural value, much of which is lost if it is cut off short as is done in the large majority of cases.

William Morris liked limp vellum bindings and often used them, but without head bands. Instead of the short strips of vellum used for bands in old books Morris used specially prepared silk tapes, and brought them through the vellum at the joint in the old-fashioned way. Instead of being cut short and pasted down as the old limp vellum bands were, Morris continued his tapes and brought them out again near the front edge of the vellum, where they could be tied. This is an excellent arrangement and keeps such a book together in an admirable and effective way.

Vellum bindings required flat backs because the material would not yield sufficiently to be tucked round the projecting bands of a normal flexible bound book, in which the leather back is firmly fixed over the bands. Nevertheless in several instances old vellum books, in boards, which have been sewn on raised bands show traces of these bands in low relief across the back. In such cases it is common to find that some padding has been put in the spaces between the bands so as to level the back up. Books treated in this way are usually stiff to open and uncomfortable to consult.

The flat back, which was necessary for the same reason in the case of books bound at a later date in velvet, cloth, silk or canvas, necessitated some modification of the thick projecting bands of the ordinary book, and the requisite flatness was attained by using strips of vellum or tape instead of cords of hemp. Then it was found that it was not necessary to fasten the back of the cover to the back of the sections, so the “hollow back” came into existence.

The majority of modern books are bound with hollow backs; it can be recognised by opening a book to the full and seeing if the back is separate, and it has one real virtue as well as several vices. If a book should happen to be printed too far back, a hollow back binding will enable it to “throw up,” and show the printing right down to the inner edge of the paper, whereas a flexible binding always tends to open less freely, especially in the middle.

For books that are likely to be much used hollow backs are unadvisable as they are sooner worn out, but for small light books there is no doubt much to be said in their favour.

Among English binders of note I think Charles Lewis was the first to use hollow backs extensively. But for fine books there is no doubt that the old-fashioned flexible sewing on raised bands is the best in the long run. Books bound with hollow backs often have the bands “sawn in,” that is to say, a trench is cut for each band across the back of the sections. In these trenches the bands are laid, and the sewing is of a simpler and quicker sort than it is when the thread is brought round each band as it is in the flexible style. Moreover, there is a weak point where the thread touches the edge of the saw cut, and at this place the paper is always apt to give. It is obvious that to cut away paper from the back of a section must always be not only a barbarism but also structurally wrong. Such a method of sewing a book can only be excused on the score of cheapness, and it may be urged that in this case it does not matter.

Towards the end of the fifteenth century it was found that the extreme tops and bottoms of the backs of the sections of bound books not only looked untidy and unfinished, but also that they tended to gape, in fact they were weak points both structurally and artistically. When it was once realised it was soon rectified, and a small additional band was made of a strip of leather or vellum, to fit along the outer edge of the top and bottom at the back. This slip was then sewn on by means of thread and button-hole stitching, being caught in at intervals by a long stitch drawn through the centre of one or other of the sections of the book. The loose ends of the headband itself were drawn in to the boards, forming in fact an additional band. The headband is a point in the forwarding of a book which has not received much notice as yet, but it is of some importance, as there is no doubt that much attention has been paid by binders to the ornamentation of headbands from mediæval times until now, and it is the first point in the binding of a book in which ornament is considered.

Some of the early headbands are sewn over with strips of soft leather, and at a later time they are cut in distinctive ways, flat or round, and sewn over with silks of particular colours or combinations of colours.

Of late years the vice of cutting off the ends of the headband has come into being; these ends ought to be drawn properly into the boards, as Mr. Douglas Cockerell has indeed done for me in several instances. But no words can adequately condemn the miserable ready-made coloured slips which are often found simply stuck on in the proper place.

Vellum books were the first to be properly bound. The papyri which were stabbed are so rare that they may be passed by. As I have already said, vellum is apt to be curly; for this reason the boards used for binding in early days were made of thick wood, the heaviness of which, even when unaided by clasps, tended to keep the vellum flat.

Beech wood was largely used for these old covers, and from the German word “Buch,” meaning beech, we derive our word “book.” Beechen boards were light, decorative, and very carefully dried and seasoned. It is remarkable how flat such old boards are, and they were no doubt very highly valued as they often have upon them the stamp of the monastery in which they were used.

The edges of the boards are sometimes bevelled, from the upper side in English or French books, and from the under side in German books—but such a distinction must not be taken as invariable. Boards often show signs of having been used more than once, and it is rarely that any decoration shows on the wood. In a few cases of German books may be found outline drawings of an heraldic nature.

The first bound books were made with flat backs, and the boards fitted close upon the outer sheets of vellum, papyrus or paper. In this formation, however, it was found that there was a strong liability for the back to fall in and the foredge to project outwards.

As early as the fifteenth century, in the case of printed books on paper, this fault of the back falling in led some few binders to neutralise it by giving the back of the book a rounded form by means of hammering, and this quite prevented the falling in of the back. The exact extent of the rounding can easily be seen by looking at the front edge of a book, because the curves of the back and front correspond. The boards, however, remained in their first position, flat on the outer sheets.

But another trouble was apparent in both these cases, namely, that when such books were opened, the joint between the boards and the back showed a tendency to pull up the few pages next adjacent. In time these pages became torn and injured, and constructively there was something wrong with the principle of attachment.

Paper is soft, and when a “rounded” book was fitted with hard boards and strongly pressed there would be a certain tendency for the boards to sink into the mass of the paper and to throw up a small ridge along the edge of the back. Such a small accidental ridge is often found on old paper books.

But there is no doubt that the actual process of making an intentional groove for the boards to fit in was practised by a few fifteenth century binders in England. This groove is made by an extension of the process of rounding the back, and it is produced by hammering the back over two hard boards carefully placed in the proper position. The shape of a back thus treated is theoretically as shown in Fig. 44, and it will be seen that the actual joint now falls away from the body of the book and is removed to the artificial line along the outer edge of the groove, and from this line the projecting bands are drawn in to the boards. If this operation of “backing” is properly done it is almost impossible for the back of a book to fall in. It will always open easily and return to its original form, and if the bands are properly attached to the boards, the latter will never fall off.

Although the principle of backing was known at the early time mentioned, it was not universally understood and practised until quite recent times. Now, however, it is fully recognised as one of the most important processes in the binding of books, especially large ones.

We have seen that the book sewn in leather bands has the ends of these bands left loose, projecting in the case of a large book some two or three inches. Holes were now carefully cut along the back edge of the board to fit the ends of the bands, sloping upwards, and some little way in other slots were cut from the upper surface of the board to meet them. Into these holes the ends of the bands were drawn, and when in proper position they were pegged down with one or more small wooden pegs. Sometimes the bands were drawn right through the boards and fastened inside. Numbers and numbers of instances of this work exist and are quite sound to-day. But such books do not open satisfactorily, as there is a disagreeable pull upon the outer sections when the book is opened. In fact, the junction between the bands and the boards is not scientifically correct, because the backs are not rounded.

The vulnerable part of the binding will now be seen to be the soft threads which cover the bands where they adjoin the back of the sections of the book, and to protect these delicate threads a strip of leather was cut, damped and pressed over the bands so as to fit quite closely, and fastened on with glue, projecting a short way over on to the board itself so as to cover up the holes used for the bands. This is called a “half-binding.” On the leather of such half-bindings there is usually some blind tooling, lines or rolls or even small cameo stamps.

Mediæval bindings are commonly provided with clasps. The original reason of this was to help to keep the vellum leaves flat, but of course artistic binders saw that clasps might be made very ornamental, and so many of them are. They have survived as ornamental adjuncts to a binding until the present day, although there is no necessity for them.

Bosses in the centre and at the corners of mediæval bindings were of structural use, as they protected the actual boards of the book from wear. In early days books were kept upon their sides and probably had flat boards between each volume. So the bosses took the wear and preserved the books. Many of the mediæval bosses and corners have ornamental settings; they are generally of brass, but sometimes of silver or wood. The titles of such books were sometimes written on the front edges and sometimes on a slip of paper fastened under horn on the upper board, and in the late fifteenth century they were sometimes lettered in gold or blind, also on the boards.

It happened that the art of Byzantium was in its prime just about the same time that vellum began to be arranged in sections—in book form—in contradistinction to the roll form. So we find some of the finest examples of Byzantine art on bookbindings. Of course it must be understood that these richly worked covers are only ornamentally added to the true binding underneath, and the actual work of the binding is more or less as I have already described it. But the decorative art in these cases is so predominant that they are generically known as “Byzantine” bindings.

The large majority of the known specimens of this kind are no longer in company with their original manuscripts. Many of them are preserved simply as loose covers, and alien manuscripts have been inserted in others. I expect that there are still considerable numbers of these bindings preserved in remote churches and monasteries in Central Europe, and some day perhaps some traveller with sufficient knowledge and time at his command will make search for them. But for the present fine and genuine examples of Byzantine bindings are of great rarity. It is said that there are not more than about three hundred of them known. The life history of many of these is well known, and so are many of the additions and alterations to which they have been subjected during their long life. Luckily some of the finest are now safely housed in our London museums, and others are, or are likely to be, safe in a great local museum.

To be able to judge whether these bindings are genuine, a librarian must be an antiquary, a goldsmith and silversmith, skilled both in repoussé work, engraving and cast work, a gem cutter, an enameller, a skilled bookbinder, and an artist, and if he fails in one of these knowledges his judgment cannot be relied on. An ordinary art critic who possesses “Flaire” can pick up a certain amount of knowledge by reason of having made many mistakes and profited by them, and if he has been fortunately able to associate himself only with true and known fine examples for a long time, he may eventually be a good guesser. But no true judgment can be given without actual technical knowledge.

The result of all this is, that collectors of such bindings have been freely victimised. It is not really difficult to copy or to imitate a Byzantine binding—it is not nearly so difficult as to counterfeit a fine Le Gascon or a fine Roger Payne. It does not so much concern the binder as the metal worker, and in metal a little skill goes a long way, and so it does in the cloisonne enamel work which alone is found in these bindings. The present market price of a very fine Byzantine binding—a known one—will run into five figures, so it is little to be wondered at that it pays the clever faussaires of the Continent to direct their attention to covers of this kind. They need not even take the trouble to provide manuscripts for the inside; loose covers will sell with equal readiness.

Ivory carving played an important part on early bindings. Apart from diptychs, the earliest known decorative binding consists of plaques of ivory carved with Biblical scenes, and having in the centre a lamb within a wreath of cloisonne work, with inlays of coloured glass. It is now kept in the cathedral treasury at Milan, and is without its original inside; it is attributed to the fifth century.

Early Irish manuscripts were usually bound in quite simple limp leather covers, but these were kept in book boxes or “Cumdachs,” all of which still existing are of the greatest interest.

One of the finest is that of the Stowe Missal, dating from about the eleventh century. It belonged originally to the monastery of Lorrha in Tipperary, whence it was taken to the Irish monastery of Ratisbon. In 1784 it was found by Mr. John Grace in Austria, and afterwards belonged to the Stowe Library, and Lord Ashburnham, from whom it was purchased by the nation in 1883 with the rest of his library, and then deposited in the museum of the Royal Irish Academy.

The lid is ornamented with a large cross set with pearls and metal bosses, having in the centre a large oval crystal over a pink ground, and enclosed with recent enamels, no doubt replacing old ones. At each end of the cross is a large jewel. The spaces between the arms of the cross contain silver gilt plates engraved with figures of Saints, and on the silver edge plates is an inscription in Irish inviting the reader to “Pray for Dunchad descendant of Taccan of the family of Cluan who made this.” This may therefore be called the earliest signed binding in the world, if a cumdach can be considered a binding. No doubt much of the work now on this lid, as well as some of the jewels and their settings, have been added since the original work was made.

The base of the box has a similar ornamentation upon it, but the jewels and enamels are all gone, the spaces between the arms of the cross being filled with silver plates cut into an open pattern set over bronze gilt.

The sides of the box are much destroyed, but they have fortunately not been restored. The ornamentation consists of enamelled bosses or curious castings in the centre of each side, flanked by panels of open designs cut out of thin silver over gilt, and corner ornaments of cast work. There are small bosses of blue enamel.

The remarkable open work over gilt bronze occurs again in the base of the cumdach containing Molaise gospels, it also belongs to the eleventh century, and this sort of ornamentation seems to be characteristic of Irish work of that time. The lid of the box is ornamented in a modification of the same idea, but in this case the bronze showing through the silver is beautifully worked. The main idea is a cross, and between the arms are the evangelistic emblems, and with their names, Leo, Aquila, Homo, Taurus. An inscription asks us to “Pray for Cenn, for the successor of Molaise for whom this case was made by Gillabaithain the artisan.” It is possible that this may be a little earlier than the Stowe cumdach.

A few more cumdachs are known, about ten of them altogether, and others are recorded only.

There is that of Dimma’s book, of the Cathach of the O’Donnells, and several others, some of which are quite plain.

Other book cases were made of beautifully worked leather. The satchel of the Book of Armagh is one of these, and leather cases or “Forels” were made of cuirbouilli in the fifteenth and sixteenth centuries, particularly in Italy. These forels were in the shape now known as “slip off” cases, and they generally had attachments for a strap. They are always charmingly chased, and often bear heraldic ornamentation.

Cuirbouilli was calf or hide, worked with knife, style, or hammer, and then probably boiled in oil, but the exact process that was used is now unknown. It is very strong, and its value was known in England, where it was used to make wrist guards for archery, but as far as I know it was never used here for bindings or covers for bindings. I have heard the calf or sheep bindings of the early sixteenth century, bearing panel stamps, described as cuirbouilli, but they are not so.

The earliest cover still containing its original manuscript is probably that given with other treasure to Theodelinda, Queen of the Lombards, by Pope Gregory the Great, in the seventh century. It is now kept in the cathedral treasury at Monza.

The manuscript is described in a contemporary document as being “theca persica inclusam,” and, so far, its cover resembles the Irish cumdachs. The case is of gold, ornamented with a large cross, outlined with lines of small flat garnets, with sapphires and emeralds, set in cloisons. The body of the cross is richly jewelled with cabochon cut stones and pearls, and at the extremity of each arm is a curious glass bead of millefiori glass, green with a little red flower in the centre.

Between the arms of the cross are four gamma shaped ornaments, each set with small flat jewels. Across the sides are bands of gold on which are engraved the words De Donis Dei offerit Theodelenda reg. gloriosissima sancto Joanni Baptiste in Basilica quem ipsa fundavit in Modicia prope palatium suum. A handsome border with flat garnets enclosed the whole, and the four cameos set near the gammas are a recent addition. Byzantine art is generally considered as a decadent form of Roman art, but in this particular instance there is much Greek feeling, and it is in all respects, excepting perhaps the inscription, a magnificent piece of work.

The binding of the Gospels of Lindau has now two sides of different dates, each of which has been repaired and added to in recent times. The earlier of the two, far the finer, is probably contemporary with the manuscript, having been made about the later half of the ninth century, as the Abbey of the Noble Canonesses at Lindau was founded by the Emperor Lewis the Pious in A.D. 834. A large golden cross pattée is the chief motive of this beautiful piece of work, and it is ornamented with rare enamels showing bust figures wearing stoles, and jewels in great variety, the borderings being inlaid with small flat pieces of garnet. The spaces between the arms of the cross are filled with bronze plaques elaborately ornamented with Celtic interlacings in chiselled work. The workmanship is probably Irish, and was most likely done abroad, possibly enough at Lindau itself, by some Irish artists who had emigrated. Irish jewellers and enamellers of this early time were justly celebrated, and their services were secured whenever possible.

The later side of the binding is one of the finest existing specimens of Carlovingian art. It is radically different from the earlier side and was probably added about the eleventh or twelfth century.