bound in full morocco 'extra,' at a cost of, perhaps, twenty shillings apiece, the work would have realised, probably, seven or eight pounds. But by good judgment (and, in our opinion at least, good taste) his expenditure would not exceed fifteen shillings for the three, his profit being four times as great. Not long ago two copies of the first edition of Keats' Endymion' appeared at an auction-sale in London. Both were 'uncut,' but one was in the original form in which it issued from the press, the other was bound in morocco. The former realised £41, the latter £17, 5s. Dictum sapienti sat est.

Old books, by which we intend sixteenth and early seventeenth century volumes, are always best left alone as regards the binding. If they be at all dilapidated, it is as well to have a case made for them which can be lettered on the back, and they can then stand upon the shelf among one's other books.1 Nothing is more unseemly and incongruous than an ancient volume in a modern cover, and, try as the most skilful binder may, it is impossible to imitate an ancient binding so closely as to deceive the eye even momentarily. Do not seek to make them presentable by patching and

1 With regard to these cases, the collector will use his own judgment as to whether they be of the slip-in ' variety, by which means the binding is rubbed every time that he withdraws and inserts his volume; whether such cases be lined with velvet, and roomy enough to obviate this friction; or whether they shall open with a flap at the side.

repairing, unless they be too far gone for their value to be of any consideration.

In the case of early-printed books and works of great rarity, never, upon any account, tamper with your copy or seek to improve it in any way. Not only, as we have said, is it quite impossible to impart a contemporary appearance to a fifteenthcentury book however famous and skilful the binder, but age leaves its mark upon the constitutions of books as surely as it does upon mankind. No volume of that age will stand the handling of a casual reader, still less the pulling, patting, and pressing that re-sewing and re-covering necessitates, however gently such processes be carried out.

There is a terrible story (we hope it is untrue) told of a certain peer who decided to send to the auction-room the six or seven Caxtons which had descended to him with a noble library from his ancestors. As, however, the volumes were bound in fifteenth-century sheepskin (probably in Caxton's house) he thought that their appearance would be rendered rather more attractive if they were rebound first of all. So he sent them forthwith to the local binder; and on their return, now gorgeously clothed in 'calf gilt extra' (à la school prize), he despatched them to the London saleroom. The result may be imagined. His foolishness must have robbed him of a sum running well into four figures!

There is another point also to be considered, and that is the pedigree of a volume. The solitary impression of a binder's tool upon a fragment of binding may identify a volume and its previous owners. Some years ago the writer purchased an ancient folio without title-page and colophon, bound in tattered fragments of ancient calf covering stout oak boards. There was, apparently, nothing to indicate when, where, or by whom the volume was printed or bound, or whence it came. But from a certain peculiarity in the type (which he noticed when studying the early printers of Nürnberg) he now knows the name of the printer and the town in which he plied his trade; while from a certain woodcut which that printer used also in two other dated works only, both printed the same year, he discovered when the volume in all probability was printed.

A scrutiny of the remains of the binding revealed the blind impressions of four different stamps. As these occur frequently in conjunction upon the bindings executed by the monks at a certain monastery in Germany in the sixteenth century, there is little difficulty in assigning a provenance to the volume. Furthermore the initial H in a heart-shaped impression identifies the binder as a monk whose initials H.G. (on two heart-shaped tools) are of frequent occurrence on contemporary volumes at that time in the possession of the monastery.

Needless to say, it has not been rebound. The tattered pieces of skin have been carefully pasted down, and a case-lettered on the back-now contains the book upon his shelf. 1

In the case, however, of more recent books bound in tattered or perished calf, books of which one may obtain duplicates at any time, except they be works of extreme rarity and value there is no reason why they should not be re-bound. Even here, however, the collector must tread warily; for should he send his copy of Tim Bobbin's Lancashire dialogue of Tummus and Meary to the binders with brief instruction that

1 If you are interested in the pedigrees of your volumes (by which we mean the identification of their previous owners) you will find M. Guigard's 'Nouvel Armorial du Bibliophile,' octavo, Paris, 1890, useful where armorial bindings are concerned. It is an interesting volume, and appeared first of all in four parts (large octavo, Paris), between 1870 and 1872. Each part of this edition should have a separate title-page and half-title. An additional titlepage was issued with Part I., on which appears TOME I., and at the end of the Table Héraldique issued with Part IV., appears FIN DU TOME II. ET DERNIER. There are cuts of every coat of arms identified, but these are almost entirely French. Mr. Cyril Davenport's English Heraldic Book-stamps' was published in large octavo, in 1909. For early book-plates you must consult the numerous works upon this subject that have appeared in recent years. Mr. W. J. Hardy's 'Book-Plates' was issued first in 1893, but a second edition was published in 1897. There is also 'A Bibliography of Book-Plates,' by Messrs. Fincham and Brown, in which the plates are arranged chronologically. The Ex-Libris Society issues a journal, and there are numerous other volumes upon this subject, which you will find mentioned in Mr. Courtney's 'Register of National Bibliography.'

it is to be bound in full morocco, it may be returned to him in all the splendour of a sixteenthcentury Florentine binding.

With regard to books published in cardboard covers with paper backs and paper labels, what is to be done with these when the backs are dirty or torn off, the labels of some volumes missing? Must they be re-bound in leather or cloth? Not necessarily, and for our part we maintain that the delightful ease which one experiences in handling them when reading the early editions of Byron, Scott, or Irving, and those writers who flourished in the first few decades of the nineteenth century when books were commonly issued in this form, is sufficient excuse for retaining them in their original shape. Such volumes may easily be made presentable at the cost of a little time and trouble, as we shall presently show.

An appearance of antiquity is never a desideratum to the honest book-collector. We say 'honest' advisedly, for there have been-and doubtless are -persons so misguided as to stoop to the fabrication of certain small and excessively valuable books. To such, an appearance of age is no doubt indispensable in their wares. But these are torments which afflict the wealthy only; and for this we at least are sincerely thankful.

There is no doubt, however, that in the collection of many things antiquity in appearance is desirable: witness the modern fabrication of