Palaeography

Palaeography (Gr. iraXauos, ancient, and 7pa4av, to write), the science of ancient handwriting acquired from study of surviving examples. While epigraphy is the science which deals with inscriptions (q.v.) engraved on stone or metal or other enduring material as memorials for future ages, palaeography takes cognisance of writings of a literary, economic, or legal nature written generally with stile, reed or pen, on tablets, rolls or codices. The boundary, however, between the two sciences is not always to be exactly defined. The fact that an inscription occurs upon a hard material in a fixed position does not necessarily bring it under the head of epigraphy. Such specimens of writing as the graffiti or wall-scribblings of Pompeii and ancient Rome belong as much to the one science as to the other; for they neither occupy the position of inscriptions set up with special design as epigraphical monuments, nor are they the movable written documents with which we connect the idea of palaeography. But such exceptions only slightly affect the broad distinction just specified.

The scope of this article is to trace the history of Greek and Latin palaeography from the earliest written documents in those languages which have survived. In Greek palaeography we have a subject which is self-contained. The Greek character, in its pure form, was used for one language only; but the universal study of that language throughout Europe and the wide diffusion of its literature have been the cause of the accumulation of Greek MSS. in every centre of learning. The field of Latin palaeography is much wider, for the Roman alphabet has made its way into every country of western Europe, and the study of its various developments and changes is essential for a proper understanding of the character which we write.

Handwriting, like every other art, has its different phases of growth, perfection and decay. A particular form of writing is gradually developed, then takes a finished or calligraphic style and becomes the hand of its period, then deteriorates, breaks up and disappears, or only drags on an artificial existence, being meanwhile superseded by another style which, either developed from the older hand or introduced independently, runs the same course, and in its turn is displaced by a younger rival. Thus in the history of Greek writing we see the literary uncial hand passing from early forms into the calligraphic stage, and then driven out by the minuscule, which again goes through a series of important changes. In Latin, the literary capital and uncial hands give place to the smaller character; and this, after running its course and developing national characteristics in the different countries of the West, deteriorates and is superseded almost universally by the Italian hand of the Renaissance.

Bearing in mind these natural changes, it is evident that a style of writing, once developed, is best at the period when it is in general use, and that the oldest examples of that period are the simplest, in which vigour and naturalness of handwriting are predominant. On the other hand, the fine execution of a MS. after the best period of the style has passed cannot conceal deterioration. The imitative nature of the calligraphy is detected both by the general impression on the eye, and by uncertainty and inconsistencies in the forms of letters. It is from a failure to keep in mind the natural laws of development and change that early dates, to which they have no title, have been given to imitative MSS.; and, on the other hand, even very ancient examples have been post-dated in an incredible manner.

Down to the time of the introduction of printing, writing ran in two lines - the natural cursive, and the set book-hand which was evolved from it. Cursive writing was essential for the ordinary business of life. MSS. written in the set book-hand filled the place now occupied by printed books, the writing being regular, the lines generally kept even by ruling or other guides, and the texts provided with regular margins. The set book-hand disappeared before the printing press; cursive writing necessarily remains.

In the study of handwriting it is difficult to exaggerate the great and enduring influence which the character of the material employed for receiving the script has had upon the formation of the written letters. The original use of clay by the Babylonians and Assyrians as their writing material was the primary cause of the wedge-shaped symbols which were produced by the natural process of puncturing so stiff and sluggish a substance. The clinging waxen surface of the tablets of the Greeks and Romans superinduced a broken and disconnected style of writing. The comparatively frail surface of Papyrus called for a light touch and slenderly built characters. With the introduction of the smooth and hard-surfaced vellum, firmer and heavier letters, with marked contrasts of fine and thick strokes, became possible, and thence became the fashion. In the task which lies before us we shall have to deal mainly with MSS. written on the two very different materials, papyrus and vellum, and we shall find to how great an extent the general character and the detailed development of Greek;and Latin writing, particularly for literary purposes, has been affected by the two materials.

The history of the ancient papyrus roll and of its successor, the medieval vellum codex, and the particulars of the mechanical arrangement of texts and other details appertaining to the ,evolution of the written book are described in the article Manuscript. In the present article our attention is confined to the history of the script.

The papyrus period of our subject, as regards literary works, ranges generally from the end of the 4th century B.C. to the 4th .century of our era, when the papyrus roll as the vehicle for literature was superseded by the vellum codex. The vellum period extends from the 4th century to the 15th century, when the rise of the art of printing was the doom of the written book. Yet it must not be imagined that there is a hard and fast line separating the papyrus period from the vellum period. In the early centuries of our era there was a transitional period when the use of the two materials overlapped. The employment of vellum for literary purposes began tentatively quite at the beginning of that era; nor did the use of papyrus absolutely cease with the 4th century. But that century marks definitely the period when the change had become generally accepted.

In the case of non-literary documents, written in cursive hands, the papyrus period covers a still wider field. These documents range from the 3rd century B.C. down to the 7th century, and a certain number of examples even extend into the 8th century. The survival of cursive papyrus documents in large numbers is due to the fact that they are chiefly written in Egypt, where papyrus was the common writing material and where climatic conditions ensured their preservation. On the other hand, early cursive documents on vellum are scarce, for it must be borne in mind that, even allowing for the loss of such documents attributable to the perishable nature of that material in the humid climates of Europe, papyrus and waxen tablets were also the usual writing materials of the Greeks and Romans. The importance of the survival of Greek cursive papyri to so late a period is very great, for it enables us to trace the development of the Greek literary minuscule handwriting of the 9th century in a direct line from the cursive :script of the papyri centuries earlier.

Greek Writing. I.-The Papyri In no branch of our subject has so great a development been effected since about 1875 as in that of the palaeography of Greek papyri. Before that time our knowledge was very limited. The material was comparatively meagre; and, though its increase was certainly only a matter of time, yet the most sanguine would hardly have dared to foretell the remarkable abundance of documents which the excavations of a few years would bring to light.

The history of Greek writing on papyrus can now be followed with more or less fullness of material for a thousand years. Actual dated examples range from the late years of the 4th century B.C. to the 7th century A.D. We have a fair knowledge of the leading features of the writing of the 3rd and 2nd centuries B.C.; a less perfect acquaintance with those of the ist century B.C. For the first four centuries of the Christian era there is a fairly continuous series of documents; of the 5th century only a few examples have as yet been recovered, but there is an abundance of material for the 6th and early 7th centuries. Thus it will be seen that, while for some periods we may be justified in drawing certain conclusions and laying down certain rules, for others we are still in an imperfect state of knowledge. But our knowledge will no doubt almost yearly become more exact, as fresh material is brought to light from the excavations which are now continually proceeding; and those periods in which the lack of papyri breaks the chain of evidence will sooner or later be as fully represented as the rest. The material certainly lies buried in the sands; it is our misfortune that the exact sites have not yet been struck.

The first discovery of Greek papyri was made in Europe in 1752, when the excavations on the site of Herculaneum yielded a number of charred rolls, which proved to be of a literary character. All subsequent discoveries we owe to Egypt; and it is to be observed that the papyri which are found in that country have come down to us under different conditions. Some, generally of a literary nature, were carefully deposited with the bodies of their owners in the tomb with the express intention of being preserved; hence such MSS. in several instances have come to our hands in fairly perfect condition. On the other hand, by far the larger number of those recently brought to light have been found on the sites of towns and villages, particularly in the district of the Fayum, where they had been either accidentally lost or purposely thrown aside as of no value, or had even been used up as material for other purposes besides their original one. These are consequently for the most part in an imperfect and even fragmentary condition, although not a few of them have proved to be of the highest palaeographical and literary importance.

The date of the first find of Greek papyri in Egypt was in 1778, when some forty or fifty rolls were discovered by some native diggers, who, however, kept only one of them. After this scarcely anything appeared until the year 1820, when was found on the site of the Serapeum at Memphis, as it was reported, a group of documents of the 2nd century B.C. Then followed a fruitful period, when several important literary papyri were secured: in 1821, the Bankes Homer, containing the last book of the Iliad; in 1847, the roll containing the Lycophron and other orations of Hypereides; in 1849 and 1850, the Harris Homer, bk. xviii. of the Iliad, and a MS. of bks. ii. - iv.; and, in 1856, the Funeral Oration of Hypereides.

But the great bulk of the Greek papyri from Egypt is the result of excavations undertaken during the last quarter of the 19th century and down to the present day. Within this time four very important discoveries of documents in large quantities have taken place. In 1877 a great mass of papyri was found on the site of Arsinoe in the Fayum, being chiefly of a nonliterary nature, and unfortunately in a very fragmentary.state; they are also late in date, being of the Byzantine period. The greater number passed into the possession of the Archduke Rainer, and are now at Vienna; the rest are divided between London, Oxford, Paris and Berlin. After an interval this find was followed by the recovery in 1892, in the same neighbourhood, and chiefly on the site of a village named Socnopaei Nesus, of an extensive series of documents of the Roman period, ranging from the 1st century to the middle of the 3rd century. These papyri, being of an earlier date and in better condition than the Arsinoite collection, are consequently of greater palaeographical value. Most of them are now in Berlin; many are in the British Museum; and some are at Vienna, Geneva and elsewhere. The third and fourth great finds, and the most important of all, were made by Messrs Grenfell and Hunt when excavating, in the seasons1896-1897and 1905-1906, for the Egypt Exploration Fund, at Behnesa, the ancient Oxyrhynchus. Thousands of papyri were here recovered, including, among the non-literary material, a number of rolls in good condition, and comprising also a great store of fragments of literary works, among which occur the now well-known " Logia," or " Sayings of Our Lord," and fragments of the Scriptures, and in some instances of not inconsiderable portions of the writings of various .classical authors. This great collection ranges in date from the 2nd century B.C. to the 7th century A.D.; but in what proportion the documents fall to the several centuries cannot be determined until the series of volumes in which they are to be described for the Graeco-Roman branch of the Egypt Exploration Fund shall have made some substantial progress.

These four great collections of miscellaneous documents have been supplemented by finds of other groups, which fit into them and serve to make more complete the chronological series. Such are the correspondence of a Roman officer named Abinnaeus, of the middle of the 4th century, shared between the British Museum and the library of Geneva in the year 1892; a miscellaneous collection, ranging from the 2nd century B.C. to the 3rd or 4th century A.D., acquired for the Egypt Exploration Fund and published by that society (Fayum Towns and their Papyri, 1900); another collection obtained for the same society from the cartonnage of mummy-cases dating back to the 3rd century B.C. (The Hibeh Papyri, 1906); and a series recovered from excavations at Tebtunis for the University of California (The Tebtunis Papyri, 1902, 1906), generally of the 2nd century B.C. But of these lesser groups by far the most interesting is that which Mr Flinders Petrie extracted, in 1889-1890, from a set of mummy-cases found in the necropolis of the village of Gurob in the Fayum. In the manufacture of these coffins numbers of inscribed papyri had been employed. The fragments thus recovered proved to be some of the most valuable documents for the history of Greek palaeography hitherto found, supplying us with examples of writing of the 3rd century B.C. in fairly ample numbers, and thus carrying back our fuller knowledge of the subject to a period which up to that time had remained almost a blank. Besides miscellaneous documents, there are included the remains of registers of wills entered up from time to time by different scribes, and thus affording a variety of handwritings for study; and, further, the value of the collection is enhanced by the presence of fragments of the Phaedo and Laches of Plato,° and of the lost Antiope of Euripides and of other classical works.

The last decade of the 19th century was also distinguished by the recovery of several literary works of the first importance, inscribed on papyri which had been deposited with the dead, and had thus remained in a fairly perfect condition. In 1889 the trustees of the British Museum acquired a copy of the lost 'AO vaiwv Ifo?ummia of Aristotle - a papyrus of the mimes of the poet Herodas, and a portion of the oration of Hypereides against Philippides; and in 1896 they had the further good fortune to secure a papyrus containing considerable portions of the odes of Bacchylides, the contemporary of Pindar. And to the series of the orations of Hypereides the Louvre was enabled to add, in 1892, a MS. of the greater part of the oration against Athenogenes.

But the most valuable discovery, from a palaeographical point of view, took place in the present century. In 1902 a papyrus roll containing the greater portion of the Persae, a lyrical composition of Timotheus of Miletus, was found at Abusir, near Memphis, and is now at Berlin. It is written in a large hand of a style which had hitherto been known from a document at Vienna entitled the " Curse of Artemisia," and assigned to the early part of the 3rd century B.C.; and from one or two other insignificant scraps. The new papyrus, however, appears to be even older, and may certainly be placed in the later years of the 4th century B.C.: the most ancient extant literary MS. in the Greek tongue. The ascription of this papyrus to the 4th century B.C. has received confirmation from the welcome discovery, in 1906, at Elephantine, of a document (a marriage contract) of the year 311-310 B.C., which is written in the same style of book-hand characters (Aegypt. Urkunden d. kgl. Museen in Berlin, Elephantine Papyri, 1907). Of quite recent date also is the recovery of a considerable part of a commentary on the Thaetetus of Plato,. written in a fine uncial hand of the 2nd century, now in Berlin. Considerable fragments also of the Paeans of Pindar of the 1st or 2nd century; a papyrus containing an historical work attributed to Theopompus or Cratippus, perhaps of the early 3rd century; a copy of Plato's Symposium of the same period; and a portion of the Panegyricus of Isocrates, written in an uncial hand of the 2nd century, are printed in Part V. of the OxyrhynchusPapyri. Further, many leaves of a papyrus codex containing fragments of four comedies of Menander were found in 1905 at Isom Ishkaou, the ancient Aphroditopolis. The recovery of so many great classical works within a few years may be accepted as an earnest of further finds of the same nature, now that excavations are being carried on systematically in Egypt.

From a study of the material thus placed at our disposal certain conclusions have been arrived at which satisfy us that the periodical changes which passed over the character of Greek writing as practised in Egypt coincide pretty nearly with the changes in the political administration of the country. The period of the rule of the Ptolemies from 323 to 30 B.C. has,. in general, its own style of writing, which we recognize as the Ptolemaic; the period of Roman supremacy, beginning with the conquest of Augustus and ending with the reorganization of the empire by Diocletian in A.D. 284, is accompanied by a. characteristic Roman hand; and with the change of administration which placed Egypt under the Byzantine division of the: empire, and lasted down to the time of the Arab conquest in A.D. 640, there is a corresponding change to the Byzantine class of writing. These changes must obviously be attributed to the influence of the official handwritings of the time. A change. of government naturally led to a change of the officials employed,. and with the change of officials would naturally follow a change in the style of production of official documents. In illustration. of this view, it is enough to call to mind the instances of such variations to be met with in the history of the palaeography of medieval Europe, due in the same way to political causes. It is interesting, too, to observe that in our own time the teaching in schools of a particular type of handwriting which finds favour in clerical examinations for the public service has not been without its influence on the general handwriting of the people.

Classifying, then, the writing of the papyri into the three groups - the Ptolemaic, the Roman, and the Byzantine - the next step is to determine, by a closer examination of the documents,. the changes which characterize the several centuries traversed by those groups. In doing this, we cannot apply the exact terms which are employed in describing the MSS. of the middle ages. We have to do with writing which has not yet been cast into the formal literary moulds of the later times; and it has, therefore been found necessary, as well as convenient, to divide the papyri simply into two series, representative of their contents. and not of their style of production - namely literary papyri and non-literary papyri. Neither series, however, it is to be remembered, has a style of writing peculiar to itself. While the extant literary works are, as a rule, written with more or less formality, no doubt by professional scribes for the bookmarket, not a few of even the more valuable of them are copies. in the ordinary cursive hands of the day. Conversely, while we find non-literary documents generally written in ordinary cursive hands, whether by official scribes or by private individuals, yet occasionally we meet with one produced in the formal style more proper to literary examples. Again, while applying to particular letters in papyri such technical terms as capitals, or uncials, or minuscules, we cannot convey by those terms the exact ideas which we convey when thus describing the individual letters of medieval manuscripts. For the letters of the papyrus period were not cast in finished moulds, while the uncial writing and the minuscule writing of the middle ages were settled literary hands. As will presently be seen, the early medieval uncial hand of the vellum codices de - veloped directly from the literary writing of the papyri; the minuscule book-hand of the 9th century was a new type moulded from the cursive into a fixed literary style.

Necessarily, the non-literary papyri are much more numerous than the literary documents, and present a much greater variety of handwriting, being in fact the result of the daily transactions of ordinary life; and how very widespread was the knowledge of writing among the Greek-speaking population of Egypt is sufficiently testified by the surviving examples, coming as they do from the hands of all sorts and conditions of men. We will first examine these specimens of the current handwriting of the day before passing to the review of the more or less artificial book-writing of the literary papyri.

Non-Literary, Cursive, Hands

As already stated, the oldest material for the study of Greek cursive writing is chiefly con - tributed by the papyri discovered at Gurob. Among them are not only the fragments of official registers, which have been mentioned, but also a variety of miscellaneous documents relating to private affairs, and in various hands of the 3rd century and early 2nd century B.C. The non-literary cursive papyri bear actual dates ranging from 270 to 186 B.C. But the discovery (1906) of papyri at Elephantine takes our dated series of cursive documents back to 285-284 B.C.; and in this collection also is the oldest dated Greek document yet found - the marriage contract of the year 311-310 B.C., already mentioned. In this instance, however, the writing is not cursive, but of the literary type.

The leading characteristic of Greek cursive writing of the 3rd century B.C. is its strength and facility. While it may not compare with some later styles in the precise formation of particular letters, yet_' its freedom and spontaneous air lend it a particular charm and please the eye, very much in the same way that a scholar's practised and unconscious handwriting of a good type is more attractive than the more exact formality of a clerk's hand. The letters generally are widely spread and shallow, and, particularly in the official hands, they are linked together with horizontal connecting strokes to such an extent that the text has almost the appearance of depending from a continuous horizontal line. The extreme shallowness or flatness of many of the letters is very striking. A significant indication of the antiquity of Greek cursive writing is found in connexion with the letter alpha, which is, even at this early period, in one of its forms reduced to a mere angle or wedge.

A few lines from an official order (fig. I) of the year 250 B.C. will serve to convey an idea of the trained cursive style of this century: - FIG. 1. - Official Order, 250 B.C.

(- Et.3 T01rOVX0V Kat --rfTpa Kat Elkoms --Oels Togs encravpots Or- -Tv - pa Kat ECK0673 - 0/11X0/1 Kat Tovs EpThuo o - Opaytaaye i os curoo--) As a contrast to this excellent hand, we give a facsimile of a section from a roughly written letter from a land steward to his employer, of about the same date: - FIG. 2. - Letter of a Land Steward, 3rd century ems Expna-a,unv & Kat irapa evvecos apra (3as b icpcOmrupcov avrov eray-yeXoyEvou Kat c/xXortpov (Otos'y be Kat OTC ubc.op exaoros TWV owl) rnv awreXov R-porEpov) Here there is none of the linking of the letters which is seen in the other example: every letter stands distinct. But while the individual letters are clumsily written, the same laws govern their formation as in the other document. The shallow, wide-spread mu, the cursive nu, the small theta, omikron, and rho, are repeated. Here also is seen the tau, with its horizontal stroke confined to the left of the vertical instead of crossing it, and the undeveloped omega, which has the appearance of being clipped - both forms being characteristic of the 3rd century B. C.

The trained clerical hands of the 2nd century B.C. (fig. 3) differ generally from those of the earlier century in a more perfect and less cursive formation, the older shallow type gradually disappearing, and the linking of letters by horizontal strokes being less continuous. But the Ptolemaic character marks the handwriting well through the century; and it is only towards the close of that period and as the next century is entered, that the hand begins to give way and to lose altogether its linked style and the peculiar crispness of the strokes which give it its distinctive appearance. The cursive hand in its best style (e.g. N. et Extr. pls. xxviii., xxix.) is very graceful and exact: - FIG. 3. - Petition, 163-162 B.C.

mutv xpn,uarq'o,ueva cvXa(etav 7rpoopcomucov 77,ucov OE) Towards the end of the Ptolemaic period material greatly fails. There are very few extant cursive documents between the years 80 and 20 B.C. But marks of decadence already appear in the examples of the beginning of the 1st century B.C. The general character of the writing becomes slacker, and the forms of individual letters are less exact. These imper - fections prepare us for the great change which was to follow.

With the Roman period comes roundness of style, in strong contrast to the stiffness and rigid linking of the Ptolemaic hand. Curves take the place of straight strokes in the individual letters, and even ligatures are formed in pliant sweeps of the pen. This transition from the stiff to the flexible finds some - thing of a parallel in the development of the curving and flexible English charterhand of the 14th century from the rigid hand of the 13th century; following, it would seem, the natural law of relaxation. Roundness of style, then, is characteristic of Greek cursive writing in the papyri of the first three centuries of the Christian era, however much individual hands, or groups of hands, might vary among themselves.

A specimen (fig. 4) of cursive writing of the general Roman type is selected from a papyrus (Brit. Mus. No. cxxxi.) which is of more than usual interest, as it is on the verso side of the rolls of which it is composed that the text of Aristotle's Constitution of Athens has been transcribed. It contains the farming accounts of the bailiff of Epimachus, son of Polydeuces, the owner of an estate in the nome of Hermopolis in the 9th and 10th years of the reign of Vespasian, that is 78-79: - Fig. 4. - Farm Accounts, 78-79.

(Etovs EV5EKaTOV a oveo-ractavou o-E0ao-rovOctravat Tot) AnVOS X TO St avrov ortitaxou In the second half of the ist century two styles of handwriting predominate in the cursive papyri. There is the clear and flowing hand, which may be termed the ordinary working hand; and there is also a small and very cursive style which appears in private correspondence and in legal contracts. The 2nd century follows on the same lines as the ist century; but with the 3rd century decadence sets in; the writing begins to slope, and grows larger and rougher and tends to exaggeration.

This exaggeration of the writing of the later Roman period leads the way to the pedantic exaggeration and formalism characteristic of the Byzantine period. In this period the general style of writing is on a larger scale than in the Roman; exaggeration in the size of certain letters marks the progress of the 4th century. Material is wanting for full illustration of the changes effected in the 5th century; but the papyri of the 6th century show a further advance in formalism, the common style being upright and compressed and full of flour - ishes. In the 7th century the hand assumes a sloping style, which always seems to accompany decadence, and grows very irregular and straggling. A specimen of the fully developed Byzantine hand of a legal type is here shown in a few lines from a lease of a farm (fig. 5) in the 6th century (Brit. Mus. pap.

cxiii Fig. 5. - Lease of a Farm, 6th century.

Ovtow To y StKatov Kat avrris TOW- -Eptcpco rpEta Ta rpos T7111 KaXXtEp - TV 6E(TLI, TOV In the long range covered by the Greek papyri the formation of individual letters necessarily varied under different influences; but in not a few instances the original shapes were remarkably maintained. From those which thus remained conservative it is rash to attempt to draw conclusions as to the precise age of the several documents in which they occur. On the other hand, there are some which at certain periods adopted shapes which were in vogue for a limited time and then disappeared, never to be resumed. Such forms can very properly be regarded as sure guides to the palaeographer in assigning dates. We may therefore take a brief survey of the Greek cursive alphabet of the papyri and note some of the peculiarities of individual letters. The incipient form of the alpha which gradually developed into the minuscule letter of the middle ages may be traced back to the Ptolemaic documents of the 2nd century B.C., but the more cursive letter, which was a simple acute angle, representing only two of the three strokes of which the primitive letter was com - posed, was characteristic of the 3rd century B.C., and seems to have gone out of use within the Ptolemaic period. The develop - ment of the cursive beta is interesting. At the very beginning we find two forms in use: the primitive capital letter and a cursive shape somewhat resembling a small n, being in fact an imperfectly written B in which the bows are slurred. This form lasted through the Ptolemaic period. Then arose the natural tendency to reverse the strokes and to form the letter on the principle of u; but still the capital letter also continued in use, so that through the Roman and Byzantine periods the u-shape and the B-shape run on side by side. Analogously the letter kappa, formed on somewhat the same lines as the beta, runs a similar course in developing a cursive u-shaped form by the side of the primitive capital. Delta remained fairly true to its primitive form until the Byzantine period, when the elongation of the head into a flourish led on to the minuscule letter which is familiar to us in the medieval and modern alphabet. Epsilon, the most frequently recurring letter in Greek texts, departs less from its original rounded uncial form that might have been expected. Frequent and varied as its cursive formations are, yet the original shape is seldom quite disguised, the variations almost in all instances arising from the devices of the scribe to dispose swiftly and conveniently of the cross-bar by incorporating it with the rest of the letter. The tendency to curtail the second vertical limb of eta, leading eventually to the h-shape, is in evidence from the first. But in the development of this letter we have one of the instances of temporary forms which lasted only within a fixed period. In the 1st century, side by side with the more usual form, there appears a modification of it, somewhat resembling the contemporary upsilon, consisting of a shallow horizontal curve with a vertical limb slightly turned in at the foot, '0. Its development from the original H is evident: the first vertical limb is slurred, and survives only in the beginning of the hori - zontal curve, while the cross-bar and the second vertical are combined in the rest of the letter. This form was in general use from the middle of the 1st to the middle of the 2nd century, becoming less common after about 160, and practically dis - appearing about The letters formed wholly or in part by circles or loops, theta, omikron, rho, phi, in the earlier centuries have such circles or loops of a small size. Just as there is an analogy between beta and kappa in their developments, as already noticed, so also do mu and pi advance on somewhat similar lines. From the earliest time there is a resemblance between the broad shallow forms of the two letters in the 3rd century B.C., and particularly when they adopt the form of a convex stroke the likeness is very close; and again, in both Roman and Byzantine periods an n-shaped development appears among the forms of both letters. There is also one phase in the development of sigma which affords a useful criterion fot fixing the date of documents within a fixed limit of time. In the Ptolemaic period the letter, always of the C-form, is upright, with a flattened horizontal head; in the Roman period a tendency sets in to curve the head, and in the course of the ist century, by the side of the old stiffer form of the letter, another more cursive one appears, in which the head is drawn down more and more in a curve, C. This form is in common use from the latter part of the ist century to the beginning of the 3rd century. The cursive form of tau, in which the horizontal stroke is kept to the left of the vertical limb, without crossing it, is one of the early shapes of the letter. The formation of the letter Xi in three distinct horizontal strokes is characteristic of the Ptolemaic period, as distinguished from the later type of letter in which the bars are more or less connected. Lastly, the early Ptolemaic form of the W-shaped omega is noticeable from having its second curve undeveloped, the letter having the appearance of being clipped.

Lit irary Hands. - Literary papyri written in book-hands, distinct from the cursive writing which has been under consideration (and in which literary works were also occasionally written), may be divided into two classes: those which were produced by skilled scribes, and therefore presumably for the market, and those which were written less elegantly, but still in a literary hand, and were probably copied by or for scholars for their own use.

Standing at the head of all, and holding that rank as the only literary papyrus of any extent which may be placed in the 4th century B.C., is the famous lyrical work of Timotheus of Miletus, entitled the Persae, which has already been referred to and of which a section of a few lines is here reproduced: A o H M E N FIG. 6. - The Persae of Timotheus, late 4th century B.C.

(- pa giaTO SE -V UELpLaU EXX -TE rav VEwv rroAUa - - aovatp, rrvpos Se 'auk- ' -3 UTOVOEVTa S€ aA'yn- -a a p Es EAAa5a r)-yav- -'yvvTE I TETpao -) The hand, as will be seen, is rather heavy and irregular, but written with facility and strength, and, though the papyrus, perhaps, is not to be classed among the calligraphic productions for the book market, it must rank as a well-written example of the literary script of the time. Capital forms of letters which afterwards assumed the rounded shapes known as uncial are here conspicuous. The exactly formed alpha, the square epsilon with projecting head-stroke, the irregular sigma, the small theta and omikron are to be remarked. Indeed, the only letter which departs essentially from the lapidary character of the alphabet is the omega, here a half-cursive form but still retaining the principle of the structure of the old horse-shoe letter and quite distinct from the w-shape which was soon to be developed. Of this type of writing are also the two non-literary documents already mentioned above, viz. the " Curse of Artemisia " at Vienna, and the marriage contract of the year 311-310 B.C., found at Elephantine. In the latter the sigma appears in the rounded uncial form.

By rare good fortune important literary fragments were recovered in the Gurob collection, which yielded the most ancient dated cursive documents of the 3rd century B.C., so that, almost from the beginning, we start with coeval specimens of both the cursive and of the book-hand, and we are in a position to compare the two styles on equal terms, and thus approximately to date the literary papyri. Palaeographically, this is a matter of the first importance; for while cursive documents, from their nature, in most instances bear actual dates, the periods of literary examples have chiefly to be decided by comparison, and often by conjecture.

The literary fragments from Gurob fall into the two groups just indicated, MSS. written for sale and scholars' copies. Of the former are some considerable portions of two works, the Phaedo of Plato and the lost Antiope of Euripides. Both are written in carefully formed characters of a small type, but of the two the Phaedo is the better executed. As the cursive fragments among which they were found date back to before the middle of the 3rd century B.C., it is reasonable to place these literary remains also about the same period. Their survival is a particularly interesting fact in the history of Greek palaeography, for in them we have specimens of literary rolls which may be fairly assumed to differ very little in appearance from the manuscripts contemporary with the great classical authors of Greece. Indeed, the Phaedo was probably written within a hundred years of the death of the author.

In the facsimile (fig. 7) of a few lines from this papyrus here placed before the reader, the characteristics of the Ptolemaic cursive hand are also to some extent to be observed in the formal book-hand: - r f 1 ,1 f" F f d' Y c'A a El l% r-'10-1' .1 A NA p .p F et e M M?-(A r3 Aric ? P7-te alAi -rt-{ NA E(c i A v.T.MNI c1')4 ' A FIG. 7. - The Phaedo of Plato, 3rd century B.C.

- rrELBODUa SE - avaKwpav ohoµ j .er i ava-yxr) xprl6[B]ad amp/ 5 as auX AE-yEOaOac Kat aOpoq-EcrOaL rrapaKe [[Aeueo"[B]]]at pnSEvL aXAWL) The general breadth of the square letters, the smallness of the letters composed of circles and loops, and the particular formation of such letters as pi and the clipped omega, are repeated. But the approach also of many of the letters to the lapidary capital forms, like those in the papyrus of Timotheus, is to be remarked, such as the precisely shaped alpha, and the epsilon in many instances made square with a long head-stroke. This mixture of forms seems to indicate an advance in the development of the book-hand of the 3rd century B.C., as contrasted with the archaic style of the older Timotheus.

Of the and century B.C. there are extant only two papyri of literary works written in the formal book-hand, and both are now preserved in the Louvre. The one, a dialectical treatise containing quotations from classical authors, has long been known. The other is the oration of Hypereides against Athenogenes, which is an acquisition of comparatively recent date. The dialectical treatise must belong to the first half of the century, as there is on the verso side of the papyrus writing subsequently added in the year 160 B.C. The period of the Hypereides cannot be so closely defined; but the existence on the verso of later demotic writing, said to be of the Ptolemaic time, affords a limit, and the MS. has been accordingly placed in the second half of the century. While the writing of the earlier papyrus is of a light and rather sloping character, that of the Hypereides is firm and square and upright.

Passing to the ist century B.C., the papyri which have been recovered from the ashes of Herculaneum come into account.

Many of them, the texts of which are of a philosophical nature, are written in literary hands, and are conjectured to have possibly formed part of the library of their author, the philosopher Philodemus; they are therefore placed about the middle of the century. To the same time are assigned the remains of a roll containing the oration of Hypereides against Philippides and the third Epistle of Demosthenes (Brit. Mus. papp. cxxxiii., cxxxiv.). But the most important addition to the period is the handsomely written papyrus containing the poems of Bacchylides (fig. 8), which retains in the forms of the letters much of the character of the Ptolemaic style, although for other reasons it can hardly be placed earlier than about the middle of the century: > t -1 ¦c i f r-i r-j 4; 1‹ E-c a 1" Aziz -' T 14 0 I a A` t 2,4 r1 D-p .,4 M r i 'ft---? ' '0-Tc 4 -r 0, E t t c 0 c{ g o -r t ' r p-e '-40' 1C.* = -T f 1 FIG. 8. - Bacchylides, 1st century B.C. (XELpas avravwv 7rpos avyas c7r7rwKEOS a€Xtou TEKVa Svoravoto Xvc' as 7rapOpovos E ayayEty ODUw SE TOL [[Elkouc 00us]] avyas couvLKoTptxas) With the latter half of the ist century B.C. we quit the Ptolemaic period and pass to the consideration of the literary papyri of the Roman period; and it is especially in this latter period that our extended knowledge, acquired from recent discoveries, has led to the modification of views formerly held with regard to the dates to be attributed to certain important literary MSS. As in the case of non-literary documents, the literary writing of the Roman period differs from that of the Ptolemaic in adopting rounded forms and greater uniformity in the size of the letters.

Just on the threshold of the Roman period, near the end of the 1st century B.C., stands a fragmentary papyrus of the last two books of the Iliad, now in the British Museum (pap. cxxviii.), which is of sufficient extent to be noted. Then, emerging on the Christian era, we come upon a fine surviving specimen of literary writing, which we have satisfactory reason for placing near the beginning of the 1st, century. It is a fragment of the third book of the Odyssey (fig. 9), the writing of which closely resembles that of an official document (Brit. Mus. pap. cccliv.) which happens to be written in a formal literary hand, and which from internal evidence can be dated within a few years of the close of the 1st century B.C. There can be no hesitation, therefore, in grouping the Odyssey with that document. The contrast between the round Roman style and the stiff and firm Ptolemaic hands is here well shown in the facsimiles from this papyrus (fig. and the Phaedo and Bacchylides: ekt GW.. 1 C C,u Y ? ? ??1a iJKaJ FIG. 9. - The Odyssey, beginning of 1st century.

(7rat(ES ayE DriXci,2aXwc Ev aB v4 apµar a-yovTEs cva ws ECt'aO 0c S apa Toy, aXa 11,EV Kap7raXtyces S E-EV av v4) ap av SE 'yvvrt Taµcr) (71,701 1 Kac - otJ/a TE ota ESOVat ScoTp€4)e- av S apa TfXE,uaXos 7rEptKaX- 7rap S apa v€ rropu5rjs 7reto to Es Suchpov S avEf3atve Kau r]v- In a similar style of writing are two fragments of Hesiodic poems recently published, with facsimiles, in the Sitzungsberichte (1900, p. 839) of the Berlin Academy. The earliest of the two, now at Strassburg, may be assigned to the first half of the ist century; the other, at Berlin, appears to be of the 2nd century.

At this point two MSS. come into the series, in regard to which there is now held to be reason for revising views formerly entertained. The papyrus known as the Harris Homer (Brit. Mus. pap. cvii.), containing portions of the eighteenth book of the Iliad, which was formerly placed in the ist century B.C., it is thought should be now brought down to a later date, and should be rather assigned to the 1st century of the Christian era. The great papyrus, too, of Hypereides, containing his orations against Demosthenes and for Lycophron and Euxenippus, which has been commonly placed also in the 1st century B.C., and by some even earlier, is now adjudged to belong to the latter part of the ist century A.D.

Within the ist century also is placed a papyrus of great literary interest, containing the mimes of the Alexandrian writer Herodas, which was discovered a few years ago and is now in the British Museum. The writing of this MS. differs from the usual type of literary hand, being a rough and illformed uncial, inscribed on narrow, and therefore inexpensive, papyrus; and if the roll were written for the market, it was a cheap copy, if indeed it was not made for private use. Of the same period is a papyrus of Isocrates, De pace (Brit. Mus. pap. cxxxii.), written in two hands, the one more clerical than the other; and two papyri of Homer, Iliad, iii. - iv. (Brit. Mus. pap. cxxxvi.), and Iliad, xiii. - xiv. (Brit. Mus. pap. dccxxxii.), the first in a rough uneducated hand, but the latter a fine specimen of uncial writing. To about this period also is the Oxyrhynchus Pindar to be attributed, that is to the close of the ist or beginning of the 2nd century. Then follows another famous papyrus, the Bankes Homer, containing the last book of the Iliad, which belongs to the 2nd century and is also written in a careful style of uncial writing. To these is to be added the beautiful papyrus at Berlin, containing a commentary on the Theaetetus of Plato, written in delicately formed uncials of excellent type of the 2nd century; and of the same age is the Panegyricus of Isocrates from Oxyrhynchus, in a round uncial hand. Three important papyri of the Iliad, written in large round uncials, of the 2nd century, are noticed below.

With regard to the later literary works on papyrus that have been recovered, the period which they occupy is somewhat uncertain. The following are, however, placed in the 3rd century, during which a sloping literary hand seems to have been developed, curiously anticipating a similar change which took place in the course of development of the uncial writing of the vellum MSS., the upright hand of the 4th to 6th centuries being followed by a sloping hand in the 7th and 8th centuries: a MS., now in the British Museum, of portions of bks. ii. - iv. of the Iliad, written on eighteen leaves of papyrus, put together in book-form, but inscribed on only one side; on the verso of some of the leaves is a short grammatical treatise attributed to Tryphon: portion of Iliad v., among the Oxyrhynchus papyri (No. ccxxiii.): a fragment of Plato's Laws (Ox. pap. xxiii.): a papyrus of Isocrates, in Nicoclem, now at Marseilles: a fragment of Ezekiel, in book-form, in the Bodleian Library: a fragment of the " Shepherd " of Hermas at Berlin: and a fragment of Julius Africanus, the Hellenica of Theopompus or Cratippus, and the Symposium of Plato, all found at Oxyrhynchus.

Of the 3rd century also are some fragments which are palaeographically of interest, as they are written neither in the recognized literary hand nor in simple cursive, but in cursive characters moulded and adapted in a set form for literary use - thus anticipating the early stages of the development of the minuscule book-hand of the 9th century from the cursive writing of that time.

With the 3rd century the literary hand on papyrus appears to lose most of its importance. We are within measurable distance of the age of vellum, and of the formal uncial writing of the vellum MSS. which is found in some existing examples of the 4th century and in more abundant numbers of the 5th century. We have now to see how the connexion can be established between the literary handwriting of the papyri and the firmer and heavier literary uncial writing of the vellum codices. The literary hands on papyrus which have been reviewed above are distinctly of the style inscribed with a light touch most suitable to the comparatively frail material of papyrus. In the Bankes Homer, however, one may detect some indication of the fullness that characterizes the vellum uncial writing. But it now appears that a larger and rounder hand was also employed on papyrus at least as early as the 1st century. In proof of this we are able to cite a non-literary document (fig. Io) bearing an actual date, which happens to be written in characters that, exclusive of certain less formally-made letters, are of a large uncial literary type. This writing, though not actually of the finished style familiar to us in the early vellum MSS., yet resembles it so generally that it may be assumed, almost as a certainty, that there was in the 1st century a full literary uncial hand formed on this pattern, which was the direct ancestor of the vellum uncial. The tendency to employ at this period a calligraphic style, as seen in the fragments of the Odyssey and one of the Hesiodic poems mentioned above, supports this assumption. The document now referred to is a deed of sale written in the seventh year of Domitian, A.D. 88 (Brit. Mus. pap. cxli.). The letters still retaining a cursive element are alpha, upsilon, and in some instances epsilon. '[[Ccctcu ' I W Fig]]. io. - Deed of Sale, A.D. 88. (- E) 7rT0XE/1atlt EvEp'yE s tat Km 77 Tovtov 'yVV- v TOV 7rEOEwS wS ETw - - eTE7rtypacw77v airo T77S- -avrov 7rE0Ea EXatwv -) As evidence in support of this view that the uncial hand of the vellum MSS. is to be traced back to the period of the document just quoted, we have the important papyrus found by Mr Flinders Petrie at Hawara in Egypt, and now in the Bodleian Library, which contains a portion of the second book of the Iliad. The writing is of the large uncial type under consideration; and there is now full reason for assigning it to the 2nd century at latest. Before the discovery of the document of the year 88 there was nothing to give a clue to the real period of the Homer; and now the date which has been 'suggested is corroborated by a fragment of papyrus from Oxyrhynchus inscribed with some lines from the same book of the Iliad (fig. II) in the same large uncial type (Ox. Pap. vol. i. no. 20, pl. v.). In this latter instance there can be no question of the early date of the writing as on the verso of the papyrus accounts of the end of the 2nd century or of the beginning of the 3rd century have been subsequently added. Yet a third example of the same character has more recently been found at Tebtunis (Tebt. Pap. vol. ii. no. 265, p1. i.): again a considerable fragment of the second book of the Iliad. Thus, then, in the 1st and 2nd centuries there was in use a large uncial hand which was evidently the forerunner of the literary uncial hand of the early vellum codices. It is also to be noted that the literary examples just mentioned are MSS. of Homer; and hence one is tempted to suggest that, as in the production of sumptuous copies on papyrus of a work of such universal popularity and veneration as the Iliad this large and handsome uncial was specially employed, so also the use of a FIG. 11. - The Iliad, 2nd century.

( - wv 'yXwaaa 7roXv07rEp- -03 av77P (777,5 Item) 0- -EtUBw K0071,770"a/.tEVO -) similar type for the early vellum copies of the sacred text of the Scriptures naturally followed.

Greek Writing. Ii. - The Vellum Codices Uncial Writing. - It has been shown above how a round uncial hand had been developing in Greek writing on papyrus during the early centuries of the Christian era, and how even as early as the 2nd century a well-formed uncial script was in use, at least for sumptuous copies of so great and popular an. author as Homer. We have now to describe the uncial hand as it appears in Greek MSS. written on vellum. This harder and firmer and smoother material afforded to the scribes better scope for a calligraphic style hardly possible on papyrus. With the ascendancy of the vellum codex as the vehicle for literature, the characters received the fixed and settled forms to which the name of uncial is more exactly attached than to the fluctuating letters of the early papyri. The term uncial has been borrowed from the nomenclature of Latin palaeography' and applied to Greek writing of the larger type, to distinguish it from the minuscule or smaller character which succeeded it in vellum MSS. of the 9th century. In Latin majuscule writing there exist both capitals and uncials, each class distinct. In Greek MSS. pure capital-letter writing was never employed (except occasionally for ornamental titles at a late time). As distinguished from the square capitals of inscriptions, Greek uncial writing has certain rounded letters, as a, 'c,' w, modifications in others, and some letters extending above or below the line.

It is not probable that vellum codices were in ordinary use earlier than the 4th century; and it is in codices of that age that the handsome calligraphic uncial above referred to was developed. A few years ago the 4th century was the earliest limit to which palaeographers had dared to carry back any ancient vellum codex inscribed in uncials. But the recovery of the Homeric papyri written in the large uncials of the 2nd century has led to a revision of former views on the date of one early vellum MS. in particular. This MS. is the fragmentary Homer of the Ambrosian Library at Milan, consisting of some fifty pieces of vellum cut out of the original codex for the sake of the pictures which they contain; and all of the text that has survived is that which happened to be on the back of the pictures. The Ambrosian Homer has hitherto been generally placed in the 5th century, and the difference of the style of the writing from that of the usual calligraphic type of uncial MSS. of that time, which had been remarked, was thought rather to indicate inferiority in age. But the similarity of the character of the writing (taller and more slender than is usual in vellum codices) to that of the large uncials of the papyrus Homers of the 2nd century from Hawara and Oxyrhynchus and Tebtunis is so striking that the ' St Jerome's often quoted words, " uncialibus, ut vulgo aiunt, litteris " in his preface to the book of Job, have never been explained satisfactorily. Of the character referred to as " uncial " there is no question; but the derivation of the term is not settled.

Ambrosian Homer must be classed with them. Hence it is now held that that MS. may certainly be as early as the 3rd century. But, as that century was still within the period when papyrus was the general vehicle for Greek literature, it may be asked why that material should not in this instance also have been used. The answer may fairly be ventured that vellum was certainly a better material to receive the illustrative paintings, and on that account was employed. The Ambrosian Homer may therefore be regarded as a most interesting link between the papyrus uncial of the and century and the vellum uncial of the 4th and 5th centuries.

With the introduction, then, of vellum as the general writing material, the uncial characters entered on a new phase. The light touch and delicate forms so characteristic of calligraphy on papyrus gave place to a rounder and stronger hand, in which the contrast of fine hair-lines and thickened down-strokes adds so conspicuously to the beauty of the writing of early MSS. on vellum. And here it may be remarked, with respect to the attribution to particular periods of these early examples, that we are not altogether on firm ground. Internal evidence, such, for example, as the presence of the Eusebian Canons in a MS. of the Gospel, assists us in fixing a limit of age, but when there is no such support the dating of these early MSS. must be more or less conjectural. It is not till the beginning of the 6th century that we meet with an uncial MS. which can be approximately dated; and, taking this as a standard of comparison, we are enabled to distinguish those which undoubtedly have the appearance of greater age and to arrange them in some sort of chronological order. But these codices are too few in number to afford material in sufficient quantity for training the eye by familiarity with a variety of hands of any one period - the only method which can give entirely trustworthy results.

Among the earliest examples of vellum uncial MSS. are the three famous codices of the Bible. Of these, the most ancient, the Codex Vaticanus, is probably of the 4th century. The writing must, in its original condition, have been very perfect as a specimen of penmanship; but nearly the whole of the text has been traced over by a later hand, perhaps in the 10th or 11th century, and only such words or letters as were rejected as readings have been left untouched. Written in triple columns, in letters of uniform size, without enlarged initial letters to mark even the beginnings of books, the MS. has all the simplicity of extreme antiquity (Pal. Soc. pl. 104). The Codex Sinaiticus (Pal. Soc. pl. 105) has also the same marks of age, and is judged by its discoverer, Tischendorf, to be even more ancient than the Vatican MS. In this, however, a comparison of the writing of the two MSS. leads to the conclusion that he was mistaken. The writing of the Codex Sinaiticus is not so pure as that of the other MS., and, if that is a criterion of age, the Vatican MS. holds the first place. In one particular the Codex Sinaiticus has been thought to approach in form to its possible archetype on papyrus. It is written with four columns to a page, the open book thus presenting eight columns in sequence, and recalling the long line of columns on an open roll. With regard to such general outward resemblances between the later papyrus literary rolls and the early vellum uncial MSS., we may cite such papyri as the Berlin commentary on the Theaetetus of Plato of the and century and the Oxyrhynchus fragment of Julius Africanus of the 3rd century as forerunners of the style in which the two great codices here mentioned were cast.

The Codex Alexandrinus (fig. 12) is placed in the middle of the 5th century. Here we have an advance on the style of the other two codices. The MS. is written in double columns only, and enlarged letters stand at the beginning of paragraphs. But yet the writing is generally more elegant than that of the Codex Sinaiticus. Examining these MSS. with a view to ascertain the rules which guided the scribes in their work, we find simplicity and regularity the leading features; the round letters formed in symmetrical curves; E and C, &c., finishing off in a hair-line sometimes thickened at the end into a dot; horizontal strokes fine, those of E, H, and e being either in the middle or high in the letter; the base of A and the cross-stroke of II also fine, and, as a rule, kept within the limits of the letters and not projecting beyond. Here also may be noticed the occurrence in the Codex Alexandrinus of Coptic forms of letters (e.g. „alpha and mu) in the titles of books, &c., confirmatory of the tradition of the Egyptian origin of the MS.

<NWN ?' _ ?rr;,  ?? ?-a E ??S CoMCo FIG. 12. - The Bible (Cod. Alex.), 5th century.

(TEKvwv UOV 7rEporaTovv Tas EP aX776(La KaOcos EvTo X7]v EXa(30µEv arc. Tou 7r[ar]p[o]s).- 2 John 4.

To the 5th century may also belong the palimpsest MS. of the Bible, known from the upper text as the Codex Ephraemi, at Paris (ed. Tischendorf, 1845), and the Octateuch (Codex Sarravianus), whose extant leaves are divided between Paris, Leiden and St Petersburg - both of which MSS. are probably of Egyptian origin. Perhaps of the end of the 5th or beginning of the 6th century is the illustrated Genesis of the Cottonian Library, now unfortunately reduced to fragments by fire, but once the finest example of its kind (Cat. Anc. MSS. i. pl. 8). And to about the same time belong the Dio Cassius of the Vatican (Silvestre, pl. 60) and the Pentateuch of the Bibliotheque Nationale (ibid. pl. 61).

In the writing of uncial MSS. of the 6th century there is a marked degeneration. The letters, though still round, are generally of a larger character, more heavily formed, and not so compactly written as in the preceding century. Horizontal strokes (e.g. in A, II, T) are lengthened and finished off with heavy points or finials. The earliest example of this period which has to be noticed is the Dioscorides of Vienna (fig. 13), which is of particular value for the study of the palaeography of early vellum MSS. It is the first uncial example to which an approximate date can be given. There is good evidence to show that it was written early in the 6th century for Juliana Anicia, daughter of Flavius Anicius Olybrius, emperor of the West in 472. Here we already notice the characteristics of uncial writings of the 6th century, to which reference has been A'r7' rG 1 N I ?

FIG. 13. - Dioscorides, early 6th century.

(- La 7rpoµfKf Xpcoyart - [a]vTwv Vtet/.17]TaL [ X a]7raOou 8L7r7 7 K7 7 KaL E X ovTa 7roXXas co[v] - TW 7rEpLcEpEL) made. To this century also belong the palimpsest Homer under a Syriac text in the British Museum (Cat. Anc. MSS., i. pl. 9); its companion volume, used by the same Syrian scribe, in which are fragments of St Luke's Gospel (ibid., pl. 10); the Dublin palimpsest fragments of St Matthew and Isaiah (T. K. Abbot, Par Palimpsest, Dubl.), written in Egypt; the fragments of the Pauline Epistles from Mount Athos, some of which are at Paris and others at Moscow (Silvestre, pls. 63, 64; Sabas, pl. A), of which, however, the writing has been disfigured by retracing at a later period; the Gospels (Cod. N) written in silver and gold on purple vellum, whose leaves are scattered in London (Cott. MS., Titus C. xv.), Rome, Vienna, St Petersburg, and its native home, Patmos; the fragmentary Eusebian Canons written on gilt vellum and highly ornamented, the sole remains of some sumptuous volume (Cat. Anc. MSS. i. pl. II); the Coislin Octateuch (Silvestre, pl. 65); the Genesis of Vienna, and the Codex Rossanensis, and the recently recovered Codex Sinopensis of the Gospels, instances of the very few early illustrated MSS. which have survived. Of the same period is the Codex Marchalianus of the Prophets, which, written in Egypt, follows in its style the Coptic form of uncial.

Reference may here be made to certain early bilingual Graeco-Latin uncial MSS., written in the 6th and 7th centuries, which, however, have rather to be studied apart, or in connexion with Latin palaeography; for the Greek letters of these MSS. run more or less upon the lines of the Latin forms. The best known of these examples are the Codex-Bezae of the New Testament, at Cambridge (Pal. Soc. pls. 14, 15), and the Codex Claromontanus of the Pauline Epistles, at Paris (Pal. Soc. pls. 63, 64), attributed to the 6th or 7th century; and the Laudian MS. of the Acts of the Apostles (Pal. Soc. pl. 80) of the 7th century. To these may be added the Harleian Glossary (Cat. Anc. MSS. i. pl. 13), also of the 7th century. A later example, of the 8th century, is the Graeco-Latin Psalter, at Paris, MS. Coislin 186 (Omont, Facs. des plus anciens MSS. grecs, pl. vii.).

An offshoot of early Greek uncial writing on vellum is seen in the Moeso-Gothic alphabet which Ulfilas constructed for the use of his countrymen in the 4th century, mainly from the Greek letters. Of the few extant remains of Gothic MSS. the oldest and most perfect is the Codex Argenteus of the Gospels, at Upsala, of the 6th century (Pal. Soc. pl. 118), written in characters which compare with purely written Greek MSS. of the same period. Other Gothic fragments appear in the sloping uncial hand seen in Greek MSS. of the 7th and following centuries.

About the year 600 Greek uncial writing passes into a new stage. We leave the period of the round and enter on that of the oval character. The letters E, e, instead of being symmetrically formed on the lines of a circle, are made oval; and other letters are laterally compressed into a narrow shape. In the 7th century also the writing begins to slope to the right, and accents are introduced and afterwards systematically applied. This slanting style of uncials continues in use through the 8th and 9th and into the 10th centuries, becoming heavier as time goes on. In this class of writing there is again the same dearth of dated MSS. as in the round uncial, to serve as standards for the assignment of dates. We have to reach the 9th century before finding a single dated MS. in this kind of writing. It is true that sloping Greek uncial writing is found in a few scattered notes and glosses in Syriac MSS. which bear actual dates in the 7th century, and they are so far useful as showing that this hand was firmly established at that time; but they do not afford sufficient material in quantity to be of really practical use for comparison (see the tables of alphabets in Gardthausen's Griech. Paldog.). Of more value are a few palimpsest fragments of the Elements of Euclid and of Gospel Lectionaries which occur also in the Syriac collection in the British Museum, and are written in the 7th and 8th centuries. There is also in the Vatican a MS. (Reg. 886) of the Theodosian code, which can be assigned with fair accuracy to the close of the 7th century (Gardth. Gr. Pal. p. 158), which, however, being calligraphically written, retains some of the earlier rounder forms. This MS. may be taken as an example of transitional style. In the fragment of a mathematical treatise (fig. 14) from Bobio, forming part of a MS. rewritten in the 8th century and assignable to the previous century, the slanting writing is fully developed. The formation of the letters is good, and conveys the impression that the scribe was writing a hand quite natural to him: t. ` /. FIG. 14. - Mathemat. Treatise, 7th century.

- µrEv] y[ap] 7ravT[os] urEp€ov oXmµ[aror] 7rpos T6 /.IETEWpOP EUXEpEO-TEp-) It should be also noticed that in this MS. - a 'secular one there are numerous abbreviations (Wattenbach, Script. gr. specim. tab. 8). An important document of this time is also the fragment of papyrus in the Imperial Library at Vienna, which bears the signatures of bishops and others to the acts of the Council of Constantinople of 680. Some of the signatures are in slanting uncials (Wattenb., Script. gr. specim., tabb. 12, 13; Gardth., Gr. Pal. tab. 1). Of the 8th century is the collection of hymns (Brit. Mus., Add. MS. 26, 113) written without breathings or accents (Cat. Anc. MSS. i. pl. 14). To the same century belongs the Codex Marcianus, the Venetian MS. of the Old Testament, which is marked with breathings and accents. The plate reproduced from this MS. (Wattenb., Script. gr. specim., tab. 9) contains in the second column a few lines written in round uncials, but in such a laboured style that nothing could more clearly prove the discontinuance of that form of writing as an ordinary hand. In the middle of the 9th century at length we find a MS. with a date in the Psalter of Bishop Uspensky of the year 862 (Wattenb. Script. gr. specim., tab. 10). A little later in date is the MS. of Gregory of Nazianzus, written between 867 and 886 (Silvestre, pl. 71); and at the end of the 9th or beginning of the Toth century stands a lectionary in the Harleian collection (Cat. Anc. MSS. i. pl. 17). A valuable series of examples is also given by Omont (Facsimiles des plus anc. MSS. grecs. de la Bibl. Nat.). But by this time minuscule writing was well established, and the use of the more inconvenient uncial was henceforth almost entirely confined to church-service books. Owing to this limitation uncial writing now underwent a further calligraphic change. As the Toth century advances the sloping characters by degrees become more upright, and with this resumption of their old position they begin in the next century to cast off the compressed formation and again become rounder. All this is simply the result of calligraphic imitation. Bibles and service-books have always been the MSS. in particular on which finely formed writing has been lavished; and it was but natural that, when a style of writing fell into general disuse, its continuance, where it did continue, should become more and more traditional, and a work of copying rather than of writing. In the 10th century there are a few examples bearing dates. There are facsimiles from three of them, viz. a copy of the Gospels (fig. 15), in the Vatican, of 949 (New Pal. Soc. pl. 105), the Curzon Lectionary of 980, and the Harleian Lectionary of 995 (Pal. Soc. pls. 154, 26, 27). The Bodleian commentary on the Psalter (D. 4, i) is likewise of great palaeographic value, being written partly in uncials and partly in minuscules of the middle of the 10th century (Gardth., Gr. Pal. p. 159, tab. 2, col. 4). This late form of uncial writing appears to have lasted to about the middle of the 12th century. (Omont. Facs. pl. xxii.). From it was formed the Slavonic writing in use at the present day: / T? FIG. 15. - The Gospels (Vatican), A.D. 949. (AE'ywv + K[UpG]E iew s

bbvaaa(,2€ KaOa piaac + KaL EKTEivas  xEZpa* i /iaro aUTOU b l{ o oU]S Mycov)  Under the head of late uncial writing must be classed a few bilingual Graeco-Latin MSS. which have survived, written in a bastard kind of uncial in the west of Europe. This writing follows, wherever the shapes of the letters permit, the formation of corresponding Latin characters - the purely Greek forms being imitated in a clumsy fashion. Such MSS. are the Codex Augiensis of Trinity College, Cambridge, of the end of the 9th century (Pal. Soc. pl. 127) and the Psalter of St Nicholas of Cusa (pl. 128) and the Codex Sangallensis and Boernerianus of the 10th century (pl. 179). The same imitative characters are used in quotations of Greek words in Latin MSS. of the same periods.

Minuscule Writing

The beautifully formed minuscule bookhand, which practically superseded the uncial book-hand in the 9th century, did not spring into existence all at once. Its formation had been the work of centuries. It was the direct descendant of the cursive Greek writing of the papyri. It has been shown above, in tracing the progress of the non-literary, cursive writing on papyrus, how the original forms of the letters of the Greek alphabet went through various modifications, always tending towards the creation of the forms which eventually settled down into the recognized minuscules or small letters of the middle ages and modern times. The development of these modifications is apparent from the first; but it was in the Byzantine period especially that the changes became more marked and more rapid. All the minuscule forms, as we know them in medieval literature, had been practically evolved by the end of the 5th century, and in the course of the next two hundred years those forms became more and more confirmed. In the large formal cursive writing of the documents of the 6th and 7th centuries we can pick out the minuscule alphabet in the rough. It only needed to be cast in a calligraphic mould to become the book-hand minuscule, the later development of which we have now to trace. This calligraphic mould seems to have been found in the imperial chancery, from whence issued documents written in a fine round minuscule hand on an ample scale, as appears from one or two rare surviving examples attributed to the 8th and 9th centuries (see the facsimile of an imperial letter, dated variously A.D. 756 or 839, in Wattenbach, Script. graec. specim., pls. xiv., xv., and in Omont, Facs. des plus anc. MSS. grecs. pls. xxvi., xxvii.; and Brit. Mus. papyrus xxxii.). The fine hand only needed to be reduced in scale to become the caligraphic minuscule book-hand of the vellum MSS.

Thus, then, in the 9th century, the minuscule book-hand came into general use for literature, and, with the finely prepared vellum of the time ready to receive it, it assumed under the pens of expert calligraphers the requisite cast, upright, regular and symmetrical, which renders it in its earliest stages one of the most beautiful forms of writing ever created.

Greek MSS. written in minuscules have been classed as follow: (I) codices vetustissimi, of the 9th century and to the middle of the 10th century; (2) vetusti, from the middle of the 10th to the middle of the 13th century; (3) recentiores, from the middle of the 13th century to the fall of Constantinople, 1453; (4) novelli, all after that date.

Of dated minuscule MSS. there is a not inconsiderable number scattered among the different libraries of Europe. Gardthausen (Gr. Pal. 344 seq.) gives a list of some thousand, ending at A.D. 1500. But, as might be expected, the majority belong to the later classes.' Of the 9th century there are not ten which actually bear dates and of these all but one belong to the latter half of the century. In the 10th century, however, the number rises to nearly fifty, in the iith to more than a hundred.

In the period of codices vetustissimi the minuscule hand is distinguished by its simplicity and purity. The period has been well described as the classic age of minuscules. The letters are symmetrically formed; the writing is compact and upright, or has even a slight tendency to slope to the left. In a word, the beauty of this class of minuscule writing is unsurpassed. But in addition to these general characteristics there are special ' In Omont's Facs. des MSS. grecs dates de la Bibl. Nat. will be found a useful list of upwards of 300 facsimiles of dated Greek MSS. (including uncials).

distinctions which belong to it. The minuscule character is. maintained intact, without intrusion of larger or uncial-formed letters. With its cessation as the ordinary literary hand the uncial character had not died out. We have seen that it was still used for liturgical books. It likewise continued to survive in a modified or half-uncial form for scholia, rubrics, titles, and special purposes - as, for example, in the Bodleian Euclid (fig. 16) - in minuscule written MSS. of the 9th and 10th centuries. These uses of the older character sufficed to keep it in remembrance, and it is therefore not a matter for surprise that some of its forms should reappear and commingle with the simple minuscule. This afterwards actually took place. But in the period now under consideration, when the minuscule had been cast into a new mould, and was, so to say, in the full vigour of youth, extraneous forms were rigorously excluded.

(et7ro 0 MN 2 TT Tpeywvcov E7rc Oetoc tc10141r) apa Ta 7rpL07 a[Taj Ta AF PAZ Tpiycova. a Ta 0 MN / TT' Zjute [ac]Ta fTEpEa 7r- Ta et7ro T('v EtprluEvcwv 5rpC6µaT[cov]- va IQoiit,irt Tv'y(avovTa 7rpos 6XX[77Xa] -) The breathings also of this class are rectangular, in unison with the careful and deliberate character of the writing; and there is but slight, if any, separation of the words. t: addition, as far as has hitherto been observed, the letters run above, or stand upon, the ruled lines, and do not depend from them as at a later period. The exact time at which this latter mechanical change took place cannot be named; like other changes it would naturally establish itself by usage. But at least in the middle of the 10th century it seems to have been in use. In the Bodleian MS. of Basil's homilies of 953 A.D. (Pal. Soc. pl. 82) the new method is followed; and if we are to accept the date of the 9th century ascribed to a MS. in the Ambrosian Library at Milan (Wattenb., Script. gr. specim., tab. 17), in which the ruled lines run above the writing, the practice was yet earlier. Certain scribal peculiarities, however, about the MS. make us hesitate to place it so early. In the Laurentian Herodotus (W and V., Exempla, tab. 31), which belongs to the 10th century, sometimes the one, sometimes the other system is followed in different parts of the volume; and the same peculiarity happens in the MS. of Gregory of Nazianzus of A.D. 9 72 in the British Museum (Pal. Soc. pl. 25; Exempla, tab. 7). The second half of the 10th century therefore appears to be a period of transition in this respect.

The earliest dated example of codices vetustissimi is the copy of the Gospels belonging to Bishop Uspensky, written in the year 835. A facsimile is given by Gardthausen (Beitreige) and repeated in the Exempla (tab. 1). Better specimens have been photographed from the Oxford Euclid of A.D. 888 (Pal. Soc. pls. 65, 66; Exempla, tab. 2) from a MS. of Saints' Lives at Paris of A.D. 890 (Omont, Facs. des MSS. gr. dates, pl. 1), and from the Oxford Plato (fig. 17) of A.D. 895 (Pal. Soc. pl. 81; Exempla, tab. 3). Sabas (Specim. Palaeograph.), has also given two facsimiles from MSS. of 880 and 899.

Of dated examples of the first half of the 10th century about a dozen facsimiles are available.

After the middle of the 10th century we enter on the period of the codices vetusti, in which the writing becomes gradually vi Q M C'T' V 'T t ' Y'm cc, f 1 634 ' 6 " r" l erO 4 ^ ^ ' *Tararre lela-cis p-6--y Q-t -rw X P 2 -rp t ¦v-120 11 dJ .

.?

r i b - s 4 Mt N V 03 crirr E--du 'TtV d? o ?! ? P /141-611 m P "'ITT t °r-?-'DJT^ 644/1 !/ ct1 era "T"/1C't,-{' l'tc "-tcrp 0 o-°tt FIG. 16. - Euclid (Oxford), A.D. 888.

less compact. The letters, so to say, open their ranks; and, from this circumstance alone, MSS. of the second half of the century may generally be distinguished from those fifty years earlier. But alterations also take place in the shapes of the letters. Side by side with the purely minuscule forms those of the uncial begin to reappear, the cause of which innovation already been explained. These uncial forms first show s p UUt. 00 cr y 6-&w QUO: = p .per}J-aracri ' 'A/ ' '4 r Zrrou. 13 P o cr N hero  ? s ? ?.

FIG. 17. - Plato (Oxford), A.D. 895.

(- [tdX]Xeis 7rapa 4nX17(30v SEXEaOac vvvt. [aµdi]ca 1 37 7Eiv. My µ1 j COL KaTa vouv '7c - [ovyKE4aXac]wac.) €Oa EKammpov: 7ravv ,Lev ouv: [Ei]vac cp 7 ai. To xaipecv raac wcocs. - [O]aa TOU 'yEvous MTi Toi rov cuf??wva' Ears. to Taura. dXX t TO 4 povE6v. Kai re) - [Ted Toutwv aU EU'y'y€Pi 60 01) TE 6p) themselves at the end of the line, the point at which most changes first gained a footing, but by degrees they work back into the text, and at length become recognized members of the minuscule characters. In the iith and 12th centuries they are well established, and become more and more prominent by the large or stilted forms which they assume. The change, however, in the general character of the writing of this class of codices vetusti is very gradual, uniformity and evenness being well maintained, especially in church books. On the other hand, a lighter and more cursive kind of minuscule is found contemporaneously in MSS. generally of a secular nature. In this hand many of the classical MSS. of the 10th or 11th centuries are written, as the MS. of Aeschylus and Sophocles, the Odyssey and the Apollonius Rhodius of the Laurentian Library at Florence, the Anthologia Palatina of Heidelberg and Paris, the Hippocrates of Venice (Exempla, tabb. 32-36, 38, 40), the Aristophanes of Ravenna (Wattenb., Script. gr. specim., tab. 26), the Strabo of Paris (Omont, Facs. des plus anc. MSS. grecs, pl. 40), a Demosthenes (fig. 18) at Florence (Pal. Soc. ii. pl. 88, 89), &c. In a facsimile from a Plutarch at Venice (Exempla, tab. 44), the scribe is seen to change from the formal to the more cursive hand. This style of writing is distinguishable by its light and graceful character from the current writing into which the minuscule degenerated at a later time.

ccp A r --TIP 19 4 4, Boo ye ?ploy /1" :44411U 14 (45144(941 f /GEr ! oVAl ore: «p o, f °° cam ' -1Gen FIG. 18. - Demosthenes (Florence), early 11th century. (Clvexei.V 5E7 XEyOVTWV TeVWV EB[EXEtv] irpoXECpos Aoyos. cis &pa Kai 7rap'- X'aryaoa Et'pyao*voc Tcvis. ou5[Evos]- V'aya7r0T17s E7rcypa,u,uaros 'iv- ' Tov6' uµcv avayvsoa€rac 're) i rL[ypa,uµa]- TOV X yov W livSpEs athivaiot) The gradual rounding of the rectangular breathings takes place in this period. In the ti th century the smooth breathing, which would most readily lend itself to this modification, first appears in the new form. In the course of the 12th century both breathings have lost the old square shape; and about the same time contractions become more numerous, having been at first confined to the end of the line.

When the period of codices recentiores commences, the Greek 0.1 30, cots.IT°°e(?c "° b `ut'a rt o Yoav"tcre0. vpcI_ Kt$6,k`ocr1) cior FIG. t9. - The Odyssey, 13th century. (i) Was OTL tpov ivtKfaas Tov ax17T77v ws &pa xovi 7 cas 04 4 Xac g XXa(Ev avrap 63vaaEUs aµ)cvoµov rrpos youva KaM"ETo SouXtxtijos) minuscule hand undergoes extensive changes. The contrast between MSS. of the 13th century and those of a hundred years earlier is very marked. In the later examples the hand is generally more straggling, there is a greater number of exaggerated forms of letters, and marks of contraction and accents are dashed on more freely. There is altogether a sense of greater activity and haste. The increasing demand for books created a larger supply. Greater freedom and more variety appear in the examples of this class, together with an increasing use of ligatures and contractions. The general introduction of paper likewise assisted to break up the formal minuscule hand. To this rougher material a rougher style of writing was suited. Through the 14th and 15th centuries the decline of the set minuscule rapidly advances. The writing becomes even more involved and intricate, marks of contraction and accents are combined with the letters in a single action of the pen, and the general result is the production of a thoroughly cursive hand. In some respects, however, the change was not so rapid. Church books were still ordinarily written on vellum, which, as it became scarcer in the market (owing to the injury done to the trade by the competition of paper), was supplied from ancient codices which lay ready to hand on the shelves of libraries; and in these liturgical MSS. the more formal style of the minuscule was still maintained. In the 14th century there even appears a partial renaissance in the writing of Church MSS., modelled to some extent on the lines of the writing of the 12th century. The resemblance, however, is only superficial; for no writer can entirely disguise the character of the writing of his own time. And lastly there was yet another check upon the absolute disintegration of the minuscule book-hand in the 15th century exercised by the professional scribes who worked in Italy, and who in their calligraphical productions reverted again to the older style. The influence of the Renaissance is evident in many of the MSS. of the Italian Greeks, which served as models for the first Greek printing types.

The Greek minuscule book-hand had, then, by the end of the 15th century, become a cursive hand, from which the modern current hand is directly derived. We last saw the ancient cursive in use in the documents prior to the formation of the set minuscule book-hand, and no doubt it continued in use concurrently with the book-hand. But, as the latter passed through the transformations which have been traced, and gradually assumed a more current style, it may not unreasonably be supposed that it absorbed the cursive hand of the period, and with it whatever elements may have survived of the old cursive hand.

Latin Writing. I. - The Roman Cursive The course of Latin palaeography runs on the same lines as that of Greek palaeography. In regard to the former, as in regard to the latter, the documents fall into two main divisions: those which are written in the ordinary cursive hand of everyday life, and those which are written in the formal book-hand of literature. But Latin palaeography covers a wider ground than Greek. Greek writing being limited to the expression of the one has language of a single people has a comparatively narrow and simple career. On the other hand, the Latin alphabet, having been adopted by the nations of western Europe, underwent many transformations in the course of development of the national handwritings of the different peoples, and consequently had a wide and varied career. But in one respect Latin palaeography is at a disadvantage as compared with the sister branch. As we have seen, Greek documents are extant dating back to the 4th century B.C., and the development of Greek writing can be fairly well illustrated by a series of examples of the succeeding centuries. There is no such series of Latin documents available to afford us the means of tracing the growth of Latin writing to the same remote period. No Latin document, either of a literary or of a non-literary character, has yet been recovered which can be placed with certainty earlier than the Christian era. Egypt, while giving up hundreds and thousands of documents in Greek, has hitherto yielded but little in Latin, even of the 1st century, and little too of the next following centuries. Indeed, for our knowledge of Latin writing of the 1st century we still have to depend chiefly upon the results of excavations at Pompeii and Herculaneum and in the Roman catacombs, upon the wall-scribblings which have been laid bare, and upon the waxen tablets and the few papyri which have thence been recovered.

At the time when we come into touch with the first extant examples of Roman writing, we find a few instances of a literary or book-hand as well as a fairly extensive variety of cursive hands. It will be convenient in the first place to examine the Roman cursive writing during the early centuries of our era. Then, for the moment suspending further research in this branch of our subject, we shall proceed to describe the literary script and to trace the development of the large form of book-hand, or majuscule writing, in its two divisions of capitals and uncials, and of the intermediate styles composed of a mixture of large and small letters, or consisting of a blend of the two classes of letters which has received the name of half-uncial. Then we shall turn to follow the development of the national hands, when it will be necessary to come into touch again with the Roman cursive, whence the western continental scripts were derived; and so we shall proceed to the formation of the minuscule writing of the middle ages.

The materials for the study of the early Roman cursive hand have been found in the wall-scribblings, or graffiti, of Pompeii and Herculaneum and Rome (collected in the Corp. inscr. lat. vol. iv.); in the series of 127 libelli or waxen tablets, consisting of perscriptiones and other deeds connected with sales by auction and tax receipts discovered in the house of the banker L. Caecilius Jucundus at Pompeii, and bearing dates of A.D. 15, 27, and 53-62 (published in C.I.L. iv., supplement); in a few scattered papyri from Egypt; and in a set of four-andtwenty waxen tablets bearing dates ranging from A.D. 131 to 167, which were found in ancient mining works in the neighbourhood of Alburnus Major (the modern Verespatak) in Dacia (C. I. L. iii.).

It will have been observed that in the case of the above documents there are three different kinds of material on which they have been inscribed: the plaster surface of walls, the waxen coatings of tablets, and the smooth surface of papyrus. The two former may be classed together as being of a nature which would offer a certain resistance to the free movement of the stilus; while in the case of papyrus the writing-reed or pen would run without impediment. Hence, in writing on the former materials there was a natural tendency to form the letters in disconnected strokes, to make them upright or even inclined to the left, and to employ vertical strokes in preference. The three following specimens from the graffiti and the two sets of tablets will demonstrate the conservative character of this kind of writing, covering as they do about a century and a half. This conservativeness may suggest the probability that the hand seen in the graffiti and the Pompeian tablets had not changed very materially from that practised a century or more earlier, and that it is practically the hand in which the Roman classical writers composed their works. When examining the alphabet of this early Roman cursive hand, we find (as we found in the early Greek cursive) the first beginnings of the minuscule writing of the middle ages. The slurring of the strokes, whereby the bows of the capital letters were lost and their more exact forms ` FIG. 20. - Wall inscription, 1st century. (censio est nam noster magna habet pecuni [am]).

L (?l L (L N IV ¦ «r t???`< ‘4, ?

/ FIG. 2 1. - Pompeian Tablet, A.D. 59.

(quinquaginta nummos nummo libellas quinque ex reliquis ob fullonica. .. anni L. Verani Hupsaei et Albuci Justi d.v.i.d. solut.) a?? ?.SC?  ?

FIG. 22. - Dacian Tablet, A. D. 167.

(descriptum et recognitum factum ex libello erat Alb[urno] maiori ad statione Resculi in quo scriid quod i[nfra] s[criptum] est) modified, led the way to the gradual development of the small letters. With regard to the particular forms of letters employed in the waxen tablets, compare the tables in Corp. inscr. lat., vols. iii., iv. The letter A is formed by a main stroke supporting an oblique stroke above it and the cross bar is either omitted, or is indicated by a small vertical stroke dropping, as it were, out of the letter.

The main stroke of B dwindles to a slight curve, and the two bows are transformed into a long bent stroke so that the letter takes the shape of a stilted a or of a d. The D is chiefly like the uncial the E is generally represented by the old form I found in inscriptions and in the Faliscan alphabet. In the modified form of G the first outline of the flat-headed g of later times appears; H, by losing half of its second upright limb in the haste of writing, comes near to being the small h. In the Pompeian tablets M has the four-stroke form I I!, as in the graffiti; in the Dacian tablets it is a rustic capital, sometimes almost an uncial). The hastily written 0 is formed by two strokes both convex, almost like a. As to the general character of the writing, it is close and compressed, and has an inclination to the left. There is also much combination or linking together of letters (Corp. inscr. lat. iii. tab. A). These peculiarities may, in some measure, be ascribed to the material and to the confined space at the command of the writer. The same character of cursive writing has also been found on a few tiles and potsherds inscribed with alphabets or short sentences - the exercises of children at school (Corp. inscr. lat. iii. 962).

In writing with the pen upon the smooth and unresisting surface of papyrus, the scribe would naturally write a more fluent hand. The disjointed writing of the graffiti and the tablets was changed for one which gradually became more consecutive and which naturally tended in course of time to slope to the right in the effort to be more current and to write letters in connexion without lifting the pen. One of the earliest available examples of Latin writing on papyrus to which an approximate date can be assigned is a fragment at Berlin containing portions of speeches delivered in the senate, said to be of the reign of Claudius, A.D. 41-54 (Steffens, Lat. Pal. taf. 101). The writing, though still somewhat restrained and admitting but little linking of the letters, is yet of a more flowing character than that of the contemporary tablets and graffiti.

We have to pass into the second century before finding the most perfect Latin document on papyrus as yet discovered (fig. 23). This is now in the British Museum, and records the /711 CN.-,Lv -T y ' 7 J X1 71 I' S "? ' fi to c" Jk -1 FIG. 23. - Sale of a slave, A.D. 166.

(- et si quis eum puerum - cerit simplam pecuniam - te dare stipulatus est Fabul - Julius Priscus id fide sua - C. Julius Antiochus mani -) purchase of a slave-boy by an officer in the Roman fleet of Misenum stationed on the Syrian coast, A.D. 166 (Pal. Soc. i. 190; Archaeologia liv. p. 433). The writing of the body of the document is in a formal cursive, generally of the same formation as the inscriptions on the Dacian waxen tablets of the 2nd century, as will be seen from the accompanying facsimile of a few lines (fig. 23).

With this example of legal handwriting of the 2nd century it is interesting to compare two specimens of more ordinary cursive in different styles found in private letters of about the same time. The first (fig. 24) is taken from a fragmentary letter of the year 167 (Grenfell and Hunt, Greek Papyri, 2nd series, cviii.) and is a typical example of a hurried style.

o(7-005 u%'^ L'r%? 07  ?,,,z / ?- N FIG. 24. - Letter, A. D. 167.

(Octobrium ad Puluinos adinterueniente Minucium et Apuleium nepotem scribamnonis Octobris imp. Uero ter -) The second (fig. 25) is from a letter written by one Aurelius Archelaus to Julius Domitius, tribunus militum, recommending a friend named Theon, of the 2nd century (Oxyrhynchus Papyri, i., xxxii.), an instance apparently of slow and imperfect penmanship, every letter painfully and separately formed, yet not in the detached strokes characteristic of the writing of the graffiti and the tablets.

In the examples above we recognize practically the same alphabet as in the graffiti and tablets, but with certain exceptions, particularly in the shape of the letter E,. which is either normal or written very cursively as an acute-angled tick, and in the reversion of other letters to the more normal capital forms.

There is not sufficient material to trace step by step the development of the Roman cursive hand between the 2nd and the 5th centuries; but still, with the few scattered examples at hand, there seems to be reason for conjecture that Latin writing on papyrus passed through phases not very dissimilar to those of Greek writing on the same material. For, when we emerge from the 3rd century, we find an enlarged flowing hand, as in the Latin translation of the fables of Babrius in a fragmentary papyrus of the Amherst collection (No. xxvi.), ascribed to the 3rd or 4th century, and in a letter of recommendation from an N `?ol??n A 74 froN f A A i‘AklevAA. / 'vt vnA. y ' v 'T' y v/v?/?/v C ?(01? vAAA  ?.A T A in' FIG. 25. - Letter, 2nd century.

(Jam tibi et pristine commen daueram Theonem amicum meum et mod[o qujoque puto domine ut eum ant oculos habeas tanquam me est e nim tales omo ut ametur a te) Egyptian official of the 4th century, now at Strassburg (Archiv. fir Papyrusforschung, iii. 2.168); the handwriting of the latter recalling the large style of the Greek cursive of the Byzantine period (fig. 26). That there should be an affinity between the writing of Greek and that of Latin papyri emanating from Egypt is naturally to be expected.

f 7 r 1 d/YO - k ' 1 C 1 1)171 FIG. 26. - Letter of recommendation, 4th century. (Cum in omnibus bonis benigni[tas] etiam scholasticos et maxime qui [honolrificentiae tuae traduntur quod -) This example shows what an immense advance had by this tim e been made in the formation of the minuscule hand, and but little more is required for its completion. It is to be noted, however, that the peculiar old form of letter B with the loop on the left still persists. But only a short time was now needed to bring this letter also in a new shape into line with the other members of the growing minuscule alphabet.

At this point must be noticed a very interesting and important class of the Roman cursive hand which stands apart from the general line of development. This is the official hand of the Roman Chancery, which is unfortunately represented by only two fragmentary papyri of the 5th century (fig. 27), and proves to be a curious moulding of the cursive in a calligraphic style, in which, however, the same characters appear as in other Roman cursive documents, if somewhat disguised. The papyri contain portions of two rescripts addressed to Egyptian officials, and are said to have been found at Phile and Elephantine. Both documents are in the same hand; and the fragments are divided between the libraries of Paris and Leiden. For a time the writing remained undeciphered, and Champollion-Figeac, while publishing a facsimile (Chartes et MSS. sur papyrus, 1840, pl. 14), had to confess that he was unable to read it. Massmann, however, with the experience gained in his work upon the waxed tablets, succeeded without much difficulty in reading the fragment at Leiden (Libellus aurarius, p. 147), and was followed by M. de Wailly, who published the whole of the fragments (Mein. de l'Institut (1842), xv. 399). Later, Mommsen and Jaffe have dealt with the text of the documents (Jahrbuch des. gem. dent. Rechts (1863), vi. 398), and compared in a table the forms of the letters with those of the Dacian tablets.

FIG. 27. - Deed of the Imperial Chancery, 5th century. (portionem ipsi debitam resarcire nec ullum precatorem ex instrumento) The characters are large, the line of writing being about three-fourths of an inch deep, and the heads and tails of the long letters are flourished; but the even slope of the strokes imparts to the writing a certain uniform and graceful appearance. As to the actual shape of the letters, as will be seen from the reduced facsimile here given, there may be recognized in many of them only a more current form of those which have been described above. The A and R may be distinguished by noticing the different angle at which the top strokes are applied; the B, to suit the requirements of the more current style, is no longer the closed d-shaped letter of the tablets, but is open at the bow and more nearly resembles a reversed b; the tall letters f, h, 1, and long s have developed loops; 0 and v-shaped U are very small, and written high in the line. The letters which seem to differ essentially from those of the tablets are E, M, N. The first of these is probably explained correctly by Jaffe as a development of the earlier H quickly written and looped, and may be compared with the tick-shaped letter noticed above. The M and N have been compared with the minuscule forms of the Greek mu and nu, as though the latter had been adopted; but they may with better reason be explained as merely cursive forms of the Latin capitals M and N. That this hand should have retained so much of the older formation of the Roman cursive is no doubt to be attributed to the fact of its being an official style of writing which would conform to tradition.

To continue the development which we saw attained in the letter of the 4th century above (fig. 26) we turn to the documents on papyrus from Ravenna, Naples, and other places in Italy, which date from the 5th century and are written in a looser and more straggling hand (fig. 28). Examples of this hand will be found in largest numbers in Marini's work specially treating of these documents (I papiri diplomatici), and also in the publications of Mabillon (De re diplomatica) ChampollionFigeac (Chartes et MSS. sur papyrus), Massmann (Urkunden in Neapel and Arezzo), Gloria (Paleografia), as well as in Facs. of Ancient Charters in the British Museum, pt. iv., 1878, Nos. 45, 46, and in the Facsimiles of the Palaeographical Society.

FIG. 28. - Deed of Sale (Ravenna), A.D. 572. (huius splendedissimae urbis) The letter a has now lost all trace of the capital; it is the open u-shaped minuscule, developed from the looped uncial (6., (); the b, throwing off the loop or curve on the left which gave it the appearance of d, has at length developed one on the right, and appears in the form familiar in modern writing; minuscule n, and u are fully formed (the last never joining a following letter, and thus always distinguishable from a); p, q, and r approach to the long minuscules, and s, having acquired an incipient tag, has taken the form y which it keeps long after.

This form of writing was widely used, and was not confined to legal documents. It is found in grammatical works, as in the second hand of the palimpsest MS. of Licinianus (Cat. Anc. MSS., pt. ii., pls. 1, 2) of the 6th century, and in such volumes as the Josephus of the Ambrosian Library of the 7th century (Pal. Soc. pl. 59), and in the St Avitus of the 6th century and other MSS. written in France. It is indeed only natural to suppose that this, the most convenient, because cursive, hand, should have been employed for ordinary working MSS. which were in daily use. That so few of such MSS. should have survived is no doubt owing to the destruction of the greater number by the wear and tear to which they were subjected.

Latin Writing. Ii.-Literary Hands We have now to return to the 1st century, the date from which we started in the investigation of the Roman cursive writing, and take up the thread of the history of the book-hand of literature, a few rare examples of which have survived from the ruins of Herculaneum. That a Roman book-hand existed at a still earlier period is quite certain. The analogy of the. survival of very ancient examples of a Greek literary hand. is a sufficient proof; and it is a mere truism to say that as soon as there was a literature, there was likewise a book-hand for its vehicle. No work could be submitted for sale in the market that was not written in a style legible to all. Neatly written copies were essential, and the creation of a formal kind of writing fitted for the purpose naturally resulted. Such formal script must, however, be always more or less artificial as compared with the natural current hand of the time, and there must always be an antagonism between the two styles of script; and, as we have seen in Greek palaeography, the book-hand is always subject to the invading influence of the natural hand.

Capital Writing

Among the Herculaneum fragmentary papyri, then, we find our earliest examples of the Roman literary hand, which must be earlier than A.D. 79, the year of the destruction of the city; and those examples prove to us that the usual literary hand was written in capital letters. Of these letters there are two kinds - the square and the rustic. Square capitals may be defined as those which have their horizontal. lines at right angles with the vertical strokes; rustic letters are not less accurately formed, nor, as their title would seem toimply, are they rough in character, but, being without the exact finish of the square letters, and being more readily written, they have the appearance of greater simplicity. In capital writing the letters are not all of equal height; F and L, and in the rustic sometimes others, as B and R, overtop the rest. In the rustic alphabet the forms are generally lighter and more slender, with short horizontal strokes more or less oblique and wavy. Both styles of capital writing were obviously borrowed from the lapidary alphabets employed under the empire. Both styles were used for public notices inscribed on the walls of Pompeii and other places. But it has been observed that scribes with a natural conservatism would perpetuate a style some time longer in books than it might be used in inscriptions. We should therefore be prepared to allow for this in ascribing a date to a capital written MS., which might resemble an inscription older by a century or more. Rustic capitals, on account of their more convenient shape, came into more general use; and the greater number of the early MSS. in capitals which have survived are consequently found to be in this character. In the Exempla sodium latinorum of Zangemeister and Wattenbach are collected specimens of capital writing.

The literary fragments of papyrus from Herculaneum are written generally in rustic capitals, either of the firm, solid character used in inscriptions, or of the lighter style employed in the fragments of a poem on the battle of Actium (fig. 29). As this poem is the earliest literary work in Latin, of any extent, written in the book-hand, a specimen of the writing is here given. Its period must necessarily lie between the year 31 B.C. the date of the battle and A.D. 79; and therefore we may place it at least early in the 1st century.

That the rustic capital hand was generally adopted for finely written literary MSS. from the period of our earliest examples. onwards through the centuries immediately following may be assumed from the fact of that character being found so widely in favour when we come down to the period of the vellum MSS. Unfortunately no examples have survived to fill the gap between the first century and the oldest of the vellum codices written in rustic capitals of the 4th century. Of the three great MSS. of Virgil preserved in the Vatican Library, which are written in FIG. 29. - Poem on the Battle of Actium, early 1st century. (praeberetque suae qualis ad instantis signa tubae classesq est facies ea visa loci -) this character, the first in date is that known as the Schedae Vaticanae (Exempla, tab. 13; Pal. Soc. pl. 116, 117), a MS. famous for its series of well-finished illustrative paintings in classical style; it is ascribed to the 4th century. The other two MSS. are known as the Codex Romanus and the Codex Palatinus (Exempla, tab. r1, 12; Pal. Soc. pl. 113-115), and are now generally assigned to the 5th century. All three MSS. no doubt must always have been regarded as choice works; and the large scale of the writing employed, particularly in the case of the Romanus and the Palatinus, and the consequently magnificent size of the MSS. when complete, must indicate an unusual importance attaching to them. They were editions de luxe of the great Roman poet. The writing of the Codex Palatinus (Fig. 30) especially is most exact, and is manifestly modelled on the best type of the rustic hand as seen in the inscriptions.

If SIAIUAQVf Mosiii.Rvasia.Di IIOf 111COG I 1AM1 IAIOStiQS11StC1.Lcalil Rr/01DiR11 FIG. 30. - Virgil (Cod. Palatinus) 5th century. (Testaturque deos iterum se ad proelia cogi Bis iam Italos hostis haec altera foedera) In assigning dates to the earliest MSS. of capital-writing, one feels the greatest hesitation, none of them bearing any internal evidence to assist the process. It is not indeed until the close of the 5th century that we reach firm ground - the Medicean Virgil of Florence having in it sufficient proof of having been written before the year 494. The writing is in delicatelyformed letters, rather more spaced out than in the earlier examples (Exempla, tab. 10; Pal. Soc. pl. 86). Another ancient MS. in rustic capitals is the Codex Bembinus of Terence of the 4th or 5th century (Exempla, tab. 8, 9; Pal. Soc., p1.135), a volume which is also of particular interest on account of its marginal annotations, written in an early form of small hand. Among palimpsests the most notable is that of the Cicero In Verrem of the Vatican (Exempla, tab. 4).

Of vellum MSS. in square capitals the examples are not so early as those in the rustic character. Portions of a MS. of Virgil in the square letter are preserved in the Vatican, and other leaves of the same are at Berlin (Exempla, tab. 14). Each page, however, begins with a large coloured initial, a style of ornamentation which is never found in the very earliest MSS. The date assigned to this MS. is therefore the end of the 4th century. In very similar writing, but not quite so exact, are some fragments of another MS. of Virgil in the library of St Gall, probably of a rather later time (Exempla, tab. 14a; Pal. Soc. p. 208).

In the 6th century capital-writing enters on its period of decadence, and the examples of it become imitative. Of this period is the Paris Prudentius (Exempla, tab. 15; Pal. Soc. pls. 29, 30) in rustic letters modelled on the old pattern of early inscriptions, but with a very different result from that obtained by the early scribes. A comparison of this volume with such MSS. as the Codex Romanus and the Codex Palatinus shows the later date of the Prudentius in its widespread writing and in certain inconsistencies in forms. Of the 7th century is the Turin Sedulius (Exempla, tab. 16), a MS. in which uncial writing also appears - the rough and misshapen letters being evidences of the cessation of capital writing as a hand in common use. The latest imitative example of an entire MS. in rustic capitals is in the Utrecht Psalter, written in triple columns and copied, to all appearance, from an ancient example, and illustrated with pen drawings. This MS. may be assigned to the beginning of the 9th century. If there were no other internal evidence of late date in the MS. the mixture of uncial letters with the capitals would decide it. In the Psalter of St Augustine's Canterbury, in the Cottonian Library (Pal. Soc. pl. 19; Cat. Anc. MSS. ii. pls. 12, 13), some leaves at the beginning are written in this imitative style early in the 8th century; and again it is found in the Benedictional of Bishop Aethelwold (Pal. Soc. pl. 143) of the 10th century. In the sumptuous MSS. of the Carlovingian school it was continually used; and it survived for such purposes as titles and colophons for some centuries, usually in a degenerate form of the rustic letters.

Uncial Writing

There was also another majuscule form of writing, besides capitals, employed as a literary book-hand at an early date, but not coeval with the early period of capital writing. This second book-hand was the so-called Uncial hand, a modification of the capital form of writing, in which the square angles of the original letters were rounded off and certain new curved shapes were. introduced, the characteristic letters of the uncial alphabet being a, a, E, h, cn. The origin of some of these rounded letters may be traced in certain forms of the Roman cursive letters of the graffiti and the tablets. But a considerable length of time elapsed before the fully developed uncial alphabet was evolved from these incipient forms. In fact it is only in the vellum MSS. that we first find the firmly written literary uncial hand in perfect form. No doubt the new material, vellum, with its smooth hard surface, immediately afforded the means for the calligraphic perfection with which we find the uncial writing inscribed in these codices.

From the occurrence of isolated uncial forms in inscriptions, the actual period of growth of the finished literary hand has been determined to lie between the later part of the 2nd century and the 4th century. Uncial letters are especially prevalent in Roman-African inscriptions of the 3rd century; but certain letters of the uncial alphabet are not as yet therein matured; minuscule forms of a few letters, particularly b and d, are employed. The discovery also, at Oxyrhynchus, of a fragmentary papyrus of the 3rd century, containing a portion of an epitome of Livy, presents us with an example of the uncial hand in progress of formation for literary purposes, the text being composed mainly of letters of the uncial type, but including a certain proportion of letters, as b, d, m, r, of the minuscule or small character. At length in the 4th century, as already stated, the perfected uncial literary alphabet is found in the vellum codices.

There are still extant a very large number of Latin uncial MSS., a proof of the wide use of this form of literary writing in the early middle ages.

The Exempla of Zangemeister and Wattenbach, so often quoted above, contains a series of facsimiles which illustrate its progress through its career. ' The letter rn has been adopted by the editors as a test letter, in the earlier forms of which the last limb is not curved or turned in. The letter E also in its earlier and purer form has the cross stroke placed high. But, as in every style of writing, when once developed, the earliest examples are the best, being written with a free hand and natural stroke. The Gospels of Vercelli (Exempla, tab. 20), said to have been written by the hand of Eusebius himself, and which may indeed be of his time, is one of the most ancient uncial MSS. Its narrow columns and pure forms of letters have the stamp of antiquity. To the 4th century also is assigned the palimpsest Cicero De republica in the Vatican (Exempla, tab. 17; Pal. Soc. pl. 160), a MS. written in fine large characters of the best type; and a very ancient fragment of a commentary on an ante-Hieronymian text, in three columns, has also survived at Fulda (Exempla, tab. 21). Among the uncial MSS. of the 5th century of which good photographic facsimiles are available are the two famous codices of Livy, at Vienna (fig. 3 1) and Paris (Exempla, tab. 18, 19; Pal. Soc. pl. 31, 32, 183).

FIG. 31. - Livy (Vienna MS.), 5th century. (Iam tibi illa quae igno rantia saecularis bo na opinatur ostendam) To distinguish between uncial MSS. of the 5th and 6th centuries is not very easy, for the character of the writing changes but little, and is free from sign of weakness or wavering. It may, however, be noticed that in MSS. which are assigned to the latter century there is rather less compactness, and occasionally, as the century advances, there is a slight tendency to artificiality.

When the 7th century is reached there is every evidence that uncial writing has entered on a new stage. The letters are more roughly and carelessly formed, and the compactness of the earlier style is altogether wanting. From this time down to the age of Charlemagne there is a continual deterioration, the writing of the 8th century being altogether misshapen. A more exact but imitative hand was, however, at the same time employed, when occasion required, for the production of calligraphic MSS., such as Biblical and liturgical books. Under the encouragement given by Charlemagne to such works, splendid uncial volumes were written in ornamental style, often in gold, several of which have survived to this day.

Mixed and Half-uncial Writing

It is obvious that the majuscule styles of literary writing, viz. the square capital, the rustic capital and the uncial, were of too elaborate and too stately a character to serve all the many requirements of literature. The capital hands, as we have seen, appear to have been employed, at least in many instances, for codices produced on a grand scale, and presumably for special occasions; and if the uncial hand had a longer and wider career, yet in this case also there must often have been a sense that the employment of this fine character gave a special importance and value to the MS. It is not improbable that the survival of so large a number of uncial MSS. is due to the special care that they received at the hands of their owners. Other more manageable styles of writing were necessary, and concurrently with the majuscule hands other forms were developing. The hand which bears the name of Half-uncial was finally evolved, and had itself an important career as a book-hand as well as exercising a large influence on the medieval minuscule hand of literature.

From the first, as we have seen in the case of the graffiti and the tablets, a mingling of capital forms and minuscule forms was prevalent in the non-literary style of writing. There are indications that the same mingling of the two streams was allowed in writing of a literary character. It appears in a rudimentary state in a papyrus fragment from Herculaneum (Exempla, tab. 2 b); and it appears in the epitome of Livy of the 3rd century found at Oxyrhynchus, in which minuscule letters are interspersed among the uncial text. From the regularity and ease with which this MS. is written, it is to be assumed that the mixed hand was ordinarily practised at that time. It is often employed for marginal notes in the early vellum codices. It is used for the text of the Verona Gaius (Exempla, tab. 24) of the 5th century, in which, besides the ordinary uncial shapes, d is also found as a minuscule, r as the transitional r, and s as the tall letter v. Again, in the uncial Florentine Pandects of the 6th century appears a hand which contains a large admixture of minuscule forms (Exempla, tab. 54). From these and other instances it is seen that in uncial MSS. of a secular nature, as in works relating to law and grammar, the scribe did not feel himself restricted to a uniform use of the larger letters, as he would be in producing a church book or calligraphic MS. But the mixed hand, although partaking something of the nature of the Half-uncial hand, was not actually that form of writing. The Half-uncial hand was not only a mingling of uncial and minuscule forms, but also a blending of them, the uncial element yielding more or less to the minuscule influence, while the minuscule element was reacted upon by the uncial sentiment of roundness and sweeping curves. In its full development the Half-uncial, or Roman Half-uncial as it is also called, were it not for a few lingering pure uncial forms, might equally well be described as a large-type minuscule hand. It has, in fact, been sometimes styled the pre-Carolingian minuscule. An early form of this writing is found in the papyrus fragment of Sallust's Catiline, perhaps of the early 5th century, recently recovered at Oxyrhynchus. In vellum codices of the 5th, 6th and 7th centuries Half-uncial writing of a very fine type is not uncommon. It is used for the marginal scholia of the Bembine Terence, of the 5th century. The MS. of the Fasti consulares, at Verona, brought down to 494 A.D. (Exempla, tab. 30), is also in this hand. But the earliest MS. of this class to which a more approximate date can be given is the Hilary of St Peter's at Rome (fig. 3 2), which was written in or before the year 509 or 510 (Exempla, tab. 52; Pal. Soc. pl. 136); the next is the Sulpicius Severus of Verona, of 517 (Exempla, tab. 32); and of the year 569 is a beautifully written MS. at Monte Cassino containing a Biblical commentary (Exempla, tab. 3).

FIG. 32. -St Hilary, A.D. 509-510. (episcopi manum innocente[m] [lin] guam non ad falsiloquium coeg [isti]- nationem anterioris sententi[ae] -) Other examples, of which good facsimiles may be consulted, are the Corbie MS. of Canons, at Paris (Exempla, tab, 41, 42), the St Severianus at Milan (Pal. Soc. pl. 161, 162), the Ashburnham St Augustine (Pal. Soc. ii. 9), and the Paris St Augustine (New. Pal. Soc. pl. 80), of the 6th century; and the Cologne MS. of Canons (Exempla, tab. 44), and the Josephus (Pal. Soc. pl. 138) and St Ambrose (Pal. Soc. pl. 137) of Milan, of the 6th or 7th century.

The influence which the Half-uncial literary hand exercised upon the minuscule book-writing of the 7th and 8th centuries may be traced in greater or less degree in the continental MSS. of that period. We shall find that it formed the basis for the beautiful national handwritings of Ireland and Britain; and it played an important part in the Carolingian reform of the bookhand of the Frankish Empire.

Latin Writing. Iii.-The National Hands We have now to follow the rise and development of the national handwritings of western Europe, all of which were derived from the Roman hand, but from different phases of it. While the Roman Empire was the central power controlling its colonies and conquests, the Roman handwriting, however far apart might be the several countries in which it was current, remained practically one and the same. But, when the empire was broken up and when independent nationalities arose upon its ruins and advanced upon independent paths of civilization, the handwriting inherited from Rome gradually assumed distinctive characters and took the complexions of the several countries, unless from some accident the continuity of the effects of the Roman occupation was disturbed, as it was in Britain by the Saxon invasion. On the continent of western Europe, in Italy, in Spain, in Gaul, the Roman cursive hand had become the common form of writing, and it remained the framework on which the national hands of those countries developed. Thus grew the Lombardic hand of Italy, the Visigothic hand of Spain, and the Merovingian and, later, the Carolingian hand of the Frankish Empire. The earliest charters of the three national divisions, written in cursive hands directly descending from the Roman cursive, and dating generally from the 7th century, still remained related in their general style. It was in the book-hands, elaborated from the cursive character, that the lines of national demarcation became more clearly defined, although naturally there occur also many examples in mixed styles which it is difficult to assign to one or another country.

Lombardic Writing

The national handwriting of Italy did not follow one and the same lines of development throughout the peninsula. In ordinary documents the cursive hand which is seen in the Ravenna deeds, the direct descendant of the Roman cursive, continued in use, growing, in course of time, more and more intricate and difficult to read, the earliest examples, down to the middle of the 9th century, being in the large straggling character of their prototype. The illegible scrawl into which the hand finally degenerated in notarial instruments of southern Italy was at length suppressed by order of Frederick II. (A. D. 1210-1250). But at an early date the Lombardic book-hand was being formed out of this material. In northern Italy new influences were brought into action by Charlemagne's conquest, the independent growth of the native hand was checked, and a mixed style in which the Merovingian type was interwoven with the Italian was produced, to which the name of Franco-Lombardic has been given (see below, fig. 39). But in the Lombardic duchies of the south the native Lombardic book-hand had an unimpeded growth. In such centres as the monasteries of Monte Cassino near Naples and La Cava near Salerno, it took in the 9th century a very exact FIG. 33. - Exultat roll (Lombardic, 12th century). ([H]ec nox est de qua scriptum est Et nox ut dies illuminabitur) and uniform shape. From this date the attention which it received as a calligraphic form of writing, accompanied with accessory ornamentation of initial letters, brought it to a high state of perfection in the 11th century, when by the peculiar treatment of the letters, they assume that strong contrast of light and heavy strokes which when exaggerated, as it finally became, received the name of broken Lombardic. This style of hand lasted to the 13th century.

Papal Documents

A word must be said in this place regarding the independent development of the hands used in the papal Wk d,,r- - a?n,w?uw {, a, r a 4T:, FIG. 34. - Bull of Pope John VIII. (much reduced, A.D. 876). (Dei genetricis mariae filib haec igitur omnia quae huius praecepti) chancery, that great centre which had so wide an influence by setting the pattern for the handsome round-hand writing which became so characteristic of the Italian script. Specimens of a special style of writing, founded on the Lombardic and called littera romana (fig. 34) are to be seen in the early papal documents on papyrus dating from the latter part of the 8th century. In the earliest examples it appears on a large scale, and has rounded forms and sweeping strokes of a very bold character. Derived from the official Roman hand, it has certain letters peculiar to itself, such as the letter a made almost like a Greek w, t in the form of a loop, and e as a circle with a knot at the top. This hand may be followed in examples from A.D. 788 through the 9th century (Fats. de chartes et diplomes, 1866; Ch. Figeac, Chartes et doc. sur papyrus, i - xii.; Letronne, Diploni. merov. cetat., pl. 48. In a bull of Silvester II., dated in (Bibl. l'Ec. des chartes, vol. xxxvii.), we find the hand becoming less round; and at the end of the next century, under Urban II., in 1097 (Mabillon, De re dipl. suppl. p. 115) and 1098 (Sickel, Mon. graph. v. 4), it is in a curious angular style, which, however, then disappears. During the i i th and 12th centuries the imperial chancery hand was also used for papal documents, and was in turn displaced by the exact and calligraphic papal Italian hand of the later middle ages.

Visigothic Writing

The Visigothic writing of Spain ran a course of development not unlike that of the Lombardic. In the cursive hand attributed to the 7th century the Roman cursive has undergone little change in form; but another century developed a most distinctive character. In the 8th century appears the set book-hand in an even and not difficult character, marked by breadth of style and a good firm stroke. This style is maintained through the 9th century with little change, except that there is a growing tendency to calligraphy. In the 10th century the writing deteriorates; the letters are not so uniform, and, when calligraphically written, are generally thinner in stroke. The same changes which are discernible in all the handwritings of western Europe in the 11th century are also to be traced in the Visigothic hand - particularly as regards the rather rigid character which it assumes. It continued in use down to the beginning of the 12th century. Perhaps the most characteristic letter of the book-hand is the q-shaped g. The following specimens (figs. 36) illustrate the Visigothic as written in a large heavy hand of the 9th century (Cat. Anc. MSS. ii. Plate 37), and in a calligraphic example of 1109 (Pal. Soc. Plate 48) .

4t- FIG. 35. - Prayers, 9th century. (tibi dulcedine proxi morum et dignita to operum perfectorum) turn &pf fuauin ?f ee, us- 'c l entm? S &- crofinenawcienbAh 1 t ftu ' '15c11) ttuL' ufque9uo ruccri er flet8tlicu FIG. 36. - Beatus on the Apocalypse, A.D. 1109. (patrum et profetarum et sanctorum et apostolorum que gemitibus et tormenta desiderii sui habuit usquequo fructum ex plebe sua) Merovingian. - The early writing of the Frankish Empire, to which the title of Merovingian has been applied, had a wider range than the other two national hands already described. It had a long career both for diplomatic and literary purposes. In this writing, as it appears in documents, we see that the Roman cursive is subjected to a lateral pressure, so that the letters received a curiously cramped appearance, while the heads and tails are exaggerated to inordinate length.

Facsimiles of this hand, as used in the royal and imperial chanceries, are to be found scattered in various works; but a complete course of Merovingian diplomatic writing may be best studied in Letronne's Diplomata, and in the Kaiserurkunden of Professors Sybel and Sickel. In the earliest documents, commencing in the 7th century and continuing to the middle of the 8th century, the character is large and at first not so intricate as it becomes later in this period. The writing then grows into a more regular form, and in the 9th century a small hand is established, which, however, still retains the exaggerated heads and tails of letters. The direct course of this chancery hand may then be followed in the imperial documents, which from the FIG. 37. - Merovingian diploma, A.D. 679-680.

(dedit in respunsis eb quod ipsa de annus triginta et uno inter ipso -ondam semper tenuerant et possiderant si -) second half of the 9th century are written in a hand more set and evidently influenced by the Carolingian minuscule. This form of writing, still accompanied by the lengthened strokes already referred to, continued in force, subject, however, to the varying changes which affected it in common with other hands, into the 12th century. Its influence was felt as well in France as in Germany and Italy; and certain of its characteristics also appear in the court-hand which the Normans brought with them into England.

The book-hand immediately derived from the early Merovingian diplomatic hand is seen in MSS. of the 7th and 8th centuries in a very neatly written but not very easy hand (Cat. Anc. MSS. ii. Plates 29, 30; Arndt, Schrifttaf. 28).

FIG. 38. - St Gregory's Moralia, 7th century. Merovingian Writing, 7th century.

(- dam intra sinum sanctae eclesiae quasi uicinos adpositos increpant. Saepe uero arrogantes -dem quam tenent arrogantiam se fugire osten -) But other varieties of the literary hand as written in France are seen to be more closely allied to the Roman cursive. The earliest example is found in the papyrus fragments of writings of St Avitus and St Augustine of the 6th century (Etudes paleogr. sur des papyrus du VI me siecle, Geneva, 1866); and other later MSS. by their diversity of writing show a development independent of the cursive hand of the Merovingian charters. It is among these MSS. that those examples already referred to occur which more nearly resemble the Lombardic type.

rr>I FIG. 39. - Ecclesiastical Canons (Franco-Lombardic), 8th century. (propter unitatem salua propriaetate na non sub una substantia conuenientes, neque -itam sed unum eundem filium. Unicum deum) The uncial and half-uncial hands had also their influence in the evolution of these Merovingian book-hands; and the mixture of so many different forms accounts for the variety to be found in the examples of the 7th and 8th centuries. In the Notice sur un MS. Merovingien d'Eugyppus (1875) and the Notice sur un MS. Merovingien de la Bibl. d'Epinal (1878), Delisle has given many valuable facsimiles in illustration of the different hands in these two MSS. of the early part of the 8th century. See also Exempla Codd. Lat. (tab. 57), and autotypes in Cat. anc. MSS. ii. There was, however, through all this period a general progress towards a settled minuscule writing which only required a master-hand to fix it in a purified and calligraphic form. How this was effected will be described below, after disposing of the early national writing of our own islands.

Irish Writing

The early history of the palaeography of the British Isles stands apart from that of the continental schools. As was noticed above, the Roman handwriting which was used by the Roman settlers in Britain and was imparted by them to the native Britons was swept out of existence when the Saxon invasion abruptly destroyed the continuity of Roman civilization in these islands. Britain had to wait a long time for the reappearance of Roman writing in the country; but it was destined to reappear, though in a different phase, in book-form, not in cursive form; and not directly, but through another channel. That channel was Ireland.

It is evident that the civilization and learning which accompanied the establishment of an ancient Church in Ireland could not exist without a written literature. The Roman missionaries would certainly in the first place have imported copies of the Gospels and other books, and it cannot be doubted that through intercourse with England the Irish would obtain continental MSS. in sufficient numbers to serve as models for their scribes. From geographical and political conditions, however, no continuous intimacy with foreign countries was possible; and we are consequently prepared to find a form of writing borrowed in the first instance from a foreign school, but developed under an independent national system. In Ireland we have an instance how conservative writing may become, and how it will hand on old forms of letters from one generation to another when there is no exterior influence to act upon it. After once obtaining its models, the Irish school of writing was left to work out its own ideas, and continued to follow one direct line for centuries. The subsequent English conquest had no effect upon the national handwriting. Both peoples in the island pursued their own course. In MSS. in the Irish language the Irish character of writing was naturally employed; and the liturgical books produced in Irish monasteries by Irish monks were written in the same way. The grants and other deeds of the English settlers were, on the other hand, drawn up by English scribes in their then national writing. The Irish handwriting went on in its even uninterrupted course; and its consequent unchanging form makes it so difficult a matter to assign accurate dates to Irish MSS.