HANDBOOK OF GREEK & LATIN PALAEOGRAPHY by Sir E M Thompson. 3rd edition, published 1906. Kegan, Paul, Trench, Trübner & Co Ltd, London.

| HOME | title page | preface | note to 3rd edition | contents | tables of alphabets |

Book label

Honorary fellow of University College, Oxford; correspondent of the Institute of France, and of the Royal Prussian Academy of Sciences; and Director and Principal Librarian of the British Museum

Third edition with additions and corrections.

London. Kegan Paul, Trench, Trübner & Co. Ltd. Dryden House, 43, Gerrard Street. 1906.

To my friend Leopold Delisle
Member of the Institute and Administrator General of the National Library of France


This Hand-book does not pretend to give more than an outline of the very large subject of Greek and Latin Paleography. It must be regarded as an introduction to the study of the subject, indicating the different branches into which it is divided and suggesting the lines to be followed, rather than attempting full instruction. It in no way supersedes the use of such works as the collections of facsimiles issued by the Palasographical Society and by other societies and scholars at home and abroad; but it is hoped that it will serve as an aid to the more intelligent and profitable study of them.

Our conclusions as to the course of development of the handwritings of former ages are based on our knowledge and experience of the development of modern forms of writing. Children at school learn to write by copying formal text-hands in their copy-books, and the handwriting of each child will bear the impress of the models. But as he grows up the child develops a handwriting of his own, diverging more and more from the models, but never altogether divesting itself of their first influence. Thus, at all times, we have numerous individual handwritings, bat each bearing the stamp of its school and its period; and they, in their turn, react upon and modify the writing of the next generation.

In this way have arisen the handwritings of nations and districts, of centuries and periods, all distinguishable from each other by the trained eye, And the faculty of distinction is not entirely, but to a very great degree, dependent on familiarity. Anyone will readily distinguish the handwritings of individuals of his own time, and will recognize his friend's writing at a glance as easily as he recognizes his face; he has more difficulty in discriminating between the individual handwritings of a foreign country. Set before him specimens of the writing of the last century, and he will confuse the hands of different persons. Take him still farther back, and he will pronounce the writing of a whole school to be the writing of one man; and he will see no difference between the hands, for example, of an Englishman, a Frenchman, and a Fleming. Still farther back, the writing of one century is to him the same as the writing of another, and he may fail to name the locality where a MS. was written by the breadth of a whole continent.

Palaeographical knowledge was formerly confined to a few, chiefly to the custodians or owners of collections of manuscripts; works of reference on the subject were scarce and expensive; and facsimiles, with certain exceptions, were of no critical value. In these days, when photography has made accurate reproduction so simple a matter, the knowledge is within the reach of all who care to acquire it. The collections of facsimiles which have been issued during the last twenty years have brought into the private study materials which the student could formerly have gathered only by travel and personal research. And more than this: these facsimiles enable us to compare, side by side, specimens from manuscripts which lie scattered in the different libraries of Europe and which could never have been brought together. There is no longer any lack of material for the ready attainment of pal geographical knowledge.

Abroad, this attainment is encouraged in various countries by endowments and schools. In our own country, where the development of such studies is usually left to private exertion and enterprise, Palaeography has received but little notice in the past. In the future, however, it will receive better recognition. In the Universities its value has at length been acknowledged as a factor in education. The mere faculty of reading an ancient MS. may not count for much, but it is worth something. The faculty of assigning a date and locality to an undated codex; of deciding between the true and the false; in a word, of applying accurate knowledge to minute points—a faculty which is only to be acquired by long and careful training is worth much, and will give a distinct advantage to the scholar who possesses it.

I have to thank my colleague, Mr. G. P. Warner, the Assistant-Keeper of the Department of MSS., for kind help in passing this work through the press.

E. M. T.

British Museum, 14th December, 1892.


After an interval of thirteen years, a third edition of this Handbook is called for.

I am very conscious of the imperfections of the work; and it seems to me that the time has arrived for a wider and fuller treatment of the subject. The publishers, however, are satisfied that the Handbook is still found useful in its present stereotyped form. Therefore, at their request, I have merely printed some additions and corrections at the end of the text, together with a revised list of the principal palseographieal and diplomatic works brought down to the present time.

E. M. T.

British Museum, 24th February, 1906.


Chapter I.—History of the Greek and Latin Alphabets . page 1.
Chapter II.—Materials used to receive writing: Leaves— Bark—Linen—Clay and Pottery—Wall-spaces—Metals —Lead—Bronze—Wood—Waxen and other Tablets— Greek Waxen Tablets—Latin Waxen. Tablets. page 12.
Chapter III.—Materials used to receive writing (continued): Papyrus—Skins—Parchment and Vellum—Paper. page 27.
Chapter IV.—Writing implements : Stilus, pen, etc.—Inks—Various implements. page 48.
Chapter V.— Forms of Books: The Roll—The Codex—The Text—Punctuation—Accents, etc.—Palimpsests. page 54.
Chapter VI.—Stichometry—Tachygraphy—Cryptography. page 78.
Chapter VII.—Abbreviations and Contractions—Numerals. page 86. ← Note: These are PDF pages (6.4mb). [The Ancient Greek abbreviations are just too complicated to attempt unicode equivalents!]
Chapter VIII.—Greek Paleography : Papyri—Antiquity of Greek writing—Divisions of Greek Palaeography. page 107.
Chapter IX.—Greek Palaeography (continued): The Literary or Book-Hand in Papyri. page118.
Chapter X.—Greek Palaeography (continued): Cursive writing in Papyri, etc.—Forms of cursive letters. page 130.
Chapter XI.— Greek Palaeography (continued): Uncial writing in vellum MSS. page 149.
Chapter XII.—Greek Palaeography (continued) : Minuscule writing in the Middle Ages—Greek writing in Western Europe. page150.
Chapter XIII.—Latin Palaeography: Majuscule writing—Square Capitals—Rustic Capitals—Uncials. page183.
Chapter XIV.—Latin Palaeography (continued): Mixed Uncials and Minuscules—Half-uncials. page 196.
Chapter XV.—Latin Palaeography (continued): Roman Cursive writing. page 203.
Chapter XVI.—Latin Palaeography (continued) : Minuscule writing — Lombardic writing — Visigothic writing — Merovingian writing—The Caroline reform. page 217.
Chapter XVII.—Latin Palaeography (continued): Irish writing—English writing before the Norman Conquest. page 236.
Chapter XVIII.—Latin Palaeography (continued) : The Literary or Book-Hand in the Middle Ages—The English Book-Hand in the Middle Ages. page 257.
Chapter XIX.—Latin Palaeography (continued): Cursive writing—The Papal Chancery—The Imperial Chancery —English Charter-hand — English Chancery-hand — English Court-hand. page 293.
Additions and Corrections. page 321.
Works of Palaeography and Diplomatic. page 339.
Index. page 353.


Derivation of Greek and Latin Alphabets . To face page 10.
Greek Cursive Alphabets. page 148.
Latin Cursive Alphabets. page 216.


| top | contents | the greek & latin alphabets | the greek alphabet | the latin alphabet | table: derivation of Greek & Latin Alphabets from Egyptian | >> | systems of dating | additions & corrections |

The Greek and Latin Alphabets.

Although the task which lies before us of investigating the growth and changes of Greek and Latin palaeography does not require us to deal with any form of writing till long after the alphabets of Greece and Rome had assumed their final shapes, yet a brief sketch of the origin and formation of those alphabets is the natural introduction to such a work as this.

The alphabet which we use at the present day has been traced back, in all its essential forms, to the ancient hieratic writing of Egypt of about the twenty-fifth century before Christ. It is directly derived from the Roman alphabet; the Roman, from a local form of the Greek; the Greek, from the Phoenician; the Phoenician, from the Egyptian hieratic.

The hieroglyphic records of Egypt extend through a period of from four to five thousand years, from the age of the second dynasty to the period of the Roman Empire. Knowing the course through which other primitive forms of writing have passed, we must allow a considerable period of time to have elapsed before the hieroglyphs had assumed the phonetic values which they aIready possess in the earliest existing monuments. Originally these signs were ideograms or pictures, either actual or symbolical, of tangible objects or abstract ideas which, they expressed. From the ideograms in course of time developed the phonograms, or written symbols of sounds, first as verbal signs representing entire words, then as syllabic signs of the articulations of which words are composed. The last stage of development, whereby the syllabic signs are at length taken as the alphabetical signs representing the elementary sounds into which a syllable can be resolved, has always proved the most difficult. Some forms of writing, such as the ancient cuneiform and the modern Chinese, have scarcely passed beyond the syllabic stage. The Egyptians curiously went more than half-way in the last perfecting stage; they developed alphabetical signs, but failed to make independent use of them. A phonogram was added to explain the alphabetically-written word, and an ideogram was added to explain the phonogram. It has been truly said that this cumbrous system seems almost inconceivable to us, who can express our thoughts so easily and so surely by six-and-twenty simple signs, The fact, however, remains that the Egyptians had unconsciously invented an alphabet; and they had been in possession of these letters for more than four thousand years before the Christian era. The oldest extant hieroglyphic insciiption is engraved on a tablet, now in the Ashmolean Museum at Oxford, which was erected to the memory of a priest who lived in the reign of Sent, a monarch of the second dynasty, whose period has been variously given as 4000 or 4700 b.c. In the cartouche of the king's name three of the alphabetical signs are found, one of which, n, has descended and finds a place in our own alphabet. The age of our first letters may thus be said to number some six thousand years. In addition, it is a moderate computation to allow a thousand years to have elapsed between the first origin of the primnsval picture-writing of Egypt and the matured form of development seen in the hieroglyphic characters of the earliest monuments. We may without exaggeration allow a still longer period and be within bounds, if we carry back the invention of Egyptian writing to six or seven thousand years before Christ.

To trace the connection of the Greek alphabet with the Semitic is not difficult. A comparison of the early forms of the letters sufficiently demonstrates their common origin; and, still further, the names of the letters and their order in the two alphabets are the same. But to prove the descent of the Semitic alphabet from the Egyptian has been a long and difficult task. Firstly, in outward shape the Egyptian hieroglyphs of the monuments appear to be totally different from the Semitic letters and to have nothing in common with them. Next, their names are different. The names of the Semitic letters are Semitic words, each describing the letter from its resemblance to some particular object, as aleph an ox, beth a house, and so on. When the Greeks took over the Semitic letters, they also took over their Semitic names; by analogy, therefore, it might be assumed that in adopting. the Egyptian letters the Semites would also have adopted the Egyptian names. Thirdly, the order of the letters is different. All these difficulties combined to induce scholars to reject the ancient, though vague, tradition handed down by Greek and Roman writers, that tho Phoenicians had originally obtained their letters from Egypt. By recent investigation, however, the riddle has been solved, and the chain of connection between our alphabet and the ancient Egyptian hieroglyphic writing has, beyond reasonable doubt, been completed.

The number of alphabetical signs found among the inscriptions on Egyptian monuments has been reckoned at forty-five. Some of these, however, are used only in special cases; others are only alternative forms for signs more commonly employed. The total number of signs ordinarily in use may thus be reduced to twenty-five—a number which agrees with the tradition handed down by Plutarch, that the Egyptians possessed an alphabet of five-and-twenty letters. Until lately, however, these hieroglyphs had been known only in the set and rigid forms as sculptured on the monuments. In 1859 the French Egyptologist de Rouge made known the results of his study of an ancient cursive form of —not, however, the original, but a copy, which, having been found in a tomb of the eleventh dynasty, is anterior to the period of the Hyksos invasion, and may be assigned to the period about 2500 B.C. The old hieratic cursive character which is employed in this most ancient document is the style of writing which was no doubt made use of in Egypt for ordinary purposes at the time of the Semitic conquest, and, as de llouge has shown, was taken by the new lords of the country as material wherewith to form an alphabet of their own. But, as has already been remarked, while adopting the Egyptian forms of letters, the Semites did not also adopt their Egyptian names, nor did they keep to their order. This latter divergence may be due to the fact that it was a selection that was made from a large number of ideograms and phonograms, and nota complete and established alphabet that was taken over. In the table which accompanies this chapter the ancient hieratic character of the Prisse papyrus may be compared with the early Semitic alphabet of some sixteen hundred years later, and, in spite of the interval of time, their resemblance in very many instances is still wonderfully close.

This Semitic alphabet appears to have been employed in the cities and colonies of the Phoenicians and among the Jews and Moabites and other neighbouring tribes at a period not far removed from the time when the children of Israel sojourned in the land of Egypt. Bible history proves that in patriarchal times the art of writing was unknown to the Jews, but that, when they entered the promised land, they were in possession of it. All evidence goes to prove its acquisition during the Semitic occupation of the Delta; and the diffusion of the newly-formed alphabet may have been due to the retreating Hyksos wheu driven out of Egypt, or to Phoenician traders, or to both. See Isaac Taylor, The Alphabet, chap. ii. § 8.

The most ancient form of the Phoenician alphabet known to us is preserved in a series of inscriptions which date back to the tenth century b.c. The most important of them is that engraved upon the slab known as the Moabite stone, which records the wars of Mesha, king of Moab, about 890 B.C., against Israel and Edom, and which was discovered in 1868 near the site of Dibon, the ancient capital of Moab. Of rather earlier date are some fragments of a votivo inscription engraved on bronze plates found in Cyprus in 1876 and dedicating a vessel to the god Baal ol Lebanon. From these and other inscriptions of the oldest type we can construct the primitive Phoenician alphabet of twenty-two letters, as represented in the third column of the table, in a form, however, which must have passed through many stages of modification since it was evolved from the ancient cursive hieratic writing of Egypt.

The Greek Alphabet.

The Greeks learned the art of writing from the Phoenicians at least as early as the ninth century B.C.; and it is not improbable that they had acquired it even one or two centuries earlier. Trading stations and colonies of the Phoenicians, pressed at home by the advancing conquests of the Hebrews, were established in remote times in the ishinds and mainlands of Greece and Asia Minor ; and their alphabet of two-and-twenty letters was adopted by the Greeks among whom they settled or with whom they had commercial dealings. It is not, however, to be supposed that the Greeks received the alphabet from the Phoenicians at one single place from whence it was passed on throughout Ilellas; but rather at several points of contact from whence it was locally diffused among neighbouring cities and their colonies. Hence we are prepared to find that, while the Greek alphabet is essentially one and the same in all parts of Hellas, as springing from one stock, it exhibits certain local peculiarities, partly no doubt inherent from its very first adoption at different centres, partly derived from local influences or from linguistic or other causes. We cannot, then, accept the idea of a Cadmean alphabet, in the sense of an alphabet of one uniform pattern for all Greece.

Among the two-and-twenty signs adopted from the Phoenician, four, viz. aleph, he, yod, and ayin, were made to represent the vowel-sounds a, e, i, o, both long and short, the signs for e and o being also employed for the diphthongs ei and ou. The last sound continued to be expressed by the omikron alone to a comparatively late period in the history of the alphabet. The fifth vowel-sound u was provided for by a new letter, the upsilon, which may have been either a modification or "differentiation" of the Phoenician waw, or derived from a letter of similar form in the Cypriote alphabet. This new letter must have been added almost immediately after the introduction of the Semitic signs, for there is no local Greek alphabet which is without it. Next was felt the necessity for distinguishing long and short e, and in Ionia, the aspirate gradually falling into disuse, the sign Η, eta, was adopted to represent long e, probably before the end of the seventh century B.C. About the same time the long o began to be distinguished by various signs, that used by the Ionians, the omega, Ω, being apparently either a differentiation of the omikron, or, as has been suggested, taken from the Cypriote alphabet. The age of the double letter Φ and of Χ and as they appear in the Ionian alphabet, must, as is evident from their position, be older than or at least coeval with omega.

With regard to the sibilants, their history is involved in great obscurity. The original Semitic names appear to have become confused in the course of transmission to the Greeks and to have been applied by them to the wrong signs. The name zeta appears to correspond to the name tsade, but the letter appears to be zayn. Xi, which seems to be the same word as shin, represents the letter samekh. San, which is probably derived from zayn, represents tsade. Sigma, which may be identified with samekh, represents shin. But all these sibilants were not used simultaneously for any one dialect or locality. In the well-known passage of Herodotus (i. 139), where he is speaking of the terminations of Persian names, we are told that they "all end in the same letter, which the Dorians call san and the Ionians sigma." There can be little doubt that the Dorian san was originally the M-shaped sibilant which is found in the older Dorian inscriptions, as in Thera, Melos, Crete, Corinth and Argos. It has also been identified with a T-shaped sign wliich was used for a special sound on coins of Mesembria, and. at Halicarnassus in the fifth century b.c. This sibilant is now known to have been derived from the Phoenician letter tsade. In a Greek abecedarium scratched upon a small vase discovered at Formello, near Veii, this letter is seen to occupy the eighteenth place, corresponding to the position of tsade in the Phoenician alphabet. In the damaged Greek alphabet similarly scrawled on the Galassi vase, which was found at Cervetri in 1836, it is formed more closely on the pattern of the Phoenician letter. In the primitive Greek alphabet, therefore, san existed (representing tsade) as well as sigma (representing shin), but as both appear to have had nearly the same sibilant sound, the one or the other became superfluous. In the Ionian alphabet sigma was preferred.

But the disuse of the letter san must date far back, for its loss affected the numerical value of the Greek letters. When this value was being fixed, the exclusion of san was overlooked, and the numbers were calculated as though that letter had not existed. The preceding letter pi stands for 80; the kappa for 90, the numerical value of the Phoenician tsade and properly also that of san. At a later period the obsolete letter was re-adopted as the numerical sign for 900, and became the modern sampi (i.e. san + pi ), so called from its partial resemblance, in its late form, to the letter pi.

With regard to the local alphabets of Greece, different states and different islands either adopted or developed distinctive signs. Certain letters underwent gradual changes, as eta from closed to open Η, and theta from crossed to the dotted circle , which forms were common to all the varieties of the alphabet. The most ancient forms of the alphabet are found in Melos, Thera, aud Crete, which moreover did not admit the double letters. While some states retained the digamma or the koppa, others lost them; while some developed particular differentiations to express certain sounds, others were content to express two sounds by one letter. The forms for beta and Epsilon for epsilon are peculiar to Corinth and her colonies; the Argive alphabet is distinguished by its rectangular lambda; and the same letter appears in the Boeotian, Chalcidian, and Athenian alphabets in the inverted form .

But while there are these local differences among the various alphabets of ancient Greece, a broad division has been laid down by Kirchhoff, who arranges them in two groups, the eastern and the western. The eastern group embraces the alphabet which has already been referred to as the Ionian, common to the cities on the western coast of Asia Minor and the neighbouring islands, and the alphabets of Megara, Argos, and Corinth and her colonies ; and, in a modified degree, those of Attica, Naxos, Thasos, and some other islands. The western group includes the alphabets of Thessaly, Euboen, Phocis, Locris, and Poeotia, and of all the Peloponnese (excepting the states specified under the other group), and also those of the Achaean and Chalcidian colonies of Italy and Sicily.

In the eastern group the letter Ξ has the sound of x ; and the letters Χ, Ψ, the sounds of kh and ps. (In Attica, Naxos, etc., the letters Ξ and Ψ were wanting, and the sounds x and ps were expressed by ΠΣ, ΦΣ. In the western group the letter Ξ is wanting, and Χ, Ψ have the values of x and kh ; while the sound ps was expressed by ΠΣ or ΦΣ, or rarely by a special sign . In a word, the special test-letters are :—
The Greek and Latin Alphabets.

Eastern: Χ = kh. Ψ = ps.
Western: Χ = x. Ψ = kh.

How this distinction camo about is not known, although several explanations have been hazarded. It is unnecessary in this place to do more than state the fact.

As the Semitic languages were written from right to left, so in the earliest Greek inscriptions we find the same order followed. Next came the method of writing called boustrophedon, in which the written lines run alternately from right to left and from left to right, or vice versa, as the plough forms the furrows. Lastly, writing from left to right became universal. In the most ancient tomb-inscriptions of Melos and Thera we have the earliest form of writing. Boustrophedon was commonly used in the sixth century B.C. A notable exception, however, is found in the famous Greek inscription at Abu Simbel—the earliest to which a date can be given. It is cut on one of the legs of the colossal statues which guard the entrance of the great temple, and records the exploration of the Nile up to the second cataract by certain Greek, Ionian, and Carian mercenaries in the service of Psammetichus. The king here mentioned may be the first (B.C. 654—617) or the second (B.C. 594 — 589) of the name. The date of the writing may therefore be roughly placed about 600 B.C. The fact that, besides this inscription, the work of two of the soldiers, the names of several of their comrades are also cut on the rock, proves how well established was the art of writing even at this early period.

The Latin Alphabet.

Like the local alphabets of Greece, the Italic alphabets varied from one another by the adoption or rejection of different signs, according to tlio requirements of language. Thus the Latin and Faliscan, the Etruscan, the Umbrian, and the Oscan alphabets are sufficiently distinguished in this way ; but at the same time the common origin of all can be traced to a primitive or so-called Pelasgian alphabet of the Chalcidian type. The period of the introduction of writing into Italy from the great trading and colonizing city of Chalcis must be carried back to the time when the Greeks wrote from right to left. A single Latin inscription ← On a small vase found in Rome in 1880. See L'Inscription de Duenos in the Mélanges d'Archéologie et d'Histoire of the Ecole Française de Rome, 1882, p. 147. has been found which is thus written; and in the other Italic scripts this ancient system was also followed. We may assume, then, that the Greek alphabet was made known to the native tribes of Italy as early as the eighth or ninth century B.C., and not improbably through the ancient Chalcidian colony of Cumae, which tradition named as the earliest Greek settlement in the land. The eventual prevalence of the Latin alphabet naturally followed the political supremacy of Rome.

The Latin alphabet possesses twenty of the letters of the Greek western alphabet, aud, in addition, three adopted signs. Taking the Formello and Galassi abecedaria as representing the primitive alphabet of Italy, it will be seen that the Latins rejected the letter san and the double letters theta, phi, and chi (Ψ), and disregarded the earlier sign for xi. ← Some of these letters are generally accepted as the origin of certain of the symbols used for the Latin numerals. But a different origin has been lately proposed by Professor Zangemeister: Entstehung der römisehen Zahlzeichen (Sitzber. d. k. Preuss. Akad., 1887). In Quintilian's time letter X was the " ultima nostrarum" and closed the alphabet. The sound z in Latin being coincident with the sound s, the letter zeta dropped out. But at a later period it was restored to the alphabet, as Z, for the purpose of transliteration of Greek words. As, however, its original place had been meanwhile filled by the new letter G, it was sent down to the end of the alphabet. With regard to the creation of G, till the middle of the third century B.C. its want was not felt, as C was employed to represent both the hard e and g sounds, ← The sound represented by C in Latin no doubt also gradually, but at a very early period, became indistinguishable from that represented by K. Hence the letter K fell into general disuse in writing, and only survived as an archaic form in certain words, such as kalendae. a survival of this use being seen in the abbreviations C. and Cn. for Gaius and Gnaeus; but gradually the new letter was developed from C and was placed in tho alphabet in the position vacated by zeta. The digamma had become the Latin F, and the upsilon had been transliterated as the Latin V; but in the time of Cicero upsilon, as a foreign letter, was required for literary purposes, and thus became again incorporated in the Latin alphabet—this time without change of form, Y. Its position shows that it was admitted before Z.

Table to face page 10 (Derivation of the Greek and Latin Alphabets from the Egyptian.)