Verrius Flaccus and his epitomists, Festus and Paulus, have preserved many treasures of early glossographers who are now lost to us. He copied Aelius Stilo (Reitzenstein, “Verr. Forsch.,” in vol. i. ofBreslauer philol. Abhandl., p. 88; Kriegshammer,Comm. phil. Ien.vii. 1. 74 sqq.), Aurelius Opilius, Ateius Philol., the treatiseDe obscuris Catonis(Reitzenstein,ib.56. 92). He often made use of Varro (Willers,De Verrio Flacco, Halle, 1898), though not of hisling. lat.(Kriegshammer, 74 sqq.); and was also acquainted with later glossographers. Perhaps we owe to him theglossae asbestos(Goetz,Corpus, iv.;id., Rhein. Mus.xl. 328). Festus was used by Ps.-Philoxenus (Dammann, “De Festo Ps.-Philoxeni auctore,”Comm. Ien.v. 26 sqq.), as appears from theglossae ab absens(Goetz, “De Astrabae Pl. fragmentis,”Ind. Ien., 1893, iii. sqq.). The distinct connexions with Nonius need not be ascribed to borrowing, as Plinius and Caper may have been used (P. Schmidt,De Non. Marc. auctt. gramm.145; Nettleship,Lect. and Ess. 229; Fröhde,De Non. Marc. et Verrio Flacco, 2; W. M. Lindsay, “Non. Marc.,”Dict. of Repub. Latin, 100, &c.).Thebilingual(Gr.-Lat., Lat.-Gr.) glossaries also point to an early period, and were used by the grammarians (1) to explain the peculiarities (idiomata) of the Latin language by comparison with the Greek, and (2) for instruction in the two languages (Charis. 254. 9, 291. 7, 292. 16 sqq.; Marschall,De Q. Remmii P. libris gramm. 22; Goetz,Corp. gloss. lat.ii. 6).For the purposes of grammatical instruction (Greek for the Romans, Latin for the Hellenistic world), we have systematic works, a translation of Dositheus and the so-calledHermeneutica, parts of which may be dated as early as the 3rd centuryA.D., and lexica (cf. Schoenemann,De lexicis ant.122; Knaack, inPhil. Rundsch., 1884, 372; Traube, inByzant. Ztschr.iii. 605; David,Comment. Ien.v. 197 sqq.).The most important remains of bilingual glossaries are two well-known lexica; one (Latin-Greek), formerly attributed (but wrongly, see Rudorff, inAbh.. Akad. Berl., 1865, 220 sq.; Loewe,Prodr.183, 190; Mommsen,C.I.L.v. 8120; A. Dammann,De Festo Pseudo-philoxeni auctore, 12 sqq.; Goetz, Corp. ii. 1-212) to Philoxenus (consulA.D.525), clearly consists of two closely allied glossaries (containing glosses to Latin authors, as Horace, Cicero, Juvenal, Virgil, the Jurists, and excerpts from Festus), worked into one by some Greek grammarian, or a person who worked under Greek influence (his alphabet runs A, B, G, D, E, &c.); the other (Greek-Latin) is ascribed to Cyril (Stephanus says it was found at the end of some of his writings), and is considered to be a compilation of not later than the 6th century (Macrobius is used, and theCod. Harl., which is the source of all the other MSS., belongs to the 7th century); cf. Goetz,Corp.ii. 215-483, 487-506, praef.ibid.p. xx. sqq. Furthermore, the bilingual medico-botanic glossaries had their origin in old lists of plants, as Ps.-Apuleius in the treatiseDe herbarum virtutibus, and Ps.-Dioscorides (cf. M. Wellmann,Hermes, xxxiii. 360 sqq., who thinks that the latter work is based on Pamphilus,q.v.; Goetz, Corp. iii.); the glossary, entitledHermeneuma, printed from theCod. Vatic.reg. Christ. 1260, contains names of diseases.Just as grammar developed, so we see the original form of the glosses extend. Ifmassucum edacemin Placidus indicates the original form, the allied gloss of Festus (masucium edacem a mandendo scilicet) shows an etymological addition. Another extension consists in adding special references to the original source, ase.g.at the glossOcrem(Fest. 181a. 17), which is taken from Ateius Philol. In this way collections arose like thepriscorum verborum cum exemplis, a title given by Fest. (218b. 10) to a particular work. Further theglossae veterum(Charis. 242. 10); theglossae antiquitatum(id.229. 30); theidonei vocum antiquarum enarratores(Gell. xviii. 6. 8); thelibri rerum verborumque veterum(id.xiii. 24. 25). L.Cincius, according to Festus (330b. 2), wroteDe verbis priscis; Santra,De antiquitate verborum(Festus 277a. 2).Of Latin glossaries of the first four centuries of the Roman emperors few traces are left, if we except Verrius-Festus. Charis, 229. 30, speaks ofglossae antiquitatumand 242. 10 ofglossae veterum, but it is not known whether these glosses are identic, or in what relation they stand to theglossemata per litteras Latinas ordine composita, which were incorporated with the works of this grammarian according to the index in Keil, p. 6. Latin glosses occur in Ps.-Philoxenus, and Nonius must have used Latin glossaries; there exists aglossarium Plautinum(Ritschl,Op.ii. 234 sqq.), and the bilingual glossaries have been used by the later grammarian Martyrius; but of this early period we know by name only Fulgentius and Placidus, who is sometimes called Luctatius Placidus, by confusion with the Statius scholiast, with whom theglossae Placidihave no connexion. All that we know of him tends to show that he lived in North Africa (like Fulgentius and Nonius and perhaps Charisius) in the 6th century, from whence his glosses came to Spain, and were used by Isidore and the compiler of theLiber glossarum(see below). These glosses we know from (1) Codices Romani (15th and 16th century); (2) theLiber glossarum; (3) the Cod. Paris. nov. acquis. 1298 (saec. xi.), a collection of glossaries, in which the Placidus-glosses are kept separate from the others, and still retain traces of their original order (cf. the editions published by A. Mai,Class. auct.iii. 427-503, and Deuerling, 1875; Goetz,Corp.v.; P. Karl, “De Placidi glossis,”Comm. Ien.vii. 2. 99, 103 sqq.; Loewe,Gloss. Nom.86; F. Bücheler, inThesaur. gloss. emend.). His collection includes glosses from Plautus and Lucilius.(Fabius Planciades) Fulgentius (c.A.D.468-533) wroteExpositio sermonum antiquorum(ed. Rud. Helm, Lips. 1898; cf. Wessner,Comment. Ien.vi. 2. 135 sqq.) in sixty-two paragraphs, each containing a lemma (sometimes two or three) with an explanation giving quotations and names of authors. Next to him come theglossae Nonianae, which arose from the contents of the various paragraphs in Nonius Marcellus’ work being written in the margin without the words of the text; these epitomized glosses were alphabetized and afterwards copied for other collections (see Goetz,Corp.v. 637 sqq.,id.v. Praef. xxxv.; Onions and Lindsay,Harvard Stud.ix. 67 sqq.; Lindsay,Nonii praef.xxi.). In a similar way arose theglossae Eucheriiorglossae spiritales secundum Eucherium episcopumfound in many MSS. (cf. K. Wotke,Sitz. Ber. Akad. Wien, cxv. 425 sqq.; = theCorpus Glossary, first part), which are an alphabetical extract from theformulae spiritalis intelligentiaeof St Eucherius, bishop of Lyons,c.434-450.7Other sources were theDifferentiae, already known to Placidus and much used in the medieval glossaries; and theSynonyma Ciceronis; cf. Goetz, “Der Liber glossarum,” inAbhandl. der philol.-hist. Cl. der sächs. Gesellsch. d. Wiss., 1893, p. 215;id.inBerl. philol. Wochenschr., 1890, p. 195 sqq.; Beck, inWochenschr., p. 297 sqq., and Sittls,ibid.p. 267;Archiv f. lat. Lex.vi. 594; W. L. Mahne, (Leid. 1850, 1851); also various collections ofscholia. By the side of the scholiasts come the grammarians, as Charisius, or an ars similar to that ascribed to him; further, treatisesde dubiis generibus, thescriptores orthographici(especially Caper and Beda), and Priscianus, the chief grammarian of the middle ages (cf. Goetz inMélanges Boissier, 224).During the 6th, 7th and 8th centuries glossography developed in various ways; old glossaries were worked up into new forms, or amalgamated with more recent ones. It ceased, moreover, to be exclusively Latin-Latin, and interpretations in Germanic (Old High German, Anglo-Saxon) and Romanic dialects took the place of or were used side by side with earlier Latin ones. The origin and development of the late classic and medieval glossaries preserved to us can be traced with certainty. While reading the manuscript texts of classical authors, the Bible or early Christian and profane writers, students and teachers, on meeting with any obscure or out-of-the-way words which they considered difficult to remember or to require elucidation, wrote above them, or in the margins, interpretations or explanations in more easy or better-known words. The interpretations written above the line are called “interlinear,” those written in the margins of the MSS. “marginal glosses.” Again, MSS. of the Bible or portions of the Bible were often provided with literal translations in the vernacular written above the lines of the Latin version (interlinear versions).Of such glossed MSS. or translated texts, photographs may be seen in the various palaeographical works published in recent years; cf.The Palaeogr. Society, 1st ser. vol. ii. pls. 9 (Terentius MS. of 4th or 5th century, interlinear glosses) and 24 (Augustine’s epistles, 6th or 7th century, marginal glosses); see further, plates 10, 12, 33, 40, 50-54, 57, 58, 63, 73, 75, 80; vol. iii. plates 10, 24, 31, 39, 44, 54, 80.From these glossed or annotated MSS. and interlinear versions glossaries were compiled; that is, the obscure and difficult Latin words, together with the interpretations, were excerpted and collected in separate lists, in the order in which they appeared, one after the other, in the MSS., without any alphabetical arrangement, but with the names of the authors or the titles of the books whence they were taken, placed at the head of each separate collection or chapter. In this arrangement each article by itself is called a gloss; when reference is made only to the word explained it is called thelemma, while the explanation is termed theinterpretamentum. In most cases the form of the lemma was retained just as it stood in its source, and explained by a single word (tesca:sancta, Varro vii. 10;clucidatus: suavis, id.vii. 107; cf. Isid.Etym.i. 30. 1, “quid enim illud sit in uno verbo positum declarat [scil.glossa] ut conticescere est tacere”), so that we meet with lemmata in the accusative, dative and genitive, likewise explained by words in the same cases; the forms of verbs being treated in the same way. Of this first stage in the making of glossaries, many traces are preserved, for instance, in the late 8th century Leiden Glossary (Voss. 69, ed. J. H. Hessels), where chapter iii. contains words or glosses excerpted from theLife of St Martinby Sulpicius Severus; chs. iv., v. and xxxv. glosses from Rufinus; chs. vi. and xl. from Gildas; chs. vii. to xxv. from books of the Bible (Paralipomenon; Proverbs, &c., &c.); chs. xxvi. to xlviii, from Isidore, the Vita S.Anthonii, Cassiodorus, St Jerome, Cassianus, Orosius, St Augustine, St Clement, Eucherius, St Gregory, the grammarians Donatus, Phocas, &c. (See also Goetz,Corp.v. 546. 23-547. 6. and i. 5-40 from Ovid’sMetam.; v. 657 from Apuleius,De deo Socratis; cf. Landgraf, inArch.ix. 174).By a second operation the glosses came to be arranged inalphabeticalorder according to the first letter of the lemma, but still retained in separate chapters under the names of authors or the titles of books. Of thissecondstage the Leiden Glossary contains traces also: ch. i. (Verba de Canonibus) and ii. (Sermones de Regulis); see Goetz,Corp.v. 529 sqq. (from Terentius), iv. 427 sqq. (Virgil).The third operation collected all the accessible glosses in alphabetical order, in the first instance according to the first letters of the lemmata. In this arrangement the names of the authors or the titles of the books could no longer be preserved, and consequently the sources whence the glosses were excerpted became uncertain, especially if the grammatical forms of the lemmata had been normalized.A fourth arrangement collected the glosses according to the first two letters of the lemmata, as in the Corpus Glossary and in the still earlierCod. Vat.3321 (Goetz,Corp.iv. 1 sqq.), where even many attempts were made to arrange them according to the first three letters of the alphabet. A peculiar arrangement is seen in theGlossae affatim(Goetz,Corp.iv. 471 sqq.), where all words are alphabetized, first according to the initial letter of the word (a, b, c, &c.), and then further according to the firstvowelin the word (a, e, i, o, u).No date or period can be assigned to any of the above stages or arrangements. For instance, the first and second are both found in the Leiden Glossary, which dates from the end of the 8th century, whereas the Corpus Glossary, written in the beginning of the same century, represents already the fourth stage.For the purpose of identification titles have of late years been given to the various nameless collections of glosses, derived partly from their first lemma, partly from other characteristics, as glossaeabstrusae; glossaeabavus majorandminor; g.affatim; g.ab absens; g.abactor; g.Abba Pater; g.a, a; g.Vergilianae; g.nominum(Goetz,Corp.ii. 563, iv.); g. Sangallenses (Warren,Transact.Amer. Philol. Assoc.xv., 1885, p. 141 sqq.).A chief landmark in glossography is represented by theOrigines(Etymologiae) of Isidore (d. 636), an encyclopedia in which he, like Cassiodorus, mixed human and divine subjects together. In many places we can trace his sources, but he also used glossaries. His work became a great mine for later glossographers. In the tenth book he deals with the etymology of many substantives and adjectives arranged alphabetically according to the first letter of the words, perhaps by himself from various sources. His principal source is Servius, then the fathers of the Church (Augustine, Jerome,Lactantius) and Donatus the grammarian. This tenth book was also copied and used separately, and mixed up with other works (cf. Loewe,Prodr.167. 21). Isidore’sDifferentiaehave also had a great reputation.Next comes theLiber glossarum, chiefly compiled from Isidore, but all articles arranged alphabetically; its author lived in Spain c.A.D.690-750; he has been called Ansileubus, but not in any of the MSS., some of which belong to the 8th century; hence this name is suspected to be merely that of some owner of a copy of the book (cf. Goetz, “Der Liber Glossarum,” inAbhandl. der philol.-hist. Class, der kön. sächs. Ges.xiii., 1893;id., Corp.v., praef. xx. 161).Here come, in regard to time, some Latin glossaries already largely mixed with Germanic, more especially Anglo-Saxon interpretations: (1) the Corpus Glossary (ed. J. H. Hessels), written in the beginning of the 8th century, preserved in the library of Corpus Christi College, Cambridge; (2) the Leiden Glossary (end of 8th century, ed. Hessels; another edition by Plac. Glogger), preserved in the Leiden MS. Voss. Qo. 69; (3) the Épinal Glossary, written in the beginning of the 9th century8and published in facsimile by the London Philol. Society from a MS. in the town library at Épinal; (4) theGlossae Amplonianae,i.e.three glossaries preserved in the Amplonian library at Erfurt, known as Erfurt1, Erfurt2and Erfurt3. The first, published by Goetz (Corp.v. 337-401; cf. also Loewe,Prodr.114 sqq.) with the various readings of the kindred Épinal, consists, like the latter, of different collections of glosses (also some from Aldhelm), some arranged alphabetically according to the first letter of the lemma, others according to the first two letters. The title of Erfurt2(incipit II. conscriptio glosarum in unam) shows that it is also a combination of various glossaries; it is arranged alphabetically according to the first two letters of the lemmata, and contains theaffatimandabavus maiorglosses, also a collection from Aldhelm; Erfurt3are theGlossae nominum, mixed also with Anglo-Saxon interpretations (Goetz,Corp.ii. 563). The form in which the three Erfurt glossaries have come down to us points back to the 8th century.The first great glossary or collection of various glosses and glossaries is that of Salomon, bishop of Constance, formerly abbot of St Gall, who diedA.D.919. An edition of it in two parts was printedc.1475 at Augsburg, with the headlineSalemonis ecclesie Constantiensis episcopi glosse ex illustrissimis collecte auctoribus. The oldest MSS. of this work date from the 11th century. Its sources are theLiber glossarum(Loewe, Prodr. 234 sqq.), the glossary preserved in the 9th-century MS.Lat. Monac.14429 (Goetz, “Lib. Gloss.” 35 sqq.), and the great Abavus Gloss (id., ibid.p. 37;id., Corp.iv. praef. xxxvii.).TheLib. glossarumhas also been the chief source for the important (but not original) glossary of Papias, ofA.D.1053 (cf. Goetz inSitz. Ber. Akad. Münch., 1903, p. 267 sqq., who enumerates eighty-seven MSS. of the 12th to the 15th centuries), of whom we only know that he lived among clerics and dedicated his work to his two sons. An edition of it was published at Milan “per Dominicum de Vespolate” on the 12th of December 1476; other editions followed in 1485, 1491, 1496 (at Venice). He also wrote a grammar, chiefly compiled from Priscianus (Hagen,Anecd. Helv.clxxix. sqq.).The sameLib. gloss.is the source (1) for theAbba PaterGlossary (cf. Goetz,ibid.p. 39), published by G. M. Thomas (Sitz. Ber. Akad. Münch., 1868, ii. 369 sqq.); (2) the Greek glossaryAbsida lucida(Goetz,ib.p. 41); and (3) the Lat.-Arab. glossary in theCod. Leid. Scal. Orient.No. 231 (published by Seybold inSemit. Studien, Heft xv.-xvii., Berlin, 1900).The Paulus-Glossary (cf. Goetz, “Der Liber Glossarum,” p. 215) is compiled from the second Salomon-Glossary (abacti magistratus), theAbavus majorand theLiber glossarum, with a mixture of Hebraica. Many of his glosses appear again in other compilations, as in the Cod. Vatic. 1469 (cf. Goetz,Corp.v. 520 sqq.), mixed up with glosses from Beda, Placidus, &c. (cf. a glossary published by Ellis inAmer. Journ. of Philol.vi. 4, vii. 3, containing besides Paulus glosses, also excerpts from Isidore; CambridgeJourn. of Philol.viii. 71 sqq., xiv. 81 sqq.).Osbern of Gloucester (c.1123-1200) compiled the glossary entitledPanormia(published by Angelo Mai asThesaurus novus Latinitatis, from Cod. Vatic. reg. Christ. 1392; cf. W. Meyer,Rhein. Mus.xxix., 1874; Goetz inSitzungsber. sächs. Ges. d. Wiss., 1903, p. 133 sqq.;Berichte üb. die Verhandl. der kön. sächs. Gesellsch. der Wiss., Leipzig, 1902); giving derivations, etymologies, testimonia collected from Paulus, Priscianus, Plautus, Horace, Virgil, Ovid, Mart. Capella, Macrobius, Ambrose, Sidonius, Prudentius, Josephus, Jerome, &c., &c. Osbern’s material was also used by Hugucio, whose compendium was still more extensively used (cf. Goetz, l.c., p. 121 sqq., who enumerates one hundred and three MSS. of his treatise), and contains many biblical glosses, especially Hebraica, some treatises on Latin numerals, &c. (cf. Hamann,Weitere Mitteil. aus dem Breviloquus Benthemianus, Hamburg, 1882; A. Thomas, “Glosses provençales inéd.” inRomania, xxxiv. p. 177 sqq; P. Toynbee,ibid.xxv. p. 537 sqq.).The great work of Johannes de Janua, entitledSumma quae vocatur catholicon, dates from the year 1286, and treats of (1) accent, (2) etymology, (3) syntax, and (4) so-called prosody,i.e.a lexicon, which also deals with quantity. It mostly uses Hugucio and Papias; its classical quotations are limited, except from Horace; it quotes the Vulgate by preference, frequently independently from Hugucio; it excerpts Priscianus, Donatus, Isidore, the fathers of the Church, especially Jerome, Gregory, Augustine, Ambrose; it borrows many Hebrew glosses, mostly from Jerome and the other collections then in use; it mentions theGraecismusof Eberhardus Bethuniensis, the works of Hrabanus Maurus, theDoctrinaleof Alexander de Villa Dei, and theAuroraof Petrus de Riga. Many quotations from theCatholiconin Du Cange are really from Hugucio, and may be traced to Osbern. There exist many MSS. of this work, and the Mainz edition of 1460 is well known (cf. Goetz inBerichte üb. die Verhandl. der kön. sächs. Gesellsch. der Wiss., Leipzig, 1902).The gloss MSS. of the 9th and 10th centuries are numerous, but a diminution becomes visible towards the 11th. We then find grammatical treatises arise, for which also glossaries were used. The chief material was (1) theLiber glossarum; (2) the Paulus glosses; (3) theAbavus major; (4) excerpts from Priscian and glosses to Priscian; (5) Hebrew-biblical collections of proper names (chiefly from Jerome). After these comes medieval material, as thederivationeswhich are found in many MSS. (cf. Goetz inSitzungsber. sächs. Ges. d. Wiss., 1903, p. 136 sqq.; Traube inArchiv f. lat. Lex.vi. 264), containing quotations from Plautus, Ovid, Juvenal, Persius, Terence, occasionally from Priscian, Eutyches, and other grammarians, with etymological explanations. Thesederivationeswere the basis for the grammatical works of Osbern, Hugucio and Joannes of Janua.A peculiar feature of the late middle ages are the medico-botanic glossaries based on the earlier ones (see Goetz,Corp.iii.). The additions consisted in Arabic words with Latin explanations, while Greek, Latin, Hebrew and Arabic, interchange with English, French, Italian and German forms. Of glossaries of this kind we have (1) theGlossae alphita(published by S. de Renzi in the 3rd vol. of theCollect. Salernitana, Naples, 1854, from two Paris MSS. of the 14th and 15th centuries, but some of the glosses occur already in earlier MSS.); (2)Sinonoma Bartholomei, collected by John Mirfeld, towards the end of the 14th century, ed. J. L. G. Mowat (Anecd. Oxon.i. 1, 1882, cf. Loewe,Gloss. Nom.116 sqq.); it seems to have used the same or some similar source as No. 1; (3) the compilations of Simon de Janua (Clavis sanationis, end of 13th century), and of Matthaeus Silvaticus (Pandectae medicinae, 14th century; cf. H. Stadler, “Dioscor. Longob.” inRoman. Forsch.x. 3. 371; Steinmeyer,Althochd. Gloss.iii.).Of biblical glossaries we have a large number, mostly mixed with glosses on other, even profane, subjects, as Hebrew and other biblical proper names, and explanations of the text of the Vulgate in general, and the prologues of Hieronymus. So we have theGlossae veteris ac novi testamenti(beginning “Prologus graece latine praelocutio sive praefatio”) in numerous MSS. of the 9th to 14th centuries, mostly retaining the various books under separate headings (cf. Arevalo,Isid.vii. 407 sqq.; Loewe,Prodr.141; Steinmeyer iv. 459; S. Berger,De compendiis exegeticis quibusdam medii aevi, Paris, 1879). Special mention should be made of Guil. Brito, who lived about 1250, and compiled aSumma(beginning “difficiles studeo partes quas Biblia gestat Pandere”), contained in many MSS. especially in French libraries. ThisSummagave rise to theMammotrectusof Joh. Marchesinus, about 1300, of which we have editions printed in 1470, 1476, 1479, &c.Finally we may mention such compilations as theSumma Heinrici; the work of Johannes de Garlandia, which he himself callsdictionarius(cf. Scheler inJahrb. f. rom. u. engl. Philol.vi., 1865, p. 142 sqq.); and that of Alexander Neckam (ib.vii. p. 60 sqq., cf. R. Ellis, inAmer. Journ. of Phil.x. 2); which are, strictly speaking, not glossographic. TheBreviloquusdrew its chief material from Papias, Hugucio, Brito, &c. (K. Hamann,Mitteil. aus dem Breviloquus Benthemianus, Hamburg, 1879;id.,Weitere Mitteil., &c., Hamburg, 1882); so also theVocabularium Ex quo; the variousGemmae;Vocabularia rerum(cf. Diefenbach,Glossar. Latino-Germanicum).After the revival of learning, J. Scaliger (1540-1609) was the first to impart to glossaries that importance which they deserve (cf. Goetz, inSitzungsber. sächs. Ger. d. Wiss., 1888, p. 219 sqq.), and in his edition of Festus made great use of Ps.-Philoxenus, which enabled O. Müller, the later editor of Festus, to follow in his footsteps. Scaliger also planned the publication of aCorpus glossarum, and left behind a collection of glosses known asglossae Isidori(Goetz,Corp.v. p. 589 sqq.; id. inSitzungsber. sächs. Ges., 1888, p. 224 sqq.; Loewe,Prodr.23 sqq.), which occurs also in old glossaries, clearly in reference to the tenth book of theEtymologiae.The study of glosses spread through the publication, in 1573, of the bilingual glossaries by H. Stephanus (Estienne), containing, besides the two great glossaries, also theHermeneumata Stephani, which is a recension of thePs.-Dositheana(republished Goetz,Corp.iii. 438-474), and theglossae Stephani, excerpted from a collection of theHermeneumata(ib.iii. 438-474).In 1600 Bonav. Vulcanius republished the same glossaries, adding (1) the glossaeIsidori, which now appeared for the first time; (2) theOnomasticon; (3)notaeandcastigationes, derived from Scaliger (Loewe,Prodr.183).In 1606 Carolus and Petrus Labbaeus published, with the effective help of Scaliger, another collection of glossaries, republished, in 1679, by Du Cange, after which the 17th and 18th centuries produced nofurther glossaries (Erasm. Nyerup published extracts from the Leiden Glossary, Voss. 69, in 1787,Symbolae ad Literat. Teut.), though glosses were constantly used or referred to by Salmasius, Meursius, Heraldus, Barth, Fabricius and Burman at Leiden, where a rich collection of glossaries had been obtained by the acquisition of the Vossius library (cf. Loewe,Prodr.168). In the 19th century came Osann’sGlossarii Latini specimen(1826); the glossographic publications of Angelo Mai (Classici auctores, vols. iii., vi., vii., viii., Rome, 1831-1836, containing Osbern’sPanormia, Placidus and various glosses from Vatican MSS.); Fr. Oehler’s treatise (1847) on theCod. Amplonianusof Osbern, and his edition of the three Erfurt glossaries, so important for Anglo-Saxon philology; in 1854 G. F. Hildebrand’sGlossarium Latinum(an extract fromAbavus minor), preserved in a Cod. Paris. lat. 7690; 1857, Thomas Wright’s vol. of Anglo-Saxon glosses, which were republished with others in 1884 by R. Paul Wülcker under the titleAnglo-Saxon and Old English Vocabularies(London, 2 vols., 1857); L. Diefenbach’s supplement to Du Cange, entitledGlossarium Latino-Germanicum mediae et infimae aetatis, containing mostly glosses collected from glossaries, vocabularies, &c., enumerated in the preface; Ritschl’s treatise (1870) on Placidus, which called forth an edition (1875) of Placidus by Deuerling; G. Loewe’sProdromus(1876), and other treatises by him, published after his death by G. Goetz (Leipzig, 1884); 1888, the second volume of Goetz’s own greatCorpus glossariorum Latinorum, of which seven volumes (except the first) had seen the light by 1907, the last two being separately entitledThesaurus glossarum emendatarum, containing many emendations and corrections of earlier glossaries by the author and other scholars; 1900, Arthur S. Napier,Old English Glosses(Oxford), collected chiefly from Aldhelm MSS., but also from Augustine, Avianus, Beda, Boethius, Gregory, Isidore, Juvencus, Phocas, Prudentius, &c.There are a very great number of glossaries still in MS. scattered in various libraries of Europe, especially in the Vatican, at Monte Cassino, Paris, Munich, Bern, the British Museum, Leiden, Oxford, Cambridge, &c. Much has already been done to make the material contained in these MSS. accessible in print, and much may yet be done with what is still unpublished, though we may find that the differences between the glossaries which often present themselves at first sight are mere differences in form introduced by successive more or less qualified copyists.Some Celtic (Breton, Cornish, Welsh, Irish) glossaries have been preserved to us, the particulars of which may be learnt from the publications of Whitley Stokes, Sir John Rhys, Kuno Meyer, L. C. Stern, G. I. Ascoli, Heinr. Zimmer, Ernst Windisch, Nigra, and many others; these are published separately as books or in Zeuss’sGrammatica Celtica, A. Kühn’sBeiträge zur vergleich. Sprachforschung, Zeitschr. für celtische Philologie, Archiv für Celtische Lexicographie, the Revue celtique, Transactions of the London Philological Society, &c.The first Hebrew author known to have used glosses was R. Gershom of Metz (1000) in his commentaries on the Talmud. But he and other Hebrew writers after him mostly used the Old French language (though sometimes also Italian, Slavonic, German) of which an example has been published by Lambert and Brandin, in theirGlossaire hébreu-français du XIIIesiècle: recueil de mots hébreux bibliques avec traduction française(Paris, 1905). See furtherThe Jewish Encyclopedia(New York and London, 1903), article “Gloss.”Authorities.—For a great part of what has been said above, the writer is indebted to G. Goetz’s article on “Latein. Glossographie” in Pauly’sRealencyklopädie. By the side of Goetz’sCorpusstands the great collection of Steinmeyer and Sievers,Die althochdeutschen Glossen(in 4 vols., 1879-1898), containing a vast number of (also Anglo-Saxon) glosses culled from Bible MSS. and MSS. of classical Christian authors, enumerated and described in the 4th vol. Besides the works of the editors of, or writers on, glosses, already mentioned, we refer here to a few others, whose writings may be consulted: Hugo Blümner;Catholicon Anglicum(ed. Hertage); De-Vit (at end of Forcellini’sLexicon); F. Deycks; Du Cange; Funck; J. H. Gallée (Altsächs. Sprachdenkm., 1894); Gröber; K. Gruber (Hauptquellen des Corpus, Épin. u. Erfurt Gloss., Erlangen, 1904); Hattemer; W. Heraeus (Die Sprache des Petronius und die Glossen, Leipzig, 1899); Kettner; Kluge; Krumbacher; Lagarde; Landgraf; Marx; W. Meyer-Lubke (“Zu den latein. Glossen” inWiener Stud.xxv. 90 sqq.); Henry Nettleship; Niedermann,Notes d’étymol. lat.(Macon, 1902),Contribut. à la critique des glosses latines(Neuchâtel, 1905); Pokrowskij; Quicherat; Otto B. Schlutter (many important articles inAnglia, Englische Studien, Archiv f. latein. Lexicographie, &c.); Schöll; Schuchardt; Leo Sommer; Stadler; Stowasser; Strachan; H. Sweet; Usener (Rhein. Mus.xxiii. 496, xxiv. 382); A. Way,Promptorium parvulorum sive clericorum(3 vols., London, 1843-1865); Weyman; Wilmanns (inRhein. Mus.xxiv. 363); Wölfflin inArch. für lat. Lexicogr.; Zupitza. Cf. further, the various volumes of the following periodicals:Romania;Zeitschr. für deutsches Alterthum;Anglia;Englische Studien;Journal of English and German Philology(ed. Cook and Karsten);Archiv für latein. Lexicogr., and others treating of philology, lexicography, grammar, &c.
Verrius Flaccus and his epitomists, Festus and Paulus, have preserved many treasures of early glossographers who are now lost to us. He copied Aelius Stilo (Reitzenstein, “Verr. Forsch.,” in vol. i. ofBreslauer philol. Abhandl., p. 88; Kriegshammer,Comm. phil. Ien.vii. 1. 74 sqq.), Aurelius Opilius, Ateius Philol., the treatiseDe obscuris Catonis(Reitzenstein,ib.56. 92). He often made use of Varro (Willers,De Verrio Flacco, Halle, 1898), though not of hisling. lat.(Kriegshammer, 74 sqq.); and was also acquainted with later glossographers. Perhaps we owe to him theglossae asbestos(Goetz,Corpus, iv.;id., Rhein. Mus.xl. 328). Festus was used by Ps.-Philoxenus (Dammann, “De Festo Ps.-Philoxeni auctore,”Comm. Ien.v. 26 sqq.), as appears from theglossae ab absens(Goetz, “De Astrabae Pl. fragmentis,”Ind. Ien., 1893, iii. sqq.). The distinct connexions with Nonius need not be ascribed to borrowing, as Plinius and Caper may have been used (P. Schmidt,De Non. Marc. auctt. gramm.145; Nettleship,Lect. and Ess. 229; Fröhde,De Non. Marc. et Verrio Flacco, 2; W. M. Lindsay, “Non. Marc.,”Dict. of Repub. Latin, 100, &c.).
Thebilingual(Gr.-Lat., Lat.-Gr.) glossaries also point to an early period, and were used by the grammarians (1) to explain the peculiarities (idiomata) of the Latin language by comparison with the Greek, and (2) for instruction in the two languages (Charis. 254. 9, 291. 7, 292. 16 sqq.; Marschall,De Q. Remmii P. libris gramm. 22; Goetz,Corp. gloss. lat.ii. 6).
For the purposes of grammatical instruction (Greek for the Romans, Latin for the Hellenistic world), we have systematic works, a translation of Dositheus and the so-calledHermeneutica, parts of which may be dated as early as the 3rd centuryA.D., and lexica (cf. Schoenemann,De lexicis ant.122; Knaack, inPhil. Rundsch., 1884, 372; Traube, inByzant. Ztschr.iii. 605; David,Comment. Ien.v. 197 sqq.).
The most important remains of bilingual glossaries are two well-known lexica; one (Latin-Greek), formerly attributed (but wrongly, see Rudorff, inAbh.. Akad. Berl., 1865, 220 sq.; Loewe,Prodr.183, 190; Mommsen,C.I.L.v. 8120; A. Dammann,De Festo Pseudo-philoxeni auctore, 12 sqq.; Goetz, Corp. ii. 1-212) to Philoxenus (consulA.D.525), clearly consists of two closely allied glossaries (containing glosses to Latin authors, as Horace, Cicero, Juvenal, Virgil, the Jurists, and excerpts from Festus), worked into one by some Greek grammarian, or a person who worked under Greek influence (his alphabet runs A, B, G, D, E, &c.); the other (Greek-Latin) is ascribed to Cyril (Stephanus says it was found at the end of some of his writings), and is considered to be a compilation of not later than the 6th century (Macrobius is used, and theCod. Harl., which is the source of all the other MSS., belongs to the 7th century); cf. Goetz,Corp.ii. 215-483, 487-506, praef.ibid.p. xx. sqq. Furthermore, the bilingual medico-botanic glossaries had their origin in old lists of plants, as Ps.-Apuleius in the treatiseDe herbarum virtutibus, and Ps.-Dioscorides (cf. M. Wellmann,Hermes, xxxiii. 360 sqq., who thinks that the latter work is based on Pamphilus,q.v.; Goetz, Corp. iii.); the glossary, entitledHermeneuma, printed from theCod. Vatic.reg. Christ. 1260, contains names of diseases.
Just as grammar developed, so we see the original form of the glosses extend. Ifmassucum edacemin Placidus indicates the original form, the allied gloss of Festus (masucium edacem a mandendo scilicet) shows an etymological addition. Another extension consists in adding special references to the original source, ase.g.at the glossOcrem(Fest. 181a. 17), which is taken from Ateius Philol. In this way collections arose like thepriscorum verborum cum exemplis, a title given by Fest. (218b. 10) to a particular work. Further theglossae veterum(Charis. 242. 10); theglossae antiquitatum(id.229. 30); theidonei vocum antiquarum enarratores(Gell. xviii. 6. 8); thelibri rerum verborumque veterum(id.xiii. 24. 25). L.Cincius, according to Festus (330b. 2), wroteDe verbis priscis; Santra,De antiquitate verborum(Festus 277a. 2).
Of Latin glossaries of the first four centuries of the Roman emperors few traces are left, if we except Verrius-Festus. Charis, 229. 30, speaks ofglossae antiquitatumand 242. 10 ofglossae veterum, but it is not known whether these glosses are identic, or in what relation they stand to theglossemata per litteras Latinas ordine composita, which were incorporated with the works of this grammarian according to the index in Keil, p. 6. Latin glosses occur in Ps.-Philoxenus, and Nonius must have used Latin glossaries; there exists aglossarium Plautinum(Ritschl,Op.ii. 234 sqq.), and the bilingual glossaries have been used by the later grammarian Martyrius; but of this early period we know by name only Fulgentius and Placidus, who is sometimes called Luctatius Placidus, by confusion with the Statius scholiast, with whom theglossae Placidihave no connexion. All that we know of him tends to show that he lived in North Africa (like Fulgentius and Nonius and perhaps Charisius) in the 6th century, from whence his glosses came to Spain, and were used by Isidore and the compiler of theLiber glossarum(see below). These glosses we know from (1) Codices Romani (15th and 16th century); (2) theLiber glossarum; (3) the Cod. Paris. nov. acquis. 1298 (saec. xi.), a collection of glossaries, in which the Placidus-glosses are kept separate from the others, and still retain traces of their original order (cf. the editions published by A. Mai,Class. auct.iii. 427-503, and Deuerling, 1875; Goetz,Corp.v.; P. Karl, “De Placidi glossis,”Comm. Ien.vii. 2. 99, 103 sqq.; Loewe,Gloss. Nom.86; F. Bücheler, inThesaur. gloss. emend.). His collection includes glosses from Plautus and Lucilius.
(Fabius Planciades) Fulgentius (c.A.D.468-533) wroteExpositio sermonum antiquorum(ed. Rud. Helm, Lips. 1898; cf. Wessner,Comment. Ien.vi. 2. 135 sqq.) in sixty-two paragraphs, each containing a lemma (sometimes two or three) with an explanation giving quotations and names of authors. Next to him come theglossae Nonianae, which arose from the contents of the various paragraphs in Nonius Marcellus’ work being written in the margin without the words of the text; these epitomized glosses were alphabetized and afterwards copied for other collections (see Goetz,Corp.v. 637 sqq.,id.v. Praef. xxxv.; Onions and Lindsay,Harvard Stud.ix. 67 sqq.; Lindsay,Nonii praef.xxi.). In a similar way arose theglossae Eucheriiorglossae spiritales secundum Eucherium episcopumfound in many MSS. (cf. K. Wotke,Sitz. Ber. Akad. Wien, cxv. 425 sqq.; = theCorpus Glossary, first part), which are an alphabetical extract from theformulae spiritalis intelligentiaeof St Eucherius, bishop of Lyons,c.434-450.7
Other sources were theDifferentiae, already known to Placidus and much used in the medieval glossaries; and theSynonyma Ciceronis; cf. Goetz, “Der Liber glossarum,” inAbhandl. der philol.-hist. Cl. der sächs. Gesellsch. d. Wiss., 1893, p. 215;id.inBerl. philol. Wochenschr., 1890, p. 195 sqq.; Beck, inWochenschr., p. 297 sqq., and Sittls,ibid.p. 267;Archiv f. lat. Lex.vi. 594; W. L. Mahne, (Leid. 1850, 1851); also various collections ofscholia. By the side of the scholiasts come the grammarians, as Charisius, or an ars similar to that ascribed to him; further, treatisesde dubiis generibus, thescriptores orthographici(especially Caper and Beda), and Priscianus, the chief grammarian of the middle ages (cf. Goetz inMélanges Boissier, 224).
During the 6th, 7th and 8th centuries glossography developed in various ways; old glossaries were worked up into new forms, or amalgamated with more recent ones. It ceased, moreover, to be exclusively Latin-Latin, and interpretations in Germanic (Old High German, Anglo-Saxon) and Romanic dialects took the place of or were used side by side with earlier Latin ones. The origin and development of the late classic and medieval glossaries preserved to us can be traced with certainty. While reading the manuscript texts of classical authors, the Bible or early Christian and profane writers, students and teachers, on meeting with any obscure or out-of-the-way words which they considered difficult to remember or to require elucidation, wrote above them, or in the margins, interpretations or explanations in more easy or better-known words. The interpretations written above the line are called “interlinear,” those written in the margins of the MSS. “marginal glosses.” Again, MSS. of the Bible or portions of the Bible were often provided with literal translations in the vernacular written above the lines of the Latin version (interlinear versions).
Of such glossed MSS. or translated texts, photographs may be seen in the various palaeographical works published in recent years; cf.The Palaeogr. Society, 1st ser. vol. ii. pls. 9 (Terentius MS. of 4th or 5th century, interlinear glosses) and 24 (Augustine’s epistles, 6th or 7th century, marginal glosses); see further, plates 10, 12, 33, 40, 50-54, 57, 58, 63, 73, 75, 80; vol. iii. plates 10, 24, 31, 39, 44, 54, 80.
From these glossed or annotated MSS. and interlinear versions glossaries were compiled; that is, the obscure and difficult Latin words, together with the interpretations, were excerpted and collected in separate lists, in the order in which they appeared, one after the other, in the MSS., without any alphabetical arrangement, but with the names of the authors or the titles of the books whence they were taken, placed at the head of each separate collection or chapter. In this arrangement each article by itself is called a gloss; when reference is made only to the word explained it is called thelemma, while the explanation is termed theinterpretamentum. In most cases the form of the lemma was retained just as it stood in its source, and explained by a single word (tesca:sancta, Varro vii. 10;clucidatus: suavis, id.vii. 107; cf. Isid.Etym.i. 30. 1, “quid enim illud sit in uno verbo positum declarat [scil.glossa] ut conticescere est tacere”), so that we meet with lemmata in the accusative, dative and genitive, likewise explained by words in the same cases; the forms of verbs being treated in the same way. Of this first stage in the making of glossaries, many traces are preserved, for instance, in the late 8th century Leiden Glossary (Voss. 69, ed. J. H. Hessels), where chapter iii. contains words or glosses excerpted from theLife of St Martinby Sulpicius Severus; chs. iv., v. and xxxv. glosses from Rufinus; chs. vi. and xl. from Gildas; chs. vii. to xxv. from books of the Bible (Paralipomenon; Proverbs, &c., &c.); chs. xxvi. to xlviii, from Isidore, the Vita S.Anthonii, Cassiodorus, St Jerome, Cassianus, Orosius, St Augustine, St Clement, Eucherius, St Gregory, the grammarians Donatus, Phocas, &c. (See also Goetz,Corp.v. 546. 23-547. 6. and i. 5-40 from Ovid’sMetam.; v. 657 from Apuleius,De deo Socratis; cf. Landgraf, inArch.ix. 174).
By a second operation the glosses came to be arranged inalphabeticalorder according to the first letter of the lemma, but still retained in separate chapters under the names of authors or the titles of books. Of thissecondstage the Leiden Glossary contains traces also: ch. i. (Verba de Canonibus) and ii. (Sermones de Regulis); see Goetz,Corp.v. 529 sqq. (from Terentius), iv. 427 sqq. (Virgil).
The third operation collected all the accessible glosses in alphabetical order, in the first instance according to the first letters of the lemmata. In this arrangement the names of the authors or the titles of the books could no longer be preserved, and consequently the sources whence the glosses were excerpted became uncertain, especially if the grammatical forms of the lemmata had been normalized.
A fourth arrangement collected the glosses according to the first two letters of the lemmata, as in the Corpus Glossary and in the still earlierCod. Vat.3321 (Goetz,Corp.iv. 1 sqq.), where even many attempts were made to arrange them according to the first three letters of the alphabet. A peculiar arrangement is seen in theGlossae affatim(Goetz,Corp.iv. 471 sqq.), where all words are alphabetized, first according to the initial letter of the word (a, b, c, &c.), and then further according to the firstvowelin the word (a, e, i, o, u).
No date or period can be assigned to any of the above stages or arrangements. For instance, the first and second are both found in the Leiden Glossary, which dates from the end of the 8th century, whereas the Corpus Glossary, written in the beginning of the same century, represents already the fourth stage.
For the purpose of identification titles have of late years been given to the various nameless collections of glosses, derived partly from their first lemma, partly from other characteristics, as glossaeabstrusae; glossaeabavus majorandminor; g.affatim; g.ab absens; g.abactor; g.Abba Pater; g.a, a; g.Vergilianae; g.nominum(Goetz,Corp.ii. 563, iv.); g. Sangallenses (Warren,Transact.Amer. Philol. Assoc.xv., 1885, p. 141 sqq.).
A chief landmark in glossography is represented by theOrigines(Etymologiae) of Isidore (d. 636), an encyclopedia in which he, like Cassiodorus, mixed human and divine subjects together. In many places we can trace his sources, but he also used glossaries. His work became a great mine for later glossographers. In the tenth book he deals with the etymology of many substantives and adjectives arranged alphabetically according to the first letter of the words, perhaps by himself from various sources. His principal source is Servius, then the fathers of the Church (Augustine, Jerome,Lactantius) and Donatus the grammarian. This tenth book was also copied and used separately, and mixed up with other works (cf. Loewe,Prodr.167. 21). Isidore’sDifferentiaehave also had a great reputation.
Next comes theLiber glossarum, chiefly compiled from Isidore, but all articles arranged alphabetically; its author lived in Spain c.A.D.690-750; he has been called Ansileubus, but not in any of the MSS., some of which belong to the 8th century; hence this name is suspected to be merely that of some owner of a copy of the book (cf. Goetz, “Der Liber Glossarum,” inAbhandl. der philol.-hist. Class, der kön. sächs. Ges.xiii., 1893;id., Corp.v., praef. xx. 161).
Here come, in regard to time, some Latin glossaries already largely mixed with Germanic, more especially Anglo-Saxon interpretations: (1) the Corpus Glossary (ed. J. H. Hessels), written in the beginning of the 8th century, preserved in the library of Corpus Christi College, Cambridge; (2) the Leiden Glossary (end of 8th century, ed. Hessels; another edition by Plac. Glogger), preserved in the Leiden MS. Voss. Qo. 69; (3) the Épinal Glossary, written in the beginning of the 9th century8and published in facsimile by the London Philol. Society from a MS. in the town library at Épinal; (4) theGlossae Amplonianae,i.e.three glossaries preserved in the Amplonian library at Erfurt, known as Erfurt1, Erfurt2and Erfurt3. The first, published by Goetz (Corp.v. 337-401; cf. also Loewe,Prodr.114 sqq.) with the various readings of the kindred Épinal, consists, like the latter, of different collections of glosses (also some from Aldhelm), some arranged alphabetically according to the first letter of the lemma, others according to the first two letters. The title of Erfurt2(incipit II. conscriptio glosarum in unam) shows that it is also a combination of various glossaries; it is arranged alphabetically according to the first two letters of the lemmata, and contains theaffatimandabavus maiorglosses, also a collection from Aldhelm; Erfurt3are theGlossae nominum, mixed also with Anglo-Saxon interpretations (Goetz,Corp.ii. 563). The form in which the three Erfurt glossaries have come down to us points back to the 8th century.
The first great glossary or collection of various glosses and glossaries is that of Salomon, bishop of Constance, formerly abbot of St Gall, who diedA.D.919. An edition of it in two parts was printedc.1475 at Augsburg, with the headlineSalemonis ecclesie Constantiensis episcopi glosse ex illustrissimis collecte auctoribus. The oldest MSS. of this work date from the 11th century. Its sources are theLiber glossarum(Loewe, Prodr. 234 sqq.), the glossary preserved in the 9th-century MS.Lat. Monac.14429 (Goetz, “Lib. Gloss.” 35 sqq.), and the great Abavus Gloss (id., ibid.p. 37;id., Corp.iv. praef. xxxvii.).
TheLib. glossarumhas also been the chief source for the important (but not original) glossary of Papias, ofA.D.1053 (cf. Goetz inSitz. Ber. Akad. Münch., 1903, p. 267 sqq., who enumerates eighty-seven MSS. of the 12th to the 15th centuries), of whom we only know that he lived among clerics and dedicated his work to his two sons. An edition of it was published at Milan “per Dominicum de Vespolate” on the 12th of December 1476; other editions followed in 1485, 1491, 1496 (at Venice). He also wrote a grammar, chiefly compiled from Priscianus (Hagen,Anecd. Helv.clxxix. sqq.).
The sameLib. gloss.is the source (1) for theAbba PaterGlossary (cf. Goetz,ibid.p. 39), published by G. M. Thomas (Sitz. Ber. Akad. Münch., 1868, ii. 369 sqq.); (2) the Greek glossaryAbsida lucida(Goetz,ib.p. 41); and (3) the Lat.-Arab. glossary in theCod. Leid. Scal. Orient.No. 231 (published by Seybold inSemit. Studien, Heft xv.-xvii., Berlin, 1900).
The Paulus-Glossary (cf. Goetz, “Der Liber Glossarum,” p. 215) is compiled from the second Salomon-Glossary (abacti magistratus), theAbavus majorand theLiber glossarum, with a mixture of Hebraica. Many of his glosses appear again in other compilations, as in the Cod. Vatic. 1469 (cf. Goetz,Corp.v. 520 sqq.), mixed up with glosses from Beda, Placidus, &c. (cf. a glossary published by Ellis inAmer. Journ. of Philol.vi. 4, vii. 3, containing besides Paulus glosses, also excerpts from Isidore; CambridgeJourn. of Philol.viii. 71 sqq., xiv. 81 sqq.).
Osbern of Gloucester (c.1123-1200) compiled the glossary entitledPanormia(published by Angelo Mai asThesaurus novus Latinitatis, from Cod. Vatic. reg. Christ. 1392; cf. W. Meyer,Rhein. Mus.xxix., 1874; Goetz inSitzungsber. sächs. Ges. d. Wiss., 1903, p. 133 sqq.;Berichte üb. die Verhandl. der kön. sächs. Gesellsch. der Wiss., Leipzig, 1902); giving derivations, etymologies, testimonia collected from Paulus, Priscianus, Plautus, Horace, Virgil, Ovid, Mart. Capella, Macrobius, Ambrose, Sidonius, Prudentius, Josephus, Jerome, &c., &c. Osbern’s material was also used by Hugucio, whose compendium was still more extensively used (cf. Goetz, l.c., p. 121 sqq., who enumerates one hundred and three MSS. of his treatise), and contains many biblical glosses, especially Hebraica, some treatises on Latin numerals, &c. (cf. Hamann,Weitere Mitteil. aus dem Breviloquus Benthemianus, Hamburg, 1882; A. Thomas, “Glosses provençales inéd.” inRomania, xxxiv. p. 177 sqq; P. Toynbee,ibid.xxv. p. 537 sqq.).
The great work of Johannes de Janua, entitledSumma quae vocatur catholicon, dates from the year 1286, and treats of (1) accent, (2) etymology, (3) syntax, and (4) so-called prosody,i.e.a lexicon, which also deals with quantity. It mostly uses Hugucio and Papias; its classical quotations are limited, except from Horace; it quotes the Vulgate by preference, frequently independently from Hugucio; it excerpts Priscianus, Donatus, Isidore, the fathers of the Church, especially Jerome, Gregory, Augustine, Ambrose; it borrows many Hebrew glosses, mostly from Jerome and the other collections then in use; it mentions theGraecismusof Eberhardus Bethuniensis, the works of Hrabanus Maurus, theDoctrinaleof Alexander de Villa Dei, and theAuroraof Petrus de Riga. Many quotations from theCatholiconin Du Cange are really from Hugucio, and may be traced to Osbern. There exist many MSS. of this work, and the Mainz edition of 1460 is well known (cf. Goetz inBerichte üb. die Verhandl. der kön. sächs. Gesellsch. der Wiss., Leipzig, 1902).
The gloss MSS. of the 9th and 10th centuries are numerous, but a diminution becomes visible towards the 11th. We then find grammatical treatises arise, for which also glossaries were used. The chief material was (1) theLiber glossarum; (2) the Paulus glosses; (3) theAbavus major; (4) excerpts from Priscian and glosses to Priscian; (5) Hebrew-biblical collections of proper names (chiefly from Jerome). After these comes medieval material, as thederivationeswhich are found in many MSS. (cf. Goetz inSitzungsber. sächs. Ges. d. Wiss., 1903, p. 136 sqq.; Traube inArchiv f. lat. Lex.vi. 264), containing quotations from Plautus, Ovid, Juvenal, Persius, Terence, occasionally from Priscian, Eutyches, and other grammarians, with etymological explanations. Thesederivationeswere the basis for the grammatical works of Osbern, Hugucio and Joannes of Janua.
A peculiar feature of the late middle ages are the medico-botanic glossaries based on the earlier ones (see Goetz,Corp.iii.). The additions consisted in Arabic words with Latin explanations, while Greek, Latin, Hebrew and Arabic, interchange with English, French, Italian and German forms. Of glossaries of this kind we have (1) theGlossae alphita(published by S. de Renzi in the 3rd vol. of theCollect. Salernitana, Naples, 1854, from two Paris MSS. of the 14th and 15th centuries, but some of the glosses occur already in earlier MSS.); (2)Sinonoma Bartholomei, collected by John Mirfeld, towards the end of the 14th century, ed. J. L. G. Mowat (Anecd. Oxon.i. 1, 1882, cf. Loewe,Gloss. Nom.116 sqq.); it seems to have used the same or some similar source as No. 1; (3) the compilations of Simon de Janua (Clavis sanationis, end of 13th century), and of Matthaeus Silvaticus (Pandectae medicinae, 14th century; cf. H. Stadler, “Dioscor. Longob.” inRoman. Forsch.x. 3. 371; Steinmeyer,Althochd. Gloss.iii.).
Of biblical glossaries we have a large number, mostly mixed with glosses on other, even profane, subjects, as Hebrew and other biblical proper names, and explanations of the text of the Vulgate in general, and the prologues of Hieronymus. So we have theGlossae veteris ac novi testamenti(beginning “Prologus graece latine praelocutio sive praefatio”) in numerous MSS. of the 9th to 14th centuries, mostly retaining the various books under separate headings (cf. Arevalo,Isid.vii. 407 sqq.; Loewe,Prodr.141; Steinmeyer iv. 459; S. Berger,De compendiis exegeticis quibusdam medii aevi, Paris, 1879). Special mention should be made of Guil. Brito, who lived about 1250, and compiled aSumma(beginning “difficiles studeo partes quas Biblia gestat Pandere”), contained in many MSS. especially in French libraries. ThisSummagave rise to theMammotrectusof Joh. Marchesinus, about 1300, of which we have editions printed in 1470, 1476, 1479, &c.
Finally we may mention such compilations as theSumma Heinrici; the work of Johannes de Garlandia, which he himself callsdictionarius(cf. Scheler inJahrb. f. rom. u. engl. Philol.vi., 1865, p. 142 sqq.); and that of Alexander Neckam (ib.vii. p. 60 sqq., cf. R. Ellis, inAmer. Journ. of Phil.x. 2); which are, strictly speaking, not glossographic. TheBreviloquusdrew its chief material from Papias, Hugucio, Brito, &c. (K. Hamann,Mitteil. aus dem Breviloquus Benthemianus, Hamburg, 1879;id.,Weitere Mitteil., &c., Hamburg, 1882); so also theVocabularium Ex quo; the variousGemmae;Vocabularia rerum(cf. Diefenbach,Glossar. Latino-Germanicum).
After the revival of learning, J. Scaliger (1540-1609) was the first to impart to glossaries that importance which they deserve (cf. Goetz, inSitzungsber. sächs. Ger. d. Wiss., 1888, p. 219 sqq.), and in his edition of Festus made great use of Ps.-Philoxenus, which enabled O. Müller, the later editor of Festus, to follow in his footsteps. Scaliger also planned the publication of aCorpus glossarum, and left behind a collection of glosses known asglossae Isidori(Goetz,Corp.v. p. 589 sqq.; id. inSitzungsber. sächs. Ges., 1888, p. 224 sqq.; Loewe,Prodr.23 sqq.), which occurs also in old glossaries, clearly in reference to the tenth book of theEtymologiae.
The study of glosses spread through the publication, in 1573, of the bilingual glossaries by H. Stephanus (Estienne), containing, besides the two great glossaries, also theHermeneumata Stephani, which is a recension of thePs.-Dositheana(republished Goetz,Corp.iii. 438-474), and theglossae Stephani, excerpted from a collection of theHermeneumata(ib.iii. 438-474).
In 1600 Bonav. Vulcanius republished the same glossaries, adding (1) the glossaeIsidori, which now appeared for the first time; (2) theOnomasticon; (3)notaeandcastigationes, derived from Scaliger (Loewe,Prodr.183).
In 1606 Carolus and Petrus Labbaeus published, with the effective help of Scaliger, another collection of glossaries, republished, in 1679, by Du Cange, after which the 17th and 18th centuries produced nofurther glossaries (Erasm. Nyerup published extracts from the Leiden Glossary, Voss. 69, in 1787,Symbolae ad Literat. Teut.), though glosses were constantly used or referred to by Salmasius, Meursius, Heraldus, Barth, Fabricius and Burman at Leiden, where a rich collection of glossaries had been obtained by the acquisition of the Vossius library (cf. Loewe,Prodr.168). In the 19th century came Osann’sGlossarii Latini specimen(1826); the glossographic publications of Angelo Mai (Classici auctores, vols. iii., vi., vii., viii., Rome, 1831-1836, containing Osbern’sPanormia, Placidus and various glosses from Vatican MSS.); Fr. Oehler’s treatise (1847) on theCod. Amplonianusof Osbern, and his edition of the three Erfurt glossaries, so important for Anglo-Saxon philology; in 1854 G. F. Hildebrand’sGlossarium Latinum(an extract fromAbavus minor), preserved in a Cod. Paris. lat. 7690; 1857, Thomas Wright’s vol. of Anglo-Saxon glosses, which were republished with others in 1884 by R. Paul Wülcker under the titleAnglo-Saxon and Old English Vocabularies(London, 2 vols., 1857); L. Diefenbach’s supplement to Du Cange, entitledGlossarium Latino-Germanicum mediae et infimae aetatis, containing mostly glosses collected from glossaries, vocabularies, &c., enumerated in the preface; Ritschl’s treatise (1870) on Placidus, which called forth an edition (1875) of Placidus by Deuerling; G. Loewe’sProdromus(1876), and other treatises by him, published after his death by G. Goetz (Leipzig, 1884); 1888, the second volume of Goetz’s own greatCorpus glossariorum Latinorum, of which seven volumes (except the first) had seen the light by 1907, the last two being separately entitledThesaurus glossarum emendatarum, containing many emendations and corrections of earlier glossaries by the author and other scholars; 1900, Arthur S. Napier,Old English Glosses(Oxford), collected chiefly from Aldhelm MSS., but also from Augustine, Avianus, Beda, Boethius, Gregory, Isidore, Juvencus, Phocas, Prudentius, &c.
There are a very great number of glossaries still in MS. scattered in various libraries of Europe, especially in the Vatican, at Monte Cassino, Paris, Munich, Bern, the British Museum, Leiden, Oxford, Cambridge, &c. Much has already been done to make the material contained in these MSS. accessible in print, and much may yet be done with what is still unpublished, though we may find that the differences between the glossaries which often present themselves at first sight are mere differences in form introduced by successive more or less qualified copyists.
Some Celtic (Breton, Cornish, Welsh, Irish) glossaries have been preserved to us, the particulars of which may be learnt from the publications of Whitley Stokes, Sir John Rhys, Kuno Meyer, L. C. Stern, G. I. Ascoli, Heinr. Zimmer, Ernst Windisch, Nigra, and many others; these are published separately as books or in Zeuss’sGrammatica Celtica, A. Kühn’sBeiträge zur vergleich. Sprachforschung, Zeitschr. für celtische Philologie, Archiv für Celtische Lexicographie, the Revue celtique, Transactions of the London Philological Society, &c.
The first Hebrew author known to have used glosses was R. Gershom of Metz (1000) in his commentaries on the Talmud. But he and other Hebrew writers after him mostly used the Old French language (though sometimes also Italian, Slavonic, German) of which an example has been published by Lambert and Brandin, in theirGlossaire hébreu-français du XIIIesiècle: recueil de mots hébreux bibliques avec traduction française(Paris, 1905). See furtherThe Jewish Encyclopedia(New York and London, 1903), article “Gloss.”
Authorities.—For a great part of what has been said above, the writer is indebted to G. Goetz’s article on “Latein. Glossographie” in Pauly’sRealencyklopädie. By the side of Goetz’sCorpusstands the great collection of Steinmeyer and Sievers,Die althochdeutschen Glossen(in 4 vols., 1879-1898), containing a vast number of (also Anglo-Saxon) glosses culled from Bible MSS. and MSS. of classical Christian authors, enumerated and described in the 4th vol. Besides the works of the editors of, or writers on, glosses, already mentioned, we refer here to a few others, whose writings may be consulted: Hugo Blümner;Catholicon Anglicum(ed. Hertage); De-Vit (at end of Forcellini’sLexicon); F. Deycks; Du Cange; Funck; J. H. Gallée (Altsächs. Sprachdenkm., 1894); Gröber; K. Gruber (Hauptquellen des Corpus, Épin. u. Erfurt Gloss., Erlangen, 1904); Hattemer; W. Heraeus (Die Sprache des Petronius und die Glossen, Leipzig, 1899); Kettner; Kluge; Krumbacher; Lagarde; Landgraf; Marx; W. Meyer-Lubke (“Zu den latein. Glossen” inWiener Stud.xxv. 90 sqq.); Henry Nettleship; Niedermann,Notes d’étymol. lat.(Macon, 1902),Contribut. à la critique des glosses latines(Neuchâtel, 1905); Pokrowskij; Quicherat; Otto B. Schlutter (many important articles inAnglia, Englische Studien, Archiv f. latein. Lexicographie, &c.); Schöll; Schuchardt; Leo Sommer; Stadler; Stowasser; Strachan; H. Sweet; Usener (Rhein. Mus.xxiii. 496, xxiv. 382); A. Way,Promptorium parvulorum sive clericorum(3 vols., London, 1843-1865); Weyman; Wilmanns (inRhein. Mus.xxiv. 363); Wölfflin inArch. für lat. Lexicogr.; Zupitza. Cf. further, the various volumes of the following periodicals:Romania;Zeitschr. für deutsches Alterthum;Anglia;Englische Studien;Journal of English and German Philology(ed. Cook and Karsten);Archiv für latein. Lexicogr., and others treating of philology, lexicography, grammar, &c.
(J. H. H.)
1The history of the literary gloss in its proper sense has given rise to the common English use of the word to mean an interpretation, especially in a disingenuous, sinister or false way; the form “gloze,” more particularly associated with explaining away, palliating or talking speciously, is simply an alternative spelling. The word has thus to some extent influenced, or been influenced by, the meaning of the etymologically different “gloss” = lustrous surface (from the same root as “glass”; cf. “glow”), in its extended sense of “outward fair seeming.”2See Matthaei,Glossaria Graeca(Moscow, 1774/5).3See Labbé,Veteres glossae verborum juris quae passim in Basilicis reperiuntur(1606); Otto,Thesaurus juris Romani, iii. (1697); Stephens,Thesaurus linguae Graecae, viii. (1825).4See Biener,Geschichte der Novellen, p. 229 sqq.5Irnerius himself is with some probability believed to have been the author of the Brachylogus (q.v.).6Thus Fil. Villani (De origine civitatis Florentiae, ed. 1847, p. 23), speaking of the Glossator Accursius, says of the Glossae that “tantae auctoritatis gratiaeque fuere, ut omnium consensu publice approbarentur, et reiectis aliis, quibuscumque penitus abolitis, solae juxta textum legum adpositae sunt et ubique terrarum sine controversia pro legibus celebrantur, ita ut nefas sit, non secus quam textui, Glossis Accursii contraire.” For similar testimonies see Bayle’sDictionnaire, s.v. “Accursius,” and Rudorff,Röm. Rechtsgeschichte, i. 338 (1857).7The so-calledMalbergglosses, found in various texts of the Lex Salica, are not glosses in the ordinary sense of the word, but precious remains of the parent of the present literary Dutch, namely, the Low German dialect spoken by the Salian Franks who conquered Gaul from the Romans at the end of the 5th century. It is supposed that the conquerors brought their Frankish law with them, either written down, or by oral tradition; that they translated it into Latin for the sake of the Romans settled in the country, and that the translators, not always knowing a proper Latin equivalent for certain things or actions, retained in their translations the Frankish technical names or phrases which they had attempted to translate into Latin. E.g. in chapter ii., by the side of “porcellus lactans” (a sucking-pig), we find the Frankish “chramnechaltio,” lit. a stye-porker. The person who stole such a pig (still kept in an enclosed place, in a stye) was fined three times as much as one who stole a “porcellus de campo qui sine matre vivere possit,” as the Latin text has it, for which the Malberg technical expression appears to have beeningymus, that is, a one year (winter) old animal,i.e.a yearling. Nearly all these glosses are preceded by “mal” or “malb,” which is thought to be a contraction for “malberg,” the Frankish for “forum.” The antiquity and importance of these glosses for philology may be realized from the fact that the Latin translation of the Lex Salica probably dates from the latter end of the 5th century. For further information cf. Jac. Grimm’s preface to Joh. Merkel’s ed. (1850), and H. Kern’s notes to J. H. Hessels’s ed. (London, 1880) of the Lex Salica.8Anglo-Saxon scholars ascribe an earlier date to the text of the MS. on account of certain archaisms in its Anglo-Saxon words.
1The history of the literary gloss in its proper sense has given rise to the common English use of the word to mean an interpretation, especially in a disingenuous, sinister or false way; the form “gloze,” more particularly associated with explaining away, palliating or talking speciously, is simply an alternative spelling. The word has thus to some extent influenced, or been influenced by, the meaning of the etymologically different “gloss” = lustrous surface (from the same root as “glass”; cf. “glow”), in its extended sense of “outward fair seeming.”
2See Matthaei,Glossaria Graeca(Moscow, 1774/5).
3See Labbé,Veteres glossae verborum juris quae passim in Basilicis reperiuntur(1606); Otto,Thesaurus juris Romani, iii. (1697); Stephens,Thesaurus linguae Graecae, viii. (1825).
4See Biener,Geschichte der Novellen, p. 229 sqq.
5Irnerius himself is with some probability believed to have been the author of the Brachylogus (q.v.).
6Thus Fil. Villani (De origine civitatis Florentiae, ed. 1847, p. 23), speaking of the Glossator Accursius, says of the Glossae that “tantae auctoritatis gratiaeque fuere, ut omnium consensu publice approbarentur, et reiectis aliis, quibuscumque penitus abolitis, solae juxta textum legum adpositae sunt et ubique terrarum sine controversia pro legibus celebrantur, ita ut nefas sit, non secus quam textui, Glossis Accursii contraire.” For similar testimonies see Bayle’sDictionnaire, s.v. “Accursius,” and Rudorff,Röm. Rechtsgeschichte, i. 338 (1857).
7The so-calledMalbergglosses, found in various texts of the Lex Salica, are not glosses in the ordinary sense of the word, but precious remains of the parent of the present literary Dutch, namely, the Low German dialect spoken by the Salian Franks who conquered Gaul from the Romans at the end of the 5th century. It is supposed that the conquerors brought their Frankish law with them, either written down, or by oral tradition; that they translated it into Latin for the sake of the Romans settled in the country, and that the translators, not always knowing a proper Latin equivalent for certain things or actions, retained in their translations the Frankish technical names or phrases which they had attempted to translate into Latin. E.g. in chapter ii., by the side of “porcellus lactans” (a sucking-pig), we find the Frankish “chramnechaltio,” lit. a stye-porker. The person who stole such a pig (still kept in an enclosed place, in a stye) was fined three times as much as one who stole a “porcellus de campo qui sine matre vivere possit,” as the Latin text has it, for which the Malberg technical expression appears to have beeningymus, that is, a one year (winter) old animal,i.e.a yearling. Nearly all these glosses are preceded by “mal” or “malb,” which is thought to be a contraction for “malberg,” the Frankish for “forum.” The antiquity and importance of these glosses for philology may be realized from the fact that the Latin translation of the Lex Salica probably dates from the latter end of the 5th century. For further information cf. Jac. Grimm’s preface to Joh. Merkel’s ed. (1850), and H. Kern’s notes to J. H. Hessels’s ed. (London, 1880) of the Lex Salica.
8Anglo-Saxon scholars ascribe an earlier date to the text of the MS. on account of certain archaisms in its Anglo-Saxon words.
GLOSSOP,a market town and municipal borough, in the High Peak parliamentary division of Derbyshire, England, on the extreme northern border of the county; 13 m. E. by S. of Manchester by the Great Central railway. Pop. (1901) 21,526. It is the chief seat of the cotton manufacture in Derbyshire, and it has also woollen and paper mills, dye and print works, and bleaching greens. The town consists of three main divisions, the Old Town (or Glossop proper), Howard Town (or Glossop Dale) and Mill Town. An older parish church was replaced by that of All Saints in 1830; there is also a very fine Roman Catholic church. In the immediate neighbourhood is Glossop Hall, the seat of Lord Howard, lord of the manor, a picturesque old building with extensive terraced gardens. On a hill near the town is Melandra Castle, the site of a Roman fort guarding Longdendale and the way into the hills of the Peak District. In the neighbourhood also a great railway viaduct spans the Dinting valley with sixteen arches. To the north, in Longdendale, there are five lakes belonging to the water-supply system of Manchester, formed by damming the Etherow, a stream which descends from the high moors north-east of Glossop. The town is governed by a mayor, 6 aldermen and 18 councillors. Area, 3052 acres.
Glossop was granted by Henry I. to William Peverel, on the attainder of whose son it reverted to the crown. In 1157 it was gifted by Henry II. to the abbey of Basingwerk. Henry VIII. bestowed it on the earl of Shrewsbury. It was made a municipal borough in 1866.
GLOUCESTER, EARLS AND DUKES OF.The English earldom of Gloucester was held by several members of the royal family, including Robert, a natural son of Henry I., and John, afterwards king, and others, until 1218, when Gilbert de Clare was recognized as earl of Gloucester. It remained in the family of Clare (q.v.) until 1314, when another Earl Gilbert was killed at Bannockburn; and after this date it was claimed by various relatives of the Clares, among them by the younger Hugh le Despenser (d. 1326) and by Hugh Audley (d. 1347), both of whom had married sisters of Earl Gilbert. In 1397 Thomas le Despenser (1373-1400), a descendant of the Clares, was created earl of Gloucester; but in 1399 he was degraded from his earldom and in January 1400 was beheaded.
The dukedom dates from 1385, when Thomas of Woodstock, a younger son of Edward III., was created duke of Gloucester, but his honours were forfeited when he was found guilty of treason in 1397. The next holder of the title was Humphrey, a son of Henry IV., who was created duke of Gloucester in 1414. He died without sons in 1447, and in 1461 the title was revived in favour of Richard, brother of Edward IV., who became king as Richard III. in 1483.
In 1659 Henry (1639-1660), a brother of Charles II., was formally created duke of Gloucester, a title which he had borne since infancy. This prince, sharing the exile of the Stuarts, had incensed his mother, Queen Henrietta Maria, by his firm adherence to the Protestant religion, and had fought among the Spaniards at Dunkirk in 1658. Having returned to England with Charles II., he died unmarried in London on the 13th of September 1660. The next duke was William (1689-1700), son of the princess Anne, who was, after his mother, the heir to the English throne, and who was declared duke of Gloucester by his uncle, William III., in 1689, but no patent for this creation was ever passed. William died on the 30th of July 1700, and again the title became extinct.
Frederick Louis, the eldest son of George II., was known for some time as duke of Gloucester, but when he was raised to the peerage in 1726 it was as duke of Edinburgh only. In 1764 Frederick’s third son, William Henry (1743-1805), was created duke of Gloucester and Edinburgh by his brother, George III. This duke’s secret marriage with Maria (d. 1807), an illegitimate daughter of Sir Edward Walpole and widow of James, 2nd Earl Waldegrave, in 1766, greatly incensed his royal relatives and led to his banishment from court. Gloucester died on the 25th of August 1805, leaving an only son, William Frederick (1776-1834), who now became duke of Gloucester and Edinburgh. The duke, who served with the British army in Flanders, married his cousin Mary (1776-1857), a daughter of George III. He died on the 30th of November 1834, leaving no children, and hiswidow, the last survivor of the family of George III., died on the 30th of April 1857.
GLOUCESTER, GILBERT DE CLARE,Earl of(1243-1295), was a son of Richard de Clare, 7th earl of Gloucester and 8th earl of Clare, and was born at Christchurch, Hampshire, on the 2nd of September 1243. Having married Alice of Angoulême, half-sister of king Henry III., he became earl of Gloucester and Clare on his father’s death in July 1262, and almost at once joined the baronial party led by Simon de Montfort, earl of Leicester. With Simon Gloucester was at the battle of Lewes in May 1264, when the king himself surrendered to him, and after this victory he was one of the three persons selected to nominate a council. Soon, however, he quarrelled with Leicester. Leaving London for his lands on the Welsh border he met Prince Edward, afterwards king Edward I., at Ludlow, just after his escape from captivity, and by his skill contributed largely to the prince’s victory at Evesham in August 1265. But this alliance was as transitory as the one with Leicester. Gloucester took up the cudgels on behalf of the barons who had surrendered at Kenilworth in November and December 1266, and after putting his demands before the king, secured possession of London. This happened in April 1267, but the earl quickly made his peace with Henry III. and with Prince Edward, and, having evaded an obligation to go on the Crusade, he helped to secure the peaceful accession of Edward I. to the throne in 1272. Gloucester then passed several years in fighting in Wales, or on the Welsh border; in 1289 when the barons were asked for a subsidy he replied on their behalf that they would grant nothing until they saw the king in person (nisi prius personaliter viderent in Anglia faciem regis), and in 1291 he was fined and imprisoned on account of his violent quarrel with Humphrey de Bohun, earl of Hereford. Having divorced his wife Alice, he married in 1290 Edward’s daughter Joan, or Johanna (d. 1307). Earl Gilbert, who is sometimes called the “Red,” died at Monmouth on the 7th of December 1295, leaving in addition to three daughters a son, Gilbert, earl of Gloucester and Clare, who was killed at Bannockburn.
See C. Bémont,Simon de Montfort, comte de Leicester(1884), and G. W. Prothero,Simon de Montfort(1877).
See C. Bémont,Simon de Montfort, comte de Leicester(1884), and G. W. Prothero,Simon de Montfort(1877).
GLOUCESTER, HUMPHREY,Duke of(1391-1447), fourth son of Henry IV. by Mary de Bohun, was born in 1391. He was knighted at his father’s coronation on the 11th of October 1399, and created duke of Gloucester by Henry V. at Leicester on the 16th of May 1414. He served in the war next year, and was wounded at Agincourt, where he owed his life to his brother’s valour. In April 1416 Humphrey received the emperor Sigismund at Dover and, according to a 16th-century story, did not let him land till he had disclaimed all title to imperial authority in England. In the second invasion of France Humphrey commanded the force which during 1418 reduced the Cotentin and captured Cherbourg. Afterwards he joined the main army before Rouen, and took part in subsequent campaigns till January 1420. He then went home to replace Bedford as regent in England, and held office till Henry’s own return in February 1421. He was again regent for his brother from May to September 1422.
Henry V. measured Humphrey’s capacity, and by his will named him merely deputy for Bedford in England. Humphrey at once claimed the full position of regent, but the parliament and council allowed him only the title of protector during Bedford’s absence, with limited powers. His lack of discretion soon justified this caution. In the autumn of 1422 he married Jacqueline of Bavaria, heiress of Holland, to whose lands Philip of Burgundy had claims. Bedford, in the interest of so important an ally, endeavoured vainly to restrain his brother. Finally in October 1424 Humphrey took up arms in his wife’s behalf, but after a short campaign in Hainault went home, and left Jacqueline to be overwhelmed by Burgundy. Returning to England in April 1425 he soon entangled himself in a quarrel with the council and his uncle Henry Beaufort, and stirred up a tumult in London. Open war was averted only by Beaufort’s prudence, and Bedford’s hurried return. Humphrey had charged his uncle with disloyalty to the late and present kings. With some difficulty Bedford effected a formal reconciliation at Leicester in March 1426, and forced Humphrey to accept Beaufort’s disavowal. When Bedford left England next year Humphrey renewed his intrigues. But one complication was removed by the annulling in 1428 of his marriage with Jacqueline. His open adultery with his mistress, Eleanor Cobham, also made him unpopular. To check his indiscretion the council, in November 1429, had the king crowned, and so put an end to Humphrey’s protectorate. However, when Henry VI. was soon afterwards taken to be crowned in France, Humphrey was made lieutenant and warden of the kingdom, and thus ruled England for nearly two years. His jealousy of Bedford and Beaufort still continued, and when the former died in 1435 there was no one to whom he would defer. The defection of Burgundy roused English feeling, and Humphrey won popularity as leader of the war party. In 1436 he commanded in a short invasion of Flanders. But he had no real power, and his political importance lay in his persistent opposition to Beaufort and the councillors of his party. In 1439 he renewed his charges against his uncle without effect. His position was further damaged by his connexion with Eleanor Cobham, whom he had now married. In 1441 Eleanor was charged with practising sorcery against the king, and Humphrey had to submit to see her condemned, and her accomplices executed. Nevertheless, he continued his political opposition, and endeavoured to thwart Suffolk, who was now taking Beaufort’s place in the council, by opposing the king’s marriage to Margaret of Anjou. Under Suffolk’s influence Henry VI. grew to distrust his uncle altogether. The crisis came in the parliament of Bury St Edmunds in February 1447. Immediately on his arrival there Humphrey was arrested, and four days later, on the 23rd of February, he died. Rumour attributed his death to foul play. But his health had been long undermined by excesses, and his end was probably only hastened by the shock of his arrest.
Humphrey was buried at St Albans Abbey, in a fine tomb, which still exists. He was ambitious and self-seeking, but unstable and unprincipled, and, lacking the fine qualities of his brothers, excelled neither in war nor in peace. Still he was a cultured and courtly prince, who could win popularity. He was long remembered as the good Duke Humphrey, and in his lifetime was a liberal patron of letters. He had been a great collector of books, many of which he presented to the university of Oxford. He contributed also to the building of the Divinity School, and of the room still called Duke Humphrey’s library. His books were dispersed at the Reformation and only three volumes of his donation now remain in the Bodleian library. Titus Livius, an Italian in Humphrey’s service, wrote a life of Henry V. at his patron’s bidding. Other Italian scholars, as Leonardo Aretino, benefited by his patronage. Amongst English men of letters he befriended Reginald Pecock, Whethamstead of St Albans, Capgrave the historian, Lydgate, and Gilbert Kymer, who was his physician and chancellor of Oxford university. A popular error found Humphrey a fictitious tomb in St Paul’s Cathedral. The adjoining aisle, called Duke Humphrey’s Walk, was frequented by beggars and needy adventurers. Hence the 16th-century proverb “to dine with Duke Humphrey,” used of those who loitered there dinnerless.