tes on Abbreviations in Greek Manuscripts BY T. W. ALLEN QUEEN'S COLLEGE, OXFORD CRAVEN FELLOW WITH ELEVEN PAGES OF FACSIMILES BY PHOTOLITHOGRAPHY Oxford AT THE CLARENDON PRESS M DCCC LXXXIX Price Five Shillings. TO ALFRED GOODWIN Digitized by the Internet Archive in 2013 http://archive.org/details/notesabbrevigrOOalle Notes on Abbreviations in Greek Manuscripts BY T. W. ALLEN queen's college, oxford craven fellow WITH ELEVEN PAGES OF FACSIMILES BY PHOTOLITHOGRAPHY Oxford AT THE CLARENDON PRESS M DCCC LXXXIX [ All rights reserved] HENRY FROWDE Oxford University Press Warehouse Amen Corner, E.C. PRINTEO IN ENGLANQ. NO TE S ON ABBREVIA TIONS IN GREEK MANUSCRIPTS. -M- \ sentence that occurs iii the late Charles Graux' review ■*■-*- of Lehmanns Die tachygraphischen Abkilrzungen will explain the purpose of this pamphlet. He says, 41 est absolu- ment indispensable que beaucoup d' observations exactes soient d'abord publiees avant que, a l'exemple du natural- iste qui dresse comparativement un tableau de la faune ou de la flore des divers regions du globe, les paleographes puissent un jour etablir l'etat des abreviations usitees dans le monde byzantin de siecle en siecle V The science of Greek Palaeography, or at least that portion of it which deals with contractions, has reached a point at which what is necessary for its progress is not the production of all- inclusive handbooks with an immediate practical aim, where conclusions are laid down with all possible definiteness and width, but rather a series of observations of the actual usage of manuscripts, noted with all available accuracy by investigators whose occupations have given them familiarity with the ways of scribes and the possibilities of Greek writing. Such observations indeed must be classified and brought into relations with one another, and hypotheses may be suggested to explain the facts observed ; but, as with any other yet undeveloped science, the first and main object must be the collection of fresh evidence. How little has been done towards such collection is known to anyone who has tried to find definite information upon the 1 Revue Critique, 1880, Notices bibliographiques, etc., p. 168. B 2 ABBREVIATIONS IN GREEK MANUSCRIPTS. origin and history of any single Greek abbreviation. Com- paratively little stress is laid on the subject of compendia in Montfaucon's great book, and Bast's well-known Commentatio Palaeographica, though stimulating in the highest degree, and the work of one whose palaeographical knowledge can seldom have been surpassed, has a directly practical inten- tion, and moreover hardly recognises the principle of chrono- logical development in the history of a compendium. The recognition and application of such principles to compendia forms the cardinal merit of Lehmann's handbook, a work that with all its industry and system is in point of actual palaeographical expertness far behind the Commentatio Palaeograjohica ; both Graux and Vitelli have pointed out, with indisputable truth, the impossibility of deriving general conclusions in palaeography from the compilation of fac- similes. Professor Vitelli however has at length given us the first instalments of a treatise, that, independent of the actual value of its results, is a model in all respects of what such a work should be. The Museo Italiano, part I. pp. 9-15, 32 ; II. 168-173 contains a great number of his observations upon codices in the Laurentian and else- where. I shall be more than gratified if this pamphlet, which owes so much to the Spicilegio Fiorentino, be thought a not unworthy companion to it. The bulk of the material presented here is taken from manuscripts in the Bodleian and the British Museum ; but a tour in France and Italy in the early part of last year (1888), undertaken under the Craven Trust, has enabled a number of additional examples to be added to those already collected. In arranging the compendia in alpha- betical order I have thought to consult the convenience of those looking for examples of any one in particular; at the end are collected some instances of tachygraphy which are grouped under the several manuscripts. I have endeavoured in discussing the various forms to avoid the faults urged against others ; where hypotheses are ventured, it is as hypotheses that they are given, and with the A. 3 knowledge that a little increase in our information may overset them ; I shall feel no particular shame if such a fate befalls one or two of my ' combinations ' : caedimus inque vicem jpraebemus crura sagittis is a line that every palaeo- grapher should accept. The plates have been produced by a photographic process at the University Press, and will, it is hoped, be thought more successful than previous reproductions of drawings. I have finally the pleasant duty of thanking many librarians, at home and abroad, to whose kindness the possi- bility of making this collection has been due : to some of them my thanks have been already elsewhere given : here I gladly record my obligation to the Kev. Gio. Bollig at the Vatican, the Rev. Antonio Ceriani at the Ambrosiana, the Abate Anziani at the Laurenziana, Conte Soranzo at Venice, M. Henri Omont at Paris, Bodley's Librarian and Mr. Madan in Oxford, and lastly to the Principal Librarian of the British Museum, for constant encouragement and advice. A. Alpha, it is well known, is properly represented by a horizontal stroke, improperly by a horizontal stroke dotted ; the latter sign properly denotes to, and in this sense is often found, though probably not so often as the dotted t; this abuse of the two dots, most frequent in combination with t, still occurs freely with other letters ; lastly, there are many mss. which make use of the notation both in the original and in the illegitimate sense. I give examples of these four cases : (i) the simple horizontal stroke appears to be exclusively used in the scholia of the Arethas-mss. * ; 1 By the 'Arethas-mss.' I mean the manuscripts which are known to have belonged to Arethas, deacon of Patrae, and afterwards archbishop of Caesarea, and which contain large quantities of scholia, apparently in the same hand : I here refer to five — the D'Orville Euclid (888), Clarke Plato (896), Lucian Harleian 5694 (undated), Aristotle Urbinas 35 (undated), Clement Paris grec 451 (914); cf. generally the Observationes Palaeographicae of E. Maass in the 'Melanges Graux,' Paris, 1880, p. 749 sq., and Vitelli Collezione Fiorentma fasc. iv. pt. 1, where it is shewn that Laur. 60, 3 (Aristides) is in the hand of the Clarke Plato. A certain resemblance also, so far as the scholia are con- cerned, is to be seen in the mss. Mutin. 126 (Clement Alex.) and Vallicell. F. 10 (Canones eccl.). B 2 4 ABBREVIATIONS IN GREEK MANUSCRIPTS. of so large an extent of scholia it is impossible to speak certainly, but at least the tachygraphic dots, whether as roc or a, are very rare in these mss. : cf. oxHjua eujudeeia from Lucian. The dots may be more certainly said not to appear in the Gospels Bodl. Auct. E. 5. 11 (s. X-XI), though the total amount of abbreviation here is comparatively small : cf. paoiAeiav napa 1 . (2) mss. in which the dotted stroke occurs and is always tcx, are Grotta Ferrata B. a. iii. (s. XII) navia Ta koto, unooTcxoeooc, juaTCuoTHTa ; Epistles, Bodl. Auct. E. 5. 9. (s. X) KaTa, dnaj-opeuovTa, with pdGKavoc, dvTiKaeiGTajuevoov : Philo, Selden supr. 12 (s. X) exovTa, toOto. (3) The most frequent case is that in which a manuscript uses at one and the same time the dotted stroke and the dotted t to ex- press Ta : such are the Paris Demosthenes 2 (Z, grec 2934, s. X) GujunepaajucfTa, Kara, but npoc Ta j Clement Mutin. 126 (s. X) evsTaaeic, to; Nonnus Add. mss. 1 8231 (a. 972) to €TaEev Ta; Vat. 1982 3 (s. X) to ogtcx ; Iliad Venetus A enerra (schol.) eujuoAeovTa (text) ; Psalter, Bodl. Auct. T. 4. 19 (s. X) exovTa, to; Aristides Vat. 1298 (s. X-XI) 4 touto toHic ; Hermogenes Paris grec 1983 (s. XI-XII) 5 rpajujuaTa, to; S. Maximus Angelic. T. 1. 8 (s. XI) ndvTa, ovto ; S. Maximus Mutin. 12 (s. XII) voHjuaTa bo£a^ovTa, Ta nparjuaTa ; Palladius Aed. Christ. 70 (a. 1104) 6 neipd^eTai, jueTa. Clarke 12 (Greg. Naz. s. X) combines the simple stroke for alpha with the dotted t for Ta, cf. koto Ta£iv, napaGaAacoiov. It is in anyone's power to extend this list. (4) The use of the double dots in combination with other consonants than t is fairly common ; it is most thoroughly carried out perhaps in the school of S. Nilo ; cf. AaoG from B. a. xix (a. 965, hand 1 A rather late instance of the simple stroke is noWaKis from the ms. Turin B. 1. 22 (a. 1 149). Cf. also fyikayaOov from Galen H. 45 Arch. S. Petr. (s. XII-XIII). 2 See the facsimile, Pal. Soc. 1886, 2nd series, pt. 3; the examples come from the scholia. 3 See the description of this ms. infra p. 34. 4 I am inclined to put this ms. earlier than M. de Nolhac (Fulvio Orsini, p. 171). Not X, as Bast. 6 Not 1 107 as the Catalogue. Al— All. 5 of S. Nilo), jBapei oafBpaTOO, aAAa, napd, djuapTHjuaTCov, 90apTHv from B. a. i (a. 986, hand of Paul), dnoedvei onepjua from Angel. B. 3. 11 lm . the use of the dots for the syllable ap (already fully rendered by the compendium) in such words as juapTuc, emeappHoeic from the London Nonnus, is a characteristic of the school ; the ring in the compendium is turned indifferently up or down. Other mss. where the dots are thus freely combined with consonants are Nonnus Paris suppl. grec 469 A (a. 986) to beojua ; Vat. 1982 evrauGa, naAaiaavTec paGoc ; Vat. 1 298 napabeifjuanKooc ; and similarly the Hermogenes Par. grec 1983. Al. The ordinary sign for the diphthong cu hardly needs fresh exemplification 2 ; it is constant in the Arethas- scholia, and, among other tenth-century mss., in Clarke 1 2 and Mutin. 126. The t achy graphic form has not met me outside the Grotta Ferrata school (q. v.) 3 . A curious form occurs repeatedly in Barocc. 26 (Canons, &c, s. XI ineunt.) ; cf. eEaj-opd^ovTca f. 209 r., nopeuovTai f. 346 r., KccrexovTai f. 347 r. : it is not unlike the sign which Vitelli (p. 12 n. 8) explains as a prolongation of the tachygraphic symbol ; but I think it as likely that it may be merely the sign for e used by itacism ; cf. binnoTe from the same ms. All. Lehmann's remarks upon the comparatively late origin of the double apostrophe for otic are just ; the xirethas-mss. use no single sign for the syllable, but render it by the sign for ai with sigma attached : so dpedic, tcxTc from the D'Orv. Euclid. At the same time the sign must have been in existence by the beginning of the 1 Lehmann well illustrates the use from Nonnus, Add. ms. 18231. 2 The separation of the sign from its preceding consonant, which Prof. Vitelli (p. 172 n. 3) seems to think rare, is common in the Arethas-mss. : cf. yeoopeTpais, nXcvpai D'Orv. Euclid, (paiverai Plato, v Par. 3032. 3 Cf. also dno o-reprja-is [sic] Par. 990, dno Vat. 1316 (s. XIII). 8 ABBREVIATIONS IN GREEK MANUSCRIPTS. 35 I would not be so positive K The Grotta Ferrata school on the other hand have the sign passim, with and without the superfluous dots, (v. under A), with the ring above or below, the stroke horizontal or slanting, in the middle or at the end of words : cf. pdppapov, napeAAHoiv from Nonnus, 9apjiiaKov 9GapTHv from Isidore, HjuapTHKooi 9peap from Angel. B. 3. 11, dcpeapTov from G. F. B. a. iii (s. XII) ; the simple form occurs also in Vat. 1982, ajuapTiccv, auTapKcoc, that with the dots in Mutin. 12 (s. XII) oapKoc, KaeapeevTec. Other examples are ovap, e£ dprouc from Mutin. 126 (s. X), dvd£apxoc, Hnap from Clarke 12, ndpeevou from Bodl. Auct. E. 5. 11 (s. X), djuapToAouc Kapnov from Laud. gr. 1 (s. XI-XII), 6 judpKoc from Vallicell. E. 40 (Oaten, s. XI), avajuapTHTOuc from Epictetus Bodl. Misc. 251 (s. XII), beAeap, undpxeiv from Neap. II. A a. 22 (s. XI-XII). APA. Vitelli's remark that this compendium, though possibly rare on the whole, is frequent in mss. of a particular class, is illustrated by the examples that I have here to give. In the D'Orville Euclid it is naturally frequent, without accent as a rule in the text, whose scribe Stephanus very largely omits both accents and breathings ; with aocent and breath- ing in the scholia : cf. nos. 1, 2 ; it is found also in the scholia of the other Arethas-mss., e.g. no. 3 from the Clarke Plato, no. 4 from Urbinas 35 : among mathematical mss. it occurs in the mathematical scholia to the Anthology (Paris suppl. grec 384, f. 639, &c, cf. no. 5), in Euclid Laur. 28, 3 (no. 6), both round and angular forms in Euclid Bologn. Archigin. A. 1. 18 (nos. 7, 8). The form is frequent in Bodl. Misc. 251, and in the sense of dpa, cf. no. 9. As a part of napd, both shapes are found in the Paris Plato, grec 1807, e.g. If. 7 r., 20 r. (nos. 10, 11). Vat. 191 (varia math., s. XV) has the sign occasionally, nos. 12, 13. I have not seen it in mss. of the Grotta Ferrata school 2 . 1 'AfxapTiait Vallicell. F. 10. 2 1 may notice here that the Alpha with crossed downstroke which Belger, Hermes XVI. p. 278 (Frag. math. Bobiense f. 114. 1. 30) imagines to be apa is APA— rAP. 9 AS. I give a few examples of this compendium used other- wise than at the end of words: dnapouoidoTooc Clarke Plato, XpHoaoeai Clarke 12, x a pd<3C30VTec Nonnus Add. 18231, dsee- voGvtcc doGeveicxc ecpaoKev Vat. 1982, bibdoKei HvarKaojuevoc Bodl. Auct. T. 4. 19, rasTpijuaprouc Eoe 16 (s. X). AY. Of this diphthong I can only produce instances from a few mss. more or less tachygraphical in character: viz. Vat. 1982 evTcxuea (bis) toutcx and passim, Grott. Ferr. B. a. iii. (s. XII) tccuthc, toicu'th eauTov, and very often. The sign occurs most abundantly in the London Nonnus, but ap- parently only in the strictly tachygraphical portion 1 . AYTOZ. A ligature for this pronoun worth recording occurs in some of the Grotta Ferrata mss. ; it consists of the a and u run together with the case-ending added : cf. cxCtoc auTole auTHv eau-roov cooauToac from Gr. Ferr. B. a. i. and Angel. B. 3. 11. A similar combination of a and u occurs in cxutoG from Aed. Christ. 70 (a. 1 104), and the ligature is probably common. TAP. I give a selection of more noticeable forms assumed by rdp. Nos. 1 and 2 are from the D'Orville Euclid and Harleian Lucian respectively, and this uncial form is usual in the Arethas-scholia : no. 3 is from the text of the D'Orv. Euclid. Forms with the uncial Gamma are 4 from Genuens. 7 2 (a. 1057), 5 fr° m Barocc. 196 (a. 1042), 6 from Mutin. 230 (a. 105 1 ), 7 from Aed. Christ. 70 (a. 1 104), 8 from Bodl. Auct. T. 4. 19 (s. X) : of minuscule forms, 9 is from Laur. 32, 15 (Iliad D s. X), 10 from Vat. 1982 (s. XI), 11 from Grott. Ferr. B. a. i (a. 986), 12 from Angel. B. 3. 11, 13 from the Paris Demosthenes £ (from the text), 14, 15 from Vat. 1298 (Aristides, s. X-XI), 16 from Vallicell. E. 40 (s. XI), 17 from Bodl. Eoe 16, 18 from Bodl. Auct. E. 5. 9 in reality eorat. I hope to call attention elsewhere to the mis-deciphering of this ms. 1 Examples from Par. grec 990 will be found in the account of that ms., P. 37. 2 Biblioteca della Congregazione di san Carlo. io ABBREVIATIONS IN GREEK MANUSCRIPTS. (s. X), 19, 20 from Laud. 89 (s. XI), 21 from Koe 1 (s. XI— XII) 1 , 22 from Turin B. 1. 22 (a. 1149). The abbreviation for rpa, which seems one of the peculiarities of the Grotta Ferrata school, may be illustrated by rpo^eiv from Isidore (a. 986), eeia rpacpn from Angel. B. 3. 11. It may be doubted whether sufficiently precise evidence as yet exists to prove Lehman n's conclusions (p. 92), as to the origin and relation of the two modes of representing the syllables ap and ep. TINETAI. The simple contraction for p'veTai acquires some interest from the varying forms under which it appears. It is most common perhaps in mathematical mss. Cf. 1, 2 from the D'Orville Euclid (text), 3 from mathematical scholia to the Paris Anthology, 4 from the Paris Demos- thenes 1, 5, 6, 7 from Laur. 5, 3 (Clem. Alex. s. X). AE. The ordinary usage for the particle be probably needs no illustration ; I may however add one or two examples to that given by Prof. Vitelli (p. 169) of the syllable be at the beginning or in the middle of a word ; viz. eibeac from the British Museum Nonnus, be\eTai bebooKooc bexa from Vat, 1982. Rather unusual ligatures are nos. 1, 2 from Roe 16. E. Lehmann's article on e is thorough and practically sufficient. The waved line which in minuscule represents both a 1 and e is universal in the verbal endings -juevoc and -jueea, and in other combinations of the syllable jue (e. g. juevei from the London Nonnus). But the abbreviation of e attached to other consonants is comparatively rare, and I can only cite examples from Laud. 37 (s. XII) : TeAeiv TeAei^v k69oAh oejueAHv. The spelling of the ms. is itacistic (cf. oHjuaivcovTec s> €C ^ and this particular usage may possibly be so explained 2 . 1 This form perhaps is analogous to those examples of cp given from Laur. 72, 5 by Vitelli, p. 15. 2 The same sign in Laud. 2>7 represents r) ; v. necpaXr] above, cruciTr) drjfxrjTpas. Cf. also hepyrjo-aaav Tip(opi)6i](rovTai from a ms. of New College, Oxford (No. 59, s. xii), rfjv from Vat. 587 (Cyril, s. XII) 5' ^et'/3ero Laur. 32, 15 (man. 2, s. XII-XIII). It may be doubted whether this mode of abbreviation is found HNETAI— EIN. il EIN, HN, IN. The proposition that the three syllables hv, eiv, and iv had originally a common sign, and that at a later period they were differentiated by the doubling of the sign, for eiv, the adding of diacritic points, for iv, the original sign being appropriated to hv, is correctly stated by Lehmann, p. 55. At what time and under what cir- cumstances these steps took place cannot at present be determined ; the statistics here presented may advance the question somewhat \ Manuscripts in which hv, eiv, iv are represented by the single sign are : the five Arethas-mss. , Euclid (a. 888), Plato (896), Lucian, Aristotle Urbih. 35, Clement (914); Clarke 12, Bodl. Auct. T. 4, 19, Laur. 28, 3, Angelica B. 1. 7 (Caten. in Matth. s. X), Vat. 1298, Vallicell. C. 41 (Caten. in lob. s. X), Grott. Ferr. B. a. iii. (s. XII). The following two differentiate eiv, but have a common sign for hv and iv: Angel. B. 1. 5 (Caten. in Evang. s. XI); Angel. C. 4. 14 (Liban. epp., s. X-XI) ; the following four have a common sign for hv and iv, while eiv is apparently always written in full : Iliad Ven. A, Bodl. Canon, no (s. X ineunt.), Grott. Ferr. B. a. i (986), ib. B. a. iv (992) 2 . A late example of the use of the original sign for eiv is Bodl. Auct. E. 2. 4 (1 106), q>epeiv. On the other side, the earliest dated ms. that I know of where the duplicated sign is used for eiv is the Nonnus Paris suppl. grec 469 A (a. 986) 3 ; cf. dnoGTd^eiv mveiv; another ms., undated but before the twelfth century. A very remarkable usage occurs on a page of tenth-century minuscule bound up with the ms. Vallicell. F. 10. It is illus- trated in the words irepi apapTrjparoiv mi KkrjpiKav, and consists in the ordinary tachygraphical sign for 77. I do not know an exact parallel. 1 Prof. Vitelli thinks (p. 10, n. 3) that Lehmann's observations upon the mode of forming the compendium for eiv at different periods — whether the strokes were drawn up or drawn down — need modification. I have not paid attention to the point. I may here say once and for all, that Lehmann's statements of, and a fortiori inferences from, the usage of Nonnus, Brit. Mus. Add. mss. 1 823 1, whether in regard of tachygraphy or the ordinary system of abbreviation, are entirely erroneous. A correct account of the ms. is given on p. 33 sq. 2 The two Grotta Ferrata mss., where they abbreviate hv, use the proper tachygraphical symbol ; for iv they use both modes of abbreviation. 3 The abbreviations of this ms. are strictly limited in number, but those that are used recur constantly. I did not find instances of iv, ar, 19. 12 ABBREVIATIONS IN GREEK MANUSCRIPTS. certainly not late in the tenth century, Bodl. Auct. E. 5. 9, has the later usage throughout for the syllables eiv, iv, eic ; cf. biacpepeiv, eAecooiv, ujuelc. Later dated examples are Mutin. 230 1 (a. 105 1), pHGeloiv; Genuens. 7 (1057), Koojuelv eoTiv; Genuens. 2 (1075), aTevi^eiv; Aed. Christ. 70 (1104), oneubeiv. From these instances it is plain that no more particular conclusion can be drawn than that the old system lasted long, and the new system began early ; but perhaps it may be said (1) that the abbreviation of iv was, relatively to hv and eiv, rare (a similar remark is made by Lehmann, p. 67), and (2) that the differentiation of eiv from hv began earlier than that of iv from hv. Ell, HZ, II. Vitelli's observation (p. 169, n. 2) that the double sigma in the sense of eic was in use in the first half of the tenth century is more than confirmed by the Clarke Plato, where in the scholia that come from the hand of Arethas the word evoTdoeic occurs with the final syllable thus represented ; in the rest of the ms. however the single sigma seems always to be found. Manuscripts in which 9 stands for all three syllables are the Clarke Plato, Clarke 12, and Clement Mutin. 126: it stands for hc and ic in the D'Orv. Euclid, Harl. Lucian, Vat. 1982, Iliad Ven. A, and doubtless in many other tenth-century mss. ; the last four mss. appear to write the syllable eic in full 2 , and I think Lehmann's remark (p. 57) well-founded, that the abbreviation is, in early mss., comparatively rare ; neither eic nor ic is abbreviated (as it would seem) in Nonnus Par. suppl. grec 469 A (986). The Grotta Ferrata school use as a rule the proper tachy graphic sign for eic, but the single sigma occurs occasionally, e. g. eKepeyeic from Nonnus ; the syllable ic in these mss. is usually written in full. Early examples of the ordinary double sigma are cpooKeic Demosth. Par. I, Hjuelc Euclid Laur. 28, 3 (s. X-XI) ; two dated instances are buvdjueic from Mutin. 230 (a. 105 1), navHpjpeic from Bodl. Auct. T. ii. 2 (a. 1066). Lastly, the 1 My notes do not give an instance from this ms. of eiv abbreviated. 2 I now find an example in Vat. 1982 of s for eiy : f. 218 v. Orjaavpi&is creaurej). Ell. I3 syllable is often enough abbreviated in the middle of a word ; cf. K€KAeiajuevai KiveTaBai from Clarke 12. Prof. Vitelli (12 n. 1, 172 n. 2) challenges the explanation of a ligature for eic, of which he gives examples ; I have found the form in Clarke 12, etc tc\, Laud. gr. 1, etc tov, etc touc, etc thv, Mutin. 12 (s. XII) juiGoc eic rov, Bodl. Misc. 251 (s. XII) eic tov (bis), D'Orv. x. 1. 1, 2 (Etym. Magn. s. XIV) etc to, and it is evidently the sign found by Ludwich in the Hamburg ms. of Odyssey-scholia (Aristarchs Horn. Texi- hritik, ii. 698) 1 . In one instance, Clarke 12 f. 189 r. eioiv (no. i),the ligature occurs not in combination with the article, and with the sigma expressed ; and this I think gives some ground for explaining the form itself as the ordinary ligature for ei vas] Vat. 2 (s. XI), els tovto Vat. 1456 (s. XI), els ttjv Grott. Ferr. Z. a. xxv. (schol. min. in Iliad, s. XI). The form is in fact fairly common. 2 Cf. Martin. Les scolies du manuscrit d'AristopJiane a Ravenne, p. xvii. 14 ABBREVIATIONS IN GREEK MANUSCRIPTS. use of the compendium in the middle of a word is illus- trated by evHOKouai from Clarke 12. EINAI. To the various forms of the sign for eivca given by Lehmann and Vitelli I add the following : nos. 1, 2 from Plato Paris 1807, which are apparently a near approach to the original form 1 , 3 a similar form from Vat. 1298 (Aristides, X-XI) ; the shape the sign assumes in the Arethas-mss. is shewn by 4 from Euclid, 5 from Plato, 6, 7 from Lucian ; Demosthenes Z and the Anthol. Pal. (Paris portion) offer 8 and 9, without accent or breathing, Vat. 1982 no. 10 ; here and in 1 1 from the London Nonnus, occurs the stroke for abbreviation. The horizontal type occurs in Clarke 12, no. 12 passim, Laur. 28, 3 no. 13, Grott. Ferr. B. a. iv. (a. 992) no. 14, Angelica T. 1. 8 (s. XI) no. 15, Vat. 1298 no. 16, Hermog. Paris grec 1983 (s. XI-XII) no. 17, Bodl. Misc. 251 no. 18 passim. EN. The form which this syllable takes in the Arethas-mss. is illustrated by juevToi, eoiKev, eicbeajuev from the Harl. Lucian, ev from Plato f. 395 v. For other tenth-century mss., cf. pouAneevTd Clarke 12, rreveepac Auct. E. 5. 11, ev eveev Nonnus Add. 18231 ; cf. also ev evToAdc from Barocc. 235 (s. XII), and the odd ligature oubev Mutin. 230 (a. 105 1 ). It will be observed that these forms are either right angles or slight departures therefrom. The varia- tion in which the downstroke is prolonged occurs in Vat. 1982 ev, eAerev (Vitelli, p. 9 n. 2). The strictly tachygraphical sign is by no means rare, and is found in mss. that are not otherwise particularly tachygraphic : cf. e. g. ooc^ojuev Auct. E. 5. II (s. X-XI) oooeHeijuev [sic], juev, 6eev Roe 16; further juev (bis) ev (bis) unojuvHaojuev Nonnus Add. 1 823 1, ojuiAoujuev Grott. Ferr. B. a. iv. (992), juev Aerojuev KctTexojuevoi Vat. 1982 2 . 1 In explaining the genesis of the original form it is difficult to accept either Lehmann's view that the two dots come by false analogy from the sign for Arri, or that of Graux (Rev. Crit. 1878, Notices Bibliographiques, etc. p. 66) by which they are due to the desire for symmetry. 2 And e?x ev > Angel. B. 3. 11, second hand. EINAI— EI. 15 EP. Of both the methods of representing ep there are abundant examples in older minuscule. The more common perhaps is that consisting of a stroke, inclined to the left, but above or below the line, with a ring attached ; this is to be found in the scholia of the D'Orv. Euclid (onep 1, tfnep), the Clarke Plato (onep 2), Plato Paris 1807 (unep bis), Laur. 28, 3 (onep 3), Clarke 12 (uneppoAHv), Mutin. 126 (unep 3, onep 4), Vat. 1298 (unep 4), Mutin. 193, Lucian s. XI-XII. (ei'nep), Mutin. 12, s. XII. (coonep, anep), and throughout in the Grotta F errata school, e. g. ouvepj-01 Nonnus, onepjua G. F. B. a. i„ oepjuom'vouc Angel. B. 3. 11, unep q>uoiv G. F. B. a. iv., onep coonep ib. B. a. iii. The second mode, the simple cross-stroke, of whatever origin, may be seen in the text of the D'Orv. Euclid (anep), often in the scholia to the Clarke Plato (coonep onep unep), in those to Demosthenes Z (coonep 2, onep 2), and it occurred in the now lost * codex Vallae' of Archimedes 1 . I have suggested under rap a parallel to the form quoted by Vitelli, p. 15. EI. The normal use of the sign for ec (e. g. as in viKcovTec from the Harl. Lucian) is well established ; it is less com- mon to find it either in the middle of a word or upon the line. Of the former case reveoecu dpeo«eiv from Clarke 12, eSeonv from Auct. E. 5. 11 are examples ; for the latter one may compare juaSaviec Plato Par. 1807, Aeoviec Mutin. 126, beonoTcu Angel. B. 3. II, ouveipcxvTec G. F. B. a. i., evi'^ovTec ib. B. a. iv., eoTiv ib. B. a. iii., cpBdoavTec dnoAei\j/eo9ai Vat. 1982. The itacism by which the double apostrophe, ordinary representative of cue, is employed for ec, of which Vitelli gives some instances, p. 1 2, is more widely spread than is commonly supposed, and occurs in mss. of a good age and often otherwise carefully written ; such are the well-known Laur. D of the Iliad (32, 15) juevov/Tec, Angel. C. 4-14 (Liban. epp. s. X-XI) boKouvTec dvaAcboavTec, Bodl. 1 As we are told by the writer of the Angelica C. 2. 6, who gives a table of the abbreviations used in his archetype : I take thence nep and the explana- tion. Cf. Heiberg, Philologus 42, p. 421 sq., and my own notes on the Biblio- teca Angelica, forthcoming in the Classical Berieic. 1 6 ABBREVIATIONS IN GREEK MANUSCRIPTS. Auct. T. 4. 19 (s. X exeunt.) bebooKOTec passirn, Laud. gr. I baijuovec ndvrec, Laud. gr. 37 (s. XII) oHjuaivoovTec kciAoGvt€c, Laud. gr. 89 (s. XI) anccvTec, Barocc. 235 (s. XI) tbovTec, Bibl. Corsini 41 G\ 16 (Evang. s. XII) pAenovTec baijuovoov-rec, ValHcell. E. 29 (S. X) eKTeTHKOTGC. EITAI. One of the most interesting of Greek compendia, and that may almost be said to have been discovered since the appearance of Lehmann's handbook, is the sign that represents eoTcu — almost technical in geometrical mss., but occurring with a certain frequency in mss. of other subjects. The sign was originally found by Bast in the ms. S. Germ. 249 (Comm. Pal., p. 810), and this is the only instance that Lehmann has before him ; Prof. Vitelli (p. 168) has found it in the Laur. Aeschylus, the Aristotle Paris grec 1853, and the Euclid Laur. 28. 3. The oldest instance that I am able to give of it is the Fragmentum mathematicum Bobiense, f. 1 1 4 v. 30 in Belger's copy, Hermes XVI, where Belger misreads it d'pa 1 . Next it is used in most of the Arethas-mss. ; the Bologna Euclid Archiginnasio A. 1. 18 2 has it, and doubtless most other mathematical mss. before the twelfth century ; lastly, it is one of the many compendia used by the scribe of the Bodleian Epictetus Misc. 251 (s. XII). Hence we get the following chain : 1 Frag. Bobiense, 2-4 D'Orv. Euclid (text), 5 ib. (scholia), 6 Plato, 7 Lucian, 8 Urbin. Aristotle, 9-12 forms from Laur. 28, 3 rather different to the one given by Vitelli, 13-16 from the Bologna Euclid, 17-20 from the Bodl. Epictetus. I enclose in brackets Prof. Vitelli's no. 53 for greater completeness. On comparing these forms with those given by Bast and Vitelli, it appears (1) that the dots signifying t are absent and present indifferently in mss. of the same age ; (2) that, with this qualification, the original form is best represented by the type given by the Bobbio fragment and the text 1 I have to defer the proof of this statement, but its truth will be evident to anyone who tries to read the passage grammatically. 2 Heiberg's b, saec. xi. EITAI— K. 17 of the D'Orville Euclid, and that the letters contained in the sign are therefore (t) + a + the tachygraphic t either attached to or crossing the downstroke of a 1 . (3) The late forms 17-20 are direct descendants of no. 1 ; the transition from one type to the other will be plain if we imagine such a form as Vitelli's no. 53 written with the curve open instead of closed. His no. 39 I should be inclined to explain as due to carelessness on the part of the scribe, but in any case it does not disturb the general result 2 . I. The curious expedient of representing, in late manu- scripts, iota by two dots on a level with each other, is well known, but not so universal that it may not be worth while illustrating from four dated mss. ; viz. koju- juc(tik6v from Coll. Nov. 258 (a. 1298 written by Demetrius Triclinius) where the usage is frequent, juvHoei-m from a note in Roe 1 that bears the date 141 7, recopj-ico from Mutinensis 118 (a. 14 ?8) 3 , and KovTeAeovn from Vat. Ottobon. 58 (a. 1538). IN A. A simple contraction for i'va, consisting of an iota with a mark of abbreviation beneath it, occurs in three mss. of my observation: Angel. T. 1. 8 (s. XI), Vallicell. E. 40 (s. XI), and the Bodleian Epictetus, Misc. 251 (s. XII). The form is practically the same in all three 4 . K. The article in Lehmann shows well how the waved line, descendant of the original tachygraphic k, represents final 1 Lehmann's analysis (p. 104) comes near to this, though in the single form given by Bast it was impossible to perceive the direct presence of the a. (Since the article on carat was written, I have found instances of both the plain and the dotted form in the Aristotle Ven. 201 of a.d. 955. In either case the form was open, and the example is important as an indication of the age at which this tendency manifested itself.) 2 To Lehmann's account of iari little exception is to be taken ; I doubt how- ever his statement (p. 102) that the horizontal line over ./. in Vat. 1809 denotes v — surely it is the general sign of omission. Curious representations of the word are no. 1 from G. F. B. a. iii, no. 2 from Laud. gr. 1 (s. XII). 3 On the date I must refer to my notes on the Estense in the Classical Review for February, 1889. 4 It is to be seen also in Vallicell. E. 63 (s. XII), a ms. in a hand not unlike that of the Bodleian Epictetus, and is probably common. C 1 8 ABBREVIATIONS IN GREEK MANUSCRIPTS. syllables beginning with k, chiefly but not exclusively ter- minations in -koc, etc. The usage is very constant in mss. that I have seen : cf. iaTpiKH, dpiejuHTiKH, emeujUHTiKcp Plato Clarke 39, iajupiKoc Clarke 12, ttoAitikov Auct. E. 5. 9, eeoopHTiKoO Nonnus Add. 1823 1. Nonnus however usually exhibits the case-termination, e.g. eKKAHoiao-riKHc, dneAao-riKHv, ouk embeiKTiKok ; so also jueptKHv Auct. E. 5. 9 l . Instances of terminations other than -koc are rdAcxKToc Clarke Plato, juiKpov Auct. E. 5. 9, ei'pHKev, ej-fvcooKov, kcxkcx Laud. gr. I, jua- Kp6v Laur. 32. 15 (Iliad D), £(3&ojuhkovto[kic Demosth. Paris Z 2 . KATA. Neither Lehmann nor Prof. Vitelli (p. 15) has given much illustration of the modes of representing kcitcx. One may distinguish (1) partial abbreviation, where (a) the k is tachygraphically rendered ; so Kaia 1 KaTcxxpcoaac often in the London Nonnus, koto 2 KaieKaei (sic) Clarke 12, koto; 3 and 4 Laur. 28, 3, KaTd 5 Vat. ] 298 3 , or (b) the abbreviation is confined to the other letters. A few examples will suffice to illustrate this ; cf. KOdd 9 KaTa&icoKeiv Bodl. Auct. E. 5. 9, KaTd 10 Demosthenes 1, koto 11, 12 Vat. 1298, KaTcmiv Angel. T. 1, 8 (s. XI), where the pair of dots appear to do double duty. (2) Complete abbreviation, the proper tachy graphical sign; e.g. KotTabuojuevH Clarke 12, KcxTc^eeipei Vat. 1982. The sign is universal in the more tachygraphical parts of Nonnus, but does not occur in as much of the ms. as is written in minuscule 4 . A. Lehmann's account of the origin and usage of the con- traction 5 for syllables beginning with Lambda is de- servedly commended by Graux, Kev. Crit. 1880, Notices Bibliographiques, etc., p. 165. His examples however admit 1 Rather individual forms of the k with case -termination are dnoo-ToXiKr)?, €-yK\r]ij.aTiKco, (niaKOTrwu from Vallicell. F. 47 (s. X). 2 Prof. Vitelli (p. 172) considers the waved line in the sense of the syllable Kai a rarity. I have found it in at least four mss. ; Roe 16 diKatoavvr], Auct. E. 5. II Kaicrapeia, Laud gr. 39 Kaipou, Knianpos, viKida. 3 Kara 6 Vat. 587 (s. XII), Kara 7 and 8 Vat. 1 3 16 (s. XIII). 4 Cf. also KaTaXapfidva Par. 990. How uncertain the use of the symbol is appears from KareXa&v from the same ms. I. e. a superimposed Lambda that afterwards assumes various shapes. KATA— OMOY. 19 largely of multiplication. The use occurs, though rarely, in the Arethas-ms., e.g. okcxAhvov, tooGKeAec Plato, noAic Lucian ; other tenth-century examples are nauAoc, noAic, ou/ipoAov Clarke 12, KecpdAaiov djuneAov Demosth. 2, oxoAhv, paoiAeiov Nonnus Paris suppl. grec 469 A, napapoAHv Iliad Ven. A ; in all of these instances it will be noticed how the right stroke of the A is prolonged. Freer examples from the same century are eniGToAHv Auct. E. 5. 9 (but 6 dnoGToAoc ib.) TeAoc, nauAoc, paoiAeuc Auct. E. 5. 11. The Grotta Ferrata school use the contraction not unfrequently : so kukAoc, emoToAHc, KoovoTavTivonoAeooc Nonnus, dnooToAoi G. F. B. a. i. Lastly, a few eleventh-century mss. may be cited: Laud. gr. 39, Selden supr. 11, Laud. gr. 1 (the eccentric form tou biapoAou). I add a somewhat more interesting example, eeoAdrou, from Phot. Bibl. Ven. 450 (s. X) 1 . MEN. The Bodleian ms. of the poems of Gregory Nazianzen, Clarke 12 saec. X 2 , so often already cited, has not un- frequently the tachygraphical symbol for juev, whether the particle or a syllable in a word ; so pev f. 1 57 v., juevouv ib. et saep., eeHpdoajuev 176 v. 3 For other tachygraphical usages of this ms. v. under Kcoc Grott. Ferr. B. a. iv. (a. 992), iKTepoc Bodl. Auct. T. 1. 2 (s. XI ineunt.). These examples may go some way toward resolving Lehmann's doubt (p. 75) about the meaning of the sigma in the two words that he quotes from Sabas 1 . OYN. Illustration of this common compendium, whether as the particle or the syllable, is hardly needful. I take almost at random koAouvtcxi, oneoc ouv from the D'Orv. Euclid, ouv 1 from the Clarke Plato, ouv 2 from Par. grec suppl. 469 A, and the unusual ligature pepcuouvToc, ouv from Barocc. 21 (s. XII). Of the genesis of the symbol I do not remember to have met with any account ; Lehmann (p. 76) leaves the question open. If however we compare this ordinary sign for ouv with the tachygraphic symbol for ev (v. supra), it is plain that they have one part in common, namely the crooked stroke that concludes either compendium : this stroke therefore, in either case, may be taken to repre- sent v, and while the open curve in ev will stand for e, that which is closed in ouv may similarly be inferred to represent 2 . OYI. Of the ordinary form of the compendium for this syllable illustration is needless ; somewhat remarkable forms however are dAAouc Par. suppl. grec 469 A (Nonnus a. 986), touc Tonouc ValHcell. C. 41 (s. X); a combination frequent in Grott. Ferr. B. a.iv. (a. 992) is perhaps worth recording, TTveujuaTiKouc. The double waved line (Vitelli pp. 9, n. 2, 169) occurs in Boe 16 (s. X) toutouc touc niGTeuovTac, Laud, gr. 39 touc xpovouc, Theophrastus Urbinas 61 (s. X) toic 1 I give a few examples of the sign for on ; no. I, from Clarke 12 f. 180 v., resembles the primitive form as given in Vat. 1809, no. 2, from the Paris Greek Anthology, is an instance of the omission of the dots ; cf. also nos. 3, 4 from Par. 3032, no. 5 from Vallicell. C. 61 (s. XV). 2 Compare ow as represented in Vat. 1809, e.g. a-vvecrrcoTos, f. 195 r. B. v. 22. 22 ABBREVIATIONS IN GREEK MANUSCRIPTS. tottouc 1 ; the single waved line (the proper tachygraphic sign) in Barocc. 26, eAeuGepoujuevouc, au-rouc xpiGTiavouc 2 . OYTOZ, etc. To Vitelli's examples (p. 1 1, n. 3) of outoc, outooc, etc., I may add a few others. The abbreviation is almost a feature of the Arethas-mss., cf. 1 outoo Euclid (text), 2 outooc ib. (scholia), 3, 4, 5 outoc, outoi, outooc Lucian ; further 6 outooc Vat. 1982, 7 outooc Vat. 1298, 8 outooc Mutin. 193 (Lucian, s. XI-XII). FTAPA. The word napd may be compendiously expressed (1) by n -f- the sign for dpa, (2) n with the topstroke crossed by a slanting line. ( 1) The former is by far the more common, and may be thought not to need exemplification ; I give however some instances from mss. that are for any reason noticeable : napd 1, 2 from Plato Paris 1807, napdKAHGiv napacpuAaKH Clarke 12, napaAmoov Demosth. Z, napd 3 Nonnus Par. suppl. grec A 469, napd 4 Iliad Ven. A, napd tov Auct. E. 5. 9, napd 5 napaboEoov Laur. 28, 3, napdox^vTai from so tachygraphic a ms. as Vat. 1982 (2) The other method in its nature, as Lehmann rightly observes, tachygraphic, is frequent in the Arethas-scholia ; cf. napd 6 napapAH0H D'Orv. Euclid, napajuueHodjuevoc Plato, napd 7 Lucian, napd toIc Urbin. 35 s . At the same time the Plato-scholia use the contrary system as well, e. g. napabeirjuaTa, napd tov. The cross-stroke is the usual method followed by the Grotta Ferrata school ; in addition to Vat. 1809, cf. napabebojuevov from Nonnus, napabpajuoov Isidore, napdpaaiv Angel. B. 3. 11. The twelfth-century ms. however, G. F. B. a. iii, has the more usual system : so napd 10. (3) A certain number of mss. offer instances of both forms at once. Beside the Clarke Plato quoted above, cf. napaKaAco napd 1 1 from Bodl. Auct. T. 4. 19, napd jueTpoov napapoAfic Mutin. 12. (4) I have 1 Par. 3032 \6yovs, Angel. B. 3. 11 (man. sec.) avrovs, Turin B. vii. 30 (s. X-XI) aWovs. 2 The form of the sign for ov, in which it is not round but angular, occurs in Bodl. Auct. T. 4. 19 (s. X exeunt.) oXiaSalvovaip, tov Xaov, Par. 990 o-vveXOovarjs ; see further under Tachygmphy. 3 Cf. also naph 8 Valliccll. F. 10 (s. X), napa 9 Vallicell. F. 47 (s. X). OYTOI— T. 23 noticed some variations in the more usual form which cannot be explained as coalescence of accent (Lehmann p. 91). The difference consists in a hook at the top of the upstroke, cf. napaGKeuHv, napabeij-juaTiKooc, napd 12 from Vat. 1298 ; it is curious that these forms, if the semi-circle were wanting, would be almost exactly like those given by Prof. Vitelli (p. 14) from Laur. 32, 9 and Laur. 59, 9 (plate nos. 7, 24, 25). An exaggeration of this variation appears to be the form napd 13 which I take from Angel. C. 4. 15 (Liturgiae, a, 1 165). Lastly, the singular form napd 14 Angel. T. 1. 8 (s. XI) must apparently be explained as an individual error of the scribe's, who had (v. s, T) some acquaintance with tachygraphy. TTPOZ. I have not seen the sign for npoc in the Arethas- scholia, and it will probably be found not to occur ; its place is taken by up. It is constant however in the text of the D'Orville Euclid (Stephanus) — a large and characteristic form: cf. nos. 1-4. A few instances of its occurrence elsewhere are npoooonov, npoc to Demosth. £, npoc 5 Anth. Pal. (scholia to the Paris portion) npooTaTai npoc 6 Vat. 1298, npoQKAoojucvoc, npoc 7, 8 Epictetus Bodl. misc. 251. A degraded form is npoc 9 from Barocc. 235 (Caten. in Psalm. s. XI) \ T. The representation of t by two dots placed over or across the following vowel or syllable is one of the most characteristic and consistently carried out practices of Greek tachygraphy, and is found in mss. otherwise of the ordinary type of writing far more often than is usually supposed. It is in fact often the only trace of tachygraphy that a ms. will show. The scanty account in Lehmann has been greatly added to by Prof. Vitelli (to, p. n, 170, toc n, 32, 1 It is extraordinary that a doubt can exist as to the origin of this sign. The slightest reflection upon the forms given above leads us back to the type that is in use in Vat. 1809, and of which not a bad example will be found under the xiith century ms., Grotta Ferrata, B. a. iii, p. 34. Each of the four letters is represented. The speculations in Lehmann (p. 87) must be read to be believed. 24 ABBREVIATIONS IN GREEK MANUSCRIPTS. re ib., 173, toIc ii, tou ib., tco 32, toov 170, tooc 11), and I am able to offer here some further collection of instances. The first example of each syllable, by way of type, is taken from the British Museum Nonnus; cases where the example comes from the strictly tachy graphic part of that ms. are enclosed in brackets. The notation appears entirely absent from the Arethas-scholia. T&J 1 : hgmvhtcu Nonn., cpopouvTcxi Par. 990, eoTcu Par. 3032. T&lC : 1 from Nonn., 2 Laur. 28, 3, T&N : OTav 1 Nonn., ot6;v 2 Par. 3032. T<\C : ttcwtcxc Nonn., exoviac Demosth. 2, to c Vat. 1982, eairrdc Angel. T. I. 8, dnejunoAouvTac Vallicell. E. 40, noAejuoOvTcxc Barocc. 138 (s. XII), Tdc Bodl. Misc. 251, KcxTacpopoGvTac Par. 990, tottc^ovtcxc Turin B. 1. 22 (a. 1 149). T&Y : TCdJTCxic Nonn., evTauOa Vat. 1982, tcxuthc Grott. Ferr. B. a. iii, tcxOtc< Par. 990. T€ : cooTe Nonn., totg ttot€ Clarke 12, outc (1, 2) cootg Vat. 1298, out€ 3 Bologn. Archigin. A. 1. 18, cootg 3 Thucydid. Brit. Mus. Add. 11,727 (s. XI), ttot€ 2, out6 4, evioie cooTe 4 Hermog. Par. grec 1983, out€ 5, tuttoutc Bodl. Auct. E. 5. 9, cooTe 5, Vat. 191 (circ. I4O4), T6KV01C, HJUeT6p0lC Par. 990, TT0T6 TOT6 C00T6 JUHT6 Neap. II. A a. 22, T6C : 900TiG9evT€c Nonn,, AeovTec Mutin, 126 (Clement), dvapdvTec Clarke 12, ovtcc exovTec Laur. 28, 3, evi^ovTec Grott. Ferr. B. a. iv (992), ndvTec Iliad Ven. A, (pedoavTec naAaiaavTec Vat. 1982, ndvTec dnoAauoovTec Angel. T. 1. 8, KcupooKonouvTec Mutin. 12 (s. XII), nepiAapovTec Hermog. Par. grec 1983, juoaxonoinaavTec Barocc. 138, bpdoavTec Angel. B. 3. 11 (man. 2) 2 . 1 For to. v. s. A. 2 I have not concerned myself, here or under es, with the illegitimate use of the two dots in the sense of es. It is worth recording however that the use occurs passim in the Ravenna Aristophanes, and is most remarkable in a ms. of T— YTTEP. 25 THN : 1 Noun., 2 Grott. Ferr. B. a. iv, 3 Laud. gr. 1 passim, 4 Vallicell. D. 43. Tl : napeon Vat. 1982. TIN : eo-riv Grott. Ferr. B. a. iii. TO : 6ktc(to Nonn., npcoTOKdeebpoi Par. 990, touto Par. 303 2 ' TOIC : toutoic Nonn., toIc Vat. 1982, toIc btKoaoic Par. 990. TON : auTov Nonn., eauTov ecpeauTov Vat. 1982, tov, q>0apTOv Par. 990. TOC : auToc Nonn., goojucxtoc outoc Vat. 1982, bia navTOc Par. 990. TOY : auTou Nonn., tou xp i gtou Par. 990, tou cpiAooocpou Arch. S. Petr. H. 45 (Galen, s. XII-XIII). TOYC : touc Nonn., cxutouc touc Vat. 1982. TCO : outoo Nonn ? , cxutoj equToo outoo Vat. 1982, djuuHToo Par. 990. TCON : ndvTOov to>v veooy Vat. 1982, twv ovtcov Angel. T. 1. 8, Toov Mutin. 12, ndvToov D'Orvill. X. 1. 1, 2 (Etym. magn. s. XIV), cpoudiVTcov Par. 990. TCOC : outooc Nonn., ekoTooc outooc Vat. 1982, ndvTooc dbiaoTdTooc Angel. T. 1. 8, AeAHGoTcoc Bodl. misc. 251, outooc Par. 990. > YfTEP. The tachy graphic abbreviation for unep cannot be said to occur frequently outside of the Grotta Ferrata school; I have not found it in the Arethas-scholia. Cf. however unep toov unep 90oKeoov from Demosth. 2, unep 1 from Grott. Ferr. B. a. xix (a. 965), unep 2 unep oou Nonnus, unep Aorov Gr. Ferr. B. a. i, unep 3 uneppdc Angel. B. 3. 11. For the partial-abbreviation, which is frequent, v. s. EP. that age and style. Cf. dieKnepcovTes (text), rexv^vres (scholia, man. pi.). This coincidence between the text and the first hand of the scholia may serve as another proof of the identity of the hands, in addition to those already brought together by M. Albert Martin in his admirable study upon this ins. 26 ABBREVIATIONS IN GREEK MANUSCRIPTS. YJ70. A tachy graphical contraction for uno occurs more frequently than for unep but without being itself common. There appear to be two distinct symbols in use : ( i ) one, already known from tachygraphy, and frequent in the Grotta Ferrata mss. : so e n6 i G. F. B. a. xix, uno 2 uno thc unovoiaic Nonnus, uno^upov Angel. B. 3. 11, uno 3 Par. 990 *. (2) Another sign, quite unlike the former, already quoted by Bast (p. 794) from S. Germ. 249, is identical in shape with the symbol for and (q. v.) so largely used by the Grotta Ferrata school, and only distinguished from it by the breathing, or in fact, where the breathing is ambiguous or incorrect, by the context. Certain instances are unobeiKvuovra, uno kcxkoG, uno 4 from the Bodl. Epictetus, Misc. 251 ; the same form is probably given by Vat. 1982 in unobeee'ioo: (sine spir.), and further by a late mathematical ms., Vat. 191 circ. 1404, uno 5 (sic). The fact of one compendium standing for both and and uno at once suggests that the letters actually denoted by the sign must be those common to both words, viz. n + ; and though I do not hold this conjecture proved, especially as regards the 0, it may at all events strengthen the con- clusion arrived at by Lehmann (p. 84) from consideration of ano alone. QN. Under this head I have to notice the tachygraphic sign, rightly described by Lehmann at the beginning of his article as a waved line ; it is in fact not unlike an open omega inverted. In this form it is used by an ancient hand 1 The argumentation between Gitlbauer and Lehmann (Lehm. p. 88) upon the origin and relation of the signs for vnep and 11770, does not perhaps admit of proof or disproof; but it may be allowable to suggest another hypothesis that appears at least as probable. Lehmann justly doubts Gitlbauer's half- revolution of the sign for vnep as a preliminary to explaining it ; but his own analysis of it, and especially his theory of the original identity of the two signs, appear to me no less arbitrary. A simple comparison of the common letters in vnep, vtto, with the common strokes in the two compendia suggests that (1) the left-hand stroke in vrrip is p, while the left-hand stroke in vno is o (the tachygraphic o is a line bent into two curves) ; (2) that the right- hand cross-stroke in either case is a mere mark of abbreviation, to erect the letters p and into substantive symbols for tnep and vno. For an analogy I may refer to my own analysis of the strange sign for ets (p. 12). yno— QS. 27 among the Plato Arethas-scholia, e. g. twv ovtcov, apxoov, ttAcctoov (as a rule the Arethas-mss. use the ordinary form, and that as often within the word as at the end). Cf. also toov veoov Vat. 1982, toov SeoopHjudroov, toov dricov Mutin. 12 1 . Early dated examples of the coalescence of the circumflex accent with the ordinary sign (Vitelli p. 10, n. 2) are toov G-enuens. 2 (a. 1075), V U X^ V Angel. C. 4. 15 (a. 1165). OP. A few instances of the compendium for this syllable, to be added to those given by Vitelli (pp. 15, 32, 171) are : pHToop Clarke Plato, reoopj-iKoov Harl. Lucian (both in the scholia), eKToop Iliad Laur. 32, 15 (text), juHTponcxToop Laur. 5, 3 (not cited, I think, by Prof. Vitelli), cc 1, outooc, dTTiKok), but it is found also in the text of the D'Orville Euclid f. 1 20 v. (xutooc, and it existed in the ' codex Vallae ' of Archimedes from which Angel. C. 2. 6 was copied; cf. the passage in the plate from f. 222 v. explaining coc and nooc 2 . (2) The syllable is found written on the line most constantly in mss. more or less connected with tachygraphy, e. g. outooc Nonnus, cogtc Angel. B. 3. 11, cpuoeooc Grott. Ferr. B. a. iv, ooonep ib. B. a. iii, outooc Kapeooc Vat. 1982, but not unfrequently elsewhere, so ooore toorrep often in the Arethas-scholia, 00c 2 Laur. 28, 3 3 . 1 A few more examples are tg>v i Vallicell. F. 10, to>v 2 Vat. 1456 (s. XI), TrXaraiv Par. 3032, navraiv Par. 990, t£>v alpenicav Ven. 450 (Phot. Bibl. s. X). 2 I make bold to explain in this way the sign given by Prof. Vitelli, plate II. no. 40, p. 172, n. 2 : it is cos + ep, i.e. eoo-n-ep, which, as Prof. Vitelli says, is demanded by the context. Another instance of this form of cos is 6'cpecos Vat. 2 (s. XI). 3 The beginner may with profit contemplate the fourth example of cos in 28 ABBREVIATIONS IN GREEK MANUSCRIPTS. QZTTEP. I do not remember to have seen or read of a single sign for obanep ; the separate abbreviation of either syllable is of course frequent enough. A ligature, consisting of the tachygraphical signs for both syllables interlaced, is given by the ms. Etym. Magn. D'Orville x. i. i, 2 s. XI V, of which a facsimile is prefixed to Gaisford's edition of the E. M.; cf. no. 1 from f. 289 v., no. 2 from 288 r. ; both are at the end of the line. The context of the first is, ia-reov be on coonep AereTCti Kpoujua Kai Kpouojua, outco Aererai XP l M a K0( i XP^ - The formation of the symbol is clear if we compare coonep, onep from the tachy graphic part of Nonnus \ Tachygbaphy. It has been often pointed out that in Greek minuscule and late uncial writing there are two systems of abbreviation in use at once : one, of rare oc- currence and of obviously tachygraphic origin, the other, far commoner, and though also ultimately tachygraphic in source, so familiar as to be known by contrast as the 6 ordinary ' system. Facts as to the coincidences and di- vergencies of the two systems are well given by Lehmann in his introductory chapters, and a masterly sketch of the subject, with illustrations, will be found in Graux' review of Gardthausen, Journal des Savants 188 1, p. 312 sq. The extent to which the ' tachygraphic ' system entered into the writing of ordinary books is one of the questions in palaeo- graphy which most stand in need of additional evidence. That the system was far more widely spread and more generally used in books than was commonly supposed, there Lehmann, § 47. He will not get light from Gardthausen, p. 258, nor even from Diels' explanation of the Fragmentum Bobiense, Hermes 1877, p. 421 sq. 1 I have in this tract hardly touched the large and interesting province of mathematical signs. I may however here mention one that is quoted by Hultsch ap. Gardthausen from Vat. 211, but that has not hitherto found an explanation. It represents ^copt'or, and is found with or without case- ending. Examples 1 and 2 are ^coptoi/, 3 xopiW, 4 x.pia. All these come from Euclid Laur. 28, 3. The usage occurs also in the D'Orv. Euclid, but at the moment of writing I am without examples. It consists of x an d P rendered tachygraphi- cally, upon the same system as that employed in Vat. 1809; cf. any page of Gitlbauer's facsimile. The second cross-stroke is doubtless a mark of abbrevia- tion. TA CHYGRA PHY. 29 can be no doubt ; but whether any principle governed its employment, and whether any place, persons or style of author can be connected with it, must for the present remain an open question. One of the few facts known for certain is that the later tachygraphical system was practised by monks of the order of S. Basil, and in especial by the Basilian school of Grotta Ferrata near Eome. Of the eleven manuscripts whose usage I proceed to summarise, seven were certainly written at Grotta Ferrata, one may have been, and another, though written elsewhere, was the work of a Basilian *, (1) An account of the history of the school of S. Nilo at Grotta Ferrata, the monastery and village between Frascati and Marino on the lower slopes of the Alban Hills, is to be looked for in the Prolegomena that are to complete the catalogue not long since published by the learned librarian of the Abbey, P. Antonio Kocchi 2 . In the mean time it may be convenient to say that San Nilo, the founder of the monastery at the close of the tenth century, established therewith a school and style of writing. The school may be said to continue, at least in the person of the Biblio- thecarius, to the present day ; the distinctive characteristics of the handwriting of S. Nilo 3 may be traced, in mss. written by his disciples, for more than a generation. Manuscripts with which I am acquainted that exhibit this type of writing are, beside the three books in the hand of S. Nilo himself (B. a. xix, xx, xxi), two mss. still in the monastery, B. a. i and B. a. iv, one in the Biblioteca Angelica at Borne, B. 3. 11, the London Nonnus 4 , and 1 For examples of tachygraphy published since the appearance of Lehmann's book cf. Vitelli, Spicilegio Florentine, Desrousseaux, Melanges de V ' Ecole Frmifaise de Rome, 1886, p. 544 sq., Gitlbauer, Ph'dologische Streifziige, 1886, p. 387 sq. 2 Codices Cryptenses, Tusculani, 1883. 3 Cf. the extract from the Vita Nili Rom. 1624, p. 28, quoted by Rocchi under B. a. xix : literarum forma utens densa et minuta. Facsimiles of the three Grotta Ferrata mss., and the one from the Angelica are shortly to be published by the Palaeographical Society. 4 When I was at Rome the celebrated Vat. 1809 was temporarily inacces- 30 ABBREVIATIONS IN GREEK MANUSCRIPTS. the fragment Vallicell. D. 43. The type loses its chief characteristics in the twelfth century, but still exists as a particularly neat and close minuscule ; authentic examples are B. a. iii, and the latter part of Angelica B. 3. 11. I have imagined resemblances in Mutinensis 12 and Bodl. Raw]. 156. It has always been well known that S. Nilo's disciples were tachygraphers, and the two great examples of their production in this direction, Vat. 1809 and Brit. Mus. add. 18,231 have been more or less carefully examined ; but for our knowledge of the codices that still remain in the Abbey we still depend practically upon Montfaucon and Piacentini. A stay of five days at Grotta Ferrata in May of last year (1888) gave me opportunity to inspect the library with this purpose. I have to thank Padre Bocchi's benevolence not only for access to the manuscripts that he knows so well, but for the arrangements that he was good enough to undertake for my entertainment in the village. May he accept an im- perfect acknowledgment of one of the pleasantest weeks that have fallen to the writer's experience. The three mss. in the hand of S. Nilo (a. 965) are un- abbreviated. On the last four pages however of B. a. xix occur a considerable number of compendia ; most noticeable tachygraphically are unep (imep tou AaoG), uno, and 10 in h anAn rvoaaic. The last sign is one of the rarest of those in use by the school, and will probably hardly exist elsew T here than in the purely tachy graphical parts of Vat. 1809 and Nonnus and, in ordinary writing, in Vallicell. D. 43. The other mss. appear to give the syllable always in full. Vat. 1982 gives' the common sign, v. s. Ell (Caoic). (2) The ms. of the British Museum, add. mss. 18,231, sible ; but to judge from Gitlbauer's facsimile of the tachygraphical part, the style of S. Nilo is to be recognised there also: it is of course well known, independently of the hand, that the ms. came from Grotta Ferrata. Lastly, from the description given by Graux (Arch, des Missions, etc. 3 . ser. V. p. 123) of the ms. 0. 74 of the Biblioteca Nacional at Madrid it would seem that it also belonged to the 'Scuola di san Nilo.' TA CHYGRA PHY. 3 1 containing works by Gregory Nazianzen with the com- ment of Nonnus, has for some years past been known to oifer extensive examples of tachygraphy ; I need not refer to the passages in the handbooks where it is noticed, nor to the facsimiles of pages in the various collections. It may be said to exhibit three styles of writing : (a) min- uscule : the text, only rarely abbreviated, and various introductions, indices, etc. (ff. 4 v., 12, 13, 14 v., 1 5 r., 318 r.- 330 v.) which on the contrary are very closely contracted; (b) scholia in large semi-uncial, which, beginning almost without contractions, gradually increase the percentage of signs till they almost reach pure tachygraphy ; (c) purely tachygraphical marginal remarks, glosses and corrections. In this article I deal with tachygraphy only in so far as it is introduced into ordinary writing, and therefore it is only the first of these three classes that I here notice. It is to be hoped indeed that the whole tachygraphical contents of the ms. may some day be made public, but I offer here nothing beyond a collection of the tachygraphical signs that are found in the minuscule part of it. The ms. was written, it is well known, in 972, but the writer has not given his name. Lehmann (p. 53) has rightly concluded, from a comparison of facsimiles, that it belongs to the Grotta Ferrata school, and the resemblance is obvious to anyone who has been both at London and Grotta Ferrata ; but he is certainly wrong in identifying the scribe with Paul who wrote the Isidore. The editors of the Palaeogra- phical Society, who in their forthcoming fasciculus publish several facsimiles of Grotta Ferrata mss., decide that the hand of Nonnus is the same as that of the Angelica Theodoretus, to be noticed below ; and there is a clear difference of writing between these mss. and the Isidore. A graver fault, however, with regard to this ms. has been committed by Lehmann, than the wrong identi- fication of its hand. Belying upon the evidence of the facsimile of a single page, he has in various places of his book made general statements of the usage of the entire 32 ABBREVIATIONS IN GREEK MANUSCRIPTS. ms. — which are wrong ; and based upon them theories as to the history of Tachygraphy, which are necessarily even further from the truth. It is sufficient to warn anyone who uses the hook that the statements in pp. 21, 22, 53, 54, 57, 67, as to the representations in the Nonnus of the syllables eiv, iv, eic, are incorrect. The matter is the most serious blot in a meritorious handbook, and a conspicuous example of the results of ' Palaeography from Facsimiles.' The manuscript then, in this part of it, exhibits specifically tachy graphical signs for the following syllables : ai, aic, ano, ap, eiv, ev, em, iv, ou, -napa, ra, thv, -t eon, tooc, unep, utto. Cf. the words cpaibpoTaTe, euxalc, ano, undpxeiv, oujunpaT- Teiv, KaTHverKev, eniKaGioac, eTeoiv, oubeic, toubaloc, thv tc, tout' eon, outo3c, unep, unovoiaic. Here the frequent use of at, aic, eiv, iv and especially ou, is noticeable ; the last compendium occurs in ' ordinary writing,' only in the mss. Vallicell. D. 43 and Par. 990. The curiously consistent contraction for tout eon also deserves notice ; in the semi- uncial scholia the words are still farther abbreviated. Otherwise the use of the symbol for Tau is singularly limited. The other signs are more or less characteristic of the school. For instances of Td v. ante s. A ; the usage of the manuscript for ap, efvai, ep, koto, napa has also been previously illustrated. Lehmann's statement that ic in this ms. is always written in full, is true to the extent that the tachy graphical sign does not occur in this part of the ms. ; in the scholia and the pure tachy graphic glosses it is common, and in the minuscule part itself the ordinary sign sometimes represents the syllable : v. ante s. €IC. The sign for 10, as I have already noticed, is very rare in all the Grotta Ferrata school. (3) The Biblioteca Angelica at Kome possesses one ms. of the school of Grotta Ferrata, Theodoretus B. 3. 11. The book consists of two parts, of which the first is of the tenth century, the second of the eleventh or twelfth. A description of it will be found in my notes on the Angelica TA CHYGRA PHY. 33 shortly to appear in the Classical Review, and in the current fasciculus of the Palaeographical Society, where a facsimile of a page from the first part is given. This earlier hand is the same, Mr. Maunde Thompson tells me, as that of the London Nonnus of a.d. 972 ; it is very con- siderably abbreviated. The second hand, though later, is not without traces of tachygraphy, v. ante s. EN. The tachygraphical signs used by the first hand are as follows : ai (aHjuaivei), cue (-njuoopiaic), drro (ante), ap (id.), eiv (tuxgIv), etc (hjucTc, oubeic), em (erriKaAeGoovTai), ep (ante), iv (uaKivSivov), napd (ante), unep (id.), t>no (id.). (4) The well-known Isidore, B. a. i, written in 986 by Paul, second Abbot of Grotta Ferrata, offers the following dis- tinctively tachygraphical signs : eu (gkoAiou), cue (biotGHKcuc) ap, and combinations of a with other letters (ante), and (ante), ei (baud^ei) ^ eiv (ucpcu'veiv), em (emeujuiaic), ep (ante), iv (doKHGiv), napd (ante), unep (unep Aorou). It is curious that neither here nor in Angel. B. 3. 11 is there any employment of the double dots for t which are so frequent in most mss. that are at all tachygraphic. The scribe uses the ordinary system of abbreviation freely, and the total proportion of contracted words is large. (5) It was my good fortune to have brought to me in the Biblioteca Vallicelliana at Borne, one day in February of this year, a Latin ms. numbered D. 43, of the Dialogues of S. Gregorius Magnus, saec XI, at the end of which were bound up two leaves of a Greek ms. in the writing of Grotta Ferrata. A photograph of one leaf was sent to Pre. Bocchi at the Monastery, and he was afterwards good enough to examine the ms. itself. His conclusion as to the hand is, I am happy to say, a confirmation of what sug- gested itself to me at first sight : that the writer is Paul, monk of Grotta Ferrata, the scribe of the Isidore B. a. i. 1 An unusual way of rendering «. The diacritic point is appended to the sign for at. In Vat. 1809, as one sees in Gitlbauer's facsimile, it is iota and the dia- critic point that together give et. D 34 ABBREVIATIONS OF GREEK MANUSCRIPTS. The subject-matter is harder to determine ; the leaves con- tain, as a librarian's note on the first of them and at the beginning of the ms. says, ' fragmentum indicis codicis antiqui/ but the work to which the index was prefixed Pre. Eocchi was unable to discover ; it was probably, in his opinion, a commentary on part of the New Testament, possibly S. Paul's Epistles. The interest however of the fragment is palaeographical. The leaves, numbered %% and 89, measure iof x 8 in., are in double columns with 44 lines in each ; they together form one sheet of vellum, and, the text being continuous, must therefore have been the middle leaves of a quire ; the flesh-side of the vellum is outward, the rulings are on the hair-side. The writing is above the line. The abbrevia- tions are extraordinarily numerous ; both in this respect and for the rarity of some of the symbols used, no other piece of f ordinary writing' at present known to have come from Grotta Ferrata can compare with these two pages. The following list gives the tachy graphical signs that I have found: a, ai, cue, a v a, ano, ap, eie, eK, eni, ep, goti, eiai, eivcu, iv, ic, ou, rrapa, -roOjfeoTi, thv, unep, uno : cf. the words Kae'oTi, 9CuveTca, vecpeAcuc, dvarevvoovTcn, aTTOOHjuel, cxnap)(H, Gc^elc, eKOHjue? Iktoc, en^aivojuevou, jueTepxojuevouc, eon, eioi, elvcu, cutougiv, ee'juic, ouk, vooujuevoc, napa, tout€gti, napapdTHv, unep, uno^uj-iov. Of these signs four are of great rarity, dva, 4k, ic, ou : of dva I do not know another instance in ordinary bookhand ; Nonnus has it, but in the tachy graphical por- tion : ck occurs also in Vat. 1982 and Par. grec 990 : tc, as I have already observed, is found elsewhere only in G. F. B. 1. xix, and ou only in Nonnus and Par. grec 990. It is instructive to compare the amount of abbreviation used by the same scribe in these two mss., G. F. B. a. 1 and Vallicell. D. 43. (6) The ms. of S. Maximus, Grott. Ferr. B. a. iv written in 992 by Neophytus, is far less widely contracted than the Isidore, while on the other hand it offers some examples of TA CHYGRA PHY. 35 signs which that ms. does not use. Cf. ai (on vai), ck (a>c ek toO), ev (ojuiAoCjuev), iv (eeAHoiv), Tec {ante), thv (id.) : a small piece of more continuous tachygraphy occurs on the margin Off. 39O; pHTCOV JUUOTHpiOOV. (7) Again ms. B. a. iii (s. XII according to Bocehi), in a good calligraphic hand, has, with an abundance of compendia of the ordinary sort (cf. s. A, AP, CIN, €P, etc.), the peculiarity of the tachygraphic sign for au, unknown elsewhere in the ordinary writing of the school, but very frequent here, whether as au or tchj ; cf. s. AY and the examples TauTHc, TauTHv, eauTov, cm thv : I take a longer piece of tachy- graphy from £ 6 J v. Geqeai be movoo to) rrpoc aAueoiav [?] gAenovji. (8) Montfaucon Pal. Graec. p. 283 gives a facsimile of a manuscript in the house of the Basilian order at Borne. The entire collection once belonging to this order is well known to be now in the Vatican, where it is incorporated under the ' Yaticani graeci,' beginning at 1962. The par- ticular ms. intended by Montfaucon is, as I have found out, Vat. 1982 or Basil. 21. Its description is as follows: membr. 8 x 5J in., ff. 223 ; ff. 1-189 are in ordinary script, 36 lines on a page, in a rather small good eleventh-century hand below the line, only slightly contracted; fT. 190 r.,. 190 v., 191 r., being blank leaves of the same book, are filled with contemporary semi-tachy graphic writing, con- taining on 190 r. and v. various medical receipts, on 190 v. 191 r. an extract from Chrysostom (inc. to rdp nAeiGTcc twv djuapTHjudTwv), 191 v. is blank, Ff. 192-223 are apparently a different book, in a much smaller but contemporary hand, much abbreviated, 29 lines to the page, containing S. Basil's Adroi on various portions of scripture, the first being that on Psalm VII which Montfaucon facsimiled ; it is defective at the end. The book may have come from Grotta Ferrata, but was certainly not written there, for on a modern fly-leaf at the beginning is the inscription Em libris MS. Monasterii d 2 36 ABBREVIATIONS IN GREEK MANUSCRIPTS. S. Helie Carbonensis * Nulliiis Dioecesis, Anglonen. Or din is S. Basilii Magni, and the hand bears not the slightest re- semblance to that of the famous Isidore of Grotta Ferrata ; it is in fact hard to see what can have suggested their identity to Montfaucon 2 . The ms. appears to use the following tachygraphical signs : otic (napoijuiaic), au (evTauGa), ano (ano twv), be (ante), €K (no. i), ev (ante), em (emGujuiac), kcx (both with a dot, as KctTe)(6juevoi, and without, as KaAAicpcoviac), Kap (Kapbia), koto (KOTacpOeipei), jua (onepjua), juap (djuapTHjuonroov), oa (pimoOeioa), to, Tac, Tec, toic, tov, toc, tou, touc, too, tcov, tcoc (ante), uno (imobeGeloa), oov (ante); see also under A, A€, 6INAI, €IC, TTAPA. The most noticeable points are the use of some quite rare forms, such as e«, kcx, Kap, jua, juap, sa (}xa and Qa very frequently), the series of T-syllables, and the absence of signs 3 for en, eiv, iv, eic, ic, unep, uno (in the more usual form, cf. ante), syllables which are so common in the Grotta Ferrata school. I add a phrase that I cannot at present decipher, from f. 190 r ; the words immediately preceding are TauTcx Ae£ac 6 oooTHp. (9) A manuscript in the Estense at Modena, ii. A. 12, S. Maximus de caritate, shows some tachygraphical in- fluence : it is of the eleventh to twelfth century, membr, 5I x 4m., ff. 135, in two hands, of which it is the former (ff. 1-56) that is partly tachygraphical. The non-tachy- graphic abbreviations are uniformly of the later type : e. g. those for eic, ic, eiv, iv (q. v. ante). Tachygraphic signs occur for ap (oapKoc), iv (rvcoaiv, noitooiv), napa (ante), tcx, Tep, tcov (ante), tov (ootccov) ; the most noticeable of these is that for iv, which is very frequent. This hand is upright and well-formed, and bears some resemblance to the twelfth- 1 MSS. from this Library are now to be found at Grotta Ferrata, e.g. A. a. xiii sq. 2 Montfaucon's mistake had already been perceived, from a comparison of facsimiles, by Lehmann, p. 54. That is, tachygraphic signs. TA CHYGRA PHY. 3 7 century type of Grotta Ferrata hand, as shown in B. a. iii. The second hand is quite dissimilar. (10) I have next to mention a ms. that has long been famous in the history of Greek tachygraphy — that of Hermogenes at Paris, which is now numbered grec 3032. From this book Montfaucon took his 'notae rhetoricae et oratoriae omnium lectu difficillimae,' which for more than a century were the only published examples of Greek shorthand ; they reappeared, as is well known, with an improved interpretation in Kopp's ' Palaeographia Critica,' in 181 7c It is not however with this venerable material that I have to concern myself. The book falls into two parts, the text and the marginal annotations, and it is the latter that Montfaucon published and Kopp revised. Of the text, on the other hand, no account hitherto has been taken ; it offers nevertheless considerable palaeographical interest. The ms, is a small vellum book, pp. 152, of no doubt the tenth century, in quaternions, written, mostly below the line, in a small upright ornamental minuscule, con- siderably ligatured ; in the margin, not very constantly, are annotations in pure tachygraphy in the same hand as that of the text, and in characters quite as large. We have therefore an instance of what may be called the normal case for the introduction of tachygraphy into book- hand ; the case namely where a scribe accustomed to prac- tise the tachy graphical system sets himself to write a book for the purposes of an ordinary reading public. For much the greater proportion of his text, abbreviation of any sort is excluded ; but at times, and especially at the ends of lines, it is legitimate, and it is inevitable that here, a scribe who is cognisant of tachygraphy will borrow signs from it as well as from the system of abbreviation ordinarily in use. Mss. of the sort are the London Nonnus, where the text is as a rule written out in full, not however without a diligent inspection of ends of lines yielding a fair return of 38 ABBREVIATIONS IN GREEK MANUSCRIPTS. tachy graphical symbols, and the second part of Angel. B. 3. ii, where the tradition of the school manifests itself in the same manner in a twelfth-century hand. This is precisely what has happened in the ms. in question, Par. grec 3032. The text is on the whole little contracted, but frequently at the ends of lines, and occa- sionally in the body of the paragraph, the scribe allows himself to shorten a word, adopting the same system as that in which he afterwards wrote his marginal comments. A considerable number therefore of signs usually considered tachygraphical may be taken from the text of this ms. I enumerate those that I have found, adding that as my study both of this and the following ms. was short, the list must not be considered exhaustive. Syllables represented are : ai (buvajucu), ano (dnopAencov), apa (apa), vou (ataxivou), 01 (boOAoi), ov (tov, cpiAinnov), oc (ttAhGoc), oti (bis), a (napabeir- juaToc), tcu (Iotcu), Tav (otcxv), to (touto), oov (nAaTOOv). V. also s. OYC . (11) Another tachygraphical Paris ms., grec 990, has been indicated and in part described by Ch. Graux, in the bril- liant sketch of Greek Tachy graphy to which I have already referred, Journal des Savants, 1881, pp. 316, 317. I learn from M. Omont that there is no likelihood of the complete study of the ms., which is there promised, being published, and I therefore take the opportunity of giving here such further particulars as I was able to gather during a short inspection of the ms. I lament that my account is not fuller, and does not better take the place of the authori- tative study that was to have been expected of the re- gretted French palaeographer ; but, as tachygraphy now stands, the addition of even a handful of new forms is worth making, especially when, as in the present instance, there are at hand the means for reproduction. The ms. is dated 1030, and is of the ordinary minuscule of that period ; there is nothing in the character of the hand to suggest the tachygraphical knowledge of the TA CHYGRA PHY. 39 scribe. The ordinary compendia that are used, e. g. for aic, civ, exhibit the later stage of the forms. The ms. con- tains the poems of S. Gregory Nazianzen with a prose paraphrase, in parallel columns ; it is in the paraphrase, according to the necessities of space, as Graux clearly describes, that the abbreviations occur. I have observed the following, to which must be added the forms already quoted by Graux, the most interesting of which are no and boc : ai (juoopaivovTec), aic (dvoiaic), aA (6cp0aAjiiouc), dno (dno OTepHoic), au (nauaeTai), eA (peATioTOv), em (no. 1), ep (napep- XGTai), eu (nveujua), iv (KaAoOaiv), Ka (KOTeAa^ev), Kai (no. 2), KaTa (KaTOtAajupdvoo), ju e (no. 3 jueTci), 01 (dvejuoi), 6juoC (no. 4), ov (6'vtooc, tov), ou (ouvgAGoughc), nav (ndvTOoc), pau (no. 5 Kepauvoc), Ta (ndvTa, touto, unepetboojuaTo:), Tai (cpopouvToa), Tac (KaTacpo- poOvTac), iau (v. toutcx above), tg (tekvoic, HMeTepoic), thc (thc OaAaooHc), to (npooTOKaOebpoi), toic (toIc biKcuoic), tov (cpGapTOv, tov), toc (KAanevTOc), tou (toO xpiaTOu), too (djuoHTOo), toov (cpoiTOOVTOuv), Tooc (outooc), uno (uno thv), oov ((ieoav). A remark or two upon some of these forms may be in place ; to discuss the ms. as a whole will need more complete examination. The syllables aA, eA, jue, pau are of very great rarity. I do not know if parallels can be quoted from any ms. but Yat. 1809 and Add. mss. 18231. The form 01 is illustrated by Graux 1. c, and by Yitelli Spic. Fiorent. pp. 13, 168 : it must still be called rare. The form eu has been found by Martin in the scholia to the Ravenna Aristophanes ; the tachygraphical n, as in ndvTcoc, is very uncommon in ordinary bookhand ; the sign for ov is not generally found on the line ; Kai represented by three dots occurs elsewhere in Vat. 1809, Add. mss. 18231, in some Laurentian mss. quoted by Vitelli, pp. 15, 32, and in the late ms. Vat. Pal. 73 discussed by M. Desrousseaux (v. p. 28, n. 1), With regard to the series of Tau-abbreviations, there is to be noticed the freedom with which the dots are superimposed instead (their more usual position), of being arranged one on either side of a stroke of the sign for the following syllable ; 40 ABBREVIATIONS IN GREEK MANUSCRIPTS. the toov of this ms. compared with the same syllable in Vat. 1982 will illustrate the difference. It is obviously a mere convenience of the scribe's, and not a distinct usage. A somewhat erroneous idea of the origin of these two dots is given by Gardthausen Griech. Palaogr. p. 199. In conclusion it is convenient to notice here some mss. which offer one or two tachygraphical signs only. One such is a ms. of Galen, H. 45 in the Capitular Library of S. Peters at Eome, of the xii-xiiith century, written on palimpsest vellum ; the text offers no peculiarities, but at the end is written in red the sentence toC cpiAooocpou KupoG qnAaj-deou, in the manner represented in the plate. Not only the tou, but the plain strokes for the alpha in a ms, of this age, and the purely tachygraphical form of the 9, are very noticeable. A single tachygraphical form, sac, is given in the word e^oooooac (the context is hc kcu tov ncuba e^ooooaac) by Vallicell. E. 55 of the xith century. Other mss. of this sort that have already been described are Clarke 12 (v. s. juev, re), Angelica T. 1. 8 (to:, Tec, toov, tooc), Roe 16 (ev), Laud. gr. 1 (thv), Neapol. II. A a. 22 (tc, tooc). Plate I ■V7R> cr ere ■fr t^.^ -art f 3 k<, \y \* *l' i f -\ y A \ rCA^€TpS c - -^evp^ #$/Vt C$s"fA ^ r ^^ KKo**f*** 03 r eiJ'-tiVw U^>^f Trrofr^fW »'^ prf^ T - /J ^ , ? t7 ../ s VV Abbreviations in Greek MSS. Clarendon Press. Plate III, TO\i if cV ox y1~~ 4: T^ >V JF 10 _ \\2Z >x ^ i.A yv r ~ } a 5 Aj S- ji * fr £-1^ 7'* T * Abbreviations in Greek MSS. Clarendon Press. Plate IV ^V ^ T V s ^ GOO" ytf^TT* ItoS^c-npT* Pa 'o " , i 3 ...ft .7 ' ^ » r 3 ^ V i /, _ / 12. 13 ^ eN D-TT7 H Tp ° TT] 7 - 1 - O / my, $>««',* ^."""l go, N.2-? r«N-^ &K e Y + ™^\~3> x " Nj; v^i+pJi J^hjcpr i\?* /"f^'-V" rU> ^ "«p7^« v , . f a ;' 3 INk L L L «^^ ^^ <^ K 1 j 2. 3 ^ 7T y^ 0y tyr 1y x^ <:^"UO«NA Abbreviations in Greek MSS. Clarendon Press. Plate VI. Aoroc ,*lfy? i£tf* *^°v ^ c ^ N a ^ / M6N "L ^4 g«VlF?z 2-3 4 5" 6 7 2 OMOIWC 6a/\ ON ft>vX"-ry*T^ mV < * / f'X' r ^^ ^ T " OYN «*J^i o^-cj *$ 9 a€ua4omr7DC o rmr Abbreviations in Greek MSS. Clarendon Press. Plate VII. / & * >y (touj 4- V S r i 8 2. 3^5" ^ b \y\ \r\ ottoc y r vty y^ Y » PAP* rV nV /KX./ rTf-VK-' ^ fCT7K tt or r rr npoc -^ i . l, ^_ ^^^ ^ < \ 2. t I/ 1 J T&sC inv/j A°^ * / '' crew/ ^4h c fejn>XoMJ TMC[h\] V^ '£+, L a *~ 3 3 TA.IC H,l 0' Te I- s> , St- Abbreviations in Greek MSS. Clarendon Press. Plate VIII. TUN A / / / T | ^ TIN 7/ roc \-vl ?fi° f y ^^ <^p\ (Vj T^-iJtS*' ^P 6 ^ pe^frf 1 y^y"yij H^K x<^ f*o V Vf Cvn; br ^ h ^ ^^' f.t 7 -£^*M 7^r4-Ul uAeWpii Aiirematims in Greek MSS. Clarendon Press. Plate XI j.kjo r ^ .9. <2- '^*Z-~ V ' f|X) tf T^w\ 1*>l«> 7X ^t T3VM^ ^WJl^T ckj ~^y / (x|) £pW? ^ o/ Y^ t/rfc^ 7T o / L . . £ «■ •- / Cv • — ■ • S-F/rJ- TTjUjh ^7 f r \ * f -2? ov^ ^T cof- g-3 °0 co < L/ Abbreviations in Greek MSS. Clarendon Press. Kantian HENRY FROWDE Oxford University Press Warehouse Amen Corner, E.C.