W. M. Lindsay, An Introduction to Latin Textual Emendation

Errors of Emendation

---

Deliberate emendation a feature of Renaissance MSS, not early minuscule MSS

It is in the MSS. of the fifteenth century, the period of the Renaissance of classical studies, that deliberate emendation has most usurped the place of faithful reproduction of an original. Manuscripts were at that time to a great extent written by scholars themselves or under their supervision; and it was regarded as the first duty of the preparer of a MS. to furnish his readers with a text which was correct in sense and grammar. The corruptions, real or imaginary, of the original were silently emended; and lacunae were filled up with words or whole lines, suggested by the ingenuity of the scribe himself or of the scholar whose directions he followed. The fifteenth-century MSS. of Plautus which contain the last twelve plays afford examples to the full. We know them to have been copied from the “Codex Ursinianus” (p. 6), and from no other source; and since their original is preserved to us, we can estimate exactly the deviations of each copy. Thus in Pseud. 1063 “viso quid rerum meus Ulixes egerit”, the archetype of our minuscule MSS. had qui instead of quid a common mistake (ch. vii. § 2). The corruption quirerum is faithfully preserved in B, but in the original of CD a new confusion ensued—the substitution of s for r (ch. vi. § 1), quiserum. This quiserum the late copy boldly alters to quid servus: “viso quid servus meus Ulixes egerit.
”

Again, the missing scenes of the Amphitruo were supplied by a Renaissance scholar, Hermolaus Barbarus, with verses of his own; and these scenae suppositae appear both in MSS. of the time and in the early printed editions.¹

It is these practices of the fifteenth and sixteenth centuries which have brought MSS. of this period under suspicion, and which attach uncertainty to any text, such as the Silvae of Statius, for which we have no older authority than Renaissance MSS., or the poem on Prosody of Terentianus Maurus, for which we have only the “editio princeps” (Milan, 1497).²

---

The Carolingian copyists

The period of Charlemagne and his successors was another period of revival of learning. But, as a rule, the Carolingian monk-copyists confined their emendation to matters of spelling and punctuation, and have not played havoc with the text of Latin authors to the extent of the Renaissance scribes. Interpolations are scarcely to be found, except in grammatical texts, like Priscian and Nonius; and even there they are to be regarded more as marginal jottings of additional examples of a word or fresh illustrations of a grammatical rule, which accidentally found their way in subsequent copies out of the margin into the text.

---

Wrong correction of spelling

The first duty of a Carolingian monk-copyist was to correct the barbarous spelling of his original. In late Latin e and i, o and u were in certain circumstances pronounced alike, ct had assumed the sound of tt, x of s, so that in manuscripts of the period preceding the Revival of Learning under Charlemagne we find barbarisms like littoris for lectoris, creaturem for creatorem, auxerint for hauserint. An i was prefixed in late Vulgar Latin to words beginning with sp, sc, st, and some other consonants; a final -um was pronounced as o; and these mispronunciations, which have left their mark on the Romance languages (e.g. Fr. école from Vulg. Lat. i-schola, Ital. vino from Lat. vinum), perverted the spelling of the illiterate Merovingian copyists. For the purpose of weeding out these barbarisms from MSS. the Carolingian scribes were provided with text-books on orthography, one of the most widely used of which was the De Orthographia of Alcuin, the head of the famous monastery-college of Tours. This task they performed on the whole extremely well; and we owe it to these monks that the texts of Latin authors have come down to us in a readable state. But mistakes were unavoidable. Occasionally a scribe left a misspelling uncorrected, when it had assumed the form of a different word; as, for example, littoris, a misspelling of lectoris, has the form of the genitive singular of littus (on errors of this kind see ch. v). Occasionally in excess of zeal he altered a perfectly correct form under the idea that it was a misspelling. Thus o, we have seen, was the Late Latin equivalent of -um of the accusative singular masculine and nominative/accusative singular neuter of the second declension, and would be corrected by a Carolingian copyist to -um. But in Bacch. 463 “cave malum” of the MSS. may be a wrong correction of the Plautine cave malo (cf. Pers. 369 “malo cavere meliust te”), the scribe having wrongly supposed malo to be the same late Latin misspelling of malum as vino etc. (Ital. vino), of vinum. Similarly the common Plautine adverb facēte has been altered to facite (2nd plural imperative) in Mil. 1141, 1161 etc.

---

Wrong separation of words

Besides correcting the spelling of his original, a Carolingian scribe had to attend to the proper separation of the words and to the punctuation of the sentence. Punctuation, unless in a very rudimentary form,³ can hardly be said to be known in Latin MSS. till the Carolingian period; and although in the more carefully written ancient MSS.⁴ we find the words ticked off from each other by dots, the separation of words in Latin MSS. may be put down to the credit of minuscule copyists.⁵ A thorough separation, indeed, is not found in the earlier minuscule MSS., for small words, such as prepositions, pronouns, and particles, are usually joined to longer neighbouring words, a practice which is exemplified on every page of the Plautus MSS. (e.g. Capt. 10 “inalide” BDE, 34 “depreda” DEJ, dequestoribus BDE, for in Alide, de praeda, de quaestoribus).⁶ This has often led to the loss of these small dependents. For example, in Cas. 854 “i belle bellalula” the first word has been lost in our existing minuscule MSS., probably because ibelle, so written, was taken for the same barbarous spelling of belle as ischola, quoted above, of schola; in Rud. 875 “obsecro me”, written obsecrome, has become in B obsecrom and in CD obsecro.

Keller (Epilegomena zu Hora:) explains the loss of i in a class of Horace MSS. in

“dum favet nox et Venus, i secundo
omine,

”

C. iii. 11. 50:

by the supposition that i was written with a point before and after it, .i., and that these points were mistaken for puncta delentia (ch. iv. § 1, below).

---

Erroneous emendations

The wrong separation of words in MSS. can sometimes be remedied by a stroke of the pen. Madvig's brilliant restoration of a passage of Seneca's Epistles (89. 4) is a well-known example: “philosophia unde dicta sit, apparet: ipso enim nomine fatetur quid amet. Sapientiam ita quidam finierunt, ut dicerent divinorum et humanorum sapientiam”. In this passage the QUIDAMET of the Archetype had been wrongly broken up by a Carolingian scribe into quidam et, so that the MSS. offered: ipso enim nomine fatetur quidam et sapientiam ita quidam finierunt; and modern editors printed: ipso enim nomine fatetur. Quidam et sapientiam ita * * quidam finierunt, supposing a sentence to have dropped out after ita. Similarly in

“quid nos dura refugimus
aetas?

”

Hor. C. i. 35. 34:

the durare fugimus of certain MSS. is easily emended. But usually the wrong separation of words brought other mistakes in its train. In the Persa of Plautus, v. 587, we have seen (p. 3) that aequom hic orat, “he talks justice,” written AEQVOMHICORAT, was in the Carolingian archetype of our minuscule MSS. wrongly broken up into aequo mhi (mihi) corat. The natural consequence was that copyists fancied this corat to be a misspelling of curat, so that we have in the manuscripts C and D aequo mihi curat. Another example has been already quoted from v. 546 of the same play. Quia specie quidem, written QVIASPECIEQVIDEM, was broken up into qui aspeci equidem, and was changed — how could it be otherwise? — to qui aspexi equidem, the aspeci being taken for a misspelling of aspexi. A curious instance is found in v. 288 of the Menaechmi, where opsonatu redeo, “I come back from catering for dinner,” appeared first as opso nature deo, then as ipso naturae deo.

---

“Ghost words.”

The point to notice about all these wrong emendations of the text, or rather wrong corrections of what seemed to be misspellings, is the extreme facility with which they were produced. A Carolingian copyist, accustomed to correct errors like littoris for lectoris, vino for vinum etc. in every other line of a Merovingian original, could hardly do anything else than change qui aspeci equidem into qui aspexi equidem, opso nature deo into ipso naturae deo, or in Pseud. 1173 “quo tum odie” (for quofumo die) into quo tum hodie. Once that the first error had been made, the error of not “visualising” properly QVOTVMODIE as quotumo die, the further development of the corruption followed as a matter of course. If, on the other hand, a miswriting in the original was not recognised to be a miswriting, but was copied faithfully as it stood by successive generations of scribes, a non-existent form might become perpetuated in MSS. A good example of this is a miswriting of the Vulgate in the book of Job, ch. xix. 23-24: “quis mihi det ut (sermones mei) exarentur in stylo ferreo et plumbi lamina vel celte sculpantur in silice?” Here celte is apparently nothing but a miswriting of the adverb certe. There seems to be no such word as celte in Latin. And yet this miswriting not only survived successive generations of copyists, but has actually established itself as a word of our language. In archaeological books the term “celt” is of frequent occurrence, meaning a particular kind of cutting instrument which is found among the remains of prehistoric man (see The Oxford English Dictionary s.v.)

Other familiar examples of “ghost-words,” as these are called, are—“Grampian,” from a miswriting of Mons Graupius in Tacitus Agr. 29; “Boadicea,” a scribe's error for Boudicca; Virgil's Inarime (Aen. ix. 716), from a misapprehension of Homer's εἰν Ἀρίμοις (Il. ii. 783） as one word, Εἰναρίμοις. An example from Plautus MSS. is found in Poen. 1301, where the word bajulus, “a porter,” written in the older spelling baiiolus, became in the archetype of our minuscule MSS. baliolus, through the common confusion of i with l (ch. vi. § 1). This miswriting baliolus has found its way into a good many Latin dictionaries, and has been provided not only with more than one meaning, but also with more than one etymology!

---

Modernising of archaic forms

The text of Plautus, with its numerous archaic forms and constructions, has unfortunately suffered greatly at the hands of mediaeval scribes. For example, in the Latin of Plautus' day illīc, as well as illi, was dative singular of ille; illi, as well as illīc, was the adverb (originally locative singular of ille) “there.” But when a Carolingian copyist found the O. Lat. illic dative singular in his original, he naturally took it for a barbarous misspelling; and relying on the rule of his text-book of orthography that illi was the dative singular, illīc the adverb, he would usually substitute illi without hesitation. For the O. Lat. adverb illi he would similarly write illic in his copy. In a few cases the metre shows us that a change of this kind has been made. Thus in Amph. 249: “namque égo fui illi in ré praesenti et méus quom pugnatúmst pater,” “for I was there at the action itself, and so was my father, when the fight took place.” Here the minuscule MSS. offer illic; but since the metre (iambic tetrameter acatalectic) requires the form illi, we can be sure that illic is a scribe's substitution for the illi of his original (cf. Capt. 277-9). This is in fact the leading principle of Ritschl's treatment of the text of Plautus—the restoration to the text of the archaic forms of Plautine Latin, which had been silently changed to the classical forms by mediaeval scribes. Thus in Merc. 46, where the minuscule MSS. agree in reading objurgare, Ritschl made the line (an iambic trimeter) metrical by reading objurigare: “objúrigare páter haec noctes ét dies
”, “--u-u w - -u - u-
” “my father used to censure this night and day,” the old form objūr[icaron]go being attested by MSS. in other passages, e.g. by A in Trin. 70: “nemóst. Quid tu igitur rógitas, tene objúrigem?” “No one. Why ask then if it is you I censure?” When we look at the passages of Plautus which the Latin grammarians quote in illustration of archaic forms, we see to what an extent these archaic forms have been modernised in the course of transmission. Festus cites termentum, a derivative of tero, from the Bacchides; but in Bacch. 929, clearly the line referred to, non pedibus termento fuit, “was not rough enough to hurt the feet,” we find tormento substituted in all the minuscule MSS., though A, the Ambrosian Palimpsest, retains termento. The Palimpsest, however, is as guilty as the other MSS. in other passages—such as Epid. 10, quoted both by Festus and by Donatus for the archaic adjective habitus, “in good condition, stout”: “córpulentiór videre atque hábitior. Huic grátia”, “you seem stouter and in better liking. Thanks to this (thieving hand of mine),” where A has abilior. To a fourth-century scribe, like the scribe of A, these old forms, habitus, illi adverbial, illic dative, etc., would be almost as unknown as to a Carolingian scribe. As a matter of fact we have in the minuscule MSS. quite as many archaic forms as in A. Sometimes they are preserved by all the minuscule MSS., e.g. Men. 405 semul (class. simul); sometimes by B alone, e.g. Stich. 383 poste (post haec CD, postea A), Pseud. 386 ecfecta (haec facta CD, effecta A). And in a large number of instances a trace, sometimes a very faint one, is preserved of a lost archaism: e.g. in Cas. 380, quoted by Priscian for the old nominative singular sortis, “a lot,” the words alia sortis, “another lot,” are represented in BVE by alias oris; and we can see that the old form aiio (class. aio) stood in the archetype in Cas. 71: “at ego aiio id fieri in Graécia et Carthágini”; for B has alio, which in VE is changed to alia (to agree with Graecia), and in J (as in A) is aio. The Noctes Atticae of Aulus Gellius (second century a.d.) are full of complaints about scribes of the time, who altered, e.g., majores vestrum (Sallust Cat. xxxiii. 3) into majores vestri, vadimonium stitisses (Cato) into vadimonium stetisses, quadrupes eques (Ennius) into quadrupes equus; so that we can see that this form of error in MSS. was of an early date, and that mediaeval scribes are not the only offenders in this respect.

The modernising of archaic forms is carried out to a great extent in the MSS. of Plautus, and the student of Plautine textual criticism must bestow great attention on this point. But since most of the examples that might be quoted in illustration (for a fuller list of them see Appendix A) are peculiar to archaic Latin authors, it will be sufficient to mention those which are commonly found in MSS. of Cicero, Virgil, and the like.

ii for i.—An early grammarian (Velius Longus) tells us that Cicero was in the habit of writing with double i words like Maia, Ajax, where an i (j) stood between two vowels. In MSS. of Cicero we find that this archaic spelling has proved a stumbling-block to scribes. Thus eiius for ejus has become in Milo. 7. 16 ei jus, in Fam. vi. 2. 1 et jus, in Att. viii. 4. 1 ei vis. (Leo ad Plaut. Mil. 1274 cites melius as a corruption of eiius.）
st for est.—Early grammarians (e.g. Marius Victorinus) recommend the curtailed spelling of the substantive verb in audiendust for audiendus est, audiendast for audienda est, audiendumst for audiendum est, etc. Spellings of the kind have usually been altered by mediaeval scribes, who sometimes rightly understand the contraction and expand the -st to the full form est, but often pervert st to sit or si, or even sim (written in MSS. sī), or sunt (written in MSS. st with a line above). Thus in Pseud. 448 insipientiast, “it is folly,” is rightly written in the Palimpsest INSIPIENTIAST, but in the minuscule MSS. is insipientia (-cia) si (cf. Most. 701).
• -is Accusative Plural—The usual form of the accusative plural of I-stems in good authors has -is, e.g. civis, navis. This is very often altered by mediaeval copyists to -es; though the remark of Aulus Gellius (Noct. Att. xiii. 21. 3-5) that Virgil wrote urbis in Georg. i. 25: “urbisne invisere, Caesar” etc., but urbes in Aen. iii. 106: “centum urbes habitant magnas”, shows us the danger of correcting each and every form in -es to the form in -is.
-os, -om for -us, -um.—In Late Latin, as we have seen, u and o were in certain cases interchanged in pronunciation and writing; e.g. vinum was pronounced and written vino (cf. Italian vino). A Carolingian scribe might easily mistake a genuine form like equom or servos, nominative singular, for a Merovingian misspelling, and substitute equum, servus. In classical Latin o, the older vowel, was used after another u or after the consonant v, but in the Latin of Plautus' time the old spelling was retained in many other situations beside. It is not always easy to feel certain that an o which takes the place of u in the MSS. of Plautus is a genuine survival of the Old Latin form (e.g. opos sit for opus sit, which has become corrupted to possit in Stich. 573; see above, p. 3), or a mere instance of the Late Latin use of o for ŭ. Similarly e for i in Plautus MSS. is a survival of the early form in semul, “simul,” Men. 405, and the like, but may sometimes date only from the Late Latin period when ĭ had become e (cf. ch. v. § 9).
quoi, the spelling in vogue in the youth of Quintilian (i. 7. 27), is, if understood by the scribe, corrected to cui; if not understood, it often becomes quo or qui (cf. Bacch. 126, 225, 617).
Unassimilated Prepositions in Compounds.—A scribe was always prone to correct forms like subpeto, adpeto etc. to suppeto, appeto etc., and in this respect has often effaced the traces of the spelling of his original. On the other hand, he may often by this change have unconsciously reverted to the older spelling; for the assimilated forms are often those used by the ancient writer himself (e.g. Plautus used the form assum for adsum, and makes a pun with it on assum, “roasted,” in Poen. 279), while the unassimilated forms have been foisted into the text by grammarians and editors of the Empire. The same is true of spellings like quot for quod, set for sed—spellings which are generally corrected by Carolingian scribes, and which may in many cases not be really ancient. (For examples in A see Studemund's Index.) The assimilation of prepositions in compound verbs often led to corruptions. Thus conjecti in Livy xxxvi. 12. 4, misread as conlecti (ch. vi. § 1), has become collecti; conjecere in Livy xxx. 5. 4, misread conlecere, has become collegere (see Heraeus, l.c.)
ec- for ex- in Compounds before f.—The spellings ecfero, ecfugio etc. were a puzzle to mediaeval scribes, who generally replace them with haec fero, et fero, or the like. Thus in Pseud. 386, as we have seen, ecfecta is retained by B alone, while C and D have haec facta. (On the similarity of ec to et in minuscule see ch. vi. § 1.)
• iis for eis.—The dative and ablative plural of the demonstrative is, when in the form iis or is, are often changed to his. The correction in MSS. written in Caroline minuscules is commonly made by writing above the initial i the sign of the Greek rough breathing (see ch. ii. § 7).⁷
• -umus for -imus.—The change was made at the close of the Republic, e.g. maxumus (-imus), vicensumus (-cesimus).
• -undus for -endus in gerundive.
-rier for -ri in inf. pass., e.g. Most. 117.

---

Change of unfamiliar to familiar word

Not only an archaic form but any unfamiliar word was liable to be changed by a scribe. Taking it for a mere misspelling, he would substitute for it some similar word which was familiar to him. In fact most textual corruptions might be included under the category of change of unfamiliar to familiar words; and it is a cardinal law of textual emendation that the form found by the scribe in his original must be supposed to have been a form less easy to understand or construe than the form which he substituted for it in his copy. In Mil. 831 the unfamiliar word hēmīnas, “pints,” has become feminas in our minuscule MSS.; in Mil. 1178 scŭtŭlam, “an eye-shade,” has become cultura; and so on. The Latin interjections often puzzled mediaeval scribes. Thus heus has been changed to ejus in Men. 673 (cf. Men. 836; Mil. 1358). Greek words too, which were often written in Latin letters, are curiously transformed. The salutation χαῖρε appears as care in MSS. of Cicero Fin. i. 3. 9; ναὶ γάρ, written necar (Bacch. 1162), has been emended by a scribe to ne carpe. And names of persons and places suffer in the MSS. of Plautus as they do in all MSS.; e.g. in Men. 1112 “ut abii Tarentum” has been changed by one scribe to ut habitarem tum.

To guard against mistakes with a proper name, a horizontal line was often drawn above it in MSS. This line seems to have been mistaken by the scribe of the Ambrosian Palimpsest for a contraction-sign in Cas. 994 “Hector Ilius”, where the Palimpsest has ecastor illius, the scribe supposing that ECTOR with over-line was a contraction of ecastor. The confusion of Ilium with illum, Ilio with illo, etc., is of frequent occurrence in MSS. (cf. ch. vi. § 1; ch. v. § 12). Examples of this and other confusions of proper names are given by Madvig in his Adversaria Critica i. pp. 71, 125 sqq.

---

Grammatical corrections

Sometimes it is the grammatical construction of the sentence which has been altered through being unintelligible to the scribe. The use of the ablative with facio, “offer sacrifice,” in Virg. E. iii. 77 “cum faciam vitula”, was not understood by copyists, who have changed vitula to vitulam. Similarly the Plautine use of the adverb was not recognised in Asin. 807, where quot pure habuerit has been altered to quot puras habuerit, with puras made to agree with the word in the previous part of the line, noctes.

---

Assimilation of ending

We may include under the head of grammatical corrections the assimilation of one ending to another. For instance, Mil. 631 albicapillus, written in the archetype albi capillus, has become albus capillus. But this is often a mere clerical error. Thus Mil. 630 “clare oculis video” was wrongly written claris oculis video, evidently because the writer's mind wandered to the termination of the neighbouring word, and not because he thought clare to be a miswriting of claris.

---

Change of objectionable words

An error of deliberate emendation, not exemplified, I believe, in the MSS. of Plautus, is the change of objectionable words. Many examples of this are found in a ninth-century Paris MS. containing extracts from Martial; on which see the preface to Schneidewin's edition of Martial p. lxxxv.

---

Metrical emendations

As instance of a metrical emendation we may take

“ipse rotam adstringit sufflamine mulio consul.
”

Juvenal viii. 148:

The word mulio having been miswritten multo (ch. vi. § 1), the line was emended in our MSS. to multo sufflamine consul to save the metre. Such emendations are rare in Plautus MSS., and where they do occur, cannot be referred to mediaeval scribes; for these were ignorant of Plautine metre. They belong rather to ancient editions or recensions of the plays.

---

Completion of a quotation

Nor can we exemplify from Plautus MSS. a not infrequent case of interpolation—namely, the addition of words for the purpose of completing a quotation. For instance, a passage of Virgil (A. vi. 577) is quoted by Nonius Marcellus (229 M. 28) in illustration of the word Tartarus: “tum Tartarus ipse
bis patet in praeceps tantum.

” This is the form in which Nonius seems to have quoted the passage, and the form in which it is found in the ninth-century Leyden MS. of Nonius. But in the Laurentian MS., belonging to the end of the ninth or the beginning of the tenth century—a MS. which seems to be a direct copy of the Leyden codex—the second line is completed with the letters t. q. s. a., standing for tenditque sub auras.⁸

---

Adaptation to parallel passage

Very similar to this is another error—namely, the adaptation of a passage of an author to a parallel passage found in the same or another author. St. Jerome complains that in some MSS. of the Gospels the description of an event in one Gospel was altered or amplified in order to bring it into exact agreement with the description of the same event in another Gospel:

magnus hic in codicibus error inolevit, dum quod in eadem re alius evangelista plus dixit, in alio, quia minus putaverint, addiderunt. vel dum eundem sensum alius aliter expressit, ille qui unum e quattuor primum legerat, ad ejus exemplum ceteros quoque aestimaverit emendandos. unde accidit ut apud nos mixta sint omnia, et in Marco plura Lucae atque Matthaei, rursum in Matthaeo plura Joannis et Marci, et in ceteris reliquorum, quae aliis propria sunt, inveniantur.

The error has often arisen through the practice of transcribing parallel passages on the margin, whence the divergent reading has found its way into the text. Often too it is a mere slip of memory on the part of the scribe, who in transcribing one passage had let his thoughts wander to the diction of a similar passage. On this form of error see ch. v. § 5.

---

Completion of argument or construction

Of additions to complete the argument, or to complete the construction of the sentence, a passage of the De Natura Deorum may serve as example (i. 31. 86): “dubium est enim, utrum dicat aliquid beatum esse et inmortale, an, si quod sit [id esse mortale].” Here the words id esse mortale are a scribe's addition of this kind. The change of et spiritu to sed quae spiritu in a passage of Cicero (Marcell. 9. 28), “nec vero haec tua vita ducenda est quae corpore et spiritu continetur”, is clearly the emendation of a monk-copyist.

The British Museum MS. of Plautus (J), a twelfth-century MS., is a copy of a lost codex which was professedly an “emended” version. The writer of this lost MS., perhaps an abbot of a monastery, wrote at the end an epigram of his own composition, which has been reproduced in the British Museum copy: “exemplar mendum(Corrupt.) tandem me compulit ipsum
cunctantem nimium Plautum exemplarier(To copy.) istum
ne graphicus mendis proprias idiota repertis
adderet, et liber hic falso patre falsior esset.

” The result of the learned abbot's interference has been that J is the most worthless of the twelfth-century MSS. of Plautus. Occasionally, especially in the Epidicus, he has made the right correction of his exemplar mendum; e.g. in v. 329 he has rightly corrected fere to facere (fere E, ferre B). But in the great majority of cases he has added his “propriae mendae” to the “mendae repertae”; e.g. in Capt. 274, “eúgepae! Thalém talento nón emam Milésium”, “bravo! I would not give a talent for Thales of Miletus,” the first two words, written in his original euge potalem (a corruption of euge petalem of BD), have been corrected by him to euge ob talem.

---

List of Examples

Here is a fuller list of examples from Plautus MSS. of the various errors of emendation:—

Wrong separation of words:
• Poen. 557 “nos ratu's” (nostratus B, nostra CD).
• Pseud. 529 “lepide lenonem” (lepidele nomen B, lepidule nomen CD).
• Pseud. 831 “in patinas indidi” (inpatina scindidi BD).
• Asin. 684 “me amantem egere” (meam ante megere BD, meam ante mergere EJ).
• Aul. 807 “anu ea rem” (an veram BDE, an vera V (2nd hand), J).
• Mil. 685 “suave ductust” (sua deductust CD).
• Truc. 437 “quae mihi credidit” (quem hic redidit B (1st hand), quem hinc reddidit CD).
Modernising of archaisms:
In Aul. 748 occurs the old phrase (attested by Nonius) luci claro, “in broad daylight,” in which luci, being regarded as an adverb and therefore independent of distinctions of gender, is joined with the neuter of the adjective. Our minuscule MSS. agree in reading luce clara.
• In Mil. 74 “ībus,” the old dative plural of is, is attested by the grammarians: “latrónes, ibus dínumerem stipéndium.” Two of our MSS. have latronis bus, the third latronibus.
Rud. 357 inposuit for inposivit.
Mil. 1181 sis for sies.
Amph. 445 collum for collus (attested by Nonius).
Amph. 554 tu autem for tuatim (attested by Charisius and Nonius).
Asin. 942 voluptatis for volup.
Aul. 93 extemplo for extempulo.
Bacch. 445 attingas for attigas (attested by Nonius).
Capt. 30 inde audivit for indaudivit.
Amph. 673 occepto for occepso (attested by Nonius).
Change of unfamiliar form:
• Merc. 441 “licitere” (liceret B, litigere CD).
Trin. 798 Athenas aurum for ad thensaurum.
Trin. 934 eo for eho.
The interjection st, “hush!” is changed in the minuscule MSS. to sta Cas. 148, to sit Pseud. 129, to si Pseud. 600, 952, Most. 489, etc., to ast Mil. 985, to est Cas. 212, to sed Curc. 155.
Grammatical correction:
• Capt. 159 “multigeneribus” (multi generibus P, multis generibus DE, 1st hand).
• Mil. 1106 “ecquid fortis visast” (fortius P).
• Mil. 1045 “viden tu ignavom, ut sese infert?” (inferat CD).
Assimilation of ending:
• Bacch. 956 “paria item tria eis tribus sunt fata nostro huic Ilio” (nostra P).
• Men. Arg. 4 “avos paternus facit Menaechmum e Sosicle” (paternos P).

The following passages of Plautus should, I think, be emended in accordance with the above considerations:—

Mil. 1005. The MSS. offer: “PAL. prius nequam iliam oculis tuis. PYRG. Video id (ideo) quod credo tibi.” If we look at the whole passage, we shall see that the emendation of this line is mainly a matter of punctuation, like the famous emendation of a line of Shakspeare (Oth. v. 2. 7), “Put out the light and then—. Put out the light!
” for which in the old copies we find, “Put out the light and then put out the light.
” The soldier Pyrgopolinices is talking with the slave Palaestrio about the pretended widow who has sent him a message: “PAL. tum aútem illa ipsast nímium lepida nímisque nitida fémina.
PYRG. hércle vero jam ádlubescit prímulum, Palaéstrio.
PAL. príusne quam illam oculís tuis —
PYRG. Videón id quod credó tibi?

”

Palaestrio is interrupted before he can finish his question with the word vidisti, “Are you in love with her without having yet seen her?” by the impatient rejoinder of the soldier, “Your description of her is so satisfactory that it is as good as seeing her for myself,” lit. “Do I not see what I believe from you?” with ne, as often, for nonne.

Cases of interruption of this kind are not uncommon in Plautus, e.g. Truc. 504: “ST. Sálve. AS. Salvom— ST. Scío: sed peperitne, óbseero, Phronésium?” where Astaphium is interrupted in the middle of the sentence salvom te advenisse gaudeo (other examples are Epid. 117, 128; Poen. 607). Mil. 1319 has been thus restored by Niemeyer: “PH. Íbo, quamquam invíta facio: hómini pietas— PL. Scío: sapis.”

Stich. 629 should, I think, be so explained. The parasite's bid for a dinner non ego isti apud te (cenabo)? is interrupted by Epignomus: “Nón ego isti apud te—. Sátis spectatast míhi jam tua felícitas” (ista for isti A).

Most. 419. Sphaerio brings the key of the house to Tranio. The line assigned to Tranio by the MSS. is quite intelligible, if we suppose Sphaerio to hold up the key after the first sentence: “sed quíd tu egredere, Sphaério? jamjam. óptume
praecéptis parulsti,

”

“what do you come out for, Sphaerio? (Seeing the key) Ah! now I understand (cf. “jamjam novi,” Curc. 233). You have carried out my instructions capitally.”

I would resolve acto of the MSS. in Most. 1134 into two words, ac tu (a common case of substitution), and read: “áge, mitte ista: ac tu ád me in cenam—. Díc venturum:
quíd taces?
”

Archaic forms should, I fancy, be restored to the following lines: jusses (with Bugge) to

“SA. Pernám quídem
jussés adponi frígidam postrídie (jus est BCD).
TO. Ita fieri jussi,

”

Pers. 106,

“you should have ordered the ham to be served up cold next day. So I did”; eaepsae (on this old feminine plural of ipse see my Latin Language ch. vii. § 20) to

“eaepsaé se patinae férvefaciunt ílico (eaepsese patinae A, eae ipsae sese patinae BCD).
”

Pseud. 833,

The Greek word ἀνίαν may be the cause of the corruption in

“postrémo modús muliebrís nullust núnquam;
lavándo et fricándo scimús facere ἀνίαν,

”

Poen. 231,

where the MSS. offer nníam and enjam.

The Plautine construction of the adverb with sum, a feature of colloquial Latin, seems to me to have been wrongly changed by scribes in

“potuít: plus jam sum líbcre quinquénnium (libera, MSS.)
”

Epid. 498,

and by modern editors in Truc. 172, where aliter should not be emended: “ego fáteor, sed longe áliter est amícus atque amátor.”

Wrong assimilation of endings⁹ by the scribe of P appears to me in Mil. 894, “mala mílle meres. St! né pave; pejóribus convéniunt”, where P seems to have had mala milla merest. Translate: “you deserve a thousand punishments. Hush! don't alarm yourself; they (sc. the punishments) befit worse women than me.”