the scribes of R have attempted wrong emendations in the belief that words in the original were wrongly spelled. The following is a list of the confusions in spelling to be found in R. Each case of confusion of letters or sound is followed by a list of errors of which it is the cause.

(1) The confusion of e, ae, oe. This is the most common of all. ae and e, in particular, are used without discrimination, as, indeed, is also the case in the Puteanus. The scribes write predae = praedae, aequitum = equitum, poenatibus and paenatibus = penatibus, praelium, saenatum, gregae = grege, caedae = caede, urbae urbe, etc.

=

Examples of possible corruptions due to this cause: XXVI, 34, 2 operae praetium P, opere praetium R. XXVI, 15, 10 infestis signis ire ad urbem pergit P, signis irae R. — XXII, 33, 8 aedem in arce faciendam locauerunt P, arcae R. - XXIIII, 20, 2 eos effuse palatos hanno adortus P, effusae R. — XXVIII, 25, 6 suis recte factis gratia qui exsoluat P, rectae R. — XXVI, 49, 15 nunc ut id curem impensius, vestra quoque uirtus dignitasque facit P, curae R. The scribe did not notice the abbreviation for m.-XXIII, 49, 12 suppletis copiis ex prouincia, ut quae maxime omnium belli cupida P, prouinci aut quem axime R1. The scribe saw quaem, and supposed that it was an error in spelling on the part of the scribe of P. XXVIIII, 2, 9 dum cedenti duodecimae legioni . . . tertiam decumam legionem . . . firmamentum ducit P, caedenti R. XXVIII, 30, 6 cum lelius et ipse in quinqueremi portu carpeiae . . . invehitur P, ipsae R. This confusion of ipse, ipsae is one frequently made by the scribes of R. — XXVIIII, 9, 9 pleminius impotens irae, neclectam ab scipione . . . suam iniuriam ratus P, iraena et lectam R. Here the scribe has seemingly resolved the first e of neclectam into ae, and then attached one letter to the first word, and the other to the second. Other examples of this nature have been already given in the chapter on wrong word-division.

As the result of writing e for ae, que is, in manuscripts of all periods, almost as common a spelling of the relative pronoun as

quae. The spelling que is, indeed, so common that few errors are likely to arise from it; but, on the other hand, scribes who wished to employ the spelling quae for the relative often went too far, and wrote quae for the enclitic que. This creates an opportunity for corruption, especially if the relative suits the context, and if, as was the case in R, the words were so divided that quae was written separately, and not attached to the preceding word. Examples of this kind of error are to be found on every page. I shall give only a few.

XXIIII, 1, 7 aduocataque extemplo contione P, aduocata quae extemplo contione R. XXVI, 33, 10 sciretque (scisceretque Luchs) plebs P, sciret quae plebs R. XXVI, 33, 13 deuina (= diuina) humanaque, utensiliaque siue quid aliud dediderunt P, utensilia quae siue R. — XXVII, 39, 6 non enim receperunt modo aruerni eum deincepsque aliae gallicae atque alpinae gentes P, deinceps quae aliae R.-XXVIII, 25, 5 quae causa ire consternationisque subitae foret P, consternationis quae subitae R. - XXIII, 39, 7 et nolae . . . plebs hannibalis erat, consiliaque occulta . . inibantur P, consilia occulta R. quae - In XXV, 11, 15 Luchs's text has: suas cum claustra portus hostis haberet, quem ad modum inde in apertum mare euasuras. For the words in italics P has: haberet et que ad modum. The et is probably due to dittography, and the que to leaving off the sign for the nasal. The scribe Nauto has added to the difficulty by writing haberet et quae ad modum.

...

(2) o written for u, and vice versa. Spellings such as tulerantes, pupulatione, furtuna, uicture (= uictore), ducendo (= docendo), expugnatoros (= expugnaturos), incolomi, ligorum (= ligurum), totum (= tutum), luxoria, moros (= muros), syracosanos are not uncommon. The errors due to this cause consist chiefly in the confusion of the ending us with the ending -08, e.g. maximus for maximos, medius for medios, Romanus for Romanos. Errors of this nature have already been treated under Emendation.

(3) Some of the other vowels are confused, though not so frequently.

diadema, luciria Luceria, delegestis

=

=

delegistis, itenere = itinere, uergis uirgis, seculorum siculorum. Spelled seculorum, this word is likely to be confused with saeculorum.

=

=

a and e are occasionally confused in unaccented syllables owing to mental pronunciation. Examples are: XXII, 20, 4 escensio ab nauibus in terram facta P, ascensio R.— XXVI, 51, 3 ipse

paucos dies .. exercendis . . copiis absumpsit P, exarcendis R. XXVII, 41, 5 id modo romanum quaerere apparebat P, apperebat R. A corrector has changed apperebat to appetebat.

(4) The wrong insertion or omission of h. This creates many possibilities for error, owing to the confusion of such words as aut, haut; ac, hac; iis, his (the correctors in the manuscript have, in almost all cases, changed iis to his); ostium, hostium; anno, hanno; etc. Examples: XXVIIII, 19, 8 quae piacula quibus deis quibus hostiis fieri placeret P, de his quibus R. — XXIIII, 13, 4 hac cum spe dimissi tarentini P, ac cum spe R.-XXII, 24, 6 quia haud dubie hostis breviore uia praeuenturus erat (Luchs), haud quia dubiae hostibus P, aut quia R.— XXIIII, 16, 6 metu poenae collem haut procul castris ceperunt P, aut procul R. This confusion is quite common.-XXIIII, 20, 2 eos effuse palatos hanno adortus P, effusae palatos anno R.-XXIIII, 14, 1 parte altera hanno... altera ti. gracchus P, parte altera anno . . . alterati gracchus R.XXIIII, 40, 8 ad ostium fluminis P, ad hostium fluminis R.XXII, 39, 6 ominis etiam tibi absit c. flamini mememoria P, hominis R. - XXVI, 26, 8 dilectum . . . per totam siciliam habitum P, abitum R. XXII, 12, 7 et modo citato agmine ex conspectu abibat P, habebat R.

...

(5) Final d written for final t, and vice versa. This gives rise to such confusions as at, ad; it, id; quot, quod; aliquot, aliquod; idem, item; etc.

Examples of errors due to this cause: XXIIII, 39, 13 ipse hibernacula. . . communiit aedificauitque P, communi id R.XXIIII, 33, 3 iouis it (= id) templum est P. In R the scribe did not recognize it as the other spelling of id, and wrote: iouis

sit templum est, which was, however, afterward corrected. — XXVIII, 26, 15 sedit tacitus paulisper P, sed id tacitus paulisper R.— XXII, 16, 1 quod uiae inter bina castra erat P2. The scribe of R first wrote quod, and then altered so as to read: quot uiae inter bina castra erat. XXII, 34, 10 quia inuitis iis dictator esset dictus P, dictatores sed R.- XXVIII, 24, 2 quantam excitatura molem uera fuisset clades P, uerafuis sed clades R.

- XXVI, 38, 7 at ille . . . rem hannibali aperit P, ad ille R. —XXIIII, 14, 9 ad quae clamor cum ingenti adsensu est sublatus is the reading of the texts. P has: at quae clamor, etc. The at quae of P has in R become atque.

of

The two errors which follow are due to a similar confusion Ρ and b: XXVI, 29, 8 conlege optionem dare prouinciae P, obitionem dare R. Compare, also, pueri liber ludibrium in XXIIII, 4, 2, where P has puerili per ludibrium.

(6) b written for u, and u for b. Spellings such as uoleuat, pleuis (= plebis), occasionally occur, but no corruptions have arisen in R from this cause.

(7) ci for ti before a vowel, and vice versa. This confusion in spelling was so common that few errors were likely to arise from it. I have found but one example of anything amounting to a real corruption from this cause: XXVIIII, 12, 1 neclectae eo biennio res in graecia erant P, ingratia erant R.

(8) i for g before a vowel, and g for i. XXIIII, 35, 4 aiebat P, agebat R.—XXII, 35, 7 gessisset P, iessisset R.

(9) The following errors are due to the confusion of the sound of s with that of x, and that of c before i: XXVIIII, 2, 2 ausetanum P, auxetanum R. XVIIII, 21, 12 onus

demptum erat de scipione cognoscendi P, decipioni R.

(10) In XXVIIII, 29, 5, the texts have arcessit, P has accessit, and R has accensit. The error is due to the fact that n was not pronounced in the combination ens.

(11) The following error seems to be due to a confusion of the sounds of 7 and r, which gives rise to the spelling fraglantia for flagrantia: XXVIII, 28, 10 nihil ultra facile creditam mortem meam a nobis uiolatus sim P, facere R.

(12) The use of the double letter for the single, and vice versa, has already been treated under Dittography and Haplography. This is the cause of confusions such as dimisi, dimissi; uelit, uellit; etc. A good example of an error growing out of the repetition of a single letter is XXV, 41, 10 ut claudius comitia perficeret P, ut claudius commitia perficeret R. The unusual spelling caused the scribe of M to emend, and write ut claudius commitius perficeret.

XIII. MISCELLANEOUS EXAMPLES ILLUSTRATING THE CUMULATIVE GROWTH OF ERROR

The examples given in the preceding pages have been chosen with the view of illustrating, in each case, a single stage of error in the process of transcription. For this purpose I have used, for the most part, only the errors actually made by the scribes who copied the Reginensis, disregarding the alterations made by the correctors, who corrected each quaternion as soon as it was finished. These corrections are, however, interesting in themselves, and show how the trifling mistakes of the scribes frequently grew into more serious errors in the superficial attempts of the correctors to make sense of the passages thus affected and to emend the blunders. The Reginensis, thus emended, was copied in turn, in the eleventh century, to produce the Mediceus, and this manuscript was subsequently corrected by several hands. The Puteanus itself had, before the Reginensis was copied, received corrections from two correctors, P2, P3, in addition to those which were made in the manuscript by the scribe who copied it. Consequently, between the first readings of P, which were sometimes already corrupt, and the last corrections in M, there is sufficient latitude to exhibit several distinct stages in the growth of a corruption. It is the purpose of the following list of examples to illustrate this cumulative growth. Each of the errors here given has, in the hands of the scribes and the correctors of these three manuscripts, passed through at least two stages in the process of corrup