Some XV-Century Truths in Apollonius Speake, Graham Greek, Roman and Byzantine Studies; Spring 1975; 16, 1; ProQuest pg. 103 Some XV -Century Truths in Apollonius Graham Speake THIS ARTICLE concludes my discussion of the D-manuscripts of Apollonius Rhodius. In an article entitled "The So-called D­ Manuscripts of Apollonius"1 Francis Vian and I considered the integrity of the D-group, its relation to the rest of the transmission, and the editorial techniques employed by its scribe, Demetrius Moschus. We established the independence of each MS. and concluded with the possibility that any single MS. may be the unique witness to an otherwise lost reading of either of the first two families. I followed this with a paper on "The Scribal Habits of Demetrius Moschus"2 in which I classified and discussed the degenerative changes that occurred in these MSS. as a result of Moschus' tampering with the text: he emerged a reckless fellow but by no means a stupid one. In this final episode I shall attempt to penetrate beyond the tangled web of corruption and contamination that has grown over these Mss. in order to lay bare what is most valuable to us in our reconstitution of the text, namely the unique preservations of the truth, readings which may be described as ancient or Apollonian and which are not preserved or conjectured in any other established line of the trans­ mission. In the pages that follow I list the readings from Mss. of the D-group that I would print in a text of the Argonautica. We have already seen that Moschus, fortified by his good knowledge of Homer and his own attempts at hexameter poetry, felt no hesitation in changing and correcting-or even improving, to his mind-the copy before him. He was thus a more significant force in the transmission of the text than the traditionally selfless 'Byz. anon: Nevertheless he was the scribe of four Mss. of the D-group and I shall continue to refer to him as 'the scribe', having thus warned the reader not to be misled by my use of the term. In assessing the good readings carried by Mss. of the D-group I shall be concerned primarily with whether or not they provide us with a text that Apollonius wrote (or at least might have 1 GRBS 14 (1973) 301-18. 2 GRBS 15 (1974) 113-33. 103 104 SOME XV-CENTURY TRUTHS IN APOLLONIUS written); but mindful of the warning issued above about Moschus as 'scribe', I shall not hesitate to speculate on how these readings come to be found in our MSS., even if it may appear to the reader that the inevitable and sometimes insoluble contest between textual authority and scribal interference is of purely academic interest. As before the Mss. discussed are: M Milan, Ambros. 426 (H.22 sup.) (Books 1 and 2), early XVI century R Vatican, gr. 1358, ca 1505, Demetrius Moschus Q Vatican, gr. 37, ca 1491-1514, Demetrius Moschus C Rome, Casan. 408 (G.III.5), 149Q-1510, Demetrius Moschus D Paris, gr. 2729, 149Q-1510, Demetrius Moschus d Collective siglum for the group M R Q CD The lemmata are taken from Frankel's Oxford Classical Text (Oxford 1961). Other editions cited are those of Brunck (Strasburg 1780), Wellauer (Leipzig 1828), Mooney (Dublin 1912; repr. Amster­ dam 1964), Vian (Book 3, Paris 1961), and Livrea (Book 4, Florence 1973). BooK 1 148 Tovc 7}yt: M R Q, Tovc 8' 7}yt: cett. The relative Tovc is supported by Frankel's references (23, 35, 119, 180, 212). It is surely a conjecture in our MSS., one that was first made in modern times by Herwerden. Vian argues for the parataxis.3 285 Kt:vt:oict d, Kt:v eoict cett. This is the reading of schol. P and has stood in the text since Brunck's edition. Perhaps it is a happy accident in d, though the scribe's eye could easily have strayed to the scholia. None of the d-Mss. in fact contains scholia, and this is perhaps an argument for suggesting that Moschus' working copy did. 335 brt7Tvt:vcovctv] Jm7Tv£vcwctv R Q CD. The subjunctive is supported by PSI 1478. There is a close parallel (but with Kt:) at Odyssey 9.138f: , ., ' HC 0 K£ VCXV'T€WJI (J ' ' , ' , , ,_ VJLOC £7TO'TpVvrJ Kat €7TL7TJI£VCWCtJI CXTJ'TCXL. We have seen so many instances of change for the sake of Homeri­ cism4 that this is most probably a conjecture. 403 lTT&Knov M, J7TaKTlov cett. The genitive is absurd here when followed by aKTlov in 404. Brunck's parallels (359, 2.689) would suggest that this is a conjecture; but it could be a mere accident. 8 REA 75 (1973) 93 (ad 2.376). 'GRBS 15 (1974) 118-22. GRAHAM SPEAKE 105 565 xevav R Q, coni. Brunck, xevov cett. The aorist is clearly required here. This verb regularly causes confusion, cf. Frankel's references. 625 86avn] ylpovn M R Q. It is easier to consider this a preservation of the truth which in other MSS. has been displaced by the intrusive gloss 86avn than to account for corruption in the opposite direction. 643 o olD, oc ol cett. We may assume that Apollonius had some rudimentary knowledge of the digamma5 which precludes the reading ofoc here; the corruption is common in MSS. ofHomer (cf Wellauer ad loc. and Leaf on Iliad 6.90). We may with equal certainty assume that Moschus knew nothing of it except what he had observed from reading Homer and that any Mss. of Homer he may have seen are likely to have contained the corruption. It follows, then, that this would be a conjecture beyond his powers and must be a case of genuine preservation. 6 802 p:ijvtc C, p.7Jnc L, p.rrhc P.Oxy. 2698, p.ijnc cett. Commenting on the superscript letter in the papyrus, Kingston writes: "If this letter is v, fLfjvtc, as conjectured by Frankel, was at least a variant in anti­ quity. It is also interesting to note that, although the coincidence is no doubt fortuitous, L, a tenth-century MS., not only reflects an ancient pair of variants, but also reproduces, almost letter for letter, the word+ variant arrangement of ll."7 The superscript letter in L has been satisfactorily explained by Campbell,8 who at the same time commends Frankel's conjecture. It is not so easy to account for the appearance of the truth in C. The scribe might be thinking of the first line of the Iliad, where Achilles' fLfjvtc is ovAofL€VYJ. It is tempting to class this reading as a genuine preservation, but our knowledge of the scribe's tendency to Homericize at every possible opportunity prevents our considering any origin for it other than degenerative change. 805 a1T£cc£vovTo D, br£ccEvovTo cett. Even Wellauer, who has little respect for D, accepts this reading. In Homer the word is used in the passive to mean 'to flee'; the active occurs at Nic. Ther. 77 and AP 9.642. The unfamiliarity of the word suggests a genuine preserva­ tion, but it could be accidental confusion of the prepositional prefix. 811 Kovpat] 7'€ Kovpat C, -r€ Kopot LA, -r€ Kopat R Q D S E. With the 5 See Mooney, Appendix ll, p.416. • Vian retains oc, cf his Recherches sur les Posthomerica de Quintus de Smyrne (Paris 1959) 154. 1 P. Kingston, ed., Oxyrhynchus Papyri 34 (1968) 70-71. 8 M. Campbell, CQ N.S. 19 (1969) 271. 106 SOME XV-CENTURY TRUTHS IN APOLLONIUS exception of H.Cer. 439 Homer and Apollonius use the form KovpTJ exclusively. C's retention of TE suggests that, if not an accident, this reading is a conjecture, perhaps influenced by the occurrence of the word at 801 and 818.9 944 ~eplflov'To M R Q C, ~Epl.8oVT(n cett. Past tenses are needed both here and in 942. Emendation here is easier than two lines above. 955 elpvcco:vTec] £pvco:vTec C, £>..~..vco:VTEc cett. C' s £pvcavTEc may be closer to the truth than the quotation from Heliodorus' scholia to Dionysius Thrax accepted by Frankel: £gEpv­ cavTEc would explain the compound in EKAvco:vTEc, the reading of most Mss. Alternatively it could be a corruption of elpvc(c)o:VTec. But particularly in view of D's behaviour at 987 (£gl.pvco:v for £g?j>..o:co:v) and 1276 (EK a· apo: VTJOC by tmesis with EpVCO:V'TEc), the reading of c here seems most likely to be a conjecture. Vian argues persuasively10 for the retention of the vulgate EKAvco:vTec. 1019 c/>o:Ti~eTo:t .;;a· E'n] c/>o:Ti,eTo:t t:lci.T, D. Was this taken from the scholia Parisina to 1.1109, or an independent conjecture, or a genuine preservation? Perhaps most likely it is an ancient variant which has found its way into the text. 1233 p.o>..te M, p.oytc cett. p.oyte is the Homeric form (which we are not surprised to see retained by the Moschan Mss.), but our texts of Apollonius generally show a preference for the later form. Through M it reached Vat.Pal. 150 and the Aldine edition, but its appearance in M is less easy to account for. Perhaps the scribe is reminded of 3.634, but since he does not even copy Book 3 we should not discount the possibility of a genuine preservation.11 1323 avo:7TA?}cetv] aV0:7TA~CO:t M. Platt was probably right12 to prefer the aorist after EK7TA~co:t in 1318 and EKTEMco:t in the scholia Parisina. But the proximity of EK7TAfjco:t also prevents our considering any possible source other than conjecture by the scribe of M. BooK2 34 dptTpecf>loc] dpt:t'Tpt:cf>l.oc RQD. Purely a matter of orthography. 108 aeet'TEpjj CKO:t~C MRCD, at:et'TEpfi CKO:tjj QLAG k. This may • Vian retains T£ KOpa.' on the grounds that its correction introduces too many con­ secutive spondees. 1o REA 72 (1970) 94-95. u Vian believes that here Apollonius uses ,_uSy,c exceptionally because he is quoting II. 21.417. u JP 35 (1919) 84. GRAHAM SPEAKE 107 be the reading of SPc, with which MS. d not infrequently shares an interesting reading.13 On the other hand, the truth would be obvious to any scribe who gave a moment's thought to what he was copying. Either this is further evidence of a link between d and w or Moschus is thinking for himself. 119 tf'€Aav TETaywv] f'aAa TETaywv D, f'Eyav (p.£Aav) TETaywv cett. P.Oxy. 2697 preserves avTeTay(wv) in the margin but not, unfortu­ nately, the reading of the text. "The conjecture of Sanctamandus seems to be confirmed," writes Kingston. "But what did the text read? To write aVTeTaycfw in the margin is not a wholly effectual way of correcting f'EAav TETaycfJV to 1-'&X avTeTaywv, if this is what the corrector had in mind."14 I suggest that the text read not f'EAav TETaywv but the unmetrical f'aAa TETaywv. The rarity of the word avTeTaywv is not, as Kingston thinks it is, the reason for its presence in the margin, but it is the reason for its corruption, which involves no more than the removal of one letter. Now al!fa f'aAa is a common Homeric combination (cf e.g. Iliad 4.70), substitution of which would be by no means beyond the powers of Moschus. How are we to decide whether the reading is the result of genuine preservation of an ancient error or another Homericism no older than the pen of Demetrius Moschus which brought him within a v of the truth? 243 xepl d, XELP;, LAG k. Vian confirms that this is the reading of S and perhaps of BPc, but the possibility of independent metrical con­ jecture cannot be excluded.15 342 Kat Q, KE cett. A phonetic 'error' on which one would hesitate to base any argument.16 389 Kat CD, Ke(v) cett. Brunck adopted this reading from D and not, as Frankel's apparatus suggests, by conjecture. Moschus, how­ ever, could well have conjectured it or even hit on it by phonetic ·error'. 399 KVTadSoc M Q D, KVTalSoc cett. Purely a matter of orthography facilitated by a glance at 1267 and 4.511. The alternatives are weighed by Campbell, who "would read KvTalSoc and label D's KvTadSoc as •fort. recte'. "17 u Cf. GRBS 14 (1973) 311. u Oxyrhynchu.s Papyri 34 (1968) ~7. u Cf. GRBS 15 (1974) 114-18. u It is also the reading ofK and W (= Vrat. Rehdig. 35) and the L scholia. 11 RevPhil 47 {1973) 73-74. 108 SOME XV-CENTURY TRUTHS IN APOLLONIUS 460 ava~Ec0aL R Q, ava~acOaL cett. An obvious conjecture-and I am not convinced that it represents the truth, cf. 172 and 1.1343. 474 -nc MRQD, -n cett. The truth is so obvious that one cannot consider any source other than conjecture by the scribe. 498 E7'~CLUL avpaL d, E7'~ct0L avpa£ cett. Cf. 525, where LAGMRQ read €-r~na' and S k CD €-r~not. Frankel's explanation of the corrup­ tion is attractive: "commixtis oi ET"fJdat-accentum nota-et ai E7'~CLOL avpaL." But if ET"f]ciaL avEp.OL is admitted (Hdt. 2.20), why not E7"'fJc£a, avpaL? At all events the likelihood of attraction to the ending of avpaL restricts our verdict to one of conjecture by the scribe. 499 avwyfj M R Q, apwyfj cett. This is the rarer word and is surely more suitable when it comes to Zeus' role in connection with the winds. Pace Mooney, there seems to me no reason why 524ff should have any bearing on this word and, pace Frankel, 556 cannot be the source, as there M R Q read apwyfj. Conjecture is possible, but in view of the rarity of the word it seems most likely to be a genuine pre­ servation. For similar confusions cf. 1.1134 avwyfj m, apwyfj W, 2.556 &vwyfj V 1 A s G' apwyfj k. 705 7Te-rpatv v7To DELpacn] 7Te-rpalfic v7To DEL pan D. Also read by Etym.Gen., according to Frankel.I8 In his introduction to the fragments published in part 34 of Oxyrhynchus Papyri, Kingston writes that "coincidence in superior readings between D and the Etymologicum Genuinum suggests that D had access to a line of the transmission which bypassed the archetype." My collation of d has yielded no addition to the three examples listed by Frankel in his Praefatio Critica (the others being 3.201 and 278, q.v.). Here there seems no logical reason for alteration from singular to plural (or vice versa),19 but we can never be sure that we are dealing with a logical scribe. Kingston may yet be proved right, but it is with some hesitation that I would classify this reading as a preservation of the truth. 718 KE'Lv• M R, KE'Lc• cett. Ke'LcE is often dubious in our texts of Apollonius, cf. 1.955, 1224, 4.1217, 1239. For KE'Lvo cf. 4.1153 ..... ' , I ,.. C' ' \ 'Y Jl KEtvO Kat ELC€7'£ VVV LEpOV KI\T}L':>E7'UL av-rpov M'TJOE{TJC. 18 Professor Vian tells me: "Selon Ia collation ordinairement tres sure qui m'a ete communiquee par J.-M. Jacques, le Genuinum en 2.705 ecrit: 7TETpat.,., v,o 3E,pa3,. n semble done que Frankel se soit trompe." He also tells me, however, that 3E'pa" is clearly the reading of Eac here. u Professor Lloyd-Jones draws my attention to Soph. Aj. 697 '1T£Tpala.c &,o 3E,p&8oc, but an isolated parallel from tragedy can hardly affect our decision here. GRAHAM SPEAKE 109 It is perhaps a conjecture, but more probably an ancient variant (in w? cf 1.1224). 795 Jcf>' D, vcfo' cett. v1ro is defended by Campbell.20 If brl is right, it can be no more than a scribal conjecture or chance confusion of prepositions. See Campbell (loc.cit.) for further examples of confusion involving prepositions in Apollonius. 829 alyav€n C, alyav€'Y/v cett. When op€yof-LCX£ is followed by a geni­ tive of the thing aimed at, the instrument must be in the dative, cf 1.1313 and Iliad 13.190, though, as Frankel points out, 2.1212 is an exception.21 The scribe might know that, or it could be a lucky slip. 843 v~toc M, v'Y/lov cett. v'Y/lov produces a not impossible hypallage, but the proximity of JK KoTlvo£O makes the corruption from v~toc to v'Y/lov more likely than the reverse. 874 &A.A.ot MD, ~,\A.ot cett. This word is regularly confused in our MSS. Cf 1.1101 and Frankel ad loc. 940 o' M C, T' cett. Clearly the preferable connective and a simple enough conjecture. 987 ltr~Tu:c D, lm]Tvec M, Etr'Y/TEEc cett. "Die vielleicht richtige Form," says Frankel of D's Jtr~nec,22 but we can make no assertions about the source of a reading the truth of which depends merely on orthography. BooK 3 81 aloe D, aZye cett. This was conjectured by Platt, accepted by Frankel, and correctly ascribed to D by Vian. Did Moschus know that oye in Homer is substantival? 201 lp,1Tecfovactv R C D,eKtrEcfov(K)actv cett. Another reading coincident with Etym.Gen., cf above on 2.705. lf-Lcfovw is so much commoner than lKcfovw from Homer onwards that this could be the conjecture of an intelligent scribe. 264 EtrETEAAeT' R Q, etrenlAaT' cett. Also read by P.Oxy. 874. An imperfect suits the tense of BvflcKwv much better than an aorist. But what scribe would think of that? After all, no editor did before the publication of the papyrus. 278 trpooof.Lcp E'vt] trpoOofLov lv'i Q CD, trpooop,cp E'vt cett. This reading, also found in Etym.Gen.,23 gives more support to Kingston's theory 20 CQ 21 (1971) 414. 21 Cf also Bi.ihler on Mosch. EltT. 112 and Fraenkel on Aesch. Ag. 1111. 22 Einleitung zur kritischen Ausgabe der Argonautika des Apollonios (Gottingen 1964) 91. 23 The actual reading of Etym.Gen. is €vt cpAtfi 1rpo86p.ov. 110 SOME XV-CENTURY TRUTHS IN APOLLONIUS (cf above on 2.705) than either of the other two. Wilful alteration to the genitive here is improbable and, although it could just be a slip, genuine preservation seems highly likely. 291 xeva:ro Q C, SeveTo S, xeveTo cett. A common confusion, cf 1.565 and Frankel ad loc. 316 o1T1T'[1 TE R, o7TYJ -re Q CD k, o1rmhe cett. Vian now denies that 01T1T'[1 TE was the reading of sac,u but it is in H and N. Palaeographi­ cally this is the most likely solution, and it is so easy a conjecture that there is no difficulty about its appearance in unrelated MSS. 379 a1To1TpolYJKa. D, l:rrLTTpolYJKa. cett. Both words are Homeric, and a1ro- appears to give the better sense. A conjecture. 401 ayopEVOLC Q, ayopEVELC cett. The optative is read by H and N. See above on 316. 548 aOep{getv R Q D, aOepl,ELV cett. The praesens propheticum of most MSS. is the lectio difficilior (cf Kuhner-Gerth 1.195ff), but the reading of R Q D is attractive, especially in view of €ccec(}a., in 550. Another conjecture. 606 8wwylpovct D, 8wwnfpotn cett. Pace Lloyd-Jones,25 we do not need a word here to mean 'chiefs, nobles' but rather one for the common people. Such a word is provided by most Mss., and Frankel and Vian are right to print it in the text. 8'YJfLOYlpwv is Homeric (Iliad 3.149), as 8YJJ.Lchepoc is not; once again Moschus is tampering with the text, and this alteration should have been printed in my previous article on his scribal habits under the heading 'Homer­ icism'. 879 Dtege>.a'YJCL R Q C, 8tege>.acvct cett. This conjecture, which appears also in Vat.Pal. 150, is no doubt influenced by €>.a.ev in 872. 909 fLETa ccf>lctv R Q D, Ka.Ta ccf>lctv cett. P.Oxy. 691 has JL[ eTa, which was conjectured by Stephanus and, in view of Iliad 1.368, is obviously correct. Here again Moschus is either turning his knowledge of Homer to excellent use or offering another genuine preservation. 1034 €vl Q CD, £1ri cett. If this is the truth, the scribe probably hit on it by conjecture or by simple confusion of prepositions. The passage is elucidated by Campbell, who supports €1rL26 1086 £vppetT6c d, £vppvToc k, £vppYJvoc LAS G. The reading of d seems to have been a variant in the archetype, cf Frankel ad loc. The 24 He believes that 01T7TO'T(i has been corrected to cm1TTJTE. zs CR 13 (1963) 157. 2e CQ 19 (1969) 280. GRAHAM SPEAKE 111 scribe may have found the correct spelling either by genuine pre­ servation or by phonetic 'error'. 1155 avwvcav R Q D, ap' loiicav cett. "Est ea correctio librarii, quali­ bus plenus est deterrimus ille codex, eaque minime necessaria." So comments Wellauer, who, had he but cast his eye rwo lines above to 7fi o' ap', would have seen an excellent reason why this should not be a 'minime necessaria correctio', and had he but collated D for himself, would have been less hasty in his condemnation of it. 1172 EJLEAovTo CD, J.LEAoVTo k Qac, EjL€AAovTo R, jL£AAovTo LAG QPc, IL{!LvovTo S. The reading of CD is also the lemma of the scholia Parisina; but Frankel is right to suspect the passage.27 1310 lpmovTa D lac, bnovTa cett. "Eine schone Verbesserung eines hoffnungsvollen liebenswi.irdigen Schi.ilers Breidenbach," wrote Wilamowitz.28 This is the stock epic phrase for falling on one's knees, but it seems too bold a change for Moschus to have made himself. We have already discussed the significance of the agreement with tac.29 BooK 4 94 4>cfw1Jc€v TE R QD, Bapcvv/.v TE cett. Frankel's reasons for accepting 4>dw1Jdv are rejected by Vian30 and Livrea. 17pocrrTvgaTo 4>dw1Jc/.v TE occurs at 1.1331, and the last two feet are so common in Homer for the introduction of a speech that we cannot but suspect Moschus yet again of Homericism. 142 l.Atyyotctv Q, elAtyyotctv cett. Purely a matter of orthography. 182 ~~ R Q D, ~8~ cett. Sense demands ~~, and Moschus could easily have found it by conjecture. 233 €vmA~cH R, €vt7TA~at (sic) D,31 E7TmA~cH cett. This is obviously the truth, cf 3.1350 and Odyssey 19.117. Wellauer attributed it to "praepositionum confusio frequenrissima," but it could be a con­ jecture. 247 ev'YJA~v d, ev'YJAfi cett. The accusative is clearly required and easily conjectured. 345 tTo] ToSe R Q, Toye S, To cett. The reading of R Q is not neces­ sarily ancient, but it is at least a possibility, which is more than can be 27 Noten zu den Argonatttika des Apollonios (Munich 1968) 433. 28 Hellenistische Dichtung II (Berlin 1924) 251. 2t GRBS 14 (1973) 309-10. 3° REA 75 (1973) 88 n.5. 31 Livrea's report is inaccurate. 112 SOME XV-CENTURY TRUTHS IN APOLLONIUS said for To which Frankel prints (with an obelus). His attitude to this crux is uncharacteristically defeatist, though he does propose Toye 8~, which is accepted into the text by Livrea. At best R Q present the truth by conjecture; but there would be a case for considering the reading as a degenerative change for metrical reasons. 400 ayoL'TO R c D, ayoLV'TO cett. The singular is preferable as referring just to Apsyrtus, and corruption to the plural is much more likely than the reverse. But the scribe is meddlesome enough for conjecture to be as likely as genuine preservation.32 636 &fNccpaToL R Q, &OI.ccpa.Tov LAS G C, &OI.ccpa.Tat k D. Apollonius is concerned not with the size of continents but with the dangers his heroes faced on the waterways of Europe. So it is the lakes that need an awe-inspiring epithet (cf {Ja.Ovv 627, KVKWfLEVov 629, aTTelpova 633); but corruption to &OI.ccpa.Tov after ifTTnpov would be hard to resist. k D retain vestiges of the truth with &OI.ccpaTaL assimilated to atTE in 635 (or even to TTITTTavTat ?). I do not believe this is a correction that would occur to a scribe. 873 &ve7T(XAfLEVoc d, &vecTTUAfLEVoc G, &vaTTaAfLEVoc cett. This reading is correctly ascribed to D by both Brunck and Wellauer and is not a conjecture by the latter as stated in Frankel's apparatus. The verb is &vecpaAAofLaL, and, pace Livrea, the reading of dis supported by 2.825 and Quintus Smyrnaeus 1.140. G reads &vecTTaAfLEvoc, and in view of the close link established between d and w,33 this is most likely a genuine preservation. 1429 TTaAw EfLTTaAtv D, TTaAtv EfLTTE8ov cett. Frankel questions the propriety of EfLTTE8ov with an aorist verb34 and suggests TotaL TTI8ov EfLTTaAtv as a possibility. D's TTaAtv EfLTTaAtv is clearly a palaeographical slip, but once again we have to decide at which end of the trans­ mission corruption occurred. If it represents a corruption of TTaAtv EfLTTe8ov, it is mediaeval (or rather Moschan), but if it is a corruption of EfLTTaAw, then it must be ancient and a genuine preserva­ tion. I am inclined to favour the former but do not entirely discount the possibility of the latter. 1570 ltavayet d, ltevayet k, ltavl.xE£ cett. A lucky slip, as proved by the fact that it is also in N, which we now know to be an unrelated Ms.ss 31 D spoils his good work by writing 1Tetrr/p for 1TetTpl in 399, cf GRBS 15 (1974) 131. aa Cf GRBS 14 {1973) 311-14. 34 Noten 601. But see Vian, REA 75 {1973) 97 n.l. as Cf GRBS 14 {1973) 303--04. GRAHAM SPEAKE 113 1615 iTT"i vw:)B, R Q C, €mvt:t68, cett. Chance again. e7Ttv€L68, would be a a:rra~ here, but the corruption hardly surprises us. The contribution made by d to the text of the .A.rgonautica is small but not insignificant. The ancient readings which it offers us are over­ shadowed by a greater abundance of corruption and degenerative change. They are there none the less, and now that we know more about d's position as an independent witness in the transmission, we may adopt them without hesitation. As a further result of this newly established independence, d, while perhaps not adding a great deal to the text, should become a more frequent visitor to the apparatus criticus, taking its place alongside m, w and k; and the fact of inde­ pendence is an eloquent demonstration of how many gaps remain in our knowledge of the transmission of ancient texts. This discussion of the d group has sidelighted the working habits and editorial technique of a previously little-known Renaissance scribe and added a new name to the history of classical scholarship. But the discovery both of his erudition and of his audacity in tamper­ ing with the text has a direct relevance to the readings discussed in this article, and our decision as to how they come to appear in d must be considered and cautious.36 CHRIST CHURCH, OxFORD july, 1974 as I am grateful to Professor F. Vian for criticism of this paper in typescript.