The Extant Greek Works of Apollonius of Perga, Vol. II

In that first location, not pros ekplo towards the sailing out, but only ek- out- is omitted; in the third, not eschaton epe- the final subsequent, but hos eschaton as the final. On p. 2, 15, after the second lacuna in codex 16, tharanes sweeping/clearing is read (corrected by second hand), and on p. 4, 25, after de but/and, peri about is added. Now, since the same readings are found in codex 20, at least one copy must intervene between V and codices 16 and 20; for it is not possible that one was copied from the other, because codex 20 on p. 2, 15 omits only ek- and on p. 2, 17, instead of eschaton epeleusomenoi the final ones who will arrive, it has edia ton eteleusomenoi a corruption of the original. Furthermore, in codex 16, Serenus's booklets lack an inscription, whereas in codex 20, they are inscribed Serene peri kylindrou tomes Serenus on the section of a cylinder and Serenou Antinseos philosophou peri kylindrou tomes of Serenus of Antinoe the philosopher on the section of a cylinder. From this it simultaneously appears that Nuremberg codex 20, century V, appendix 6 codex 20 was not copied from codex 6, as one would have expected, because it belonged to Regiomontanus Johannes Müller, a famous 15th-century mathematician; for there those booklets are inscribed Serenou Antinseos philosophou peri kylindrou tomes alpha Serenus of Antinoe the philosopher on the section of a cylinder, book 1 and Serenou Antinseos philosophou peri kylindrou (konou Bessarion) tomes beta Serenus of Antinoe the philosopher on the section of a cylinder (cone, Bessarion) book 2 (deleted by Bessarion); in V, the former booklet is inscribed Serenou peri kylindrou tomes Serenus on the section of a cylinder, the other has no inscription, but at the end of the former one reads Serenou Antisseos philosophou peri kylindrou tomes Serenus of Antissa the philosopher on the section of a cylinder—which subscription a recent hand changed into the title of the other work, adding to beta the second at the end and inserting telos tou alpha end of the first before it. That codex 9 is connected by a close relationship with codex 20 Turin B I 14 is apparent from the fact that it offers edia ton eteleusomenoi on p. 2, 17 (p. 2, 15 ek- and ou diaka- omitted in the lacuna), but since it does not have peri on p. 4, 25, neither was copied from the other; furthermore, on p. 4, 13, for syneidomen we have seen/understood, codex 9 has synoi.

Nothing therefore remains, unless we think that codices 9, 16, and 20 were copied from the same copy of codex V, in which pros ekplo and ou diaka- on p. 2, 15, eschaton epe- on p. 2, 17, and peri on p. 4, 25 had been omitted from the beginning, later supplied as pros plo on p. 2, 15 and, by an error of reading, edia ton ete- on p. 2, 17, perhaps from V itself.

Munich 576, Uppsala 48 Copies of codex 20 are codices 19 and 24, as these readings show: p. 2, 4 echoi might have] echei has 19, 20, 24; p. 2, 8 euarestesomen we may please] euarestesomen we will please 19, 20, 24; p. 2, 15 ou diakatharantes not having cleansed] tharanes 19, 20, 24; p. 2, 17 eschaton epeleusomenoi] edia (alpha written so that it becomes similar to the letter omega) ton eteleusomenoi 20, edio ton eteleusomenoi 19, 24; p. 4, 6 axonas axes] axonas 19, 20, 24; p. 4, 25 de] de peri 19, 20, 24. These locations demonstrate that neither was copied from the other: p. 4, 5 tas the] tous the by abbreviation 19, 20, tas corrected from tou or ton 24; p. 4, 9 kalo I call] 19, kalo followed by erasure of 1 letter 20, kalos beautifully 24; p. 4, 11 te and] 19, 20,