¹ In his introduction to Aristotle's Physics (Oxford, 1936)Google Scholar, 79. References to Ross will be to that volume. I also follow his numbering of the pages and lines in Book 7, as does Wardy, Robert in The Chain of Change (Cambridge, 1990)Google Scholar.

² Shute comes to these conclusions in On the History of the Process by which the Aristotelian Writings arrived at their Present Form (Oxford, 1888)Google Scholar. Indeed, he goes so far as to claim that both were inauthentic, devised by ‘rival schools’, the second after Alexander's commentary, since he only mentions one (p. 123). Shute gives a more moderate account in ‘Aristotle's Physics, Book 7’, Anecdota Oxoniensia 1.3 (1882)Google Scholar, where he gives his analysis and recension of the text. Ross knew only the latter, and it may even have been Shute's more mature account, since the 1888 volume was published posthumously by friends.

³ He does, in the end, opt for his ‘weaker’ thesis that the α version was written by Aristotle, the β version probably by one of his immediate students. See Wardy (n. 1), ch. 5.

⁴ Indeed, Lang, Helen had already in ‘God or soul: the problem of the First Mover in Physics 7’, Paideia (1978)Google Scholar, Special Aristotle Volume, effectively argued for the unity and coherence of Book 7 without any hint of the Wardy thesis. In a more recent work, Aristotle's Physics and Its Medieval Varieties (Buffalo, 1992)Google Scholar, she has called our attention to reading carefully what at the beginning Aristotle tells us he is inquiring into. In this spirit, all that Book 7 tells us is that ‘The moved is necessarily moved by something’.

⁵ This is Ross's dating, early in the century. Gercke, A., ‘Aristoteleum’, WS 14 (1892)Google Scholar, suggests a date of late 10th, early 11th c. More recently, Irigoin, Jean, ‘L'Aristote de Vienne’, JÖByz 6 (1957)Google Scholar has given convincing argument, on the basis of scholia relating it to other datable texts, for placing this MS in the 9th c., perhaps as early as 850. This would make it clearly our earliest extant copy of some of Aristotle's major texts. Herbert Hunger, in his more recent catalogue of the library's MSS, Katalog der griechischen Handscriften I (Vienna, 1961)Google Scholar, accepts Irigoin's dating.

⁶ Though not discrete ones. For penetrating and accessible accounts of the two families of MSS and their lineage and interaction, see Harlfinger, D. (ed.), Griechische Kodikologie und Textüberlieferung (Darmstadt, 1980).Google Scholar

⁷ Much of the research for this paper was work done at the Aristotelesarchiv in the Institut für Altertumskunde of the Freie Universität Berlin in the summer of 1990. I am grateful to my hosts at the institute for their kind hospitality, and to the University of Kentucky Research Foundation and the Southern Regional Educational Board for their assistance in travel expenses.

⁸ Ross (n. 1), 15.

⁹ This story is made more complex by the history of handwriting and copying. The minuscule style of MSS (which became the standard style for medieval codices) began only in the eighth century. By the tenth century, many of the older MSS from which the minuscules were copied had themselves been worn out or discarded because they were both unhandy in volume and too difficult to read in style (though Robert Browning, ‘Recentiores non deteriores’, BICS 7 [1960] offers examples from the Byzantine Renaissance of the 13th c. when copies were still being made from uncial texts from as early as the mid 8th c.). While this transliteration saved many texts from extinction, it did so at the price of cutting off access to earlier copies, so that our MSS J and E may be very close to as far back as we could expect to go.

¹⁰ At this point it becomes necessary to differentiate between the numbering and labelling of ancient Greek books. Works that had a continuous history of ordering, like Plato's Republic and Laws, received letter designations that were both labels and numbers. Thus, the letter digamma serves as a label for sixth in the list, and also for the numeral six. In Hellenistic and Imperial times, matters became more complicated. Beginning with certain standardization of size of books for the Alexandrian library, this led to a match with certain works like the dialogues of Plato, but required that others, such as the works of Homer, become subdivided. Works that became subdivided, like the books of Homer, and works that became assembled, like Aristotle's Metaphysics, had their books labelled according to the current alphabet, and so the digamma/stigma dropped out of the order; works whose number of books exceeded the number of letters in the alphabet had their books ordered by number, and so in these cases the digamma/stigma was used as a label, not as a number, and those for Plato's works as numerals. Needless to note, such complications could lead to confusion as well as to manipulation on the part of copyists and editors.

¹¹ That the titles are in small uncial script is standard for the period of the MS, and that the alterations are in the same form of script seems clear. D. Danielis de Nessel, Breviarium & Supplementum Commentariorum Lambecianorum, in which he gives his 1690 catalogue of the MSS then held in the Augustissima Bibliotheca Caesarea Vindobonensis, offers a brief account of our MS in vol. 4.58. His only specific notes are to the Theophrastus portion of the codex, including a variation in title and a ligature variation with a related MSS. That he makes no comparable note of our title alterations in Physics 7 should not lead us to suppose that they were not there. Neither Gercke nor Irigoin nor Hunger make note of them either. Irigoin, in his dating argument, notes that the scholia on this MS are by the same hand as MSS in Paris and Venice known to be from a scholarly community in the latter half of the 9th c. It would be sheer speculation to suppose that the changes are by that hand (since the evidence may be too slight to judge the uncial hand of the changes), and more speculative yet to suppose that the changes were prompted by knowledge of other MSS and traditions available to that group of scholars.

¹² Ross (n. 1), 15 interprets Simplicius as maintaining that Eudemus ‘omitted book 7’ on the basis of 1036.13 ff., that he ‘passed over’ it (p. 3). Runner, H.E., The Development of Aristotle Illustrated from the Earliest Books of the Physics (Amsterdam, 1951)Google Scholar, 81 complains of this interpretation, translating τοῦτο παρελθὼν ὡς παριτεόν as ‘passed it by as being superfluous’. This he takes to presuppose that Eudemus not only had it to take into account, but that it was already in its present position, and Eudemus merely regarded it as not worthy of comment. However, we could just as well take παριτεόν to mean here ‘beyond the regular number’ or ‘extraordinary’ and not conclude with Runner that it had already found its place in the corpus as we know it. This would give us a very different sense of ‘passing by’ from Runner's reading of it. Indeed, the whole discussion presupposes that Eudemus is following a single treatise already arranged something like ours. The more likely account is that Andronicus is using Eudemus’ Physics as his model for the arrangement for that of Aristotle, and that it was he who inserted at this point our seventh book in its place because it was Andronicus’ judgement that the subject matter was similar to the last book.

¹³ He does not follow Simplicius in his counting of books of principles and books of motion, but he follows Porphyry and the other major commentators in numbering four each.