^{page 154 note 1} The fullest record of MSS. readings that I have at present is in book 26, c. 30 to the end of the book, but without Bλ in three chapters only A has alere (c. 30. 10) with PC⁴M² (see below), uelit (c. 31. 6) with some dett. against uelitet PR etc., exornarem (c. 31. 9) with M²D¹ against exortem PR etc.; and scio with P⁴ for socio PR; reliquum (c. 32. 6), relicum P¹C, relictum PRM. To these may be added e.g. talis uiri usa, (27. 34. 8) with P (before change). These scarcely suggest a descent from R or anything else; nor do they belong to such a class as the reading of 26. 31. 11 dimissi Siculi C: dimissis Siculi PRM: dimissis Siculis M²DA dett., a reading which only shows the superior intelligence of C at this point. According to Drak. M omits ipsa (25. 21. 5) and id (25. 37. 7); A has these; and there are other such instances. In 30. 30. 26, according to Luchs and my friend the late A. H. Kyd, B omits terra, which CA retain; in 30. 30. 30 B has quod ad id but CA quoad id, and in 30. 31. 9 quandam B, quondam CA; on the other hand we find portendisse deos C but portendisset eos BA. The gentle art of making stemmata on insufficient evidence is not yet lost. Like iron, verily it doth draw a man. Madvig's coupling of H and L in the first decade practically as twin-brethren always reminds me of the boy's translation of Ovid's genito cum fratre creatus etc., ‘I was born with a twin-brother who had been born twelve months previously.’

^{page 155 note 1} I.e. to where Rhenanus begins his citations from Spirensis.

^{page 155 note 2} The facts about Spirensis (eleventh century as is known from its one surviving folio at Munich) can be gathered best from Rhenanus in ed. Frob., Basle, 1535, Madvig's preface to his Emendations, Mommsen and Studemund's Analecta (with an account also of the Turin MS. now destroyed), and Luchs' Prolegomena.

^{page 155 note 3} They must be ascribed to A² until an earlier claim is established.

^{page 155 note 4} Valla says in offering variations (adn. on 23. 45) that he follows Hegio's example in the Captiui of buying up all the slaves so as to get the man he wanted; this plan is excellent, and atal to ‘jackals.’

^{page 155 note 5} That is Harl. 2684 (Luchs' H), on which later.

^{page 155 note 6} It is remarkable how difficult it is to obtain this work, at least in this country; also the minor edition of 21–30 (2 vols., 1888–9). Consequently, starting on the facts recorded in the Analecta, I retraversed some no small portion of his work on AαβγδεFV, and can testify to his wonderful accuracy.

^{page 155 note 7} By an unfortunate slip MrShipley, (in Journ. of Philology IV. 411, 1909)Google Scholar has confused these two MSS. The Harleian MSS. of Livy are numerous; but ne timete, commilitones; they are gradually decreasing: Frigell laid Harl. 2672: 70b nicely to rest, and now Mr. Shipley has rolled two into one.

^{page 156 note 1} Not E with Luchs, for Luchs has not dealt with the textual corrections; also he cites the addition 30. 41. 6 (classibus) to the end by the symbol E. A Spirensian scribe no doubt wrote it, but he played the game most nobly, and did not write Spirensian.

^{page 156 note 2} Luchs, , Prolegomena p. xxxxiGoogle Scholar., has rightly stated the relations of these MSS. A and λ. Apparently quoting from memory MrShipley, has misrepresented this statement and made A an in-and-out copy of λ (Journ. of Philology IV. 411–2, 1909)Google Scholar; alas! poor ghost, A, being a mere phantom of λ, he passes his insubstantiality into 2684 as though the leanest of Pharaoh's kine. That A^s is in any way a copy of L is sufficiently disproved by the omission by L of all Spirensian notes or additions in book 26, by the character of the Spirensian corrections given by either (on which further on): that A is a copy of λ can be also disproved by e.g. the retention in A of passages omitted by λ, such as 27. 14. 2 ad demendam … Romanis, and 28. 7. 7 petit.. naues. Can Mr. Shipley be thinking of Gebhard's Pal. 3 (Pal. Vat. 876) which has not the Spirensian section of 26? Anyhow, he couples A in his stemma with Vat. 1847 mentioned by Luchs, p. xxxxii.

^{page 156 note 3} My notes show the influence of A thus corrected on that beautiful Florentine MS. (called F by Luchs, which however he has overrated through his neglect of A^s) and on αβ;γδ;ε, also on Vat. Pal. 879 (Gebhard's Pal. 2), which all contain the supplement 26. 41–44. With A^s they all have the spurious link at the end of this, but A^s has written it in smaller letters and paler ink, as if from consciousness of guilt or desire not to obtrude.

^{page 157 note 1} The special characteristic of the corrections of A⁴ apart from the others is that he sometimes restores Puteanean readings (Professor Clark tells me he is supposed to have used Vat. Reg.); but he also supplies some Spirensian readings and a large number of purely Vallan readings, e.g. those at the beginning of 22.

^{page 157 note 2} In c. 45, fol. 172, he read, according to the published emendations, sederent, but in A he has, as might have been expected from his scholarship, left siderent (fol. 172); this is one of the notes he has signed. The words as emended are ee, cymbis, uadis: uadis is restored by erasure of a g written on the top of d: the y is written on an erasure (probably of i or u): ee is written on an erasure so clumsily done with a blunt knife that V made a mess and hadto write ee again above.

^{page 158 note 1} I had no reason to record all when I was doing the last half of the decade, which was before working on the first half. A characteristic of the A^v emendations is that they are mostly preceded by c' and that again often preceded by d. This d is very frequent also where there is no emendation: this again is plainly the mark of a man at work, and I have little hesitation in saying that in all places that man was Valla himself.

^{page 158 note 2} Though he did attack Livy on the Tarquins. A has lost the folios (long ago gone, probably to make laxas scombris tunicas) which contained the greater part of the first three books;and with this loss may have gone a draft of Valla's dissertation.