Erasmus was bluffing?
No reason to impugn his integrity, he had seen mss. all over Europe over many years, some of them are even in Grantley's footnote. On p. 320 he does not throw out the bluffing red herring.
BCEME - 19
Erasmus claimed to have inspected more manuscripts than Valla ever did. (In fact Erasmus was bluffing here. His text of the Catholic Epistles was as yet based only on three manuscripts: GA 1, 2815 and 2816.)17
17 P.-Y. Brandt 1998, 121–122; ASD VI-3:1–12; Krans 2006, 335–336; and ASD VI-4:1–6, 484, identify
the manuscripts used by Erasmus. For the first edition he consulted the following manuscripts: GA
1eap (used for proofreading and annotations); GA 2e (printer’s copy); GA 817e (proofreading and
annotations [Theophylact]); GA 2814rK† (the second volume of GA 1, borrowed from Reuchlin);
GA 2815ap (the second volume of GA 2; used for printer’s copy and corrections, and as principal
source for Acts, Heb 12:18–13:25, Catholic Epistles); GA 2816ap (used for corrections; Estienne
used the first volume as his γʹ [= GA 4e]); GA 2817p† (printer’s copy, used for corrections and
annotations). Since Erasmus counted manuscripts comprising multiple volumes as a single manuscript,
his claim to have used five manuscripts at Basel (a claim questioned by Krans 2006, 335
n. 1) is correct. Erasmus also drew on notes taken from manuscripts he had examined in England,
and his edition consequently contains readings taken from GA 69 (text and annotations; see ASD
VI-3:10–11) and 2105p† (annotations [Theophylact; not noted by Brandt]). For the 1519 edition,
Erasmus integrated some corrections from GA 3eap and a manuscript of the Gospels from St Agnes
at Zwolle; see ASD IX-2:191, note to line 461; and Erasmus, Epist. 504, 516. For the 1522 edition,
Erasmus added material from the Aldine edition and Montfortianus (GA 61eapr). The 1527 edition
integrated variants from the Complutensian edition. In the 1535 edition of the Annotations,
Erasmus recorded variants from codex Vaticanus (GA B/03), supplied to him by Juan Ginés de
Sepúlveda, in a letter sent in 1533: see the annotations on Mk 1:1, Lk 10:1, Lk 23:46, and Acts 27:16
(Erasmus 1535b, 113, 187, 216, 331). Erasmus found the agreements between the readings in GA 03
and the Latin Vulgate suspicious, though in the case of Mk 1:1 he thought that Vaticanus had the
correct reading. See also his comments on Vaticanus in 1535a, β3v; and the commentary in Semler
1764, 75–76. In his Apologia to Stunica, Erasmus mentions having used manuscripts in England,
Brabant and Basel; on these manuscripts (one Greek, one Latin), see ASD IX-4:55, note to line
855; and 327, note to line 250. Tregelles, in Horne 1856, 4:108–111; Rummel 1986, 35–42, 195; Elliott
2009c, 244.
BCEME - p. 320
22 Lee had argued that if this variant was so important, it would have been mentioned by Valla, who had seen seven codices of the Greek text; see Erasmus 1520, 200–201. In his reply to Lee (ASD IX-4:323, 326), Erasmus points out that Valla was a fallible human, and that he himself had seen more than Valla’s seven codices, all of them lacking the comma.
Plus, do we even know if Valla's seven Greek mss. all were complete NT texts?