The Consolation of Philosophy

I would have preferred to use the complete L; but L is somewhat inferior to codex T see 32, 6 and has its own orthography, to which we now descend.

In establishing the orthography, besides the codices P which preserved certain things faithfully, such as coherceo to restrain and urgueo to urge, while generally exhibiting more recent spellings like iocundus pleasant, T whose copy F does not entirely agree in orthographic matters, and V written around the middle of the 9th century, just like P, one must consider what Cassiodorus noted or excerpted in his De orthographia On Orthography. 1) Forms such as illuminatio illumination, irrisio mockery, 3) immutabilis unchangeable, impius wicked, improbus dishonest are recommended by the assimilated preposition before m and n, which is observed in the Lyons and Verona codices of Augustine's City of God (CSEL. XXXIX, p. VIIII). However, since Boethius prefers to use different forms, 4) and in certain places such as 28, 10 inlaesum inviolatumque unharmed and inviolate, 42, 7 inlatas iniurias inflicted injuries, cf. 92, 28 he preferred unassimilated forms, 5) I have retained the primary forms of the prepositions ad and in here and there with the codices, though it cannot be entirely denied that the uncial archetype, just like the Puteanus codex of Livy, 6) exhibited assimilated forms everywhere. Likewise, Cassiodorus prefers exsilium exile, exsors devoid of (P¹TV¹), exspectare to wait for (L¹ more often, T twice), exstirpare to root out (P²TL¹V), and similar ones.

1) See the Index of words s.v. orthographia orthography (p. 224 sq.). I have not concerned myself with corrupted forms in the codices such as conditio condition, herere to stick, sepe often, sepire to enclose, suspitio suspicion.
2) Gramm. Lat. VII 163, 3; 200, 18; 201, 11; 202, 11. 18. 25; 203, 14; 211, 2 (Instit. c. 15 LXX 1129 Migne).
3) 7, 1 in PTV it is corrected to irridemus we mock. It is difficult to judge whether errors born during the act of writing were removed by the corrections of the single codex T or if new readings were intruded.
4) Cf. what Mommsen set forth in Mon. Germ. Auct. ant. XII, p. CXVI.
5) Rand, On the hist. of the De Vita Caesarum of Suetonius (Harv. Stud. of Class. Phil. XXXVII 1926) 35, 5, believes that editors, by overestimating the codices of the 9th century, admit too many unassimilated forms. Linderbauer, in his edition of the Rule of St. Benedict (Florileg. Patrist. XVII. Bonn, 1928), p. 7 sq., argues that assimilation was always used in pronunciation but generally omitted in writing.
5) Cf. Shipley, Certain sources of corruption in Latin Mss. Am. Journ. of Arch. 2. Ser. VII (1903), 413 sqq.