¹ Another important discussion of the MSS. is in H. Grisar, Analecta Romana, 1889.

² de Rossi, II, 174.

³ ibid. 193.

⁴ de Rossi, II, 304, 316. Silvagni ascribes it to the latter date, Nova Series I, xxx.

⁵ Sabinus, de Rossi, II, 407; Alciatus, C.I.L. VI, p. xviii, IX, p. xxvi; Panvinius, C.I.L. VI, p. liii, IX–X, p.

Doni, C.I.L. VI, p. lviii, IX–X, p. xxxviii.

⁶ Earlier schemes and more detailed announcements of the plans for this corpus are to be found in articles by Gatti, G., Arch. R. Soc. Rom. di Stor. Pat. XXXI, 1908Google Scholar; before 1300. A. Beccaria, Arch. Stor. Ital. 1909, with a bibliography of numerous previous articles; Silvagni, A., Raccolta delle iscrizioni medioevali italiane di civiltà cristiana anteriori al sec. XIIIGoogle Scholar. Resumés des communications presentées au Congrès de Varsovie, 1933.

⁷ Since this essay was originally written the first parts of Professor Silvagni's Monumenta Epigraphica Christiana have appeared, consisting of plates illustrating the material before 1300.

⁸ V. Forcella, Iscrizioni di Milano dal sec. VIII ai giorni nostri 1889–93; C. Gazzera, Iscrizioni cristiane antiche del Piemonte, 1851; de Villard, U. Monneret, Iscrizioni cristiane della provincia di Como anteriori al secolo XIGoogle Scholar; L. Billo, ‘Le iscrizioni Veronesi dell' alto Medioevo,’ Arch. Ven. 1934.

⁹ The ‘complete’ texts referred to are texts of what now survives. Where more is preserved from early transcriptions I have given references which also include the text of these records.

¹⁰ Rome, Milan, Pavia, Lucca, Como, Naples and Benevento.

¹¹ Cf. the lives of these popes in the Liber Pontificalis.

¹² Toesca, I, 216.

¹³ v. no. 15 for a possible exception.

¹⁴ This form occurs in half-uncial script.

^14a W. de Grüneison, S. Marie-Antique, 1911, p. 416, gives alphabets of contemporary painted inscriptions.

^{page 52 note 1} Galletti, Del Primicero, 41. His epitaph (now lost) is dated mense deceb indiction VII temp, dn stefani pp. It would seem therefore that Theodotus could not have been primicerius while Zacharias was still living. G. B. Tatum (Art Bulletin 1944, 28) produces numerous examples of people represented with square nimbi who could not have been living at the time of the portrait and suggests that in the IX century it was used to distinguish people who had risen above the common level without attaining sainthood. In this case the representation would imply that Zacharias, though perhaps already dead at the time of their execution, had been interested in Theodotus' paintings.

¹⁵ For the inscription of Leo III of Cimitile see no. 126.

¹⁶ Cf. the inscriptions on the tombs at Jouarre (Le Blant, I, no. 199) of 680 or on the altar at Ham of the late VII century (La Blant, I, no. 91). On the other hand the epitaph of Theophilact of 671 at Naples has the same motive and is not unlike (Silvagni vol. IV. fasc. I). The most similar script letters are also Merovingian; cf. the capitals used in Luxeuil script. Zimmermann, , Vorkarolingische Miniaturen, vol. I, pl. 48Google Scholar.

¹⁷ Another example is reproduced by Silvagni, II, fasc. 1, pl. xi, 2.

¹⁸ An even later example of the same border pattern still being used is no. 123 (834).

¹⁹ It is noteworthy that diagonal arms to letters such as F, L, which are found combined with interest in serifs in some Merovingian inscriptions, are not used.

^19a Haseloff, pl. 44.

²⁰ v. conclusions p. 56. By abstract I mean sculpture of the same aesthetic nature as ‘popular’ inscriptions, not the panels of interlacing and other decorative motives which abound.

²¹ It may be classed as an undeveloped example of the style of carving associated with nos. 32, 33, 36 and 43.

²² v. M. Deloche, Lex Anneaux Sigillaires, 1900, e.g., nos. CXXVII, CLXXXVI, CCLXVII, etc.

²³ No. 30: Ursus, Juvintinus, Juvianus, Domnicus, Vidalianus with Tancol, Refol, Vergondus, Teodoalfo, Gondelme. No. 31: Ursus with Hildericus Dagileopa. No. 52: Bonus, Julianus, Martinus, Lupicus with Romald.

²⁴ Gregorius (2), Vincentius, Martinus and Agnellus, Venerius, Barbatus, Antonius (3), and Adrianus, Anastasius, Stephanus.

²⁵ Förstemann, Altdeutsches Namenbuch, suggests that German vulfo and beta root forms were sometimes translated lupus and ursus. In which case a classical name ceases to he any indication of race, but not of course vice versa; v. also no. 26, where we have a Lombard duke of royal blood called Gregory.

²⁶ Ursus, Juvintinus, Juvianus, Julianus, Martinus.

²⁷ Both in Le Blant and in de Rossi there are reproductions of inscriptions dated to the VI or VII centuries which in some letter-forms and in a general irregularity foreshadow our style, inscriptions often described as litteris rudis, but, though it is difficult to be sure without a representative collection of good reproductions, the style does not seem to have been fully formed, nor the irregularity become transformed into design, till the VIII century.

²⁸ But cf. no. 26; there seems little doubt that this dux Gregorius was Lombard.

²⁹ P. Deschamps, ‘Paléographie des inscriptions de la fin de l'epoque mérovingienne au XII siécle,’ Bull. Mon., 1919, Pl. 2, fig. 2.

³⁰ The lettering on the Wolvinus altar in S. Ambrogio is in the same style, v. alphabet p. 87, though certain letter-forms A, L, M, R have serif terminations more resembling MS letters. The small section by a different hand noted by Tatum, G. B., The Paliotto of Sant' Ambrogio, Art Bulletin, 1944, p. 27Google Scholar, fig. 5, is not, I think, necessarily later, considering the variety of IX century paleography.

³¹ Also v. Silvagni, II, fasc. I, pl. v 2.

³² In these the relation to contemporary Frankish inscriptions is shown in contrast to the strict classicism of the late IX century Roman inscriptions.

³³ Another dated example is reproduced in Silvagni, II, fasc. II, pl. v. 2 (893).

³⁴ Perhaps also Silvagni, I, pl. XXXVI, 3, and II, fasc II, pl. V, 1, commemorating the bringing of relics from Rome to Legguino; was the inscription also brought from Rome? It is dated 846.

^34a cf. Silvagni, I; Diehl, pl. 36 b.

³⁵ M.G.H. Scriptores, I, p. 302Google Scholar, and p. 36, ad an. 795.

³⁶ Cf. Silvagni, I.

³⁷ Very similar lettering is, however, shown on the arch from Knin, reproduced by Strzygowski, Altslavische Kunst, p. 93 (fig. 66), apparently dated 892–910.

³⁸ Strzygowski, op. cit., fig. 65, well of Duke Viceslav, c. 800.

³⁹ Cf. also nos. 61–65 and 69.

⁴⁰ Reproduced in Haseloff, pl. 77.

^40a Cf. Haseloff, pl. 47.

⁴¹ The epigraphical evidence therefore agrees with Giovenale's view, based largely on the history of S. Maria in Cosmedin, that Joannes' work was done in the reign of Nicholas I (858–867).

^41a The epitaph of David, bishop of Benevento, of 796, reproduced by Silvagni, IV, fasc. 11, pl. 1, is, however, in beautiful classical lettering. The development of the native style may therefore have been later.

⁴² In 1938, when this paper was written. This statement applies also to other inscriptions noted below as unpublished.

^42a I am indebted to Miss E. M. Jamison for kindly searching for references to Audoalt.

⁴³ So probably does the lettering on the Roman sarcophagus which was used as a tomb for Otto II, in the Vatican grottos; but the text is too short and the surface on which it is cut is so rough that it is difficult to be sure.

^43a Neues Archiv. ix, 514.

^43b v. no. 148, which is dated February of the same indiction but falls within the succeeding pontificate.

⁴⁴ It is, however, similar to some contemporary Frankish inscriptions, cf. Deschamps, Bull. Mon., 1929, pl. VII.

⁴⁵ Hodgkin, , Italy and her invaders, VI, 238Google Scholar.

⁴⁶ But unlike the Anglo-Saxons they produced no vernacular literature.

⁴⁷ Among which the inscribed verses at Pavia and Benevento are prominent, v. M.G.H., Poetae Latini, I and IV, pt. II.

⁴⁸ Of the MSS. probably produced in the Lombard kingdom which survive none contain illustrations beyond ornamental initials, arcades, etc. There are no full-page illuminations of figure subjects.

⁴⁹ If the stucco figures are then dated after the formation of the style we have an explanation of the difference between them and the altar.

^49a Toesca, I, 281, fig. 174.

^49b E. Schaffran, Die Kunst der Langobarden, 1941, pl. 24 c.

^49c Haseloff, pl. 51.

^49d Toesca, I, 278, fig. 169.

⁵⁰ The only signature is that of Johannes magister at Bobbio—but that name tells us nothing certain of his nationality.

⁵¹ This does not exclude the possibility that these carvers are the magistri Comacini mentioned in the laws of Rotbari and Liutprand.

⁵² Cattaneo, Ch. II; Rivoira I, 104.

⁵³ Cattaneo, 53.

⁵⁴ Haseloffi pl. 43.

⁵⁵ v. Wroth, Catalogue of Coins of the Vandals, Ostrogoths and Lombards in the British Museum, 1911.

⁵⁶ N. Åberg, Die Goten und Langobarden in Italien, 1923.

⁵⁷ It is remarkable that the coins of Stephen III and Hadrian I are barbaric in lettering and design. Perhaps their inscriptions would have become so. Those of Leo III are transformed after the Carolingian model.

⁵⁸ In the same style are the silver reliquaries of Paschal I (Haseloff, pl. 57, 58).

⁵⁹ Toesca, I, 404.

⁶⁰ The Dietrich carving in Bingen, however, seems to me from the reproduction in Bauer, Mainzer Epigraphik, 24, to be popular both in epigraphy and sculpture. It is the only known inscription in old high German and is dated by philologists to the X century.

⁶¹ In particular, P. E. Schramm, Kaiser, Rom und Renovatio.

⁶² De Rossi, II, 324.

⁶³ Cf. epitaph of Adelberga, 840, reproduced by Deschamps, Bull. Mon., 1929, pl. II.

⁶⁴ v. p.124, no. 42.

⁶⁵ N. Gray, ‘Dark Age Figure Sculpture in Italy,’ Burlington Magazine, Nov. 1935.

⁶⁶ Bertaux, 251.

⁶⁷ Haseloff, pl. 67; M. T. Tozzi, ‘Di Alcune Sculture medioevali della Compania’, Boll. d'Arte, 1931, 272.

⁶⁸ Toesca, I, 435.