¹ A further list will be found in Delbrueck, p. 3.

² Il Campo Marzio dell' Antica Roma, Rome, 1762, pp. 29–30, pl. XLGoogle Scholar.

³ See also Ashby-Fell, p. 137. The Peutinger Table puts ad Pontem Iulii iii, which is probably a slip, but may record some rebuilding.

⁴ Frank, op. cit., p. 140; Blake, p. 44.

⁵ Mattingly, H., Coins of the Roman Empire in the British Museum, vol. I, London, 1923, p. 75Google Scholar, nos. 432–6.

⁶ Tomassetti, loc. cit.; Delbrueck, I, p. 4; Martinori, loc. cit.

⁷ For the probable extent of this rebuilding see Piranesi, op. cit., pl. XXXIX.

⁸ Delbrueck gives the variation between the highest and the lowest as 1·92 m. Some of this may be due to the subsidence of pier III.

⁹ Delbrueck, II, pl. II, places an arch at either end of the bridge. Parallels can be found for both arrangements, and perhaps that with two arches is the commoner.

¹⁰ E.g. Pons Fabricius, Pons Aemilius, Ponte Nomentano, Ponte Salario; see Frank, op. cit., pp. 140–2; Blake, pp. 44–5. There seems to be no reason for believing that the outer facing of tile Pons Fabricius was of travertine, as stated in Platner–Ashby, p. 400.

¹¹ Piers I, II, III and IV have possible remains of original cut-waters, but these are only visible at low water and are much obscured by weed. That of pier I seems unusually elongated and has been suppressed by a wall of travertine blocks, perhaps belonging to a later embankment.

¹² Delbrueck, I, p. 9. It seems to be satisfactorily bonded into the rest of the pier.

¹³ At least in the case of pier III. Delbrueck, I, p. 9, figs. 8a, 8b.

¹⁴ There is some reason for believing with Delbrueck (I, pl. II) that the bridge-head was pierced by a small arched flood-passage, similar to those used in this position in the Pons Fabricius.

¹⁵ Frank, op. cit., p. 141.

¹⁶ This is certain in the case of die travertine, but not in that of the tufa, which is badly weathered. The lowest course of the vault at the south end of arch 2 is unusual in having a heavily bossed surface.

¹⁷ Delbrueck, I, p. 7.

¹⁸ Three courses in the vault.

¹⁹ The fifth course of the later vault is entirely of travertine, the sixth partially so.

²⁰ The Pons Cestius (Gratian) of 369; and the bridge of Valentinian, built in 365–6 on the site now occupied by Ponte Sisto.

²¹ The defence of Rome by the Vitellians in A.D. 69 is the most obvious occasion; but from Tacitus, ' account (Hist. III, 82Google Scholar) it appears that the bridge was in use during the fighting, and is therefore not likely to have been seriously damaged.

²² The slots that held the centering for the brick arches erected at this time are above, not below, the latest stone work. See also Piranesi, op. cit., pl. XXXIX.

²³ See Perkins, J. B. Ward, JRS, XXXVIII, 1948, p. 66, fig. 10Google Scholar.

²⁴ Martinori's reference to its fine state of preservation is certainly no longer applicable.

²⁵ This is presumed from the fact that the clamp-slots that remain intact are often so placed that metal clamps could not have been extracted without damage to the stone. In Rome iron clamps were never used for joining tufa to tufa (except peperino), though they are usual from the Augustan period onwards in joints between travertine and tufa. A good example of this is to be found in the Temple of the Castores, where wooden clamps occur freely in the tufa of the podium-wall, and iron was used only to join the blocks of the wall to the travertine piers that take the weight of the columns. The same system is to be found in the Colosseum.

²⁶ Florescu, G., Ephemeris Dacoromana, III, 1925, pp. 22–7Google Scholar, figs. 11–4; Blake, pp. 105, 212, pls. 22, fig. I; 52, fig. 1.

²⁷ Martinori, E., Via Salaria, Rome, 1931, pp. 81–2Google Scholar. The slope is somewhat steeper than would appear from Martinori's two illustrations.

²⁸ Ashby and Fell report signs of the foundations of the far bridge-head, but these are no longer visible.

²⁹ Ashby-Fell, p. 161, n. 3. The suggestion that these grooves were intended for sluices seems unlikely on account of their large number; they occur only at the joints of the blocks and are probably only an unusually deep form of drafting.

³⁰ The best example in buildings other than bridges is probably the gate known as the Basto del Diavolo at Ariccia, generally considered to be Augustan (Blake, p. 203, pl. 22, fig. 2). For an arch at Cortona, see Blake, p. 187; Modona, A. Neppi, Cortona Etrusca e Romana, Florence, 1925, pl. IVGoogle Scholar, b (where there is no sign of the alternation of headers and stretchers in the vault that is suggested by fig. 5 of the same work). The arches in the ramps leading to the upper sanctuary at Palestrina have outer voussoirs of peperino laid in this manner, in conjunction with concrete vaults.

³¹ Delbrueck (II, p. 70) dates it tentatively to 142 B.C., the date of the completion of the Pons Aemilius, with which it was connected. This seems to be borne out by the level of the buildings around it. See also Blake, p. 172, n. 126, for bibliography.

³² Delbrueck (loc. cit.) puts it at the end of the second century B.C.; Riis (p. 84) and Blake (p. 216) suggest an Augustan date. See also Persichetti, N., Röm. Mitt., XVIII, 1903, p. 299Google Scholar, fig. 5, for description, measurements, and a good photograph. The main arch was largely destroyed during the late war.

³³ Blake, p. 212, pl. 21, fig. 2.

³⁴ Photo Gardner. I have not seen this bridge and do not know whether it still exists.

³⁵ Danielli, L. Rossi, Bollettino Storico-archeologico Viterbese, Viterbo, Feb. 15, 1908, p. 9Google Scholar; Blake, pl. 212, pl. 13, fig. 4. It appears to be an aqueduct.

³⁶ See above, no. 7, n. 26. The small arch at the extreme west end is of normal construction, perhaps only on account of its small size, though it may also be earlier. For other examples, see Blake, pp. 211–2: two of these are incorrect; in the arch under Ponte di Nona the ‘divided voussoirs’ are due to natural cracks in the stone; and Ponte Amato has suffered from a comparatively modern rebuilding, and there is no evidence for divided voussoirs in the ancient parts of the arch. Ponte Scutonico and Ponte San Giorgio I have not seen.

³⁷ Personal observation, not entirely conclusive. See Nibby, II, pp. 578–9.

³⁸ Concord, the Castores, Mars Ultor, etc.

³⁹ An early example of the use of wooden clamps is to be seen in the Aqua Marcia, inside Tor Fiscale. This section is believed to belong to the original building of the Marcia in 144 B.C.; Ashby, T., The Aqueducts of Ancient Rome, ed. Richmond, I. A., Oxford, 1935, p. 136Google Scholar; Van Deman, E. B., The Building of the Roman Aqueducts, Washington, 1934, p. IIIGoogle Scholar.

⁴⁰ See below, no. 40, n. 118.

⁴¹ Blake, p. 199, giving bibliography. Säflund, G. (“Le Mura di Roma Repubblicana,” Acta Instituti Romani Regni Sueciae, I, 1938, p. 234Google Scholar, and Eranos, XXXVIII, 1930, p. 197Google Scholar) proposes a date at least a century later; but the traditional date is generally accepted.

⁴² Nibby, II, pp. 593–6.

⁴³ Promis, C., Le Antichità di Aosta, Turin, 1862, pp. 192–7Google Scholar, pl. XIV; CIL V, 6899; Barocelli, P., Ricerche e Studi sui Monumenti Romani della Val d' Aosta, Ivrea, 1934Google Scholar; PP. 59–62, figs. 41–2, and Forma Italiae, XI, i, Rome, 1948Google Scholar, zone II, no. 20, fig. III.

⁴⁴ Blake, pp. 215–6, pl. 23, fig. 2; CIL XI, 367.

⁴⁵ Km. 16·7 of the modern road.

⁴⁶ ad Att. I, 1; see RE XV, 1965Google Scholar, s.v. Minucius, 60.

⁴⁷ Eroli (p. 24) does not state in which part of the bridge the clamps occur. It is difficult to be certain of the existence of mortar, since there is a tendency for unmortared joints to become sealed with a limy deposit from the water that percolates through the structure.

⁴⁸ CIL XI, ii, 4121, a–b. The condition of both, and especially of b, is poor.

⁴⁹ The voussoirs of this course are not tapered at all. Those of the second course have double the normal taper, though their width at the extrados is the same as that of the rest.

⁵⁰ The information given here is based partly on personal observation from a range of 25 m., but mainly on Anderson photographs 363–4, taken before the collapse. See Noack, op. cit., pl. 131 and Blake, pl. 24.

⁵¹ For a diagram, see Choisy, AB, pl. XVI, 2.

⁵² Compare the more ornate but less ambitious Ponte di Augusto at Rimini, which took at least six years to build (CIL XI, 367).

⁵³ Dr. Blake points out that the alternation of courses of headers and stretchers belongs properly to solid stone construction, whereas less regular masonry implies a concrete core. This is undeniable, but may not be applicable in the case of Ponte d'Augusto. Pier II, at least, has almost certainly a concrete core throughout, though both types of facing are used in it. For another violation of this principle, see Durm, fig. 211.

⁵⁴ Restored thus by Choisy (AB, pl. XXI).

⁵⁵ AA. SS. Maii, vol. 1, p. 388 (May 3rd.).

⁵⁶ The original bridge was destroyed by the Romans during the Second Punic War (Zonaras, VIII, 25).

⁵⁷ Among others, the Forum of Augustus, the podia of the temples of the Castores, of Mars Ultor and of Concord, and the Aqua Claudia.

⁵⁸ Examples of opus quadratum in travertine exist in the town walls of Anagni and Paestum (personal observation) and of Lucca (Blake, pp. 107–8). All these are probably rather too early to have much bearing on the case in point.

⁵⁹ See p. 90, n. 41.

⁶⁰ Noack, F., Röm. Mitt., XII, 1897, pp. 174–82Google Scholar, figs, vi-xi, pls. VIII-IX; Richmond, I. A., JRS, XXIII, 1933, pp. 161–3Google Scholar; Riis, pp. 65–98; Blake, pp. 199–201.

⁶¹ Frothingham, p. 192, pl. XXXIII; Richmond, I. A., PBSR, XII, 1932, pp. 56–62Google Scholar, figs. 8–11; Blake, p. 200, pl. 17, fig. 2.

⁶² This has been quoted as a parallel (Blake, p. 201). It is difficult to be certain whether or not the moulding is separate from the voussoirs. The single arch and both piers are original. The piers are decorated with pilasters 0·60 m. wide, which return under the arch and support a moulded cornice similar to that of Ponte d'Augusto. The arch-moulding is of the same form as the cornice. The facing is of smooth-dressed limestone blocks of irregular length. See also Blake, p. 216, pl. 23, fig. 3.

⁶³ It is possible that the bridge was damaged by the Vitellians in their half-hearted attempt to hold Narnia against the advancing Flavian army in A.D. 69 (Tac, . Hist., III, 58 ff.Google Scholar). It is perhaps hardly likely that such damage would have been sufficient to warrant the rebuilding of a whole pier and much of the upper walls.

⁶⁴ See below, under Ponte Centesimo, no. 19.

⁶⁵ For bridges at Sommières and Vermonton, see Choisy, HA, p. 584. For Pont d'Ambroix, see below, p. 99, n. 73. For the Pont du Gard, Choisy, AB, p. 128, fig. 79; Durm, figs. 271, 335; Espérandieu, E., Le Pont du Gard, Paris, 1926, pp. 42, 61Google Scholar. A closer parallel to Ponte d'Augusto is probably El Kantara in Algeria, between Biskra and Constantine, which has three parallel but not juxtaposed rings, connected by a solid stone extrados, but without an overall bonding system: diagram, Choisy, HA, p. 571; description and photograph, Gsell, S., Monuments Antiques de l'Algérie, Vol. II, Paris, 1901, p. 7Google Scholar, pl. LXXIII.

⁶⁶ Promis, C., Le Antichità di Aosta, Turin, 1862, pp. 92–5Google Scholar, Pl. II; Baroccelli, P., Ricerche e Studi sui Monumenti Romani della Val d'Aosta, Ivrea, 1934, pp. 38–9Google Scholar, figs, 12–16. This is an interesting bridge, the lower parts of which are probably of early Imperial date. The upper parts of the arch are unusual in that the voussoirs are joined by external iron clamps.

⁶⁷ Jacopi, G., Bollettino d'Arte, Anno XXX, Ser. III, 1936, pp. 166–9Google Scholar. The span is 34·20 m. For a rather similar bridge over the Gök Su, with a span of over 30 m., see Dörner, F. K. and Naumann, R., Forschungen in Kommagene (Istambuler Forschungen, vol. 10), Berlin, 1939, pp. 61–6Google Scholar, pls. 7, 21, 22.

⁶⁸ See below, under Ponte di Traiano, no. 36. There seems to be considerable disagreement as to the exact measurements. It was almost entirely rebuilt in 1859–60, but without serious changes. An interesting feature is the ring of small stones round the extrados, which suggests faintly the ring of slabs used in Ponte d'Augusto.

⁶⁹ The use in Rome of roughly dressed travertine with well defined marginal bevel or draft seems to begin with the podium of the Temple of Saturn, probably built by L. Munatius Plancus in 42 B.C. (Platner-Ashby, p. 464; Blake, p. 156). Previously travertine was normally dressed to a smooth face whenever it was intended to be seen. Wide drafting of the intrados of arches appears frequently in conjunction with rough-dressed masonry. Early examples in Rome include the door-arch of the building behind the Temple of Romulus, now generally attributed to the Augustan Temple of the Penates (Platner-Ashby, p. 389; Lugli, G., Roma Antica, il Centro Monumentale, Rome, 1946, p. 226Google Scholar), and the back wall of the Forum of Augustus, where it is used in the Gabine tufa arches, but not in the travertine Arco dei Pantani. Under Claudius and the Flavians it becomes quite a feature of Roman travertine construction, being used in the arch of Aqua Virgo in Via del Nazareno (A.D. 46), in Porta Maggiore (A.D. 52) and in the inner arcading of the Colosseum. Bridges in which it occurs include two near Santa Marinella on the Via Aurelia (Blake, p. 210, pl. 20, figs. 2–3); Ponte Manlio (below, no. 33), the bridge over the Valchetta (above, no. 3) and Ponte Toro (below, no. 39) on the Flaminia; and several examples on the Appia.

⁷⁰ One of the earlier examples is in the so-called Tempio di San Manno near Perugia (Durm, fig. 52; Riis, p. 82; Blake, p. 197). Compare the Severan bridge over the Bölau Su (above, no. 12, n. 67).

⁷¹ Blake, p. 146.

⁷² It seems to be generally agreed that the one remaining ancient pier of the Ponte Rotto belongs to an Augustan reconstruction (Delbrueck, I, p. 22; II, pl. II; Platner Ashby, p. 398; Blake, p. 178). This is not to say that the original Pons Aemilius of 142 B.C. did not also have flood arches.

⁷³ In Gaul, examples include Pont d'Ambroix and the bridges at Boisseron and St. Thibéry; see A. Grenier, Manuel d'Archéologie Gallo-Romaine (Dechelette, J., Manuel d'Archéologie, Vol. VI), part II, i, Paris, 1934, pp. 190–4Google Scholar, figs. 61–2. For Spain, see E. Hübner, Bullettino dell' Instituto di Corrispondenza Archeologica, 1862, p. 170 (Bridge at Mérida); Almagro, M., Ampurias, II, 1940 pp. 176–7Google Scholar, pl. II (Bridge at Luco); Taracena, B. in Ars Hispaniae, Vol. II, Madrid, 1947, p. 15Google Scholar, fig. 2 (Luco); p. 17, fig. 5 (Merida); de Almeida, J. A. Ferreira, Fasti Archaeologici, III, 1948, no. 263Google Scholar, figs. 14–15 (Bridge at Vila Formosa).

⁷⁴ See below, no. 40. In Ponte di Pietra at Verona very narrow piers were used in conjunction with high, narrow flood-passages.

⁷⁵ Meomartini, A., I Monumenti e le Opere d'Arte della Città di Benevento, Benevento, 1889, pp. 267–8Google Scholar, pl. XXXVII; Blake, p. 211. It is just possible that the inscription, CIL IX, 2122, of A.D. 198, recording the complete rebuilding of a bridge by Severus and Caracalla, belongs to Ponte Apollosa, though its provenance is not certain. The bridge itself was badly damaged during the war, and has since been rebuilt, not from the original blocks. Part of one of the embankment walls was uncovered on that occasion, and is in a fair state of preservation, as are the piers and some of the walls supporting the river-banks. Montecchini's measurements and Mr. Gardner's photographs, however, are the principal evidence on which a description of it can be based.

⁷⁶ Ponte Lebbroso, Meomartini, op. cit., pp. 274–85, pls. XXXIX–XLI; Ponte Corvo, ibid., p. 270, pl. XXXVIII; Ponte Tufaro, ibid., pp. 264–7, pl. XXXVI; Blake, p. 211. All three were badly damaged during the war: Ponte Tufaro has not been rebuilt.

⁷⁷ The presence, low down in the left-hand wall of the far bridge-head, of some masonry remarkably like that of Ponte Apollosa is perhaps a hint that the irregular masonry belongs to a later rebuilding. This point should not be pressed. The dowels appear in the arches, and it seems that clamps were not used in the walls.

⁷⁸ Blake, pp. 274–5 et al. The tesserae are about 0·08 m. square on the surface and up to 0·20 m. deep.

⁷⁹ Three Augustan mile-stones on the Appia, , CIL IX, 5986, 5988 and 5989Google Scholar, give no particulars of the work done.

⁸⁰ CIL X, 6846. See below, under no. 15. The inscription does not mention any specific repair, but is said to have come from an ancient bridge near Ponte Maggiore. To judge from old photographs, there were quite a number of these in the vicinity of Ponte Maggiore.

⁸¹ CIL X, 6853. The bridge is recorded in one of Mr. Gardner's photographs.

⁸² Lugli, G., Forma Italiae, I, i, part 2, Rome, 1928Google Scholar, Zone Il, no. 36, fig. 12; Blake, p. 211; see also p. 100, n. 80.

⁸³ See also p. 100, n. 76.

⁸⁴ Ponte della Catena, near Cori, which is in parts of almost polygonal construction, is generally considered to be early, but has other unusual features such as a multi annular arch. See Accrocca, A., Cori, Storia e Monumenti., Rome, 1933, pp. 95–6Google Scholar, fig. 22; Blake, p. 211, with bibliography.

⁸⁵ Possibly the upper parts of this bridge and of that over the Titerno below Faicchio (A. Maiuri, Not. Scav., 1929, PP. 211–3, fig. 2; Blake, p. 193) were always of wood. Here there are no signs either of a stone arch or of slots for beams.

⁸⁶ The polygonal masonry of the walls of Cesi (Ashby Fell, pp. 172–3, fig. 12; Blake, p. 103) is of a very different type.

⁸⁷ This is the width of the embankment at the crossing of the Treia (above, no. 7, p. 87) and is that normally found in small culverts.

⁸⁸ Compare the somewhat similar device used in the arches of the Colosseum (Choisy, AB, fig. 78; Durm, fig. 248).

⁸⁹ Ponte Corvo on the Appia south of Benevento has the somewhat similar marks P and IIII.

⁹⁰ Ovinius, L. Rusticus Cornelianus, curator viae flaminiae probably in the early third century, RE, XVIII, 2, 1995Google Scholar.

⁹¹ For other examples, see below, no. 20.

⁹² For parallels, see p. 89.

⁹³ Compare Durm, J., Die Baukunst der Griechen, Leipzig, 1910, figs. 70Google Scholar; 71, iii.

⁹⁴ There was another example of it under the now destroyed bridge over the Fosso dell'Acqua Traversa, on the Appia near Fondi. (Photo. Gardner.)

⁹⁵ For similar culverts on the Via Salaria, see Persichetti, N., Viaggio Archeologico sulla Via Salaria, Rome, 1893, p. 67Google Scholar, and Not. Scav., 1896, pp. 536–7.

⁹⁶ See above, p. 100, n. 76.

⁹⁷ Ashby, T., PBSR, III, 1906, p. 151Google Scholar, fig. 15.

⁹⁸ See above, p. 96, n. 66. Only the lower parts of the arch are involved.

⁹⁹ See above, p. 99, n. 69.

¹⁰⁰ Blanchet, A., Carte Archéologique de la Gaule Romaine, Fasc. VIIGoogle Scholar (Département de Vaucluse), Paris, 1939, P. 50, pl. IGoogle Scholar.

¹⁰¹ The Forum of Augustus was dedicated in 2 B.C.; the arch of the Aqua Marcia bears an inscription of 5 B.C. (CIL VI, 1244).

¹⁰² Km. 219·6 of the modern road; see Montecchini, pp. 19–10; Ashby-Fell, p. 182.

¹⁰³ José de, C. Serra Ráfols, in Carta Arqueológica de España, Barcelona, Madrid, 1945, pp. 127–8Google Scholar, fig. 19; Taracena, B., in Ars Hispaniae, vol. II, Madrid, 1947Google Scholar, fig. 1. Only the bridge-heads are ancient.

¹⁰⁴ Montecchini loc. cit. For a conjectural dating, see below, pp. 110, 117.

¹⁰⁵ The style of masonry is similar to that of Ponte Etrusco (no. 24) and of the third phase of Ponte Manlio (no. 33).

¹⁰⁶ Ponte Spiano has been omitted from this list on account of the radical difference in the construction of its arch. The pietra corgnola abutments of the first phase of Ponte dei Ciclopi have disappeared, but enough remains of the pietra grigna masonry of the second phase to show that it was of the same type as that of the third period of Ponte Manlio.

¹⁰⁷ See above, p. 106, n. 102.

¹⁰⁸ For more notable examples of buttresses in the early second century A.D., see below, no. 36, p. 112.

¹⁰⁹ Ponte del Diavolo on the Via Latina (above, p. 89, n. 34) also has projecting key-stones, but they are of very tall, narrow form, unlike those of Ponte Cardaro and Ponte Manlio.

¹¹⁰ For parallels, see above, p. 98, n. 69. The absence of this drafting in the smaller arch may indicate that it was executed after the latter had been buried, either by gravel from the river-bed or by the successive lengthening of the substruction walls.

¹¹¹ Ponte Spiano seems in any case to have been rather an unnecessary luxury, in view of the size of the stream, which is now contained in a two-foot drainpipe running under the road. As such, it is likely to have been built at a time when the rest of the road was already in a good state of repair.

¹¹² Anderson–Spiers–Ashby, pl. LXVI.

¹¹³ E. Hübner, Annali dell'Instituto di Corrispondenza Archeologica, 1863, pp. 173–94, with Monumenti dell' Instituto, VI–VII, 1857–1863, pls. LXXIII–LXXVGoogle Scholar; Merckel, C., Die Ingenieurtechnik in Alterthum, Berlin, 1899, p. 299Google Scholar, figs. 102–3; Taracena, B., in Ars Hispaniae, vol. II, Madrid, 1947, pp. 17Google Scholar, 20, fig. 6. Two other bridges in Spain, those at Mérida and Salamanca, also have but tresses of this type, but are not certainly dated; Taracena, op. cit., p. 17, figs. 4–5: Merckel, op. cit., p. 301, considered both to be Trajanic.

¹¹⁴ Measurements taken from Sansi, op. cit., pp. 217–8.

¹¹⁵ Frothingham and Pietrangeli; Sansi, however, states that occasional blocks run right through the pier. The space, unfortunately, has had to be filled in, in the interest of stability.

¹¹⁶ Livy, XLI, 27. ‘Censores vias sternendas silice in urbe, glarea extra urbem substruendas marginandasque primi omnium locaverunt, pontesque multis locis faciendos’.

¹¹⁷ Pietrangeli rejects the inscription and proposes an Augustan date for Ponte Sanguinario on the basis of its resemblance to Ponte Calamone and Ponte Cardaro.

¹¹⁸ Richmond, I. A. and Holford, W. G., PBSR, XIII, 1935, pp. 69–76Google Scholar. Ponte di Pietra does not conform to the Augustan plan of Verona, and one end of it was blocked by the circus. It is, therefore, likely to be earlier than either. Marconi, P. (Verona Romana, Bergamo, 1937, pp. 26–33Google Scholar, figs. 14–19), gives a description and suggests a pre-Roman date. This suggestion is rendered improbable by the core of the bridge, which was partially of concrete (visible in Works of Art in Italy, Losses and Survivals in the War, Pt. II, H.M.S.O., 1946, p. 61Google Scholar).

¹¹⁹ Polcastro, Giandomenico, Notizia della Scoperta in Padova d'un Ponte Antico, con una Romana Iscrizione, Padua, 1773Google Scholar. The inscription (CIL V, 2845) records that a board of adlegatei pontem faciendum D.D.S. locarunt, eidemque probarunt, and resembles that of the Pons Fabricius (62 B.C.) in being cut on one of the arches and not, as later became the rule, on a separate slab.

¹²⁰ Scarpellini, A. in Emilia Romana, vol. I, Florence (Istituto di Studi Romani), 1941, p. 195Google Scholar, fig. 7. The bridge was destroyed during the war.

¹²¹ Durm, figs. 251, 335; Blake, pp. 212–3, pl. 26, fig. 4.

¹²² Durm, fig. 757; Blake, p. 223. It can hardly be earlier than the foundation of die Augustan colony.

¹²³ Durm, fig. 335.

¹²⁴ Durm, fig. 251; Scriattoli, A., Viterbo nei suoi Monumenti, Rome, 1915–1920, figs. 646–7Google Scholar. Dr. Blake (pp. 221, 224, n. 2, 240) regards it as belonging to the middle of the first century B.C.

¹²⁵ CIL XI, 6619 records the rebuilding of a substruction wall near Massa Martana in A.D. 124.

¹²⁶ For others, see p. 110.