^{page 128 note 1} Livy, iii. 52. 3; C.I.L. xiv. p. 447.

^{page 128 note 2} Livy, ii. 11. 7, iii. 6. 7, v. 49. 6.

^{page 128 note 3} C.I.L. xiv. pp. 435, 447, and No. 4003.

^{page 128 note 4} As far as Bovillae the road must have already existed: for it seems clear that this town was in some respects the successor of Alba Longa. After the destruction of the latter, the sacra gentilicia of the Iulii were celebrated at Bovillae; and the inhabitants called themselves Albani Longani Bovillenses (C.I.L. xiv. p. 231Google Scholar).

^{page 130 note 1} Strabo v. 3. 5, p. 231, Χωρία τὰ κατὰ τὴν παραλίαν ὅσα έλώδη καὶ νοσϵρά, οÎα τὰ των Ἀρδϵα-τϖν καὶ τὰ μϵταξὺ Ἀντίον καὶ Λανουίον μέΧρι Πωμϵντίνου κ.τ.λ Virg. Aen. vii. 412, et nunc magnum tenet Ardea nomen; sed fortuna fuit.

^{page 130 note 2} Cic., Pro Plancio, 8. 21Google Scholar.

^{page 131 note 1} The hilly districts on the right bank of the Tiber were, to judge from the comparative scarcity of remains of buildings, in ancient times, as now, mainly forest lands.

^{page 131 note 2} No praetorian milestones are as a fact known.

^{page 131 note 3} Hülsen, Notizie degli Scavi, 1896, 87 sqq.

^{page 131 note 4} The inscriptions relating to these curatores viarum have been brought together by Cantarelli in Bull. Comm. Arch. 1891, 81 sqq.

^{page 133 note 1} The arrangement did not, it is true, remain long in force: for in 63 B. C. the ager Praenestinus was once more in the hands of large proprietors. Cic., De Leg. Agr. ii. 28, 78Google Scholar: Nam si dicent per legem id non licere, ne per Corneliam quidem licet: at videmus, ut longinqua mittatnus, agrum Praenestinum a pancis possideri. Perhaps this was partly owing to the extravagance of the new coloni. Cf. In Catilin., ii. 9, 20.

^{page 134 note 1} C.I.L. xiv. 169: P. Martio Quir (inia tribu) Philippo curatori viae Praenestinae, aedilicio curuli, v(iro) q(uaestorio) ab aerario, tribuno fabrum navalium Portens(ium), corpus fabrum navalium Ostiens(ium) quibus ex s.c. coire licet, patrono Optimo, s(ua) p(ecunia) p(osuit). The inscription is dated 11th April, 195 A.D.

^{page 135 note 1} Cf. Horace, , Sat. i. 7, 28Google Scholar: tum Praenestinus …. durus vindemiator et invictus.

^{page 135 note 2} See De La Blanchère, Un Chapitre de l'Histoire Pontine.

^{page 138 note 1} I shall not enter here into the somewhat difficult question of the topographical history of the Viae Tiburtina and Collatina within the walls of Aurelian. It is probable that both originally started from the Porta Viminalis, the distances being therefore reckoned from that gate, and not from the Porta Esquilina. Hülsen, however (Forma Urbis Romae, tab. i.), considers that in Republican times these two roads started from the Porta Esquilina, and names the strip of road between this gate and their bifurcation, a little way beyond the later Porta Tiburtina, “Via Gabina (Tiburtina Vetus),” by which he probably means that in early times the first few miles of the Via Collatina served as the first portion of the road to Gabii and Praeneste. Kiepert similarly marks a road (in his wall map of Latium) running direct from the Porta Esquilina to the second milestone of the Via Praenestina. The object in both cases is to explain the passage of Strabo quoted below (p. 150, note 1). Of Kiepert's road, however, there are no traces at all (Lanciani, Forma Urbis, 24, 25): nor do I know of any road connecting the Via Collatina with the Via Praenestina within the first mile or two of the city.

^{page 140 note 1} It seems probable that long before the construction of the railway it was not easy to trace, for, though Ameti (1693) and Fabretti (De Aquis et Aquaeductibus, Diss. i. tab. i. 1st ed. 1680) mark it perfectly correctly, subsequent writers do not. The sudden turn at right angles of the Aqua Virgo, so as to run parallel to the road, is strong evidence that it took this line.

^{page 141 note 1} For the use of this and some other photographs (Figs. 4, 7, 12) I am indebted to the kindness of Miss Dora Bulwer.

^{page 144 note 1} C.I.L. vi. 20609 was seen there in the seventeenth century, but existed in Rome itself, near S. Angelo in Pescheria, in the sixteenth.

^{page 144 note 2} The references are to the second edition (the one generally met with) published in 1788. The first edition (1680) does not contain the second of the two maps cited.

^{page 144 note 3} Nor is there, so far as I know, any trace of the prolongation of this road to Ponte Lucano or to Le Cappannelle which Fabretti and Ameti (in his map published in 1693) both show. The theory is probably due to confusion with the road described below on p. 177.

^{page 147 note 1} A little E. of the place where the road crosses the railway, a mediaeval cemetery belonging to some domus culta near Lunghezza was found in making the railway in 1886 (Not. Scav. 1886, 55).

^{page 149 note 1} Dessau, (C.I.L. xiv. p. 279Google Scholar) states that the Antonine Itinerary gives the distance from Rome to Gabii as fifteen miles, which would of course be erroneous; but he is apparently misquoting the Itinerary, which, according to Parthey and Pinder's edition (1848), p. 143, gives the distance correctly.

^{page 149 note 2} The slip in C.I.L. xiv. p. 457: Viae Praenestinae nulli extant cippi milliarii is an unfortunate one. One (C.I.L. x. 8306) is fragmentary, of the time of Maxentius, noted as situated “nel pozzo in una masseria fuori porta Maggiore.” Another (C.I.L. x. 6886) also belongs to the time of Maxentius, and bears the number seven (upon the reverse is C.I.L. vi. 1342), but was found in a garden within the Aurelian walls, certainly, therefore, not in its original position. The last is described below (p. 198).

^{page 150 note 1} Strabo (v. 3, 9, p. 237), though he is writing of the state of affairs after the construction of the tomb of Eurysaces, speaks as if the bifurcation took place at the Porta Esquilina: ἡ Λαβικανὴ ἀρΧομένη μὲν ἀπὸ τῆς Ἠσκυλίνης πύλης, ἀΦ᾽ ᾗς καὶ ἡ Πραινϵστίνη, ἐν ἀριστϵρᾶ δ᾿ ἀΦϵῖσα καὶ ταύτην καὶ τὸ πϵδίον τὸ Ἠσκυλῖνον, πρόϵισι κ.τ.λ.

But it is impossible to suppose that the two roads separated, and then reunited after a mile at the Porta Maggiore, to separate once more just outside it. The passage of Strabo, therefore, remains inexplicable. Cf. Jordan, , Topographie, i. 1, 358, 362Google Scholar, and supra, p. 139n.

^{page 150 note 2} See Lanciani, Commentari di Frontino, 36.

^{page 152 note 1} In Bull. Com. 1891, 321, the existence is recorded of two sepulchral cippi, which very probably belonged to this road (and if so, go far to prove its antiquity) in the Vigna Serventi. The inscription of both is identical—one is used as a step in the wine-cellar, the other was found close to the Vicolo del Pigneto.

There are indications of an ancient road having diverged E.N.E. at the Vigna Pulini, following the boundary between the Tenuta dell' Acqua Bollicante and the vineyards, which belong to the Basilica of St. Peter; but where the boundary stops all traces of the road are lost. At the Vigna Rocchi are several fragments of marble columns and capitals, some broken selce, and a sarcophagus ornamented with undulating channels, bearing the following unpublished inscription upon a tabula ansata in the centre.

The sarcophagus is 2.13 m. in length, and 41 cm. in height, while the tablet measures 39 by 31 cm., the letters being 3 cm. in height.

^{pagev 152 note 2} This detail is due to the Diario di Roma, Jan. 2, 1819 (reprinted in Fea, , Varietà at Notizie xiii. p. 144Google Scholar), where the date is given as 1702.

^{page 154 note 1} One niche contained a leaden box (diam. 15 cm.) in which were some eggs.

^{page 159 note 1} The map is not on a sufficiently large scale to render it possible to indicate these remains with absolute accuracy—in fact, a good deal has been omitted from it at this and other crowded points in order to avoid confusion.

^{page 161 note 1} Nibby's map is incorrect, but he knew the truth (Analisi, iii. 627).

^{page 162 note 1} Said to be in the Palazzo Casali.

^{page 165 note 1} See Winnefeld, , Jahrbuch des Instituts, Ergänzungsheft iii. p. 26Google Scholarsqq.

^{page 168 note 1} Cecconi (Storia di Palestrina, p. 17) notes the discovery of a fine sarcophagus near Tor Angela not long before he wrote, about 1820.

^{page 174 note 1} Other examples of altars of this type are mentioned in Bull. Com. 1897, 164.

^{page 174 note 2} It is in my opinion quite certain that the ancient road ascended almost straight from Pontedi Nona, not curving, as the modern road does, round the hill at the E. end of the bridge.

^{page 177 note 1} The reference is to a quantity of MS. notes, containing Nibby's diaries of excursions in the Campagna, which I purchased at the sale of the library of Count Virginio Vespignani in 1900.

^{page 178 note 1} See also the same author's New Tales of Old Rome, p. 192.

^{page 178 note 2} This star is not shown in the engraving, as it did not appear in the photographs from which this was made.

^{page 181 note 1} In January 1902 I was able, after a good deal of rain had fallen, to find the point at which the Via Praenestina crosses the Osa stream itself, a little way further W. This is 67·80 m. to the S. of the bridge of the modern highroad, but there are no indications of the existence of a bridge on the ancient road, and the course of the stream has very likely changed. One or two pavingstones are to be seen in each bank of the stream, and remains of tombs on the S. side of the road. It was apparently running 10° S. of E.

^{page 181 note 2} Cf. Ann. Inst. 1840, 33.

^{page 182 note 1} In these same works (according to Bull. Inst. 1845, 53) the conduit which conveyed the water to the baths was actually discovered.

^{page 182 note 2} Compare, however, p. 185, n. 1.

^{page 182 note 3} Edifizi, v. p. 92: cf. vi. tav. 110 for plan.

^{page 183 note 1} See also Tacitus, , Ann. xv. 43. 4Google Scholar.

^{page 183 note 2} See plan, Fig. 8.

^{page 183 note 3} A white marble from Asia Minor (marmor coralliticum, Plin., H. N. xxxvi. 62Google Scholar).

^{page 185 note 1} It is curious that the boundary line should not follow the Via Praenestina here as elsewhere, and an examination of the fieldwall which marks it shows that it is full of pavingstones. Probably, therefore, a road ran E. from the temple, parallel at first to the highroad, then crossing it S. of the church, and then turning S.S.E. The existence of the section S. of the highroad is certain, for its pavement, 2·90 m. in width, with crepidines 45 cm. in width on each side, can still be followed for some way. We were told that on the further side of Pantano it could be seen E. of Monte Falcone going towards Colonna. It, or that mentioned p. 194, n. 2, may be the road spoken of by Fea (Gabio, 10) as crossing the basin of Pantano.

^{page 185 note 2} Numbered respectively 2 and 3 in Visconti's plan.

^{page 185 note 3} The wife of Domitian.

^{page 188 note 1} According to Dionysius, the text of the treaty, written on a bullock's skin, was preserved to his day in the temple of Semo Sancus on the Quirinal. Cf. Mommsen, i. 280.

^{page 188 note 2} Its mention in the treatise De Coloniis as muro ducta colonia has not, in view of the character of that treatise, any historical value (C.I.L. xiv. p. 278 n. 5).

^{page 194 note 1} This stamp dates from the beginning of the second century A. D.

^{page 194 note 2} Before the ascent begins, a road, the pavement of which is still well preserved for the most part (though in places it is covered by soil, and in others the stones have been removed for fieldwalls), diverges to the S., and runs in the direction of the Aqua Alexandrina, which it should cross a little to the W. of its springs, if indeed it does not turn down to them. I have not yet followed its whole course, but could see no place where it could pass under the aqueduct. To the prolongation of this, or to that mentioned in the footnote to p. 185, Ficoroni (Labico, 30) probably refers—“al quale (i.e. all' acquedotto) al fine dov' è una scaturiggine d'acqua, l'antica selciata passa contigua.” (He is propounding a theory that the ancient road from Osteria del Finocchio to S. Cesareo ran in an absolutely straight line to the N. of the present road.)

^{page 195 note 1} A wall of opus quadratum 90 cm. in width can be traced on the S. side—it was intended to support the earth above the road.

^{page 199 note 1} The variations in the text here given are due to a more recent examination of the inscription The gentile names seem to be unknown hitherto: N. is an abbreviation for the praenomen Numerius (Mommsen, , Röm. Forsch. i. 19Google Scholar).

^{page 200 note 1} In the Tenuta della Pallavicina was found a lead waterpipe, [Claudius Fel]icissimus fecit (C.I.L. xiv. 2777 = xv. 7837 b)

^{page 202 note 1} The sketch was made from the description given us; we did not see the object itself.

^{page 203 note 1} In this neighbourhood (on the Colle di Quadraversa, before reaching Colle Linaro) was found the sepulchral inscription of Sex. Baebianus, Pompeius, scriba quaestorius et aedilicius (C.I.L. xiv. 2839Google Scholar).

^{page 203 note 2} Fabretti (map opp. p. 90) makes it run to the eighteenth milestone of the Via Labicana, while Ameti takes it back to the Villa Strozzi on the hill W. of Zagarolo.

^{page 204 note 1} In this section some very large paving stones are observable. One measured no less than 1·30 × 1·00 m. The pavement, too, is extremely well preserved, and measures 4·40 m. in width. There is a crepido on each side, of blocks of selce, one higher than the rest being placed every 4½ paces (p. 181).

^{page 207 note 1} A plan of this amphitheatre was made by Palladio (Portfolio xv. f. 8′, of his drawings in the library of the Society of Architects): in his time the vineyard in which it stands belonged to Camillo Colonna.

^{page 212 note 1} From the Valle Vigesimo, according to Cecconi (Storia di Palestrina, p. 18, n. 34), an ancient road ascended to the Colle Cappelle and Le Tende. Perhaps it went on along the Colle S. Rocco to Gallicano, though he does not say so.

^{page 212 note 2} The inscriptions on these pipes are republished in C.I.L. xv. 7888, 7881, 7886, 7887.

^{page 213 note 1} There are no traces of pavement in situ in this road, but several loose pavingstones, and at Casale S. Pietro there are a great number; while I was credibly informed that pavement still exists in situ in the road from Torrione Frocina to S. Pietro.

^{page 213 note 2} Here were found C.I.L. xiv. 2851, 2895, 2918, 3326, 3327; also the brickstamps C.I.L. xv. 2344, 2363 and another, which I believe to be unpublished, /EPPOC There is here, besides the remains of the church of S. Pietro, a very large and well preserved water reservoir, also traces of a villa decorated with very fine marbles.

^{page 213 note 3} Nibby and Gell in their maps make this road keep along the Colle d' Oddo and the Colle Tondo, crossing the stream W. of the latter.

^{page 214 note 1} C.I.L. xiv. 2960 was found here (cf. Petrini, Memorie Prenestine, p. 78).

^{page 216 note 1} According to Strabo (loc. cit.) about 26 miles—210 stadia—by which road we are not told.

^{page 216 note 2} The distance is really only 18 or 19 miles (p. 218 n.).

^{page 217 note 1} It seems to have been usual in the first century A.D. (see C.I.L. x. p. 684, Strabo v. 3. 6. p. 233, Horace, , Sat. i. 5Google Scholar), though the road was apparently in existence from what Strabo says: πλησίον τῆς Ταρρακίνης βαδίζοντι ἐπὶ τῆς 'Pώμης παραβέβληται τῆ Ἀππίᾳ διῶρυξ …. πλϵῖται δὲ μάλιστα μὲν νύκτωρ ὥστ' 'Pἐμβάντας ἀΦ'ἑσπέρας ἐκβαίνϵιν πρωίας καὶ βαδίζϵιν τὸ λοιπὸν πῆ ὁδψ, ἀλλὰ καὶ μϵθ' ἡμέραν. ῥυμουλκϵῖ δ' ἡμιόνιον. Why this troublesome mode of progression was adopted does not seem clear, unless because the road was often impracticable, as De La Blanchère thinks (Terracine, in the Bibliothèque des Écoles Françaises, p. 80 sqq.).

^{page 218 note 1} This distance is calculated either along the modern road from Palestrina to Valmontone (which follows the line of an ancient road) and thence along the present Via Casilina to Piombinara (where the Via Labicana probably rejoins the modern road) and on to the Osteria della Fontana, or along the Via della Selva (an ancient road according to Fabretti, De Aquis, map, opp. p. 90, Chaupy, , Maison de Campagne d'Horace, iii. 467Google Scholar), which diverges at S. Bartolomeo from the Cave road, thence to the modern Via Casilina and on as before. I have not, however, been able to work out the problem thoroughly. The modern Via Casilina from S. Cesareo to Piombinara follows in all probability the line of an ancient road.

^{page 219 note 1} I do not know how old the name is—it certainly does not go back very far; Nibby does not seem to know it, so that it may be an invention of the last fifty years.

^{page 221 note 1} This would correspond with the road marked by Kiepert (Carta dell' Italia Centrale) as leaving the Via Latina at the Vigna Aquari, and running E. N. E. past the Molino S. Pio to the Via Labicana, which it reaches a little to the W. of Torre Pignattara. There are a few ancient pavingstones (not in situ) in the section between the Via Appia Nuova and the Via del Mandrione; but I am inclined to think that it falls better into the line of the Vicolo dello Scorpione, which (though retaining no traces of antiquity) seems to correspond to the line of an ancient road, which would have passed through the Porta Metrovia of the Aurelian wall

^{page 222 note 1} It is possible that this was the specus of the Anio Vetus: for Lanciani (Commentari di Frontino, p. 49, note 1) states that in 1880 he saw in the Vigna Marescotti, on the left hand side of the modern road, the specus of a large aqueduct constructed in opus reticulatum with its sides covered with deposit, and which he believed to be the Anio Vetus. The same aqueduct was, however, discovered in 1882 near Porta Furba, in making the military road (Not. Scav. 1882, 271, cf. 66), and again in 1890, both in the cutting of the Albano railway (Not. Scav. 1890, 12) and close to the point where the railway to Naples crosses the Via Latina. (Its specus is still visible in the short tunnel which carries the railway under the Marrana Mariana.) It was, further, found 450 m outside the Porta Maggiore crossing the Naples railway at right angles, and then turning sharply S. E. at 45.4° m. above sea level. In this case it would have required arches to cross the valley of the Marranella, which it would have had to do twice, a proceeding for which there seems to be no sufficient reason, though it must be confessed that the direction which it takes (Lanciani, Forma Urbis, 32) does certainly warrant this supposition.

^{page 223 note 1} The common tradition is represented by the Vita Silvestri in the Liber Pontificalis, by Bede (De sexta aetate Mundi), and by Nicephorus Callistus (viii. 31). The version according to which Helena was buried in Constantinople (Socrates, i. 17) rests on a misinterpretation of the expression used by Eusebius (vita Constantini, iii. 47),ἐπὶ τὴν βασιλϵύουσαν πόλιν. The Liber Pontificalis has (i. 65, ed. Mommsen): Eisdem temporibus fecit Augustus Constantinus basilicam beatis martyribus Marcellino presbitero et Petro exorcistae inter duas lauros et mysileum, ubi mater ipsius sepulta est Helena Augusta, Via Lavicana, miliario III … qui sepulchrum est ex metallo purphyriticus exculptus sigillis.

The name “inter duas lauros” as the name of an imperial domain occurs as early as the time of Tertullian, who (Apol. 35) speaks of those qui inter duas lauros obsident Caesarem. Marucchi (Guide des Catacombes romaines, 213) publishes a graffito (the date of which is not stated) in which Helena is actually mentioned. It probably belongs to the sixth century or thereabouts.

^{page 225 note 1} It is stated by him in a volume of his MS. notes now in my possession.

^{page 225 note 2} It is identical with the Vigna della Certosa, which lies on the S. side of the road, between the second and third (ancient) milestones. It is in this vineyard that the last arches (towards Rome) of the Aqua Alexandrina are seen (Lanciani, Commentari di Frontino, 171).

^{page 228 note 1} In 1755 the inscriptions C.I.L. vi. 631, 632, belonging to the collegium Silvani Aureliani (177 A.D.), the members of which were gladiators of different kinds, were discovered here. Their names and special performances are given in the first of the two inscriptions, which is the album collegii.

^{page 228 note 2} This name is that given to the statue by Helbig: it is generally known as Eros.

^{page 228 note 3} From a letter preserved in the Archives at Modena, and published in Bull. Com. 1898, 28, which bears date March 28th, 1787, we learn for the first time that these two busts were also found here.

^{page 229 note 1} This stamp had previously been found only in the Alban Hills, and is therefore classed among the stamps belonging to bricks from that district in the Corpus. The discovery of this copy (which is upon a flange-tile) shows that these bricks may have been made nearer Rome. Another stamp giving the name of the same factory owner—(Arescusa, Annia), C.I.L. xv. 1141Google Scholar—has been several times found in Rome itself.

^{page 234 note 1} Giostra means a large inclosed space (literally a ousting place or tilting ground).

^{page 235 note 1} This is the orthography adopted by most Latin writers (C.I.L. xiv. p. 274).

^{page 235 note 2} Cluverius, Italia antiqua, p. 947; Kircher, Latium, p. 120.

^{page 235 note 3} Ficoroni, Labico, Bertarelli, Labico. The name Labico was only given to Lugnano in the year 1880, Ficoroni's erroneous theory being thereby sanctioned.

^{page 235 note 4} Alberti, Italia, p. 144; Biondo, Italia Illustrata (ed. 1543), P. 102; Kircher, Latium, pp. 71, 72 (though on p. 120 he rejects this site in favour of Zagarolo).

^{page 235 note 5} See (among others) Holstenius, Ad Cluver., p. 194; Fabretti, De Aquis, 175; Nibby, , Analisi, ii. 157Google Scholar; Gell, Environs of Rome, 280; Westphal, Römische Kampagne, p. 79.

^{page 235 note 6} Capmartin de Chaupy, Maison de Campagne d' Horace, ii. 174; Rosa, Bull. Inst. 1856, 154; C.I.L. xiv. p. 275.

^{page 235 note 7} Vitale, , De oppido Labici dissertatio (Rome, 1778Google Scholar).

^{page 240 note 1} Here was found the fragmentary inscription published in Bull. Com. 1899, 36.

^{page 241 note 1} I owe my knowledge of the existence of the last section of this road to Padre Grossi-Gondi.

^{page 244 note 1} Padre Grossi-Gondi informs me that he has traced this road onwards in a S.S.E. direction, passing between the Villa Mondragone and the large Roman villa known as Le Cappellette (not to be confounded with the villa mentioned below on p. 260, which lies S. of Monte Compatri). The inscription on a waterpipe found here (C.I.L. xv. 7822), Matidiae Aug. fil. leaves it uncertain whether this villa belonged to the older or the younger Matidia.

^{page 244 note 2} I also saw here the lip of a dolium bearing the following stamp, which I published in Rendiconti dei Lincei loc. cit.

On the side of the dolium the following numbers were scratched: XXXIIII K VII, showing that the capacity of the dolium was 34 amphorae 7 heminae.

^{page 247 note 1} Truglio or Trullo means something circular: hence the name Lo Trullo, given to the great round tower where the Aurelian walls start from the left bank of the Tiber, to run towards the Porta Flaminia. These tombs were originally square, at any rate at the base, but have now become rounded masses.

^{page 248 note 1} The photograph reproduced was taken from a paper squeeze. The inscription has been published in Not. Scav. 1901, 327, from a copy of the same squeeze, which I gave to Prof. Tomassetti.

^{page 249 note 1} The name (“broken aqueduct”) is significant.

^{page 249 note 2} While these papers have been in progress, further agricultural operations have led to the discovery and destruction of remains of various buildings and of the specus of the Aqua Claudia and Anio Novus. The brickstamps C.I.L. xv. 2343 (1st century A.D.) and a fragment of an unpublished stamp were discovered. The lettering of the latter stamp is good and of an early type: the points are triangular.

^{page 250 note 1} The pavement of both these roads, which I saw in 1900 in fair preservation, has recently been removed.

^{page 251 note 1} To the N. of the tower, lying by the path, I found C.I.L. xv. 1408: I do not know the precise point at which it was discovered.

^{page 251 note 2} On my last visit I found here another stamp (see Not. Scav. 1899, 50)

also part of the side of a dolium marked LXVIII in letters 85 mm. high.

^{page 251 note 3} It is worthy of remark in this connection that Michele Stefano de Rossi (Secondo Rapporto sugli studi paleoetnologici, p. 16 from Giornale Arcadico, 1878, vol. lviii.Google Scholar) mentions the existence of a neolithic station in the neighbourhood of Torre Iacova. I acquired in 1901 a small axe-head of greenstone, said to have been found at the mouth of the Fosso di Fontana Candida a mile further W., and was told that many similar fragments had been seen in the vineyard. It is also noteworthy that in the E. bank of the Fosso della Lite, close to the bridge of the Aqua Marcia, a rock-cut tomb is to be seen.

^{page 253 note 1} Upon the N. slope of Monte Mellone there is a very large water reservoir constructed in opus reticulatum, sunk below ground, 41·85 m. in length, divided into two aisles, each 5 m. in height and 3·30 m. in width, by a partition wall O·89. m. in thickness, pierced by eleven arches, each 2·35 m. in height and span. Further down the hill, on the north-western slope, are the remains of the villa which it supplied—a platform supported on the W. side only by a wall, in front of which is a large cryptoporticus in opus incertum.

On the E. side of the road are the remains of a villa, broken up to facilitate cultivation, and some fragments of a very large inscription (the letters are well but not deeply cut, and are 17 to 18 cm. in height), cut upon white marble blocks 20 cm. in thickness. Too little remains to give any clue to the subject of the inscription as a whole—not even one whole word can be restored.

^{page 253 note 2} See Lanciani, Bull. Com. 1884, p. 210; C.I.L. xiv. 2925. These ruins are upon the N. of the modern road from Frascati to Colonna; but the house on the S. of it is also built upon the remains of some ancient structure. It is doubtful whether the ancient road from the Macchia di Fontana Candida to Monte Compatri ran E. of this house, as the modern path does, or W. of it; the latter alternative seems more probable. It may even have descended straight to C. Statuti, taking up the line of the boundary of the Agro Romano (cf. p. 195). The further question arises whether the remains on each side of the modern road from Frascati to Colonna belong to the same group, which would make the antiquity of this latter road extremely doubtful. The present road is, in many places, of recent construction; the older road having degenerated into a path. At the point under discussion, however, the two coincide. Interesting remarks on the subject by Stevenson are to be found in his MS. notes—vol. cit. f. 23—of August 21st, 1890. He considers its antiquity improbable, as it appears to him to pass between buildings belonging to a single group, not only at Pallotta, but a little further W., to the S. of Casale Statuti, where there are the remains of a large water reservoir on the S. of the road, and of the platform at a villa on the N., both orientated in the same direction. In a field-wall near the reservoir Stevenson found part of a rectangular brickstamp DVARI. It seems possible that the ancient road may have run higher up. Stevenson himself discovered such a road running parallel to the modern one, but could not trace it further than the large villa known as Le Cappellette (p. 260). Its direction, however, is such that, if prolonged, it would have joined the Via Labicana at Ad Quintanas. The existence of some artery of communication, corresponding to, though perhaps not identical with, the modern road from Frascati to Colonna, is practically certain.

^{page 253 note 3} Capmartin de Chaupy mentions the pavement of the road as existing in this wood in his day.

^{page 254 note 1} The distance between the 14th and 15th milestones is a trifle too great on the map. This is due to some inaccuracies in the military map, which were only discovered when my map was already drawn, so that complete correction was impossible.

^{page 254 note 2} There was also on this road a Catacomb of the SS. Quattuor Coronati, the locality of which is unknown (Stevenson in Kraus's, Realencyclopädie, ii. 113Google Scholar).

^{page 256 note 1} With 2770 was found the brickstamp, C.I.L. xv. 462 c (from the praedia Quintanensia, which were situated not far from the station Ad Quintanas: see C.I.L. xv. p. 8), also the Greek inscription Kaibel I.G.I. 1011, which mentions a grove sacred to the Muses, a statue of Venus, a statue of Domitian (Villa Albani), and a bust of Lucius Verus, and three others. See Vitale, De oppido Labici dissertatio (1778), p. 36; Letters di Winckelmann, ed. Fea, , iii. 247–251Google Scholar; Cavaceppi, , Raccolta d'Antiche Statue, i. 2Google Scholar, Stevenson, Cimitero di Zotico, p. 93 (who cites a letter of Lami, dated May 15th, 1758a). The inscriptions C.I.L. xiv. 2773, 2783, were also found here, and the fragments ibid. 2767, 2771, 2778. In the Lettere di Winckelmann, ed. Fea, , iii. 246Google Scholar, there is a notice of excavations in the Borghese property at Torre Verde (which I have not been able to locate) in which were found many fluted columns of marble and granite.

^{page 256 note 2} Dessau, (C.I.L. xiv. p. 275Google Scholar, note 5) denies, but on insufficient grounds, the theory, advanced first by Ficoroni, that the site of Ad Quintanas was different from that of the old town of Labici.

^{page 257 note 1} Tomassetti (Bull. Com. 1899, 288, Not. Scav. 1900, 50) gives a fragment of an inscription in travertine, with letters 22 cm. high, which may have belonged to this tomb.

^{page 257 note 2} This is most probably the meaning of the abbreviation Q.Q.

^{page 258 note 1} Tomassetti ascribes the T in the first line to the first inscription.

^{page 258 note 2} In the time of Capmartin de Chaupy it would seem that some of the pavement was preserved in situ just on the W. of the large tomb. Maison de Campagne d'Horace, ii. 174: “une des traces plus manifestes traversant un petit chemin à côté de la villa Pazzolini [la Pasolina]…dans la vigne dans laquelle il se perd…un Tombeau qui n'étoit pas de la dernière classe.”

^{page 259 note 1} Another deverticulum ran from about this point northwards, passing through the vineyards of Le Marmorelle to the modern Via Casilina (p. 237), while another, probably diverging from it and not from the Labicana directly, ran eastwards, skirting the northern slopes of the hill on which the village of Colonna stands (ib.).

^{page 259 note 2} There is much brick lying about. I also saw a drum of a tufa column 0·46 m. in diameter.

^{page 260 note 1} Vitale, (op. cit. pp. 22–24) refers to this road as a Roman road, and states that it starts from Le Marmorelle.

^{page 262 note 1} Further N.W. are traces of another villa in opus reticulatum.

^{page 263 note 1} It is, however, given in C.I.L. xiv. (No. 221^*) among the Inscriptiones falsae vel alienae, as belonging in reality to Velitrae, and is published among the inscriptions of the latter city. (C.I.L. x. 6608.)

^{page 264 note 1} In Stevenson's MS. notes voi. cit. f. 18) the following is inserted— “Luigi Moscatelli sotto la Colonna dice d' aver trovato un sotterraneo con iscrizioni e monete circa 600, pitture, dei tegoli con bolli. Pare che la vigna sia dal lato verso Monte Compatri.” The note is dated Nov. 1894, and signed M. Pasquale.

^{page 265 note 1} See Stevenson, MS. cit. f. 19, where another brickstamp found in a field wall below La Colonna on this side is also given, A NN I AE CO.

^{page 265 note 2} Here, in July 1890, Stevenson saw a brickstamp (of which a rubbing is given in MS. cit. f. 19).

^{page 266 note 1} The antiquity of this road, which goes direct from S. Cesareo into the modern road to Monte Compatri (the Via Maremmana inferiore), is extremely doubtful, though Westphal (Römische Kampagne, p. 80) maintains it, arguing from its straight direction and the large number of broken pavingstones which were in his time to be found in it. He makes it a continuation of the road from Frascati to Monte Compatri, which probably is ancient, though the modern road winds far more than the old road can have done (see map). In the map of Ameti a road is shown as ancient which seems to run from this point northwards, where it becomes lost. Before this it is crossed by a road from the N. side of Colonna to Zagarolo. There is probably some confusion between the roads that run from the Osteria della Colonna to Pallavicina and Cavamonte respectively, that which we have supposed to run from the 17th mile of the Via Labicana to the 14th of the Via Praenestina, and the path which runs, leaving the deverticulum last mentioned on the right, past Casale la Vetrice to the hill on the W. of Zagarolo.

^{page 268 note 1} The same photograph (which I took in January, 1900) will be found reproduced in Lanciani's New Tales of Old Rome, p. 33.

^{page 269 note 1} It is possible that a path crossing the Colle della Casa Romana in a S.W. direction, and coming from the so-called Via Praenestina Nuova, may follow the line of an ancient road, which would have fallen into the Via Labicana not far from this point, but the evidence is inadequate.

^{page 270 note 1} It is, again, possible that an ancient road ran along the Valle degli Appesi, coming perhaps rom the Via Praenestina Nuova, and following the communal boundary line between Zagarolo and Palestrina, and going on thence up to Algidus. But there is no definite evidence of its existence, and the configuration of the ground is such as to lend itself to the construction of imaginary lines of road.

^{page 270 note 2} So Fabretti, Incriptiones, p. 416 and map, Chaupy, , Maison de Campagne d'Horace, iii. 465Google Scholar.

^{page 270 note 3} Fabretti (Inscriptiones, p. 415) speaks of it as Lo Cimmero, and under this name it appears in his and Ameti's maps. Ficoroni (Labico, 37) supposes it to have been the site of the station of Ad Quintanas: he gives (ib. 40) an engraving of a ring found there, and (ib. 86) of a glans plumbea, with the inscription FIR. (C.I.L. ix. 6086, 40).

^{page 271 note 1} Whether these inscriptions were found in the ruins E. of the Villa Rospigliosi is quite doubtful: but if so, it would tend to show that the villa of Julius Caesar had remained a part of the imperial domain ever since his day.

^{page 271 note 2} So Westphal, Römische Kampagne, p. 77, and Kiepert, , C.I.L. xiv.Google Scholar, map.

^{page 272 note 1} It is to be noted, however, that the remains of pavement along the paths in the district of Praeneste, whether in situ or in field-walls, are extremely scanty; and very often the only piece of positive evidence for the antiquity of a road is the statement of some one of the writers on the topography of Praeneste, whose works date back a century or more. Cecconi's work is especially valuable in this respect, and seems to be trustworthy, though not complete in some cases, while in others he seems to admit the existence of too many ancient roads (supra, p. 214).

^{page 272 note 2} He very probably derived his knowledge of its existence from the maps of Fabretti (1680) and Ameti (1693), who mark it perfectly correctly, which Cingolani (1704) does not. Holstenius (ad Cluv. pp. 193 sqq.) who died in 1661, knew the truth also.

^{page 273 note 1} As I have already stated, the name Labico has only belonged to this village since 1880, up till which time it was called Lugnano (i.e. fundus Longianus ?). Nibby identified the place with the ancient Bola, but without sufficient grounds. The site is certainly a fine one for an ancient city, being isolated except for a narrow neck on the W. The rock has been perpendicularly scarped, and on the S. side is honeycombed with small caves, which may have been originally tombs or habitations (Ficoroni, Labico, 66), or, perhaps, never served for anything else than pigsties —their present use. Traces of antiquity are however wanting, so far as I know, and the scarping may date from the Middle Ages, for the place is known to have belonged to the Counts of Tusculum in the eleventh century.

^{page 273 note 2} Valmontone has similarly been identified by Nibby, (Analisi, iii. 369Google Scholar) with Tolerium, one o. the ancient Latin cities, but without adequate reason. Its site is even stronger than that of Labico, the rock on which it stands being isolated except on the N.W. No traces of earlier fortifications than those of the Middle Ages are to be seen, unless the blocks of tufa used in the houses belonged to the ancient walls, as Nibby thinks. He notes indeed that some of them seem to be in situ, and he further remarks the existence of some remains of opus reticulatum and of a sarcophagus of the third century used as a fountain basin. The rock on which the place stands is full of small caves, now used as pigsties, as at Lugnano.

^{page 273 note 3} Westphal (Römische Kampagne, 77, 81) states that he saw the “Unterlagen” or foundation blocks of the ancient road in the modern one between S. Cesareo and Lugnano, and pavingstones (not in situ) E. of Valmontone. There are several in the modern bridge just to the E. of the village, and a large number are to be seen in use in the modern pavement in front of the Osteria a little urther on.

^{page 274 note 1} The antiquity of this road was proved in 1899, by the discovery of pavement in situ about two miles from Velletri (Not. Scav. 1899, 338). It may be noted that Kiepert, (C.I.L. xivGoogle Scholar. map) prolongs this road to Valmontone, not to Labico—I do not know on whose authority.

^{page 274 note 2} See Fernique, op. cit. 123, Cecconi, op. cit. p. 43, n. 10, as to the antiquity of this last section.

^{page 274 note 3} The reference here (as in p. 275, n. 1) is to the hill to the E. of the Fontanile delle Macere, and not, as elsewhere, to the hill N. of Colle dei Quadri.

^{page 275 note 1} If this is so, we may agree with Rosa in placing Ad Pictas on the hill to the E. of the Fontanile delle Macere—the Colle Treare. The distance from Rome would then be 24 miles by the Via Latina, and just over 26 by the Labicana. This agrees better with Strabo, but not with the Itineraries.

^{page 275 note 2} The Colle della Lite, on which Bertarelli (Labico, 16) placed the site of Labici, is to be identified with this same hill.

^{page 275 note 3} Two reliefs in marble, of late date, and several copies of the brick-stamp C.I.L. xv. 2340 were found here in 1878 (Not. Scav. 1878, 68; according to which C.I.L. xiv. 2987, 3324, 3382, 3399 were also found here. These inscriptions are now preserved in the Palazzo Doria at Valmontone, and are said to have been found in 1789 in a place called La Cavalla; but this is certainly not true of C.I.L. xiv. 3416, 3418, 3423, which are placed with them (see p. 279, n. 1). In Not. Scav. l. c. the place is called Il Monumento).

^{page 276 note 1} Cecconi (op. cit. p. 88, n. 29) would make an ancient road run along this path to I Casali, and thence northward to La Marcigliana. The only objection lies in the steepness of the descent to the E. of I Casali.

^{page 276 note 2} The wood bore the name La Cacciata, i.e. the covert or preserve.

^{page 276 note 3} The measurements are as follows: total length, 17·66 m.; width 01 chambers, 3·88 and 3·96 respectively; width of dividing wall, ·95; span of arches, 2·40 to 2·62. To the S. of the reservoir, between it and the road, are traces of the villa which it supplied.

^{page 278 note 1} Holstenius (Aa. Cluv. p. 196, quoted in C.I.L. x. 6883) says: “ad ecclesiam d. Ioannis Baptistae columna milliaria extat litteris ferme exesis, quam ego xxvii ab urbe lapidem fuisse existimo.” If, as is probable, he is referring to the church mentioned on p. 277, the milestone he saw would have been either the 25th or 26th—not in its original position. At the Casale del Re near Artena (belonging to Dr. Cesare Caputi), I copied the following inscription from a milestone—

Imp(cratori) [Caesari] Dom(ino) [Nostro M. Aurelio Va.erio] Maxen(tio) p(io) f(elici) perpetuo [invicto Aug(usto).]

The fourth line I was unable to decipher satisfactorily. The number I made to be 23, but I was told it had been read as 24. The milestone was a marble column 0·34 m. in diameter. I was told that it had been found in the Quarto della Pescara, on the boundary between the communes of Giulianello and Artena, where it had been long in use as a boundary stone. If (as is most probable) it belonged to the Via Latina, it must have been brought from a distance of 3 miles at least to the N. of the point where it was found.

^{page 278 note 2} Possibly the remains of the Church of Nostra Donna in Selci, which, however, Cingolani and Ameti place on the N. side of the valley in which the Naples railway now runs, only a little to the E. of Valmontone. In this they are probably wrong. See Holstenius, l.c.

^{page 279 note 1} C.l.L. xiv. 3416, 3418, 3423.

^{page 280 note 1} De la Blanchère also cites Serangeli (whose MS., Notizie istoriche della terra di Monte-Fortino, was written in 1717) as speaking of a road which ascended from La Cacciata (p. 274) to the N. gate of La Civita—the name by which are known the remains of a large circuit of “Cyclopean” walls on the mountain above Monte Fortino (or Artena, as it is now called); but its course is not very clearly described.