ence of ten miles, included 379 turrets, and opened into the country by thirteen gates'." This serves for the second date. Lastly, "whatever fancy may conceive, the severe compass of the geographer defines the circumference of Rome within a line of twelve miles and three hundred and forty-five paces." These words of the same historian apply to the third point of time.

Now it is quite clear that all these measurements differ, and yet it is equally clear that the historian avers they are all the same. He says, in another place, speaking of them in the age of Petrarch, the walls "still described the old circumference 3." It is true he cites authori ties; but he speaks without reserve, and has not attempted to account for the difference between the three above-given dimensions. We shall find no help, therefore, from the Decline and Fall of the Roman Empire, unless we follow

'Decline and Fall, cap. lxxi. tom. xii. oct. p. 398.

2 Ibid. сар. xli. p. 227.

3 Ibid. cap. lxxi. p. 411. tom. xii. Mr. Gibbon has failed to observe that the walls were dilated after Aurelian and Probus, by Constantine, who took down one of the sides of the Prætorian camp and made the remaining three serve for the fortifications of the city, whose circuit thereby became necessarily somewhat enlarged.

only one of these various accounts, and believe in the third computation, which is that assigned by D'Anville from Nolli's map, and which coincides with the experience of two of our countrymen, who made a loose calculation' of the circuit by walking round the walls in the winter of last year, (1817).

'The writer was one these. The following is a note of their walk. They set out from the banks of the Tyber, near the Flaminian gate (Porta del Popolo); their rate of walking was 592 paces in five minutes, and they noted the time from gate to gate. To the Porta Pinciana (shut) 18 minutes-Porta Salara 8-Porta Pia 3-a shut gate (Querquetulana) 12-St. Lorenzo 8-Maggiore 74-Lateran, or Porta St. Giovanni, 12-Porta Latina (shut) 17-Porta Capena, or St. Sebastiano, 41-a shut gate 33-Porta di St. Paolo (Ostian) 144—– delay 4-within the wall, the outer circuit not being accessible, 42-delay 7-within the walls down to the Tyber 61-delay 4-bank of the Tyber within ruined wall 10-delay occasioned by going across the Tyber to the opposite corner 381-from bank of the Tyber to Porta Portese -Porta Aurelia, or S. Pancrasio, 181-Porta Cavalli leggieri 141—a shut gate (Porta delle Fornaci) 21-Porta Fabbrica (shut) 6—Porta Angelica 141-Porta Castello (a shut gate) 51-round to the corner of the bastion of St. Angelo, on bank of the Tyber 73along the bank of the Tyber where there are no walls, to the ferry at the Ripetta 74-delay 10-crossing the Tyber and walking along the bank to the corner of the walls whence they set out, 63.-The time employed in walk was 4 hours, 38 minutes; the delays amounted to one hour, four minutes, and a quarter. The time taken in walking round the actual circuit

Poggio's measurement was probably nearly exact, for he did not reckon the ramparts of Urban, and, perhaps, not the Vatican; but it is singular, that the pilgrim of the thirteenth century, who undoubtedly saw the same walls, and enumerates very nearly the same quantity of turrets, should' give to them a circumference double that of the Florentine, and nearly coinciding with that of the time of Alaric, that is, twenty-one miles. If, however, they were so accurately measured at that time, the present walls cannot possibly stand on the site of those of Aurelian; for, since the Vatican has been included, and also the ramparts of Urban VIII, which Mr. Gibbon has overlooked, or falsely confounded with the Vatican, the modern circuit being larger on one side the Tyber, and the same on the other, it is evident that the whole circumference at present must be greater than it was under Aurelian. That is to say,

twelve miles, three hundred and fifty-five paces, of the city was three hours, thirty-three minutes, and three quarters. Supposing the rate of walking to be about three miles and a half an hour, the measurement is twelve miles and a quarter.

"Murus civitatis Romæ habet turres 361. Castella id est merulos 6900, portas 12, pusterulas (portæ minores) 5. In circuitu vero sunt milliaria 22. exceptis Transtiberim et civitate Leoninâ id est porticu St. Petri." Lib. de mirabilibus Romæ, in loc. citat. p. 283.

are more than twenty-one miles" which is absurd."

The present walls may touch at points and take in fragments, but they cannot include the same circumference as the twenty-one miles accurately measured by the mathematician Ammonius. Some assistance might be expected from the examination of the walls themselves: but here again it may be necessary to warn the reader in what manner he is to understand an assertion which he will find in another work, lately published, of the same author1 "Those who examine with attention the walls of Rome, still distinguish the shapeless stones of the first Romans, the cut marbles with which they were constructed under the Emperors, and the ill-burnt bricks with which they were repaired in the barbarous ages." Now the whole of the modern walls are of brick, with the following exceptions. There are some traces of the arched work on which the walls of Aurelian, perhaps, were raised, about the Porta Pia and the Porta Salara. There are buttresses of travertine, and, in one

* "Ceux qui examinent avec attention les murailles de Rome distinguent encore les pierres informes des premiers Romains, les marbres bien travaillés dont on les construisit sous les Empereurs, et les briques malcuites dont on les reparoit dans les siecles barbares." Nomina gentesque antiquæ Italiæ, p. 209.

case (the Porta Capena), of marble, about the gateways, which are of the same imperial date. There are single shapeless fragments of marble here and there, mixed up with the more modern work, and occasionally laid upon the top of the walls. This is all that can apply to Mr. Gibbon's description; for as to the shapeless stones of the first Romans, they cannot be discovered, except in those scarcely distinguishable mounds which are within the walls, a little beyond the Therma of Diocletian, and are usually thought part of the Tullian rampart'. It must be remarked also, that there is no evidence that the walls of the Emperors were of cut marble. The authority of Cassiodorus has been followed by Marlianus * and others, as affording a proof that they were composed of square blocks. But it has been noted by Nardini3, on another occasion, that the Gothic minister, in making use of the word mania, does not always allude to the walls of the city, but of other structures; and in that sense we have before interpreted, in a preceding

2

1 The plan in the last edition of Venuti lays down the Agger Tarquinii in the space between the Lateran and Santa Croce in Gerusalemme: repeated search may fail in finding any trace of this Agger. Donatus positively says there is none. Lib. i. cap. xiii.

2 Urbis Romæ topographia, lib. i. cap. ix.

3 Roma Vetus, lib. i. cap. viii.