vitality of the god of vegetation. See A. G. Bather, 'The problem of the Bacchae,' Journal of Hellenic Studies, 14 (1894), pp. 244-263.

2. 5. Plataea. The ruins of Plataea occupy a triangular or fanshaped plateau which projects northward into the Boeotian plain from the foot of Mt. Cithaeron. The base of the triangle is turned towards the plain and the north, while the apex almost touches the declivity of Mt. Cithaeron, from which it is divided by a shallow ravine some 150 yards wide at the narrowest point and about 16 feet deep. The plateau, which is nearly a mile long from north to south by rather less than three-quarters of a mile wide at the widest, rises little above the level of the surrounding plain. It is highest at its southern extremity, from which it slopes at first rather sharply and then gently to the north, where it terminates in a steep but short descent to the plain. On its eastern and western sides it is bounded by the beds of watercourses. The sides of the plateau are rocky and precipitous only at a few points, as at the south-western end, where the wall runs along the edge of a rough and jagged cliff nearly vertical though only a few yards high, and again about the middle of the west side, where the wall overhangs a steep rocky cliff from 10 to 26 feet high. But as a rule the sides slope gently down to the plain below; and although the slope has been diminished since the classical age by the washing down of soil from the surface of the plateau, it was probably never high or steep enough to constitute a sufficient natural defence without the aid of walls. The general surface of the plateau is not broken by any notable depression except in the north-east corner, where it is split by two small ravines or hollows, formed by a couple of tiny brooks running north. The higher part of the plateau, towards the south, presents for the most part a rocky uneven surface, which does not admit of cultivation; but the lower part, towards the north, which is also the larger, is covered with soil to a considerable depth (about 10 feet in places) and is cultivated throughout, except where the foundations of ancient or mediaeval buildings prevent the passage of the plough. In many places, however, the

soil is very thin and the rock crops up, especially towards the west and south. The rock of the plateau as well as of the ridge of Mt. Cithaeron to the south is a coarse grey marble, which corrodes into deep holes and channels when it is exposed to the weather. Of this rock all the existing walls of the ancient city are built. The soil is a clayey lime earth, very loose when dry, but exceedingly clinging and sticky when wet; indeed in rainy weather the mud makes the surrounding plain almost impassable, sticking to the traveller's boots in great clotted masses and driving him to walk in the beds of running streams as the only tolerable pathways through the slough.

The ancient walls may be traced all round the edge of the plateau ; in some places they are standing to a considerable height. The length of the circuit is about 2 miles. Inside of the circuit-wall are two cross-walls, an upper and a lower. The upper cross-wall extends, roughly speaking, from east to west, and cuts off the apex of the triangle. The lower cross-wall cuts off the north-west corner of the site, and appears to have formed, at some date, an acropolis.

Five different styles of masonry may be distinguished in the existing walls, pointing to as many different periods of construction. (1) The earliest part of the walls is built of well-jointed polygonal masonry, the blocks being of fairly uniform size, often pentagonal and occasionally hexagonal in shape. The pieces of wall built in this style are the worst preserved of all, the stones being much corroded and weatherworn. (2) The walls of the second period are built in an intermediate style,

better than the walls of the first period but not so well built as the walls of the third period; polygonal blocks hardly occur in them at all. (3) The walls of the third period are the best built and best preserved. To this period belong the upper cross-wall and the wall at the north-east corner. The blocks are large, measuring from 3 to 10 feet in length by 3 feet in height and 2 feet in thickness. They are four-sided and are laid in horizontal courses, with the edges neatly and accurately fitted. The vertical joints are commonly slanting, not perpendicular. The outer surface of the blocks is somewhat rounded or

(4)

It is

bulging (bull-nosed'), and is cut vertically into wide and rather deep grooves. Altogether the workmanship is careful and excellent. The fourth period is represented by the lower cross-wall alone. carelessly built of materials taken from earlier buildings, including architectural fragments and blocks of the walls of the second and third period. The joints are loose, and little trouble has been taken to fit the stones into their new positions; blocks taken from walls of the third period, for example, have their grooved side turned in just as often as out. At some time mortar and tiles were employed to fill up the chinks; but whether this was done when the wall was built or subsequently, does not appear. (5) Latest and worst of all are some fragments and stretches of Roman or, more probably, Byzantine wall, built of rubble and tiles laid in mortar. They may be seen here and

there on the north and west sides.

The walls of the first four periods are very uniformly about 11 feet thick. The outer face is built of larger blocks and in a rather better style than the inner, but the difference is not great. In all the walls the space between the outer and the inner face was filled up with smaller rough stones and earth. Their height and the way in which they were finished at the top cannot be determined. In many places, especially at the north-west and south-west, the wall has wholly disappeared, and its line can only be traced by the smoothing of the natural rock as a bed in which to lay the first course of masonry. All the walls were defended at intervals by projecting towers, of which there are many remains. Four of these towers are round; all the rest are square.

The walls at the southern end of the site, beyond the upper crosswall, belong to the earliest period. From the greater age of these walls and the height of this part of the site above the rest of the plateau it has been conjectured that the southern end of the site, forming the apex of the triangle, was the original acropolis or even the original city, which may have been a good deal smaller than the later city. No trace, however, has been found of an old wall enclosing this part of the site on the north. Such a wall may very well have extended east and west nearly on the line of the existing upper cross-wall and may have been taken down when that cross-wall was built. After the construction of the cross-wall the part of the site to the south, enclosed by the oldest walls, would seem to have been left outside of the city walls. For the cross-wall, strengthened by projecting towers, has all the appearance of being an outer wall; and of the older walls to the south of it not much is left, little more indeed than a single course. At the extreme southern point of the circuit there are the remains of a square tower, measuring 18 feet on the sides. From this point there is a fine view over the whole site and the plain of the Asopus away north to the line of low hills beyond which, hidden from sight, lies Thebes.

By far the best preserved as well as best built portion of the whole circuit is the upper cross-wall. Starting from the west wall it runs south-east for 445 yards in a nearly straight line, then bends round at an angle and runs north-east to join the eastern circuit-wall, It is

everywhere 11 feet thick; both faces are carefully finished, and the intermediate space is filled with earth and rubble. The height of the existing wall varies greatly; in some places it barely appears above ground, while in one place (the third tower from the west) it is standing to a height of 12 ft. 6 in. Along the outer or southern side of the wall are remains of eight quadrangular towers projecting from the curtain at intervals of about 140 feet from each other. Each tower measures about 22 feet on the face by 16 feet on the sides, and appears to have been built solid, the interior having been filled up with earth and rubble. A peculiarity of this cross-wall is that on its inner side there are several platforms or thickenings, each about 33 feet long and 3 feet thick. These were probably buttresses to strengthen the wall, but too little of them is left to allow us to be positive on this head. Both wall and towers are constructed on their outer and inner faces of very massive blocks, three courses of which make a height of about 7 feet. At one point in the wall I counted eight courses standing. The towers form an integral part of the wall; they were obviously built at the same time and continuously with it, not added at a later time.

The lower cross-wall, which cut off the north-western corner of the site so as to form an inner fortification, has seven quadrangular towers projecting from its outer side. They measure about 20 feet on the face by 18 feet on the sides. The third tower from the west is standing to a height of about 12 ft. 6 in. Leake and Vischer regarded this north-western enclosure as an acropolis or citadel. Mr. Grundy holds that the city which was besieged and taken by the Peloponnesian army in 429-427 B.C. was confined to this north-western corner of the site. He argues that this corner is strategically the strongest part of the whole site and must always have been surrounded by a fortificationwall. The slopes on the north and west are still steep, in places precipitous, and may have been much steeper in antiquity; and the foundation of a wall placed upon their edge would be from 60 to 90 feet above the plain at their base. The total circumference of the enclosure is 1430 yards; and a circuit-wall of this extent, aided by the natural strength of the ground, might very well (Mr. Grundy thinks) have been defended by 480 men, which we know to have been the strength of the garrison (Thucydides, ii. 78), whereas it is difficult or impossible to believe that so small a garrison could have successfully defended a circuit-wall of 2 miles in length, which is approximately the length of the existing circuit-wall of Plataea. We may agree with Mr. Grundy in thinking that the city which the 400 Plataeans and 80 Athenians defended for two years against the enormously superior numbers of the Peloponnesian army was much smaller than the later city of which the remains cover the plateau; but his theory that this earlier city coincided with the fortified enclosure at the north-western corner of the site is not borne out by architectural evidence. For, as we have seen, the cross-wall which cuts off this corner is of later date than almost all the rest of the walls. It may even belong to the Byzantine period, and would seem to have been constructed at a late time when the town had shrunk to these narrow limits.

The original city which was besieged and taken by the Peloponnesians may have occupied, as has already been suggested, the southern part of the site, the walls of which are the earliest of all. To this view, however, it is objected by Mr. Grundy that a town of this area, measuring perhaps from three-quarters of a mile to a mile in circumference and unaided, except to a small extent on the west side, by natural defences, could not have been held by 480 men against a large army. However that may be, it would seem that at some time subsequent to the Peloponnesian war the city was enlarged. This may have been done either when the Plataeans were restored after the peace of Antalcidas in 387 B.C. (Paus. ix. 1. 4) or at the time of their second restoration by Philip of Macedon in 338 B.C. (Paus. ix. 1. 8). That the ancient Plataea did not comprise the whole of the area afterwards occupied by the city is moreover made probable by the fact that in 479 B.C., at the time of the battle of Plataea, the temple of Hera was in front of the city (Herodotus, ix. 52), whereas when Pausanias visited Plataea the temple appears to have been within the city. It is true that the old temple which existed at the time of the Persian war was destroyed some time after 427 B.C. by the Thebans, who built a new temple in its place (Thucydides, iii. 68); but probably the new temple occupied the site of the old one. The remains of a large temple, which was probably that of Hera, were discovered a few years ago by the Americans within the area of the later city (see below, p. 17 sq.).

Inside of the walls of Plataea, or rather of the northern portion of it which is cut off by the upper cross-wall, lies a mass of ruins, including the remains of a number of Byzantine churches, some of which were excavated by members of the American School of Archaeology in 1889 and 1890. The number of these churches (there were at least eight of them within the walls) proves that in Byzantine times a populous town must have existed on the site. Some of the churches contain ancient architectural fragments of both the Greek and Roman periods, including some blocks of white marble with the 'egg and dart' pattern executed in good style. Most of these fragments (architraves, capitals, and bases) belong to the Roman age. Two springs, one immediately to the west of the walls where the women of the neighbouring village of Kokla wash their clothes, the other some 250 yards to the east of the plateau, have walls made of ancient fragments of white marble. The marble of all the architectural pieces found on the site resembles Pentelic, but probably came from a nearer quarry. Within the northern portion of the site, a little to the east of the lower cross-wall, are seven or eight small rectangular holes or niches cut in the rock, probably for the reception of votive or other tablets. Just outside of the west wall, and close to the spring, are some half-dozen sarcophaguses perched on the rocks; each of them is hewn out of a single piece of the common coarse grey marble of which the rock consists. To the south of the sarcophaguses are some graves in the shape of rectangular pits hewn in the rock. Amongst the inscriptions found by the American archaeologists on the site were a long fragment of the Latin preamble to Diocletian's edict On prices, a fragment of the Greek text of the same