The place known as Chauchitsa lies some sixty kilometres to the north of Salonika (Fig. 1). It was identified first as an ancient site at the end of the year 1917, during military operations. During the digging of trenches and other works antiquities were discovered which were recognised as belonging to the various stages of culture of a period extending from Neolithic to Roman times. The majority of the objects so discovered I published in the Annual of the British School for 1918–1919.

On April 15th this year I paid a preliminary visit to the site. While considerable damage had been done by the cutting of trenches, I felt satisfied that sufficient of the original ground remained undisturbed to justify excavation. I therefore commenced excavation on the 28th of April, 1921. As the isolation of the site made the conditions of a small test excavation of this nature somewhat difficult I decided to continue only, for such time as would enable me to ascertain the general nature of the site.

The dead of Pharsalos sleep sound; but more than a century ago their battlefield began to walk; and since Leake a dozen scholars in succession have found it in as many different places, over an area of forty miles by ten.

I must apologise for introducing yet another claimant theory, but I have at least aimed also at reducing the total by the abolition of one or two of its wilder predecessors. I propose, then, to summarise the ancient evidence, to state and criticise the theories based on it, and to conclude with a new attempt at solution, checked by walking over the plains of Pharsalos.

Far the first, then, among our authorities stands of course Caesar himself (Bellum Civile, III. 81–98). After the occupation of Gomphoi and Metropolis, all the states of Thessaly, except Larissa, had declared for him. Accordingly, ‘having found a convenient position in the Thessalian plain,’ he sat down there to wait for Pompey and the final decision. A few days later Pompey marched south to Larissa and joined Scipio. When next heard of, their army is encamped opposite Caesar on a hill.

In the last volume of this Annual (xxiii. 206 ff.) I attempted to restate the arguments for the view that the provincial era of Macedonia dates from 1st Dios, 148 B.C., and the Actian (or, to give it the name which it bore in Macedonia, the Augustan) era from 1st Dios, 32 B.C.

I mentioned (p. 210) Prof. M. Holleaux as having expressed in 1914 a doubt of the continued existence of separate Macedonian and Achaean eras running concurrently and a desire for further investigation of the question. This was, indeed, the starting-point of my own inquiry. I regret, however, that I overlooked a later statement of the same scholar (Rev. Et. Anc. xix. 81), who in 1917 accepted as fully established the Macedonian era of 148 B.C., and pointed out that the Oxyrhynchus Epitome of Livy proves that the defeat of Andriscus and the submission of Macedonia took place in that year (cf. Kornemann, Die neue Livius-Epitome aus Oxyrhynchus, 91 f., 113 f.). Further, in an article which became accessible to me only after my own discussion was already in print, Pomtow himself has declared in favour of the earlier starting-point of the Macedonian era, on the ground both of the Macedonian and also of the Delphian documents.

No self-respecting Cairene dragoman omits to point out to his clients among the curiosities of the mosque of Amr at Fostat two columns near the South door, which are endowed, according to popular superstition, with the miraculous power of discriminating between true Moslems and Unbelievers. Placed at such a short distance apart (some ten inches) that the passage between them can with difficulty be negotiated by a man of average build, the columns none the less allow a true Moslem, however stout, to pass between them, while an Unbeliever, however slim, finds passage impossible. In other words, the space is supernaturally widened if necessary to accommodate the former and contracted to exclude the latter class.

The columns actually used for this purpose at Cairo do not seem long to have been associated with the superstition. Visitors to the mosque in the sixties do not mention it, though they refer to the companion marvel of the column miraculously transported from Mecca. The superstition itself, however, is of great antiquity and relatively well documented. The purpose of the rite, a spiritual test, distinguishes it sharply from the many similar ‘passing through’ rituals universally current and generally considered ‘lucky’ acts practised with a view to the healing of disease, etc. Its symbolism, as we shall see, suggests a Christian origin.

The torso of this type published by Arndt in 1893 was found at Athens in the Theatre of Dionysos in 1865 and put into the magazine of the National Museum. In the register of the Acropolis Museum shortly after 1891 is entered the fragment of a bearded head, consisting of the front upper part to the lips, again found on the South Slope, which was placed in the Acropolis Museum magazine. To this in 1918 Dr. Keramopoullos joined the lower part of the beard, which had also been among the fragments there. At the end of 1920 I was fortunate enough to notice the connexion between the torso of the National Museum and the head of the Acropolis Museum. The touching surface leaves no doubt that they originally formed part of the same statue, and the combination (unrestored except for plaster in the irregular joins of head with beard and beard with body) appears to represent a straightforward copy of about the beginning of the first century B.C. (Pl. III.). That, though accurate in essentials, it was not highly finished, may be seen by examining the line where the hair leaves the side of the forehead, the ends of the curls on the chiton, and the summary, but confident, treatment of material both in chiton and himation: yet the work can at once be seen to be softer, fresher, and more valuable stylistically than the conventional herms of Naples, Palermo, and the Uffizi.

The authors' aim is to include in this article all the inscriptions on minor votive objects found at Sparta during the excavations, by the British School, of the Sanctuaries of Orthia (in 1906–10) and of Athena Chalkioikos (in 1907–08). They have been arranged according to material, those from each of these two sites being described separately, in the following order: stone, metal, ivory, pottery.

A few of these inscriptions, which have been published already in the Annual and subsequently in the Laconian volume of the Corpus (I.G. v. 1), are re-published here, without facsimiles, for the sake of completeness. Mr. Woodward made copies of almost all these inscriptions in 1907–08, but never completed the material for publication, owing to the claims of other work. Mr. Hondius, in 1921, re-copied and took impressions of all those inscribed on the carvings in soft stone, except Nos. 20 and 21, and with these exceptions his drawings of the unpublished examples are here reproduced. When Mr. Woodward's reading differs from his, the two versions are given. The inscriptions on the ivory flutes (Nos. 26, 27) are from Mr. Hondius's copies, as are Nos. 66 and 69 of those on the bronze objects; Mr. Woodward is responsible for those on Nos. 24, 67, 68 of the same material, which seem to have escaped Mr. Hondius's notice, and for all those on the pottery fragments from both the sites.

The following notes are intended to supplement the topographical articles published in Volumes XIV. and XV. of this Annual. In addition to the classical remains mentioned below, we wish to call attention to the interesting Byzantine churches at Chrysapha, Zaraphon, Gouves, Apidia and the monasteries of Daphni, north of Geraki, and of H. Ioannes Prodromos, south of that village.

Early in 1921 the Neuchâtel Asphalte Company, in beginning the construction of the new coast road from Old Phaleron to Vouliagmene(see map, Fig. 1), found not far inland from the small promontory of Hagios Kosmas some ancient foundations, and by them the inscription, which is here published. The discovery was communicated by Mr. Bottomley, Manager of the Company, to the British School, and to him I am indebted for permission to publish it, and also for the sketch plan of the ancient foundations here shown (Fig. 2). On the top of a low hill the line of a wall (about 1·50 m. thick) of the late fifth or fourth century B.C., built of rectangular blocks of poros, appeared, and on the top of this the inscription was found as indicated in the plan (Fig. 2). Slightly north of the inscription was a large block running out westwards at right angles to the main wall. If the main wall is the boundary of the Shrine of Dionysos mentioned in the inscription, it is possible that this big block formed part of the shrine itself. In any case it seems likely that we should recognise in these remains, scanty though they are, the ruins of the Dionysion in which the inscription was to be erected. On the south side of the road a double tomb came to light (Fig. 2). In the upper tomb, which was covered by a large tile so curved that it was in section half an ellipse, a small aryballos with a simple anthemion ornament in a very late red-figure style was found with a few much-decayed bones, perhaps those of a child; but in the lower tomb, which was made of two shallow rectangular troughs of terra-cotta, there were only a few bones. The tomb, which cannot be earlier than the late fifth century B.C., would not be likely to have any connection with the Dionysion, and was in any case outside it.

The results of recent prehistoric excavations in Greece show that there is no evidence to support, still less to prove, the widespread assumption that the round hoop-roofed house is the original type from which all forms of human houses have been evolved. The truth is that there is a greater variety of primitive house types than is usually supposed, and allowance must be made for the possibility that in course of development these exercise reciprocal influence on one another.

Dr. Bulle has pointed out that from the earliest times both rectangular and round huts and houses occur contemporaneously and that both forms are to be found to-day in use among primitive peoples. This is exactly what the archaeological evidence shows for Greece. Neither by stratification, nor by transitional forms does the Greek evidence support the view that in Greece the rectangular house developed from a round hut like those of Orchomenos I. Dr. Mackenzie, to whose paper I would refer, has collected most of the material for this question, and the case has also been clearly stated by Fimmen.

As Sir Arthur Evans had suggested that in view of the great discoveries in Crete, which have thrown entirely new light on the origin and development of the Mycenaean civilisation, it was desirable that supplementary excavations should be undertaken at Mycenae in the hope of being able to solve some of the problems connected with the Grave Circle, the Treasury of Atreus and other monuments, the British School applied to the Greek Ministry of Public Instruction for a permit. The permission was readily granted, as Professor Tsountas most courteously and generously gave up his rights on the site for the time being in favour of the British School. The excavation began on April 23rd, 1920, and lasted till July 9th, but the number of men employed never rose above twenty-five. The headquarters of the party were at the ‘Fair Helen Hotel,’ which was found as usual clean and comfortable, while its proprietor, Demetrios Ntases, himself a veteran digger, gave us much sage advice. A house was also rented to serve as a temporary museum and store-room.

The Greek Ministry of Public Instruction readily renewed our permit, and as usual gave us every facility they could. Mr. Theophaneides was specially sent to act as Government representative during the excavations, which were also visited by Dr. Keramopoullos, Ephor for Argolis, and after the close of the work Mr. Balanos came to inspect the walls and tholos tombs and consult with the Director about the measures necessary to repair and conserve them. Other visitors included Dr. Chatzidakis of Candia and Professor Kurt Mueller.

The actual excavation began on May 22nd, though Miss Lamb and Miss Herford arrived two days before and occupied themselves in arranging the workroom and other preparations. The number of men employed was at first twenty-five, and gradually rose to forty-seven, and then diminished again by degrees to six, when work ended on July 27th. The headquarters of the party were in the Metrovgenes house, which proved convenient and comfortable, while accommodation for the lady members was retained at the ‘Fair Helen Hotel.’

The pieces of fresco here described come from the so-called ‘Ramp House,’ the excavation of which was begun by Schliemann in 1876, and finished by ourselves in 1921. It lies south-east of the Grave Circle, immediately west of the Little Ramp, and east of the House of the Warrior Vase. In date it belongs to the Third Late Helladic period. Of the frescoes, many came from the ‘Megaron,’ where they were found below a floor level in a deposit dating from Late Helladic II. or earlier times; the details of the stratification will be given in a later section. Some fine fragments came from the narrow passage between the ‘Megaron’ and the Little Ramp, where they were found at approximately the same level and in a deposit of the same date and character; indeed, one or two pieces of pottery found there join pieces from the ‘Megaron.’The deposit seems to have extended to the north of the Ramp House, for it was found immediately outside the north wall, and also at the foot of the Little Ramp: both places produced fresco. That such variety and quantity of fresco should have decorated any of the buildings that preceded the Ramp House on this site seems improbable.

During the campaign of 1920 we sank a few trial pits within the acropolis walls to test the stratification and to try to find, if possible, earlier remains, after it had become clear to us that the buildings immediately round the Grave Circle were not earlier in date than Late Helladic III. Two such pits were made on the slope about fifty metres south-east of the top of the Ramp. Near by we noticed what was at first taken to be a rectangular pit enclosed by a rough stone wall measuring 1·00 by 1·00 m. and ·45 m. thick. The interior of this was cleared out, and then we found the walling was merely the coping (about 2·22 m. high at the point where it is best preserved) to a circular well-shaft about 1·15 m. in diameter, driven down through the hard limestone rock.

As we excavated the well-shaft nothing worth noting, except some miscellaneous potsherds and the remains of a crushed leaden vessel, came to light till a depth of about six metres below the highest point of the coping, when some fragments of a funnel-shaped rhyton in Lapis Lacedaemonius (A, below) were found. From this down to a depth of 7·75 m. the other objects here described were recovered.