southern extremity of Africa, the rounding of which by ancient navigators must have required many months.

Rennell, in endeavouring to prove the possibility of the circumnavigation, insists upon the benefit which the Phoenicians might have derived from the monsoons on the eastern, and the trade-winds on the western side of Africa, and has pointed out the manner in which that benefit would accrue; we have still to consider, however, that in the catacolpic navigation of the ancients, a prevailing wind might be fair at a certain distance from the coast, but made adverse by the windings of that coast, and prove an obstinate impediment to the doubling of some of the promontories.

If the experiment of circumnavigating Africa had ever been successfully tried, one cannot conceive that it should have been so entirely forgotten in Egypt, that the geographers of Ptolemaic and subsequent times were not agreed in opinion whether Africa was or was not a peninsula. It matters little what was believed by such Latin writers as Cornelius Nepos, or Pomponius Mela, or by Pliny, who says not only that Africa had been circumnavigated by the Phoenicians of Necho from the east, but also in the opposite direction by Hanno the Carthaginian, though we know from the journal of Hanno that he never reached beyond, or at least far beyond, Serra Leona on the western coast of Africa. One of the most learned of the Greeks, Eudoxus, the friend of Plato, could not have placed any confidence in the story of Herodotus, when, instead of referring to it, he supposed that Africa had been circumnavigated by

Hanno, beginning from the Red Sea. Eratosthenes and Strabo, though persuaded that Africa was a peninsula, appear to have been ignorant of any practical proof of the fact, and consequently must have disbelieved the report of Herodotus. The judicious and inquiring Polybius, who had himself been employed on a voyage of discovery on the western coast of Africa, expressly states that it never had been ascertained whether Libya was surrounded by the sea2. Finally, Ptolemy, whose work comprehends every thing that was known of Africa in the Greek and Imperial times of Egypt, and whose information has of late years been verified in some remarkable instances, denied the junction of the Atlantic and Indian seas, and must therefore have believed that Africa was not a peninsula 3.

As to the sun having been seen, as Herodotus informs us, on the right hand of the navigators in sailing round Africa, that is to say, in the northern portion of the heavens, the Egyptians could hardly have been ignorant that such a phænomenon would be the consequence of a prolonged southern course, a part of Egypt itself having been within the tropic. This part of the story, therefore, the only part which Herodotus could not believe, was an easy and natural addition, if we suppose the whole to have been a fiction, like so many others which Herodotus received from the Egyptian priests.

2 Plin. 5, 1; Polyb. iii. 38. 59.

3

On the whole of this question and the cited authorities, see Rennell, Geography of Herodotus, cc. 24, 25; Gosselin, Recherches sur le tour de l'Afrique.

ELANA, by the EDITOR.

The Editor remarks that "Elana was situated 10 miles east of Petra (Euseb. Onom. s. v. 'Hλá0), and 150 miles S.E. of Gaza (Plin. 5, 11)," and that "the site is now occupied by a fortress called Akaba, in which a garrison is stationed, because it lies on the route of the Egyptian pilgrims to Mecca." But the ancient description of the position of Ælana is not correct. Akaba is 60 G. M. south-westward of Petra, and 120 G. M. due south of Gaza.

APAMEIA PHRYGIÆ, by G. L.

G. L. is not correct in saying that "Arundell was the first who clearly saw that Apameia must be at Denair." The same is shown in my Asia Minor (p. 158), which was published ten years before the journey of Arundell. Arundell was the first who visited and described the place after Pococke.

APOLLONIA, by the EDITor.

The Editor, in reference to four cities of this name, which were frequently confounded, namely those of Mygdonia, Chalcidice, Acte of Athos, and Thrace, remarks that they "are correctly distinguished by Leake, who errs however in making the passage of Athenæus, 1. 8, c. 3, p. 334 e, refer to No. 6 (Políghero) instead of to No. 5 (Póllina). (Northern Greece, vol. iii. p. 457.)"

I cannot assent to this correction. The words of Athenæus are . . τὴν Βολύκην λίμνην, περὶ ἧς Ηγή

σανδρος ἐν τοῖς ὑπομνήμασί φησιν οὕτως" [ ̓Απολλωνίαν] τὴν Χαλκιδικὴν δύο ποταμοὶ ῥέουσιν, ̓Αμμίτης καὶ Ὀλυνθιακός· ἐμβάλλουσι δὲ ἀμφότεροι εἰς τὴν Βολύκην λίμνην. Bolyce was the name of the lake or lagoon at the head of the Toronaic gulf, near which Olynthus was situated. Boλúkŋ λíμvŋ means indeed lake of Olynthus, B being a common Macedonian prefix. Apollonia Mygdoniæ, the No. 5 of the Editor, was situated to the north of the range which separated Mygdonia from Chalcidice. It preserves its ancient name, and stands, as he justly observes, on the great road from Thessalonica to Amphipolis. In the article Bolbe by the same writer, he says that the Ammites and Olynthiacus were streams falling into the lake Bolbe. It is evident that he believed Bolbe and Bolyce to have been one and the same lake instead of having been 25 G. M. apart. Vide Travels in Northern Greece, iii. p. 155.

ARGOS ORESTICUM, by the EDITOR.

"Leake confounds the Macedonian Argos with Argos Oresticum." The reasons given by the Editor for this opinion are, 1. That Stephanus mentions an Argos in Macedonia as well as Argos Oresticum. 2. That the same Stephanus on the authority of Hecatæus describes the Orestæ as a people of Epirus (Moλosoikov Élvos). 3. That Strabo also places the Orestæ in Epirus. 4. That Ptolemy distinguishes the Epirote from the Macedonian Orestis. But the text of Strabo when interpreted by an autopsia of the countries in question, leaves no doubt that the Orestæ, whose capital was called Argos, were one

of the tribes of upper Macedonia bordering northwestward upon those of Illyria, and south-westward upon the Epirotic tribes (rà 'Hepwτika Ovn), among whom the Oresta were reckoned until they were subdued by the kings of Macedonia together with the Lyncestæ, Pelagones and Elymiotæ, and with these constituted the country called upper Macedonia (ἡ ἄνω Μακεδονία). We find Argos named together with Stobi and Pelagonia among the cities of the second Macedonia as late as the tenth century (Const. Porphyrog. de Them.). Under such circumstances it is not surprising that the Orestæ should have been attributed sometimes to Epirus and sometimes to Macedonia. The supposed mention by Stephanus of an "Argos in Macedonia as well as Argos Oresticum" arises from an obvious error in the text of Stephanus, in voce "Apyos, where he enumerates the cities of this name, and where instead of ἑβδόμη κατὰ Μακεδονίαν ὀγδόη "Αργος Ορεστικὸν, ἣ ἐν Σκυθίᾳ, we ought to read ἑβδόμη "Αργος Ορε στικὸν κατὰ Μακεδονίαν, ὀγδόη, ἡ ἐν Σκυθίᾳ. As to the double mention of Orestis in Ptolemy's chapter on Macedonia, it cannot refer to two different districts, as in both instances Amantia is named as the only town. It is clearly an erroneous repetition in the text of Ptolemy. There can be no doubt as to the position of Amantia in the valley of the Celydnus near the Acroceraunian mountains and the Ionian gulf (Tr. in N. Greece, i. p. 375). Its being placed in Macedonia by Ptolemy is accounted for by Strabo, who says (p. 326), Ενιοι δὲ καὶ σύμπασαν τὴν μέχρι Κορκύρας Μακεδονίαν προσαγορεύουσιν. How the district of Amantia, which is a part of Chaonia in