Full text of "A history of ancient geography among the Greeks and Romans, from the earliest ages till the fall of the Roman Empire"

y...-. . J.,
,>■ -j.^K',"'. '.'.;,■; ■■■ ,

r"*^'

y?2.0b^A

HISTOEY

ANCIENT GEOGEAPHY

AMONG THE GKEEKS AND EOMANS

FEOM THE EARLIEST AGES TILL THE FALL OF

THE EOMAN EMPIRE.

By E. H. BUNBUEY, F.E.a.S.

WITH TWENTY ILLUSTRATIVE MAPS.

IN TWO VOLUMES. — VOL. IL

LONDON:
JOHN MURKAY, ALBEMARLE STREET.

1879.

The right of Translation is reserved.

LONDON :

PRINTED BY WILLIAM CLOWES AND SONS,

STAMFORD STREET AND CHAEING CROSS.

CONTENTS OF VOL. 11.

CHAPTEE XVII.

f GEOGRAPHERS AETER ERATOSTHENES.

Section I. — Hi-pparchus.

§ 1. Authority of Eratosthenes attacked by Polemon. § 2. Hipparchus : his
criticisms. § 3. Conceives the idea of a map of the world, based on
nstronomical observations of latitude and longitude. § 4. Adopted for
the most part the conclusions of Eratosthenes. § 5. His division of
the world into dimata. §§ 6, 7. Enumeration of these as given us by
Strabo. Accepted the existence of Thule. § 8. Erroneous views of
geography of .sia ; exaggerated extent assigned to India. § 9. His
criticisms o^ j^ratosthenes frequently erroneous. § 10. Indicates mode
of determ.'.iing longitudes by ecl^es. § 11. Eefuses to admit the
continuity of the Ocean .. .. .. ,. .. ,. Page 1

Section 2. — Polybius.

§ 1. His work marks an important epoch in geographical knowledge. In-
creased knowledge of the West, arising from extension of Eoman power.
§ 2. Life and travels of Polybius. His great interest in geography.
§ 3. General views. Little acquaintance with the western and
northern shores of Europe. § 6. His account of Spain, and of Gaul.
§ 5. Greatly increased knowledge of the Alps. Passes across tbem. § 6.
Imperfect geographical notion. Statement as to their height. § 7. Geo-
graphy of Italy. \|§ 8. Of the countries on the Adriatic. § 9. The
Via Egnatia : value of Eoman roads. § 10. Description of Byzantium,
the Euxine, and the Palus Mseotis. § 11. Few notices concerning
Asia. § 12. Valuable information respecting Asia Minor. § 13.
Africa : increased knowledge of, from the Punic Wars. § 14. Ex-
ploring expedition commanded by Polybius : very imperfectly known
to us, § 15. Doubts the continuity of the oceans. § 16. His state-
ments of distances, as reported by Pliny t,.. .. .. 16

NOTE A.
„ B.

PAGE

Hannibal's Passage of the Alps .. 37
Western Coast of Africa, according to

Polybius 40

iv CONTENTS OF VOL. II.

CHAPTEE XVIII.

POLTBITJS TO POSIDONIUS.

Section 1. — Progress of Roman Conquests.

§ 1. Progress of geograpliical knowledge from Polybius to Strabo dependent
on that of Eoman arms. Conquest of Southern Gaul, and of almost
all Spain. § 2. "Wars with Dalmatians, Pannonians, Thracians, &c.
The Danube first reached. § 3. No progress in Asia. Hellenization
of Bactria and adjacent provinces. § 4. Kise of Parthian monarchy an
obstacle to extension of knowledge eastwards ,. .. Page 43

Section 2. — Oreeh Writers.

§ 1. Apollodorus of Alexandria : his treatise on the Homeric Catalogue of the
Ships. His views of Homeric geography. § 2. Demetrius of Scepsis :
his work on the Trojan Catalogue. His views concerning the site of
Troy, and of Homeric geography generally. § 3. Agatharchides of

"* Cnidus : his geographical work on the Eed Sea. §4. His description

of tribes on the coast : the Ichthyophagi : the Chelonophagi. § 5. In-
land tribes : the Ehizophagi, &c., the Elephantomachi. § 6. Curious
information concerning their manners and customs: and the wild
animals. No notice of gold. § 7. The Troglodytes. Geographical
details of coast to the Straits. § 8. The Arabian coast. § 9. The
Sabjeans : their wealth and luxury. § 10. Their trade with India.
Inland trade by caravans across Arabia. § 11. Time occupied by
voyage from Palus Moeotis to Ethiopia .. .. .. 48

Section 3. — Artemidorus.

§ 1. Artemidorus: his date: his geographical work : comprised a full Periplus
of the Mediterranean and Euxine Seas. Imperfect information con-
cerning the western shores of Europe. His Periplus of the Eed Sea.
§ 2. Account of India and Taprobane. § 3. His statements of dis-
tances : and length of habitable world : compared with Eratosthenes.
§ 4. Another computation, by land routes. § 5. Breadth of the world.
§ 6. Eoute from Ephesus to the Euphrates. § 7. Metrodorus of Scepsis
and Alexander Polyhistor .. .. ,. ,. .. 61

Section 4. — Scymnus Chius.

§ 1. Geographical treatise in verse ascribed without foundation to Scym-
nus Chius — author really unknown : considerable fragments remain.
§ 2. Has very little value: a confused medley made up from
different sources. § 3. Description of the Euxine the best part. § 4.
Confused accounts of Western Europe. Accurate notices of Greek
colonies .. ,. .. ,. .. ,. ,. ., 69

CONTENTS OF VOL. II. V

Section 5. — Voyage of Eudoxus.

§ 1. The voyage of Eudoxus of Cyzicus a real voyage of discovery. Narrative
of it by Posidonius. His first voyage. § 2. His attempt to circum-
navigate Africa : its unsaccessful result. § 3. Truthful character of
the narrative, misrepresented by later writers .. .. Page 74

Section 6. — Roman Wars : the Jugurthine and Mithridatic Wars.

§ 1. Increased knowledge of Northern Africa from the Second Punic War.
§ 2. Operations against Jugurtha. § 3. Sertorius : the Fortunate
Islands. § 4. Extensive trade of Gades. Conquest of Spain by Pom-
pey. § 5. "War in the East against Mithridates and Tigranes. Power
and dominions of Mithridates. § 6. Campaigns of LucuUus. § 7. He
crosses the Euphrates and invades Annenia. § 8. Imperfect knowledge
of geography. § 9. Campaign of Pompey. § 10. Eetreat of Mithri-
dates to the Bosporus : his projects and death. § 11. Progress of
Roman arms in Thrace and neighbouring provinces. § 12. Relations
with Parthia. § 13. Expedition of Crassus ,. .. .. * 79

Section 7. — Posidonius.

§ 1. Posidonius : his historical writings : geographical liotices. His work on
the Ocean. § 2. His determination of the circumference of the earth.
§ 3. His estimate of length of the " habitable world." Suggestion of
circumnavigation from Spain to the Indies. § 4. Admits the possi-
bility of sailing round Africa. § 5. Clear notions on tides. Notices of
physical phenomena. § 6. Scattered notices cited from him by Strabo :
erroneous views. § 7. Astronomical writers probably belonging to this
period: Cleomedes. §8. Greminus .. .. .. .. 93

PAGE

S^OT]

E A.

Greek Monarchs of Bactria

102

Measurements of Mediterranean accord-

ing to Artemidorus

103

Tigranocerta

104

River Arsanias . .

105

?»

Gauls on the Danube

ih.

Zeugma .-.

107

CHAPTEE XIX.

c^sak's waks.

§ 1. Roman province in Gaul. Cimbri and Teutones. § 2. Cfesar determines
on conquest of Gaul.' The Helvetians: their migration and defeat.
§ 3. Campaign against Ariovistus in Alsace. § 4. War with the
Be]ga3. P. Crassus reduces the Armoricans. Revolt of the Alpine

VI CONTENTS OF VOL, II.

tribes. § 5. Eevolt of the Armoricans. Naval war with the Veneti.
Crassus subdues Aquitania. § 6. War witb the Germans. C^sar
crosses the Ehine. § 7. His first expedition to Britain, § 8. His
second expedition. § 9, Subsequent operations : he crosses the Khine
a second time. § 10. General revolt of the Gaulish nations, and final
subjection, § 11. Cassar's conquest of Gaul led to an accurate know-
ledge of the country. § 12. Tribes and towns : natural strongholds,
§ 13. Imperfect knowledge of Britain and Germany, § 14. Geogra-
phical account of Britain : no mention of Cassiterides. § 15, Imperfect
acquaintance with Germany. § 16, Ethnographical notices of Ger-
mans, § 17. Civil Wars. March of Cato from Cyrene to Utica.
§ 18. Wars in the East. Expedition of Antony against the Parthians.
Difficulty of following it in detail ., ,. ,. ., Page 109

PAGE

NOTE A. Belgian Tribes 135

B. Caesar's Passage from Gaul to Britain 136

C. Landing of Csesar in Britain ,. ., 137

D. Passage of the Thames .. ., 138

E. The Capital of Cassivellaunus .. .. 139

F. British Tribes ib.

CHAPTEE XX.

BOMAN EMPIRE TJNDEE AUGUSTUS.
Section 1, — Eoman Empire under Augustus.

§ 1. Eoman Empire completed by the annexation of Egypt. Eeview of the
provinces and dependent States, Spain, Gaul. § 2. Britain and
Germany still independent. § 3. Annexation of provinces south of
the Danube : Ehsetia, Vindelicla, Noricum, and Pannonia. § 4. Sub-
jugation of Alpine tribes in Italy. Cottian Alps independent. § 5. II-
lyricum and Dalmatia, § 6. Moesia. The Danube the northern fron-
tier, § 7. Nations north of the Danube : the Dacians, Bastarnte, and
Sarmatians, § 8. Sarmatian tribes from the Danube to the Bory-
sthenes. Greek cities. Kingdom of Bosporus. § 9. Macedonia and
Thrace, Province of Achaia, § 10, Asia, Eoman province of the
name, § 11, Bithynia, Pontus, Galatia, § 12. Cappadocia, Lycia, and
Pamphylia, Cilicia, Commagene, § 13, Syria : native local dynasties
preserved, § 14, Parthian monarchy, Atropatene, Armenia, Peace-
ful relations of Augustus with the Parthians, § 15. Establishment
of Parthian Empire unfavourable to extension of geography, ApoUo-
dorus of Artemita. § 16, Isidorus of Charax : his work on the Statlimi
Parthici. § 17. Eelations of Eome with Scythians and Indians. §18,

CONTENTS OF VOL. II.

Vague notions of tlie Seres and production of silk. § 19. Arabia still
wholly independent. Egypt. § 20. Cyrenaica — Koman province of
Africa. § 21. Numidia and Mauretania .. .. .. Page 141

Section 2. — Eoman Writers: Juba.

§ 1. Very few Eoman writers on geography. Varro Atacinus. Cornelius
Nepos : his tale of the Indian navigators. § 2. Sallust : his geography
of Africa. § 3. Statius Sebosus : the Fortunate Islands. § 4. Juba :
his knowledge of Africa. Strange theory of the Nile. § 5. Account
of the Fortunate Islands. § 6. Extension of Koman roads and itiner-
aries. § 7. M. Agrippa: his map. Supposed measurement of the
world by J. Cassar .. .. .. .. .. .. 171

Section 3. — Military Expeditions,

§ 1. Expedition of iEIius Gallus into Arabia. Narrative of it by Strabo.
§ 2. Geographical difficulties. Wealth of the Sabajans. § 3. "Expe-
dition of Petronius into Ethiopia. § 4. Of Cornelius Balbus against
the Garamantes .. .. .. .. .. -. 179

Section 4. — Wars in Germany.

§ 1. Increased knowledge of Germany after the time of Caesar. Drusus crosses
the Pihine: his first campaign in Germany. § 2. His second and third
cami)aigns. Advances to the Elbe. § 3. Campaigns of Tiberius. § 4.
Changes among the German nations. Migration of the Marcomanni
§ 5. Defeat of Quintilius Varus. § 6. Campaigns of Germanicus

186

SECTipN 5. — Diodorus.

§ 1, Diodorus Siculus : his historical work : a mere compilation. § 2, The
first five books. § 3. Account of islands in the Mediterranean : and of
Britain. § 4. Of the tin trade 194

PAGE

NOTE A. Alpine Tribes .. 199

B. Galatia 200

C. Tigers 201

D. Juba's Account of the Nile . . . . «&•

E. The Fortunate Islands 202

F. Expedition of jElius Gallus into Arabia 204

G. Aliso 206

H. Defeat of Varus 207

Vlll CONTENTS OF VOL. II.

CHAPTEE XXI.

STBABO.

Section 1. — General Views.

§ 1. Strabo : his date and life : notices found in his works. § 2. His travels :
his historical work : his geography written in advanced age. § 3. Rela-
tion of his geographical work to that of Eratosthenes. § 4. Historical
and mythological digressions. § 5. His neglect of Herodotus. Dis-
cards the statements of Pytheas. § 6. Neglect of Roman writers.
§ 7. His work intended for the general reader. Difference of character
from those of Pliny and Ptolemy. § 8. Great advance in physical
geography. § 9. Inferior in regard to mathematical geography to
Eratosthenes and Posidonius. His general views on this subject.
§ 10. Increase of materials for geography in his time. § 11. The first
two books. His views on Homeric geography. § 12. Review of Era-
tosthenes. Greological speculations. § 13. Discussion of length and
breadth of inhabited world. § 14. Defends Eratosthenes against Hip-
parchus: follows him generally in regard to Asia. § 15. Want of
observations of latitude and longitude. The Sphragides. § 16. Review
of Posidonius and Polybius. Division of the earth into zones. § 17.
Outline of Strabo's own views : assumes general notion of figure of the
earth, &c. § 18. Boundaries of habitable world : error as to breadth :
his map of the Mediterranean inferior to that of Eratosthenes. § 19.
Origin of erroneous conclusion. § 20. Northern and southern limits of
the world. § 21. Mode of constructing a map. § 22. G-eneral descrip-
tion of the MediteiTanean and the countries surrounding it, § 23. Er-
roneous ideas of figure and position of Spain, Gaul, and Britain. § 24.
Imperfect notions of that of Italy, Sicily, and the other islands. § 25.
General want of geographical accuracy. Vagueness as to distances.

Page 209
Section 2. — Descriptive Geography — Europe.

§ 1. Description of Spain. § 2. Its civilization and wealth. Its mines, § 3.
His account derived solely from Greek authorities. § 4. Islands adja-
cent to Spain. § 5. Geography of Gaul. § 6. Description of the
provinces. §7. Aquitania. Manners and customs of the Gauls. §8.
Britain : his knowledge derived almost entirely from Cjesar. lerne.
§ 9, His account of the Alps greatly in advance of Polybius. § 10.
Passes through them. § 11. Description of Italy and adjacent islands :
its imperfections. The Apennines well described, Campania and its
volcanic phenomena. Greek colonies. § 12. Sicily, ^tna. The
^ohan Islands, § 13. Sardinia and Corsica. § 14. North of Europe.
Germany. Shores of the Ocean, Very imperfect knowledge of all this
part of Europe. § 15. Defective information concerning European

CONTENTS OF VOL. II. ix

Scythia. Conquests of Mithridates. The Tauric Chersonese. § 16.
The Getfe or Dacians. Illyricum, Pannonia and Moesia. § 17. Mace-
donia and Thrace. § 18. Three books devoted to Greece and the
islands. Their unsatisfactory character. Chiefly occupied with Ho-
meric geography. § 19. Erroneous geographical idea of Greece. § 20.
Notices of physical geography. § 21. Islands of the ^gean. Page 239

PAGE

NOTE A. Age of Strabo 272

„ B. Volcanic Eruptions — ^Thera and Methone 274
„ C. Distances of Chorographer .. .. 275

CHAPTER XXII.

STRABO — (continued).
Section 1. — Asia.

§ 1. Strabo in general followed Eratosthenes in respect to Asia. Its division
into two parts by Mount Taurus. § 2. The Tanais and Palus Majotis.
Scythians and Sarmatians. § 3. Caucasian tribes on the east of the
Euxine. Albanians, Iberians and Colchians. § 4. Detailed descrip-
tion of Caucasus. § 5. The Caspian Sea : regarded as an inlet of the
Ocean. § 6. Nations east of the Caspian: the Daas, the Massageta?
and Sacas. § 7. Sogdiana, Bactriana, Aria, Margiana. The Seres.
§ 8. Termination of Asia to the east : its supposed extent. § 9. Ar-
menia and Media. Description of the chain of Mount Taurus. § 10.
Of the course of the Euphrates and Tigris. § 11. Armenia and Atro-
patene. Confusion respecting the lakes. The Greater Media. § 12.
Cappadocia : full desci-iption derived in part from personal knowledge.
§ 13. Pontus : his description one of the most valuable parts of his
work. Obscure mountain tribes in the interior. § 14. Imperfect
acquaintance with interior of Asia Minoi*. Galatia. § 15. Phrygia.
The provinces on the ^gean : Caria, Ionia, Lydia and the Troad. He
follows Demetrius of Scepsis in respect of the Troad. § 16. Lycia,
Pamphylia and Cilicia. § 17. Want of definite geographical data :
erroneous statements of distances. § 18. Line of route taken from
Artemidorus. Width of Asia Minor, § 19. His account of India :
taken from Megasthenes and earlier writers. No later sources of
knowledge. § 20. Imperfect geographical idea of the country : adopts
that of Eratosthenes without alteration. The rivers of India. § 21.
The great mountain chain to the north. No information as to the
peninsula of India or Taprobane. § 22. Ariana : his use of the term ;

a 3

X CONTENTS OF VOL. II.

countries comprised in it: follows Eratosthenes throughout. §23.
Persia proper : well described. § 24. Susiana : its rivers : difticulty
of determining them. § 25. Assyria and Babylonia. Mesopotamia.
§ 26. Syria : its description full and satisfactory : but contains some
strange errors. § 27. Arabia. Increased knowledge of the peninsula.
Expedition of Ji^lius Gallus .. .. .. .. .. Page 276

. Section 2. — Africa.

§ ] . Full and accurate description of Egypt. The voyage up the Nile. § 2.
Canal across the isthmus of Suez. § 3. Particulars of Indian trade.
Prosperity of Egypt, but decayed condition of Thebes and other towns,
§ 4. The Oases : the Upper Nile. § 5. General idea of form of Africa:
same as that of Eratosthenes. § 6. Mauretania. § 7. Mount Atlas.
The GtetuliaDs : tribes of the interior. § 8. Provinces on the Medi-
terranean very briefly described. § 9. The Cyrenaica. Silphium,
Native tribes of the interior. § 10. The work of Strabo little known
till long after his death. Its great celebrity among the Byzantines.

321

PAGE

NOTE A. TheChalybes 336

CHAPTEE XXIII,

STBABO TO FLINT.

Section 1. — Straho to Pliny.

§ 1. Extension of the Pioman Empire in Britain and Mauretania. § 2. Inva-
sion of Britain under Claudius. Conquest of southern part of the
island. § 3. Wars continued imder Nero : conquest extended to the
Tyne. § 4. No progress made in Germany. § 5. Expedition of a
Eoman knight in quest of amber. Little additional information ac-
quired. § 6. Wars in Armenia. § 7. Exploration of the Nile by two
Roman centurions sent by Nero. § 8. Mauretania. Expedition of
Suetonius Paulinus across Mount Atlas. § 9. Voyage of Hippalus to
India 338

Section 2. — Poviponius Mela,

§ 1. His work the only regular geographical treatise in Latin. Its date. § 2.
Peculiar arrangement. Division of the continents. The Antichthones.
§ 3. Popular character of his work. His statements often taken from
early writers. § 4. Want of critical judgement : fables and errors.
§ 5. His geography of Western Europe : in advance of the Greek

CONTENTS OF VOL. II. XI

writers. § 6. His account of Gaul and Britain. 'The Orcades. § 7.
Imperfect account of Germany. Sarmatians. § 8. Caspian Sea.
Eastern extremity of Asia. § 9. India. § 10. Western Asia. § 11.
Arabia and interior of Libya. § 12. External coast of Africa Page 352

PAGE

NOTE A. British Pearls 369

„ B. The Dumnonii . . . . . . . . ih.

„ G. The Age of Mela 370

CHAPTEE XXIV.

PLINY.

Section 1. — General Vieivs.

1. Pliny's Natural History : its encyclopasdic character. § 2. His life and
works : mode of composition. § 3. Defective character of geographical
portion of his work. § 4. Its political and statistical value. § 5.
Statements of distances : crude mode of dealing with them. § 6. His
second book. Meteorology. Ignorance of astronomy. § 7. Notices of
astronomical phenomena. § 8. Statement of proofs of the continuity
of the Ocean. Misrepresentation of his authorities. § 9. Notices of
earthquakes. Height of mountains. § 10. Measurement of inhabited
world : of circumference of the earth .. .. .. 371

Section 2. — Descriptive Geography — Europe.

1. Peculiar ari'angement of his descriptions of countries. § 2. Spain : im-
perfect geographical account. Correct notion of the Pyrenees. § 3.
Gaul : the old Eoman province well described : the rest very imperfectly.
§ 4. His account of Italy : its statistical value : but geographically
almost worthless. § 5. The Apennines : the Padus : the Tiber. § 6.
Confusion of historical geography. Circeii. § 7. Sardinia and Corsica.
Sicily. § 8. Eoman provinces south of the Danube. § 9. His descrip-
tion of Greece : its confused and useless character. § 10. The Euxine
and its European shores. Confused account of Scythian tribes. § 11.
Germany : its northern shores. Scandinavia. Nations of the interior.
§12. Meagre notices of the British Islands .. .. .. 387

Section 3. — Geography of Asia and Africa.

1. More interesting than his account of Europe. Contains additional in-
formation. Asia Minor. § 2. Syria. Palmyra. § 3. The Euphrates
and the Tigris. § 4. Armenia and the Caucasus. Vague account of
the Caspian and lands beyond. Margiana. § 5. Parthian Empire:

Xll CONTENTS OF VOL. II.

its divisions. Parthia Proper. § 6. Meagre account of Upper Asia :
and of Scythian tribes : the Seres. § 7. India : considerable advance
in the knowledge of it in the time of Pliny. The Ganges and its
tributaries. The Indus. General conformation of India. § 8. Trade
with India. Voyage from Arabia direct to India : and back. § 9.
Examination of his account : its authentic character. § 10. Tapro-
bane : additional information recently obtained concerning it. Trade
with the Seres. § 11. Description of Arabia : apparent fulness, but
defective in reality. § 12, Periplus of the coasts : enumeration of
tribes of the interior. Extent of Koman knowledge of the country.
§ 13. Erroneous estimate of its size. § 14. The fied Sea and coast
of Africa outside the Straits. § 15. Interior of Africa : Eoman explora-
tions. Confused account of Ethiopia. The southern ocean. § 16.
The western coast of Africa. Confused accounts from different sources.
§ 17. North of Africa. Mauretania. Mount Atlas. § 18. Mediterra-
nean coast. Vague notices of tribes of the interior. § 19. His men-
tion of the river Niger or Nigris. § 20. His account of the Nile and of
Egypt. § 21. Comparison of the size of the continents. His enumera-
tion of the cZ^m(Ito .. ,. .. ., .. .. Page 405

PAGE

NOTE A. Sources of the Tigris 439

„ B. Writers on Ethiopia . . . . . . 440

„ C. Pliny's Account of the CUmata . . 441

CHAPTER XXV.

PERIPLUS OF THE ERYTHE^AN SEA.

§ 1. Peculiar character of the document known by this name : its accuracy.
§ 2. Its date : may be assigned to the reign of Domitian. Not the
authority used by Ptolemy. § 3. Description of voyage down the
lied Sea : Auxuma. § 4. From Adulis to Cape Aromata. § 5. Trade
and productions. § 6. Cape Aromata, Taba), Opone. § 7. Coast from
Opone to Rhapta, § 8. Identification of localities : Rhapta the limit
of knowledge. § 9. Arabian coasts of the Eed Sea, from Leuce Come
to Muza. § 10. From Muza to Cane. § 11. From Cane to the
Zenobian Islands. § 12. The Persian Galf. § 13. Omana : Scythia.
§ 14. Mouths of the Indus : Gulfs of Eirinon and Barace. § 15.
Barygaza. § 16. Tidal phenomena. § 17. Configuration of coast :
the Dekkan. § 18. Coast of Ariace. Melizigara. § 19. Coast of
Limyrice. Muziris and Nelkynda. § 20. Imports and exports :
arrangements of trade. § 21. Account of navigation from Arabia
' direct -to the Indian ports. § 22. Nelkynda the limit of ordinary

CONTENTS OF VOL. II. XI 11

trade. Imperfect information beyond. Colchi and the pearl fishery.
§ 23. Taprobane. The mouth of the Ganges. Chryse. § 24. Trade
from the Ganges with Thinse , . . . . . . , Page 443

PAGE

NOTE A. Destruction of town of Arabia Felix .. 478
„ B. Minnagara .. .. ., ., ib.

CHAPTER XXVI.

PLINY TO PTOLEMY.

Section 1. — Dionysius Pariegetes.

§ 1. Poetical work of Dionysius : its unpretentious character. § 2. Its date :
probably belongs to the reign of Domitiaii : its want of poetical merit.
§ 3. His general views of geography. § 4. The great Indian pro-
montory : the Seres. § 5. The Scythian tribes : the Huns. § 6.
Imperfect knowledge of western nations. British Islands. Chryse and
Taprobane. § 7. India and Arabia : the Nile. § 8. Great reputation
enjoyed by his little work : paraphrases and translations .. 480

Section 2. — Tacitus.

§ 1. Agricola : his extension of Roman conquests in Britain. § 2. Subdues
the Caledonians and advances to north of the island. § 3. His Life
by Tacitus : ethnological and geographical notices in it. § 4. His
Germania : has little geographical value. § 5. Its great ethnograiihical
value. Division of the Germans into three tribes. § 6. Description
of the tribes in geographical order. The Chatti, Chauci, Cherusci,
Cimbri. § 7. The Suevi, Semnones and Langobardi. § 8. The
Hermunduri, Marcomanni and Quadi. Tribes on the northern Ocean.
§ 9. Islands in the Ocean : the Suiones and Sitones. § 10. Diversity
of different accounts of German tribes .. .. .. .. 490

Section 3. — Progress of Roman Arms : Extension of the Empire.

§ 1. The Eoman Empire at its height from Domitian to Trajan. Wars with
the Dacians. § 2. Trajan : his conquest of Dacia : reduces it to a
Eoman province. § 3. His wars in the East : crosses the Euphrates
and Tigris : navigates the Persian Gulf. § 4. Peaceful reign of
Hadrian : his administrative talents : makes the tour of the empire :
his visit to Britain : his wall. § 5, His travels: imperfect account of
them. § 6. Arrian : his Periplus of the Euxine : a report addressed
to the emperor. § 7. Contains a minute and detailed Periplus of the

XIV CONTENTS OP VOL. II.

south coast : with a brief account of the rest. § 8. Eeign of Antoninus
Pius : his wall in Britain. § 9. Extended knowledge of distant
countries. Eoman embassy to China .. .. .. Page 502

PAGE

NOTE A. Dacia .. 516

„ B. Travels of Hadrian .. .. .. 517

CHAPTEK XXVII.

MAKINUS TYRIUS.

§ 1. The work of Marinus known to us only through Ptolemy. Impossible
to judge of his merits. § 2. Solely engaged in collecting materials for
the map of the world. § 3. Determination of length and breadth of
inhabited world: great extension of both. § 4. Breadth. Koman
expeditions in the land of the Ethiopians : Agisymba, § 5. Enormous
exaggeration of result. § 6. Extension of coast of Africa to the south :
equally exaggerated. Cape Prasum. § 7. Thule the limit to the
north. § 8. Length of the world. Position of the Fortunate Islands.
Length of the Mediterranean. § 9. Increased trade with the Seres.
Itinerary from the Euphrates to Sera. § 10. Calculation of distances :
the result vastly in excess. § 11. Supposed confirmation by calculation
of extension by sea. § 12. Increased knowledge of lands beyond Cape
Comorin. The Gangetic Gulf: coast from thence to Cattigara, § 13.
I*eriplus of the coasts : unskilfully dealt with by Marinus. § 14. The
result utterly erroneous. § 15. Disbelief in connection of Oceans.
Probably the result of recent discoveries. § 16. Marinus adopts the
measurement of the earth by Posidonius : and reckons only 500 stadia
to a degree. § 17. Special importance of this eiTor from nature of his
work. § 18. Attempts to realise the idea of Hipparchus of a map
founded on astronomical observations. § 19. His map of the Mediter-
ranean generally adopted by Ptolemy. Defective arrangement of his
work. § 20. Imperfect mode of drawing maps. § 21. Impossible to
criticize his results in detail. His work wholly superseded by that of
Ptolemy 519

CONTENTS OF VOL. II. XV

CHAPTER XXVIII.

PTOLEMY.

Part 1. — Eis Oeographical System.

Ptolemy : Ms date and that of his work. § 2. His obligations to Marinus
of Tyre. § 3. More of an astronomer than a geographer. Took up
the idea of Hipparchus. § 4. Attempt to construct a map, based on
observations of latitude and longitude. Unable to carry out the idea,
but retained the form. § 5. His six books of tables : advantage of
this form. § 6. Calculated to mislead by a deceptive appearance of
accuracy. § 7. Erroneous estimate of their value in consequence.
§ 8. Ptolemy's ovra. account of his mode of proceeding. § 9. His cor-
rections of Marinus. § 10. His positions for the Mediterranean and
the Roman Empire devoid of all scientific authority. Paucity of
astronomical observations. § 11. Attempt to correct longitudes. § 12.
Examination of his map of the Mediterranean. Erroneous position of
Sardinia and Corsica : and of Carthage. § 13. Massilia and Byzantium :
Alexandria and Rhodes. § 14. His longitudes still more erroneous.
Source of this error. § 15. Reckons only 500 stadia to a degree.
Effect of this false graduation on his map. § 16. Erroneous position
of the Fortunate Islands : and therefore of his prime meridian. § 17.
Correction of his longitudes according to true graduation. Amount of
error remaining. § 18. Tendency to exaggerate distances. § 19. Con-
tinuation of map eastwards of the Mediterranean: vitiated by the
same causes. Corrections of Marinus, § 20. Estimates of distance to
Sera and the Sin«. Indefinite limits of the world. § 21. Breadth of
the world from Prasum to Thule. § 22. The eighth book : its peculiar
character. § 23, Its supposed scientific value. § 24. Its real purpose.
Not based upon real observations, § 25, Ptolemy's mode of con-
structing his maps. His theoretical skill, § 26. The maps appended
to his work : probably copied from the originals .. .. Page 546

CHAPTER XXIX.

PTOLEMY — (continued).

Part 2, — Detailed Geography.

§ 1. Extent of geographical information possessed by Ptolemy. Increased
knowledge of the British Islands. § 2, Description of Ireland, § 3.
Britain : his accurate knowledge of the southern portion. Strange
error with regard to position of Scotland, § 4. The neighbouring
islands misplaced: the Orcades. Thule. §5. Gaul and Spain: His

XVI CONTENTS OP VOL. II.

materials : errors in geographical application of them. § 6. Elvers in
Gaul. § 7. Germany : defective character of his map. § 8. Sarmatia :
the nations that inhabited it. § 9. Confusion as to rivers and moun-
tains : false idea of the Palus M^otis. § 10. Accurate notion of the
Caspian : acquainted with the Volga. § 11. Countries bordering on the
Mediterranean : inaccurate positions assigned on his map. § 12.
Improved map of Greece. § 13. Scythia : its division by Mt. Imaus.
§ 14. Vagueness of notions concerning Northern Asia : confusion of
names from different sources. § 15. Account of Serica. § 16. And of
the Sinai. Extended knowledge of south-east of Asia. § 17. Erroneous
view of position of India. § 18. Erroneous exaggeration of size of
Ceylon. § 19. Misconception of the countries east of the Gangetic
Gulf. § 20. Possessed valuable information: but his geograjDhical
arrangements altogether wrong. § 21. Confusion of whole subject :
impossible to identify Cattigara or Thinse. § 22. labadius. Supposed
continuous land connecting Asia with Africa. § 23. His account
of Arabia: improved periplus of the coast: tribes and towns of the
interior. § 24. Africa : his knowledge did not extend beyond that of
Marinus. Difficulties attending his account of the sources of the
Nile : its correctness proved by recent discoveries. § 26. His know-
ledge derived fi-om the east coast. Trade with Ehapta. § 27. Moun-
tains of the Moon. § 28. The Gir and Nigir. Ptolemy's account of
them. Its difficulties. § 29. Division of modern geographers upon
the subject. § 30. Connection of Ptolemy's two rivers with the Atlas
and Northern Africa. § 31. Peculiar conformation of this part of
Africa: its wadies and lakes. § 32. Erroneous position in latitude
assigned to the rivers. Its coincidence with that of the Quorra acci-
dental. § 33. No mention of the. great desert of Sahara. § 34.
Difficulty of identifying the Gir and Nigir with any known rivers.
§ 35. Ptolemy's account of the west coast of Africa. Pervading
errors with regard to latitudes. Probable identifications. § 37. Erro-
neous position of the Fortunate Islands : and of Cerne. § 38. Un-
trustworthy character of his geography of Africa. § 39. Deficiency of
materials concealed by the scientific form in which they are arranged.
§ 40. Ptolemy's want of conception of physical geography. Imperfect
indication of mountains and rivers. § 41. His erroneous longitude for
India contributed to the discovery of America .. .. Page 580

NOTE A.

Rate of marching in Africa . .

PAGE

637

„ B.

Ptolemy's Longitudes iu the Mediter-
ranean

638

„ C.

Latitude of Thule

639

„ D.

„ E.

Ptolemy's Map of Scotland
Oxiana Palus

640
641

„ F-

Ptolemy's Map of India

642

„ G.

labadius..

643

CONTENTS OF VOL. II. XVll

CHAPTEE XXX.

GEOGBAPHY AFTER PTOLEMY.

Section 1. — Historical Events.

§ 1. Decline of geographical science after Ptolemy. § 2. Few events that
had any bearing on its extension. Wars of M. Aurelius. Severus in
Britain. § 3. Wars of the Eoruans in the East. Fall of the Parthian
monarchy. § 4. Wars with the Persians. Zenobia. § Expedition of
Julian : his death. § 6. Northern frontier of the Eoman Empire.
Dacia abandoned : and the Agri Decumates in Germany. § 7. Con-
quests of Theodosius in Britain .. ., .. .. Page 645

Section 2. — Oreek Writers.

§ 1. Pausanias : his Description of Greece : its object archgeological, not geo-
graphical. § 2. His digressions : account of the Ethiopians and Mount
Atlas. § 3. Notice of the Seres and production of silk. § 4, Other
Greek writers after Ptolemy very poor and meagre. § 5. Marcianus
of Heraclea. His Periplus of the Outer Sea : adds nothing to the
information of Ptolemy. § 6. Exaggerates still more the size of
Taprobane. § 7. His account of the western lands and of Britain.
§ 8. His epitome of Menippus. § 9. Anonymous Periplus of the
Euxine. § 10. The Stadiasmns of the Great Sea. Its peculiar charac-
ter and value. §11. Agathemerus : his treatise on geography. §12.
Stephanus of Byzantium : his Geographical Dictionary : its purpose
in reality grammatical, not geographical : the extant work a mere
abridgement .. .. .. .. .. .. .. 655

PAGE

NOTE A. Province of Valentia in Britain . . 672

„ B. The Stadiasmus of the Great Sea . . ib.

CHAPTER XXXI.

GEOGRAPHY AFTER PTOLEMY — (continued).

Roman Writers.

1. Miserable character of the literature of this period. Solinus : his work
taken almost wholly from Pliny. § 2. His notice of the British
Islands : and of the Seres. Popularity of his work in the middle ages.
Use of the word " Mediterranean." § 3. Ammianus Marcellinus : his
historical work : its merits. His geographical episodes. § 4. Their
defective execution. Account of nations adjoining the Pioman Empire.

XVlll CONTENTS OF VOL. II.

The Huns and the Alani : the Saracens. § 5. Avicnas : his poetical
Description of the World : a translation of Dionysius. § 6. Another
translation of the same work by Priscianus. § 7. Th« Ora Maritima
of Avienus : a clumsy compilation : his authorities. § 8. His account
of the voyage of Himilco, and of the tin islands. § 9. Ausonius : his
poem of the Mosella: its geographical notices. His Oi'do Nobilium
Urbium. § 10. Rutilius : his poetical account of his voyage. § 11.
Orosius : summary of geography prefixed to his history. § 12. Julius
iEthicus and Julius Honorius. § 13. Itineraries. Itinerary of An-
toninus. § 14. The Jerusalem Itinerary. § 15. The Tabula Peutin-
geriana. § 16. The Notitia Dignitatum : its account of the Eoman
Wall in Britain. § 17. The anonymous Geographer of Eavenna.
§ 12. Dicuil : his account of Thule .. .. .. Page 675

PAGE

NOTE A. Voyage of Himilco .. .. ... 703

„ B. The Cosmography of jEthicus . . . . ib.

„ C. Measurement of Roman Empire . . 706

LIST OF MAPS IN VOL. II.

1. Map to illustrate Cesar's Wars

2. Map to illustrate the Wars of the Eomans in

Germany

3. The World according to Strabo..

4. The World according to Pomponius Mela

5. The Eoman Empire ..

6. Map to illustrate the Periplus of the Erythb^an

Sea .. ..

7. The World according to Dionysius Periegetes

8. Map of the World according to Ptolemy

9. Map of the British Isles according to Ptolemy
10. Map of North- Western Africa according to Ptolemy

PAGE

face

130

194

238

368

442

476

490

578

584

632

HISTORY

ANCIENT GEOGRAPHY.

CHAPTEE XVII.

GEOGKAPHEKS AFTEB ERATOSTHENES.

Section 1. — Hip^parchus.

§ 1. Gee AT as were undoubtedly the merits of Eratosthenes as
a geographer, when we compare the state of the science in his
hands with what it had previously been, they were far from
being generally acknowledged in his own day. The philo-
sophers and writers of the Alexandrian school appear indeed
to have been pre-eminently distinguished by that spirit of
jealousy and disposition to find fault with their contemporaries
and predecessors which has too often characterised men of
letters and learning in all ages. Accordingly we find that
Eratosthenes was assailed with severe criticisms, at the same
time that even his adversaries were obliged to adopt many of
his conclusions. Polemon especially, a Stoic philosopher of
considerable reputation, who flourished about the beginning of
the second century B.c.,^ and who devoted much attention to
the detailed examination of Greece itself, and its most cele-
brated localities, attacked Eratosthenes in the most vehement
manner, and even ventured to assert that he had never visited
Athens, — a paradox justly treated with contempt by Strabo.^

' According to Snidas (s. v.), he was He was therefore about a generation

a contemporary of Aristophanes of younger than Eratosthenes.
Byzantium, and flourished in the reign ^ Strabo, i. p. 15. .

of Ptolemy Epiphanes, B.C. 205-181.

VOL. II. B

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVII.

Of his other criticisms we have no account, but as Polemon
appears to have occupied himself with researches of an anti-
quarian and topographical character,^ rather than with ques-
tions of general geography, it is probable that they would
have had little bearing on our immediate subject.

§ 2. A far more formidable adversary presented himself
in the great astronomer Hipparchus, who though he did
not write any geographical work himself, devoted a whole
treatise to the criticism of that of Eratosthenes, and the
refutation of some of his conclusions.* Unfortunately this
treatise is lost to us, in common with the one against which
it was directed, and we know the views of Hipparchus,
as well as those of Eratosthenes, only by the very imperfect
report of them in Strabo, who was himself little versed in
astronomy, and hence particularly ill qualified to judge in
such a controversy. The general character of his criticisms
is however not difficult to discern. Hipparchus regarded
geography from an astronomer's point of view, and was
strongly impressed with the necessity of basing all geogra-

^ Polemon appears to have travelled
about Greece very much in the same
spirit that Pausanias did, nearly four
centuries later ; but he did not com-
pose any general work on the subject,
and contented himself with a number
of small treatises on special subjects,
such as one on the offerings in the
Acropolis of Athens, another on those
of the treasuries at Delphi, &c. Several
of these are cited by Athenaeus and
other authors. His treatise against
Eratosthenes is cited by more than
one grammarian, and appears to have
borne the special title Trepl ttjs 'A^??-
vljcnv 'Eparoffdevovs eindr)fiias (Schol.
ad Aristoph. Av. s. II). From the
character of his researches Polemon de-
rived the title of 6 irepiiiyltTris (applied
in a very different sense from that in
which it is commonly given to Diony-
sius), and from his diligence in col-
lecting inscriptions from sepulchral
and other monuments he was nick-
named o ffr-r\\oK6tTas (Athenaeus, vi. p.

234, d). He appears, indeed, to have
been the first instance of what we
should call in modern days an anti-
quarian traveller. Concerning his
works, see Clinton, F. H. vol. iii. p.
524. The extant fragments are pub-
lished by G. Miiller, in his Fragmenta
Historicormn Grxcorum, vol. iii. pp.
108-148.

■* His work is referred to by Strabo
as TO. irphs rhv ^EpaToffdevri (i. J). 7), but
whetiier this was its precise title does
not appear. He however tells us dis-
tinctly that it did not profess to be a
treatise on geography, but merely an
examination of that of Eratosthenes
(^'linrdpxv M^'' oSv /xr] yecDypacpovi'Ti aW'
i^era^ovTi to, Xexdevra iy rfj yeci>ypa(pia
Tjj ^Eparocrdevovs, ii. 1, § 41, p. 93).
Hipparchus flourished from about 162
to 125 B.C. (See Clinton, F. H. vol. iii.
p. 532.) His birth is therefore placed
conjecturally at about 190 B.C. He
was a native of Nictea in Bithynia.

Sect. 1.

HIPPARCHUS.

phical science upon astronomical observations :^ a truth which
Eratosthenes indeed had been the first to point out, though, as
we have seen, he failed in carrying it into effect from the want
of such observations. Hipparchus was very little better pro-
vided in this respect : hence his means of really correcting the
conclusions of his predecessor were very small, while he appears
to have been led, either by a love of controversy or from
attaching undue weight to authorities of little value, to dispute
many of the views of Eratosthenes which were in reality well
founded.

§ 3. Hipparchus indeed appears to have clearly conceived
the idea, which was afterwards adopted by Ptolemy, of a map
of the earth's surface, or rather of the habitable portion of the
earth, according to the views then generally entertained, in
which every important point should be distinctly laid down
according to its latitude and longitude, determined by astro-
nomical observations.^ But even in the time of Ptolemy, as
we shall hereafter see, the construction of such a map, how-
ever theoretically desirable, was in practice wholly impossible,
and the great geographer was obliged to content himself with
positions calculated from itineraries and other such materials
as he could command.^ Still more was this the case in the
time of Hipparchus, and his attempt to rectify the map of
Eratosthenes, though more correct in a few points, was
disfigured on the other hand by still graver errors.

§ 4. On most of the fundamental points indeed he was con-
tented to adopt the conclusions of his predecessor. Thus he
accepted the calculation of Eratosthenes for the measurement
of the earth's circumference,^ and consequently regarded every

^ Strabo, i, p. 7. ^ Strabo, I. c.

' Ptolem. Geogr. i. c. 2.

' This is distinctly stated by Strabo
(ii. p. 113), and repeated again in
another passage (ii. p. 132), where it is
moreover confirmed by the addition
that every 360th part will therefore
contain 700 stadia. Pliny on the con-
trary tells us (H. N. ii. c. 108, § 247)
that Hipparchus added somewhat less

than 25,000 stadia (stadiorum paullo
minus xxv millia) to the measurement
of Eratosthenes : a singularly vague
statement, which we have no means of
explaining, but there seems no doubt
that it is founded on some miscon-
ception. Strabo's testimony is too
explicit to be set aside, and he un-
doubtedly wrote with the work of
Hipparchus before him.

B 2

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVII.

one of the 360 parts, or degrees, into which he divided this
great circle, as comprising 700 stadia.^ He adopted moreover
his notion of a principal parallel of latitude extending from
the Strait of the Columns to the Gulf of Issus, and passing
through the island of Khodes. But instead of placing the
Sicilian Strait on the same parallel, he justly described the
line in question as passing somewhat to the south of Syracuse •}
an important rectification, which however was unfortunately
not received by succeeding geographers. In like manner he
took the meridian passing through Alexandria as a kind of
principal meridian, from which longitudes were calculated
towards the east and west. This line he supposed, in common
with Eratosthenes, to pass through Meroe, Alexandria, Rhodes,
Alexandria Troas, and Byzantium, as well as the mouth of the
Borysthenes. It was doubtless from its passing through so
many known points that it was taken by Hipparchus as the
basis of his calculation for the division of the known or habit-
able world into climata or zones of latitude,^

§ 5. This division was undoubtedly the most important con-
tribution of Hipparchus to scientific geography. Unfortunately
it is known to us only through the abstract given by Strabo,
who himself tells us that he was contented with a rough abridge-

' Hipparchus was apparently tlie
first to divide the circle into 360°.
Eratosthenes, as we have seen, did not
carry the division further than into
sixty parts.

> Strabo, ii. p. 134. The words ra
'ZvpaKOvcriuv voriwrepa rerpaKoaiois
(TTaSiois are indeed somewhat vague,
but they probably mean to refer to the
city, rather than the territory. Both
Groskurd and the Latin translator
render them as if the reading were
^vpaKovffuiv.

- Hipparchus appears to have been
the first who applied the term KXifxara
in this sense, wldch was subsequently
adopted by Ptolemy and later geo-
graphers. Eratosthenes had, as we
have seen, anticipated him in drawing
parallels of latitude thiough a certain
number of points upon his chief meri-

dian, and determining the regions
through which they would pass. But
these lines were drawn at irregular
intervals. Hipparchus introduced the
important modification of fixing these
intervals with reference to astronomical
phenomena, especially to the length of
the solstitial day. The manner in
which his statements are reported by
Strabo would at first seem to leave it
doubtful whether he applied the term
of climata to the circles themselves, or
to the spaces bounded by them ; but as
the latter use of the word was that
generally adopted in subsequent times,
it is probable that it originated with
Hipparchus. Strabo however certainly
desi.-ribes the parallels or circles that
formed the limits of each cliina, not
the spaces comprised between them.

Sect. 1.

HIPPAECHUS.

ment, as sufficient for tlie purposes of the geographer.^ It
appears indeed that Hipparchus had calculated the celestial
appearances and the changes they underwent for every degree
of latitude/ proceeding north along the meridian of Alexandria
from the equator to the pole. This was of course theoretically
possible for an astronomer, even in his day, but he certainly
possessed no materials for connecting these results with geo-
graphy, even in regard to the portions of the earth then known,
and it is not probable that he attempted to do so. But he
regarded the whole habitable world as divided by eleven
parallels of latitude (that is, lines parallel with the equator)
for each of which he indicated the length of the longest day —
the simplest and most obvious mode of determining the latitude,
though of course giving but a rough approximation — together
with certain other celestial appearances such as were easily
observed.^ He added at the same time the names of the regions
and places, which, according to his calculation, lay under
these parallels, and the distances from the one to the other.
We must here briefly enumerate them as recorded to us by
Strabo,^ though there is unfortunately great reason to suppose
that they are very imperfectly reported by that geographer.

§ 6. The first parallel passed through the Cinnamon Eegion,
and this was regarded by Hipparchus, as it had been by
Eratosthenes, as the southern limit of the habitable world. It
was placed by him 8800 stadia from the equator,^ and was

3 Strabo, ii. 5, p. 132.

* It may be as well to mention, in
order to avoid misconception, tliat
though 1 here and elsewhere use the
terms " latitude " and " longitude " in
tlie sense familiar to all modern readers,
they were not employed in this tech-
nical sense either by Hipparchu!^, or
by any subsequent geographer until
the time of Ptolemy. At least Ptolemy
is the first extant writer in wliom they
are found. They were certainly un-
known to Strabo.

^ Thus, for example, the inhabitants
of the Eegion of Cinnamon were tlie
first for whom the Little Bear was con-

stantly within the arctic circle; that
is, never set ; at Syene the greater
part of the Great Bear was in the same
condition ; to the north of Byzantium
Cassiopeia fell wholly within the
arctic circle, &c. These notices are
very imperfectly given by Strabo, and
have been omitted in the summary
given in the text.

6 Lib. ii. c. 5, §§ 35-43.

' This is distinctly stated by Strabo
(ii. 5, p. 132), who himself adopts this
measurement. Eratohthenes, as we
have seen, made the interval only 8300
stadia (Chapter XVL p. 639, and Note
C), but he reckoned the parallel through

HISTORY OP ANCIENT GEOGRAPHY.

Chap. XVH.

situated, according to his computation, midway between the
equator and the tropic. Towards the west it passed through
the southernmost portions of Libya, and towards the east
through the southern extremity of Taprobane, or even a little
to the south of it.^

The next parallel lay through Meroe — a point which
assumed a special importance in the eyes of all the Alex-
andrian geographers — and Ptolemais Epitheras on the coast
of the Troglodytes. It was distant 3000 stadia from the
preceding. Along this line the longest solstitial day was of
thirteen hours. The same line prolonged to the east passed
through the southern extremity of India.

The third line was drawn through Syene, and was considered
as coinciding with the summer or northern tropic. It passed
about 5000 stadia to the south of Cyrene ; and towards the
east traversed the land of the Ichthyophagi on the coast of
Gedrosia, and was continued through India. For all places
along this line the sun was vertical at the summer solstice, and
the longest day was of thirteen hours and a half. It was
distant 5000 stadia from the parallel of Meroe.

The fourth parallel was drawn (apparently for convenience'
sake) through Alexandria and Cyrene, at a distance of 5000
stadia from the preceding, though the line which had a sol-
stitial day of fourteen hours lay about 400 stadia farther south.^
Alexandria and. Cyrene were thus assumed to be on the same
parallel, though they really differ by a degree and a half of
latitude, while the same circle was supposed to pass only 900
stadia to the south of Carthage — which really lies more than

the Land of Cinnamon as H400 stadia
to the south of Meroe, while Hippar-
chu3 made tlie difference only 3000

« Strabo, ii. 5, § 35, pp. 132, 133.
This expression certainly seems to be
at variance with the statement of Pom-
ponius Mela that Hipparchus doubted
whether Taprobane was an island of
vast size, or the beginning of a new
world (prima pars orbis alterius, Mela,

ii. § 70). But the authority of Mela is
worth very little, and the silence of
Strabo on so bold an hj^pothesis seems
conclusive against it. Hipparchus in
this instance doubtless followed Era-
tosthenes, who had already (as we have
seen) assumed it as one of the cardinal
points in his map of the world that
this first parallel of latitude passed
through Taj^robane.

» Strabo, ii. 5, § 38, p. 133.

Sect. 1.

HIPPARCHUS.

five degrees and a half to the north of Alexandria — and was
thence produced through the middle of Maurusia (Mauretania)
to the Western Ocean. Such grave errors with regard to the
position of places which might be naturally supposed to be
well known, show how far Hipparchus was from being able to
attain in practice that accuracy on which he laid so much stress
in theory. The same line produced to the east passed through
Lower Egypt, a part of Syria/ Babylonia/ Susiana, Persia,
Carmania, and the interior of Gedrosia to India.

He next mentioned, as it were in passing, that for Ptolemais
in Phoenicia, Sidon and Tyre^ the longest day was fourteen
hours and a quarter. This circle was about 1600 stadia north
of Alexandria, and 700 north of Carthage.

§ 7. The next parallel, which corresponded to a solstitial
day of fourteen hours and a, half, and was distant 3640 stadia
from Alexandria,* passed through the middle of the island of
Ehodes, and just to the south of Xanthus in Lycia, as well as
through the southern extremity of the Peloponnese, and 400
stadia to the south of Syracuse. This was the same parallel
which, according to Eratosthenes, if prolonged eastward, passed
through Caria, Lycaonia, Cataonia, Media, the Caspian Gates
and the foot of the Indian Caucasus.

Hipparchus does not appear to have noticed in treating of
these climata the position of Athens, but we learn from other

* Our editions of Strabo (ii. 5, § 38)
have KoiAijs Svpi'as Koi rrjs Svcd 'Svplas ;
but these words are certainly corrupt.
See Groskurd's and Kramer's notes.
It is however impossible to restore the
true reading.

' There is little doubt that we should
here read BafivXwvias for Ba^vAajvos, as
we know that Hipparchus himself
placed Babylon at 2500 stadia farther
north than Pelusium, which was on
the same parallel with Alexandria
(Strabo, ii. p. 88). Hence it could
only be the southernmost portions of
the province that could be in the same
latitude with Alexandria (see Gros-
kurd's note on Strabo, ii. p. 134).

^ The mention of these places suffi-

ciently shows how little Hipparchus
himself was pretending to scientific
accuracy. For he could not have been
ignorant that Ptolemais, Tyre and
Sidon followed one another at con-
siderable intervals from south to north,
the last being in fact more than forty
minutes (400 stadia) to the north of
the first.

* This statement again differs from
Eratosthenes, who had calculated the
difference in latitude between Alex-
andria and Khodes at 3750 stadia (see
Chapter XVI. p. 639) ; but this m all
probability referred to the eity of
Ehodes, while Hipparchus especially
specifics that the line passed through
the middle of the island.

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XVII.

passages^ that he placed that city 37 degrees north of the
equator ; just about a degree farther south than its true
position. So little did he possess trustworthy observations
even for the best known localities.

His next parallel was drawn through Alexandria Troas (near
the entrance of the Hellespont), Amphipolis on the Thracian
coast, ApoUonia in Epirus, and across Italy, passing to the
south of Eome and north of Naples. The longest day was of
fifteen hours. As this line derives special interest from its
connecting G-reece with Italy, it is worth while to point out
that while Alexandria Troas is situated about 15 G. miles
south of the parallel of 40°, Amphipolis and ApoUonia lay more
than 40 miles north of the same parallel, which again instead
of passing between Kome and Naples, runs about 50 miles
south of the latter city. Thus his positions as compared with
one another deviated from the truth by more than a degree.^
He placed this parallel about 7000 stadia from that of Alex-
andria, or 28,800 from the equator.''

The next parallel was that through Byzantium and Nicaea —
the latter place being obviously mentioned in consequence of
its being the birthplace of Hipparchus himself. He had more-
over himself made an observation of the gnomon at Byzantium,
from whence he concluded the latitude of that place to be the
same with that of Massilia, as determined by Pytheas : an
unaccountable error, which had the effect of distorting his
map of all the surrounding regions. Yet this erroneous con-
clusion was unfortunately followed by all succeeding geo-
graphers to the time of Strabo.^

* This is repeatedly stated in his
commentary on the Phenomena of
Aratus.

" The actual line on which the sol-
stitial day is fifteen liours corresponds
to 41° 21'. All the points indicated
were therefore materially too far to the
south.

' This corresponds very nearly with
the truth; as 28,800 stadia, at 700
stadia to the degree, would give 41°

10'. As might be expected, Hipparchus
was correct in his astronomy ; it was
his geography that was deficient.

* So unaccountable indeed does this
error appear to M. Gossellin that he
endeavours to prove that it was due to
Pytheas, and that Hipparchus only
adopted his observations in botli cases.
But the testimony of Strabo is precise
that Hipparchiis "found the same re-
lation of the gnomon to its shadow at

Sect. 1.

HIPPARCHUS,

The parallel of Byzantium was placed by Hipparclius at a
distance of 1500 stadia from that of Alexandria Troas : the
longest day was of fifteen hours and a quarter. It was not till
one had sailed 1400 stadia farther north that a point was
reached where the longest day was fifteen hours and a half.
This line, which was apparently an arbitrary one, not marked
by any place of sufficient importance to be noticed, was re-
garded by Hipparchus as just midway between the equator
and the pole. It therefore corresponded according to his
calculation with 45° of north latitude.^ This coincides very
nearly with the sum of his measures in stadia, which give
31,700 stadia from the equator, while 45° would give 31,500,
if we reckon, as Hipparchus undoubtedly did throughout this
calculation, 700 stades to a degree.^

§ 8. The next parallel, which was distant 3800 stadia from
Byzantium, passed through the regions at the mouth of the
Borysthenes, and the southern portion of the Palus Mseotis.
Here the solstitial day was sixteen hours in length, but during

Byzantium that Pytheas had done at
Massilia " (Strab. i. p. 63 (hy yap \6-
you etprjKe [Tlvdeas] rov iv MacrffaAia
yvMfjLOvos TTphs Trjv (TKidv, rhv ahrhv koI
"linrapxos Kara rhv 6ix(iivvfji.ov Kaiphv
evpe7y eV Tq3 Bv^avTiqj <p7i(ni/\ ii. p. 115),
and the very unusual exactness with
which that proportion is stated (that it
was in the ratio of 120 to 42 minus
one-fifth) points clearly to personal
observation.

The proximity of Byzantium to
Nicsea, the birthplace of Hipparchus,
also explains why he should have taken
special pains to make observations at
the former city.

" The parallel which really gives a
day of precisely 15 J liours is 45° 39',
again showing the near approach to
accuracy of these calculations of Hip-
parchus. It was in the application of
these mathematical inferences to prac-
tical geography that he failed, from
the want of correct observations.

' There is, however, a difficulty with
regard to the numbers of Hipparchus,
which we have no means of explaining ;

as it is evident that the sum of the
several distances he has given ought
to correspond exactly with the measure-
ment of half the distance from the
equator to the pole. Gossellin has
introduced various arbitrary changes in
the numbers given by Strabo, but some
of these, as shown by Ukert, are not
only unnecessary, but erroneous, and
such a mode of solving a difficulty is
at best very hazardous. It is clear
that there is an error somewhere, but
where, we are unable to decide.

We are distinctly told by Strabo,
with reference to this very part of his
work, that Hipparchus reckoned 700
stadia to a degree (ii.'p. 132). But we
must not suppose that he was therefore
employing a different stadium from
that in common use. His error, in
common with Eratosthenes (whom he
followed), was that of over-estimating
the quadrant of a great circle, which,
he reckoned at G3,000 stadia instead
of 54,000, its true value in round
numbers.

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVII.

the middle of the summer the twilight lasted almost all
through the night, the sun being only a short distance below
the horizon.^ According to the calculation of Hipparchus this
latitude coincided with that of the northern parts of Gaul ; ^
which he thus placed nearly 5^° to the north of Massilia, or
just about the latitude of Paris.

His next parallel was drawn 6300 stadia to the north of
Byzantium, and passed to the north of the Palus Mseotis, while
to the west it lay still through the northernmost parts of Gaul.
Here the solstitial day was of seventeen hours' duration, and
the phenomena of the twilight nights were still more remark-
able. At the winter solstice in the same latitudes the sun did
not rise more than 6 cubits, or 12^, above the horizon.

Beyond this again he pointed out that at a distance of 9100
stadia from the parallel of Massilia, the sun would rise only
4 cubits, and the longest day be of eighteen hours : and beyond
that again would come a circle where the longest day was of
nineteen hours, and the sun in winter would rise only 3 cubits.
Both these circles he appears to have considered as passing
through different parts of Britain : but this part of his system
is very imperfectly known to us, Strabo, by whom alone it is
reported, having considered it as of little importance to geo-
graphy, as these extreme northern regions were (in his opinion)
unknown, and uninhabitable from cold.*

It appears certain, however, that Hipparchus, in common
with Eratosthenes, adopted the leading statements of Pytheas,
and admitted the existence of an island named Thule, where
the solstitial day was twenty-four hours long.^ As an astro-

2 Strabo, ii. p. 13.5. Of course this
statement, thougli partly true of the
supposed parallel of latitude, is wholly-
false as applied to the northern shores
of the Black Sea, which are really in
just about the same latitude as Geneva.
As applied even to the latitude of Paris,
which is situated in lat. 48° 50', very
nearly on the line of sixteen hours'
day, it is considerably exaggerated.

3. 11. 1, pp. 72, 75.

^ Strabo, ii. p. 135. He here stops
with the circle of G300 stadia nortli of
Byzantium, but in another passage (ii.
p. 75) he cites some furtlier observations
of Hipparchus, from which the account
given in the text is derived.

^ This is indeed not distinctly stated
by Strabo ; but as that author through-
out censures Hipparclnts, in common
with Eratosthenes, for altacliiiig cre-
dence to the fables of Pytheas, and for

Sect. 1.

HIPPAECHUS.

nomer Hipparchus would know, that in proceeding north, this
phenomenon would really occur on the Arctic Circle, and hence
probably he more readily admitted the statement that it had
been actually observed.

§ 9. It must be admitted that, notwithstanding many grave
errors, Hipparchus had really made considerable progress
towards laying down a correct map of the countries bordering
on the Mediterranean, and had introduced some important
corrections into that of Eratosthenes. But the case was alto-
gether otherwise with regard to Asia. Here we have seen that
the principal parallel of latitude, assumed by Eratosthenes to
be continued from the Gulf of Issus through Thapsacus, the
Caspian Gates, and the foot of the Indian Caucasus to the
Indian Ocean, was really a very fair approximation to the
truth, as far at least as the Indian frontier. This parallel
coincided generally with the southern foot of the great range
that under the names of Taurus, Paropamisus, and Imaus, was
considered by Eratosthenes as traversing Asia from west to
east. Hipparchus, however, rejected this view — on what
grounds we do not know — and carried up almost all the points
on this line to latitudes far exceeding the truth, placing even
Thapsacus not less than 4800 stadia to the north of Babylon,
or more than three degrees and a half beyond its true latitude,
and the Caspian Gates nearly on the same parallel.^ But
from this point he conceived the great central chain of Taurus
— the existence of which, as a kind of backbone of Asia, he
did not dispute — to have a direction to the north-east, so as to
remove Hyrcania, Margiana, and Bactria successively farther
and farther to the north. So far indeed did he carry this dis-
placement, as to remove Bactria proper (the environs of Balkh)
which is really in the same latitude with the southernmost
part of the Morea, to a level with the northern portions of
Britain.''

following him in the high latitudes,
•which he assigned to the northern
parts of Britain, there can be little
doubt that he accepted his statement
with regard to Thule. Had it been

otherwise Strabo could hardly have
failed to mention the confii-mation of
his own doubts by so high an authority.

« Strabo, ii. pp. 78, 81.

' Id. pp. 71-75. According to Strabo's

HISTOKY OF ANCIENT GEOGKAPHY.

Chap. XVII.

So far as we can discern from the account given by Strabo
(which is by no means clear) this astounding error arose prin-
cipally from his having adopted the erroneous and greatly
exaggerated estimates of the dimensions of India, given by
some earlier writers, especially Daimachus. As Hipparchus
had correctly fixed the extent of India towards the south, in
accordance with the views of Eratosthenes,® he was unable to
gain space for the enormous length which he assigned to it
(from north to south) without removing the mountain barrier
of the Hindoo Koosh (which all admitted to be its northern
boundary) much farther to the north than its true position.
Hence the countries to the north of this, Bactria, Sogdiana, &c.,
were in like manner transported into the far regions of Northern
Asia, a supposition which, as Strabo observes, is sufficiently
negatived by the fact of the great fertility and productiveness
of those provinces.^

Another error into which Hipparchus fell was probably con-
nected with the preceding. He maintained that the river
Indus had its course towards the south-east, instead of flowing
from north to south, as had been the received view of all
geographers from the time of Alexander.^ Of his views with

own geography, wliicli brought down
Britain far below its true latitude,
the parallel of Bactria (as assumed by
Hipparchus) would pass altogether to
the north of that island, and even of
lerne (Ireland), which he supposed to
lie considerably farther north.

* The comparatively correct position
assigned to the south of India and
Taprobane by Eratosthenes and Hip-
parchus was doubtless based upon the
report that those regions lay within
the tropic ; an observation erroneously
applied by others to the parts of India
visited by Nearchus and Onesicritus,
but whicli was of course perfectly true
with regard to the more southern por-
tions of the peninsula.

n Strabo, ii. pp. 73, 75. Strabo
indeed greatly exaggerates the degree
of cold that must belong to the parallel
in question, which, as he observes,
would fall far to the north of Ireland,

which is itself barely inhabitable ( ! ),
but the luxuriant growth of vines in
Hyrcania and Bactria, on which he
lays especial stress, is undoubtedly a
sure criterion of a more southerly
climate. The great fertility of the
environs of Balkh and Merv (Bactria
and Margiana) is attested by all modern
travellers.

Throughout this discussion it may
be observed that Strabo argues solely
from the probabilities of tLe case, such
as climate and natural productions.
It is evident that he had no astro-
nomical observations to appeal to,
which, however rude and imperfect,
would have been decisive of the ques-
tion.

'Strabo, ii. p. 87. In this instance,
as in several others, Hipparchus, out
of opposition to Eratosthenes, returned
to the views embodied in what Strabo
calls " the old maps " (/cafiaTrep eV to7s

Sect. 1.

HIPPAECHUS.

regard to the river Ganges we have no distinct statement, but
he doubtless -considered it as falling into the Eastern Ocean,
and apparently as having a course about parallel with that of
the Indus.

§ 10. This fundamental misconception as to the direction of
the great mountain chain, led him also to bring up the coasts
of Carmania and Persia, and the Persian Gulf, far above their
true position in latitude, and thus distorted his whole map of
Asia. But be'sides this great error, his minor criticisms of
Eratosthenes, with regard to the distances and relative positions
of many points on his map, specimens of which have been
preserved to us by Strabo, certainly seem to warrant the obser-
vation of that author, that they were dictated by a captious
disposition to demand an amount of accuracy that belonged
rather to the geometer than to the geographer.^ Hipparchus
indeed was justified in demanding the most perfect accuracy
of which the subject would admit, and his theoretical concep-
tion of what geography ought to be was in advance of all his
predecessors : but he overlooked the imperfect nature of the
means at his command, which gave him in reality very little
power of rectifying their conclusions.

In several instances indeed he criticized the arguments of
Eratosthenes, and rejected his conclusions in order to return
to the views of earlier writers, where these (as Strabo points
out) were much more erroneous, and those of Eratosthenes
substantially correct.^ To take a single instance : in regard
to the regions bordering on the Euphrates, which might be
supposed to have been better known to the Greeks in the days
of the Seleucidan monarchy, Eratosthenes placed Babylon
more than 2000 stadia to the east of Thapsacus, while Hip-
parchus assumed that it was not more than 1000.* The real

apxo-iois Tviva^i icaTayeypaTrrat) : those

namely before the time of Eratosthenes.

(See also ii. 1, p. 90.)

2 Strabo, ii. pp. 79, 87, &c._

^ This was the case especially with

the stranore error with regard to the

position of India. This had been in
great measure corrected by Eratos-
thenes, while Hipparchus returned to
the old view, and carried it to an exag-
gerated extent (Strabo, ii. 1, §§ 2-4).
* Strabo, ii. p. 90.

14 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVII.

interval is more than 5^° of longitude or about 240 Gr. miles,
(2400 stadia), so that the estimate of Eratosthenes was in
reality considerably leloiv the truth. Throughout the long
discussion which Strabo has preserved to us, of the geography
of these regions, and the points connecting them with Alex-
andria, it would appear that Hipparchus was generally wrong,
while the views of Eratosthenes were approximately correct.
It is difficult, however, to pronounce judgement with con-
fidence in such a controversy without being able to consult
and compare the original authorities.

One thing, however, is curious to observe : how both
Eratosthenes and Hipparchus are aiming at something like
geometrical correctness, and applying geometrical arguments
where they had no accurate observations to go on, or even
approximately correct measurements of distances. But with
all this, it is impossible not to see that they had an idea,
though dim and vague, of a kind of triangulation analogous
to that by which a modern geographer would endeavour to
connect distant points with which he was but imperfectly
acquainted.

§ 11. With regard to longitudes it does not appear that
Hipparchus was able to make any considerable advance on the
results obtained by his predecessor. He was indeed, as we
have already pointed out, the first to indicate the true method
of determining longitudes by the comparative observation of
eclipses,^ but no such observations were at his command, nor
have we any account of his having attempted to institute
them. The very imperfect means at the command of the
ancients for the measurement of time would indeed have suf-
ficed to prevent their being made with any approach to cor-
rectness ; but even such rough approximations as they could
give would have been a valuable assistance.

Hipparchus indeed wrote throughout as an astronomer,
rather than a geographer. Hence he does not seem to have

^ Strabo, i. 1, § 12, p. 7. Ptolem. Geogr. i. 4. See preceding chapter, p.

633.

Sect. 1.

HIPPAECHUS.

obtained, or even sought to obtain, any additional information
concerning tile western regions of Europe, beyond what was
known to his predecessors, though the course of events had in
his day had the effect of opening out new sources of knowledge,
of which he might readily have availed himself. In one
instance we find him recurring to an error which had been
generally received in earlier times, but had apparently beeu
rejected by Eratosthenes,^ in making the Danube flow with
one arm into the Adriatic, and with the other into the Euxine.'
This strange misconception continued indeed to be repeated
by many Greek writers long after his time.

§ 12. On one of the fundamental conceptions of geography
Hipparchus departed from the view which was generally
adopted in his time, as well as by most succeeding writers.
He refused to admit that the habitable world was surrounded
on all sides by sea, or that the Atlantic Ocean was continuous
with the Indian Ocean, and that again with the sea to the
north of Scythia,^ This scepticism appears to have been
based, not, as in the case of Herodotus, upon the mere absence
of proof, but upon certain observations of Seleucus (a Baby-
lonian author otherwise unknown) with regard to the tides,
which appeared to Hipparchus to be incompatible with the
hypothesis of a continuous circumfluent ocean.

In regard to the question, so much discussed among the
Alexandrian writers, of the Homeric geography, Hipparchus
altogether rejected the views of Eratosthenes, and adopted
the popular explanation, according to which the localities
visited by Ulysses were identified with well-known places on
the shores of the Mediterranean.^

" This is not indeed clearly stated by
Strabo, but may probably be inferred
from his expression that Hipparchus
shared this erroneous opinion " with
some of his predecessors." It was ap-
parently one of the notions of the

earlier geographers, which he revived
in opposition to Eratosthenes.

M. 3, § 15, p. 57.

8 Strabo, i. 1, § 9.

' Ibid. i. 1, § 2, p. 2.

l6 HISTOKY OF ANCIENT GEOGRAPHY. Chap. XVII.

Section 2. — Polyhius.

§ 1. Just about contemporary with Hipparchus was an author
of a very different character — the historian Polybius. Of his
merits as a political or historical writer it does not belong to
our present subject to speak, but the publication of his great
historical work may be considered also as marking an important
epoch in the progress of geographical knowledge. Polybius
was the first to avail himself of the new sources of information
that had been opened out to him by the wars and conquests
of the EomanS in Western Europe, and which had placed the
knowledge of those countries on an entirely new footing. As
he himself remarks, while Alexander had opened the way to a
more complete knowledge of the East, it was the conquests of
the Eomans that had first led to a similar acquaintance with
the West, and had afforded the means of access to regions
hitherto almost unknown to the Greeks.-^

These new sources of information had been hardly beginning
to be available in the days of Eratosthenes, and although the
rapid extension of the Roman power during the half century
that followed his death (U.c. 196-146), and the repeated wars
that brought it into collision with the Glreek monarchies both
in Europe and Asia, would seem likely to have awakened the
interest of the Greeks in general in all that their formidable
neighbour was doing elsewhere, there is no trace of their
having taken advantage of the opportunity thus afforded
them. We have seen how imperfect was the knowledge pos-
sessed by Eratosthenes of the western countries of Europe,
Spain, Gaul, and the regions north of the Adriatic : and
although his successors could hardly fail to have acquired an
increased acquaintance with these regions, it does not appear
that this had been yet embodied in any methodical form, so as
to render it available to the literary public in general.

§ 2. Polybius himself had indeed enjoyed peculiar advantages

' Polvb. iii. 59.

Sect. 2. POLYBITJS. 1 7

in this respect from the circunistances of his life and political
career. Born at Megalopolis in Arcadia about B.C. 204, he
was the son of Lycortas, one of the most distinguished leaders
of the Achaean League, and was early initiated in political and
military affairs. After the Second Macedonian War and the
defeat of Perseus (b.c. 167) he was one of the Achseans selected
as men of rank and influence to be sent as hostages to Eome,
where he remained seventeen years; and during this period
he had not only the opportunity of studying the political
institutions and history of Eome, which were still very im-
perfectly known to the Greeks in general, but he contracted
close personal friendships with many of the leading Roman
statesmen ; among others with the younger Scipio Africanus,
whom at a later period he accompanied during the Third
Punic War. Polybius was himself present at the destruction
of Carthage in B.C. 146, and was employed by Scipio in the
command of a Eoman squadron to explore the coasts of Africa.
It is probable also that he accompanied Scipio during the war
against Numantia (b.c. 134) : at all events it is certain that he
not only visited Spain and Gaul, as well as Africa, but under-
took, according to his own account, long and dangerous
journeys through those countries, extending even to the shores
of the Atlantic, with the express view of making himself
acquainted with their geographical position, as well as their
natural characters and productions.^ At what period of his
life Polybius undertook these extensive travels we are not
distinctly told : but it is difficult to place them before the fall
of Carthage, as, during the whole period of his compulsory
residence at Eome, he was in a certain sense a prisoner, and
after he was set at liberty, he returned in the first instance to
Greece, where he took an active part in public affairs, until
summoned by Scipio to attend him to the war. His historical
work was certainly not completed till after the same period ;
and ended with the destruction of Corinth, B.C. 146. Polybius

2 PoJyb. iii. .59.
VOL. II. C

HISTORY OP ANCIENT GEOGRAPHY.

Chap. XVII.

liimself survived that event by more tlian twenty years, having
lived to the advanced age of 82 ; his death may probably be
placed about B.C. 122.^

No historian of antiquity was more fully alive to the im-
portance of geography, as an aid to history, than Polybius.
This is not only apparent from the geographical remarks inter-
spersed through his whole work, but, like his predecessor
Ephorus, he had set apart one whole book for a systematic
treatise on geography, in which he had fully developed his
views upon that subject. Unfortunately this book — the thirty-
fourth of his voluminous work — is one of those lost to us : the
fragments preserved to us by Strabo and others being very
inconsiderable. From these fragments, however, combined
with the remarks introduced in earlier parts of his history, we
are enabled to gather the following leading outline of his
geographical views.

§ 3. He considered, in accordance with the views generally
received in his time, that Europe was bounded by the Tanais
on the east : and Asia was separated from Africa by the Nile :
the strait at the Pillars of Hercules of course forming the
boundary between Europe and Africa.* The whole of the north
coast of Africa from the Columns to the Altars of the Philseni,
on the shore of the Great Syrtis,^ was subject to the Cartha-
ginians, who had also extended their rule (previous to the
Second Punic War) over the whole coast of Spain from the
Columns to the headland where the range of the Pyrenees
descended to the Mediterranean.^ With the Pyrenees them-
selves he was well acquainted, and rightly conceived them as
extending from sea to sea, and sejDarating the Kelts or Gauls
from the Spaniards. Both these nations he knew to extend

^ For the chronology of the life of
Polybius, see Clinton's Fasti Hellenici,
vol. iii. p. 526 ; and the excellent
article on his life in Dr. Smith's Diet,
of Biogr. vol. iii. pp. 443-448.

1 Polyb. iii. 39.

* The name is here found for the

first time; but it is mentioned inci-
dentally as the well-known limit of the
Carthaginian dominions to the east
(iii. 39, X. 40). Tlie legend connected
with it is first related by Sallust
{B. Jug. 79).
« Id. iii. 39.

Sect. 2.

POLYBIUS.

across to the outer sea, or Atlantic Ocean, but of the regions
bordering on that Ocean (with the exception of part of Spain)''
he seems to have had very imperfect information. As we have
already seen he rejected altogether the authority of Pytheas
and his statements concerning the western coasts of Spain,
Gaul, and Britain, as well as his account of Thule and the
remoter lands towards the north. At the same time he had
very little to substitute in their place, and though he had
evidently some information concerning the British Islands, and
the lands from whence tin was brought,^ as well as concerning
the northern regions of Europe, it was evidently of a very
vague and general character. Indeed he himself tells us that
Scipio was unable to obtain any trustworthy information con-
cerning Britain from the merchants of Massilia or Narbo, or even
from those who came from Corbilo, an important emporium
of trade, situated apparently at the mouth of the Loire.* Un-
fortunately the part of his work in which he treated specifically
of these countries is lost to us : and the same thing is the case
with regard to the few notions he professed to have picked up
concerning the northern regions of Europe, extending from
Gaul to the Tanais.^ We may, however, infer from the total
silence of Strabo, that they contained little, if anything, of
importance. Polybius was indeed fully conscious of his
ignorance of these regions, and was content (like Herodotus)
to leave it in uncertainty whether there was continuous sea
to the north of Europe or not.^

7 He was certainly acquainted with
the mouth of the Tagus, and seems to
have obtained particular information
concerning the western shores of Lusi-
tania. See a passage cited from his
34th book by Athenseus, vii. p. 302 c.

* Id. iii. 57. This passage is im-
portant as being the first where men-
tion is found of the production of tin,
in connection with Britain. It is also
remarkable for the use of " the British
Islands " in the plural, but this ex-
pression apparently relates to Britain
and the adjoining Tin Islands (Casei-
terides), rather than to the more distant

and outlying lerne or Hibernia, with
which it is uncertain whether Polybius
was acquainted.

^ xxxiv. 10. The name of Corbilo is
not mentioned by any later writer, and
its site cannot therefore be fixed.
Strabo, who cites it from Polybius, says
only that it was an emporium on the
river Liger, but from the connection in
which he mentions it, we may probably
infer that it was at the mouth of that
river

» Polyb. iii. 37.

2 Ibid. iii. 38.

c 2

20 HISTORY OF ANCIENT GEOGEAPHY. Chap. XVII.

§ 4. His knowledge of Spain, which he had himself visited,
and which had been to a great extent opened up by the wars
of the Komans in that country, undoubtedly far exceeded that
which any other Greek had ever possessed. His geographical
account of the peninsula is indeed unfortunately lost, but the
number of names of towns, as well as of the native tribes, which
he incidentally mentions, suiBciently attests the extent of his
knowledge. If, indeed, this part of his work was not written
till after he had accompanied his friend Scipio to the Numan-
tine war, he must have had ample opportunities of informing
himself concerning the political, as well as the physical,
geography of Spain. He was well acquainted with the great
rivers : the Bsetis, the Anas, and the Tagus, and even attempted
an estimate of the length of the last ; which he considered to
have a course of 8000 stadia from its sources to the Ocean.^
He gave an account also of the great fertility of Lusitania,
which has every appearance of being derived from personal
observation ; as well as of the silver mines near New Carthage,
which were still extensively worked in his time, giving em-
ployment, it was said, to not less than 40,000 persons.* Their
produce was estimated at 25,000 drachmae a day, which was
probably an exaggeration.

Of Gaul he apparently knew much less : though the Eomans
had now established permanent footing in its south-eastern
portions and the commercial relations of Massilia with different
parts of the country had doubtless opened out new sources of
information concerning the interior, and even the external
coasts, which were before inaccessible. But the more accurate
and complete knowledge of Gaul, which had been acquired
in the time of Strabo, led him to attach less value to the
statements of Polybius, and consequently we rarely find his
authority cited, and are left much in the dark as to the actual
extent of his knowledge. From a passage of his work still
extant, however,^ it would appear that he knew the names at

3 xxxiv. 7, ap. Strab p. 106 ' Pnlyb. ap. Strab. p. 147. ' iii. 38.

Sect. 2. POLYBIUS. 2 1

least of the principal nations adjoining the Western Ocean,
which he promises to give in detail elsewhere. It seems
certain also that he was acquainted with the outlet of the
Loire (Liger) into the Atlantic Ocean, and even with the
existence of the Morini, as a people divided only by a narrow
strait from the island of Britain.^ He appears to have con-
sidered this as the most northern part of the continent of
Europe. This circumstance would alone show how complete
was his ignorance of the adjoining regions. The most remote
people of the interior of whom we find mention are the Arverni : ''
but there is little doubt that his knowledge in this direction
was really more extensive.

§ 5. In no respect was Polybius more in advance of all his
predecessors than in his knowledge of the Alps. The import-
ance of this great chain of mountains as one of the main
geographical features of Europe could not fail indeed to force
itself upon the attention of all observers as soon as the Eomans
had extended their conquests to the foot of the great barrier,
which encircled Italy on the north, and appeared to cut off all
communication with the nations beyond. But already before
the birth of Polybius, Hannibal had shown that it was possible
to conduct an army across this formidable mountain barrier :
and the experiment was subsequently repeated by Hasdrubal
with comparatively little difficulty. The Eomans indeed do
not appear to have followed their example until a much later
period, so far as the passage of armies was concerned, as their
troops could be transported with much greater facility by sea
to Massilia and the mouths of the Rhone. But there can be
no doubt that the passes across them were already well known
and frequented by the neighbouring tribes, and Polybius him-

See Pliny, M. N. iv. 23, § 122. i ' Tlieii- name is found in the account

At the same time it must be admitted
that Pliny's expression (unde per Lug-
dunum ad portum Morinorum Britan-

of the march of Hasdrubal throu<
Gaul, to join his brother Hanuibal in
Italy, as given by Livy (Liv. xxvii. 39 ) ;

niciim, qua videtur mensuram agere and thougli this part of the work of
Polybius) is ambiguous, and Pliny may \ Polybius is lost, we may safely assume
have himself supplied the name. | that it was derived from him by Livy.

22 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVII.

self had acquired accurate information concerning- them. He
states indeed expressly that he had himself followed in detail
the route of Hannibal over the mountains : ^ his description of
the localities is lively and picturesque, and as he had himself
no doubt of the route taken by the great Carthaginian general,
we may feel confident that he would have set the question at
rest for all subsequent inquirers had he not unfortunately
adopted the plan of omitting almost all proper names, whether
of tribes or places, as barbarous in sound and utterly unknown
to his Greek readers. To us on the contrary they could
hardly have failed to convey most valuable information.^

In the time of Polybius it would appear that only four
passes across the main chain of the Alps were generally known
and used : the first, through the Ligurians, following the coast
of the Tyrrhenian Sea ; next that through the Taurini, which
was supposed to be the one followed by Hannibal ; thirdly,
that through the Salassians (an expression applicable to either
the Great or the Little St. Bernard) ; fourthly, that through
the Rhsetians — the modern Tyrol — which from its great com-
parative facility must have been frequented in all ages.^ Un-
fortunately the statements of Polybius on this subject are
preserved to us only at second-hand by Strabo, and we have
thus no means of estimating the amount of information which
he really possessed concerning the passes thus enumerated.

§ 6. But there can be no doubt that his knowledge of the
Alpine nations that bordered Italy on the north, and extended
from thence to the Danube must have been very imperfect.
It was not till the reign of Augustus that the Eoman arms
were carried into these wild regions : and there was probably
but little commercial intercourse with the tribes north of the
Alps until those inhabiting the mountains had been brought
under the dominion of Rome. But the foundation of the
Roman colony of Aquileia, as early as B.C. 181, doubtless led
to more extensive commercial relations with the neighbouring

iii. 48. '' Note A, p. 37. ' Polyb. ap. Strab. iv. 6, p. 209. Sec Note A.

Sect. 2.

POLYBIUS.

tribes in this quarter : and the discovery soon after of gold
mines of extraordinary richness in the land of the Taurisci, a
Norican tribe,^ must have tended greatly to increase the com-
munication with Italy. Such was the effect produced by the
sudden influx of gold from this source, that we are assured the
price fell directly throughout Italy by one-third.^

Notwithstanding the valuable information acquired by
Polybius concerning the Alps, and its great superiority to that
of his predecessors, we must not suppose that he had anything
like a clear geographical acquaintance with the course and
configuration of, that great chain. He was indeed familiar — as
any one who had visited the north of Italy must needs be —
with its great extent, which he estimated at 2200 stadia,
measured along the plains at the foot of the mountains (a
statement very much beneath the truth), and he rightly judged
them to be of much greater elevation than any of the moun-
tains in Greece or the neighbouring countries. But his mode
of estimating their height was singularly rude and imperfect :
for (he said) it was possible for any active man to ascend the
most lofty of- the Greek mountains — Taygetus, Parnassus,
Olympus, &c. — within a single day, or thereabouts, while it
would take any one more than five days to ascend the Alps.*
Whether this refers to the journey across the passes, or was
founded upon some vague stories he had heard from the moun-

^ iv Tois TavplcTKOis to7s 'NaipiKo7s.
This is the first mention of the name of
the Noricaos, afterwards so familiar to
the Koman writers. Pliny (iii. 20,
§ 123) appears to regard the Tauriscans
and Noricans as synonymous terms ;
while Strabo calls the Tauriscans a
Norican tribe (iv. 6, § 9), and this
appears to have been the general view,
but their position is not fixed by any
ancient writer.

* Polyb. ap. Strab. iv. 4, p. 208.
From his description it is evident that
this gold was derived from " diggings,"
very much in the same manner as that
now found in Austrnlia. It often
occurred in lumps or " nuggets " of the
size of beaus or peaa. The works were

doubtless soon exhausted, and seem to
have ceased to exist long before the
time of Sti'abo, who refers to them
quite as a thing of the past.

* xxxiv. 10. It is curious to compare
this with the statement of Strabo that
the mountains in the land of the Me-
dulli, which he regarded as the highest
in the* whole range, were 100 stadia in
the direct ascent of the highest part :
a very moderate estimate, as his words
clearly imply that he is not speaking
of their perpendicular height {rh yovv
bpOidiTarov avrSiv vi]ios araSicov eKarhv
'^X^'-^ </>aa'I Ti]v avd^affLV, KayOei/Se irdAiu
T-,)v eirl Tohs (jpovs rf/s 'IraXias Kard-
^acnv. Strabo, iv. 6, § 5).

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVII.

taineers of the time actually required to climb the highest
peaks, it is in any case a palpable exaggeration. At the same
time his description of the Rhone (Rhodanus) as having its
sources "in the most northerly parts of the Alps above the
inmost recess of the Adriatic," and flowing from thence towards
the south-west,^ sufficiently indicates how vague, or rather how
utterly erroneous, was his conception of the general configura-
tion of the Alpine chain.

§ 7. His knowledge of the geography of Italy was doubtless
greatly superior to that of any of his predecessors : the more
northern parts of that country, which had been previously very
imperfectly known to the Greeks, having been opened out by
the conquests of the Romans in Cisalpine Gaul, Liguria and
Yenetia. His account of the great lakes of Northern Italy, and
the rivers that flowed from them, though distorted in our ex-
isting copies of Strabo, was evidently originally quite correct,^
and shows an accurate knowledge of the countries in question.
But his general description of the peninsula, and the nations
that inhabited it, which undoubtedly found a place in his
thirty-fourth book, has been unfortunately lost to us. The
expression in one passage '^ where he speaks of Italy as a kind
of triangle, having the Alps for its base, and the promontory
of Cocinthus for its vertex, would seem at first to indicate a
very imperfect idea of its figure ; but it is clear that this
similitude is to be taken only in a very rough and general
sense, and we learn from another passage cited by Strabo,^
that he was well acquainted with the peculiar configuration of
the southernmost part of the peninsula, and its two long pro-
jecting arms separated by the deep gulf of Tarentum. In like
manner his description of Cisalpine Gaul, and the relations
between the two great cliains of the Alps and the Apennines

5 iii. 37.

« Polyb. ap. Strab. iv. 6, § 12. It
is admitted by all recent editors, that
the error in our existing copies, by
which the Addiui is described as flow-
ing from the Lake Vcrbanus, and the

Ticinus from the Lake Larins, is a mere
fault of the copyists and cannot be
imputed either to Polybius or Strabo,

' ii. 14.

8 Ap. Strab. v. i. § 3.

Sect. 2.

POLYBIUS.

that bounded it, though falling far short of the requirements
of modern geography, still presents a clear and intelligible
picture, exceeding almost any similar passage to be found in
earlier writers. In describing these regions he evidently does
so in the belief that they were still imperfectly known to his
contemporaries in general, and takes occasion in passing to
censure the ignorance that Timseus especially had shown con-
cerning them. That writer, as having been a native of Sicily,
might naturally "have been presumed to have possessed better
means of information concerning Italy and the adjoining
islands than most other Greeks, and probably enjoyed a repu-
tation on that account ; for which reason Polybius especially
selected him for criticism, and showed up at considerable
length the ignorance he had displayed with regard both to
Italy and the adjoining islands of Sardinia and Corsica.^

§ 8. Another quarter in which the wars of the Eomans had
first led the way to a more accurate geographical knowledge
was in regard to the countries bordering on the Adriatic.
Nowhere does the ignorance that prevailed down to a late
period among the Greeks appear more inexplicable than in
regard to this inland sea. From a very early period the Greek
colonies of Apollonia and Epidamnus (or Dyrrachium), situ-
ated just within its entrance, had risen to considerable import-
ance and become the seats of a flourishing commerce : ^ while
soon after the beginning of the fourth century B.C. the founda-
tion of new colonies in the islands of Pharos and Issa, and that
of Ancona on the opposite coast of Italy, must have led to
greatly increased commercial relations with the adjoining
nations, and to a more familiar knowledge of its shores.^ Yet

" xii. 3, 4. It is amusing to find
Polybius himself, after stating that
Corsica produced no wild animals,
except foxes, rabbits, and moufflons, or
wild sheep, proceeding to give a par-
ticular account of rabbits, as an animal
with which he evidently presumed his
readers to be unacquainted.

' This is sufficiently attested by their
coins, which are found in large numbers

in the countries adjoining the Adriatic,
and even far into the interior, and some
of which belong to quite an early period.
Dyrrachium (or Epidamnus, as it is
called by the historian) figures as a
place of importance in Thucydides, at
the outbreak of the Peloponuesiun War
(Thucyd. i. 24).

^ The colonies of Pharos and Issa
were founded by the Parians about b.c.

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XYII.

we find the Greek writers long afterwards continuing to trans-
mit the old erroneous notions concerning the countries at the
head of the Adriatic which they had received from their pre-
decessors. Thus Timseus, as we learn from Polybius, while
he repeated the old fables concerning the Eridanus, and
the amber tears of the poplars on its banks, showed the
greatest ignorance of the real geography of these regions.^
We have seen also that even Hipparchus, the contemporary of
Polybius himself and an enlightened geographer, accepted the
popular notion that the Ister or Danube discharged a part of
its waters by one arm into the head of the Adriatic : * while
writers subsequent to Polybius were still content to repeat the
strange misconception that the Hyllic Chersonese on the coast
of Dalmatia was about as large as the Peloponnese ! ^ The
wars of the Romans with the lUyrian queen Teuta, and at a
later period with the Dalmatians, being of necessity carried on
principally by sea, must have led to a comparatively accurate
knowledge of the eastern shores of the Adriatic : but the
description of them by Polybius, which was contained in his
special geographical treatise, is unfortunately lost to us. It
appears however that he was still very imperfectly acquainted
with its geographical form and dimensions, and had a very
exaggerated idea of its extent.^

387, -with the assistance of Dionysius
of Syracuse, who was desirous of esta-
blishing his power over the whole of
the Adriatic (Diodor. xv. 13). Ancona,
on the contrary, according to Strabo,
was founded by fugitives from Syra-
cuse, who sought to escape from tlie
despotism of the tyrant (Strabo, v. p.
241). Hence Juvenal calls it "Dorica
Ancon." (Sat. iv. 40). It was the only
Greek colony in this part of Italy.

3 Polyb. ii. 16.

* Strabo, i. 3, § 15.

^ Scymnus Chius, vv. 773-776. Ac-
cording to this author, who is doubtless
following earlier authorities, the
Danube flowed by five mouths into the
Euxine, and by two into the Adriatic.

° DicsBarchus had previously stated
that the distance from the Peloponnese

to the head of the Adriatic was greater
than that from the same point to the
columns of Hercules, which he esti-
mated at 10,000 stadia ! (ap. Strab. ii.
4, p. 105.) Polybius himself gives the
length of the Adriatic from the Pelo-
ponnese (opposite to Leucadia) to the
head of the Gulf of Quarnero at 8250
stadia (ap. Strab. I. c), still an enormous
over-statement ; the real distance being
less than 600 G. miles. Hence, as
usual, Gossellin tries to explain his
error by supposing him to have made
use of a smaller stade. But as the
distances given from the Peloponnese
to Leucadia, thence to Corcyra, and
thence again to the Acroceraunian
promontory, are all approximately cor-
rect, while that along the Illyrian coast
is greatly in excess, he is driven to the

Sect. 2.

POLYBIUS.

§ 9. Another point on which the Eomans had contributed a
material addition to the accuracy of geographical knowledge
was by the construction of the celebrated Via Egnatia, leading
direct from ApoUonia to the shores of the Propontis. The
value of such a line of high-road, with measured distances, con-
necting the Adriatic with the ^Egean and the Propontis, would
have been an invaluable resource to the ancient geographers,
had they known how to avail themselves of it, in the manner
that a modern geographer would have done. But even as it
was, it became an important means of rectification of the
notions previously existing, Eratosthenes, by a strange and
unaccountable error, had estimated the interval between the
two seas (the Adriatic and ^gean) at only 900 stadia (90 G.
miles) while Hipparchus, correcting him, had correctly asserted
that it was more than 2000 stadia.' Polybius, following the
line of the Egnatian Way, gave the distance from ApoUonia
to Thessalonica as 267 Roman miles, or 2136 stadia. From
thence to Cypsela on the river Hebrus he reckoned 268 miles :
apparently the road had not at that time been carried any
farther.^ These distances agree almost exactly with those
found at a much later period in the Antonine Itinerary,^ thus
showing the great value of the new source of information now
for the first time introduced into geography. Of this Polybius
was fully aware, and he repeatedly alludes to the great advan-
tage derived from the lines of Roman roads, " with the miles
measured and marked along them."^ But he does not appear,

expedient of supposing Polybius to
have employed one kind of stade in the
first part of the passage and another in
the latter ! It would seem much more
simple to admit that the former dis-
tances were correct, because they were
well known to the Greeks, the latter
was exaggerated because it was still
imperfectly known. The very broken
and irregular conformation of the Dal-
matian coast would also contribute
greatly to the difficulty of estimating
the length of the partqjlus, and tlie
strange misconception concerning the

Hyllic peninsula probably arose from
the same cause.

' Strabo, ii. 1, § 40, p. 92.

8 Polyb. ap. Strab. vii. p. 322.

3 Itin. Ant. pp. 329-332. In the
itinerary tlie line of road is naturally
continued to Byzantium, but when
first constructed it was undoubtedly
directed to the Hellespont, not to the
Bosphorus.

' ravra yap vvv ^e^T)ixaTiffTai kolI
crecfTjyUeicoTai Kara ffTa^iovs oktw Sia Poo-
fxaiaiv 61T1/X6A&JS, iii. 39. See also xxxiv.
12, concerning the Egnatian Way.

HISTORY OF ANCIENT GEOGEAPHY. Chap. XVII.

any more tlian preceding geograpliers, to have pointed out any
mode of correcting these itinerary distances, in order to apply
them to the determination of the geographical intervals, and
the true position of the points thus connected. The want of
all power of taking observations of longitude deprived the
ancient geographers of the most ready and important means
of correction; but we find no trace of their applying sys-
tematically the simple and obvious expedient of deducting a
given portion of the itinerary distances for the windings of the
road, in order to arrive at a nearer approximation to the direct
distances between any given points.

§ 10. Polybius had himself visited Byzantium, and his de-
scription of the peculiar site of that city, and the advantages
it derived from its position, in regard to the trade with the
Euxine and the Palus Mseotis, is one of the most valuable that
has been left us from antiquity. It is worthy of remark that
he prefaces these details with the excuse that they were not
generally known, on account of the place lying rather out of
the way of those parts of the world that were generally visited.^
It does not appear that he had himself penetrated any farther
within the Euxine, but he had clearly obtained good informa-
tion concerning that sea, as well as the Palus Mseotis,^ and
the connecting strait of the Cimmerian Bosporus : and his
observations on the currents of the two straits, and the phy-
sical changes that were going on in the two seas are a very
interesting specimen of early speculations on physical geo-
graphy.* His inference, that from the great amount of alluvial
deposit brought down by the numerous rivers flowing into the

* iv. 38. Sia Th fiiKphy e^ai Ke7a6ai twv
iTnffKOTroviu.ei'wv /xepwv ttjs olKov/xfvrjs.
Yet he directly afterwards speaks of
his own times as a period when all
countiies were visited and become ac-
cessible, both by land and sea (jwv vvv
KaipSiv, iv ols TrdvTwv TrAwrciv Kal iropev-
tSiv y^yovSruv, iv, 40).

^ One proof of this is that he was
well aware of the small size of the
Palus Mffiotis as compared with the

Euxine, as he reckons the circum-
ference of the one at 8000 stadia, of
the other at 22,0U0 (iv. 39). Earlier
writers, as we have seen, had supposed
the Palus Mjeotis to be half as large
as tlie Euxine : and even long after the
time of Polybius very exaggerated
notions were entertained of its size and
extent.

^ Polyb. iv. 39-42.

Sect. 2.

POLYBIUS.

Euxine and Palus Meeotis, the former sea would gradually
become shoal, as the latter was already in his day,^ and that
both would eventually be filled up — was unquestionably cor-
rect in theory ; but he seems to have greatly overrated the
rapidity of the process, chiefly from not having made sufficient
allowance for the great depth of the Black Sea. Even the
shallow Palus Mseotis, though growing continually shallower,
is still far from being filled up, and Polybius would doubtless
be much surprised, could he compare its present condition, at
the small amount of change that has actually taken place in
2000 years.^

§ 11. Of the knowledge possessed by Polybius of Asia we
have very imperfect means of judging. But there is no reason
to suppose that he had any important sources of information
concerning the more distant regions of that country, which
were not available to Eratosthenes. Indeed he himself in one
passage spoke of Eratosthenes as the best authority concerning
the Asiatic provinces from the Euphrates to the Indus.'' At the
same time there can be no doubt that had his account of the
campaigns of the Seleucidan kings in the remoter provinces of
their empire been preserved to us entire, we should have derived
from it many valuable contributions to the more detailed
knowledge of those countries. The most important of these
operations was the expedition of Antiochus the Great with a
view to reduce the revolted provinces of Upper Asia again to
submission. From the extant fragments of this part of his his-
tory we learn that Antiochus, after defeating Arsaces, king of

* Aristotle had already pointed out
thisfact (Meteorologica, i. 14, § 29), and
drawn the same inference as Polybius.
That philosopher observes that the
Palus Mseotis -was no longer navio:able
in his day for the same sized vessels as
it had been sixty years before, a very
curious fact, which be had doubtless
learnt from Greek traders. Polybius
tells us that iu his day the greater part
of it was only from 7 to 5 fathoms deep,
and could not therefore be navigated

hy large ships without a pilot (iv. 40), a
clear proof of the size of the trading
sliips that frequented the Euxine. At
the present day the greater part of it
has still a depth of about 6 or 7 fathoms
(Admiralty chart).

^ See on this subject the observations
of Dr. Goodenough in the Journal of
the Geographical Society, vol. i. pp.
105-107.

' Polyb. ap. Strab. xiv. p. 663.

30 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XVII.

Partllia, and reducing Euthydemus, king of Bactria, to sub-
mission — though, leaving him the title of king — crossed the
(Indian) Caucasus, and descended into India, where he re-
newed with Sophagasenus, the Indian king, the relations of
friendship contracted by Seleucus I. with Sandracottus, about
100 years before, and received from him a number of addi-
tional elephants. He afterwards accomplished his return
through Arachosia, and after crossing the river Erymanthus
(evidently the Etymander, or Helmund) proceeded through
Drangiana into Carmania, where he wintered.^ Some further
details with regard to this latter part of his march Avould have
been peculiarly interesting, as the route from Drangiana into
Carmania lies across an arid and perilous desert, which has
only very recently been traversed by any modern traveller.^

An incidental notice of the town of Gerrha on the Persian
Gulf is interesting as showing the extensive commercial rela-
tions maintained by the inhabitants with other parts of Arabia,
from whence they derived large quantities of myrrh and frank-
incense, as well as with the Greeks of Seleucia. Antiochus
appears to have intended to reduce the city and neighbouring
tribes, but was ultimately content to leave them in enjoyment
of their liberty ; a concession which they however purchased
by magnificent presents.^

§ 12. From the few portions that remain to us of this part
of his history it seems probable that Polybius followed, in
regard to the remoter provinces of Asia, the same rule that he
had laid do^^^l to himself with respect to Gaul and the Alpine
tribes, of introducing as few proper names of places as possible,
except such as might be supposed already familiar to Greek
ears : like Hecatompylus and Zariaspa. Fortunately it was
otherwise with regard to Asia Minor, with which the Greeks in
his day would in general be sufficiently well acquainted ; and
his accounts of the campaigns of the Eoman and the Syrian

' Polyb. X. 49, xi. 34. These opera-
tions may apparently be assigned to
the years 206 and 205 B.C.

3 See Chapter XII. p. 521.
' Id. xiii. 9.

Sect. 2.

POLYBIUS.

monarchs in that country are among the most valuable ma-
terials for the determination of its geography. These have
indeed been preserved to us chiefly at second-hand ; but the
few fragments of the original that remain are sufficient to
show how closely the narrative of Polybius has been followed
by Livy ; and we may rely with confidence on the geogra-
phical details furnished by the latter, in this part of his work,
being derived immediately and wholly from his Greek au-
thority. Nowhere is this more conspicuous than in his relation
of the campaign of the consul On. Manlius against the Gala-
tians or Gauls then recently established in Phrygia, which is
at once an interesting piece of military history, and a most
valuable addition to our detailed geographical knowledge of
the countries in question.^ The same remark applies to the
important geographical details, as well as the graphic descrip-
tions of localities, found in the narrative of the wars of the
Eomans in Greece and Macedonia, as preserved to us by Livy.
There can be no doubt that all these details — the value of
which is acknowledged by all modern topographers — are
derived directly from Polybius.^

§ 13. With regard to Africa — at least to the northern parts
of that continent, bordering on the Mediterranean — there can
be no doubt that Polybius possessed much more ample means
of information than had been accessible to previous Greek
geographers. The wars of the Romans with Carthage, and
their alliance with the Numidian king Masinissa, had opened
out the knowledge of regions and countries in this direction,
which had been previously almost a sealed book to the Greek
writers. There is little doubt that commercial jealousy ex-
cluded foreign traders from Carthaginian ports, with the

2 See the remarks of Colonel Leake
{Asia Minor, p. 145), who derived im-
portant assistance from this source in
his able attempt to arrange the geo-
graphy of Asia Minor with the very
defective materials at his command.

* To the same source is undoubtedly
owing the unusually detailed and

graphic account by Plutarch of the
defiles of the river Aous, where Flami-
ninus defeated Philip V., king of
Macedonia ; though Plutarch has com-
mitted the strange mistake of con-
founding the Aous with the neighbour-
ing river Apsus (Pint. Flaminin. c. 3 ;
Leake's Northern Greece, vol. 1. p. 389).

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVII.

exception of tlie capital, and perhaps one or two other points :
and the scanty information possessed by most Greek writers
upon the extensive regions subject to the Carthaginian rule is
a remarkable feature in all the earlier geographical treatises.
But from the time of the Punic Wars the names of the Mas-
sylians, the Masssesylians, and the Maurusians or Mauretanians
had become familiar to the Romans, and had doubtless reached
the ears of the Greeks before they were introduced to them by
Polybius.*

§ 14. But that historian not only possessed all the informa-
tion that had thus become available to the Romans in general,
but he had enjoyed special opportunities of surveying and
examining in detail the coasts of Africa, having been ap-
pointed by his friend Scipio, during the Third Punic War,
to the command of a squadron, with the express purpose of
carrying on such investigations. In pursuance of this object
he not only visited the Carthaginian coasts along the Medi-
terranean — which gave him occasion to describe the island of
Meninx, near the Lesser Syrtis, and to enter into a detailed
account of the Lotus-tree, and the manner in which it was
employed as food ^ — but he extended his explorations beyond
the Pillars of Hercules, and proceeded to a considerable dis-
tance along the western coast of Africa. Unfortunately the
results of this last voyage — of which the narrative, had it been
preserved to us in the original, would have been one of the
most interesting and valuable contributions to our geogra-
phical knowledge — have been transmitted to us in a form so
imperfect and obscure that they add almost nothing to the
information we derive from other sources. The narrative of
Polybius himself is utterly lost : and strange to say, no men-
tion is found in Strabo of this remarkable voyage, which we

* Together with these familiar names
we find mentioned by Polybius (iii. 33),
a people called the Maccsei, a name
unknown to later writers. The Maca3
of Herodotus (iv. 175), called by Ptolemy
(iv. 3, § 27) Macaji, who dwelt on the

Great Syrtis, can hardly be the people
meant.

^ Polyb. xii. 2. See pn this subject
Bavth, Wanderimrien, pp. 259-265, and
above, Chapter VIII. Note P.

Sect. 2. POLYBIUS. 33

know only from the notice of it in Pliny, who has cited from
it a number of names and distances ; but these are given in so
confused a manner that it is impossible to arrange them in any
intelligible order, or to arrive at any satisfactory conclusion
concerning them.® We are unable even to discover to what
distance his actual voyage extended, and what was the farthest
point reached ; but it seems probable that his own explorations
did not proceed much, if at all, beyond Cape Noun (about
600 miles from Cape Spartel); and that the names of head-
lands, mountains, and rivers, which belong to more distant
regions, and some of which we recognize from their occurrence
in the voyage of Hanno, were derived from other authorities.
The loss of this portion of the work of Polybius is the more to
be regretted, as there is no doubt from all our knowledge of
this careful and conscientious author that it was given in the
original in a thoroughly trustworthy form : and would have
thrown much light upon the earlier voyage of Hanno, as well
as upon the geography of the western coast of Africa in the
time of Polybius.

§ 15. With regard to the extension of Africa towards the
south, he did not acquiesce in the opinion generally adopted
in his time of its being surrounded by the Ocean ; but dis-
tinctly tells us that with regard to the extreme parts of
Ethiopia, where Asia and Africa joined, no one was able to
say with certainty whether there was continuous land, or it
was surrounded by the sea. The same doubt existed with
regard to these regions as with respect to those in the extreme
north of Europe.^ He did not therefore adopt the theory of a
circumfluous ocean, but was content (like Herodotus) to leave
the matter in a state of doubt.

It is singular that Polybius, while in this instance refrain-
ing so carefully from any theoretical inference, should have
adopted without hesitation a view suggested by some previous
geographers, that the immediate neighbourhood of the equator

« Note B, p. 40. ' Polyb. iii. 37.

VOL. II. D

34 HISTOKY OF ANCIENT GEOGEAPHY. Chap. XVH.

was mucii less hot than the torrid zones on each side of it, so
as to be habitable, and that it was in fact inhabited. This
would appear to point to the existence of some dim and
floating traditions of the populous and fertile regions of
Soudan, south of the great desert of Sahara. But he appears
to have based it principally upon some supposed astronomical
causes. No mention is found in Strabo of such a theory as
ascribed to Polybius,^ and it was therefore probably not noticed
in his history ; but he is reported to have written a special
treatise on the subject.^

§ 16. Although Poly bins, writing as a historian and poli-
tician, would naturally give his attention rather to physical
geography, and the natural boundaries and leading geo-
graphical features of countries, than to the more technical
parts of the science, he yet seems to have rightly compre-
hended the necessity of laying down a correct map of the
different countries with which he was concerned; and thus
bestowed considerable pains upon the determination of dis-
tances, as well as the configuration of lands and seas. He is
even cited by Pliny, together with Eratosthenes, as one of the
most diligent inquirers into this branch of the subject.^ But
from the loss of that portion of his work, which contained a
formal treatise on geography, the greater part of his state-
ments of this description are lost to us. We however learn
from Pliny that he reckoned the distance from the Strait of
the Columns to Carthage at 1100 Eoman miles (8800 stadia),
and from thence to the Canopic mouth of the Nile at
1528 miles, or 12,224 stadia; while he estimated the total
length of the Mediterranean from the Straits to Seleucia in
Syria at 2440 miles or 19,520 stadia;^ a calculation con-

" Strabo, however, elsewhere inci-
dentally refers to such a view as being
entertained by some of the earlier geo-
graphers (koI yap el olKriffifjia ravrd
iffTLv, Siffirep olovrai rives, ii. 5, § 34, p.
195).

cited by Geminus, an astronomical
writer of the first century B.C., EUm.
Astvon. c. 14. ,

* " Polybius et Eratosthenes, dili-
gentissimi existimati." Plin. if. N. v.
6, § 40.

irepl TTjs irio\ rhv la-r]iJ.epLvhvolKriaeoi}s, \ ^ Plin. H.N. vi. 33, § 206. This

Sect. 2.

POLYBIUS.

siderably nearer to tlie truth than that of Eratosthenes ; only
falling short of the real length by about 500 stadia, while
that of Eratosthenes exceeded it by more than ten times
that amount.^ On the other hand he materially underrated
the width of the Mediterranean, asserting that its greatest
breadth — which he erroneously conceived to be in a line due
south from Narbo to the coast of Africa — did not exceed 3000
stadia.* In consequence of this error he had to bring down the
coasts of G-aul and Liguria much too far to the south, and give
a much narrower form to the Mediterranean, than had been
supposed by Eratosthenes and his followers, who, as we have
seen, placed Massilia almost exactly in its true position. He
estimated the whole width of Europe from south to north, —
from Italy to the Ocean, at 1150 miles; drawing his line,
according to Pliny, through Lugdunum to the port of the
Morini opposite to Britain.^ But in this case, as in several

sum total was made up of the following
details : — From the Straits in a straight
line due east to Sicily he reckoned 1260
miles and a half,

thence to Crete . . . 375 miles
„ Khodes . . 1831 „
„ the Chelidonian

Islands. . the same
distance,
„ Cyprus . . 322 miles
„ Seleucia . . 115| „

Here the affectation of precision im-
plied by the notice of half-miles (d
passu s) contrasts curiously with the
absence of any indication of the par-
ticular points in the islands of Khodes,
Crete, and even Sicily, from which his
measurement was taken. The intro-
duction of Rhodes and the Chelidonian
Islands between Crete and Cyprus, as
if they all lay on the same parallel of
latitude, is also a grave error.

3 See Chapter XVI. p. 685.

* Polyb. ap. Strab. ii. 4. § 2, p. 105.
The effect of this error was greatly
enhanced by its being coupled with
another, which was evidently held by
Polybius in common with all his pre-
decessors, that the line of the African
coast, from the Straits to Carthage, lay

much to the south of its true position ;
so that, according to his conception,
the direct distance from the angle near
Narbo to the point where a meridian
line drawn from thence would inter-
sect the parallel from the Straits of
Gades to the Sicilian Strait, did not
exceed 2000 stadia, while there re-
mained 1000 more to the coast of Africa
(Strabo, Z.c). But in point of fact the
African coast, at the point opposite to
the inmost bight of the Gulf of Narbo
(which is near Montpellier), lies more
than 50 G. miles (500 stadia) to the
north of the parallel of 36°, which
passes through the Straits of Gibraltar.
Polybius therefore did not err so much
in his estimate of the width of the
Mediterranean at this point (which,
however, really amounts to 6J degrees
of latitude or 3900 stadia), as in bring-
ing down both the coast of Africa and
that of Gaul far below their true posi-
tion in latitude. Yet this error, which
was the more inexcusable, as it was a
departure from the more correct views
previously entertained, was, as we shall
hereafter see, not only adopted, but
carried considerably faither by Strabo.
^ Plin. H. N. iv. 22, § 121. "Polybius

D 2

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVII.

others, we are left mucli in doubt from the loss of the original
work, and the careless manner in which his statements are
reported by Pliny and others. The whole distance hy sea from
the Straits of Gibraltar to the mouth of the Palus Mseotis he
reckoned at 3437 miles, following the most direct course that
was possible.

latitudinem Europse ab Italia ad Ocea-
num scripsit imdecies centena et quin-
quaginta M. esse, etiam turn incom-
perta magnitudine ejus." Pliny adds
that the true distance from the frontier
of Italy at the Alps " per Lugdunum
ad portum MorinorumBritannicum, qua
videtur mensuram agere Polybius," was
not less than 1318 miles, an enormous

exaggeration, the origin of which we
are unable to explain. The distance
according to the Itineraries does not
exceed 650 Roman miles.

The expression " portum Morinorum
Britannicum " for the port from which
men traded with Britain, is singular;
but this probably belongs to Pliny and
not to Polybius.

Note A. POLYBIUS. 37

NOTE A, p. 22.

Hannibal's passage of the alps.

It is impossible within tlie limits of a note to attempt the discussion
of the much disputed question of the passage of the Alps by Han-
nibal. The route which he followed, and the particular pass by
which he crossed the mountains were a subject of doubt in the
days of Livy, and have continued to be so down to our own time.
But it may at least be asserted that the question has been brought
within much narrower limits by the course of modern investiga-
tions. No one will any longer be found to maintain the claims
either of the Great St. Bernard or the Mont Genevre : though the
former was adopted by Cluver, and the latter by DAnville and
Gibbon. The choice may be said to lie between the Little St.
Bernard — the one supported by De Luc, and in a more elaborate
form by Wickham and Cramer, as well as in the recent work of
Mr. Law (The Alps of Hannihal, 2 vols. 8vo. Lond. 1866) — and the
Mont Cenis, which was maintained by Ukert, and with a slight
modification by Mr. Ellis (^Treatise on Hannibal's Passage of the Alps,
8vo. Camb. 1854). I can here only state very briefly the reasons
which appear to me decisive in favour of this latter hypothesis.

1. It is agreed on all hands that the question must be decided
by the authority of Polybius alone : neither Livy nor any later
writer having any clear understanding of the subject. Now it is
certain that Polybius does not intimate that there was in his day
any doubt about the matter : he describes the march in consider-
able detail, and notices the special natural features and obstacles
which had any marked influence on its incidents, evidently assuming
that these were well known. But Polybius (as we have seen in the
text) was acquainted with only four passes across the AIjds, and
after mentioning the second of these, that through the Taurini, he
added (as his words are reported to us by Strabo) " which was that
by which Hannibal crossed " (eTra rr]v Sta Tavplvoiv rjv 'Avvb/3a<s
SirjXOev, ap. Strab. iv. 6, p. 209). It is true that we only have this
passage at second hand : and the advocates of the Little St. Bernard
theory reject the words just cited, as being an addition of Strabo's.
But no one (I think) reading the passage for the first time would

38 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVII.

doubt their forming part of the statement derived from Polybius ;
and I may venture to add that a long familiarity with the mode in
which Strabo cites his authorities, confirms the conviction in my
own mind that they were so derived.

2. If the authority of Polybius be really as expressed in the
above words, it appears to me almost decisive in favour of the
Mont Cenis route. The only two that could reasonably be de-
scribed as passing through the land of the Taurini, would be this,
and that over the Mont Genevre, which may safely be pronounced
untenable, as not agreeing in detail with any of the circumstances
recorded of the passage. This latter route, which was always
described in later times as passing through the Cottian Alps, was
well known and frequented by the Eomans : but it appears to
have been first followed, and as it were discovered, by Pompey
when marching from Italy into Spain in B.C. 74, when, as he himself
states in a letter to the senate, he opened out a route different from
that of Hannibal, but more convenient for the Eomans (" per eas
[Alpes] iter, aliud atque Hannibal, nobis opportunius patefeci."
Pompeii Epist. ap. Sallust. Hist. Fr. iii. 1). This new route was
almost certainly the Mont Genevre, which was in fact much the
most direct line into the Eoman province of Gaul and Spain : and
for that reason the route of the Mont Cenis seems to have fallen
into disuse after this time.

3. Without attaching too much importance to the dramatic
incident of Hannibal's address to his soldiers, and pointing out to
them the plains of Italy (Polyb. iii. 54), it is told by Polybius —
the most unpoetical of historians — in a manner that has altogether
the air of truth. Such a scene would readily find a place on the
Mont Cenis, which descends directly into the broad valley of Susa
and in full view of the plain of the Po : while there is no part of
the Little St. Bernard from which anything else could be seen than
the upper part of the valley of Aosta, nearly 50 miles from its
opening into the plains.

4. The descent of the valley of Aosta, just referred to, appears to
me to present an insuperable objection to the route by the Little
St. Bernard. Polybius states distinctly that from the time when
Hannibal had overcome the difficulties caused by the precipices and
the steep slopes of snow — all of which must have occurred on the
upper part of the pass — he descended in three days' march to the
plains (iii. 56). But as Dr. Arnold, who first appears to have felt

Note A. POLYBIUS. 39

the force of this difficulty, justly observes, no army could, according
to any ordinary rate of marching, get in three days from the Little
St. Bernard to the plains of Ivrea. (Arnold's History of Borne,
vol. iii. p. 481.) The actual distance is not less than 64 English
miles. But moreover the march would have lain for the whole
way through the country of the Salassians, the most untameable
of robbers, who rendered the passage of the valley insecure for an
armed force, even in the days of Csesar. (Strabo, iv. G, § 7, p. 205.)
Yet no allusion is found to any such difficulties, and though we
learn from Strabo that Polybius was familiar with the name of the
Salassians (see the passage quoted in the text) he has never once
mentioned them in connection with Hannibal's passage of the
Alps.

5. In comparison with these more general considerations, I am
not inclined to lay much stress upon any of the details that are
related concerning the march up the valley and the passage of the
actual heights. The general character of the two valleys of the
Isere and the Arc is much the same : and Dr. Arnold, who accepts,
though with considerable reserve, the passage by the Little St.
Bernard, at the same time remarks : " In some respects also I
think Mont Cenis suits the description of the march better than
any other pass." He adds also : " I lay no stress on the Koche
Blanche ; it did not strike me when I saw it as at all conspicuous " :
a remark in which all unprejudiced observers will concur. But
moreover there is no reason to suppose that by the word XevKOTrerpov
Polybius meant to designate any remarkably white cliff, but simply
one of those cliffs of bare white limestone so common both in the
Alps and the Apennines. It is certainly used by him in this
sense in the only other passage in which the word occurs. (Polyb.
X. 30.)

6. In regard to the difficulties caused by the snow in descending
from the summit of the pass (which are described in a very graphic
and characteristic manner) these certainly seem to imply the ex-
istence of a greater amount of snow than is found at the present
day either on the Mont Cenis or the Little St. Bernard. But the
supposition that the snow-line descended in those days to a lower
level than it does at present, is by no means improbable : and a
very small fluctuation in this respect would produce a considerable
change in either of the passes in question. The difference in their
elevation is only about 400 feet.

40 HISTOEY OF ANCIENT GEOGEAPHY. Chap. XVII.

7. But the new theory suggested by Mr. Ellis in 1853, that the
route actually followed by Hannibal was the one now known as
the Little Mont Cenis, — a lateral pass, which deviates from the
high road about 7 miles below Lanslebourg, and rejoins it on the
•plateau at the summit — seems to meet several difficulties and agree
with the details related by Polybius better than either of the alter-
native routes, while it of course possesses all the same advantages
in the argument from general considerations as the well-known
high road over the Mont Cenis. So far as it is possible to judge
without personal examination of its details, this little known pass
seems to meet all the requirements of the narrative of Polybius.

For the earlier literature of the subject I must refer my readers
to the work of Ukert {QeograpMe der Griechen u. Bomer, vol. ii.
pt. ii. pp. 562-566). All the more recent theories have been fully
discussed by Mr. Law, whose elaborate work has nevertheless done
but little to advance our real knowledge. *

NOTE B, p. 33.

WESTEEN COAST OF AFEICA ACCOEDING TO POLYBIUS.

It will be worth while to give the passage of Pliny at full.
After relating the marvellous accounts of Mount Atlas, and
alluding to the commentaries of Hanno, which he had evidently
not himself seen, he continues :

" Scipione .^miliano res in Africa gerente Polybius annalium
conditor ab eo accepta classe scrutandi illius orbis gratia circum-
vectus prodidit a monte eo [Atlante] ad occasum versus saltus
plenos feris, quos generat Africa ; ad flumen Anatim ccccLXXxv
M. P. ; ab eo Lixum ccv M. P., [Agrippa Lixum] a Gaditano freto
cxii M. P. abesse ; inde sinum qui vocatur Saguti, oppidum in pro-
montorio Mulelacha ; flumina Subur et Salat portum Eutubis a
Lixo ccxiii M. P. ; inde promontorium Solis, portum Eisardir,
Gaetulos Autololes, flumen Cosenum, gentes Sclatitos et Masatos,
flumen Masathal, flumen Darat in quo crocodilos gigni. Dein
sinum Dcsvi M. P. includi mentis Baru promontorio excurrente in
occasum, quod appellatur Surrentium ; postea flumen Salsum, ultra

Note B. POLYBIUS. 4 1

quod -^thiopas Perorsos, quorum a tergo Pharusios : iis juugi
mediterraneos Gsetulos Daras. At in ora ^thiopas Daratitas,
flumen Bambotum crocodilis et hippopotamis refertum. Ab eo
montes perpetuos usque ad eum quem Theon Ochema dicemus,
inde ad promontorium Hesperium navig-atione dierum ac noctium x ;
in medio eo spatio Atlantem locavit, ceteris omnibus in extremis
Mauretanise proditum." (Lib. v. c. 1, §§ 9, 10, ed. Sillig.)

The slightest examination of the above extract will suffice to
show the complete confusion in which it is involved. The greater
part of the names are indeed otherwise unknown, but some are
readily recognized, or may be identified with reasonable certainty.
Thus there can be no doubt that Lixus is the well-known town of
the name, which afterwards became a Eoman colony, and occupied
the site of the modern Al Araisch, but the distance given from the
Fretum Gaditanum or Straits of Gibraltar is greatly in excess of
the truth. Here the entirely different statement of Agrippa, so
strangely intercalated in the midst of those of Polybius, is much
more nearly correct, though still considerably in excess. The
river Anatis is otherwise wholly unknown : and it is impossible to
determine what point the author took as the commencement of his
measurements. If the point where the ridge of Mount Atlas first
descends to the sea be supposed to be designated by the words " ab
eo monte," we must fix on Cape Ghir, which is about 400 E. miles
(or 430 Eoman) from Al Araisch : thus falling short of the dis-
tance given by more than 200 miles. Hence M. Vivien de St.
Martin, who has analysed the passage of Pliny with great care,
supposes Cape Noun to be meant, which may be considered as the
last termination of the offshoots of the Atlas : but the point is not
really susceptible of determination. Again the river Darat, in
which crocodiles were found, must doubtless be the same with the
Daradus of Ptolemy, which is still called the Draa, and is the
largest river in this part of Africa. In like manner the Bambotus
is in all probability the same as the large river mentioned by
Hanno (under the name of Chretes) as abounding with crocodiles
and hippopotami, and this, as we have seen, may probably be iden-
tified with the Senegal. The Theon Ochema also doubtless refers
to the mountain of that name mentioned by Hanno : but it is quite
uncertain what headland is designated by the Western Promontory
(Hesperium Promontorium) ; and the statement that Polybius
placed Mount Atlas in this part of Africa is entirely inexplic-

42 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XVII.

able, and seems wholly at variance with his previous statements
concerning it.

If we attempt to apply the few points thus determined to the
explanation of the rest of the passage, we shall find that they only
serve to show more clearly the hopeless confusion in which the
whole is involved. But one thing is clear : that the distances
from the extremity of Atlas to Lixus and the Straits are given
from south to north, while the names and details that follow are
given in order of succession from north to south, though Pliny him-
self had evidently no idea of the difference. This has been clearly
shown by M. Vivien de St. Martin, who has entered into an elabo-
rate examination of the whole passage, and has thrown as much
light on it as it is capable of receiving. (Le Nord de VAfrique dans
VAntiquite, pp. 337-342.)

The probability is, that if we possessed the original narrative of
Polybius, we should find it present as strong a contrast to the above
confused and unintelligible statement, as does the authentic account
of the voyage of Nearchus to the abstract of it given by Pliny.
(See Chap. XIII. Note A, p. 542.)

It is remarkable that in this extract no mention is made of the
island of Cerne, which confirms the inference, that the voyage of
Polybius did not extend so far. But if we can depend on the
accuracy of another passage of Pliny (vi. 31, § 199), Cerne was
mentioned by Polybius, who placed it at the extremity of Mauretania,
opposite to Mount Atlas, a description wholly at variance with its
true position.

( 43 )

CHAPTEE XVIII.

POLYBIUS TO POSIDONIUS.

Section 1. — Progress of Boman conquests.

§ 1. The progress of geographical knowledge from the time
of Polybius to that of Strabo was in great measure dependent
on the progress of the Eoman arms. As province after province,
and kingdom after kingdom, were successively reduced under
the all-absorbing dominion of the great republic, and tribes
that had hitherto enjoyed a wild and lawless independence
were brought under a regular administration, or compelled to
acknowledge fixed boundaries, and render at least a nominal
submission to their powerful neighbour, the regions they
occupied became better known, and assumed a more definite
character in the mind of the geographer. The materials for
the construction of a map, or for that accurate geographical
description of a country which is really impossible without a
map, were still wanting ; but the strong administrative turn of
the Eomans, as well as their habit of constructing high roads
in all the newly acquired provinces of the empire, tended
materially to promote the acquisition of a more distinct and
detailed knowledge of the countries successively added to
their dominions, while they were at the same time continually
carrying their arms farther and farther among the semi-
barbarous nations that encircled their frontiers.

Thus we find that after the time of Polybius the Eomans
gradually extended their conquests across the Alps into the
southern parts of Gaul. Here the Salyans or Salluvians, a
people of Ligurian origin, inhabiting the tract from the Var
to the neighbourhood of Massilia, were the first to succumb to

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

the Eoman arms. In B.C. 125 they were defeated by the
consul M. Fulvius Flaccus ; and two years later (b.c. 123)
their subjugation was completed by the consul C. Sextius
Calvinus, who established in their territory the Eoman colony
of Aquae Sextiee, which has retained to the present day the
appellation of Aix.^ The Vocontii, who adjoined them on
the north, occupying the country between the Durance and
the Isere, soon followed their example, while the AUobroges
— a more powerful people, who held the mountain districts
of Savoy and northern Dauphine — were first defeated by
Cn. Domitius in B.C. 122, and having again renewed the
contest in the following year, with the support of their neigh-
bours the Arverni, sustained an overwhelming defeat from the
consul Q. Fabius Maximus, who assumed the surname of AUo-
brogicus in celebration of his victory.^ The AUobroges were
now reduced to the condition of subjects, but the Arverni were
left in possession of their independence, which they retained
to the time of Caesar. Meanwhile the Romans secured a
footing in Central Gaul by maintaining a steady alliance with
the ^Edui, who occupied the southern portions of Burgundy.

A few years later the Romans extended their dominion
across the Rhone, and the foundation of the Roman colony of
Narbo Martins (Narbonne), in B.C. 118, secured the possession
of this part of the province.^ Tolosa, the capital of the Tec-
tosages, appears to have been at this time on terms of friendly
alliance with Rome, but having afterwards joined the Teutones
and Cimbri, on their irruption into Gaul, was taken and
plundered by the consul Q. Servilius Caepio in B.C. 106;* and
from this time the Tectosages, as well as the more eastern
tribe of the Volcae, the Arecomici, became subject to Roman
rule. The Roman province in Gaal had now become definitely
organized, and had acquired the same limits which it retained
to the time of Caesar.

^ Livii Epit. Ix. Ixi. ; Florus, iii. 2 ;
Veil. Pat. i. 15.
^ Livii Epit. Ixi. ; Florus, I. c.

3 Veil. Pat. i. 15.

* Orosius, V. 15 ; Justin, xxxii. 3.

Sect. 1. POLYBIUS TO POSIDONIUS, 45

In Spain on the other hand the subjection of the Lusi-
tanians after the death of Viriathus (B.C. 140), and the reduc-
tion of Numantia by Scipio Africanus, had already brought the
greatest part of the peninsula under the Eoman dominion at
an earlier period ; the wild tribes that inhabited the mountains
in the north — the Cantabrians, the Asturians, and Gallicians —
alone retaining their independence, which they preserved
almost unimpaired till the time of Augustus.

§ 2. The Dalmatians, on the east coast of the Adriatic, were
defeated and reduced at least to nominal submission by
L. Metellus in B.C. 119 ; the lapydes or lapodes, also an lUyrian
people, situated in the modern Croatia, had been already
defeated by the consul Sempronius Tuditanus ten years before
(B.C. 129) ; neither people however became really subject to
the Eomans before the time of Augustus. The Scordiscans,
who are called by some Eoman writers a Thracian people, but
were more probably a Celtic race, settled at this time in the
south of Pannonia,^ first came in contact with the Eoman
arms as early as B.C. 175 ; and again in B.C. 135 ; but they
attracted little attention till B.C. 114, when they inflicted a
severe blow on the Eoman arms, having defeated the consul
C. Porcius Cato and destroyed his whole army ; after which
they extended their ravages over the whole of Macedonia and
Thessaly, until they were first checked by T. Didius, and
ultimately driven across the Danube by the consul M. Livius
Drusus in B.C. 112.® "We however find them again mentioned
a few years afterwards, in conjunction with the Triballi — an
old name that here reappears after a long interval — as carrying
on hostilities within the limits of Thrace.^ The wild tribes
that inhabited that country were indeed still unsubdued, and
continued for more than a century afterwards to trouble the
Eoman governors of Macedonia, or give them occasion to dis-
tinguish themselves by military successes. But C. Scribonius

^ Concerning these Celtic tribes in I * Liv. Epit. Ixlii. ; Eutrop. iv. 24.
Pannonia and the neighbouring regions, ' Eutrop. iv. 27.

Sec Sect. 6, Note E, p. 105. '

46 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

Curio in B.C. 75 appears to have been the first Eoman general
that penetrated to the banks of the Danube.^

§ 3. While the Eomans were thus extending the limits of
geographical knowledge in Europe, they had as yet made no
progress in Asia beyond the regions already well known to
the Greeks. Nor had the latter been able on their side to
contribute any important additions to the knowledge already
available in the time of Eratosthenes and his immediate suc-
cessors. In this direction indeed the course of events had
been decidedly unfavourable to the advance of geographical
science. That Hellenization of a large part of Asia, which
had followed so rapidly upon the conquests of Alexander
that it had appeared likely at one time to include the whole
continent from the Mediterranean to the Indus within the
domain of Greek knowledge and civilization, had met with a
severe check from the disruption of the Seleucidan . empire
through the revolt of the upper provinces. The kings of
Bactria, Ariana, and the provinces adjoining the Indus, were
indeed all of Greek origin,^ and probably did their best to
maintain and encourage the surviving remains of Greek civili-
zation within their dominions. Nor can it be doubted that
if they had retained unbroken connection with the more
westerly provinces of the Syrian monarchy, they would have
been the means of materially extending the knowledge pos-
sessed by the Greeks of these regions of Upper Asia, and
even of India itself. It appears certain that Menander, a
Greek monarch who reigned in the regions of the Paropamisus
(apparently about 160-140 b.c.),^ had not only established his
dominion over the whole of the Punjab, and perhaps the lower
valley of the Indus also, but had carried his arms beyond the
Hyphasis — the limit of Alexander's conquests — as far as the

* Eutrop. vi. 2. I ' The date here given is that assigned

8 This is evident from the purely j to the reign of Menander by General

Greekcharacterof their names, as found l Cunningham; but it must be confessed

on their coins, and which include such that this, in common with almost all

well-known names as Lysias, Diomedes, the other dates of the Bactrian kings,

Menander, Plato, Demetrius, &c. [ is in great measure conjectural.

Sect. 1.

POLYBIUS TO POSIDONIUS,

banks of the Jumna, if not even farther.^ The extensive trade
carried on by these Greek rulers with the adjoining regions
of India is confirmed by the curious fact that more than two
centuries later the silver coins of Menander and ApoUodotus —
another monarch of pure Greek extraction— still formed the
ordinary currency at Barygaza and other Indian ports.^ To
the discovery of these and similar coins in our own days we
are indebted for reviving our knowledge of the extent and
long duration of this detached fragment of the Hellenic
world, concerning which we find but few and scattered notices
in the extant historians.*

§ 4. It was unfortunate that the rise of the Parthian
monarchy, almost simultaneously with that of the Graeco-
Bactrian kings, and its rapid extension over the provinces of
the table-land of Iran, until it absorbed the fertile regions on
the banks of the Euphrates and the Tigris,^ had the effect of
entirely cutting off the more distant Greek settlements from
the Hellenized regions of Syria and Asia Minor. Hence
doubtless arose the very imperfect knowledge apparently
possessed by Greek writers of these outlying districts, where
a certain amount of Greek civilization still lingered ; and the
absolute want of any additional geographical information
derived from this source.^

^ Strabo, xi. p. 516. According to
his statement (apparently derived from
Apollodorns of Artemita) Menander
crossed the Hypanis, by which he
evidently means tlie Hyphasis, and
advanced as far as the Isamus. Unfor-
tunately the last name is unknown,
and its identification with the lomanes
or Jumna is merely conjectural.

^ This fact is stated by the anony-
mous author of the Periplus of the
ErythrsBan Sea (§ 47). That curious
document will be fully examined in a
future chapter.

* See Note A, p. 102.

* The reduction of Babylonia and the
adjacent provinces by Mithridates I.,
king of Parthia (Arsaces VI.), was com-
pleted, according to Professor Rawlin-
son, before b.c. 150 (Rawlinson's Sixth
Oriental Monarchy, p. 77).

" It is singular that Strabo speaks of
the extension of the Parthian monarchy
as one of the sources of increased geo-
graphical knowledge in his day (ii. 5,
p. lis). Its real effect appears to have
been precisely the contrary.

HISTOEY OF ANCIENT GEOGEAPHY.

Chap. XVIII.

Section 2. — Greek Writers.

§ 1. But though the Greeks can hardly be said to have
made any positive contributions to the extension of geogra-
phical knowledge during this period, there were several writers
on geographical subjects, whose names merit a passing notice,
and some of them at least possess the more importance in our
eyes from the preservation of portions of their works down
to modern times. Perhaps the first in order of time among
these — though his age is not exactly known — was Apollo-
DOEUS, a grammarian of Alexandria,^ and a voluminous writer
upon various subjects, of which however the only one that has
been preserved to us is his well-known mythological treatise.
He was also the author of a commentary on the Catalogue of
the Ships in the Iliad, in twelve books : a work which appears
to have been in part of a geographical character — as it could
not indeed well avoid — but mixed with much matter of a his-
torical or mythological description. Our knowledge of it is
derived almost entirely from Strabo, who repeatedly refers to it
by name, though more often to censure than to praise ; but there
can be no doubt that he made extensive use of it, where he
does not acknowledge his obligation — a large part of his own
work being occupied with discussions and examinations of the
Homeric Catalogue, similar to those which must have con-
stituted the bulk of his predecessor's treatise.

In regard to the general principles which he applied to the
investigation of the Homeric geography Apollodorus showed a
sounder judgement than Strabo is willing to allow, having
adopted the same view with Eratosthenes, that Homer, while
showing an accurate and minute knowledge of the geography,
and even the topography, of Greece itself, and the neighbour-

' Apollodorus was a native of Athens,
but he studied under the celebrated
grammarian Aristarchus, and may,
therefore, be regarded as belonging to
the Alexandrian school. His age
cannot be determined ■with accuracy,
but his great chronological work — ^com-

posed, like his Tris -/reploSos, in iambic
verse^ — ended with the year 145 B.C.,
and was dedicated to Attains II. Phila-
delphus, king of Pergamus, who died in
138 B.C. (See Clinton's F. H. vol. iii.
pp. 105, 119.)

Sect. 2. POLYBIUS TO POSIDONIUS. 49

hood of Troy, was almost wholly ignorant of the more distant
regions of the world ; ^ and that it was idle to bring to the test
of geographical accuracy such passages as those concerning
the Ethiopians or the wanderings of Menelaus. He rejected
also the commonly received identification of many of the
localities mentioned in the Odyssey, such as that of Gaulos
with the island of Calypso, and Scheria with Corcyra; and
considered that the poet had intentionally transferred the
wanderings of Ulysses to the shores of the unknown Ocean, in
order that he might be at liberty to indulge in poetic fictions
without restraint.^ But besides this work, which we find fre-
quently cited, and which appears to have become a kind of
standard authority on the subject, Apollodorus also composed
a formal geographical treatise, in iambic verse, to which he
gave the name of r^9 TreptoSo?.^ It contained a regular descrip-
tion of the three continents,^ and was probably in many
respects similar to the poetic treatise ascribed to Scymnus
Chius, to which we shall have occasion to revert hereafter:
but we have very little information concerning it : the cita-
tions in Stephanus of Byzantium, though numerous, being
confined almost entirely to mere names.

§ 2. Nearly contemporary with Apollodorus was Demetrius
OF Scepsis,^ who wrote an elaborate treatise, in not less than
thirty books,* upon the catalogue of the Trojan allies, as pre-
served in the Iliad. This is frequently referred to by Strabo,
and evidently contained a considerable amount of geogra-
phical information, though the greater part of so voluminous
a work must have been occupied with historical and mytho-

« Strabo, vii. p. 298.

9 lb. p. 44, vii. 3, § 6, pp. 298, 299.

1 Strabo, xiv. p. 677.

2 See the fragments of it collected by
C. Miiller in his Fragmenta Histori-
corum Gnecorum, vol. i. p. 449.

^ According to Strabo (xiii. 1, § 55),
Demetrius was a contemporary of Crates
and Aristarchus, which would make
him somewhat senior to Apollodorus,

VOL. II. E

He was a boy, or quite a youth (fieipa-
Kiov), at the time when the Romans
first crossed over into Asia, B.C. 190
(Id. xiii. 1, § 27). His work was pro-
bably not composed till long afterwards.
But it would appear to have been pub-
lished hejore that of Apollodorus, who
is Siiid to have borrowed largely from
him (Strabo, viii. p. 339).
^ Strabo, xiii. 1, p. 603.

HISTOKY OF ANCIENT GEOGRAPHY. Chap. XVIII

logical disquisitions.^ Its principal interest in modern times
arises from the author having been the first to raise doubts as
to the true site of the Homeric Ilium : a question upon which,
as Strabo observes, the circumstance of his birth-place having
been situated within a few miles of the localities, as well as the
pains he had bestowed upon the subject, entitled him to much
consideration. Hence his views were adopted by Strabo, and
have received the assent of many scholars in modern times, so
far at least as relates to the rejection of the claims of the
Ilium of his day to represent the Homeric city.® In his
general views on Homeric geography, and the extent of the
poet's knowledge, Demetrius appears to have agreed with
Eratosthenes and Apollodorus, and consequently incurred the
censure of Strabo for a want of due reverence for the poet's
authority. In particular he denied that Homer had any know-
ledge at all of the voyage of Jason to the Phasis:"^ a conclusion
in which most modern critics would agree with him.

§ 3. A writer whose works had much more important bearing
upon geography, properly so called, was Agathaechides, a
native of Cnidus, who was the author of several historical and
geographical treatises of considerable interest, known to us
chiefly from their mention by Photius. Among these we find
included a history of the affairs of Asia in ten books, and of
those of Europe in not less than forty -nine : besides which he
composed a separate treatise in five books concerning the Red

' This is apparent from the passages
cited, the greater part of which refer
to subjects of this class. But from
some of those quoted by Strabo it is
evident also that he went fully into
topographical details (see the references
to his work collected by Clinton, F. B.
vol. iii. p. 527, note ; and by C. Miiller,
in a note to his Fragm. Hist. Grxo.
vol. iv. p. 382). It is indeed almost
certain that the full topographical
account of the Troad, given by Strabo
in his great work (xiii. c. 1), was de-
rived principally, if not entirely, from

Demetrius.

•^ Hardly any modern writer has in-
deed adopted the view of Demetrius in
identifying the 'IxUccv Kwfxf) with the
heroic Ilium ; but the theory of M,
Chevallier, placing the site above
Bunarbashi, would scarcely have been
so hastily embraced in recent times,
had not the scepticism of Demetrius
and Strabo shaken the traditional faith
in tlie historic site.

aTro57]iJ.Lav rov 'laffovos" Oixijpov. Strabo,
i. 2, § 38, p. 45.

Sect. 2.

AGATHAECHIDE3.

Sea and the nations adjoining it.^ How far the first two works
were of a historical and how far of a geographical character we
are unable to determine with certainty, very few fragments
having been preserved to us ; but Photius has fortunately
transmitted to us an abstract of two whole books of the treatise
on the Eed Sea, from which we learn also that the account of
the countries and nations adjoining it, which we find in Dio-
dorus, is derived entirely from the same source : and the two
abridgements thus serve to supplement one another. Aga-
tharchides (as we learn from Photius) passed the latter years
of his life at Alexandria, where he enjoyed the important
position of tutor to the young king, Ptolemy Soter II. (about
116 B.c.),^ and he had thus every opportunity of acquiring the
most authentic information concerning the regions in ques-
tion. Of these he seems to have availed himself with diligence
and judgement ; and the fragments of his work, notwith-
standing the imperfect form in which they have been trans-
mitted to us, are undoubtedly among the most valuable of the
minor geographical writings that remain from antiquity.^

§ 4. The few extracts that are preserved from the first book
relate only to the capture of elephants and the arrangements
made for that purpose by the Ptolemies, and to a mythological
discussion, of very little interest, concerning the origin of the
name of the Erythraean Sea. Of the contents of the second,
third, and fourth books we have no information, but they appa-
rently contained an account of the Ethiopians, properly so
called, from which it is probable that the interesting account

8 Photius, Bibliotli. cod. 213, p. 171,
ed. Bekker. The abstracts of the two
Looks Trep! TTjs ipvdpas QaXacra-qs are
given in cod. 250, pp. 441-460.

^ For the date, and the determina-
tion of the Ptolemy to whom he thus
acted as tutor, upon which the whole
chronology of his life depend.«, see the
elaborate investigation by C. Miiller in
his edition of the Geographi Grxci
Minores, torn. i. Prolegomena, pp. liv-
Iviii.

' They are published (from the ex-
tracts given by Photius) in Hudson's
edition of the Geographi Graici Minores,
vol. i. ; but by far ilie best edition is
that given by C. Miiller {Geogr. Graiei
Minores, tom. i.), who has printed the
extracts as given by Diodorus parallel
with those given by Photius, so that
the reader can at once compare the
two ; and has added alao some valuable
notes.

E 2

HISTOKY OF ANCIENT GEOGRAPHY. Chap. XVIII.

of their manners and customs, whicli we find in Diodorus, was
for the most part derived.^ But it is with the beginning of
the fifth book that the really valuable portion of our extracts
commences. After giving a very curious and interesting
account of the gold mines, which were still worked in his day
on the borders of Egypt and Ethiopia, near the Eed Sea,^ he
proceeded to describe the habits and manner of life of the
Ichthyophagi, a tribe occupying the western shores of the Red
Sea, and who were, as he justly remarks, identical in these
respects with similar tribes that extended along the coasts of
Arabia, Carmania and Gedrosia to the frontiers of India.
These Ichthyophagi were, as he describes them, a race in the
lowest state of civilization ; living wholly upon fish, but not
possessing boats, or even nets, with which to catch them, and
dependent therefore upon what were left upon the shore by the
receding tide. They went entirely naked, and had their wives
and children in common : they were said to possess no idea of
good and evil, and to show a remarkable apathy and indif-
ference to danger or pain. It was said also that they drank
only every fifth day, when they repaired to the few sources of
water that were to be found at the foot of the nearest moun-
tains : and some tribes were even reported never to drink at
all. Some of them dwelt in caves, others formed rude huts
with the bones of the largest fishes, covered with sea-weed :
and others again formed holes or dens in the great masses of
sea-weed that were accumulated on the shore during heavy
gales.^

In connection with these Ichthyophagi he mentioned another
tribe, to whom the Greeks gave the name of Chelonophagi,
from their subsisting almost entirely upon the turtles which

^ Diodor. iii. c. 2-10. We cannot,
however, assume that this account is
wholly taken from Agatharchides, as
Diodorus claims to have derived his
information in part from Artemidorus,
and also to have had personal inter-
course with Ethiopian deputies during
the time of his stay in Eg^'pt (lb. c. ii.j.

* Agatharchid. § 23-29, ed. Miiller;
Diodor. iii. 12-14. These gold mines
were situated in the mountains near
Cosseir, now occupied by the Ababdeh
Arabs ; but they have long ceased to be
worked.

^ Agatharch. §§ 31-46 ; Diodor. iii.
15-20.

Sect. 2. AGATHAECHIDES. 53

abounded in these seas : the shells of which also served them
by way of roofs under which to shelter themselves, as well as
occasionally as boats with which to cross the sea, for short
distances.^ This tribe however did not dwell near the shores
of the Eed Sea, but inhabited a group of small islands, ad-
joining the coasts of Carmania and Gedrosia, which fronted
the Indian Ocean.

§ 5. Agatharchides next proceeded to describe the various
tribes of Ethiopians that dwelt inland, beyond the regions
occupied by the comparatively civilized race that held the
island of Meroe and the district immediately south of Egypt.
Here he first mentioned the Rhizophagi (Root-eaters), who
dwelt on the banks of the Astaboras (Atbara) above its con-
fluence with the Nile, and subsisted, as their Greek name
implied, mainly on the roots of reeds and other water plants
growing in the marshes.^ Adjoining these were the tribes called
Hylophagi and Spermatophagi, who fed not only on the fruits,
but even devoured the leaves and young shoots of trees, which
they climbed for this purpose with incredible agility.'' Next
to these came the Hunters (Cynegetae, called by other writers
Gymnetes), who inhabited a region abounding in wild beasts,
so that they were compelled always to sleep in trees. They
were excellent archers, and were thus able to slay even wild
cattle, panthers, and other formidable wild beasts, for which
they lay in wait as they quitted their watering-places.^ Again
to the west of these were the race distinguished as Elephanto-
machi or Elephantophagi, from their subsisting almost entirely
on the flesh of the elephants which they killed, with no other
weapon than a sharp axe, with which they hamstrung the
huge beasts. So devoted were they to this pursuit that
Ptolemy had in vain endeavoured to induce them to abandon
it, in order to assist his hunters in taking the elephants
alive.^

Agatharch. § 47; Diodor. iii. 21.
Agatharch. § 50 ; Diodor. iii. 23.
Agatbarch. § 51 ; Diodor. iii. 24.

8 Id. § 52 ; Diod. iii. 25.
' lb. § 53-56 ; Diod. 26.

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

Beyond these to the west were a race of Ethiopians called
Simi (flat-nosed) and towards the south were a people called
the Struthophagi (ostrich-eaters), from the ostriches which
abounded in their country, the chase of which supplied them
with their principal means of subsistence.^ Not far from these
were the Acridophagi or Locust-eaters, a very black people,
who supported themselves to a great degree upon the swarms
of locusts that periodically visited their country.^ An ad-
joining region, though fertile and abounding in pastures, was
said to be uninhabited on account of the multitude of scor-
pions and large spiders with which it swarmed.^ The remotest
people of all towards the south were the so-called Cynamolgi
or Canimulgi, a very barbarous race, but who kept numbers of
large dogs with which they hunted down the wild cattle that
roamed in vast herds over their territory.^

§ 6. The geographical position of these races is in general
very obscurely indicated, and was probably but imperfectly
known to Agatharchides himself; and the names by which
they are described were obviously of Greek invention, not
native appellations.^ But the notices of .their manners and
habits are very curious, and have been to a great extent
confirmed by the observations of modern travellers — Bruce,
Burckhardt, and others ; though it is probable that they never
were so strongly characteristic of different tribes as they were
regarded by the Greek geographers. The same statements
were copied by Artemidorus, and from him again by Strabo,
and are repeated by Pliny, .i^Elian and other later writers. In
fact it appears probable that the work of Agatharchides was
the original source of all the information possessed concern-
ing the Ethiopian tribes of the interior by either Greek or
Koman writers down to a late period.

' Agatharch. § 57 : Diod. iii. 28.
" Id. § 58 ; Diod. iii. 29.
3 Id. § 59 ; Diod. iii. 80.

* Id. § 60 ; Diod. iii. 31.

* Some of tliese are again found at a
later period in Ptolemy, who has of

course given them a more definite posi-
tion, but it is very doubtful how far we
can rely upon the names given to such
fluctuating and barbarous tribes being
applied to the same people at so long
an interval.

Sect. 2. AGATHARCHIDES. 55

He was also the first to recount many curious particulars
concerning the wild animals that were found in these remote
regions, and with which the hunting expeditions of the Ptole-
mies had brought the G-reeks of Alexandria into acquaintance.
Thus he described the camelopard, the ostrich, the rhinoceros,
and several species of apes or baboons, to which he gave the
name of Cynocephali, Cepi and Sphinxes; the Crocottas or
laughing hyasna, and the wild bulls, which he erroneously
supposed to be carnivorous, but justly described as animals
of terrible ferocity.*^ After speaking of the huge serpents Avith
which these countries were said to abound he gave a very
curious account of one which had been captured and brought
alive to Alexandria, where he had himself seen it, and which
was thirty cubits in length/

It is remarkable that among all these notices of the remote
Ethiopian tribes we find no trace of the supposed abundance
of gold among them which bears so conspicuous a part in the
fables current concerning them in the time of Herodotus ; an
omission the more singular because gold is really found in
considerable quantities in the neighbourhood of the Upper
Nile, and gold mines were actually opened under Mehemet
Ali at a place called Fazoglo in Sennaar.^ Nor do the old
fables concerning the Pygmies, the men with huge ears, and
other similar tales appear to have found a place in the pages
of Agatharchides. The extension of more accurate information
was gradually displacing all such fictitious creations; it was
found at least that they did not exist in Ethiopia.

§ 7. Eeturning from these remote regions towards the north,
and the countries bordering on the Red Sea, Agatharchides
next proceeded to describe the Trogiodytfe, a people inhabiting
the mountain ranges that border the Eed Sea on the west, at
more or less distance from the coast, throughout almost its

« Agatharcli. § 76 ; Dioil. iii. 35. i was found that the reality fell far

' Agatharch. § 78; Diodor. iii. 3G, short of the rumours current concerning

j7. them; and they were soon ahaudoucd.

* As usually happens, however, it |

56 HISTORY OP ANCIENT GEOGRAPHY. Chap. XVIII.

whole length.^ Though less barbarous than the Ichthyophagi
of the coasts, the Troglodytes were still in a very rude con-
dition ; they dwelt principally in caves, — from whence their
name — went almost naked, and had their wives in common ;
but they had extensive flocks and herds, of which they drank
the milk and blood, as well as consumed the flesh. It was
their custom to put to death the aged and infirm, as well
as those afflicted by any protracted disease. Their habits of
life were doubtless determined in great measure by local con-
ditions ; and those of the mountain tribes who inhabit the same
ranges at the present time are still very little removed from
barbarism. But ancient writers were chiefly impressed with
their habit of dwelling in caves, and hence applied the name
of Troglodytes to the people adjoining the western shores of
the Eed Sea from the frontiers of Egypt to the Straits of
Bab-el-Mandeb.

After this general description of the habits of the Troglo-
dytes Agatharchides appears to have given in some detail a
more particular account of the western coast of the Eed Sea
or the Troglodytic coast, as it was generally termed, not,
however, — so far as we can judge from our existing abstracts —
giving a regular Periplus or enumeration of the ports, islands
and headlands, but merely noticing the more interesting
and striking natural phenomena. Thus he mentioned the
hot springs near Arsinoe ; the Scarlet Mountain near Myos
Hormus,^ the extensive bay called Foul Bay (acoXtto? 'A/ca-
dapro^) from the rocks and shoals with which it abounded,
and the island of Topazes, celebrated for the gems of that
name;^ but did not even notice the important port of Bere-
nice in the same neighbourhood. From thence to Ptolemais
Epitheras (the great hunting-station of the Ptolemies) the sea

8 Agatliarch. §§ 61-63; Diodor. iii. I /^jAraiSes), see Note A, Chapter XV.

32, 33. p. 607.

' Concerning the position of this ^ Agatharch. §§ 81, 82 ; Diod. iii.

celebrated port, and the bright red 39.

mountains that serve to identify it (ppos \

Sect. 2.

AGATHAECHIDES.

was said to be very shallow and abounding in sea-weed and
sand-banks, so as to be very difficult of navigation for the
large ships that were required for the transport of the ele-
phants. But beyond Ptolemais the coast trended towards the
east, while the sea became deep and open, and presented no
difficulties to the navigator, though abounding in huge fish
and sea monsters of various kinds. The adjoining tracts were
traversed by rivers, which took their rise in the Pseboean
mountains ^ — a name by which the author must have meant
to designate the mountains of Abyssinia, though he applies
the same name to the promontories -that close in the mouth
of the galf, and form the Straits of Bab-el-Mandeb.* It is
singular that he appears to have given no particulars con-
cerning these straits, or the cinnamon and spice producing
lands beyond them.

§ 8. Instead of this he returned (according to our existing
abstracts) to the head of the Gulf of Heroopolis or the inmost
bight of the Ked Sea, and proceeded to describe in con-
siderable detail the Arabian coast of that sea.^ Many of
the points mentioned may be recognized without difficulty,
though his description, at least as transmitted to us, is un-
accompanied with any statement of distances. Thus the spot
named Phcenicon, or the palm-grove, on the Gulf of Heroo-
polis, may be safely identified with the modern Tor, still famous
for its groves of palm-trees ; the headland that separates the
Gulfs of Suez and Akabah (now called Eas Mahomed) with
the small island lying off it, called by the Greeks the Island of
Seals, are clearly described ; the barren and precipitous rocks
that bound the coast for a considerable distance in one part,
the swarm of small islands in another, and the marshy and
wooded tract that succeeds, are found well to agree with the
observations of modern voyagers.® The names of the different

8 Agatharch. § 84; Diodor. iii. 41.
" Diod. ibid.

« Agatharch. §§ 85-94 ; Diodor. iii.
42-45.

« Agatharchid. §§ 85-94; Diodor.
iii. 42-44. The same particulars are
repeated from Artemidorus by Strabo,
xvi. 4, § 18.

HISTORY OP ANCIENT GEOGRAPHY. Chap. XVIII.

Arabian tribes are of comparatively little interest, though that
of the Thamudeni is still to be traced in the modern Thamud,
a tribe dwelling in the northern part of the Hedjaz. The
Debffi, who dwelt along the coast in the southern portion of
the same region, were noted for Iheir extensive herds of camels,
their territory was said also to abound in gold, which was found
also in great quantities in that of the Gasandse or Gasanden-
ses, further towards the south.^ No gold is now known to
exist in these regions.

§ 9. South of these tribes, occupying the south-western corner
of Arabia, were situated the Sabseans, a nation of a very differ-
ent character. It is especially to Agatharchides that we
are indebted for the glowing descriptions of the wealth and
prosperity of this people, which gave origin to the name of
Arabia Felix applied to this portion of the peninsula, and
which continued to be repeated by Greek and Eoman writers
down to a late period.^ The territory of Yemen is indeed
naturally the most fertile part of Arabia, but it was the ex-
traordinary abundance of spices that constituted the great
wealth of the country. Such was the quantity of these that
we are told the inhabitants were debilitated by the excess of
fragrant odours, and compelled to burn bitumen and goats'
beards in order to counteract their influence (!). The sea-coast
produced balsam, cassia, and another fragrant plant, which was
not exported, as its perfume did not last ; in the interior were
forests of frankincense, myrrh, cinnamon and other odoriferous
trees, mixed with palms and tall reeds.^ But besides the spices
thus produced in their own country, they imported them in
large quantities from the opposite coast of Africa (the Regio
Ciunamomifera),^ and again exported them to the surround-

' Agatbarchid. §§ 92-96 ; Diod. iii.
45.

8 Id. §§ 97-102 ; ap. Diod. cc. 46, 47 ;
Artemidor. ap. Strab. xvi. 4, § 19 ;
Plin. R. N.

s These forests were said to swarm
With serpents of very small size, but
whose bite Avas deadly (lb. § 98). They

are probably the same as the " winged
serpents " described by Herodotus (iii.
108) as abounding in these regions,
and interfering with the gatliering of
frankincense.

' It is remarkable that they are de-
scribed as traversing the sea which
separated them from this part of Africa

Sect. 2.

AGATHAECHIDES,

ing nations. The great wealth of the country was in fact
derived mainly from their trade in these commodities, which
was considered as the most lucrative known, and having
been carried on by the Sabseans for centuries past had been
the means of accumulating enormous quantities of gold and
silver in the country. Hence they had magnificent palaces,
resplendent with gold, silver, and precious stones ; they drank
from goblets of gold, studded with gems, and even their seats
and couches were overlaid with silver. Their capital city
was called by Agatharchides Saba, but it appears, as we learn
from Artemidorus, to have also borne the name of Mariaba,
which is still retained in the modern name of Mareb, the
ruins of which are still extant about 80 miles E. of the present
capital of Yemen.^

§ 10. Besides this trade their ports on the Indian Ocean
were the resort of numerous ships from the neighbouring
countries to the east, as far as the Persian G-ulf, together with
many traders from the mouths of the Indus :^ and they thus
became the principal entrejpot of the trade with India. We have
already seen that it was probably through this channel that
the Alexandrian merchants obtained their Indian commodities.
The stories concerning the wealth of the Sabseans, as well as
others concerning their luxurious habits of life, are evidently
gross exaggerations, but there is no doubt that the country
was really at this time the seat of an extensive and flourishing
commerce, which was carried on partly by sea, but to a con-
siderable extent also by land caravans passing through the
territory of the Minnaeans (the Hedjaz) to the city of Petra,
which had already become a great emporium for the trade from
all parts of Arabia.* Thither also repaired caravans from

in vessels made of liides (Sep/xarlvois
ttAoIols), apparently similar to those in
use among the Britons.

^ Artemidorus ap. Strab. xvi. 4,

§19.

^ Agatharchid. § 103. These traders
are dtsuribed as coming "from the

place where Alexander foundixl a naval
station by the river Indus." Tiio name
is given in Diodorus (c. 47) as Potana :
probably Pattala is meant, thougli that
place was not really founded by Alex-
ander.

* Agatharchid. § 87 ; Diodor. iii. 42,

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVIII.

Gerrha, which had become the chief emporium on the east side
of Arabia, so that its inhabitants were said to rival the Sabseans
in opulence.^

No geographical indications are preserved by Agatharchides
(to judge from our existing extracts), of the outer coast of
Arabia, bordering on the Indian Ocean, and the vague notice
of certain islands to the south, of great fertility which may
possibly refer to Socotora and the neighbouring islets, is of the
most indefinite character. But it seems probable that an
allusion at the end of his work ® to some islands recently dis-
covered in this part of the world must refer to those just men-
tioned. His account of the astronomical phenomena observed
in these southern regions is in great part inaccurate and
absurd, and can only have been gathered from ignorant navi-
gators, who did not themselves understand the phenomena
which they reported.^

Upon the much controverted question of the cause of the
inundations of the Nile, Agatharchides entertained sound
views, attributing them to heavy and continuous rains taking
place in the mountains of Ethiopia, from the summer solstice
to the autumnal equinox ; a fact which was attested, as he
observes, by the natives of the neighbouring regions.^

§ 11. One other passage of Agatharchides^ deserves notice
from the light it incidentally throws upon the navigation of
his day. After pointing out the contrast between the Ethio-
pians in the south, and the Scythians in the north, resulting
from the extremes of climate in the two cases, he adds that,
after all, the intervals which separate these extremes are not
so very great. For that many persons sailing with vessels
carrying cargoes, and having a favourable wind, would reach
Khodes in ten days, from the Palus Mseotis : from thence it
was only four days' sail to Alexandria, and ten days' voyage up

* Agatharch. §§ 87, 102.

" Id. § 110.

' Id. § 104 ; Diod. iii. 48. But we
have seen that the same thing was the
case to a considerable extent even with

the accounts given by Onesicritus and
Nearchus as they have been transmitted
to us.

» Id. § 112.

» Agatharch. § 66 ; Diod. iii. 34.

Sect. 3.

ARTEMIDOKUS.

the Nile from thence would suffice in many cases to reach the
confines of Ethiopia. So that a continuous voyage of twenty-
four days was sufficient to pass from the coldest regions of the
world to the hottest. The shortness of the time here allowed
for the ascent of the Nile is remarkable, and could certainly
only apply to very exceptional cases. But the statement
concerning the time in which the voyage was frequently
made from the Palus Mseotis to Alexandria is curious and
instructive.

Section 3. — Artemidorus.

§ 1. Nearly contemporary with Agatharchides was Arte-
MiDOEUS, a native of Ephesus, but who, like most of his
contemporaries, studied and wrote at Alexandria. We are
told that he flourished in the 169th Olympiad (b.c. 104-101 ^),
and it is certain that his geographical work was not published
until after that of Agatharchides, of which he made great use.
His principal work was a general treatise on geography, in
which, however, special attention was devoted to the countries
bordering onifche Mediterranean and Euxine Seas, of which it
contained a full and detailed Periplus.^ Artemidorus himself
had examined a large part of the shores of the Mediterranean,
and had even visited Gades and the Sacred Promontory, which
he described from personal inspection.^ In regard to the west
of Europe generally he appears to have collected much valuable
information, and is frequently cited by Strabo as correcting

^ This is the statement of his epito-
mizer Marcian of Heraclea. 'Apre/xl-
Sciopos Se 6 'E^effios yedoypa^os Kara, r^v
€KaTO(TTT]v €^7}K0<Tr7iy evvdr7\v 'OAv/jLiridSa
yeyovcis, p. 65, ed. Hudson.

^ Marcianus of Heraclea, who had
composed an epitome of his work, speaks
of it as if it were only a Periplus of the
Mediterranean, though extending to
eleven books. But it is certain that it
contained a description of other regions

besides. Stepbanus of Byzantium cites
his statements concerning Taprobane,
which were found in his ninth book
(s. V. Tairpo^avT]).

^ Id. ibid. pp. 64, 65 ; Strabo, iii. p.
137. He correctly stated that the
Sacred Promontory (Cape St. Vincent),
which was placed by Eratosthenes at
five days' voyage from Gades, was not
in reality distant from that city more
than 1700 stadia (Strabo, iii. p. 148).

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

the errors of earlier writers, such, as TimEeus and Eratosthenes,
though he fell into others in his turn.* Throughout his work
indeed Strabo appears to have made use of Artemidorus as
one of his principal authorities : but as usual he only cites
him in peculiar and exceptional cases, where he has occasion
either to censure his statements, or refer to them as super-
seding those of his predecessors. In the absence of the
original work it is impossible for us to judge of the advance
really made by Artemidorus in the knowledge of Western
Europe, as compared with Polybius on the one hand, and with
his successor Posidonius on the other. But it is probable that
he still possessed very imperfect information concerning the
external, or Atlantic shores of Spain and Gaul, as well as still
more of Britain. Marcianus of Heraclea, a late writer, describes
Artemidorus as a very valuable authority for the geography of
the Mediterranean, his Periplus of which he regarded as so
accurate that he himself composed an abridgement of it.^
But his knowledge of the external ocean, both towards the
west and the east, is justly regarded by the same authority as
very vague and confused. With regard to the Ked Sea indeed
and the nations adjoining the Indian Ocean, as well as the
Ethiopian tribes of the interior, Artemidorus seems to have
done little more than copy Agatharchides, his account of them,
which has been preserved to us by Strabo,*' being in great part
talven almost word for word from the earlier writer. He, how-

4 strabo, iii. pp. 137, 148, 159 ; iv.
pp. 183, 185, 198.

' A fragment supposed to have
formed part of this abridgement will
be found in Hudson's GeograpM Grxci
Minores, torn. i. pp. 60-74, and is in-
cluded in the valuable edition by M.
Miller of the works of Marcianus of
Heraclea and some other minor geo-
graphers (8vo. Paris, 1839), as well as
in the collection of the GeograpM Grxci
Minores by 0. Midler (tom. i. pp. 563-
572). But it seems certain, as pointed
out by the last editor, tliat the extant
fragment does not belong to the
abridgement of Artemidorus, but to

that of Meuippus, a writer of the
Augustan age, of whose work Marcianus
had also made an epitome.

« Strabo, xvi. 4, §§ 5-19, pp. 709-778.
No mention is made of Agatharchides
by Strabo in conned ion with this de-
scription, though he elsewhere (p. 779)
cites him at second hand in respect to
the origin of the name of the Eed Sea.
It is evident that he had never seen
his original work, and knew only that
of Artemidorus. The two are printed
side by side by C. Midler, from whence
it is readily seen how closely the one
is copied from the other.

Sect. 3.

AETEMIDORUS.

ever, added a regular Peripliis of the Eed Sea, giving the
distances from port to port, and from headland to headland,
which are not found in our existing extracts from Agatharchides,
and are a valuable addition to our geographical knowledge^
He described also in detail, though without giving distances,
the coast of Africa from the Straits and the Promontory of
Deire to the Southern Horn (Noti Keras), under which name
he designated the eastern extremity of Africa, now known as
Cape Gruardafui.® He was aware that from thence the coast
trended away towards the south, but in this direction, he adds,
nothing more was known of it.^ On this side therefore
geography had made no progress since the days of Era-
tosthenes.

§ 2. In regard to India on the other hand Artemidorus
appears to have followed inferior authorities, and his account of
that country is treated by Strabo as superficial and inaccurate.
The statement quoted from him concerning the course of the
Ganges does not however bear out this censure : he described that
river as flowing from the Emodi Mountains (one of the names
under which the Greeks designated the Himalaya) towards the
south, till it reached a city called by the same name (Ganges)
and then turning to the east and holding that course as far as
Palibothra and its outflow into the sea.^ His estimate of the
size of Taprobane — which he stated to be 7000 stadia in length
and 5000 in breadth^ — though greatly exaggerated, did not

' The whole length of the Eed Sea
he estimated at 14,000 stadia (1400 G.
miles) ; which considerably exceeds
the truth, the real length being just
about 1200 geographical, or 1400
Encjlish miles.

8' Id. xvi. 4, §§ 14, 15.

^ reXevTOLOv o.KpcsiTrjpLOv rrjs izapaKias
Tavrrjs rb No'tou Kepas. Kdfj.-\\iavT i Se
Tovro &s iirl ix.ecrrjfji.^piav ovKeri, (pTjcriv,
eXo/^ev KifxevcDV a,vaypa(pa.s ovSe t6tzcov
5ia rb fit^Keri eivai yvcopLfiov r-ijv e^yjs
TrapaXiav. Id. § 14, p. 774.

1 Strab. XV. p. 719. Though of
course not strictly accurate, this de-
scription is undoubtedly an improve-
ment on the received idea that it flowed

simply from west to east. Its large
affluent, abounding in crocodiles and
dolphins, which he called Qidanes, is
otherwise unknown, but it seems pro-
bable that the true reading is Ol^dv-ns,
or 'lofjidvy)s, as suggested by the most
recent editors, Coray and Kramer (ad
loc.) ; in which case we have here the
first mention by name of the Jumna.

2 Stephan. Byzant. s. v. 'Va-Kpo^oLvi).
Our existing text of Stephanus gives
only 500 stadia for the breadth of the
island, but that is clearly erroneous,
and there is little doubt that we should
read, as suggested by Forbiger, i^evra-
/cto'xiAicoz' for ■KiVTaKocriwv.

64 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

differ materially from that given by other geographers : the
vast size of that island having become a received fact in tra-
ditional geography. Towards the north his knowledge was
bounded by the same limits as that of his predecessors :
beyond the Tanais, he stated, nothing was known, but that
Sarmatian tribes occupied the regions to the north.^

§ 3. Artemidorus is censured by his epitomizer Marcianus as
deficient in accurate geographical knowledge,* by which he
probably means that he made no attempt to determine the
position of places by their latitude and longitude (in the
manner subsequently introduced by Ptolemy) : but it is cer-
tain that he bestowed great pains upon the determination of
distances, and not only produced in this manner the most
accurate Periplus in detail that had yet been put together of
the internal seas (the Mediterranean and the Euxine), but
entered into an elaborate computation of the total length of
the habitable world, from the mouth of the Ganges to the
Sacred Promontory, which is adopted by Pliny as the most
accurate known to him. It will be worth while briefly to
compare the results with those arrived at by Eratosthenes.^

The whole distance from the Indian Ocean at the mouth of
the Ganges to Gades he computed at 68,545 stadia; a sum
total which he made up as follows. From the mouth of the
Ganges through India and Parthia to Myriandrus on the Gulf
of Issus he reckoned 41,725 stadia : thence by the most direct

3 Plin. H. N. ii. 108, § 246.

* Trjs jJLev aKpifiovs yewypacpias \ei-
irerai, Epit. p. 65. .

^ We have here the advantage that
besides the extract given by Pliny (H.
N. ii. 108, § 242, foil.), who has con-
verted the stadia into Koman miles, we
find that the measurements agree so
exactly with those given by Agathe-
merus {Geograph. i. 4) as to leave no
doubt that both are taken directly from
the same source. The two statements
thus serve mutually to correct one
another ; and Agathemerus has in
many instances supplied details which
are wanting in Pliny. The numbers

are, as usual, frequently corrupt, but
by comparing the two lists, and the
details with the sums total, they can in
most instances be restored with reason-
able certainty.

It is unfortunate that Artemidorus,
in stating these distances, appears to
have reckoned them all from Gades ;
while Eratosthenes and other writers
generally reckoned them from the
Strait of the Columns, and we do not
know exactly the distance assigned by
Artemidorus between the two. But
the interval is stated by Strabo at 750,
or 800 stadia; and the difterence is
immaterial (Strabo, iii. 1, p. 140).

Sect. 3.

ARTEMIDORUS.

course by sea, touching at Cyprus, Patara in Lycia, the islands
of Ehodes and Astypaleea, the promontory of Tsenarus in
Laconia, Cape Pachynus and Lilybaeum in Sicily and Caralis
in Sardinia, to Gades 26,820.^ To this he added 3932 stadia
from Cades by the Sacred Promontory to that of the Artabri
(Cape Finisterre) ''' which he appears to have considered — con-
trary to the received opinion among his contemporaries — as
the most westerly point of Europe.

§ 4. But besides this computation, which (as will be seen)
did not differ materially from that of Eratosthenes, though
based upon more accurate measurements in detail, Artemi-
dorus gave another estimate, founded as far as possible upon
itinerary or terrestrial measurements, which he regarded as
more trustworthy than those by sea.^ This second line pro-
ceeded across Asia from the mouth of the Ganges to the
Euphrates, a distance which he reckoned at 41,350 stadia:*
thence to Mazaca in Caj)padocia 2550 ; thence through Phrygia
and Caria to Ephesus 3320 stadia; from Ephesus across the
-^gean Sea to Delos 1600, and thence to the Isthmus of Corinth
1700 stadia. Thence he drew the line by Patrse, Leucadia,
Corcyra, to the Acroceraunian promontory and thence across
the sea to Brundusium, which he placed at 3880 stadia from
Corinth : thence to Kome by land 2880 stadia. From Eome

« Note B, p. 102.

' In regard to this number the MSS.
both of Pliny and Agathemerus vary
(see Miiller's note). As we know from
Strabo that Artemidorus correctly re-
duced the distance from Gades to the
Sacred Promontory to 1700 stadia, it
teems difficult to believe that he re-
garded the otlitr as projecting more
than 2000 stadia, or (iiccording to one
reading) more than 5000 stadia farther
west.

* The expression " alia via, qua;
certior, itinere terrene maxime i^ati t a
Gange," &c. is indeed found only in
Pliny, and not iw Agathemerus, but it
is certainly most prol able ihat it pro-
ccf ds from Arte midorus and not from
Pliny himself.

VOL. II.

^ No details are here given; but on
the former line Agatliemerus reckons
16,000 stadia from the mouth of the
Ganges to that of the Indus; from the
Indus to the Caspian Gates 15,300:
and thence to the JJuphrates 10,050;
making up precisely the sum here
stated. The point on the Euphrates
was probably the Zeugma (opposite
Bir), from whence was the shortest
overland route to Myriandrus.

Strabo states (p. 664) tliat Artemi-
dorus agreed with Eratosthenes in
regard to the dirt ct distance from the
Euphrates to India ; and we know that
Eratosthenes reckoned in round num-
bers 40,000 stadia; but doubtless the
discrepimcy was regarded by Strabo
as immaterial.

HISTORY OF ANCIENT GEOGKAPHY. Chap. XVIII.

he followed the line of the Eoman roads to Scingomagus at
the foot of the Alps, a distance of 519 Koman miles or
4152 stadia; thence to Illiberis at the foot of the Pyrenees
3747 stadia, and thence through Spain to Gades 6650 stadia.^
The whole result thus obtained Was 71,560 stadia : an excess
of more than 3000 stadia over his other estimate, which was
the natural consequence of the devious course that this second
line had in fact pursued. Of the extent of those deviations
Artemidorus had undoubtedly no idea, but it is difficult to
suppose that he really regarded Ephesus, Corinth, Brundusium,
Eome, Susa, Illiberis and Gades, as situated nearly enough
in the same latitude to allow of a line passing through those
points presenting any approximation to a measurement of
the earth's longitude.^ Yet we find no indication of his having
applied any subsequent correction to the sum total thus
attained : and it appears certain that, while Artemidorus de-
serves great credit for having seen the important advantage to
be derived from the Koman roads, with their measured dis-
tances, he still failed to see the necessity of applying to these
itinerary/ distances the corrections necessary in order to deduce
from them any geographical results.

§ 5. He next proceeded to give in like manner a measure-
ment of the breadth of the habitable world from the Ethiopian
Ocean, which he placed 5000 stadia to the south of Meroe ^ —

^ Agathemerus, § 17 ; Plin. E.N. ii.
108, § 244. The numbers in the two
last cases are corrupt and erroneous in all
our MSS. I have adopted the correc-
tions proposed by C. Miiller. The sum
total being fixed, as Agathemerus and
Pliny here coincide, the details are of
less importance. But as the route
followed by Artemidorus through Gaul
and Spain is uncertain, we cannot feel
any security as to the numbers given.

'■' It is sufficient to say roughly that
Ephesus and Corinth are in about 38°
N. lat., Eome about 42°, Susa (Scingo-
magus), above 45°, and Gades 86|°.

It appears almost more strange that
in a sea with which the Greeks were
so well acquainted as tha;t between

Greece and Italy, he could suppose
that a line from Patrss by Leucadia,
Corcyra, and the Acroeeraunian Pro-
montory to Brundusium would nearly
coincide with a prolongation of the
parallel thi-ough Ephesus and Corinth.
But when we come to consider the map
of the Mediterranean as conceived by
Strabo, we shall see that it in great
measure agrees with this assumption
of Artemidorus.

^ The mention of this Ethiopian
Ocean is very remarkable. Though
Strabo has preserved to us such copious
extracts from the part of the work of
Artemidorus relating to the Erythraean
Sea and the Ethiopian tribes, there is
no mention of anything corresponding

Sect. 3.

AETEMIDOEUS.

to the Tanais, his extreme limit to the north. This also he
reckoned along two different lines, both however proceeding
from Alexandria by Ehodes to Tenedos and the mouth of the
Hellespont, and thence by the Bosphorus to Cape Carambis,
from whence it crossed the Euxine to the entrance of the
Palus Mseotis and thence to the mouth of the Tanais. The
two calculations (into the details of which it is unnecessary to
enter) gave as their results 18,056 and 18,690 stadia respec-
tively from Alexandria to the mouth of the Tanais. Adding
to these the distance from Alexandria to Meroe, which he
estimated, in accordance with Eratosthenes, at 10,000 stadia,
and that to the Ethiopian Ocean at 5000 more, we should obtain
totals of 33,056 and 33,690 stadia, the last of which cor-
responds exactly with that given by Pliny (according to the
best MSS.) of 4212 miles or 33,696 stadia.* He thus arrived
at the same conclusion with Eratosthenes that the breadth of
the inhabited world was rather less than half its length. Their
results were however based on a very different calculation ;
Eratosthenes having extended his measurement far to the
north in Scythia in order to attain the latitude of Thule, while
Artemidorus carried his farther to the south, to the supposed
Ethiopian Ocean.

§ 6. Besides these principal lines, numerous other distances
are cited from Artemidorus by Strabo, the ipiost important of
which are those given along a line of itinerary route from
Ephesus to the Euphrates, which he expressly described as
being the high-road followed by all who travelled towards the

to this expression. It would seem as
if Artemidorus had supposed the sea
to sweep round from the Southern
Horn (which was the farthest point
known to him) at once to the south of
Ethiopia. Eratosthenes, as we have
seen, placed the Kegion of Cinnamon
in the same parallel with the Sem-
britse on the Upper Nile, at a distance
of 3500 stadia south of Meroe.

*■ For the discussion of these numbers
see the note of C. Miiller on Agathe-
merus in his Geographi Grxci Minores

(tom. ii. p. 481). The old editions of
Agathemerus all gave 8000 stadia for
the first stage in the measurement— the
distance from the Ethiopian Ocean to
Meroe— but all the best MSS. have
5000 ; and this agrees with the result
given by Pliny according to the best
MSS. ; the received text of later edi-
tions (including that of Sillig) having
been altered in accordance with a con-
jecture of Harduin's, so as to coriespond
with the erroneous reading of Agathe-
merus.

F 2

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

east.^ It ascended the valley of the Mseander to Laodicea,
thence through Phrygia, the Paroreios, and Lycaonia, to Ma-
zaca, the capital of Cappadocia, from which it proceeded to
the Euphrates, at a place called Tomisa on the borders of
Sophene in Armenia, which appears at this time to have been
a town of importance.^ From thence a line of route led across
the chain of Mount Taurus southwards to Samosata, and thence
across Mesopotamia. East of the Euphrates, according to
Strabo, the distances given by Artemidorus did not differ from
those of Eratosthenes ; probably he had in fact no fresh
materials by which to correct them.^

The work of Artemidorus appears to have continued for a
long period to enjoy considerable reputation, probably on
account of its general and systematic character, as well as the
accuracy of its details as to distances, dimensions of islands,
&c. At the same time Strabo, though frequently availing
himself of its authority, does not place it on a par with those
of Poly bins and Posidonius, nor think it necessary to enter
into any regular exposition of the system of its author, which
probably did not differ materially from that of Eratosthenes.

§ 7. To the same period with Artemidorus belong two other
writers whose names deserve a passing mention, as they left
geographical works, which, though now utterly lost, are fre-
quently cited by later authorities. One of these' is Metro-
DORUS of Scepsis, who was a rhetorician of eminence, and
played a considerable part in political life. Among his various
writings, we are told that he left a Periegesis, which must
have been a work of a distinctly geographical character, and
from which the statements quoted from him by Pliny and

^ ArtemiJor. ap. Strab. xiv. 2, p. 663.
iirel Be kolvt) tis 68hs rerpnr'rai airacn
Tails sttI Tcts avaToKas dSonropovcriv e|
^'Ecpecrov. At a later period this line of
route seems to have ceased to be fre-
quented, as it is not found ia any of
the Roman Itineraries. Its details
will be more fully examined in the
chapter on Strabo's geography of these

countries.

" It is called by Strabo <ppovpiov
a^iSXoyov (xii. 2, p. 535). and was given
over by Lucullus to Ariarathes, king
of Cappadocia, as a reward for the ser-
vices he had rendered in the Mithri-
datic War.

' Strabo, xiv. 2, p. 663.

Sect. 4.

SCYMNUS CHIUS.

Stephanus of Byzantium were probably taken. Far more
numerous are the citations from the other author above
referred to; Cornelius Alexandee, surnamed Polyhistoe,
on account of the extent and variety of his knowledge. He
appears to have left a considerable number of works treating
of the geography, as well as the history and antiquities, of
different countries, as Egypt, Syria, Lycia, Phrygia, &c. : but
unfortunately our knowledge of them is due almost entirely to
the jejune citations of Stephanus of Byzantium, which gene-
rally preserve to us nothing more than barren names.^

Section 4. — Scymnus Chius.

Of a very different character was a little work, which has
received in modern times far more attention than it deserves,
from the accidental circumstance of its having been in great
part preserved to us, while so many more valuable and im-
portant treatises on the same subject have perished. This is
the little compendium of geography in iambic verse, com-
monly known, though without any foundation, under the name
of Scymnus Chius. It belongs unquestionably to the period
we are now considering, being dedicated to a certain Nico-
medes, king of Bithynia, who, as C. Miiller has shown, must
be the third monarch of the name, who reigned from the year
91 to 76 B.c.^ But the author is in reality wholly unknown.
The principal fragment was preserved in the same MS. with
the Epitomes of Marcianus of Heraclea, and was in conse-
quence first published under the name of that author — an
attribution clearly erroneous. Holstenius and Isaac Yossius

' The fragments both of Metrodorus
and Alexander Polyhistor are collected
by C. Miiller in his Fragmenta Hisfori-
corum Grascorum, torn. iii. pp. 203-214.
It is uncertain whether the different
writings of Alexander qnoted by Ste-
phanus under the names of AlyvTrrtaKa,

Trepl Kapias, irepl AvkIus, &c., formed
separate treatises, or were parts of one
great work, but the former hypothesis
is the most probable.

° Bee his Prolegomena in his Geogr.
Grxci Minores, torn. i. p. Ixxvii.

HISTOEY OP ANCIENT GEOGRAPHY. Chap. XVIII.

were tlie first to attribute it to Scymnus Chius, a writer cited
more than once by late grammarians as author of a Perie-
gesis : ^ and it continued to pass under his name till the year
1846, when Meineke, in republishing the extant fragments,
showed clearly that there were no grounds for ascribing them
to that writer. The real work of Scymnus Chius, whose age is
wholly unknown, appears to have been in prose, and the few
statements cited from him are not in accordance with those of
our author.^

The portions of the work that have been preserved to us
consist, first, of a long continuous fragment of mor® than
700 lines, containing the introduction, a rambling and ill-
written prefatory discourse, together with the greater part of
the division of the work relating to Europe. Besides this,
there are numerous smaller fragments concerning the shores of
the Euxine, both the European and Asiatic sides, which had
been incorporated by the author of the anonymous Periplus of
the Euxine Sea, usually published as a kind of sequel to that
of Arrian.^ All that related to the rest of Asia and Africa is
wholly lost to us.

§ 2. The anonymous author professes to have composed his
little work in imitation of the chronological treatise of Apollo-
dorus, and in like manner put it in iambic verse, for the pur-
pose of its being more easily remembered.* Both his style and
his versification are very indifi'erent ; and the work is wholly

' Scymnus Chius is cited by name
several times by Stephanus of Byzan-
tium, and also by the Scholiast on
ApoUonius Khodius, iv. 284-. Nothing
is known concerning him from other
sources.

^ See Meineke's Preface to his edi-
tion of Scymnus Chius (Berolin. 1846).
His views have been adopted by C.
Miiller in his recent edition of the
Geographi Grxci Minor es (torn. i. Pro-
legom. p. Ixxiv-lxxvii). Meineke how-
ever has retained the name "utjjote
louico duorum sfeculorum usu recep-
tum," and, as a mutter of convenience,

I have done the same, when I have
had occasion to cite the little work in
question.

^ They were first recovered, and dis-
tinguished from the prose text in which
they are thus incorporated, by Holste-
nius, whose arrangement of the verses
has been generally followed by the
later editors.

' vv. 19-35. It is singular that he
does not advert to the geographical work
of ApoUodorus, which was also in
iambic verse (see above, p. 49), and
would seem to furnish a better prece-
dent.

Sect. 4.

SCYMNUS CHIUS.

destitute of anything like poetical ornament or character ; but
these defects might be excused, if the matter were more valu-
able or trustworthy. The book was indeed designed only as a
popular compendium for general use : and therefore makes no
pretence to novelty or profound research.^ But while its
writer gives a long list of authors whom he had consulted, or
professed to consult,® he was wholly without the critical skill
to compare and discriminate between his different authorities,
or to discard the statements of earlier writers, which the pro-
gress of geographical knowledge had shown to be erroneous.
Thus one of his principal authorities is Ephorus, and he often
makes unhesitating use of that author, without reference to the
results of more recent discoveries. The consequence is that
his book, instead of representing the state of geographical
knowledge in his own day, is a jumble of confused statements
belonging to wholly different periods. By far its greatest
value at the present day arises from the notices taken from
Ephorus, Timseus, and other earlier authors (cited with their
names), concerning the foundation of the different Greek
colonies and cities on the coasts of the Mediterranean. In
this respect indeed we derive from it some important accessions
to our knowledge. But in a strictly geographical point of
view it is almost wholly worthless.

§ 3. The best part of that which remains to us, is un-
doubtedly the description of the coasts of the Euxine, which
is taken principally, as he himself informs us, from Demetrius

* Its author, however, claims in one
passage (vv. 128-136) to have carefully
investigated and visited in person not
only the cities of Greece and Asia
Minor, but those of the Adriatic and
Ionian Seas, and those of Tyrihenia,
Sicily, and other western lands, as well
as Carthage and a great part of Libya.
But it is certain that no trace of any
additional information derived from
this source is to be found in the poem.

^ Unfortunately this passage of his
work (vv. 110-125) is corrupt, and the

names of several of the authors cannot
be determined. He professes to have
followed Eratosthenes most of all, then
Ephorus, Dionysius of Chalcis, who had
written five books on the Kriaeis, or
foundations of cities, the Sicilian Cleon
(a writer very little known, but cited
also by Marcianus of Heraclea) and
Timosthenes. Then follow some lines
which cannot be deciphered, after which
he adds the names of Timseus and
Herodotus.

HISTOEY OF ANCIENT GEOGRAPHY.

CiiAP. xvm.

of Callatia, a writer otherwise almost unknown,'' but who seems
to have been judicious and well-informed. Besides the Periplus
of its shores, we find many interesting notices of the founda-
tion of the numerous Milesian and other Ionian colonies that
bordered its circumference.^ But of the Scythian tribes on the
north his account is taken exclusively from Ephorus, and it
does not appear that any progress had been made in geo-
graphical knowledge in that quarter. He states indeed that
the Borysthenes was navigable for forty days from its mouth ; ^
and dwells strongly upon its commercial advantages — its
quantity of large fish, as well as the flocks and herds, and
crops of corn on its fertile banks, so that he terms it "the
most useful " of all rivers ; ^ but he describes it as inaccessible
in the upper part of its course from ice and snow. Beyond
the nations that he enumerates, none of whom were far removed
from the Euxine, he says that the country was altogether
uninhabited and unknown. The Tanais, he tells us, was
according to some a branch of the Araxes — a strange miscon-
ception, though shared as we have seen, even by Aristotle^ —
but Ephorus described it as flowing from a great lake; a
statement probably copied from Herodotus.

§ 4. With regard to the west of Europe his accounts are
very confused and erroneous, and evidently reflect the vague
notions current in the days of Ephorus and Timaeus, without
reference to the more exact information that was available at
the time he wrote. Thus he represents the Celts as inhabiting
the whole western region from the neighbourhood of Gades
and Tartessus (an old name which he borrowed from Era-
tosthenes^) to near the head of the Adriatic. He adds a

' He is probably the same as the
Demetrius, mentioned without any-
other adjunct by Agatharchides (§ 64),
as one of the writers who had given
the best accounts of the northern regions
of the earth.

« See Chapter IV. § 2.

" V. 816. This statement is un-
doubtedly copied from Herodotus, and

goes far to show that there is no error
in the existing text of that author, what-
ever we may think of the accuracy of
his information.

* ouros Se ■wa.vTwv ^(Tt\ xpejaiSecTTaTos,
V. 813.

^ Meteorologica, i. 13, § 16.

^ The name of Tartessus, which was
at iir^t applied to the south of Spain

Sect. 4.

SCYMNUS CHIUS.

strange story of there being a great northern column at the
extreme limit of their country projecting like a promontory
into a stormy sea, from the foot of which the Danube (Ister)
took its rise.* That river he conceived, in accordance with the
notion so long prevalent among the Greeks, to flow with one
arm into the Adriatic, with another into the Euxine ; and he
repeats, as usual, the fables concerning the Eridanus, and
the amber distilling from the poplars on its banks. He also ,
places the Electrides, or Amber Islands, near the head of
the Adriatic; and what is more remarkable, mentions two
other islands in the same neighbourhood as producing the
finest tin.^

On the other hand his notices of the Greek colonies on the
coast of Liguria and Spain, from Massilia and its dependencies
to Ehoda near Emporium, and even the outlying settlement of
Msenace near the columns of Hercules, is exact and instructive.
The same remark applies to his account of the Greek cities in
Sicily, concerning which he had good authorities; but he
does not even condescend to mention by name those of Punic
origin, though Panormus and Lilybseum were undoubtedly in
his time among the most important cities of the island. His
account of Italy is in like manner very confused ; here again
his information concerning the Greek colonies is valuable and

generally, gradually disappeared as the
Greeks became better acquainted with
that country, much as the name of
Cathay has done in modern times.
Eratosthenes still applied the name to
the territory near Calpe ; but Artemi-
dorus, who had himself visited the
country, denied that any such name
was found there (Strab. iii. 2, p. 148).
Others gave the name of Tartessus to a
river, which Strabo identifies with the
BsBtis or Guadalquivir ; and he con-
siders the region of Tartessus to be the
same with the land of the Turduli, the
modern Andalusia (ibid.).

It is remarkable that our author de-
scribes Tartessus as producing in abun-
dance (i.e. trading in) " the stream-

washed tin of Gaul " as well as gold
and brass.

T\ Xeyo/JLevY) TapTrjcrcro';, £7rt(^ai/T)S n-oAiy,
^OTa/jioppvTOv Ka(T<xiTepov ex T^s KeArtK^s
Xpvijov re koX \a\Kov (pepovcra nKeCova.

, vv. 164-166.

* VV. 188-195.

Svo Se KOLT avTou's eicri rijerot KeCfievat
Ka(X(rCTepov cCl SoKoOcrt Ka\Ki,(TTOv <j>ep€LV.
VV. 399, 400.

This is, as far as I remember, the
only notice connecting the tin islands
with the Adriatic. It would seem to
indicate that tin, as well as amber, was
sometimes brought overland to the head
of that gulf: but such a trade could
hardly be carried on to any considerable
extent.

74 HISTOEY OF ANCIENT GEOGKAPHY. Chap. XVIII.

interesting, while his notices of the nations inhabiting the
peninsula are a jumble of statements derived from the earliest
Greek writers, with a few that belong to a later period.^ In
general it may be said that his ethnography of Italy is not at
all in advance of that of Scylax, who wrote two centuries and
a half before him ; except that he naturally dwells emphatically
upon the power and importance of Kome ; ^ a topic that could
scarcely be omitted by a writer in his time.

With regard to Greece itself he expressly tells us that he
followed the authority of Ephorus ; ® but this part of his work
is meagre and of little value. He could not indeed be ex-
pected to add to our geographical knowledge of countries so
familiar to all. His account of the Adriatic on the other hand
was taken from Theopompus ; ^ so little did he attempt to avail
himself of the latest and best authorities ; and is in conse-
quence full of errors.'^

Section 5. — Voyage of Eudoxus.

§ 1. We have seen from the above review how little progress
was made, through the whole Alexandrian period, after the
reigns of the three first Ptolemies, in the knowledge of the
external coasts of Africa or Asia. The commercial relations
then established appear to have settled down into a regular
routine, from which there was little, if any, deviation; and
geographical explorations, or voyages of discovery in the
modern sense, were very rarely undertaken in ancient times.
There was however one notable exception during the period
in question, which well deserves our attention, though our
information concerning it is unfortunately very imperfect.

" Thus lie mentions GEnotria and the
OEnotrians — an aiDpellation used only
by the Greeks in early times, together
with the Samnites, Lucanians, and
Campanians (vv. 241-244).

ot/cou/xeVyjs, v. 233.

8 V. 472. « V. 370.

1 Thus he repeats the strange state-
ment that the Hyllic Chersonese (the
peninsula of Sabion cello) was about as

' He culls it &(TTpovTi Koivhvrrjs SAtjs large as the Peloponnese

Sect. 5.

VOYAGE OF EUDOXUS.

This was the voyage of Eudoxus of Cyzicus, our knowledge
of which is derived exclusively from the account given by
Posidonius, and preserved to us by Strabo.^

According to this narrative, Eudoxus, who was a man of
education and of an inquisitive mind, had been sent by his
native city on an honorary mission to the court of Ptolemy
Euergetes II. (Physcon), and was much occupied with inquiries
concerning the course and sources of the Nile. While he was
still at Alexandria it happened that an Indian was brought to
the king by the guards of the Arabian Gulf (the Ked Sea),
whom they reported that they had found alone in a ship which
had been wrecked on the coast. As soon as he had learnt a
few words of Greek so as to make himself understood, the
captive stated that he had set sail from India, and the ship
having been driven out of her course all his companions had
perished of hunger, leaving him the sole survivor. He offered,
moreover, if the king would fit out a ship, to direct them on
their way to India ; an offer which was accepted, and Eudoxus
among others took part in the adventure. They accomplished
the voyage successfully, and having taken with them suitable
presents, brought back in return a valuable cargo of spices
and precious stones. Eudoxus however was frustrated in his
hopes of private advantage, the king having seized and appro-
priated the whole cargo. But after the death of the tyrant (in
B.C. 117) his wife Cleopatra who succeeded him in the govern-
ment,^ sent out Eudoxus a second time with more extensive

^ Posidonius ap. Strab. ii. 3, § 4, pp.
98-100.

^ There is some little uncertainty
with regard to the chronology of these
voyages of Eudoxus, but their date may
be fixed within very narrow limits.
His first voyage took place in the reign
of Ptolemy Euergetes II., who died in
B.C. 117. Cleopatra at first reigned
conjointly with her elder son, Ptolemy
Lathyrus, and it was during this period
that she sent out Eudoxus, the second
time. But before his return Cleopatra
had quarrelled with her son, and had

been compelled to leave him in sole
possession of the sovereignty. This
took place about B.C. 112, and explains
the statement of Posidonius that when
Eudoxus returned from his second
voyage he found Cleopatra no longer at
the head of affairs, but her son, by
whom he was despoiled for the second
time (p. 99). The expression of Cor-
nelius Nepos (ap. Plin. ii. 67), that he
fled from Ptolemy Lathyrus (cum
Lathyrum regem fugeret), was correct,
though it does not refer to the reign of
Lathyrus as sole monarch after the

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

equipment. He again accomplished the voyage in safety, but
on his return was driven out of his course by adverse winds
beyond Ethiopia. He however found a friendly reception from
the natives, some words of whose language he wrote down ; and
brought away with him the sculptured prow of a ship which
had been wrecked on the coast and was reported to have
belonged to some navigators who had come from the westward.
Keturning to Alexandria in safety he was again plundered of
all the produce of his expedition. But having shown the prow
which he had brought with him to the sailors and traders in
the port, it was recognized as belonging to a class of ships
that were in ,the habit of sailing from Gades, and some even
asserted that it was that of a particular vessel that had sailed
beyond the river Lixus in Mauretania, and had never been
again heard of.

§ 2. Eudoxus now abandoned all reliance upon the Egyptian
monarchs, but being convinced that it was possible to sail
round Africa, he determined to try the experiment ; and for
this purpose, after returning to his own country, he embarked
his whole fortune on board another ship with which he sailed
first to Dicaearchia in Italy, then to Massilia and ultimately to
Gades. Everywhere he proclaimed the object of his enterprise
and obtained so much assistance that he was able to fit out
a large ship, with two light vessels, similar to those used by
pirates, to accompany it, evidently with a view to facilitate his
landing from place to place. In addition to the crews he took
on board physicians, artisans of various kinds, and dancing
girls from Gades ; and thus equipped set out " on his voyage
to India." At first he stood well out to sea, meeting with
continual westerly winds ; but after a time the discontent of
those on board compelled him to approach the land ; where he

death of his mother ; which belongs to
a later period, B.C. 89-81. We may,
therefore, suppose Eudoxus to have set
out on his second voyage about B.C. 113,
and to have returned in b.o. 112, or

111. (See this point fully discussed
by 0. Miiller ia the Prolegomena to his
Geographi Grxci Minores, torn. i. p.
Ivii.)

Sect. 5. VOYAGE OF EUDOXUS. . 7/

met with the disaster he had feared, his large ship having run
aground, so that he was unable to get her off again. His crews
however and cargo were saved, and out of the timbers of the
lost vessel he was able to construct a third bark, about the
size of a penteconter. He then resumed his voyage, until he
reached a tribe of Ethiopians who spoke the same language
with those that lie had previously visited on the eastern coast
of Africa, and whose words he had written down. From these
he learned that their territory adjoined that of Bocchus, king
of Mauretania.

At this point — for what reason we are not told, but probably
on account of the small size of his vessels — he determined for
the present to abandon the enterprise and return northwards.
On reaching Mauretania he sold his barks and proceeded
by land to the court of Bocchus, whom he endeavoured to
persuade to fit out a fresh expedition. Finding however after
a time that his efforts were unavailing and that his life was
in danger, he fled from Mauretania, and took refuge in the
iloman territory. But bis spirit was still unbroken ; he re-
turned to Gades, and there fitted out another large ship, with
a penteconter to accompany it. On his former voyage he had
discovered an island abounding in wood and water, but unin-
habited ; and he now took with him agricultural implements,
seeds and building materials, with a view to establish himself
there for the winter, if his voyage should be delayed.

§ 3. Here the narrative of Posidonius unfortunately breaks
off abruptly ; of the subsequent fortunes of Eudoxus he knew
nothing ; but he accepted his conclusion, though certainly
based on most inadequate evidence, that Africa could be
circumnavigated. Strabo on the contrary treats the whole
story with contempt, and reproaches Posidonius for his cre-
dulity in giving credit to an old wife's tale, which he regards
as on a level witb the fictions of Euhemerus and Antiphanes.*
There is certainly no foundation for this ; the story of Eudoxus,

* Strabo, ii. 3, § 5, pp. 100-102.

HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

as related by Posidonius, contains nothing either absurd or
incredible ; the most remarkable point in it — though one that
attracted least attention — his having accomplished two direct
voyages to India and back, is rendered not improbable by the
facility with which the same voyage was performed at a later
period, from the time of Hippalus onward. In regard to Africa
we do not know how far he advanced along either the eastern
or western coasts of the continent, but there is no reason to
suppose that he proceeded very far. The expression " beyond
Ethiopia " probably means no more than that he was carried
to the coast south of Cape Guardafui, at that time the limit of
the Greek knowledge in this direction; while on the other
side we are distinctly told that the farthest tribes whom he
visited bordered on the dominions of king Bocchus, and were
therefore not very remote from Mauretania.^ His only con-
tribution to the proof that Africa was really surrounded by
the ocean was derived from the idle story of the ship's prow
being one that came from Gades, on which certainly no reliance
could be placed,® and in this respect Posidonius undoubtedly
merits the censure of Strabo, for having admitted this as con-
clusive proof, while he rejected the story of the circumnavi-
gation as told by Herodotus.

It is a striking instance of the carelessness with which such
statements were repeated at second hand by ancient writers,
that Pliny quotes Cornelius Nepos as relating that Eudoxus,
in making his escape from Ptolemy Lathyrus had sailed round
from the Arabian Gulf (the Red Sea) all the way to Gades!^
The readiness with which such stories were credited arose not
merely from the general belief that Africa was bounded by a

* Strabo, I. c. p. 100. This statement
would seem to exclude the supposition
of his liaving advanced beyond tlie
Great Desert, and come in contact
■with the negro tribes on tlie coast
visited by Hanno.

^ It is indeed true that fragments of
wrecks have been occasionally carried
for very long distances (see the case
quoted by Humboldt, Cosmos, vol. ii.

note 163, Engl, transl.). But the cur-
rents on the east coast of Africa would
be altogether unfavourable ; and the
evidence of identify (the chief point of
all) appears to have been very ques-
tionable.

' Plin. H. N. ii. 67, § 169. The same
statement is made by Pomponius Mela
(iii. 9, § 90), also on the authority of
Cornelius Nepos.

Sect. 6.

THE JUGURTHINE AND MITHEIDATIC WARS.

circumfluous ocean, but from the erroneous idea that that
ocean was far less distant to the south than it really was. Had
the continent possessed the form supposed by Eratosthenes
and Strabo, its circumnavigation would have been a com-
paratively easy affair.^

Section 6, — Roman Wars : the JugurtJiine and Mithridatic

Wars.

§ 1. Meanwhile the Roman arms were opening the way for a
more accurate knowledge of Northern Africa. We have seen
that the acquaintance possessed by the Greeks with that
region was substantially bounded by the Carthaginian terri-
tory on the west, and that though they were familiar with the
2Mraplus or voyage along the coast of the Mediterranean from
Carthage to the Straits of the Columns, they had little, if any,
knowledge of the interior. The Carthaginians themselves
appear to have had but few settlements of importance along
this line of coast : and if they carried on any trade with the
tribes of the interior, from this all foreigners would doubtless
be jealously excluded. But as early as the Second Punic War,
the Romans came of necessity into contact with the Numidian
tribes which had previously been connected with Carthage
only : and the opposite part taken by the two great divisions
of the nation — the Massy lians and Masseesylians — under their
respective chiefs, Masinissa and Syphax, must have rendered
all Roman writers familiar with these two leading divisions of
the Numidian race.^ The result of the war was to place the
whole JSTumidian territory from the frontiers of the narrow
province still left to Carthage, to the river Mulucha, which
separated it from Mauretania on the west,^ under the dominion

* See the maps representing the
form of the known world as conceived
by both these geographers.

" Tlie names both of the Mussyli
and MassEesyli are found in Polybius
(iii. 33). That of the Maccssi men-

tioned by the same writer, in con-
junction with them, is not found in
any Liter author.

* The river Mulucha, called by
Strabo Molochath, which still continued
in the time of Pliny to be the limit

HISTORY OF ANCIENT GEOGEAPHY. Chap. XVIII.

of Masinissa, and subsequently of his son Micipsa. But until
the ambition of Jugurtha involved him in war with Rome, it is
not probable that anything had occurred to lead to any more
accurate knowledge of the countries in question. We learn
indeed that Eoman and Italian traders had established them-
selves in considerable numbers in a few of the towns, such as
Vaga and Cirta, which nearly adjoined the Roman province, or
were but little removed from the sea : ^ but it is probable that
they extended their relations little farther to the west, and
the statement of Sallust that before the Jugurthine War the
Romans were known only by name to the Mauretanians,^ may
be accepted a-s nearly, if not literally, correct. The Gastulians
beyond the range of Mount Atlas towards the interior were
merely wild nomad tribes, wanting the first elements of civili-
zation, and apparently as yet unvisited by all but native
traders. They were for the first time enlisted by Jugurtha,
and rendered him valuable assistance as irregular cavalry.

§ 2. The operations of the war were not of a nature to throw
much light upon the geography, at least of the more remote
parts of Numidia, and they are very obscurely related by
Sallust. The farthest point to which the Roman arms were
carried towards the south was Capsa, a short distance to the
north of the Tritonian Lake, and surrounded by barren deserts,
notwithstanding which it was attacked and taken by Marius.
That general appears also in his last campaign to have carried
his arms as far as the frontiers of the kingdom of Bocchus,
near the river Mulucha : but this part of his operations is very
imperfectly described. No part of Numidia was at this time
permanently annexed to the Roman dominion : after the final
defeat of Jugurtha in B.C. 106, it was placed under the govern-
ment of Hiempsal, and it did not become a Roman province

between the Mauretanians and MassEe-
sylians, may be clearly identified with
the Wad el Maloiish.or Maloya, which
falla into the Gulf of Melillah, near
tlie present boundary between Algiers
and Morocco.

2 Sallust. Bell. Jugurth. c. 21, 26, 47.

' Id. ibid. 19. "Mauris omnibus rex
Bocchus imperitabat, prsetf-r nomen
cetera ignarus populi Eomani, itenique
nobis neque bello neque pace antea
cognitus."

Sect. 6. THE JUGUKTHINE AND MITHRIDATIC WAES. 8 1

till the time of Caesar, B.C. 46. Mauretania still retained its
independence, but it was brought into more frequent relations*
with Kome during the civil wars of Sertorius in Spain.

§ 3. It is in connection with the wars last alluded to, that
we find the first distinct mention of a discovery which ever
after retained its place in ancient geography. We have seen
above (p. 60) that Eudoxus of Cyzicus was reported to have
discovered in his voyage along the west coast of Africa an
uninhabited island, abounding in wood and water, of which he
conceived so favourable an opinion that he set out on his
second voyage with the idea of establishing himself there for
the winter. The account is too vague to enable us to deter-
mine whether this might be Madeira or one of the Canary
Islands : it is certain that there is no small island nearer to the
mainland at all answering this description. Other indications
also point to the fact that the existence of these outlying
islands in the Atlantic was at this time well known to the
traders of Gades. Among these the most familiar is the story
told by Plutarch * of Sertorius having fallen in near the mouth
of the Beetis with some sailors who had just arrived from
"the Atlantic Islands," which they reported to be two in
number, separated only by a narrow strait, and distant about
10,000 stadia from the coast of Africa. They enjoyed the
most perfect of climates — warm, without excessive heat — with
only as much rain as was desirable, but soft and damp winds
continually blowing, so as to maintain an equable temperature,
and produce an unexampled degree of fertility. Not only
were the islands well adapted for tillage and the growth of
fruit-trees, but they produced of their own accord such abun-
dance of fruits of various kinds as would suffice to support a
whole population without toil or labour. These fortunate con-
ditions led the mariners of Gades to identify the islands in
question with the Islands of the Blest, of which the Greek

* Plut. Sertorius, c. 8. The same I book of his Histories, fr. 67, 68 : from
story was related by Sallust in the jSrst | whom Plutarch probably copied it.

VOL. ir. Q

HISTOKY OP ANCIENT GEOGKAPHY. Chap. XVIII.

poets had sung : an assumption subsequently adopted by
almost all Greek and Eoman writers. Sertorius, wlio was at
the time with difficulty making head against his enemies, was
disposed to sail away at once to these happy abodes and settle
himself permanently there : but his crews were unwilling to
follow him, and he was compelled to abandon the idea. Hence
" the Islands of the Blest " continued until a much later period
to float in a state of dim geographical vagueness bordering on
the mythical.^ The name of the Fortunatse Insulae was after-
wards applied more specially to the group of the Canary
Islands ; but the account given in this — the earliest notice of
the discovery of any of these outlying Atlantic islands —and
especially of their great distance from the mainland of Africa,
certainly seems to point to Madeira and Porto Santo as the
two islands in question.®

§ 4. At this period Gades was undoubtedly one of the most
important emporiums of trade in the world : her citizens having
absorbed a large part of the commerce that had previously
belonged to Carthage. In the time of Strabo they still
retained almost the whole trade with the Outer Sea, or Atlantic
coasts, both of Africa and Europe : and carried on extensive
fisheries on the coast of Mauretania, while they sent large
ships on long voyages both in the Ocean and the Mediter-
ranean.^ The terms in which Strabo speaks of the extent of
the trade of Gades and the opulence of its merchants, as com-
pared with the narrow limits and natural disadvantages of the
islet on which it stood, remind one strongly of the parallel

^ See especially the ■well-known
poetical description of them by Horace
in one of his Epodes {Epod. 16, w.
41-66), a passage which was in great
probability suggested by this very inci-
dent in the life of Sertorius.

^ Not only does the description of
the peculiarly damp and equable cli-
mate apply much better to Madeira
than to any of the Canary Islands, but
the whole account especially described
the islands as situated in the midst of
the Ocean, and though the distance of

10,000 stadia from the continent is in
any case a gross exaggeration, it is
impossible to believe that it could have
been applied to islands like Lanza-
rote or Fuerteventura lying only about
50 or 60 miles from the mainland.

The incidental notice of them by
Strabo f iii. 2, § 13, p. 150), as situated
" not far from the promontory of Mau-
retania opposite to Gades" seems to
point to a similar conclusion.

^ Strabo, iii. 5, § 3, p. 168.

Sect. 6. THE JUGURTHINE AND MITHRIDATIC WARS. 83

case of Venice in the Middle Ages. And we learn from inci-
dental notices in the history of Eudoxus, already related, that
this was already the case at least a century before the time of
Strabo.^ The general insurrection of the native tribes of the
Iberian peninsula, which had been aroused by Sertorius, led to
extensive military operations for their reduction, and when
Pompey returned from thence to Italy he erected on the
summit of the pass across the Pyrenees a monument as a
trophy, which recorded that he had reduced to subjection not
less than 876 towns in that country.^ The number is doubtless
an exaggeration, more especially if we regard it, as Pliny
certainly did, as confined solely to the eastern province or
Hispania Citerior. But we have no details of these campaigns,
and are therefore unable to judge how far they contributed
to extend or improve the geographical knowledge of the
peninsula.

§ 5. During the same period the Koman arms were actively
employed in the East, and here the operations of LucuUus,
and afterwards of Pompey, against Mithridates and Tigranes,
undoubtedly added largely to the geographical information of
the Komans in regard to portions of Asia that were previously
very imperfectly known either to them or to the Greeks. Mith-
ridates, the sixth monarch of the name, who had succeeded in
B.C. 120 to his paternal kingdom of Pontus,^ including portions
of Paphlagonia and Cappadocia, had gradually extended his
arms over the neighbouring nations towards the east and north,
including the region known as Lesser Armenia (west of the
Euphrates, between that river and Cappadocia) and the whole of
Colchis, with the other wild tribes that extended from thence to

» Id. ii. 3, § 4, p. 99.

8 Plin. iii. 3, § IS.

^ The kingdom of Pontus had been
founded, soon after the death of Alex-
ander (apparently about 318 B.C.), by
Mithridates the son of Ariobarzanes,
■who is usually styled Mithridates II.,
though he was really the first mouarch
of the dynasty who had any claim to

be regarded as an independent sove-
rei;j;n. It extended from the frontiers
of Colchis on the east to the Halys on
the west ; but its limits on the south
were probably never very clearly de-
fined, and its rulers were almost per-
petually engaged in hostilities with
those of Cappadocia, a country of which
Pontus had originally formed part.

G 2

HISTORY OP ANCIENT GEOGEAPHY.

Chap. XVIII.

the Caucasus. So extensive indeed had his power and influence
become, that Parisades, the ruler of the petty Greek kingdom
that had so long maintained itself on the shores of the Cim-
merian Bosporus, was induced to place himself under the
sovereignty of Mithridates, in order to obtain the protection
of his arms against the northern barbarians, — the Sarmatians
and Eoxolani, who were now pressing hard upon the Greek
settlements in this quarter.^ The same course was subse-
quently adopted by the free cities of Chersonesus and Olbia.
The generals of Mithridates, Diophantus and Neoptolemus,
fully answered the expectations entertained from them : they
defeated the barbarians in several battles, and carried their
victorious arms to the Tanais on the one side, and to the Tyras
(Dniester) on the other, where a fort called the Tower of
Neoptolemus, near the mouth of the river, served to mark the
limit of his temporary dominion.^ But even beyond these
limits he concluded alliances with the Bastarnse and the Getse,
who at this period occupied the tracts from thence to the
Danube, and both these nations are said to have contributed
auxiliary contingents to the forces which he brought into the
field against the Eomans.*

§ 6. Meanwhile the extension of his power towards the west

2 Strabo, vii. 4, § 3, p. 309. The
name of the Eoxolani here appears for
the first time. They evidently dwelt
at this period in the steppe country of
Southern Kussia ; but Strabo himself
admits that his notions of their position
were very vague (vii. pp. 294, 306).
The kingdom of the Bosporus, of
which Panticapseum was the capital,
and which is familiar to all scholars
from the intimate relations with Athens
maintained by its kings, Leucon and
Parisades, in the days of Demosthenes,
disappears from history during the
intermediate period, until we find it
again mentioned upon this occasion.
But from the recurrence of the name
there can be no doubt that this last
Parisades belonged to the same dynasty
with the earlier monarchs.

3 Id. vii. pp. 306, 307, 309-312. It
was during these operations that
Neoptolemus was said to have de-
feated the barbarians in a combat of
cavalry on the ice, on the very same
spot in which he had the summer before
defeated them in a naval engagement
(Strabo, vii. p. 307).

^ Appian, Mithridat. 69. The
lazyges, who are termed by Appian, as
well as by Strabo (vii. p. 306), a Sar-
matian race, also figure among these
auxiliaries. This is their first appear-
ance in history. They were at this
time among the tribes nortli of the
Euxine, apparently iu the plains be-
tween the Tyras and Borystlienes, but
their exact position is not clearly indi-
cated.

Sect. 6.

THE JUGURTHINE AND MITHEIDATIC WAES.

and south was checked by the petty sovereignties that hemmed
him in on that side, and which, though individually unable to
oppose him, were supported by the power and influence of
Eome.^ It was not till B.C. 90 that he ventured, by dispos-
sessing Nicomedes III. of his kingdom of Bithynia, and
Ariobarzanes of that of Cappadocia, to provoke a collision
with the great republic : and in B.C. 88 he overran, and made
himself master almost without opposition of the Roman pro-
vince of Asia. The operations of the war that followed (b.c.
88-84) were however confined to Greece and the nearer parts
of Asia Minor, and have therefore little geographical interest :
it was not till the war was renewed in B.C. 74, and Lucullus
was appointed to the command, that they assumed a different
character. That general indeed proceeded at first with great
care and caution, and when after defeating Mithridates before
Cyzicus, he followed him into his own dominions, he was
delayed for a considerable time by the protracted siege of
Amisus. It was not till the spring of B.C. 72 that he attacked
Mithridates himself in his head-quarters at Cabeira, and de-
feated him in a great battle which compelled him at once to
seek refuge in the dominions of his son-in-law Tigranes, king
of Armenia.^

§ 7. That monarch had on his part become the founder of a
power that for the moment appeared to rival,, if not to surpass,
that of Mithridates himself. Beginning only as the ruler of
the cold and barren highlands of Armenia, he had successively
annexed several of the neighbouring provinces, including
Sophene, Atropatene, and Gordyene — the last of which he had

^ At the time -when Mithridates VI.
ascended the throne, the dominions of
Rome in Asia Minor comprised, besides
what they termed the province of Asia,
Phrygia, Lycaonia, and Cilicia Trachea.
Cappadocia and Bithynia were still
ruled by independent monarchs, as was
Paphlagonia also, but the petty dynasts
of that country held only the interior —
the kiogs of Pontus having already
extended their dominion over the sea-

coast as far as the confines of Bithynia,
including the flourishing city of Sinope,
which under Mithridates became the
capital of his kingdom. The Galatians,
who had been settled in Asia since the
time of Attains I. of Pergamus, still
maintained their independence under
their native rulers.

•^ For tlie history of this war see
Appian (MUhridatica, 71-83), and
Plutarch (Lucull. 7-24).

86 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

wrested from the Parthian kings : and had afterwards taken
advantage of the divided state of the Syrian monarchy, to
overrun the whole of the provinces that still remained subject
to the Seleucidan kings, from the Euphrates to the sea, together
with the northern part of Mesopotamia. After the fashion
of so many oriental kings, he had sought to immortalize
himself by the foundation of a new capital, to which he gave the
name of Tigranocerta, and which he peopled in great measure
with captives carried off from his new conquests, including
many Greeks from the cities of Cilicia and Cappadocia.''
As soon therefore as the Armenian monarch had decided to
espouse the cause of Mithridates, and declared war against
Kome, LucuUus determined at once to strike a blow at the
heart of his dominions, by marching direct upon Tigranocerta.
In order to effect this he had to cross the Euphrates — which
was now for the first time passed by the Roman arms — as well
as the rugged chain of Mount Taurus : but all these obstacles
were successfully surmounted : and LucuUus, advancing
through Sophene,^ and crossing the Tigris in the upper part
of its course, appeared before Tigranocerta, and laid siege to
that city before Tigranes had completed his preparations for
defence, or Mithridates had arrived to support him. The
Armenian monarch in vain endeavoured to avert the fall of
his capital : he was defeated in a great battle, Tigranocerta
was taken ; and after halting for some time in Gordyene,
Lucullus advanced northwards with the view of penetrating
into the heart of Armenia itself. He was again encountered
by the combined forces of Mithridates and Tigranes at the
passage of the river Arsanias, where he again defeated them
with great slaughter. But when he wished to follow up his
advantage by pushing on at once to Artaxata, the ancient
capital of Armenia, the discontent of his troops compelled him

' Appian, Mithridat. c. 67; Plut.
LucuU. 21, 26; Strabo, xii. 2, p. 539.
See Note C, p. 104.

Sophene was the district bounded

by the Euphrates on the west, by the occasion.

Murad Chai on the north, and extend-
ing on the east to beyond the western
source of the Tigris. Its name is
mentioned for the first time on this

Sect. 6.

THE JUGURTHINE AND MITHRIDATIC WAES.

to abandon the idea, and to return southwards across the
Taurus, where he laid siege to Nisibis in Mygdonia, as the
Greeks now called the north-eastern district of Mesopotamia.^
The reduction of this important fortress was the last of the
brilliant exploits of LucuUus : the discontent and mutinous
disposition of his troops not only prevented him from following
up his successes in the ensuing summer, but compelled him to
retreat into Pontus, and remain there in a state of inactivity,
while Mithridates, supported by Tigranes, recovered a great
part of his former dominions.^

§ 8. These campaigns of Lucullus in Asia derive a special
interest from their being the first occasion on which the Roman
arms were carried across the Taurus into the regions adjoining
the Euphrates and the Tigris, which subsequently became for so
long a period the constant battle-field between them and their
Oriental neighbours the Parthians and Persians. They were
also the first to bring either Glreeks or Romans into somewhat
closer acquaintance with the upland regions of Armenia, a
cold and dreary tract, concerning which we have scarcely any
information since it was traversed by Xenophon and the Ten
Thousand in their marvellous retreat from the banks of the
Tigris to the Euxine.^ Unfortunately our accounts of these
operations are extremely imperfect : neither Plutarch nor
Appian, from whom our information is principally derived,
troubled themselves much with the geography of the countries
in question, and several of the leading points are subject to
much doubt. Even the position of Tigranocerta cannot be
considered as satisfactorily determined^ : and the river Ar-
sanias must be included in the same category.* Talaura also,
a mountain fortress of Mithridates, where he deposited a great

^ The name was given to it (as we
are expressly told by Strabo (xvi. 1, p.
747), by the Macedonians, evidently
with reference to the region of the same
name in Macedonia, but its origin is
not explained.

Nisibis, which afterwards played so
important a part in the wars between

the Romans and Parthians, here ap-
pears in history for the first time.

1 Appian, Mithridat. 84-91; Pint.
Lucull. 24-33.

^ Xenophon, Anab, iv. See Chapter
X. p. 351.

3 See Note 0, p. 104.

* See Note D, p. 105.

88 HISTOKY OF ANCIENT GEOGKAPHY. Chap. XVIIL

part of his treasures,^ has not yet been identified. Artaxata,
the former capital of Armenia, which was reported, by a very
strange tradition, to have been founded by Hannibal,^ was situ-
ated in the valley of the Araxes, about 15 miles below Erivan.
It probably became again the capital of Tigranes, when he was
compelled by Pompey to withdraw within his original frontiers,
and still retained that dignity down to a later period.

§ 9. When Pompey succeeded LucuUus in the command,
B.C. 66, he found that Mithridates had recovered the greater part
of his original dominions, while Tigranes also had reoccupied
Armenia Minor and great part of Cappadocia. But the move-
ments of the Eoman general were rapid and decisive. He
quickly defeated Mithridates, drove him out of Pontus, and com-
pelled him to take refuge in Colchis, while he himself struck into
the heart of Armenia and was advancing directly upon Artaxata,
when he was met by Tigranes, who laid his tiara at his feet, and
purchased favourable terms of peace by this abject submission.
He was, in consequence, left in possession of Armenia Proper,
while the provinces of Sophene and Gordyene were erected
into a separate principality for his son. Pompey next advanced
northwards in pursuit of Mithridates, passed the river Cyrus or
Kur, and defeated in succession the two warlike nations of the
Albanians and Iberians, who had sought to oppose his progress
in arms. Both these tribes appear on this occasion for the
first time in history : they had in all probability hitherto main-
tained a wild independence in their mountain homes, without
acknowledging the rule either of the Persian or the Macedonian
kings. The Iberians occupied the upper valley of the Kur,
and the mountain slopes that separated it from the basin of the
Euxine, while the Albanians held the lower valley of the same
river, and the mountain tract from the foot of the Caucasus to
the shores of the Caspian.' Pompey himself was desirous, after
defeating the army of the Albanians, to have penetrated in
person as far as the Caspian Sea, and actually advanced within

* Appian, Mithridat. 115. * Strabo, xi. p. 528 ; Plut. ImcuU. 31.

' Plut. Pomp. 34.

Sect. 6. THE JUGUKTHINE AND MITHRIDATIC WARS. 89

three days' march of its shores ; but was deterred, we are told,
by the number of venomous reptiles he encountered,^ an idle
tale evidently got up by the natives.^

§ 10. Meanwhile Mithridates had effected his retreat in
safety to the Bosporus. After wintering at Dioscurias — the
extreme limit of Greek civilization in this quarter — he had
forced his way at the head of an army all along the eastern
coast of the Euxine, from thence to Phanagoria on the Asiatic
shore of the Bosporus — a march of unprecedented difficulty,
both from the rugged and mountainous character of the
country and the warlike and lawless disposition of the in-
habitants. The whole of this tract between the Caucasus and
the Euxine was occupied by a number of wild tribes, speaking
different dialects, and regarded by the Greeks as different
nations, to whom they gave the appellations (evidently much
modified to suit a Greek ear) of Heniochi, Zygi, and Achsei.^
With some of these different tribes, Mithridates had already
entertained relations, partly of a peaceable, partly of a hostile
character ; and on the present occasion also he appears to have
effected his passage in part by negotiation and the influence of
his name, as well as by force of arms.^ But under all circum-
stances the accomplishment of this march of more than 300
miles through a country that presented the greatest natural
difficulties, and that had never previously been traversed by an
army, is certainly one of the exploits that redounds the most to
the credit of the Pontic king.

Pompey did not attempt to pursue the fugitive monarch
beyond the Phasis ; and contented himself with sending his
fleet to watch his proceedings at the Bosporus, while he himself
proceeded southwards to complete the subjugation of Syria. It

Plut. Pomp. 36. I in summer from the multitudes of ser-

^ It is curious to find that the same
fable is still current in this neighbour-
hood, and is gravely repeated by an
intelligent German traveller, who was
assured that the Mughan Steppe (be-
tween the lower course of the Araxes
and the Caspian) was wholly impassable

pents with which it swarmed. (Kohl.
Reisen in Sud-Bussland. vol, ii. p. 170.
See the remarks of Petzholdt, Der
Kauhasus, vol. i. p. 198.)

1 Strab. xi. p. 497.

^ Appian, Mithridat. c. 102 ; Strab.
xi. p. 496.

90 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

was on this occasion that for the first time he carried the
Eoman arms into Judsea, and entered Jerusalem at the head of
an army, B.C. 63. Before that time it was probably very little
known to the Eomans. He appears to have considered that
no further danger was to be apprehended from Mithridates,
who had been driven to the shores of the Palus Mseotis, while
it would be a difficult and hazardous enterprise to attempt
to follow him into these remote regions. The aged monarch
was, however, very far from being disposed to acquiesce in his
defeat, and after having established himself at Panticapseum,
which still continued to be the chief centre of civilization and
trade in these parts, and confirmed his power over the sur-
rounding districts, he began to extend and strengthen the
alliances which he had already formed with the Scythian and
other tribes that occupied the steppes extending from the shores
of the Borysthenes to those of the Danube. His object was to
combine these various nations, including the Bastarnse and the
Getse, as well as the Gaulish tribes, who were at this time
settled between the Danube and the Adriatic, the Scordiscans,
Boians,^ &c., into one great mass, and putting himself at their
head to cross the Alps and throw himself at once upon Italy,
where he conceived the Komans to be still the most vulnerable.
But this gigantic scheme was nipped in the bud by the defec-
tion of his own troops, as well as of the recently annexed Greek
cities. His son Pharnaces put himself at the head of the revolt,
and Mithridates was compelled to put an end to his own life
(B.C. 63). With him perished all his projects, and we hear
little of any incursions of the barbarians from this quarter
until a much later period.*

§ 11. On the side of the Danube indeed the Eoman arms had
made but little progress, though the successive Eoman generals
who were appointed to the province of Macedonia repeatedly
endeavoured to earn the distinction of a triumph by hostilities
against the barbarians that adjoined them on the north. It has

See Note E, p. 105. * Appian, Mithridat. 107-111.

Sect. 6.

THE JUGURTHINE AND MITHEIDATIC WARS.

been already stated, that C. Scribonius Curio in B.C. 75 was the
first Koman general who penetrated to the banks of the great
river. But he appears to have merely made a hasty expedition
thither and returned. The nation over whom he celebrated a
triumph was the Dardanians, who at this period appear to have
held the mountain tract at the head of the valley of the Morava,
on the confines of Moesia and lUyricum. His successor,
M. LucuUus (the brother of the conqueror of Mithridates),
turned his arms against the Bessi, a Thracian tribe who occu-
pied the mountain range of Hsemus (the Balkan) and the
upper valley of the Hebrus, and after subduing them, overran
the open country of Moesia, as far as the right bank of the
Danube. He did not indeed attempt to secure the possession
of this region, which was not reduced to the form of a Eoman
province until long after; but turned his arms against the
Greek cities ou the shores of the Euxine, and took or reduced
to submission Tomi, Istrus, Odessus, Callatia, Mesembria, and
Apollonia.^ For these successes he was rewarded with a
triumph on his return to Eome in B.C. 71. But no real progress
was made at this period towards the subjugation of any of the
tribes north of Mount Haemus; while these continual petty
wars tended to keep up a feeling of irritation and hostility
towards the Komans among the barbarians adjoining their
northern frontier, of which the design of Mithridates was ably
conceived to take advantage.

§ 12. Both Lucullus and Pompey had entered more or less
into relations, partly hostile, partly amicable, with the king
of Parthia, and Lucullus is even said to have at one time
meditated turning his arms against that monarch, leaving

* Appian, lUyr. c. 30 ; Eutrop. vi.
10. But there is much confusion in
the lists of these towns. Appian dis-
tinctly states that there were six of
them ; and proceeds to enumerate Is-
trus, Dionysopolis, Odessus, Mesembria,
Callatis, and Apollonia. (The text is
corrupt, but the two last names may be
safely restored.) Butropius, on the
other hand, gives Apollonia, Callatis,

Parthenopolis (a name otherwise un-
known), Tomi, Histrus, and Burziona,
evidently the Bizone of Strabo (vii. 6,
p. 319), but an obscure town. I have
chosen the six most important names ;
but there is no reason to doubt that
he captured the smaller towns also.
Apollonia alone was destroyed; the
others were probably admitted to
favourable terms.

92 HISTORY OF ANCIENT GEOGEAPHT. Chap. XVIII.

Mithridates and Tigranes to themselves.^ Plutarch indeed
does not hesitate to ascribe to that general the blame of the
subsequent disasters of the Eoman arms in this quarter; it
being, as he asserts, the trophies erected by him in Armenia
close to the Parthian frontier, and the capture of Tigranocerta
and Nisibis, that excited Crassus to emulate his glory by
assailing the Parthians themselves.'' But it is certain that
from the time that Syria was reduced to a Roman province
and their frontier was thus carried to the Euphrates, the out-
break of hostilities between the two powers became inevitable.
The Parthians at this time held the whole of Mesopotamia, up
to the frontiers of Armenia, having recovered possession of the
districts of which they had been deprived by Tigranes.

§ 13. The ill-fated expedition of Crassus (b.o. 53) was not
calculated to throw any light upon the geography of these
regions, and our knowledge of its details is, as usual, during
this period of Eoman history, very scanty and imperfect. It
is certain however that he crossed the Euphrates at the point
known as the Zeugma,® nearly opposite the modern Bir, where
a bridge of boats had been permanently established by the
Seleucidan kings, in consequence of which it appears to have
already become the customary place of passing the river, instead
of Thapsacus.^ From hence he might either have advanced
through the north of MesojDotamia, by Edessa and Nisibis,
keeping near to Mount Masius on his left, till he reached the
Tigris, and then descended the left bank of that river, or he
might at once have descended the valley of the Euphrates
(as the younger Cyrus had done), keeping the river on his
right, so as to secure him supplies of provisions. Instead of
adopting either of these courses he appears to have taken an
intermediate line, and marched through the open plains of
Mesopotamia where his troops suffered severely from heat and

^ Plut. Lucull. c. 30. I poses it to have been the same place

' Id. ibid. c. 36. where Alexander had crossed the

* Dion Cass. xl. 17. His statement i Euphiates. See the next note.

of this fact may be received without | '■' Sec Note F, p. lOti.

question, though he erroneously sup- i

Sect. 7.

POSIDONIUS.

drought. He did not however advance many days' march
before his progress was arrested by the Parthian army ; and
it seems probable that the Roman historians exaggerated
the natural difficulties that he met with, in order in some
degree to extenuate his disaster. The farthest point that he
reached was a river, called by Plutarch the Balissus,^ which
must in all probability be the same as that called by other
writers, the Beliche, or Balichas, and still known by the name
of Nahr Belik, which falls into the Euphrates near Eakka. It
was here that his first defeat took place, and that his son was
killed ; ^ after which he fell back upon Carrhse, a fortified town
that was at this time held by a Eoman garrison. Of the site
of this place, which was the scene of his final disaster and
death, there is no doubt ; it is still called Haran, and is situated
about 60 miles from Bir, where Crassus crossed the Euphrates.^
The statement of Plutarch that when Crassus set out on this
ill-omened expedition, he looked forward not only to the total
defeat of the Parthians, but to carrying his arms beyond their
frontiers " to the Bactrians and Indians and the external ocean "*
shows the fixed impression that that sea was to be found imme-
diately beyond the two nations in question.

Section 7. — Posidonius.

§ 1. Contemporary with the events which we have been
passing in review, was an author, who though, like Hippar-
chus, he did not compose any strictly geographical treatise.

» Plut. Crass. 23.

^ This first battle seems to have
been fought in the neighbourhood of a
place called Ichnse, which we know
from Isidore of Charax to have been
situated on the river Balichas (the
Belik), only 5 schoeni (150 stadia) from
Nieephorium on the Euphrates (Isidor.
Mans. Parth. § 1).

3 Plut. Crass. 27-31 ; Dion Cass. 1.
25-27 ; Oros. vi. 13. The defeat and
death of Crassus are mentioned by all
these writers in connection with

Carrhse; but the spot where he was
finally surrounded and slain, was near
a place called Sinnaca, situated on the
heights bordering the plain of Mygdo-
nia. Crassus himself, with the re-
mains of his army had broken up from
before Carrhse with the view of gaining
these heights, when he was intercepted
by the treachery of Surenas. (Plut.
Crass, c. 29 ; Strabo, xvi. 1, § 23, p.
747.)

* Id. ibid. c. 16.

HISTORY OP ANCIENT GEOGRAPHY.

Chap. XVIII.

yet exercised considerable influence upon the progress of geo-
graphical science, and whose writings are among those most
frequently referred to by Strabo, either as authorities, or for
the purpose of criticism. This was PosiDONius, a philosopher
of the Stoic school, who was born at Apamea in Syria about
the year B.C. 135, and became the successor of Pansetius at
Ehodes, where he enjoyed the reputation of being one of
the most distinguished philosophers of his day, and formed
friendly relations both with Cicero and Pompey. He was sent
ambassador to Eome on the part of the Khodians, during the
last illness of Marius, B.C. 86 ; and visited that city again for
the last time in B.C. 51, when he was not less than 84 years of
age.^ Besides several works of a purely philosophical charac-
ter, he left a great historical work in not less than 52 books,
forming a continuation of that of Polybius, and extending
to the end of the Mithridatic Wars, of the whole course of
which he had been himself a contemporary witness. This
history, like that of Polybius, appears to have contained many
notices of a geographical character, accounts of the manners
and customs of the various nations mentioned, and observa-
tions on physical phenomena, which were in some instances
at least the result of his own experience, as he had travelled
extensively, and visited in person the coasts of Spain, Gaul,
and Liguria.® His only work that was more specially con-
nected with geography was one " On the Ocean ;" in which he
appears to have treated fully — as was indeed almost inevitable
in connexion with such a subject — of the figure and dimen-
sions of the earth, and the general principles of mathematical
geography.

* For the dates and facts of the life
of Posidonius, see Clinton {F. H. vol.
iii. p. 540), and the account of his life
prefixed to the fragments of his works
by C. Miiller in his Fragmenta Histori-
corum Grsecorum, vol. iii. All the ex-
tant fragments of his writings are col-
lected by Bake (Posidonii Bhodii
Beliquix Doctrinx, 8vo. Lngd. Bat.
1810) ; those that have any bearing on

history or geography will be found in
the work of Miiller alre^idy cited.

" The period of his life at whicli lie
travelled thus extensively is uncertain ;
but it was probably in his earlier years.
At all events it was certainly before
the publication of his historical work,
in which he embodied many notices
that were the result of his personal ob-
servations and inquiries.

Sect. 7.

POSIDONIUS.

§ 2. It was apparently in this work ^ that Posidonius intro-
duced his new attempt to determine the circumference of the
earth, which he undertook independently of that of Eratos-
thenes ; and arrived at a widely different result. His method
however like that of his predecessor was scientific and sound
in theory. Having observed that the star Canopus, which
from its brilliancy had attracted the especial attention of the
Greeks from the time they settled in Alexandria, was only just
visible on the horizon at Rhodes, while at Alexandria it rose
to a height equal to a fourth part of one of the signs of the
zodiac, or a forty-eighth part of a great circle ; assuming more-
over (as Eratosthenes had done before him) that Rhodes and
Alexandria were on the same meridian, and that the interval
between them was 5000 stadia, he arrived at the conclusion
that this distance was a forty-eighth part of the whole meri-
dian circle, and therefore the circumference of the earth was
240,000 stadia.^ This conclusion did not differ very widely
from that of Eratosthenes, while it was somewhat nearer to
the truth. Yet it was based upon two matei'ial errors. In
the first place he accepted without enquiry the rough popular
estimate of 5000 stadia for the distance between Rhodes and
Alexandria ; though this was greatly in excess of the truth,
and founded merely on the vague estimate of navigators.^
But on the other hand his supposed observations of Canopus
were grossly inaccurate, the real difference in latitude between
Alexandria and Rhodes being only about 5° 15', or less than
a sixtieth part of a great circle, instead of a forty-eighth ! It
was only because these two enormous errors in great part
counteracted one another that the result was apparently so
fair an approximation to the truth.

Unfortunately Posidonius seems to have subsequently become
aware of the error in his estimate of the distance between

' I say " apparently," because we are
told that Posidonius wrote also a treatise
Trepi KdcTfiov, and another irepl fieredptav,
in either of which such a subject might
well have been introduced ; and Cleo-

medes does not mention the name of
the work from which his account is
taken.

* Cleomed. Cycl. Theor. i. c. 10, p. 62.

9 See Chapter XVI. p. 639.

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVIII,

Rhodes and Alexandria, and adopted the reduced computation
of Eratosthenes — which was itself, as we have seen, founded
on calculation, not on measurement, — of 3750 stadia for the in-
terval between the two. But as he retained his own assump-
tion, — founded on supposed observations, but of so rude and
vague a character as to give utterly erroneous results — that
the difference of latitude between the two was equal to ^V of a
great circle, or 7^ degrees, it followed that he reduced the whole
circumference to only 180,000 stadia, or just three-fourths of
his former computation.^ Yet, by a strange fatality, this con-
clusion, the result of such a complication of errors, came to be
generally accepted by the later Greek geographers in pre-
ference to that of Eratosthenes, and was even adopted by the
great astronomer Ptolemy.

§ 3. While Posidonius thus reduced the supposed cir-
cumference of the globe, his estimate of the length of the
" habitable world," — which, as we have seen, was considered
in his day as the only proper subject of geography — did not
materially differ from that of Eratosthenes. He made it indeed
somewhat less, reckoning it at about 70,000 stadia,"-^ which
he considered as being just about half the circumference of
the globe, as measured on that circle of latitude : i.e. on the
circle passing through Rhodes and the Straits of Gibraltar,
along which the greatest length was always measured. Thus,

* Strabo, ii. 2, p. 95. It is true that
the account here given of the mode by
which Posidonius arrived at this second
result, is mainly conjectural. For
Cleomedes, to whom we are indebted
for the details — perfectly clear and
intelligible in themselves— of the pro-
cess by which Posidonius attained his
first result, of 240,000 stadia, does not
say anything of his having subse-
quently altered it ; while Strabo and
other writers, who refer to him as the
author of the measurement making
the circumference 180,000 stadia, do
not allude to his having come to any
other result. The mode of reconciling
the two, adopted in the text, was first

suggested by Eicciolo (an Italian astro-
nomer of the seventeenth century), who
was followed by Lalande and other
writers on astronomy, as well as by
Gossellin ; and the exact coincidence
of the numbers (48 times 3750 being
p]-ecisely equal to 180,000) is such as to
leave hardly a doubt of its correctness
(see Ukert, Geogr. vol. ii. p. 48).

^ Eratosthenes, as we have seen,
made it more than 77,000 stadia. But
this included an addition at each end,
which were probably omitted by Posi-
donius, who expressly called the Sacred
Promontory the westernmost point of
the world.

Sect. 7. POSIDONIUS. 97

be observed, tbere would be only so many myriads of stadia
(70,000) to be traversed by any one wbo, setting out from the
west, with an east wind, would sail to India.^ It is curious to
find him, like Eratosthenes, again speculating on the circum-
navigation of the world, and anticipating by so many centuries
the project of Columbus of sailing direct from Spain to the
Indies.

§ 4. On another point, which could be merely the result of
speculation, we find Posidonius concurring with Eratosthenes
and Polybius in assuming that the tract immediately under
the equator enjoyed a milder temperature than the burning or
torrid zones to the north and south of it.* In regard to the
unity of the ocean, and the possibility of circumnavigating
Africa, Posidonius also held the same view with Eratosthenes
and Strabo: maintaining its theoretical possibility, but not
admitting that it had been ever actually accomplished. It
was in connection with this subject that he related the curious
history of the voyage of Eudoxus of Cyzicus, which has been
already examined.^

§ 5. Another subject, on which Posidonius appears to have
been the first Greek writer that arrived at clear ideas, was that
of the tides. For this he was indebted to his journey to Spain,
where he spent some time at Gades, and from his own obser-
vations, coupled with the information he received from the
natives, acquired a distinct knowledge not only of the diurnal
recurrence of the tides, but of their monthly cycles of vari-
ation, which he correctly ascribed, to the influence of the moon,
and its different positions with regard to the sun : so that the
highest tides, as he observed, always coincided with the full
moon, and the lowest with the half-moon, or intermediate
quarters.^

^ 'TTTOVoe? Se (o UoffetScivtos) rh rrjs
oiKovfiePTis jxTjKOS 67rTa iron /j.vpiadooi> ffra-
Sicov vTT&pxov ^fxicrv eivai rod '6\ov kvk\ov
Kad' t)v elK-riTTTOLi, &(TT€, (prjcrlv, airh ttjs
SvcTfws evpcf) trXioov iv roffavrats fivpidaiv
eA0ot &j/ els 'IvSovs. Strabo, ii. 3, p.

102.

^ Posidon. Fr. 72, ed. Muller.

* Posidon. ap. Strab. ii. 3, p. 98. For
the voyage of Eudoxus, see Chapter
XVIII. p. 74.

" Id. lb. iii. 5, pp. 173, 174.

VOL. II. H

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XVIII.

Posidonius, who enjoyed a high reputation as a physical
philosopher, was naturally led to pay particular attention to
the phenomena connected with physical geography, especially
those of volcanoes, earthquakes, &c. He was the first to record
the appearance of a new islet, among the Lipareean Islands,
thrown up by volcanic agency, between Hiera and Euonymus :
and his account of the circumstances attending the outbreak
is almost precisely in accordance with those of modern obser-
vers, who have witnessed similar phenomena.'^ He was indeed
strongly impressed with the changes in the earth's surface
that had been produced by movements of elevation or subsi-
dence (as they would be termed by a modern geologist) owing
to earthquakes and similar causes : and even went so far as
to suggest that the Atlantis' of Plato might not be a mere
fiction, but that an island equal to a continent in size might
really have disappeared and sunk into the depths of the
ocean.^

§ 6. It is unnecessary to refer more particularly to the
numerous scattered notices that are found in Strabo, cited
from Posidonius, especially in regard to Spain and Gaul, for
which he appears to have been one of the principal authorities
followed by the later geographer.^ One curious statement may
be mentioned, that the greatest depth of any sea thai had been
measured was that of the Sardinian Sea which was not less
than 1000 fathoms.'^ He stated that tin was found in the
interior of Spain, north of Lusitania, and among the Artabri,

' Posidon. ap. Strab. vi. p. 276. The
date of this event is fixed by Pliny (U.
N. ii. 88, § 203) to the year 126 b.c,
which coincides with the expression of
Posidonius, that it had occurred within
his own memory.

^ <Jt£ eVSe'xeTat koI fir) izXacTfia eivai
rb TTEpl T7JS vfjffov rris 'ArXavrlSos.
Strabo, ii. 3, § 6, p. 102.

" It was from Posidonius that Strabo
derived his strange story of an island
at the mouth of the Loire, inhabited by
a race of women whom he calls Sam-
nitfe, who were wholly devoted to the

celebration of Bacchic rites (iv. 4, § 6,
p. 198).

The same tale is repeated by Diony-
sius Periegetes (vv. 560-569), who
writes the name Amnitse.

' Posidon. ap. Strab. i. 3, § 9, p. 54.
The fact is true that there is " a pro-
digious depth of water around Sardinia
and Corsica," to lase the words of
Admiral Smyth {Mediterranean, p.
137) ; but that it had ever been sounded
in ancient times to a depth of 1000
fathoms may well be doubted.

Sect. 7. POSIDONIUS. 99

but was brought also from the islands of the Cassiterides and
the British Islands to Massilia.^ This is the first instance in
which we find mention of the overland trade in tin, which
was certainly" an established practice in Caesar's time : but
Posidonius, who had himself visited Massilia, had doubtless
acquired his information in that city.

One point on which Posidonius entertained strangely erro-
neous views was in regard to the distance between the Euxine
and the Caspian, which he estimated at only 1500 stadia.^
Strabo, who reckons it double this width, or 3000 stadia,* justly
remarks that this error is the more inexcusable, as the recent
campaigns of Pompey against the Iberians and Albanians had
opened the way to better information concerning these
countries. It is remarkable that Posidonius added, that he
believed the distance from the Palus Mseotis to the Ocean
was not much greater.^ This notion of the proximity of the
northern ocean to the Mseotis and the Caspian seems to have
become gradually implanted in the minds of the Greek geo-
graphers at this period from their belief in the direct commu-
nication of this external sea with the Caspian. Their knowledge
of the geography of these regions was too imperfect to lead them
to see the necessity for a large tract of land to the north, to
supply the waters of the Tanais and other great rivers.

In one instance Posidonius appears to have been misled by
his own experience. For, having encountered contrary winds
on his return voyage from Gades to Eome, and been driven
about by them between the Gymnesian Islands and Sardinia
and the opposite coast of Africa, so that he did not reach Italy
till after a voyage of three months, he erroneously assumed
that the east winds blew with the same regularity and violence
in this part of the Mediterranean that the Etesian winds did
in the seas more familiar to the Greeks.^

^ Posidon. ap. Strab. iii. p. 147.
« Ibid. xi. 1, § 5.
* Ibid. § 6. " Ibid. § 5.

" Strabo, iii. 2, p. 144. It was during
this voyage that lie on one occasion

approached near enough to the coast of
Africa to observe the apes on the shore,
of whose appearance and habits he gave
a lively description, which is cited by
Strabo (xvii. p. 827).

11 2

100

HISTORY OF ANCIENT GEOGEAPHY. Chap. XVIII.

Posidonius appears to have been the first who compared the
form of the inhabited world to that of a sling, broader in the
middle and tapering at the two ends : a simile adopted by
several later geographers, though conveying but little idea to
a modern reader.'

§ 7. To the period immediately following Posidonius may
probably be assigned two astronomical writers whose works have
been preserved to us, though their names are otherwise unknown
and their date cannot be fixed with any certainty. Of these
Cleomedes, the author of a treatise called "The Circular
Theory of the Heavenly Bodies," ^ containing an exposition of
the received views concerning the magnitude and figure of
the earth, its position in the centre of the universe, and the
motions of the sun, moon, and stars, expressly disclaims all
pretension to originality, and says that the doctrines which it
sets forth are not the opinions of the author himself, but are
compiled from various writers ancient and modern, but are
chiefly derived from Posidonius.^ It is probable therefore
that the little work in question is in fact an epitome of the
views of Posidonius. Its principal interest arises from his
having preserved to us the accounts which we have already
examined of the mode of proceeding adopted by Eratosthenes
and subsequently by Posidonius to determine the circum-
ference of the earth.-^ Cleomedes, as well as Posidonius him-
self, belonged to the Stoic sect of philosophers, who' were at
this time the upholders of scientific astronomy, as opposed to
the Epicureans, who adhered to the popular notions on astro-
nomy, and ridiculed the notion that the earth was spherical
and situated in the centre of the universe. Thus we find
Lucretius in his well-known poem arguing strongly against
the idea of the tendency of all things to a centre, which

' Agathemerus, i. § 2.

* KvkAikti @eoopia rSiv Merewpaiv, in
two books. It is publisherl by Bake,
8yo, Liigd. Bat. 1820. (Concerning its
author and his probable date, see Sir
G. Lewis's Hid. of Ancient Aatronomy,

p. 216; and the article Cleomedes, in
Dr. Smith's Diet, of JBiogr. vol. i. p. 792.

" ii. ad fin.

' See above, p. 95, and Chapter
XVI. p. 621.

Sect. 7. GEMINUS. 1 01

had been the basis of cosmography ever since the time of
Aristotle.^

§ 8. Geminus, of whom like his supposed contemporary
Cleomedes nothing is known concerning his life or date, has
left a work under the title of " An Introduction to the Celestial
Phaenomena," ^ which is in fact an elementary introduction to
astronomy, very similar in its scope and character to that of
Cleomedes. It is therefore in like manner of interest as em-
bodying what were then the generally received notions con-
cerning those cosmographical relations of the world which are
necessarily connected with the general principles of geography.
Both these writers may probably be placed between the time
of Posidonius and that of Strabo, and the system enunciated
by them is obviously the same as that which is summarized by
the great geographer as generally recognized in his time by
all persons who had received a liberal education.*

"^ Lucretius de Berum Natura, i. yv.
1052-1082.

^ 'Elffayo}'/^ els ra ^aivofxiva. It IS
printed by Petavius in his Uranologium.

Concerning its author, see Sir G. Lewis
I. c. and the article Geminus in Dr.
Smith's Diet, of Biogr. vol. ii. p. 238.
* Strabo, ii. 5, § 2.

102 HISTORY OF ANCIENT GEOGRAPHY. Chap. XVIII.

NOTE A, p. 47.

GREEK MONARCHS OF BACTRIA.

The researches of recent archaeologists in India and the ad-
jacent regions towards the north-west, have thrown a great deal
of light upon the obscure notices which were found in ancient
writers in regard to the existence of this Greek monarchy in
Bactria, Ariana, and the upper valley of the Indus. They have
revealed to us the names of not less than thirty monarchs, who
ruled in this part of Asia, during the interval between the founda-
tion of the monarchy, about 250 B.C., and its destruction or subju-
gation by the Scythian tribes : a fact recorded by Strabo (xi. 8, § 2),
without any indication of its date, but which may probably be
placed about 120 B.C. Unfortunately the total absence of dates
upon these coins leaves their chronological arrangement in a state
of hopeless uncertainty. Two points alone seem to be clearly
established : it may be inferred from the purely Greek character of
the names, that the descendants of the original Greek settlers con-
tinued to retain the government in their own hands till the down-
fall of the monarchy ; and secondly, that there must frequently have
been two or more contemporary sovereigns ruling at the same time
in different parts of the widely-extended dominions, which were
occasionally united in the hands of one powerful monarch. This
is indeed sufficiently evident from the number of kings, whose
existence is established by their coins, as compared with the limited
period to which they can be assigned. But as soon as we attempt
to trace the division in detail, and determine what kings or
dynasties reigned in one province, and what in another, we are
met by difficulties almost equal to those attending the chronology.
The inferences drawn by the latest numismatical inquirers point
to the fact, which is indeed sufficiently probable in itself, that the
range of the Hindoo Koosh frequently formed the limit between
two separate kingdoms, and that one Greek ruler established
himself in Cabul and the Punjab, while another maintained the
provinces of Bactria and Sogdiana north of the great mountain
range. It appears also probable that the latter provinces were
overrun by the Scythian invaders to whom Strabo refers, and their
Greek rulers expelled, some time before the last Greek monarchs
were dispossessed of the Punjab and the adjacent regions.

Noteb. posidonius. 103

The reader who wishes for more information concerning the
Greek monarchies in this part of Asia, will find the subject fully dis-
cussed by Prof. Wilson in his Ariana Antiqua (4to. Lend. 1841), who
has brought together all that was known in his time. The results
of more recent researches are embodied in a series of papers by
Gen. Cunningham in the Numismatic Chronicle {N. S. vols, viii., ix.,
X., and xii.) ; and still more recently the whole subject has been
examined anew by Dr. A. von Sallet in the Zeitschrift filr Numismatih,
vol. vi. pp. 165-209. But it must be confessed that these dis-
coveries, interesting as they are to the numismatist, have con-
tributed very little to our positive knowledge, either in a historical
or geographical point of view.

NOTE B, p. 65.

MEASUKEMENTS OF MEDITEREANEAN ACCORDING TO
ARTEMIDOEUS.

The details as given by Agathemerus for this part of the line
are: —

From Myriandrus to the Cleides off Cyprus .. 1400 stadia.

thence to the' promontory of Acamas 1300

to the Chelidonian Islands off Lycia .. 1300

to Patara in Lycia 800

toEhodes 700

to Astypalgea across the Carpathian Sea .. 940

to the headland of Tsenarus 1450

to Cape Pachynus in Sicily 3600

to Lilybseum 1520

to Caralis in Sardinia 2800

and thence to Gades, passing by the Gymnesian

Islands 10,000 „

Here the large round number and the want of details show that
the estimate was but a vague one.

In regard to the above numbers I have adopted the corrections
proposed by C. Miiller for those marked with asterisks, not that
I consider them free from doubt, but the discussion of these details
is of little importance : the sum total being clearly established by
the agreement of the numbers as given in Pliny and Agathemerus.

104 HISTOEY OF ANCIENT GEOGKAPHY. Chap. XVIII.

NOTE C, p. 87.

TIGRANOCEETA.

The site of the capital of Tigranes has been the subject of much
doubt. It was placed by D'Anville and other early writers at Sert
or Sort on the Bohtan Tschai, the Centrites or eastern arm of the
Tigris ; but this is clearly an error arising from the apparent re-
semblance of name, which is a mere fallacy, the termination certa
or Jcerta signifying merely a city (Appian, Miihr. 67). There are
no ruins at Sort, nor does the site accord with the few particulars
recorded of Tigranocerta. St. Martin, on the other hand, follows
the Armenian historians, who are unanimous in identifying Tigra-
nocerta with Amida (the modern Diarbekr), the celebrated fortress
on the Tigris, which plays so conspicuous a part in the later wars
of the Eomans with the Persians. The same view is adopted by
Mr. Ainsworth, but is open to the unanswerable objection that
LucuUus is distinctly described as crossing the Tigris hefore he
advanced upon Tigranocerta ; and no mention of that river is found
in connection with the siege of the capital (St. Martin, Memoires sur
I'Armenie, vol. i. p. 171-173; Ainsworth's Travels in Asia Minor,
vol. ii. p. 361). Moreover, the manner in which the name is intro-
duced by Pliny, who thus describes the cities of Armenia, " In
majore (Armenia) Arsamosata Euphrati proximum, Tigri Carcathio-
certa, in excelso autem Tigranocerta ; at in campis juxta Araxem
Artaxata" {H. N. vi. 9, § 26), certainly indicates its occupying a
lofty situation, not on either of the great rivers. On the other
hand, Tacitus, in whose time Tigranocerta retained its ancient
name and was still a place of importance, places it on a river
which he calls Nicephorius, and at a distance of only 37 Eoman
miles from Nisibis (Tacitus, Annal. xv. 4, 5). The latter statement
is at variance with all other inferences. Perhaps the most plausible
conjecture is that which identifies Tigranocerta with the city after-
wards called Martyropolis, the site of which seems to be clearly
fixed at a place called Meja Farkin, about 27 miles N. of the Tigris
and N.E. of Diarbekr (see Kiepert's Map of Asia Minor). St.
Martin identified Martyropolis with Carcathiocerta, a city described
by Strabo (xi. p. 527) as the capital of Sophene, but this may pro-
bably be sought farther westward. Strabo, however, in one passage
(xvi. p. 747) distinctly places Tigranocerta in the region called

Notes D, E. POSIDONIUS. I05

Mygdonia, south of Mount Masius ; and in another (xi. p, 522),
though less definitely, he affirms the same thing. In both these
passages it is mentioned in connection with Nisibis. Unfortu-
nately a third passage (xi. p. 532), where he meant to describe its
position more accurately, is corrupt, and cannot be restored with
any certainty. Indeed the whole comparative geography of these
regions is still a mass of confusion.

NOTE D, p. 87.

KIVER AESANIAS.

The river Arsanias, though not noticed by Strabo, is mentioned
by Pliny (v. 24, § 84) as an affluent of the Euphrates, and it again
played an important part in the operations of the Eoman generals
in Armenia against the Parthian king Vologeses (Tacit. Annal.
XV. 15 ; Dion Cass. Ixii. 21). But the only clue to its identification
is that suggested by St. Martin, that it is the same with the Aradzani
of Armenian writers, a name applied by them to the river otherwise
known by its Turkish name of Murad Tchai, which is in fact the
southern arm of the Euphrates, and in some respects the more im-
portant of the two. In this case the military importance of the
river is fully accounted for (St. Martin, Mem. de VArmenie, vol. i. pp.
52, 171). It may be observed that no Greek writer mentions the two
arms of the Euphrates. Xenophon, where he speaks of crossing
the Euphrates on his retreat through Armenia (^Anah. iv. 5, § 2)
must undoubtedly refer to the southern branch, or Murad Tchai :
but he has no mention of the northern branch, which flows by
Erzeroum. Yet it is this latter river which is exclusively desig-
nated under the name of Euphrates by Strabo (xi. p. 527), as well
as by Pliny (v. 24, § 83).

NOTE E, p. 90.

GAULS ON THE DANUBE.

We find repeated mention in the wars of Mithridates of these
Keltae or Gauls, who are represented as bordering upon the Danube,
or as situated near the head of the Adriatic. They may very

I06 HISTOKY OF ANCIENT GEOGRAPHY. Chap. XVUI.

probably have been a remnant of those who invaded Greece and
Macedonia after the death of Alexander, who had remained behind
in Pannonia and the adjacent regions. Strabo distinctly tells us
that the Boii and Scordisci, who at one period occupied this tract,
from which they had been in his time expelled by the Dacians,
were tribes of Celtic or Gaulish origin (vii. pp. 293, 313). He
ascribes also the same origin to the Tauriscans, who occupied a
part of the regions called by the Eomans Noricum (lb. p. 293).
This position sufficiently explains the importance attached to their
co-operation by Mithridates, in regard to his proposed invasion of
Italy (see Appian, Mithridat. c. 119). He is not likely to have
entertained any relation with the Celts of Gaul itself, in the
Eoman sense of the term ; nor would their alliance have been of
material assistance to his plans. The ethnology of these regions
at this period is extremely obscure, but there seems no doubt that
there existed to a considerable extent Gaulish races, intermixed
with the Illyrian and Thracian tribes (Strabo, I. c). All these
tribes were, at a later period, included under the name of Pan-
nonians, when the Eoman province of that name was constituted ;
but they appear to have in reality belonged to diJBferent races, and
composed a very heterogeneous assemblage.

At this period, so far as we can gather from Strabo, it would
seem that the plains of Wallachia and Southern Hungary, north
of the Danube, were occupied by the Getse ; while the Bastarnae,
a Sarmatian tribe, held the tract extending from the Tyras
(Dniester) to the mouths of the Danube.

It is certain that Mithridates from an early period — probably as
soon as he had extended his own nominal dominion as far as the
Dniester — had begun to form diplomatic relations with the nations
beyond that river to the west, and to raise auxiliary levies among
them. Some of these were undoubtedly Gauls : and we find
Bitsetus, a leader of Gaulish mercenaries, in immediate attendance
on Mithridates at his death, when he was called on by the fallen
monarch to despatch him, on the express ground of his long and
faithful services (Appian, Mithridat. c. 111).

Note F. POSIDONIUS. IO/

NOTE F, p. 92.

ZEUGMA.

This fact is more than once mentioned by Strabo, and is essen-
tial to a clear comprehension of the geography of Mesopotamia. But
unfortunately that geographer has himself fallen into error by con-
necting the Zeugma, which was in his day become well known as the
established place of passage of the Euphrates, with that at Samosata,
considerably higher up the river ; hence he repeatedly designates
it as the Zeugma of Commagene (to Kara KofA,[xayr]vr)v ^eiSy/^a — and
in one place distinctly states that it was close to Samosata — ('Atto
%afji.oa-dT(iiv tijs T^OfXfJLayrjvrjs ^ Trpos rr} StaySacret kol tw Zevyixari Ketrat,
xiv. 2, § 29, p. 664). The distance also which he gives, of 2000
stadia from Thapsacus, would agree much better with Samosata
than with the other Zeugma.

But the position of the true Zeugma is fixed by the distances
given by Isidorus (nearly a contemporary of Strabo) in proceeding
from thence to Nicephorium on the Euphrates, as well as by that
given by Strabo himself (p. 749) of 1400 stadia (140 G. miles) to
the Gulf of Issus, which is very nearly correct. Pliny also gives
the distance of the Zeugma from Samosata at 75 Eoman miles,
which is a very fair approximation.

The fact is that there appear to have been in all three places of
passage in habitual use. 1. That at Thapsacus, which continued
to be in use from the earliest ages till after the time of Alexander,
but was disused long before the time of Strabo. 2. That opposite
to the modern Bir, which came to be generally adopted under the
Seleucidan kings, on account of its being the nearest and most
direct line from their capital of Antioch, and to which the name of
" the bridge " (Zeugma) came in consequence to be applied. 3.
That at Samosata itself, which was apparently the customary place
of passage for persons coming from Cappadocia and Asia Minor
(see Strabo, xiv. p. 664). But this must always have been a cir-
cuitous and inconvenient route from Antioch.

A considerable town gradually arose at the second of these three
places, which retained the name of Zeugma, and appears to have
risen to importance under the Eoman Empire, as numerous coins
are found with this name. There had previously been a city

I08 HISTOEY OP ANCIENT GEOGRAPHY. Chap. XVIII.

founded by Seleucus on the opposite bank, doubtless with a view
to secure the passage of the river. This is called by Strabo
Seleucia, by Isidorus Apamea, and must have occupied the site of
the modern Bir, or Birehjik, which is still the habitual place of
passage of caravans. (See Chesney's Euphrates.')

( 109 )

CHAPTEE XIX.

c^sak's wars.

§ 1. While the Eoman arms were extending the dominions
of the great Eepublic both towards the east and the west : and
her subject provinces now formed an uninterrupted chain
from the Euphrates to the Atlantic coasts of Spain, towards
the north they were still confined within extremely narrow
limits ; and even the range of geographical knowledge in this
direction was still surprisingly scanty. We have already
marked the successive steps by which the Eomans had esta-
blished themselves in the south-east of Gaul/ and formed in
that country a regular province extending from the Alps to
the Cevennes and the Garonne, and northward to the Lake
Lemannus (the Lake of Geneva), which marked at this period
the most northerly limit of the Eoman Empire.^ Within
these ' boundaries there can be no doubt that a strong tincture
of Italian civilization was rapidly diffusing itself, as it had
already done through the kindred province of Cisalpine Gaul
on the other side of the Alps : Eoman traders had established
themselves in large numbers in the principal towns of "the
province," ^ and the two Eoman colonies of Aquse Sextise and
Narbo became, as in so many similar cases, important centres
of civilization, as well as strongholds in support of the Eoman
dominion. But since the subjection of the Allobroges and
the annexation of Toulouse and its territory no attempt was
made to extend the actual limits of the Eoman territory in
this direction.

> See Chapter XVIII. sect. 1. I of Caesar, of " the Province," without

2 Cajs. de B. G. i. 2. any further distinction, should be still

3 Cicero, Orat. pro M. Fonteio, c. 1. retained as a proper name by the re-
It is singular that the name thus gion so well known even at the present

applied by the Komans before the time ' day as Provence,

no HISTORY OF ANCIENT GEOGRAPHY. Chap. XIX.

The Senate indeed pursued to some extent the same policy
as elsewhere, by entering into relations more or less friendly
with the different tribes or nations bordering upon the Eoman
province : more especially with the two powerful tribes of the
Arverni and the ^Edui, both of whom were honoured with the
title of allies and " friends " of the Eoman people.* Even the
German chieftain Ariovistus, who had crossed the Ehine with
a numerous body of his countrymen, and established himself
in the territory of the Sequani (Franche Comte) received a
similar appellation.^ But while the Eoman Government
could not regard with indifference the movements and political
relations of the barbarian nations on its immediate frontier, it
appears to have refrained from anything like active inter-
ference : a policy which was observed with unusual moderation
by successive governors of the province for a period of about
half a century.

During this interval however the newly established province
in Gaul had been almost overwhelmed by the passing tempest
caused by the invasion of the Cimbri and Teutones — a move-
ment which for a time was sufficient to strike terror into the
hearts of the Eomans themselves. Whatever may have been
the original abodes or ethnic relations of these two allied
nations — a question upon which there is much difference
of statement among ancient writers, and equal diversity of
opinion among modern ethnographers^ — there can be no doubt
that they invaded the Eoman province from the north, in con-
junction with the Helvetians and Ambrones — both of them
undoubtedly Gaulish tribes : and after defeating successive
Eoman consuls, who had in vain endeavoured to stem the
tide of invasion, they poured like a flood over the province
from the Alps to the Pyrenees. But this formidable inroad,
like many similar irruptions of the northern barbarians under
the Eoman Empire, seems to have swept over the country

Ctcs. de B.G. i. 33, 43. Tacit. Annal xi. 25. * Cses. B.G. i. 43.

" See Latham's Germania of Tacitus, p. 133.

Chap. XIX.

Cesar's wars.

Ill

without leaving any permanent traces : and after the great
defeat of the Teutones by Marius in the plains near Aquae
Sextiee (b.c. 102), the Eoman province appears to have speedily
relapsed into a state of tranquillity.

§ 2. The whole aspect of affairs was altered by the appoint-
ment of Csesar in B.C. 59 to the government of Transalpine
Gaul, which he for the first time combined with that of the
Cisalpine province of the same name.^ His object in accepting
the command was undoubtedly to rival the fame which Pompey
had earned by his successes in the East, and he made no
secret of his intention to reduce the whole of Gaul under
the dominion of Eome.^ In the course of nine years that
he remained in command, B.C. 58-50, he carried on hostilities,
either in person or by his lieutenants, in almost every part of
the country, from the Rhone to the coast of Brittany, and
from the Pyrenees to the mouths of the Rhine ; and when he
finally quitted the province to carry on the Civil War, he
could fairly boast of having reduced the whole of Transalpine
Gaul to a state of subjection as complete as that in which he
found the original Eoman province.

The first occasion for the exercise of his arms was furnished
him by one of those national movements that appear to have
been so common alike among the Gaulish and German races.
The Helvetians, who had previously taken part in the great
invasion of the Cimbri and Teutones, had determined to emi-
grate in a mass from the abodes they then occupied, with a
view to establishing themselves in a more fertile region in the
west of Gaul. Their superiority in arms to the greater part
of the Gaulish tribes was an acknowledged fact, due, according
to Caesar, to the continual wars with the neighbouring Germans,

' The provinces of Cisalpine Gaul
and lUyricum had been in the first
instance conferred upon Caesar by the
people, for a term of five years, at the
instigation of the tribune Vatinius, to
which the senate afterwards added
Transalpine Gaul also.

^ This is distinctly stated by Cicero,

in his oration De Provinciis Consulari-
bus, held as early as the spring of b.c.
56. " C. Cassaris longe aliam video
fuisse rationem. Non enim sibi solum
cum lis, quos jam armatos contra popu-
lum Romanum videbat, bellandum esse
duxit, sed totam Galliam in nostram
ditionem esse redigendam," c. 13, § 32.

112

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XIX.

in whicli they were involved by their exposed situation in
immediate proximity with those formidable warriors.^ Count-
ing on this superiority and on the terror of their name, they
had hoped to traverse the central districts of Gaul with little
difficulty. But the intervention of Csesar completely changed
the state of the case. He fell upon their rear-guard as they
were passing the Arar (Saone), and cut to pieces a large
number of them, belonging to the tribe or pagus of the
Tigurini.^ Following up the march of the main body, he
pursued them through the land of the ^duans, defeated them
a second time in the neighbourhood of Bibracte, and completed
their destruction within the borders of the Lingones, where
the remnant of this formidable host was compelled to sur-
render at discretion.

This first campaign is memorable not only for the ability
and boldness with which Caesar pushed forward into the heart
of Gaul, to a distance of more than 150 miles from the frontier
of the Eoman province ; but still more from the picture it
presents to us, by an eye-witness and an observer of first-rate
intelligence, of one of those great national movements, such
as we read of in earlier times among the Gauls, and which
became so frequent among the German and Slavonian nations
before the close of the Roman Empire. The Helvetians left
home (according to documents found in their camp and cited
by Ceesar) to the number of 368,000 souls, including men,
women and children. Less than a third of these (about
110,000) returned to their native country after their final

9 Cjbs. b. g. i. 1.

' The Helvetians were at this time
divided into four principal pagi or
cantons, of which the Tigurini were
one ; another, the pagus Verbigenus,
is also mentioned by Caesar {B. G. i.
27); the names of the other two are
unknown, but the hypothesis adopted
by Walckenaer (Geogr. cles Gaules,
vol. i. p. 311), and favoured by Dr.
Long, that they were the tribes men-
tioned by Strabo (iv. p. 183, vii. p. 293),
under the names of the Ambrones and

Toiigeni or Tugeni, as taking part in
the invasion of the Cimbri, is certainly
plausible enough. Both names sub-
sequently disappear from history.

The neighbouring tribes of the
Eauraci, Boii, Tulingi, and Latovici
joined the Helvetians in this move-
ment, but are expressly distinguished
from them (ibid. v. 29). Of these the
Eauraci certainly dwelt in the neigh-
bourhood of Basle : tlie Boii came from
Noricum (Caes. I. c.) ; the other two
are unknown.

Chap. XIX.

CiESAR's WARS.

113

defeat.^ Yet their chiefs appear to have found no difficulty
in moving this unwieldy mass across a large part of Gaul, and
had it not been for the intervention of Csesar, there is no
reason to doubt that they would have successfully effected
their original project of a national emigration.

§ 3. Caesar's second campaign (in the same year with the
preceding, B.C. 58) brought him in contact with a still more
formidable enemy. For some time previously to his arrival
the Gauls had suffered severely from the incursions of the
German tribes beyond the Ehine : and one of these German
leaders, a chieftain named Ariovistus,^ had succeeded in esta-
blishing himself in possession of a large portion of the territory
of the Sequani, while fresh swarms of invaders had either
already crossed the Ehine, or were assembling on its banks
with a view to occupy the more fertile and better cultivated
lands on the other side of it. In this state of things the
Gaulish chiefs implored the succour of Csesar, who imme-
diately took advantage of the opening : and after a fruitless
pretence of negotiation, attacked and defeated Ariovistus in
the plains of Upper Alsace, about 50 miles from the Ehine.*
The slaughter was immense ; the remains of the German army
were driven across the river ; and for some time all attempts
on their part to recross it in this direction were effectually
checked.^

2 Ca3s. B. G. i. 29. The statement
that only 92,000 of the original emi-
grants — ^just a fourth of the whole-
were capable of bearing arms, shows
how large a proportion of women and
children took part in a migratory move-
ment of this character.

^ Ariovistus is called by some modern
writers, king of the Suevi, but he is
never so termed by Csesar, who styles
him simjaly a " king of the Germans "
(rex Germanorum, B. G. i. 31). The
forces under his command were a mixed
multitude from a number of German
tribes, who appear to have flocked
around his standard for the sake of
gain. The cnumci'ation of these tribes

VOL. II.

in c. 51 does not seem to imply that
the Suevi were in any predominant
numbers.

■* The site of this battle cannot be
determined with any precision. The
statement that the routed Germans
fled from the field for about fifty miles
to the banks of the Ehine (c. 53 extr.),
does not prove that this was the shortest
distance to that river. But it renders
it improbable that it was fought so
near the Ehine as the site between
Miilhausen and Thann which is fixed
upon by the Emperor Napoleon (^His-
toire de Jules Cesar, vol. ii. p. 86).

' Gees. B. G. i. 37-54,

1 14 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XIX.

The political consequences of this victory were immense.
It may be said to have practically decided for centuries the
question whether Gaul should become a Roman province or
be subjected to its German neighbours. It is at the same time
interesting to the geographer as the first occasion on which a
Roman army ever came in sight of the Rhine — that great
river that was to form for so long a period the much contested
barrier between them and the barbarians.

§ 4. The campaigns of the following year (b.c. 57) were
spread over a much wider field. After the close of his opera-
tions in the preceding year Caesar had established the winter
quarters of his legions in the land of the Sequani, instead of
withdrawing them within the limits of the Roman province :
a step which was justly regarded by the Gauls as a sign of his
intention to subdue the whole country.^ The consequence
was that all the tribes of the Belgse, who at this period
occupied the whole region north of the Seine and Marne,
combined together to expel the Roman intruders.^ The Remi
alone adhered to the Roman cause, and with the assistance of
the intelligence furnished by them, and the support in arms of
the iEduans, Ceesar was able to carry the war at once into
the enemy's territory. He crossed the river Axona (Aisne),
defeated the combined forces of the Belgians who had attacked
his camp, reduced in succession the Suessiones, Bellovaci and
Ambiani ; and followed up his advantage by a decisive defeat
of the Nervii, the most formidable and warlike of all the
Belgian tribes. The Aduatuci, who occupied the country
about the confluence of the Sambre and Mouse, were the last
to oppose the Roman general in arms, but their capital city
or stronghold was taken after a short siege, and its fall was
followed by the submission of all the remaining Belgian
tribes.

While Ceesar himself was thus engaged in the north of
Gaul, he had detached P. Crassus with a single legion to the

« Cces. B. G. ii. 1. ' See Note A, p. 135.

Chap. XIX. CiESAK S WAKS. II5

regions bordering on the Western Ocean. His lieutenant was
apparently favourably received, and was able to announce the
submission and friendly disposition of the tribes known to the
Gauls by the name of Armoricans, who occupied Brittany and
the west of Normandy. These comprised the Yeneti, Osismii,
Curiosolitse, Unelli, Esuvii, Aulerci and Eedones. The nations
along the course of the Loire, the Namnetes, Andes, Turones,
and Carnutes were equally favourable : P. Crassus took up his
winter quarters among the Andes,^ while other legions were
stationed among the Turones and Carnutes, and Csesar himself
repaired to Italy for the winter in the belief that the whole of
Gaul was effectually subdued.®

A few months sufficed to show the futility of this confidence.
Already before the middle of the winter Servius Galba, who
had been dispatched with a single legion to keep in order the
Alpine tribes in the upper valley of the Khone— the Nantuates,
Veragri and Seduni, was attacked in his winter quarters, and
though he repulsed the assailants with loss, he was compelled
to abandon the intention of wintering at Octodurus (Martigny)
where he had first taken up his quarters, and descend into
the more tranquil regions on the Lake of Geneva. The
object of Caesar in posting one of his legions in these moun-
tain regions was to secure the passage of the Pennine Alps
(the Great St. Bernard) — a pass which was already frequented
by traders, though exposed to many dangers and subject to
heavy exactions from the tribes who occupied it.^ This pur-
pose was for the time wholly frustrated : and we hear nothing
of its resumption by Csesar during his Gaulish wars.

§ 5. In the following spring, B.C. 56, a much more formidable
danger arose among the Armorican tribes in the north-west of
Gaul, which had been lately reduced to submission by P. Crassus.

* It was doubtless on this occasion
that P. Crassus collected the infor-
mation concerning the Cassiterides or
Tin Islands, which is referred to by
Strabo (iii. 5, § 11).

" Cscs. B. G. ii. 35.

' " Causa mittendi fuit, quod iter per
Alpes, quo magno cum periculo, mag-
nisque cum portoriis mercatores ire
consuerant, patefieri volebat" (B. G.
iii. 1).

I 2

Il6 HISTOKY OF ANCIENT GEOGEAPHY. Chap. XIX.

These tribes, finding that what the Romans aimed at was not
merely nominal submission, but real subjection, conspired
together with tbe view of expelling the invader. They were
essentially a maritime people, especially the Veneti, who beld
the chief control over the neighbouring seas, having many ships
of large size, with which they traded to Britain, and by the
habit they had acquired of navigating these stormy seas, had
attained a complete monopoly of the commerce of the adjoining
regions.^ With them were united the auxiliary squadrons
of the other Armorican states already alluded to ; but besides
these they obtained assistance from the more distant Morini
and Menapii, and drew succours from the opposite island of
Britain — tbe name of which here appears for the first time in
Roman history.^

The war was long protracted and laborious ; principally owing
to the nature of the country, which is admirably described by
Csesar. The land of the Veneti — the present department of the
Morbihan — was low and rocky, intersected by shallow inlets of
the sea, on the promontories and tongues of land between which
their towns were situated, in positions almost inaccessible by
land, and difficult of approach by sea on account of the shoals
and rocks, and rapid alternations of the tides, to which the
Romans were little accustomed. It was not till towards the
end of the summer that Csesar was able to assemble a fleet
capable of coping with that of the Veneti and their allies, who
were able to put to sea with not less than 220 ships, fully
manned and equipped, and far superior in size and strength
to those that the Romans could bring against them.* The
account given by Caesar of the ships used by the Veneti is
remarkable, and shows an advance in navigation far beyond
that usually ascribed to these semi-barbarous nations. They
were of large size, rising so high out of the water that
the Romans could hardly assail them with missiles, and even
when they raised turrets on their galleys these did not equal

2 Cces. B. G. iii. 8. ^ Ibid. iii. 9. ■• Ibid. iii. 14.

Chap. XIX. CiESAR S WAES. tl/

in height the poops of tlie Gaulish ships. At the same time
they were built wholly of solid oak, so that the beaks of the
galleys made very little impression upon them. Their sails
were made of hides or leather to withstand the violence of the
gales in these seas, and their anchors were fastened with chains
instead of cables.^ The Roman fleet was composed in part of
long ships or galleys built in the Loire expressly for the occa-
sion, partly of vessels furnished by the Gaulish tribes, of the
Pictones and Santones who adhered to the Roman alliance.®
But the courage and skill of Decimus Brutus, who commanded
it, triumphed over all the advantages of the enemy, and he
defeated the allied fleet in a decisive action with such loss,
that it was immediately followed by the submission of the
Veneti and of all their maritime allies.

The other operations of the year were of comparatively little
importance. But while Csesar was engaged in the war with
the Armoricans, his lieutenant P. Crassus had reduced to sub-
jection almost the whole of Aquitania, extending — as the term
is used by Caesar — from the Garonne to the Pyrenees. A few
mountain tribes alone remained in arms, whom he was deterred
by the lateness of the season from following into their rugged
fastnesses.'^ Csesar himself before the close of the season made
an expedition against the distant nations of the Morini and
the Menapii, who had made no signs of submission, but they
retreated before his approach into the vast forests and marshes
with which their country was almost wholly covered, into
Avhich he found it impracticable to pursue them.^

§ 6. At the close of this third year's campaign, as remarked
by Dr. Merivale, " the only members of the Gaulish race
who retained their liberty were the mountain tribes of the
Pyrenees and the amphibious wanderers of the Waal and the
Scheldt."^ But Csesar did not want employment for his
legions, and the campaign of the following year (b.c. 55) was

* Csesar, B. G. iii. 13, 14.
« Ibid. 9, 11.
' Ibid. 20-27.

8 Ibid. 28, 29.

^ History of Rome, vol. i. p. 358.

ii8

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XIX.

rendered memorable as the first occasion on which the Eoman
arms were carried across the Ehine into Germany, and across
the sea into Britain.

The occasion for the former enterprise was furnished in the
first instance by the Germans themselves. The Usipetes and
Tencteri, two German tribes that had occupied a territory on
the right bank of the Ehine, in the lower part of its course,^
finding themselves hard pressed by the powerful nation of the
Suevi, who were gradually extending themselves to the west,
and subduing or expelling all the tribes that lay between them
and the Ehine, had crossed that river and established them-
selves in the territory of the Menapii. Hence they spread
without difSculty into the adjoining regions, and were ex-
tending their incursions on every side when Caesar arrived in
that part of Gaul, and after a brief negotiation attacked and
defeated them between the Mouse and the Ehine, driving
them with great slaughter into the one river or the other.
But a large body of their cavalry succeeded in making their
escape across the Ehine, and took refuge in the land of the
Sugambri,^ who made common cause with the fugitives, and
refused the demands of Csesar to deliver them up, alleging
that the Eomans had no right to interfere beyond the Ehine.
At the same time the Ubians, the only people on the right
bank of that river who had entered into friendly relations with
Eome, entreated Csesar to cross the Ehine with a view to strike
terror into the Suevi and their allies. He in consequence con-
structed a bridge across the river — a work which was accom-
plished with marvellous celerity, being completed within ten
days from its first commencement^ — and passed it with his

* Cses. B. G. iv. 1. Both these names
here appear for the first time, but
are afterwards repeatedly mentioned
during the wars of the Eomans with
the Germans. They were on this oc-
casion effectually driven out of Gaul,
and never again crossed the Ehine.

^ See note to next page.

^ B. G. iv. 18. Tlie exact place
where Csesar constructed this cele-

brated bridge cannot be determined,
but it may be fixed within definite
limits. The history of the campaign,
combined with the natural geography
of the country, and the course of the
river, leave no doubt that it must be
placed between Coblentz and Ander-
nach, probably in the neighbourliood
of Neuwicd (see Ukert, Germania,
p. 18, note).

Chap. XIX.

CiESAK S WARS.

119

whole army. But the Suevi and Sugambri, on the first intelli-
gence of the construction of the bridge, had withdrawn into
the interior of the vast forests with which their territories
abounded. Thither Caesar did not think fit to follow them,
and after laying waste the lands of the Sugambri near the
river he contented himself with this demonstration, and re-
crossed the bridge, after having spent only eighteen days on
German soil.*

§ 7. He had previously made up his mind to undertake an
expedition against Britain ; an enterprise to which he was pro-
bably urged more by the desire of the fame to be earned by
being the first Roman general to set foot in that remote and
little-known island, than by any hope of real advantage. The
summer was indeed so far advanced that he could not look for
any great results, and he took with him only two legions,
intending this first expedition, if we may trust his own account,
rather as a mere reconnaissance than as a serious invasion.
Sailing from the Portus Itius, a seaport on the coast of the
Morini,^ which afforded the shortest passage into the island,
he crossed the Straits of Dover, and succeeded in effecting a
landing in the face of the enemy, probably in the neighbour-
hood of Deal.^ But he scarcely penetrated at all into the
interior, his cavalry, which had sailed from a different port,
having failed in accomplishing their passage, while his fleet
suffered severely from a storm, accompanied by the unwonted
phenomenon of spring tides. Under these circumstances he
determined to return to the mainland ; and contented himself
with repulsing an attack of the Britons on his camp, which was
followed by a pretence of submission on the part of some of the
neighbouring tribes.

§ 8. The very imperfect success of this first attempt only
stimulated him to make greater efforts in the following spring

* Ibid. 19. This is the first occasion
ou which the Sugambri or Sicambri,
the name of whom was afterwards so
familiar to the Romans — the " c£ede
gaudentes Sicambri" of Horace —

appear in history. They dwelt at this
period on the right bank of the Ehine,
north of the Ubii.

5 See Note B, p. 136.

» See Note C, p. 137.

I20 HISTOEY OF ANCIENT GEOGRAPHY, Chap. XIX.

(B.C. 54). He caused a large number of ships to be built for the
express purpose of the invasion, and assembled not less than
600 such vessels, besides 28 ships of war.^ With this great
fleet, on board of which he carried five complete legions and
2000 cavalry,^ he effected the passage of the Straits without
any loss, and landed at the same place as the year before, but
this time without opposition, the natives having withdrawn
into the interior at the sight of so formidable an armament.
His first action was fought on the banks of a river about twelve
miles from the sea-coast :^ thence he pushed on into the interior
of the island as far as the more important river Tamesis
(Thames), which bounded the territory of Cassivellaunus, king
or chief of the Trinobantes. This chieftain had been appointed
to the supreme command of the British forces, but appears to
have abandoned the defence of the maritime districts as hope-
less, and withdrew at once across the Thames, the line of which
he hoped to d€fend against the invader. Ceesar reached that
river at a distance of about 80 miles from the sea, at a spot
which, according to the information which he had received,
was the only one where the stream was fordable : ^ he here suc-
ceeded in forcing his passage and capturing a " town " or
stronghold of Cassivellaunus not far distant.^ This was the
farthest point to which he penetrated. Several of the neigh-
bouring tribes hastened to send deputies and make overtures
of submission,^ and their example, after a short time, was fol-
lowed by Cassivellaunus himself. Csesar, apprehensive lest the
war should be protracted until the close of the summer, and
desirous of returning to Gaul, admitted them to favourable
terms, and contented himself with demanding hostages and the
imposition of a nominal tribute.*

The military operations in Gaul during the remainder of the
season had no especial interest in a geographical point of view.

' Cffis. B. a. V. 2.
« lb. 8.

" Probably the Stour, which flows
by Canterbury and Eichborough.

^ See Note D, p. 138.

2 Sec Note E, p. 139.

3 See Note F, p. 139
' Ibid. V. 22.

Chap. XIX. C^SAKS WARS. 121

The revolt of the German, or semi-German, tribes of the
Eburones and Treveri, though they succeeded in cutting ojQf
one division of his army under Titurius Sabinus and Aurun-
culeius Cotta, and gravely endangering two others, was un-
successful, and failed in producing any permanent result.
Caesar had judiciously posted all his legions, with a single
exception, in the territory of the Belgians (in the wider sense
of the word) : this being apparently the only part of Gaul
where he expected any outbreak. He himself took up his
winter quarters at Samarobriva (Amiens), which appears to
have been one of the most considerable towns in those
regions.

§ 9. The conquest of Gaul was however still far from being
complete, and the three following campaigns (b.c. 53, 52, and
51), were all employed in putting down insurrections of the
native tribes that combined in defence of their liberties before
they were finally compelled to acquiesce in their subjection to
the Eoman yoke. The first of these was principally confined
to the Belgian, or rather German, tribes of the Treveri and
Eburones, and the assistance sent them from beyond the
Ehine, from the powerful nation of the Suevi, led Csesar to
cross that river for the second time. His passage was effected
on this occasion a little higher up than before, but in the same
part of its course. He did not penetrate any farther into the
interior than on the previous occasion. He was received in a
friendly manner by the Ubii, whose territory immediately
adjoined the Ehine ; but found that the Suevi had retired on
his approach to the farthest limits of their territory, where a
vast forest, called by Csesar the Silva Bacenis, separated them
from the Cherusci ^ on the east ; and hither he judged it im-

* B. G. vi. 10. This is the first
mention of tlie name of the Cherusci,
afterwards so familiar to the Komans
during their long wars in Germany.
They appear to have dwelt at this time
between the Weser and the Elbe. The
forest called by Osesar Bacenis (a name

not found in later writers), which he
describes as "silva infinita magnitu-
dine" would therefore correspond to
the Harz and the range of the Teuto-
burger Wald, on the borders of West-
phalia.

122

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XIX.

prudent to follow them. He therefore withdrew a second time
across the Ehine, after a brief stay on the German side of the
river.^ Nevertheless, he availed himself of the opportunity to
collect many interesting particulars concerning the nations,
inhabitants, and natural productions of Germany, with which
the Eomans now for the first time became acquainted.'^

§ 10. The following campaign (b.c. 52), in which the
standard of revolt was raised by Vercingetorix, at the head of
the Arverni, and was followed by a general defection of almost
all the Gaulish tribes — even the faithful -^dui being carried
away by the contagion of example to join in the movement —
was in a political sense one of the most important of all, and
never did the military genius of Csesar show itself more con-
spicuously : but his movements were confined within the limits
of Gaul itself, and he had no occasion to carry his arms beyond
the districts with which he was already acquainted. Hence the
operations of this year, interesting as they are in a military and
topographical point of view, cannot be considered as having con-
tributed materially to the extension of geographical knowledge.

The same remark applies to the desultory hostilities of the
following year (b.c. 51), which were confined to successive par-
tial revolts in different parts of Gaul — among the Bellovaci
and Treveri in Belgium, and among the Pictones and adjacent
tribes in the west. The last blow was given to this final move-
ment by the reduction of Uxellodunum, a fortress of great
natural strength,^ in which the last of the rebel leaders had
taken refuge. The capture of this stronghold may be said to
have completed the conquest of Gaul. From this time the
whole country from the Ehone and the Ehine to the Western

« C£es. B. G. vi. 9, 10, 29.

' Ibid. 21-28.

* The position of Usellodimimi may
be now considered as established be-
yond a doubt. It occupied a hill, now un-
inhabited, called the Puy d'Issolu, near
the north bank of the Dordogne, within
the limits of the district still called
Quurcy, a name derived from that of

the Cadurci, its ancient inhabitants —
but near the frontiers of the Limousin.
This site, which was fii-st suggested by
D'Anville (Notice de la Guide, p. 729),
and adopted by Thierry {Hist, des
Gaulois, vol. iii. p. 220), has been fully
coutirmed by researches made on the
spot by order of the Emperor Napoleon
III. (See his Vie de Ce'sar, vol. ii. j). 3-lo.)

CuAP. XIX. CiESAKS WARS. 1 23

Ocean, passed, without any further attempt at resistance, under
the ordinary administration of a Roman province, and rapidly
acquired, in all but the remotest districts, a strong tincture
of Eoman civilization.

§ 11. With regard to Gaul itself the effect of these succes-
sive campaigns of Julius Csesar was to bring the whole of that
great country within the domain of definite geographical
knowledge. For the vague ideas and arbitrary assumptions of
previous authors were substituted the distinct and clear state-
ments of an able observer and remarkably lucid writer. Our
good fortune in possessing the original work in its integrity
renders this contrast still more striking to us, and in esti-
mating the results thus obtained we must bear in mind that
had we possessed in like manner the complete works of some
of the Greek writers, especially Polybius and Posiddnius, we
should probably have found that they possessed, though in a
somewhat vague and imperfect form, a knowledge of many
nations, as well as physical features of the country, that are now
for the first time found mentioned in the pages of Csesar.

But whatever allowance may require to be made on this
account, it is certain that Caesar's own record of his observa-
tions and operations in Gaul must have formed for the Eomans,
as well as for ourselves, the first foundation of all accurate
knowledge of that country. The brief geographical summary
with which he opens his work states clearly the ethnological
division of the country into three portions, the inhabitants of
which, as he distinctly tells us, differed from one another in
language, institutions, and laws. These were the Aquitani
in the south, the Belgse in the north, and the Celts or Gauls
proper in the intermediate portion. Their boundaries also are
clearly marked, the Gauls being separated from the Aquitani
by the river Garumna or Garonne, and from the Belgse by
the Seine and Marne, and here, as well as in other passages,
we find him well acquainted with all the principal rivers
which, in the case of Gaul especially, form the leading
features in the physical geography of the country. Not only

124

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XIX.

are tlie Seine, Loire, and Garonne repeatedly mentioned in
his Commentaries, but their tributaries, the Marne (Matrona),
the Aisne (Axona), and the Allier (the Elaver), were equally-
well known to him. He describes also clearly the course of
the mountain ranges of the Cevennes (Mons Cebenna), the
Jura, and the Vosges (Yosegus), as well as the great forest
tract of the Ardennes (Silva Arduena), which at that period
constituted so important a natural feature in the north of
Gaul.^ His repeated campaigns in Belgium rendered him
familiar not only with the course of the Ehine and the Meuse
(Mosa),^ but with those of the Sambre (Sabis), the Scheldt
(Scaldis), and even the Waal (Vacalus), which he correctly
describes as a branch of the Ehine, flowing into the Meuse.^
But his knowledge of the island of the Batavi, which he con-
ceived to be intercepted between the two rivers, was derived
only from hearsay, and was necessarily imperfect.

§ 12. Still more complete and accurate was his knowledge
of the different nations and tribes that inhabited the country
at the time of its conquest. Here his position gave him ad-
vantages which no ordinary geographer would have possessed :
and where he enumerates the nations that on different occa-
sions combined in arms against him, with the force of their
several contingents, or that successively submitted to his yoke,
we may feel confident that his lists are based on authentic
materials. Such lists are found, of the Helvetians and their
allies in the first book, of the Belgic tribes in the second, of
the Armorican nations and the Aquitanian tribes in the third,
and a more general enumeration, comprising all the principal
populations of Gaul in the seventh book, when they formed a
general league under Yercingetorix. An examination of these
lists is one of the most satisfactory things in ancient geo-

" His statement of its extent — that
it covered a space of more than 500
miles in length, from the Rhine and
the borders of the Treveri, to the con-
fines of the Nervii and the Remi — is
indeed in any case a great exagge-
ra,tiun ; but he here doubtless followed

some vague popular estimate. B. G.
V. 3, vi. 29.

' It must be mere chance that the
name of the Moselle does not occur iu
the Commentaries.

" B. G. iv. 10.

Chap. XIX. CiESAR's WARS. 1 25

graphy : the greater part of the names are recognized at once
from their having continued down to a recent period to give
name to the provinces or districts where they dwelt, or being
still retained in those of their chief towns. A comparatively
small number only have disappeared, and these for the most
part were either obscure or insignificant tribes, or, as in the
case of the G-erman races in Belgium, have been effaced by
the continual waves of invasion that have swept over that
part of Gaul.

In many cases also the towns may readily be identified
from their preserving the names of the tribes to which they
belonged, while in others they retained the same names under
the Eoman Empire, and are therefore well known. Such was
the case (among others) with Avaricum (Bourges), Agedincum
(Sens), Genabum (Orleans), and Lutetia (Paris), the position
of which upon an island in the Seine is distinctly noticed.^

It is a fact peculiar to the geography of Gaul that in the
great majority of cases the chief towns of the several tribes
gradually lost their own separate appellations, and were known
only by those of the tribes to which they belonged. Thus
Samarobriva, the capital of the Ambiani, became Ambiani,
whence its modern name of Amiens ; Avaricum of the
Bituriges in like manner passed into Bourges ; Noviodunum
of the Suessiones into Soissons, and so in numerous other
instances. In all these cases, however, the change can be
readily followed : and no doubt can exist as to the identifica-
tion of the cities, which have continued to occupy the sites of
the original capitals.

The case is otherwise with the great strongholds of the
Gauls at Gergovia and Alesia, both of which sites were sub-
sequently abandoned, when their strength as fortresses had
ceased to be of value. Both of them, however, can fortu-
nately be identified beyond a doubt : the hill of Gergovia
having always retained its original name, though uninhabited ;

' B. G. vii. 57.

126

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XIX.

while that of Alesia is still marked by the village of Alise
Ste. Eeine, in a commanding position on Mont Auxois, about
12 miles from Montbard. It is probable also that Bibracte,
repeatedly mentioned by Csesar as the capital of the ^Edui,
and commonly identified with the Roman city of Augnsto-
dunum (Autun), really occupied a much stronger and more
elevated position on the hill called Mont Beuvray, some dis-
tance further west.* Uxellodunum, though a site of great
natural strength, does not appear to have ever been a town of
much importance.

§ 13. But if we are struck with the accuracy, as well as the
extent, of Caesar's information concerning Gaul — a country
which he had traversed in all directions during a space of
ten years — the case is very different with regard both to Britain
and Germany. In some respects indeed the information ob-
tained by Csesar with respect to these two countries was even
a more valuable addition to the stock of geographical know-
ledge previously existing than his contributions to that of
Gaul. For the notions concerning them to be derived from
any earlier sources were so utterly vague and unsatisfactory,
that the amount of knowledge he was able to collect upon the
subject — imperfect as it was — was of the highest value, as
supplying at least a certain portion of definite and trustworthy
fact. He himself tells us that when he attempted to gather
information concerning Britain from the Gaulish traders who
were in the habit of visiting the island, he was unable to learn
what were its magnitude and dimensions, by what nations it
was inhabited, or even what were the largest and most com-
modious ports.^ Considering that an extensive trade was
undoubtedly carried on between the two countries, and

* See D'Anville, Notice cle la Gaule,
p. 156 ; and a note to Napoleon's Vie
de Ce'sar, vol. ii. p. 67.

^ B. G. iv. 20. "Itaque vocatis ad
se undique mercatoribus, neque quanta
esset insulaj magnitudo, neque qua)
aut quanta) nationes incolereut, neque
qucm usum belli hal^erent aut quibus

institutis uterentur, neque qui essent
ad majorum navium multitudincm
idonei portus, reperire poterat."

A passage that is instructive as show-
ing the diflSculty of procuring inform-
ation from such sources, and the con-
sequent uncertainty of all statements
derived from them.

Chap. XIX.

CiESAES WARS.

127

that the southern or maritime districts of the island were
inhabited by tribes of Belgian origin, who retained the names
of the parent races from which they had sprung,^ and pre-
served at least some degree of political connection with them,
it is impossible to doubt that this ignorance was in part
assumed; but it serves clearly to prove the difficulty of
obtaining such information, and fully accounts for the vague
character of the reports circulated by previous writers.

§ 14. CcGsar himself did not, as we have seen, advance far
beyond the Thames : he landed on both occasions at the same
point, and returned to it again to re-embark for Gaul. His
opportunities of personal observation were therefore very
limited, and he does not appear to have held personal inter-
course with any of the more important nations of the island,
except the people of Cantium or Kent — who were, as he remarks,
by far the most civilized people in the country, and differed but
little from their neighbours in Gaul — and the Trinobantes,
who occupied a tract north of the Thames, probably com-
prising the modern counties of Essex and Hertfordshire. His
information concerning the tribes of the interior was there-
fore derived chiefly from hearsay ; as was necessarily the case
with his general geographical notices. He describes the
island^ as of triangular form, one of the angles being formed
by the projecting point of Kent (Cantium), another by a pro-
montory extending towards the south, in the direction of
Spain. The coast between these two, which faced that of
Gaul, was about 500 miles in length. The west coast, opposite
to which lay Hibernia, was said to be about 700, miles in
extent; while the third, which faced the north (north-east)
was not less than 800 miles. Hibernia^ was estimated at about

6 B. G.Y.12.

' Ibid. V. 13.

* " Alterum vergit ad Hispaniam at-
que occidentem solem ; qua ex parte est
Hibernia, dimidio minor, ut sestimatur,
quam Britannia." Ibid. This is the
first mention in any extant author of
the name of Hibernia, though there
can bo no doubt that the name at least

was known to the Komans long before,
as was that of lerne to the Greeks.

The expression of " vergit ad Hispa-
niam" is very singular; but would
seem to imply that he conceived the
position of Britain somewhat in the
same manner that Strabo did ; though
he distinctly placed Ireland to the west
of it, and not to the north.

128

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XIX.

two-thirds of the size of Britain, from which it was separated
by a strait of about the same width as that from Britain to
Gaul. Midway between the two was an island called Mona :
besides which numerous other islands were scattered around the
principal one, in some of which it was asserted that at the
winter solstice there was continuous night for thirty days.^ Of
this Csesar could get no definite account,, but he ascertained by
observations instituted on purpose with water-clocks that even
in the parts of Britain visited by himself, the nights at that
season (the late summer) were shorter than in Gaul.-^ The
climate was also more temperate, and the cold in winter less
severe.

It is remarkable that no allusion is found in the Commen-
taries, either in this passage or elsewhere, to the celebrated
Cassiterides or Tin Islands, in connection with Britain, though
it cannot be doubted that it was the richness of Cornwall in
this respect that led to the extensive trade with Britain carried
on by the Veneti from Bretagne ; who probably transported
the ore from thence to the mouth of the Loire,^ Csesar was
erroneously informed that tin (plumbum album) was found in
the interior of Britain^ — a statement which, if it were not
intended to mislead, can only be ascribed to the ignorance of
the Belgian tribes in the south-east of the island concerning
the remote corner in the south-west. Pearls, which had been
supposed to be produced in Britain in large quantities, were
found to be in fact neither large nor of fine quality.*

§ 15. Still more imperfect was the acquaintance possessed

^ Ihid. It is evident that this is
only a reappearance of the confused
traditions alDOut Thule; but from the
expression of Csear (de quibus insulis
nonnulli scripserunt) it seems that lie
is here referring to the statements of
earlier authors (Greek or Latin) rather
than to anything he heard in the
country.

' " Nos nihil de eo percontationibus
reperiebamus, nisi certis ex aqua men-
suris brcviorcs esse quam in continenti

noctes videbamus." Ibid.

' The information on this subject
collected by P. Crassus has been already
referred to (see Note 8, to p. 115).

^ " Nascitur ibi plumbum albiun in
mediterrnneis regionibus,' v. 12.

* The British pearls are not alluded
to by Csesar, though asserted by some
later writers to be one of the tempta-
tions that induced him to attempt the
conquest of the island. (Suet. Crcs.
47. See Chapter XXIII. Note A.)

Chap. XIX. CAESARS WARS. 1 29

by Caesar with Germany, so far as it rested on personal obser-
vation. Though he twice crossed the Ehine with an army,
and might undoubtedly boast of being the first Roman general
who ever set foot on German soil, he penetrated on each occa-
sion but a very small distance into the interior, and has fur-
nished us with no details of his operations. But from his
alliance with the Ubians, who at this period occupied the right
bank of the Rhine, as well as from the numerous German
prisoners taken from Ariovistus, he appears to have had the
means of' obtaining information concerning the neighbouring
tribes and nations, as well as the character of the country in
general, of a more trustworthy character than would have been
within the reach of any ordinary geographer. Thus we find all
the principal tribes that he mentions — the Suevi, the Sugambri
or Sicambri, the Usipetes and Tencteri, as well as the Ubii
themselves, and the more distant Cherusci, all noticed under
the same names by which they shortly afterwards reappear in
history ^ : and though their limits and places of abode cannot
be said to be distinctly indicated, this was inevitable, at a
time when the leading geographical features of the country
were as yet unknown, and there were no towns or fixed points
to determine the locality of each tribe.^ A very large part of
Germany was undoubtedly at this period covered with primeval
forests. Of these Caesar mentions two by name, the great
Hercynian Forest, which had already been known by name at
least to Eratosthenes and Posidonius, and was reported to extend
over a space of nine days' journey in width, and more than
sixty days' journey in length : its extension in that direction
being unknown. It began on the confines of the Helvetii and

^ Of the tribes that bad furnished I Sednsii are supposed to have come from
their contingents to the army of Ario- more distant regions to the north.

vistus (-B. G. i. 51) — the names of which
were doubtless learnt from the cap-
tives — the Tribocci, Vangiones, and
Nemetes, were petty tribes dwelling on
the left bank of the Khine : the Suevi
and Marcomaiini are well-known Ger-
man nations ; while the Harudes and

VOL. II.

Cfesar did not penetrate far enougli
into the interior to become acquainted
with any of the great rivers— the Ems,
the Weser, and the Elbe — which natu-
rally figure so prominently in the
subsequent wars of the Romans in
Germany.

130

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XIX.

Kauraci (witli the modern Black Forest) and thence continued
along the northern bank of the Danube, to the confines of the
Dacians and Anartians/ where it quitted the course of the
river and turned to the north, into regions which had never
been visited.^ The other, to which he gives the name of Bacenis
Silva, he describes as separating the Suevi from the Cherusci :*
this evidently corresponds with the forest of the Harz, and that
subsequently known as the Teutoburger Wald.

§ 16. Caesar's account of the manners and habits of the
Germans,^ as distinguished from the Gauls, brief as it is, is
clear and characteristic, and agrees well in its general features
with that collected at a later period by Tacitus. His notices
of the wild animals on the contrary that were found in the
vast forests of Germany, were necessarily derived from hearsay,
and are not unmixed with the same fables which we find still
current in the days of Pliny,

Of his ethnographical observations undoubtedly the most
important is that in which he remarks that while the German
races in his day were perpetually pressing upon the Gauls
and tending to establish themselves across the Khine, the
contrary had previously been the ease, and Gaulish tribes had
formerly crossed the Rhine and established themselves on
German territory. This accords well with the fact that we
find at an earlier period races of Gaulish origin, the Boii,
Taurisci, and others extending down the valley of the Danube
even to the frontiers of Dacia and Illyricum.^

§ 17. The Civil Wars of the Romans, that preceded the
final establishment of the Empire, from their being confined
within the limits of the Roman dominions, were naturally
little calculated to promote the extension of geographical
knowledge. The only exception was the remarkable march
of Cato from Cyrene to Utica, which would possess much

' This mention of so obscure a peoijle
as the Anartians is very singular. A
tribe of that name is found in the list
given by Ptolemy (iii. 8, § 5) of the
Daeian tribes, but they are not men-

tioned by any other author.
' Cres. B. G. vi. 23.
» lb. vi. 10. .
1 lb. vi. 21-24.
- See Chapter XVIII. p. 90.

IMAP TO ILLUSTRATE CvESAR'S WARS.

Sranford:-' GMpr.JZstalL.lond.n

London ; John Murray.

Chap. XIX.

CiESAKS WARS.

131

interest for the geographer, had its details been preserved to
us. But unfortunately these are wholly wanting. We learn
only from Strabo that he marched round the Great Syrtis in
thirty days from Berenice ; ^ and if any reliance can be placed
upon the poetical statement of Lucan, that he completed the
whole march to the fertile districts of the Carthaginian terri-
tory within two months : * a marvellous proof of the endurance
and hardiness of the Eoman soldiers. He was at the head of
an army of more than 10,000 men, but we are not told what
proportion of these he led in safety to join the army of Scipio
in Africa. The exploit is certainly one of the most remarkable
of its kind on record, and may well be compared with the march
of Alexander throug'h the deserts of G-edrosia.^

* 'Ek TavTt\s TTjs irSXeccs (BepevLKris SG.')
TpiaKoaraios Tt'e^fj irepiddevcre rT]v 'Xvpnv
MdpKOs Karcov, KUTaycav ffrparLav irXeio-
pcov rj /jLvplcov avdplav, els fj-ep-q SieXwv roov
vSpeictiv X'^P^"' SSei/ce Se Tre^ox eV ajj-jjicp
^adeia, Koi Kavfx.acn. Strabo, xvii. 3, p.
836. This is the only deiinite and
trustworthy information that we pos-
sess concerning this remarkable march.
Plutarch furnishes scarcely any par-
ticulars, while tlie bombastic descrip-
tion of it in Lucan, to which it chiefly
owes its celebrity, is as vague as it is
inflated. The real distance from Bere-
nice (Benghazi) to Cape Mesurata,
which forms the western boundary of
the Syrtis, is, according to Dr. Barth,
who himself performed the journey,
more than 105 German or 420 geo-
graphical miles. Captain Beechey
estimates it at 426 Gr. miles. Strabo
himself in another passage states the
circumference of tlie Great Syrtis at
3930 stadia (Barth, Wanderungeii, p.
358 ; Beechey's Tripoli, p. 256). It is
most probable that Strabo would reckon
his march from one city to another, or
from Berenice to Leptis Magna, which
is about 50 miles farther westward.
The march from thence to the Cartha-
giniau territory would offer compara-
tively little difficulty.

Mr. Merivale has been misled l)y the
coiifu.sed narrative of Lucan into sup-

posing that it was the Lesser Syrtis
which alone was the scene of this
perilous march, but the testimony of
Strabo is clear and explicit, and per-
fectly C(msistent with the natural
features of the country. This lias
unaccountably been overlooked by Mr.
Long.

■* Lucan, ix. 940. When Plutarch
speaks of his traversing the sandy
desert for seven days continuously
{Cato, 56) he must clearly refer to
some special portion of the march.

* The same enterprise had indeed,
been successfully accomijlished at a
much earlier period by Ophelias, ruler
of Cyrene, who in B.C. 308 conducted
an army of Greek mercenaries from
that city to the support of Agathocles
in his war against Carthage. He
also took two mouths on the mai-ih
(Diodor. xx. 41, 42). His army was
reported, as well as that of Cato, to
have suffered severely from venomous
serpents. Absurdly exaggerated as are
the tales concerning these found in the
Greek and Roman writers, they are
not altogether without foundation.
Several species of snakes whose bite is
of a most deadly description wre found
in the sands of Northern Africa, espe-
cially the African Cobra aud the
Cerastes or Horned Vi])er. Otlitrs
attaiij to a large size. Dr. Bartli in

K 2

132

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XIX,

§ 18. Hostilities were also carried on upon the eastern
frontier of the Roman Empire during the interval of repose
that preceded the final contest between Antony and Octavian,
and the operations of the former against the Parthians were
attended with some successes and deserve a passing notice.
The history of this war was written by his friend and com-
panion Dellius,^ whose work was used by Strabo, and appears
to have thrown some additional light on the countries bor-
dering on Armenia and Mesopotamia. But our knowledge of
these campaigns is too imperfect to estimate their value in
this respect. We learn however that, after the way had been
cleared for him by the successes of his lieutenants, Ventidius
and Canidius — the first of whom defeated the Parthians and
drove them back across the Euphrates, while the second re-
duced the Armenian king, Artavasdes, to submission, and even
carried the Roman arms for the second time against the Iberians
and Albanians^ — he himself advanced at the head of a great
army through Armenia, into Atropatene, a province hitherto
unknown to the Roman arms, and which constituted a subordi-
nate kingdom dependent upon the Parthian monarchy. Here
he laid siege to a city called by Plutarch Phraata, and by
Dion Cassius Praaspa, which is described as a great city, in
which the king of Media (Atropatene) had deposited his wives
and children for security.^ It was a fortress of great strength,
and the efforts of Antony to reduce it proved ineffectual.
Unfortunately its site is very imperfectly indicated. But it
appears to be certainly the same place which is called by
Strabo Vera,^ and if this be the case it may probably be iden-
tified with the remarkable mountain fortress now known as

one instance killed a snake between
8 and 9 feet in length (p. 268) ; but no
such gigantic monsters as the Pythons
and Boas of India are known in Africa
at the present day.

•^ Strabo, xi. p. 523 ; Pint. Anton.
c. 25, 59. This is the same Dellius to
whom Horace has addressed the well-
known ode (Carm. ii. 3).

' Pint. Anton. 34 ; Dion Cass. slix.
24. Plutarch even asserts that Cani-
dius on this occasion advanced as far
as the Caucasus i&xpi rod Kavicda-ov
■wporiXQev), a statement that must doubt-
less be received with some allowance.

8 Plut. Anton. 38.

^ Ouepa. Strabo, xi. 13, p. 523. Un-
fortunately the text of Strabo in this

Chap. XIX.

C^SAKS WAES.

^33

Takht-i-Suleiman, for which Sir H. Rawlinson has claimed the
name of the Atropatenian Ecbatana.^ Be this as it may, it is
certain that Antony on this occasion carried the Eoman arms
in this direction farther than any j)receding, or indeed than
any subsequent, generaL

Having been compelled to abandon the siege, he commenced
his retreat towards Armenia, but suffered severely from drought
and thirst in traversing the arid plains of Atropatene (the
modern Azerbijan), as well as from the continual harassing
attacks of the Parthians. It was not till after twenty-seven
days' march, during which they were engaged in almost per-
petual hostilities, and are said to have lost not less than 24,000
men, that the Eoman army reached the river Araxes, after
crossing which they found themselves in safety, within the limits
of a friendly country.^ The distance from Phraata or Vera to the
Araxes is given by Strabo, on the authority of Dellius, at 2400
stadia, or 240 G. miles. Sir H. Rawlinson, who was himself
well acquainted with the country, points out the accuracy with
which the details of this march are given by Plutarch, evidently
following the authority of Dellius : among other incidents the
sufferings of the Eoman soldiers were on one occasion greatly
augmented by their coming to a stream of salt water which is
undoubtedly the Aji, a river flowing a few miles to the north
of Tabriz, the only one of this nature in all Azerbijan.^

On the other hand the route by which Antony had advanced
into Atropatene is very obscurely indicated. Strabo indeed
represents him as being purposely misled by the king of

passage is corrupt, and it is impossible
to determine the connexion of this
name with the preceding clause of the
sentence in which is found that of
Gazaca, a ^well-known name, corre-
sponding to the Armenian Gandsak, a
treasury. Groskurd, Kramer and C.
Miiller consider the two names as
referring to two distinct places — the
one being the winter, the other the
summer residence of the kings of Atro-
patene — and it is difficult to resist this
conclusion. Sir H. Rawlinson, on the

contrary, identifies the two, as merely
different appellations of the same place
(Geogr. Journ. I. c). The name of
Gazaca occurs in Ptolemy and Am-
mianus Marcellinus, as well as in
Stephanus of Byzantium : that of Vera,
I believe, is not found in any other
writer.

' Kawlinson in Geogr. Journal, vol.
X. p. 65, &c.

2 Plut. Anton. 41-49; Dion Cass,
xlix. 28-31.

3 Id. ibid. pp. 113-117.

134

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XIX.

Armenia, who caused him to take a circuitous and laborious
route from the Euphrates instead of the more direct and easy
one.* But this may well be doubted. The direct route from
the Zeugma (at Bir) would have led him through the same
country as had been traversed by Crassus, and exposed him to
a repetition of the same disasters. By keeping to the moun-
tains through Commagene, Sophene, and the southern provinces
of Armenia, he avoided exposing himself to the attacks of the
Parthian cavalry, while the alliance of the Armenian king
secured his northern flank. Artavasdes however, though at
first acting as the ally of Antony, abandoned him in the time
of his need — a defection for which the Koman triumvir at a
later period punished him by invading his kingdom and
depriving him of the sovereignty.^

It is during this period that we find the first notice of the
city of Palmyra, against which Antony detached a predatory
expedition, on account of the wealth which the inhabitants
were reported to have accumulated by their extensive com-
mercial relations, with the Syrians on the one side, and the
Parthian provinces on the other.®

* Strabo, xi. 13, p. 524.

* Plutarch, Ant. 50 ; Dion Cass. xlix.
39, 40. On this occasion Antony
advanced from Nicopolis in Lesser
Armenia, and pushed on direct upon

Artaxata, the Armenian capital, of
which he made himself master.

^ Appian, B. C. v. 9. This expe-
dition appears to have taken place in
B.C. 41.

Note A. O^SAk'SWAKS. 135

NOTE A, p. 114.

BELGIAN TRIBES.

The enumeration on this occasion of the various Belgian tribes,
and the forces they were able to muster, is a document of the
highest interest, and furnishes the foundation for all inquiries into
the geography of this part of Gaul.

The nations mentioned by Caesar, most of whom can be deter-
mined, and their site fixed with the greatest clearness, are as
follows :

The Bellovaci, whose capital city still retains the name of
Beauvais.

The Ambiani, whose name is still found in that of Amiens.

The Atrebates who gave name to Artois and its capital of Arras.

The Caletes whose name is still found in that of the Pays de
Caux, the part of Normandy adjoining the sea, from the mouth of
the Seine to that of the Bresle.

The Veliocasses in the Vexin, the district between the Beauvaisis
and the Seine.

The Veromandui in the Vermandois, a portion of Picardie around
St. Quentin.

The Suessiones in the diocese of Soissons.

The Eemi, who were in alliance with Caesar, in that of Eeims.

In all these cases the names alone suffice to fix the locality
beyond dispute. In the case of the Nervii, the most powerful and
warlike of all the Belgic tribes, the name has disappeared, though
still mentioned by Tacitus and Ptolemy : but their position is
certain : they occupied the region of Hainault and the diocese of
Cambrai, extending eastward to the Sambre. In like manner the
Morini held the sea-coast adjoining the Straits of Calais from the
mouth of the Somme to the Scheldt, and the Menapii the still more
northerly district about the mouths of the Scheldt and the Meuse.

The Aduatuci were situated to the north of the Nervii, about the
confluence of the Sambre and Meuse : their chief city is supposed,
though on doubtful evidence, to have occupied the site of Nairtur.

The Eburones must be placed to the north of these last, apparently
in the district hubsecpiently occupied by the Tungri (Tongros) ;

136 HISTOKY OF ANCIENT GEOGRAPHY. Chap. XIX.

while the three nations associated with them by C^sar as dis-
tinctly German tribes (qui tino nomine Germani appellantur), the
Condrusi, Cserasi, and Pasmani, are otherwise wholly unknown.

The powerful nation of the Treveri (certainly a German tribe)
did not take part with the Belgians on this occasion, but is
repeatedly mentioned elsewhere in the Commentaries : they held
apparently the whole of the subsequent diocese of Treves, on each
side of the Moselle, and extending to the left bank of the Ehine.
The same was the case with the Mediomatrici, (Cass. B. G. iv. 10,
vii. 75) whose name survives in the much abbreviated form of
Metz, the ancient diocese of which probably coincided with the
limits of their territory. The Ubii at this time dwelt on the right
bank of the Ehine, opposite to the Treveri.

NOTE B, p. 119.

Cesar's passage from gaul to Britain.

Both the point of departure, from which Caesar sailed on his
expedition to Britain, and that where he landed in the island, have
been of late years made the subject of much controversy. Mr.
Long, who is the most recent writer that has examined the ques-
tion, arrives at the conclusion " that it will never be settled
whether CfBsar sailed from Wissant or from Boulogne." (Decline
of the Boman Bepuhlic, vol. iv. p. 433.) Without presuming to
" settle " the question, I may briefly state the reasons which in my
opinion are decisive in favour of Wissant ; the conclusion adopted
by D'Anville, Gossellin, Walckenaer, as well as more recently by
M. de Saulcy. Csesar tells us that he selected the Portus Itius, as
his point of departure, because it was the most convenient passage
to Britain, about thirty miles from the continent (quo ex portu
commodissimum in Britanniam trajectum esse cognoverat, circiter
milium passuum xxx a continenti. B. G. v. 2). Now Wissant is the
Clearest port to Britain, and was on that account much used in the
middle ages. This was a point that could be readily ascertained
by a mere inspection of the coast. It is true that the distance is
less than the 30 (Koman) miles stated by Caesar; but we have
repeatedly had occasion to observe that the ancients had absolutely
no means of determining distances at sea with any aj)proach to

Note C. C^SAE S WARS. 137

accuracy. The difference is inconsiderable : Wissant being about
22 English or 23.^ Roman miles from Dover : while Gessoriacum
or Boulogne is nearly (if not quite) 30 English miles from Folke-
stone, the nearest point of the British coast. It is certain that
after the Eomans had permanently established themselves in Gaul,
and came to have frequent intercourse with Britain, Gessoriacum
came to be the customary port of communication between the two :
but if we suppose this to be the Portus Itius of Caesar we have to
account for the change of name, of which we have no similar
instance in regard to any other name mentioned in the Com-
mentaries,

I entirely concur with Mr. Long in believing the Icium or
Itium Promontorium of Ptolemy (ii. 9, § 2) to be Cape Grisnez, the
only headland of importance along this whole line of coast, and
which must in all ages have attracted attention; though the
geographer has in this case much misplaced its position. But if
Cape Grisnez be the promontory of Itium, the Portus Itius would
be naturally looked for in its immediate neighbourhood : and the
name would suit much better with Wissant, which is barely 3 miles
from Cape Grisnez, than with Boulogne which is nearly ten.

For a fuller discussion of this subject I must refer my readers to
Mr. Long's article Itius Portus in Dr. Smith's Dictionary of Ancient
Geography, and to his Appendix to his History, vol. iv. already cited,
as well as to the work of M. F. de Saulcy (Les Campagnes de Jules
Cesar dans les Gaules, Paris, 1862, pp. 125-224). The arguments on
the other side are ably brought forward by Mr. Lewin {Invasion of
Britain hy Julius Ccesar, 8vo., London, 1859).

NOTE C, p. 119.

LANDING OF CiESAR IN BRITAIN.

This point, like that discussed in the preceding note, after
having been regarded as a settled question by most English his-
torians and topographers from Camden down to our own day, has
of late been much disputed. It would far exceed the limits of a
note to enter into the details of the controversy, for which I must
again refer my readers to Mr. Long's valuable History of the Decline
of the Boman Bepuhlic, vol. iv. Appendix I., who has, in my opinion,

138 HISTORY OF ANCIENT GEOGEAPHY. Chap. XIX.

successfully refuted the arguments of those who contend that C^sar
must have landed to the westward of Dover, in the neighbourhood
of Hythe or Lympne. The contrary seems certainly implied in
the expression of Caesar that, on his second voyage, when his ships
had drifted with the tide, he found at dawn of day that he had left
Britain behind him on his left (" longius delatus ^stu orta luce sub
sinistra Britanniam relictam conspexit," v. 8). This passage is in
my opinion decisive of the whole question. It is not only clearly
intelligible, but appropriate and graphic (as Cesar's language
generally is) on the supposition that the fleet was carried through
the Straits of Dover beyond the South Foreland, where the coast
trends away to the north. The advocates of the opposite theory
fail to give any intelligible explanation of it in accordance with
their views. I will only add that the distance of 12 miles from
the place of his landing to the river where the enemy first disputed
his advance would just about bring him to the banks of the Stour.
This question has also been fully investigated by M. F. de Saulcy,
in the work cited in the previous note, who arrives at the conclusion
that Cassar sailed from Wissant and landed at Deal.

NOTE D, p. 120.

PASSAGE OP THE THAMES.

The precise spot at which Ceesar crossed the Thames has been a
subject of much controversy, and cannot yet be said to be deter-
mined with certainty. But it may be placed with reasonable
assurance within narrow limits. It could not have been lower
down than Kingston, because the tide comes up as far as Teddington,
just below that town : and there is no reason to place it higher up
than Chertsey. At the present day the river is fordable at many
points between these towns, the most practicable of such fords
being at Sunbury. But the name of Coway Stakes, still given to a
spot on the north bank of the river, near the mouth of the Wey,
and the tradition preserved by Bede, that the stakes still visible in
his day in the river-bed were those which had been driven in to
prevent the passage of Caesar, certainly give a strong probability to
the supposition, adopted by Camden and others, that this was the
^ery spot where he crossed the river. Its distance from the sea

Notes E, F. CiESAR S WARS. 1 39

would also accord sufficiently well with the statement of Cassar
that the Tamesis was about 80 Eoman miles from the sea (5. G. v.
11). This estimate could obviously have reference only to his
own march from the neighbourhood of Deal. The direct distance
from the Thames to the nearest part of the coast of Sussex he had
no means of knowing. (Caesar, B. G. v. 18 ; Orosius, vi. 9 ; Bede,
Hist. Eccles. i. 2 ; Camden's Britannia, vol. ii. p. 168. See also the
Archceologia, vol. ii. pp. 141-158, and a note to the Emperor Napo-
leon's Hist, de Cesar, vol. ii. p. 191.)

NOTE E, p. 120.

THE CAPITAL OF CASSIVELLAUNUS.

Mr. Merivale supposes this "oppidum" of Cassivellaunus to
have been on the site of Verulamium, but there seems to me no
foundation for this. It is precisely in reference to this " oppidum "
of Cassivellaunus that Caesar explains what was meant by the term
among the Britons — a mere stockade or enclosed space in the midst
of a forest where they took refuge with their flocks and herds in
case of an invasion. " Ab his cognoscit non longe ex eo loco oppi-
dum Cassivellauni abesse silvis paludibusque munitum, quo satis
magnus hominum pecorisque numerus convenerit. Oppidum autem
Britanni vocant, cum silvas impeditas vallo atque fossa munierunt,
quo incursionis hostium vitandse causa convenire consuerunt " (J5. G.
vol. V. 21). There would be little reason why such a temporary
stronghold should become converted into a Eoman town.

Other writers place it in the neighbourhood of Wendover, a diver-
gence which sufficiently shows the titter absence of any real clue to
its position.

NOTE F, p. 120.

BRITISH TRIBES.

The names of these tribes as given by Csesar (JB. G. v. 21) are
the Cenimagni, Segontiaci, Ancalites, Bibroci, and Cassii : none of
which are mentioned by any later writer or are found in Britain
under the Eoman dominion. Hence they cannot be placed with

140 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XIX.

any certainty, though there are some reasons for supposing the
Segontiaci to have occupied a part of Berkshire, of which Silchester
was the capital. (Beale Poste, Britannic Hesem-ches, p. 15 5.) It
has been proposed by some editors to read " Iceni, Cangi," for the
unknown name of the Cenimagni, but there is no authority for so
arbitrary a change. It was first proposed by Lipsius in a note on
Tacitus {Annal. xii. 32), and has been adopted by the recent editors
Nipperdey and Oehler : but it is improbable that so powerful a
tribe as the Iceni should have submitted so readily. It is much
more likely that the names thus enumerated by Ceesar should have
been comparatively unimportant tribes on the banks of the Thames
(say in Buckinghamshire and Berkshire) which at a later period
had been absorbed into the more important tribes, or were too
insignificant to attract notice.

The Cassii are assumed by Dr. Latham (Did. of Anc. Geogr. s. v.)
to be the people of whom Caspivellaunus was king, but this is not
stated by Caesar, and is certainly at variance with this incidental
notice of their submission, while Cassivellaunus still held out.
Cassar does not mention over what people that chieftain originally
ruled : he Lad established himself on the throne of the Trinobantes
by the murder of the previous king — the father of Mandubracius
(B. G. V. 20) — but it is not clear whether this was his original
kingdom, or an addition made to his previous dominions. The name
of Cashiobury (near Watford, about 7 miles S.W. of St. Alban's)
may possibly, as suggested by Dr. Latham, retain some trace of that
of the Cassii, but the evidence of a single isolated name is very
precarious.

( HI )

CHAPTER XX.

EOMAN EMPIRE UNDER AUGUSTUS.

Section 1. — Roman Empire under Augustus.

§ 1. The annexation of Egypt as a Roman province (b.o. 30)
completed in great measure the fabric of the Roman Empire,
in the form which it retained with comparatively little alter-
ation during a period of three centuries. The whole extent
of the Mediterranean Sea, which still continued to be the centre
of the ancient world, was now encircled by an uninterrupted
chain of provinces, either directly subject to the Roman ad-
ministration, or held by tributary and dependent kings, who
enjoyed their nominal sovereignty only at the pleasure of their
all-powerful neighbour. The number of these vassals or pro-
tected states still continued to be more considerable in the time
of Augustus than at a later period, the greater part of them
having afterwards been gradually absorbed into the vast
monarchy of Rome. It will assist us in considering the state
of geographical science under the Roman Empire, its progress
and its limits, if we take a brief preliminary survey of that
Empire itself, as it was first constituted under Augustus, as
well as of its relations with its immediate neighbours.

Commencing with the West, the whole of Spain had been
reduced to a state of subjection, and was divided into three
provinces. Some of the northern tribes, indeed, the Cantabri
and the Astures, who held the rugged mountain regions
adjoining the Bay of Biscay, had still maintained their inde-
pendence, until after the accession of Augustus, and were not
finally subdued until the year 22 b.c.^

' Dion Cass. liii. 25, liv. 5. Strabo, I with the allusions to these wars in
iii. p. 1.56. All scholars are familiar | Horace (" Cantaber noii ante doma-

142

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XX.

The conquest of Gaul had been completed by Julius Cjesar,
and the whole country, from the Pyrenees to the Rhine and
the Ocean, passed without difficulty under the dominion of
Augustus. The foundation of numerous colonies, and the con-
struction of roads in all directions tended rapidly to dissemi-
nate Eoman civilization through all parts of the country ; and
while the Eoman armies on the Rhine were kept in almost con-
tinual hostilities with their neighbours, the Germans, on the
other side of that river, Gaul itself appears to have enjoyed
almost undisturbed tranquillity.

It was especially to Agrippa that Gaul was indebted for
much that contributed to promote its prosperity. It was he
that first laid out and constructed four great lines of road, all
proceeding from Lugdunum (Lyons) as a centre, of which one
traversed the central provinces as far as the Santones on the
Western Ocean, another led to the Rhine ; a third to the
Northern Ocean, adjoining the Bellovaci and Ambiani ; and
the fourth southwards to the province of Narbo and Massilia.^
It is from the same period that dates the distribution of Gaul
into four provinces ; the old Roman province of Gallia Nar-
bonensis in the south, Belgica in the north, and Gallia Lug-
dunensis, which extended from Lugdunum to the farthest
extremity of Armorica, but was bounded by the Loire to the
south : the whole territory from that river to the Pyrenees
being included under the name of Aquitania, though the people
of that name, as described by Csesar, did not extend north of
the Garonne.^

§ 2. No attempt was made either by Augustus or his immediate
successor to follow up tlie imperfect designs of Julius Csesar,
by renewing the invasion of Britain. We are told indeed that
on two occasions — once before the downfall of Antony, and

bilis," Carm. iv. 14, 41 ; " Cantabrum
indoctiun juga furre nostra," Ibid. ii.
G, 2, etc.). The name of the Astures
appears to have excited less attention,
and is not found in the Eoman poets
of this period ; lliongli it has survived

to our own days in that of the Asturias.

" Strabo, iv. p. 208.

^ This division continued in use till
the time of Constantine, and is recog-
nised both by Pliny and rtolemy.

Sect. 1. ROMAN EMPIRE UNDER AUGUSTUS. I43

again at a subsequent period (B.C. 27) — the emperor enter-
tained the project of an expedition to the British Islands ; * but
he contented himself with the more practical and easier task of
settling the administration of Gaul, and accepted friendly over-
tures from the princes and chieftains of the island, without
insisting on the payment of a regular tribute.^

With Germany, on the other hand, the relations of the
Romans were becoming continually more frequent, and though
they were generally of a hostile character, they could not but
add materially to the knowledge previously possessed of these
wild and thinly-peopled regions, hitherto so little known either
to Greek or Roman writers. The expeditions of successive
Roman generals, who carried their arms as far as the Weser
and the Elbe, will deserve to be noticed in their chronological
sequence. But no part of Germany beyond the Rhine was per-
manently added to the Roman dominions under Augustus.
It was not till a considerably later period that the Roman
frontier was carried to the line stretching across from the
Rhine to the Danube, so as to include almost the whole of
Suabia.

§ 3. On the southern side of Germany the case was very
different. It was here that the Roman Empire received by far
its most important accession under Augustus, by the conquest
of what may be briefly called the Danubian provinces, including
Rhsetia, Vindelicia, Noricum, and Pannonia. Strange as it
appears to us at the present day, it is an undoubted fact that
while Italy was extending its power to the Western Ocean on
the one side, and to the Euphrates and Araxes on the other,
the wild tribes on its own northern frontier had never been
subdued, and the valleys and defiles of the Alps were still held
by races of hardy and vigorous mountaineers, who defied the
power of Rome and disdained even the semblance of submission.
It was not till long after the accession of Augustus to the
imperial power that he turned his attention in earnest to the

' Dion Cass. xlix. 3S, liii. 22, 25. * Straho, iv. p. 200.

144 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

subjugation of these tribes, who had recently provoked his
interference by lawless incursions into Cisalpine and Helvetian
Gaul, in which they had displayed even more than their accus-
tomed barbarity. They were, however, effectually reduced to
subjection (in B.C. 15) by the two step-sons of the Emperor,
Drusus and Tiberius, their strongholds in the mountains
stormed, and a considerable part of the population compelled
to emigrate. The Bheetians, who held the mountains adjoining
Tridentum (Trent), and extended from thence through the
Tyrol into the Grisons, were the first to succumb ; but the
Yindelicians, who occupied the northern slopes of the Alps,
were subdued within the same summer, and the Koman frontier
was carried at once to the Lake of Constance and the Danube.^
The foundation in the newly acquired territory of the colony
of Augusta Yindelicorum (Augsburg) which speedily rose to
be one of the most flourishing and important colonies of the
empire,'' tended materially to consolidate the new conquest.
Noricum, a district which had previously maintained friendly
relations with Eome,^ shared the same fate, apparently on very
slight grounds of provocation.^

The Pannonians, on the other hand, did not succumb without
a vehement and long-continued struggle with the Roman power.
On their south-western frontier they immediately adjoined

« Dion Cass. liv. 22 : Strabo, iv. 6,
p. 206 ; Veil. Pat. ii. 96. These were
the campaigns which are celebrated by
Horace in two of the finest odes of his
fourth book, which was published
within a few years afterwards (Oarm.
iv, 4, and 25). The name of the
Vindelici appears on this occasion for
the first time. That of the Rhasti
was known to Polybins (np. Strab.
iv. p. 209) ; bi;t it is hardly likely that
tliat author had any real acquainlance
with the tribes on the other side of
the Alps. The two nations appear
thro^ighout as intimately connected
with one another, and were probably of
common origin.

' It is termed by Tacitus (Germania,

c. 41), " splendidissima Rsetite pro-
vincial colonia."

^ A king of Noricum is mentioned
by C8el^ar (Bell. Civ. i. 18) as sending
an auxiliary force of 300 cavahy to his
support at the outbreak of the Civil
War. He. must therefore have esta-
blished friendly relations with him
during the time that he held the com-
mand in Cisalpine Gaul. But the ex-
tensive use among the Romans of Noric
iron, whicli appears to have been the
principal source of their supply of that
indispensable metal, implies the exist-
ence of extensive commercial relations.

" Dion Cass. liv. 20. See Note A,
p. 144.

Sect. 1.

ROMAN EMPIRE UNDER AUGUSTUS.

145

that of the Eomans, and their native hardihood led them to
molest their richer neighbours by continual incursions. An
extensive commerce was already carried on from Aquileia, in
the land of the Yeneti — one of the most flourishing cities of
Northern Italy — over the pass of Mount Ocra into the valley of
the Save, and thence to the Danube : ^ and this it became an
important object with the Eoman government to secure.
Hence we find Augustus, as early as B.C. 35, conducting an
expedition in person into Pannonia, which ended with the
capture of their strong city of Siscia on the Save,^ a blow which
was followed for a time by the submission of the whole people.
They were however far from being eiFectually subdued : we
find them again in arms in B.C. 15, when they invaded the
Eoman province of Istria : and it was not till a.d. 8 that they
were finally reduced to subjection.^

§ 4. But if it is remarkable to find provinces so nearly
adjacent to Italy retaining their independence to so late a
period, it is still more surprising, according to our modern
notions, to learn that this was the case to a great extent even
with tribes on the south side of the Alps, and which we are
accustomed to consider as altogether included within the limits
of Italy. Yet it is certain that many of these mountain tribes
were, at the time when Augustus first ascended the throne.

' Strabo, vii. p. 314. Strabo cor-
rectly points out that this pass was the
lowest part of the Alpine chain which
extended from the 'Rhsetian Alps to
the country of the lapodes, where it
rose again to a more considerable ele-
vation. The lapodes occupied a part
of the modern Croatia, extending from
the Save and the Kulpa, to the Gulf
of Quarnero at the head of the Adriatic.
They were a wild and warlike race,
who were first reduced to subjection by
Augustus.

2 Dion Cass. xlix. 36, 37. Accord-
ing to the boast of Augustus himself,
this was the first occasion on which
the Pannonians were assailed by the
Eoman arms (Monum. Ancyr. p. 35).

Siscia was situated at the junction

VOL. IT.

of the Kulpa (Colapis) with the Save,
a position which gave it an importance
analogous to that of Belgrade in modern
days. The site is still marked by a
village named Siszek.

3 Dion Cass. Iv. 29-34; Veil. Pat.
ii. 110-116; Suet. Tib.l6; Mon. Ancyr.
p. 35. The alarm at Eome on this last
occasion was great, principally no
doubt on account of the proximity of
the enemy, who were actually design-
ing to invade Italy by the pass of
Mount Ocra ; and could thus have
been at the gates of Eome, it was said,
within ten days ! Hence also Sueto-
nius (I.e.), with obvious exaggeration,
calls the war "gravissimum omnium
exlernorum bellorum post Punica."

146

HISTORY OF ANCIENT GEOGKAPHY.

Chap. XX.

either altogether independent of the Eoman authority, or, if
nominally tributary, yet exempt from all practical control, and
ready at any time to break out into hostilities. Such was the
position of the Salassi, who occupied the great valley of Aosta,
from its entrance at Ivrea to the foot of the mountain passes at
its head. They had indeed been attacked and defeated as
early as B.C. 143 by the Consul Appius Claudius, and in
B.C. 100 the Roman colony of Eporedia (Ivrea) was settled at
the mouth of the valley. But they continued to retain their
lawless and predatory habits, and committed constant depre-
dations upon the neighbouring colonists, as well as upon all
who had occasion to pass through their country. In one
instance they plundered the baggage of a part of Ceesar's
army ; and compelled Decimus Brutus on his retreat from
Mutina in B.C. 43 to purchase his passage by the payment of a
large sum of money.* It was not till after repeated campaigns
that they were finally reduced to subjection by Terentius
Yarro in B.C. 25 ; a result that was attained only by the almost
total extirpation of the tribe.^ At the same time the founda-
tion of the Roman colony of Augusta Praetoria (Aosta) at the
point of junction of the two passes of the Great and Little
St. Bernard, served to secure those two important lines of
communication.®

Very much the same state of things existed also in regard
to other Alpine tribes in somewhat similar situations, such as
the Camuni — whose name is still retained in that of the Val
Camonica — the Lepontii at the head of the Lacus Verbanus or
Lago Maggiore ; — the Triumpilini, in the Val Trompia, &c. :
all of which were seated on the Italian slope of the Alps.

* Strabo, iv. p. 205.

^ Id. ibid. Dion Cass. liii. 25.

" It is certain that from this period
onward the only three passes of the
Alps from Italy into Ganl that were in
habitual use were : 1, that through
the Cottian Alps (the Mont Genevre),
which led down the valley of the
Durance into that of the Rhone ; 2, the

Little St. Bernard (per Alpes Graias),
which led from Augusta direct to Lug-
dunum, and was much the sliortest
route into central Gaul ; 3, the Great
St. Bernard (per Alpes Peuninas),
which was tlie direct road to the Lake
Lemannus and the country of the
Helvetians.

Sect. 1.

ROMAN EMPIRE UNDER AUGUSTUS.

147

The Maritime Alps were in like manner held by Ligurian
tribes, who retained their independence till the year 14 B.C.,
when they were for the first time reduced to subjection.''
Augustus after this carried a high road through their country,
and to commemorate the final conquest of the Alpine tribes
erected a monument at the highest point of the pass, record-
ing the names of not less than forty-four "Gentes Alpinee
devictae." The monument is still standing, at a place called
Turbia (a corruption of Tropsea Augusta), though in ruins:
but the inscription has fortunately been preserved to us by
Pliny,^ and is one of our most important authorities for the
topography of the Alpine tribes. Many of them, however,
are of course obscure names, otherwise unknown, and which
cannot be determined with any certainty.

But even thus the whole of the Alpine regions were not yet
incorporated with the Eoman Empire. There still remained
twelve petty tribes, placed under a native king named Cottius,
who having been uniformly friendly to the Eomans, had given
no pretext for hostilities, and continued to enjoy the nominal
sovereignty of a small mountain territory, standing in much
the same relation with the Roman Empire that the "pro-
tected " native princes hold with the British Empire in India.
His capital was Segusio, now Susa, and an inscription still
extant records the names of the " civitates " subject to his
rule.^ It was not till the reign of Nero that this petty princi-
pality was formally included in the Roman dominions : and
the name of the Cottian Alps attached to the portion of the
range adjoining the Mont Genevre, continued to perpetuate
until long after the memory of their obscure ruler.

§ 5. On the other side of the Adriatic, lUyricum and DaJ-

' Dion Cass. liv. 24.

» Plin. iii. 20, s. 24, § 136. The in-
scription bears the titles Imp. xiii.
Tr. Pot. xvii., which prove that this
monument was not set up till B.C. 7.

^ It is still visible on a triumphal
arch at Susa erected by Cottius himself
in honour of Augustus ; and is given
in Orelli's Inf^criptionen Latinpe Selectx,

No. 626. Most of these names are
otherwise unknown, and were doubt-
less those of obscure and petty tribes ;
but among them are those of the
MeduUi and Caturiges, who occupied
the valleys on the Gaulish side of the
Mont Genevre, and we thus learn that
the dominions of Cottius extended on
both sides of the Alps.

L 2

148

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

matia were for tlie first time reduced to permanent subjection
as a Roman province under the reign of Augustus. lUyricum
liad been indeed conquered as early as B.C. 167, wben after the
defeat of its king Gentius, it. was reduced to a condition
analogous to that of Macedonia, though it did not then receive
the formal organization of a province. But at a later period we
find it placed under the government of Julius Caesar, at the
same time with the two Gauls : and at this date it seems to
have already passed into the ordinary condition of a Roman
province. Nor do we hear of its causing him any trouble ;
except a plundering incursion of a people called the Pirustse on
its south-eastern frontier.^ But the Dalmatians, a race of hardy
and stubborn mountaineers in the northern part of the province,
were certainly still unsubdued at this time : and when, in
B.C. 34, Augustus (then only triumvir) undertook their re-
duction in person, he encountered an obstinate resistance, and
his efforts were attended with but partial success.^ The Dal-
matians, as well as their neighbours the Pannonians, appear in
arms again and again : it was not till a.d. 10 that they were
finally reduced to subjection by Tiberius, who received the
honour of a triumph over them, at the same time as for his
more celebrated victories over the Germans.^ From this time
the Roman province of Illyricum extended from the frontier of
Epirus to the Save : it was often called, at least in official
language, Dalmatia, from the prominent part assumed by the
people of that name : but the general designation of Illyricum
was commonly retained by geographers and historians.*

> Cffisar, B. G.y.l.

2 Dion Cass. xlix. 38. This out-
break on the part of the Dalmatians
seems to have originated during tlie
Civil War, when the generals of
Pompey and Caesar were striving with
one another for the possession of Illy-
ricum. On this occasion Gabinius, the
lieutenant of Caesar, was defeated by
" the barbarians " with a loss of more
than 2000 men, and compelled to fall
back upon Saloua (Hirt. B. Alex. c. 43).
The towns of the sea-coast, Salona and
ladera especially, were at this time

flourishing and civilized places (lb. c.
42, 43), and faithful subjects of Rome,
though the barbarians of the interior
were still unsubdued.

^ Dion Cass. Iv. 34; Veil. Pat. ii.
116 ; Suet. Tib. 16.

■* Thus Suetonius, in the passage just
referred to, says : " Toto lUyrico.quod
inter Italiam, regnumque Noricum,
et Thraciam, et Macedoniam, intcrque
Dauubium flumen et sinum maris
Adriatici patet, perdomito et in ditio-
non redacto." See also Tacitus (Annul.
ii. 44 ; Hht. i. 9, 76).

Sect. 1.

ROMAN EMPIRE UNDER AUGUSTUS.

149

§ 6. Immediately adjoining Illyricum on the east lay the
extensive province of Moesia, including the whole of the
countries now known as Servia and Bulgaria, extending from
the mountain barrier of the Balkan (Mt. Hsemus) to the
Danube. This tract also was a recent addition to the Roman
Empire. The conquest appears to have taken place in B.C. 29,
when Marcus Crassus (the grandson of the triumvir), having
been led across the Hsemus in pursuit of the Dacians and
Bastarnae, who had attacked the Roman allies, was not content
with driving those nations back across the Danube, but re-
duced the Mcesians themselves to subjection.^ It does not
appear that McBsia was at this time converted into a Roman
province, but this must have taken place not long afterwards.
It had certainly assumed the character of an ordinary province
before the accession of Tiberius.^

At this time therefore the Danube formed the northern
boundary of the Roman Empire, from its sources in the Her-
cynian forest almost to its mouth. The Peninsula of the
Dobrutscha, formed by the abrupt deflection of the river to
the north, when it has reached within 50 miles of the Euxine,
was alone excepted; this barren and pestilential region was
not annexed to the Roman dominion till a later period.'' On
the sea-coast the town of Tomi— so well known as the place
of banishment of the unfortunate Ovid-«-marked the extreme
limit of their power, and might be justly regarded as the
farthest outpost of civilization.^ The wild barbarians of the

s Dion Cass. li. 23-27.
On this occasion the name of the
Triballi, which had disappeared from
history for nearly three centuries; is
once more mentioned.

'' Tacit. Annal. i. 80. Augustus
himself in the Monumeutum Ancyra-
num (p. 35) speal^s of having not only
carried the frontiers of the Empire to
the Danube, and driven back the
Dacian army across that river, but that
his own armies had followed them
across the Dfinube and compelled the
Dacians to submission ; a circumstance
not mentioned by any historian.

' At the time when the Itineraries
were compiled, the Roman territory
was extended to the Danube : the
frontier town of Noviodunum was ap-
parently situated near Tultcha {Itin.
Ant. p. 22(j).

* Thus the poet's exclamation was
no exaggeration :

Longius hac nihil est, nisi tantum fiigiis et
hostis
Et maris adstricto qua3 coit unda gelu.
Hactenus Euxini pars est Romana sinistri ;
Proxima Basternte Sauromatteque tenent.

Tristia, ii. 195-19S.

The Greek colony of Istrus or Istro-

ISO

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

plains beyond were in the habit, as the poet tells us, of
carrying their depredations up to the very walls of the city.^

§ 7. North of the Danube there were only nomad or half-
civilized tribes, inhabiting the vast plains of Hungary and
Southern Eussia; and known to the Eomans only by their
occasional irruptions into the adjoining provinces, and the
hostilities to which these incursions gave rise by way of re-
prisals. The principal of these nations at the period we are
now considering were the Dacians, the Bastarnse, and the
Sarmatians, more commonly known to the Eomans by the
vague term of Scythians. The Dacians occupied the whole of
what now forms the southern part of Hungary, the Banat and
Transylvania : they appear to have been at this time gathering
strength under a king named Boerebistas, and had reduced or
exterminated some of the neighbouring tribes,^ but the more
prominent part which they henceforth assumed in Eoman history
was probably owing principally to the immediate proximity
in which they now found themselves to the Eoman frontier.^
The question of the relation in which the Dacians stood to the
Getge, whom we find in possession of these same countries at
an earlier period, was one on which there existed considerable
difference of opinion among ancient writers : but the prevailing
conclusion was that they were only different names applied to
the same people.^ Even Strabo, who describes them as distinct,
though cognate tribes, states that they spoke the same lan-
guage.* According to his distinction the Getse occupied the

polls was indeed situated 250 stadia
farther north ; and was certainly still
in existence, but it seems to have been
at this time a place of little import-
ance (it is called a izoxixviov by Strabo,
vii. p. 319), and was not occupied by
the Eomans.

The site of Tomi seems to be now
clearly established at Kustendje (the
Constant! ana of Procopius), one of the
best ]iorts along this line of coast.

» Ovid. Tristia.

• Strabo, vii. 3, p. 304.

'-' All readers of Horace are familiar

with the prominent manner in which
the " rugged Dacian " (Dacus asper)
figured in the imagination of the
Romans of his day among the fierce
barbarians still hanging on the out-
skirts of the Empire (Herat. Carm. i.
35, 9; ii. 80, IS; iii. 6, 14, 8, IS).
Hostilities were carried on with them
on several occasions during the reign
of Augustus, but with no important
result.

^ Strabo, vii. p. 301 ; Dion Cass. li.
22.

■* Strabo, I. c.

Sect. 1. KOMAN EMriRE UNDER AUGUSTUS. I5I

more easterly regions adjoining the Euxine, and the Dacians
the western, bordering on the Germans. Probably the name
of Geta3, by which they were originally known to the Greeks
on the Euxine, was always retained by the latter in common
usage : while that of Dacians, whatever be its origin, was
that by which the more western tribes, adjoining the Panno-
nians, first became known to the Romans.^

The Bastarnse, who had already become known by name at
least to the Romans as early as B.C. 168, when they furnished
an auxiliary force to Perseus, king of Macedonia, and who now
reappear as the neighbours and allies of the Dacians, are a
people of whom very little is really known. They are dis-
tinctly termed by Dion Cassius a Scythian (meaning probably a
Sarmatian) race ; but Strabo says of them that they adjoined the
Germans and were almost of German race themselves : and the
same conclusion is adopted by Tacitus, who though including
them among German nations intimates some doubt as to whether
they were really Germans or Sarmatians. Other writers describe
them as Gauls, or of Celtic race.® Their place of abode at this
period is not clearly defined ; but they appear to have been
situated east of the Carpathians, between them and the Sar-
matian people called by Strabo the Tyrigetse.^

§ 8. The tract adjoining the lower course of the Danube on the
north, and extending from thence along the coast of the Euxine
to the Borysthenes, seems to have been held at this period en-
tirely by Sarmatian tribes. These pressed closely on the

^ The uame of Dacians is first found 1 them, see Zeuss, Die Beutsclien, pp

in Csesar {B. G. vi. 2.5), where he
speaks of the Hercynian Forest as ex-
tending along the Danube to the con-
lines of the Dacians and Anartians
He therefore appears to have considered

127-130; Ukert, Geogr. \ol. iii. pt. ii.
pp. 427, 428 ; Schafarik, Slavische
AUerthiimer, vol. i. p. 393.

' Strabo, I. c. But the Peuciui,
whom all writers agree in associating

them as immediately adjoining the j with the BastarnaB, are described by

Germans on the east.

" Dion Cass. li. 23; Strabo, vii. p.
306 ; Tacit. Germ. c. 46. For a full
discussion of the question concerning
the Bastarnas, and the diiferent pas-
sages of ancient writers relating to

him as inhabiting, and deriving tiieir
name from, the island of Peuce at the
mouths of the Danube. Ovid also
speaks of the Bastarnse and Sauroraati
as inhabiting the country immediately
beyond Tomi {Tridla, ii. 198).

152 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

Eoman outposts in this quarter : Tomi, as we learn from the
unhappy Ovid, was assailed by Sarmatian as well as Getic
marauders, and even the population of the town was com-
posed in great part of Sarmatians as well as Getse.^ Beyond
the mouths of the Danube the Eomans seem to have had little
intercoiu'se, and certainly exercised no political influence over
the population. At the mouth of the Borysthenes indeed the
Greek colony of Olbia or Olbiopolis still maintained its posi-
tion, and must have continued to carry on a considerable
amount of trade with the interior. In like manner the little
kingdom of the Bosporus still subsisted at the entrance of the
Sea of Azov and retained a nominal independence, though
acknowledging the supremacy of the Eoman Emperors.^ It
continued to preserve some tincture of Greek civilization
down to a late period.

§ 9. Eeturning to the south of Moesia, the two important
regions of Macedonia and Thrace were very differently situated
in their relations to the Eoman Empire. Macedonia had long
been reduced to the condition of a Eoman province ; the ex-
tent of which however considerably exceeded the limits of
Macedonia properly so called. On the west it comprised a
considerable part of what had previously been reckoned as
lUyricum, so as to extend to the Adriatic and include the two
important points of Dyrrhachium and Apollonia : while to-
wards the east the coast-line of Thrace along the -iiEgean, as
well as the Thracian Chersonese, were also annexed to the
province of Macedonia. The whole of Thessaly also was
subject to the same jurisdiction. The province was traversed

8 Ovid, Tristia, ii. 191. Thelazyges,
a Sarmatian tribe, who are first men-
tioned among the barbarian nations in
this quarter, with whom Mithridates
entered into alliance (Appian, Mithfi-
dat. c. 69), reappear in Ovid, and were
apparently at this period one of the
most powerful branches of the Sar-
matians.

recurs iu Ovid in one passage only, in
which he adds the epithet " ilavi "'
(Ex Ponto, iv. 2, 07).

'' This is clearly proved by their
coins, which form an unbroken series
from the time of Augustus to that of
Constantine. They bear the head of
the reigning Eoman emperor on the
one side, and that of the king of Bos-

Another name found in Appian, iu j porus (with the title of BatnAeus) on
the fcame passage, the Coralli, also : the other.

Sect. 1.

IIOMAN EMPIRE UNDER AUGUSTUS.

153

from one extremity to the other by the Egnatian way, one of
the most important highways in the Empire, leading from
Dyrrhachium and Apollonia to the Hellespont, and thus
forming the main line of communication between Italy and
the Asiatic provinces. It had not however as yet been regu-
larly constructed any farther than Cypsela on the river
Hebrus.^

Thrace on the contrary had not at this period been regularly
incorporated with the Eoman Empire. The southern coast, as
we have just seen, had been annexed to the government of
Macedonia : the rest of the country continued under the rule
of native princes, who acknowledged the supremacy of the
Eoman Emperors, while retaining the title of king, and the
control of their internal administration. Before the accession
of Augustus, and especially during the Civil Wars, we find
the Eoman governors of Macedonia engaged in almost con-
tinual hostilities with some of the Thracian tribes, among
whom the Bessi and the Odrysae seem to have at this time
held the predominant place. Ultimately the latter obtained
the upper hand, and having had the sagacity to attach them-
selves to the Eoman alliance became masters for a time of all
Thrace. Their king Ehcemetalces, who was established on the
throne by Augustus, retained his power for a considerable
number of years, and appears to have reduced the Thracian
tribes to a state of comparative tranquillity.^ But the dis-
sensions between his successors led to repeated interference
on the part of Eome, and Thrace was ultimately reduced to a
Eoman province under the reign of Claudius. The flourishing
city of Byzantium was never subject to these Thracian princes

* Strabo speaks of the Egnatian Way
as ^e^'qfia/T Iff fx^vr} Kara jxiKiov ical Koreff-
TfiKwixtvq fJ-^XP'- Kuif'eA.coi' KoX "E0pov iro-
rafxov (vii. 7, p. 322), in a manner
■which must refer to his own time.
Cicero indeed at a considerably earlier
periodj calls it " via ilia nostra, quse
per Macedoniani est usque ad Helles-
pontum militaris " (Orat. de Provinc.

Consular, c. 2, § 4). But it may have
been opened for military purposes
■without having received the complete
finish to ■which Strabo refers : or the
words of Cicero may not be intended
to be construed strictly.

2 Dion Cass. liv. 20, Iv. 30 ; Tacit.
Aniud. ii. 6i.

154

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

and still retained under the Eoman Empire its nominal in-
dependence and autonomy.^

Greece, witk the exception of Thessaly (which, as we have
seen, was united with Macedonia), constituted a Roman province
under the name of Achaia.* A large part of the cities indeed
enjoyed a nominal autonomy, of which Athens affords a well-
known example, but they were subject to Eome for all but
municipal purposes. The greater part of the country had
already fallen into a state of depopulation and decay, which
afforded a melancholy contrast with its former greatness. The
newly founded Roman colonies constituted almost the only
exceptions, among which Corinth held the chief place, and
after its restoration by J. Caesar rose rapidly a second time
to be one of the most flourishing commercial cities in the
Mediterranean. Delos, which had for a time taken its place
and become an important emporium of trade, especially as the
great central mart of the slave trade, apj)ears never to have
recovered the blow it sustained during the Mithridatic War,
and was still in a decayed condition in the time of Strabo.^
Some of the other small islands of the ^gean were rendered
familiar by name to the Romans from their being frequently
used as places of banishment for political exiles.^

3 It retained this position till the
reign of Severus, when, having sided
with his rival Pescennius Niger, it sus-
tained a memorable siege of three
years, after which its walls were de-
stroyed, and it was reduced to a state
of comparative insignificance till the
time of Constantino (see Gibbon, c. 5).

* It is singular that Greece did not,
for a long time after its conquest, con-
stitute a separate province, but was
either united with, or treated as a mere
dependency of, Macedonia. It was not
definitely organised as a separate pro-
vince till the reign of Augustus. See
Marquardt, Handbuch der Bomischen
Alterthiimer, vol. iii. pp. 121-128;
Hertzberg, Gesch. Griechenlands uiiter
der Bonier, vol. i. p. 504.

^ Strabo, x. p. 486. Delos had been
made a free port by the Romaus after

the defeat of Perseus in b.c. 187 ; appa-
rently with a view to injure the trade
of Rhodes. Polyb. xxxi. 7, § 10. But
its great commercial prosperity did not
begin till after the fall of Corinth.
Concerning its great importance and
prosperity at tliis period, see Cicero,
Or at. pro Leg. Manil. 18, § 55.

" Juvenal, Sat. i. 73 ; vi. 563 ; x.
170. Tacit. Annal. ii. 85 ; iii. 68 ; iv.
21, 30, &c. The province of the islands
(Insularum provincia) was not consti-
tuted till a much later period. At
this time the Cyclades were apparently
included in Achaia, and the Sporades
and Asiatic Islands in the province of
Asia. The important island of Crete
was, by a singular anomaly, annexed
for administrative purposes to the
Cyrcua'ica.

Sect. 1. EOMAN EMPIRE UNDER AUGUSTUS. I55

§ 10. The Roman dominions in Asia had received no consider-
able addition, since they were first extended to the Euphrates and
the frontiers of Armenia, by the arms of Lucullus and Pompey.
But within those limits many changes had taken place, and
their political relations with the native princes still continued
in an unsettled state. Many of these changes had resulted
from the Civil Wars of the Eomans : princes and dynasts were
dethroned or restored, according as they favoured the one
side or the other, and provinces transferred from one petty
sovereign to another at the will of the victorious leader. But
it is unnecessary here to follow these successive arrangements,
which for the most part had a mere transitory political eifect,
without permanently affecting the geographical boundaries
of the countries in question. A very brief glance at these
relations, as they subsisted in the reign of Augustus, will
suffice for our present purpose.

The Eoman province of Asia was far from including the
whole of what we are now in the habit of designating as Asia
Minor. As originally constituted, it corresponded to the
dominions of the kings of Pergamus, in the enlarged form
that these had assumed after the defeat of Antiochus the
Great, when the Eomans had rewarded the support of Eumenes
in the war, by extending his limits to the Taurus. The
monarchy thus created was left by the will of Attains III.
to the Roman people (b.c. 133), and after the defeat of
Aristonicus was incorporated as a Roman province, B.C. 129.
It included the whole of Mysia and Lydia, with ^olis, Ionia
and Caria, except a small part which was subject to Rhodes,
and the greater part, if not the whole of Phrygia. A portion
of the last region was however detached from it, and after
various fluctuations of boundaries, that of the Roman province
of Asia was fixed so as to comprise the three districts of which
Laodicea, Apamea, and Synnada were the capitals, excluding
the eastern and south-eastern portions, which were annexed to
Galatia. ''

' For a more detailed view of the I I must refer my readers to tlie exotl-
proviiiees of Asia Minor at this period | lent maps by Dr. C. Miiller of tlie

156

HISTOEY OF ANCIENT GEOGEAPHY.

Chap, XX.

§ 11. Bithynia had in like manner been formed in the first
instance out of the kingdom of the same name, which had
passed, after the death of its last monarch, Nicomedes III., in
B.C. 76, into the hands of Eome. It received, however, after the
defeat of Mithridates the Great, a material accession of terri-
tory, and as constituted at that period by Pompey, it extended
along the shores of the Euxine as far as Themiscyra, thus
including the whole sea-coast of Paphlagonia, with a part of
that of Pontus. A petty dynasty of princes still continued to
rule over the interior of Paphlagonia, which was first united
to the Koman province by Augustus.

The rest of the kingdom of Pontus Avas not incorporated
with the Roman dominions after the defeat of Mithridates, or
even after that of Pharnaces by Caesar. It still continued to
be subject to the rule of a dynasty of princes, originally
selected by the Roman Emperors, and virtually dependent
on them, but still retaining full powers of local administration.
At the time of the accession of Augustus it was governed by
a Greek named Polemon, who had been appointed by Antony,
but was retained in his power by Augustus. This he trans-
mitted after a tranquil reign to his widow Pythodoris, from
whom it passed to their son, Polemon II., at whose death, in
the reign of Nero, a.d. 63, this part of Asia was for the first
time organized as a Roman province under the name of Pontus
Polemoniacus. The two last kings had materially extended
their dominion towards the east and north, and had reduced
the wild tribes that inhabited Colchis and the eastern coasts
of the Euxine to a nominal submission : the first Polemon had
also made himself master of the Bosporus, and the Greek
cities at its entrance. But no attempt was made by the
Romans to retain these conquests ; their dominion along the
Euxine never appears to have extended farther than the
confines of Colchis.^

Kingdoms of the Successors of Alex-
ander in Dr. Smith's Atlas of Ancient
Geography, pi. 5, 6.

' Slrabo, xi. 2, p. i'J6. The tribes

along tlie coast from the borders of
Colchis to those of the Greek settle-
ments on the Bosporus (the region
occupied in modern times by the Cir-

Sect. 1. ROMAN EMPIRE UNDER AUGUSTUS. 1 5/

South of Bithynia lay the province of Galatia, also one of
the most recent additions to the Roman Empire. This region
had continued, from the time of its first occupation by the
Gauls to that of Csesar, to be governed by chiefs with the title
of tetrarchs, each presiding with quasi-regal authority over a
portion of the country. But after the death of Csesar, Deiotarus
made himself king of the whole country, and his successor,
Amyntas, who was appointed by Antony, received from the
triumvir a large accession of territory, including Lycaonia,
Isauria, Pisidia, a part of Phrygia, and Cilicia Trachea. Having
conciliated the favour of Augustus, Amyntas remained in
possession of these dominions till his death in B.C. 25, but on
that event his kingdom was put an end to, and the provinces
subject to his rule were incorporated with the Empire under
the general name of Galatia, with the exception of Cilicia
Trachea, which was handed over to Ariobarzanes, king of
Cappadocia. Thus the province of Galatia, in the Roman
sense of the term, was far more extensive than the limited
region previously known by that name, and extended from the
confines of Bithynia and Paphlagonia to the range of Mount
Taurus.^

§ 12. Cappadocia, an extensive province occupying the
eastern portion of the great interior table-land of Asia Minor,
still continued to be ruled by its native dynasty, who had
earned the favour of the Romans by their steady support in
the wars against Mithridates and Tigranes : an alliance which
was however almost forced upon them by circumstances, those
monarchs being their most dangerous enemies. Cappadocia
at this time extended eastward to the Euphrates so as to
include the fertile district of Melitene, between that river and
the chain called Anti-Taurus.^ Armenia Minor, lying also

cassians) are described by him as a race I Maris Euxini, §26), Dioscurias was

of lawless pirates, whose depredations
the Roman governors took little jiains
to restrain. In the reign of Hadrian,
as we learn from Arrim {Periplim

still the limit of the Roman dominion
on this side.

9 See Note B, p. 200.

^ Strabo, xii. I, p. .534.

158

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

on the west of the Euphrates between Cappadocia and Pontus,
was not reckoned to belong to the kingdom of Cappadocia,
though actually held by the last king Archelaus. After the
death of this monarch in a.d. 17, his hereditary dominions
were converted into a Koman province :^ but Armenia Minor
remained a separate and nominally independent sovereignty
at least till the reign of Vespasian.

In the south-west corner of Asia Minor the districts of Lycia
and Pamphylia may be regarded as practically forming one
province, though they were not formally united as such until
the reign of Claudius.^ The cities of Lycia indeed continued
in the time of Augustus to retain in name at least their
independence, and to form a league for their self-govern-
ment : but they were under the protection of the Eoman
authorities, and doubtless in great measure subject to their
control.*

Cilicia, a country clearly marked out by nature, and in-
habited by a people who formed a separate nation from the
time of Herodotus, had nevertheless undergone strange vicissi-
tudes in its political condition. It first came in contact with
the Roman arms on account of the piratical incursions of its
inhabitants. These gave occasion to a Eoman prsetor, M.
Antonius, being sent against them as early as B.C. 103, and we
subsequently find repeated mention of Cilicia being assigned
as a province to Roman generals. But this was merely as the
theatre of hostilities : a province of the name was first consti-
tuted in B.C. 75, by P. Servilius Isauricus, who subdued the
Isaurians, and followed up his victory by reducing to submission
the rugged mountain country known as Cilicia Trachea. The
rich and fertile tract forming the eastern portion of Cilicia,
and known as Cilicia Campestris, was at this period still

^ Tacit. Annal. ii. 42 ; Strab. Z. c.

^ Sueton. Claud. 25.

* The condition of tlie Lycian cities
at this period, as forming a federal
league bnt a Roman dependency, is
well illustrated by their coins, wliich

bear the head of Augustus, but with-
out his name or imperial title ; while
on those struck under Claudius the
full imperial titles immediately appear.
(See Warren, On Greek Federal Coinage,
p. 38.)

Sect. 1. EOMAN EMPIRE UNDER AUGUSTUS. 159

subject to the kings of Syria, who retained possession of it till
the time of Pompey. That general wrested it from the hands
of Tigranes, king of Armenia, and united it with the portion
already occupied by the Eomans (B.C. 64). At this time the
province of the name comprised in addition numerous out-
lying districts — -Pamphylia, Pisidia, Isauria, Lycaonia, a large
part of Phrygia, and the island of Cyprus. Such was the
extent of the Eoman province of Cilicia, when Cicero was
appointed to the charge of it as Proconsul, B.C. 51. But these
arrangements were broken up by M. Antony : the extraneous
districts were finally separated from Cilicia, and that province
reduced within its natural limits. But the western portion, or
Cilicia Trachea, was handed over by Augustus to Archelaus,
king of Cappadocia : and was not reunited to the Roman
Empire till the reign of Vespasian. A petty dynasty of
native kings, of whom the names of Tarcondimotus and
Philopator alone are known in history, still maintained its
nominal sovereignty in the mountain tracts of Amanus, on
the eastern frontier of Cilicia, but the boundaries of their
territory are very imperfectly known. Cyprus, which had for
a time been united with Cilicia, was constituted by Augustus
a separate province, and retained its distinct government from
that time forwards.

The province of Commagene, on the west bank of the
Euphrates, which had been subject to the Seleucidan kings
of Syria, was at this period still governed by a native dynasty,
though under the protection of Rome. It was united to the
Empire for a time by Tiberius, but again placed under a
native ruler by Caligula, and finally reduced to a province by
Vespasian in a.d. 73. Its capital was Samosata, a strong town
on the Euphrates, in a position commanding the passage of
the river, which rendered it an important point in the wars
between the Romans and Parthians.^

'^ It is first mentioned in history
during the campaign of M. Antony
against the Parthians (b.c. 36) ; and its

military importance is attested both by
Strabo and Josephus (Strabo, xvi. p.
749; Joseph. Ant. xiv. 15, § 8; Bell.

i6o

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XX.

§ 13. The extensive province of Syria had been subject to
Eome, and ruled by Roman governors, ever since its conquest
by Pompey in B.C. 64. But though it was organized as a
Eoman province, and is repeatedly mentioned as such, it was
far from being brought under one uniform and regular
administration ; and its condition at this period was extremely
complicated. Judaea, though it had been twice conquered,
first by Pompey, and a second time by M. Antony, was at this
time still governed by its own king, Herod, commonly called
the Great, though tributary to the Roman Emperor, and
acknowledging his supremacy.^ Damascus again belonged to
a native prince of the name of Aretas, of Arabian origin, who
held also Bostra, with the surrounding district, and Petra, but
was certainly tributary to Rome. Native dynasties also reigned
in Chalcis, Emesa, and the district of Abilene. The Roman
policy indeed appears to have been at this time to maintain all
these petty princes in nominal sovereignty, but practically
dependent upon Rome. It was only by slow degrees that they
were successively absorbed under the imperial administration,
and it was not till the reign of Trajan that the province of
Syria could be considered as forming a complete organic
whole.'' The outlying city of Palmyra — the name of which is
first mentioned during the wars of M. Antony in Syria — was
certainly at this period independent and preserved a position
of neutrality between the Romans and Parthians, while it
carried on trade with both.^ It does not appear however to
have as yet risen to a place of great importance, as its name is

Jud. vii. 7, § 1). It had a bridge over
tlie Euphrates, apparently the only
one between the Zeugma and the nar-
row gorges of Mt. Taurus, from which
the river emerges about 30 miles above
Samosata. This came to be one of
the most frequented ]iassages of the
Euphrates under the Roman Empire.
The modern town still retains the name
of Samsat.

" Herod reigned over Judsea, from
the time of Ids confirmation on tlio

throne by Augustus in b.c. 30, till his
death in B.C. 4.

' For the details, see Marquardt,
Handbuch der Bomischen AUerthiimer,
Th. iii. pt. i. pp. 17.5-194; who has
worked them out with great care aad
diligence.

8 Appian, B. Oh. v. 9 ; Plin. JS. N.
V. 25. This is clearly implied by
Pliny as continuing to be the case even
in his time.

Sect. 1. ROMAN EMPIRE UNDER AUGUSTUS. l6l

not mentioned by Strabo. The period of its prosperity dates
only from the time of Hadrian, by whom it was united to the
Eoman province.

§ 14. Beyond the Euphrates to the east lay the extensive,
and still powerful, monarchy of the Parthians, who at this
period held undisputed rule over Mesopotamia and Syria, as
well as the broad regions beyond, which had previously formed
part of the Persian Empire. Media Atropatene (the modern
province of Azerbijan) was, as we have already seen,® held by
a separate dynasty of rulers, who were however the constant
allies of the Parthians, and probably to a great extent de-
pendent upon the greater monarchy. Armenia on the contrary
may be considered as being in some degree dependent upon
the Eoman Empire, though always retaining its own kings.
But after the expedition of M. Antony to Artaxata, and the
deposition of Artavasdes, the succeeding monarchs were for
some time appointed by the Eoman emperors, and though these
were repeatedly expelled by their own subjects, or by the
neighbouring kings of Parthia, it is clear that both Augustus
and Tiberius not only claimed, but repeatedly exercised, the
right of nomination to the vacant throne.^ The national
feeling of the Armenians however inclined rather to the Par-
thians than to the Eomans, and the possession, or rather the
supremacy over this important province continued to be
the subject of repeated contests between the Eoman and the
Parthian monarchs, from the time of Augustus to that of
Trajan.

With the Parthians themselves Augustus had the wisdom to
avoid any open collision, and while shrinking from taking
up the project of the dictator Csesar to avenge the defeat of
Crassus by a war with Parthia, he was able to gratify the pride
of the Eoman people by procuring the return of the standards
taken on that occasion. This result was greatly facilitated by

" See the preceding Chapter, p. 132.

' See a summary of these relations in Tacitus (Annal. ii. 3, 4, and 56).

VOL. II. M

1 62 HISTOEY OF ANCIENT GEOGEAPHY. Chap. XX.

the domestic dissensions of the Parthian royal family, and the
competition of rival claimants for the throne. They even con-
sented to accept as their king a Parthian prince who had spent
many years at the court of Augustus, and might therefore be
looked upon in some degree as appointed by that emperor.^
The circumstance was celebrated with triumph by the Eomans,
and from this period it became a favourite policy with suc-
ceeding emperors to interfere whenever it was possible, in the
disputes that were continually arising with respect to the Par-
thian succession. But no attempt was made till the reign of
Trajan to extend the Koman frontier on this side by any
permanent conquests.

§ 15. It is singular that Strabo^ speaks of the establishment
of the Parthian empire as one of the causes that had con-
tributed to the extension of geographical knowledge in his day.
The real eifect would appear to have been just the reverse.
By destroying the Greek monarchy in the provinces of Central
and Upper Asia, they cut off to a great extent the communi-
cations of all the interior of that vast continent with the Greek
world, and isolated almost completely the provinces in the far
East, on the borders of Bactria and India, which still retained
some traces of Hellenic civilization. The Greek element
indeed, with its characteristic vitality, continued to maintain
itself under the Parthian monarchy, as it did subsequently
under the Turks. Seleucia on the Tigris, a city of Greek origin,
and in great part peopled with Greeks, was an opulent and
thriving commercial city, and doubtless maintained relations
more or less frequent with the distant provinces of the empire.
The names of two Greek writers are also preserved to us, who
were born in cities subject to the Parthian rule, and who
undoubtedly contributed something to geographical knowledge.
Apollodoeus of Aetemita (a town of Assyria) wrote a
history of Parthia, which is repeatedly quoted by Strabo,^ and

2 Tacit. Annul ii. 1, 2.

^ Strabo, i. 2, p. 14 ; xi. 6, p. 508.

■* There is no clue to the date of this

Apollodorus. When Forbiger (Geogr.
vol. i. p. 356, note) calls him a contem-
porary of "Posidoniiis, tliis is a mere

Sect. 1.

ROMAN EMPIEE UNDER AUGUSTUS.

163

from which we should doubtless have derived much interesting
information had it been still extant : but though Strabo refers
to him as having thrown much additional light upon the
geography of Upper Asia, especially of Hyrcania, Bactriana,
and the neighbouring countries/ it does not appear that Strabo
himself had derived much benefit from his work. The only
instance in which we distinctly learn that he had improved
upon the knowledge of previous authors, was in regard to the
river Ochus, the modern Attrek, which had been unknown to
the earlier geographers, while others had confounded it with the
Oxus. As it flowed through Hyrcania, and in the immediate
neighbourhood of the original province of Parthia, it was
repeatedly mentioned by Apollodorus.^ The same author
appears to have given, incidentally at least, an account of the
Greek monarchy in Bactria, which would have been of especial
interest to us •? but there is no evidence of Strabo having any
additional geographical information concerning those coun-
tries, beyond what he derived from Eratosthenes and the histo-
rians of Alexander.

§ 16. Another writer, not referred to by Strabo, but fre-
quently cited by Pliny, is Isidoeus of Chaeax, a city of
Babylonia, near the head of the Persian Gulf. He appears to

conjecture ; but as it is not very likely
that he would have written a history
of the Parthians before they had at-
tained to their great power, he may be
fairly presumed to have lived within
the first century B.C. His native place
of Artemita was situated on a river
called Silla, at a distance of 500 stadia
from Seleucia. It is mentioned as a
place of consideration both by Strabo
(xvi. p. 744) and by Isidore of Charax
(§ 2), who distinctly terms it a Greek
city.

' Strabo, ii. 5, § 12, p. 118.

« Strabo, xi. pp. 509, 515. It is
not to be wondered at, that the Ochus
should have been unnoticed by earlier
writers, as it is really a stream of no
great importance. In modern times
the Attrek, though it has been sud-
denly brouglit into notoriety in con-

nexion with the advance of the Eussians
in Central Asia, was until lately very
imperfectly known to geographers, and
was confounded with the Tejend (the
river of Meshed) which does not flow
into the Caspian (see Wilson's Ariana,
p. 146). The Attrek rises in the same
range of hills as the Tejend, but has
from the first a westerly course, and
falls into the south-east corner of the
Caspian, about fifty miles north of
Astrabad. The recent travels of Mr.
V. Baker in this region have thrown
much light on its geography, and at
the same time have shown how imjoer-
fectly it was previoiisly known (Baker's
Clouds in the East, 8vo, 1877). See
also a valuable paper by Capt. G.
Napier in the Journal of Oeogr. Soe.
vol. xlvi.

' Id. xi. p. 516.

M 2

164

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XX.

have written in the reign of Augustus, though the statements
on this subject are very confused and contradictory.^ He is
quoted by Athenaeus as having written a work called UapOla'i
7repir]'yr]TCK0'?, in which he gave full details concerning the
pearl fishery,^ so that it could not be a mere dry geographical
treatise. But he must also have composed a more general
work on geography, as Pliny repeatedly cites him as an
authority concerning measures and distances in other parts of
the world,^ and regards him as having more recent and trust-
worthy information than Artemidorus and other earlier writers.
There is still extant under his name a very brief treatise called
in our manuscripts 'traOfMol HapOiKol, which, as its title implies,
is nothing more than an itinerary, giving the distances, first,
from Zeugma on the Euphrates to Seleucia on the Tigris, then
from that city across the passes of Mount Zagrus to Ecbatana,
and thence by Ehagee and the Pylae Caspiae, through Hyrcania,
Parthia, Margiana, Aria, and Drangiana to Alexandria or
Alexandropolis in Arachosia,^ where the Parthian empire
ended. It has every appearance of being an extract, or
abridged summary from his larger work : and in the great
dearth of information concerning the countries it traverses, is a
document of some value to us for the correction of our other
authorities. But it is in itself very meagre and slight, and
the paucity of details concerning the route eastward from

^ They are fully considered and
examined by C. Miiller in the Prolego-
mena to his GeograpM Gneci Minores
(pp. Ixxx-lxxxv), who supposes him
to be the writer meant by Pliny, who
was sent forward by Augustus to ex-
plore the eastern regions, when his
grandson Gains Cajsar was about to
lead an expedition into those countries.
Tlie name of this author, wlio is called
by Pliny " terrarum orbis situs recent-
issimum auctorem," is given in oiir edi-
tions as Dionymim, and has been erro-
iieously supposed to be the same with
Dionysius Periegetes. Tiiis error had
been already exposed by Bernharcly
(ad Dionys. Pcrieg. p. 49(j), who sug-

gested that Isidorus was the person
really meant, a conclusion adopted by
C. Miiller. Be this as it may, all in-
dications seem to concur in placing our
author in the reign of Augustus, or at
all events very little after it.

^ Athena3us, iii. p. 93 d.

1 Phn. H.N. ii. 242, 246; iv. 102,
121 ; v. 40, &c. All these citations
refer to questions of general geography,
such as the dimensions of Europe,
Africa, &c.

^ This Alexandria was undoubtedly
identical with the modern Candahar,
which is still the capital of the adja-
cent regions.

Sect, 1.

ROMAN EMPIRE UNDER AUGUSTUS.

165

Seleiicia, as compared with the portion west of the Tigris,
seems to prove how very imperfect was the information avail-
able at this period concerning the provinces of Upper Asia.
Nor do we find that the details contained in this little
treatise — such as they are — were made use of by subsequent
geographers. Strabo was certainly unacquainted with the
writings of Isidorus, whom he never mentions, and Pliny,
though repeatedly referring to his larger work, never cites his
authority in regard to the Parthian provinces.

§ 17. The Greek dynasties that had subsisted for a period
of nearly two centuries in Bactriana and the adjoining pro-
vinces south of Paropamisus, had long before this been over-
whelmed by the irruption of a race of barbarians from the
north,^ and probably the last traces of Greek civilization were
by this time extinct. Beyond the Parthian Empire to the
north and east there was nothing to be found but Scythians
and Indians. With the Asiatic tribes included by the Greeks
under the former appellation, it is not likely that the Eomans
at this period held any communication whatever : the Scythians
who are mentioned by historians as well as poets* as sending
embassies with offerings of friendship to Augustus, were in all
probability European Scythians from the neighbourhood of the
Euxine. 'But it was otherwise with the Indians. What was at
this period the extent and character of the commercial relations
with India carried on by the Romans, or rather by the Alex-
andrine Greeks under their authority, is not very clearly
known, though it was certainly not inconsiderable. But we
learn from Dion Cassius and Strabo that during the sojourn of
Augustus at Samos in B.C. 20 among the numerous embassies

^ The date of this event, as of almost
all others connected with tlie history of
these Grseco-Bactrian dynasties, is very
uncertain ; but the conquest of Bactria
proper by the Scythian tribes, who are
termed by Strabo Sac£e (xi. 8, § 2),
appears to have taken place about b.c.
128. Greek rulers, however, certainly
maintained themselves in the districts
south of the Hindoo Koosh to a con-

siderably later date ; and it is probable
that their dominion in these provinces
was not finally overthrown till about
90-80 B.C.

* Sueton. Oct. 21 ; Horat. Carm. iv.
14, 42. In the Monumentum Ancyra-
num (p. 36) the Scythians are men-
tioned in conjunction with the Bastarnse
and Sarmatte.

1 66 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

that came to do him honour was one from " the Indians," sent
by a king named Porus, who professed to be lord over six
hundred (!) other kings. They brought, among other gifts, a
gigantic serpent, and some real tigers, which were said to be
the first ever seen by the Komans.^ One of the deputies also
astonished the Greeks at Athens by burning himself alive on a
funeral pile, in the same manner as Calanus had done when
returning with Alexander from India. We have no clue to the
part of India from whence the embassy was sent, or the real
name of the prince who sent it — that of Porus being evidently
only a Greek appellation : but it was probably the reduction of
Egypt under the Koman authority that had made the Indians
acquainted with the fame of the new ruler of so large a portion
oftheEast.^

§ 18. But although the Eomans had acquired no real geo-
graphical knowledge of the far East in addition to that long
possessed by the Greeks, they had yet heard the name at least,
though in a very vague manner, of a people who gradually
assumed an important place in geography — the Seres. There
can be no doubt that this was owing to the extension of com-
mercial relations, which had been the means of making them
acquainted with silk, an article ever after in great request with
Koman ladies, and which gradually became one of' the most
important objects of trade. But its real nature, as well as the
country from whence it came, were still equally unknown. The
well-known line of Yirgil —

Velleraque ut foliis depectant tenuia Seres ^

* Dion Cass. liv. 9 ; Strabo, xv. p. I had beea sent to him from Indian

720. See Note C, p. 201. monarchs. "Ad me ex India regum

^ Mr. Merivale seems inclined to legationes siepe missse sunt, nunquam

doubt the reality of this embassy, but antea visas apud quemquam Roman-

I can see no ground for such a suspicion.
Nicolaus of Damascus, who is referred
to by Strabo, had himself seen and con-
versed with the envoys at Antioch, and
is a trustworthy authority.

It would appear, indeed, as is not
improbable, that this embassy was fol-
lowed by others ; as we find Augustus
himself boasthig that repeated missions

orum prinoipem" (Monum. Ancyran.
p. 36, ed. Zumpt.).

' Georg. ii. 121. It is remarkable
that this line is the earliest notice of
the Seres, and the jjroduction of silk,
that is found in any ancient writer,
though the name is here introduced as
one that would be familiar to his
readers. Horace also uses the term

Sect. 1.

EOMAN EMPIRE UNDER AUGUSTUS.

167

represents the popular notion, which continued prevalent in
the time of Pliny, that it was stripped from the leaves of trees.
The Seres from whom it came were probably regarded as an
Indian nation by those at least who troubled themselves to
have any definite idea upon the subject. But whether it was
brought by sea to Alexandria, or overland through the
Parthian dominions, we have no information.® The latter
hypothesis is, however, the most probable.

§ 19. With Arabia the Eomans had as yet had very little
intercourse. Even the district immediately south of Palestine,
commonly known as Arabia Petrsea, which was subsequently
annexed as a Eoman province, was at this period still subject
to a native prince. It was occupied by a tribe named by the
Greeks and Eomans Nabateei, and their capital city of Petra was
already a place of considerable trade, the resort of numerous
merchants, both Eomans and Greeks,® but its great commercial
prosperity belongs, like that of Palmyra, to a later period.
The expedition made by ^lius Gallus in the reign of Augustus,
with the view of reaching the fertile districts of Arabia Felix,
deserves a separate notice.

Egypt passed under the Eoman dominion, and was con-
verted into a Eoman province without any change in its
boundaries. These are indeed so strongly marked by nature
as to be hardly susceptible of alteration. Towards the south
alone the limit between Egypt and Ethiopia might admit of
some doubt ; but here also the Eomans acquiesced in the esta-

" Sericse sagittse " (Carm. i. 29, 9), and
repeatedly alludes to the Seres as one
of tlie barbarian races hanging on the
skirts of the Eoman Empire [Carm. i.
12, 56 ; iii. 29, 27; iv. 15, 23); but all
these notices evidently refer to the
nomad tribes of Central Asia. It is
singular that Horace has no mention of
silk.

Strabo alludes to the textures called
Serica, which he describes as carded off
the bark of certain trees (^roiavTa Se
Kal TO, '2,'iipiKa, e(c t'ivojv <p\oicai/ ^ulvo-
/ueV'/js ^vcraov, XV. i. p. G'Jo). The

account of Pliny (vi. 17, § 54) is more
precise, but not moie accurate.

* According to Florus (iv. 12, extr.),
the Seres actuall}' sent an embassy to
Rome at the same time as the Indians,
but this circumstance is not mentioned
by any writer of authority ; and the
silence of Augustus, where he is enu-
merating all similar cases that had
added lustre to his reign (Mon. Ancyr.
I.e.), seems conclusive against its
authenticity.

» Strabo, xvi. 4, § 21.

i68

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

blislied line of demarcation, and Syene became the frontier
fortress of the Eomans, as it had been under the Persian
governors. Ethiopia was still entirely independent, and was
governed by a queen named Candace, whose capital was at
Napata, just below the fourth cataract, but who doubtless
reigned also over the adjoining regions of Meroe.^

§ 20. West of Egypt, the fertile though secluded region of the
Cyrenaica, which still retained to a great degree the prosperity
that it had enjoyed ever since the foundation of the first Greek
colonies in that favoured land, had been long before united
with the Roman Empire. Having been separated from the
Egyptian monarchy after the death of Ptolemy Physcon
in B.C. 117 it constituted a separate kingdom under his son.
Ptolemy Apion, who, at his death in B.C. 96, left it by his will
to the Eoman people.^ It was not however formally reduced
to a province till B.C. 67, when, after the conquest of Crete by
Metellus Oreticus, that island was united for administrative
purposes with Gyrene, and the two together constituted one
province. Strange as this arrangement appears, it continued
unchanged till the time of Constantino. The valley of the
Catabathmus formed the limit between Cyrenaica and Egypt,
while towards the west the province extended to the Altars
of the Philseni,^ which marked the frontier on the side of
Africa.

The Roman province of Africa, as constituted under Augustus,
was one of great extent, and included not only the district
usually known under that name, from the Lesser Syrtis to
Carthage and Utica, but the long tract of coast extending
from the Lesser to the Greater Syrtis, where it met the

^ Strabo, xvii. p. 820. The investi-
gations of Lepsius have clearly esta-
blished the site of Napata, which was
situated at a place called Merawi, a
few miles below the conspicuous iso-
lated mountain called Jebel Barkal, at
the foot of which are extensive re-
mains of temples and pyramids, that
belonged to the Ethiopian capital,

though detached from it (Lepsius,
JBriefe aus Aegypten, p. 240).

- Liv. Epit. Ixx. ; Eutrop. vi. 9.

3 Sallust. B. Jugurth. 19, 79. The
legend from which this spot derived
its name is first told by Sallust, but the
name is already mentioned by Polybius
(iii. 39) as marking the limit of the
Carthaginian dominion.

Sect. 1. KOMAN EMPIRE UNDER AUGUSTUS. 1 69

Cyrenaica at tlie Altars of the Philseni. It thus included not
only the modern territory of Tunis, but great part of that of
Tripoli also.* The whole of this latter district had formed
part of the dominions of Carthage until after the Second Punic
War, when it was wrested from their hands and consigned to
Masinissa, together with Numidia and Gsetulia. But after the
defeat of Jugurtha it was again reunited with the province of
Africa, which thus comprised almost exactly the former Car-
thaginian territory. Towards the south, the desert formed
the natural boundary : but the outlying oasis of Cydamus
(Ghadamis) appears to have acknowledged the Roman supre-
macy, and been regarded as belonging to the province :^ while
even the more remote Garamantes (the people of Fezzan) were
made to feel the force of the Roman arms and compelled to a
nominal submission.^

§ 21. The extensive regions of Numidia and Mauretania
had undergone repeated changes in their territorial division.
There is indeed no natural separation between the two : and
the name of JSTumidians, which is only a corruption of the
Greek JSTomades, though adopted by the Romans and used as
a proper name, could never have been a true ethnic appellation.

At the time of the Second Punic War the Numidians were
divided into the two great tribes of the Massyli and the Mas-
ssesyli. The former occupied the territory from the river
Tusca, which formed the limit of the Roman province of
Africa, to the Ampsaga on the west;^ while the Masssesyli

* The city of Leptis Magna, origi-
nally a Phcenician colony (Sallust,
B. J. 78), was the capital of this part of
the province, and held much the same
prominent position as that of Tripoli
at the present day. The only other
towns in the region of the Syrtes, as it
was sometimes called, were OEa, on the
site of the modern Tripoli, and Sabrata,
the ruins of which are still visible at a
place called Tripoli Vecchio (Barth,
Wunderungen, p. 277). Tlie three
together gave the name of the Tripolis

from the Pentapolis of Cyrenaica.
Hence the modern appellation.

^ Roman inscriptions have been
found at Ghadamis (see Barth, Wande-
rungen, p. 249) ; and it appears to have
been connected by established caravan
routes with Tacape and Leptis.

^ See Tacitus (Annals, iv. 23, 26).
The expedition of Cornelius Balbus
against the Garamantes will be con-
sidered hereafter. (See p. 184.)

^ The Tusca was a small stream,
flowing into the sea at Tabraca, the

of Africa to this region, as distinguished ' site of which still retains the name of

170

HISTORY OF ANCIENT GEOGEAPHY.

Chap, XX.

extended from the latter river to the Mulucha, whicli separated
them from Mauretania. As late as the Jugurthine War this
river still formed the boundary between the kingdoms of
Jugurtha and of Bocchus.^ But in consequence of the part
taken by Juba, the last king of ISTumidia, in the Civil War
between Pompey and Csesar, his kingdom was confiscated by
the victor, and became a Eoman province in B.C. 46. The
historian Sallust was its first governor. But Juba II., the
son of the preceding, having gained the favour of Augustus,
was reinstated by him in his paternal dominions, in the
general settlement of affairs after the death of Antony,
B.C. 30. A few years afterwards however the emperor altered
this arrangement and united the eastern portion of Numidia
with the Koman Empire, while he gave Mauretania in its place
to Juba, who thus ruled over the whole tract from the Atlantic
Ocean to the river Ampsaga. The kingdom thus constituted
naturally took the name of Mauretania, which was thus ex-
tended over the larger part of what had previously been called
Numidia. It was not till the reign of Claudius (a.d. 42) that
Mauretania, in this new acceptation of the term, was incorporated
with the Roman Empire : it was then divided, on account of its
great extent, into the two provinces of Mauretania Tingitana
and Mauretania Csesariensis, which were separated by the river
Mulucha, that had previously formed the boundary between
Mauretania and Numidia.

The extent of Mauretania along the Atlantic coast is not
clearly defined, but it seems to have extended, nominally at
least, as far as the point where the chain of Mount Atlas
descends to the sea,^ in about 30° N. latitude. The southern
slopes of Mount Atlas, and the fertile, date-producing tract
that intervenes between the foot of that range and the great

Tabarkali, It is about 50 miles east
of Bona. The Ampsaga was tbe river
■which flowed by Cirta (Constantina),
and entered the sea west of the con-
spicuous promontory called Tretum,
now the Seven Capes.

8 Sallust, B. Jug. 19, 92, 110.

^ Pliny, however, has no account of
the province beyond Sala, opposite the
modern town of Sallee (in hit. 34°),
and it is probable that there Averc no
towns or settlements farther south.

Sect. 1. EOMAN EMPIKE UNDER AUGUSTUS. I7I

desert of the Sahara, was the native abode of the Gsetulians,
an aboriginal race/ who had never owned more than a pre-
carious and nominal allegiance to the Mauretanian and Numi-
dian kings. This did not hinder their breaking out into
frequent revolts, one of which, under the reign of Juba, was so
serious that he was compelled to call in the aid of the Komans,
and the Eoman general Cornelius Cossus, who was sent against
them, assumed the surname of Gsetulicus in honour of his
victory.^ Unfortunately we have no details concerning this
campaign, which might otherwise have thrown some light
upon the very little known geography of these regions.

Section 2. — Roman writers. — Juba.

§ 1. It would have seemed natural to suppose that while the
Bomans were thus extending their dominion, or carrying their
arms into almost every part of the known world, their atten-
tion would have been strongly attracted towards the study of
geography, and that we should have found numerous writings
upon this subject. But so far from this being the case we find
hardly any Eoman author of note, who had either earned or
deserved any reputation as a geographer. P. Terentius Yaeeo
Ataoinus, a younger contemporary of his more celebrated
namesake, wrote a free translation of the Argonautica of
Apollonius Ehodius, which obtained a considerable reputation
in his day : he was also the author of a poem on geography
and cosmography in general, which is cited by Pliny among

1 Sallust, writing from Punic autho-
rities, distinctly tells us that the Gsetu-
lians and Libyans were the earliest
inhabitants of Africa (JB, Jugurth. c. 18),
a statement that doubtless deserves
more credit than the absurd fables by
which he follows it up, in order to con-
nect them with Hercules.

The relations of the Gsetulians to
the Numidian kings are well illustrated
by the part they played during the

Civil War in Africa (Hirt. B. Afr. 32,
35, &c.), as well as during that with
Jugurtha (Sallust, B. J. 19, 80).

There can be little doubt that they
were the same people whose descend-
ants, under the name of Berbers and
Tuaricks, still occupy the valleys of
Mount Atlas, as well as the oases of the
Great Desert.

2 Dion Cass. Iv. 28 ; Flor. iv. 12.

172

HISTORY OF ANCIENT GEOGEAPHT.

Chap. XX,

the authorities of whidi he had made use for this part of his
work,^ but we know almost nothing about it. Other treatises,
cited under the names of Libri Navales, de Ora Maritima and
Littoralia, or de Littoralibus, are by some ascribed to this
Yarro Atacinus, by others to M. Yarro, the friend of Cicero
and author of so great a variety of works, that they may well
have comprised some of a geographical character : but nothing
more is known concerning them.* Coenelius Nepos also, who
is frequently quoted by Pliny among his authorities for the
geographical portion of his work, must have given consider-
able attention to geography, though we do not learn that he
composed any treatise especially devoted to that subject.
Among the statements for which he is cited as responsible
is the strange story of the Indian navigators, who had been
carried all round the north of Asia and Europe till they found
themselves on the coast of Grermany, and were sent by a king
of the Suevi to Metellus Celer, who was at that time proconsul
of Gaul.^ He stated also that " a certain Eudoxus, sailing from
Egypt in order to escape from king Ptolemy Lathyrus, and
setting out from the Red Sea, had effected the navigation to
Gades." ^ This entire perversion of a case, the real facts of
which are known to us from Posidonius, is a sufficient proof
how little value can be attached to these random stories. '^
But if such tales do little credit to the judgement of Cornelius
Nepos, it must be added that he appears to have bestowed
considerable pains on the collection of measurements and
estimation of distances, for which he is frequently cited by
Pliny. ^ Unfortunately we have no means of judging of the

^ Plin. H. N. lib. i. in the Catalogue
of authorities for books iii. iv. and v.

■• On this subject,, see the article
Varro in Dr. Smith's Diet, of Biogr.
vol. iii. p. 1227.

* Plin. E. N. ii. 67, § 170. The same
tale is more fully told by Mela, on the
same authority, and he adds " Cornelius
Nepos ut recentior, auctoritate sic
certior " (iii. 5, § 45).

•* " Ncpos Cornelius auctor est,

Eudoxum quendam sua ajtate, cum
Lathurum regem fugeret, Arabico sinu
egressum Gades usque pervtctum."
Plin. I. c. § 169.

' Pliny himself in one passage cen-
sures him for the greediness with which
he caught up idle talcs (quajque alia
Cornelius Nepos avidissime credidit, v.
1, § 4).

8 if. A^. iii. §4; iv. 12, § 77.

Sect. 2. EOMAN WRITERS. 1 73

sources from wliicb. his conclusions were derived, where they
differed from those of earlier writers.

§ 2. The historian Sallust (C. Sallnstius Crispus) who, as we
have seen, was the first governor of Numidia after it was
reduced to a Koman province, took advantage of the oppor-
tunity to collect information, geographical as well as historical,
concerning the adjoining regions, which enabled him to give,
in his history of the Jugurthine war, an interesting sum-
mary of the geography of that part of Africa. But he tells
us himself that he could obtain no satisfactory information
concerning the nations of the interior.^ He learnt only that
above the Numidians {i. e. farther inland) were the Gsetulians,
part of whom were a settled race and dwelt in huts, others
were uncivilized and without any fixed habitations : beyond
them were the Ethiopians, and after that desert regions
parched up by the burning sun.^ From the citations of later
authors he appears to have introduced similar geographical
notices in the lost books of his Histories, but nothing has been
preserved to us beyond a few isolated fragments. It appears
that he described the Tigris and Euphrates as flowing from
the same source, in the mountains of Armenia.^

§ 3. Another writer, apparently a contemporary of Sallust,
but known to us only from the citations of Pliny, was Statius
Sebosus, who appears to have made careful inquiries con-
cerning the Oceanic coast of Africa, and the islands which lay
off it. According to the information he was able to collect,
the first of these was Junonia, situated 750 miles from Gades,
to the west of which were Pluvialia and Capraria, and 250 miles
farther, lying off the coast of Mauretania towards the south-
west, the Fortunatse, which bore the names of Convallis (or
Invallis according to some MSS.) and Planaria.^ It is im-
possible to identify these with certainty, but it is clear that

» San. R. Jug. 17.

1 Ibid. 19.

2 Sail. Hid. iv. Fragra. 194, ed.
Gerlacli. But it may well be doubted

whether the words " uno fonte " are to
be taken in a literal sense.
^ Plin. E. N. vi. 32, § 202.

174

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XX.

some of them at all events must be referred to the group of
the Canary Islands, to which the name of the Fortunatae
Insulse was afterwards generally applied. Those previously
known to Sertorius, as we have already seen, appear rather to
have been the two islands of Madeira and Porto Santo.*

§ 4. A much more important contribution to geographical
knowledge was the work of the younger Juba, the second of
the two Numidian monarchs of the name. Having been
carried to Eome by the dictator Caesar as a mere child, after
the death of his father in B.C. 46, he received his education
there, and devoted himself to study with such success that he
soon came to be esteemed one of the most learned men of his
day.^ He became also the intimate friend of the young
Octavian, who ultimately restored him to his father's kingdom,
which he subsequently exchanged (as already mentioned) for
the more extensive dominion of Mauretania. His principal
work was a history or description of Africa,^ a task for which his
position gave him unparalleled advantages. It is frequently
cited by Pliny, and appears indeed to have been one of the
chief authorities upon which that author relied, both for the
geographical description of the country, and for the natural
history of the wild beasts with which Africa abounded. But
as it is only in a few instances that Pliny directly cites his
authority, we cannot justly estimate how much Juba really added
to what was previously known. Of the interior of Africa in-
deed he appears to have had little more knowledge than hi§
predecessors. But he was certainly the first author of the
extraordinary theory concerning the origin of the Nile, which
was adopted and received by several ancient authors. Accord-

■* other statements of Sebosus con-
cerning the Gorgacles or Islands of the
Gorgons — those which were visited by
Hanno — and the islands of the Hespe-
rides, which he placed at 40 days'
voyage (!) beyond them, are treated by
Pliny himself as very vague and uncer-
tain (Plin. H. N. vi. .SI. § 201).

* He is called by Avienus (fZe Ora
MariUma, v. 279),

Octaviano principi acceptisslmus
Et litterarum semper in studio Juba.

Plutarch also terms him 6 irdvTwv laro-
piKWTaTOS fiaffiXeaiv (Vit. Sertor. 9), and
Athenseus (iii. p. 83 b.) dW;p -rroAvfia-
Oeararos.

« Ai^vKo,. Pint. Parallel. Mhwr. 23 ;
Athenjeus (/. c.) cites it more generally
as TO. Trepl Ai^v-qs avyypa.^iJ.aTa.

Sect. 2. JUBA. 1 75

ing to this account, which was given as the result of his special
inquiries, the Nile had its source in a mountain of Lower
{i. e. western) Mauretania not far from the Ocean : it almost
immediately formed a lake, which contained crocodiles, and
fish of the same species as were found in Egypt.'' On issuing
from this lake it buried itself for some days' journey, and flowed
underground through a sandy and desert tract, till it reap-
peared in Mauretania Csesariensis,® where it formed another
and larger lake, distinguished by the same animals. It then
lost itself again in the sands and pursued an underground
course for twenty days' journey to the nearest Ethiopians, when
it broke out again by the source which was called Niger.
Thenceforth it separated Africa from Ethiopia, and its banks
became fertile, covered with forests, and abounding in wild
beasts ; it then cut through the midst of the Ethiopians and
assuming the name of Astapus, flowed onwards past Meroe and
other islands, into Egypt.^

§ 5. Juba also made diligent inquiries concerning the
Fortunate Islands, which, according to the information he
obtained, were five in number, which he named Ombrios,
Junonia, Capraria, Nivaria and Canaria, all of them obviously
names given by voyagers or geographers, and all but one of
Latin origin. No mention is made of their being inhabited,
but they abounded in all kinds of fruits and birds, and some
traces of buildings were found. Large dogs also were said to
be found in great numbers on the island, called from that
circumstance Canaria,^ two of which were brought to the king.
Date-palms also abounded on this island, though not in any of

' He mentioned by name " alabetse,
coracini, et siluri," of which the first
name is not found elsewhere in Pliny's
work ; the coracini and siluri were both
abundant in the Nile, and the former
even said to be peculiar to it. Plin.
n. N. ix. §§ 44, 68; xxxii. § 56. The
preciseness of this statement is very
remarkable, whatever we may think of
tlie inference drawn from it.

* This must of course refer to the
inland portion of the province, south of
the range of Atlas : though the name
was usually applied only to the region
adjoining the Mediterranean.

9 See Note D, p. 201.

^ This is obviously one of the etymo-
logical fancies, or fictions, so common
among both Greek and Eoman writers.

1/6

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XX.

the others. Besides these Jiiba was the discoverer (if we may
construe the expression of Pliny literally) of some islands off
the coast of Mauretania, where he established a factory of
purple, from which circumstance they derived the name of
Purpurarise.^

Besides his work upon Africa, Juba was author of a treatise
on the geography of Arabia, which was also regarded by Pliny
as one of the most trustworthy accounts of that country, and
was the authority that he principally followed in describing it.
It was dedicated to Caius Caesar (the grandson of Augustus)
when he was about to proceed on his expedition to the East
(b.c. 1).^ It is remarkable that neither this work, nor that
on Africa, is ever alluded to by Strabo, who appears to have
been wholly unacquainted with the writings of Juba.*

§ 6. But if the Eoman writers contributed but little to the
progress of geography, so far as related to the knowledge of
distant countries and nations, the mere extension of the
Eoman Empire, and the gradual introduction of their im-
proved system of organization and administration into all
parts of their dominions must have added greatly to the
accuracy and completeness of the knowledge already possessed
of the vast portion of the world which was subject to their
sway. The construction of roads was one of the special objects
of attention with all Eoman governors : these were care-
fully measured and marked with milestones : itineraries of the
distances along them were preserved and carefully recorded,
and they thus obtained a means of geographical measurement,
defective indeed according to the requirements of modern
science, but still far superior to anything previously possessed.^
The Itineraries that have been transmitted to us are of much

2 Plin. H. N. vi. 31, § 201. See Note
E, p. 202.

3 Id. xii. 31, § 56.

'' This may perhaps have arisen from
the work of Juba having been com-
posed in Latin. Dr. C. Miiller indeed
supposes Juba to have written in Greek,
and has included his remains amonp;

his Fragmenta Historicorum Graecorum
(vol. iii.), but I am not aware of any
proof of this, and considering the cir-
cumstances of his life, it would appear
much more probable that he should
have written in Latin.

^ See the use made by Strabo (vii. 7,
p. 322) of the Egnatian Way.

Sect. 2. AGllIPPA. 1 7/

later date, but it cannot be doubted that similar compilations
existed from the earliest periods of the Eoman Empire.

§ 7. M. Ageippa, the friend and supporter of Augustus, who
was a man of first-rate administrative talent, and had, as we
have already seen, been the first to introduce a system of high
roads into Gaul," was particularly active in promoting this
branch of geographical knowledge. He caused a map of the
whole world, as then known — the Eoman Empire and the
adjacent countries — to be set up in the portico of Octavia at
Rome, and accompanied it with a detailed commentary, stating
the distances from one important point to another, and the
length and breadth of the different provinces.'^ These were
doubtless derived from the itineraries, wherever such existed ;
and Pliny speaks in the highest terms of the diligence and
care which he bestowed upon this compilation, which he
quotes as the highest authority in all cases where he was
likely to have had official information.^ Agrippa, however,
did not confine himself to such limits : he not only gave the
distances in countries, such as the northern shores of the
Euxine, which were not indeed subject to the Roman govern-
ment, but were still familiar and accessible; but he added
those concerning countries of which he had nothing but the
vaguest knowledge. Thus he stated the distance from the
mouth of the Danube to the Northern Ocean at 1000 Roman
miles, and the breadth of the tract between the Scythians and
Germans, from the desert of Sarmatia to the river Vistula at
400»miles.^ He even ventured to estimate the length of the

* See above, p. 142.

' It appears very probable that the
anonymous work repeatedly cited by
Strabo under the title of "the choro-
grapher" (o xwpo7P«<l>os), was either
the commentary thus appended by
Agrippa to his map, or was directly
derived from it.

* Thus with regard to Spain, after
pointing out the discrepancies between
ditferent statements as to the measure-
ment of the provinces, Pliny adds:
" Agrippam quidem in tanta viri dili-

gentia, prseterque in hoc opere cura,
cum orbem terrarum orbi spectandum
propositurus esset, errasse quis credat,
et cum eo Divum Augustum ? Is
namque complexam earn porticum ex
destinatione et cum conimentariis M.
Agrippse a sorore sua inchoatam
peregit " (H.N. iii. 2, s. 3, § 17).

9 " Agrippa totum eum tractum ab
Istro ad Oceanum bis ad decies centena
M. pass, in longitudinem, quadringentis
in latitudinem ad flumen Vistulam a
desertis Sarmatise prodidit" (Plin. H.N.

VOL. II. N

178

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

nortliern coast of Europe along the sliores of the Ocean, but
this Pliny himself pronounces to be altogether vague and
uncertain •} and Agrippa himself, in reporting the supposed
length and breadth of Britain and Ireland, appears to have
done so with an expression of doubt. ^

The map of Agrippa was evidently painted upon the wall of
the portico. That this was not an uncommon practice, we learn
from an incidental statement of Varro, that he found his friend
Fundanius and others contemplating a map of Italy painted
on the wall of the temple of Tellus.^ Maps also, though of
course on a less extensive scale, were painted on boards, and
geography was thus made a matter of j)opular instruction.*

The construction of this map and the accompanying com-
mentaries may probably have given rise to the popular notion,
which we find in later writers, that Julius Csesar had ordered
a general measurement of the world to be made, which was
completed under the reign of Augustus.^ This notion also
was undoubtedly connected with the supposed decree "that
all the world should be taxed," and with the real fact that
a census of the population of the Empire was ordered and
accomplished by Augustus f a measure which must have been

iv. 12, s. 25, § 81). The numbers are
differently given in the MSS., and
admit of much doubt. This passage
also deserves remark as the fiist in
-which the name of the Vistula occurs.

> Plin. H. N. iv. 14, s. 28, § 98.

- Id. ih. 16, s. 30, § 102. " Agrippa
longitudinem dccc m. pass, esse :
latitudinem ccc m. credit. Eandem
Hibernise latitudinem, sed longitudinem
cc M. passuum minorem." It is re-
markable that the length and breadth
of Britain thus given are really very
fair approximations : those of Ireland
are greatly in excess, but the know-
ledge of that island possessed by the
Eomans was always very imperfect.

^ " Offendi ibi C. Fundanium socerum
meum,et C. Agrium equitemEomanum,
et P. Agrarium publicanum, spectautes
in pariete pictam Italiam." Varro de
Re Bust i. 2. Tlie custom of painting

such maps upon walls was continued
in Italy down to modern times ; and
they may still be seen on the walls of
several palaces.

*• Thus Propertius :

Conor et e tabula pictos edlscere mundos.

V. 3, V. 37.

^ See the detailed account of this
pretended measurement in the Proce-
mium to the Treatise on Cosmography
ascribed to Julius Aethicus, appended
by Gronovius to his first edition of
Pomponius Mela, Lugd. Bat. 1722. The
question will be further considered in
a future chapter.

" Concerning this census, see Mar-
quardt, Handbuch der Bomische Alter-
tliumer, vol. iii. p. 56 ; Pluschke, ilher
den Census zur Zeit der Geburt Ghristi,
p. 13, and Iloeck, Bom. Qesoli. vol. ii.
pp. 392-426.

Sect. 3. MILITARY EXPEDITIONS. 1/9

the means of accumulating a vast mass of geographical as well
as statistical information. The care bestowed by the Romans
upon the administration of their provinces must have been
constantly tending in the same direction, and though the
means of anything like a mathematical survey were still
almost wholly wanting, the political geography of the Empire
was undoubtedly making continual progress.

Section 3. — Military Exjpeditions.

§ 1. Whatever positive additions were made to geographical
knowledge under the reign of Augustus were the result of
military operations and expeditions into the countries bordering
on the Roman Empire. One of the most important of these,
though in great measure baulked of its intended object, was
the expedition of ^lius G-allus into Arabia, a full account of
which has been preserved to us by Strabo.'' Gallus was prsefect
of Egypt under Augustus, and his enterprise was undertaken
at the command of the emperor himself, who was not only
stimulated by curiosity, but tempted by the accounts that he
had heard of the great wealth of the southern Arabians. He
was encouraged also by the promises of the Nabataean Arabs,
who were at this time, as we have already seen, in a state of
semi-dependence upon Rome ; but their chief Syllaeus, upon
whose guidance Gallus mainly relied, proved faithless, and the
failure of the expedition was in great part attributable to his
treachery.

The Roman general began by transporting his whole army,
amounting to ten thousand men, including contingents fur-

7 Strabo, xvi. 4, §§ 22-24, pp. 780-
782. Strabo himself had been on inti-
mate terms Tvith ^Elius Gallus, who
was prsefect of Egypt at the time that
he visited the country, and had made
tlie ascent of the Nile in his company
(xvii. p. 816). Hence his narrative
possesses special claims to our con-

sideration.

Julius Gallus must not be confounded
with Cornelius Gallus, who was tlie
first governor of Egypt appointed by
Augustus (Strab. ih. p. 819). The
date of the expedition in question is
assigned by Dion Cassius to the year
24 B.C. (Dion Cass. liii. 29).

N 2

l80 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

nished by the Jews and Nabataeans, by sea from Cleopatris in
Egypt (at the head of the Gulf of Suez), down the Eed Sea to
a port called Leuce Come, which was a place of importance,
carrying on a considerable trade with Petra by caravans.
Here he was compelled to halt the whole summer and the
following winter, in order to recruit his troops, which were
suffering severely from an epidemic arising from the hardships
and discomforts encountered by them on their passage. This
had lasted fifteen days, and he had lost many of his ships
upon the rocks and shoals with which this part of the Eed
Sea abounds. When at length he was able to resume his
march, he advanced for many days through a country extremely
deficient in water to the territory of Aretas, a kinsman of
Obodas the king of the Nabatseans, by whom he was received
in a friendly manner; and furnished to some extent with
supplies. After traversing this comparatively fertile district
for 30 days, he entered upon a tract which was completely
desert, and inhabited only by wandering Arabs ; this was
called Ararene, and occupied him not less than 50 days till
he arrived at a city called Negrana, in a fertile and settled
region. This he took without difficulty, but after six days'
march from thence he was met by the barbarians, who encoun-
tered him at the passage of a river, but were defeated with
great slaughter, while the Eomans lost only two men. After
this he took a city named Asca, and another named AthruUa,
from whence he proceeded to attack a city called Marsiaba,
belonging to a tribe named the Ehammanitas, who were subject
to a king of the name of Ilasarus. But being frustrated in
his attack, and suffering from want of water, he determined to
retreat, though he was assured by captives that he was within
two days' march of the Land of Spices.^ On his return march
he found out how much he had been misled by his faithless
guide, and accomplished with little difficulty in sixty days

* Svh /ifv ovv Tjfiepooy oShy &iTf<Txe Trjs apufxaTO(p6pov, KadaTrep tmv aixP-nXcoroov
aKoveiv ^v. Strabo, p. 782,

Sect. 8.

MILITARY EXPEDITIONS.

I8l

the same distance that had occupied him six months on his
advance; returning in the first instance to Negrana, nine
days' march ; thence in eleven days to a place called the
Seven Wells (Hepta Phreata), and thence through a peaceful
country by two villages called Chaalla and Malotha to a place
named Egra, situated on the sea-coast, and subject to the king
of the Nabatseans, where he was able to embark his troops and
transport them by sea to Myos Hormus.

§ 2. Detailed as is this account in comparison with many
similar notices,^ it is almost impossible to extract from it any
definite geographical information. Even the point from whence
Gallus set out on his march, Leuce Come, though described as
an important emporium of trade — a position which we find it
still occupying near a century later ^ — has been much disputed,
but the probabilities certainly preponderate in favour of a
place called Howara, situated in exactly 25° of N. latitude,
and distant about 200 miles from Kosseir ; Egra, the place to
which he returned, must apparently have been not far from
Leuce Come, as the manner in which Strabo contrasts the
time occupied on his advance and his return implies that he
must have come back (approximately at least) to the same
point from which he set out.^ Hence we may infer that the
farthest point reached, Marsiaba, could not have been much

* Thus Pliuy, tliougli dwelling upon
tlie importance of the expedition of
Gallus as the only one which had
penetrated into the interior of Arabia
(Romana arma solus in earn terrain
adhuc intulit jSilius Gallus ex equestri
ordine), contents himself with giving
us the names of the towns which he
destroyed, without any indication of
their geographical sequence or position,
except that Caripeta (a name not found
in Strabo) was the farthest point to
which he penetrated (Plin. if. N. vi.
28, § 160). Dion Cassius, who also
gives a brief history of the campaign
(liii. 29), mentions uo name except
Athloula {"AdXovAa), evidently the same
with the Athrulla of Strabo, but which
he regards as the farthest point at-

tained.

* See 'the Periplus Maris Erythrsei,
§19.

^ Nor could he have found the means
of transport across the gulf, till he re-
turned to the friendly country of the
Nabatseans. But the position of Egra
is as uncertain as that of Leuce Come.
Ptolemy indeed places a town of the
name in lat. 26° ; but little reliance
can be placed oq his latitudes, and as
he does not mention Leuce Come at all,
he gives no clue to their relative posi-
tion. Egra must have been situated
to the south of Leuce Come, as Gallus
could never have returned, without
knowing it, to a point beyond that from
which he started.

1 82 HISTORY OF ANCIENT GEOGEAPHY. Chap. XX.

more than 60 days' march — about 700 miles at the utmost —
distant from Leuce Come. This would bring us to the borders
of Yemen, which would accord with the account of their having
reached a comparatively fertile country, and with the state-
ment (on which however very little dependence can be placed)
of their having approached within a few days' march of the
Land of Spices. But Marsiaba itself cannot be determined
with any approach to certainty. The name is probably identi-
cal with Mariaba, but it is very uncertain whether it is the
same with the celebrated city of that name, the capital of the
Sabseans, for it is certain that there were several towns of
the same name. In the present imperfect state of our know-
ledge of the interior of Arabia all further investigation seems
hopeless ; the name of the Rhammanitse is otherwise wholly
unknown, and all such appellations of the subordinate tribes
of Arabs must in all ages have been very fluctuating and
uncertain.^

The accounts brought back by Gallus, as derived from
hearsay information, of the great wealth of the Arabians,
especially the SabEeans, coincided with the notions previously
entertained upon the subject.* But it is remarkable that
both Strabo and Pliny, while dwelling upon the abundance of
spices and precious stones produced in their country, ascribe
the great accumulation of wealth in their hands to the circum-
stance that while they were continually receiving gold and
silver in exchange for their own productions, they spent
nothing in return upon goods imported from other countries.^

§ 3. Immediately connected with the expedition of ^lius
Gallus into Arabia and more successful in its results, was that
of Petronius into Ethiopia. That country, as has been already
mentioned, was at this time governed by a queen named
Candace, said to have been a woman of masculine energy of
character, who took advantage of the Roman forces in Egypt

3 See Note F, p. 204. I = Stmbo. xvi. 4, § 22 ; Pliu. H. N.

* See Chapter XVIII. p. 58. | vi. 28, § 162.

Sect. 3. MILITAIIY EXPEDITIONS. 1 83

being weakened, as she conceived, by the absence of a large
part of them in Arabia, to attack the frontier fortress of Syene,
which she took by surprise, as well as Elephantine and Philse.
The Koman governor, C. Petronius, however, quickly recovered
possession of these towns and followed up his advantage by
penetrating into Ethiopia, where he defeated the array of
Candace in a great battle, and took the city of Pselchis. From
thence he advanced as far as Premnis, called by Strabo a
strong city, which he took, and then proceeded to attack
Napata, the capital of Candace. This he also took, and
destroyed, carrying off the inhabitants into captivity ; but he
did not think it expedient to advance farther, and having
refortified Premnis and left there a Roman garrison with
provisions for two years, he himself returned to Alexandria.
Candace soon after assembled another large force, with which
she attacked the Eoman garrison at Premnis, but Petronius
was able to relieve it, and the Ethiopian queen again sued for
peace. This time the Eoman governor compelled her to send
an embassy to Augustus himself, who was at this time wintering
at Samos ; where her envoys were received with distinction
and admitted to favourable terms, even the tribute previously
imposed being remitted.^

According to Strabo, in advancing from Pselchis to Premnis,
Petronius traversed the sandy desert in which the army of
Cambyses had perished. This is clearly a mistake, for it was
the army sent against the Oasis of Ammon that was lost in the
sands — not that which Cambyses himself led into Ethiopia,
of which Strabo was evidently thinking. But the fact that
Petronius traversed a sandy desert of considerable extent, in
which it was thought possible that such a catastrophe could
have occurred, seems to indicate that he took the road, gene-
rally followed by modern caravans, directly across the desert
from Korosko to Abu Hamed, so as to cut off the great bend of
the Nile. In this case Premnis was probably situated near

« Strabo, xvii. 1, pp. 820-821 ; Dion Cass. liv. 5.

1 84

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

the angle of that river, in the neighbourhood of Abu Hamed,
which is distant about 100 miles from Meraue, the site of
Napata. It is probably the same name with the Primis of
Pliny ^ and Ptolemy ; but the latter author.mentions two places
of the name, which he calls Primis Magna and Parva.^ One
of these may probably be identified with Ibrim, a place
between Pselchis and the Second Cataract, but this can have
nothing to do with the Premnis of Strabo, which appears to
have been not far from Napata, and it is clear that, in order
to reach it Petronius quitted the Nile and traversed a desert
tract. Pselchis is clearly identified with Dakkeh, and it is but
a little distance above that place that the caravan route strikes
off from Korosko. This route would indeed present great
difficulties to the passage of an army ; ^ but not such as can
be pronounced insuperable by troops possessing such hardiness
and powers of endurance as the Romans displayed on several
other occasions. And the site of Napata being clearly esta-
blished, in the immediate neighbourhood of Jebel Barkal, the
road across the desert must have been at this period well
known.^

§ 4. Another expedition that attracted considerable attention
about the same period was that of Cornelius Balbus against
the Garamantes in the interior of Africa. The name of the
Garama'ntes, as we have seen, was already known to Herodo-
tus,^ and was doubtless familiar to the Greeks of Cyrene, who

' Plin. H. N. vi. 29, s. 35, § 181.
That author enumerates the towns
taken by Petronius in the following
order: Pselcis, Primis, Aboccis, Pth-
thuris, Oambusis, Atteva, Stadisis
(where there was a cataract of the
Nile), and finally Napata. He asserts
that Petronius had proceeded, in all,
970 Roman miles from Syene, a great
exaggeration. The towns mentioned
by Pliny would appear, from a com-
parison with Ptolemy, to have been
situated on the west or left bank of the
Nile. But Pliny seems to have followed
quite a different account of this expe-
dition from Strabo.

* TlpifiLS ^ Tlprjfxis jxiKpa, and Tlpifiis
^ UpTifiis ixeyoX-n. Ptol. iv. 7, § 19. He
places the latter two degrees to the
south of the former, and Napata be-
tween the two.

^ See the description of it by Hoskins
(Travels in Ethiopia, pp. 19-32), and
Lepsius {Brief e aus Aegypten, pp. 124-
136).

' It appears that this route was cer-
tainly frequented in ancient times, as
inscriptions in hieroglyphics are found
at the wells. Hoskins, p. 24. Concern-
ing the site of Napata, see above, p. 168.

^ See Chapter VIII. p. 278.

Sect. 3.

MILITARY EXPEDITIONS.

185

held extensive commercial relations with the barbarian tribes
of the interior. But we hear no more of them until the whole
extent of the Carthaginian territory in Africa, including the
coast regions adjoining the two Syrtes, had passed under the
power of the Eomans. It was then that Balbus, being ap-
pointed governor of the province of Africa, determined (in
B.C. 20) to carry his arms against these independent tribes of
the interior, over whom he obtained sufficient successes to
entitle him to claim a triumph ; ^ but we have very little
information concerning the extent or character of his expedi-
tion. We learn only from Pliny that he took the towns of
Cydamus (Ghadamis) and Garama, which is termed by Pliny
" clarissimum oppidum," and was evidently at that period the
capital of Fezzan, as Mourzuk is at the present day.* There
is therefore no doubt that he actually penetrated as far as
Fezzan; the name of which, Phazania, became henceforth
known to the Eoman geographers ; but of course no permanent
conquest of these outlying regions was attempted.^ Com-
mercial intercourse was however continued, though with diffi-
culty, the wandering tribes obstructing it by filling up the
wells with sand. The route taken by Balbus was apparently
from the neighbourhood of the Lesser Syrtis (the Gulf of
Cabes), through Ghadamis to Fezzan ; it was not till the reign
of Vespasian that the direct route from (Ea (Tripoli) was dis-
covered and found to be shorter by four days' journey.^ A
long list is given by Pliny of the names of places which
figured in the triumph of Balbus, but these were of course
mere villages, and none of them can be identified, except the

3 Plin. fi". N. V. 5, § 36 ; Fast. Capit,
This triumph attracted especial atten-
tion as being the first ever celebrated
by one who was not a native Eoman
citizen, Balbus being a native of Gades
in Spain, from which circumstance he
derived the surname of Gaditanus
(Plin. Z. c. ; Veil. Pat. ii. 51 ; Strabo,
iii. p. 169).

* It still retains the name of Germa,
or Germ, and is about 70 miles to the

N.W. of Mourzuk.

* During the insurrection of Tacfa-
rinas, in Numidia, in the reign of
Tiberius, the Gara mantes appear as an.
independent people, who for a time
supported the rebel chief, and after his
death sent envoys to Eome to sue fpr
pardon. They attracted attention as a
people rarely seen ("Garamantumlegati,
raro in urbe visi." Tac. Ann. iv. 23).

« Plin. I. c.

HISTOKY OF ANCIENT GEOGEAPHT.

Chap. XX.

two already mentioned. It is remarkable that Strabo, though
he incidentally alludes to the triumph of Balbus, seems to
have had no detailed information concerning his expedition,
and his knowledge of the Garamantes was very vague and
imperfect, even the names of Cydamus and Garama being
apparently unknown to him. Nor is any notice of the campaign
found in Dion Cassius. But the excitement caused at Rome
by the triumph over a people in so remote a situation is suffi-
ciently shown by the well-known introduction of their name
in Virgil,^ among the conquests destined to adorn the reign
of Augustus.

Section 4. — Wars in Germany.

§ 1. It was not till a later period of the reign of Augustus
that the Eoman arms became the means of adding largely to the
knowledge previously possessed of Germany and the north of
Europe. We have already seen how imperfect was the geo-
graphical information of Csesar concerning Germany. Though
he had obtained correctly the names of several of the chief
tribes into which the nation was at that time divided : — the
Suevi, Marcomanni, Cherusci and Sigambri — as well as the
more neighbouring tribes of the Ubii,® the Usipetes and Tenc-
teri, all of which immediately adjoined the Rhine, and he was
acquainted in a vague way with the vast extent of the forests
which stretched into the interior of the country, he does not
mention the names of any of the great rivers which form so
important a feature in the geography of northern Germany,

' super et Garamantas et Indos

Proferet imperium.

Mn. vi. 795.

If these lines refer, as there is every
probability, to tlie exploits of Balbus,
they must have beeu written in the
last year of the poet's life, as the
triumph of Balbus took place in the
spring of B.C. 19, and Virgil died in the
autumn of the same year.

* The Ubiaus, who, in the time of

Csesar, were on the right bank of the
Khine, opposite to the Trcviri (iv. 18),
had been allowed by Agrippa to cross
the river, and establish tliemselves ou
its western bank (Strab. iv. p. 194).
Their chief town (civitas Ubiorum,
Tacit. Annal. i. 37) subsequently re-
ceived a Eoman colony, and became
the celebrated Colonia Agrippiua,
which still retains the name of Cologne.

Sect. 4. WAES IN GERMANY. 1 8/

and assume a prominent part in the subsequent campaigns of
the Eoman generals.

The Roman governors of Gaul were for the most part content
with maintaining the frontiers of the province, and defending
the line of the Rhine against the incursions of their German
neighbours. Drusus, the step-son of Augustus, was the first
(after the dictator Csesar) who crossed the Rhine and carried
his arms into the enemy's -country (b.c. 12). Setting out from
the Island of the Batavi — as the Romans called the island
formed at that time by the confluent streams of the Rhine, the
Waal and the Meuse,^ — he crossed the Rhine into the territory
of the Usipetes, and from thence into that of the Sigambri,
both of which he laid waste to a considerable extent. Returning
thence to the Island, where he had assembled a considerable
fleet, he sailed by an artificial channel or canal, which had
been dug under his directions, into the great lake that at that
time occupied a part of what is now the Zuyder Zee, and from
thence by an arm of the Rhine into the Ocean.^ Here he con-
tinued his voyage along the coast of the North Sea, as far as
the mouth of the Ems (Amisia), effecting the submission of the
Frisians, whose name appears for the first time on this occasion.
They seem to have received him in a friendly manner, and
even rendered him service as allies. It was otherwise with the
Chauci, who dwelt on the east bank of the Ems, and whom he
consequently attacked, but with heavy loss, owing in part to
the imperfect knowledge possessed by the Romans of the tides,
a frequent source of disaster to them in these northern seas.
The approach of winter compelled him to return, but he had

9 It is first mentioned under this
name by Caesar (i?. G. iv. 10).

1 The course pursued by Drusus on
this occasion is best known to us by
that pursued by his son Germanicus,
who is said to have followed precisely
in hia father's footsteps (Tacit. Annal.
ii. 6, 8). The topography of these
countries is extremely obscure, on ac-

the formation of the Zuyder Zee, which
did not take place till the 13th century.
But it would appear that there were
several lakes, the largest of them bear-
ing the name of Flevo, or Flevus, which
had a navigable outlet into the North
Sea. They were fed probably by the
Yssel, and the canal dug by Drusus
may have opened the communication

count of the physical changes that have between the Ehine and the Yssel. But
subsequently occurred, and especially the whole subject is very difficult.

HISTORY OF ANCIENT GEOGRAPHY.

CHAr. XX.

earned the glory of being the first Eoman general who had
ever navigated the Northern Ocean.^

§ 2. The next year (b.c, 11) he returned to the customary land-
warfare ; and again crossing the Ehine into the territory of the
Usipetes and Tencteri, threw a bridge over the river Lupia
(Lippe), by which he advanced into the country of the Sigam-
bri, and thence into that of the Cheriisci, which he traversed
successfully till he reached the banks of the Visurgis (Weser).
Here he was checked by want of provisions and bad weather,
and determined to retreat, which he accomplished with some
difficulty, leaving however a fortified post with a garrison at a
place called Aliso, on the banks of the Lippe, and another in
the land of the Chatti, on the right bank of the Ehine.^

In his third campaign (b.c. 9) Drusus began his invasion by
the land of the Chatti, where his newly erected fortress secured
his base of operations, and passing through them to the Suevi,
he forced his way, not without continual combats, through their
territory also into that of the Cherusci, which he again tra-
versed as far as the Weser. This time he crossed that river
also, and pressed onwards as far as the Albis (the Elbe), where
he was met (it is said) by a vision, which compelled him to
retreat, after erecting a trophy on the banks of the river to
mark the extreme term of his advance.^ The portent was
fulfilled by his untimely death, the result of a fall from his
horse. But young as he was — he died in his thirtieth year —
Drusus had marked his government not only by these daring

"^ " Oceanum septentrionalem primus
Komanorum ducum navigavit." Sueton.
Claud, i.

The only intelligible account of
these campaigns of Drusus is contained
in Dion Cassius (liv. 32, 33), but is un-
fortunately very concise. They were
described by Livy in the three last
books of his history (which ended with
the death of Drusus), and the loss of
these is much to be regretted. The
few notices found in Florus (iv. 12) and
Orosius (vi. 21) are doubtless derived
from that source, but they are so rhe-

torical and unconnected as to be utterly
worthless.

^ Dion Cass. liv. 33.

^ These Tropsea Drusi are again
mentioned by Ptolemy (ii. 11, § 28),
but there is no clue to their situation.
Strabo (vii. p. 291) speaks of Drusus as
carrying on the war, just before his
death, between the Ehine and the
Saale (Salas), which would point to his
having reached the Elbe a long way
from its mouth ; but this seems at
variance with all else that we know of
his operations.

Sect. 4. WAES IN GERMANY. 1 89

expeditions into the heart of the enemy's country, but by
measures of a more permanent character — such as the con-
struction of the canal from the Rhine to the Lake Flevus,
which always continued to bear the name of the Fossa Dru-
siana, and was an important means of military communication.^
He erected also a chain of forts along the line of the Rhine,
most of which by degrees grew into flourishing towns,^ and
threw a permanent bridge over that river at Bonn.^ He may
be considered therefore as in great measure the founder of
that Roman civilization on the Rhine, which gradually attained
to so remarkable a development.

§ 3. The next year (b.c. 8) Tiberius succeeded his brother on
the Rhine, and in his turn crossed that river, but was met in a
peaceful spirit, and the neighbouring tribes of G-ermans, in-
cluding even the fierce Sigambri, submitted to give hostages
for their good behaviour. Some years later (a.d. 4) Tiberius
resumed the offensive, and after reducing to submission the
Cherusci, as well as some less important tribes, crossed the
Weser and pressed forward to the banks of the Elbe.

The next year (a.d. 5) was marked by more important ope-
rations, Tiberius himself having for the second time advanced
at the head of his army to the Elbe, while his fleet sailed
round to the mouth of the same river, and ascended it far
enough to form a junction with the land forces. But this
combined operation was not attended with any really im-
portant results ; ^ beyond the submission of the Chauci, who

* Tacit. Annul, ii. 8 ; Sueton. Claud, i.
« Florus, iv. 12, § 26.

' This appears to be the meaning of
Florus (I. c), though the passage, as it
stands in the MSS., is certainly corrupt.

* This is the distinct statement of
Dion Cassius (Iv. 28) concerning both
these campaigns of Tiberius. Kod iJ-expt
ye rod iroTa/jLOv, irpSTepov faev rod Ovl-
(Tovpyov, jxera Se rovro koI tov 'A\piov,
TrpoexaJpTjo'ej'' oi) fj-evroi Kol a^LOiJi.V7iiJ.6vev-
r6v TL tJtc y" iirpaxB-n. There is obvi-
ously great exaggeration in the account
of them given by Velleius Paterculus

(ii. 104-107), unfortunately the only
author from whom we derive any
details. His extravagant flattery of
Tiberius was combined in this instance
with the desire to extol the importance
of military exploits in which he had
himself taken part.

It is much more strange that Dean
Merivale should not only adopt the
exaggerated views of Velleius, but
should actually speak of this second
campaign of Tiberius — which was
marked by no decisive action, and pro-
duced no lasting result — as " the most

190 HISTOEY OF ANCIENT GEOGEAPHY. Chap. XX.

at this period dwelt between the Weser and the Elbe; and a
defeat of the Langobardi, a nation whose name here occurs for
the first time in history. They appear to have been at this
period settled on the left bank of the Elbe, though we soon
afterwards find them established beyond that river.*

It was undoubtedly the voyage of the Eoman fleet that on
this occasion attracted the most attention; and it would be
interesting to know what it really accomplished. The navi-
gation of the Northern Ocean was at that time regarded as in
itself so remarkable an event that even the voyage to the
mouth of the Elbe would be considered a great exploit, and
would easily become the subject of much exaggeration. If
indeed we could trust to Pliny, it would seem that the fleet
had advanced northwards as far as the Cimbrian Promontory —
the extreme northern point of Jutland ^ — but this seems highly
improbable. Such a voyage, more than double the length of
that already accomplished, through seas wholly unknown, and
without any adequate motive, — all possibility of combination
with the land forces being here out of the question — would be
wholly at variance with the ordinary practice of Roman com-
manders. Nor could it have failed to attract general notice ;
but we find no other mention of it. It is not even alluded
to by Velleius Paterculus, who has shown every disposition to
magnify and exaggerate all that took place under the com-
mand of Tiberius. Augustus himself in the celebrated inscrip-
tion of Ancyra speaks of his fleet as sailing " towards the east
to the extreme parts of the world," - but no definite meaning

remarkable for the success of its far- Divi Augusti, Germaniam classe cir-
sighted arrangemeiitrf of any recorded cumvecta ad Cimbroi'iim promontoriuin,

in ancient military history." {History
of the Romans under the Empire, vol.
iv. p. 310.)

9 Augustus, it is said, had prohibited
the Koman armies from crossing the
Elbe ; and Tiberius turned back from
that river, wliich was never again
reached by a Eoman general.

' " Septemtrionalis vero Oceanus ma-
jore ex parte navigatus est, auspiciis

et iude immeuso mari prospccto, aut
fama cognito. ad Seythicam plagam et
humore nimio rigentia." (Plin. if. N.
ii. 67, § 167.)

^ " Classis Romana ab ostio Eheni ad
solis orientis reg'onem usque ad orbis
extrema navigavit, quo neque. terra
neque mari quii-quam Romaiiorum ante
id temi^us adit." (Alomim. Ai)cijranum,
ed. Zumpt. p. 34.)

Sect. 4.

WAES IN GEEMANY.

191

can be attached to such rhetorical expressions. It would how-
ever seem probable that the Eomans first brought back from
this voyage some vague information concerning the existence
of the great promontory or peninsula extending far towards
the north, and of an extensive bay, to which they gave the
name of Codanus Sinus, beyond it to the east. The Cimbri,
who inhabited the modern Holstein as well as the peninsula
itself, sent envoys to make their submission,^ and the Eomans
may readily have acquired the information from them.

§ 4. While the Eoman generals were thus occupied in the
north of Germany, great changes had taken place in the more
southern portions of that country. The Marcomanni, one of
the most powerful and warlike of the German tribes,* who
appear to have been at one time settled between the Ehine,
the Main and the Danube, had been persuaded to migrate
from this region, and establish themselves in the midst of the
forests of Bohemia, where they considered themselves secure
from attack. Here they gradually strengthened themselves,
under the command of a native leader named Maroboduus
until they were able to subdue or to reduce to submission all
the neighbouring tribes, so as to extend their power from the
Danube to the Elbe and the Saal. It was probably in con-
nection with these movements that L. Domitius Ahenobarbus
(the grandfather of the emperor Nero) had some years before
(e.g. 2) been led to take part in favour of the Hermunduri,
who had been expelled from their native abodes, and took
advantage of the opportunity to penetrate into the heart of
Germany. He even crossed the Elbe without opposition, and
after concluding a treaty with the natives immediately beyond
it, erected an altar in honour of Augustus.^

* Ibid. p. 35. It may well be sus-
pected that the Roman fleet really ad-
vanced no further than one of the pro-
montories of Sloswick, and would then
readily accept the idea tliat this was
tlie great northern promontory of which
they had undoubtedly heard from the
Cimbri.

■* TJie Marcomanni are mentioned by

Caesar among the tribes that composed
the army of Ariovistus (-B. G. i. 51).
At this time they dwelt nearer the
Rbine : their migration to Bohemia
under the direction of Maroboduus is
attested by Strabo (vii. 1, p. 290) and
Velleius (ii. 108).

^ Tacit. Annal. iv. 44 ; Dion Cass. Iv.
10 a.

192 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

But this expedition, like so many others, had produced no
permanent impression. Maroboduus had continued to consoli-
date his power, and was become the most formidable adversary
of the Eomans in Germany. Hence Tiberius, after having, as
he supposed, put down all opposition in the north, determined
to direct his arms against the Marcomanni, and for this
purpose concerted a well-arranged plan of attack, according
to which he was himself to advance from Carnuntum on the
Danube, while Sentius Saturninus, a general of proved ability,
was to lead another large army from the banks of the Rhine,
through the Hercynian forest, to the frontiers of Bohemia.^
This plan was however frustrated by the great revolt of the
Pannonians, Dalmatians and lUyrians already alluded to,
which compelled Tiberius to draw off his legions into those
provinces ; a treaty was concluded with Maroboduus, and the
Marcomanni were for the present left in undisturbed possession
of their newly acquired home.

§ 5. Though the accounts which represent the north of
Germany, from the Rhine to the Weser, as at this time com-
pletely subdued and rapidly approximating to the condition
of a Roman province, are probably much exaggerated, it
seems certain that Roman habits of life and Roman civiliza-
tion were really beginning to make considerable progress
among the tribes beyond the Rhine, as they had already done
on the left side of the river. And had not this gradual opera-
tion been abruptly checked, the same process of transformation
that took place in Gaul, would probably have extended to
Germany also. But an end was suddenly put to all such
progress, and the advance of the Roman arms permanently
arrested at the Rhine, by the great defeat of Quintilius Varus
in A.D. 9.'' That memorable disaster, involving the total de-
struction of an army of three legions, and the loss of the
important frontier post of Aliso,^ changed the whole aspect of

® Veil. Pat. ii. 109. 1 perium, quod in littore Oceani non

^ Florus, who wrote under Hadrian, steterat, in ripa Rheni fluminis staret"

remarks, after relating the defeat of (Florus, Epit. iv. 12).

Varus : " Hac clade factum ut im- | * See Note G,»p. 20G.

Sect. 4. WARS IN GERMANY. 1 93

affairs in Germany, and was never retrieved. The actual scene
of its occurrence cannot be definitely determined, and from
the nature of the engagement, or rather series of engagements,
that ended in the annihilation of the Roman force, there is
little clue to its identification. It undoubtedly occurred within
the limits of the Teutoburger Wald — the Teutoburgensis
Saltus of Tacitus — a tract of rugged forest country extending
from S.E. to N.W. between the Lippe and the Weser, from the
neighbourhood of Paderborn to that of Osnabriick. But the
precise locality that was marked by the death of Yarns and
his surviving officers cannot be identified.^ No tradition
attaches to any particular spot, nor have the researches of
modern antiquaries succeeded in discovering any remains that
might enable us to trace the movements of the Roman army,
or determine the scene of the final disaster. A few years later,
while the traces were still recent, the locality was visited by
Germanicus, who paid funeral honours to the fallen Romans,
but the tumulus erected on this occasion to mark the site was
shortly afterwards purposely destroyed by the Germans.^

§ 6. The subsequent campaigns of the Romans in this quarter
had comparatively little geographical interest. Tiberius, in
A.D. 11, again crossed the Rhine and made a show of invading
Germany, but his movements were slow and cautious, and led
to no practical result. After his accession to the throne the
command of the army on the Rhine devolved on the young
Germanicus, who earned a great military reputation in three
successive campaigns (a.d. 14-16), but did not actually pene-
trate into Germany so far as his father Drusus and Tiberius
had already done. Though he boasted, on a trophy that he
erected on his most distant battle field, that he had subdued
all the nations between the Rhine and the Elbe,^ he certainly

9 See Note H, p. 207.

' Tac. Ann. ii. 7.

^ " Csesar congeriem armonim struxit,
superbo cum titulo : debellatis inter
Rhenum Albiraque nationibus exer-
citum Tiberii Csesaris ea monimenta
Marti et Jovi et Aus;usto sacravisse."

Tacit. Annul, ii. 22.

In like manner he celebrated his
triumph in the following year "de
Cheruscis Chattisque et Angrivariis,
quseque alise nationes usque ad Albim
colunt" (Id. ibid. 41). Of these the
Angrivarii, whose name appears for

VOL. II. O

194

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX.

never approached the latter river in person, and had in fact
only just crossed the Weser. His fleet, which had sailed round
as far as the mouth of the Ems, sustained great loss from a
storm on its homeward voyage, and Germanicus himself with
difficulty escaped. The Eomans however at this period seem
to have still maintained some garrisons in these maritime
districts, and the two tribes of the Frisians and the Chauci
continued faithful to the Eoman alliance.

Meanwhile the southern nations of Germany had been
divided by internal discord, and Maroboduus, who had at one
time assumed so threatening a position, was ultimately driven
from his throne, and having taken refuge in the Koman
dominions, lived and died an exile at Eavenna.^ The Eoman
arms made no progress in this quarter, but they appear to
have already begun to occupy, or at least to extend their
jurisdiction over the south-western corner of Germany, lying
between the Ehine, the Neckar and the Danube; a district
subsequently known as the Agri Decumates, and which was
for a considerable period incorporated with the Eoman Empire.

Section 5. — Diodorus.

§ 1, Among the Greek writers contemporary with the reign
of Augustus there is one who deserves a brief notice in this
place, though his work was not directly of a geographical
character. This is the well-known historian Diodorus, com-
monly known as Diodorus Siculus, from his having been a
native of Agyrium in Sicily. His voluminous historical work,
to which he gave the name of Bibliotheca Historica, as if it
formed a complete historical library in itself, extended to forty
books (fifteen of which have been preserved to us) and was

the first time in this campaign, dwelt
on both banks of the Weser, between
the Oherusci and the Chauci. They

are again mentioned by Tacitus in tl
Germania (c. 33).
» Tac. Ann. ii. 62, 63.

MAP TO ILLISTRAIE THE WARS OF THE ROtlANS l\ (.IK\i\N> ,\\|> THE GERMANIA OF TACITITS

Sect. 5. DIODORUS. I95

intended to comprise a general history of the world, including
that of the barbarians and Eomans as well as the Greeks, from
the commencement of historical tradition to the beginning of
the Gaulish wars of Julius Csesar.* Though he himself tells
us that this last date (b.c. 59) was the appointed terminus of
his undertaking, it is certain from his own expressions that
he survived the death of Caesar, and that his work was not
published till after that event.^ On the other hand the
absence of all allusion, in his elaborate description of Egypt,
to that country having passed under the government of Kome,
renders it probable that it was published before B.C. 30, when
Egypt was annexed to the Roman Empire.

Of the historical merits or qualifications of Diodorus it does
not fall within the province of the present work to speak ; it
is enough to say that the uncritical character of his work,
which disfigures it throughout, is not less apparent in the few
notices that have a direct bearing upon geography, than in
the more strictly historical portions of his narrative. Though
he himself tells us that he spent more than thirty years in
the preparation of his great work, and that he undertook
many laborious and dangerous journeys in order to visit those
parts of Asia and Europe which were of the chief historical
interest, and make himself personally acquainted with the
localities,^ it is certain that very little trace is to be found of
any advantage resulting from this cause. Much as we owe to
the laborious compilation of Diodorus for having preserved
to us an outline of many periods of ancient history, which
would otherwise have been almost totally lost, it is impossible
not to feel that it is a mere farrago of materials of very
unequal value, jumbled together without any attempt at
critical judgement or selection.

§ 2. Nowhere are these defects more apparent than in the

* Diodor. i. 4. I eeJs), which he repeatedly associates

* This is evident from the title of with his name, i. 4, v. 21.
DivUS (6 Sia Ttts irpci^ets irpocraYopeuflels | " i. 4.

196

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XX,

first five books, which contain brief notices of the early history
of the Egyptians, Assyrians, Chaldseans, &c., together with
accounts of the Ethiopians, Arabians, Indians, Scythians and
other nations, which would have been of considerable value
and interest to us had they been compiled with somewhat
more judgement and critical sagacity. As it is they con-
tribute very little to our positive knowledge of the nations
to which they relate, and almost nothing to our geographical
acquaintance with the countries they inhabited. Among the
few portions of this part of the work that are really trust-
worthy and valuable is the long and detailed account of the
Ethiopian nations adjoining the Red Sea, which is taken — as
has been already pointed out, and as we are indeed informed
by Diodorus himself — directly from Agatharchides and Artemi-
dorus,'' and which would have been an important accession to
our geographical knowledge had we not happened to possess
the same information from another source.^

§ 3. The fifth book contains notices of the various islands in
the Mediterranean, some of which are not without interest,
though his account of Sicily is singularly poor and meagre,
considering that he was a native of the island.^ To these are
appended, as it were in passing, a brief notice of an island in
the Atlantic Sea, beyond the Pillars of Hercules, several days'
sail from the coast of Africa, which would appear to be based
upon the accounts of the recently discovered island already
noticed, though greatly exaggerated, representing it as of large
size and containing navigable rivers ; ^ and a tolerably full
account of the British Islands. This last is evidently derived
in great part from the information obtained by Caesar, to whose
expedition he expressly refers, as the first time that Britain had
ever been visited by a foreign invader. In one respect indeed

^ Diodor. iii. 11.

8 See Chapter XVIII. §§ 2, 3.

* On the other hand his account of
tlie Balearic, or Gymnesian Islands (as
they were termed Ly the Greeks), is

one of the best that is preserved to us
from antiquity (v. IS).

' V. 19. Concerning this island, see
Chapter XVIII. p. 81.

Sect. 5.

DIODOEUS.

197

his geographical information was in advance of that furnished
us by Csesar himself, or any later geographer till the time of
Ptolemy — that he gives us the names of the three promon-
tories forming the angles of the triangular island, which he
names Cantium, Belerium (evidently the Bolerium of Ptolemy,
the Land's End) and Horcas, the most northern headland, to
which Ptolemy also gives the name of Orcas, evidently in con-
nection with the adjacent group of the Orcades.^

§ 4. To this he adds the most circumstantial account found
in any ancient writer of the production of tin in Britain.^ This,
as he correctly tells us,* was found only in the part of the
island adjoining the promontory of Belerium, the inhabitants
of which were the most hospitable and civilized of all the
Britons, on account of the extensive commerce resulting from
this cause. The tin extracted from these mines was fused
into ingots of a peculiar shape, and carried to a small island
adjoining Britain of the name of Ictis. Here it was purchased
by traders, who carried it to G-aul, where it was transported over
land on horses in about thirty days to the mouths of the Rhone.
The island of Ictis is described as surrounded by the sea at
high water, but connected with the main land by a tract of
sand, which was left bare at low water, so as to render it a
peninsula, to which the tin was carried in waggons. This cha-
racteristic account leaves no reasonable doubt that the locality
indicated was St. Michael's Mount, to which the description
precisely answers, and which contains a small port such as
would have been well suited to ancient traders.^ From whence

2 Diod. V. 21.

' It is remarkable that while he here
correctly describes the tin-producing
mines as situated on the main island
of Britain, he has no mention in this
place of the name of the Cassiterides or
Tin Islands, which he in common with
almost all other writers considered as
connected with Spain and describes
them elsewhere accordingly (v. 38),
where he however repeats the statement
that a great quantity of tin was trans-
ported from Britain to tlie opposite
shores of Gaul, and from thence overland

to Massilia and Narbo. * v. 2.

* The resemblance of the name to
that of Vectis — the Isle of Wight — has
led some modem writers to suppose
that to be the island meant; but in
such cases the resemblance of physical
characteristics outweighs enormously
that of mere name. Nor could the Isle
of Wight have been by possibility at
any time the centre of the tin- trade,
which, as Diodorus himself points out,
was confined to the district near the
Land's End, to which he gives the
name of Belerium.

198 HISTORY OF ANCIENT GEOGEAPHY. Chap. XX.

Diodorus received this information we have no knowledge.
There can be little doubt that his island of Ictis is the same
with the Mictis of Timeeus, though that writer, as we have
seen, had only a very confused idea of its position : but his
detailed account of the mode in which the tin was carried
across Gaul to the mouths of the Ehone (i.e. to Massilia) seems
to point to some much more recent source of information. It
may not improbably have been derived from that obtained by
P. Crassus, the lieutenant of Csesar, which is recorded to us by
Strabo.^

It is eminently characteristic of the uncritical character of
mind of Diodorus, that in the same book with this account
of Britain, he has given at considerable length a description
of the island of Panchaia in the Erythraean Sea,' taken from
Euhemerus, whose work is justly treated by Strabo and other
authors as a pure and absolute fiction.®

« Strabo, iii. 5, § 11, p. 176. ' Diodor. v. 42-46.

» See Strabo, i. p. 47, ii. p. 102, &c.

Note A. KOMAN EMPIRE UNDEll AUGUSTUS. 199'

NOTE A, p. 144.

ALPINE TRIBES.

The ethnology of these Alpine tribes is still very obscure. But it
would appear certain that none of them were Germans. The well-
known statement of Livy that the Khseti and other Alpine nations
were of kindred origin with, the Etruscans (Liv. v. 33 : " Alpinis
quoque ea gentibus hand dubie origo est, maxime Eeetis : quos loca
ipsa efferarunt, ne quid ex antiquo preeter sonum linguae, nee eum
incorruptum, retinerent ") is one of those assertions of the value of
which it is almost impossible to judge : but supported as it is by the
statement of their retaining a similarity of language — a fact of which
the Eomans were well able to judge — we should be hardly justified
in rejecting it altogether. But this Etruscan element was pro-
bably confined to some of the more southerly tribes, occupying the
slopes of the Alps adjoining Italy, into which they had been driven
when expelled by the Gauls from the valley of the Po. The pre-
ponderance of argument appears to be in favour of the Ehsetians
(and with them the Vindelicians, who are always described as being
a kindred tribe) being of Celtic 'or Gaulish extraction, (see Zeuss,
Die Deutschen, pp. 228-238 ; and Diefenbach, Geltica, vol. i. pp. 133-
137). The same thing may be asserted more confidently of the
Tauriscans, who formed the bulk of the population of Noricum,
and of the Scordiscans, a decidedly Celtic people (Strabo, vii.
pp. 313, 315), who were at this period still settled in Pannonia.
But the lapodes, a tribe who inhabited the Julian Alps between the
Save and the Adriatic, are called by Strabo a mixed people, partly
Gaulish and partly Illyrian ('IctTroSes, ^St) tovto l-Trip.iKTov 'lXXvpLOi<s
Koi KcA,Tots Wvo?. Strabo, iv. p. 207) : and there certainly seems
reason to believe that the bulk of the Pannonians were an lliyrian
race.

Dion Cassius, who had himself been governor of the province of
Dalmatia and uj)per Pannonia, has given us a graphic sketch of
the Pannonians, whom he describes as KaKojBmraroL dvOpwTrwv oVres :
inhabiting a cold and barren country, producing neither oil nor
wine, and compelled to make their drink as well as food from

200 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

barley and millet. But this very poverty rendered them also the
hravest and most pugnacious of mankind. (Dion Cass. xlix. 36.)
It is evident that this description could apply only to the trihes
inhabiting the mountain districts on the borders of Dalmatia and
lUyria, corresponding to the modern Bosnia and Herzegovina ; not
to those that occupied the fertile plains on the banks of the Save
and the Drave.

Dion Cassius (I. c.) correctly points out the error committed by
many Greek writers in confounding the Pannonians with the
Pgeonians, a people inhabiting the mountains in the north of
Macedonia, with whom they had nothing in common : but he makes
no statement with regard to the ethnic affinities of the Pannonians.
The name first appears in history when the people came in contact
with the Koman arms.

NOTE B, p. 157.

GALATIA.

Galatia, in this sense, included the cities of Iconium, Antioch of
Pisidia, Derbe and Lystra, so well known from the part they bear in
the travels of St. Paul, and M. Eenan has well pointed out that it
was to the inhabitants of these cities, and not to the Galatians,
properly so called, that the Epistle to the Galatians was in all pro-
bability addressed. (Eenan, St. Paul, pp. 48-60.)

The original people of the name, who continued to inhabit the
province where they had been settled ever since the time of
Attains I., king of Pergamus, between the Sangarius and the
Halys, retained their nationality with striking pertinacity. They
continued to be divided into three tribes, the Tectosages, Trocmi,
and Tolistoboii ; all of them distinctly Gaulish names, and the first
still borne in the time of Strabo by a tribe in the south of Gaul.
(Strabo, xii. p. 567.) They retained also their native language,
which they continued to speak with very little change, as late as
the time of Hieronymus, in the fourth century of the Christian era.
(Hieronym. Comment, in Epist. ad Galat. ii. 3, p. 430). Their chief
city at this time was Pessinus, but Ancyra became the capital of
the Eoman province, and soon rose to the important position which
it has ever since retained.

Notes C, D. JUBA. 201

NOTE C, p. 166.

TIGERS.

It was doubtless one of these tigers that Augustus afterwards
exhibited in a cage on the occasion of the dedication of the theatre
of Marcellus (Plin. E. N. viii. 17, § 65). This was the first tiger
seen at Borne, as we are expressly told by Pliuy. But Dion Cassius
certainly goes too far in supposing that those presented to Augustus
were the first ever seen by the Greeks. Tigers must have been seen
by the companions of Alexander in India : besides which they were
found in Hyrcania and the adjoining provinces : every one is
familiar with the expression of " Hyrcanse tigres" in Virgil
{^n. iv. 367). Pliny also says : " Tigrin Hyrcani et Indi ferunt "
(/. c.) ; and Mela has a full notice of them in reference to Hyrcania
(iii. 5, § 43). They are still found not uncommonly on the west
side of the Caspian in the dense forests and jungles near the mouth
of the Araxes.

NOTE D, p. 175.

JUBA's account of the NILE.

Plin. V. 9, s. 10, §§ 51-53. It is hardly worth while to discuss
in detail a statement which is so obviously a mere string of un-
founded inferences and assumptions. But it deserves a passing
notice as the first suggestion of that supposed connection between
the Niger and the Nile, which continued so long to be a favourite
theory even with modern geographers. If we can rely upon the
mention of the river Niger being found in Juba, he was certainly
the earliest author that was acquainted with that celebrated name,
and the fuller notice of it found elsewhere in Pliny (v. 8, § 44) may
probably be derived from the same source. The statement that its
banks were clothed with forests is also interesting, as the first indi-
cation in any ancient author of the existence of the fertile regions of
Soudan, beyond the broad desert tract of the Sahara.

The supposition that the two lakes mentioned in the first part
of the account were fed by the same river, and had a subterranean
communication with one another and with the Niger, is of course a
mere fancy ; but the statement that they contained crocodiles and

202 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX.

large fish., such as siluri, &c., is remarkable, as this was a fact likely
to be within his own knowledge, and he even adds that a crocodile
sent from thence was preserved in the temple of Isis at Ctesarea.
(lol.) (Crocodilus quoque inde ob argumentum hoc Cgesarese in Iseo
dicatus ab eo spectatur hodie. § 51.) The lakes now found at the
foot of the Atlas are all shallow, and nearly dry in the summer; but
they may well have been more extensive in ancient times. The
assertion also made (§ 51) that ihe rise and fall of the Nile was
coincident with the fall of rain and snow in Mauretania was cer-
tainly erroneous : the rains of tropical Africa, upon which the
inundation of the Nile really depends, having no connection with
those of Mauretania.

NOTE E, p. 176.

THE FORTUNATE ISLANDS.

The account of the group of islands in question given by Juba
and reported by Pliny (vi. 32, §§ 203, 204) deserves a more
careful examination, as the only one with any pretension to accu-
racy transmitted to us from antiquity. That of Ptolemy, as we shall
hereafter see, is a mere confused jumble of different reports. Pliny
begins with telling us, after giving the different statements con-
cerning the islands of the Gorgons, Hesperides, &c., and showing
their utter uncertainty, that there was no more certain information
concerning the islands of Mauretania. It was only ascertained
(constat) that there were a few, opposite to the land of the Autololes,
which were discovered by Juba (a Juba repertas), and in which he
had established a factory for dyeing the Ga^tulian purple (§ 201).
There can be no doubt that these are the same islands to which he
gives the name of Purpuraria^, a few lines further on, where he
tells us, as the result of the researches of Juba concerning the
Fortunate Islands, that these were situated towards the south and
west, and were distant 625 miles from the Purpurarias, " sic ut ccl
supia occasum navigetur, dein per ccclxxv M. P. ortus petatur."
It is very difficult to know what sense Pliny attached to these
words : the most probable explanation is that suggested by Gossellin,
that his authority was really describing the double voyage, to and
fro, and that he has erroneously combined the two into one dis-
tance, lu this case, if we suppose the Purpuvariic Insula? to bo the

Note E. JUBA. 203

two easternmost of the Canary Islands, Lanzarote and Fuerte-
ventura — the conclusion adopted by D'Anville and Gossellin, as well
as more recently by Mr. Maj.or, — the distance of 250 miles is not
far from the truth, as the direct distance to the outermost of the
group, while the larger number may be accounted fur by supposing
it to be the aggregate of the separate distances from one island to
another, a frequent source .of error in similar computations. On
this supposition the description of the islands would follow the
course of the return voyage, beginning with one of the outermost,
and this is in accordance with the fact that Ombrios, which he
names first, is described as having a lake or pool (stagnum) in the
mountains, a statement which probably refers to the celebrated
crater or caldera in the island of Palma ; one of the two most
westerly of the group. The island of Nivaria, perpetually shrouded
in snow or mist, is clearly Teneriffe, with its mighty snow-clad
peak : while Canaria, the most fertile of the group, is equally cer-
tainly that now known as Grand Canary. There remain Junonia
and Capraria, of which the former might readily be identified with
Gomera, and the latter with Ferro, the smallest of the whole group.
But a difficulty arises from the circumstance that a second and
smaller island of the same name is said to exist in the neighbour-
hood of Junonia. No such island now exists, and it is almost certain
that there is some misconception on this point. With this excep-
tion the identifications are satisfactory enough.

The chief difficulty arises with regard to the Purpurarise Insulas,
which are certainly not mentioned by Pliny in a manner that would
lead us to suppose they were so nearly connected with the Fortunate
Islands as are Lanzarote and Fuerteventura with the rest of the
Canaries. His expressions would rather seem to imply that they
were small islands on the coast of Mauretania. But no such islands
_are to be found, and the circumstance that these two, in common
nth. the rest of the Canaries, abound in orcMl, a kind of lichen
/■ielding a beautiful purple dye, raises a strong presumption that
this was the " Purpura Gastulica " mentioned by Pliny, and from
which the islands derived their name.

Humboldt supposes the Purpurari^ to have been the group of
Madeira and Porto Santo, but these islands produce no orchil :
besides which they lie so far out to sea that they can hardly be
supposed to have been those described by Pliny as islands of
Mauretania, opposite to the Autololes. Moreover, unless we

204 HISTORY OF ANCIENT GEOGRAPHY. Chap. XX,

suppose Lanzarote and Fuerteventura to be those designated as
the Purpurariai, these two important islands remain unaccounted
for, and it is impossible to explain why Juba, in describing the
Fortunate Islands, proceeding from west to east, should have
stopped with Canaria and not noticed the two lying between it and
the mainland. The actual distance of Fuerteventura from the
nearest point of the coast of Africa does not exceed 50 G. miles.

If we compare the list cited from Statins Sebosus with that of
Juba, we find the names of Junonia and Capraria the same, while
his Pluvialia is obviously identical with the Ombrios of Juba ; but
his distances and positions are altogether unintelligible, and it is
evident that his information was mere hearsay. That of Juba, on
the contrary, was clearly the result of careful inquiry, and is in
general perfectly correct. But there is no foundation for the state-
ment that he sent out an expedition for the express purpose of
exploring the Fortunate Islands. (Major's Prince Henry, p. 136.)
Pliny's words (" Juba de Fortunatis ita inquisivit ") cannot be held
as affirming anything of the kind.

NOTE F, p. 182.

EXPEDITION OF iELIUS GALLUS INTO ARABIA.

The geography of this expedition has been investigated with
much diligence by Mr. Forster in his ' Geography of Arabia,' as well
as by Dr. Vincent, Gossellin, and several earlier writers; and more
recently by Mr. AVilliams in Dr. Smith's ' Dictionary of Ancient
Geography' (ait. Maesyabte); but the wide divergence between
their views and results sufficiently shows the uncertaintj" of the
subject. No conclusion can ftxirly be drawn from the march in
advance, as we are distinctly told by Strabo that the Roman army
was purposely misled, and wandered about without occasion, so as
to waste much time. Whether it was actually led too far into the
interior, to Nejd and other inland districts, and then out again to
the borders of Yemen, or only strayed within more moderate limits,
we have no means of judging; though the former supposition is the
most probable, if we can place any reliance on the statement of the
time occupied on the march. But none of the names of places men-

Note F. MILITAHY EXPEDITIONS. 205

tioned by Strabo during the advance can be identified. These
names themselves vary much in our MSS., while those given by Pliny
are equally uncertain. That author tells us only : " Gallus oppida
diruit non nominata, auctoribus qui ante scripserant, Negranam,
Nestum, Nescam, Masugum, Caminacum, Labeciam, et supra dictam
Mariabam circuitu vi mil. passuum; item Caripetam quo longissime
processit" (vi. 28, § 160. The readings adopted by Sillig in his
latest edition are here followed, but the names vary much in the
earlier editions and MSS.). Here we find the name of Negrana in
both authors, and the Nesca of Pliny may reasonably be identified
with the Esca or Asca of Strabo, but his other four names mentioned
in the same sequence, as well as Caripeta, which he represents as the
terminus of the expedition, are wholly unknown. His Mariaba
is doubtless the same place as the Marsiaba or Marsyabse of Strabo,
though he erroneously represents it as having been taken by
Gallus ; but he appears to have confounded it with another Mariaba,
which was situated in the land of the Calingii, and indeed it is
clear that there were several places of the name in Arabia. Pliny
himself mentions three towns of the name, which he certainly sup-
posed, whether correctly or not, to be situated in different parts of
Arabia. The most important and best known of these was un-
doubtedly the capital of Sabaea, which still retains the name of
Mareb : and this has been generally supposed to have been the
place besieged by Gallus. The land of spices (Jj apwiJiaT6(f)opos)
could hardly have been any other than Hadramaut, and this they
are supposed to have approached within two days' journey ; but
very little reliance can be placed on this statement, which is said
to have been derived from captives. The distance also from Leuce
Come (supposing that place to have been at Howara) exceeds what
any army could reasonably be supposed to have marched within
60 days. '

The position of Leuce Come at Howara seems to me well esta-
blished, notwithstanding the counter arguments of Mr. Williams :
and the opinion of D'Anville, who first pointed out that the modern
Arabic name has the same signification as the ancient one, has
been adopted and confirmed by Eitter and C. Miiller. (Eitter,
Geographie von Asien, xii. p. 123, &c. ; C. Miiller in his edition
of the GeograpM Greed Minores, torn. i. p. 272, note; D'Anville,
Memoires sur VEgypte, p. 243.) The objection that it is too far
south to have been included in the territory of the Nabataeans

206 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XX.

has little weight, as the limits of these Arabian tribes are little
known, and were doubtless subject to great fluctuations. The
author of the 'Periplus of the Erythr^an Sea,' in whose time
Leuce Come was still a place of considerable trade, places it at two
or three days' voyage across the Gulf (i. e. the Eed Sea) from Myos
Hormus (Peripl. § 19) ; an estimate considerably less than the truth
(the real distance being about 250 miles) ; but this part of his
treatise is given in a very vague and general manner, and evidently
does not aim at great accuracy.

The subject has recently been investigated afresh by M. Sprenger,
in a paper inserted in the Journal of the Asiatic Society (N. S. vol. vi.
1873, pp. 121-141), who has thrown considerable light on several
points, though some of his identifications rest upon dubious grounds.
The most important is that of Negrana, which mskj safely be con-
sidered as represented by the modern Nejran, situated in lat. 17° 20',
and about 150 miles n.n.w. of Mareb, which would suit well with
the nine days' march from the one city to the other on the retreat.
He therefore identifies the Marsyaba of Strabo with the well-known
city of Mariaba in Sabsea, still called Mareb. And he finds the
name of the Ehammanitse represented by a town called Ehadman in
the same neighbourhood, and the Caripeta of Pliny in a place still
called Kharibeh. He supposes Gallus on his advance to have
been led through the district of Nejd and Hajr (of which Eiadd is
the capital), and from thence to the borders of Yemen, where
Negrana would be the first place they met with. All the other
towns mentioned by Strabo and Pliny were probably within the
same fertile district of Yemen.

In these general conclusions we must be content to acquiesce, as
the account given by Strabo is not sufficiently detailed to admit of
more accurate determination, and the " towns " in the interior of
Arabia have seldom left any vestige of their existence.

NOTE G, p. 192.

ALISO.

The position of Aliso, which bears so important a part in these
campaigns, unfortunately cannot be identified with any certaintj^.
There can be no reasonable doubt that it was the fortress erected

Note H. WARS IN GERMANY. 20/

by Drusus in B.C. 11, on the banks of the Lippe, in order to secure
possession of the territory between the Ehine and the Weser (Dion
Cass, liv. 33). This was placed, according to Dion Cassius, at the
confluence of the Lippe with a small stream which he names 'EXtcrwi/,
but this streamlet cannot be identified ; and accordingly Aliso has
been placed at almost every point along the course of the Lippe,
from its sources to its confluence with the Ehine. The pre-
dominance of opinion among German writers appears to be in
favour of a place called Elsen, a few miles west of Paderborn, at
the juncture of the Lippe with the Alme, and this view has been
strenuously maintained by the most recent inquirers into the
subject, Von Weitersheim and Abendroth, as well as by M. Schieren-
berg (Die Bomer im Glieruskerlande, p. 27). Dean Merivale, how-
ever, considers this as too far from the Ehine, and is disposed to
adopt Hamm, about 35 miles farther west, as a more plausible
locality; the same view is sanctioned by Niebuhr, while Ukert
inclines in favour of Haltern, still considerably further towards the
west, and only about 25 miles from the Ehine. The point does not
seem susceptible of any positive decision ; the trifling Eoman
remains that have been found in different localities being incon-
clusive, as there were certainly other Eoman forts and military
stations in this part of the country. The subject is full}^ discussed
and the older authorities reviewed by Ukert (Geogr. vol. iii. pt. i.
p. 489). See also Merivale's History of the Bomans, vol. iv. pp. 232,
360 : and the article Aliso in Smith's Geogr. Diet. vol. i. p. 103.

The river Else, the name of which might readily suggest its
identity with the Elison of Dion Cassius, is out of the question, as
it flows into the Werra instead of the Lippe, and is on the further
side of the Teutoburger Wald.

NOTE H, p. 193.

DEFEAT OF VARUS.

This is the conclusion in which the most recent German his-
torians find themselves compelled to acquiesce. Great ingenuity
and much labour have been expended upon the subject by German
antiquaries, and the result of their researches is fully summed up
by Ukert {Geogr. vol. iv. pp. 123-136). But that judicious and

208 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XX.

cautious writer, after giving the substance of all tliat we learn from
ancient authors concerning this memorable event, points out how
imperfect is the information that we derive from them concerning
any of its details. We do not know the situation of the camp of
Varus, from whence he set out, nor the direction of his march in
the first instance : and though it seems certain that when he found
himself attacked on all sides, he directed his march towards Aliso,
the position of that fortress, as already stated, is itself uncertain.
The nature of the country also is such as, while agreeing com-
pletely with the general descriptions of the ancient authorities, can
hardly admit of any more accurate determination, Forests, marshes,
and a succession of ridges of hills of no great elevation, are found
throughout the tract in question, and afford no clue to the distinction
between one locality and another.

The only narrative that gives us any details is that of Dion
Cassius (Ivi. 18-22) : but some interesting facts are furnished by
Tacitus in his account of the visit of Germanicus to the same
localities a few years afterwards (Annal i. 61, 62). The rhetorical
flourishes of Velleius (ii. 117-119) and Florus (iv. 12) convey no
definite information. The most recent investigation of the subject
by E. von Wietersheim (Gescliichte der VdlJserwanderung, vol. i.
pp. 425-433) has been supplemented by Major Abendroth in his
Terrainstudien zu dem BucTczuge des Varus und den Feldziigen des Ger-
manicus (8vo. Leipzig, 1862), who has examined the ground from a
military point of view. He fixes the position of Aliso at Elsen
near Paderborn, and places the scene of the three days' fighting and
the destruction of the army of Varus north of the ridge of the
Teutoburger Wald, in the neighbourhood of Lemgo and Detmold,
and on the line of retreat from thence to Elam. But his arguments,
though plausible, are hardly conclusive. Another late writer on
the subject (Schierenberg, Die Edmer im Cheruskerlande, Frankfort,
1862) fixes the site of the battles between Feldrom and Driburg,
a few miles further south, but on the southern slope of the moun-
tain ridges near the sources of the Lippe. After reading them both,
one still feels disposed to acquiesce in the dictum of Niebuhr, more
than fifty years ago : " Die Gegend wo Arminius den Varus schlug
ist nimmermehr zu ergriinden." (Vortrdge iiher Bomische Gescliichte,
vol. iii. p. 156.)

( 209 )

CHAPTEE XXI.

STEABO.

Section 1. — General Views.

§ 1. We are now come to tlie period when we are able for the
first time to obtain a complete and satisfactory view of the
state of geographical science. For this advantage we are
indebted to the comprehensive work of Strabo, which, as
Humboldt has justly remarked, "surpasses all the geogra-
phical writings of antiquity, both in grandeur of plan, and in
the abundance and variety of its materials." ^ Its author
flourished during the whole of the reign of Augustus and the
early part of that of Tiberius, and his great geographical work
could not have been completed earlier than the year a.d. 19 :
so that it may be taken as representing the state of geo-
graphical science, as well as the political organization of the
Empire, as it existed after the death of Augustus and the
completion of his task in the construction of that vast system
of government.

Of the author's life and personal history we know little ; but
as that little is derived entirely from incidental notices and
statements in his own work, it may be relied on as perfectly
authentic. Some modern writers however have endeavoured
to derive from these notices a number of inferences and con-
clusions, which are, to say the least, very dubious, and it is
safer to disregard them altogether. He was a native of the
city of Amasia in Pontus, which, though situated in the in-
terior of the country, and at one time the residence of a
dynasty of barbarian kings, had imbibed a strong tincture of

' Humboldt's Cosmos, vol. ii. p. 187. Engl, transl.
VOL. II. P

2IO

HISTORY OF ANCIENT GEOGEAPHT,

Chap. XXI.

Greek civilization, and had probably a large Greek popula-
tion.^ It is certain at all events that Strabo received a good
Greek education, which fitted him for subsequently pursuing
his studies under rhetoricians and philosophers of the highest
reputation. Of his father's family we know nothing, but that
of his mother occupied a distinguished position, and different
members of it had held important military and political posts
under Mithridates Euergetes, and his more celebrated son,
Mithridates Eupator. During two generations they had settled
at Cnossus in Crete, and from this circumstance Strabo derived
connections with Crete, to which he adverts in his description
of that island.^

The year of his birth cannot be determined with certainty :
but he tells us himself that he was quite young, when he was
sent by his father to Nysa in Caria, to prosecute his studies
under Aristodemus, a native of that place, who at this time
enjoyed a considerable reputation as a grammarian.* He sub-
sequently studied philosophy under the Aristotelian Xenar-
chus, who was a native of Seleucia in Cilicia, but it is probable
that this took place either at Alexandria or Eome, in which
cities Xenarchus resided a great part of his life.^ Beyond the
journey necessary on this occasion we know nothing of the cir-
cumstances that led to his travels, but these appear to have
been commenced while he was still young, and we learn from
himself that he visited Corinth at the time that Augustus was
there on his return from Egypt to celebrate his triumph at
Eome, a circumstance which fixes his visit to Greece in B.C. 29.
From thence he proceeded to Eome, where he certainly spent
a considerable time, and probably remained several years. But
the farthest part of Italy to which his travels extended was
Etruria ; where he visited the headland of Populonium — from
whence travellers were told that they could see Corsica and

^ Amasia, which still retains its
name, and the curious tombs of the
kings there, are fully described by Mr.
Hamilton in his Hesearches in Asia

Minor and Pontns, vol. i. pp. 366-372.
3 Strabo, x. 4, ^ 10, p. 477.
* Id. xiv. i. § 48, p. 650.
' Id. xiv. 5, § 4, p. 670.

Sect. 1. STEABO : GENERAL VIEWS. 211

Sardinia ** — and apparently also the Port of Luna, or Gulf of
Spezia. It was probably on bis return from Kome that he
repaired to Alexandria, where he resided a considerable time,
and took the opportunity to accompany the Eoman governor
-331ius Gallus on his voyage up the Nile to Syene and Philse.^
This expedition took place in B.C. 24.

§ 2, Though Strabo boasts of the extent of his travels as
qualifying him for the task he had undertaken, and asserts
that they comprised a wider range than any previous geo-
grapher had done, " for that those who had penetrated farther
towards the West, had not gone so far to the East, and those
on the contrary who had seen more of the East had seen less
of the West:" it must be admitted that they were not really
in any way remarkable, nor is there any evidence that they
were undertaken in a scientific spirit, or carried out in a
systematic manner. Though he had visited several distant
points — according to the ideas of his age — and could assert
with truth that. he had travelled from the frontiers of Armenia
on the east to the shores of the Tyrrhenian Sea on the west,
and from the Euxine Sea to the borders of Ethiopia,^ he was
far from having seen, even in the most superficial way, the
different countries that lay within these limits. His personal
acquaintance with Italy was by no means extensive : and even
of Greece itself he saw very little : apparently only Corinth
(where he ascended the Acro-Corinthus)— Athens, Megara and
perhaps Argos. He speaks of having seen Gyrene from the
sea ^ (probably on his voyage from Italy to Egypt), but he did
not land there, or take the trouble to visit so celebrated a city :

« Id. V. 2, § 6, p. 223. It is a popular
error, though one repeated by many
writers, in modern as well as ancient
times, that Sardinia, as well as Corsica,
is visible from this point of the Tyr-
rhenian coast. Strabo himself remarks
that it is "a long way off and seen
with difficulty " (irSppccOev fxkv koI fiSXis).
But it is in fact wholly concealed by
the intervening lofty mass of Elba,
even if the distance, of above 120

miles, were not too great (Dennis's
Etruria, vol. ii. p. 239). Eratosthenes,
though he had certainly never been
there, denied that either Corsica or
Sardinia could be seen, for which he is
justly censured by Strabo, as the former
is plainly visible on a fine day.

7 Strabo. ii. 5, § 12, p. 110; xvii. 1,
§ 50, p. 818.

8 Id. ii. 5, § 11, p. 117.

3 Id. xvii. 3, § 20, p. 837.

P 2

212

HISTOEY OF ANCIENT GEOGKAPHY.

Chap. XXI.

and he describes Tyre in terms that prove he had not seen it/
and consequently could not have coasted along the sliores of
Phoenicia. He probably returned from Alexandria direct to
Ehodes. With Asia Minor he was naturally better acquainted,
from its proximity to his native country : but even there the
very unequal character of his descriptions shows how imperfect
was his acquaintance with many parts of that great peninsula.
Though a native of Pontus, his description of the neighbouring-
countries of Armenia and Colchis is but vague and superficial,
while of the lands beyond the Phasis, between the Caucasus
and the Euxine, he knew no more than what he derived from
the historians of the Mithridatic wars.^

On his return to his native city Strabo appears to have
devoted himself to the composition of a great historical work,
to which he gave the title of 'Historical Memoirs' {'laropiKa
vTTOfjivrjiJiara), extending to not less than 43 books, and com-
prising the period from the fall of Carthage and Corinth
(B.C. 146), with which Polybius had closed his great work, to
the death of Caesar, or perhaps even to the Battle of Actium.^
It was not till after he had completed this, that he undertook
the composition of his geographical treatise, which he himself
calls a colossal work.* He must therefore have been already
in advanced age, and it required no little energy to enter upon
such a task : but we are not reduced to the necessity of
adopting the paradoxical conclusion of Groskurd, that he did
not commence it till he was in his eighty -third year ; a state-
ment which would require much better evidence than we
possess on the subject to entitle us to receive it. We know
indeed with certainty, from historical facts incidentally men-

' Id. xvi. 2, § 23. ivTavda 5e cpaffi
TToXvcrreyovs ras oiKias, uiare ical toov ev
'Pd/xri /xaWov. The fact that the houses
were lofty and of many stories, was one
which he must have known from per-
sonal observation, had he even passed
by sea within sight of Tyre.

2 xi. 2, p. 407.

2 Strabo, i. 1, § 23, p. 13. It is cited

by Plutarcli (Lucull. c. 28, Sylla, c. 26),
and he himself refers to it in xi. 9, § 3.
The statement that it was in forty-three
books rests upon the somewhat dubious
authority of Suidas (e. v. UoXv^ios, see
Bernhardy's note).

■* Ibid. p. 14. KoXoffffovpyia yap t<s
Kol avn'].

Sect. 1. STKABO : GENERAL VIEWS. 213

tioned, that it was not altogether completed, in its present
form, until the year 18 or 19 A.D. : but the period at which it
was commenced and the time occupied in its composition are
wholly unknown to us. Moreover the birth-year of Strabo, as
already mentioned, is itself uncertain, and the assumption of
Groskurd that he was born as early as B.C. 66, is a mere infer-
ence, and rests upon no satisfactory evidence.^

§ 3. The Geography of Strabo is not only the most im-
portant geographical work that has come down to us from
antiquity ; but it is unquestionably one of the most important
ever produced by any Greek or Eoman writer. It was indeed,
so far as we know, the first attempt to bring together all the
geographical knowledge that was attainable in his day, and
to compose what would be called in modern times a general
treatise on geography. It would be a great mistake to regard
it (as some German writers have done) as merely a new edition
of that of Eratosthenes, with additions and corrections. The
general outline of his system was indeed adopted by Strabo,
though not without considerable alterations — some of them, as
we shall see, very far from improvements : but this could
hardly have been otherwise, as the great Alexandrian geo-
grapher had been the first to lay the foundations of scientific
geography on a basis on which his successors could not but
continue to build. But the work of Eratosthenes, which was
comprised in only three books, was limited to an exposition of
his general geographical system, together with statements of
distances and directions, that might serve to determine the
configuration of the several countries described. It was merely
a technical geographical treatise in the strictest sense, and its
small extent alone proves that it could not have contained any
such full or detailed description of each country, and its
natural productions and peculiarities, as Strabo justly con-
ceived to fall within the domain of the geographer.^ Still
less could it have admitted of those historical and incidental

See Note A, p. 272. « See Chapter XVI. p. 653.

214 HISTOEY OF ANCIENT GEOGKAPHY. Chap. XXI.

notices whicli form one of the great sources of interest in the
work of the later author. Strabo indeed appears to have been
the first who conceived the idea of a complete geographical
treatise, as comprising the four divisions that have been called
in modern times, mathematical, physical, political, and histo-
rical geography, and he endeavoured, however imperfectly, to
keep all these objects in view, in the execution of his extensive
but well-considered plan.

§ 4. His historical digressions, though in themselves valuable
and interesting, especially to us, who have lost so many of the
original sources from which they were derived — are sometimes
longer than can well be deemed suitable to a geographical
work ; and this is still more the case with his mythological
ones : but to a Greek all the early mythical legends had not
only a charm from association, but possessed a vivid reality
which we can hardly appreciate at the present day. Strabo
discusses questions of the heroic ages of Greece, and the ex-
ploits of Hercules and Jason, with as much earnestness as he
would those of Alexander and his successors. To him the
voyage of the Argonauts to Colchis was as real as that of
Columbus or Vasco de Gama to ourselves •? and with regard
to the Homeric geography he adopted in their fullest extent
the views of those who regarded the poet as the source of all
wisdom and knowledge, whose statements might require to be
explained or accounted for, but could not possibly be discarded
as erroneous. The blind reverence paid by most Greeks of his
day to the works of the great poet was little short of that with
which many other nations are accustomed to regard their
sacred books — as an authority paramount to all others, which
it was rank heresy to dispute or question. Eratosthenes, as we
have seen, had indeed led the way to a more cautious criticism,
in this respect : but he appears to have found few followers in

' See the passage (i. 2, § 38, p. 45)
^vhere he speaks of tcSj/ irepl rhv 'idcrova
avfi^avroiv koI TTiv'Apylii koI roiis ^Apyo-
vairas, r w i/ 6 fio Xoyov/xefaiy irapa

traffiv : and indignantly rejects the
idea that Homer could have been igno-
rant of what everybody knew.

Sect. 1. STRABO : GENERAL VIEWS. 215

these opinions, and Hipparchus, Polybius and Posidonius all
accepted the ordinary and received identifications of the
localities in the Odyssey without scruple or hesitation.^

§ 5. It is remarkable that while Strabo was thus ready to
adopt the mythical legends of the earlier days, and even the
forms into which they had been worked up by Ephorus and
other logographers, he treated the work of Herodotus with
altogether undeserved contempt, and classes him with Ctesias
and other compilers of fables, whose statements are wholly
unworthy of consideration.^

On some points, as we have seen, Herodotus had really
correct information, where Eratosthenes and other later
writers were misled into error — as with regard to the Caspian
Sea : on others, his scepticism, though not well founded in
fact, was certainly not unphilosophical. But the full informa-
tion that we possess at the present day, which enables us to
discriminate the true from the false, among the conflicting
statements on these and other subjects, was wanting in the
time of Strabo : and even had he brought to the task more
critical sagacity than he actually possessed, it would have
been difficult for him without such assistance to have arrived
at sound conclusions. In like manner he may be censured for
discarding without reserve the accounts of Pytheas concerning
the western and northern regions of Europe : but here he was
evidently led away by the example of Polybius, for whose
judgement and authority he entertained — and not without
reason — a high respect. Some of the statements of Pytheas
were undoubtedly such as to inspire great doubts of his vera-
city : and it must be added that they did not correspond with
the geographical system of Strabo, in regard to the points on
which he differed from Eratosthenes. The love of system
was carried to an extreme by almost all the Greeks, and our
geographer was certainly not exempt from that failing.

* ApoUodonts, as we have seen, was
an exception, and Demetrius of Scepsis
had to some extent adopted the same
view. Strabo, i. 2, §§ 35, 38.

9 i. p. 43, xi. p. 508. On both these
occasions he associates the name of
Herodotus with those of Ctesias and
licllauicus and other retailers of fables.

2l6 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXI.

Another instance in which he was led to reject the state-
ments of Eratosthenes without sufficient reason was in regard
to the island of Cerne on the west coast of Africa, the very-
existence of which he treats as a fable,^ though as we have seen
there is no reason to doubt that it was long occupied by the
Carthaginians as an emporium of trade.

§ 6. Strabo may be still more deservedly censured for the
nesdect he showed for Latin writers, and the information to be
derived from that source. Though he himself points out the
great increase in the knowledge of the western parts of Europe
that had resulted from the extension of the Koman arms in that
quarter, he certainly availed himself to a very small extent of
the materials thus placed at his disposal. It is true that no
Boman writer of eminence had as yet put forth any professed
geographical work ; but their historical writings undoubtedly
contained much that was of the greatest value to the geo-
grapher. Yet in regard even to the west of Europe — Spain,
G-aul, and Britain, — Strabo continued to follow principally
the Greek authorities ; and though he refers in one passage
directly to Caesar's Commentaries,^ and evidently derived other
information from the same source, yet he was far from availing
himself of that valuable work to the extent that he might well
have- done. He appears also to have been unacquainted with
the works of his contemporary Juba, from whom, as we have
seen, Pliny gathered so much information ; otherwise he could
scarcely have failed to cite him in regard to Mauretania and
Western Africa. But we cannot wonder if Strabo, writing at
Amasia, was ignorant of literary works that were well known
at Rome, when we find that his own great work, notwith-
standing its importance and its great merits, remained for a
long period comparatively unknown, and is not even once
cited by Pliny in the vast array of authorities which he
has brought together.

§ 7. The geographical treatise of Strabo was designed, as he
himself tells us,^ as a kind of sequel to his historical work,

' i. 3, p. 47. 2 iv. 1^ p i77_ M. 1, § 2o, p. 13.

Sect. 1. STKABO : GENERAL VIEWS. 21/

already mentioned ; and was intended for the same class of
readers, tliat is to say, for politicians and statesmen rather than
for regular students of philosophy. In modern phrase it was
meant for the general reader, and not for the mere geographer.
It is this purpose which has given to the book its peculiar
character, and to which it owes a great part of its merits. The
author has refrained from giving us long and dry catalogues
of names, such as we find in Pliny and Ptolemy; and has
endeavoured to furnish us with a general picture or descrip-
tion of each country, its character, physical peculiarities and
natural productions, as well as its geographical configuration.
The minute topographical details, and enumeration of obscure
places, belong, as he justly observes, to the chorographer,
rather than to the general geographer, and must be supplied
in each instance according to the point of view of the writer,
and the requirements of his readers. It must be admitted
that the execution of his plan has fallen far short of the
justness of its conception ; that in endeavouring to select the
more prominent and important names he has often omitted
others of at least equal interest; and not unfrequently the
scantiness of his notices probably arises in reality from his
want of knowledge. But in comparing his geographical
details with those of Pliny and Ptolemy we must always bear
in mind the essential difference in the character of their works,
and must not hastily assume that the earlier geographer was
ignorant of names of towns, rivers, or headlands, simply
because he has not thought fit to mention them.

§ 8. It is otherwise with regard to the physical geography
of the several countries described. In this respect it cannot
be doubted that the work of Strabo was a great advance u]3on
all that had preceded it, and it possesses a great superiority
over all other geographical writings that have been preserved
to us from antiquity. But its deficiencies are not the less
glaring, when tested by the requirements of modern science.
The directions of mountain chains, the courses of great rivers,
and the other natural features, which constitute the peo-

2l8 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXI.

graphical framework of every country, are indeed for tlie most
part briefly indicated, but often passed over in a very sum-
mary way, and very rarely described in anything like a regular
and systematic manner. Great allowance must be made for
defective information, and for the want of instruments with
which to make observations ; but even after admitting these
deficiencies it can hardly be denied that the work of Strabo
in this respect falls short of what we might reasonably have
expected.

§ 9. In regard to the mathematical portion of his task also it
is evident that the qualifications of Strabo were by no means of
a high order ; and there can be no doubt that in this respect he
was inferior to his predecessors Eratosthenes and Posidonius.
But as he had the advantage of availing himself of their
labours, as well as those of the great astronomer Hipparchus,
this was of comparatively little moment. His work, as he
himself repeats, was not designed for professed astronomers or
mathematicians, and the leading conclusions of those sciences
with regard to the figure and dimensions of the earth, its
relation to the heavenly bodies, and the great circles of the
globe — the equator, the ecliptic, and the tropics — were in his
day considered as so well established as to be familiar to
every one who had received a liberal education. He accepts
also the division into five zones as one generally recognized,*
though on this point there was considerable difference of
opinion among earlier writers, some dividing the torrid zone
into two, one on each side of the equator, so as to make six in
all. He quotes with approval the assertion of Hipparchus that
it was impossible to make any real progress in geography
without having recourse to astronomical observations for the
determination of latitudes and longitudes;^ and even gives us
in considerable detail the succession of the climates as observed
by that astronomer — an important contribution to our know-
ledge of ancient geography, which has been already examined

' ii. 5, § 3, p. 111. M. 1, § 12, p. 7.

Sect. 1. STKABO : GENERAL VIEWS. 219

in a preceding chapter.^ In his criticisms of Eratostlienes also
he discusses at great length the views of that writer with
regard to the latitude of Thule and the position of the northern
portions of Europe in relation to it ; and censures his errors (or
supposed errors) with regard to some other points in his map
of the world. But after having once discussed these subjects
he scarcely ever adverts to them again, and in determining
the extent and dimensions of the countries he describes, accord-
ing to the varying estimates of different authors, he never
attempts to fix them by reference to latitude and longitude.

§ 10. Strabo begins with pointing out,' as a reason for his
having undertaken anew that which had been already done by
many writers before him, that the extension of the Eoman
Empire and that of the Parthians had added largely to the
knowledge previously possessed of the inhabited world; just
in the same manner as the conquests of Alexander had done
shortly before the time of Eratosthenes. The Romans (he
says) had opened out all the western parts of Europe, as far as
the river Albis (the Elbe) which divides Germany through the
midst, and the regions beyond the Danube as far as the river
Tyras (the Dniester). The tracts on the north of the Euxine
from thence to the Palus Mseotis, and again along the eastern
coast to the borders of Colchis had been first made known by
the campaigns of Mithridates and his generals f while the
Parthians had furnished more accurate knowledge of Hyrcania,
Bactriana, and the Scythian tribes beyond those nations. It
has been already pointed out that very little additional informa-
tion had really been derived from this last source : and the
knowledge possessed by Strabo of the Scythian races either in
Europe or Asia is singularly meagre and defective.

§ 11. The first two books of his Geography constitute a kind
of general introduction to the whole, and while they are much

« See Chapter XVII., pp. 4-10, | butions of Herodotus to this portion of
' i. 2, § 1, p. 14 ; and compare ii. 5, geography ; but he appears to have
§ 12, p. 118. treated that writer with such unmerited
* It is especially singular that he contempt that he did not even in pass-
has overlooked the valuable coutri- | ing refer to his fourtli book.

220

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXI.

the most difficult portion of the work, they are at the same
time much the most unsatisfactory. A great want of order and
method reigns throughout. They comprise, or appear intended
to comprise, a historical review of the progress of geography
from the earliest days to his own time, but this is done in such
an unmethodical and irregular manner as in great measure to
fail of his object. We are indeed indebted to this part of the
work of Strabo for almost all that we know concerning the
geographical systems of his predecessors ; especially for that
of Eratosthenes. But we have already seen how imperfect
that knowledge is, Q,nd how defective are our materials for
estimating the real merits of the founder of geographical
science. Instead of giving us a systematic review of the work
of Eratosthenes, or that of any of his successors, Strabo
contents himself with criticising individual points, and dis-
cusses these at great length, often breaking off in the midst
into the discussion of collateral questions, which have no
immediate bearing on his subject. Thus, after opening his
treatise by justly claiming for the study of geography a place
among those included under the name of philosophy, he
proceeds to fortify this position by citing the names of men
distinguished as philosophers, who had given their attention
also to geography, and at the head of the list he places that
of Homer, whom he distinctly terms the founder of all geogra-
phical knowledge, and no less eminent in this respect than for
his poetical excellence and his political wisdom.^ He then
enumerates Anaximander of Miletus and Hecatseus as having
followed in the same track,^ as well as Eratosthenes, Polybius
and Posidonius in later times; after which he returns to

8 i. 1, p. 2.

^ Of these lie tells us only that
Anaximander was the first to publisli
a geographical map, while Hecatajus
had left behind him a written treatise
(ypdij.fj.a), which was believed to be his
by comparison with his other writings
(iriffTovfjievov iKeivov elvai e/c ttjs &Wris
avTov ypa((>ris, i. 1, § 11). From this

expression it is evident that doubts
had been entertained concerning the
authenticity of the worlj extant under
the nnme of Hecatasus. On this point
see Chapter V. p. 135.

In another passage (i. 1, § 1) he men-
tions DemocrituSjEudoxus, Dicaiarchus,
and Ephori;s, as having paid attention
to the study of geography.

Sect. 1. STRABO: GENEEAL VIEWS. 221

Homer, and sets forth at considerable length the proofs of his
extensive geographical knowledge — his acquaintance with the
Ocean surrounding the earth, the Ethiopians, the Nomad
Scythians to the north, &c. In the course of this exposition
he notices the view of Eratosthenes that we ought not to look
for philosophical accuracy in a poet, or to attempt to reconcile
the wanderings of Ulysses and Menelaus with the true details
of geography — a suggestion which he indignantly repudiates,
and taking up the subject again, a few pages further on, argues
against it at such length that more than half the first book
is taken up with the discussion of this subject of Homeric
geography.

§ 12. Passing over without further notice the earlier geo-
graphers — of whom it would have been interesting to have
heard more — and not condescending even to mention the name
of Herodotus in connection with this part of his subject, Strabo
comes at once to Eratosthenes, whom he censures for having
frequently relied upon untrustworthy authorities, especially
Damastes ; as well as for the doubts he had cast upon the
voyage of Jason, and those of other early navigators. He then
proceeds to discuss at considerable length the physical views
of Eratosthenes concerning the changes that had taken place
in regard to the earth's surface : especially the hypothesis, in
which he had followed the physical philosopher Straton, of the
straits connecting the Euxine with the Mediterranean and the
latter with the Atlantic having been formed by sudden dis-
ruptions, which had materially lowered the level of these
interior seas. In proof of this they appealed to the existence
of shells and other marine remains at places remote from the
sea, and even at a considerable elevation above it. But Strabo
justly rejects the theory in question, and maintains that the
phenomena referred to could be better explained by changes
in the earth's surface, such as are continually in operation, and
producing alternate subsidences and elevations of different
portions of the land.^ In proof of this he cites numerous

2 i. 3, § 5, p. 51.

222

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXI.

instances of the engulfment of towns by earthquakes, the
disappearance of islands in the sea, and the throwing up of
others, as in the case of one which had recently been thus
elevated in the neighbourhood of Thera^ and another near
Methone in Argolis. Such things as thus take place on a
small scale, he argues, might equally occur on a large one :
and not only is it possible that the Liparsean Islands and the
Pithecusse (Ischia and Procida) may have thus been thrown
up above the sea, but it is probable that Sicily itself, instead
of being broken off from the mainland (as was the general
belief in antiquity^) may have been elevated from the depths
of the sea by the fires of ^tna.^ The geological speculations
contained in this portion of his work show a soundness of view
very unusual among ancient writers on these subjects, and are
referred to with well-merited eulogy by Sir C. LyelL^

§ 13. Strabo next proceeds to examine the second book of
Eratosthenes, in which that author had laid the foundations of
his geographical system ; and discusses the length and breadth
of the inhabited world, and the division into three continents.
He severely censures him, both here and in other passages, for
having given credence to the fables of Pytheas, an author
whom he considers altogether unworthy of credit ; and he is
led in consequence to reject entirely the existence of Thule,
and the latitude assigned to it by Eratosthenes, who had taken
the parallel of Thule for the northernmost limit of the inhabited
world. Strabo on the other hand assumes it to be clearly made
out by recent investigations that lerne (Ireland), which was
situated to the north of Britain, was the farthest land in that direc-
tion, and as he supposed Britain itself to be extended lengthwise
opposite to Gaul, so that its greatest length was about 5000
stadia (500 G. miles), and its breadth considerably less, while
the island of lerne was not more than about 4000 stadia (400

M. 3, § 16, p. 57. See Note B,
p. 274.

■• This was supposed, with that per-
verse etymological ingenuity so com-
mon among the Greeks and Konians, to

be the origin of the name of Ehcgium,
the city next the stvait ('P/^yioj').

^ Ibid. § 10, p. .54.

^ Principles of Geology, vol. i. pp. 23,
24, 10th edit.

Sect. 1.

STKABO: GENERAL VIEWS.

223

miles) from the centre of Britain, he arrives at the result that
the most northern limit of the inhabited world must be brought
down very much farther to the south than the position assigned
to it by Eratosthenes. As at the same time he adopts his
southern limit — the parallel through the Cinnamon Kegion
and Taprobane — the necessary conclusion is that Eratosthenes
had greatly overrated the whole breadth of the world.

It is a striking instance of that love of system and persistent
adherence to theoretical conclusions once supposed to be estab-
lished, so characteristic of the Greeks, that Strabo, after proving,
as he conceives, the error committed by Eratosthenes in this
respect, immediately adds, that having been thus mistaken with
regard to the breadth of the known world, he was necessarily led
into error with respect to its length ; for that all the best writers
were agreed that the length was more than double the breadth.
As if the proportion between the two were not a simple matter
of fact to be determined by measurement and calculation !
It is certain indeed that Eratosthenes had started from the
same assumption, and had even made additions to the length
at each end with the express view of bringing out this result.''
Of these Strabo rejects the addition at the western extremity,
where Eratosthenes had supposed the projecting part of Europe
to extend beyond the Sacred Promontory towards the west,^
but retains that belonging to India, concerning which he had
no better information than what he derived from Eratosthenes.
His estimate of the length of the known world does not there-
fore after all differ materially from that of his predecessor.

It is in the course of this discussion that Strabo throws out
the remarkable suggestion, that besides the world known to
the Greeks and Romans, and inhabited by them, or by races
with which they were acquainted, there might be other con-
tinents or other worlds unknown to them. The length of the
Inhabited World (77 olKovfjuevrj) was, as he had shown, not more

' See Chapter XVI. p. 643.
* Posidonius, as we have seen, re-
turned to the old view that the Sacred

Promontory was the westernmost point
of Europe, and Strabo doubtless in this
instance followed his authority.

224

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXI.

than about a third part of the total circumference of the glohe
in the temperate zone ; it was therefore possible that there
might be within this space two or even more inhabited
worlds.^ But these, as he points out in another passage^
would be inhabited by different races of men, with whom the
geographer had no concern. The manner in which he in-
troduces this speculation as something possible, and even
probable,^ is a striking proof of the philosophic character of
Strabo's mind. The well-known passage in one of the tragedies
ascribed to Seneca^ is evidently derived from some such
suggestion as this, adopted and amplified by the imagination
of the poet.

§ 14. In his second book Strabo continues the examination
of the work of Eratosthenes, and discusses the various changes
introduced by him into the map of the world. Here he
judiciously takes his part in opposition to many of the attacks
of Hipparchus, especially to that preposterous distortion of
India, and the adjoining parts of Asia, which Hipparchus had
introduced anew into the geography of those regions. In
regard to the whole of Asia indeed Strabo adopted the map of
Eratosthenes with very little alteration. Little or nothing
had in fact been added to the knowledge of those countries in
the interval, which could affect the general geographical
outline. It was only with regard to the countries bordering
on the Caucasus and the Caspian that Strabo had acquired
any more detailed information than his great predecessor,
and even this was of such an imperfect character that he still
believed the Caspian to communicate with the northern ocean,
as had been asserted by Patrocles.

Equally little change was he able to introduce in the
general conception of the continent of Africa, though he

^ i. 4, § 6, p. 65. KaKovjxev yap oiKOv-
fxfvriv ^v olKOVfjLSV Kal yvccpiCofJ-eV eV5e-
Kerai 5e Kol iv ttj avTij eiiKpaTw Cwvrj Koi
Svo olKovfjLtvas elvai ?/ Kal irAeiovs.

> ii. 5, § 13, p. 118.

2 oTtep iarl -KiQavov, lie says, in the

second of the two passas^es referrred to.
3 Seneca, Medea, vv. 376-380.
Venieiit annis srecula seris,
Quibus Oceanus vincuUi verum
Laxet, et ingens pateat tcllus,
Tethysque novos detegat orbes.
Nee sit terris ultima Thule.

Sect. 1. STRABO : GENERAL VIEWS. 22$

undoubtedly possessed much more information in detail con-
cerning all the portions of that country which, had been subject
to the Carthaginians, as well as Numidia and Mauretania.
But of the western or Atlantic coast he knew nothing more :
and while he rejected the statements of Eratosthenes con-
cerning Cerne and other Carthaginian settlements on that
coast, he neglected (strangely enough) to avail himself of the
valuable new materials, which the voyage of Polybius must
certainly have furnished him.

It was principally with respect to Europe, and especially
the western and northern parts of that continent that the
knowledge possessed by Strabo was greatly in advance of that
of the Alexandrian geographer. This he has himself pointed
out to us ; but while it is perfectly true so far as relates to
the geographical details of the several countries described,
and the nations that inhabited them, he was so far from having
acquired a correct geographical idea of their position and rela-
tions, that his general map of Europe is even more faulty than
that of his predecessor.

§ 15. We are greatly indebted to the lengthened examin-
ation into which Strabo enters of the geographical positions
assumed by Eratosthenes, and the criticisms of his successor
Hipparchus, for the information thus afforded us concerning
their rival geographical systems. The results of this have
been already considered. Some interesting notices are also
introduced parenthetically in the course of the discussion.
But the discussion itself is eminently unsatisfactory, and
serves to show all the more strongly how little real progress
could be made in scientific geography so long as all accurate
observations were wanting. Strabo himself observes that not
only were there no observations of latitude — as determined by
the shadow of the gnomon, and the length of the longest
day — for any part of the mountain chain supposed to extend
across Asia from Cilicia to the frontiers of India, but that
there was the same want of accurate knowledge with regard
to the Alps, the Pyrenees, and the mountains of Thrace, lUyria,

VOL. II. Q

226 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXI.

and Germany.* Even where observations existed, they were
often so defective as to be calculated to mislead rather than
to correct, and we have seen that the great astronomer Hip-
parchus himself had been the means of introducing a grave
error, by assigning to Byzantium the same latitude as Massilia.^
The want of observations of longitude was still more complete :
and the conclusions adopted by Eratosthenes with regard to
the distances from east to west across the continent of Asia
were the result, as Strabo repeatedly tells us, of the examin-
ation and comparison of various itineraries.'' This is un-
doubtedly the only means open to the geographer under such
circumstances, but the liability to error which must always
exist in the computation of distances from itinerary routes
uncorrected by observations, was greatly increased in this case
by the want of any correct bearings.

Still more unsatisfactory is the tedious discussion that follows
of the division adopted by Eratosthenes into Sphragides or
" Seals " — a discussion which after all leaves us (as already
remarked^) almost wholly ignorant as to the purpose and
meaning of the divisions in question. It is in fact not so
much an examination of that part of the system of Eratos-
thenes, as of the objections brought against certain portions
of it by Hipparchus, and an attempt — in some instances cer-
tainly successful — to refute these objections. But such a
criticism of a criticism, where the original work is lost to us,
naturally becomes extremely obscure, and Strabo has taken no
pains to put his readers in possession of the subject matter of
the controversy. This is the more to be regretted, as the
division in consideration, which appears to have been peculiar
to Eratosthenes, was certainly of a systematic character, and
would therefore have had considerable importance in its bear-
ings on scientific geography.

§ 16. Strabo next proceeds to consider the geographical
views of Posidonius and Polybius, and in the course of this

* ii. 1, p. 71. * See Chap. XVII. p. 8. « ii. 1, pp. 69, 79.

' Chapter XVI. p. 6.54.

Sect. 1. STEABO : GENEKAL VIEWS. 22/

examination has fortunately preserved to us the account given
by the former of the voyage of Eudoxus of Cyzicus. This has
been already fully considered.^ With this exception these two
sections contain very little of any real value. A considerable
space is occupied with a discussion of the division of the
terrestrial globe into zones — a suggestion said to have origin-
ated with Parmenides, but which was developed in a more
systematic form by Aristotle. It was the latter who first
defined them in the sense in which they are understood by
modern geographers. He regarded the torrid zone as com-
prising the space on each side of the equator as far as the
tropics : and the two temperate zones as extending from the
tropics to the arctic circles.^ It would certainly seem as if
the great philosopher had here used the term "arctic circles"
in the same sense as that assigned to them by modern geogra-
phers, as two fixed and definite circles on the sphere, analogous
to the tropics. But the ancients in general used the term in a
different sense, so that every different latitude had its different
arctic circle : ^ and hence both Posidonius and Strabo agreed
in censuring Aristotle for adopting as the limit of the tempe-
rate zone a boundary that was itself fluctuating and variable.
The former writer fixed as the limit the circle where the visible
arctic circle coincided with the tropic, which is in fact the
same thing that is meant by the modern use of the term Arctic
Circle, and is probably what was really naeant by Aristotle,
however he may have expressed himself.

Polybius had departed from the established division of the
earth into five zones, and had maintained that there ought
to be six, regarding those on each side of the equator, extend-
ing from thence to the tropics, as two separate zones. This
innovation is justly rejected by Strabo, who however evidently

« Chapter XVIII. p. 75.

» Strabo, ii. 2, p. 94.

^ The term " arctic circle " was gene-
rally used by the Greeks to denote the
circle in the heavens parallel to the
equator which just touches the horizon,

and which therefore separates those
parallels which are always above, from
those which are partly above and partly
below the horizon. Of course in this
sense every different latitude had a
different arctic circle.

Q 2

228

HISTOEY OP ANCIENT GEOGEAPHY.

Chap. XXI,

failed to see that all such divisions were j)nrely arbitrary, and
merely fixed as a matter of convenience. He argues also at
considerable length against the extension of the term " torrid
zone " to the whole space comprised between the equator and
the tropics, a considerable part of which, as he points out,
from Syene south to the Land of Cinnamon, was not only
habitable, but known to be inhabited. The whole question
here arises from his insisting on the term " torrid " (StuKeKav-
fjievrj) as implying a region so burnt up with heat as to be
absolutely uninhabitable : and the only real interest in this
discussion is derived from the manner in which it illustrates
the fixed conviction of geographers in the time of Strabo, that
there was such a zone of the earth, rendered uninhabitable by
excess of heat, just as the arctic regions were by excessive
cold ; and which in consequence formed an insuperable barrier
to all exploration in that direction.

§ 17. Having thus disposed of the geographers that had pre-
ceded him, Strabo at length proceeds^ to explain the outline
of his own views, which is much the most interesting part of
his introduction. The astronomical and mathematical part of
his subject indeed he passes over very briefly, remarking that
on these subjects the geographer may content himself with
taking for granted the conclusions of physical philosophers
and mathematicians, and that he does not write for persons
unacquainted with the elements of those sciences.^ Thus he
begins with assuming that the earth is spherical, and situated
in the centre of the universe : he assumes also the division into
five zones, and the circles upon the sphere, which as he points
out have been derived from the motion of the celestial bodies —
the equator, the ecliptic or zodiac, the tropics, and the arctic
circles.* He adopts also the measurement of the earth's cir-

2 ii. 5, §2, p. no.

3 Hence, as has been already men-
tioned (Chapter XVI. p. 619), he cen-
sures Eratosthenes for dwelling at
unnece&sary length upon the proof
that the earth was a sphere : a fact

which Strabo himself assumes as gene-
rally admitted.

^ It is remarkable that he here uses
the term " arctic circles " as something
fixed and definite, just as modern geo-
graphers do. He evidently employs

Sect. 1. STEABO : GENERAL VIEWS. 229

CTimference, as determined by Eratosthenes : and consequently
liis division of each great circle into sixtieth parts, each con-
taining 4200 stadia, which is equivalent to reckoning 700 stadia
to a degree.® He then points out that the whole of the habit-
able world (rj oLKov/juevr]), with which alone the geographer has
to deal, is comprised within a portion of the globe bounded by
two parallels of latitude, and two meridians of longitude, so as to
constitute a quadrilateral space within the northern hemisphere,
nearly coinciding with the temperate zone of that hemisphere,
but occupying little more than a third of its whole extent.*'

§ 18. The form of the habitable world he compares to that
of a cloak (chlamys), a comparison which appears to have been
generally adopted in his time,^ on account of its upper or
northern portions being supposed to be much more contracted,
while it spread out in proceeding southwards. Its greatest
length he estimates at 70,000 stadia, and its breadth at less
than 30,000. The great diminution that he introduces into
this last dimension proceeds from two causes : first, that, as
already stated, he discards altogether the existence of Thule,
or of any habitable land so far north as the Arctic Circle, and
regards Ireland as the most northerly of all known lands : next,
that he rejects the latitude assigned by former geographers to
Massilia, and brings it down much farther to the south than
its true position. We have seen that Eratosthenes, as well as
Hipparchus and other geographers, agreed in placing Massilia
and Byzantium on the same parallel of latitude — a gross error,
inasmuch as the former city lies more than two degrees to the

the term as equivalent to what he calls
elsewhere " the drcle that bounds the
frigid zone." kvkKos aWos tovtw ttu-
pdWr)\os bpl^wy ttjj/ Karetf/vjfx^vyjv iv
Tcfi Popeicp rifj.i(T(paiplcii, p. 112. It is evi-
dently tlie same as he elsewhere (p.
114) refers to as that ottov b Oepivhs rpo-
■KiKhs apKTiKhs ylyerai, which is, in fact,
tlie same with what we call in modern
usage the Arctic Circle.

^ Strabo, ii. 5, § 7, p. 113.

" Ibid. § 6. He here applies to the
quadrilateral space thus measured off as

the boundary or frame enclosing the
inhabited world, the descriptive epithet
of aTr6i/dv\os : a term used for the
weight employed in spinning, and for
other circular and conical bodies ; but
which would seem hardly suitable to a
definite jjortion cut off from a conoidal
surface.

' 5j d'oiKov/iievr] xAa^wSoeiST/s ej/ rovrai
vTJaos. Ibid. Earlier writers, especially
PosidoniuB, had compared tlie form of
the inhabited world to a slins:.

230 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXI.

north of the latter. But Strabo, while rejecting 'the observa-
tions on which this conclusion was founded, fell into the
strange mistake of bringing down Massilia still farther to the
south, so as actually to place it as much to the south of Byzan-
tium as it really is to the north. Of course the effect of this
error is to distort, to a strange extent, the whole map of the
Mediterranean. But its influence upon the portion of the map
of Europe to the north is not less unfortunate. As Massilia
was a kind of cardinal position from whence he measured the
breadth of Gaul across the continent to the Northern Ocean,
the effect was to bring down the northern coast of Gaul and the
mouths of the Rhine to the same latitude with the Euxine and
the mouths of the Danube ! As at the same time he adhered to
the position erroneously ascribed to Byzantium by Hipparchus,
and to the received notion that the mouth of the Borysthenes
was nearly due north from Byzantium, he placed the mouth of
that river (which is really situated in about 46J degrees of
N. latitude) in the same parallel with the northern extremity
of Britain, and supposed the Eoxolani, who in his time
inhabited the tracts adjoining the Borysthenes and the Palus
Meeotis, to be, in common with the inhabitants of lerne
(Ireland) the most northerly people in the known world.^

So far therefore was positive geography from having kept
pace with the increased knowledge of nations and countries
which had been undoubtedly acquired in the interval of two
centuries from Eratosthenes to Strabo, that it had actually
receded ; and a comparison of the maps drawn according to the
two systems will show that that of the older geographer was, in
regard to Europe in general, and especially the basin of the
Mediterranean Sea, a much nearer approximation to the truth
than that of his successor.

§ 19. The reasoning by which Strabo is led to this unfor-
tunate conclusion is a striking proof of the vagueness of the
data on which geographers were accustomed to rely, in the

« ii. 5, § 7, \u ll'i-

Sect. 1.

STEABO : GENERAL VIEWS.

231

absence of trustworthy observations. Taking for granted the
correctness of the parallel of latitude drawn by Eratosthenes
through the whole length of the Mediterranean, from the Strait
of the Columns to the Gulf of Issus, which, as we have seen, was
assumed to pass through the Sicilian Strait and the Island of
Ehodes,^ he says that " it is generally agreed " that the course
from the Columns to the Sicilian Strait lies "through the
middle of the sea." Navigators were also " generally agreed "
that the greatest width of the sea from the bight of the
Gaulish Gulf to the African coast did not exceed 5000 stadia.
Massilia therefore, which was somewhat to the south of the
inmost recess of the gulf, must be less than half this distance
(2500 stadia) from the parallel in question. But the distance
from Rhodes to Byzantium is not less than 4900 stadia : the
latter city, therefore, must be situated far to the north of
Massilia.-^

It is hardly necessary to point out in detail the complication
of errors involved in this argument. The main point is the
assumption that the voyage from the Columns to the Sicilian
Strait was a direct course from west to east, and lay through the
middle of the sea in so literal a sense that it was equidistant
from the shores of Europe and Africa. This great error was
undoubtedly combined in Strabo's mind with another, which
he held in common with all his predecessors, that the north
coast of Africa did not deviate very widely from a straight line,
instead of advancing, as it really does, so far to the north
between the Strait of the Columns and Cape Bon, as to render
it impossible in fact to hold a straight course eastward from
the Straits.

§ 20. He considers the habitable portion of the world to
extend 4000 stadia (400 G. miles) to the north of the mouth
of the Borysthenes — •mea.ning probably the city of that name ^
— and this parallel he conceives to pass to the north of

See Chapter XVI. p. 629.

ii. 5, § 8, p. 115.

See Chapter XVI. p. 631.

It is a

striking instance of the vagueness with
which all these discussioiis and calcu-
lations were carried on, that Strabo,

232 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXI.

lerne. That island he describes as "barely habitable on
account of the cold,^ and he applies the same remark to the
Scythians inhabiting the regions farthest known to the north
of the Borysthenes. The great cold of these countries, which
was familiar to the Greeks from the time their colonies first
settled to the north of the Euxine, naturally led to the belief
of their being situated much farther north than they really
are, and tended strongly to confirm the erroneous idea of their
position derived from mistaken observations.

The southern limit he places about 3000 stadia beyond
Meroe, or 8000 from Syene, which he takes as situated on the
tropic. This line he considers as passing through the land of
the Automoli or Sembritse,* and the Land of Cinnamon ; and
regards the lands farther south as uninhabitable from excess
of heat. In regard to this limit therefore he followed Era-
tosthenes, though for some reason, which is not very clearly
explained, he placed it at 8800 stadia from the equator instead
of 8300, which was the distance assigned by the older geo-
grapher.^

§ 21. Having thus considered the position and extent of the
habitable world with reference to the terrestrial globe, Strabo
proceeds to show how a map of it is to be laid down.^ This, as
he points out, is a matter of no difficulty upon a globe, where
the curved lines of the parallels of latitude and meridians of
longitude, are represented by similar curves. But as such an
artificial globe, in order to admit of the countries being repre-
sented in sufficient detail, would require to be not less than
ten feet in diameter,'^ and after all, but a small part of it would
be occupied by the geographical representation of known lands.

like Eratosthenes, always contents
himself with using the expression " the
Borysthenes," without defining more
accurately the point referred to.

^ a.d\iws Sta ^vxos o'lKovfjiivriv, ii. 1,
p. 72. Agnin in another passage (p.
]15) he speaks of it as occupied by
wypiwv re\i(iis avOpunroiv Koi KaKws ol-

* See Chapter XVI. Note 0, p. 664.

« ii. 5, § 10, p. 116.

' It appears that a globe of this size
had actually been constructed by a
writer named Crates — probably the
Stoic philosopher of the name, better
known for his commentaries on Homer
and Hesiod — to which Strabo refers as

Hovpraiv 5ia i|/uxos. to something well known ((r(pa(pav «a-

^ ii. 5, § 14, p. 118, I 9a.nep ti-jv KpaTriTiiou, I. c).

Sect. 1. STEABO : GENERAL VIEWS. 233

few people could procure one, and in general they must be
content with a map on a plane surface. In this case, as he
justly observes, all the lines that are circles on the globe must
be represented by straight lines, parallel with the equator and
the meridian respectively ; a proceeding inevitably productive
of error, but which as he conceives " will not make much dif-
ference ; " ^ an assertion somewhat startling to modern geo-
graphers, who are accustomed to have recourse to a variety of
ingenious contrivances to avoid, or at least diminish, the error
resulting from such a process, but which was not far from the
truth in Strabo's time. For the greatest error arising from
such a mode of plane projection (to use the modern phrase),
would really be trifling as compared with those resulting from
erroneous estimates of distance, and the want of any accurate
observations of latitude and longitude. He therefore proceeds,
as he expressly tells us,^ throughout the rest of his work, to
consider the countries as represented on a plane surface in the
manner above described. The error of the result would be
indeed comparatively unimportant in separate maps of each
country, and would only assume any prominence in the general
map of the Inhabited World. ^ Even in regard to this, if the
parallel of 36° were assumed (as it certainly was by Strabo, in
imitation of Eratosthenes) as the fundamental parallel, with
reference to which the map was constructed, the error would
have comparatively little influence upon the countries border-
ing on the Mediterranean ; while with regard to the more
northern parts of Europe, the knowledge of them was as yet
much too imperfect for such a cause to produce any appreciable
effect.

^ Stoimi yap fiiKpov, iay avrl tUv kvk-
Xoiv, tSiv re TrapaWriXcov /cat rHv fiearip.-
fipLVWV . . . evdeiaa ypoLKfyco/xev, I. c. He
afterwards suggests that the meridian
lines might be drawn " a little inclined
towards one another ; " but again adds
that it is of little consequence.

» Ibid. § 12, p. 117.

' It would be the more conspicuous
in this case, if, as Strabo himself recom-

mends, such a map were not less than
seven feet in length. (I. c.)

As we shall hereafter see, even
Ptolemy, while giving an elaborate
mode of projection for his general map,
was contented in his special maps of
countries with laying down his parallels
of latitude and longitude as straight
lines crossing one another at right
aniarles.

234

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXI.

§ 22. Before proceeding to describe the different parts of the
world in detail, Strabo gives a general outline of the whole,
which is instructive and clearly expressed. The "Inhabited
World " he considered, in common with all preceding geo-
graphers, except Hipparchus, to be a vast island, surrounded
on all sides by the ocean, of which the Caspian Sea, as well as
the Persian and Arabian Gulfs, were inlets or arms. But by
far the most important and extensive of these inlets, as he tells
us, was that which extended from the Columns of Hercules on
the west to the Euxine and the Gulf of Issus on the east. For
this sea, — so familiar to modern geographers by the name of
the Mediterranean,— the Greeks had no distinctive name,
because it had so long been practically the only one known to
them ; and Strabo can only distinguish it as " the Inner " or
" Our " Sea.^ But he was fully alive to its importance in a
geographical point of view, as affording the key to the con-
formation and arrangement of all the countries around its
shores, and giving rise by its numerous arms and inlets to the
broken and irregular character for which Europe was dis-
tinguished from Asia and Africa, and which he justly regards
as one of the principal causes of its superior civilization and
political importance.^ He accordingly proceeds to describe at
considerable length the form and dimensions of this sea, as
well as of its subordinate portions, the Adriatic, ^gean, Pro-
pontis and Euxine. We have already seen how erroneous
were his conceptions of the general form of the Mediterranean
especially in regard to its breadth, and the relative position of
the coasts of Gaul and Africa. With respect to its length he
was better informed ; he reckoned 12,000 stadia from the
Columns to the Sicilian Strait, 9000 from thence to the coast
of Caria opposite to Ehodes, and 5000 from thence to the head

^ T) eprhs Kol Kad' rifias Xejo/xivr) 6d-
Aarra, ii. 5, § 18, p. 121. la the sub-
sequent exposition he never calls it
otherwise than ^ Kad' 'o/xas OdAaTra,
even the appellation of f] ivrhs QaKarTa,
corresponding to the Internum Mare of

the Eomans, not having apparently
passed into established use, as was the
case with the Latin appellation.

^ Ibid. p. 122. irphs 'diravTa Se rd
TOLadra, i>s e<)i?7J', V nap' tj/xlv (JaAarra
Tr\eoveKTriiJ,a ex^' fteya.

Sect. 1. STEABO : GENERAL VIEWS. 235'

of the Gulf of Issus. This gives 26,000 stadia for the whole
length of the Mediterranean; a computation in which he
appears to have followed Eratosthenes, and which, as we have
seen, differs from the truth by little more than a fifth.*

Many others of his distances present in like manner a very
fair approximation to the truth, though almost invariably
erring more or less on the side of excess. The fact appears
to be, that deficient as the ancients were in any mode of
measuring distances at sea, the rough estimates formed by
navigators came to be a reasonable approach to the truth
where the distances were habitually traversed. Hence their
approximate correctness forms a striking contrast with the
strange misconceptions entertained even by such a geographer
as Strabo of the relative position and configuration of the
countries that surrounded the Mediterranean.

§ 23. The same remark applies still more strongly to his
description of the countries themselves. Even with regard
to such provinces as Spain and Gaul, which he himself
describes as being well known to the geographer in his time,
since they had been brought under the dominion of Eome, his
geographical conceptions were strangely wide of the truth.
Thus he describes the Pyrenees as forming a continuous chain
from one sea to the other, in a line from north to south : and
having their direction parallel to that of the Ehine (!) ^ Again, .
he not only supposed the Sacred Promontory (Cape St. Vin-
cent) to be the most westerly point of Europe, ignoring
altogether the manner in which the coast of Portugal projects
to the westward, near the mouth of the Tagus : but he was
equally ignorant, or negligent, of the still greater projection
of the coast of Gaul between the mouths of the Seine and
Loire, forming the great Armorican peninsula, and he describes

* See Chapter XVI. p. 634.

* In defining the boundaries of Gaul
he tells tis that it was limited on the
west by the Pyrenees, and on the east
by the Ehine, which was parallel with
the Pyrenees (aTrb fiev Svaews opi^ei ra

Tlvprjvaca oprj, Trpo(raTrr6iJ.€va rrjs eKare-
pQidey OaAdrrrjs . . . airh Se raiv avaro-
Xwv 6 'Prjvos TrapdWrjXos &>v rp Tlvprivr},
iv. 1, p. 177. He again repeats the
same statement, ii. 5, p. 128.

236

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXL

the Gaulish. Gulf on the Ocean (the Bay of Biscay) " as looking
towards the north and towards Britain." He appears indeed
to have conceived the northern coasts of Gaul as preserving
much the same general direction from the Pyrenees to the
mouth of the Rhine, and that the four great rivers, the
Garumna (Garonne), the Liger (Loire), the Sequana (Seine) and
the Rhine, pursued parallel courses from S. to N.^ The mouths
of these rivers he describes as being all of them opposite to
Britain, and consequently affording the most convenient places
of passage to that island.'^ Britain itself was a great triangle,
having its longest side opposite to Gaul, and extending about
5000 stadia (500 G. miles) in length, so that its south-eastern
promontory of Cantium, or Kent, was opposite to the mouth
of the Rhine, and the south-western to Aquitania and the
Pyrenees.®

§ 24. Even his ideas of the geographical position of Italy
and Sicily were still very imperfect. Though in one passage
he describes Italy, according to the popular notion, as ex-
tending from north to south,^ it is evident that in accordance
with his views concerning the position of Massilia and the
northern shores of the Mediterranean, he could not have found
space for it,^ without extending it much more to the east
than it really does, so that he must have given it an elongated
form, somewhat similar to that which it assumed on the Ptole-
maic maps, but extending still more from west to east, and
then curving round at its extremity, so as to descend to the
Sicilian Strait. His notion of Sicily was still more erroneous.
He was of course well aware of the triangular form of the
island — an idea familiar to the Greeks from the earliest period —
but he supposed the coast from the promontory of Pelorus
adjoining the Sicilian Strait to that of Pachynus to have a
general direction from east, to west instead of from north to

® This he expressly states with re-
gard to the Garoune and the Loire,
that they were both parallel with the
Pyrenees (iv. 2, p. 190), and again (p.
192) tliat the Seine was parallel with

the Rhine.
' iv. 5, p. 199.

* Ibid., and see ii. 5, p. 128.
= ii. 5, § 28, p. 128.

Sect. 1,

STEABO : GENEEAL VIEWS.

237

south, while that of Lilybseum was brought down much to the
south of Pachynus, in order to bring it within about the real
distance of the coast of Africa. The interval between the two
he estimates at 1500 stadia, which considerably exceeds the
truth, but not more than do most of his maritime distances.-^
This was of course a passage frequently made, and no great
error could exist with regard to it : but as Strabo had an
entirely erroneous notion of the latitude of Carthage, which he
placed more than 5^ degrees to the south of its true position,
he was compelled altogether to distort the form of the island
in order to make it fit with his assumed hypothesis, and at the
same time with the well-known fact of its proximity to the
coast of Africa.

He must have found himself in a somewhat similar difficulty
with regard to the islands of Sardinia and Corsica, the dimen-
sions of which he gives with tolerable accuracy, and correctly
tells us that they extended in a line towards the south and
Africa.^ At the same time he estimates the distance from the
southern point of Sardinia to the coast of Africa at 2400 stadia,
a statement greatly exceeding the truth,^ and which, if it be
added to the measurements given of the two islands and the
intervening strait, would give a result considerably greater
than the whole breadth which he has assumed for the Medi-
terranean. It became therefore necessary to place the two
islands in a direction deviating materially from that which he
has himself stated them to occupy.

^ He here (vi. p. 267) speaks of this
distance as rovXaxiTTov Siapfj.a iirl
Aifiv7]v, but adds the words irepl Kapxi)-
S6va ; and it is to be suspected that he
was really thinking not of the shortest
passage to Caj>e Bon, the nearest point
of Africa (the distance of which from
Lilybfeum is less than 80 G. railes),
but of that to Cartilage itself, which is
nearly 40 miles further.

^ iirifiriKeLS S'elal koI TrapaAA.7jAoi <TX^'
Sov at Tpe'ts, irrl votov kou Ai^vrjv rerpa^-
jxivai, V. 2, § 8. But of the three islam is
here mentioned, one is iEthalia (Elba),

which cannot, with any reasonable
allowance for vagueness of expression,
be described as parallel to the other
two.

^ The error is in this instance the
more surprising as the distance was
already much more correctly given by
the early geographer Scylax, who states
that it was a day and a night's voyage,
equivalent, accordin'j: to his mode of
computation, to 1000 stadia, an esti-
mate very near tlie truth. (Scylax,
Periplus, § 7.) See Chapter XI. p. 387,
note.

238

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXI.

§ 25. So difficult is it to reconcile the different statements of
Strabo with one another, and to represent the Mediterranean
and the countries bordering its shores in the positions which he
appears to have conceived them as occupying, that it has been
supposed he did not himself attempt to represent them on a
map, and was content to adopt the map of the world as settled
by Eratosthenes, pointing out and correcting certain errors
in it by verbal criticism. But this seems most improbable,
especially when we consider the manner in which he gives
directions for the construction of such a maj?, so as to enable
any of his readers to frame one for themselves. It is far more
natural to suppose that he drew out a map of the world, based
for the most part upon that of Eratosthenes (from whose system
as a whole he did not deviate widely), but introducing such
corrections and alterations as he deemed necessary, where he
had, or supposed that he had, better information than the older
geographer. But in so doing where he found the data fur-
nished him by previous authors at variance, or even contra-
dictory with one another, he would probably settle the matter
in a summary manner by some compromise satisfactory to his
own mind.

How little indeed he aimed at anything like geographical
accuracy, and how vague were the statements upon which he
had to rely can hardly be better shown than by taking an
island so well known as that of Crete, the position of which
might be supposed to be familiar to all. He tells ns, first, that
according to Sosicrates, who was considered by ApoUodorus
to be the most accurate writer concerning the island, it was
more than 2300 stadia in length : while Hieronymus gave the
length as only 2000 stadia, and Artemidorus, who reckoned the
circumference as only 4100 stadia, must have made the length
much less.* Again, in regard to its position with relation to

* Strabo, x. pp. 474, 475. All these
statements greatly exceed the truth.
The actual length of the island, as
measured on a map in a direct line,
does not exceed 140 G. miles, or 1400
stadia. The estimates cited by Strabo

were doubtless founded on itinerary
measurements, which would have been
particularly fallacious in the case of an
island of so rugged and moimtainous a
character.

Tin: woiJiJ) ACcoKDixc; to strabo.

ndoo - .Ioh.1 Ml

Sect. 2.

STRABO: EUROPE.

239

the surrounding countries, he tells us that its south-western
extremity, Cape Kriu Metopon, was distant from the Cyrenaica
two days and two nights' voyage, while Cape Sammonium, the
eastern promontory, was four days and four nights' voyage from
Egypt, hut others said only three (!). This distance was
reckoned by some at 5000 stadia, hy others at less (!).^ He
however tells us more definitely in one place that Cape
Cimarus, the N.W. promontory of the island, was 700 stadia
from Cape Malea, in another that Cape Sammonium was
1000 stadia from Ehodes.^ Here he appears to have been
following some better authorities and these two last statements
are not far from the truth : and yet so little real idea had he
of the true position of the island that he says, its eastern pro-
montory. Cape Sammonium, does not project far to the east of
Sunium (!).'^ It is evident either that he never attempted to
reconcile these varying and conflicting statements so as to
represent his own geographical ideas upon a map, or that he
solved the difficulties thus arising by some process of which
he has left us no indication.

Section 2. — Descriptive Geograjphy. — Europe.

§ 1. In his third book Strabo commences the particular
description of the different countries of Europe, beginning
with Spain, to which the whole of this third book is devoted.
His description of the Iberian peninsula is marked at once
by the chief merits as well as the chief defects that characterize
his work in general. We have already seen how imperfect
was his idea of its geographical form and position, and how
distorted his conception of its appearance on a map. But he

= Strabo, x. p. 475.

« Ibid. p. 475; ii. 4, p. 106.

' Ibid. p. 474, virepTrnrTov tov "Sovviov
oi) iro\v TTphs eco. The difference in
longitude between Ihe two points really

exceeds 21 degrees. It would have
been nearer the truth to have said that
the ivestern extremity of Crete was not
far to the west of Sunium,

240 HISTOEY OF ANCIENT GEOGKAPHY. Chap. XXL

was well acquainted with, its leading geographical features : the
great rivers that traversed it from east to west, the Beetis
(Guadalquivir), the Anas (Guadiana), the Tagus, the Durius
(Douro), and the Minius (Minho) : as well as the Iberus or
Ebro, which however he considered as having its course
parallel with the chain of the Pyrenees, and consequently
flowing from N. to S. On the other side of the valley of the
Ebro, and parallel with the Pyrenees, was a chain of mountains
to which he gives the name of Idubeda, and which he describes
as containing the sources of the Tagus and Durius. From the
middle of this range branches off another called Orospeda
which trends to the westward, and ultimately takes a turn to
the south. Beginning at first with hills of moderate elevation,
it gradually rises in height till it joins the range that separates
the valley of the Bsetis from the coast near Malaca (the Sierra
Nevada), which he regarded as the main continuation of this
central chain, while other parallel ridges on the north side of
the Bsetis contained the mines for which Spain was so famous.
The Anas and the Bsetis had their sources near one another
in the range of Orospeda: they are correctly described as
flowing at first to the west and then turning off more towards
the south. Imperfect as is this outline of the physical geo-
graphy of Spain, it shows a general acquaintance with the
leading features of the country, and a correct appreciation of
the manner in which those features determine the character
and conformation of its different regions.

§ 2. The whole of the northern part of the peninsula,
adjoining the Ocean, he correctly describes as occupied by
a tract of mountainous country, extending from the headland
of Nerium (Cape Finisterre) to the extremity of the Pyrenees :
and the nations inhabiting this quarter, the Callaici, Astures
and Cantabri, which had but lately been brought under the
dominion of Eome, were still lawless and predatory tribes,
living in a semi-barbarous condition. The account given of
their habits of life and customs, which must have been taken
by Strabo from previous writers, may probably refer to a period

Sect. 2. STRABO : EUROPE. 24 1

somewhat earlier than that at which lie wrote, but it is at all
events curious and interesting. Some of their peculiarities
were indeed, as he himself remarks, common also to the Gauls
as well as to the Thracians and Scythians,^ and were probably-
inherent in their mode of life and the stage of semi-civilization
in which they found themselves, rather than belonging to them
as a race. The Lusitanians on the west, from the promontory
of the Artabri to the mouth of the Tagus, partook to a great
extent of the same characteristics, even the inhabitants of the
plains and fertile districts having gradually been compelled
by the continued incursions of their ruder neighbours to adopt
their warlike and desultory habits : but the inhabitants of the
Hither province, as it had long been called,^ occupying the
eastern portion of the peninsula, were in a much more civilized
state, and even the Celtiberian tribes of the interior, which
had cost the Komans such repeated and long continued efforts
to subdue them, were gradually settling down under the in-
fluence of Koman civilization and of the numerous Roman
colonies that had been established among them. The province
of Bsetica on the other hand, which was occupied principally
by the Turdetani in the valley of the Bsetis, and the Bastelani
between them and the sea coast, was not only completely
tranquil and civilized, but had become Romanized to such an
extent as to have almost entirely laid aside the use of the
native language, and adopted Latin in its stead.^

This result was mainly owing to the great natural fertility of
the country. Strabo can indeed hardly find words to express his
admiration of the richness of Turdetania, the modern Andalusia,
which had from the earliest times been proverbial for its wealth,
under the name of Tartessus,^ and had continued to enjoy the

^ iii. 4, p. 165.

^ The distinction between the two
provinces had been established from an
early period, and still subsisted in the
time of Strabo (iii. 4, p. 166), though,
as he observes, the political limits of
the divisions fluctuated from time to
time.

> Strabo, iii. 2, p. 151.

^ See the well-known passage of
Herodotus (i. 163), and those quoted
from other writers by Strabo (iii. 2,
§ V6, pp. 150, 151). In Strabo's time
the name had become quite obsolete,
and he himself points out its fluctuating
and uncertain use by earlier writers,

VOL. II. R

242 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXI.

same pre-eminence under the Phoenicians, Carthaginians, and
Eomans. It not only produced corn, wine, and oil in great
abundance, but wool of first-rate excellence,^ honey, wax, pitch,
kermes, and vermilion (cinnabar) ; while the sea-coast furnished
salt-fish in quantities equal to that of the Euxine. The mouths
of the rivers and the estuaries formed by the action of the tides
gave peculiar advantage for the export of these various com-
modities : hence an active and constant trade was carried on,
and the ships of Turdetania that sailed from thence to Dicee-
archia and Ostia — the two ports of Rome— were the largest of
all that were seen in those great centres of commerce.*

But in addition to all these varied sources of wealth, Strabo
dwells above all upon the extraordinary mineral riches of this
favoured tract. In this respect indeed the south of Spain
enjoyed a reputation in ancient times similar to that of Mexico
or Peru down to our own day. Gold, silver, brass (copper),
and iron were found in quantities, as well as of a quality,
unsurpassed in any other part of the world. Gold was not only
obtained by digging, but by simple washing. The other metals
were all derived from mines ; and these were worked prin-
cipally in the mountains near the sources of the Bsetis, and
extending from thence towards New Carthage : the most
valuable of all the silver mines being in the immediate neigh-
bourhood of that city. In the time of Polybius these had
given employment to 40,000 workmen, and were said to have
yielded 25,000 drachms (about £900) a day ; but in Strabo's
time the mines had passed into the hands of private persons,
and the produce had apparently fallen off.^

some of whom applied it to the whole i * Strabo, iii. 2, §§ 8-10, p. 146. It

country, some to a town, some to a , is amusing to find him noticing among

river. The last was unquestionably j other advantages of Turdetania, its

the same with the Bsetis or Guadal- j freedom from all destructive wild beasts

quivir. j {jSiv oKidpioiv driplcov), except rabbits (!),

^ So highly was the wool of this I which abounded so much in all parts

part of Spain valued, that, as he assures | of Spain as to do great damage to the

us, rams for breeding purposes had ', crops. ' They were killed by means of

been known to fetch as much as a talent
each.

■• iii. 2, § G, p. 145.

ferrets {yaAas ayptds), the use of which
he describes exactly as it is practised
at the present day (lb. § G). He relates

Sect. 2.

STRABO : EUROPE.

243

§ 3. It is remarkable, that throughout his description both
of the natural productions and physical peculiarities of Spain,
and of the manners and customs of its inhabitants, Strabo
appears to have relied almost exclusively upon Greek autho-
rities, his statements being derived principally from Polybius,
Artemidorus, and Posidonius. He indeed speaks in one
passage^ in very disparaging terms of the Eoman writers in
general, whom he accuses of doing little but copy the Greeks ;
but it seems impossible that their historians, in relating their
long-continued wars with the Spaniards, should not have con-
tributed many facts to the geography of the country. The
construction of roads in all directions through Spain, and the
itineraries which must certainly have existed in his day of the
stations and distances along these, would also have furnished
most valuable materials to a geographer that was able to appre-
ciate them. But no attempt is made by Strabo to turn to
account these sources of information. The only instance in
which he especially refers to the Eoman campaigns is that of
D. Brutus Callaicus against the Lusitanians, and the particulars
of this he probably learnt from Polybius.^ Even where he
adverts to the construction by the Eoman s of a great highway
from the Pyrenees through Tarraco and Saguntum to the
frontiers of Bsetica, and thence on to Corduba and Gades, he
gives no account of the distances ; and contents himself with
telling us that Julius Csesar accomplished the journey from
Eome to his camp at Obulco on his way to Munda in twenty-
seven days.^

In his enumeration of the names of towns and of the native

on this occasion the well-known story
of the inhabitants of the Gymnesian
Islands (Iviza and Formentera), having
sent a deputation to the Eoman govern-
ment to represent that they were abso-
lutely driven out of their liomes by the
multitude of these animals. What hap-
pened in the island of Porto Santo
after its discovery by the Portuguese
shows that this may have been no
exaggeration.

« Ibid. 4, p. 166.

' iii. 3, p. 152. The campaigns of
D. Brutus against the Lusitanians and
Callaici or GallsBcians, by which he
earned the surname of Callaicus, took
place in e.g. 138-135, a few years only
before Polybius was in Spain with his
friend Scipio during the Numantine
War.

« iii. 4, § 9, p. 160.

E 2

244

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXI.

tribes in Spain, Strabo has made a judicious selection, and
must have followed good authorities, as almost all the names
he mentions are well known from other sources, and must have
been places of some importance. At the same iime he avoids
the error into which Pliny and Ptolemy subsequently fell, of
loading their pages with obscure and insignificant names. He
indeed adds some judicious remarks^ on the proneness of
geographical as well as historical writers to bestow the title of
towns and cities on places that were, in fact, mere villages. It
was thus that some writers asserted that there were more than
a thousand cities {irokei'i) in Spain ; and even Polybius affirmed
that Tiberius Gracchus took or destroyed three hundred cities
in Celtiberia alone. This exaggeration, as he points out, was
the more inexcusable in the case of Spain, as the inhabitants of
the interior for the most part lived wholly in villages, and the
barren and rugged character of the country was ill adapted to
the formation of towns.^

§ 4. In the last section of the third book Strabo treats of the
islands adjacent to Spain, and describes at some length the
Balearic Islands, as well as the neighbouring Pityusse, both of
which were in his day well known: the former especially
having received two Roman colonies. He then gives a long
account of Gadeira (Gades), which was still at this period one
of the most important emporia of commerce in the world ; and
enters into somewhat idle disquisitions as to its relations with
the fabulous island of Erytheia, the abode of king Geryones.
In conclusion, he mentions the celebrated Cassiterides, which
he describes as ten in number, lying close together, but far out
to sea to the north of the port of the Artabri, from which they
were separated by a wider extent of sea than that between Gaul

« Strabo, iii. 4, § 13, p. 163.

' It seems strange to us at first to
find the broad corn-growing plains of
Old and New Castile included in this
character, as wild and barren regions,
supporting but a scanty population.
Bur, their inclement climate was alone
sufficient to produce this impression to

a Greek or Italian, and even at the
piesent day a recent writer speaks of
" the trackless, lonely, wind-blown
plains " of Castile as much exposed to
drought, notwithstanding the fertility
of the soil, and thinly-peopled in con-
sequence.

Sect. 2. STRABO : EUROPF. 245

and Britain.^ The inhabitants are described as wearing long
black garments, and walking about with long wands in their
hands, looking like the Furies of tragedy. They traded in tin
and lead, in exchange for which they received pottery, salt, and
bronze vessels. The trade with these islands had for a long
time been confined to the Phoenicians from Gades, but had
been opened out to the Eomans by P. Crassus, who visited
them in person, and from that time the intercourse was carried
on briskly.^

It is remarkable that he says nothing, either here or else-
where, of the proximity of the Tin Islands to Britain : * he
seems to have regarded them only with reference to Spain, and
in connexion with Gades, from whence the trade with them had
originally been carried on.

§ 5. The fourth book is devoted to Gaul, Britain and the
Alps. His ideas concerning the form and position of Gaul
have already been explained, and we have seen how widely
they departed from the truth. But erroneous as were his
notions in a strictly geographical sense, he was, as in the case
of Spain, well acquainted with the general character of the
country, the nations that inhabited it, and the main geogra-
phical features that determined its conformation. Besides the
Alps and Pyrenees he describes the Cemmenus (Cevennes)
as a chain of mountains, branching off from the Pyrenees, at
right angles, and extending to the centre of Gaul, where it
gradually sank into the plain. He mentions also the Jura,

2 iii. 5, § 11, p. 175. at Se Karrne-
p'lSes SfKa jj-ep eiai, Keivrai S' iyyvs aWij-
\u>v Trphs (xpKTOv oiTrb rov ra>v 'AprajSpco)/
AifjLevos ir^Kdyiai..

* Ibid. p. 176. It can scarcely be
doubted tbat this Publius Crassus is
tlie same as the lieutenant of Csesar,
who subdued the Armorican tribes in
Gaul, and visited the shores of the
Western Ocean (Osesar, B. G. ii. 34) ;
but it is strange that if Strabo had
access to the information which he col-

Britain.

* It may be remembered that no men-
tion of the Cassiterides is found in
Csesar. In another passage indeed (ii.
5, p. 120), Strabo speaks of them as
lying in the open sea north of the
Artabri, "in about the same latitude
as Britain : " vrjffoi ireXdyiai, Kara, rb
BpeTTavLKov irais Kkijia iSpvixivai, the
south-western angle of which he re-
garded as facing the Pyrenees ! But
this very phrase evidently excludes

lected, he should still connect the Cas- ' the supposition that they were in its'
siteridcs with Spain, instead of (^nn\ or immediate vicinity.

246 HISTORY OF ANCIENT GEOGBAPHY. Chap. XXI.

under the name of lourasios, and describes it as separating the
Helvetii from the Sequani, who inhabited the region known in
modern times as Franche Comte.^ With the Khone and its
tributaries he was well acquainted, and describes very cor-
rectly the confluence of the Dubis (Doubs) with the Arar
(Saone), and that of the latter river with the Rhone, but he
erroneously supposed both the Arar and the Dubis — as well as
the Sequana (Seine)— to take their rise in the Alps — showing
how vague was his knowledge of the relations of the different
mountain-chains in this part of Graul.® He was familiar also,
as already mentioned, with the great rivers that flowed into
the Ocean — the Garonne, Loire, and Seine — all of which he
conceived to flow, in a general way, from south to north,
parallel with the Rhine and the Pyrenees. And he was fully
alive to the remarkable advantages derived by Gaul from the
facilities of internal communication afforded by these rivers,
which approached so near to each other that a very short
passage over land was needed from the Saone to the Seine on
the one hand, and from the Rhone to the Loire on the otherJ
These facilities were in his time turned extensively to account :
and a flourishing transit trade was carried on from the ports on
the Ocean to those of Narbo and Massilia on the Mediter-
ranean. Burdigala (Bordeaux) at the mouth of the Garonne
was already an important emporium of trade. The names of
the sea ports at the mouths of the other two great rivers, the
Loire and the Seine, Strabo has unfortunately omitted to men-
tion. Corbilo, which had formerly been the chief port on the
Loire,^ had in his time ceased to exist.

§ 6. His description of the Roman Province, or Gallia Nar-
bonensis, as it was now beginning to be called, is minute and
accurate, and he clearly points out the difference of its climate,
which distinguished it from other parts of Gaul, and more
nearly approached to that of Italy. ^ With the rest of Gaul his

* iv. 3, § 4, p. 193. I See Chapter XVII. p. 19.

« iv. 3, § 2, p. 192. » iv. p. 178. The whole of the Nar-

'' iv. 1, § 14, p. 189. bonitis (he says) produces the same

* Polyb. ap. Strab. iv. 2, § 1, p. 190. ' fruits as Italy. But as you advance to

Sect. 2. STRABO : EUROPE. 247

acquaintance was comparatively superficial, but he had here
an excellent authority before him in Csesar, of whose Com-
mentaries he made great use, and whom he generally follows
in regard to the names and divisions of the Gaulish tribes. A
more recent authority was indeed available in his time in the
inscription on the altar at Lyons erected in honour of Augustus
by the combined nations of Gaul, and which bore the names of
sixty tribes or states (civitates)/ But Strabo does not appear
to have derived any assistance from the materials furnished
by this document. Nor do we find him making any use,
for the purposes of his geographical description, of the lines of
road which the Eomans had already constructed through the
country : though he himself tells us that Agrippa had made
four such lines of highway, all proceeding from Lugdunum
(Lyons) as a centre. The first of these proceeded through the
Cevennes to the Santones and Aquitania ; the second led to
the Rhine ; the third to the shores of the Ocean, adjoining the
territory of the Bellovaci and Ambiani ; and the fourth to the
Narbonitis and the neighbourhood of Massilia.^ From thence
another line branched off by Tarasco to Nemausus and Narbo,
and thence to the passage of the Pyrenees. This last he
describes minutely, as well as another branch proceeding from
Tarasco through the land of the Vocontii to Ebrodunum
(Embrun), and thence over the Mont Genevre to Ocelum in
Italy .^ This was in his day one of the most frequented passes
over the Alps. But his accurate details concerning these
roads through the Eoman province, which had existed long
before, render the absence of them in regard to the great
central lines the more striking.

There can be no doubt that his knowledge of the parts of

the north, and to Mount Cemmenus, | altar is again referred to by Dion Cas-

the country ceases to produce olives sius (liv. 32), by Suetonius {Claud. 2),

and figs, though other things still grow. ; and by Juvenal {Sat. i. v. 44). Unfortu-

As you advance farther, the vine also j nately no record has been preserved to

does not readily ripen its fruit. But | us of the inscribed names,

the whole country bears corn, millet, i ^ iv. 6, § 11, p. 208.

acorns, and all kinds of cattle. ; ^ iv. 1, § 3, pp. 178, 179. -

' Strabo, iv. 3, § 2, p. 192. This | •

248 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXI.

Gaul adjoining the Ocean was very imperfect : the vagueness
and generality of his notices of this part of the country con-
trasts strongly with the detailed accuracy of his description of
the regions adjacent to the Mediterranean and the Pyrenees.
The only exception is with regard to the Veneti, of whose
naval power and the construction of their ships he gives a full
account ; but this is taken directly from Csesar.* Of the other
Armorican tribes he mentions only the Osismii, whom he iden-
tifies with the Ostimii of Pytheas, and states that they dwelt
upon a promontory projecting a considerable distance into the
sea, but not to the extent maintained by that writer, and those
who followed him. It is evident that Strabo had here no
correct information, and had no idea of the real extent and
magnitude of the Armorican promontory. He apparently con-
ceived the Yeneti, who, as he learned from Caesar, carried
on an extensive trade with Britain, to be situated opposite to
that island.^

§ 7. With regard to the division of Gaul he begins by stating
in accordance with Csesar that it was divided into three
nations, the Aquitanians, the Celts or Gauls properly so called,
and the Belgse. The Aquitanians were, as he justly observes,
a wholly distinct people from the Celts, and more nearly
resembled the Iberians. In this ethnographical sense they
were bounded by the Garonne to the north : but in the reor-
ganization of Gaul by Augustus, that emperor had extended
the limits of Aquitania to the Loire, thus uniting fourteen
tribes of Celtic origin with the Aquitanians properly so called.
The rest of Gaul was divided into the provinces of Gallia Lug-
dunensis and Belgica : but Strabo differs from all other writers

* iv. 4, § 1, p. 194. Compare Caesar, ' annia, qum contra eas regiones posita
B. G. V. 13. est, arcessunt " (B. G. iii. 9). There is

* Strabo, iv. 4, § 1. He may in part ' nothing in Csesar to show whether he
have been misled by Caesar's expression, was acquainted with the configuration
■where, after enumerating the maritime of the coasts of Brittany and Nor-
nations that sent auxiliary forces to the mandy ; he twice mentions the name
Veneti, including the Osismii and of the Osismii among the Armorican
Lexovii, as well as the Morini and tribes, but with no further indication
Monapii, he adds, "auxilia ex Brit- of their position (5. (?. ii. ?>4 ; vii. 75).

Sect. 2. STEABO : EUROPE. 249

in extending the latter province along the shores of the Ocean
from the mouths of the Ehine to those of the Loire, so as to
include the Yeneti and Osismii among the Belgic tribes.® This
is probably an error, but Strabo himself remarks that the geo-
grapher does not require to take much pains with regard to
the merely political and administrative divisions of countries
where these do not coincide with natural boundaries.

His account of the manners and customs of the Gauls, as
well as of their religious rites and ceremonies, is taken almost
entirely from Csesar, but with the addition of some circum-
stances of more dubious authenticity, which he derived from
Posidonius, Artemidorus, and other Greek authorities. He
adds however that the Gauls were rapidly becoming civilized,
and imitating the Roman manners, as well as adopting their
language. This change had already taken place to a great
extent in the Roman province, or Narbonitis, where the native
tribes had been stimulated by the example of the Massaliots,
and begun even to devote their attention to literature and
study : and it was from thence extending itself by degrees into
the neighbouring parts of Gaul.'^

§ 8. Of Britain he had very little knowledge beyond what
he derived from Csesar. We have already seen that he erro-
neously conceived the south coast of Britain to extend opposite
to that of Gaul, from the mouths of the Rhine to the Pyrenees,
and that the interval was throughout much the same, so that
the distance was not much greater from the mouths of the
Garonne and the Loire than from those of the Seine and
Rhine. But the nearest point, he correctly adds, was from the
Portus Itius, in the land of the Morini, from whence Csesar
sailed on his expedition to the island : the distance at this
point being only 320 stadia. It is strange however that he
altogether rejects the statements of Caesar with regard to the
dimensions of the island, and regards the side opposite to
Gaul — the length of which he estimates at the utmost at

* Strabo, I. c. ' iv. 1, § 12, p. 186.

250

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXI.

5000 stadia — as the longest side of Britain, instead of being,
as Caesar had described it, and as it really is, by much the
shortest.^ He consequently gave to the island a very incon-
siderable extension towards the north, so as to bring its most
northerly portions into the same latitude as the mouths of the
Borysthenes, and only 8700 stadia, or 14^ degrees of latitude
north of the Strait of the Columns.^

No attempt had been made since the time of Caesar to
subjugate Britain, but the native princes had entered into
friendly relations with the Eoman Emperors, and a consider-
able commerce was carried on with the island. Among the
products exported from thence Strabo enumerates gold and
silver as well as iron, but makes no mention of tin : besides
these, he says, it furnished corn, cattle, hides and slaves, and
dogs for the chase of a very fine breed. The climate was
milder than that of Gaul, but very subject to mists, so that
even in bright weather the sun was only visible for three or
four hours in the day.^

lerne or Ireland he conceived, as has been already men-
tioned, and as he himself repeatedly states, to be situated to
the north of Britain. Its length was greater than its breadth,
but he does not give an estimate of either : nor does he in this
place say anything of its distance from Britain. But he else-
where states that the interval was not known with any cer-
tainty.^ He however regarded it as the most northern of
all known lands, and as barely habitable on account of the
cold.^ Of its inhabitants little was known : they were said
to be mere savages, addicted to cannibalism, and holding
promiscuous intercourse with their Avomen. But Strabo himself

« strabo, i. 4, § 3, p. 63 1 iv. 5, § 1,
p. 199. See Chapter XIX. p. 127.

» Id. ii. 5, § 78, pp. 114, 115.

' Id. iv. 5, §§ 2, 3. lu another pas-
sage (ii. 5, p. 115) he tells us that the
Romans purposely refrained from con-
querincij the ■ island, in order to avoid
the expense of maintaining it.

- ii. 5, § 8, p. 115. TO S' inuOiv iirl

T^v 'lepvriv ovKeTt yvwpijjLOV, tr6aov &v tis
Oeirj. Again, ia another passage vii.
p. 72), he tells us that lerne was not
more than 5000 stadia distant from
Gaul : a statement that he must have
found some difBculty in reconciling
with his own system.

^ See the passages cited in note to
p. 232.

Sect. 2. STRABO : EUEOPE. 2$ I

adds that he had no trustworthy authorities for these facts.
The other islands around Britain he treats as unworthy of
notice, and mentions Thule only to repeat his disbelief of the
account of it that had been given by Pytheas.*

§ 9. He next returns to speak of the Alps, his knowledge of
which shows, as might be expected, a great advance upon that
of Poly bins. Indeed the recent subjugation of the Alpine
tribes under Augustus, and the frequent communication held
by the Eomans with their Transalpine provinces, had necessarily
led to a much more familiar acquaintance with these mountains.
Hence Strabo is not only able to give us many interesting
particulars concerning the different nations inhabiting the
Alps and a correct description of their localities, but his
account of the mountain chain itself shows a clear idea of its
general form and configuration, and of the rivers that flowed
from it. Thus he describes the Alps as forming a great curve
having its concave side turned towards the plains of Italy, its
centre in the land of the Salassians, and its two extremities
bending round, the one by Mount Ocra, and the head of the
Adriatic, the other along the sea coast of Liguria to Genoa,
where they join the Apennines.^ In another passage ^ he fixes
the termination of the Maritime Alps with more precision at
Vada Sabbata (Vado), 260 stadia from Genoa, which almost
exactly coincides with the view generally adopted by modern
geographers. The highest summits of the whole range he
supposes to be those in the land of the MeduUi (between the
Mont Genevre and the Petit St. Bernard), where the direct
ascent of the mountains was said to be not less than 100 stadia
and the descent on the other side into Italy the same distance.
Here among the hollows of the mountains was a lake, and two
sources, from one of which flowed the Druentia (Durance) into
Gaul to join the Rhone ; from the other the Durias (Dora) to

' Strabo, iv. 5, § 5, p. 201. Of the
Cassiterides he had already spoken, in
connection with Spain, and evidently
did not regard them as belonging to

the group of the Britannic Islands (see
above, p. 245).

* V. 1, § 3, p. 211.

" iv. 6, § 1, p. 202.

252 HISTOEY OF ANCIENT GEOGKAPHY. Chap. XXI.

join the Po.'^ That river itself had its sources in the same neigh-
bourhood, but at a lower level, and was swelled in its course
by the junction of many tributaries.^ In like manner he tells
us correctly that the Rhone and the Ehine had their sources
near to one another in the Mount Adula (^ABovXaq) f — the only
distinctive appellation of any particular group which he men-
tions — and that they each formed a large lake in their course
lower down.-^ He was also aware of the true source of the
Danube, which he well describes as lying in a detached ridge
of mountains, beyond the Ehine and its lake, adjoining the
Suevi and the Hercynian Forest.^

With the eastern extremity of the Alps, where the chain
sweeps round the head of the Adriatic he was also well
acquainted, and gives a curious account of the commerce that
was carried on in his day over the Mount Ocra — which he
correctly describes as the lowest part of the Alps — from
Aquileia to a place called Nauportus or Pamportus on the
Save. It was by this route that Italian goods were conveyed
into Pannonia and the other countries on the banks of the
Danube.^ The other mountaineers of the xilps also carried on
some trade with Italy, bringing down resin, pitch, wax, honey,
and cheese. In his time they were become tranquil subjects
of Eome, and had laid aside the predatory habits which they
had practised for centuries.

§ 10. Augustus, who had completed the subjugation of the
mountain tribes, had also, he tells us, bestowed great pains
upon the construction of roads through their country : and had
rendered these practicable for carriages, wherever the natural
difficulties were not too great.* Still the number of high

' iv. 6, § 5, pp. 203, 204. j formed by the Rhiue, though it had

* Ibid. p. 204. ' recently attracted attention during the

" Ibid. § 6, p. 204. He adds that tlie | campaign of Tiberius against the Vin^

Aduas (Adda), which formed the Lake
Larius, had its som-ces in the same
mountain group.

' Of these he was well acquainted
with the name of the Lake Lemanus,
through which the Ehone flowed ; but
apparently knew no name for the lake

delici, who had actually established a
naval station on the only island it con-
tains (Strabo, vii. 1, § 5, p. 292).

2 iv. G, § 9 ; vii. 1, § 5,

3 iv. 6, § 10, p. 208.
' iv. G, § 6, p. 204.

Sect. 2. STEABO : EUROPE. 253

roads thus opened was but small. Of the two passes leading
from the valley of the Salassi to Lugdunum, the one through
the Centrones (the Little St. Bernard) which was the longer
and more circuitous was available for carriages, the other
across the Pennine Alps (the Great St. Bernard) was more
direct, but narrow and steep, and not practicable for carriages.
The road through the Graian Alps, and the petty kingdom of
Cottius (the Mont Genevre)^ was apparently also open to
carriages, and was one of the most frequented passes in the
Roman times. No mention is found of any other pass between
the Great St. Bernard and that through the Eheetians (the
Brenner pass in the Tyrol) which from its comparative facility
must have been frequented in all times. But Strabo, with a
want of method often found in his work, while censuring
Polybius for noticing only four passes across the Alps, has
omitted to give us any regular enumeration of those known
and frequented in his own day. He describes in strong terms
the natural difficulties of these passes, the frightful chasms and
giddy precipices along which the narrow roads had to be
carried, as well as the avalanches of snow, which were capable
of carrying away whole companies of travellers at once. These
he ascribes with remarkable precision to the sliding of great
masses of snow, congealed by successive frosts, one over the
other.®

§ 11. The fifth and sixth books are devoted to the description
of Italy and the adjoining islands, with which he was of course
well acquainted, and for the topography of which he had abun-
dant materials at his command. We have already seen how
erroneous was his conception of the true position and configu-
ration of the peninsula, as it would be represented on a map ;
but with its general features he was naturally familiar, and his
outline of its physical geography is on the whole clear and
satisfactory. The leading natural features of Italy are indeed
so strongly marked by nature that it would be difficult not to

° This route is described in detail, iv. 1, § 3, p, 179.
« iv. 6, § 6, p. 20i.

254

HISTOEY OF ANCIENT GEOGRAPHY,

Chap. XXI.

seize them correctly. Such is in the first place the broad
valley, or rather plain, of the Po, bounded by the great chain
of the Alps on the north and by the inferior, but scarcely less
marked, range of the Apennines to the south, and gradually
jDassing into the lagunes and marshes of Venetia and the low
country near Kavenna. The Apennines also are well described
by Strabo as extending directly across the whole breadth of
the land, from the frontiers of Liguria and Tyrrhenia on the
one sea to the neighbourhood of Ariminum and Ancona on the
other, and then turning inland so as to divide the peninsula
into two through its whole length, but keeping nearer to the
Adriatic till they turn off again in Lucania, and after passing
through Lucania and Bruttium end in the promontory of
Leucopetra not far from Ehegium.^ He compares the penin-
sular portion of Italy — excluding the two projecting spurs or
promontories of lapygia and Bruttium — with that of the
Adriatic Sea adjoining it :^ rather a singular comparison and
rendered more so by his adding that the length of each is not
much less than 6000 stadia (600 G-. miles), a great exaggeration,
as the distance from Ariminum to the extremity of the lapygian
peninsula (thus including the latter, which Strabo excludes) is
little more, as measured on the map, than 360 Gr. miles.^

This last statement is probably copied from some of his
earlier G-reek authorities : and indeed throughout this portion
of his work we find him fluctuating between two sets of autho-
rities — the earlier Greek writers, to whose statements he clings
with a strange tenacity, even in regard to matters on which
much better sources of information were open to him, and the
more recent statements of Koman writers, based upon more
accurate measurements and itineraries. Among the latter
especially we find him repeatedly citing an anonymous author
whom he calls " the chorographer," and of whom all that we

' V. 1, § 3, p. 211.

8 Ibid.

^ strabo had apparently, in common
with the earlier geographers, an exag-
gerated notion of the length of the

Adriatic, and adapted his ideas of
Italy to it. The Autonine Itinerary
gives the distance (by road) from Ari-
minum to Brundisiiim at 524 Roman
miles, or 420 G. miles.

Sect. 2.

STRABO : EUROPE.

255

know is that from his giving the distances in miles it may be
fairly inferred that he was a Latin, not a Greek, author.^
Whether this anonymous work was based mainly on the itine-
raries and consequently confined chiefly to distances, cannot
be affirmed with certainty, nor do we know from what sources
Strabo derived his knowledge of the topography of those parts
of Italy which he had not himself visited, but it is certain that
these topographical details are for the most part very correct,
and the order in which the numerous towns mentioned are
enumerated is generally systematic and well chosen. It is
clear indeed, as has been already shown, that maps of Italy
were well known, and probably not uncommon, in the time of
Strabo, and the clear and methodical character of his descrip-
tion certainly gives the impression of having been written
with such a representation before him. At the same time the
more lively and graphic manner in which he describes par-
ticular localities — as for instance the Port of Luna, Volterra,
Populonium, and the greater part of Campania — points clearly
to being the result of personal observation. His account of
Northern Italy on the other hand, in which he gives many
interesting details concerning the marshes and lagunes of
Venetia and the coast of the Adriatic from Altinum to
Kavenna, and his description of the site of the latter city —
a position almost exactly resembling that of Venice at the
present day^ — must probably have been derived at second-
hand from some other writer. He follows the popular Eoman
notion that the Padus was the largest river in Europe except
the Danube : ^ but rejects without hesitation its identification
with the famous Eridanus, which he treats as a wholly fabulous
stream.*

In describing Campania he takes occasion to give us some

* Oq this subject see tlie note to p.
177, Chapter XX.

2 V. 1, § 7, p. 213. It is remarkable
also that he notices the tides in this
part of the Adriatic, which is, as he ob-
serves, the only part of " our sea," which
is affected in this respect like the

Ocean. Ibid. § 5, p. 212.

^ els 5e Ti]v ' ASpiWT LK^jv OdXarrav e/c-
TriTTTet (0 ndSos sc), ixeyiaros yey6fievos
Twv Kara, rrjv EupcoTTTjj/ irorafJLWV itX^v
Tov "larpov, iv. 6, § 5, p. 204.

* rhv 'Hpt5af(5j', Thv iJ,7}Safj.ov yrjs ovTa,
V. 1, § 9, p. 215.

256 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXI.

curious particulars concerning the volcanic eruptions of which
it had been the scene, particularly of one of Mount Epomeus
in the island of Pithecusa (Ischia), which had been described
by the historian Timseus, having happened not long before
his time.^ On the other hand he has recorded his sagacious
observation of Mount Vesuvius, that it had every appear-
ance of having once been a burning mountain, but which had
gone out for want of fuel.^ It was little suspected by any one
how soon it was destined to resume its activity.

The account of the south of Italy — Lucania, Bruttium,
lapygia, and Apulia, which occupies the greater part of the
sixth book, is mainly derived from Greek authorities, and
taken up to a great extent with historical particulars con-
cerning the Greek colonies which bordered the whole of these
shores. Many of these are of much interest and would be
otherwise unknown to us, but not strictly of a geographical
character. In describing the Apulian coast he treats of the
distances along the Adriatic generally, and points out the
discrepancy between " the chorographer " and his Greek au-
thorities, Artemidorus especially ; and takes the opportunity
of commenting on the diversity frequently found in this
respect among different authors.'' When he has no means of
determining between them, he adds, he contents himself with
repeating the conflicting statements : but it does not seem to
have occurred to him that the Eoman authorities, having the
advantage of measured roads, were in most cases, if not in
all, entitled to the greater credit.* In describing Brundusium
he notices briefly the course of the Appian Way — the great
highway from Kome to the provinces of the East, which in his

« V. 4, § 9, p. 248.

® ws reKfiaipoir' &v rts rh xcopfov rovro
KuieaBai irpSrepou koI fX^"' Kpam^pas
Trvp6s, (TfieaQrjvai S' iiriAnrovcTts ttjs
Sx-ns, V. 4, § 8, p. 247.

' vi. 3, § 10, p. 285.

* In this instance the chorographer
gave the distance from Brundusium to
the Garganus at 165 (Eoman) miles,
and from thence to Ancona at 254 miles.

The first distance is almost precisely
correct, according lo the Itineraries,
which give 167 miles from Brundusium
to Sipontum (Manfredonia) ; but the
second falls considerably abort of that
given in the Itineraries, which amounts
to 281 miles. It is probable, however,
that in this instance the latter exceed
the truth.

Sect. 2. STRABO : EUROPE. ' 257

day consisted of two main branches, the one, practicable for
carriages, leading from Brundusium to Tarentum, and thence
direct through Venusia to Beneventum : the other, practicable
for mules only, proceeding through Egnatia, Canusium, and
Herdonea, and rejoining the main line at Beneventum.^ He
gives the whole distance from Eome to Brundusium as 360
miles, Avhich is almost precisely correct, the distance by the
first of the two roads described being 358 miles according to
the Antonine Itinerary .'^

§ 12. He describes Sicily at considerable length, and on this
occasion gives us the distances furnished him by " the cho-
rographer " in detail, showing the nature of the materials on
which that author relied. Nor does his estimate of the dimen-
sions of the island differ widely from the truth ; though, as we
have already seen, he had such a distorted idea of its position,
and the bearings and directions of its three sides. But he had
never himself visited the island, and his description is neither
very complete nor very accurate. He draws indeed a lamentable
picture of the state of decay to which it was reduced in his
time, notwithstanding its great natural fertility, so that many
of the towns had altogether disappeared, while the interior was
abandoned almost entirely to shepherds ; ^ and on this account
he dwells the less carefully upon topographical details.

Of the physical geography of the island he does not attempt
to give any general view, but dwells at considerable length
upon the peculiar characters of ^tna, and the volcanic pheno-
mena to which it was subject. Of the streams of lava especially
he gives an accurate and philosophical account, pointing out
how the burning matter that overflows from the crater in a
liquid state gradually hardens into a compact and hard rock,
like a mill-stone.^ He notices also the great fertility of the
soil produced by the volcanic ashes for the growth of vines : a

vi. 3, § 7. It would seem that in

the two branches.

Itin. Ant. pp. 107-111, 120.
vi. 2, § 6, p. 272.

his time the name of the Appian Way
wtis confined to the portion from Bene-
V( ntiim to Rome, after tlie junction of j ^ vi. 2, § 3, p. 269.

VOL. II. S

258

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXI.

circumstance that lie had already observed in regard to
Vesuvius.* Much of this description appears to be taken from
Posidonius : but Strabo adds an account of the appearance of
the summit and the actual condition of the crater, as he had
heard it from persons who had recently made the ascent.^ It is
evident therefore that in his time it was not uncommon for
inquisitive travellers to make the ascent, which really offers
no difficulties. He gives us also many interesting particulars
concerning the volcanic phenomena of the ^olian Islands,
especially of the remarkable eruption that had thrown up a
small islet or rock out of the sea in the neighbourhood of the
island called by the Greeks Hiera, as being above all others
sacred to Vulcan, and the scene of his subterranean operations.^
Both this island and that of Strongyle (Stromboli) seem to
have been at this period in a state of constant volcanic
activity.

§ 13. Of the other two great islands of the Mediterranean,
Sardinia and Corsica, Strabo has given but a very brief and
imperfect account.'' Both of them indeed were in ancient
times, as they have continued almost to our own day, in a state
of semi-barbarism little corresponding with their size and
their natural resources ; and though they had long been
brought under the direct authority of Rome, the mountaineers
of the interior continued in both islands to lead a lawless and
barbarous life, plundering their neighbours in the plains, and
only checked from time to time by the Roman governors, who
would make a razzia for the purpose of carrying off slaves, but
never attempted to exercise any permanent authority over
these wild districts.^ Some parts of Sardinia, however, as Strabo

■• V. 4, §8, p. 247.^

^ ot 5' ovv ceoxTTi ava^dvres Sir^yovvTO
■nixlv, vi. 2, § 8, p. 274.

« vi. 2, § 11, p. 277. This outbreak,
which was related by Posidonius as
occuiriug within his own memory (/cara
T^v favToC i.iv4fx-r)v), was almost certainly
the same event as that mentioned by
Pliny, and referred by liim to tlie 3rd
year of the 103rd Olympiad (b.c. 126).

It is noticed also by Orosius and Julius
Obsequens ; but Orosius describes a
similar phenomenon, the emergence
from the sea of an island not previously
existing, as taking place 60 years earlier
(B.C. 186). It is probable that the
small island, now called Vulcanello, is
due to one or other of these eruptions.

' v. 2, § 7, pp. 224, 225.

^ V. 2, § 7, p. 225.

Sect. 2. STRABO : EUROPE. 259

observes, were fertile and produced abundance of corn, but they
suffered much from unhealthiness, as well as from the depre-
dations of their neighbours in the mountains. The only towns
of any importance were Caralis and Sulci.^

It has already been pointed out that Strabo committed a
strange error with regard to the geographical position of
Sardinia and Corsica, as well as that of Sicily : and it is a
striking instance of his disregard for real geographical accu-
racy, that he repeats the measurements given by the anony-
mous chorographer, both for the length and breadth of the two
islands, and for the distance from Sardinia to Africa, without
perceiving, or at least without noticing, how entirely they
were at variance with his own system and arrangement.^

§ 14. Eeturning to the north of Europe, Strabo proceeds, in
the seventh book, to give a brief general account of the coun-
tries extending from the Khine eastwards to the Borysthenes
and the Tanais, and situated to the north of the Danube,
which he describes as cutting the whole of this eastern half of
Europe into two divisions.^ No part of his work is more defec-
tive than this. Imperfect as was the knowledge actually pos-
sessed of these regions, he was far from turning to account all
the information concerning them that was really available in
his day. In regard to Germany, indeed, he did not fail to make
use of the new discoveries that had been opened out by the
campaigns of Drusus and Germanicus, which had extended, as
he observes, the knowledge of Germany from the Rhine to the
Elbe.^ He mentions also the intermediate rivers ; the Amisia
(Ems), and the Yisurgis (Weser), as well as the minor con-
fluents the Lupia (Lippe) and the Salas (Saale). All these
streams, as we have seen, had attracted attention in the Roman
wars. He mentions also the principal names of German tribes
and nations, with which the same wars had rendered his con-
temporaries familiar, though with very little attempt to explain

Ibid.

Note C, p. 275. _

oiaipe7 yap oZros airxa'a.v ojs SyyvTarw

Sixa TT/i/ \exde7crav y7iv, vii. 1, § 1.
3 vii. 1, § 4, p. 291.

S 2

260 mSTOEY OF ANCIENT GEOGEAPHY. Chap. XXT.

their topograpliical relations, of whicli lie liad probably very
imperfect knowledge. He describes at some length the
Hercynian Forest, which according to his conception consti-
tuted one of the main physical features of the country, ex-
tending from the Lake of Constance and the sources of the
Danube* to the northern frontier of Bohemia and Moravia,
including within it {i.e. between it and the Danube) a tract
of fertile country, occupied in part by the Quadi and Mar-
comanni, who had recently taken possession of the district
previously known as Boiohemum (Bohemia).^

But beyond the Elbe he tells us that everything was entirely
unkno^va ; an ignorance which he ascribes in great part to the
policy of Augustus in preventing his generals from carrying
their arms beyond that river :" and while he rejects as fables
the tales that were related by G-reek writers of the Cimbri,
who had long been known by a kind of vague tradition as
dwelling on the northern Ocean,' he has nothing to substitute
in their place. Even of their geographical position he had no
clear notion, and would have led us to imagine that they dwelt
on the west side of the Elbe ; he only notices the belief that
they inhabited a peninsula in connection with the tradition
(which he rejects as a fable) that they had been expelled from
it by an irruption of the sea.^ Of the great Cimbric Cher-
sonese or Promontory, as a geographical feature, he had
evidently no idea, or of the bay beyond it (the Codanus Sinus
of Latin writers), though some vague accounts of both had
certainly reached the Roman authorities in his day.* Nor
does he condescend to notice the tradition adopted by some
earlier authors of the existence of a great island in this part of

* vii. 1, § 5. ' and apparently at an earlier perioil, as

*. Ibid. § 3. This name appears in that author found fault with the fables

Strabo in the slightly distorted form related concerning tliem. The irrup-

Bouioiuo;/: but he termsit the )3a(riAefoj/, tion of the Cimbri and Teutones into

or royal residence, of Maroboduus. und Gaul and Italy (about IfiO B.C.), must

appears therefore to have conceived it as have made the Komans familiar with

a town, rather than a country or region. the riame, and would naturally excite

^ Ibid. 5J 4. p. 291. curiosity as to their original abodes.

' The Cimbri were certainly known ^ Strabo, vii. 2, § 1. p. 21i2.

to Posidonius (,ap. Strab. vii. 2, p. 293), ' See Chapter XX. p. 191.

Sect. 2. STEABO : ErEOPE. 261

the northern Ocean — the Basilia of Timseus and the Ahalus of
Pytheas. Even the name of the Vistula, which was certainly
known to the Eomans in his day, finds no place in his geo-
graphy, and it is strange that he does not even allude to the
trade with these regions for amber, which attracted so much
attention both among Greeks and Eomans.

The whole coast of the Ocean beyond the Elbe, he expressly
tells us,^ was utterly unknown, nor had any one made the journey
by land, so that it was only by geographical inference from the
comparison of parallels of latitude that one could arrive at the
conclusion that proceeding eastward from the Elbe would bring
one to the Borysthenes, and the regions north, of the Euxine.
But who were the nations inhabiting this extensive tract,
whether Bastamse, lazyges, Eoxolani, or other Scythian tribes,
no one could say : nor whether any of these tribes extended to
the northern Ocean, or there was a space beyond, uninhabit-
able from cold or other causes. On one point alone he was
correctly informed : for he tells us that the whole country
eastward from the frontiers of Germany to the Caspian was
one vast plain ■? thus discarding altogether the vague notions
as to the Khipeean ^lountains in the north, which had so long
lingered among Greek geographers, and which still retained a
hold on popular belief down to a later period.

§ 15. It is remarkable that Strabo's acquaintance with the
regions to the north and north-west of the Euxine was almost
as imperfect as with those on the Baltic. Eegarding Hero-
dotus as altogether unworthy of confidence, he evidently
neglected to avail himself of the interesting materials col-
lected by that historian, and he had no means of supplying
the deficiency. It would appear that the increasing pressure
of the northern barbarians upon the Greek cities of the Euxine
had limited their commercial relations with the interior : and
though Strabo himself tells us that the campaigns of the
generals of Mithridates had been the means of opening out

' Id. TiL 2, § 4, p. 294. I luofias y-exp't "rrjs Kacrrias redias icnv,

- 71 yap TrpoijdpKrias -raffa dxo Vep- \ rjr lafuew. vii. 3, § 17, p. 306.

262

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXI.

a more accurate knowledge of these countries, it is certain
that these did not carry their arms far from the coast, and the
interior seems to have remained virtually unknown. Thus he
tells us that the sources of the Tanais, like those of the Nile,
were wholly unknown :^ and the same was the case with those
of the Borysthenes, Hypanis and Tyras.* The Borysthenes he
describes as navigable for 600 stadia,^ and he apparently re-
garded it as not known any higher up. The most northerly
people known to our geographer in this part were the Eoxolani,
who in his time dwelt between the Tanais and the Borysthenes,
and were known in history from their having taken part in
war against Diophantus, the general of Mithridates.® They
were regarded by Strabo as a Sarmatian tribe, as were also the
lazyges ; while the Bastarnse, a powerful nation who at this
time occupied the tract between the Tyras (Dniester) and the
Carpathians, so as to adjoin the Germans on the west, are
described as pretty nearly of German race themselves.'^

The Tyras had formed the limits of the conquests of Mithri-
dates on the west, which was marked by the erection of a fort
at the mouth of that river bearing the name of his general
Neoptolemus.^ The Eomans in the days of Strabo had not
attempted to extend their power beyond the mouths of the
Danube. The interval between the two was occupied by an
unpeopled tract, called by Strabo " the desert of the Getse,"
and which he describes as the scene of the expedition of
Darius, of which Herodotus has left us so exaggerated an
account. But it is difficult to adopt Strabo's suggestion lite-
rally, and suppose that the Persian king never even reached
the Dniester, a distance of less than a hundred miles from the
Danube.^

' xi. 2, § 2, p. 493. He, however,
justly maintains that the Tanais falls
into the Palus Mseotis from the north ;
and rejects the wild hypotheses which
derived it either from the east and the
Caucasus, or from the far west, near
the sources of the Danube (!)
' ' ii. 4, § 6, p. 107.

^ vii. 3, § 16, p. 306.

" vii. 3, § 17, p. 306. ' Ibid.

8 vii. 3, § 16. See Chapter XVIII.
p. 84.

» vii. 3, § 14, p. 305. The same
tract was, according to Strabo, also tlie
scene of the expedition of Lysunachus,
in which that monarch was defeated

Sect. 2.

STRABO: EUROPE.

263

The accurate and detailed account which Strabo gives us of
the Tauric Chersonese forms a striking contrast with the vague
and unsatisfactory knowledge he possessed of the countries to
the north. Here he was in a land which had long been occu-
pied by Greek colonies, to whom it had thus become known
in detail. It had more recently passed under the dominion of
Mithridates, and it was probably from the historians of that
monarch that Strabo derived his particulars. His statement
that the peninsula as a whole resembled the Peloponnesus
both in form and size is more correct than such general
comparisons usually are •} he was aware that it was really
joined to the mainland only by a narrow isthmus, and has
given a correct and curious account of the peculiar character
of the Putrid Sea which separates it from the Palus Mseotis.^

§ 16. Of the Getse or Dacians, who at this time occupied the
extensive tract north of the Danube, Strabo had very little
real knowledge, and the greatest part of the section devoted
to this people is in fact occupied with a very unsatisfactory
discussion as to the ethnographical relations of the people
called by Homer Mysians, and the " illustrious mare-milking "
tribes of the same poet. As already mentioned, he describes
the Getse and Dacians as two distinct nations, or at least dis-
tinct branches of the same nation, for he adds that they speak
the same language, which was the case also with the Getse and
the Thracians.^ He appears not to have noticed or compre-
hended the fact that the one name was originally applied to
them by the Eomans, the other by the Greeks. Yet he

and taken prisoner by Dromichsetes,
king of the Getse, bnt this naay more
probably be placed farther west, be-
tween the Danube and the Carpa-
thians.

' 7) Se fxeydXr] Xepp6i/7](Tos rfj TleXoTTOV-
v/iffo: wpoaeoLKe Kal rb tr^rjiua Kal rh 1x4-
yeBos. vii. 4, § 5, p. 310.

2 Ibid. § 1, p. 308. The precise
agreement of Strabo's account of this
curious natural feature of the country
with its present condition is a striking

proof that no considerable physical
changes have taken place in this part
of the Euxine since the time of the
geographer: as a depression or ele-
vation of a few feet would suffice en-
tirely to change the character of the
Putrid Sea.

* bfx.6y\()}TT0L S' el(r\v 01 AaKol rols
Terais, vii. 3, § 13, p. 305. And in
another passage (§ 10, p. 303) he calls
the Getse ouSyAwTTou ro7s @pai)v
edvos.

264

HISTOKY OF ANCIENT GEOGKAPHY.

Chap. XXI.

recognized a similar fact with regard to the Danube, which
as he points out was called by the Eomans Danubius down to
the cataracts, while below that point they adopted the Greek
appellation of Ister.*

With regard to the regions south of the Danube, Illyricum,
Pannonia and Moesia, which as we have already seen had been
lately brought under the dominion of Rome, Strabo had not
much that was new to relate, though by availing himself of the
results of the Eoman wars and conquests in these parts, he was
doubtless enabled to bring together a more complete and con-
sistent view of these nations, than had been presented by any
previous Greek writer. But their ethnological relations were
then, as they have always continued to be, extremely obscure,
and his notices with regard to them, though not without value,
show but an imperfect insight into the subject, while he has
unfortunately neglected in almost all cases to cite the authori-
ties from whom he has derived them. In respect to the
geography of this part of Europe he has correctly seized the
main fact of the chain of Mount Heemus (the Balkan) branch-
ing off from the great Illyrian ranges which descend in a
continuous mass from the Alps along the east coast of the
Adriatic, and extending in a line parallel with the Danube
(from west to east) to the shore of the Euxine.^ He has
indeed an exaggerated notion of its importance, both in height
and extent, but this was the case even with modern geo-
graphers down to a very recent period ; and he justly rejects
the statement of Polybius that from the summit both the
Euxine and the Adriatic Seas were visible at the same time.^

^ Strabo, vii. 3, § 13, p. 304. This is
the first mention in any ancient author
of the cataracts or rapids, popularly
known as the Iron Gates, which con-
stitute so serious an impediment to the
navigation of the Danube.

^ vii. 5, § 1, p. 313.

•* This had indeed been an article of
the popular creed long before the time
of Polybius. It was already asserted
by Theoporapiio, and the expedition of

Philip v., king of Macedonia, to the
summit, was evidently originated in
consequence of this belief. The de-
tailed account of that expedition, given
by Livy (xl. 21, 22), is doubtless de-
rived from Polybius, but the Roman
historian shares in the doubts so rea-
sonably expressed by Strabo. Ponipo-
nius Mela at a later period repeats the
ordinary story (ii. 2).

It is impossible to determine what

Sect. 2.

STRABO : EUROPE.

265

§ 17. His account of Macedonia and Thrace is preserved
only in a fragmentary form, this part of his work being wanting
in all the extant manuscripts, but the deficiency is in great
part supplied by the two Epitomes, and it is not probable that
we have lost much that is really valuable. In regard to both
countries, his minute and accurate account of the portions
adjoining the coasts of the -^gean contrasts strongly with his
vague and general information concerning the mountain dis-
tricts of the interior, and the wild tribes that inhabited them.
The latter were still very imperfectly subdued, and their rela-
tions with the Eomans were chiefly those of mutual hostility.'^
This portion of the Turkish Empire was even down to the
present day one of the least known parts of Europe, and the
same thing appears to have been the case in the time of
Strabo. The geographer had however here the great advan-
tage of the Eoman military highway, the Via Egnatia, which
traversed the whole country from west to east : the importance
of which, with its measured distances, had been already recog-
nized and pointed out by Polybius ; from whom indeed Strabo's
account of it is almost entirely derived.®

§ 18. Three books are devoted by Strabo to the geography
of Greece and the neighbouring islands. Here of course
nothing new was to be expected, while materials for the topo-
graphical description of the country could not fail to be forth-
coming in profusion. Yet there is hardly any part of his work
which in a geographical point of view is more unsatisfactory.

was tlie particular suramit actually
ascended by Philip ; we are told only
that he marched thither from Stobi
through the country of the Msedi.
Leake supposes him to have selected
one of the lofty group near the head
waters of the Strymon, between Sofia
and Kiiistendil, but the orography of
this part of Turkey is still too imper-
fectly known to enable us to form even
a pldu&ible coDJecture.

It is strange tliat Strabo, who shows
on this occasion a very reasonable scep-
ticism, should accept without hesitation

the popular story that the rising sun
was visible from the summit of Mount
Athos three hours before it made its
appearance to those at the foot of the
mountain ! (vii. Fr. 33, 35).

' The account given by Tacitus
(Annal. iv. 46-51) of the outbreak of
the Thracian tribes in a.d. 26, some
years after the death of Strabo, shows
how far these wild mountaineers were
from being effectually reduced to sub-
jection.

« Strabo, vii. 7, § 4, pp. 322, 323.
See Chapter XVII. p. 27.

266

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXI.

That tendency to digression upon mythological and poetical
topics, which, as we havp already pointed out, is one of the
leading defects of Strabo's whole wo*rk, is here developed to
the greatest extent, and has had the effect of converting all
these three books into a desultory and rambling commen-
tary upon the Homeric Catalogue of the Ships, together
with some other passages of the ancient poets, rather than a
systematic geographical treatise. Nor is this commentary
illustrated with local details and topographical identifications,
which would have had some real interest, and conveyed much
inform.ation to the modern student. He had himself visited
only a few points of Greece,^ and was therefore compelled to
collect his information at second hand : and unfortunately he
sought this more in the works of the logographers and gram-
marians, than in those of the topographers and local historians.
It is but justice to him to remark on the contrast which his
account of Corinth — which he had seen — presents with the
rest of this portion of his work.^ Here his description — though
not very full or detailed — is clear, intelligent and character-
istic. Of Sparta and Argos on the contrary he gives us no
description at all — or nothing worthy of the name — contenting
himself with remarking that they were too well known to
require it : and the reader would gather from his pages no
idea of the striking character and natural features of the plain
of Argos, or the valley of the Eurotas. Nor does he anywhere
give us a clear outline of the grouping and connexion of the
mountain chains, which in so remarkable a manner constitute
as it were the skeleton of the Peloponnese, and determine the
physical geography of the country.

One circumstance that appears to have contributed to pre-
vent him from dwelling more fully ifpon the actual geography
of Greece in his own time was the state of decay to which it
was then reduced : a circumstance to which he recurs again

^ Athens, Megara, and Corinth, are
the only points in continental Greece,
which ho can be proved to have actually

visited. Groskurd adds Argos, but I
can see no evidence of this.
' viii. 6, §21, p. 379.

Sect. 2. STRABO : EUROPE. 26/

and again. Even the fertile district of Messenia was in great
part desolate and abandoned : Laconia retained a few towns
which were tolerably flourishing, though its population had
much declined : but the upland plains of Arcadia were almost
wholly depopulated, the towns had ceased to exist or were
lying in ruins, and even the agricultural labourers had quitted
the country, leaving the fertile arable lands to siipport nothing
but herds of cattle, horses, and asses.^ The case was little
better, if at all, with Northern Greece. Boeotia especially had
never recovered from the ravages of the Mithridatic War:
Thebes was reduced to a mere village,^ and Tanagra and
Thespise alone could still claim the appellation of to^vns.* In
other passages he points out the state of depopulation and
decay of Acarnania, ^tolia, Locris and the adjoining territory
of the ^Enianes.^ The new colonies founded by Augustus —
Mcopolis, Patrae, and Corinth — were indeed flourishing settle-
ments, but their prosperity was to a great degree at the cost
of the neighbouring districts. Such a state of things might
be some excuse for not entering minutely into topographical
details, but it is none for going back to the heroic ages, and
wasting time in idle discussions on the obscure towns men-
tioned only by Homer, whose names and sites were alike
unknown in the flourishing ages of Greece, as well as in the
days of Strabo.^

This unfortunate mode of treating his subject appears to
have arisen in great measure from his following the example
of the writers who had composed professed commentaries upon
the Homeric Catalogues, ApoUodorus and Demetrius of Scep-
sis, rather than the authors of strictly geographical or topo-
graphical works, which were certainly not wanting in his day.'^

2 viii. 8, § 1.

3 ix. 2, § 5, p. 403. e| iKeivov S" ^Stj
irpoLTTOvTes evSeecrrepov ael fiexpi eis
7]fj.as ovSe Kcifirfs a^w\6yov rinrov crdi-

^OVffl.

* Ibid. § 25, p. 410.

^ ix. 4, § 11 ; X. 2, § 23.

^ At the same time he took so little

interest in tracing the remains of these
extinct cities, that he does not even
notice the gigantic ruins of Tiryns and
Mycense, but adds with regard to the
last that not a trace of it was visible '
(viii. 6, § 10, p. 372.)

7 The fragment of a Description of
Greece, commonly ascribed to Dicse-

268

HISTORY OF A^NCIENT GEOGRAPHY.

Chap. XXI.

But his blind reverence for the great poet, whom he regarded
as the first and best of authorities/ was the original source of
this defective method. Besides the two writers already cited,
his principal, and by far his most valuable, authority was
Ephorus, from whom he derived the greater part of his in-
formation in regard to the historical facts which he relates
concerning the foundation of cities, the changes of population,
&c. ; much of which is really valuable and interesting.^ But
though Ephorus, like Polybius, had devoted a portion of his
work to a separate and regular geographical treatise, it is
remarkable that he is hardly ever cited by Strabo for any
statement of a distinctly geographical character.

§ 19. In this respect indeed, strange as it may appear to us,
the knowledge of Greece possessed by Strabo was scarcely less
defective than that of the more western portions of Europe.
Familiar as was the general notion of the Peloponnese, as
resembling a leaf of the plane-tree, as well as the leading pro-
montories and bays that determined its configuration, it will
be found that its orientation (if the word may be allowed) was
wholly erroneous : and when Strabo tells us that its length
and breadth were about equal (1400 stadia in each direction),
he adds that its greatest length was from Cape Malea to
^Egium, and its greatest breadth from west to east from Cape
Chalonatas in Elis to the Isthmus.-^ He must therefore have
regarded the Isthmus as nearly, if not quite, the most eastern
point of the Peloponnese, ignoring the extent to which the
coast of Argolis runs out in an easterly direction to Cape
Scyllseum, or rather supposing the great promontory thus
formed to have a southerly instead of a south-easterly direction.
The effect of this is to give to the whole map of the Pelo-

arohus, -wlietber or not it be jiistly
attributed to that author, shows that
such topographical works were in
existence long before the time of Strabo,
and we can hardly doubt that there
were many such. (See Chapter XVI.
p. C17.)

* See especially viii. pp. 337, 349.

^ Besides numerous otlier citations,
he says expressly in one passage : "Ec^o-
pos, S> rh TrXilffrov irpoffxp^/^eOa Sia r-^v
irepl TavTU iTn/.L€\eiav. ix. 3, § 11, p.
422.

' viii. 2, § 1, p. 335.

Sect. 2. STRABO : EUROPE. 269

ponnese a sleiv round whicli greatly distorts its general appear-
ance. At the same time this brings Cape Malea much to the
west of its true position, and explains why Strabo, in measuring
the length of the peninsula from north to south, drew his line
from Cape Malea to the Corinthian Gulf, instead of from Cape
Tsenarum.

Still more erroneous was his conception of the configuration
and position of Northern Greece. We have already seen that
he considered Cape Sunium, the extremity of Attica, as hut little
farther north than Cape Malea,^ so that a line drawn from
thence to the Isthmus of Corinth would present but a slight
curve, while a straight line (or nearly so) might be drawn from
the Isthmus through the Gulf of Corinth to the straits at its
entrance, and thence to the Acroceraunian Promontory.^ This
conclusion he derived from Eudoxus of Cnidus, a man (as he
justly observes) of mathematical knowledge, and acquainted
with the observations of latitude, as well as familiar with the
countries in question ; and whose authority he consequently
accepts as unexceptionable.^ That such a man should have
arrived at conclusions so wide of the truth in regard to countries
so well known, is indeed a striking proof how little geography
could yet be regarded as based upon any sound and satisfactory
foundation. Yet we shall find — as in so many similar cases —
the influence of this error once introduced into systematic
geography continuing to pervade the works of successive
writers, and even materially affecting the Ptolemaic map of
Greece.

Again, while he points out correctly the manner in which
continental Greece is cut into by a succession of deep bays and
inlets, so as to constitute in a manner a series of successive
peninsulas, his notions of the distances between these bays and
their relative position to one another, are often strangely
erroneous, and it is not always easy to reconcile his statements
with one another.^

2 ii. 1, § 40, p. 92.

3 ix. 1, §1, p. 390.
•> Ibkl. § 2, p. 391.

^ See viii. 1, § 3. His description
of Greece as constituting four suc-
cessive peninsulas is in great measure

270

HISTOKT OF ANCIENT GEOGRAPHY.

Chap. XXI.

§ 20. Concerning the physical geography of Greece he gives
us very little information. He notices indeed, as he could
hardly fail to do, the remarkable formation of parts of Arcadia
and Boeotia, and the manner in which the streams found sub-
terranean channels, and the lakes were discharged by similar
outlets, the stoppage of which from time to time gave occasion
to great inundations or to the extension of the lake- waters far
beyond their ordinary limits. In regard to the Lake Copais in
particular he gives us some curious details, based apparently
on good authority. He adopts also the popular notion that the
river Erasinus in Argolis derived its sources from the Lake of
Stymphalus," and that the Alpheius and Eurotas had their
origin from two fountains close together, the waters of which
pursued their course for some distance underground, and then
issued forth again, the one in Laconia, the other in the
Pisatis.' JSTeither of these facts has been verified by modern
observers, but the last is certainly not without a foundation of
truth ; and that the rivers of Greece frequently pursue a sub-
terranean course for considerable distances is undoubtedly true :
the same phenomenon occurs in other countries composed of
similar cavernous limestones, such as Carniola and Dalmatia.

The mountains of Greece were of course familiar by name to
all men of letters in the days of Strabo, whether geographers
or not. But no attempt is found in his description of the
country to arrange them in groups or point out the geo-
graphical relations of the different ranges. He states, in
accordance with the generally received notion in his day, that
Cyllene was the highest mountain in the Peloponnese, but adds
that, "some said" it was 20 stadia (12,000 feet) in perpen-
dicular height, and others only fifteen.^ This is the only

fanciful, though his conception of the
largest of these, as boimded by a line
drawn from tlie Ambi-acian Gulf on the
west, to the Maliac Gulf on the east,
corresponds to a natural division, which
has been taken as the basis in the
limitation of the modern kingdom of
Greece. His estimate of the width of
this so-called isthmus between the two

gulfs at 800 stadia (80 G. miles) is not
greatly in excess of the truth; the
direct distance in a straight line being
just about 70 G. miles.

« viii. 8, § 4, p. 389.

' viii. 3, § 12, p. 343.

' viii. 8, § 1. jxeyiffTOV S' opos qv aiirjj
KvXK-fjVT)' T^v yovu Koiderov oi fj.iv e'ucoffi
OTaTiiwv (parrlv, ol 5f I'iCdv ■KevTeica'i'Seica..

Sect. 2.

STEABO: EUROPE.

271

instance in which he attempts to give the height of any of the
mountains mentioned : he does not even allude to the different
estimates or alleged measurements that had been made of
Mount Olympus and its neighbours Ossa and Pelion.

§ 21. Of the islands in the ^Egean his account is very
meagre, and their geographical positions are but obscurely
indicated. They were for the most part in a state of great
poverty and decay : ® even Delos having never recovered from
the blow it sustained in the Mithridatic War. His description
of Crete is fuller and more interesting than usual ; and he cor-
rectly points out in this instance the distinct character of the
White Mountains, the most westerly group in the island,
forming a ridge 800 stadia in length, and not inferior to
Taygetus in height, and the isolated mass of Mount Ida, of still
greater elevation, and having a circumference of not less than
600 stadia.^ Yet we have already seen how imperfect was his
notion of the position or dimensions of the island, and how
erroneous and conflicting are his statements concerning its
distance from the nearest points of the mainland.

The words iv avrrj here refer to Arcadia
only, but the lofty mountam group in
the north-east of that region was gene-
rally regarded as the highest in thePelo-
ponnese ; no one apparently suspecting
that it was exceeded in elevation by
Taygetus. The real height of Cyllene,
according to the French commission, is
7788 feet.

It is singular that Strabo does not
refer to the more moderate estimate of
Apollodorus, an author of whom he
made such frequent use. (See Chapter
XVI. p. 618.)

8 The only one of which Strabo dis-
tinctly speaks from personal observation
is the rocky islet of Gyaros, where he
found only a fishing village, whose in-
habitants weie so poor that they
deputed one of their number to repre-
sent to the emperor Augustus their

inability to pay a tribute of 150
drachms ! (Strabo, x. p. 485.) But
Gyaros is one of the smallest and
poorest of the islands, and when visited
by Dr. Eoss in 1841 had no permanent
inhabitants (Eoss, Beise aii/ den
Griechischen Inseln, vol. ii. p. 171).
Yet it was frequently used under the
Eoman Empire as a place of banish-
ment or confinement for criminals. (See
Juvenal, i. 73 ; Tacit. Ann. iii. 68, &c.)
' X. 4, § 4. According to the recent
measurements of Captain Spratt, the
highest summit of the White Moun-
tains and Mount Ida are very nearly
of the same height, both of them ex-
ceeding 8000 feet, and thus s-omewhat
higher than Taygetus, which is in
reality the highest mountain in the
Peloponnese, but does not exceed 7900
feet.

272 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXI.

NOTE A, p. 213.

AGE OF STRABO.

The conclusion of Groskurd that Strabo mnst have "been born as
early as B.C. 66, rests on the assumption that he was not less than
thirty-eight when he was at Corinth in b.c. 29 : an argument that
there is nothing to support, except the idea that his extensive
travels were undertaken with a view to the composition of his
geographical work, and that he was not likely to have conceived
so comprehensive a plan at an early age. But this is all pure
conjecture. Strabo does not tell us that he travelled with a view
to his geography, but that his having seen a considerable part of
the world and visited distant countries, gave him advantages for
such a work. This would rather point to the opposite conclusion,
that he conceived the idea of writing a geographical work because
he had travelled, and therefore after his travels, and not before.
At all events it seems inexplicable that he should have travelled
for the purpose of collecting materials for his geographical work,
and then on his return to Amasia devoted himself to the compo-
sition of a long and elaborate historical work, and delayed com-
mencing the other, which had been his main object, until a period
of life when he could hardly have hoped to complete it.

Clinton places his birth not later than B.C. 54, and is disposed to
put it a few years earlier (perhaps B.C. 60, F. H. vol. iii. p. 553) :
and this is the nearest approximation we can make to its determi-
nation. He was a pupil or hearer of Tyrannion, a grammarian of
Amisus (Strab. xii. p, 548), who was carried off by Lucullus to
Eome ; but this probably took place at Eome, not in Asia. He
mentions in one passage also (lb. p. 568) having seen P. Servilius
Isauricus, the conqueror of the pirates and freebooters of Isauria
and Pisidia, who died in B.C. 44, a statement that we cannot
account for, but this is little to be wondered at. The old general
may well have been in Asia again at a late period of his life,
without our having any record of the circumstance. With regard
to the date of the composition of his work (the most important
point for us) we have the following data : —

1. In the fourth book (p. 206) he says that the Noricans and

Note A. STRABO : EUROPE. 273

Carnians were reduced to subjection by Tiberius and Drusus in
one campaign, and had since then been quietly paying tribute for
thirty-three years. The campaign in question took place in B.C. 1.5
(see Clinton, F. H. vol. iii. ad ann.). This passage therefore could
not have been written before a.d. 18.

2. At the close of the sixth book (p. 288) he speaks of Germanicus
and Drusus, the two Caesars, as both living. As Germanicus died
in A.D. 19, this passage must have been written before that date.

3. On the other hand, at the beginning of the seventh book
(p. 291) he distinctly refers to the triumph of Germanicus after his
victories over the Geimans, in which he had avenged the defeat of
Varus. This triumph was celebrated in a.d. 17 (Clinton, F. B. ad
ann.) : and therefore the passage in question must be subsequent
to that date.

4. In the twelfth and thirteenth books he repeatedly notices the
great earthquake which had lately (recoo-Ti) destroyed or damaged
so many cities of Asia (xii. 8, p. 579 ; xiii. 3, p. 621 ; 4, p. 627).
This took place, as we learn from Tacitus, in a.d. 17 (Tac. Ann. ii.
47) : and as Strabo particularly notices the pains taken by Tiberius
to restore and repair the damaged cities, these passages could not
have been written till the following year (a.d. 18).

5. Again in the twelfth book (c. 1, p. 534) he tells us that
Archelaus, king of Cappadocia, was recently dead, and his kingdom
had been reduced to a Eoman province, but its definite organiza*
tion as such was not yet known. Now Archelaus died at Eome in
A.D. 17 (Tac. Ann. ii. 42), but a considerable time may well be
supposed to have elapsed before the details of the provincial admin-
istration were fully settled. Hence this passage also may well have
been written in a.d. 18.

6. In the seventeenth book, the last of the whole work, he
mentions the death of Juba IT., king of Mauretania, and the suc-
cession of Ptolemaeus, as a recent occurrence (xvii. 3, p. 828). The
exact date of the death of Juba is uncertain : but it did not take
place before a.d. 18 or 19 (see Eckhel, D. N. V. vol. iv. p. 157); it
is therefore probable that the work of Strabo was not finJKhed till
the latter year.

All these indications point to very nearly the same period ; and
may be taken as proving that it could not have been completed in
its present form before the year 18, and most probably was not
published till the following year, a.d, 19. If we adopt Clinton's

VOL. II. T

274 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXI.

date for the author's Mrth, and suppose him to have "been horn
before b.c. 54, he must have been more than seventy-three years of
age before he completed his geographical work. This advanced
period of the author's life, and his residence in a remote provincial
town like Amasia, must have thrown great obstacles in the way of
its extensive publication : and may tend to explain the comparative
neglect with which it was received by his contemporaries.

NOTE B, p. 222.

VOLCANIC ERUPTIONS — THERA AND METHONE.

The volcanic phenomena in the group of islands of which Thera
(now called Santorin) is the principal, have been celebrated in all
ages. The islet thrown up in the centre of the bay, to which
Strabo here alludes, made its appearance in the year B.C. 186.
Another similar eruption took place in the lifetime of the geogra-
pher (a.d. 19) though probably after the date at which this passage
was written. Other outbreaks have taken place at intervals, down
to our own time, the last having occurred as recently as 1866. A
full account of them will be found in Lyell's Principles of Geology
(vol. ii. pp. 65-73, 10th edit.).

One of the other instances cited by Strabo has received less
attention than it deserves. " Near Methone on the Hermionic
Gulf (he tells us, i. 3, § 18, p. 59) a mountain seven stadia in
height was thrown up, after a violent fiery eruption ; it was un-
approachable by day on account of the heat and the smell of
sulphur, but at night there was no bad smell, but a bright light
and great heat, so that the sea around was boiling for a distance
of five stadia, and turbid for not less than twenty stadia. The
shore was piled up with huge fragments of rocks as large as towers."
This was evidently the same eruption referred to by Pausanias (ii.
34, § 1) as having occurred in the reign of Antigonus the son of
Demetrius (b.c. 277-239), so that it must have been described by
competent observers. The whole peninsula of Methon e (or Methana,
as it is more commonly called) is clearly of volcanic origin, but this
is the only recorded instance of volcanic action, within the historical
period.

Note C.

STEABO: EUROPE.

275

NOTE C, p. 259.

DISTANCES GIVEN BY THE CHOROGKAPHER,

The chorographer reckoned Corsica as 160 Eoman miles in length
and 70 in breadth : and Sardinia as 220 miles long by 98 broad.
(Strab. p. 224.) The measurements of modern geographers give to
the former island 116 English miles (124 Eoman) by 51 in its
greatest breadth : while Sardinia measures about 140 G. miles by
60, or 175 Eoman miles by 75. The distances given by the cho-
rographer are therefore largely in excess : but this is still more
the case with his statement that the shortest interval from the
African coast to Sardinia amounted to 300 miles, an estimate
more than double the truth, as the southernmost point of Sardinia,
Cape Spartivento, is really little more than 100 G. miles or 125
Eoman miles from Cape Serrat in Africa. So enormous an error,
in regard to a distance that might be supposed so well known, is
very difficult to account for. Some of the editors of Strabo have
proposed to read 200 for 300, which would accord with the estimate
of Pliny (S N. iii. 13, § 84), but in any case the distance is greatly
over-stated : and it is hazardous to make such arbitrary changes
without authority. It may be added that the distances cited by
Strabo from the chorographer do not in general agree with those of
Pliny.

T 2

{ 2/6 )

CHAPTEE XXII.

STEABO.

Section 1. — Asia.

§ 1. With the eleventh book Strabo commences the descrip-
tion of Asia, which occupies the whole of the following six
books. Throughout this part of his work he in general follows
Eratosthenes very closely, having adopted, as we have seen,
all his leading conclusions in regard to the configuration of
that great continent. Thus he begins with assuming that the
chain of Mount Taurus traverses it continuously from west to
east, preserving approximately the same latitude and direction
from Lycia and the Rhodian Persea, where it abuts upon the
jiEgean, to its eastern termination in the Indian Ocean. But
as he assigns to this range or mountain belt a width in many
places of as much as 3000 stadia, it cannot of course be
considered as a mere chain of mountains, but comprises within
its own extent various tribes and nations, some of them obscure
and insignificant, others of considerable importance, such as
the Armenians, Medians, &c. He then proceeds to describe
the various nations of Asia, according to their position with
reference to this great mountain barrier, dividing them into
those within the Taurus, according to the phrase in use among
the Greeks,^ that is to the north of the chain, and those without,
or to the south of it. Those nations that, as just pointed out.

1 ii. 5, § 31, p. 129. Strabo himself
refers to the phrase as one in general
use (Jx Stj Kol ii/rhs rod Tavf)ov KaKovffiv,
xi. 12, § 1), rather than of his own
selection. It is evident that it must
have originated with the Greeks at an
early period, with reference to the
nations of Asia Minor, who were limited

to the south by the range of the Taurus,
properly so called. When this appel-
lation came to be extended by geo-
graphers (as it was by Eratosthenes
and Strabo) to a great mountain cliaiu
traversing the whole length of Asia,
the expression became singularly inap-
propriate.

Sect. 1. STRABO : ASIA. 2//

lay wholly, or in great part, within the limits of the mountain
tract, he classes with the northern or southern group according
to their proximity and connection with the one or the other.

Northern Asia, or Asia north of the Taurus, he considers as
naturally divided into four portions : first the countries bor-
dering on the Tanais (which he assumes as the boundary
between Europe and Asia) and extending from thence to the
Caspian Sea, and the isthmus that separates the latter from
the Euxine : secondly, the regions extending eastward from
the Caspian to the ScythianS;, who adjoined the Indians to the
north; thirdly the nations that extended from the isthmus
already spoken of to the Caspise Pylge and the range of Mount
Taurus, on the one hand, and to the Halys on the other, thus
comprising the Medians, Armenians, Cappadocians and neigh-
bouring tribes : and lastly the country now called Asia Minor,
extending westward from the Halys to the ^gean, and forming
a kind of peninsula bounded by the isthmus between the
Cilician Gulf and the Euxine.

The portion of Asia south of the Taurus comprised India,
Ariana (a term which he uses in its widest sense), Persia, and
all the nations that extend from the Persian to the Arabian
Gulf, the Nile, and the part of the Mediterranean adjoining
Egypt and Syria. Under this general appellation he includes
Assyria, Babylonia, Mesopotamia, Syria and Arabia.

§ 2. Beginning with the Tanais he tells us that that river
flowed from north to south, but was not, as commonly supposed,
diametrically opposite to the Nile (i. e. on the same meridian
with it), but farther to the east.^ Like the Nile, its sources were
unknown, but while the course of the latter river was known
for a long distance, the Tanais was known only for a short
way above its mouth, on account of the cold, and the natural
difficulties of the country, and still more of the obstacles
opposed by the wild and nomad nations that occupied its

* (peperai jjikv ovv onrh ruv apKrucZv
fuepwp, oh ijl)]v &1S h.v Karh Sidp-erpov avrip-
povs rw NeiAw, Kaddircp uojxi^ovtriv ot

iroWol, aWa kwdiviSirepos fKfivov, xi, 2,
§2.

278

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

banks. On account of this uncertainty, some writers supposed
it to have its sources in the Caucasus, and then to make a great
bend round, so as to fall into the Palus Mseotis from the north.
Others still more absurdly connected it with the Ister. Strabo
justly rejects all these suggestions, and regards it as probable
that it came from sources in the north and at no great dis-
tance.^ The Palus Mseotis he considered, in common with
most other geographers, to have its principal length from
north to south, so that the direct course of navigation from
the strait at its entrance (the Cimmerian Bosporus) to the
mouth of the Tanais, would be from south to north. Its length
in this direction he estimates at 2200 stadia.^

Of the nations north of the Palus Mseotis he appears to
have had no knowledge at all, and only tells us in a vague
and general way that the northern regions towards the Ocean
were inhabited by Scythian tribes, of nomad habits and
dwelling in waggons. South of these were the Sarmatians
(also a Scythian tribe),^ and between these and the Caucasus
the Aorsi and the Siraci, partly nomads, and partly agricul-
tural : besides which the Aorsi carried on a considerable trade,
bringing Indian and Babylonian wares, which they received
from the Armenians and Medians, and transported on the
backs of camels from the Caspian to the Palus Mseotis. By
this means they had amassed considerable wealth, and wore
ornaments of gold.^

Strabo's account of the Greek settlements on the Asiatic
side of the Cimmerian Bosporus — Phanagoria, Corocondame,
Hermonassa, &c., is unusually minute and precise, and his
detailed enumeration of the petty tribes in the vicinity has

^ xi. 2, § 2.

* Ibid. § 3. This is a very moderate
estimate — the real length being about
160 G. miles, or 1600 stadia — and pre-
sents a remarkable contrast with the
exaggerated notions generally current
concerning the vast extent of the Palus
Mseotis.

* ivhoTepw Se rovroov 'XapfidraL, Kol
oItol ^Kvdai. Ibid. § 1, p. 492. It

■would be hasty to draw any ethnological
inference from these words. Strabo
is probably here using the term " Scy-
thians " in the vague and general sense
in which, as he himself tells us, it was
often employed by the Greeks to de-
signate all the nomad nations of
Northern Asia.
8 xi. 5, § 8, p. 506.

Sect. 1. STRABO : ASIA. 2/0

the air of being derived from good information. This was
also the case, as we have already remarked/ with his descrip-
tion of the Tauric Chersonese. Unfortunately he does not in
either case indicate the authority from which his materials
are derived.

§ 3. It is otherwise with his account of the Caucasian tribes
inhabiting the eastern coast of the Euxine, from the Greek
colonies last spoken of to Dioscurias and the mouth of the
Phasis. Here he distinctly refers to the historians of the
Mithridatic wars as furnishing the most recent and accurate
information.^ We have already pointed out how remarkable
a military exploit that prince had accomplished in conducting
his army through so rugged and difficult a country, peopled
by such wild tribes: and we cannot wonder that it should
have attracted so much attention among Greek writers. But
this passage had naturally produced no permanent effect upon
the inhabitants of this coast, who were never really reduced to
subjection by the Romans, and continued in the time of Strabo
to carry on piratical expeditions with light barks, with which
they scoured the coast of the Euxine, and committed great
depredations.^ The nations he places in order along the coast,
proceeding eastward from Sindica (the name under which he
comprises the tract extending from the Cimmerian Bosporus
to Gorgippia), are the Acheei, Zygi and Heniochi, the last of
whom adjoined the Colchians, who occupied the rich and
fertile lands on the banks of the Phasis. The broad valley
of that river formed even in those days one of the chief natural
highways into the heart of Asia, and the nations occupying
the broad tract of comparatively level and fertile country ex-
tending from thence to the Caspian — the Iberians and Alba-
nians — were far more advanced in civilization than their
neighbours on either side. The Iberians in particular, who
inhabited a considerable part of the modern Georgia, are
described as a settled agricultural people, with towns and

' See above, p. 263. « Strabo, xi. 2, § 14, p. 497. » Ibid. § 12.

28o

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXU.

villages, houses with tiled roofs, and some pretension to
architectural effect, and possessing also an organized political
constitution.^ The Albanians, farther east, between the Ibe-
rians and the Caspian, were more devoted to pastoral occu-
pations, and partook in some degree of a nomad character, but
were a tranquil and peaceable people, presenting a great
contrast to the wild and fierce tribes of the mountain districts.^
These three nations, the Colchians, Iberians and Albanians,
occupied what Strabo regards as the isthmus between the
Euxine and the Caspian. The width of this intervening tract
had indeed been greatly underrated by earlier geographers, thus
giving it much more the character of an isthmus than it really
possessed, and even Posidonius had estimated it at only 1500
stadia from sea to sea.^ Strabo on the contrary, though he
continues to designate it as an isthmus, assigns it a breadth
of 3000 stadia, and even this is considerably below the truth.*
It is probable that he had more accurate information concern-
ing these regions, in addition to the historians already cited,
from the circumstance that Moaphernes, who was his mother's
uncle, had held the government of Colchis under Mithridates
the Great.^ That monarch derived from thence the greater
part of the timber which he required for building his fleets.
In addition to this Colchis furnished flax, hemp, and pitch in
abundance, as well as all kinds of fruit, while the numerous
rivers by which it was traversed afforded every facility for
conveying its produce to the coast. Strabo indeed appears
to have been fully alive to the richness and natural impor-

» xi. 3, § 1.

' xi. 4, § 1.

^ Posidon. ap. Strab. xi. 1, § 5, p.
491. He even compared it with the
isthmus from Pelusium to the Red Sea ;
and added that he believed it was
much about the same distance from tlie
Mseotis to the Ocean.

■' Strabo, ■ihid. The direct distance
from the mouth of the Phasis to the
Caspian near the mouth of the Cyrus is
about 380 G. miles, or 3800 stadia.

The actual shortest line as measured
on the map from sea to sea does not
exceed the 3000 stadia given by Strabo ;
but as such a line crosses the chain of
the Caucasus obliquely, it could never
have suggested the idea of an isthmus ;
and Strabo himself tells us that his
statement refers to the distance from
the mouth of one river to the other.
This is indeed the only line by which
it is practicable to pass from sea to sea.
' xi. 2, § 18, p. 499.

Sect. 1.

STRABO : ASIA.

2«I

tance of this country — one of the fairest regions of the world —
though in modern times so little known until a very recent
period.^

§ 4. Of the mountain chain of the Caucasus itself he gives
a clear and unusually full account. He justly describes it as
extending like a wall across the isthmus which separates the
Euxine from the Caspian, and impending over the eastern
coast of the former sea through its whole extent from the
confines of Sindica to Dioscurias. At the same time it throws
out offshoots of a lower elevation, by means of which it is con-
nected with the mountains of the Moschi, and through them
with the Armenian mountains, and the ranges that belong to
the system of the Taurus. The lower ranges and slopes of the
Caucasus were covered with extensive forests, inhabited by
mountain tribes, who subsisted principally on game, wild fruits,
and milk. The higher summits were covered with snow and
ice, and inaccessible in winter, but in summer the inhabitants
ascended them, wearing broad snow-shoes of raw hide furnished
with spikes, and brought down their burdens from thence, by
sliding down on hides. As one descended the slope to the
north, the climate became less severe, notwithstanding the
more northern latitude, on account of its proximity to the
great plain of the Siraci.'' In another passage he describes
particularly the pass leading into Iberia from the northern
nomad nations :^ evidently the same as that now called the
Pass of Dariel, which is indeed the only practicable pass
across the whole range, and must therefore, though presenting
great natural difficulties, have been more or less frequented in
all ages. It was, he says, a steep and difficult ascent for
three days (coming from the north) and after that a narrow
pass for four days' journey along the valley of the Aragus,
so narrow as only to allow one person to pass at a time and
guarded at its entrance by a very strong fortress. The river

^ No mention is found of gold among
the productions of Colchis, notwith-
standing it3 supposed connection with

the fables of the golden fleece.
' xi. 5, i}§ 6, 7, p. 506.

« lb. 3, § 5, p. rm.

282

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

Aragus still preserves the name Aragwa, and Strabo's acquain-
tance with the name of this unimportant stream, as well as with
other minor tributaries of the Cyrus, shows the accuracy of his
information.^ In describing the Cyrus itself and the Araxes
as flowing into the Caspian by separate mouths, it is not un-
likely that his statement was correct, though the Araxes now
joins the Cyrus more than 70 miles from its mouth.^ But the
whole of this country is a swampy delta, and the alluvial
accretions of land proceed with such rapidity that great
changes may have taken place since the time of Strabo.

Dioscurias, which he, in common with almost all other
ancient geographers, regarded as the easternmost point of the
Euxine,^ was a considerable emporium of trade, and resorted
to by all the neighbouring nations, who even • in those days
spoke so great a variety of languages and dialects, that it was
said not less than seventy distinct languages were spoken
there.^ The trade with the interior of Asia was carried from
the mouth of the Phasis, where there was a city of the same
name, up the river by water as far as a fort called Sarapana,
from whence it was four days' journey overland, by a road
practicable for vehicles, to the Cyrus.* The pass of Suram
across the watershed uniting the two mountain systems of the
Caucasus and the opposite range is indeed one of very mo-
derate elevation, and presenting little natural difficulty.

Strabo deservedly rejects the appellation of Caucasus given
by the Macedonian soldiers to the lofty range of the Hindoo
Koosh, between Bactria and India, and ascribes it to the desire
of flattering Alexander by associating his conquests with the
name of the mountain chain that had the reputation of being
the loftiest in the world, and was celebrated in the Greek
fables in connection with Prometheus.^

3 xi. 3, § 2.

' lb. 4, § 2. Strabo himself notices
the great amount of alluvium brought
down by the river Cyrus.

^ xi. 2, § 16, p. 497. See Chapter
XVI. p. 636.

' Ibid. p. 498. Some even increased
the number to three hundred ! Pliny
ascribes this last statement to Timos-
thenes. (Plin. H. N. vi. 5, § 15.)

^ Ibid. § 17, p. 498.

* xl. 5, § 5, p. 505.

Sect. 1.

STRABO: ASIA.

283

§ 5. With regard to the Caspian he shared, as we have
already seen, in the opinion of Eratosthenes, and all other
geographers since the time of Herodotus, that it was an inlet
from the northern ocean, similar to the Persian Gulf on the
south. So clearly indeed was this idea fixed in his mind that
he describes the sea and the nations on its banks, as they
would present themselves to a person sailing in from the north.^
At first he tells us the gulf is rather narrow, but afterwards
widens out as one advances, until in its innermost (i. e.
southern) portion, it is about 5000 stadia in width. The
length from the entrance to the inmost bight is much about
the same,^ but slightly more. These dimensions he has
probably taken from Eratosthenes : he expressly cites that
author as his authority for the distances around the shores of
the Caspian to the mouth of the Oxus and from thence to that
of the laxartes.® Patrocles was evidently the original source
from which both derived their information,^ as he was also for
the statement advanced with confidence by Strabo as well as
Eratosthenes that both the Oxus and laxartes fell into the
Caspian Sea, after pursuing separate courses from their sources
to their mouths.^ It does not appear that Strabo had any
further information concerning these regions than what he
derived from these earlier geographers, with the exception of
some particulars respecting Hyrcania which : he cites from
Apollodorus of Artemita. He was indebted to this writer
especially for clear notions respecting the river Ochus, which

« xi. 6, § 2, p. 507.

'^ lb. § 1. The breadth is greatly-
exaggerated— that of the southern por-
tion of the Caspian being really less
than 2'40 G. miles : while the length
is in reality nearly three times the
breadth. Strabo himself adds, after
citing the statements of Eratosthenes,
that allowance must be made for con-
siderable vagueness in regard to regions
so little known, especially in respect to
distances {pel Se Trept Twv iv rfj fieplSt.
ravrrj koI rots tirl rocrovrov eKreroTncr-
(xevoLS o.irKovffrepov aKOveiv, Koi jUaAicTTa
■jrepl Tuy Sia(TTrjfJi.dTwv). He had in

reality much greater reason than he
was aware of, for this caution.

8 See note to Chapter XVI. p. 644.

3 He is cited by Strabo (xi. 7, § 1,
p. 508) as describing the Caspian as
just about equal in size to the Euxiae,
which, as a rough approximation, is a
correct estimate.

1 xi. 7, §4, p. 510 1 11, §5, p. 518.
6 fj-ivroi 'la^dpTTis air' apxfjs fj-expi- reAous
eTep6s iffTi rod ''D.^ov koL els fief rrjv
avrjjv nXsurSiv QaKarrav, al S' i/j.fio\al
SiexovtTiv aWrJKaiv, lis <{>7](n UarpoKXijs,
irapaadyyas us oySorJKovra.

284. HISTOEY OF ANCIENT GEOGKAPHY. Chap. XXII.

had been ignored by previous geographers or supposed to be
a mere tributary of the Oxus, but according to Strabo had
a distinct course and fell into the Caspian by a separate
mouth. ^

§ 6. Proceeding eastward from Hyrcania and the south-east
corner of the Caspian, Strabo tells us that one still has the
chain of Taurus on the right hand, which forms a continuous
range from Armenia to this point, and is known generally by
the native name of Parachoathras. It was not till after passing
the land of the Arians that the great chain assumed the name
of Paropamisus, while it was erroneously termed by the Mace-
donians the Caucasus. It was this same chain which was
prolonged without interruption to the Indian Ocean, though
known by different names, as the Emoda, Imaus, &c.^

On the left or towards the north, were situated the Dase,
nearest to the Caspian Sea, and beyond them the Massagetae
and Sacse. All these nations are included by Strabo under
the general name of Scythians,* though as he justly observes
the earlier Greek writers only gave this name to the European
Scythians and those adjoining the Tanais and Palus Maeotis,
and distinguished the Asiatic nomad tribes as the Massagetae
and Sacse. Hence we find these names occurring as those of
nations with which Cyrus made war on his extreme frontier.
The laxartes was the boundary which separated the Sacae, or
nomad nations included under that name, from the Sogdians,
who as well as the Bactrians, were a comparatively civilized
people, even before they had shared in the Greek civilization
introduced by the Bactrian kings. That monarchy had been
already overthrown before the time of Strabo, and he distinctly
tells us that the barbarians who had wrested the fertile pro-
vinces of the Bactrians and Sogdians from their Hellenic rulers
were tribes from beyond the laxartes, to which he gives the

2 xi. 7, p. 509. The Ochus of Arte- j until quite- recently, as imperfectly

miclorus may be certainly identified known as that of the Ochus among the

with tlie modern Attrek; the course ancients. ' xi. 8, p. 511.

of which was even in modern times, I ' xi, 8, § 2, p. 511.

Sect. 1.

STRABO: ASIA.

285

names of Asii, Pasiani^ Tochari, and Sacarauli, but apparently
includes them all under the general term of Sacse.^ The name of
Sac8B or Sakas was in fact the Persian appellation for the nomad
nations on their northern frontier,® and doubtless applied with
as little regard to their ethnic affinities or subdivisions as was
that of Scythians by the Greeks, or Tartars in modern times.

§ 7. Of the countries south of the laxartes, Sogdiana, Bac-
triana, Aria and Margiana, Strabo gives but a brief account,
and appears to have known very little, if at all, more than
what he learned from Eratosthenes and the historians of Alex-
ander. He has indeed given us a few interesting historical
particulars concerning the growth and extension of the Greek
kingdom of Bactria, which at one time extended over a con-
siderable part of the north-west of India, down even to the
mouths of the Indus, while they carried their arms eastward as
far as the Seres and Phryni.'' This notice is taken from Apol-
lodorus of Artemita, and is therefore the first mention of the
Seres in any ancient writer.^ But Strabo evidently did not
see its importance, as he brings it in merely in passing. It is
curious indeed that although the use of silk was, as we have
seen, already familiar to the Eomans in his day, and the name
at least of the Seres was well known to them, he never alludes
to their existence, except in this incidental manner, nor does
he even mention the trade in silk, which must have already
assumed considerable importance.^ He had certainly no idea

* xi. 8, § 2, p. 511. This event took
place about b.c. 126, though the exact
date cannot be fixed. The only other
writer who mentions it is Trogus Pom-
peius, of whom unfortunately only the
epitome is preserved. He terms the
Scythian nations who occupied Bactria
and Sogdiana, Sarancse, and Asiani
(Prolog, lib. xli.); but in another pas-
sage mentions also the Thocari or
Tochari. (lb. xlii.)

° Til is we are distinctly told by
Herodotus (vii. 64, 01 yapXiipffai irdyTas
rovs 'ZKvdas KoK^ovcri 2a/cas), and his
statement is fully confirmed by the

Persian inscriptions. (See Kawlinson's
note, on the passage.)

' xi. 11, § 1, p. 516.

* Concerning the age of Apollodorus^
see Chapter XX. p. 162.

^ In the only passage (p. 694) where
he mentions the 'S.ripLKa — a kind of
woven stuffs {vcpdafiaTo) made of a sort
of thread scraped from the bark of trees
(e/c Tivcov (pKoMv ^aivofievris ^vaffov — he
regards them as an Indian product,
analogous to cotton. But this passage
is tdken from Nearchus ; and the men-
tion of the Seres as one of the most
long-lived of the Indian tribes, attain-

286

HISTOKY OF ANCIENT GEOGKAPHY. Chap, XXII.

of their real geographical position, and supposed them to be
merely a nomad race of Scythians.

The north of Asia, as well as the regions east of Sogdiana,
was indeed, as he expressly tells us,^ a mere blank to him :
and it was only by conjecture that he inferred them to be
occupied by nomad nations, resembling the Scythians in their
habits of life. It was not certain, though alleged by some,
that the sea extending around from India to the Caspian had
ever been navigated, though it was believed, on the authority
of Patrocles, to be possible.^

§ 8. It seems to have been also in his day a received con-
clusion in geography,^ though in fact resting upon mere
conjecture, that the great mountain chain which traversed the
whole continent of Asia from west to east, and was called
Imaus in its easternmost continuation, ended in the Indian
Ocean without projecting in any material degree beyond the
rest of India. From the promontory thus formed (to which he
gives the name of Tamarus, adopted from Eratosthenes), which
separated India from Scythia, the coast towards the north
trended rapidly away, so that this portion of Asia assumed
something of a pyramidal form, having the vertex at the
promontory already mentioned.

It is singular that a notion so utterly devoid of foundation
should have assumed so consistent and definite a form. It is
almost more singular to find a geographer like Strabo, though
admitting his entire ignorance of this part of Asia, proceeding
to define its length and breadth ; the former of which along
the chain of the Taurus from the Caspian Sea to the Indian
Ocean he determines to be about 30,000 stadia, or 3000 Gr.
miles, while the breadth was less than ten thousand stadia.*

ing the age of more than 200 years
(pp. 701, 702), is derived from Onesi-
critus, and cannot be regarded as show-
ing any real acquaintnnce with the
nation liearing that name.

' xi. 11, § 6, p. 518.

* ovx o^oAoyovffi Se, '6ti TrepieirXevcrdv

Tives airh Trjs 'Ii/SiKrjs iir\ tt}v 'TpKavlau,
Hn 5e Svvarhv, TlaTpoKATJs eipTj/ce, xi. II,
§ 6, p. 518. Pliny and other later
writers, as we shall see, asserted that
the passage had been actually made.

« Ibid. § 7, p. 519.

* Ibid. p. 519.

Sect. 1.

STKABO: ASIA.

287

It is unnecessary to point out that these conclusions are mere
inferences, based upon the assumed length and breadth of the
whole continent and other assumptions equally unfounded.

§ 9. Returning from these little known regions towards the
west, Strabo proceeds to describe two countries — Armenia and
Media — which he considers as situated rather in the range of
Mount Taurus than either to the north or south of it, being so
intersected and encircled by the various ramifications of that
great chain that they could not be assigned to the group of
nations on either side of it.^ Armenia especially he correctly
describes as being almost entirely a land of mountains and
high table-lands,® which contained the sources of several great
rivers, especially the Euphrates and the Tigris, and the Araxes
towards the Caspian, as well as others of minor importance
which flowed to the Euxine Sea.

On this occasion he gives a general outline of the direction
and conformation of the chain of Taurus and its subsidiary
ranges, which shows a considerable acquaintance with the
orography of this part of Asia.' Mount Taurus (he tells us)
takes its rise in Caria and Lycia, but does not at first attain
any considerable height or breadth. It first rises to a great
elevation opposite the Chelidonian islets on the frontiers of
Lycia and Pamphylia,^ and from thence extends eastwards to
the north of Cilicia, a great part of that country being formed
by the valleys intercepted between the offshoots of the great
mountain range. Beyond that it throws off two great arms or
branches, the one called Anti-Taurus, towards Cappadocia and
Armenia Minor, the other, Amanus, towards the south, ex-
tending to Syria and the Euphrates. The main chain itself,
though cut through by the Euphrates, is continuous with the

* xi. 12, § 1, p. 520.

8 dpoTreSia, xi. 14, § 4, p. 528.

' xi. 12, §§ 2-4.

' On this account many writers con-
sidered that the headland opposite to
these islands was the beginning of the
chain of Taurus, but Strabo properly
points out that the mountain ridge

which separated Lycia from the dis-
tricts of the interior (the Cibyratica)
was in fact a continuation of the
Taurus, wliich was thus prolonged into
the Ehodian PersBa, and might be con-
sidered as ending in the mountain pro-
montories opposite to Khodes (xiv. 2

§ 1 ; 3, § 8).

288 HISTORY OP ANCIENT GEOGRAPHY. Chap. XXII.

mountains of Armenia on the other side of that river, and it is
here that it rises into a great mountain mass, sending out
offshoots in different directions, known by the name of Pary-
adres, and other local appellations, and forming the boundary
of Armenia on the side of the Iberians and Albanians. From
this great central mass was continued another chain towards
the east, known by the name of Parachoathras, bordering on
the Caspian Sea and extending through Media Atropatene and
the Greater Media to the Caspian Gates, whence it was con-
tinued still farther east along the confines of Aria. It was
this east and west prolongation of the chain that was regarded
by Greek geographers from Eratosthenes to Strabo as the true
continuation of the Taurus, which served to connect it with
the great ranges of the Paropamisus or the Hindoo Koosh.
But besides this there were several subsidiary ranges to the
south of the Euphrates in its upper course, and it was to the
most elevated portion of these that the Greeks gave the name
of Niphates, in which according to Strabo the Tigris took its
rise.^ From thence there branched off towards the south
another great ridge called Zagrium or Zagros, extending a
long way, and forming the separation between Media on the
one hand and Assyria and Babylonia on the other, till it joined
on to the mountain ranges of Susiana and Persia.

§ 10. In connection with this subject Strabo gives an account
of the course of the Euphrates and Tigris, both of which rivers,
as he correctly tells us, rose in the mountains of Armenia.^
The lower part of their courses, where they encompassed
Mesopotamia, had long been familiar to the Greeks, and even
the upper part of that of the Euphrates was well known to
Strabo, who correctly describes it as rising in the northern
portion of the Taurus, and flowing in the first place from east
to west through the Greater Armenia to the frontiers of the
Lesser : then separating that province from Acilisene and
making a sudden turn to the south where it reached the

xi, 12, §4. » Ibid. 12, § 3; 14, § 2.

Sect. 1. STRABO : ASIA. 289

confines of Cappadocia ; thenceforward leaving Cappadocia
and Commagene on the right, and Acilisene and Sophene on
the left, till it issued into the plains of Syria, and took another
great bend towards Babylonia and the Persian Gulf. In this
description Strabo clearly had in view only the northern branch
of the Euphrates — that which flows near Erzeroum, and has its
sources in the mountains not far from that city. It is this
river which alone was regarded both by Greek and Roman
writers as the true Euphrates, and which is still distinguished
by the name of Frat. The southern arm or affluent, called at
the present day the Murad Tchai, which is considered by
modern geographers, as well as by native Armenian writers, as
one of the main sources of the Euphrates, and is in fact the
larger river of the two, was treated by ancient geographers
as a mere tributary, and is not even noticed by Strabo in
describing the geography of Armenia.^

The Tigris he describes as rising in the southern slopes of
Mount Taurus, and says that its sources were distant from
those of the Euphrates about 2500 stadia. This river also rises
from two different and distant sources, forming two different
arms, which, after holding separate courses, unite between
Diarbekr and Mosul. But there seems no doubt that Strabo
regarded as the main source of the Tigris the stream that rises
in Mount Niphates, and flows from thence due south until it
joins the river of Diarbekr.^ From the terms in which he
speaks, both here and elsewhere, of the outflow of the two

* It was, however, in all probability
this river, which under the nanie of
Arsanias .had acquu'ed celebrity in the
Mithridatic Wars by the defeat of

^ He has elsewhere a strange story
of the Tigris flowing through the Lake
Arsene (the Lake of Van) without
mingling its waters, which fell into a

Tigranes on its banks by LucuUus ' great chasm at one end of the lake, and
(Plut. Lucull. 81), and which figures ' after llowing for a long distance under-

again in the wars of the Eomans with
tiie Armenian kings (Tacit. Annal. xv.
15). It is described by Pliny as a tribu-
tary of the Euphrates (Plin. H. N. v. 24,
§ 84). It is also, as has been already
observed, the river to which Xeuophon
gives the name of Euphrates. (See
Cliapter X. p. 353.)

ground, reappeared in tlie district of
Chalonitis (xi. p. 529). The last addi-
tion is utterly unintelligible, the dis-
trict known fis Chalonitis being far
away in the eastern part of Assyria, at
the foot of Mount Zagros. There is
probably some mistake in the name.

VOL. II. U

290 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXII.

great rivers into the Persian Gulf, it is clear that each of them
in his day had still its separate outlet to the sea, instead of
uniting their streams into one as they do at the present day.*

§ 11. Armenia had, as we have already seen, been iSrst
opened out to the knowledge of geographers by the campaigns
of Lucullus and Pompey in the Mithridatic Wars, while the
expedition of M. Antony against the Parthians had first made
them acquainted with Media Atropatene or Azerbijan. Strabo
availed himself of the materials thus furnished him, and
there was probably no part of Asia of which his knowledge
was more in advance of that of Eratosthenes. But the rugged
and mountainous character of the two countries, and the
intricate and complicated relations of the mountain chains by
which they are traversed, opposed great difficulties to an
accurate geographical knowledge of them — and Strabo's in-
formation was still very imperfect. Thus we find him describing
three lakes, one to which he gives the name of Spauta,^ in
Media Atropatene, remarkable for its excessive saltness ; the
other two, which he calls Mantiane and Arsene or Thopitis in
Armenia. Both of these had also salt or brackish water, the
former especially, which he calls the largest lake next to the
Maeotis, and which had salt-works on its shores. There are in
fact only two lakes to which his description can possibly apply :®
the Lake of Van, which is that called by him Arsene or
Thopitis — it is the Arsissa of Pliny and Ptolemy — and the

■* It may, however, be questioned
whether he is not here simply follow-
ing Eratosthenes, without enquiry as
to what changes might have taken
place in the interval.

^ It is a very plausible suggestion
of M. St. Martin ( Mem. sur VArmem'e),
and the recent editors of Strabo, that
this name, which is written STrat^ra in
all our MSS., should really be Kairavra,
and is a corruption of the Armenian
name Kapotan, signifying the blue lake.
Strabo himself tells us that this was

such meaning. (See Kramer's note, ad
loc.) The Lake of Urumiah, also
called Shahi, which is without a doubt
the lake meant by Strabo, is remark-
able at the present day for its excessive
saltness. • According to Col. Monteith
(Journal of Geogr. Soo. vol. iii. p. 56) it
contains nearly twice as much salt as
the sea.

" There is indeed a third lake, in
the north of Armenia, of considerable
size, now called the Lake Goukcha,
but this is out of the question, as its

the signification of the Armenian name i waters are perfectly fresh and conse-

(Kuafn €pfj.7]V€vde'L(ra), tiiough he erro- j qucntly abound in fish. (Sec Col. Mou-

neously connects tlii.s interpretation i toilh, I.e. p. 41.)

with the name MafTianh which has no 1

Sect. 1. STRABO : ASIA. 29I

Lake of Urumiah on the borders of Armenia and Atropatene,
whicli is clearly the one that he describes under the name of
Spauta in the one country, and of Mantiana in the other.''
Such a confusion might easily arise in writing from different
materials, but it shows how far he was still removed from
possessing a correct geographical idea of the countries in
question.

In other respects his account of Armenia and the neigh-
bouring province of Atropatene is generally accurate enough ;
and he had a clear knowledge of the topographical relations of
the various provinces and districts into which Armenia was
divided, or by which it was surrounded ; though modern geo-
graphers have much difficulty in determining their position
and extent. Of the greater Media, or the country generally
known by that name, he treats in connection with Media
Atropatene, although in a geographical point of view it would
certainly have been more properly classed with the pro-
vinces of the Persian Empire south of the Taurus. His
geographical account of this province is brief and summary,
and we perceive immediately that he had here no recent
sources of information, and was forced to fall back entirely
upon Eratosthenes and other authorities of the Macedonian
period. Media, indeed, was in all times a country imperfectly
known, and its boundaries seem never to have been very accu-
rately defined. Strabo's own account is by no means clear,^
and there can be little doubt that the relations and limits of
the mountain tribes, that were by some regarded as belonging
to Media, by others to the adjoining provinces, were in reality
subject to frequent changes. The mountaineers of the lofty
ranges of Zagros were evidently as little really subject to the
Persian or Macedonian rulers, as the Koords of the present

' Strabo, xi. 13, § 2 ; 14, § 8. In But the Oossseans, according to his own

point of fact every one of these lakes is
known at the present day by at least
two different names.

8 He tells us (xi. 13, § 6) that Media
was bounded on the east by Parthia
and the mountains of the Cossseans.

account, as well as that of other
writers, inhabited the ranges of Mount
Zagros, on the west of Media, and ad-
joining the district of Elymais; nor
have we any trace -of the existence of
such a people farther east.

u 2

292 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXII.

day to the Turks or Persians. The Cossseans in particular
were in the habit of levying tribute from the Persian kings,
when they moved with their court from Babylonia to Ecbatana,
their usual residence in summer.^ But according to Strabo
Media might be properly considered as extending from the
pass called the Median Grate, leading from Ecbatana into
Babylonia, on the west, to the Caspian Gates on the east ;
a distance which he estimates at 4100 stadia.^ He justly
describes it as a cold and upland country, almost entirely
mountainous, with the exception of the portion near the
Caspian Gates — the environs of the modern Teheran — which
was a fertile and productive plain. Even in the mountain
districts also there were some fertile valleys, and both Media
and Armenia were renowned for their breed of horses, vast
numbers of which were reared in both countries, and furnished
annually as tribute to the Persian kings.

§ 12. Strabo now returns nearer home, and his twelfth book
is occupied with the description of Cappadocia and Pontus,
and the northern provinces of Asia Minor, along the coast of
the Euxine. Here he derived great advantages from the
proximity of these countries to his native city : and he had
himself travelled through a considerable part of the interior.
Unfortunately he has not thought fit to record the extent or
course of his travels, but as he distinctly tells us that he had
visited in person the Cappadocian Comana,^ which was situated
quite in the interior, in the upper valley of the Sarus, he must
have traversed a considerable portion of that province. It is
not unlikely that he returned to his native city by this route.

9 Strabo, I. c. \ dorus should not have been better

1 This is greatly over-estimated, I informed.

though probably taken from Apollo-
dorus of Artemita, the historian of the
Parthian Wars, whom he cites else-
where (xi. p. 519), for the total dintauce
from his native city to the Caspian,
which he estimated, still more erro-
neously at SOOO stadia. As Artemita
lay on the high road from Seleucia to
Ecbatana, it is strange that ApoUo-

The pass across Mount Zngros, to
which he gives the name of Median
Gate (M7)5i/c); ttvKti, xi. 13, § 8), is
clearly that leading from Hamadan by
Kermanshah to Bagdad, which must in
all ages have formed one of the prin-
cipal passes across the great mountain
chain.

2 xii. 2, § 3, p. 535.

Sect. 1.

STEABO: ASIA.

293

across Cilicia and Cappadocia from Tarsus to Amasia, His
description of Mazaca (better known by its later name of
Ca^sarea) at the foot of Mount Argaeus, and his information
concerning the ascent of that mountain, also point apparently
to the result of a personal examination.^ The general cha-
racter that he gives of the country as an open upland tract,
almost wholly bare of wood, but not devoid of fertility, and
producing abundance of corn, as well as supporting immense
quantities of sheep and an excellent breed of horses, is fully
confirmed by the descriptions of recent travellers. He notices
also various mineral productions of the country, the most im-
portant of which was the red earth, commonly known as the
Sinopic, from its being exported from that city, but which was
really found in Cappadocia.* The vestiges of volcanic pheno-
mena at the foot of Mount Argeeus had also attracted his atten-
tion, and he describes the plain below Mazaca as impregnated
with fire, which was visible in holes and chasms for an extent
of many stadia.^ If this account be not greatly exaggerated,
there must have been volcanic outbreaks of the mountain at a
period much more recent than is generally supposed. Strabo
however does not mention any tradition of such an event.

He gives a distinct account of the course of the two im-
portant rivers, the Sarus and the Pyramus, which took their
rise in Cappadocia, and thence bursting their way through

^ xii. 2, § 7. He tells us that Mount
Arira3us is the most lofty of all (in
Asia Minor ?), and its summit is
covered with perpetual snow : that
those who ascend it, who are few in
number, assert that in fine weather
both seas, the Euxine and the Gulf of
Issus, might be seen from its summit.
This story has every appearance of
being gathered from the inhabitants of
Mazaca. The first traveller in modern
times who made the ascent was Mr.
William Hamilton, who met with
cloudy weather, but did not believe it
possible that the two seas could be seen
in any case, on account of the high
mountains which intervene both to the

N. and the S. (Hamilton's Asia Minor,
vol. ii. p. 280). Its elevation he cal-
culates at 13,000 feet, and there is not
only much snow on the summit, but
extensive glaciers descend from thence
on its northern and eastern flanks.
Since that time the mountain has been
again ascended by M. TchihatcheflF.

* xii. p. 540. He terms this fj.i\ros,
and evidently considers it as identical
with that of Spain, which is true cin-
nabar ; but the Sinopic fxixros was only
a kind of bright red earth, of an
ochreous nature.

'* xii. 2, § 7. irvpiXrfKTa ireSia Kal
fj.effTa, ^oQpoiv irvphs iirl araSiovs iroA-
Kovs.

294

HISTOKY OP ANCIENT GEOGRAPHY.

Chap. XXII.

the lofty ranges of Mount Taurus flowed through Cilicia to the
sea.^ Of the remarkable gorge by which the Pyramus forced
its way through the mountains he has given us a particular
description, which was evidently derived from personal obser-
vation.''

§ 13. With Pontus he was of course familiar, and it is an
important fact in the ethnography of Asia that he distinctly
confirms the statement of Herodotus and other writers that the
Cappadocians, who had originally extended from the chain of
Taurus to the Euxine, were of Syrian extraction, or belonged
to the great Aramaean race, in common with the Syrians and
Assyrians.^ At the same time he speaks of the various tribes
inhabiting the mountain ranges near the Euxine, the Moschi,
the Tibareni, and the Chaldseans, — in terms which seem to
imply that they were separate tribes, and they may probably
have been of a different race, perhaps more connected with
their Armenian and Caucasian neighbours. But the ethno-
graphy of these mountain tribes is a problem of hopeless
perplexity. It is certain however that the separation of
Pontus from Cappadocia was a purely artificial one, arising in
the first instance from the division of the great province of
Cappadocia under the Persian Empire into two satrapies,
which after the Macedonian conquest gradually became con-
solidated into separate kingdoms.^ The boundary as finally
established was one of the mountain ridges parallel with the
Taurus, which traverse this part of Asia Minor, but it cannot
now be identified.^

Strabo has left us a detailed enumeration of the districts
into which Cappadocia was divided in his time, eleven in all,
but several of these are otherwise unknown, and cannot be

« xii. 2, §§ 3, 4, p. 536.

' The words ourais Mojxev are con-
clusive on this point.

It is only quite of h\te years that
these defiles of the Taurus, which con-
nect the uplands of Cataonia with
Cilicia, have been explored by modern
travellers.

* This is disputed by Sir H, Eawlin-
son [Herodotus, vol. i. pp. 653-4), but
his arguments appear to me far from
convincing ; and it is difficult to see
how Strabo could have been mistaken
upon such a pcnnt.

" Strabo, xii, 1, § 4, p. 534.

1 xii. 2, § 10, p. 540.

Sect. 1.

STRABO: ASIA.

29s

determined with any certainty. There were only two cities in
the whole country, which he considers worthy of the appel-
lation, Mazaca and Tyana : the other districts contained only
a scattered population, with a few strongholds or fortresses,
among which that of Nora, so long defended by Eumenes, was
one of the most celebrated. Even the fertile district of
Melitene, adjoining the Euphrates, which presented a great
contrast to the rest of Cappadocia from its abounding in vines
and fruit-trees, did not contain a town of any importance.^

The description of Pontus by Strabo ^ is one of the most
complete and satisfactory portions of his work, and is by far
the best account that we possess from any ancient writer of
a country that until very recently was but imperfectly known
to modern geographers. With it he associates the Lesser
Armenia, obviously on account of its situation west of the
Euphrates, as that district was politically connected either
with Armenia properly so called, or with Cappadocia.

On the other hand the mountain tribes of the Tibareni and
Chaldaeans who inhabited the ranges of Paryadres, between
the confines of Armenia Minor and the Euxine, were under
the rule of Pythodoris, who bore the title of Queen of Pontus.
These mountaineers, as well as their neighbours the Moschi,
who more immediately adjoined the confines of Colchis, were
still in a very rude and barbarous condition, dwelling in great
forests, and subsisting on wild fruits and the flesh of animals
procured by the chase.* Some of them even lived in trees ;
others in high towers. These last were the Mosynoeci of
Xenophon, but Strabo does not recognize the name as one
existing in his time.^ He tells us however that the people to
whom he gives the name of Chaldaeans, were the same who
had been formerly called Chalybes,^ and had been renowned

2 xii. 2, § 6, p. 537. The strong fort-
ress of Tomisa, which figures promi-
nently in the Mithridatic Wars, was
situated on the eastern side of the
Euphrates, and consequently belonged
properly to Sophene. lb. § 1, p. 535.

3 xii. 3. The geography of this part
of Asia Minor was but little known in
modern times before the travels of Mr.
William Hamilton in 1836 (published
in 1842). * Ibid. § 18, p. 549.

* Ibid. « xii. 3, § 19.

296

HISTORT OF ANCIENT GEOGEAPHY.

Chap. XXII.

from the earliest ages as workers in iron : and he then enters
into a long and tedious discussion^ to show that these were
the same people termed by Homer Halizones, who dwelt about
a place called by the poet Alybe, " where was the birth-place
of silver." ^ The connection of the names Alybe and Chalybes
would be probable enough, were it supported by any other
arguments : but it is not ; and had the poet ever heard of so
distant a people as the Chalybes, it would doubtless have been
as workers in iron, the natural abundance of which in the
region in question must have attracted attention from a very
early period,^

Mithridates the Great having extended his dominion along
the shores of the Euxine from the borders of Colchis to
Heraclea, thus including all the sea-coast of Paphlagonia, and
a part of that of Bithynia, Strabo has adopted the same exten-
sion, and has described under the head of Pontus the Avhole
southern coast of the Euxine, beginning from Heraclea. This
long line of sea-board was studded throughout with Grreek
colonies, some of which, as Heraclea, Sinope, Amisus, Phar-^
nacia and Trapezus, were flourishing and important commer-
cial cities ; while many smaller settlements are noticed in
detail by Strabo, who was evidently well acquainted with the
whole line of coast, and has given a careful enumeration of its
rivers and headlands, as well as of the towns which lined its
«hores.^

§ 14. It is quite otherwise with the interior of the country.
So far as the province of Pontus, properly so called, is con-

' xii. 3, §§ 20-24.

avTap 'A\L^uiv<av 'OSi'os Kol 'Etrio'Tpo^o?

VPXO"' ,^

TrjS.60ev ef AAv^r)?, bOev apyvpov cctti
yeueBKr),

Homer, Iliad, II. v. 857.

" See the interesting account of the
mode in which iron is worked at the
present day in this district, in Hamil-
ton's Travels in Asia Minor, vol. i. pp.
271-277. Silver mines are now worked
at Gumisoh Khana in the interior,
south of Trcbizond, but these are not

mentioned by Strabo ; and it appears
that in his time there were no silver
mines in the land of the Chalybes,
though he assumes that there were in
the time of Homer (e/c S? rrjs yT]s ra
/j-iTaWa, vvv /xej/ CTLdripov, TrpSrepou Se
Kol apyvpov. xii. 3, § 19). See Note A,
p. 336.

> xii. 3, §§ 7-18, pp. 543-5i8. In
this instance we have the advantage of
comparing the details furnished by
Strabo, with tlie equally minute par-
ticulars in the Periplus of Arriaa.

Sect. 1.

STRABO : ASIA.

297

cerned, the knowledge of our author was complete and definite,
as might naturally have been expected with regard to his
native country. His description of the fertile valleys and
plains of Western Pontus is highly characteristic, and almost
all the localities which he describes have been readily iden-
tified by modern travellers. The picture which he gives us
of his native city Amasia, and its very peculiar and striking
position, was found by Mr. Hamilton to be at once clear and
satisfactory, though it had been imperfectly understood by
persons who had not visited the locality.^ But the personal
knowledge of Strabo evidently extended very little, if at all,
beyond the Halys, and with the interior of Paphlagonia and
Bithynia, as well as the great provinces of Galatia, Phrygia,
and Mysia, his acquaintance was apparently very imperfect.
The brief and perfunctory manner in which he describes these
interior regions of Asia Minor affords a strong contrast with
the fullness and clearness of his account of Pontus, as well as
with the copious details which he furnishes concerning the
provinces on the western coast.

He tells us indeed expressly ^ that it was difficult to define
the limits of the different nations that occupied the interior of
Asia, and even those of Bithynia and Mysia, on account of the
frequent changes and fluctuations, ethnographical as well as
political, to which they had been subject. In the former
point of view he distinctly inclines to regard the Mysians,
Bithynians and Phrygians as cognate races, probably all alike
of Thracian origin.* The Galatians were of course well known
as a historical fact to be Gauls, and the three tribes into which
they were divided still retained the purely Gaulish names of
Trocmi, Tolistobogii, and Tectosages.^ South of Galatia, on
the confines of Cappadocia and Phrygia, he places the great
salt lake of Tatta, which constitutes in fact one of the leading
physical features of the interior.^ Immediately to the south

^ Hamilton's Researches in Asia
Minor, vol. i. pp. 366-370.

3 xii. i, § 4, p. 564 ; 8, § 2, p. 571.
' Ibid. p. 564. The same view was

taken by Herodotus (vii. 73, 74).

5 See Chapter XX. Note B, p. 200.

" xii. 5, § 4, p. 568. His description
of it has been fully confirmed by recent

298

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

of this again lay the cold and upland tracts of Lycaonia and
Isauria, bare of wood and deficient in water, but furnishing
pasturage to innumerable flocks of sheep : a description exactly
corresponding to their present condition. Iconium was the
only city of importance in Lycaonia. Isauria immediately
adjoined the foot of the Taurus, and within the rugged ranges
of that mountain chain were the Pisidians, whose lofty and
inaccessible strongholds had afforded them shelter during the
piratic wars, so that they were with great difficulty reduced by
P. Servilius, who derived from his exploits the surname of
Isauricus.^ The Pisidians again adjoined on the south the
fertile maritime district of Pamphylia, with its flourishing
cities of Side and Aspendus. Notwithstanding the rugged and
difficult nature of their country the Pisidians seem to have
been well known to the neighbouring Greeks, and Strabo cites
from Artemidorus the names of thirteen of their cities, the
most important of which were Sagalassus and Selge. Of the
latter of these, and its extraordinary position, he has given a
minute account (probably derived from the same authority),
which has been confirmed by the researches of recent tra-
vellers.^

§ 15. While he describes the great inland province of Phry-
gia, as already mentioned, very briefly and imperfectly, he was
well acquainted with that portion of it which adjoined the
frontiers of Caria, through which led the great high-road from
Ephesus to Apamea. The latter city, the position of which he
describes very fully and with remarkable accuracy, was in his
day become one of the principal centres of trade in all Asia,
being in this respect second only to Ephesus itself.^ It was

travellers, though not mimixed with
exaggeration. It is now called by the
Turks Tuzla, or the Salt Pan, from the
extent to which it is saturated with
salt.

' In the time of Xenophon as we
have seen, the Pisidians, though nomi-
nally subject to the Persian Empire,
were practically a race of independent
freebooters. (Sec Chapter X. p. '3i5.)

8 Strabo, xii. 7, p. 570. The site of
Selge, which is still called Serghe, was
first identified by Mr. Daniell in 1843.
(See Spratt and Forbes's Lycia, vol. ii.
pp. 17-32.)

" xii. 8, § 15, p. 577. It is not im-
probable that Strabo's description of
Apamea may be derived from personal
observation. Wo learn distinctly that
he had himself visited the city of

Sect. 1.

STBABO: ASIA.

299

from thence that the most frequented line of route led through
Antiochia, Philomelium and Mazaca in Cappadocia (Csesarea)
to the Euphrates, and thence into the interior of Asia.^

The whole of the thirteenth, and the greater part of the
fourteenth book of Strabo are devoted to the description of
the western provinces of Asia Minor, from the Propontis to the
frontiers of Lycia ; including the Troad, Ionia, Lydia and
Caria, with the adjacent islands. All these regions were of
course well known to the Greeks, and Strabo could no more
attempt to add to the previously existing information than in
regard to Greece itself. But his account of them, considered
as a geographical description of a well-known country, stands
on a very different footing from that of European Greece. He
had here the advantage of extensive personal acquaintance,
having been sent, as we have seen, when quite a young man,
to study at Nysa in Caria, and having visited Ephesus and
other cities of Ionia and Caria,^ besides having necessarily
seen, on his passage thither, a great part of the coasts and
islands of this side of the ^gean. But besides this he had
evidently for this part of his work the use of much better
materials and authorities than any of which he availed himself
in his description of Greece.

This is particularly the case with regard to the Troad, under
which name he comprises the whole of the north-western angle
of Asia, from the Propontis to the Gulf of Adramyttium. Here
he had the advantage of following Demetrius of Scepsis, who,
as we have seen, had devoted a special treatise to the dis-
cussion of the Homeric Catalogue of the Trojan allies, in which
he had naturally examined with minute care the localities and
names in the neighbourhood of Troy itself, and his investiga-

Hierapolis in the valley of the MBeander
ou the confines of Lydia and Phrygia
(xiii. 4, § 14), and only 60 miles distant
from Apamea. This circumstance suf-
ficiently explains the accurate know-
ledge he shows of Laodicea (which he
reckons the second city in importance
in Phrygia), Hierapolis, and the smaller
towns in the same neighbourhood,

Eumenia, &c. (xii. 8, §§
13, 16).

1 xiv. 2, § 29, p. 663.

^ The exlont of his travels in this
part of Asia cannot be determined;
but he appears during his residence at
Nysa to have visited several of the
neighbouring cities, including Mylasa
in Caria, and Hierapolis in Phrygia.

300 HISTOKY OF ANCIENT GEOGEAPHY. Chap. XXII.

tion had been materially aided by the situation of his birth-
place of Scepsis in the very centre of the region in question.
The consequence is that Strabo, who devotes, as he himself
acknowledges, a somewhat disproportionate space to the ex-
amination of this small portion of Asia Minor, has in this
instance presented us with a chorographical description of the
country, superior to any other that Ave find in his whole work :
while the incidental discussions and controversies in regard
to the Homeric names of localities and nations, though neces-
sarily arising in connection with this subject, are far from occu-
pying the disproportionate amount of attention which they do
in the case of European Greece. The most interesting of these
controversies at the present day is undoubtedly that relating to
the true position of Troy itself, or the Homeric Hium, a ques-
tion first raised by Demetrius of Scepsis, but for our knowledge
of which we are wholly indebted to Strabo, who adopted in
their full extent the views of his much valued authority.^

We are not clearly informed what authors he followed in
respect to the neighbouring countries of Ionia, Lydia and
Caria, or how much may have been derived from his own
personal observation : but there is nothing to exclude the sup-
position that he had himself visited the principal cities of this
part of Asia, and we know that he had extended his travels for
some distance into the interior. His notices of Sardis and the
tombs of the Lydian kings on the lake Coloe, and still more
his account of the curious volcanic district called Katakekau-
mene — the Burnt Land — have every appearance of being the
result of actual inspection.* He justly points out the con-

^ strabo, xiii. 1, pp. 595, 597. How I -which he here refers is of coiTise the
little attention these sceptical views ' city so colled in Lis day, and which
attracted in ancient times is sufficiently enjoyed immunity from tribute, as the
shown by tlie fact that they are not j reputed parent of Eome.
even thought worthy of mention by i * xiii. 4, § 11. For a full description
Pliny, who dismisses the far-famed \ of this interesting geological district,
city with tlie brief and passing notice : j see Hamilton's Travels, vol. ii. pp. 128-
" Est tamen et nunc Scamandria civitas | 138 ; and TchihatcheiF, Asie Mineurc.
parva, ac md passus remotimi a portu j Strabo describes three distinct era-
Ilium immune, undo omuis rerum cla- i ters, about 40 stadia distant I'rom one
ritas " (v. o3, § 124). The Ilium to j another, and surmounted by rugged

Sect. 1.

STRABO : ASIA.

301

nection between these extinct volcanic phenomena, and the
earthquakes to which all this part of Asia was eminently sub-
ject, especially Philadelphia, the city nearest to the Burnt
Country, where earthquakes were so frequent that Strabo
expresses his wonder how the inhabitants could be induced to
live there.^ The great earthquake which a few years before
(a.d. 17) had destroyed, or seriously damaged, twelve of the
chief cities in this part of Asia, especially Sardis and Magnesia
ad Sipylum,^ was, as he tells us, only one among many similar
calamities, from which they had repeatedly suffered. Full
information concerning other places in the interior must have
been readily obtainable from the Greeks in the cities nearer
the coast : and we cannot doubt that it was Strabo's early per-
sonal acquaintance with these regions that led to his collecting
the materials concerning them, which he has put together in
so clear and satisfactory a manner.

§ 16. Proceeding along the southern coast of Asia Minor, he
describes in succession Lycia, Pamphylia and Cilicia, with the
neighbouring island of Cyprus. His account of Lycia, though
brief, is very clear and distinct in a geographical point of view,
and we are indebted to him for giving us on this occasion an
account of the constitution of the Lycian League, which has
been regarded by some political writers in modern times as the
model of a well-constituted federation.'' For this, as well as
for the geographical description of the country he was appa-
rently indebted to Artemidorus : ^ and the same writer was
probably one of his chief authorities in respect to Pamphylia
and Cilicia also. But as Strabo had attended the lectures of
Xenarchus, a Peripatetic philosopher, who was a native of

hills, which he reasonably infers to
have been formed of the heated matter
ejectir-d from them. He notices also
that this volcanic district, like that of
Catania in Sicily, was specially favour-
able to the growth of vines.

= xii. 6, §' 18 ; xiii. 4, § 10.

" Ibid. xii. 8, § 18 ; xiii. 3, § 5 ; 4,
§ 8, Tacit. Annul, ii. 47.

' See the remarks of Mr. Freeman,
in his History of Federal Government
(vol. i. pp. 208-216), who cites also the
observation of Montesquieu {Esprit des
Loix, liv. ix. c. 3), that if he were called
upon to choose a model of a federal
republic, he would take that of Lycia.

8 See xiv. 3, § 3, p. 665.

302

HISTOEY OP ANCIENT GEOGRAPHY. Chap. XXII.

Seleucia on the Calycadnus, he may well have derived some
particulars from that master.^ His description of Tarsus also
is of a character to lead very strongly to the inference that he
had visited that city in person. But be this as it may, it is
certain that he possessed very good information concerning
the whole of this line of coast, and that his enumeration of
the cities and towns, as well as of the rivers and headlands that
formed its marked natural features, is found to be at once
copious and trustworthy. Of the interior there was of course
little to tell, the lofty and rugged ranges of Mount Taurus
impending over the sea at so short a distance that there had
never been any Greek settlements or civilized towns at any
distance from the sea-board; except in the extreme east of
Cilicia, where the mountains receded from the shore and the
broad alluvial plain formed by the deposits of the rivers Sarus
and Pyramus extended from the foot of the Taurus to that of
the Syrian Amanus.^

§ 17. While Strabo's general description of Asia Minor is on
the whole thus full and satisfactory, it is remarkable how little
pains he has taken to furnish us with positive data as to dis-
tances and positions, such as would enable a geographer to
construct a map of the country. He has indeed given such
distances by sea along the coast of the Euxine from Trapezus
to the entrance of the Bosphorus, as well as for the west coast,
adjoining the ^gean, but in regard to the latter he himself
points out that the extremely irregular configuration of the
coast, and the number of the projecting headlands and penin-
sulas, rendered the periplus or coasting voyage from one point
to another disproportionately long as compared to the direct

^ Groskurd assumes that Strabo
studied under Xenarchus at Seleucia,
but there is no proof of this, and as he
himself tells us that Xenarchus lived
but little at home (eV oficM fj.eu ov iroAv
St4rpi\f/ev), but spent the greater part of
his life at Alexandria and Athens, and
finally at Rome, as a teaclier (xiv. 5,
§ 4), it is much more probable that

Strabo followed his lectures in the
latter city.

' The alluvial character of this plain,
and its rapid extension by the accumu-
lations of the rivers, could not fail to
attract attention ; and an oracle was
said to have foretold that the deposits
of tlie Pyramus would one day reajh to
Cyprus (Strabo, xii. 2, p. 53G).

Sect. 1. STRABO : ASIA. 303

distance.^ But notwithstanding this he considers the whole
line of coast from the south-west extremity of Caria to the
Propontis as preserving a general direction from south to north
" as it were on a meridian line," ^ and measuring in direct dis-
tance about 5000 stadia, or but little less. This same line he
considered as prolonged (according to the erroneous conception
to which we have already more than once adverted) in the
same direction to Cyzicus and Byzantium. As might be
expected under the circumstances his estimate of 5000 stadia
is greatly exaggerated, the distance from Ehodes to the Hel-
lespont by the nearest course which was possible for a navi-
gator to pursue being little more than 300 G. miles (3000
stadia), while that measured along a meridian line would not
exceed 4 degrees of latitude or 2400 stadia. But such a line
instead of falling, as he supposed, at the entrance of the
Hellespont, would in reality strike the Propontis east of
Cyzicus.

He also, in common with all his predecessors, exaggerated
the degree to which the promontory of Carambis projects
into the Euxine towards the north ; while on the south coast
on the contrary he does not seem to have been aware of the
extent. to which Lycia projects towards the south beyond the
southernmost point of Caria. ^ But his incidental notice that
the Chelidonian Islands were opposite to {i.e. on the same
meridian with) Canopus in Egypt, is remarkably accurate,^
and supplies an important point in constructing the map of
this part of the Mediterranean.

§ 18. For the interior of the country he was possessed of

^ xiv. 1, § 2, p. 632. I have had any clear idea of the configu-

^ Kol \onrhv iw eufieias o ttXovs fiixp^ ^ ration of this part of Asia Minor, which

rris UpoirovTiSos, ojs &j' /jLea-qix^piv^v riva \ is indeed so complicated that we cannot

■jroLZv ypa1j.1j.7jv offov irevTaKiffxi-^'-o'v (Tra- ; wonder at any one unprovided with a

Siaiv 71 fjiKphv aiToxdiTovcrav, xiv. 2, § 14. i good map, on which the bearings were

It would be difficult to find a coast to correctly laid down, failing to under-

which such a characteristic was less j stand it.

applicable than to the west coast of i ^ xiv. 3, § 8, p. 666. ^oKovffi 5e at

Asia Minor. I XeXiSoviai Kara Kdvw^6v ttchx -K'nneLV.

"* xiv. 5, § 22, p. 677. He is here j The actual difference of longitude

arguing against Apollodorus, but I does not exceed 20'.

neither one nor the other appears to '

304 HISTOEY OF ANCIENT GEOGEAPHT. Chap. XXII.

very inferior materials : here he has given us, besides some
distances by land in Ionia and Caria, only one main line of
route, which he has taken from Artemidorus ; ^ and even in
this instance he has not given us the distances in detail. The
route in question, which was that habitually followed in
Strabo's time by all travellers proceeding from Ephesus towards
the East, led from that city through Magnesia, Tralles, Nysa,
and Antiochia to a place called Carura, on the confines of Caria
and Phrygia, a distance of 740 stadia : thence through Phrygia,
passing through Laodicea, Apamea and Metropolis to a place
called Holmi, on the frontier of the district known as the
Paroreius, 920 stadia : then across the Paroreius to TyriaBum
on the confines of Lycaonia, a little more than 500 stadia :
across Lycaonia, passing through Laodicea (called for dis-
tinction's sake Katakekaumene) to Coropassus 840, and from
Coropassus to Garsaura, a small town on the confines of Cap-
padocia, 120 : thence to Mazaca, the capital of Cappadocia,
680 stadia ; and from Mazaca to the Euphrates at Tomisa,
1440. From thence a road led across the chain of Taurus to
Samosata on the Euphrates, a distance of 450 stadia.^

It is singular that he has not furnished us with a single line
of route, or detail of distances across Asia Minor from the
Cilician or Syrian Sea to the Euxine, though he repeatedly
discusses the question of the so-called isthmus which united
the peninsula of Asia Minor to the continent of Asia. We
have seen that from the time of Herodotus downwards a very
erroneous notion had prevailed of the breadth of this isthmus,
or the interval from sea to sea, from the Gulf of Issus to that
of Amisus. That historian had described it as five days'
journey for an active man, and even Artemidorus (according to
Strabo) had estimated it at only 1500 stadia. Our geographer,
on the contrary, following the statement of Eratosthenes, con-
siders it as not less than 3000 stadia, which is actually in
excess of the truth, if measured in a direct line, as was certain] v

See Chapter XVIII. p. 67. ' Strabo, xiv. 2, p. G(i3.

Sect. 1. STRABO : ASIA. 305

intended in this instance. But he correctly judges that the
line should be drawn across either from the mouth of the
Cydnus below Tarsus, or from the Gulf of Issus to Amisus, and
not to Sinope, as had been done by several preceding geo-
graphers.^ Supposing the distance thus measured from sea to
sea, nearly along a meridian line, the interval is really about
4° 20' of latitude, or 260 G. miles (2600 stadia), so that the
estimate of Eratosthenes and Strabo is not very wide of the
truth. But the admission that the supposed isthmus was really
so broad as this in great measure destroyed the idea of the
peninsular character of Asia Minor, which had come to be a
received article of faith among ancient geographers.^

§ 19. Strabo's account of India, which occupies the greater
part of his fifteenth book, is in some respects one of the most
interesting parts of his work, and must have been still more so to
his contemporaries, from the numerous particulars that he has
brought together with regard to the natural productions and
physical peculiarities of the country, as well as the singular
political institutions and customs of its inhabitants. These
are taken almost entirely either from Megasthenes, or from the
still earlier writers, Nearchus, Onesicritus, and Aristobulus,
who had accompanied Alexander on his expedition down the
Indus, and had collected much hearsay information concerning
other parts of India which they had not themselves visited.
Later sources of knowledge he appears to have had absolutely
none. Though a considerable trade was carried on in his day by
way of the Eed Sea with India, and some of the traders were even
said to extend their voyages as far as the Ganges, they were for
the most part (he tells us) ignorant men, from whom no informa-
tion could be obtained concerning the countries they visited.^

* Amisus was in fact situated more
than 40 Gr. miles, or 400 stadia farther
to the south than Sinope; and the
distance therefore by so much the less.

' The broadest part of Asia Minor,
from Cape Anemurium on the S. to Cape
Carambis on the N. is only about 6°, or
360 G. miles, across from sea to gea.

» Strabo, xv. 1, § 4, p. 686. His
statement, that hut few of them made the
voyage round India to the mouth of
the Ganges {ffirdvioi ^ej/ kol TnpmeirXev-
Kaffi /j-expL rod Tayyov), must certainly
be meant to imply that some of them
dhl or were said to have done so ; but
it may well be doubted whether Strabo
VOL. 11. X

3o6

HISTOKY OF ANCIENT GEOGEAPHY.

Chap. XXII.

He is indeed careful to impress upon his readers the vague
and uncertain character of the materials which he had at his
command, and upon which he was forced to rely. India had
from a very early period taken a strong hold upon the imagina-
tion of the Greeks, and had thus become the subject (as Strabo
points out) of almost endless exaggerations and fables. For
this reason he dismisses at once with contempt all the mar-
vellous tales of Ctesias and other early writers, and justly
regards the expedition of Alexander as having for the first time
opened out trustworthy information concerning this far-famed,
but little-known, region. But even the writers of this period he
found far from agreeing among themselves, sometimes varying
even with regard to facts which had come within their own
observation, and still more concerning such as they could only
have learnt by hearsay.^ The care which Strabo takes to
excuse himself on account of discrepancies and probable errors
arising from these causes is sufficient proof that he had no
means of correcting them from any later authorities. But, as
we have seen in discussing the information collected by
Megasthenes and his contemporaries,^ their statements con-
cerning the natural productions of India, which must have
come under their own personal observation, are generally accu-
rate and trustworthy, while those relating to the manners and
customs of the inhabitants and the peculiar social polity, which
was calculated in an especial degree to arrest the attention
of an intelligent G-reek traveller, if not in all respects correct,
contained much that was really valuable and interesting.

§ 20, Unfortunately the case was far otherwise in regard
to the purely geographical knowledge of the country. In
this respect Strabo does not pretend to have made any advance

had met with any one who had really
made the voyage. His report of the
embassy of the Indian king Porus to
Augustus, already noticed (see Chapter
XX. p. 166), is derived from Nicolas
of Damascus (xv. I, § 73).

2 Id. XV. 1, §§ 2, 10. Even those wlio

had themselves visited the country, as
he truly observes, had only seen a
small part of it, along certain lines of
march or route, and must describe all
the rest at second hand.
3 See Chapter XIV. sect. 1.

Sect. 1.

STRABO : ASIA.

307

upon his predecessors, and he adopts without modification the
conclusions of Eratosthenes upon these points, while he admits
the untrustworthy character of his materials, and his conse-
quent liability to error.* In one instance only had Eratos-
thenes possessed more definite and trustworthy information,
which was doubtless derived from Megasthenes. This was with
regard to the so-called " royal road " to Palibothra, which the
Greek envoy had undoubtedly travelled, and the distances along
which were measured.^ Relying upon this, he had reckoned
the distance from the Indus to Palibothra at 10,000 stadia
(1000 G. miles), to which he added 6000 more for the distance
from thence to the mouth of the Ganges, and thus obtained
16,000 in all for the total length of India. Patrocles, as Strabo
tells us, diminished this estimate to 15,000 ; but he gives us
no account of the grounds of this correction, and does not take
upon him to decide between them.^ He adopts also the view
of Eratosthenes with regard to the orientation of India, and its
greatest length being from west to east, in opposition to the
more correct conclusions of Megasthenes. Hence he considers
the promontory of the Coniaci (Cape Comorin) to project to
the south-east, so that its extreme point was 3000 stadia farther
east than the mouth of the Ganges. His conception of the
map of India did not therefore differ in any material particular
from that of Eratosthenes.

* Strabo, xv. 1, §§ 10, 11, p. 688.

^ Ibid. § 11. rovTov Se rh fj.ev f^^XP^
HaXi^SOpojv exoi ris iiv fiefiaiorepais
eiTreiy ; KaTa^e/xeVpTjTat yap axoiviois, /cat
tffTiv o5bs ^acriXiKT] ffTaSiccv ixvpiciiv.

This must undoubtedly be the same
route, the measurements along which
are given by Pliny, but in so confused
and corrupt a manner as to be of no
real value (see Chapter XIV. p. 557 ).
The sum total of his distances would
give 1611 Eoman miles, or 12,888 stadia
from the Hypliasis to Palibothra, while
Eratosthenes reckoned only 10,000
stadia /rom the Indus to the same city,
and even this is considerably beyond
the truth.

^ As no Greek had been heynnr]

Palibothra, it is clear that the estimate
of the distance from thence to the sea
must have been founded on mere hear-
say, and from the nature of the country
this must have been of the vaguest
description. But the estimate of 5000
stadia (500 G. miles) adopted by Pa-
trocles is a very fair approximation for
the distance from Palibothra to the mouth
of the Ganges. The distance to the
sea at the mouth of tlie Hoogly is of
course much less, but of this the Greeks
had evidently no notion. It was a
received idea among them, and is dis-
tinctly repeated by Strabo himself (xv.
1, § 13), that the Ganges had but one
mouth !

X 2

308 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXII.

In proceeding to describe the rivers of India, lie justly
remarks that while the principal rivers of any country formed
one of its most important geographical features, this was pre-
eminently the case with India, where the rivers, as in the case
of Egypt and the Nile, were essential to the fertilization of the
country, which was only rendered habitable through their
means/ This was strictly true with regard to the countries
watered by the Indus and its tributaries, and regarding those
on the banks of the Ganges and its affluents Strabo had very
imperfect information. This he himself repeatedly acknow-
ledges, and while he describes in detail the rivers flowing into
the Indus, and the lands that lay between them, he tells us
that the others were rather unknown than known,^ The name
of the Ganges was indeed familiar to all : and though very
exaggerated statements were current as to its size and width,
it was generally agreed that it exceeded the Indus in mag-
nitude, and was in fact the largest river in the known
world.® But Strabo not only does not attempt to enumerate
the numerous great tributaries that flow into it,^ but he does
not specifically notice any of them, merely observing in passing
that Artemidorus calls one of them by the name of CEdanes,^
and that another fell into the Ganges under the walls of
Palibothra.^ He cites also from Artemidorus the correct

' XV. 1, § 26, p. 697.

^ tSiv Ss aWtav iffrlv &jvoia, izXeiuiv t)
•yviiXTis. Ibid.

* oTi fxiv yap jxiyiffTos rwv fxpriy-ovevo-
(xivojv Kara ras Tpe7s Tjireipovs, kol fier'
aiirhv 6 "luSos . . , iKavSis (Tvjx(pu>ve7rai.
XV. 1, § .S5.

This had been already stated by
Megasthenes, and probably adopted
from him by succeeding writers.

' The absence of all attempt at snch
an enumeration is the more remarkable

is found in our text of Strabo (xv. 1,
§ 72) : but as no such name is men-
tioned by any other writer, tlie latest
editors have proposed to read Olfjidvris,
or 'lo/xdyris. It is certain that the
omission of all mention in Strabo of
the lomanes, or Jumna, the most im-
portant of all the tributaries of the
Ganges, is very singular, bnt even if
its name were here introduced, its mere
passing mention would show that
Strabo was wholly unaware of its real

as Megasthenes had given a list of no I importance,

less than nineteen affluents or tribu- I ^ xv. 1, § 36, p. 702. The name_ of

taries of the Ganges (Arrian, Indica, this river has dropped out of our exist-

c. 4). Apparently Strabo hnd no j ing text of Strabo, but it is probable

means of selecting the most important, | tlmt the author wrote Erannoboas,

and did not choose tn burden his text which we find in the parallel passage

with such a number of unknown names. of Arrian (Indica, c. 10, § 5).
^ OiSdv-ns. It is thus that the name

Sect. 1. STRABO : A8IA. 309

statement that the Granges had its source in the Emodi Moun-
tains (one of the many names by which the Himalaya was
known to the Greeks), and flowed at first to the south, after-
wards taking a turn to the east, which course he supposed
it to pursue to Palibothra, and from thence to the Eastern
Sea.*

§ 21. Of the great mountain chain that formed the northern
boundary of India, and which, in accordance with the system
of Eratosthenes, he regarded as a prolongation of the Taurus,
and extending from west to east, he had no detailed knowledge,
and merely tells us that its different portions were known by
the native appellations of Paropamisus, Emodus, Imaus, and
other names, without attempting to define or localise them
further.^ But it appears from another passage ^ that he applied
the name of Imaus to the extreme eastern portion of the range,
which ended, according to his ideas, in the Eastern Ocean;
while that of Paropamisus we know to have belonged to the
mountain ranges north of Afghanistan, now called the Hindoo
Koosh. It remains therefore to apply the name of Emodus or
Emodi to the great central chain of the Himalayas, in which the
Ganges as well as the Jumna and Sutledge takes its rise :
and this appears to be the sense in which Strabo understood
the term, though differing materially from its use by later
geographers.^

Of the great peninsula of India, to the south of a line drawn
from the mouths of the Indus to those of the Ganges, he gives us
no particulars at all. Altogether it may safely be asserted that
while Strabo in his account of India has shown much judgement
in the collection of his materials from preceding writers, and a

XV. 1, § 72. OaXcLTTTi ^vva-n-rov.

T7]v 'IvSiKTjv TTfpicipiKev cLirh /iev Twv ' '' In accordance with this, as we have

seen, Artemidorus described the Ganj^es
as rising in the Emodian mountains (e/c
Tcov 'lifi(eSa>v opS>v, I. c), and Strabo
speaks of tlie forests between the Hy-
daspes and Acesines as at the foot of
the Emodian mountains (17 irphs to7s

dpKTcov Tov Tavpov to, eff^ara a,Trh rrjs
' hpiavrjs (tie'xpi T'^j hcxias QaXa/n-qs., airep
ol lirLX<^pi-oi Kara fxepos Uapoira/xLaop re
Koi ' HixaiShv koI "ifiaov Kol aKXa bvofJLd^ov(Ti.
MaKeSoves Sh KavKaaov. XV. 1, § 11, p. 689.
'' xi. 11, § 7, p. 519. ToO Tavpov rh

reKevrawy o KaXovffiu "l/xaiov, Ty 'IvSiKii 'H/xodSo7s opeatv vKri, xv. 1, § 29, p. 698).

310

HISTOEY OP ANCIENT GEOGEAPHY.

Chap. XXII,

sound spirit of criticism in rejecting many fables and ex-
aggerations, there is hardly any part of his work which shows
less progress in real geographical knowledge beyond that
already possessed by Eratosthenes and his other predecessors.

In regard to the island of Taprobane also, which in common
with Eratosthenes he regarded as situated at the southern
limit of the known world, he had nothing to add to what he
derived from the Alexandrian geographer, and adopted his
erroneous ideas of its position and extent. Of the great and
wealthy islands farther east, or of the vast extension of portions
of the Asiatic continent beyond the mouths of the Ganges,
not the faintest rumour had reached his ears. He had found
indeed in his authorities the name of the Seres, of whose
longevity marvellous tales were related,^ but evidently sup-
posed them to be merely an Indian tribe.

§ 22. Very much the same remark as applies to Strabo's
description of India may be made also with regard to the next
great division of Asia — the countries which he comprises under
the general name of Ariana. Under this head he includes all
the provinces extending from the frontiers of India westward
to those of Persia, and from the Taurus and the Paropamisus
southward to the Persian Gulf and the Erythraean Sea. As
employed in this comprehensive sense, the term comprised the
provinces of Gedrosia, Arachosia, the Paropamisadse, Drangiana
and Carmania, and extended over the greater part of the great
central plateau or table-land of Iran, exclusive however of
Persis or Persia Pro23er, and of Media, of which he had already
treated separately : but including apparently the great salt
desert which occupies the whole central portion of ih\Q plateau,
extending from the frontiers of Seistan (Drangiana) to those of
Yezd and Kerman. Of the vast extent and importance of this
great natural feature of the tract in question ^ Strabo seems to

» XV. 1, § 34, p. 701 ; § 37, p. 702.

" The Khubeer or Great Salt Desert
in the uoitli of Persia is itself in length
about 400 miles, and 250 in breadth

(Kinneir's Persian Empire, p. 19), but
tliis joins on to the deserts of Kerman,
Seistan, and others of scarcely inferior
extent.

Sect. 1.

STEABO: ASIA.

311

have had a very inadequate idea, and only briefly mentions it
as the desert portion of Carmania, extending to Parthia on the
one side and to Paraetacene on the other. But of all these
regions he had no further knowledge than that which had
been derived from the historians of Alexander, and had been
already put into a definite geographical form by Eratosthenes,
to whom he distinctly refers as the best authority, upon whose
information he was not able to make any improvement.^

He describes at considerable length, though with very little
geographical detail, the celebrated march of Alexander through
Gedrosia ; but though this portion of his work is interesting
for comparison with the narrative of Arrian, it contributes
very little to clear up the grave geographical difficulties with
which, as we have seen, the accounts of that march are com-
plicated : ^ while we are left almost entirely in the dark as to
the march of Craterus with one main division of the army
through Arachosia and Drangiana to Carmania — a line of
route which must have contributed much to elucidate the
geography of Ariana.^

Of the other countries included in this section of his work
he has given us only a very brief and summary account : but
we are indebted to him for one important ethnographical
notice— that the name of Ariana was sometimes employed in a
wider sense, as comprising a part of Persia and Media, as well
as Bactria and Sogdiana to the north, for that these nations also
spoJce nearly the same language ;* a statement which, as Prof.
Wilson observes, there is every reason to believe correct.^ It
is remarkable that in regard to all these countries he appears
to have derived his information almost exclusively from Era-
tosthenes or still earlier writers : we find no reference to the
existence even of such itineraries as that which is still preserved
to us under the name of Isidore of Charax. The knowledge of

' XV. 2, § 8. irepl wv ''EpaToffdevTjs
oiirais e^p7}Kev ov yap exofi^v ti xiyeiv
^4Xtiov Trepl avraiv.

^ See Chapter XII. NotcXx, p. 519.

* Ibid. Note Y y, p. 521.

* XV. 2, § 8, p. 724. elffl yap iroos Ka\
bjjL6y\()3TTOi TTapa ixtKpSy.

^ Wilson's Ariana, p. 121.

312 HISTOEY OP ANCIENT GEOGRAPHY. Chap. XXII.

all Upper Asia still remained in almost precisely the same
condition which it had attained under the successors of
Alexander.

§ 23. The last section of his fifteenth book is devoted by
Strabo to Persis, or Persia properly so called, with the adjacent
proYince of Susiana. Both of these districts were of special in-
terest to the Greeks as having been so long the seat of the great
Persian Empire that had extended its dominions from the banks
of the Indus to the shores of the Mediterranean : and their
leading geographical features had long been familiar to all.
The characteristic division of Persia into three parallel tracts
of very different physical character and climate, and yielding
in consequence wholly different productions, is well described.*^
The first of these, a band extending along the sea-shore from
the frontier of Carmania to the river Oroatis, was parched with
heat, of a sandy soil and producing little else except dates.
This is the tract now called the Ghermsir, or hot region, and
which fully corresponds with the description of Strabo. Above
this was a fertile district capable of producing all kinds of
crops and especially favourable to the pasturage of sheep :
while above this again to the north was a rugged and cold
mountain region. The character of these separate tracts is in
fact determined by their difference of elevation, the traveller
proceeding towards the interior of Persia rising, as it were, by
successive steps from the low sandy plains adjoining the sea,
to an elevation of more than 5000 feet in the table-land of the
interior. Of this Strabo had, as usual, but an imperfect com-
prehension, from the want of any means of estimating altitudes
above the sea, but the contrast of the different climates was too
marked to escape observation. He notices also^ the occurrence
of numerous straits or narrow passes through ' these successive
ranges of mountains, which had borne an important part in
the operations of Alexander, who had insisted upon forcing
his way through them, instead of contenting himself, as the

•^ XV. 3, § 1. ' XV. 3, § (J, p. 72».

Sect. 1.

STEABO : ASIA.

31.

Persian monarchs had done, with paying a sum of money to
the mountaineers that guarded them.^

§ 24. His account of Susiana — the modern Khuzistan — is
less satisfactory than that of Persia, especially with regard to
the rivers which traversed the province, concerning which he
found conflicting statements in his authorities, and had no
means of reconciling them. There are indeed few problems in
ancient geography more difficult than the determination and
identification of the rivers of Susiana, which take their rise in
the lofty ranges of Mount Zagros, and after traversing the
fertile tracts of the plains, end in the marshy, muddy, alluvial
tract that lines the whole extent of coast from the mouth of
the Oroatis to that of the Euphrates. The difficulty arises not
merely from the different, and apparently conflicting, state-
ments of ancient authors, but from the changes in the country
itself at the mouths of the rivers in question, which have been
undoubtedly considerable, though we have no exact informa-
tion as to their extent and nature. It is indeed only in very
recent times that we have obtained anything like an accurate
knowledge of the geography of Khuzistan ; the site of Susa
itself was long a subject of dispute,^ and cannot be considered
as having been established beyond a doubt till the excavations
carried on in 1852 by Mr, Loftus at Sus or Shush, proved the
identity of that locality with the celebrated city of which it
retained the name, and brought to light the magnificent ruins
of the palace of the Persian kings.^ The determination of the
site of the capital establishes beyond a doubt the identity of
the celebrated river Choaspes with the modern Kherkah,
which flows near the ruins of Susa, while the Pasitigris of
Nearchus and Strabo may be identified with equal certainty
with the river now called Karun or Kuran, which flows under

« See Chapter XII., Note I, p. 475.

^ Susa was indeed correctly identi-
fied with the modern Sus or Shus by
Major Eennell ( Geography of Herodotus,
pj). 2(J3, 33i) ; but Dr. Vincent returned
to the opinion previously entertained

that it occupied the site of the modern
Shuster on the Karun (^Commerce and
Navigation of the Ancients, vol. i. p.
449).

* See Ivof tus's Ghaldsea and Susiana,
8vo, Lond. 1857, chap. 24-31.

314

HISTOEY OF ANCIENT GEOGRAPHY.

Chap. XXII.

the walls of Shuster ; but if these two conclusions be admitted,
it becomes almost impossible to find a place for the EulaBus,
which is described both by Strabo and other writers as one of
the principal rivers of Susiana.^ According to the statement
of an author named Polycleitus, — one of the historians of
Alexander the Great who is repeatedly cited by Strabo in this
part of his work — the Choaspes, Eulseus, and Tigris, all flowed
into the same lake, from which they had their common outflow
to the sea. The existence of such a lake, which has been long
filled up by the continual advance of alluvial deposits, is
attested by several other writers, and appears to admit of no
doubt.^ It seems probable also that it communicated with the
Euphrates, and received a portion of the waters of that river,
though Strabo still regarded the main waters of the Euphrates
as flowing into the sea by an independent channel.*

In the passage of Polycleitus just referred to, it seems almost
certain that the river designated by him as the Eulseus was
the same with the Pasitigris of Nearchus and Strabo, and other
authorities represent the Eulseus as flowing into the Pasitigris,
or vice versa. On the other hand there are not wanting strong
arguments for identifying the Eulseus with the Choaspes,
which flowed by Susa, and which must have discharged its
waters either into the Tigris or the lake at its mouth. It
seems impossible to determine the question without supposing
that the name of Eulseus was applied to one or the other of
the two rivers known also as the Pasitigris and Choaspes ; but

^ The Karun in the upper part of
its course receives a tributary, now
known as the river of Dizful, nearly
equal in volume to its eastern arm,
which is apparently the Coprates of
Strabo (xv. 3, p. 729), and of Diodorus
(xix. IS), which the last author de-
scribes as falling into the Pasitigris.

^ The existence of this lake is dis-
tinctly attested by Nearchus ; but his
statement concerning it is reported
somewhat differently by Strabo and by
Arrian, and the result is far from clear.
He appears, however, to have sailed

from Diridotis at the mouth of the
Euphrates to that of the Pasitigris,
and in so doing to have passed by the
lake which received the waters of the
Tigris. According to this account
therefore it would seem that the Pasi-
tigris did not in his time flow into the
lake (Strabo, xv. 3, § 5, p. 729 ; Arrian,
Indica, c. 42).

•* This he distinctly states on the
authority of Nearchus and Onesicritus,
but (as has been already observed) it
is by no means certain that they still
did so ill his own day.

Sect. 1.

STRABO: ASIA.

315

even if this be admitted, we are still unable to reconcile the
statements of ancient authors without supposing some of
them to have confounded the two streams. It is indeed not
strange that they should have done so, when we consider the
extremely complicated nature of the water systems of these
countries,^ and that none of them, with the exception of
Nearchus (whose statements we only possess at second hand)
wrote from any personal knowledge of the localities.^

§ 25, Proceeding to the westward Strabo next describes the
country which he terms Assyria, a name that he employs in
a much more general sense than it is used by other authors,
including not only the province east of the Tigris, to which
the appellation was commonly confined, but the whole of
Babylonia and Mesopotamia also ; so that Assyria, according
to his use of the term, comprised the whole extent of country
from the chain of Mount Zagros on the east to the Euphrates
on the west. It is still more singular that he should not even
designate the province beyond the Tigris as Assyria properly
so called, but while he gives the name of Aturia to the par-
ticular district in which Ninus or Nineveh was situated, he
includes all the other provinces on the east of the Tigris in
Babylonia, a name usually restricted to the region between
the two rivers. The reason of this deviation from established
usage is unknown to us ; but it was probably connected with
the historical confusion prevalent in his day, which regarded
the Assyrian and Babylonian empires as identical.^ He tells
us indeed — and no doubt correctly — that the Syrians and
Assyrians were in reality the same people, though the name

^ A glance at one of the most recent
maps, since this region has been really
examined and surveyed, will suffice to
show how impossible it must have been
to comprehend its geography, without
the assistance of any map at all.

^ It is remarkable that no mention
occurs in Strabo of Charax, which, ac-
cording to Pliny, was one of the most
important trading towns in this part of
the country. The omission may in part

be explained by the circumstance that
Strabo seems to have made no use of
the work of Isidore of Charax, which
Pliny undoubtedly did ; but it tends
strongly to confirm the conclusion that
Strabo had no other information con-
cerning these countries than what he
derived from Eratosthenes and the
historians of Alexander.
' xvi. 1, p. 737.

3i6

HISTORY OF ANCIENT GEOGEAPHY.

Chap. XXII.

of Syrians had come to be confined in the common usage of
the Greeks to the people occupying the countries between the
Euphrates and the Mediterranean.^ The Cappadocians also,
he adds, were originally the same race, and were still called in
his time Leuco-Syrians or White Syrians ; so that the same
people had at one time extended from Babylonia to the shores
of the Euxine.^

With the provinces which extended from the Euphrates
eastward to Mount Zagros, the Greeks were well acquainted.
They had remained under the Macedonian government after
the death of Alexander for nearly two centuries ; numerous
cities had been founded in them by the Syrian monarchs,
some of which had risen to great opulence and prosperity, and
the whole country was traversed by frequented lines of com-
mercial traffic. Hence Strabo must have had at his command
ample materials for the description of these regions, and ac-
cordingly we find that his geographical account of them is
clear, consistent, and intelligible, though not entering very
much into detail. Of the great cities that had once rendered
this region so celebrated, he tells us briefly that Nineveh had
altogether disappeared, but adds (of course from mere tradi-
tion) that it was much larger than Babylon ; ^ while of Babylon
itself he gives a pretty full account, though he adds that the
greater part of its site was desolate and uninhabited.^ Its
decay was mainly owing (as usual in such cases) to the rise of
the neighbouring city of Seleucia, which had become a great
emporium of trade, and was so populous and flourishing as to
surpass even the metropolis of Syria, Antioch, and was the
largest city in the East, after Alexandria in Egypt.^ The
Parthians had indeed transferred the royal residence to Ctesi-

« xvi. 1, § 1, p. 736.

" Ibid. § 2, p. 737.

> xvi. 1, § 3, p. 737.

: xvi. 1, § 5, p. 738. It is strange
tliat lie describes the walls of the city,
and their vast height and extent, as it'
thuy were still standing in his time (to;/

Se kvkAov 6X6( tov reixovs k.t.\.). It
is scarcely possible that this \yas the
case ; but he probably copied from
Aristobulus or some other of the histo-
rians of Alexander, without any refer-
ence to subsequent changes.
^ Ibid. iSee also xvi. 2, § 5, p. 750.

Sect. 1.

STRABO: ASIA.

317

plion on the opposite bank of the Tigris, but this had not
interfered with the prosperity of the commercial city, which
was still regarded as the capital of all this part of Asia,*

In describing Mesopotamia Strabo is careful to point out
the change that had taken place before his time in the trans-
ference of the customary passage of the Euphrates from
Thapsacus, which, as we have seen, derived so much geo-
graphical importance from this circumstance in the days of
Eratosthenes, to a place much higher up the Euphrates, which
was called in consequence Zeugma or "the Bridge." This
was situated just opposite to the modern town of Bir, which
occupies the site of a G-reek city called Apamea, founded by
Seleucus Nicator,^ and is still the usual place at which
travellers proceeding from Antioch or Aleppo towards Bagdad
cross the Euphrates. The change is one of great importance
in tracing the routes given by ancient writers. There was also
another passage much frequented in his time at Samosata in
Commagene, where the line of route through Asia Minor, that
he has given us from Artemidorus, crossed the Euphrates.^

§ 26. Syria was of course familiar to the Gre-eks from its
having so long been the seat of empire of the Seleucidan
dynasty, under whom it had attained to great opulence and pro-
sperity. Hence we find the description of it in Strabo at once
full and satisfactory. That of the Phoenician coast especially is
so detailed that we might readily have supposed it to be derived
from personal examination, were it not that an expression in
his account of Tyre points to the opposite conclusion.'^ Yet
his ideas concerning the interior, especially of Palestine, were
in some respects strangely inaccurate. Thus, although he was

* xvi. 1, § 16, p. 743.

= Plin. H. N. V. 24, § 87. According
to Pliny, Seleucus was also the founder
of Zeugma, but it does not follow that
the passage at Thapsacus was aban-
doned at so early a period.

« Strabo, xiv. 2, § 29, p. 654 ; xvi. 2, §
3,p.749. Strabohimself appearsto have
been in some confusion between these

two places of passage, which were in
fact 72 Eoman miles apart (Plin. I. c).
The one was the most convenient for
travellers from Antioch, the other for
those coming from Asia Minor.

' ivravOa Se (pacri TToKvari-yovs ras
o'lKlas, ware koI tu>v 4v 'VciifM) fmWov.
xvi. 2, § 23. See note to p. 212.

3i8

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

acquainted both with the Lake of Gennesareth and the Dead
Sea, of which last and its natural peculiarities he gives a full
description (taken apparently from Posidonius), he by a strange
mistake confounds it with the Sirbonian Lake or Marsh, on
the frontiers of Palestine and Egypt. At the same time he
distinctly connects its peculiar character with the other signs
of volcanic action observable in the country, and adds that
" according to the traditions of the natives " it had been
formed by a catastrophe which had overwhelmed thirteen
cities, of which Sodoma was the capital, the greater part of
which had been swallowed up in the lake.^ By another not
less singular error, he supposes the Jordan, which he justly
terms the largest river of this part of Syria, to flow into the
Mediterranean (!) ; and even tells us that it was habitually
navigated upwards from that sea.^ At the same time he
correctly describes the river that flowed by Damascus (the
Chrysorrhoas) as being for the most part absorbed by canals
for irrigation,^ and even notices the tivo peculiar rugged
regions, which gave name to the district of Trachonitis, east
of the Jordan.^ Of the natural productions of Judeea, besides
the asphalt of the Dead Sea, he dwells especially upon the
palm-groves of Jericho, and the balsam grown there, as well
as on the banks of the Lake of Genuesareth.^

In describing Jerusalem he speaks principally of the great
strength of the city as a fortress : a circumstance which had
been brought prominently forward on occasion of its siege and
capture by Pompey. It was this event which had especially
directed the attention of the Greek and Eoman world to the

« xvi. 2, § 44, p. 764.

" xvi. 2, § 16, p. 755. -rhv Se hvKov
Koi rhv 'lopSdj'Tji' avaTrX^ovcrt (poprlots,
'ApdStoL de jxaKiffTa. The mention of
the Lycus, a very trifling stream, while
no notice is found of the much more
considerable river Leontes, that flows
into the sea N. of Tyre, is calculated to
raise a suspicion that Strabo has con-
founded the latter river with the Jordan.

' Ibid. § 16, p. 755.

* Ibid. p. 756. vtripKeivrai 5' avrris
(ttJs Aap.a(TKOXj) 5ub KeySfxevoi TpdxoiVis.
It is only quite of late years that modern
travellers have become well acquainted
with this singular region, and have
recognised the fact that it is really
composed of two distinct mountain
tracts of the same singular character,
the Lei ah find tlie Jebel Hauran.

3 ll.'id. § 41, p. 763.

Sect. 1. STRABO : ASIA. 319

sacred city of the Jews, and Strabo was probably indebted for
the materials of this part of his work to Posidonius, who had
written the history of the campaigns of Pompey. It was
perhaps from the same source that he derived the curious
summary that he has given us of the traditions and rites of
the Jews, the institution of which he ascribes to Moses, an.
Egyptian priest, who came thither out of Egypt, and founded
the temple on a rocky and barren site, which was on that
account neglected by the neighbouring tribes.* As Posidonius
was himself a native of Apamea in northern Syria, it is highly
probable that he was one of Strabo's chief authorities through-
out his description of that country.

The whole of the desert tract extending from the confines of
Coele Syria and Judaea to the Euphrates is assigned by Strabo
to Arabia, and was inhabited only by wandering tribes, whom
he called Scenitse from their dwelling in tents. It is strange
that he has omitted all mention in this place of the one im-
portant exception in the case of Palmyra, which was certainly
at this period a flourishing city and emporium of trade, and to
which attention had lately been directed by the attempt of
M. Antony to plunder it of the wealth which its citizens had
thus accumulated.^

§ 27. Of the great Arabian peninsula he has given a long
account, probably the most complete that had as yet been
brought together. The greater part of it was indeed derived
from sources with which we are already acquainted. Thus he
begins® with a general description of the peninsula and the
nations that inhabited it, according to Eratosthenes, who, as we
have seen,^ was the first to bring together any satisfactory
information concerning this country. He next follows this up
with a long extract from Artemidorus, describing in detail
both shores of the Eed Sea, or Arabian Gulf, as it was termed
by the Greeks: an account which we know to have been

* Ibid. §§ 35-37, pp. 760-762.
^ Appian, B. C. v. 9. See Chapter
XIX. p 134.

« Strabo, xvi. 4, §§ 2-4.
' See Chapter XVI. p. 646.

320

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

derived by Artemidorus from the earlier treatise of his con-
temporary Agatharchides, and which has already been fully
examined.^ It is remarkable that notwithstanding the great
increase in the trade to India, which had taken place in the
days of Strabo, he had obtained no additional information
concerning the coasts of the Indian Ocean, either on the
African or Arabian side. He still regards the Noti Keras or
Southern Horn (Cape Guardafui), as the extreme limit of
knowledge on the one side, and while he describes in general
terms the land of the Sabseans and the Chatramotitse in the
south of Arabia, he gives no details either of distances or
of the natural features of the coast outside of the Straits of
Bab-el-Mandeb. It is evident that the outer coast of Arabia
was still practically unknown to geographers.^

But with regard to the interior of the country Strabo had a
new source of information, unknown to any of his predecessors,
in the recent expedition of vElius Gallus, the details of which
have been already given.^ Unfortunately, as we have seen,
the circumstances of this expedition were such as in great
measure to prevent it from throwing the light that might
have been expected upon the geography of the regions that
were traversed by the Eoman general, and we are almost
wholly unable to trace his line of route, or determine the limit
to which he advanced. It is evident that Strabo was himself
very much in the same position : he had no means of con-
necting the localities of which he learnt the names from the
Romans who had accompanied Gallus with those described by
the earlier Greek geographers, and he makes no attempt to do
so. The manner in which he defines the position of Marsiaba
(the turning-point of the expedition) as being said to be only
two days' journey from " the Land of Spices," is certainly not
calculated to give any trustworthy information. It is clear

« See Chapter XVIIT. sect. 3.
" The absence of all notice of so
remarkable a natural feature as tlie

isolated mountain promontory of Aden
is a strong evidence of tliis.

' See Chapter XX. p. 179, foil.

Sect. 2. STRABO : AFRICA. 32 1

that he at least had no idea of identifying it, as has been done
by so many modern writers, with the celebrated city of Mari-
aba, the capital of the Sabseans, which was well known to him
from Eratosthenes and from Artemidorus.^

With regard to the distances from one point to another of
the peninsula, which necessarily determined its form, Strabo
adds nothing to the information already obtained by Era-
tosthenes, concerning the time employed by caravans from the
distant provinces to Petra and Gerrha,^ which still continued
to be the two great emporiums of the trade of Arabia. In like
manner his account of the eastern coast of Arabia, and the con-
figuration of the Persian Gulf, is derived exclusively from
Eratosthenes, who had himself drawn his materials from the
voyage of Nearchus, and that of Androsthenes of Thasos, which
has been already noticed.* So little progress had been made
in real geographical knowledge during a period of more than
three centuries with respect to a country so close to Alex-
andria ! Both Eratosthenes and Strabo had an exaggerated
idea of the size of the Persian Gulf, which they supposed to be
nearly as large as the Euxine.^

Section 2.— Africa.

§ 1. The seventeenth and last book of Strabo's great work is
devoted to Africa, and fully two-thirds of it are occupied with
the description of Egypt. Here there was of course no room
for the extension of geographical knowledge, that country
having been familiarly known to the Greeks from an early
period, while the Alexandrian writers had doubtless possessed
the amplest materials for a full statistical and topographical
account of it. Moreover Strabo himself, as we have seen, had
not only visited Egypt, and ascended the Nile as far as the

2 svi. 4, § 2, p. 768 ; § 19, p. 778.
» Ibid. p. 778. See Chapter XVI.
p. 647.

' Chapter XII. p. 461.

VOL. II.

5 Strabo, xvi. 3, § 2, p. 766. SiaTe
3!j\of e/c rovTwv eli/ai, Swti fiiKphv oltto-
KiiiTSTaL T(f /iieyeOei rris Kara rov Ev^eivov
Oa\dTT7)s avrri 7) OdAarra.

322

HISTORY OF ANCIEJJT GEOGRAPHY. Chap. XXH-

First Cataract, but he had resided for a considerable time at
Alexandria, and had thus every means of obtaining the best
information. At the same time the physical peculiarities of
the country are so strongly marked, and its geographical cha-
racters at once so extraordinary and so simple, that it was
hardly possible to fail to seize them. He aptly compares the
inhabited part of Egypt above the Delta, which as he justly
remarks was merely the valley of the Nile, to a narrow band
stretched out lengthwise, extending about 4000 stadia in
length, by an average breadth rarely exceeding 300 stadia.^
He describes with considerable minuteness the Delta itself, as
well as the different mouths of the Mle, of which the most
important in his day were the Canopic and the Pelusian, and
next to them the Phatnitic, which was nearly midway between
the other two main arms.^ He gives also a graphic description
of the inundation of the Nile, and the appearance of the low
country under these circumstances. With regard to the cause
of the inundation, which had been a subject of so much dis-
cussion and curiosity among the early Greeks, he tells us that
it was in his day well known to be produced by the heavy
rains that fell in the summer on the mountains of Ujoper
Ethiopia ; a cause which, he observes, had been long suspected
by the earlier philosophers, but had been afterwards ascer-
tained to be true by personal observation,^ especially by the
expeditions sent by Ptolemy Philadelphus into these remote
regions for the capture of elephants. The real difficulty, as he
justly adds, was not to account for these copious rains in that
region, but for their entire absence in the Thebaid and neigh-
bourhood of Syene.^

6 xvii. 1, § 4, p. 789.

' xvii. 1, § ] 8, p. 801. The Phatnitic
mouth is the one now known as that of
Damietta, from the town of that name.
It is still one of the principal mouths
of the river.

* Oi fxeu oZv apxouoL crroxaCfJi''? Tt)
irXfov, 01 S' vffrepov avrdTrrat yevT]9evTes
■pcrOovTo virh ofifipwv depivSiv TrATipoiifxevoi/
rhv tiilKov, etc. xvii. 1, § 5, p. 789.

^ Ibid. p. 790. He here refers to
two works specially devoted to the
Nile, one by Budorus, the other by a
Peripatetic philosopher of the name of
Ariston. Both authors are otherwise
totally unknown. According to Strabo
the one treatise was copied almost
entirely from the other, but he was not
clear which was the plagiarist.

Sect. 2.

STEABO : AFRICA.

323

His description of the voyage up the Nile is especially-
interesting, as being derived principally from his own personal
observations. He saw the ruins of Thebes, which already in
his time had ceased to exist as a city,^ and was merely occu-
pied by a group of villages, with the vast ruins of temples and
other sacred edifices spreading over a space of 80 stadia in
extent. Among these he especially notices the celebrated
vocal statue of Memnon, the sound proceeding from which he
himself attests that he heard, but expresses a very sound scep-
ticism as to how it was produced. He was at this time travel-
ling in company with j^lius Gallus, the Eoman governor, and
the whole party were no doubt duly lionized wherever they
went.^ They ascended the river as far as Syene, saw the
Nilometer there, and the well down which the sun shone ver-
tically at the summer solstice, and then proceeded by land to
a point above the First Cataract, whence they visited the
island of Philse.^ This was the term of their expedition, as it
is still that of most modern travellers. They appear also to
have visited the Lake Moeris, and the celebrated Labyrinth,
which Strabo calls a work equal to the Pyramids. He describes
only from hearsay the important commercial route that had
been opened by the first Ptolemies from Coptos to Berenice
on the Eed Sea, but which had been in his day superseded by
that to Myos Hormus, which had become the principal
emporium of trade with Arabia and India.*

' It had been destroyed in B.C. 86 by
Ptolemy Lathyrus who, aecording to
Pausanias (i. 9, § 3), reduced it so com-
pletely to ruin, as to leave no trace of
its former wealth and prosperity. This
is of course a great exaggeration.
Strabo more correctly says : " The
remains of its former greatness are
still shown, extending for a space of 80
stadia : most of them are buildings of
a religious character. It is now inha-
bited only in scattered ■villages ; one
part in Arabia (i.e. on the right bank
of the Nile), where the city was;
another on the opposite side, where
stood the Memnonium " (xvii. p. 816).

2 They were attended by a profes-

sional i^TjyTiT^s, or interpreter (a sort
of upper laquais de place) who professed
not only to be acquainted with the
monuments, but to be able to explain
the inscriptions and hieroglyphics ; but
he was ridiculed as an impostor by the
governor's suite, whether with or with-
out reason we have no means of judging
(Strabo, xvii. 1, § 29, p. 806). Not
long after the time of Strabo the
monuments of Thebes were visited by
Germanicus, to whom the inscriptions
were interpreted by one of the chief
priests (Tacit. Annal. ii. 60).

s Strabo, xvii. 1, §§ 48-50.

^ Ibid. § 45, p. 815.

Y 2

324

HISTORY OF ANCIENT GEOGIIAPHY.

Chap. XXII.

§ 2. Anotlier point on which his testimony is curious, is
with regard to the canal that traversed the Isthmus of Suez,
and had its outlet at the city of Arsinoe at the head of the
Arabian Gulf.^ This did not, like the one recently con-
structed, proceed directly across the Isthmus, but quitted the
Nile, by which it was supplied with water, at a place called
Phaccusa on the Pelusian branch, traversed the Bitter Lakes
and entered the sea at Arsinoe, but was provided with locks at
its mouth, so as to exclude the sea-water, and hence not only
were its waters perfectly fresh, but the Bitter Lakes were ren-
dered so by their admixture. The canal itself was 100 cubits
(150 feet) in width, and deep enough to admit of the passage
of .ships of the largest burden.^ The object of this great work
had obviously been to conduct the commerce of the Ked Sea
direct to Alexandria, but the difficulties of the navigation of
the upper part of that sea had prevented this route from being
generally adopted, and, as has been just mentioned, the Arabian
and Indian trade in the days of Strabo passed by way of Myos
Hormus to Coptos on the Nile, and thence down the river to
Alexandria.

§ 3. With regard to that trade Strabo has given some
interesting information, which he probably collected at Coptos.
Comparing the commerce of Alexandria in his day with what
it had been under the Ptolemies, he tells us that in former
times not twenty ships in a year ventured to traverse the
Arabian Gulf, so as to show themselves beyond the Straits :
but in his time large fleets made voyages to India and the
extremities of Ethiopia,'' and brought back from thence cargoes
of the most valuable merchandise, which contributed twofold

* Strabo, xvii. 1, § 26, p. 805.

" fidOos S' offov aptcelv fxvpiocpdpcp vrji.
Ibid. This same expression, " a sbip
capable of carrying 10,000 amphoraj,"
is used also by Strabo in speaking of
the mouth of the Tagus (iv. p. 151),
and is evidently intended to designate
a ship of the largest class. (See Thucy-
dides, vii. 25; and Lobeck's note on
Phrynichus, p. 662.)

Tlie course of this canal must have
in great measure coincided with the
Sweet Waters Canal, recently opened
in connection with that of Suez.

' Trpdrepov (xev ye ovK elKoai ■7r\o7a
eddppei rhv ' Kpa^LOV k6Kt70v Siarrepav,
Sxrre e|co tUv crrevaiv virepKinrreiv, vvv Se
Kal ffrdhoi /xeydXoi (TTeWovTai /ue'^pf rrjs
'IvSi/cijs Koi rSiv &Kpo>v rHv klQi.0TTLKU>v.
xvii. 1, § 13, p. 798.

Seot. 2. STEABO : AFRICA. 325

to the revenue by paying import duties on its entrance into
Egypt, and again export duties when sent out from Alexandria.
That city had in fact a monopoly of these costly wares, so that
other countries were compelled to derive them from thence.^
In another passage he states the number of ships sailing from
Myos Hormus to Ii-Ila at not less than a hundred and twenty.^

But so imperfect was the statistical information that he was
able to collect, notwithstanding his intimate association with
JElius Gallus, that when he wishes to give some idea of the
revenues actually derived from these sources of wealth, he
goes back to a speech of Cicero's, in which that orator esti-
mates the annual revenue of Egypt in the time of Ptolemy
Auletes at 12,500 talents : and then adds, if such was the
income under the government of such worthless rulers as the
last of the Ptolemies, what must it have become in the pro-
sperous condition to which it had attained under its Koman
governors ? ^ It is remarkable also that though he gives us
many interesting particulars with regard to the provincial
administration of Egypt, its division into nomes, the military
force maintained there by the Eomans, &c., he has nowhere
given us any hint of the estimated population either of the
country itself, or of the city of Alexandria,^ of which in other
respects he has given a full and minute description.

Prosperous as Egypt was in general, many of the famous
ancient cities had already fallen into decay. Thebes, as we
have seen, lay in ruins, while a modern city, Ptolemais, had
become the capital of the Thebaid, and was the third city of

* Ibid. The commercial position of
Alexandria at this time must have
closely resembled that of Venice in the
middle ages.

9 ii. 5, § 12, p. 118. He here also
uses the expression of whole fleets sail-
ing to India (rHv iK ttjs ' AKe^avSpelas
ifiiropcav ffToXois ■^Srj izKeSvTuiv Sia toD
NeiAou Ka.\ rod 'Apa^iov kSXttov ix^xpi ttjs
'IcSiKTJs).

xvii. 1, p. 798.

us that the population of Alexandria
amounted to 300,000 free souls ; while
he roughly estimates the total popu-
lation of Egypt in his day at not less
than seven millions (i. 31, with Wesse-
ling's note, showing that this is the
true meaning of the passage). Josephus,
about half a century later, gives the
population of Egy^rt at 7,500,000 people,
exclusive of Alexandria (Joseph. B. Jud.
ii. 16, § 4), a statement which he pro-

* This omission is fortunately sup- fesses to derive from official documents,
plierl by Diodorus (xvii. .52), who tells \

326

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

Egyj)t in point of population.^ Memphis retained tlie second
place, and was still a great and jBourishing city, but the royal
palace there was in ruins, and the Serapeum was already half
buried in sand.* Heliopolis was altogether deserted, while
Abydos, the sacred city of Osiris, and at one time one of the
most important cities of Egypt, had sunk into a mere village.^

§ 4. Strabo is the first extant writer who distinctly notices
the Oases, those remarkable features of the geography of the
Libyan desert, of which Herodotus, as we have seen, had but an
indistinct idea. He describes them briefly, but yery correctly,
as inhabited districts, surrounded on all sides by vast deserts,
just as islands are by the sea.^ There were three of them (he
adds) in the immediate neighbourhood of Egypt; the first
(that now called the Great Oasis) opposite to Abydos, from
which it was distant seven days' journey through the desert ;
the second (the Lesser Oasis), opposite to the Lake Moeris;
the third that adjoining the Temple of Ammon, so celebrated
for its oracle, which had however fallen into neglect in the
days of Strabo.'' The position of this last he fixes at five days'
journey south of Parsetonium on the Libyan coast.^

He closes this account of Egypt — on the whole one of the
most complete and satisfactory portions of his work — with a
brief notice of the campaign of the Eoman general Petronius
against the Ethiopians, which has been already discussed.^ His
account of that people in general is derived partly from Era-
tosthenes, partly from Artemidorus, who, as we have seen, was in
this part of his work a mere copyer of Agatharchides, and de-
scribed the different wild tribes in the interior, in connexion
with the ports of the Eed Sea, from which the explorers sent out
by the Ptolemies had visited them.^ But of Meroe itself and

3 xvii. 1, § 42, p. 813.

■' Ibid. § 31, 32, p. 807.

s Ibid. § 27, p. 805, § 42, p. 813.

« xvii. 1, § 5, p. 791.

' Ibid. §§ 42, 43, p. 813.

8 Ibid. § 14, p. 799.

" Sec Chapter XX. p. 182.

' This part of the desciiption of

Ethiopia is givfn by Strabo in his six-
tceuth book, where he describes both
shores of the Red Sea, according to
Artemidorus (xvi. 4, §§ 5-18). It has
been already pointed out that this
agrees almost entirely with tliat given
by Agatharchides (Chapter XYIH.
p, 62).

Sect. 2.

STRABO: AFEICA.

327

the comparatively civilized people of which it was the capital
Strabo gives a pretty full account, derived probably from
information collected by the Eomans during the expedition to
which we have just referred.^ With regard to the Upper Nile
and its tributaries he had no information beyond that collected
by Eratosthenes, and contents himself with copying, or at least
giving the substance of, that given by the earlier geographer.^
But it is singular that he notices the existence of a large lake
above Meroe, named Psebo, containing an island which had a
considerable population,* a statement that can hardly refer to
any other than the Lake Tzana or Dembea in the heart of
Abyssinia, which is the source of the Blue Nile — yet he does
not appear to have any idea of its connexion with the Nile.
In the passage elsewhere extracted from Eratosthenes indeed
he refers to the notion, somewhat vaguely reported, that the
main and direct stream of the Nile flowed from certain lakes to
the south ; but it seems probable that this really related only
to the expanse of marshy waters formed by the White Nile in
its course above its junction with the Sobat.^

§ 5. With regard to the rest of Africa, Strabo had sur-
prisingly little to add to the knowledge already possessed by
Eratosthenes. His conception of the form of the continent did
not differ materially from that of the Alexandrian geographer.
He describes it as in a general way resembling a right-angled
triangle, having for its base the sea-coast extending from
Egypt to the Pillars of Hercules; the shorter side perpen-
dicular to this being formed by the Nile up to Ethiopia, and
by a line artificially produced from thence to the southern

2 xvii. 2, §§ 2, 3.

^ xvii. 1, § 2. This passage has
been already examined in the chapter
on Eratosthenes, Chapter XVI. p. 650.

'' Ibid. § 3. inrepKeirai 5e rf/s M.ep67}s
7) "Vefidi, \i/j.vr] fx.eyd\7] vr\(Tov ^x'^ucra
olKov/xtuTiy Ikuvws. This is tlie iirst
mention of a lake of the name. But as
we have seen, Agatharehides described
the torrents flowing into the Bed Sea,
as rising in the Psebaian mountains

(§ 84), a name by which he evidently
meant to designate the mountains of
Abyssinia, in which the Lake Tzana is
situated.

^ More definite information concern-
ing these, as we shall see, was first
acquired in the reign of Nero ; but it
is not improbable that an obscure
notion of them had already reached
Alexandria in the time of Eratosthenes.

328

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXII.

Ocean; while the hypothenuse was constituted by the shore
of the Ocean, extending the whole way from the land of the
Ethiopians to the extremity of Mauretania.^ We see here that
Strabo assumed, as had been the case with almost all geo-
graphers since the time of Eratosthenes,'^ that the southern
shores of Libya were surrounded by a circumfluent ocean ; but
having no real information upon the subject, and no concep-
tion of the vast extension of the African continent towards the
south, he naturally drew the line at no great distance beyond
the limit of the known regions, so as to connect the farthest
points actually known to him, and thus reduced the continent
of Africa to less than a third of its real dimensions. He
himself tells us indeed that the southern extremity of the
triangle was wholly unknown, being unapproachable, or at
least having never been visited, on account of the burning
heat, and that it was merely by conjecture that he placed the
limit of the inhabited world in this direction about 3000 stadia
to the south of Meroe,^ while he assumed without any kind of
proof that the line of the southern coast was not more than
1000 stadia further south. He thus arrived at the conclusion
that the greatest breadth of the Libyan continent was about
13,000 or 14,000 stadia; and its length, from Alexandria to
the Straits, somewhat less than double.^

§ 6. He begins the description of Libya, in the sense in
which he conceives the term («, e. as excluding Egypt) Avith
its western extremity, or Mauretania ; a land which he justly
describes as rich and fertile, and containing many valuable
natural productions — among others the beautiful wood which
supplied the celebrated tables so much sought after by the
Romans in his time.^ Notwithstanding this, the inhabitants

« xvii. 3, § 1.

' The two exceptions, as we have
seen, were Hipparchus and Polybius.
See Chapter XVII.

* It is strange, at all events, that he
should have drawn the conjectural
line so immediately beyond the limits
of the known regions. He had himself

placed the Sembritpc on the Upper
Nile 8000 stadia south of Meroe, yet
he here assumes tlie southern limit of
Africa to be only 3000 or 4000 stadia
beyond Meroe.

» Ibid.

' xvii. 3. § 4, p. 826.

Sect. 2. STRABO : AFRICA. 329

were still very uncivilized, and led for the most part the life
of mere nomads. He dwells at some length upon the wild
animals with which the country abounded,^ and assures us
that besides lions, panthers, and other wild beasts it produced
abundance of elephants, and the rivers contained crocodiles
similar to those in the Nile.^ It was apparently this circum-
stance that had given rise to the absurd notion (adopted and
developed, as we have seen, by Juba) * that the Nile really
took its rise in the mountains of Mauretania.

It is singular that Strabo never alludes to the work of Juba,
of which he appears to have been totally ignorant, though it
certainly contained the best and fullest information concerning
Africa that was available in his time. Many of his statements
indeed concerning the wild animals and natural productions
of Africa, coincide with those cited by Pliny from the work of
the Numidian monarch ; but these had been doubtless men-
tioned by other writers also. The only authority referred to
hi/ name is Iphicrates, an author otherwise unknown. Con-
cerning the western coast of Mauretania he had evidently very
little knowledge, and tells us that the subject had been so
much disfigured by fables, that it was difficult to know upon
what information to rely. He mentions, though not without
an expression of doubt, the number of colonies that the Cartha-
ginians were said to have established on this coast, of which
(he says) not a trace remained.^ Their number, which was
reported at three hundred (!) was certainly a great exaggera-
tion ; but there is no reason whatever to doubt the fact that
such colonies, or trading stations, had been established outside

2 Ibid. §§ 4, 5, pp. 826, 827.

3 xvii. p. 826, 827. Camelopards
also were mentioned by an author
named Iphicrates, as being found in
the land of the Western Ethiopians
which adjoined the Atlantic, as well
as animals that he calls piC^is, a name
otherwise unknown.

* See Chapter XX. p. 174.
^ ^oivLKLKas 3e •iroAeis Trafj.Tr6\?\as TLvds,
uv ov5iv I5up iariv '[x''°^- xvii. 3, § 8.

This statement is taken from Artemi-
dorus, who censured Eratosthenes for
having believed in their existence. It
is very strange that Strabo never refers
in this part of his work to the voyage
of Polybius along this western or
Atlantic coast of Africa (see Chapter
XVII. p. 32). On such a point as this,
for instance, his testimony would have
been conclusive.

330

HISTOKY OF ANCIENT GEOGEAPHY.

Chap. XXII.

the Straits of the Columns for a considerable distance along
the western coast of Africa. In Strabo's time there appears to
have been no permanent settlement (or at least he knew of
none) further south than Lixus, the modern El Araish, only
about 40 G. miles south of Cape Spartel.^

§ 7. The name of Mount Atlas was of course long familiar to
the Greeks as that of the mountain range so conspicuous as one
sailed through the Straits ; and Strabo was well aware that the
same range was prolonged through the whole extent of Maure-
tania, and in a certain sense as far even as the Syrtes.'' Beyond
this first, or coast, range dwelt the Gsetulians, whom he describes
as the greatest people in Africa, comprehending obviously
under that name all the different but cognate tribes, which
under the name of Berbers, Tuaricks, and other appellations,
actually extend from the neighbourhood of the Atlantic to the
borders of Cyrenaica. All the Libyan tribes indeed, as he
expressly tells us, resembled one another in their dress and
habits of life, which were in great measure the same with those
of the Moors and Numidians.^

Farther inland were situated two nations to which he gives
the names of Pharusians and Nigretes, or Nigritae, who adjoined
the Western Ethiopians, with whom they appear to have had
something in common.^ But he gives us no further clue to

* Even ■SN'itli regard to this, his state-
ments are strangely confused ; he has
certainly confounded the Lixus of Era-
tosthenes, which was called Linx by
Arteuiidorus, with Tingis, the modern
Tangier, which was situated, as he
correctly tells us, very near to the
Promontory of Cotis (Cape Spartel).
Hence he places it opi^osite (avrl-
iropQfjiov) to Gades, the distance between
them being 800 stadia (80 G. miles),
about the sume, he adds, as the dis-
tance of each from the Straits (xvii.
3, § 2). All this is strangely inaccurate,
and shows that he had no trustworthy
information :\i all, as to distances, even
concerning the immediate neighboui-
hood of the Straits.

' xvii. pp. 825-827. He tells us that

Atlas was the Greek name for the
mountain tliat was seen on the left hand
on passing through the Straits ; but
the native name was Dyris (Aupis). He
subsequently adds that the same range
of mountains extended from Cotes to
the frontier of the Massajsyli ; and
afterwards (p. 829) states that the
mountain district in the interior was
prolonged as far as the Syrtes. But he
does not attempt any descrij^tion of the
mountain range, which was doubtless
very little explored.

« xvii. 3, ^ 7, p. 828. He through-
oiit calls the inhabitants of Mauretauia
Maiuusians (Jslavpovaioi), a name appa-
rently adopted by the Greeks as equiva-
lent to the Latin Mauri.

9 Ibid.

Sect. 2. STRABO : AFRICA. 33 I

their geographical position ; except that they were clearly
separated from the settled portions of Mauretania and Numidia
by a considerable extent of desert, as he describes them as
occasionally visiting those countries, " crossing the desert with
skins full of water hung under the bellies of their horses." It
is probable therefore that the tribes thus designated were
really situated to the south of the Great Desert or Sahara.
This is confirmed by the mention of their country being
subject (like the south of Ethiopia) to tropical summer rains.^
But Strabo's own idea of their position was evidently extremely
vague. In another passage he tells us that it was these same
tribes that had destroyed the Carthaginian colonies on the
west coast ; and that they were situated thirty days' journey
from Lixus.^

§ 8. Of the countries bordering on the Mediterranean,
Mauretania Caesariensis and Numidia, Strabo's account is very
brief and perfunctory. He appears to have had but little
information concerning them, and does not even allude to the
numerous colonies which, as we learn from Pliny, had been
settled along this line of coast by Augustus, and must therefore
have been already in existence when Strabo wrote. Even of
the province of Africa, comprising the immediate territory of
Carthage, his description, though correct, is succinct and
summary. This province, as well as the adjacent Numidia,
had suffered severely in successive wars, and the period of the
great wealth and prosperity, to which it attained under the
Koman Empire, does not appear to have yet begun. But
the new colony founded by Julius Csesar on the site of
Carthage was already rising rapidly into importance, and was
become the most populous city in Africa.^

He describes in considerable detail the coast from Carthage
to the Cyrenaica, with the two Syrtes, and the islands of
Cerciua and Meninx — the latter of which, he tells us, was

* AiyeruL Se Kavravda tovs Qepivovs I ^ Ibid. § 3, p. 826.
ofx^povs iTriTToKd^etv. Ibid. p. 828. | ^ xvii. 3, § 15, p. 833.

332 HISTORY OF ANCIENT GEOGEAPHY. Chap. XXU.

generally believed to be the land of the Lotophagi, described
by Homer.* It in fact abounded with the tree bearing a sweet
fruit, to which the Greeks gave the name of Lotus. For this
information he was probably indebted to Polybius, who, as we
have seen, had conducted an exploring voyage along this
coast of Africa;^ and it is not improbable that the same writer
was his chief authority for his description of this coast in
general. But it is curious, and characteristic of the sort of
geographical knowledge possessed by the Greeks, even at this
period, that while his details of the -paraplus are in general
very correct, and his distances at least fairly accurate, he was
still so ignorant of the general form and configuration of the
coast, as to have no clear conception of the great projection
formed by the Carthaginian territory, and the deep bay to the
east of it. Hence he tells us that Automala, a port in the
innermost bight of the Great Syrtis, was on a parallel of lati-
tude about 1000 stadia south of Alexandria, and less than
2000 south of Carthage.^ The first statement is not very
far from correct, while the difference between the parallel in
question and that of Carthage is not less than six degrees and
a half of latitude, or 3900 stadia ! Yet his account of the
Great Syrtis itself is very fairly accurate, and free from the
gross exaggerations which had been accumulated by earlier
writers.

§ 9. Of the Cyrenai'ca he gives a pretty full, and very correct,
account. With the coast of this region he was acquainted
by personal observation, as he distinctly tells us that he had
seen the city of Cyrene from the sea.'' This must probably
have been on his voyage from Italy or Sicily to Alexandria.
He states also that from a headland named Phycus (now Eas
Sem) near Cyrene, which was the most northerly point of
this coast, to Cape Tsenarus (Matapan) in the Pelopon-
nese was 2800 stadia; a remarkably correct estimate, which

' xvii. 3, § 17, p. 834. I « xvii. 3, § 20, p. 83G.

^ Soc Chapter XVII. p. 32. j ' Ibid. p. 837.

Sect. 2.

STRABO: AFRICA.

333

affords us valuable assistance in constructing his map of the
Mediterranean.^

The region producing the siljohium, for which Cyrene had
so long been famous, was situated immediately beyond the
inhabited district, towards the interior ; it was a barren tract
extending about 1000 stadia in length by 300 in breadth.^
The interior beyond this was inhabited by the Marmaridse, a
nomad tribe who extended as far as the Oasis of Ammon.
Strabo was acquainted also with the Oasis of Augila,^ which
he places at four days' journey to the south-east of Automala,
and correctly describes as resembling that of Ammon, having
good water and abundance of palm-trees. But of the native
tribes his knowledge was evidently very imperfect ; he men-
tions the Nasamones and Psylli as dwelling around the Great
Syrtis; and elsewhere notices the Garamantes as occupying
the interior beyond the Gsetulians. He adds that they were
distant about nine or ten days' journey from the Ethiopians
on the Ocean, and fifteen from the Oasis of Ammon.^ It is
strange that he makes no allusion to the recent expedition of
Balbus into their country, or to the more definite information
which he had brought back from thence.

It is to the credit of Strabo, and in accordance with
that soberness of judgment which in general distinguishes
him, that he does not attempt, in the absence of authentic
information concerning the interior of Africa, to supply the
deficiency by repeating the fables which had been so long
current in regard to this part of the world, and which still

8 Ibid. p. 837.

9 Ibid. § 22, p. 837; § 23, p. 839.
He tells us that the silphium had at
one time been very nearly extirpated
by the barbarians. The limited area
to which it was confined may account
for its disappearance, or rather degen-
eracy, at the present day. See on this
subject Barth, Wanderungen, pp. 410,
468.

^ xvii. § 23, p. 838. The name had
fallen out of our MSS., but has without
doubt been correctly restored by recent

editors. (See Kramer's note.)

^ xvii. 3, § 19, p. 835. His concep-
tion of the Garamantes seems to have
been that of a nation extending for a
long distance from E. to W., to the
south of the Gastulians, and conse-
quently farther in the interior. But
he had evidently no definite idea of
their locality. The statement that they
were not more than ten days' journey
from the Ethiopians on the Ocean, is
wholly unintelligible.

334 HISTORY OF ANCIENT GEOGRAPHY, Chap. XXII.

continued to find their place long after in the works of Koman
writers. It is more singular that he has omitted in the
description of Africa all notice of those Islands of the Blest,
or Fortunate Islands, to the existence of which not far from
the coast of Mauretania he has incidentally alluded in the
earlier part of his work.^ This circumstance alone would be
sufficient to show that he had not made use of the work of
Juba, who, as we have seen, had collected a considerable
amount of information in regard to them.

§ 10. But if we are surprised to find that Strabo had failed to
avail himself of valuable works that had certainly been pub-
lished at the time when he wrote his geography, our wonder
may well be diminished when we find (as has been already
mentioned) that his own great work remained unknown to
most of his successors. His name is not even noticed among
the multifarious writers cited by the all-compiling Pliny ; nor
is any allusion to it found in the great work of Ptolemy. His
geographical treatise forms indeed so important an era to our-
selves in estimating the progress of geography that we find it
difficult to believe that it did not assume an equally important
part in the eyes of his contemporaries and their immediate
successors. But the silence of Pliny, half a century after-
wards, is conclusive evidence that this was not the case. If
his great work was written (as is generally supposed) at
his native place of Amasia, in a remote province of Asia,
and completed only a short time before his death, this might
in some measure account for the tardy recognition of its
merits. But it could hardly have failed to find its way to
Alexandria, where he had himself studied, and which was
still in great measure the centre of learning to all the Hellenic
world.

It was certainly in the hands of the learned as early as the
time of Athenseus (about the beginning of the third century),

^ iii. 150. MaKo.puvTLi'a,': VTjcrovs Kara- I ou rroXv a-KoBev twv &Kpo}V Trjs Mavpovalas
vofid^quTes &s Kol vvv SeiKw/xevas tcTfiev | rwu avTiK^iixivwv toIs raSeipois,

Sect. 2.

STRABO: AFRICA.

335

who refers to it in two passages,* neither of them having any
direct bearing on geography : but its geographical importance
is for the first time recognized by Marcianus of Heraclea — a
writer who cannot be placed earlier than the third century —
who mentions Strabo, in conjunction with Artemidorus and
Menippus of Pergamus, as one of the authorities most to be
relied on with respect to distances.^ With this exception we
find hardly any reference to it till the time of Stephanus of
Byzantium, towards the end of the fifth century, by whom it
is frequently cited. Among the later grammarians of the
Byzantine times on the contrary it enjoyed a high reputation,
and is continually referred to by Eustathius, who even calls
Strabo the geographer (6 'yecojpd(po<i) par excellence, notwith-
standing the commanding position then occupied by Ptolemy,^
It is certain that if we regard the science of geography as
including all its branches, historical, political, physical and
mathematical — there is no other writer upon the subject in
ancient times that can compare with Strabo.

* Athenseus, iii. p. 121 ; xiv. p. 657.
It is remarkable that his historical work,
■which v?as subsequently forgotten, is
repeatedly cited at an earlier period,
both by Josephus {Ant. Jud. xiv. 7,
§ 2), and by Plutarch {Sylla, c. 26,
Lucullus, c. 28).

^ Marciani Upit. § 3.

^ The existence of two different
Epitomes, compiled independently of
each other, of the great work, both of
which have been preserved to us, and
are of considerable use in correcting and
confirming the original text, is an ad-
ditional proof of the popularity of his

work in the middle ages. One of these
dates from the tenth century, and is
therefore considerably older than any
of our existing MSS. of Strabo. The
other, though preserved only in a MS.
of the 14th century, is also of consider-
able value. (See Kramer's Pvcefatio,
p. xlii.)

But notwithstanding /these aids, the
defective character of our MSS., and
the frequent corruptions of the text,
which it is impossible for us now to
rectify, are a source of continual em-
barrassment and regret to the student
of ancient geography.

336 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXII.

NOTE A, p. 296.

CHALYBES.

The Chalybes oi- Chalybians were a people of Asia Minor whose
name was certainly familiar to tlie Greeks from an early period.
Tiiey appear in the Prometheus of ^schylus (v. 715) as mSyjpo-
T€KTovcs, or workers in iron: and Herodotus notices them among
the nations subdued by Croesus (i. 28). They are here introduced
as if they dwelt within (i.e. to the west of) the Halys : but this
may be only a slight inaccuracy of expression, and there seems no
doubt that they were really situated to the east of that river.
ApoUonius Ehodius, who on a point of this sort probably followed
good geographical authorities, placed them beyond the Thermodon,
the reputed abode of the Amazons, and next to the Tibarenians
[Argonaut, ii. v. 1000-1008). Dionysius Periegetes, who assigns
them the same position (v, 768-771), probably followed ApoUonius.
Strabo also associates the Chaldseans (whom he distinctly identifies
with the people formerly called Chalybes) with the Tibarenians,
Macrones, and Mosynoecians, but describes them as inhabiting the
rugged mountain country above Pharnacia (Cerasus) (xii. 3, § 18).
Xenophon in his description of the retreat of the Ten Thousand
along the shores of the Euxine, places the Chalybes between the
Mosynoecians and Tibarenians, and says they were a small tribe
subject to the Mosyncecians, and subsisting principally by working
in iron. [Anab. v. 5, § 1.) Hamilton (Mesearches in Asia Minor,
&c., vol. i. p. 275) found a people whom he describes as working
iron, which was found in abundance near the surface of the soil,
without the labour of mining, and of excellent quality, in the
neighbourhood of the modern Unieh, between the mouth of the
Thermodon and the Jasonian Promontory : and these he very rea-
sonably regards as representing the ancient Chalybes, This
position would agree with that assigned to them by ApoUonius,
rather than with that' of Xenophon and Strabo. But Xenophon
could hardly have been in error in placing them east of the
Tibarenians, whose position is clearly fixed by that of the Greek
settlement of Cotyora, which was in their territory {Anah. v. 5, § 3)
and which was certainly either at or near the modern town of

Note A. STRABO : AFRICA. 337

Ordu. It was 180 stadia east of the Jasonian Promontory (Arrian,
Periplus, § 23). The probable explanation of the discrepancy
appears to be that the Chalybes were originally a more consider-
able people, occTipying the south coast of the Euxine to a greater
extent, who had been broken up and driven out of part of their
abodes by the irruption and invasion of other tribes, while detached
portions of them retained their ancient name and habits, and con-
tinued to work at their manufacture of iron in the manner that
they do to this day.

Apollonius gives a striking description of their labours, and the
appearance of their country, which must have been a poetical
exaggeration as applied to the Chalybes, but would be no untrue
picture of the " Black Country " of Staffordshire in our own days :

vvKTi r' iTTiirXofievri XaXvfiuy irapci, yaiav 'Ikovto.
TOLffi fihv oUre jiowv &poTos /xeXei, ovt4 tis &X\r)
^uTa\jr) KapTToto fn.eX'Kppouos' ov Be fiev o'lye
iroifiuas epff-fjevri vo/j.^ evi iroifiaivovo'LV.
aWa ffi5r]p6<popov cTTvcpeK^v x^Jya yaro/xdovres
Sivov CLfiei^ovTai fiioTi\(Tiov' ouSe ttotc ({(piv
rjibs dvTeAAei Ka/xdrcoy arep, aWa KeXaiyfj
MyvvC Ka\ Kairv^ KajxaTov fiapvv orXevovcnv.

Argonaut, ii. vv. 1001-1008.

VOL. IT.

( 338 )

CHAPTEE XXIII.

STRABO TO PLINY.

Section 1. — 8trdbo to Pliny.

§ 1. The period of about half a century which intervened
between the death of Strabo and the publication of the ency-
clopaedic work of the great Eoman naturalist, Pliny, was not
in general marked by any great advance in geographical
knowledge. With one important exception, to which we shall
presently return, the limits of that knowledge remained much
the same, or were extended only in a vague and uncertain
manner. The greater part of the known world, as we have
seen, was already comprised within the Eoman Empire, while
to the east the Parthian monarchy opposed a barrier to its
advance which was never permanently transgressed, and the
barbarian nations on the north were generally in a state of
hostility with Eome, which precluded to a great degree all
exploration in that direction.

In two quarters only were the limits of the Eoman Empire
extended during the interval from the death of Augustus to
that of Vespasian. These were Britain and Mauretania. The
former, as we have seen, had been unmolested by the Eoman
arms from the time of the dictator Caesar. Augustus and his
successor Tiberius had been content to leave the islanders in
the possession of their liberty, receiving honorary embassies
from time to time from the petty princes of the tribes nearest
the coast, and apparently encouraging and promoting commer-
cial relations between their Gaulish subjects and their opposite
neighbours. These relations had certainly attained to a very
considerable extent, and Londinium (London) had already

Sect. 1.

STBABO TO PLINY.

339

risen to be an important emporium of trade, and the seat of
a considerable population.^ The coins of Cunobeline, a king
of the Trinobantes, who was a contemporary of Augustus, bear
also a striking testimony to the statements of Eoman writers
concerning the resources and opulence of the island in his
time.^ Gold and silver were reported to exist in considerable
quantities, and British pearls enjoyed a reputation, which was
found to be beyond their merits when they came to be better
known, as they could never rival those of the East in lustre or
perfection.^

§ 2. The quarrels of the petty British princes among them-
selves soon led to their invoking the interposition of Eome ;
and an abortive attempt on the part of Caligula, amounting in
fact to a mere display, became the prelude to a serious inva-
sion in the reign of Claudius. In a.d. 43 Aulus Plautius
landed in the island with an army of four legions. It is a
curious proof of the kind of mysterious greatness that seems to
have still attached to the idea of Britain, — notwithstanding
the commercial intercourse of which we have just spoken, as
well as the expedition of Caesar, a century before, — that we
are told the legionaries at first refused to embark on an enter-
prise which was to lead them beyond the limits of the known
world.* But the first campaigns presented little difficulty,

^ " Londinium, cognomento quidem
coloniaa non insigne, sed copia'nego-
tiatorum et commeatuum maxime
celebre." Tacit. Annul, xiv. 33. Tacitus
indeed is speaking of a period after the
beginning of tlie Eoman occupation,
but so considerable a trade could hardly
have arisen within a few years.

^ They are found in great numbers,
both in gold and silver (see Evans's
Coins of the Ancient Britons, 8vo. Lond.
1864). The abundance of them seems
to testify to the accuracy of the state-
ment of ancient writers that both
metals were found in Britain, in such
quantities as to be a temptation to its
conquest. "Fert Britannia aurum et
argentum et alia metalla, pretium vic-
toria3." Tacit. Agric. c. 12. See also

Strabo, iv. 5, § 2, who distinctly notices
both gold and silver as articles of
export from Britain.

^ See Note A, p. 369.

* Dion Cass. Ix. 19. as yap e|co rrjs
olKov/jLevrjs aTparevcrSfieyoL TiyavaKTOvv.
The same idea is found in the rhetorical
declamation of Josephus (writing in
the reign of Vespasian) where he
makes king Herod Agrippa II. describe
the Romans as not content with the
limits of the known world, and seeking
another world beyond the Ocean, by car-
rying their arms among the unknown
Britons.. aW' virep uiKeavhv krepav i^'fl-
T-qaav olKovfj.evriv, Kal fJ-exp^ ''''^^ avLffTopi)-
TOiv TrpSrepov BpeTTavHi/ SiTipeyKav Tct
(JirAa. (Joseph, iy. Jiid. ii. 16, § 42.)

z 2

340

HISTOEY OF ANCIENT QEOGEAPHY. Chap. XXIII.

and Plautius was able to prepare the way for the emperor him-
self, who soon after followed to earn an easy, if not a bloodless,
victory. He crossed the Thames, defeated the Trinobantes,
and took Camulodunum, where Cunobeline had established
his capital.^

Claudius himself soon quitted the island,® but he left his
generals to continue its conquest, and it appears that they
quickly reduced all the southern tribes to subjection, or at
least to submission. But the extent to which the Eoman arms
were actually carried at this period we have no means of deter-
mining. Vespasian, who was afterwards emperor, served as a
legate in these campaigns, and we are told that he subdued
two of the most powerful nations (their names are not given)
and reduced the Isle of Wight (Vectis), a conquest which
probably attracted attention from its being a separate island.^
The supposition that he advanced as far as Exeter is a mere
conjecture :^ but there seems no doubt that the Eoman autho-
rity was gradually established as far as the Severn (Sabrina) :
and when in a.d. 47 the new governor Ostorius Scapula suc-
ceeded to Plautius in the command, he appears to have found
himself already master of the central as well as southern por-
tions of the island. This may be clearly inferred from the
fact that we find him engaged in hostilities, first, against the
Iceni, who occupied Suffolk and Norfolk, next, against the

s Dion Cass. Ix. 19-21.

« The statement of Suetonius {Claud.
17), that he spent only a few days in
the island (" sine ullo pra^lio aut san-
guine intra paucissimos dies parte in-
sulEe in deditionem recepta") is pro-
bably an exaggeration ; but the whole
time of his absence from Kome did
not exceed six months.

' " Duas validissimas gentes, su-
perque viginti oppida, et insulam_ Vec-
tem, Britannise proximam, in ditionem
redegit " (Suet. Vesjyas. 4).

The two nations not being named
have left free scope to the conjectures
of antiquarians. It has been supposed
by many writers that the tribes in

question were the Belgse and Dumnonii.
That the former, who inhabited Hamp-
shire, should have been subdued by
Vespasian before he attacked the Isle
of Wight may be considered certain.
But it is unlikely that he should have
advanced so far to the west, as to re-
duce the Dumnonii, who held Devon-
shire and Cornwall, without any more
distinct notice being found of so im-
portant a conquest.

« This is admitted by Mr. Merivale,
who acknowledges that there is no
authority on the subject {Hist, of the
Romans, vol. vi. p. 28). See Note B,
p. 369.

Sect. 1.

STEABO TO PLINY.

341

Cangi, not far from the Irish Sea,^ lastly, against the Silures,
who inhabited South Wales and the countries bordering on
the Severn. The Ordovices in North Wales were soon in-
volved in this last war — rendered famous by the exploits and
captivity of Caractacus (a.d. 51) : while the Brigantes, north
of the Mersey, one of the most powerful nations of Britain,
as they held almost the whole of Lancashire and Yorkshire,
entered into friendly relations with the Eoman governor. The
foundation of a Eoman colony at Camulodunum (Colchester)
was, in pursuance of the ordinary policy in such cases, the first
step to the permanent establishment of the authority of Kome,
and the southern portion of the island was reduced to the
form of a province, and brought under the usual conditions of
provincial administration.^

§ 3. This state of things continued under the reign of Nero
until the year a.d. 61, when Suetonius Paulinus, solely with
a view to obtain credit for military successes, attacked and
reduced the Island of Mona (Anglesey), which had up to this
period continued to be the chief seat of the Druids and their
religion.^ This exploit was followed by a sudden outbreak of
the Iceni under their queen Bonduca or Boudicea, who took
and plundered the newly founded colony of Camulodunum, as
well as the two flourishing towns of Verulamium (St. Alban's)
and Londinium f but they were reduced to submission by a

^ The site of the Cangi is wholly
uncertain. The only clue to their
position is derived from this passage
(Tacit. Annal. xii. 32), from which we
learn that the Roman general had
advanced through their country till he
found himself near the sea-coast, which
faced Ireland (jam ventum hand procul
mari quod Hiberniam insulam aspectat).
The supposition that they were situated
in Caernarvonshire, because Ptolemy
has a promontory named Ganganum in
that part of the island, appears to me
whoUy untenable. Mr. Beale Poste
places them "westward of the Cori-
tani," which is plausible enough, hut
rests on no authority.

' " Kedactaque paulatim in formum

provincise proxima pars Britannise"
(Tacit. Agrio. c. 14). This he describes
as taking place within the time that
Aulus Plautius and Ostorius Scapula
were governors. What were the limits
of the province as thus first constituted,
we are not told, nor is it of much im-
portance. It almost certainly did not
extend beyond the Severn to the west,
or the Mersey and Humber to the north.

* Tacit. Annal. xiv. 29, 30 ; Agric. 14.

^ Tacitus asserts that not less than
70,000 " citizens and allies " were mas-
sacred in these three towns {Annal.
xiv. 33), and Dion Cassius raises the
number to 80,000 (Ixii. 1). This is
evidently an exaggeration, but still it
may be taken as a proof that they

342

HISTORY OP ANCIENT GEOGRAPHY. Chap. XXIU.

single defeat and henceforward subsided without further re-
sistance into the condition of provincial subjects of Eome.*
We hear of no further hostilities in Britain till the reign of
Vespasian, under whom Petilius Cerialis turned his arms
against the Brigantes, and appears to have effectually broken
the strength of that powerful nation.^ His successor Julius
Frontinus (a.d. 75) imitated his example by attacking and
subduing the Silures, who, notwithstanding their defeat under
Caractacus, had evidently still retained their independence in
their rugged and mountainous country.^ Agricola, who was
appointed to the government of Britain in a.d. 78, opened his
career by a similar campaign against the Ordovices, the
inhabitants of North Wales, and carried the Roman arms for
the second time across the Menai Strait into the island of
Mona.^

At this time therefore it may fairly be said that the
whole country to the south of the Tyne was either actually
reduced under the dominion, or at least acknowledged the
authority, of Eome. Their arms had not yet penetrated
into the northern part of the island, and the name of the
Caledonians had as yet scarcely reached their ears.^ But in
some way or other they had certainly obtained authentic
information concerning the Orcades (Orkneys) as a numerous
group of islands at the northern extremity of Britain. The
statement of late writers (Eutropius and Orosius) that they
were conquered by Claudius, is certainly erroneous ; but on the

really contained a considerable popu-
lation. The statement of Dion (I. c.) of
the large sum of money that the philo-
sopher Seneca had put out to interest
among the Britons, is also an evidence
of the extensive commercial relations
that had been ah-eady established in
the province.

■• " Uuius prfelii fortuna veteri jmti-
cntiaj restituit." (Tacit. Agric. c. 16.)

For the particulars of the revolt, see
Tacitus (Annal. xiv. 31-39) and Dion
Cassius (Ixii. 1-12).

^ Tacit. Agric. 17.

» Ibid.

' Ibid. 18.

* No mention is found of the Cale-
donians in Pomponius Mela, nor does
even Pliny notice the name as that of
a nation ; but speaks vaguely of " the
Caledonian forest" as the fai'thest limit
of the Roman conquests, which it had
taken tliein thirty years to reach, " tri-
ginta prope jam annis notitiiim ejus
(Britannia) Eomanis arniis non idtra
vicinitatem silviB Caledonia3 jn'opagau-
tibus" {Hist. Nut. iv. 16, § 102).

Sect. 1.

STBABO TO PLINY.

343

other hand the assertion of Tacitus that they were first dis-
covered, as well as subdued, by Agricola, is clearly disproved
by the fact that they are distinctly mentioned both by Mela
and by Pliny.®

§ 4. On the side of Germany little, if any, advance was made
in the period of which we are now treating. The resolution
adopted by Tiberius, to recognize the Ehine as the established
limit of the Eoman Empire in this direction, was practically
followed by his successors. We hear of no more expeditions
to the banks of the Weser and the Elbe ; ^ and when on one
occasion Corbulo, the ablest general of his day, was about to
advance into the land of the Chauci, to punish them for their
incursions on the Eoman allies, he was at once recalled by a
peremptory mandate from Claudius.^ He attempted to make
up for this disappointment by constructing a navigable canal
from the Meuse to the Ehine, to obviate the necessity of trans-
porting troops by sea in moving from one river to the other.

Meanwhile numerous changes were taking place in the
interior of Germany itself, which we are very imperfectly able
to follow. Continual wars had arisen among the different
tribes, leading in some cases to the destruction or humiliation
of nations that had once been among the most powerful of the
native races of Germany; in others to their migration and
change of abode. Thus we find the power of the Cherusci,
who had played so prominent a part in the earlier wars with
the Eomans, in great measure broken by internal dissensions ;
the Chatti sustained a severe defeat from the Hermunduri;

9 Eutropius (vii. 1 3) says of Claudius,
" Quasdam insulas etiam ultra Britan-
niam in Oceano positas, Komano im-
perioaddidit, quse appellantur Orcades."
The same statement is made by Orosius
(vii. 6). Tacitus boasts that Agricola
" simuH'ncogTOtos ad id tempus insulas,
quas Orcadas vocant, invenit domu-
itque " (Agrie. c. 10). It is very pro-
bable that they were not visited by a
Eoman fleet till the time of Agricola ;
but they were certainly known by

name, and hearsay report, long before.
Probably this first knowledge of them,
though not their conquest, really dated
from the time of Claudius.

^ Hence Tacitus, writing after a.d.
100, says with some bitterness of the
Elbe : " Albis, flumen inclitum et no-
tum olim, nunc tantum auditur " ( Germ.
c. 41).

2 Tacit. Annal. xi. 20; Dion Cass.
Ix. 30.

344

HISTOKY OP ANCIENT GEOGRAPHY.

CHAP.XXin.

and the latter people joined with the Lygii and other less
known tribes to expel Vannius, a king of the Suevi, or rather
of the tribe called Quadi, who had been recently settled in the
country now called Moravia.^ Vannius was driven across
the Danube and took refuge in the Eoman territories, where
the emperor, though he had refused to interpose in the war,
afforded him a secure asylum. In another instance Claudius
consented to nominate a king for the Cherusci, at their own
request, a step which had however only the effect of increasing
their domestic dissensions. But the relations thus subsisting
between the Eomans and their German neighbours could not
but lead to increased intercourse between them, and to the
gradual diffusion of that enlarged knowledge of the country
and its inhabitants, which we subsequently find in existence,
without knowing from what source it was acquired.

§ 5. In one instance only do we find anything like systematic
inquiry, and unfortunately in this case also with very little
clefinite result. We are told by Pliny that in the reign of
Nero a Eoman knight was sent by one Julianus, who had the
charge of a gladiatorial show given by the emperor, in quest
of amber, and that in pursuit of this object he penetrated
across the continent of Germany to the shores of the Northern
Sea. This he reported to be distant 600 Eoman miles from
Carnuntum in Pannonia, from whence he set out, and he is
said to have explored the shores of the Ocean thus discovered
for some distance.* He brought back enormous quantities of
amber, so that the very nets which protected the spectators
from the wild beasts in the arena were studded with it;^ but
unfortunately we have no geographical details, and are left
wholly in the dark as to any geographical results he may have

^ Tacit. Annal. xii. 27-30.

* Plin. H. N. xxxvii. 3, §45. " Sex-
centis fere M. pass, a Camunto Panuo-
nise abest littus id Germanise, ex quo
invehitur, percognitum nuper. Vidit
enitu eques Romanus, naissus ad id com-
paranduin a Juliano ciirante gladia-

toriutn munus Neronis principis, qui
hsec commercia et littora peragravit."
The last words would have led us to
hope for geographical information
which we do not find.

* The largest single mass weighed
not less than 13 pounds. Plin. I. c.

Sect. 1.

STRABO TO PLINY.

345

brought home. It may however be considered certain that he
reached the shores of the Baltic, which have been in all ages
the great repository of this valuable product ; a journey which
would in fact present no great difficulties, if he was able to
secure a friendly reception from the different tribes that he
encountered on his route. Indeed the choice of Carnuntum, a
place so far to the eastward, as his starting-point, can only be
explained on the supposition that there was already a trade
established between Pannonia and the amber-lands, and that
he could thus obtain information from the native traders of
the situation of those lands, and the general course to be
pursued. Pliny indeed intimates distinctly that it was through
Pannonia that amber had first come to be generally known,
and it was from thence it was carried to the head of the
Adriatic, where it was so long supposed to be produced.®

But whatever additional information this solitary explorer
may have brought back concerning the amber trade, it seems
certain either that he had really acquired no geographical
information of any value, or at least that Pliny had no access
to it ; for the utterly vague and indefinite ideas, which that
writer possessed concerning the shores of the Northern Ocean,
exclude the supposition that he had consulted the authentic
statements of any person who had himself visited those shores.'^
The circumstance that neither he, nor any other Eoman writer
before the time of Ptolemy, notices so important a river as the
Oder, is sufficient proof how little acquaintance they really
possessed with these countries. It appears at first singular
that while they had apparently never heard of the Oder,

^ " Famam rei fecere proximse Pan-
nonise, id accipientes circa mare Adri-
aticum." Plia. H. N. xxxvii. 3, § 44.
He adds, plausibly enough, that the
fables which connected it with the
Padus, arose from the habit of the
women of those countries wearing neck-
laces of amber, as they still did in his
day.

^ It is suggested by Ukert (Germa-
nien, p. 181) that the very precise and

definite information given by Tacitus
{Germania, c. 45) concerning the lo-
cality and mode of collection of amber,
which he distinctly confines to the
JUstii, a tribe not mentioned by Pliny,
may be derived from the accounts
brought home by this Eoman knight.
But this would render it all the more
difficult to account for the silence of
Pliny as to that people.

346 HISTOEY. OF ANCIENT GEOGEAPHY. Chap. XXIII.

they were familiar with the name of the Vistula, but their
knowledge of this latter river was probably obtained, not
through Germany, but through Pannonia and Sarmatia ; and
it is not unlikely that its valley was the channel by which
the amber trade with Pannonia had so long been carried on.^

§ 6. On their eastern frontier the Romans were engaged in
repeated hostilities with the Parthians, of which the possession
or rather dominion of Armenia was generally the occasion, or
the prize. That country was still governed by its native
princes, and retained nominally an independent position, but
the Romans and Parthians, by supporting the claims of rival
pretenders to the throne, sought in fact to establish their own
supremacy, while the unhappy Armenians were the victims in
turn of both contending powers. Corbulo, whose successes in
the East were regarded as rivalling those of Pompey, took and
destroyed Artaxata, the capital of Armenia, and made himself
master of Tigranocerta, which was still a populous city, and a
strong fortress.^ But though these successive campaigns must
have increased the knowledge possessed by the Romans of this
rugged and mountainous country, and refreshed the memory
of that previously acquired by Lucullus, they did not extend
their acquaintance with the neighbouring regions or contribute
in any considerable degree to enlarge the sphere of their geo-
graphical knowledge. South of the mountains of Armenia the
Euphrates formed the boundary of the Roman Empire, which
was not crossed by a Roman army, from the time of Augustus
to that of Trajan.

§ 7. It is very rarely that we find in ancient times the
boundaries of geographical knowledge enlarged by an expedi-
tion intentionally undertaken for the purpose of discovery, but

^ From Carnuntum it would be easy
to ascend the valley of the March, and
thence cross the mountains to the
sources either of the Oder or the
Yistiila.

Tacit. Annal. xv. 4. " Occupa-

fensorum et magnitudine moenium
validam." It is mentioned also by
Pliny in. N. vi. 9, § 26), as one of the
chief cities of Armenia, and its name is
still found in Ptolemy (v. 13, § 22], but
from this time it disappears. Coucern-

verat Tigranoccrtam, urbem copia de- ' ing its situ see Note D, Chapter XVIII.

Sect. 1. STRABO TO PLINY. 347

we meet with one case in the reign of Nero ; by whom two
centurions were dispatched with orders to ascend the Nile
from Syene, and solve, if possible, the long disputed question
of its origin. It is needless to say that they did not really
accomplish this object, but they brought back information of
much interest and value, and undoubtedly ascended the river
to a higher point than had previously been known to either
Greek or Roman geographers.

Pliny unfortunately contents himself with giving the dis-
tances and some few details, as high up as Meroe, a point that
was already well known ; so that thus far the explorers did no
more than add to the accuracy of topographical details.^ They
reckoned the whole distance from Syene to Meroe (following
the course of the Nile) at 873 Eoman miles ; of which Napata,
the only place worthy to be called a town, and which was
already well known by the expedition of Petronius,^ was
distant 360 miles from the capital. The latter was situated
70 miles above the junction of the Astaboras with the true
Nile, a distance which is found by modern observations to be
just about correct. About Meroe itself they found a com-
paratively fertile country, with verdure and a certain extent
of wood ; traces of elephants and rhinoceroses were also seen.
But above this the country was desert, or at least uninhabited,
and no towns were to be found on either bank.^

This is all that we learn from Pliny, who gives us no means
of judging how much farther they actually penetrated. But
a valuable supplement to his account is furnished by Seneca,
who was immediately contemporary with the expedition in
question, and states that he had his information from two
centurions who had formed part of it.* After a long journey

1 Plin. E. N. vi. 29, §§ 184-186. , ^thiopicum cogitanti."

2 See Chapter XX. p. 182. * Seneca, Natural. Qumst. vi. 8. The
^ Plin. vi. 29, § 181. " Hsec (oppida) ! philosopher ascribes the expedition to

sunt prodita usque Meroen, ex quibus \ a piu-e love of inquiry on the part of

hoc tempore nullum prope ntroque j the young prince (" quos Nero Csesar, ut

latere exstat. Certe solitudiues nuper i aliarum virtutum ( ! ), ita veritatis in

renuntiavere principi Neroni missi ab 1 primis amantissimus, ad investigandum

eo milites prsetoriani cum tribuno ad j caput Nili miserat "). Whether this be

explorandum, inter reliqua bella e.t true, or, as Pliny suggests, it was under-

348 HISTOKY OF AKCIENT GEOGEAPHT. Chap. XXIII.

(he tells us) " whicli they had accomplished by the assistance
of the king of Ethiopia, and the recommendations with which
he had furnished them to the neighbouring kings, they arrived
at length at immense marshes, the exit from which was un-
known to the inhabitants, nor could any one hope to discover
. it. So entangled were the waters and the herbage, and the
waters themselves so full of mud, and beset with plants, that
it was not possible to struggle through them, either on foot, or
in a boat, unless it were a very small one containing only one
person. There they added, we saw two rocks, from which there
fell a river with a great mass of water." The last statement
is unintelligible, and must in all probability have really
referred to some other locality; but it is impossible not to
recognize in the rest of the description a correct picture of the
great marshes on the course of the White Nile, above its
junction with the Sobat, which were first rediscovered in
modern times by the Egyptian exploring expeditions in 1839
and 1840, and have recently been rendered familiar to all by
the graphic accounts of Sir S. Baker. No such marshes are
found lower down the course of the Nile, and hence we may
assume with confidence that the explorers of Nero had actually
penetrated as far as the 9th parallel of north latitude, where
the great marshes referred to commence. The friendly recep-
tion accorded them by the king of Ethiopia, and the faci-
lities furnished by him towards their farther progress, will
explain their having advanced so far, and reached a point
which was not again visited by any European for nearly
eighteen centuries.

§ 8. At the other extremity of Africa the Roman arms had
meanwhile been the means of advancing geographical know-
ledge. Mauretania which, as we have seen, had continued
during the reigns of Augustus and Tiberius to be governed by
its own kings, had subsequently been incorporated with the

taken -with an ultimate view to con- | itself was wholly of a pacific charncter.
quest, it is clear that the expedition

Sect. 1.

STRABO TO PLINY.

349

Eoman Empire (a.d. 42), and constituted as a province, being
however divided into two, which, were distinguished as Maure-
tania Tingitana and Mauretania Csesariensis, each retaining
its separate administration. The former comprised the whole
north-western angle of Africa, adjoining the Straits, and ex-
tending eastward as far as the river Mulucha, which had
formerly constituted the limit between Mauretania and Nu-
midia.^ Its extent along the Atlantic coast was probably
ill-defined, but its real boundary in this direction was the
limit of the towns and settlements, the last of which was Sala,
still called Sallee, and situated about 110 Gr. miles south of
Cape Spartel.

It was not long after the Eoman dominion was thus esta-
blished in Mauretania, that Suetonius PauKnus (the same who
afterwards distinguished himself in Britain), being appointed
governor, took occasion to penetrate into the interior, with a
view of subduing the native tribes, and was the first to carry
the Eoman arms across Mount Atlas. He reported the whole
of the lower part of the mountain to be covered with dense
forests of trees of an unknown species : but its summit was
deeply covered with snow even in summer.^ He attained the
highest point in ten days' march, and beyond that proceeded
as far as a river which was called Ger, through deserts of
black sand, out of which there rose from place to place rocks

* In the time of Jugurtha, as Sallust
points out, the Mulucha formed the
boundary between the kingdom of
Bocchus and that of the great tribe of
the Masssesylians, wlio were at that
time considered as belonging to Nu-
midia. Under the Empire, on the
contrary, the whole territory of tlje
Masssesyli was included in the province
of Mauretania Csesariensis, which ex-
tended from the river Mulucha (still
called the "Wady Muluyah) to the
mouth of the Ampsaga (Wady el Kebir).
The provincial appellation of Numidia
was thus limited to the narrow space
between the Ampsaga and the Tusca.
(See Chapter XX. p. 169.)

^ This is probably a mistake ; no

part of the range of the Atlas yet exa-
mined being permanently covered with
snow, though the highest summits
attain an elevation of 12-13,000 feet,
and this in a different part of the
range ; but the great heat and dryness
of the climate combine to prevent the
accumxilation of any great quantity of
snow. But Suetonius, as Phny tells
us directly after, made his expedition
in the winter, and could therefore report
only from hearsay that the snow re-
mained through the summer. It is
indeed at the present day generally
believed and reported by the natives
that a part of tlie range is always
covered with snow.

350

HISTOEY OF ANCIENT GEOGEAPHT. Chap. XXIII.

that had the aspect of being burnt. He found the heat of
these regions such as to render them uninhabitable, although
it was the winter season. The forests adjoining them, which
swarmed with elephants and other wild beasts and serpents of
all kinds, were inhabited by a people called Canarians.''

Interesting as is this narrative, for which we are indebted to
Pliny, who doubtless derived it from the commentaries of
Suetonius himself,^ it is obvious that it is very imperfect, and
leaves the most important geographical questions unanswered.
We do not learn by what pass he traversed the chain, or from
what point the ten days' march was computed. The most
interesting geographical fact that we learn from it, is the
existence immediately south of the Atlas of a river which bore
the name of G-er, an appellation that has given rise to much
controversy, from its being confused, or supposed to be con-
nected, with the far more celebrated Niger, the object down to
our own time of so much discussion and so many exploring
expeditions. Taking the statement of Pliny as it stands, there
seems no doubt that the Ger discovered by Paulinus, was one
of the rivers that take their rise on the southern slope of the
Atlas, and are lost after a cou»se of no great length in the
sands of the Sahara. The most considerable of these is de-
scribed both by Leo Africanus and other Arabic historians
under the name of Ghir, an appellation by which it is known
to this day. This stream has its source in the Atlas, nearly
opposite to that of the Mulucha, and hence it would appear
probable that Suetonius had ascended the valley of the latter
river, one of the most considerable in Mauretania, and crossed
the range near its head-waters. It was by this pass that the
enterprising traveller M. Gerard Kohlfs, to whom we are in-
debted for the latest information concerning this region south

' Plin. V. 1, §§ 14, 15. This mention
of a people called Canarians on the
mainland is curious. It was doubtless
connected with the name of Canaria
given to one of the Fortunate Islands.

8 Pliny himself cites Suetonius

Paulinus among his authorities for liis
fifth book, in which the above narrative
is contained. It is probable therefore
that he had left a written account of
his campaign; the loss of which is
much to be regretted.

Sect. 1.

STEABO TO PLINY.

351

of the Atlas, crossed the mountain range in 1864, and descended
into the valley of the Ghir.

§ 9. It is in all probability also to the period that we are
now considering that must be assigned a voyage, of the date
and circumstances of which we have no information, but
which in its consequences became undoubtedly one of the
most important that was made in ancient times. This was the
voyage of Hippalus, a G-reek mariner, as we may infer from his
name, who being engaged in the trade with India, and having
observed the regularity of the monsoons, was the first to take
advantage of them, and venture to steer a direct course from
the promontory of Syagrus (Cape Fartak) in Arabia to the
coast of India, thus avoiding the whole of the great circuit by
the entrance of the Persian Gulf, and the coast of Gedrosia and
the mouths of the Indus. His example was generally followed,
and the practice had become completely established in the time
of Pliny and the author of the Periplus of the Erythraean Sea.^
Neither of these writers furnish us with any date, but they
both allude to the discovery as a recent one, and as no men-
tion is found in Strabo of so important an innovation — a case
entirely exceptional in ancient navigation — it seems reasonable
to conclude that it took place after the time of that author ; at
all events after the time that he was in Egypt and collected
the notices with which he has furnished us concerning the
trade with India, and the great develoj)ment that it had
assumed during the reign of Augustus.^ But as we have no
details concerning the voyage in question, nor any means of
judging how far it directly contributed to the geographical
knowledge of India, — though it is certain that it led to a great

8 Plin. H. N. Yi. 23, §§ 100, 101 ;

Peripl. Maris JErythriei, § 57, ed.
Miiller.

1 M. Vivien de St. Martin (ie Nord
de I'Afrique dans I'Anliquite, p. 268)
regards this rapid increase of the
Indian trade as arising from the dis-
covery of Hippalus, which he conse-
quently places before the Christian
era; but that discovery itself clearly

implies the previous existence of a con-
siderable trade in that direction, which
made it an object of importance to
shorten the voyage. Nor could any
navigator have seen the expediency of
trusting himself to the monsoon to cross
the Indian Ocean, without having
already acquired a pretty clear idea of
the situation of the countries that he
was seeking.

352 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIII.

extension of the commercial relations with that country, and
therefore indirectly to that vastly increased knowledge of its
geography that we find in Pliny and succeeding writers — it will
be as well to postpone its further consideration till we come to
examine the geographical results that we find embodied in
Pliny and the Periplus.

Section 2. — Fomjponius Mela.

§ 1. The only geographical writer of any importance, who
belongs to the period we are now considering, is Pomponius
Mela, the author of a compendious treatise on geography,
which has derived the more value in our eyes from the circum-
stance of its being the only regular treatise on the subject in
the Latin language that has been preserved to us, with the
exception of that which forms a part of the much more com-
prehensive work of the elder Pliny. It is indeed such a mere
abridgement, and has so little pretension to anything like a
scientific character, that we should have supposed it to have
derived its value almost exclusively from its accidental pre-
servation, did we not find it repeatedly cited by Pliny, in the
imposing array of his authorities, in a manner that seems to
imply that it enjoyed some reputation, even in his day.^

Of the author himself we know nothing beyond his name,
and the fact, which he tells us himself, that he was born at a
place in Spain called Tingentera, the name of which is not
otherwise known, but which appears to have been situated
close to the Strait of the Columns.^ The date of his work may

It is cited among liis authorities j ' ii. 6, § 96. It seems liiglily pro-

for all the four geographical books
(from the 3rd to the 6th), and again
for the 8th, 12th, 13th, 21st, and 22nd,
for individual notices, concerning
animals, trees, &c. But as lie never

bable that Tingentera was in reality
the native name of tlie town called by
Strabo Julia Joza, and by later writers,
and on coins Julia Traducta ; which
had been peopled, as Strabo tells us,

quotes him for any special statements, ' by inhabitants transported thither from

we are unable to determine the full \ Tingis in Mauretania, This would be

extent to which he made use of his j easily reconciled with the statement of

work. ! Mela (Z c), that Tingentera was in-

Sect. 2. POMPONIUS MELA. 353

be gathered with certainty from a passage concerning Britain,
in which he speaks of that island as having hitherto been very
imperfectly known, but about to be much better and more
certainly known from the expedition of the emperor, who was
speedily going to return to Eome, and to celebrate the triumph
which he had earned by his own personal exertions.* This
can hardly be referred to any other emperor than Claudius,
and his expedition to Britain in a.d. 43, an account of which
has been already given.^ As Mela speaks of him as not having
yet returned to Eome, we may place the composition of his
little work in that very year.

§ 2. The arrangement of his materials is peculiar, and is
evidently derived rather from writers who, like Scylax and
the author of the treatise ascribed to Scymnus Chius, had
composed a periphis of the countries bordering on the Medi-
terranean than from general or systematic treatises on geo-
graphy. He begins indeed with a brief description of the
earth, its division into hemispheres (a northern and a southern
one), and into five zones, of which two only were inhabitable,
and it is remarkable that he speaks, as of an undoubted fact
of the existence of anticMJiones, inhabiting the southern tempe-
rate zone, though they were unknown and inaccessible on
account of the heat of the intervening tract or torrid zone.®
He next gives a brief outline of the three continents, Europe,
Asia and Africa, their relative position and their boundaries,
in regard to which he follows Eratosthenes, or rather perhaps
the views that had been generally adopted from the time of
Eratosthenes to his own. Thus he makes the Tanais the

habited by Phoenicians brought over
from Africa (quam transvecti ex Africa
Phcenices habitant). It probably occu-
pied the site of the modern Tarifa.

^ iii. 6, § 49.

' See Note C, p. 370.

^ i. 1, § 4. " Eeliquse (zonte) habi-
tabiles paria agunt anni tempora, verum
non pariter. Antichthones alteram, nos
alteram incolimus. lUius situs ob ar-
dorem intercedentis plagse incognitus,

hujus dicendus est."

This idea of the Antichthones appears
to have been connected with the vague
suggestion of Hipparchus, that it was
uncertain whether Taprobane was an
island, or the commencement of another
world (see Pliny, H. N. vi. 22, § 81).
But Mela does not intimate any such
connection. With him the continent
of the Antichthones seems to have been
a purely theoretical assumption.

VOL. II. 2 A

354

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIII.

boundary between Europe and Asia, and tbe Nile that between
Asia and Africa ; he accepts as an undoubted fact the doctrine
that the inhabited world was surrounded by the Ocean, from
which it received four seas, as inlets or gulfs ; one from the
north, or from the Scythian Ocean (the Caspian) ; two from
the Indian Ocean on the south, — the Persian and Arabian
Gulfs ; and one from the west, by far the most important of
all, but for which, as we have seen, neither Romans nor Greeks
had any distinctive name, and Mela, writing as a geographer,
is constrained to use the vernacular phrase of " Our Sea."^

But after this general outline, instead of following the same
arrangement for his more detailed description, and treating of
the several countries as subdivisions of the three continents
in succession — as is done by Strabo and by all modern geo-
graphers — he begins at the Strait of the Columns (the Straits
of Gibraltar) and describes in order the countries lying along
the south shore of the Mediterranean — Mauretania, Numidia,
Africa proper, and the Cyrenaica, to Egypt; then in like
manner the portions of Asia adjoining the Mediterranean, the
j3Egean and the Euxine, from the confines of Arabia to the
Tanais; and thence returns along the north shores of the
Euxine and the Mediterranean, describing European Scythia,
Thrace, Macedonia, Greece, Italy, and the southern portions
of Gaul and Spain ; thus returning to the point from whence
he set out. He next gives an account of all the islands within
this inner sea, including not only the great and important
ones, such as Sicily, Sardinia, Crete, &c. — but mere rocks
adjacent to headlands, like the Symplegades or Chelidonise.
Lastly he proceeds to make the circuit of the continents
following the shores, or supposed shores, of the external ocean,
but this time in an inverse order to the preceding, beginning

' " Id omne, qua venit, quaque dis-
pergitur uno vocabulo Nostrum mare
dicitur." (i. § 6.) He does not even em-
ploy in any case the expression of
" Internum Mare," ■wliicli is occasion-
ally found in Pliny, though hardly

used as a proper name. Tlie now
familiar appellation of Mediterranean
is in like manner first used by Solinus,
only as a convenient designation, not
as a strictly geographical name (Solin.
c. 23, § 14).

Sect. 2. POMrONIUS MELA. 355

with the Athintic coasts of Spain, then those of Gaul, Germany
and Sarmatia, and so round the northern parts of Asiatic
Scythia to the eastern extremity of Asia, and the confines of
India. Here he again pauses to describe the islands found in
this external ocean, beginning with Gades, and including
Britain and Ireland (which he calls Juverna) and Thule ; then
he returns to the extreme east, and describes India and
Arabia, the Persian and Arabian Gulfs, the Ethiopians, and
the western coast of Africa, which he regarded, as all his pre-
decessors had done, as extending direct from the land of the
Ethiopians to the north-western angle of the continent adjoin-
ing the Strait of the Columns.

The defects of such a system are obvious. Spain and Gaul
are each divided into two separate portions, described in
different parts of the work ; while the interior of Germany,
and the Alpine and Danubian provinces — Ehsetia, Vindelicia,
Noricum, and Pannonia, find no place at all ; Dacia is equally
unnoticed, the very name of the Dacians, so familiar to the
Eomans in the first century, as well as in later times, being
actually not mentioned by Mela ! The same thing is the case
with the interior of Asia, where the Medians, Bactrians and
Arians are merely mentioned by name in the preliminary
enumeration of the nations of Asia, and find no place in the
subsequent description ; while no notice whatever is taken of
countries like Drangiana, Margiana and Sogdiana, the names of
which had been so long well known to the Greek geographers.

§ 3. It is evident that Mela intended his work as a popular
compendium of geography, rather than an introduction for the
use of the student; hence he not only dismisses the whole
subject of mathematical geography with the very few words
to which we have already adverted, but he nowhere enters into
questions of measurements and distances, contenting himself
with describing as well as he can the general form and position
of countries, their boundaries and leading natural features, as
well as their physical character and climate ; adding more-
over, in regard to all those nations that were likely to be little

2 A 2

356 HISTORY OP ANCIENT GEOGRAPHY. Chap. XXIII.

known to his readers, a brief account of their manners and
customs and other national peculiarities. It is these notices,
concise and summary as they necessarily are, that constitute
the chief interest of his little work. They must not however
be received as representing in all cases the condition of the
different nations described, as they existed in the days of
Mela. There can be no doubt on the contrary that they are
taken, in most, if not in all, instances from earlier writers, and
though he gives us no hint of his authorities, we cannot fail
to recognize that many of them are derived directly from
Herodotus, and therefore represent in reality the state of
things that existed nearly five centuries before the time at
which our author wrote. This is especially the case with the
long description of the manners of the Scythian tribes inhabit-
ing the regions north of the Euxine ; almost the whole of
which is taken without alteration from the ancient historian.^
In some respects indeed Mela is actually in arrear of Hero-
dotus ; as he not only relates without question the fables of
the Arimaspians and the griffins, the Indian ants, and the
winged serpents on the borders of Arabia and Egypt, but
accepts as an undoubted fact the existence of the Ehipaean
mountains, and the Hyperboreans beyond them, on the shores
of the Northern Ocean.^ He tells us also that the Tanais,
which was described by Herodotus as rising in a lake, had its
sources in the Rhipsean mountains, and flowed down from
them with so rapid a stream that it was never frozen even in
the hardest winter, when the Meeotis and Bosphorus were a
mass of ice ; ^ a strange fiction, which is not found in any other
geographical writer.

§ 4. Mela indeed cannot claim the merit of having exercised
much critical judgement. He has repeated without scruple all
the usual fables concerning the Amazons, the Hyperboreans,
the Blemmyes in Africa without heads, and the goatfooted
^gipanes; while with regard to the Nile, after stating the

ii. 1, §§ 1-7, S-15. » ii. 1 ; iii. 5, §§ 3G, 37. ' i. 19, § 115.

Sect. 2.

POMPONIUS MELA.

357

various theories that had been proposed to account for its
periodical inundations, he seems disposed to acquiesce in the
strange suggestion that it had its origin in the southern hemi-
sphere, or land of the Antichthones, and flowed from thence
in a hidden channel under the sea, till it emerged again in
Ethiopia! Its being flooded in summer would thus be ac-
counted for, as that was the winter season in the part of the
world where it took its rise.^ Yet in a later passage of his
work ^ he inclines to the opinion of those who supposed the
Nile to have its source near Mauretania, among the Western
Ethiopians.

It is more strange that, notwithstanding the progress of the
Eoman arms in Germany and the neighbouring countries, his
ideas concerning the Danube were almost as confused and
erroneous as those of the earlier Greek geographers. While
he correctly states that it had its sources in Germany, and
was called Danubias in the upper part of its course, and Ister
where it became known to the Greeks, he gives no particulars
as to the former, nor does he name any of its great tributaries,
contenting himself with saying that it flowed for an immense
distance, and traversed great nations.^ In another place he
tells us distinctly that the Ister (which he takes care again to
identify with the Danube) flows through Istria into the Adriatic
Sea ; ^ and even adds that its stream, like that of the Padus
from the other side, poured itself with such impetuosity into
the sea that each river retained its course unbroken, until their
waters met, and were checked by their mutual action.®

§ 5. It is unnecessary to follow in detail the brief notices
that he has given of the various countries which he describes ;
it will suflice to point out the few instances in which his little

M. 9, § 54.

3 iii. 9, §§ 96, 97.

* " Per immania magiiarum gentium
diu Danubius est." ii. 2, § 8.

' ii. 3, § 57.

" ii. 4, ij 63, This absurd tale,
strange as it may seem, appears to Lave
been generally believed, and was re-

lated among others by Cornelius Nepos,
from whom perhaps Mela derived it ;
though, as Pliny justly points out, he
ought to have had better information,
as coming from the banks of the Padus
(plerique dixere falso, et Nepos etiam
Padi adcola. Plm. H. N. iii. 18, § 127).

358

HISTORY OF ANCIENT GEOGEAPHY, Chap. XXIII.

work may be considered as really showing any advance in
geographical knowledge, or adding any facts of interest to
what may be derived from earlier writers. The position of his
birthplace naturally made him well acquainted with the Straits
which derived their name from the Columns of Hercules, con-
cerning which — often as they had been visited and described
— there was great discrepancy among Greek writers.^ He
correctly points out that the real Columns {i. e. those to which
the name had been originally given) were the two lofty moun-
tains, Calpe and Abyla, the one on the European, the other on
the African coast, which rise like pillars on each side of the
Strait ; both of them projecting considerably into the sea, but
the former much the most, so as to be almost isolated.® The
narrow sea however, or the Straits in the wider acceptation of
the term, extended as far as the promontory of Juno (Cape
Trafalgar) on the European side, and that of Ampelusia, — the
same that was called by Strabo Cotes, the modern Cape Spartel
— on the African.^

With the geography of Western Europe he appears to have
been in general better acquainted than any of his Greek pre-
decessors, and his notions concerning Spain and Gaul in par-
ticular show a considerable improvement in his conception of
their figure and position, as compared even to those of Strabo.
Thus he was well aware that the western coast of Gaul, after
preserving at first a nearly straight course northwards as far
as the mouth of the Garonne, afterwards began to trend to the
west, and project so far in that direction as to be opposite to
the northern or Cantabrian shores of Spain, leaving between
them an extensive bay (the Bay of Biscay), for which he has
however no name.^ This important feature in the geography

' See the different statements given
by Strabo, iii. 5, § 5, p. 170.

* In regard to Calpe (the rock of
Gibraltar) lie notices particularly the
extensive caves by which it is almost
perforated.

His statement that the Strait in
its narrowest part was only 10 miles

wide, is almost . precisely con-ect; the
width between Tarifa and Alcazar
Point being (according to Admiral
Smith) 91 G. miles, while between
Gibraltar and Ceuta it is 12 miles
(Smyth's 'Mediterranean, p. 159).

" ii. 6, § 9G.

1 iii. 2,"§ 23.

Sect. 2.

POMPONIUS MELA.

359

of Western Europe had been, as we have seen, entirely mis-
conceived by the Greek geographers; Eratosthenes having
given an undue extension to the Graulish or Armorican pro-
montory, while he almost ignored the projection of the Spanish
coast to the north-west ; and Strabo on the other hand was
either entirely ignorant of the Armorican promontory, or
altogether underrated its importance. It was natural that the
Eoman occupation of both provinces should have led to clearer
ideas on the subject.

His description of the north-western angle of Spain (the
modern Galicia) is unusually full and detailed, and shows an
accurate and minute acquaintance with this remote corner of
his native country which is rather surprising. It may be
observed that he always terms the headland of Cape Finisterre,
which was known to the Greeks as the promontory of Nerium,
only the Celtic promontory, and that he distinctly designates
the tribes nearest to it, the Nerii and Artabri, as well as those
inhabiting the west coast, as far as the Douro, as Celtic tribes.
The Astyres and Cantabri, who were undoubtedly of pure
Iberian origin, he regards as distinct.^ It is worthy of notice
also that he had a clear conception of the true' character of
the Pyrenean chain and states that it extended at first direct
from the Mediterranean across to the Ocean, and then turning
inland into the interior of Spain, continued with an unbroken
course till it reached the western shores of that country facing
the Atlantic.^ This view is perfectly correct, the mountains
of Guipuzcoa, Biscay, Asturias and Galicia, being in reality
only a prolongation of the chain of the Pyrenees, though that
name is commonly applied only to the direct chain, which
runs across from sea to sea, and forms the boundary between
France and Spain.

§ 6. In regard to Gaul, his expression that it was divided

2 iii. 1, §§ 10, 13.

3 ii. 5, § 85. It is remarkable that
he here applies the epithet of British
(Britannicus) to the part of the Oceau

north of Spain ; and in another passage
(i. 2, § 15) describes Europe as bounded
by the Atlantic on the west, and by
the British Ocean on the north.

36o

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXIII.

into two parts by the Lake Lemannus and the mountains of
the Cevennes, is not inappropriate, if we conceive a line drawn
from one to the other ; and this limit, which nearly coincides
with that of the ancient Roman province, he takes, appro-
priately enough, as separating the part of the country which
faces the Mediterranean from that which fronts the Ocean.
On these Atlantic shores he gives a full and characteristic
description of the remarkable estuary formed by the Garonne
near its mouth (the Gironde), which must have appeared the
more peculiar to persons unaccustomed to tidal rivers ; * and
notices at considerable length a small island called Sena,
opposite to the land of the Osismii, in the British Sea, which
was the site of a celebrated oracle, consulted by Gaulish navi-
gators, and served by nine virgin priestesses.^ The peculiar
sanctity of this locality is not mentioned by any other writer,
but the fact that the name is still retained by the little islet
of Sein, off the extreme western coast of Bretagne, bears
testimony to the accuracy of Mela's geographical information.
He appears indeed to have taken some pains to inform himself
on all questions connected with the religious creed and rites
of the Gauls ; and in another passage has given us an account
of the Druids and their tenets, which is one of the most in-
teresting notices we possess on this obscure subject.*^

Of Britain, as we have seen, he announces that the world
was on the point of receiving for the first time full and authen-
tic information; but in the mean time he contents himself
with giving a few particulars concerning its natural produc-
tions and the manners of its inhabitants, which generally accord
with those given by Csesar and Strabo. With regard to its
geographical position he tells us that it extended between the
north and west (i. e. in a direction from N.E. to S.W.), and was

* iii. 2, § 21.

* iii. 6, § 48.

" iii. 2, § 19. It may be remarked
also that he mentions the Cassiterides
in connection witli Gaul, instead of
Spain, to which they were generally

referred (" in Cdticis aliquot sunt, quas
quia plumbo abundant uno omnes no-
mine Oassiteridas appellant," lb. G, §
47). Of any connection with Britain he
had evidently no suspicion.

Sect. 2.

POMPONIUS MELA.

361

of a triangular form, like Sicily, having its main angle oppo-
site to the mouths of the Rhine, and its two sides, receding
from this, facing respectively Graul and Germany ; while the
third side at the back (as he terms it) must have faced the
Ocean to the north-west.' Beyond this lay Ireland, the name
of which he writes Juverna, and describes it as nearly equal in
extent to Britain, but of the form of an oblong parallelogram ;
and tells us that its climate was ill adapted for the growth of
corn, but its pastures were so luxuriant that the cattle were
obliged to be watched, lest they should burst themselves with
over-repletion. But the inhabitants he says were quite un-
civilized, and devoid of all the virtues of other nations.®

Imperfect as is this account, it is clear that Mela had a
better idea of the position of the British Islands than Strabo,
and did not place Ireland so far to the north. He is also the
first extant writer who mentions the Orcades, which he correctly
describes as a group of thirty islands near together. They had
been discovered, as we have seen, in the reign of Claudius,
and Mela may have had authentic information concerning
them.^ Of Thule on the contrary he evidently knew nothing
beyond what he learned from the Greek writers, and merely
repeats the customary story of the short summer nights, and
that at the summer solstice there was no night at all, the sun
being always visible.

§ 7. His knowledge of Germany was evidently very im-
perfect ; but his account of it is rendered still more so, from
the form and arrangement of his work affording him hardly
any opportunity of describing the interior of that great
country.-^ But with regard to its northern shores he appears

' iii. 6, §§ 49-52.

« Ibid. § 53.

^ Ibid. § 54. He mentions also the
Hsemodse, seven in number, which
must probably be a perversion of the
name of the Hsebudes, as they are
called by Pliny {H. N. iv. 16, § 103), but
he strangely transfers them to the side
opposite to Germany (septem Hsemo-
dse contra Germaniam vectje).

' By a strange oversight, or deviation
from established usage, he extends its
southern limit to the Alps, thus in-
cluding all Vindelicia, Ehsetia, and
Noricum, countries which were cer-
tainly never regarded by the Eomans
in general as comprised in Germany.
Even the names of these provinces are
not mentioned by Mela. It is perhaps
connected with this error that he de-

362

HISTORY OP ANCIENT GEOGEAPHY.

Chap. XXIII.

to have had somewhat more information than his predecessors,
though still in so vague and imperfect a form, that it is diffi-
cult to judge to what it actually amounted. Thus, after
enumerating the well-known rivers which flowed into the
German Ocean — the Ems, the Weser and the Elbe — he tells
us that beyond the Elbe there is a very large bay, called
Codanus, full of islands, large and small, of which the largest
and most fertile was one called Codanovia.^ This was in-
habited by the Teutoni, who also, with the Cimbri, occupied
the neighbouring mainland.^ He seems also (though the
passage is corrupt and very obscure) to have had some notion
of the Cimbrian Chersonese, and the manner in which the sea
formed narrow straits between the projecting continent and
the neighbouring islands.* There can be no doubt that by the
Codanus Sinus he meant the southern portion of the Baltic,
which he of course regarded merely as a bay of the JSTorthern
Ocean ; and we cannot hesitate to recognize in the large island
of Codanovia the same which is mentioned by Pliny under the
now familiar appellation of Scandinavia.

The easternmost people of Germany, according to Mela,
were the Hermiones, whom he places on the northern ocean.
They were separated from the Sarmatians by the Vistula,
which formed the boundary between Germany and Sarmatia.^
Of the countries east of that limit he had evidently no real
knowledge. He gives indeed a full account of the manners
and aspect of the Sarmatians, whom he describes as resembling
the Parthians in their habits and mode of life : but does not

scribes the Danube as having its sources
near those of the Ehone and the Ehine
(ii. 2, § 79), so that he apparently sup-
posed them all three to rise in the
Alps.

2 iii. §§ 31, 54. This is the read-
ing of all the best MSS. ; others have
Codanonia. Some of the recent editors
have altered it into Scandinovia, in
order to approximate to the form found
in Pliny, but there is no authoiity for
this, and the name Codanovia certainly
appears to be connected by Mela with

the Codanus Sinus.

^ iii. 3, §i? 31, 32 ; 6, § 54.

* Ibid. § 31.

' Ultimi Germanise Hermiones, § 32.
As this is immediately followed by the
mention of Sarmatia and the Vistula,
there appears no doubt that the mean-
ing of Mela is that stated in the text,
but the words as they stand (" Sarmatia
intus quam ad mare latior, ab hix qux
seqtmntur Vistula amne discreta ") are
unintelligible, or would convoy a wholly
different meaning.

Seot. 2.

POMPONIUS MELA.

363

attempt to define their geographical limits in any direction,
and proceeds at once from thence to the Asiatic Scythians,
the first of whom he tells us were the Hyperboreans dwelling
beyond the Ehipasan mountains to the north. In their country
the sun rose at the vernal equinox, and set at the autumnal, so
that they had six months of day and six months of night. He
adds the usual fables concerning their happy and virtuous life,
their longevity, and the intercourse they had long maintained
with Delos.^

§ 8. Passing from this region of fable he gives a pretty full
account of the Caspian Sea and the nations that surround it,
the names of which are correctly given, and. are known from
other sources.'^ But it is remarkable that while he adheres to
the belief universally adopted in his day, of its communicating
with the northern ocean, he distinctly states that it was joined
to it only by a long and narrow strait like a river f an expres-
sion that seems to point clearly to an increased knowledge of
these regions, which would soon lead to the discovery that the
supposed inlet from the north was in reality nothing but a
river. It is strange that while the Oxus and laxartes on one
side and the Tanais on the other, had been so long familiar
both to Greeks and Romans, no notion of the great river
Volga had yet reached their ears.^

Eastward of the Caspian he himself tells us that there was
again an unknown region : and that it had long been con-
sidered uncertain whether there was sea, or continuous land

« iii. §§ 36, 37. This account of the
Hyperboreans almost exactly agrees
with that of Pliny (H. N. iv. 12, §§
89-91). Both were doubtless taken
from the same Greek authors, and pro-
bably derived, whether directly or indi-
rectly, from Hecatseus of Abdera, who
had written a special work upon the
subject.

^ On this occasion he gives (§ 43) a
long account of the Hyrcaniau tigers,
which evidently still eujoyed a great
reputation among the Romans. The
Indian tigers were probably still so

rare as to be almost unknown. (See
Chapter XX. Note 0, p. 201.)

* " Mare Caspium vit angusto, ita
longo etiam freto primum terras quasi
fiuvius irrumpit." iii. § 38.

9 The name of the Eha, under which
the Volga is mentioned by Ptolemy, is
indeed found in the ordinary editions
of Pomponius Mela (iii. 5, § 39), but it
is a mere conjecture, introduced into
the text by Pintianus, for which there
is no authority, and ha^ been justly re-
jected by the recent editors, Tzschucke
and Parthey.

364

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIII.

extending without limit, but uniniiabitable on account of the
cold. But he appeals to the story told by Cornelius Nepos of
the Indians who had been driven by sea from their native
shores to those of Germany, as decisive of the question ; as
indeed it would have been, had it had any foundation of
truth.i

His account of the eastern extremity of Asia is peculiar :
and though probably taken from earlier Greek writers con-
tains some particulars not found in any other extant authority.
He describes the coast of Asia as tending eastward from the
Scythian Promontory — a name by which he apparently means
to designate the north point of Scythia, east of the opening of
the Caspian — to the shores that faced the east, where he places
a mountain promontory, which he calls Tabis.^ This is appa-
rently the same that is called Tamarus by Eratosthenes, and
was regarded by him as the eastern extremity of the great
ridge of Mount Taurus, which traverses Asia from thence in
its whole extent. Mela, however, in another passage gives the
name of Tamus to a headland, which he appears to have re-
garded as distinct from Tabis, though the two were probably
identical : and adds that off it lay the island of Chryse, or
the Golden Island, while that of Argyre, the Silver Island,
was opposite to the mouths of the Ganges.^ Both these were
probably mere fictions : * but it is remarkable that to the south
of Tabis, between that headland and India, he places the Seres,

' After referring to the opinions of
the ancient philosophers and Homer (!),
lie adds : " Cornelius Nepos ut receutior,
anctoritate sic certior; testem autem
rci Quintum Metellum Celerem adjicit,
eumque ita retulisse comniemorat : cum
Galliaj pro consule prseesset, Indos
quosdam a rege Boiorum (?) dono sibi
datos ; unde in eas terras devenissent
roquirendo cognossc, vi tempestatum ex
Indicia sequoribus abreptos, emensosquo
quae intererant, tandem in Germanise
littora exisse." iii. § 45.

^ Mela is the first, so far as we know,
to introduce into this part of Asia, be-
tween the Scythians and the Eastern

Ocean, a nation of Anthropophagi, the
fear of whom contributed, together with
the cold and the number of wild beasts,
to keep other nations at a distance, and
render the whole tract a solitude (iii.
§ 59) : a statement repeated by Pliny
and by all the subsequent compilers.

3 iii. §§ 68, 70.

* They here make their appearance
for the first time ; but we shall find
them continually reappearing in the
works of later geographers, both Greek
and Roman, who endeavoured, with
little success, to find a place for them,
as these rctjious became better known.

Sect. 2.

POMPONIUS MELA.

365

" a nation full of justice, and well known for the silent com-
merce that they carry on in their absence with goods left
in the desert."^ It is singular that he makes no mention
of the object of this commerce, as Pliny does in a parallel
passage : but it is interesting to find that the Eomans had
by this time at least learnt, though in a vague way, that the
Seres occupied a position in the far east of Asia.®

§ 9. Of India itself his knowledge is remarkably vague and
imperfect. Though the Eomans, as we have seen, were at
this period rapidly extending their commercial relations with
that country, and in consequence acquiring increased know-
ledge of its shores, Mela has not only given no sign of any
such recent information, but his geographical statements are
so confused and erroneous as to be in great part unintelligible ;
and it is clear from the summary manner in which he dismisses
this part of his subject, that his views were very far from dis-
tinct. He gives us indeed a brief summary of the current
stories concerning the natural productions of the country — the
gold-seeking ants, the trees distilling honey, the wool-growing
woods, &c. — as well as the manners and customs of its in-
habitants : all derived from the ordinary Greek authorities :
but his general description of its geography is confined to a
short account of the Ganges and Indus, and the somewhat
obscurely worded indication that the promontory of Colis or
CoUis, between the two, was the angle where the coast turned
from the eastern to the southern sea : and therefore formed
the south-eastern angle of Asia.'' The whole extent of the
shores of India, he tells us, was a voyage of sixty days and
nights.^ In regard to Taprobane he seems disposed to adopt

^ iii. § 60. " Seres intersunt, genus
plenum justitise, et commercio quod
rebus in solitudine relictis absens per-
agit notissimum." This is evidently
the same tradition mentioned by Pliny
on the authority of the envoys from
Taprobane (vi. 22, § 88).

° Mela had already stated at the
outset of his treatise that the Indians,

/Seres and Scythians, were the most
easterly nations known, and adds that
the Seres were intermediate between
the other two. (" Primes hominum ab
oriente accepimus Indos et Seras et
Scythas. Seres media ferme Eose par-
tis incolunt, Indi ultima," i. 1, § 11.)

^ iii. §§ 68, 69.

8 Ibid. § 61.

366 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIII.

tlie paradoxical opinion of Hipparchus, that it was not merely
a large island, but the beginning of another world.^

§ 10. Of the remainder of Asia, from India to the Red Sea,
his account is very brief, and by no means clear : but he had
a distinct idea of the conformation of the coast, as forming a
great bay, from which the two deep inlets of the Persian and
Arabian Gulfs penetrated far into the interior of the continent.
But his few notices of the nations that adjoined their shores
are so imperfect and confused that he applies to the Car-
manians the accounts given by other writers of the barbarous
tribes of the Ichthyophagi on the coast of Gedrosia, and places
the Gedrosians between them and the Persians.^ It is a curious
effect of the peculiar arrangement of his work that no place is
found for the description of Persia, Media, or the other
nations of Upper Asia, or even for Mesopotamia and Assyria,
though he takes occasion in treating of the Persian Gulf to
give a brief account of the Tigris and Euphrates. It is re-
markable that he describes the latter river as no longer reach-
ing the sea by an independent channel of its own, but gradually
dwindling away and losing itself.^ It would appear therefore
that at this time the great mass of its waters was already, as at
the present day, poured into the Tigris, while the rest was
absorbed in the sands, or lost in the marshes.

§ 11. His account of Arabia, properly so called, is very
concise, and shows no signs of acquaintance with any new
sources of information, while that of the Arabian shores of
the Eed Sea — under which head he includes, in common with
Strabo and other writers, the whole tract along its ivestern
shores, between the Sea and the Nile — is filled with fables
concerning the Pygmies, the winged serpents and the phoenix.
It is remarkable that we here again, as in the case of the
Scythians, find him copying to a great extent from Herodotus,
from whom he has also derived the fabulous account of the
Ethiopians, their profusion of gold, their marvellous longevity,

" Ibid. § 70. ' iii. 8, § 75. ^ Ibid. § 77.

Sect. 2.

POMPONIUS MELA.

367

and the Table of the Sun.^ From the same source are taken
the few particulars that he has given us with regard to the
nations of the interior of Libya, the Garamantes, Atlantes,
Augilse, &c. : though in regard to these there are some con-
fusions and changes of name that would appear to indicate
their being derived from some intermediate author, and not
directly from Herodotus.* In any case it is a sufficient proof
of the uncritical character of the work in question, that our
author has taken his materials from so early an authority,
without any indication of their character or origin. It may
be said indeed with regard to the treatise of Mela in general,
that, with the exception of the countries immediately border-
ing on the Mediterranean, it was rather calculated to supply
to its readers a compendious collection of the stories current
with regard to different countries and their inhabitants, than
to furnish them with any correct geographical information.

§ 12. Nowhere is this more apparent than in regard to the
external coast of Africa, the description of which forms the
concluding section of his work. Here indeed he had the ad-
vantage of being better informed with respect to the voyage of
Hanno, than either Strabo or Pliny, and certainly had either
seen the original narrative, or some authentic abstract of it.
He correctly tells us that Hanno, after having circumnavi-
gated a great part of Africa, setting out from the Western
Straits, turned back, not from any difficulty of navigation, but
from want of provisions.^ On the other hand he accepts
from Cornelius Nepos the erroneous statement that Eudoxus,
setting out from the Arabian Gulf, had completed the voyage
from thence to Gades ; and rests upon this fact the conclusion
that Africa was really surrounded by the Ocean.^ He follows
up these statements by a number of notices, more or less mar-
vellous; some of them — such as that of the wild and hairy

3 iii. 9, §§ 85-88. See Chapter VIII.
p. 271.

" i. 4, § 23.

•^ iii. 9, § 90. " Hanno Carthagini-
ensia exploratum missus a suis, cum

per oceani ostium exisset magnam ejus
(Africse) partem circumvectus, non se
mari sed commeatu defecisse memoratu
retulerat,"
« Ibid.

368

HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXIII.

women, the burning mountain called Theon Ochema, and the
Promontory of the Western Horn — certainly taken from
Hanno; but mixed up with tales of a purely fabulous cha-
racter, of races of men without tongues, others without nostrils,
others with legs too much bent to walk, and so on. It is not
till he approaches the confines of Mauretania that he returns
to anything like geographical accuracy ; and here we find
him, like Strabo, first mentioning the Pharusii and Nigritas,
then the Gsetulians, whose shores abounded with purple of the
finest quality.'' He notices the Fortunate Islands only in a
general way, without enumerating their names, or stating their
number : and describes Mount Atlas in connection with this
western coast, in a manner that clearly shows him to refer to
the part of the mountain-chain that approaches the Atlantic,
rather than to that more familiar to the Eomans in the north
of Mauretania.^

^ Factories for the collection and
manufacture of this purple had, as we
have seen, been established by Juba ;
but no reference to his authority is
found in Mela, nor are there any state-
ments obviously derived from his work.

* iii. 10, § 101. This is the more re-
markable as his native place was di-
rectly opposite to the northern arm of
the Atlas, where it abuts on the Straits,
and forms so conspicuous an object to
all those that pass through them.

MAP OF THE WORLD ACCORDING TO POMPONIUS MELA.

Stan/crrds GeograpMcdlSsTab.

Loudon ; Jolm. Murray.

Notes A, B. POMPONIUS MELA. 369

NOTE A, p. 339.

BRITISH PEARLS.

Tacitus says with reference to them " Gignit et Oceanus marga-
rita, sed suflfusca et liventia :" and adds that their inferiority was
said by some to he owing to a want of skill in collecting them.
(Agricola, c. 12.) Pliny also says : " In Britannia parvos atque
decolores nasci certum est" (Hist. Nat. ix. 35, § 116), and mentions
that Julius Csesar had consecrated in the temple of Venus Genitrix
at Rome a corslet adorned with British pearls. Suetonius even
asserts that Caesar was induced to invade Britain for the sake of its
pearls (" Britanniam petisse spe margaritarum," Cces. c. 47), an idle
story, which may however serve to show that great expectations
were formed of them. It is well known that pearls of inferior
quality are found at the present day in the rivers of Wales and
Scotland : but they are the production of river mussels (the Unio
margaritiferus of naturalists), not of the true pearl-bearing oyster.
It was therefore an error, though a very natural one, to suppose
that they were produced in the Ocean, like the Indian and Oriental
pearls. Pomponius Mela, singularly enough, was better informed,
and expressly states that the British pearls were produced in rivers.
" Fert (Britannia) . . . preegrandia flumina, alternis motibus modo
in pelagus modo retro fluentia, et qusedam gemmas margaritasque
generantia " (iii. 6, § 51).

NOTE B, p. 340.

THE DUMNONII.

It is in any case a remarkable circumstance that the Dumnonii,
whom we find in the time of Ptolemy occupying the whole of the
south-western extremity of Britain, including both Devonshire and
Cornwall (Ptol. Geogr. ii. 3, § 30), and who must therefore have
been one of the most powerful nations in the island, are never once
mentioned in the history of the conquest of the country by the
Romans ; nor is their name found in any writer before Ptolemy.
Their name is also found in Solinus, c. 22, but in a passage of
which both the readiiig and the sense are alike obscure : and the

VOL. II. 2 B

370 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXIII.

word Dumnonii is in fact only a correction of the editors, though
a plausible one. The conjecture of Mr. Beale Poste (Britannic
Besearches, p. 332), that they were left in nominal independence
under a native king, who continued faithful to the Eoman alliance,
though wholly without authority, appears to me highly probable.
In like manner we find them for a time leaving Sussex and some
adjoining parts of England under a king named Cogidumnus or
Cogidubnus (Tacit. Agric. 14). His name is found in an inscrip-
tion discovered at Chichester, and given by Horsley (Britannia
Bomana, p. 332), and by Hixbner (Inscr. Britann. p. 18), and the
same thing was done in the first instance with the Iceni. It was
only the tribes who opposed the Eomans in arms that find a place
in history. It is worthy of notice also that no Eoman inscriptions
have been found either in Devonshire or Cornwall. (See Hiibner,
p. 13.)

It is still more remarkable that no allusion is found in any of
the Eoman writers of this period to the tin of Cornwall, as one of
the productions of the island. In the passage already cited from
the Agricola of Tacitus, where he is enumerating the valuable
products of Britain, which made it worth the conquest (pretium
victorise), while the insignificant pearls are mentioned, no notice is
taken of the far more important article of tin, except as comprised
in the vague and general expression " alia metalla."

NOTE 0, p. 353.

THE AGE OF MELA.

All modern editors and writers on geography acquiesce in this
conclusion. Some of the earlier editors supposed the expedition of
Caligula to be meant, but that abortive attempt was of too brief
duration to render it probable that it was the one referred to. The
earlier date adopted by some scholars, who supposed the expedition
of Julius Caasar to be the one referred to, is excluded by many
statements in the work of Mela, such as the name of Cassarea given
to lol, which it first received from Juba ; the division of Spain into
three provinces, first introduced by Augustus, &c. But the very
epithet applied to the conqueror of "principum maximus" would
never have been employed by any writer before the Augustan age.

( 371 )

CHAPTER XXIV.

PLINY.

Section 1. — General Vieios.

§ 1. Fak more important than tlie abridgement of Pomponius
Mela was the geographical treatise included in the compre-
hensive work of the elder Pliny, to which he gave the name
of Natural History, or as it would be more correctly translated,
a History of Nature. In this great work, — for, with all its
defects, it fully deserves that epithet — he attempted to give a
general view of all that was known in his day of the physical
constitution of the universe, and of this world in particular, as
well as of all its productions, whether animal, vegetable, or
mineral. Hence he himself speaks of it as including all those
subjects, which were comprised by the Greeks under the name
of kyKVKXoTralSeLa ; ^ in modern phraseology, it comprehended
at once physical philosophy and natural history in all their
branches. The portions which relate to the fine arts, painting
and sculpture, which are at the present day among the most
valuable parts of the work, though occupying the greater part
of three books, are mere digressions, having very little ^con-
nection with its general plan and purpose.

In this instance we have, by a rare piece of good fortune,
full information not only concerning the life of the author,
and the exact date of the publication of his work, but we
possess unusually full particulars concerning his mode of

' Prasfat. § 14. We have here the
first application of the term whicli is
so familiar to us in modern times of an
Encyclopasdia. It is strange that the
compilers of Greek Lexicons all reject
the word as a barbarous compoimd,

without adverting to its use by Pliny,
who moreover distinctly employs it as
one already familiar ("jam omnia attin-
genda, quas Grseci rrjs eyKVKXoiraLSeias
vocant"),

2 B 2

372 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXIV.

study and the manner in whicli he amassed the materials for
so vast an undertaking, which it is very important to bear in
mind in estimating the value of the results transmitted to us.

§ 2. Caius Plinius Secundus, commonly known as Pliny
the Elder, to distinguish him from his nephew of the same
name, was born in a.d. 23, either at Verona or Comum in the
north of Italy — it is uncertain which, — and filled various
public offices, among others that of procurator in Spain, which
he held during the last years of the reign of Nero. After the
accession of Vespasian he became the intimate friend of that
emperor, as well as of his son Titus, to whom he dedicated his
great work. Among other employments he was appointed by
Vespasian to the command of the Koman fleet at Misenum,
and was stationed there in a.d. 79 when the great eruption of
Vesuvius took place, which overwhelmed Herculaneum and
Pompeii, and in which he lost his life, under the circumstances
related by his nephew in a well-known letter.^ We learn from
the same authority that besides the work to which he owes his
fame, Pliny had already composed several other literary works,
some of them of a voluminous character ; the most important
of which were, a History of the Wars in Germany, in twenty
books, the loss of which is much to be regretted in a geo-
graphical as well as historical point of view ; and a History of
his own Times, in continuation of the work of Aufidius Bassus,
which extended to thirty-one books, and included apparently
a portion at least of the reign of Vespasian.^ Both these
works appear to have enjoyed considerable reputation, and to
have been frequently used by later writers, though they were
in great measure eclipsed by those of Tacitus, who possessed
those qualities of a truly great historian which were certainly
wanting in Pliny.

2 Plin. Epist. vi. 16. For fuller par-
ticulars concerning the life of Pliny,
see the article Plinius in Dr. Smith's
Diet, of Ihography ; and the intro-
duction to Urlich's Ghrestomatliia
Pliniana, Berlin, 1857.

^ Plin. Epist. iii. 5. The latter work
is alluded to by himself in the preface
to his Nahirai Bistory (§§ 19, 20). It
appears that it was then completed,
hut had not yet been published.

Sect. 1.

PLINY: GENERAL VIEWS.

373

It was not therefore till a comparatively advanced period of
life that he devoted himself in earnest to the composition of
his Natural History ; * but he had long before been occupied
with preparing the materials for it. It must have been mainly
with a view to this object that he had accumulated the mass
of notes and extracts from his multifarious reading of which
his nephew has given us so lively a picture, and to which he
himself refers in the elaborate preface which he has prefixed
to his work. He tells us himself that he had read about
2000 volumes, out of which he had collected 20,000 facts
worthy of notice; and boasts that almost all this had been
accomplished at spare hours, especially at night.^ He was in
the habit of reading or having books read at his meals, in the
bath, while travelling, or taking exercise,^ — in short, at every
available moment ; and always making notes or extracts from
all he read." Such a miscellaneous farrago of materials would
obviously require a sound critical spirit to distinguish the
valuable from the worthless, and a highly scientific turn of
mind to co-ordinate this mass of facts into any clear and lucid
arrangement. Unfortunately Pliny was almost wholly desti-
tute either of the one quality or the other. He himself boasts
with some reason that he has undertaken a task, which as a
whole had not been attempted by any previous writer, either
Greek or Koman ; ^ and he has been justly praised by Humboldt
for the grandeur of the conception that he had formed, in this
first essay towards a physical description of the Universe.^
But the same author admits how very far the execution of his
work fell short of the original idea, not only from defective
arrangement and want of method, — " the elements of a general
knowledge of nature lying scattered almost without order in

* His work was completed and pub-
lished in A.D. 77, only two years before
his death, as we learn from the preface,
§ 3, in which he dedicates it to Titus,
in his sixth consulship. How long
before it was actually commenced we
have no means of judging.

^ Prxfatio, § 17.

« Plin. J. Epist. ill. 5.

' Prxf. § 14. " Nemo apud nos, qui
idem tentaverit, nemo apud Grsecos qui
unus omnia ea tractaverit."

^ Humboldt's Cosmos, vol. ii. p. 195,
Entrl. transl.

374

HISTORY OF ANCIENT GEOGRAPHY, Chap. XXIV.

his great work" — but from tlie want also of that scientific
insight into his subject without which it was impossible to
weave his accumulated mass of materials into an organized or
harmonious whole.^ His voluminous treatise remains in con-
sequence a vast compilation, bearing testimony to the un-
wearied diligence of its author in the collection of his facts,
but showing at the same time an almost total want of critical
judgement or philosophical arrangement.^

§ 3. The great naturalist Cuvier has pronounced a strong
censure upon that part of Pliny's work which relates to what
is now commonly known as natural history; and has shown
how far inferior he was in this department to his great prede-
cessor Aristotle.^ The same remark may be applied with even
greater force to the geographical portions, which are perhaps
on the whole the most defective parts of the whole work.
When we compare them with the writings of Eratosthenes
and Strabo, we are struck with the almost total absence of any
scientific comprehension of his subject, or of those general
views which, however imperfectly developed, were certainly
present to the minds of the Greek geographers. Instead of
any geographical outlines of the general structure and com-
position of the continents, or of the several countries that
compose them, we find for the most part mere dry catalogues
of the names of cities, or tribes, rivers and mountains ; some-
times arranged with reference, more or less carefully observed,
to the lines of coast ; but generally, especially where the
interior of a country is concerned, enumerated in alphabetical
order, or jumbled together without any arrangement whatever.

The use of maps, as we have seen, was already familiar in

» Ibid. pp. 195-198.

^ How mucli he prided liimself upon
the mere accumulation of facts, without
reference to the scientific use made of
them, or the vahie of the authorities
from which they were derived is shown,
not only by tlie tone of self-complacency
with which he dwells in his Preface
on what he had accomplished in this

respect, but by the statement appended
to the summary of each book of the
number of such facts, or rather state-
ments (res et historic et observatioues)
which it contained.

2 Ciiviei in the Biograpliie Universelle,
art. Pline. See also the remarks of
Humboldt, I. c. p. 197.

Sect. 1. PLINY : GENERAL VIEWS. 375

the days of Pliny, and he had the advantage among others
of consulting that prepared by Agrippa on a large scale, to
which we have already adverted.^ It is obvious from internal
evidence that his enumeration of towns, headlands, bays and
other natural features of the coasts, was taken in many cases
from such authorities, and in these instances he often supplies
us with a para^lus of considerable value from the number of
names and details which it furnishes ; but where this guide is
wanting, we have generally no geographical indication what-
ever to point out the site of the places enumerated. In no
instance does he attempt to determine their position by refer-
ence to latitude and longitude, in the manner pointed out by
Hipparchus, and subsequently developed by Ptolemy. Nor
do we find him, except in a very few cases, making any use of
the great lines of Eoman highway, which being in his time
already extended to almost all parts of the Empire, might
have afforded to a geographer much assistance in explaining
the position of the towns and cities through which they
passed.*

Another grave defect is the want of chronological dis-
crimination in the use of his authorities. He makes use of the
earlier Greek writers, such as Eratosthenes or the historians of
Alexander, as if they stood on the same footing with recent or
contemporary authors ; and frequently mixes the two sets of
authorities together, without any attempt to distinguish them.
This is especially the case with regard to Asia, his account
of which is much like what would be produced by a modern
writer, who attempted to blend together the geography of
Marco Polo and Ibn Batuta with the results of the most recent

See Chapter XX. p. 177. I respect had (as we have seen) been

* There cannot be a stronger proof
of this than his omission of all no-
tice of the Egnatiau Way, which, be-
sides its own importance as the great
high-road between Europe and Asia,
was an invaluable assistance to the
geographer in regard to tlie confused
and difficult geogruphy of Macedonia
and Thrace. Its importance in this

fully recognised by Polyblus and Strabo
(see Chapter XYII. p. 27). It is even
more singular that where he gives the
actual distance fiom Dyrrhachium to
Byzantium (iv. § 46), which could only
have been measured along this road, he
greatly understates it, making it only
711 M. P., while the real distance was
754 miles {Itin. Ant. p. 317).

376 HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

English and Eussian researches, without any reference to the
different sources from which his statements proceeded.

§ 4. The great value of Pliny's work really lies in its im-
portant contribution to the political or statistical geography of
the countries that were in his time organized as provinces
under the Eoman Empire. We have already pointed out^
how much the extension of the imperial administration must
have tended to this end ; and the circumstance of Pliny having
himself filled important public offices, both at Eome and in
the provinces, must have secured him full access to official
documents, as well as drawn his attention to their value and
importance. It is unfortunate that, in availing himself of
these resources, he confined himself to the mere nomenclature
of geography, or to collecting scattered notices of individual
facts for his natural history : he never appears to have sought
to combine these into one organic whole, or to present such a
picture of a country, including its natural features, charac-
teristics and productions, as is essential to the politician or
historian, not less than the geographer. This deficiency is
apparent even with regard to those countries, with respect to
which he had the best means of information, such as Spain and
Gaul, of neither of which does he give us anything like a
general picture, or characteristic description, such as those
presented to us by Csesar and Strabo, any more than a clear
geographical outline.

Pliny himself indeed repeatedly apologizes for the hasty
manner in which he runs over his descriptions of countries, on
account of the necessity of brevity, and that he is hastening
on to the more essential parts of his subject. But these con-
siderations do not prevent him from filling page after page
with voluminous lists of obscure names, while he omits almost
entirely to point out the leading geographical features of each
country, or describe the natural characters that distinguish it.
It is still more remarkable that he scarcely attempts to give

See Chapter XX. p. 176.

Sect. 1.

PLINY : GENERAL VIEWS.

any account of the characteristics of the inhabitants of each
region, of their manners and customs, or even of their physical
peculiarities. Such notices, one would have thought, would
have found their place with peculiar appropriateness in a geo-
graphical treatise designed as an introduction to a general
History of Nature. But so completely has Pliny left aside
this important branch of his subject, that he is inferior in this
respect not only to the great work of Strabo, but even to the
summary compendium of Pomponius Mela. It is strange to
find an author who aspires to give a complete natural history
of the world ignoring altogether the natural history of Man,
and the distinctive peculiarities, whether physical or acquired,
of the different races that people the surface of the globe.^

§ 5. But if he thus entirely misconceived the nature of the
problem with which he had to deal, and the task that he had
undertaken, in one branch of his subject at least he sought,
though with little success, to contribute to the domain of posi-
tive geographical knowledge, by the introduction of numerous
measurements of distances. These statements are in all cases
derived from previous authorities, frequently without naming
them, and for the western parts of Europe, including Italy, are
probably for the most part taken from Agrippa, whom he him-
self in one passage extols as worthy of especial confidence.^
In many other cases they are obviously derived from ^erijjli,
or descriptions of the coasts of the Mediterranean, Euxine, &c.,
such as were so common in ancient times, and so necessary
for navigators when latitudes and longitudes were practically
unknown.® For the eastern parts of the Mediterranean and

^ The seventh book indeed contains
a large assemblage of facts concerning
the nature of man, his physical and
mental qualities; but not even an
attempt at anything like an ethno-
graphical review of the physical pecu-
liarities and characters of the difftrent
varieties of mankind. At the same
time no portion of Pliny's work con-
tains a greater accumulation of fables
and absurd stories, many of them taken
from Isigouus of Nicica, and other

authors who belonged to the class of
■irapa5o^oypd<poi or avowed collectors of
marvellous tales. (See especially c. 2,
§§ 9-32.)

' iii. 2, § 17. See Chapter XX. p. 177.

* Among Eoman writers, besides
Agrippa, he frequently cites the au-
thority of Varro, by which name he
probably means Varro Atacinus, not
the elder and more celebrated writer
of the name. See Chapter XX. p. 171.

3/8 HISTOEY OF ANCIENT GEOGRAPHY. Chap. XXIV.

for Asia, lie falls back for the most part upon Greek writers,
especially Timosthenes, Artemidorus and Isidorus, while for
the more remote provinces of Upper Asia he chiefly followed
Eratosthenes.^

It would be unjust to deny that the distances thus given by
Pliny are frequently of considerable value, especially where
they are measured along the coasts : while others are of interest
in enabling us to understand and reconstruct the geographical
systems of earlier writers. But the point that is most remark-
able throughout, is the want of any attempt on the part of
Pliny himself to reconcile, or even discuss, the discrepancies
between them. Where he met with divergent statements, he
simply contents himself with repeating them, without any
attempt to determine between them, or to pronounce in favour
of the one system or the other. Strabo indeed was frequently
compelled to do the same thing; but, as we have seen, he
often also exercised an independent judgement, and while
he in general followed the authority of Eratosthenes, whose
system he has set forth in a clear and intelligible manner, he
at other times departed from his views, and set up a scheme of
his own, frequently indeed less correct than that of his pre-
decessor, but for which he gives his reasons, and works out his
conclusions in an intelligible form. Both Strabo and Era-
tosthenes had a clear idea of what scientific geography ought
to be, however defective might be their materials, and their
conclusions in consequence erroneous. Pliny on the contrary
had no conception of scientific geography at all, and does not
attempt to enter into any discussion on the subject. It was
enough for him to take the materials that he found ready to
his hand, without attempting to frame them into one con-
sistent whole : and though he has in this manner occasionally
preserved to us passages and statements of much scientific
value, it has been without any indication that he himself
appreciated their importance, or sought to distinguish them

9 See vi. §§ 3, 36, 56, &c.

Sect. 1.

PLINY: GENERAL VIEWS.

379

from the mass of miscellaneous matter by which they are
surrounded.

§ 6. Nowhere are these defects more conspicuous than in
the second book/ in which he gives a general view of all that
was comprehended by the ancients under the name of Meteor-
ology, a term which they applied in a much more general
sense than it is employed at the present day, as including all
that was known of the celestial bodies, the sun, moon, and five
planets, as well as the phenomena of comets and falling stars,
meteors, thunder and lightning, the seasons, winds, and tem-
pests, as well as volcanoes and earthquakes. With regard to
the general questions concerning the earth itself, its position,
and relations to the other bodies of the universe, Pliny acqui-
esces in the system that was generally received in his day,
and had been clearly expounded by Posidonius : he describes
briefly but correctly the courses of the planets, and explains
the cause of the eclipses both of the sun and moon. But it is
remarkable that while he bestows well-merited praise upon
Hipparchus for the astronomical skill that had enabled him to
predict eclipses and publish tables of them for six hundred
years to come,^ he censures him for his excessive, and " almost
impious " daring, in attempting to catalogue the fixed stars,
and determine the place of each, so that future astronomers
might note whether any changes really occurred in them.^ He
applies the same epithet to the attempt of Eratosthenes to
determine the circumference of the earth, though he admits

^ The first book contains only the
table of contents of the thirty-six laooks
that follow, which was drawn up by
Pliny himself with a view to facilitate
reference to the different topics spe-
cially treated of. It was designed in
the first instance for the use of the
emperor Titus, to whom the work was
dedicated, but would serve, as Pliny
remarks, for the convenience of others
also {Prce.fat. §§ 32, 33). He has
added at the end of the summary of
each book, a list of the authors from
whom it was compiled ; a very inter-
esting and valuable addition, but it

must not be supposed that he had con-
sulted them all in the original. He
certainly often takes his facts, even
where he cites his authorities, at second
or third hand.

2 ii. 12, § 54.

^ ii. 26, § 95. " Ideoque ausus rem
etiam deo improham, adnumerare pos-
teris Stellas ac sidera ad nomen expun-
gere." Such a censure seems the more
remarkable as coming from one whose
creed was a philosophical pantheism.
See the iine passage with which he
opens the second book.

380

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXIV.

that his process of reasoning was so ingenious, that it was im-
possible not to believe it.* It was indeed (he says) generally-
adopted, though Hipparchus had corrected it by the addition
of about 26,000 stadia.^

This strange incapacity of appreciating the great scientific
conclusions of the Greek astronomers who had preceded him
was coupled with a ready and almost childish belief in such
absurd notions as that of the germs of all creatures falling
from the figures of them impressed on the outer circle of the
heavens, and that these, when they fell into the sea, frequently
became mingled together, and thus gave birth to monstrous
and unnatural forms.^ On the other hand he justly rejected the
popular notion, which appears to have been current in his day,
of the astrological influences of the stars upon the human race,
or that every man had his star, associated with him from his
birth, and that each falling star marked the decease of the
human being to whom it belonged.^

§ 7. Imperfectly as Pliny evidently understood the mathe-
matical conclusions of his predecessors, he at least clearly
comprehended those which had the most immediate bearing
upon geography, — the obliquity of the ecliptic, and its influ-
ence upon the seasons, the variation in the length of day and
night according to the latitude, the appearance of certain stars
above the horizon from the same cause, and so on. And he
correctly argues in favour of the globular figure of the earth,
from the manner in which ships, lights, and high land dis-
appear below the horizon.^ The same thing is shown, he adds,

* Oonceming the diflSculty raised by
these words, see Chapter XVII. p. 3,
note.

' ii. 3, § 7.

' ii. 8, § 28. " Nee cum sue qu£oque
homine orta moriuntur, nee aliquem
extingui decidua significant." This is
the first allusion I have found to the
beautiful superstition, of which such
poetical use has been made in the well-
known song of Bc'raugcr, " Les c'toiles
qui filcnt."

8 ii. C5, § 1G4.

* ii. 108, § 247. He terms it " im-
prohum ausum, verum ita subtili argu-
mentatione comprehensum, ut pudeat
non credere."

His statement that it was generally
adopted (quem cunctis probari video) is
confirmed by the manner in which it
is referred to by Vitruvius (de Archi-
tectura, i. 6, § 9) as a conclusion uni-
versally recognized. The diflferent
estimate formed by Posidonius would
appiar therefore to liave been either
overlooked or discredited.

Sect. 1.

PLINY: GENERAL VIEWS.

381

by the fact that certain stars and constellations are visible in
some countries and not in others. Thus the Great Bear is not
visible in the land of the Troglodytes (Ethiopia) or the neigh-
bouring parts of Egypt, nor is the bright star called Canopus
visible in Italy or the Euxine, while at Alexandria it rises the
fourth part of a sign above the horizon, but at Rhodes it
only just skirts it.^ The Great Bear in like manner began
to set at Rhodes, and still more at Alexandria, while at Meroe
it was only visible during a short period of the year.

In another passage he correctly describes the gradual
lengthening of the solstitial day, from Meroe where the
longest day was only 12-|- hours, to fourteen hours at Alex-
andria, fifteen in Italy, and seventeen in Britain, where, he
adds, the lightness of the summer nights already promises
that which is proved by reasoning, that the parts of the earth
nearest the pole have six months continual day in summer,
and in like manner six months continual night in winter.^
Here his reasoning is perfectly sound, but when he adds that
Pytheas the Massilian writes that this is actually the case in
Thule, an island six days' voyage to the north of Britain, it is
almost certain that he either misconceived or misrepresented
his authority.^

It would be interesting to know whether this was also the
case when he cites from Onesicritus and other writers state-
ments concerning the astronomical appearances in India,
which are almost as erroneous as that just quoted concerning
Thule. The shadow falling to the south — a fact which can of
course only occur within the tropics, and even there for a short

8 ii. 70, § 178.

1 Ibid. 75, § 186.

2 ii. 75, § 187 ; iv. 16, § 104. " Quod
fieri in insula Thule Pytheas Massili-
ensis scripsit." See Chapter XV.
Note H, p. 613. The still stranger
assertion that, "according to some,"
the same thing took place in the island
of Mona, "about 200 miles from Ca-
malodunum, a town of Britain," is a
striking instance of the utterly un-

critical character of Pliny's mind,
which could think such absurdities
worthy of insertion, without even a
passing word of refutation. It seems
not impossible that this misconception
may have arisen from the passage of
Caesar (J?. 6f. v. 13), where, after de-
scribing Mona, he speaks of the astro-
nomical phenomena reported to occur
in other islands not far from Britain.
See Chapter XIX. p. 128.

38:

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

period only, until one approaches the equator, is stated to have
been observed at Pattala during the stay of Alexander's fleet ;
and the same statement is repeated concerning other places in
the northern parts of India, known to the Greeks, all alike
outside the tropics. Here it appears more probable that the
erroneous or exaggerated accounts were really found by Pliny
in his original authorities : ^ but there are unfortunately
abundant proofs throughout his work how careless he was in
the use of his materials, and how little pains he took to ascer-
tain the true meaning of the authors whose works he had
consulted, and whose authority he cites. A single passage
will sufficiently exemplify this.

§ 8. After stating in accordance with the view generally
established in his time, that the earth was surrounded by a
complete belt of water, so that the inhabited portion of it was
bounded on all sides by the ocean, he adds, that this was no
longer a matter of proof by argument, but had been established
by direct investigation.* "From Gades to the Columns of
Hercules " (he tells us) " around the shores of Spain and Gaul,
the whole of the west is at the present day well known to
navigators. The Northern Ocean was also navigated for the
greater part under the auspices of Augustus, his fleet having
coasted round Germany to the Cimbrian Promontory, and from
thence looked out upon a boundless sea, which was reported
to extend to the region of Scythia and the parts chilled by
excess of moisture. For which reason " (he observes) " it is most
improbable that the sea should be wanting where there is the
greatest amount of moisture.^ Beyond that again, the whole
coast from the East, and from the Indian Sea, extending round
in the same latitude to the Caspian, was navigated by the
Macedonian fleets under the reign of Seleucus and Antiochus.
In the neighbourhood of the Caspian also many shores of the

^ See this point discussed in a note
to the Vovago of Nearchus, Chapter
XIII. NoteE, p. 535.

^ " Nee argiimentis hoc investigan-

dum, sed jam experimcutis coguitum,"
ii. 66, § 166.

^ No further development is found
in Pliny of this strange speculation.

Sect. 1.

PLINY: GENEEAL VIEWS.

383

ocean have been explored, and but little is wanting for the
whole of the north on both sides to have been visited by navi-
gators. But as if to leave no room for conjecture, the Palus
Mseotis affords a strong argument [of the proximity of such a
sea], whether it be, as many believe, an inlet of the ocean, or
a back-water (restagnatio), separated from it by a narrow strip
of land. On the other side, beginning from Gades on the west,
a large part of the southern coast around Mauretania is at the
present day frequented by navigators. The greater part of
this southern sea and of the eastern coast was made known by
the victories of Alexander, as far as the Arabian Gulf, in
which, when Caius Csesar the son of Augustus held the com-
mand, portions of wrecks are said to have been recognized as
derived from ships of Spanish origin. And while the power of
Carthage was at its height, Hanno made the passage round
from Gades to the borders of Arabia, and left a written account
of his voyage ; as did also Himilco, who was sent out at the
same time to explore the outer coasts of Europe.^ Moreover
Cornelius Nepos states that within his own time a certain
Eudoxus, seeking to escape from king [Ptolemy] Lathyrus, set
out from the Arabian Gulf and accomplished the passage to
Gades, and long before him Cselius Antipater asserts that he
had seen a merchant who had sailed from Spain to Ethiopia
for the sake of trade." He then repeats the story told by
Cornelius Nepos of the Indians who had been driven by storms
round the northern shores of Asia and Europe to the coast of
Germany.''

It would be difScult to find a stronger instance of the pro-
miscuous manner in which Pliny raked together his materials,
or of the total want of critical judgement, or even common
accuracy with which he made use of them. We have already

^ This is the first mention we find of
the voyage of Himilco, and the only
notice of it that occurs in Pliny, though
Ills name, as well as that of Hanno, is
found in the list of his authorities for
the book. The subject will be dis-

cussed wlien we come' to the work of
Avienus, to whom we are indebted for
what little knowledge we possess con-
cerning it.

' ii. 67, §§ 167-170. See Chapter
XXIII. p. 364.

384 HISTORY OP ANCIENT GEOGRAPHY. Chap. XXIV,

seen what was the real extent of the exploration of the northern
coasts of Europe under Augustus ; but on this point at least
Pliny only reflected the popular impression of his time, con-
firmed by the statement of Augustus himself.^ With regard
to the alleged voyage of the Macedonians from the Indian
Ocean to the Caspian, we have also seen what was the real
foundation of the story, and how carefully Strabo distinguishes
the assertion of Patrocles that it was possible, from the popular
idea that it had been actually accomplished. In like manner
the statements concerning the voyages of Hanno and Eudoxus
are altogether perverted and misinterpreted ; while the story
of the Indians told by Cornelius Nepos must be in great part,
if not altogether, a fiction, and the supposed discovery of the
figure-heads of Spanish ships in the Ked Sea recalls the similar
tale told by Eudoxus, of which it is probably only a repetition.
The strange argument derived from the supposed proximity '.f
the Palus Mseotis, is one of those curious instances of subtle
arguments based upon no foundation at all, which are not un-
commonly found in the later Greek writers. But it is singular
that Pliny did not see how completely this hypothesis was at
variance with the well-known fact that the Tanais flowed into
the Palus Mseotis, and with his own statement that it had
its sources in the Khipaean Mountains, far to the north of
that sea.^

§ 9. The notices collected by Pliny concerning earthquakes,
volcanic eruptions, and other physical phenomena, are not
without value, some of his facts being otherwise unknown.
But his philosophical remarks and conclusions are of the most
futile character, and we find no trace of the sagacious observa-
tion of Strabo, who pointed out the obvious signs of volcanic
action in countries where no outbreaks of the kind had been
recorded, and thus led the way to the acknowledgement of the
important part borne by these forces in remodelling the surface
of the globe.

^ See Chapter XX. p. 190. " iv. 12, § 78.

Sect. 1.

PLINY : GENEEAL VIEWS.

385

With regard to the height of mountains — an important branch
of physical geography generally neglected by ancient writers
— he quotes the statement of Dicsearchus, that Pelion, which
was the highest mountain he had measured, did not exceed
1250 paces (6250 feet) in perpendicular altitude ; but adds
that some of the highest summits of the Alps rose with a con-
tinuous slope for a distance of not less than fifty miles.^
Elsewhere he speaks of the elevation of Mount Hssmus as
attaining to six Eoman miles.^ But probably he does not in
either case mean to represent this as the perpendicular alti-
tude.^ It would be curious to know on what foundation a
writer named Fabianus (whom he cites as his authority) had
arrived at the conclusion that the greatest depth of the sea was
fifteen stadia.*

§ 10. Pliny concludes his second book with a discussion of
the various measurements that had been given of the length
and breadth of the inhabited portion of the earth. Adopting,
as we have seen, the general conclusion of the Greek geogra-
phers, that this was surrounded on all sides by the ocean, so as
to constitute in fact a great island, he adopted also their view,
that its greatest length from east to west much exceeded its
breadth from north to south. In repeating the estimates that
had been formed of its dimensions, it is remarkable that he

1 ii. 65, § 162.

''■ iv. 11, § 41. It is more strange
that he should describe Saoce, the
central peak of Samothrace, which is
really only 5240 feet high, and far
inferior to the neighbouring Athos, as
ttn miles in height. Ibid. 12, § 73.

^ Tliis appears to me certainly to be
the natural construction of the first
passage, where he says : " Mihi incerta
hsec videtur conjectatio, haud ignaro
quosdam Alpium vertices longo tractu
nee breviore quinquaginta millium
passuum adsurgere." But the words
have been frequently understood as
implying that tliis was their actual
height, or perpendicular elevation ; an
absurdity that we have no riglit to

VOL. II.

force upon our author, when his words
will fairly admit of another meaning.

* "Altissimum mare sv stadiorum
Fabianus tradit." ii. 102, § 223. This
Fabianus is doubtless the same author
wliom he quotes in one of his latest
books under the name of Papirius
Fabianus, and terms "naturae rerum
peritissimus " (xxxvi. 15, § 125). He
was a friend of the elder Seneca, and
published many works of a philo-
sophical, as well as others of a rhe-
torical character. Posidonius, as we
have seen, estimated the greatest depth
of the Mediterranean at 1000 fathoms,
equal to ten stadia. (See Chapter
XVIII. p. 98.)

2 c

386

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

altogether omits those given by Eratosthenes, which are so
fully discussed by Strabo, and contents himself with giving
the statements of Artemidorus, and comparing with them
those of Isidorus.^ The former have been already fully dis-
cussed.® The value of Isidorus as an authority we have no
means of estimating : and Pliny merely gives his general
results, without any details of the calculation on which they
were founded. It appears that he estimated the total length
of the world from India to Gades at 9818 Koman miles,
(78,544 stadia), while Artemidorus made it only 8568 miles, or
68,545 stadia: while in regard to its breadth, he made an
addition of not less than 1250 miles to the north, from the
mouth of the Tanais to the parallel of Thule, a proceeding
that is justly censured by Pliny as a mere conjecture,^ but he
adds that the extent of the territory occupied by the Sarma-
tians towards the north was undoubtedly very great.^

He concludes with referring to the measurement of the cir-
cumference of the earth, by Eratosthenes ; but cannot refrain
from adding a foolish story, which he himself discredits, of a
certain Dionysodorus, a mathematician of Melos, in whose tomb
was found a letter stating that after his death he had pene-
trated to the centre of the earth, and that the distance was
42,000 stadia. (!)* As this would be the radius corresponding
(in round numbers) to a circumference of 252,000 stadia — it is
evident that the fiction was invented in order to support the
received calculation of this measurement. It seems not im-

* This Isidorus is probably identical
with the author of the little work (2to0-
fi.o\ nap9LKoi) still extant under the name
of Isidorus of Charax, but the state-
ments in question must have been
taken from another work. See Chapter
XX. p. 164.

« See Chapter XVIII. p. 64.

' " Quae conjectura divinationis est."
ii. 108, § 246. It is evident that Isidorus
followed the same general view as Era-
tosthenes, in thus carrying the conti-
nent of Europe far to the north, to cor-
respond with the assumed latitude of

Thnle, But his addition is so large as
to be unintelligible, if the figures given
by Pliny are correct.

* He seems even to think that it
might not be less extensive than the
estimate given by Isidorus — " Ego nou
minore quam proxime dicto spatio, Sar-
matarum fines nosci intelligo." I. c. In
adopting this view he must have en-
tirely forgotten his own argument for
the proximity of 'the Palus Mseotis to
the Northern Ocean.

8 ii. 109, § 248.

Sect. 2.

PLINY: DESCRIPTIVE GEOGRAPHY.

387

probable that the statement as to the distance had really been
made by Dionysodorus, and the story afterwards perverted
into the strange form in which it is repeated by Pliny.

Section 2. — Descriptive Geography.

§ 1. Pliny next proceeds to the detailed description of the
different countries of the world. Here he follows an order
different from that of Mela, but scarcely less inconvenient.
Beginning from the Strait of Gades (as he calls that of Gib-
raltar) he follows the northern coast of the Mediterranean,
describing the parts of Spain and Gaul adjoining that sea ;
then the western portions of Italy, down to the Bruttian Pro-
montory and Locri ; after which he gives a brief account of all
the islands in this western portion of the Mediterranean,
including Sardinia, Corsica, and Sicily : then he returns along
the coast of the Adriatic, describing the eastern portions of
Italy, with Cisalpine Gaul and Venetia, and in connection with
these the Alpine nations, including the Ehsetians and Vinde-
licians. Thence he descends along the opposite coast of the
Adriatic through Liburnia and Dalmatia, both of which he
includes in Illyricum, to the Acroceraunian Promontory, which
he considers as the limit of the second section of southern
Europe : ^ he then adds a brief account of the provinces of the
interior, Noricum, Pannonia, and Moesia, and notices a few of
the islands on the lUyrian coast. The fourth book begins with
a long and detailed, but extremely unsatisfactory, description

' Here he is certainly following a
kind of geographical arrangement ; for
he begins with telling us that the sea
indents Europe with many recesses,
but especially with four principal gulfs
(sinus), iii. 1, § 5. The fii-st of these
he conceives as extending from the
promontory of Calpe in Spain to that
of Locri (Leucopetra) in Italy; the
second, from the same promontory to
that of Acroceraunia ; and the third,

from thence to the Hellespont. " Ter-
tius Europse sinus Acrocerauniis in-
cipit montibus, finitur Hellesponto."
In what sense the term " sinus "
can possibly be applied to this portion
of the Mediterranean it is diificult to
conceive. Had he taken either of
the southern promontories of the Pe-
loponnese it would have furnislied
something like a natural limit.

2 c 2

388

HISTORY OF ANCIENT GEOGEAPHY. Chap. XXIV.

of G-reece, followed by Macedonia and Thrace to the Helles-
pont : then follows a very long and minute enumeration of the
Greek islands ; after which he returns to Thrace, describing
the coasts of the Euxine and the adjoining nations, the Getse
and Scythians, as far as the Palus Mseotis and the Tanais, in
connection with which he repeats the fable of the Ehipaean
Mountains and the Hyperboreans. Thence he crosses these
mountains^ — which he evidently pictured to himself as a range
running parallel with the ocean, and bounding the European
Scythians to the north — to the shores of the JSTorthern Ocean,
and follows these westward back to Gades. His notices of the
coasts and islands of the Northern Ocean are, as might be
expected, very few and scanty, and even those of Germany
singularly meagre. The same is the case also with Britain and
Ireland, which he next mentions, while he is of course able to
give a copious list of the towns and tribes of the external pro-
vinces of Gaul and Spain, as he returns along their coasts to
Gades and the Straits from whence he set out.

§ 2. There was doubtless no province of the Eoman Empire
with which Pliny was more familiar, or concerning which he
had better means of information, than Spain, in which he had
himself filled the office of Procurator, or civil governor.^ But
for that very reason we are the more struck with the extremely
imperfect character of the description he has left us, considered
in a geographical point of view. In fact, he can hardly be said
to have given us any geographical account of it at all ; a
deficiency the more striking as the strange manner in which
he has, by the arrangement already explained, divided it into
two portions, rendered it particularly necessary to give a good

" This is his own expression : " Exe-
undum deinde est, ut extera Europse
dicantur, transgressisque Biphieos mon-
ies littus Oceani septemtrionalis . . .
legendum." iv. 13, § 94. It is evident,
therefore, that these visionary moun-
tains had as definite a place in Pliny's
conception of the geography of Europe
as the Alps or the Balkan.

^ See above, p. 372. The exact date
and duration of his government is un-
certain. But he appears not to have
returned to Eome till the reign of Ves-
pasian, about A.u. 73.

No allusion is found in his work to
his having any special sources of in-
formation on this account.

Sect. 2.

PLINY : DESCRIPTIVE GEOGEAPHY.

389

general outline of the whole. But such an outline is wholly-
wanting. The whole country had in the time of Pliny been
completely brought under the Eoman system of administration,
and had been divided for administrative and judicial purposes
into districts (conventus juridici), each of which had its chief
town or capital, to which all the surrounding towns and native
tribes were subject. This division has been made by Pliny (in
this as in many other cases) the basis of his description, and
such a choice was well adapted for a mere statistical enumera-
tion of the names of places, which is in fact for the most part
all that he has given us, accompanied with a notice of the
municipal condition of those which possessed any peculiar
privileges, as colonies, municipalities of Eoman citizens, &c.
All this affords excellent material for the political statistics of
the Roman Empire, and the great number of names that he
enumerates is of use to the topographer in modern times, who
is often enabled to identify them without any geographical
indications from their being still preserved in very little altered
form.* Nor does he omit to mention briefly, with regard to all
the great rivers of the peninsula —the Iberus, Durius, Tagus,
Baetis, &c. — both where they take their rise and the nations or
cities by which they flow. But no attempt is made to com-
bine these separate notices, or to show the connection and
boundaries of the river-systems of Spain ; while of its moun-
tain chains, with the exception of the Pyrenees, he has given
us no particulars at all. Imperfect as were the notions pos-
sessed by Strabo of the physical geography of the Iberian
peninsula, they were decidedly superior to those which are
furnished us by Pliny, though the latter had far ampler mate-
rials for the topographical and detailed description of the

* The copious lists given by Pliny of
the cities and " populi " of Spain are
examined and compared with those
furnished by other authors, by M.
Heiss (Description Gen^rale des Mon-
naies Antiques de VEspagne, 4to Paris,
1870), whose work, in conjunction with
that of M. Hiibner, published in 1869,

forming the second volume of the new
Corpus Inscriptionum Latinarum, and
containing the inscriptions found in
Spain, has for the first time placed the
comparative geography of the Iberian
peninsula on a securely established
footing.

390

HISTORY OF ANCIENT GEOGEAPHT. Chap. XXIV.

country. Even these he has presented to us in so crude a form,
and so ill-arranged, that it would be difficult, if not impossible,
to represent on a map this mass of incoherent and often con-
tradictory details.

One of the few points in which he shows a knowledge of
the peninsula in advance of that of Strabo, is in regard to the
Pyrenees, which he rightly conceived as having their direction
from east to north-west,^ instead of from south to north, and
thus rendering the northern side of the peninsula shorter than
the southern or that facing the Mediterranean. He is also the
first author who attaches due importance to the projection
formed on the west coast by the great headland north of
Lisbon, now known as Cabo da Eoca, or the Eock of Lisbon ;
though he has fallen into a strange confusion by supposing
this to have been the headland called by Greek writers
Artabrum.^ At the same time he exaggerates the case where
he makes it the limit between the two sides of the peninsula,
and reduces the Sacred Promontory (Cape St. Vincent), which
had occupied so prominent a place with all the earlier
geographers, to a merely secondary position.

§ 3. With Gaul his acquaintance is far less complete ; with
the exception of the province of Gallia Narbonensis facing the

* "Ipsa Pyrensei juga ab exortu
sequinoctiali fusa in occasum brumalem,
breviores latere septentrionali quam
meridiano Hispanias faciunt." iv. 20,
§110.

" Mela first mentions this promon-
tory under the name of Magnum (iii. 1,
§ 7). Pliny says of it : " excurrit de-
inde in altum vasto cornu promon-
torium, quod aliqui Artabrum appella-
vere, alii Magnum, multi Olisiponense,
ab oppido, terras, maria, ccehim dis-
criminans " (iv. 21, § 113). No doubt
can exist as to the promontory of which
he means to speak, from the last name
applied to it, as well as from his placing
it south of the Durius (Douro). But
it seems almost certain that Jie has
erroneously applied to it what Artemi-
dorus said of the Artabran Promontory

(Cape Finisterre, the Nerium of Strabo,
to which Pliny in common with Mela
gives the name of Celticum). The
words which follow : " Illo finitur
Hispanise latus et a circuitu ejus incipit
frons : septentrio hinc oceanusque Gal-
licus, occasus illinc et oceanus Atlan-
ticus," are certainly applicable only to
Cape Finisterre. There is evidently
some great confusion in the matter, but
in the absence of the earlier autho-
rities it is impossible to say with cer-
tainty how much is the fault of Pliny
and what may be due to the Greek
writers. At the same time, with his
improved means of information, he
ought in any case to have rectified
their errors and cleared up the question,
which he has not done.

Sect. 2.

PLINY : DESCKIPTIVE GEOGKAPHY.

391

Mediterranean, which is on the whole well described, though
as usual wanting in clearness of arrangement. But here both
the names of the principal tribes, and of the large towns that
had grown up under the long continued Eoman rule, were so
well known, that it was of little consequence in what order
they were mentioned. Yet we miss even here the description
of the peculiar characters of the coast and the vast marshes
and lagunes formed by the Ehone and other rivers, which
constitute so remarkable a physical feature of this portion
of Gaul. Of the province generally he tells us that in its
productions and culture, as well as the civilization and wealth
of its inhabitants, it was become rather a part of Italy than a
province.'

Of the remainder of Gaul, or the provinces verging on the
Ocean, he disposes in a very summary manner. After briefly
indicating the division into three provinces, or rather into
three nations — the Belgse, Celtse, and Aquitani — in which he
follows the divisions marked out by Csssar rather than those
of the Eoman provinces of his day, though he appears to
regard them as identical — he proceeds simply to enumerate
the " populi," tribes or districts, included in each division.
Of these he furnishes us with a very complete list, including
all those mentioned by Csesar, and a considerable number
more, the names of which he probably derived from Agrippa,
whose personal administration of Gaul must have given him a
thorough acquaintance with the country. But beyond this
bare list of names Pliny gives us no information at all. He
hardly mentions even any of the towns, some of which had
certainly in his time attained to considerable importance ; and
none of the rivers, except those which in his day, like the

^ " Agrorum cultu, virorum mo-
rumque dignatione, amplitudine opum,
nuUi pTovinciarum postferenda, brevi-
terque Italia verius quam provincia"
(iii. 4, § 31). He describes this pro-
vince as bounded on the north, and
separated from the other provinces of

Gaul by the Mons Cebenna and the
Jura. It would thus include the Hel-
vetians ; though Pliny himself enume-
rates that people in Gallia Belgica (iv.
17, § 106), to which they were certainly
annexed for administrative purposes.,

392

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

Sequana and Garumna, formed the boundaries of the pro-
vinces, with the single exception of the Loire (Ligeris), which
he terms " flumen clarum," ^ but without giving us any par-
ticulars as to its origin or course.^ Almost the only point of a
strictly geographical character which he condescends to notice
is the projection of the peninsula of Bretagne, occupied by the
Osismii, which he describes as running out into the Ocean, so
that its circuit was not less than 625 Eoman miles, though the
breadth of the neck or isthmus joining it to the mainland was
only 125 miles. But even here the manner in which this
statement is introduced is such as would be unintelligible,
were we not able to compare it with other authorities, and
acquainted with the real facts of the case.^ It is strange also
that he appears to apply the name of Armorica, which as we
have seen was in use in Cesar's time as a general appellation
for the nations of Bretagne, to the Aquitanians of the south-
west.^

§ 4. Pliny's account of Italy is unquestionably in some
respects one of the most valuable parts of his work, at the
same time that it affords a characteristic example of its
principal defects. He appears indeed at first to rise to the
dignity of his subject,* and breaks out into an enthusiastic
panegyric upon the natural advantages of the country, which
recalls the well-known passage in the Georgics of Virgil.^
He apologizes at the same time for the imperfect manner in
which he is compelled to treat so attractive a theme, and to
run over in a cursory way what would be a subject for volumes.
But having said this, he lapses at once into a mere enumeration
of names, resembling that which he has given us for Gaul and

" iv. 18, § 107.

•* He however mentions the Araris,
Isara and Druentia, as tributaries of
the Rhone ; whicli he describes in con-
nection with the Roman province, iii.
4, § 33.

' iv. 18, § 107. He terms it "peuin-
sulam spectatiorem excurrentem in
Oceanum a fine Osismiorum."

* " Inde ad Pyrenaei mentis excursum
Aquitanica, Aremorica ante dicta " (iv.
17, § 105). It can scarcely be doubted
that he here means the same name
with the Armoricans of Ca3sar, and
that he has erroneously transferred the
name to a diiferent part of Gaul.

3 Plin. H. N. iii. 5, §§ 39-42. Com-
pare Virgil, Georg. ii. 136-176.

Sect. 2.

PLINY: DESCRIPTIVE GEOGEAPHY.

393

Spain. Here again, as might be expected, he had excellent
materials, his description of Italy being based (as he himself
tells us) upon the official record of Augustus, when he divided
Italy into eleven "regions": an administrative division of
which we learn the particulars only from Pliny, though it
continued in use for official purposes down to the time of
Constantine.^ There can be no doubt that his lists of the
towns or communities included in each region were taken from
the same official source, and may therefore be relied on as
authentic, except in so far as their names may have been
disfigured by copyists. But such a document was of course
originally intended as a statistical, not a geographical, survey ;
and though Pliny has so far departed from it, as to describe
the regions in geographical order, beginning with Liguria,
and ending with Yenetia and Istria, and even in the detailed
enumeration of the towns, to follow as far as possible the lines
of sea-coast, he has hardly attempted to give anything like a
real geographical description, either of the peninsula itself, or
of the several portions of it.^

§ 5. Even his notice of the great chain of the Apennines —
the backbone of the peninsula, which determines its whole
configuration, is so brief and summary as to convey scarcely
any information,® and is very far inferior to the clear and
characteristic sketch given by Strabo. With regard to the

* See Marquardt's Handbuch der
Momischen Alterthilmer, vol. iii. part 1,
pp. 57-64.

* He has himself described to us ia
this instance the course that he has
pursued. "Qua in re prsefari neces-
sarium est, auctorem nos 13ivum Augus-
tum secuturos, descriptionemque ab eo
factam Italise totius in regiones xi sed
ordine eo, qui littorum tractu fiet ; ur-
bium quidem vicinitates ordiuatione
utique prsepostera servari non posse;
itaque interiori in parte digestionem
in litteras ejusdem nos secuturos, colo-
niarum meutione signata quas ille in
eo prodidit numero." iii. 5, § 46.

" After describing the coast of Li-

guria from the Varus to the Macra, he
adds : '• A tergo autem supra dictorum
omnium Apeuninus mons Italise am-
l^lissimus, perpetuis jugis ab Alpibus
tendens ad Siculum fretum " (iii. 5,
§ 48). This is literally all that he tells
us concerning the position or direction
of this celebrated chain.

He does not even fix the point where
the Alps ended and the Apennines
began, though it would appear inci-
dentally (iii. 19, § 132) that he accepted
the received view, adopted also by
Strabo, which placed the point of
junction at Vada Sabbata (Vado near
Savona).

394

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

northern provinces of Italy again, he has wholly failed to giva
us any distinct account of the great valley, or rather plain, of
the Po, with its broad extent of alluvial land, and the two
mountain chains bounding it on either side like two great
lines of rampart — a natural picture which one would have
thought no one looking at it with an observant eye could
have failed to seize. He has, however, given us a detailed
description of the river Padus itself, from its sources in the
Mons Vesulus (Monte Viso), which he calls the highest summit
of the Alps,'' to its mouths in the Adriatic, of which he has
given us some interesting particulars, not to be found else-
where.* He has also enumerated correctly its principal
affluents from both sides : and in another passage has con-
nected those on the Alpine side with the lakes from which they
flow.^ The importance of this great river in a geographical
point of view was indeed enhanced in the time of Pliny by
its having been adopted by Augustus as the boundary through-
out its whole course between the Kegions into which Northern
Italy was divided: Gallia Transpadana and Venetia on the
north, Liguria and GaLUa Cispadana on the south.

Of the Tiber in like manner he has given us a copious
account, as was naturally to be expected from its special
interest to an inhabitant of Eome. But he contents himself
on the other hand with a bare mention of the Arno, as flowing
by Florence ^ : and notices in an equally cursory manner the
Liris (Garigliano) and the Vulturnus. In like manner his

' "Padus a gremio Vesuli montia
celsissimum in cacumen Alpium elati
. . . profluens." The notion that the
Monte Viso was the highest summit of
the Alps continued to be entertained
down to a late period, and is not sur-
prising, on account of the prominent
position it assumes, when viewed from
the plains of Italy, In like manner
the Canigou was long supposed to be
the highest summit of the Pyrenees.

8 iii. 15, § 118.

" " Adduam Larius, Ticinum Ver-
bauus, Mincium Benacus, Ollium Sebi-

nus, Lambrum Eupilis." iii. 19, § 131.
It appears at first strange that while he
mentions the Lago d'Iseo (Sebinus),
and even the little Lago di Pasiano
(Eupilis), he has omitted the much
more important Lake of Lugano, but
the reason doubtless is that this lake
does not give rise to a separate river,
its waters being carried off by a short
course into the Lago Maggiore (Ver-
banus).

1 " Florentini prasfluenti Arno appo-
siti." iii. 5, §52.

Sect. 2.

PLINY : DESCRIPTIVE GEOGRAPHY.

395

account of Campania, though ushered in with a rhetorical
flourish in praise of its fertility, contains in reality no
description of the peculiar natural conformation of the pro-
vince, of the volcanic phenomena with which it had so long
been associated, or even of the beautiful gulf which had not
yet been disfigured by the eruptions of Vesuvius.^

§ 6. Altogether it must be said that although his description
of Italy — if this term can be applied at all to the bare
catalogue of names which he has furnished us — supplies useful
materials to the topographer from the great number of such
names that he has brought together, and from the certainty
that these are in the main authentic, and correctly assigned to
their respective Eegions, ^ it would have been difficult to
compile one which should throw less light upon the real
geography of the peninsula. Nor is it more satisfactory in its
relation to historical geography. It was impossible indeed to
ignore altogether the changes that had taken place in this
respect : the tribes and nations that had passed away, or been
replaced by others, and the towns that had figured as im-
portant cities in the early ages of Kome, but which had wholly
ceased to exist in the time of Pliny. But these changes are
indicated so concisely, or so mixed up in a confused mass with
others, that they have seldom any real historical value. In
the case of Latium itself, where so large a number of these
early towns had been absorbed by the increasing greatness of
Eome, he gives a list of not less than fifty-three cities (clara
oppida), which had ceased to exist, including places like
Antemnse, Csenina, and Corioli, which figure conspicuously in
the early Roman history, mixed up with names utterly un-
known, and probably derived only from ancient rituals, like
that of the Septimontium at Rome.^

^ " Pulcherrimus sinus," as it is justly
termed by Tacitus, " antequam Vesu-
vius mons ardescens faciem loci ver-
teret." (Tac. Annal. iv. 67.)

^ iii. 5, § 70. " Ita ex antique Latio
Liii populi interiere sine vestigiis."

The last expression is no doubt not
intended to imply that there were no
ruins left, but some of the sites enume-
rated could hardly have been uninha-
bited in Pliny's time.

396

HISTOKY OF ANCIENT GEOGRAPHY. Chap. XXIV.

At the same time he accepts without hesitation the assertion
of an author named Licinius Mucianus — a contemporary of his
own — that there had once been twenty-four other towns on the
site then occupied by the Pontine Marshes * : as well as the
not less astounding conclusion that because Theophrastus still
described the insulated promontory of Circeii (Monte Circello)
as an island, and stated its dimensions, therefore the whole of
the intervening space by which it was joined to the mainland
had been "added to Italy" since the year in which that
author wrote ^ (u.c. 440). Uncritical as this conclusion would
have been, had Theophrastus really made the statement, it
becomes ten times more so when we find, from the passage
which is still extant, that Theophrastus said nothing of the
kind, but correctly described "the Circeium" as a lofty pro-
montory, which was said hy the inhabitants to have been once an
island, but had become united to the mainland by the alluvial
deposits of rivers.® This is only one instance out of many of
the strange manner in which Pliny misconceived or mis-
interpreted the authorities he had so diligently collected.

§ 7. His account of the two great islands of Sardinia and
Corsica is singularly meagre. After stating with tolerable
correctness the length and breadth of Corsica and its distance
from the mainland of Etruria, he tells us that it contained
eighteen " civitates" — meaning of course tribes or communities,
not cities — and two colonies, Mariana and Aleria, the one
founded by Marius, the other by Sulla. And this is all ! Not
a word of its mountain ranges, so conspicuous to any one that
had sailed over the Tyrrhenian Sea : or of the vast forests that

* " A Circeis palus Pomptina est,
quem locum xxiv urbium fuisse
Mucianus ter consul prodidit," iii. 5,
§ 59. Some MSS. have xxxiii. It is
diflicult to understand what misconcep-
tion could have given rise to this strange
statement, no trace of which is found
in any other authority. The fact of
Mucianus having been three times
consul is curiously introduced, as if it
added to his authority upon such a

point.

^ Ibid. § 58, "Theophrastus . . .
Circeiorum insulae mensuram posuit
stadia octoginta, in eo volumine quod
scripsit Nicodoro Atheniensium magis-
tratu, qui fuit Urbis nostras ccccxl
anno. Quidquid est ergo terrarum
prseter decem millia passuum prope
ambitus, adnexum insulae post eum
annum accessit Italiao."

" Theophrast. Hist. Plant, v. 8, § 3.

Sect. 2.

PLINY : DESCRIPTIVE GEOGKAPHY.

397

rendered it " shaggy and savage," as it was forcibly termed by
Theophrastus ' : or of the wildness of its inhabitants, resulting
from these physical peculiarities. Of Sardinia he tells us little
more. Though his measurements of its dimensions are much
more accurate than those of Strabo, he gives us no general idea
of the country, and does not even mention its unhealthiness, for
which it was almost proverbial among the Romans in his time.^

With Sicily he was of course much better acquainted : and
here his detailed enumeration of the towns of the island is the
more interesting, because we have the opportunity of com-
paring it with the lists given by Cicero in his Yerrine orations,
to which we are indebted for so much interesting information
concerning the topography of the island. But as usual he
gives us little more, and even his passing allusions to the
volcanic phenomena of -i3Etna and the ^Eolian Islands, which
he could not well ignore altogether, are as meagre as possible.
At the same time he enumerates the names of all the smaller
islands near the coasts of Italy and Sicily, many of them mere
rocks, wholly unworthy of notice. Nor has this list even the
merit of accuracy, for in two instances he inserts the same
island twice over : one as Planaria, and again as Planasia : the
other under the two different names of Osteodes and Ustica,
both of which unquestionably refer to the same island.^

§ 8. It is unnecessary to follow in detail the particulars that
he has left us concerning the other countries of Europe that
were in his time subject to the Roman Empire. Those that had
been long reduced under the usual form of provincial adminis-
tration, as was the case with lUyricum and Dalmatia, furnished
him with statistical details similar to those of Gaul and Spain ;

^ TTuffav Tr}v vrjcrov Saffelav Ka\ Sxnrep
riypioofxevriv ry vA'p. Hist. Plant, v. 8,
§2.

* Thus Mela terms it " fertilis et soli
quam coeli melioris, atque ut fcecunda
ita psene pestilens " (ii. 7, § 123) and
Martial uses its name as the very type
of a deadly climate (" in medio Tibure
Sardinia est," Epigr. iv. 60) Tacitus
also tells us that a number of persons

accused of Egyptian and Jewish super-
stitions were transported to the island,
where if they perished from the climate
it would be little loss (" si ob gravita-
tem cceli interissent, vile damnum,"
Tac. Ann. ii. 85). It was thus looked
on as a kind of Cayenne.

8 See the articles Planasia and
Osteodes in Dr. Smith's Diet, of Geogr.
vol. ii.

398

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

and he has grouped together the different tribes of these wild
mountain regions, according to their division into " conventus "
for administrative purposes. This affords us at least some
approach towards a geographical arrangement ; but very few of
the " populi " thus enumerated can be identified. His geogra-
phical knowledge of these provinces, as well as those extending
from the Alps to the Danube — Ehsetia, Vindelicia, Noricum,
and Pannonia — was however decidedly in advance of that
possessed by any of the Greek writers : he was well informed
concerning the tributaries of the Danube — the Save, the Drave,
and the Colapis (Kulpa), which joined the Save at Siscia : ^
and he justly censures the writers who had represented an arm
of the Danube as flowing into the Adriatic and giving name to
the peninsula of Istria.^ Moesia on the contrary, which had
been lately incorporated in the Eoman Empire, he disposes of
in a very summary manner : and with Dacia and the provinces
beyond the Danube which had not yet been brought under
subjection, his acquaintance was so imperfect that he hardly
mentions them at all.^

§ 9. There is hardly any portion of his work, which more
strongly exemplifies all the defects of Pliny's method, and his
utter want of conception of the task he had undertaken as a
geographer, than his description of Greece — a country on
which, as he himself tells us, he dwells at considerable length,
on account of its ancient fame and literary celebrity. Hence
he could not have wanted for good materials had he chosen to
avail himself of them. But as usual he affords us no real
description of the country, either geographical or physical, and
presents us with nothing but a confused assemblage of names,

' Siscia, still called Siszek, had been
converted into a fortress by Augustus,
and for some time afterwards continued
to be one of the chief cities of Pannonia.
It afterwards gradually declined, as
Sirmiura, lower down the Danube, rose
into increasing importance.

^ iii. 18, §127. He adds with unusual
emphasis : " Nullus enim ex Danuvio

amnis in mare Hadriaticum eflfunditur."
The contrary opinion, as we have seen,
was still held by Cornelius Nepos and
by Mela. See Chapter XXIII. p. 357.
^ He does not appear to have had
any knowledge of the great river Theiss,
or of the Carpathian mountains, the
name of which appears for the first
time in Ptolemy.

Sect. 2. PLINY : DESCRIPTIVE GEOGRAPHY. 399

rendered even more confused and perplexing by the mixture of
those of different ages into one undistinguished mass. We
have seen that Strabo impaired the clearness of his geogra-
phical account of Hellas by an excess of archaeological lore,
and by needless discussions on the connection of the Homeric
geography with that of his own time. But he took care at
least to keep the two distinct, and if he devoted a dispro-
portionate amount of space to such antiquarian disquisitions,
he did not omit to give us a clear geographical outline of each
province and district of Greece. Pliny gives us no such
outline (beyond the trite comparison of the Peloponnese to a
plane leaf), while the names which he heaps together in a con-
fused jumble are some of them places that were still peopled
and inhabited, some of them derived from the Homeric geo-
graphy, that had long since disappeared, others merely obsolete
or poetical names for the same towns that he enumerated under
their later appellations. He had apparently in this instance
no official catalogue upon which to rely with regard to the
existing state of things, and hence compiled at random from
his Greek authorities, with no intelligible criterion or rule of
selection.

For the northern coasts of the ^Egean he presents us with a
tolerable paraplus : but his enumeration of the islands in that
and the Ionian Sea is again a mere dry nomenclature, inter-
spersed with occasional statements of the distances from one to
the other, but unaccompanied with any geographical indica-
tions of their position : except in the case of the Oyclades,
the arrangement of which in a kind of circle, with Delos as
its centre, had come to be regarded as one of the received
points of geography.* But even here he was unable to adhere
to any definite or intelligible order, and has confused his

■* According to Strabo (x. 5, § 2) 1 at first twelve of them, but others were

the Cyclacles were not merely a geo- subsequently added. Strabo however

graphical designation, but represented rejects the three insignificant islands

an union for sacred purposes, who used of Prepesinthus, Oliarus, and Gyarus,

to send sacrificial deputies {Oewpovs) which were admitted by Artemidorus,

and choral bands to Delos. There were 1 and thus reduces the number to twelve.

400

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

enumeration by tlie introduction of obscure islets out of their
place, and tbe omission of others of more importance where
they would naturally be looked for.®

§ 10. His accounts of the Euxine and its European shores
is tolerably full and circumstantial, but as in other cases is
obscured by the confusion arising from his mixing up names
and statements derived from Herodotus or writers who followed
him, with those of later authors who described a state of things
wholly different. Thus we find him reintroducing the Panticapes
as one of the rivers of Scythia, which he describes as separating
the agricultural Scythians (Georgi) from the nomads — a state-
ment derived from Herodotus and Ephorus : " and he adds that
some writers represented the Panticapes as a confluent of the
Borysthenes below Olbia, while those better informed (diligen-
tiores) called this confluent the Hypanis : " so great was the
error (he observes) of those who placed that river (the Hypanis)
in Asia." ' But Strabo had correctly pointed out that there
was a river Hypanis on the Asiatic side of the Euxine (the
modern Kuban) of the same name with that which fell into
the Borysthenes.® That any doubt should exist in the time of
Pliny with regard to the junction of the latter in the imme-
diate neighbourhood of such a flourishing commercial city as
Olbiopolis, is utterly impossible, and his confusion of ideas can
only be accounted for by the incoherent manner in which he
has brought together his multifarious authorities. In like
manner he introduces the rivers Hypacyris and Gerrhus, both
of which are found in Herodotus,* but as Mr. Eawlinson ob-

* Thus he names Prepesinthus — a
mere islet situated between Oliarus and
Siphnus, as if it lay between Seriphus
and Cythnus — and jumps from Myco-
nus to Siphnus, returning afterwards
to Oliarus, Paros, and Naxos.

° See Chapter VI. p. 185.

' " Quidam Panticapen confluere
infra Olbiam cum Borysthene tradunt,
diligentiores Hypanim, tanto errore
eorum qui ilium in Asise parte prodi-
dere." iv. 13, § 83.

« Strabo, xi. 2, § 9, p. 494-.

^ Herodot. iv. 55, 56. Not only does
Pliny introduce these obscure names,
which were certainly unknown in his
day, but he mentions the Hypacyris
twice over, once under the name of
Pacyris, and again under that of Hy-
pacaris, the form that is used by Mela
(ii. 1, § 4). Besides these be mentions
also two rivers, which he calls the
Acesinus and Buges, neitlier of which
can be identified. His " lacus Buges "
is apparently the Putrid Sea.

Sect. 2.

PLINY : DESCRIPTIVE GEOGEAPHY.

401

serves " defy identification with any existing stream :" and
certainly Pliny liad no better means of identifying them.
This part of his work indeed (like many others) does not
represent the geography of any period in particular, but is
a mere compilation mixed up of the past and present, and of
names huddled together without anything like a clear con-
ception of their position or geographical arrangement.

This is still more the case with the enumeration of the
Scythian tribes of the interior, where we find the names of
nations familiar to the Augustan age, such as the Geloni and
Agathyrsi, associated with others like the Thyssagetse and
Budini, which were known only from Herodotus, and had been
wholly ignored by Strabo and the other Greek geographers.
It is still more inexcusable that he not only includes in his
list the fabulous Arimaspians, but proceeds to give a full
account of the Ehipaean Mountains, and the region where the
air was perpetually filled with snow falling in great flakes like
feathers. Beyond this lay the land of the Hyperboreans, of
whom he gives a similar account to that of Mela, both in all
probability derived from the same source.^ He afterwards (as
already mentioned) crosses the Ehipaean mountains to the
Northern Ocean,^ and follows its shores westward towards Spain
and Gades. Of the nations in this part of Europe, and of the
islands that adjoined its shores he admits his almost entire
ignorance, but collects together a few scattered notices from
Greek writers of an immensely large island called by Xenophon
of Lampsacus Baltia and by Pytheas Basilia :^ of another called

> iv. 12, §§ 88-91. He indeed intro-
duces the account of the Hyperboreans
with an expression of doubt (si credi-
mus) ; but at the end adds that there
can be no doubt of their existence (nee
licet dubitare de gente ea), on account
of the fact, attested by many authors,
of their having sent sacred offerings to
Delos.

2 iv. 13, § 94. See above, p. 388.

^ " Xenophon Lampsacenus a littore
Scytharum tridui navigatione insulam

VOL. II.

esse immensse magnitudinis Baltiam
tradit, eandem Pytheas Basiliam nomi-
nat." iv. 13, § 95. But in another
passage (xxxvii. 2, § 35), where he
quotes more fully the statement of
Pytheas, he says that he called the
island Abalus, while Timseus gave it
the name of Basilia. So little can we
depend upon the accuracy of his
references.

The name of Baltia, which here
appears for the first time, is interesting

2 D

402

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

Kaunonia, where amber was cast up by tlie waves in spring :*
and others called Oonae where the inhabitants lived solely on
the eggs of sea-birds and oats, a description which has nothing
in it really marvellous, though it evidently appeared so to the
Greeks.

§ 11. The northern shores of Germany he tells us were
better known : but even here his information was really very
vague and imperfect, though we discern some glimmerings of
a better knowledge of this part of Europe. He mentions the
existence of a great bay to which he gives the name of Codanus
Sinus (a name we have already met with in Mela), which was
studded with large islands. One of these, called Scandinavia,
was of unknown extent, but so large as to be said by its inha-
bitants to form another quarter of the world.^ Another, named
Eningia, was supposed to be of equal extent. He was acquainted
with the Cimbrian Promontory and the manner in which it pro-
jected far to the north : but strangely connects this with a
range of mountains which he called Sevo, and describes as not
inferior to the Rhipsean mountains, and as forming the great
bay already referred to.^ Of course such a range had no real

as the origin of our modern term Baltic.
But the latter, as applied to the great
inland northern sea, was unknown to
the ancients.

* This name is apparently derived
from Timseus ; but the whole account
is very confused. The island intended
is in all probability the same as that
previously mentioned. All these
notices from the earlier Greek writers
point to a confusion between two dif-
ferent sets of traditions— both derived
from the amber traders to the Baltic ;
the one referring to the islands unme-
diately adjoining its southern coast,
where the amber was really found ; the
other conveying some vague notion of
immense islands to the north, including
probably the southern portion of the
Scandinavian peninsula.

^ This is the first mention in any
ancient writer of this now familiar
name. It appears, indeed, in some
editions of Mela, but is a mere arbi-

trary correction of the editors, substi-
tuted for "Codanovia," which is the
reading of the best MSS. See Parthey's
edition.

" "Mens Sevo ibi immensus nee
Eipseis jugis minor immanem ad Cim-
brorum usque promontorium eflScit
sinum, qui Codanus vocatur, refertus
insulis, quarum clarissima est Scandi-
navia, incomperta3 magnitudinis, por-
tionem tantum ejus, quod notum sit,
Hillevionum geute D incolente pagis,
qu£e alterum orbem terrarum earn
appellat : nee minor est opinione
Eningia." iv. 13, § 96.

The name of the Hilleviones is other-
wise unknown, unless they are to be
regarded as identical with the AfvUvoi
of Ptolemy. That of Eningia is also
found in no other writer; the con-
jecture that Finland is meant, is ex-
tremely far-fetched and improbable.
None of the names thus mentioned can
in fact be identified with any approach

Sect. 2.

PLINY : DESCRIPTIVE GEOGRAPHY.

403

existence, but it is curious that its mention by Pliny in this
passage is entirely isolated, and nothing corresponding to it is
found in any other author, except Solinus, who, as usual, simply
copies Pliny.^

His account of Germany in general is singularly defective
and scanty : especially when we consider that Pliny had him-
self served in that country, and had written an elaborate his-
tory of the wars of the Eomans with the Germans. The first
nation adjoining Sarmatia along the shores of the Baltic, he
tells us, were the Ingaevones, under which general appellation
he includes the Cimbri, Teutones and Chauci. He .must there-
fore have regarded them as occupying the whole of the north
of Germany, from the Vistula to the Weser. But he names
also the Vindili, whom he appears to place in the north-east of
;Germany ; the Istsevones nearer the Khine, and the Hermiones
in the interior of the country, to whom he assigns the well-
known tribes or nations of the Suevi, Hermunduri, Chatti and
Cherusci. But he gives us no statement at all of the position
or boundaries of these several nations : and merely enumerates
by name the great rivers which flow into the Ocean — the
Vistula, the Elbe (Albis), the Weser (Visurgis), the Ems
(Amisius), the Ehine and the Mouse. With these he asso-'
dates one obscure name, otherwise unknown, that of the
Guttalus, which he apparently places east of the Vistula, and
therefore not properly in Germany at all.^ He notices also

to certainty, or even probability, But
Pliny seems certainly to have had a
strong- impression of the existence of
extensive lands (which of course he
regarded as islands) in the northern
ocean. He elsewhere tells us (ii. 108,
§ 246), " Nam et a Germania immensas
insulas non pridem coguitas compertum
habeo." It is strange that he does not
seem to suspect their identity with
those vaguely mentioned by earlier
Greek writers, already referred to.
These were described by them as oppo-
site to the coast of Scythia, because all
their intercourse with the northern
ocean passed from the Euxine through

that country, while the Romans, who
heard of them through the Germans,
placed them opposite to the shores of
Germany.

' Solin. c. 20, § 1.

^ This would appear from the order
in which he enumerates them (iv. 13,
s. 28, § 100) : " Amnes clari in Oceanum
defluunt Guttalus, Vistillus sive Vistla,
Albis," etc. But Solinus, who as usual
copies Pliny, says : " de internis ejus
(Germanise) partibus Alba, Guthalus,
Vistla amnes latissimi prsecipitant in
Oceanum" (Solin. c. 20, § 2). He
therefore placed the Guttalus between-
the Elbe and the Vistula. It seems

2 D 2

404

HISTOEY OF ANCIENT GEOGBAPHY. Chap. XXIV.

the chain of islands extending along the coast of Germany
between the mouth of the Ehine and the Cimbrian Promontory,^
to one of which the name of Glesaria had been given by the
Eoman soldiers, on account of their having found amber there.-^
The name was by some writers extended to the whole group.

§ 12. Still more meagre and unsatisfactory is his notice of
the British Islands. Britain itself, or Albion as he considers
it ought more properly to be called, had in his time been in
great part subdued, but he contents himself with remarking
that " in thirty years the Eoman arms had not extended the
knowledge of it beyond the Caledonian Forests," ^ and then gives
its length and breadth according to Agrippa, in whose time it
was still comparatively unknown. He does not give the name
of a single people, town, or river: and as usual has no particulars
of its physical geography, natural productions, or the manners
of its inhabitants. Of Ireland (Hibernia) he tells us only that it
was about the same breadth as Britain, but two hundred miles
shorter : and adds that the shortest passage to it, from the land
of the Silures, was thirty miles. But while he thus gives us
absolutely no particulars as to the large and really important
islands, he enumerates a number of small ones which were
scattered around them, including the Orcades, the -^modse
(the Hsemodee of Mela), the Haebudes (the original form of the
word which has been perverted into the modern Hebrides) *

more probable that the name had been
misplaced by Pliny, and really referred
to the Oder, than that he had no notion
of that great river, and yet mentioned
the Pregel or any other obscure stream
east of the Vistula, with which the
Guttalus has been identified by Ger-
man writers. But it is strange that
no definite mention of the Oder is found
in any ancient geographer : its identi-
fication with the Viadus of Ptolemy
being very dubious.

' With this important feature of the
north coast of Germany he was well
acquainted. " Promontorium Oimbro-
rum excurrens in maria longe penin-
Bulam effioit," iv. 13, § 97.

* He states that there were in all
twenty-three of these islands, which
had been made known by the Roman
arms : among the most celebrated of
these were Burcana (see Chapter XX.)
and Glaesaria or Glessaria " a succino
niilitise appellata." It is clear that he
means to say the island was so called
by the soldiers of Germanicus because
they found amber there, which as he
elsewhere tells us (xxxvii. § 42) was
called by the Germans " gliesum " or
" glessum '' {i.e. Glas).

2 iv. 16, § 102.

* The form Hebrides is sanctioned
by one or two of the MSS. of Pliny, but
theformHebudcs(orH£ebudes),whichis

Sect. 3.

PLINY : GEOGRAPHY OF ASIA AND AFRICA.

405

and others which he enumerates by name as existing between
Britain and Ireland, among which we find (associated with Mona
and Monapia, corresponding to Anglesea and the Isle of Man)
the name of Vectis, unquestionably the Isle of Wight ! * The
most distant of all he tells us was Thule, his notice of which
has been already cited : and he then refers to the statement of
Timseus (already noticed) concerning an island called Mictis
from whence tin was brought.^ It is strange to find Pliny
still referring to an author like Timseus, who wrote more than
three centuries before, for an account of the British tin trade,
and either unable or else too careless to add any particulars
from later authorities. In common with most earlier writers he
connected the Cassiterides with Spain,® and no mention of this
celebrated name is found among the islands adjoining Britain.

Section 3. — Geography of Asia and Africa.

§ 1. Pliny's geography of Asia and Africa is in some
respects more interesting than his account of Europe. It is
indeed to a great extent characterized by the same defects,'
while his materials were on the whole very inferior. But it
nevertheless contains considerable additions to our informa-

found in all the best editions of Pliny,
and the MSS. of highest authority, is
strongly confirmed by Ptolemy, who
writes thename "EfiovSai ( Ptol . ii. 2, § 1 1 ).

* The name of Monapia first occurs
in Pliny, and must be unquestionably
identified with the Isle of Man ; though
the name of the latter would dispose
us at first to consider it as representing
Mona. But the Mona of the Komans,
which was attacked by Suetonius
Paulinus and Agricola, was certainly
Anglesea. Caesar, who places it mid-
way between Britain and Ireland (J5. G.
V. 13) probably confounded tlie two.

^ See Chapter XV. p. 603.

« iv. 22, s. 36, § 119. " Ex ad verso
Ccltiberia; complurea sunt insula3 Cas-

siterides dictse Grsecis a fertilitate
plumbi."

' A striking instance of these is
found in his description of Mount
Taurus (v. 27, §§ 97-99). Here he has
evidently taken up the idea of Era-
tosthenes of extending this name to
the whole chain of mountains, or rather
succession of chains, which extended
across Asia from west to east, from
Lycia to the Indian Ocean ; but he
has so disguised and disfigured this
by his rhetorical phrases and far-
fetched turns of expression as to be
much more calculated to confuse his
reader than to assist him in forming a
general notion of the physical structure
of Asia.

406 . ' HISTOKY OF ANCIENT GEOGEArHY. Chap. XXIV.

tion. For the provinces bordering on the Mediterranean he
had here again the advantages resulting from the Roman
administration, and hence he possessed the means of giving
a topographical review of the provinces of Northern Africa
from Mauretania to Egypt more complete and elaborate than
that of any former writer. In regard to Syria and Asia Minor
also he had ample means of information : and though in
respect to the latter country he wanted the advantage pos-
sessed by Strabo of extensive personal acquaintance, and has
failed (as usual) to give us any clear general outline of its
physical geography, he has, by the number of towns he enume-
rates, and. their arrangement under the different "conventus "
or juridical divisions, furnished us with important addi-
tions to our topographical knowledge. In this respect his
account of Asia Minor is decidedly superior to that of G-reece,
but it is disfigured to a great extent by the same accumulation
of obscure names, either of places that had long since dis-
appeared, or that had never been more than poetical or anti-
quarian . appellations for cities better known under their
ordinary names. In like manner his enumeration of the
islands that line the eastern coasts of the ^gean is a detailed
and minute list, in which he has sought to include every rock
or islet that bore a name, without any ; distinction as to their
geographical importance, and generally without any clear
indication of their position.^

§ 2. His account of Syria is one of the clearest and most
satisfactory portions of his work. While that country had
been brought wholly under the Roman administration and
was enjoying a state of the highest prosperity, the recent
wars of Vespasian and Titus in Judaea had attracted the
special attention of the Roman world to this quarter, and
Pliny doubtless enjoyed the advantage of excellent materials.

* It is a marked instance of the care- I among the islands ; but specially de-

lessness with which these lists are com- scribes it as such : " Clara vero in alto

piled that he not only includes Teos, Teos cum oppido " (v. 31, § .138).

a well-known city on the mainland, |

Sect. 3.

PLINY : GEOGEAPHY OF ASIA AND AFRICA.

407

Hence we find him not only giving us a correct, as well as
minute, description of the coast from the confines of Egypt to
the Gulf of Issus, but explaining clearly the peculiar con-
formation of the two parallel ridges of Lebanon and Anti-
Lebanon, with the rise of the Orontes between them :^ and in
like manner giving a full and accurate account of the course
of the Jordan, and of the two lakes that it formed, which he
calls Gennesara and Asphaltites.^ But he estimates the
length of the latter (the Dead Sea) at 100 Eoman miles, and
its greatest breadth at 25, both dimensions being just about
double the reality. He gives also a very exaggerated descrip-
tion of Mount Casius (Jebel Okra) near the mouth of the
Orontes, which he asserts to be so lofty that the rising sun
could be seen from its summit tliree hours before it was visible
from below.^ It was doubtless the isolated character of this
mountain, rising abruptly from the sea, that gave rise to the
notion of its great altitude, while its proximity to the cities of
Seleucia and Antioch led to its being frequently ascended.^
According to Pliny the ascent by the winding course neces-
sarily followed occupied nineteen miles, whilst its direct
height was only four.*

Pliny is also the first author who gives us any special notice
of Palmyra, which was in his time a place of considerable
importance. From its position between the two empires of
the Romans and Parthians it was an object of anxious interest
to both in time of war, but had as yet preserved its inde-

" V. 20, § 77 ; 22, § 80.

1 lb. 15,16, §§71,72.

^ " Super earn mons eodem quo alius
nomine, Casius, cujus excelsa altitude
quarta vigilia orientem per tenebras
solem aspicit, brevi circumactu corporis
diem noctemque pariter ostendens."
V. 22, § 80. The same thing is stated
by Aristotle of the Caucasus, a range
of a very different character (Meteorolog.
i. 13, § 18).

* Among others it was thus ascended
by the emperor Hadrian (Spartianua
Vit. Hadriani, c. 14),

* " Ambitus ad cacumen xix M. P.
est, altitude per directum iv." (Z. c.)
It is probable that he does not mean by
this the perpendicular height, but the
height supposed to be measured in a
direct line from the base to the summit.
The real elevation of Mount Casius is
only 5318 feet, while the highest
summit of the neighbouring Lebanon
attains to more than 10,000 feet. But
we have already seen how vague were
the notions of ancient writers con-
cerning the height of mountains in
general.

4o8

HISTORY OF ANCIENT GEOGRAPHY.

Chap. XXIV.

pendence, and carried on an extensive trade in time of peace.
He correctly describes it as surrounded on all sides by a broad
belt of sandy desert : and places it 176 Roman miles from
Damascus, which but little exceeds the triith.^

§ 3. His account of the two great rivers, the Euphrates and
the Tigris, is the fullest and most detailed that we find in any
ancient author. The former river especially, from its having
so long formed the boundary between the Roman and Parthian
empires, was become familiar to all, while recent wars in
Armenia had brought the Romans acquainted with the upper
part of its course, near its source. This is placed by Pliny,
on the authority of Domitius Corbulo, in a mountain called
Aba, in the " prsefectura " or district of Caranitis, while
Licinius Mucianus, also a contemporary writer, placed it at
the foot of a mountain called Capotes, 12 miles above a town
named Zimara.^ The first statement is unquestionably correct :
the district of Caranitis being that around the modern Erze-
roum (which was still called Karin in the middle ages) and
the sources of the northern or main branch of the Euphrates
being in fact only about 20 miles N.E. of that city. But
Pliny has no indication of the existence of the two great
arms, which are correctly regarded by the Armenian writers,
as combining to produce the main stream, a view adopted
from them by all modern geographers : he still, like Strabo
and other Greek writers, considered the northern arm only
(which still retains the name of Frat) as the true Euphrates,
and regarded the southern arm (the Murad Tchai) as a mere
affluent or tributary. It is almost certainly this river which
he designates under the name of Arsanias.^

In regard to the lower part of its course he had also very

* Plin. V. 25, § 88. He says of it,
" Palmira urbs nobilis sitn, divitiis soli
et aquis amcenis, vasto undique ambitu
arenis includit agros, ac velut terris
exemta a rerum natura, privata sorte
inter duo imperia summa Romanorum
Parthorumque, et prima in discordia
semper utrinque cura."

It did not pass under the Roman
yoke till the time of Trajan. The
earliest inscriptions are of the second
century.

« V. 24, § 83.

' V. 24, § 84; vi. 27, § 128. See
Chapter XXII. p. 289.

Sect. 3.

PLINY: GEOGRAPHY OF ASIA AND AFEICA.

409

correct information, stating distinctly that the two rivers,
the Euphrates and the Tigris, had formerly had separate
mouths, with an interval of 25 miles between them, but that
the mouth of the Euphrates had been blocked up in the
course of time, and its waters diverted for purposes of irri-
gation, what remained of them finding their way into the
Tigris and thence into the sea.^ A considerable portion of
them also was diverted at a higher point of their course, and
carried by a canal into the Tigris beneath the walls of
Seleucia. This must have nearly coincided with that still
known as the Nahr el Malcha, or royal canal.

The Tigris he describes, in accordance with the received
fable, as rising in Armenia, not far from the Euphrates,
flowing through two lakes, which he calls Arethusa and
Thospitis,^ and successively passing through two underground
channels, until it finally emerges, so near the river Arsanias
that their waters became commingled in times of inundation.
This tradition would seem to have reference to the branch
of the river that has its rise near Bitlis. But the notions
of ancient writers concerning the sources of the Tigris are
very confused, and those of Pliny are certainly no clearer
than the others.^ In describing the lower course of the river
it is remarkable that he does not notice its two important
tributaries, the Greater and Lesser Zab, known to the Greek
writers as Zabatus or Lycus.

8 V. 26, § 90 ; vi. 27, § 130. " Inter
duorum amnium ostia xxv M. P. fuere,
aut ut alii tiadunt "vii M. utroque
navigabili. Sed longo tempore Eu-
phraten prseclusere Orcheni et adcolse
agros rigantes, nee nisi per Tigrim
defertur in mare."

The Orcheni were one of the tribes
of the Chaldseans, celebrated for their
skill in astronomy (Strab. xvi. p. 739 ;
Plin. I. c. § 123) ; probably they were
skilled also in engineering, and hence
began the process of diverting the
waters of the Euphrates for purposes
of irrigation.

5 Of these, there can be no doubt
that the lake Thospitis is the same

with the Thopitis of Strabo (xi. p. 529),
which is certainly the great Lake of
Van ; but it is impossible to conjecture
what is the lake to which he has given
the purely Greek name of Arethusa.
The story of the passage of the Tigris
through these lakes (of course without
mixing its waters with them) is doubt-
less a mere fiction, but the lakes them-
selves must probably have had a real
existence ; unless the two names refer
to one and the same lake, which is not
impossible. Strabo, as we have seen,
while telling the same story, mentions
only one lake. (See Chapter XXII.
p. 289.)

1 See Note A, p. 439.

410

HISTOEY OF ANCIENT GEOGKAPHY. Chap. XXIV.

§ 4. With. Armenia and the neighbouring provinces Pliny ■
considered himself as better acquainted than any of his pre-
decessors, and boasts of the superior means of information that
he derived in regard to them from the campaigns of Domitius
Corbulo and other Eoman generals.^ But there is unfortunately
little evidence of the supposed superiority displayed in his
work. His description of Armenia itself is very concise and
summary, and he does not even notice in this place the lakes
which form so important a feature in its physical geography,
and of the existence of which he was fully aware. But his
ideas concerning the nations to the north of it, the Albanians
and Iberians, which he describes as extending to the foot of the
Caucasus, were tolerably clear and distinct; and even his
names for the wild tribes of that mountain range seem to be
derived from good authority, as we recognize among them that
of the Suani, which still gives name to the valley of Suanetia.^
He gives a detailed description of the remarkable pass of
Dariel, to which he gives the name of the Caucasian Gates
(Portse Caucasise), and takes credit to himself for pointing out
the error committed by many writers, who confounded them
with the Caspian Gates, which bore so important a part in the
works of Greek geographers.* But if this error was really pre-
valent in the time of Pliny, it was introduced by the Eomans
who had been engaged in wars in these countries, for, as we
have seen, Strabo was already well acquainted with the pass
through the Caucasus, though he did not give to it the name
of the Caucasian Gates.^

Of the countries beyond the Caucasus towards the north

^ vi. 8, § 23. See above, Chapter
XXIII. p. 346.

2 vi. 4, § 14; ii. § 30. In the former
passage he correctly describes the river
Cobus as flowing from the Caucasus
through the Suani into the land of the
Colchians. This is clearly the river
how known as the Ingur.

It is a striking instance how often
close resemblances of name may bo
merely accidental, when not supported

by other evidence, that while the name
of Suaiietia occurs in the Caucasus, that
of the Suanetes is found in the list of
the Alpine tribes given by Pliny from
the monument of Augustus (Plin.
H. N. iii. 20, § 137).

* "Ab his sunt Portao Caucasife,
magno errore multis Caspios dictije." vi.
11, §30.

^ See Chapter XXII. p. 281.

Sect. 3.

PLINY: GEOGRAPHY OF ASIA AND AFRICA.

411

Pliny had evidently nothing but the vaguest idea, and in
regard to the Caspian Sea he still adhered to the erroneous
notion of its being a mere inlet of the northern ocean, as the
Persian Gulf was from the south, though communicating with
it only by a long and narrow strait.® In regard to its dimen-
sions and extent he follows the authority of Eratosthenes and
Artemidorus, adding, however, a statement from Agrippa,
which is not very intelligible. But his description of this
sea, and still more of the nations to the east of it, is very super-
ficial and confused, and he certainly appears to have had no
clear ideas on the subject. All this part of his work is im-
measurably inferior to the description of the same countries
given by Strabo: even the great river Oxus obtains only a
passing notice, with no indication of its importance or general
course, and the additional statement that it took its rise in a
lake also called Oxus was probably a mere conjecture.'^ On the
other hand we are indebted to him for the interesting notice
which he has preserved to us from Varro, of the course taken in
the time of Pompey by the overland trade from India to the
Caspian. The merchandise conveyed by this route passed in
seven days from India (Cabul ?) into Bactria, to the river
Icarus, which flowed into the Oxus, and was carried down that
river into the Caspian, from whence it passed up the Cyrus,
and thence by overland transport of only five days into the
Phasis, and so into the Euxine.^ But of the overland trade from

" " Irrumpit autem arctis faucibus et
in longitudinem spatiosis," vi. ] 3, § 38.
His knowledge on this subject was
therefore not in advance of that of
Mela,

^ It is true that the Oxus actually
does take its rise in the lake called
Sir-i-Kol, on the central plateau of the
Pamir (Wood's Journey to the Source of
the Oxus, p. 232), but it is most unlikely
that Pliny should have any real infor-
mation concerning this secluded little
mountain lake, while his knowledge on
all other points was so very meagre.
It was a common practice with ancient
geographers to assume the existence of
a lake as tlie source of a river, of which

they had no real knowledge.

With regard to the laxartes he
refers to a certain Demodamas, as his
principal authority for these regions
(" transcendit eum amnem Demodamas,
Seleuci et Antiochi regum dux, quern
maxime sequimur in iis," vi. 16, § 49),
but nothing further is known of such
an author.

* vi. 17, § 52. The river Icarus is
otherwise unknown, and the name pro-
bably corrupt. But the river meant
must be one of the streams which iiow
northwards from the Hindoo-Koosh
through Badakshan, perhaps tlie Sur-
khund, or river of Koondooz.

412

HISTORY OF ANCIENT GEOGRAPHY. Chap. XXIV.

Bactria to China, which had attained such an important charac-
ter in the time of Ptolemy, and even in that of Marinus of
Tyre, no trace is found in Pliny.

The only exception to the imperfect and perfunctory charac-
ter of his account of this part of Asia is in regard to the little
isolated district of Margiana (Merv), of which he gives a some-
what full and characteristic description, extolling its fertility
and especially its abundance of vines, which is also noted by
Strabo. It had attracted attention among the Eomans from its
being the place to which the Parthian king Orodes had trans-
ported the Eoman captives taken after the defeat of Crassus.^

§ 5. Of the extensive provinces, which were comprised within
the Parthian Empire, as they had formerly been in that of the
Persians, including the whole of the vast table-land of Iran and
the adjoining regions, Pliny appears to have had no knowledge
beyond what he derived from the Greek writers whom we have
already considered. This is the more remarkable, as he cer-
tainly appears to have made great use of the work of Isidorus
of Charax, who from the situation of his birthplace, was likely
to have had good means of information, while he is said to have
written a work expressly devoted to the geography of Parthia.^
We are indebted to Pliny indeed for the statement that Parthia
included eighteen subordinate " kingdoms " — a title which
they seem to have employed as equivalent to what the Persians
had termed satrapies f but unfortunately he does not enume-
rate them, contenting himself with stating that eleven of them
were called the " upper " provinces, occupying the northern
portion of the empire, and seven the " lower."^

Of the original home of the Parthians — the district of Par-

9 Plin. vi. 16, § 47.

» See Chapter XX. p. 163. The
unusually circumstantial account given
by Pliny of the position and history of
Charax (vi. 27, §§ 128-130), a city
which is not mentioned by any earlier
writer, must undoubtedly have been
derived from the same authority. It
appears to have been in his time an
important place of trade.

- Hence was doubtless derived the
proud title assumed by the Parthian
monarchs on their coins of "King of
Kings " {fiafftXevs fiaffiKewv). The real
designation of these viceroys appears to
have been Vitaxse or Bistaxre, a native
term which is preserved to us by Am-
mianus Marcellinus (xxiii. 6, § 14) and
Hesychius (v. Biara^).

3 vi. 25, § 112.

Sect. 3.

PLINY: GEOGRAPHY OF ASIA AND AFRICA.

413

thia Proper — he has given a brief, but very correct, account ;*
describing it as situated at the foot of the great mountain
chain which borders all these provinces on the north, and was
regarded by the ancients as a continuation of the Taurus, con-
necting that range with the Paropamisus or Hindoo Koosh to
the east. It was bounded on the east by the Arians, on the
south by Carmania and the Ariani(?), on the west by the
Modes, called Pratitae, and on the north by the Hyrcanians ;
but he adds that it was surrounded on all sides by deserts,
which is an exaggeration, though on the north and south it
certainly adjoined the great deserts, of Kharesm on the one
hand, and that of Central Iran on the other.

§ 6. Of the other provinces of Upper Asia in general Pliny's
account is very meagre and unsatisfactory, and cannot be said
to add anything to our knowledge. Even of such well-known
countries as Persia (Proper) and Media, his notices are at once
brief and confused : while his statement that Ecbatana — the
celebrated capital of Media — was founded by king Seleucus,® is
a strong instance how little reliance can be placed upon such
notices in his work, and how cautious we should be in adopting
them where we have not the means of correcting them from
other sources. Such is also the case with his assertion in
another place, that Susa was founded by Darius the son of
Hystaspes ; a statement which may probably be correct with
regard to the royal palace there, but, as we learn from recent
researches, the city itself is of much greater antiquity.®

His enumeration of the Scythian tribes to the north of the
Parthian Empire is, as usual, a mere collection of names of the
most miscellaneous description, derived from a variety of
sources, and strung together without discrimination, or any

* Ibid. § 113.

^ " Ecbatana caput Medise Seleucus
rex condidit," vi. 14, § 43. His state-
ment in another passage (ib. 26, § 116)
that it was a city of the Magi, which
was transferred by Darius to the moun-
tains (Horum [Magorum] Ecbatana
oppidum translatum ab Dario rege ad

montes), would seem to refer to some
other place of the name connected with
Persia proper ; but the whole passage
is so confused that no reliance can be
safely placed on it.

^ See Loftus's Chaldeea and Susiana,
ch. 26.

414

HISTOEY OP ANCIENT GEOGRAPHY. Chap. XXIV.

attempt at geograpliical arrangement. With regard to the
unknown eastern shores of Asia, his account almost precisely
agrees with that of Mela, and is evidently derived from the
same source.^ But in respect to the Seres, whom he places, in
common with the earlier geographer, on the Eastern Ocean,
between a promontory called Tabis * and the confines of India, he
furnishes us with some further particulars, and not only notices
their production of silk, which he describes, like Virgil, as
combed off the leaves of tree s,^ but he mentions several rivers
and other geographical names in connection with their country,
which show a certain dawning of a better acquaintance with it.^
He places also in this part of Asia a people called the Attacori,
whom he describes as resembling the Hyperboreans in the
excellence of their climate, and having been made the subject
of a special treatise by a Greek writer named Amometus,
similar to that of Hecatseus concerning the Hyperboreans.
Both races were evidently equally fabulous.^

§ 7. Pliny now proceeds to the description of India, a
country which, as we have seen, had within his own time been

' vi. 17, § 53.

® This promontory, mentioned by
Mela also under the same name, is
almost certainly the same with the
Tamarus of Eratosthenes and Strabo,
the supposed eastern termination of tiie
chain of Taurus, which had no real
existence.

^ " Seres lanicio silvarum nobiles,
perfusam aqua depectentes frondium
canitiem, imde geminus feminis nostris
labor, redordiendi fila rursusque tex-
endi." vi. 17, § 54. From this passage
it appears thiit notwithstanding his
pretensions as a naturalist he was not
at all in advance of Virgil as to the
nature of silk and the manner of ob-
taining it. See the passage already
quoted in Chapter XX. p. 166 {Georg.
ii. 121). It appears that the Romans in
the time of Pliny only valued silk tex-
tures of a very thin gauze-like charac-
ter ; and hence when they received them
from China took the trouble of unravel-
ling them and weaving them again into

a kind of gauze. That this is the
sense of the words above cited (which
indeed can scarcely admit of any other)
is fully proved by those that follow:
" Tarn multiplici opere, tarn longinquo
orbe petitur, ut in publico matrona
transluceat." The " Sericse vestes "
were evidently regarded by tiiem as a
kind of improvement on the Coan tex-
tures, of which Horace makes almost
exactly the same remark {Satir. i. 2,
s. 101. See also Seneca, de Benefic. 7,
§9).

In another passage (xxxiv. 14,
§ 145) he speaks of the Seres as send-
ing furs and ii-on, in addition to their
silk dresses. The latter he considers
as superior to all other iron in quality.

* None of these names is otherwise
mentioned or can be identified. Indeed
in the total misconception of the geo-
graphy of all this part of Asia which
prevailed in the time of Pliny, the
attempt to do so would be absurd.

- vi. 17, § 55.

Sect. 3. PLINY : GEOGRAPHY OF ASIA AND AFRICA. 415

very much opened out by the extension of the Eoman com-
merce, and concerning which he really possessed important
additional information. No other part of his work indeed
displays so much advance upon the knowledge of his prede-
cessors. Yet even here he begins with a statement of its
geographical position and dimensions which he derives from
Eratosthenes,^ and his account of the northern portions of India,
and the interior from the Indus to the Ganges, is taken wholly
from writers of the time of Alexander or that of his successors.
His detailed statement of the distances from the Indus to the
mouth of the Ganges has been already examined ; * and with
all its imperfections is a valuable contribution to our know-
ledge, but it dates also from the same period.

With regard to the Ganges itself, it is remarkable that his
information is in some respects far less precise and definite than
that of Strabo, as that writer was well acquainted with its
sources in the Emodi Montes, while Pliny tells us that some
regarded its sources as unknown, like those of the Nile : others
said that it took its rise in the Scythian mountains — an
extremely vague designation. He quotes also another account
of its source, as breaking out at once in a violent cascade with
a loud noise, and gradually lapsing down into a gentle and
placid stream, but without any indication of the geographical
site to which this was referred.^ He adds, that it received as
tributaries nineteen other rivers, among which he notices as
navigable the lomanes, the Prinas, the Cainas, the Condochates,
Erannoboas, the Cosoagus, and the Sonus. Here we meet
with another instance of the confusion so common in Pliny,
the Erannoboas and the Sonus being, as we have already seen,
only two names for the same river.^

Contrary to his usual practice, Pliny gives various particulars

^ E. N. vi. 17, § 57. With this | may be observed that Pliny himself, a
statement Pliny compares that given
b