This story is repeated in great and varying detail in sundry books by Afghans, the oldest of which appears to be of the 16th century; nor do we know that any trace of the legend is found of older date. In the version given by Major Raverty (Introd. to Afghan Grammar), Afghanah is settled by King Solomon himself in the Sulimani mountains; there is nothing about Nebuchadrezzar or Ghor. The historian Ferishta says he had read that the Afghans were descended from Copts of the race of Pharaoh. And one of the Afghan histories, quoted by Mr Bellow, relates "a current tradition'' that, previous to the time of Kais, Bilo the father of the Biluchis, Uzbak (evidently the father of the Usbegs) and Afghana were considered as brethren. As Mahommed Usbeg Khan, the eponymus of the medley of Tatar tribes called Usbegs, reigned in the 14th century A.D., this gives some possible light on the value of these so-called traditions.

We have analogous stories in the literature of almost all nations that derive their religion or their civilization from a foreign source. To say nothing of the Book of Mormon, a considerable number of persons have been found to propagate the doctrine that the English people are descended from the tribes of Israel. But the Hebrew ancestry of the Afghans is more worthy at least of consideration, for a respectable number of intelligent officers, well acquainted with the Afghans, have been strong in their belief of it; and though the customs alleged in proof will not bear the stress laid on them, undoubtedly a prevailing type of the Afghan physiognomy has a character strongly Jewish. This characteristic is certainly a remarkable one; but it is shared, to a considerable extent, by the Kashmiris (a circumstance which led Bernier to speculate on the Kashmiris representing the lost tribes of Israel), and, we believe, by the Tajik people of Badakshan.

Relations with the Greeks.—-In the time of Darius Hystaspes (500 B.C.) we find the region now called Afghanistan embraced in the Achaemenian satrapies, and various parts of it occupied by Sarangians (in Seistan), Arians (in Herat), Sattagydians (supposed in highlands of upper Helmund and the plateau of Ghazni), Dadicae (suggested to be Tajiks), Aparytae (mountaineers, perhaps of Safed Koh, where lay the Paryetae of Ptolemy), Gandarii (in Lower Kabul basin) and Paktyes, on or near the Indus. In the last name it has been plausibly suggested that we have the Pukhtun, as the eastern Afghans pronounce their name. Indeed, Pusht, Pasht or Pakht would seem to be the oldest name of the country of the Afghans in their traditions.

The Ariania of Strabo corresponds generally with the existing dominions of Kabul, but overpasses their limits on the west and south.

About 310 B.C. Seleucus is said by Strabo to have given to the Indian Sandrocottus (Chandragupta), in consequence of a marriage-contract, some part of the country west of the Indus occupied by an Indian population, and no doubt embracing a part of the Kabul basin. Some sixty years later occurred the establishment of an independent Greek dynasty in Bactria. (See BACTRIA, MEDIA, EUCRATIDES, MENANDER of India, EUTHYDEMUS, and PERSIA, Ancient History.) Of the details of their history and extent of their dominion in different reigns we know almost nothing, and conjecture is often dependent on such vague data as are afforded by the collation of the localities in which the coins of independent princes have been found. But their power extended certainly over the Kabul basin, and probably, at times, over the whole of Afghanistan. The ancient architecture of Kashmir, the tope of Manikyala in the Punjab, and many sculptures found in the Peshawar valley, show unmistakable Greek influence. Demetrius (c. 190 B.C.) is supposed to have reigned in Arachosia after being expelled from Bactria, much as, at a later date, Baber reigned in Kabul after his expulsion from Samarkand. Eucratides (181 B.C.) is alleged by Justin to have warred in India. With his coins, found abundantly in the Kabul basin, commences the use of an Arianian inscription, in addition to the Greek, sulnosed to imply the transfer of rule to the south of the mountains, over a people whom the Greek dynasty sought to conciliate. Under Heliocles (147 B.C.?), the Parthians, who had already encroached on Ariana, pressed their conquests into India. Menander (126 B.C.) invaded India at least to the Jumna, and perhaps also to the Indus delta. Tbe coinage of a succeeding king, Hermaeus, indicates a barbaric irruption. There is a general correspondence between classical and Chinese accounts of the time when Bactria was overrun by Scythian invaders. The chief nation among these, called by the Chinese Yue-Chi, about 126 B.C. established themselves in Sogdiana and on the Oxus in five hordes. Near the Christian era the chief of one of these, which was called Kushan, subdued the rest, and extended his conquests over the countries south of the Hindu Kush, including Sind as well as Afghanistan, thus establishing a great dominion, of which we hear from Greek writers as Indo-Scythia. (See YUE-CHI.)

Buddhism had already acquired influence over the people of the Kabul basin, and some of the barbaric invaders adopted that system. Its traces are extensive, especially in the plains of Jalalabad and Peshawar, but also in the vicinity of Kabul.

Various barbaric dynasties succeeded each other. A notable monarch was Kanishka (see INDIA, History) or Kanerkes, whose date is variously fixed at from 58 B.C. to A.D. 125, and whose power extended over the upper Oxus basin, Kabul, Peshawar, Kashmir and probably far into India. His name and legends still filled the land, or at least the Buddhist portion of it, 600 years later, when tho Chinese pilgrim, Hsuan Tsang, travelled in India; they had even reached the great Mahommedan philosopher, traveller and geographer, Abu-r-Raihan Muhammad al-Biruni (see BIRUNI), in the 11th century; and they are still celebrated in the Mongol versions of Buddhist ecclesiastical story.

Turkoman Dynasties.—-In the time of Hsuan Tsang (A.D. 630-645) there were both Indian and Turk princes in the Kabul valley, and in the succeeding centuries both these races seem to have predominated in succession. The first Mahommedan attempts at the conquest of Kabul were unsuccessful, though Seistan and Arachosia were permanently held from an early date. It was not till the end of the 10th century that a Hindu prince ceased to reign in Kabul, and it fell into the hands of the Turk Sabuktagin, who had established his capital at Ghazni. There, too, reigned his famous son Mahmud, and a series of descendants, till the middle of the 12th century, rendering the city one of the most splendid in Asia. We then have a powerful dynasty, commonly believed to have been of Afghan race; and if so, the first. But the historians give them a legendary descent from Zohak, which is no Afghan genealogy. The founder of the dynasty was Alauddin, chief of Ghor, whose vengeance for the cruel death of his brother at the hands of Bahram the Ghaznevide was wreaked in devastating the great city. His nephew, Shahabuddin Mahommed, repeatedly invaded India, conquering as far as Benares. His empire in India indeed—ruled by his freedmen who after his death became independent —may be regarded as the origin of that great Mahommedan monarchy which endured nominally till 1857. For abrief period the Afghan countries were subject to the king of Khwarizm, and it was here chiefly that occurred the gallant attempts of Jalaluddin of Khwarizm to withstand the progress of Jenghiz Khan.

A passage in Ferishta seems to imply that the Afghans in the Sulimani mountains were already known by that name in the first century of the Hegira, but it is uncertain how far this may be built on. The name Afghans is very distinctly mentioned in 'Utbi's History of Sultan Mahmud, written about A.D. 1030, coupled with that of the Khifjis. It also appears frequently in connexion with the history of India in the 13th and 14th centuries. The successive dynasties of Delhi are generally called Pathan, but were really so only in part. Of the Khifjis (1288-1321) we have already spoken. The Tughlaks (1321-1421) were originally Tatars of the Karauna tribe. The Lodis (1450-1526) were pure Pathans. For a century and more after the Mongol invasion the whole of the Afghan countries were under Mongol rule; but in the middle of the 14th century a native dynasty sprang up in western Afghanistan, that of the Kurts, which extended its rule over Ghor, Herat and Kandahar. The history of the Afghan countries under the Mongols is obscure; but that regime must have left its mark upon the country, if we judge from the occurrence of frequent Mongol names of places, and even of Mongol expressions adopted into familiar language.

The Mogul Dyniasty.—-All these countries were included in Timur's conquests, and Kabul at least had remained in the possession of one of his descendants till 1501, only three years before it fell into the hands of another and more illustrious one, Sultan Baber. It was not till 1522 that Baber succeeded in permanently wresting Kandahar from the Arghuns, a family of Mongol descent, who had long held it. From the time of his conquest of Hindustan (victory at Panipat, April 21, 1526), Kabul and Kandahar may be regarded as part of the empire of Delhi under the (so-called) Mogul dynasty which Baber founded. Kabul so continued till the invasion of Nadir Shah (1738). Kandahar often changed hands between the Moguls and the rising Safavis (or Sufis) of Persia. Under the latter it had remained from 1642 till 1708, when in the reign of Husain, the last of them, the Ghilzais, provoked by the oppressive Persian governor Shahnawaz Khan (a Georgian prince of the Bagratid house), revolted under Mir Wais, and expelled the Persians. Mir Wais was acknowledged sovereign of Kandahar, and eventually defeated the Persian armies sent against him. but did not long survive (d. 1715).

Mahmud, the son of Mir Wais, a man of great courage and energy, carried out a project of his father's, the conquest of Persia itself. After a long siege, Shah Husain came forth from Ispahan with all his court, and surrendered the sword and diadem of the Sufis into the hands of the Ghilzai (October 1722). Two years later Mahmud died mad, and a few years saw the end of Ghilzai rule in Persia.

The Durani Dynasty.—-In 1737-38 Nadir Shah both recovered Kandahar and took Kabul. But he gained the goodwill of the Afghans, and enrolled many in his army. Among these was a noble young soldier, Ahmad Khan, of the Saddozai family of the Abdali clan, who after the assassination of Nadir (1747) was chosen by the Afghan chiefs at Kandahar to be their leader, and assumed kingly authority over the eastern part of Nadir's empire, with the style of Dur-i-Duran, "Pearl of the Age,'' bestowing that of Durani upon his clan, the Abdalis. With Ahmad Shah, Afghanistan, as such, first took a place among the kingdoms of the earth, and the Durani dynasty, which he founded, still occupies its throne. During the twenty-six years of his reign he carried his warlike expeditions far and wide. Westward they extended nearly to the shores of the Caspian; eastward he repeatedly entered India as a conqueror. At his great battle of Panipat (January 6, 1761), with vastly inferior numbers, he inflicted on the Mahrattas, then at the zenith of their power, a tremendous defeat, almost annihilating their vast army; but the success had for him no important result. Having long suffered from a terrible disease, he died in 1773, bequeathing to his son Timur a dominion which embraced not only Afghanistan to its utmost limits, but the Punjab, Kashmir and Turkestan to the Oxus, with Sind, Baluchistan and Khorasan as tributary governments.

Timur transferred his residence from Kandahar to Kabul, and continued during a reign of twenty years to stave off the anarchy which followed close on his death. He left twenty-three sons, of whom the fifth, Zaman Mirza, by help of Payindah Khan, head of the Barakzai family of the Abdalis, succeeded in grasping the royal power. For many years barbarous wars raged between the brothers, during which Zaman Shah, Shuja-ul-Mulk and Mahmud successively held the throne. The last owed success to Payindah's son, Fatteh Khan (known as the "Afghan Warwick''), a man of masterly ability in war and politics, the eldest of twenty-one brothers, a family of notable intelligence and force of character, and many of these he placed over the provinces. Fatteh Khan, however, excited the king's jealously by his powerful position, and provoked the malignity of the king's son, Kamran, by a gross outrage on the Saddozai family. He was accordingly seized, blinded and afterwards murdered with prolonged torture, the brutal Kamran striking the first blow.

The Barakzai brothers united to avenge Fatteh Khan. The Saddozais were driven from Kabul, Ghazni and Kandahar, and with difficulty reached Herat (1818). Herat remained thus till Kamran's death (1842), and after that was held by his able and wicked minister Yar Mahommed. The rest of the country was divided among the Barakzais—-Dost Mahommed, the ablest, getting Kabul. Peshawar and the right bank of the Indus fell to the Sikhs after their victory at Nowshera in 1823. The last Afghan hold of the Punjab had been lost long before.—Kashmir in 1819; Sind had cast off all allegiance since 1808; the Turkestan provinces had been practically independent since the death of Timur Shah.

The First Afghan War, 1838-42.—-In 1809, in consequence of the intrigues Of Napoleon in Persia, the Hon. Mountstuart Elphinstone had been sent as envoy to Shah Shuja, then in power, and had been well received by him at Peshawar. This was the first time the Afghans made any acquaintance with Englishmen. Lieut. Alex. Burnes (afterwards Sir Alex. Burnes) visited Kabul on his way to Bokhara in 1832. In 1837 the Persian siege of Herat and the proceedings of Russia created uneasiness, and Burnes was sent by the governor-general as resident to the amir's court at Kabul. But the terms which the Dost sought were not conceded by the government, and the rash resolution was taken of re-establishing Shah Shuja, long a refugee in British territory. Ranjit Singh, king of the Punjab, bound himself to co-operate, but eventually declined to let the expedition cross his territories.

The war began in March 1838, when the "Army of the Indus,'' amounting to 21,000 men, assembled in Upper Sind and advanced through the Bolan Pass under the command of Sir John Keane. There was hardship, but scarcely any opposition. Kohandil Khan of Kandahar fled to Persia. That city was occupied in April 1839, and Shah Shuja was crowned in his grandfather's mosque. Ghazni was reached 21st July; a gate of the city was blown open by the engineers (the match was fired by Lieut., afterwards Sir Henry, Durand), and the place was taken by storm. Dost Mahommed, finding his troops deserting, passed the Hindu Kush, and Shah Shuja entered the capital (August 7). The war was thought at an end, and Sir John Keane (made a peer) returned to India with a considerable part of the force, leaving behind 8000 men, besides the Shah's force, with Sir W. Macnaghten as envoy, and Sir A. Burnes as his colleague.

During the two following years Shah Shuja and his allies remained in possession of Kabul and Kandahar. The British outposts extended to Saighan, in the Oxus basin, and to Mullah Khan, in the plain of Seistan. Dost Mahommed surrendered (November 5, 1840) and was sent to India, where he was honourably treated. From the beginning, insurrection against the new government had been rife. The political authorities were overconfident, and neglected warnings. On the 2nd of November 1841 the revolt broke out violently at Kabul, with the massacre of Burnes and other officers. The position of the British camp, its communications with the citadel and the location of the stores were the worst possible; and the general (Elphinstone) was shattered in constitution. Disaster after disaster occurred, not without misconduct. At a conference (December 23) with the Dost's son, Akbar Khan, who had taken the lead of the Afghans, Sir W. Macnaghten was murdered by that chief's own hand. On the 6th of January 1842, after a convention to evacuate the country had been signed, the British garrison, still numbering 4500 soldiers (of whom 690 were Europeans), with some 12,000 followers, marched out of the camp. The winter was severe, the troops demoralised, the march a mass of confusion and massacre, and the force was finally overwhelmed in the Jagdalak pass between Kabul and Jalalabad.

On the 13th the last survivors mustered at Gandamak only twenty muskets. Of those who left Kabul, only Dr Brydon reached Jalalabad, wounded and half dead. Ninety-five prisoners were afterwards recovered. The garrison of Ghazni had already been forced to surrender (December 10). But General Nott held Kandahar with a stern hand, and General Sale, who had reached Jalalabad from Kabul at the beginning of the outbreak, maintained that important point gallantly.

To avenge these disasters and recover the prisoners preparations were made in India on a fitting scale; but it was the 16th of April 1842 before General Pollock could relieve Jalalabad, after forcing the Khyber Pass. After a long halt there he advanced (August 20), and gaining rapid successes, occupied Kabul (September 15), where Nott, after retaking and dismantling Ghazni, joined him two days later. The prisoners were happily recovered from Bamian. The citadel and central bazaar of Kabul were destroyed, and the army finally evacuated Afghanistan, December 1842.

This ill-planned and hazardous enterprise was fraught with the elements of inevitable failute. A ruler imposed upon a free people by foreign arms is always unpopular; he is unable to stand alone; and his foreign auxiliaries soon find themselves obliged to choose between remaining to uphold his power, or retiring with the probability that it will fall after their departure. The leading chiefs of Afghanistan perceived that the maintenance of Shah Shuja's rule by British troops would soon be fatal to their own power and position in the country, and probably to their national independence. They were insatiable in their demands for office and emolument, and when they discovered that the shah, acting by the advice of the British envoy, was levying from among their tribesmen regiments to be directly under his control, they took care that the plan should fail. Without a regular revenue no effective administration could be organized; but the attempt to raise taxes showed that it might raise the people, so that for both men and money the shah's government was still obliged to rely principally upon British aid. All these circumstances combined to render the new regime weak and unpopular, since there was no force at the ruler's command except foreign troops to put down disorder or to protect those who submitted, while the discontented nobles fomented disaffection and the inbred hatred of strangers in race and religion among the general Afghan population.

British and Russian Relalions.—It has been said that the declared object of this policy had been to maintain the independence and integrity of Afghanistan, to secure the friendly alliance of its ruler, and thus to interpose a great barrier of mountainous country between the expanding power of Russia in Central Asia and the British dominion in India. After 1849, when the annexation of the Punjab had carried the Indian northwestern frontier up to the skirts of the Afghan highlands, the corresponding advance of the Russians south-eastward along the Oxus river became of closer interest to the British, particularly when, in 1856, the Persians again attempted to take possession of Herat. Dost Mahommed now became the British ally, but on his death in 1863 the kingdom fell back into civil war, until his son, Shere Ali, had won his way to undisputed rulership in 1868. In the same year Bokhara became a dependency of Russia. To the British government an attitude of non-intervention in Afghan affairs appeared in this situation to be no longer possible. The meeting between the amir Shere Ali and the viceroy of India (Lord Mayo) at Umballa in 1869 drew nearer the relations between the two governments; the amir consolidated and began to centralize his power; and the establishment of a strong, friendly and united Afghanistan became again the keynote of British policy beyond the north-western frontier of India.

When, therefore, the conquest of Khiva in 1873 by the Russians, and their gradual approach towards the amir's northern border, had seriously alarmed Shere Ali, he applied for support to the British; and his disappointment at his failure to obtain distinct pledges of material assistance, and at Great Britain's refusal to endorse all his claims in a dispute with Persia over Seistan, so far estranged him from the British connexion that he began to entertain amicable overtures from the Russian authorities at Tashkend. In 1869 the Russian government had assured Lord Clarendon that they regarded Afghanistan as completely outside the sphere of their influence; and in 1872 the boundary line of Afghanistan on the north-west had been settled between England and Russia so far eastward as Lake Victoria.

Nevertheless the correspondence between Kabul and Tashkend continued, and as the Russians were now extending their dominion over all the region beyond Afghanistan on the north-west, the British government determined, in 1876, once more to undertake active measures for securing their political ascendancy in that country. But the amir, whose feelings of resentment had by no means abated, was now leaning toward Russia, though he mainly desired to hold the balance between two equally formidable rivals. The result of overtures made to him from India was that in 1877, when Lord Lytton, acting under direct instructions from Her Majesty's ministry, proposed to Shere Ali a treaty of alliance, Shere Ali showed himself very little disposed to welcome the offer; and upon his refusal to admit a British agent into Afghanistan the negotiations finally broke down.

Second Afghan War, 1878-80.—In the course of the following year (1878) the Russian government, to counteract the interference of England with their advance upon Constantinople, sent an envoy to Kabul empowered to make a treaty with the amir. It was immediately notified to him from India that a British mission would be deputed to his capital, but he demurred to receiving it; and when the British envoy was turned back on the Afchan frontier hostilities were proclaimed by the viceroy in November 1878, and the second Afghan War began. Sir Donald Stewart's force, marching up through Baluchistan by the Bolan Pass, entered Kandahar with little or no resistance; while another army passed through the Khyber Pass and took up positions at Jalalabad and other places on the direct road to Kabul. Another force under Sir Frederick Roberts marched up to the high passes leading out of Kurram into the interior of Afghanistan, defeated the amir's troops at the Peiwar Kotal, and seized the Shutargardan Pass which commands a direct route to Kabul through the Logar valley. The amir Shere Ali fled from his capital into the northern province, where he died at Mazar-i-Sharif in February 1879. In the course of the next six months there was much desultory skirmishing between the tribes and the British troops, who defeated various attempts to dislodge them from the positions that had been taken up; but the sphere of British military operations was not materially extended. It was seen that the farther they advanced the more difficult would become their eventual retirement; and the problem was to find a successor to Shere Ali who could and would make terms with the British government.

In the meantime Yakub Khan, one of Shere Ali's sons, had announced to Major Cavagnari, the political agent at the headquarters of the British army, that he had succeeded his father at Kabul. The negotiations that followed ended in the conclusion of the treaty of Gandamak in May 1879, by which Takub Khan was recognized as amir; certain outlying tracts of Afghanistan were transferred to the British government; the amir placed in its hands the entire control of his foreign relations, receiving in return a guarantee against foreign aggression; and the establishment of a British envoy at Kabul was at last conceded. By this convention the complete success of the British political and military operations seemed to have been attained; for whereas Shere Ali had made a treaty of alliance with, and had received an embassy from Russia, his son had now made an exclusive treaty with the British government, and had agreed that a British envoy should reside permanently at his court. Yet it was just this final concession, the chief and original object of British policy, that proved speedily fatal to the whole settlement. For in September the envoy, Sir Louis Cavagnari, with his staff and escort, was massacred at Kabul, and the entire fabric of a friendly alliance went to pieces. A fresh expedition was instantly despatched across the Shutargardan Pass under Sir Frederick Roberts, who defeated the Afghans at Charasia near Kabul, and entered the city in October. Yakub Khan, who had surrendered, was sent to India; and the British army remained in military occupation of the district round Kabul until in December (1879) its communications with India were interrupted, and its position at the capital placed in serious jeopardy, by a general rising of the tribes. After they had been repulsed and put down, not without some hard fighting, Sir Donald Stewart, who had not quitted Kandahar, brought a force up by Ghazni to Kabul, overcoming some resistance on his way, and assumed the supreme command. Nevertheless the political situation was still embarrassing, for as the whole country beyond the range of British effective military control was masterless, it was undesirable to withdraw the troops before a government could be reconstructed which could stand without foreign support, and with which diplomatic relations of some kind might be arranged. The general position and prospect of political affairs in Afghanistan bore, indeed, an instructive resemblance to the situation just forty years earlier, in 1840, with the important differences that the Punjab and Sind had since become British, and that communications between Kabul and India were this time secure.

Reign of Abdur Rahman.—-Abdur Rahman, the son of the late amir Shere Ali's elder drother, had fought against Shere Ali in the war for succession to Dost Mahommed, had been driven beyond the Oxus, and had lived for ten years in exile with the Russians. In March 1880 he came back across the river, and began to establish himself in the northern province of Afghanistan. The viceroy of India, Lord Lytton, on hearing of his reappearance, instructed the political authorities at Kabul to communicate with him. By skilful negotiations a meeting was arranged, and after pressing in vain for a treaty he was induced to assume charge of the country upon his necognition by the British as amir, with the understanding that he should have no relations with other foreign powers, and with a formal assurance from the viceroy of protection from foreign aggression, so long as he should unreservedly follow the advice of the British government in regard to his external affairs. The province of Kandahar was severed from the Kabul dominion; and the sirdar Shere Ali Khan, a member of the Barakzai family, was installed by the British representative as its independent ruler.

For the second time in the course of this war a conclusive settlement of Afghan affairs seemed now to have been attained; and again, as in 1879, it was immediately dissolved. In July 1880, a few days after the proclamation of Abdur Rahman as amir at Kabul, came news that Ayub Khan, Shere Ali's younger son, who had been holding Herat since his father's death, had marched upon Kandahar, had utterly defeated at Maiwand a British force that went out from Kandahar to oppose him, and was besieging that city. Sir Frederick Roberts at once set out from Kabul with 10,000 men to its relief, reached Kandahar after a rapid march of 313 miles, attacked and routed Ayub Khan's army on the 1st of September, and restored British authority in southern Afghanistan. As the British ministry had resolved to evacuate Kandahar, the sirdar Shere Ali Khan, who saw that he could not stand alone, resigned and withdrew to India, and the amir Abdur Rahman was invited to take possession of the province. But when Ayub Khan, who had meanwhile retreated to Herat, heard that the British forces had retired, early in 1881, to India, he mustered a fresh army and again approached Kandahar. In June the fort of Girishk, on the Helmund, was seized by his adherents; the amir's troops were defeated some days later in an engagement, and Ayub Khan took possession of Kandahar at the end of July. The amir Abdur Rahman, whose movements had hitherto been slow and uncertain, now acted with vigour and decision. He marched rapidly from Kabul at the head of a force, with which he encountered Ayub Khan under the walls of Kandahar, and routed his army on 22nd September, taking all his guns and equipage. Ayub Khan fled toward Herat, but as the place had meanwhile been occupied by one of the amir's generals he took refuge in Persia. By this victory Abdur Rahman's rulership was established.

In 1884 it was determined to resume the demarcation, by a joint commission of British and Russian officers, of the northern boundary of Afghanistan. The work went on with much difficulty and contention, until in March 1885, when the amir was at Rawalpindi for a conference with the viceroy of India, Lord Dufferin, the news came that at Panjdeh, a disputed place on the boundary held by the Afghans, the Russians had attacked and driven out with some loss the amir's troops. For the moment the consequences seemed likely to be serious; but the affair was arranged diplomatically, and the demarcation proceeded up to a point near the Oxus river, beyond which the commission were unable to settle an agreement.

During the ten years following his accession in 1880 Abdur Rahman employed himself in extending and consolidating his dominion over the whole country. Some local revolts among the tribes were rigorously suppressed; and two attempts to upset his rulership—the first by Ayub Khan, who entered Afghanistan from Persia, the second and more dangerous one by Ishak Khan, the amir's cousin, who rebelled against him in Afghan Turkestan—-were defeated. By 1891 the amir had enforced his supreme authority throughout Afghanistan more completely than any of his predecessors, In 1895 the amir's troops entered Kafiristan, a wild mountainous tract on the north-east, inhabited by a peculiar race that had hitherto defied all efforts to subjugate them, but were now gradually reduced to submission. Meanwhile the delimitation of the northern frontier, up to the point where it meets Chinese territory on the east, was completed and fixed by arrangements between the governments of Russia and Great Britain; and the eastern border of the Afghan territory, towards India, was also mapped out and partially laid down, in accordance with a convention between the two governments. The amir not only received a large annual subsidy of money from the British government, but he also obtained considerable supplies of war material; and he, moreover, availed himself very freely of facilities that were given him for the importation at his own cost of arms through India. With these resources, and with the advantage of an assurance from the British government that he would be aided against foreign aggression, he was able to establish an absolute military despotism inside his kingdom, by breaking down the power of the warlike tribes which held in check, up to his time, the personal autocracy of the Kabul rulers, and by organizing a regular army well furnished with European rifies and artillery. Taxation of all kinds was heavily increased, and systematically collected. The result was that whereas in former times the forces of an Afghan ruler consisted mainly of a militia, furnished by the chiefs of tribes who held land on condition of military service, and who stoutly resisted any attempt to commute this service for money payment, the amir had at his command a large standing army, and disposed of a substantial revenue paid direct to his treasury. Abdur Rahman executed or exiled all those whose political influence he saw reason to fear, or of whose disaffection he had the slightest suspicion; his administration was severe and his punishments were cruel; but undoubtedly he put down disorder, stopped the petty tyranny of local chiefs and brought violent crime under some effective control in the districts. Travelling by the high roads during his reign was comparatively safe; although it must be added that the excessive exactions of dues and customs very seriously damaged the external trade. In short, Abdur Rahman's reign produced an important political revolution, or reformation, in Afghanistan, which rose from the condition of a country distracted by chronic civil wars, under rulers whose authority depended upon their power to hold down or conciliate fierce and semi-independent tribes in the outlying parts of the dominion, to the rank of a formidable military state governed autocratically. He established, for the first time in the history of the Afghan kingdom, a powerfully centralized administration strong enough to maintain order and to enforce obedience over all the country which he had united under his dominion, supported by a force sufficiently armed and disciplined to put down attempts at resistance or revolt. His policy, consistently maintained, was to permit no kind of foreign interference, on any pretext, with the interior concerns or the economical conditions of his country. From the British government he accepted supplies of arms and subsidies of money; but he would make no concessions in return, and all projects of a strategical or commercial nature, such as railways and telegraphs, proposed either for the defence or the development of his possessions, seem to have been regarded by the amir with extreme distrust, as methods of what has been called pacific penetration —so that on these points he was immovable. It was probably due to the strength and solidity of the executive administration organized, during his lifetime, by Abdur Rahman that, for the first time in the records of the dynasty founded by Ahmad Shah in the latter part of the 18th century, his death was not followed by disputes over the succession or by civil war.

Succession of Habibullah.—The amir Abdur Rahman died on the 1st of October 1901; and two days later his eldest son, Habibullah, formally announced his accession to the rulership. He was recognized with acclamation by the army, by the religious bodies, by the principal tribal chiefs and by all classes of the people as their lawful sovereign; while a deputation of Indian Mahommedans was despatched to Kabul from India to convey the condolences and congratulations of the viceroy. The amir's first measures were designed to enhance his popularity and to improve his internal administration, particularly with regard to the relations of his government with the tribes, and to the system introduced by the late amir of compulsory military service, whereby each tribe was required to supply a proportionate number of recruits. With this object a council of state for tribal affairs was established; and it was arranged that a representative of each tribe should be associated with the provincial governors for the adjudication of tribal cases.

In the important matter of foreign relations Habibullah showed a determination to adopt the policy of his father, to whom the British government had given an assurance of aid to repel foreign aggression, on the condition that the amir should follow the advice of that government in regard to external affairs. This condition was loyally observed by the new amir, who referred to India all communications of an official kind received from the Russian authorities in the provinces bordering on Afghanistan. But toward the various questions left pending between the governments of India and Afghanistan the new amir maintained also his fatber's attitude. He gave no indications of a disposition to continue the discussion of them, or to entertain proposals for extending or altering his relations with the Indian government. An invitation from the viceroy to meet him in India, with the hope that these points might be settled in conference, was put aside by dilatory excuses, until at last the project was abandoned, and finally the amir agreed to receive at Kabul a diplomatic mission. The mission, whose chief was Sir Louis Dane, foreign secretary to the Indian government, reached Kabul early in December 1904, and remained there four months in negotiation with the amir personally and with his representatives. It was found impossible, after many interviews, to obtain from Habibullah his consent to any addition to or variation of the terms of the assurance given by the British government in 1880, with which he professed himself entirely satisfied, so that the treaty finally settled in March 1905 went no further than a formal confirmation of all engagements previously concluded with the amir's predecessor. It was felt in British circles at the time that a very considerable concession to Habibullah's independence of attitude was displayed in the fact that he was styled in the treaty "His Majesty''; but, in the circumstances, it seems to have been thought diplomatic to accede to the amir's determination to insist on this matter of style. But the rebuff showed that it was desirable in the interests both of the British government and of Afghanistan that an opportunity should be made for enabling the amir to have personal acquaintance with the highest Indian authorities. A further step, calculated to strengthen the relations of amity between the two governments, was taken when it was arranged that the amir should pay a visit to the viceroy, Lord Minto, in India, in January 1907; and this visit took place with great cordiality and success.

The Anglo-Russian Convention, signed on the 31st of August 1907, contained the following important declarations with regard to Afghanistan. Great Britain disclaimed any intention of altering the political status or (subject to the observance of the treaty of 1905) of interfering in the administration or annexing any territory of Afghanistan, and engaged to use her influence there in no manner threatening to Russia. Russia, on her part, recognized Afghanistan as outside her sphere of influence.

AUTHORITIES. —-MacGregor, Gazetteer or Afghanistan (1871); Elphinstone, Account of the Kingdom of Kabul (1809); Ferrier, History of the Afghanis (1858); Bellow, Afghanistan and the Afghans (1879); Baber's Memoirs (1844); Kaye, History of the War in Afghanistan (1878); Malleson, History of Afghanistan (1879); Heusman, The Afghan War (1881); Sir H. M. Durand, The First Afghan War (1879); Forbes, The Afghan Wars (1892); Rawlinson, England and Russia in the East (1875); Wyllie, Essays on the External Policy of India (1875). A. C. Yate, Northern Afghanistan (1888); Curzon, Problems of the Far East (1894); Robertson, The Kafir of the Hindu Kush (1896); Holdich, Indian Borderland (1901); Thorburn, Asiatic Neighbours (1895); Lord Roberts, Forty-one Years in India (1898); Lady Betty Balfour, Lord Litton's Indian Administration (1899); Hanna, Second Afghani War (1899); Gray, At the Court of the Amir (1895); Sultan Mohammad Khan, Constitution and Laws of Afghanistan (1900): Life of Abdur Rahinani (1900); Angus Hamilton, Afghanistan (1906). (H. Y.; A. C. L.)

1 We owe our knowledge of the geology of Afghanistan almost entirely to the observations of C. U. Grierbach, and a summary of his researches will be found in Records of the Geological Survey of India, vol. xx. (1887), pp. 93-103, with map.

AFGHAN TURKESTAN, the most northern province of Afghanistan. It is bounded on the E. by Badakshan, on the N. by the Oxus river, on the N.W. and W. by Russia and the Hari Rud river, and on the S. by the Hindu Lush, the Koh-i-Baba and the northern watershed of the Hari Rud basin. Its northern frontier was decided by the Anglo-Russian agreement of 1873; and delimited by the Russo-Afghan boundary commission of 1883, which gave rise to the Panjdeh incident. The whole territory, from the junction of the Kokcha river with the Oxus on the north-east to the province of Herat on the south-west, is some 500 m. in length, with an average width from the Russian frontier to the Hindu Kush of 114 m. It thus comprises about 57,000 sq. m. or roughly' two-ninths of the kingdom of Afghanistan. Except in the river valleys it is a poor territory, rough and mountainous towards the south, but subsiding into undulating wastes and pasture-lands towards the Turkman desert, and the Oxus riverain which is highly cultivated. The population, which is mostly agricultural, settled in and around its towns and villages, is estimated at 750,000. The province includes the khanates of Kunduz, Tashkurgan, Balkh with Akcha; the western khanates of Saripul, Shibarghan, Andkhui and Maimana, sometimes classed together as the Chahar Vidayet, or "Four Domains''; and such parts of the Hazara tribes as lie north of the Hindu Kush and its prolongation. The principal town is Mazar-i-Sharif, which in modern times has supplanted the ancient city of Balkh; and Takhtapul, near Mazar, is the chief Afghan cantonment north of the Hindu Kush.

Ethnically and historically Afghan Turkestan is more connected with Bokhara than with Kabul, of which government it has been a dependency only since the time of Dost Mahommed. The bulk of the people of the cities are of Persian and Uzbeg stock, but interspersed with them are Mongol Hazaras and Hindus with Turkoman tribes in the Oxus plains. Over these races the Afghans rule as conquerors and there is no bond of racial unity between them. Ancient Balkh or Bactriana was a province of the Achaemenian empire, and probably was occupied in great measure by a race of Iranian blood. About 250 B.C. Diodotus (Theodotus), governor of Bactria under the Seleucidae, declared his independence, and commenced the history of the Greco-Bactrian dynasties, which succumbed to Parthian and nomadic movements about 126 B.C. After this came a Buddhist era which has left its traces in the gigantic sculptures at Bamian and the rock-cut topes of Haibak. The district was devastated by Jenghiz Khan, and has never since fully recovered its prosperity. For about a century it belonged to the Delhi empire, and then fell into Uzbeg hands. In the 18th century it formed part of the dominion of Ahmad Khan Durani, and so remained under his son Timur. But under the fratricidal wars of Timur's sons the separate khanates fell back under the independent rule of various Uzbeg chiefs. At the beginning of the 19th century they belonged to Bokhara; but under the great amir Dost Mahommed the Afghans recovered Balkh and Tashkurgan in 1850, Akcha and the four western khanates in 1855, and Kunduz in 1859. The sovereignty over Andkhui, Shibarghan, Saripul and Maimana was in dispute between Bokhara and Kabul until settled by the Anglo-Russian agreement of 1873 in favour of the Afghan claim. Under the strong rule of Abdur Rahman these outlying territories were closely welded to Kabul; but after the accession of Habibullah the bonds once more relaxed. (T. H. Ut.v)

AFIUM-KARA-HISSAR (afium, opium), the popular name of Kara-hissar Sahib, a city of Asiatic Turkey, in the vilayet of Brusa, nearly 200 m. E. of Smyrna, and 50 m. S.S.E. of Kutaiah. Pop. 18,000 (Moslems, 13,000; Christians, 5000). Called Nicopolis by Leo III. after his victory over the Arabs in 740, its name was changed by the Seljuk Turks to Kara-hissar. It stands partly on level ground, partly on a declivity, and above it rises a precipitous trachytic rock (400 ft.) on the summit of which are the ruins of an ancient castle. From its situation on the route of the caravans between Smyrna and western Asia on the one hand, and Armenia, Georgia, &c., on the other, the city became a place of extensive trade, and its bazaars are well stocked with the merchandise of both Europe and the East. Opium in large quantities is produced in its vicinity and forms the staple article of its commerce; and there are, besides, manufactures of black felts, carpets, arms and saddlery. Afium contains several mosques (one of them a very handsome building), and is the seat of an Armenian bishop. The town is connected by railway with Smyrna, Konia, Angora and Constantinople.

See V. Cuinet, Turquie d'Asie (Paris, 1894), vol. iv.

A FORTIORI (Lat. "from a stronger [reason]''), a term used of an argument which justifies a statement not itself specifically demonstrated by reference to a proved conclusion which includes it; thus, if A is proved less than B, and is known to be greater than C, it follows a fortiori that C is less than B without further proof. The argument is frequently based merely on a comparison of probabilities (cf. Matt. vi. 30), when it constitutes an appeal to common sense.

AFRANIUS, LUCIUS, Roman general, lived in the times of the Sertorian (79-72), third Mithradatic (74-61) and Civil Wars. Of humble origin (Cic. ad Att. i. 16. 20), from his early years he was a devoted adherent of Pompey. In 60, chiefly by Pompey's support, he was raised to the consulship, but in performing the duties of that office he showed an utter incapacity to manage civil affairs. In the following year, while governor of Cisalpine Gaul, he obtained the honour of a triumph, and on the allotment of Spain to Pompey (55), Afranius and Marcus Petreius were sent to take charge of the government. On the rupture between Caesar and Pompey they were compelled, after a short campaign in which they were at first successful, to surrender to Caesar at Ilerda (49), and were dismissed on promising not to serve again in the war. Afranius, regardless of his promise, joined Pompey at Dyrrhachium, and at the battle of Pharsalus (48) had charge of Pompey's camp. On the defeat of Pompey, Afranius, despairing of pardon from Caesar, went to Africa, and was present at the disastrous battle of Thapsus (46). Escaping from the field with a strong body of cavalry, he was afterwards taken prisoner, along with Faustus Sulla, by the troops of Sittius, and handed over to Caesar, whose veterans rose in tumult and put them to death.

See Hirtius, Bell. Afric. 95; Plutarch, Pompey; Dio Cassius xxxvii., xli.-xliii.; Caesar, B.C i. 57-87; Appian, B.C ii.; for the history of the period, articles on CAESAR and POMPEY.

AFRANIUS, LUCIUS, Roman comic poet, flourished about 94 B.C. His comedies chiefly dealt with everyday subjects from Roman middle-class life, and he himself tells us that he borrowed freely from Menander and others. His style was vigorous and correct; his moral tone that of the period.

Horace, Epp. ii. 1. 57; Cicero, Brutus, 45, de Fin. i. 3; Quitilian x 1. 100; fragments, about 400 lines, in Ribbeck, Scaenicae Romanorum Poesis Fragmenta, ii. (1898).

AFRICA, the name of a continent representing the largest of the three great southward projections from the main mass of the earth's surface. It includes within its remarkably regular outline an area, according to the most recent computations, of 11,262,000 sq. m., excluding the islands.1 Separated from Europe by the Mediterranean Sea, it is joined to Asia at its N.E. extremity by the Isthmus of Suez, 80 m. wide. From the most northerly point, Ras ben Sakka, a little west of Cape Blanc, in 37 deg. 21' N., to the most southerly point, Cape Agulhas, 34 deg. 51' 15'' S., is a distance approximately of 5000 m.; from Cape Verde, 17 deg. 33' 22'' W., the westernmost point, to Ras Hafun, 51 deg. 27' 52'' E., the most easterly projection, is a distance (also approximately) of 4600 m. The length of coast-line is 16,100 m. and the absence of deep indentations of the shore is shown by the fact that Europe, which covers only 3,760,000 sq. m., has a coast-line of 19,800 m.

I. PHYSICAL GEOGRAPHY

The main structural lines of the continent show both the east-to-west direction characteristic, at least in the eastern hemisphere, of the more northern parts of the world, and the north-to-south direction seen in the southern peninsulas. Africa is thus composed of two segments at right angles, the northern running from east to west, the southern from north to south, the subordinate lines corresponding in the main to these two directions.

Main Geographical Features.—The mean elevation of the continent approximates closely to 2000 ft., which is roughly the elevation of both North and South America, but is considerably less than that of Asia (3117 ft.). In contrast with the other continents it is marked by the comparatively small area both of very high and of very low ground, lands under 600 ft. occupying an unusually small part of the surface; while not only are the highest elevations inferior to those of Asia and South America, but the area of land over 10,000 ft. is also quite insignificant, being represented almost entirely by individual peaks and mountain ranges. Moderately elevated tablelands are thus the characteristic feature of the continent, though the surface of these is broken by higher peaks and ridges. (So prevalent are these isolated peaks and ridges that a special term [Inselberg-landschaft] has been adopted in Germany to describe this kind of country, which is thought to be in great part the result of wind action.) As a general rule, the higher tablelands lie to the east and south, while a progressive diminution in altitude towards the west and north is observable. Apart from the lowlands and the Atlas range, the continent may be divided into two regions of higher and lower plateaus, the dividing line (somewhat concave to the north-west) running from the middle of the Red Sea to about 6 deg. S. on the west coast. We thus obtain the following four main divisions of the continent:—-(1) The coast plains—-often fringed seawards by mangrove swamps—never stretching far from the coast, except on the lower courses of streams. Recent alluvial flats are found chiefly in the delta of the more important rivers. Elsewhere the coast lowlands merely form the lowest steps of the system of terraces which constitutes the ascent to the inner plateaus. (2) The Atlas range, which, orographically, is distinct from the rest of the continent, being unconnected with any other area of high ground, and separated from the rest of the continent on the south by a depressed and desert area (the Sahara), in places below sea-level. (3) The high southern and eastern plateaus, rarely falling below 2000 ft., and having a mean elevation of about 3500 ft. (4) The north and west African plains, bordered and traversed by bands of higher ground, but generally below 2000 ft. This division includes the great desert of the Sahara.

The third and fourth divisions may be again subdivided. Thus the high plateaus include:—(a) The South African plateau as far as about 12 deg. S., bounded east, west and south by bands of high ground which fall steeply to the coasts. On this account South Africa has a general resemblance to an inverted saucer. Due south the plateau rim is formed by three parallel steps with level ground between them. The largest of these level areas, the Great Karroo, is a dry, barren region, and a large tract of the plateau proper is of a still more arid character and is known as the Kalahari Desert. The South African plateau is connected towards the north-east with (b) the East African plateau, with probably a slightly greater average elevation, and marked by some distinct features. It is formed by a widening out of the eastern axis of high ground, which becomes subdivided into a number of zones running north and south and consisting in turn of ranges, tablelands and depressions. The most striking feature is the existence of two great lines of depression, due largely to the subsidence of whole segments of the earth's crust, the lowest parts of which are occupied by vast lakes. Towards the south the two lines converge and give place to one great valley (occupied by Lake Nyasa), the southern part of which is less distinctly due to rifting and subsidence than the rest of the system. Farther north the western depression, sometimes known as the Central African trough or Albertine rift-valley, is occupied for more than half its length by water, forming the four lakes of Tanganyika, Kivu, Albert Edward and Albert, the first-named over 400 m. long and the longest freshwater lake in the world. Associated with these great valleys are a number of volcanic peaks, the greatest of which occur on a meridional line east of the eastern trough. The eastern depression, known as the East African trough or rift-valley, contains much smaller lakes, many of them brackish and without outlet, the only one comparable to those of the western trough being Lake Rudolf or Basso Norok. At no great distance east of this rift-valley are Kilimanjaro—with its two peaks Kibo and Mawenzi, the former 19,321 ft., and the culminating point of the whole continent—and Kenya (17,007 ft.). Hardly less important is the Ruwenzori range (over 16,600 ft.), which lies east of the western trough. Other volcanic peaks rise from the floor of the valleys, some of the Kirunga (Mfumbiro) group, north of Lake Kivu, being still partially active. (c) The third division of the higher region of Africa is formed by the Abyssinian highlands, a rugged mass of mountains forming the largest continuous area of its altitude in the whole continent, little of its surface falling below 5000 ft., while the summits reach heights of 15,000 to 16,000 ft. This block of country lies just west of the line of the great East African trough, the northern continuation of which passes along its eastern escarpment as it runs up to join the Red Sea. There is, however, in the centre a circular basin occupied by Lake Tsana.

Both in the east and west of the continent the bordering highlands are continued as strips of plateau parallel to the coast, the Abyssinian mountains being continued northwards along the Red Sea coast by a series of ridges reaching in places a height of 7000 ft. In the west the zone of high land is broader but somewhat lower. The most mountainous districts lie inland from the head of the Gulf of Guinea (Adamawa, &c.), where heights of 6000 to 8000 ft. are reached. Exactly at the head of the gulf the great peak of the Cameroon, on a line of Volcanic action continued by the islands to the south-west, has a height of 13,370 ft., while Clarence Peak, in Fernando Po, the first of the line of islands, rises to over 9000. Towards the extreme west the Futa Jallon highlands form an important diverging point of rivers, but beyond this, as far as the Atlas chain, the elevated rim of the continent is almost wanting.

The area between the east and west coast highlands, which north of 17 deg. N. is mainly desert, is divided into separate basins by other bands of high ground, one of which runs nearly centrally through North Africa in a line corresponding roughly with the curved axis of the continent as a whole. The best marked of the basins so formed (the Congo basin) occupies a circular area bisected by the equator, once probably the site of an inland sea. The arid region, the Sahara—the largest desert in the world, covering 3,500,000 sq. m.—extends from the Atlantic to the Red Sea. Though generally of slight elevation it contains mountain ranges with peaks rising to 8000 ft. Bordered N.W. by the Atlas range, to the N.E. a rocky plateau separates it from the Mediterranean; this plateau gives place at the extreme east to the delta of the Nile. That river (see below) pierces the desert without modifying its character. The Atlas range, the north-westerly part of the continent, between its seaward and landward heights encloses elevated steppes in places 100 m. broad. From the inner slopes of the plateau numerous wadis take a direction towards the Sahara. The greater part of that now desert region is, indeed, furrowed by old water-channels.

The following table gives the approximate altitudes of the chief mountains and lakes of the continent:—

Mountains. Ft. Lakes. Ft. Rungwe (Nyasa) . 10,400 Chad . . . . 8502 Drakensberg . . 10,7002 Leopold II . . 1100 Lereko or Sattima . 13,2143 Rudolf . . . 1250 (Aberdare Range) Nyasa . . . 16453 Cameroon . . 13,370 Albert Nyanza . 20282 Elgon . . . 14,1523 Tanganyika . . 26243 Karissimbi . . Ngami . . . . 2950 (Mfumbiro) . 14,6833 Mweru . . . . 3000 Meru . . . 14,9553 Albert Edward . 30043 Taggharat (Atlas) . 15,0002 Bangweulu. . . 3700 Simen Mountains, . 15,1602 Victoria Nyanza. 37203 Abyssinia Abai . . . . 4200 Ruwenzori . . 16,6193 Kivu . . . . 48293 Kenya . . . 17,0073 Tsana . . . . 5690 Kilimanjaro . . 19,3213 Naivasha . . . 61353

The Hydrographic Systems.—-From the outer margin of the African plateaus a large number of streams run to the sea with comparatively short courses, while the larger rivers flow for long distances on the interior highlands before breaking through the outer ranges. The main drainage of the continent is to the north and west, or towards the basin of the Atlantic Ocean. The high lake plateau of East Africa contains the head-waters of the Nile and Congo: the former the longest, the latter the largest river of the continent. The upper Nile receives its chief supplies from the mountainous region adjoining the Central African trough in the neighbourhood of the equator. Thence streams pour east to the Victoria Nyanza, the largest African lake (covering over 26,000 sq. m.), and west and north to the Albert Edward and Albert Nyanzas, to the latter of which the effluents of the other two lakes add their waters. Issuing from it the Nile flows north, and between 7 deg. and 10 deg. N. traverses a vast marshy level during which its course is liable to blocking by floating vegetation. After receiving the Bahr-el-Ghazal from the west and the Sobat, Blue Nile and Atbara from the Abyssinian highlands (the chief gathering ground of the flood-water), it crosses the great desert and enters the Mediterranean by a vast delta. The most remote head-stream of the Congo is the Chambezi, which flows south-west into the marshy Lake Bangweulu. From this lake issues the Congo, known in its upper course by various names. Flowing first south, it afterwards turns north through Lake Mweru and descends to the forest-clad basin of west equatorial Africa. Traversing this in a majestic northward curve and receiving vast supplies of water from many great tributaries, it finally turns south-west and cuts a way to the Atlantic Ocean through the western highlands. North of the Congo basin and separated from it by a broad undulation of the surface is the basin of Lake Chad—-a flat-shored, shallow lake filled principally by the Shad coming from the south-east. West of this is the basin of the Niger, the third river of Africa, which, though flowing to the Atlantic, has its principal source in the far west, and reverses the direction of flow exhibited by the Nile and Congo. An important branch, however—the Benue—comes from the south-east. These four river-basins occupy the greater part of the lower plateaus of North and West Africa, the remainder consisting of arid regions watered only by intermittent streams which do not reach the sea. Of the remaining rivers of the Atlantic basin the Orange, in the extreme south, brings the drainage from the Drakensberg on the opposite side of the continent, while the Kunene, Kwanza, Ogowe and Sanaga drain the west corst highlands of the southern limb; the Volta, Komoe, Bandama, Gambia and Senegal the highlands of the western limb. North of the Senegal for over 1000 m. of coast the arid region reaches to the Atlantic. Farther north are the streams, with comparatively short courses, which reach the Atlantic and Mediterranean from the Atlas mountains.

Of the rivers flowing to the Indian Ocean the only one draining any large part of the interior plateaus is the Zambezi, whose western branches rise in the west coast highlands. The main stream has its rise in 11 deg. 21' 3'' S. 24 deg. 22' E. at an elevation of 5000 ft. It flows west and south for a considerable distance before turning to the east. All the largest tributaries, including the Shire, the outflow of Lake Nyasa, flow down the southern slopes of the band of high ground which stretches across the conbnent in 10 deg. to 12 deg. S. In the south-west the Zambezi system interlaces with that of the Taukhe (or Tioghe), from which it at times receives surplus water. The rest of the water of the Taukhe, known in its middle course as the Okavango, is lost in a system of swamps and saltpans which formerly centred in Lake Ngami, now dried up. Farther south the Limpopo drains a portion of the interior plateau but breaks through the bounding highlands on the side of the continent nearest its source. The Rovuma, Rufiji, Tana, Juba and Webi Shebeli principally drain the outer slopes of the East African highlands, the last named losing itself in the sands in close proximity to the sea. Another large stream, the Hawash, rising in the Abyssinian mountains, is lost in a saline depression near the Gulf of Aden. Lastly, between the basins of the Atlantic and Indian Oceans there is an area of inland drainage along the centre of the East African plateau, directed chiefly into the lakes in the great rift-valley. The largest river is the Omo, which, fed by the rains of the Abyssinian highlands, carries down a large body of water into Lake Rudolf. The rivers of Africa are generally obstructed either by bars at their mouths or by cataracts at no great distance up-stream. But when these obstacles have been overcome the rivers and lakes afford a network of navigable waters of vast extent.

The calculation of the areas of African drainage systems, made by Dr A. Bludau (Petermanns Mitteilungen, 43, 1897, pp. 184-186) gives the following general results:—

Basin of the Atlantic . . . . . 4,070,000 sq. m. '' '' Mediterranean . . . 1,680,000 '' '' '' Indian Ocean . . . . 2,086,000 '' Inland drainage area . . . . . 3,452,000 ''

The areas of individual river-basins are:—

Congo (length over 3000 m.) . . 1,425,000 sq. m. Nile ( '' fully 4000 m.) . . 1,082,0004 '' Niger ( '' about 2600 m.) . . 808,0005 '' Zambezi ( '' '' 2000 m.) . . 513,500 '' Lake Chad . . . . . . . . . 394,000 '' Orange (length about 1300 m.) . . 370,505 '' '' (actual drainage area) . . 172,500 ''

The area of the Congo basin is greater than that of any other river except the Amazon, while the African inland drainage area is greater than that of any continent but Asia, in which the corresponding area is 4,000,000 sq. m.

The principal African lakes have been mentioned in the description of the East African plateau, but some of the phenomena connected with them may be spoken of more particularly here. As a rule the lakes which occupy portions of the great rift-valleys have steep sides and are very deep. This is the case with the two largest of the type, Tanganyika and Nyasa, the latter of which has depths of 430 fathoms. Others, however, are shallow, and hardly, reach the steep sides of the valleys in the dry season. Such are Lake Rukwa, in a subsidiary depression north of Nyasa, and Eiassi and Manyara in the system of the eastern rift-valley. Lakes of the broad type are of moderate depth, the deepest sounding in Victoria Nyanza being under 50 fathoms. Apart from the seasonal variations of level, most of the lakes show periodic fluctuations, while a progressive desiccation of the whole region is said to be traceable, tending to the ultimate disappearance of the lakes. Such a drying up has been in progress during long geologic ages, but doubt exists as to its practical importance at the present time. The periodic fluctuations in the level of Lake Tanganyika are such that its outllow is intermittent. Besides the East African lakes the principal are:—-Lake Chad, in the northern area of inland drainage; Bangweulu and Mweru, traversed by the head-stream of the Congo; and Leopold II. and Ntomba (Mantumba), within the great bend of that river. All, exceot possibly Mweru, are more or less shallow, and Chad appears to by drying up. The altitudes of the African lakes have already been stated.

Divergent opinions have been beld as to the mode of origin of the East African lakes, especially Tanganyika, which some geologists have considered to represent an old arm of the sea, dating from a time when the whole central Congo basin was under water; others holding that the lake water has accumulated in a depression caused by subsidence. The former view is based on the existence in the lake of organisms of a decidedly marine type. They include a jelly-fish, molluscs, prawns, crabs, &c., and were at first considered to form an isolated group found in no other of the African lakes; but this supposition has been proved to be erroneous.

Islands.—With one exception—-Madagascar—the African islands are small. Madagascar, with an area of 229,820 sq. m., is, after New Guinea and Borneo, the largest island of the world.

It lies off the S.E. coast of the continent, from which it is separated by the deep Mozambique channel, 250 m. wide at its narrowest point. Madagascar in its general structure, as in flora and fauna, forms a connecting link between Africa and southern Asia. East of Madagascar are the small islands of Mauritius and Reunion. Sokotra lies E.N.E. of Cape Guardafui. Off the north-west coast are the Canary and Cape Verde archipelagoes. which, like some small islands in the Gulf of Guinea, are of volcanic origin.

Climate and Health.—-Lying almost entirely within the tropics, and equally to north and south of the equator, Africa does not show excessive variations of temperature. Great heat is experienced in the lower plains and desert regions of North Africa, removed by the great width of the continent from the influence of the ocean, and here, too, the contrast between day and night, and between summer and winter, is greatest. (The rarity of the air and the great radiation during the night cause the temperature in the Sahara to fall occasionally to freezing point.) Farther south, the heat is to some extent modified by the moisture brought from the ocean, and by the greater elevation of a large part of the surface, especially in East Africa, where the range of temperature is wider than in the Congo basin or on the Guinea coast. In the extreme north and south the climate is a warm temperate one, the northern countries being on the whole hotter and drier than those in the southern zone; the south of the continent being narrower than the north, the influence of the surrounding ocean is more felt. The most important climatic differences are due to variations in the amount of rainfall. The wide heated plains of the Sahara, and in a lesser degree the corresponding zone of the Kalahari in the south, have an exceedingly scanty rainfall, the winds which blow over them from the ocean losing part of their moisture as they pass over the outer highlands, and becoming constantly drier owing to the heating effects of the burning soil of the interior; while the scarcity of mountain ranges in the more central parts likewise tends to prevent condensation. In the inter-tropical zone of summer precipitation, the rainfall is greatest when the sun is vertical or soon after. It is therefore greatest of all near the equator, where the sun is twice vertical, and less in the direction of both tropics. The rainfall zones are, however, somewhat deflected from a due west-to-east direction, the drier northern conditions extending southwards along the east coast, and those of the south northwards along the west. Within the equatorial zone certain areas, especially on the shores of the Gulf of Guinea and in the upper Nile basin, have an intensified rainfall, but this rarely approaches that of the rainiest regions of the world. The rainiest district in all Africa is a strip of coastland west of Mount Cameroon, where there is a mean annual rainfall of about 390 in. as compared with a mean of 458 in. at Cherrapunji, in Assam. The two distinct rainy seasons of the equatorial zone, where the sun is vertical at half-yearly intervals, become gradually merged into one in the direction of the tropics, where the sun is overhead but once. Snow falls on all the higher mountain ranges, and on the highest the climate is thoroughly Alpine. The countries bordering the Sahara are much exposed to a very dry wind, full of fine particles of sand, blowing from the desert towards the sea. Known in Egypt as the khamsin, on the Mediterranean as the sirocco, it is called on the Guinea coast the harmattan. This wind is not invariably hot; its great dryness causes so much evaporation that cold is not infrequently the result. Similar dry winds blow from the Kalahari in the south. On the eastern coast the monsoons of the Indian Ocean are regularly felt, and on the south-east hurricanes are occasionally experienced.

While the climate of the north and south, especially the south, is eminently healthy, and even the intensely heated Sahara is salubrious by reason of its dryness, the tropical zone as a whole is, for European races, the most unhealthy portion of the world. This is especially the case in the lower and moister regions, such as the west coast, where malarial fever is very prevalent and deadly; the most unfavourable factors being humidity with absence of climatic variation (daily or seasonal). The higher plateaus, where not only is the average temperature lower, but such variations are more extensive, are more healthy; and in certain localities (e.g. Abyssinia and parts of British East Africa) Europeans find the climate suitable for permanent residence. On tablelands over 6500 ft. above the sea, frost is not uncommon at night, even in places directly under the equator. The acclimatization of white men in tropical Africa generally is dependent largely on the successful treatment of tropical diseases. Districts which had been notoriously deadly to Europeans were rendered comparatively healthy after the discovery, in 1899, of the species of mosquito which propagates malarial fever, and the measures thereafter taken for its destruction and the filling up of swamps. The rate of mortality among the natives from tropical diseases is also high, one of the most fatal being that known as sleeping sickness. (The ravages of this disease, which also attacks Europeans, reached alarming proportions between 1893 and 1907, and in the last-named year an international conference was held in London to consider measures to combat it.) When removed to colder regions natives of the equatorial districts suffer greatly from chest complaints. Smallpox also makes great ravages among the negro population.

Flora.—The vegetation of Africa follows very closely the distribution of heat and moisture. The northern and southern temperate zones have a flora distinct from that of the continent generally, which is tropical. In the countries bordering the Mediterranean are groves of oranges and olive trees, evergreen oaks, cork trees and pines, intermixed with cypresses, myrtles, arbutus and fragrant tree-heaths. South of the Atlas range the conditions alter. The zones of minimum rainfall have a very scanty flora, consisting of plants adapted to resist the great dryness. Characteristic of the Sahara is the date-palm, which flourishes where other vegetation can scarcely maintain existence, while in the semidesert regions the acacia (whence is obtained gum-arabic) is abundant. The more humid regions have a richer vegetation —dense forest where the rainfall is greatest and variations of temperature least, conditions found chiefly on the tropical coasts, and in the west African equatorial basin with its extension towards the upper Nile; and savanna interspersed with trees on the greater part of the plateaus, passing as the desert regions are approached into a scrub vegetation consisting of thorny acacias, &c. Forests also occur on the humid slopes of mountain ranges up to a certain elevation. In the coast regions the typical tree is the mangrove, which flourishes wherever the soil is of a swamp character. The dense forests of West Africa contain, in addition to a great variety of dicotyledonous trees, two palms, the Elaeis guincensis (oil-palm) and Raphia vinifera (bamboo-palm), not found, generally speaking, in the savanna regions. The bombax or silk-cotton tree attains gigantic proportions in the forests, which are the home of the indiarubber-producing plants and of many valuable kinds of timber trees, such as odum (Chlorophora excelsa), ebony, mahogany (Khaya senegalensis), African teak or oak (Oldfieldia africana) and camwood (Baphia nitida.) The climbing plants in the tropical forests are exceedingly luxuriant and the undergrowth or "bush'' is extremely dense. In the savannas the most characteristic trees are the monkey bread tree or baobab (Adanisonia digitata), doom palm (Hyphaene) and euphorbias. The coffee plant grows wild in such widely separated places as Liberia and southern Abyssinia. The higher mountains have a special flora showing close agreement over wide intervals of space, as well as affinities with the mountain flora of the eastern Mediterranean, the Himalayas and Indo-China (cf. A. Engler, Uber die Hochgebirgsflora des tropischen Afrika, 1892).

In the swamp regions of north-east Africa the papyrus and associated plants, including the soft-wooded ambach, flourish in immense quantities—-and little else is found in the way of vegetation. South Africa is largely destitute of forest save in the lower valleys and coast regions. Tropical flora disappears, and in the semi-desert plains the fleshy, leafless, contorted species of kapsias, mesembryanthemums, aloes and other succulent plants make their appearance. There are, too, valuable timber trees, such as the yellow pine (Podocarpus elongatus), stinkwood (Ocotea), sneezewood or Cape ebony (Pteroxylon utile) and ironwood. Extensive miniature woods of heaths are found in almost endless variety and covered throughout the greater part of the year with innumerable blossoms in which red is very prevalent. Of the grasses of Africa alfa is very abundant in the plateaus of the Atlas range.

Fauna.—The fauna again shows the effect of the characteristics of the vegetation. The open savannas are the home of large ungulates, especially antelopes, the giraffe (peculiar to Africa), zebra, buffalo, wild ass and four species of rhinoceros; and of carnivores, such as the lion, leopard, hyaena, &c. The okapi (a genus restricted to Africa) is found only in the dense forests of the Congo basin. Bears are confined to the Atlas region, wolves and foxes to North Africa. The elephant (though its range has become restricted through the attacks of hunters) is found both in the savannas and forest regions, the latter being otherwise poor in large game, though the special habitat of the chimpanzee and gorilla. Baboons and mandrills, with few exceptions, are peculiar to Africa. The single-humped camel—as a domestic animal—is especially characteristic of the northern deserts and steppes.

The rivers in the tropical zone abound with hippopotami and crocodiles, the former entirely confined to Africa. The vast herds of game, formerly so characteristic of many parts of Africa, have much diminished with the increase of intercourse with the interior. Game reserves have, however, been established in South Africa, British Central Africa, British East Africa, Somahland, &c., while measures for the protection of wild animals were laid down in an international convention signed in May 1900.

The ornithology of northern Affica presents a close resemblance to that of southern Europe, scarcely a species being found which does not also occur in the other countries bordering the Mediterranean. Among the birds most characteristic of Africa are the ostrich and the secretary-bird. The ostrich is widely dispersed, but is found chiefly in the desert and steppe regions. The secretary-bird is common in the south. The weaver birds and their allies, including the long-tailed whydahs, are abundant, as are, among game-birds, the francolin and guinea-fowl. Nany of the smaller birds, such as the sun-birds, bee-eaters, the parrots and halcyons, as well as the larger plantain-eaters, are noted for the brilliance of their plumage. Of reptiles the lizard and chameleon are common, and there are a number of venomous serpents, though these are not so numerous as in other tropical countries. The scorpion is abundant. Of insects Africa has many thousand different kinds; of these the locust is the proverbial scourge of the continent, and the ravages of the termites or white ants are almost incredible. The spread of malaria by means of mosquitoes has already been mentioned. The tsetse fly, whose bite is fatal to all domestic animals, is common in many districts of South and East Africa. Fortunately it is found nowhere outside Africa. (E. HE.; F. R. C.)

1 With the islands, 11,498,000 sq. m.

2 Estimated.

3 See the calculations of Capt. T. T. Behrens, Geog. Journal, vol. xxix. (1907).

4 The estimate of Capt. H. G. Lyons in 1905 was 1,107,227 sq. mi.

5 including waterless tracts naturally belonging to the river-basin.

II. GEOLOGY

In shape and general geological structure Africa bears a close resemblance to India. Both possess a meridional extension with a broad east and west folded region in the north. In both a successive series of continental deposits, ranging from the Carboniferous to the Rhaetic, rests on an older base of crystalline rocks. In the words of Professor Suess, "India and Africa are true plateau countries.''

Of the primitive axes of Africa few traces remain. Both on the east and west a broad zone of crystalline rochs extends parallel with the coast-line to form the margin of the elevated plateau of the interior. Occasionally the crystalline belt comes to the coast, but it is usually reached by two steps known as the coastal belt and foot-plateau. On the flanks of the primitive western axis certain ancient sedimentary strata are thrown into folds which were completed before the commencement of the mesozoic period. In the south, the later palaeozoic rocks are also thrown into acute folds by a movement acting from the south, and which ceased towards the close of the mesozoic period. In northern Africa the folded region of the Atlas belongs to the comparatively recent date of the Alpine system. None of these earth movements affected the interior, for here the continental mesozoic deposits rest, undisturbed by folding, on the primary sedimentary and crystalline rocks. The crystalline massif, therefore, presents a solid block which has remained elevated since early palaeozoic times, and against which earth waves of several geological periods have broken.

The formations older than the mesozoic are remarkably unfossiliferous, so that the determination of their age is frequently a matter of speculation, and in the following table the European equivalents of the pre-Karroo formations in many regions must be regarded as subject to considerable revision.

Rocks of Archean age cover wide areas in the interior, in West and East Africa and across the Sahara. Along the coastal margins they underlie the newer formations and appear in the deep valleys and kloofs wherever denudation has laid them bare. The prevailing types are granites, gneisses and schists. In the central regions the predominant strike of the fohae is north and south. The rocks, for convenience classed as pre-Cambrian, occur as several unconformable groups, chiefly developed in the south where alone their stratigraphy has been determined. They are unfossiliferous, and in the absence of undoubted Cambrian, Ordovician and Silurian strata in Africa they may be regarded as of older date than any of these formations. The general occurrence of jasper-bearing rocks is of interest, as these are always present in the ancient pressure-altered sedimentary formations of America and Europe. Some unfossiliferous conglomerates, sandstones and dolomites in South Africa and on the west coast are considered to belong to the Cambrian, Ordovician and Silurian formations, but merely from their occurrence beneath strata yielding Devonian fossils. In Cape Colony the Silurian age of the Table Mountain Sandstone is based on such evidence.

The Devonian and Carboniferous formations are well represented in the north and south and in northern Angola.

Up to the close of the palaeozoic period the relative positions of the ancient land masses and oceans remain unsolved; but the absence of marine strata of early palaeozoic age from Central Africa points to there being land in this direction. In late Carboniferous times Africa and India were undoubtedly united to form a large continent, called by Suess Gondwana Land. In each country the same succession of the rocks is met with; over both the same specialized orders of reptiles roamed and were entombed.

The interior of the African portion of Gondwana Land was occupied by several large lakes in which an immense thickness—amounting to over 18,000 ft. in South Africa—-of sandstones and marls, forming the Karroo system, was laid down. This is par excellence the African formation, and covers immense areas in South Africa and the Congo basin, with detached portions in East Africa. During the whole of the time—-Carboniferous to Rhaetic—that this great accumulation of freshwater beds was taking place, the interior of the continent must have been undergoing depression. The commencement of the period was marked by one of the most wonderful episodes in the geological history of Africa. Preserved in the formation known as the Dwyka Conglomerate, are evidences that at this time the greater portion of South Africa was undergoing extreme glaciation, while the same conditions appear to have prevailed in India

TABLE OF FORMATIONS

Sedimentary. Igneous. Recent Alluvium; travertine; coral; sand dunes; continental } Some volcanic islands; dunes. Generally distributed } rift-valley volcanoes. Pleistocene. Ancient alluviums and } gravels; travertine. } Generally distributed. } A long-continued Pliocene. N. Africa; Madagascar. } succession in the } central and northern Miocene. N. Africa. } regions and among } the island groups. Oligocene. N. Africa. } Doubtfully represented } south of the Zambezi. Eocene. N. Africa, along east and } west coasts; Madagascar. } Cretaceous Extensively developed in } Diamond pipes of S. N. Africa; along coast } Africa; Kaptian and foot-plateaus in east } fissure eruptions; and west; Madagascar. } Ashangi traps of } Abyssinia {Jurassic N. Africa; E. Africa; K{ Madagascar; Stormberg } Chief volcanic period a{ period (Rhaeric) in S. } in S. Africa r{ Africa } r{Trias. Beaufort Series in S. } o{ Africa; Congo basin; } o{ Central Africa; Algeria; } { Tunis. } {Permian. Ecca Series in S. Africa. } Feebly, if anywhere } developed. Carboniferous. N. Africa; Sabaki Shales } in E. Africa; Dwyka } and Wittebery Series in } South Africa } Devonian. N. Africa; Angola; Bokkeveld } Not recorded. Series in S. Africa } Silurian. {Table Mountain Sandstone } { in S. Africa, Silurian(?). } Ordovician. { Doubtfully represented } Klipriversberg and { in N. Africa, French } and Ventersdorp Series Cambrian { Congo, Angola. and by } of the Transvaal (?). { Vaal River and Waterberg } { Series in S. Africa } Pre-Cambrian. Quartzites, conglomerates } phyllites, jasper-bearing } S. Africa and generally. rocks and schists. } Generally distributed. } Archeaan. Gneisses and schists of the } Igneous complex of continental platform. } sheared igneous } rocks;granites.

and Australia. At the close of the Karroo period there was a remarkable manifestation of volcanic activity which again has its parallel in the Deccan traps of India.

How far the Karroo formation extended beyond its present confines has not been determined. To the east it reached India. In the south all that can be said is that it extended to the south of Worcester in Cape Colony. The Crystal Mountains of Angola may represent its western boundary; while the absence of mesozoic strata beneath the Cretaceous rocks of the mid-Sahara indicates that the system of Karroo lakeland had here reached its most northerly extension. Towards the close of the Karroo period, possibly about the middle, the southern rim of the great central depression became ridged up to form the folded regions of the Zwaarteberg, Cedarberg and Langeberg mountains in Cape Colony. This folded belt gives Africa its abrupt southern termination, and may be regarded as an embryonic indication of its present outline. The exact date of the maximum development of this folding is unknown, but it had done its work and some 10,000 ft. of strata had been removed before the commencement of the Cretaceous period. It appears to approximate in time to the similar earth movement and denudation at the close of the palaeozoic period in Europe. It was doubtless connected with the disruption of Gondwana Land, since it is known that this great alteration of geographical outline commenced in Jurassic times.