SUMMARY

The close of the Afghan war of 1839-40 left a great deal to be desired in the matter of practical geography. It was not the men but the methods that were wanting. The commencement of the second and last Afghan war in 1878 saw the initiation of a system of field survey of a practical geographical nature, which combined the accuracy of mathematical deduction with the rapidity of plane table topography. It was the perfecting of the smaller class of triangulating instruments that made this system possible, quite as much as the unique opportunity afforded to a survey department in such a country as India for training topographers. It worked well from the very first, and wherever a force could march or a political mission be launched into such a region of open hill and valley as the Indian trans-frontier, there could the surveyors hold their own (no matter what the nature of the movement might be) and make a "square" survey in fairly accurate detail, with the certainty that it would take its final place without squeezing or distortion in the general map of Asia. This was of course very different from the plodding traverse work of former days, and it rapidly placed quite a new complexion on our trans-frontier maps. Since then regular systematic surveys in extension of those of India have been carried far afield, and it may safely be said now that no country in the world is better provided with military maps of its frontiers than India. In Baluchistan, indeed, there is little left to the imagination. A country which forty years ago was an ugly blank in our maps, with a doubtful locality indicated here and there, is now almost as well surveyed as Scotland. Afghanistan, however, is beyond our line, "out of bounds," and the result is that there are serious gaps in our map knowledge of the country of the Amir, gaps which there seems little probability of investigating under the present closure of the frontier to explorers.

By far the most important of these gaps are the uplands of Badakshan, stretching from the Oxus plains to the Hindu Kush. The plains of Balkh, as far east as Khulm and Tashkurghan, from whence the high-road leads to Haibak, Bamian, and the Hindu Kush passes, are fairly well mapped. The Oxus, to the north of Balkh, is well known, and the fords and passages of that river have been reckoned up with fair accuracy. From time immemorial every horde of Skythic origin, Nagas, Sakas, or Jatas, must have passed these fords from the hills and valleys of the Central Asian divide on their way to India. The Oxus fords have seen men in millions making south for the valleys of Badakshan and the Golden Gates of Central Asiatic ideal which lay yet farther south beyond the grim line of Hindu Kush. Balkh (the city) must have stood like a rock in the human tide which flowed from north to south. From the west, too, from Asia Minor and the Persian provinces, as well as from the Caspian steppes to the north-west, must have come many a weary band of tear-stained captives, transported across half a continent by their conquerors to colonize, build cities, and gradually amalgamate with the indigenous people, and so to disappear from history. From the west came Parthians, Medes, Assyrians, and Greeks, who did not altogether disappear. But no such human tide ever flowed into Badakshan from the east nor yet from the south. To the east are the barrier heights of the Pamirs. No crowd of fugitives or captives ever faced those bleak, inhospitable, wind-torn valleys that we know of. Nor can we find any trace of emigration from India. Yet routes were known across the Pamirs, and in due time, as we have seen, small parties of pilgrims from China made use of these routes, seeking for religious truth in Balkh when, as a Buddhist centre, Balkh was in direct connection with the Buddhist cities of Eastern Turkistan. And Buddhism itself, when it left India, went northward and flourished exceedingly in those same cities of the sandy plain, where the people talked and wrote a language of India for centuries after the birth of Christ. Balkh, however, never stayed the tide which overlapped it and, passing on, lost itself in the valleys under the Hindu Kush, or else, surmounting that range, streamed over into the Kabul basin. Whether the tide set in from north or west, the overflow was forced by purely geographical conditions into precisely the same channels, and in many cases it drifted into the hills and stayed there. What we should expect to find in Balkh, then (whenever Dr. Stein can get there), are records in brick, records in writing, and records in coin, of nearly every great Asiatic movement which has influenced the destinies of India from the days of Assyria to those of Mohamed. What a history to unfold!

Of the Badakshan uplands south and south-east of Balkh, we have but most unsatisfying geographical record. In the days preceding the first Afghan war when Burnes, Moorcroft, Lord, and Wood were in the field, we certainly acquired much useful information which is still all that we have for scientific reference. Moorcroft, as we have seen, made several hurried journeys between Balkh and Kunduz under most perilous conditions, when endeavouring to escape from the clutches of the border chief, Murad Beg. But Moorcroft's opportunities of scientific observation were small, and his means of ascertaining his geographical position were crude, and we gain little or nothing from his thrilling story of adventure, beyond a general description of a desolate region of swamp and upland which forms the main features of Northern Badakshan.

Lord and Wood, who followed Moorcroft at no great interval, and who were also in direct personal touch with Murad Beg under much the same political circumstances, have furnished much more useful information of the routes and passes between Haibak and Kunduz, and given us a very fair idea of the physical configuration of that desolate district. Lord's memoir on the Uzbek State of Kundooz (published at Simla in July 1838) is indeed the best, if not the only, authoritative document concerning the history and policy of Badakshan, giving us a fair idea of the conditions under which Murad Beg established and consolidated his position as the paramount chief of that country, and the guardian of the great commercial route between Kabul and Bokhara; but there is little geographical information in the memoir. The four fortified towns of the Kunduz state, Kunduz, Rustak, Talikhan, and Hazrat Imam, are described rather as depositories for plunder than as positions of any great importance, and the real strength of Murad Beg's military force lay in the quality of his hordes of irregular Uzbek horsemen and the extraordinary hardiness and endurance of the Kataghani horses. So highly esteemed is this particular breed that the late Amir of Afghanistan would permit of no export of horses from Kataghan, reserving them especially for the purpose of mounting his own cavalry.

We learn incidentally of the waste and desolation caused by the poisonous climate of the fens and marshes between Hazrat Imam and Kunduin, to which Murad Beg had transported 20,000 Badakshani families for purposes of colonization, and where Dr. Lord was told that barely 1000 individuals had survived; but Wood tells us much more than this in his charming book on the Oxus. From the point where he left the main road from Kabul to Bokhara (a little below Kuram north of the Saighan valley) till he reached Kunduz, he was passing over country and by-ways which have never been revisited by any European geographer. He tells us that "the plain between the streams that water Kunduz and Kuram has a wavy surface, and though unsuited to agriculture has an excellent pasturage. The only village on the road is Hazrat Baba Kamur. On the eastern side the plain is supported by a ridge of hills sloping down from the mountains to the south. We crossed it by the pass of Archa (so called from the fir trees which cover its crest), from the top of which we had a noble view of the snowy mountains to the east, the outliers of Hindu Kush. Next day we forded the river of Kunduz, and continuing to journey along its right bank, through the swampy district of Baghlan and Aliabad, reached the capital of Murad Beg on Monday the 4th Dec. (1837)." The story of Wood's travels in Badakshan has already been told; the moon-lit march from Kunduz through the dense jungle grass and swamp, often knee-deep in water; the gradual rise to higher ground above; the floating vapour screen that hovered over the fens; Khanabad and its quaint array of colleges and students, and the Koh Umber mountain, isolated and conspicuous, dividing the plains of Kunduz and Talikhan—all these are features which will indicate the general character of that part of Badakshan but leave us no fixed and determined position. The Koh Umber in particular must be a remarkable topographical landmark, as it towers 2500 feet above the surrounding plain with a snow-covered summit. Wood says of it that it is central to the districts of Talikhan, Kunduz, and Hazrat Imam, and its pasturage is common to the flocks of all three plains. But it is an undetermined geographical feature, and still remains in its solitary grandeur, a position to be won by future explorers.

From Khanabad to Talikhan, Faizabad, and Jirm (which, it will be recollected, was once the capital of Badakshan—probably the "Badakshan" of Arab geography), we have the description of a mountainous country supporting the conjectural topography of our maps, which indicate that this route borders and occasionally crosses a series of gigantic spurs or offshoots of a central range (which Wood calls the Khoja) which must itself be a north-easterly arm of the Hindu Kush, taking off from the latter range somewhere near the Khawak Pass. Here, then, is one of the most important blanks in the map of our frontier. Inconceivably rugged and difficult of access, it seems probable that it is more accessible from Badakshan than from the south. We know from Wood's account of the extraordinary difficulty that beset his efforts to reach the lapis-lazuli mines above Jirm in the Kokcha River something of the general nature of these northern valleys and defiles of Kafiristan reaching down to lower Badakshan. It would, indeed, be a splendid geographical feat to fix the position and illustrate the topography of this roughest section of Asia.

Between the Khawak Pass of the Hindu Kush which leads to Andarab, and the Mandal, or Minjan, passes, some 70 miles to the east, we have never solved the problem of the Hindu Kush divide. What lies behind Wood's Khoja range, between it and the main divide? We have the valley called Anjuman, which is believed to lead as directly to Jirm from the Khawak Pass as Andarab does to Kunduz. It is an important feature in Hindu Kush topography, but we know nothing of it. We may, however, safely conjecture that the Minjan River, reached by Sir George Robertson in one of his gallant attempts to explore Kafiristan, is the upper Kokcha flowing past the lapis-lazuli mines to Jirm. But where does it rise? And where on the southern slopes of the Hindu Kush do the small affluents of the Alingar and Alishang have their beginning? These are the hidden secrets of Kafiristan. It is here that those turbulent people (who, by the way, seem to exhibit the same characteristics from whatever valley of Kafiristan they come, and to be much more homogeneous than is usually supposed) hide themselves in their upland villages, amidst their magnificent woods and forests, untroubled by either Afghan or European visitors. Here they live their primitive lives, enlivened with quaint ceremonies and a heathenism equally reminiscent of the mythology of Greece, the ritual of Zoroaster, and the beliefs of the Hindu. Who will unravel the secrets of this inhabited outland, which appears at present to be more impracticable to the explorer than either of the poles? Yule, in his preface to the last edition of Wood's Oxus, remarks that Colonel Walker, the late Surveyor-General of India (and one of the greatest of Asiatic geographers) repeatedly expressed his opinion that there is no well-defined range where the Hindu Kush is represented in our maps, and he adds that such an expression of opinion can only apply to that part of the Hindu Kush which lies east of the Khawak Pass. Sir Henry Yule refers to Wood's incidental notices of the mountains which he saw towering to the south of him, "rising to a vast height and bearing far below their summits the snow of ages," in refutation of such an opinion; and he further quotes the "havildar's" (native surveyor) report of the Nuksan and Dorah passes in confirmation of Wood.

Since Yule wrote, Woodthorpe's surveys of the Nuksan and Dorah passes during the Lockhart mission leave little doubt as to the nature of the Hindu Kush as far west as those passes, but it is precisely between those passes and the Khawak, along the backbone of Kafiristan, that we have yet to learn the actual facts of mountain conformation. And here possibly there may be something in Walker's suggestion. The mountains to which Wood looked up from Talikhan or Kishm, towering to the south of him and covered with perpetual snow, certainly formed no part of the main Hindu Kush divide. Between them and the Hindu Kush is either the deep valley of Anjuman, or more probably the upper drainage of the Minjan, which, rising not far east of Khawak, repeats the almost universal Himalayan feature of a long, lateral, ditch-like valley in continuation of the Andarab depression, marking the base of the connecting link in the primeval fold formed by the Hindu Kush east and west of it. We should expect to find the Kafiristan mountain conformation to be an integral part of the now recognised Himalayan system of parallel mountain folds, with deep lateral valleys fed by a transverse drainage. The long valley of the Alingar will prove to be another such parallel depression, and we shall find when the map is finished that the dominating structural feature of all this wild hinterland of mountains is the north-east to south-west trend of mountain and valley which marks the Kunar (or Chitral) valley on the one side and the Panjshir on the other. The reason why it is more probable that the Minjan River takes the direct drainage of the northern slopes of this Kafiristan backbone into a lateral trough than that the Anjuman spreads its head into a fan, is that Sir George Robertson found the Minjan, below the pass of Mandal, to be a far more considerable river than its assumed origin in the official maps would make it. He accordingly makes a deep indentation in the Hindu Kush divide (on the map which illustrates his captivating book, The Kafirs of the Hindu Kush), bringing it down southward nearly half a degree to an acute angle, so as to afford room for the Minjan to rise and follow a course in direct line with its northerly run (as the Kokcha) in Badakshan. This is a serious disturbance of the laws which govern the structure of Asiatic mountain systems, as now recognized, and it is indeed far more likely that the Minjan (Kokcha) follows those laws which have placed the Andarab and the Panjshir (or for that matter the Indus and the Brahmaputra) in their parallel mountain troughs, than that the primeval fold of the Hindu Kush has become disjointed and indented by some agency which it would be impossible to explain. Who is going to complete the map and solve the question?

We are still very far from possessing a satisfactory geographical knowledge of even the more accessible districts of Badakshan. We still depend on Wood for the best that we know of the route between Faizabad and Zebak; and of those Eastern mountains which border the Oxus as it bends northward to Kila Khum we know positively nothing at all.

But beyond all contention the hidden jewels to be acquired by scientific research in Badakshan are archÆological and antiquarian rather than geographical. Now that Nineveh and Babylon have yielded up their secrets, there is no such field out of Egypt for the antiquarian and his spade as the plains of Balkh. But enough has been said of what may be hidden beneath the unsightly bazaars and crumbling ruins of modern Balkh. Whilst Badakshan literally teems with opportunities for investigation, certain features of ancient Baktria appear to be especially associated with certain sites; such, for instance, as the sites of Semenjan (Haibak), Baghlan, Andarab, and particularly the junction of the rivers at Kasan. That Andarab (Ariaspa) held the capital of the Greek colonies there can be as little doubt as that Haibak and its neighbourhood formed the great Buddhist centre between Balkh and Kabul. Again, who is going to make friends with the Amir of Afghanistan and try his luck? It must be a foreigner, for no Englishman would be permitted by his own government to pass that way at present.

The wild and savage altitudes of Badakshan and Kafiristan by no means exhaust the unexplored tracts of Afghanistan. We have the curious feature of a well-surveyed route connecting Ghazni with Kandahar, one of the straightest and best of military routes trodden by armies uncountable from the days of Alexander to those of Roberts, a narrow ribbon of well-ascertained topography, dividing the two most important of the unexplored regions of Afghanistan. North-west of this road lies the great basin of the central Helmund. South-east is a broken land of plain, ribbed and streaked with sharp ridges of frontier formation, about which we ought to know a great deal more than we do. Up the frontier staircases and on to this plain run many important routes from India. The Kuram route strikes it at its northern extremity and leaves it to the southward. The Tochi valley route, and the great mercantile Gomal highway strike into the middle of it, and yet no one of our modern frontier explorers has ever reached it from one side or the other. We still depend on Broadfoot's and Vigne's account of what they saw there, although it is only just on the far side of the rocky band of hills which face the Indus.

About midway between Ghazni and Bannu is the water-parting which separates the Indus drainage from that of the Helmund, and at this point there are some formidable peaks, well over 12,000 feet in height, to distinguish it. The Tochi passage is easy enough as far as the Sheranni group of villages near the head of its long cultivated ramp, but beyond that point the traveller becomes involved in the narrow lateral valleys which follow the trend of the ridges which traverse his path, where streams curl up from the Birmal hills to the south and from the high altitudes which shelter the Kharotis on the north. It is a perpetual wriggle through steep-sided rocky waterways, until one emerges into more open country after crossing the main divide by the Kotanni Pass. The hills here are called Jadran, and it is probable that the Jadran divide and that of the Kohnak farther south are one and the same. Beyond the Kotanni Pass to Ghazni the way is fairly open, but we know very little about it beyond the historical fact that the arch-raider, Mahmud of Ghazni, used to follow this route for his cavalry descents on the Indian frontier with most remarkable success. The remains of old encampments are to be seen in the plain at the foot of the Tochi, and disjointed indications of an ancient high-road were found on the hill slopes to the north of the stream by our surveyors.

Of the actual physical facts of the Gomul route we have only the details gathered by Broadfoot under great difficulties, and a traveller's account by Vigne. What they found has already been described, and the frontier expedition to the Takht-i-Suliman in 1882 sufficiently well determined the position of the Kohnak water-parting to give a fixed geographical value to their narratives. But we have no topography beyond Domandi and Wana. We know that the ever-present repellent band of rocky ridge and furrow, the hill and valley distribution which is distinctive, has to be encountered and passed; but the route does not bristle with the difficulties of narrow ways and stony footpaths as does the Tochi, and there is no doubt that it could soon be reduced to a very practicable artillery road. The important point is that we do not know here (any more than as regards the upper Tochi) a great deal that it concerns us very much to know. We have no mapping of the country which lies between the Baluch frontier and Lake Abistada, the land of the stalwart Suliman Khel tribes-people, and it is a country of which the possible resources might be of great value to us if ever we are driven again to take military stock of Afghanistan.

But the importance of good mapping in this part of Afghanistan is due solely to its position in geographical relation to the Indian frontier. It is different when we turn to the stupendous altitudes of the high Hazara plateau land to the north of the Ghazni-Kandahar route. With this we are not likely to have any future concern, except that which may be called academic. In spite of the reputation for sterile wind-scoured desolation which the uplands hiding the upper Helmund valleys have always enjoyed, it is not to be forgotten that there are summer ways about them, and strong indications that some of these ways are distinctly useful. Our knowledge of the Helmund River (such knowledge, that is to say, as justifies us in mapping the course of the river with a firm line) from its sources ends almost exactly at the intersection of the parallel of 34° of North latitude with the meridian of 67° East longitude. For the next 120 miles we really know nothing about its course, except that it is said to run nearly straight through the heart of the Hazara highlands.

Two very considerable, but nameless, rivers run more or less parallel to the Helmund to the south of it, draining the valleys of Ujaristan and Urusgan, the upper part of the latter being called Malistan. What these valleys are like, or what may be the nature of the dividing water-parting, we do not know, nor have we any authentic description of the valley of Nawar, which lies under the Gulkoh mountain at the head of the Arghandab, but apparently unconnected with it. Native information on the subject of these highly elevated valleys is excessively meagre, nor are they of any special interest from either the strategic or economic point of view. Far more interesting would it be to secure a geographical map of those northern branches of the Helmund, the Khud Rud, and the Kokhar Ab, which drain the mountain districts to the east of Taiwara above the undetermined position of Ghizao on the Helmund. These mountain streams must rush their waters through magnificent gorges, for the peaks which soar above them rise to 13,000 feet in altitude, and the country is described as inconceivably rugged and wild. This is the real centre and home of the Hazara communities, and, in spite of the fact that there are certain well-ascertained tracks traversing the country and connecting the Helmund with the valley of the Hari Rud, we know that for the greater part of the year they must be closed to all traffic. They are of no importance outside purely local interests. The comparatively small area yet unexplored which lies to the north of the Hazara mountains, shut off from them by the straight trough of the Hari Rud and embracing the head of the Murghab River of Turkistan, is almost equally unimportant, although it would be a matter of great interest to investigate a little more closely the remarkable statements of Ferrier which bear on this region.

When we have finally struck a balance between our knowledge and our ignorance of that which concerns the landward gates of India, we shall recognize the fact that we know all that it is really essential that we should know of these uplifted approaches. They are inconceivably old—as old as the very mountains which they traverse. What use may be made of them has been made long ago. We have but to turn back the pages of history and we find abundant indications which may enable us to gauge their real value as highways from Central Asia to India. History says that none of the tracks which lead from China and Tibet have ever been utilized for the passage of large bodies of people either as emigrants, troops, commercial travellers, or pilgrims into India, although there exists a direct connection between China and the Brahmaputra in Assam, and although we know that the difficulties of the road between Lhasa and India are by no means insuperable. Nor by the Kashmir passes from Turkistan or the Pamirs is it possible to find any record of a formidable passing of large bodies of people, although the Karakoram has been a trade route through all time, and although the Chinese have left their mark below Chitral. Yet we have had explorers over the passes connecting the upper Oxus affluents with Gilgit and Chitral who have not failed, some of them, to sound a solemn note of warning.

Before the settlement of the Oxus extension of the northern boundary of Afghanistan, something of a scare was started by a demonstration of the fact that it is occasionally quite easy to cross the Kilik Pass from the Taghdumbash Pamir into the Gilgit basin, or to climb over the comparatively easy slopes of the flat-backed Hindu Kush by the Baroghel Pass and slip down into the valley of the Chitral. There was, however, always a certain amount of geographical controversy as to the value of the Chitral or of the Kilik approach after the crossing of the Hindu Kush had been effected. Much of the difference of opinion expressed by exploring experts was due to the different conditions under which those undesirable, troublesome approaches to India were viewed. Where one explorer might find a protruding glacier blocking his path and terminating his excursions, another would speak of an open roadway.

From season to season in these high altitudes local conditions vary to an extent which makes it impossible to forecast the difficulties which may obtrude themselves during any one month or even for any one summer. In winter, i.e. for at least eight months of the year, all are equally ice-bound and impracticable, and although the general spirit of desiccation, which reigns over High Asia and is tending to reduce the glaciers and diminish the snowfall, may eventually change the conditions of mountain passages to an appreciable extent (and for a period), it would be idle to speculate on any really important modification of these difficulties from such natural climatic causes. We must take these mountain passes as we find them now, and as the Chinese pilgrim of old found them, placed by Nature in positions demanding a stout heart and an earnest purpose, determined to wrest from inhospitable Nature the merit of a victorious encounter with her worst and most detestable moods, ere we surmount them. To the pilgrim they represented the "strait gate" and "narrow way" which ever leads to salvation, and he accepted the horrors as a part of the sacrifice. To us they represent troublesome breaks in the stern continuity of our natural defences which can be made to serve no useful purpose, but which may nevertheless afford the opportunity to an aggressive and enterprising enemy to spy out the land and raise trouble on the border. We cannot altogether leave them alone. They have to be watched by the official guardians of our frontier, and all the gathered threads of them converging on Leh or Gilgit must be held by hands that are alert and strong. It is just as dangerous an error to regard such approaches to India as negligible quantities in the military and political field of Indian defence, as to take a serious view of their practicability for purposes of invasion.

Beyond this scattered series of rugged and elevated by-ways of the mountains crossing the great Asiatic divide from regions of Tibet and the Pamirs, to the west of them, we find on the edge of the unsurveyed regions of Kafiristan that group of passages, the Mandal and Minjan, the Nuksan and the Dorah which converge on Chitral as they pass southward over the Hindu Kush from the rugged uplands of Badakshan. None of these appear to have been pilgrim routes, nor does history help us in estimating their value as gateways in the mountains. They are practicable at certain seasons, and one of them, the Dorah, is a much-trodden route, connecting what is probably the best road traversing upper Badakshan from Faizabad to the Hindu Kush with the Chitral valley, and it enjoys the comparatively moderate altitude of about 14,500 feet above sea-level. A pass of this altitude is a pass to be reckoned with, and nothing but its remote geographical position, and the extreme difficulty of its approaches on either side (from Badakshan or Chitral), can justify the curious absence of any historical evidence proving it to have witnessed the crossing of troops or the incursions of emigrants. For the latter purpose, indeed, it may have served, but we know too little about the ethnography or derivation of the Chitral valley tribes to be able to indulge in speculation on the subject.

What we know of the Dorah is that it is the connecting commercial link between Badakshan and the Kunar valley during the summer months (July to September), when mules and donkeys carry over wood and cloth goods to be exchanged for firearms and cutlery with other produce of a more local nature, including (so it is said) Badakshi slaves. It has been crossed in early November in face of a bitter blizzard and piercing cold, but it is not normally open then. The Nuksan Pass, which is not far removed from it, is much higher (16,100 feet) and is frequently blocked by glacial ice; but the Dorah, which steals its way through rugged defiles from the Chitral valley over the dip in the Hindu Kush down past the little blue lake of Dufferin into the depths of the gorges which enclose the upper reaches of the Zebak affluent of the great Kokcha River of Badakshan, (about which we have heard from Wood), is the one gateway which is normally open from year to year, and its existence renders necessary an advanced watch-tower at Chitral. Like the Baroghel and other passes to the east of it, it is not the Dorah itself but the extreme difficulty of the narrow ways which lead to it, the wildness and sterility of the remote regions which encompass it on either side, which lock this door to anything in the shape of serious military enterprise.

Beyond the Dorah to the westward, following the Kafiristan divide of the Hindu Kush, we may well leave unassisted Nature to maintain her own work of perfect defence, for there is not a track that we can discover to exist, nor a by-way that we can hear of which passes through that inconceivably grand and savage wilderness of untamed mountains. Undoubtedly such tracks exist, but judging from the remarkable physical constitution of the Kafir, they are such as to demand an exceptional type of mountaineer to deal with them. It is only when we work our way farther westward to those passes which lead into the valleys of the upper Kabul River affluents, from the Khawak Pass at the head of the Panjshir valley to the Unai which points the way from Kabul to Bamian, that we find material for sober reflection derived from the records of the past.

The general characteristics of these passes have been described already—and something of their history. We have seen that they have been more or less open doors to India through the ages. Men literally "in nations" have passed through them; the dynasties of India have been changed and her destinies reshaped time after time by the facilities of approach which they have afforded; and if the modern conditions of things military were now what they were in the days of Alexander or of Baber, there would be no reason why her destinies should not once again be changed through use of them. We must remember that they are not what they have been. How far they have been opened up by artificial means, or which of them, besides the Nuksan and the Chahardar, have been so improved, we have no means of knowing, but we may take it for granted that the Public Works Department of Afghanistan has not been idle; for we know that that department was very closely directed by the late Amir, and that his staff of engineers is most eminent and most practical.[13]

The base of all this group of passes lies in Badakshan, so that the chief characteristics as gates of India are common to all. It has been too often pointed out to require repetition that the plains of Balkh—all Afghan Turkistan in short—lie at the mercy of any well-organized force which crosses the Oxus southwards; but once that force enters the gorges and surmounts the passes of the Badakshan ramparts a totally new set of military problems would be presented. The narrowness and the isolation of its cultivated valleys; the vast spaces of dreary, rugged desolation which part them; the roughness and the altitude of the intervening ranges—in short, the passive hostility of the uplands and their blank sterility would create the necessity for some artificial means of importing supplies from the plains before any formidable force could be kept alive at the front. Modern methods point to military railways, for the ancient methods which included the occupation of the country by well-planted military colonies are no longer available. All military engineers nowadays believe in a line, more or less perfect, of railway connection between the front of a field force and its base of supply. But it would be a long and weary, if not absolutely hopeless, task to bring a railway across the highlands of Badakshan to the foot of the Hindu Kush from the Oxus plains.

We have read what Wood has to say of the routes from Kunduz southward to Bamian and Kabul. This is the recognized trade route; the great highway to Afghan Turkistan. Seven passes to be negotiated over as many rough mountain divides, plunges innumerable into the deep-rifted valleys by ways that are short and sharp, a series of physical obstacles to be encountered, to surmount any one of which would be a triumph of engineering enterprise. Amongst the scientific devices which altitude renders absolutely necessary, would be a repeated process of tunnelling. No railway yet has been carried over a sharp divide of 10,000 or 11,000 feet altitude, subject to sudden and severe climatic conditions, without the protection of a tunnel. As a work of peaceful enterprise alone, this would be a line probably without a parallel for the proportion of difficulty compared to its length in the whole wide world. As a military enterprise, a rapid construction for the support of a field army, it is but a childish chimera. Yet we are writing of Badakshan's best road!

It is true that by the Haibak route to Ghori and that ancient military base of the Greeks, Andarab, the difficulty of the sheer physical altitude of great passes is not encountered, and there are spaces which might be pointed out where a light line could be engineered with comparative facility. Even to reach thus far from the Oxus plains would be a great advantage to a force that could spend a year or two, like a Chinese army, in devising its route, but this comparative facility terminates at the base of the Hindu Kush foot-hills; and it matters not beyond that point whether the way be rough or plain, for the wall of the mountains never drops to less than 12,500 feet, and no railway has ever been carried in the open over such altitudes. Tunnelling here would be found impossible, owing to the flat-backed nature of the wide divide. With what may happen in future military developments; whether a fleet of air-ships should in the farther future sail over the snow-crested mountain tops and settle, replete with all military devices in gunnery and stores, on the plains of the Kohistan of Kabul we need hardly concern ourselves. It is at least an eventuality of which the risk seems remote at present, and we may rest content with the Hindu Kush barrier as a defensive line which cannot be violated in the future as it has been in the past by any formidable force cutting through Badakshan, without years of preparation and forewarning.

For any serious menace to the line of India's north-western defence we must look farther west—much farther west—for enough has been said of the great swelling highlands of the Firozkohi plateau, and of the Hazara regions south of the Hari Rud sources, to indicate their impracticable nature as the scene of military movement. It is, after all, the highways of Herat and Seistan that form the only avenues for military approach to the Indian frontier that are not barred by difficulties of Nature's own providing, or commanded from the sea. Once on these western fields we are touching on matter which has been so worn threadbare by controversy that it might seem almost useless to add further opinions. Historically it seems strange at first sight that, compared with the northern approaches to which Kabul gives the command, so very little use has been made of this open way. It was not till the eighteenth century saw the foundation laid for the Afghan kingdom that the more direct routes between Eastern Persia and the Indus became alive with marching troops. The reason is, obviously, geographical. Neither trade, nor the flag which preceded it from the west, cared to face the dreary wastes of sand to the south of the Helmund, backed, as they are, by the terrible band of the Sind frontier hills full of untamed and untameable tribes, merely for the purpose of dropping into the narrow riverain of the lower Indus, beyond which, again, the deserts of Rajputana parted them from the rich plains of Central India. When the Indus delta and Sind were the objective of a military expedition, the conquerors came by way of the sea, or by approaches within command of the sea—never from Herat. Herat was but the gateway to Kandahar, and to Kabul in the days when Kabul was "India."

It was not, so far as we can tell, till Nadir Shah, after ravaging Seistan and the rich towns of the Helmund valley, found a narrow passage across the Sind frontier hills that any practical use was ever made of the gates of Baluchistan. Although there are ethnological evidences that a remnant of the Mongol hordes of Chenghis Khan settled in those same Sind hills, there is no evidence that they crossed them by any of the Baluch passes. It seems certain that in prehistoric times, when the geographical conditions of Western India were different from what they are now, Turanian peoples in tribal crowds must have made their way into India southwards from Western Asia, but they drifted by routes that hugged the coast-line. We have now, however, replaced the old natural geographical conditions by an artificial system which totally alters the strategic properties of this part of the frontier. We have revolutionised the savage wilderness of Baluchistan, and made highways not only from the Indus to the Helmund, but from Central India to the Indus. The old barriers have been broken down and new gateways thrown open. We could not help breaking them down, if we were to have peace on our borders; but the process has been attended with the disadvantage that it obliges us to take anxious note of the roads through Eastern Persia and Western Afghanistan which lead to them.

For just about one century since the first scare arose concerning an Indian invasion by Napoleon Bonaparte, have we been alternating between periods of intense apprehension and of equally foolish apathy concerning these Western Indian gateways. The rise and fall of public apprehension might be expressed by a series of curves of curious regularity. At present we are at the bottom of a curve, for reasons which it is hardly necessary to enter into; but it is not an inapt position for a calm review of the subject. There is, then, one great highway after passing through Herat (which city is about 60 miles from the nearest Russian military post), a highway which has been quite sufficiently well described already, of about 360 miles in length between Herat and Kandahar; Kandahar, again, being about 80 miles from our frontier—say 500 miles in all; and the distinguishing feature of this highway between Russia and India is the comparative ease with which that great Asiatic divide which extends all the way from the Hindu Kush to the Persian frontier (or beyond) can be crossed on the north of Herat. There, this great central water-parting, so formidable in its altitudes for many hundreds of miles, sinks to insignificant levels and the comparatively gentle gradients of a debased and disintegrated range. This divide is parted and split by the passage of the Hari Rud River; but the passage of the river is hill-enclosed and narrow, with many a rock-bound gorge which would not readily lend itself to railway-making (although by no means precluding it), so that the ridges of the divide could be better passed elsewhere.

We must concede that, taking it for all in all, that 500 miles of railway gap which still yawns between the Indian and Russian systems is an easy gap to fill up, and that it affords a road for advance which (apart from the question of supplies) can only be regarded as an open highway. Then there is also that other parallel road to Seistan from the Russian Transcaspian line across the Elburz mountains (which here represents the great divide) via Mashad—a route infinitely more difficult, but still practicable—which leads by a longer way to the Helmund and Kandahar. Were it not for the political considerations arising from the respective geographical positions of these two routes, one lying within Persian territory and the other being Afghan, they might be regarded as practically one and the same. Neither of them could be used (in the aggressive sense) without the occupation of Herat, and most assuredly should circumstances arise in which either of the two should be used (in the same aggressive sense) the other would be utilized at the same time.

This is, then, the chief problem of Indian defence so far as the shutting of the gate is concerned, and there are no two ways of dealing with it. We must have men and material sufficient in both quantity and quality to guard these gates when open, or to close them if we wish them shut. The question whether these western gates should remain as they are, easily traversable, or should yield (as they must do sooner or later) to commercial interests and admit of an iron way to link up the Russian and Indian railway systems is really immaterial. In the latter case they might be the more readily closed, for such a connection would serve the purposes of a defence better than those for offence; but in any case in order to be secure we must be strong.