The Cape-to- Railway and Train Ferries Author(s): H. Wilson Fox Source: The Geographical Journal, Vol. 55, No. 2 (Feb., 1920), pp. 73-101 Published by: geographicalj Stable URL: http://www.jstor.org/stable/1781582 Accessed: 27-06-2016 09:58 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Wiley, The Royal Geographical Society (with the Institute of British Geographers) are collaborating with JSTOR to digitize, preserve and extend access to The Geographical Journal

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms The Geographical Journal

Vol. LV No. 2 February I920

THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

DEVELOPNMENT OF THE GREAT LAKES OF BY TRAIN-FERRIES

H. Wilson Fox, M.P. Read at the,Meeting of the Soiety, I December 19I9. Maf, p. 160. OW that the Great War is over and the control of the territory which is still known as has passed out of German hands, never, I hope, to return, renewed interest will certainly be taken in the realization of ' dream-the Cape-to-Cairo Railway. No definite route for it has hitherto been laid out, and, broadly speaking, there is considerable choice of alternatives over the greater part of the distance which still separates the railways of from those of the Egyptian and systems. The subject of'railway extensions in Africa was discussed at some length in the comprehensive and most interesting paper on the "Railway Development of Africa Present and Future" (G./., Jan. I916) which was read before the Society by Sir Charles Metcalfe on 29 November I9I5, and the same topic was referred to, in perhaps more imaginative fashion, in a paper by Sir Harry Johnston (G./., April I9I5), whose horizon of fancy was not bounded by the limitations from which a great engineer, like Sir Charles Metcalfe, could obviously not escape. The most recent discussion before this Society of some of the special problems of the Cape-to-Cairo project arose in I9I8 in connection with Sir Alfred Sharpe's paper "The Backbone of Africa" (G.y., Sept. I918). My own long association with Mr. Rhodes, and the work I have had in connection with the development of Rhodesia and its railway system, have naturally caused me to take a special interest in these problems, but, beyohd a general conviction that as the interior of Central Africa developed our experience in Rhodesia would be repeated, and that the prospects of trade immediately in advance of the northern and southern railheads would continue to draw each of them forward until at last they would meet sooner or later, I had not been able until quite recently to visualize any method by which this natural process of extension might conceivably be accelerated. I have, however, always kept myself G

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 74 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

informed, so far as possible, in regard to the geographical features of the regions through which the line might conceivably pass, as to which a number of reports by competent observers have been received in recent years. The inspiration of the practical suggestions made in this paper flows from my recent work as Chairman of the Select Committee on Internal Transport which was set up by the House of Commons during the last session of the last Parliament. I had then an opportunity of seeing, and acquainting myself with, the conditions of working of the important train- ferries which are now plying regularly between this country and France, and of considering the possibility of an extension of the use of train-ferries to the Irish Channel. The studies which I then made of the conditions in which train-ferries are used for commercial purposes upon the Great Lakes of America and other places, soon suggested the question--Why should they not be used upon the Great Lakes of Africa? I hope that when you have heard what I have to say to-night you will agree with me that the answer to this question should be-They should ! There are probably some people in this audience who are well acquainted with train-ferries, but I venture to say that their knowledge is quite exceptional, and that not one Englishman in ten thousand, at any rate before the war, knew anything whatever about them. I think, there- fore, that before I pass to the particular problems of the Cape-to-Cairo Railway, it may interest you to hear some of the facts and conclusions which I have gleaned from my investigations upon the subject of train-ferries in general. The modern train-ferry is no fair-weather boat. It can brave seas quite as wild as any that are met with in the English or Irish Channels, and is capable-though fortunately it will not have to do so in Africa--of cutting its way through considerable thicknesses of ice. Train-ferries have for some time past played a most important part in the development of the railway system of Denmark, the distances over which trains are carried by this means across arms of the sea varying from Ii to 26 miles. On the Danish system they are regarded as quite ordinary features of railway con- struction, and are looked upon in just the same light as a bridge or tunnel. Perhaps the most ambitious marine project for their use is the proposed establishment of a train-ferry service between Gothenburg and Immingham -a distance of 500 miles; that such a project can be considered from a commercial standpoint by serious people, notwithstanding the terrors of a North Sea crossing, is, I think, sufficient to show that the train-ferry has passed completely out of the experimental stage. For my purpose to-night it is perhaps even more instructive to refer to the extensive use which has been made of train-ferries on great lakes. On Lake Baikal, on the route of the Siberian Railway, a large ice-breaking ferry, which was built in this country by Messrs. Armstrong, Whitworth & Co., was operated successfully for some years. But for purpose of comparison

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 75 with Africa, I think that American experience offers the best guidance, and particularly that which has been obtained on Lake Michigan. One of the services maintained on this lake covers a distance of 240 miles, and there are numerous services of from 50 to IOO miles. A high authority has stated " that the Ann Arbor train-ferry line on Lake Michigan has demon- strated that in this region, freight (or goods) can be transported 60 miles for about the labour cost of transferring it from cars to boat on one side of the lake, and from boat to cars on the other side." It is also stated, as the result of averages ascertained from the working results of eight different ferry lines in different parts of the world, that the mean expense per mile traversed by a ferry was, in pre-war days, about 8s. 6d. The Ann Arbor line is said to have been worked on a very profitable basis, a four-years average showing that the average percentage of total expenses to total earnings was only 26'8 per cent. The Ann Arbor Railroad Company, at the time to which these figures relate, operated four ferry lines running respectively for go, 8o, 78 and 60 miles, and owned three boats, each provided with four rail-tracks. Their carrying capacity was about 22 cars of 36 feet in length. The total number of cars ferried in I90or was 27,240, and the total goods transported in that year 4I5,600 tons. It may be of interest to note, in comparison with these figures, that the train-ferries on the River Parana, by means of which the Entre Rios Rail- way Company maintains its services between the province of Entre Rios and Buenos Aires, transported in the year ended 30 June I919 380,344 tons of goods, 66,239 passengers and 370,754 animals. The running cost of a ferry boat per mile was 7s. 2d. The Cross-Channel services between this country and France, which were established for war purposes by Sir Eric Geddes with very great success, are maintained between Richborough and Dunkirk, Richborough and Calais, and Southampton and Dieppe. A fourth between Southamp- ton and Cherbourg is also being brought into operation. Three of the ferry boats used have a length of 363 feet, a beam of 61-} feet, a tonnage displacement of 3654 tons, and a draught of I2 feet. Their sea- going speed is I2 knots. They have four rail-tracks, making a total track- length of I080 feet, and they carry 54 xo-ton trucks. Loading and unloading takes place at the after end only. The initial cost of each boat was 1I86,940, and they were designed for goods traffic only. The provision of special berths at the terminals for the reception of the ferries is obviously an essential feature of a train-ferry service. The cost necessarily varies with the conditions existing in each locality, but it generally represents a substantial proportion of the capital outlay. The cost of a single berth at the Channel ports was about ?85,ooo. A ferry boat on arrival at its berth must be moored accurately in position, and held as stationary as possible, so that the rails on the adjustable bridge may make a true connection with those on the deck of the vessel. An

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 76 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

adjustable communication-bridge has to be provided at each berth in order to overcome the changes in vertical height due sometimes to changes in tide-level and the water-level in dock basins, and sometimes to the vertical movement and heel of the vessel caused by the movements of the loads. These bridges are hinged at the shore end, and rest upon the stern of the vessel at the other end, forming a varying gradient up or down, accord- ing to the variation of the water-level. The working limit of this gradient is assumed at I in 20. The range of tide at the Channel ports varies from 8 to 12 feet. To those of you who wish to pursue the subject of train-ferries further than I can carry it to-night, I recommend perusal of a book called 'The Channel Ferry.' It is published by Messrs. Harrison & Sons, of 45, Pall Mall, to whose courtesy I am indebted for the use of some of the photo- graphs which I have shown to you. By consideration of these photographs, and of the figures which will be set out in the printed paper, I am sure you will be convinced, as I have been, that the train-ferry of to-day is no toy, or small convenience for passengers, but a commercial instrument of great capacity and utility, for the transport on a large scale of both passengers and goods at a lower cost, in suitable conditions, both in expense and time, than is involved in the construction and working of railways along the shores of great lakes. Alternatively, transhipment of goods from a train to an ordinary freight steamer and vice versd, which is always a costly process, can be avoided by their use. The distance from Capetown to Cairo, as the crow flies, is about 4200 miles. In the popular imagination the Cape-to-Cairo railway of the future no doubt presents itself as a continuous line of railway running more or less straight between these points, with branches connecting the central trunk line with the various ports upon the east and west coasts of the continent. In days to come, I think it far more probable that over the greater portion of this distance it will be more correct to speak of the Cape-to-Cairo system, than of the Cape-to-Cairo line, because there will certainly be a large number of alternative routes over which passengers proceeding between the two terminals will be able to travel. The railway map of South and Central Africa illustrates this point very clearly. As far north as , to which in no very distant future there will certainly be two alternative routes from the south, there is a regular network of railways connecting the chief centres of South Africa, and although the western line to the Cape, running vig Mafeking and Kimberley, will probably always be somewhat shorter in time from Bulawayo, there can be no doubt that large numbers of passengers coming from the north will proceed south rig Pietersburg, , and , from which in turn they will have a choice of routes to the Cape. There will, moreover, probably be nodal points on the Cape-to-Cairo system through which every traveller will pass, of which one will probably be Bulawayo, and another . If the whole route be regarded for

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms , LAKE BAIKAL

TRAIN FERRY, RICHBOROUGH

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE FERRY BERTH, RICHXBOROU GH

-w;D ; <0rD;0

DECK OF THE RICHBOROUGH FERRY

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 77

convenience as divided into three zones, Bulawayo may be taken as the northern limit of the southern zone, and the north end of Victoria Nyanza as the northern limit of the central zone. In the northern zone, which comprises the country lying to the north of Lake Albert, and so includes and the Sudan, the geographical conditions are such that there are unlikely to be any large number of alternative routes, such as already exist in the southern zone, and, as I shall show, will probably be established in the central zone. My principal aim this evening is to consider the means and the routes by which rail connection between Capetown and Cairo can be established soonest. The route will not, I think, be in the first instance an all-British route, though I hope that when through communication by railway has once been established by some route, the construction of the links needed to complete an all-British route will not be long delayed. I for one shall not consider that Rhodes' ideal has been -fully realized until this has been done. Of the three zones, the one with which I propose to concern myself primarily is the central zone, in which are situated those of the great lakes of Africa which are the mnost likely to be utilized in connection with the Cape-to-Cairo system, viz. Nyasa, , and Victoria Nyanza. The altitudes above sea-level of these lakes are respectively: Nyasa I520 feet; Tanganyika, 2550 feet; and Victoria Nyanza, 3726 feet. For the purposes of train-ferry services, these and other large lakes, in com- parison with many marine services, have the great advantage that they are not tidal. This fact does not, however, necessarily mean that provision need not be made at the landing-stages for appreciable changes of level, which may be due either to the seasonal rainfalls, or to the action of long- continued winds. Fellows of this Society will, no doubt, remember that the effect of wind upon the levels of inland lakes has frequently been referred to by previous lectures, notably in connection with Lopnor in Central Asia, Lake Chad in west Central Africa, and Lake Nyasa. The Admiralty chart of Victoria Nyanza fixes the range of change in the water-level at Kisumu (Port Florence) at 6 feet, and I am informed by Sir Alfred Sharpe that the range on Lakes Tanganyika and Nyasa probably does not exceed 4 feet. The general characteristics of these two lakes have been described by Sir Alfred Sharpe as follows:

" From the south to the north end of Tanganyika, a distance of 400 miles (as compared with 360 miles, the length of Lake Nyasa), steamers of practically any draught can run. Both these lakes are very deep, and in each case the northern section is deeper than the southern. Soundings of 360 fathoms have been found in Tanganyika, and only a little less in the northern half of Nyasa, the bottom of the latter lake being in fact considerably below sea-level. In the matter of harbours, Tanganyika is well supplied, Nyasa being not so well off in this respect. There are, however, in both lakes many places where a large

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 78 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

steamer could lie in deep water alongside wharfs which could be easily constructed."

"t When I traversed Tanganyika from south to north in I9I2, I travelled in the German steamer Hedwi~. von Wissmann, which some time after the commencement of the war was satisfactorily sunk by motor-boats sent out from England by the Admiralty for this special purpose. The voyage is most interesting, and as a rule the lake has not so much rough weather as one experiences on Nyasa. On the latter lake the south-east trades forced into the trough of Nyasa become south winds, and, blowing straight up the lake, raise a very unpleasant sea towards the north end. On Tanganyika the south-east trades are not so pronounced, and do not last so long. As a rule calm weather prevails, varied by occasional squalls and storms. The waters of Lake Nyasa are of a deep blue colour, those of Tanganyika are green. I have no reason to offer as the cause of this difference, but it is very noticeable."

Victoria Nyanza also possesses a number of harbours which are capable of accommodating vessels of large size. Owing, however, to its greater area and configuration, its waters are, generally speaking, more exposed to the action of wind, and on its western shores, owing to the prevalence of strong south-east and north-east gales, and the extent of swamp, there are comparatively few points at which ports suitable for railway termini could be constructed. I return now to the consideration of the possible railway routes from Bulawayo northwards with a view to deciding over which of them con- nection with Cairo is most likely to be established first. Looking at the railway routes to the north from this centre, the casual observer would naturally expect, seeing that the line from Bulawayo via the has already been carried to Bukama-a distance north estimated in degrees of latitude of about 780 miles-that the extension of this line further northwards from some point, notably Bukama, probably offers the best prospect of bridging speedily the long gap. which to-day separates the railways of Egypt and the Sudan from those of central and . For reasons which I shall give later, it would seem that if this is to be the earliest line of approach to the north, connection by rail will probably have to be established between Bukama and Albertville on Lake Tanganyika, vi Kabalo. The link from Bukama to Kabalo has not, however, been included as yet in the official programme, though a short extension of about. 60 miles to Kibombo is contemplated. If then we must assume that progress northward in this direction is not likely to take place for some time, what are the alternatives ? One would be to connect Rhodesia Broken Hill with the south end of Tanganyika, which would involve the construction of about 600 miles of line. As to this route, it may be remembered that Major E. A. Steel expressed the view, in a paper read before this Society (G.y., Sept. I9I7), that it will not be a practicable route. I ventured at the time to express a doubt as to the correctness of

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 79 this view. I have since conferred with a number of those who have been engaged during the war in transporting very large supplies over this route by motor lorries, and have come to the conclusion that if it is desired to

Sketch-map of Railways and Routes from Northern Rhodesia to Lake Tanganyika

construct this railway, it will be quite practicable from an engineering standpoint, and not particularly costly, to do so. Amongst those who have given me direct information on this matter is Lieut.-Colonel Murray, D.S.O.,

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 8o THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES a most gallant and distinguished officer who commanded the Rhodesia Police and other forces on the north-eastern frontiers of Rhodesia, and in German East Africa throughout the war. Colonel Murray, who is now in this country, has, I regret to say, suffered very severely in health as the result of his most arduous and trying services, and would not be able to undergo the fatigue involved in the reading of a paper before this Society. He has, however, supplied me with some valuable geographical notes which will be printed as an appendix to this paper. Colonel Murray says, upon the subject which I am now discussing:

" A motor road runs from Kashitu north of Broken Hill vid Serenje Mpika to Kasama and thence to Abercorn, a distance of approximately 500 to 600 miles. The country south of Kasama is perfectly flat, and the Chambezi river is the only large river to be bridged. To give some idea of how simple railway construction is in this country, I may state that the motor road was started in the latter end of January I916, and by May I916 motor transport was running between Kasama and Kashitu, a distance approximately of 500 miles; this work having been carried out during the wet season. From Kasama to Abercorn (distance ioo00 miles) the railway would run along the watershed; there are no rivers to be bridged, and a good, well-graded road already exists. The country along the whole route is well wooded. Labour is also plentiful. From Abercorn the railway would connect up with Kitutu .Bay, Lake Tangan- yika, a distance approximately of I 5 miles, thence another 15 miles by water to Bismarckburg."

Both the routes to Lake Tanganyika which I have mentioned involve a considerable deflection to the west from the true north, because, in order to reach either Rhodesia Broken Hill or Bukama, the stretch of line which runs for 3oo miles to the north-west from Bulawayo to the Victoria Falls has to be traversed. In my recent studies of this question, I have wondered whether a third line of approach to the north, starting from Bulawayo, may not prove to be the route by which communication between Capetown and Cairo will be first established. This route would follow the line to the north-east from Bulawayo to Salisbury, and thence to Beira. From Beira arrangements are in an advanced stage for the construction of a line over a distance of about I90 miles to the near Sena, which would inevitably lead to the early construction of a bridge over the river to connect the new railway with the Shire Highlands railway, of which the southern end is on the Zambezi at Chindio, and the northern terminus is now at Blantyre, in the Nyasaland Protectorate. The decision has been taken recently by the Colonial Office to extend immediately this line northwards from Luchenza to Pagonas on the south- eastern shore of Lake Nyasa. At this point deep water can be reached by a pier. As, however, an extension of only 3 miles further would carry the railway to Malindi, where vessels of large size could lie in deep water against wharves constructed along the shore, and all the surroundings are suitable for the requirements of a busy commercial port, it may be

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 81

^

f..A fLWAYS OPE -...... - V PROJECTE D.....JOQK /f TRAIW ' FERRIES ...... _ ...... _ - _ rF Q I R A

Sketch-map of railways and routes, Beira to Lake Nyasa

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 82 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

hoped that the penny-wise-pound-foolish policy of terminating the line at Pagonas will be reconsidered. From Malindi the train could be carried by train-ferry to the north end of the lake, either to Mwaya at the north-west corner of the lake in the territory still spoken of as German East Africa, or to Karonga, which is also on the north-west shore of the lake in the Nyasaland Protectorate Of these two ports, Mwaya, which is a far superior anchorage, and is also more accessible from the land, will probably be the one chosen for the terminus. The construction of a line from this port to Bismarckburg (Kasanga), at the south end of Tanganyika, over a distance of about 200oo miles, would present no special engineering difficulties, and thence the train could be carried by train-ferry to Kigoma, the terminus of the railway from the coast which runs from Dar-es-Salaam to the interior.

SCALt? l , , ,,ut,

RItLWAf PCU/Cr?0D....OOCO 1 /J

Sketh-map of Railway Routes between Nyasa and Tanganyika Sketch map of Railway Routes between Nyasa and Tanganyika

Writing in reference to the alternative routes by which communication could be established between Lakes Nyasa and Tanganyika, Colonel Murray states emphatically that the Mwaya-Bismarckburg, or, as I prefer to call it, the Mwaya-Kasanga, route "is the most suitable for railway construction. It runs along the north face of the watershed, a well-graded road already exists, and the only disadvantage is that there are a few rivers to be bridged. Bismarckburg would make an excellent port, with deep water right up to the shore. The crossing of the mountains vid the Igale Pass to Neu Langenburg, and thence to Mwaya, is no difficult

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 83 feat. At Mwaya it would be easy to get a suitable harbour; the shipping also would be sheltered from the storms." Of the south end of Lake Tanganyika, Colonel Murray writes, "In most parts the mountains run right down to the lake-shore; there is, however, the open harbour of Kitutu Bay at the most southern point, about 15 miles from Abercorn, whiclh can be reached by motor. This harbour is in Northern Rhodesia. About 15 miles from Kitutu Bay is the old German post, Bismarckburg, in German East Africa, native name 'Kasanga,' which is a first-rate anchorage, with deep water right up to the shore." If the programme which I have just indicated were carried out, railway connection between Bulawayo and the south end of Lake Tangan- yika would be effected by the construction of about 530 miles of new railways, and the establishment of a train-ferry service on Lake Nyasa, as against the construction of about 600 miles of railway from Rhodesia Broken Hill to Lake Tanganyika. Either of these routes, once the lake were reached, could be utilized for further progress northwards, and would give the same result as would be obtained by railway connection between Bukama and Albertville. I am inclined to think that, from the point of view of immediate commercial utility, the route I have indicated vit Lake Nyasa would be the most immediately valuable. In other words, the expectations of early remunerative traffic would be better. Comparative figures for the three alternative routes from Bulawayo are given in Fig. I. In the consideration of alternative routes northwards from Lake Tan- ganyika, I have had the advantage of being able to refer to certain notes, hitherto unpublished, by Captain J. E. Philipps, M.C., who was for some time Chief Political Officer in the area, and served during the war in the Intelligence Branch of the East Africa Expeditionary Force. Captain Philipps is now in the service of the Uganda Protectorate. His notes constitute a valuable addition to our geographical knowledge of the country lying to the north-east of Lake Tanganyika, and will be printed as an appendix to this paper. From Lake Tanganyika northwards the most obvious route for further extension, judging solely by the ordinary maps, would be rig Lakes Kivu, Edward, and Albert. Unfortunately, however, the geography of the country is such, by reason of the active volcanoes near the north end of Lake Kivu, that further consideration of this as a through route must probably be dismissed. Fellows of this Society will no doubt recollect the most vivid description of an eruption in this region which was given by Sir Alfred Sharpe in I9I5 on his return from a visit to it in I912-I3 (G.y, Jan. 1916). Captain Philipps confirms the view that this route is impracticable, and writes of the country lying to the north of Tanganyika as follows: " From Tanganyika northwards no use can be made of the valleys or lakes in the central rift Kivu-Albert. The Rusisi River rushes down a

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 84 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES deep gorge closed in by an enormous complication of volcanic mountains stretching both east and west to and the Congo respec- tively. Northward Kivu is cut off from Lakes Edward and Albert by a

Fig. I.-Routes between Bulawayo and Kigoma, Tanganyika range of active volcanoes, running up to I4,oo000 feet and thrown laterally across the divide. From Tanganyika use will have to be made of the from Kigoma to Tabora." Captain Philipps' condemnation of the Central Rift Valley for railway

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 85

construction is probably, however, too sweeping, seeing that other geo- graphers, and notably Sir Alfred Sharpe, are of opinion that the connection of Lakes Kivu and Edward by rail along the course of the Rusisi River would be quite practicable. In the face of the physical obstacles which bar through communication with the north vi8 the Central Rift Valley, Captain Philipps states that, if it is desired to link up the South African and Sudan systems by some route lying to the west and north of Tanganyika, it will be necessary to make extensive use of the existing and proposed railway systems in the Congo territory, to which I have already referred. In discussing the possibilities of extension northwards from Bukama, I assumed that the route from that centre to northern Africa would pass along the valley of the River Lualaba to Kabalo, and thence to Albertville and Kigoma. There is, however, a possible alternative. The map of projected railways in the shows that it is proposed to connect Kabalo with the administrative centre at Stanleyville by a rail- way along the Congo valley to Kongolo, and thence to Kindu. From Kindu to Ponthierville river transport is already utilized, and a railway connects Ponthierville with Stanleyville. I do not know whether railway construction between Ponthierville and Kindu is either contemplated or practicable. From Stanleyville a railway is projected to Mahagi on Lake Albert, at which point the line would enter what I have called the Northern Zone. It will be observed that this alternative route would lie entirely in Belgian territory from the point where the Rhodesian line enters it at Sakania, and that passengers by it would pass far to the westward of the great lakes until Mahagi was reached. With regard to the regions lying to the east and north of Tanganyika, Captain Philipps states that

" For many years it has been definitely laid down by authorities that in the rectangle formed by the southern boundary of Uganda; the German Central Railway; the western shore of Lake Victoria; and the line from Lake Kivu to Ujiji on Tanganyika, it will not be possible to construct a rai!way to connect the central line with the railway system of Uganda. This is not correct. " It is true that the country is very mountainous, showing a mass of extreme volcanic complications over the whole area indicated, with no clear systems of ranges or valleys. No European has hitherto had an opportunity of spending any time in this area with the exception of the few German officials. In Ruanda and Urundi there are no white traders or settlers. Before the war the country was a closed area from the time the Germans opened political stations in 191O-I12. " I have been for five months engaged there on a political mission up to March I917. It will be noticed on every existing map that the country north- ward from the terminus (Kasseke) of the projected railwayremains, by acci- dent or design, quite uncharted. British maps are recent compilations from German sources, and from the northern edge of this area up to the Uganda frontier the map is, in many instances, curiously incorrect. I was able, when my duties and the unsettled state of the country permitted, to .

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 86 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES confirm the existence in the uncharted portion lying parallel to the course of the Kagera, of a remarkable valley * which provides an unexpected avenue of communication between Kasseke and the Kakitumba Valley. Thence access is assured to the Rushenyi plains of Uganda and the Fort Portal system of rail. Thence also north-westward through the Ntungwe valley, if required, to Lake Edward, and also to Stanleyville on the main Congo communications, vid the Obo-Chodo divide. This I consider to be the key to the development and junction of the Central and Northern Railway systems of Africa, east of the Great Divide."

It will be evident from this quotation that Captain Philipps' notes on this question relate to an exceedingly interesting geographical discovery. I am inclined to doubt, however, whether it will prove to have any early value from the point of view of northward extension. A very full account of the route proposed from Tabora to Kasseke on the Kagera Bend was given by Mr. C. P. MacCarthy in his paper, "German Plans for Railway Development in East Africa" (G.y., Nov. I918). The distance by the projected line to Kasseke would be 270 miles; it would open up a very fertile and promising region, and traffic would be brought to it by four navigable rivers; construction would, on the whole, be easy, though there are some short stretches which would certainly be expensive, and apart altogether from any question of further extension to the north, it would seem that the prospects of the line, regarded solely as a branch from Tabora, would be good. On the other hand, if it is to be connected with the proposed Fort Portal system at Mubendi or Fort Portal, it will have to be carried northward for a further distance of probably not less than 230 miles, making the total distance from Tabora by this route 500 miles. My reason for thinking that this route is not promising from the point of view of speedy completion of a transcontinental through route from north to south is that, if train-ferries are employed, Victoria Nyanza offers far more tempting facilities for railway traffic. I have already mentioned that the port of Kigoma on Tanganyika is connected by railway with the sea at Dar-es-Saiaam. Tabora, to which I have just referred, is an important centre on this line, situated 230 miles to the east of Kigoma, and about I80 miles due south of the port on the south shore of Victoria Nyanza known as Mwanza. The intention of the Germans was to connect Tabora with Mwanza by a branch leaving the proposed Tabora-Kasseke line at Kahama, distant from Tabora about 83 miles. Of this distance about 4o miles had been constructed at the outbreak of the war, but the rails were subsequently taken up by the Belgians and used to complete the Bukama line. It is, I believe, certain that the country between Tabora and Mwanza is quite

* This valley was discovered and first reported by Major Grogan during the opera- tions against the Germans west of Victoria Nyanza. We hope to publish a description and map in a future number.-ED. G7.y.

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 87

Sketch-map of Railway Routes north and east of Tanganyika

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 88 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES easy for railway construction. It should, therefore, not be a specially costly matter to establish direct railway communication between the ports of Kigomo and Mwanza. There is, however, one difficulty of detail which I should mention. The new railway construction involved would not be limited to the x80 miles required to connect Tabora with Mwanza. Most unfortunately, the Kigoma-Dar-es-Salaam line is of metre gauge, whereas the rolling stock on the railways of southern and central Africa is constructed for 3 feet 6 inch gauge. It would, therefore, be necessary to make arrangements to allow this rolling stock to be used on the section between Kigoma and Tabora. For this purpose a third rail would be required. As the railways of the Sudan are also constructed with a 3 feet 6 inch gauge, it is fair to assume that the Cape to Cairo system of the future will be of this gauge throughout, as I think it most improbable that we can contemplate now its being changed to the English standard gauge of 4 feet 8? inches. On the assumption, there- fore, that 3 feet 6 inches will be the standard gauge of the Cape-to-Cairo system, it must, sooner or later, become necessary to change the gauge of the Kigoma-Dar-es-Salaam line and also that of the - which runs from to Victoria Nyanza, at Kisumu-from the metre gauge to the 3 feet 6 inch gauge. With a view to this contingency, I venture to suggest that the Colonial Office would be well advised in future to provide for the use of these metre-gauge railways only such new rolling stock as can be converted conveniently when necessity arises. Otherwise, the change of gauge, when the time comes, will involve the scrapping of unnecessarily large quantities of rolling stock. As soon as Victoria Nyanza has been reached at Mwanza by railway from Tabora, it will be possible to convey railway rolling stock between that port and any other port on the lake, and in particular to that port on the northern shore which may be found best adapted for use for further railway extension northwards. I do not, however, anticipate that the use of train-ferries on this great lake will be limited to north-to-south routes. I fully expect that in days to come there will be quite a number of other ferry services, notably between Kisumu (Port Florence) and the ports on the northern and western shores, for the purpose of developing the valuable agricultural regions in the vicinity of the lake, and particularly those lying to the west, north-west, and north, by giving them better communication with the shortest route to the east coast vim the Uganda railway (Fig. 2). It may be of interest at this stage to say something about the probable cost of construction of railways and train-ferries within this central zone. So far as railways are concerned, I can speak with considerable confidence, because the British South Africa Company has now had a very wide experience. The cost of constructing a railway of 3 feet 6 inch gauge with 6o-lb. rails, at a distance of, say, 50o miles from the coast, over a route which

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 89 would not require any specially heavy earth or bridge works, and reckon- ing railage on construction material at one penny per ton per mile, was, on the average, at pre-war prices, ?2730 per mile for bare track. With rails at .Io per ton, and allowing a somewhat higher sum for construction, the cost would be ?4770, and at the present price of rails (1i5 per ton) ?-5670, or say 6000ooo per mile. I am advised that at present prices of material and construction, the cost of widening the Kigoma-Tabora railway by the addition of a third

_?

Existing Railways New Construction Train Ferry Total Mileage Route ! 295 miles 610 miles 30mi/es 935 Route 2 237 ,, 180 ,, 200 ,, 6/7 Route 3 230 , 500 ,, 730 _-_--_--_-__--_-- ===_=.=== = =

ig. 2.-Routes from Tanganyika to the

rail so as to provide a 3 feet 6 inch track in addition to the existing metre track, would be about ?3600 per mile, or .828,00ooo for the section of 230 miles. With regard to train-ferries, the estimated cost of the very elaborate high-speed ferry which it was proposed to place on the Dover-Calais route in pre-war days was .13o,ooo. It may, I think, be reckoned safely that the cost of the far less elaborate and lower-speed boat which would be required in the first instance on the Great Lakes of Africa, would not exceed, even at the present prices, the figure of 300,000.o Similarly the H

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 90 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES quays and works at the terminal points on the great lakes should be capable of being constructed far less expensively than those which are needed for a marine service involving considerable variations of level due to tide. An expenditure of 75,00ooo at each terminal point should be an outside estimate. The initial cost, therefore, of establishing one train-ferry running between two terminal points on any one of the Great Lakes should not exceed, say, 500oo,ooo000. The construction of the ferries on the lakes'would, it is obvious, be greatly facilitated by the fact that all of them, so soon as the Beira-Malindi railway has been completed, will have direct railway communication with the sea. The lake routes indicated vidi Lakes Nyasa, Tanganyika, and Victoria Nyanza run practically north and south, and represent in the aggregate about 850 miles. At present prices it would cost about f,5,ooo,ooo to construct railways over this distance, as against a cost of I,5soo,ooo for train-ferries, or a saving of ?3,500,000. Moreover, if railway connection of the usual type were desired, the necessary lines would have to be built around the shores of the lakes over a much greater aggregate distance, and over country, some of which would undoubtedly be very costly for railway construction. The actual saving by the use of train-ferries would consequently be far greater than that indicated by these figures. This point is well illustrated by the estimated differences in cost involved in leaching Mubendi from Tabora by (a) direct rail route, or (b) by rail and train-ferry. The former would require the construction of 500 miles of track at a cost of say .?3,000,0o0. The latter would involve only 280 miles of new railroad construction at a cost of say .? 1,680,ooo, plus say ?5o?,??? for a train-ferry service, the capital saving by the use of the lake route being thus .820,000. The estimates given at the end of the preceding paragraph assumed the establishment of a train-ferry service between the ports of Mwanza and Entebbe, which is about I00 miles distant from Mubendi. The total cost of placing Entebbe in direct communication by rail and train-ferry with Malindi at the south end of Lake Nyasa over a total distance of I450 miles would be .4,548,ooo on the basis of the figures given above, which include the cost of providing an additional rail on the Kigoma- Tabora section. Any estimate of the cost of establishing direct com- munication by rail only between points so distant from one another, over country which is at many points so little known, must obviously be largely conjectural; it would, however, probably be unsafe to assume that less than 10o,ooo,ooo would be required for the purpose. It is obvious that the great saving in interest charges during the years which must elapse before heavy traffics can be built up, by utilizing train- ferries for the development of these central regions must also be a very important consideration. At 6 per cent. interest, the saving in interest charges alone through a saving of ?3.5??,??? on capital account, would be .?21o,oco per amnum, while the cost of maintaining train-ferry services

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 9I would also be considerably less than that which would have to be incurred for the maintenance in running order of long stretches of railway. If, on the other hand, a comparison be instituted with ordinary shipping services on the lakes, while the money difference in favour of the train-ferry services would not be nearly so great, the numerous disadvantages of a route on which frequent transhipment and handling must take place are too well known to require more than mere mention. In connection with the costs of working of train-ferry services on the Great Lakes, it is important to remember that supplies of coal in Southern Africa are both plentiful and comparatively cheap, the export price of Natal coal at Durban being to-day about 4os. per ton, and at Beira 50os. per ton. Good coal is known to exist in the Zambezi Valley, in the vicinity of the Shire Highlands line, and discoveries have also been reported in the neighbourhood of the north end of Lake Nyasa, and near Albertville on Lake Tanganyika.

The problems which I have to discuss next relate to the northern zone. I am not able this evening to express any decided view as to which one of the northern ports of Lake Victoria will be found to be most suitable for the purposes of further northward extension. The most obvious, judging solely by small-scale maps, is Jinja, which is already connected by a line of railway with Namasagali (61 miles), whence access to the navigable waters of Lake Kioga is secured. Whether it would be practicable to connect this port by train-ferry with some conveniently situated port on the north shore of the lake may be a matter for inquiry, but such information as I have been able to obtain points to the conclusion that it would not. Lake Kioga appears to be generally shallow, its waters are heavily choked with sudd, and the swamps which line its shores at many points are a considerable obstacle to the con- struction of ports and railways. If, therefore, the line from Jinja is to be carried further north, it would seem probable that it will have to pass round the north-west end of the lake. The prospects for such an undertaking, when studied with reference to such large-scale maps as are available, do not appear to me to be encouraging, owing to the enormous extent of the swamp areas in this region. I am inclined, therefore, to think that if the northward route must pass to the west of Lake Kioga, the most promising port on Victoria Nyanza from which to start will be found to be Port Bell near Entebbe. This port is already connected by a railway 7 miles in length with , whence, had it not been for the war, it was the intention to extend it either to Mubendi (whence connection with Fort Portal and the south-east corner of Lake Albert could be established without difficulty) or to Butiaba at the north-east end of Lake Albert, or possibly to both. At either point, connection with the existing system of land and water transport from Cairo would be established (Fig. 3).

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 92 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES

It is obvious that the advantages offered by this system for the transport of passengers and goods are unsatisfactory from the point of view of time, convenience, and expense. While the journey from Cape- town to Bukama in the southern and central zones (2600 miles) can be made by rail in eight days, that from Cairo to Lake Albert in the northern zone would occupy nearly twenty-one days under the most favourable conditions with existing facilities. There is accordingly ample

KHARTUM / el Fasher \ 400 - ...... unr'beid ..ar. \%~~~~~ -~~~~~ ~~~I

I

/ ''/rr 2 \\ P /r

/ ',

/ ,e_', K $ \I / l / ,, ~,' ', ~- 325 / ', A 8/m

$Stanleyville

?xist/ng flailway Aew Construction Toe/ Mileage Route 1 430 miles /950 miles 2380 Route 2 430 " /690 2/20 Roulte '3 /68 " 925 ^ ' /093

Flg. 3.-Routes from Lake Albert to Khartoum justification for further inquiry as to means by which they may be improved, and particularly in the regions lying to the south of Khartoum. I think it unlikely, in face of the facilities for navigation offered by the Nile, that continuous railway connection with Cairo will be completed for a considerable time, and I am inclined to think that for some time to come the northern terminus of any transcontinental railway service from north to south is more likely to be established 'at Khartoum or WVadi Halfa than at Cairo. If there is to be direct railway

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE.TO-CAIRO RAILWAY AND TRAIN FERRIES 93 communication between the Vic- toria Nyanza region and Khartoum, these connecting lines cannot, gene- rally speaking, be carried close to the river-bank, owing to the cha- racter of the country, and particu- Jarly because of the extent of swamp; they will have to traverse the slopes which bound the valley. On the west bank, the route by which a railway would run would probably be that which has been described by Major Christy (G.y., Sept. I9I7). At its northern ex- tremity it would probably connect with the railway which it is in- tended to carry from Khartoum to E1 Fasher, vid Sennar, Kosti and E1 Obeid, which latter point has already been reached. At its southern end it would either cross the Nile in the neighbourhood of Rejaf, or, keeping to the west of the Nile, connect with Stanleyville, possibly vim Mahagi on Lake Albert. But whatever course may ultimately be decided upon, it is obvious, from the point of view on north-to-south connection, that the Christy route must be an unfavour- able one, because it involves a considerable detour to the west, and must therefore be expensive both in money and time. The distance from Mahagi to Khartoum by this route would be about 2000 miles. On the east bank of the Nile, which seems to offer more pro- mising prospects, it would be neces- sary to connect either Mubendi or Kampala with Sennar, by a route which would probably run vi8 Dufile Fig. 4.-Lake Albert to Cairo by the Nile and Rejaf in the first instance, and would thus bridge the gap of the Fola rapids between the navigable

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 94 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES sections of the Nile. I am inclined to think that the commercial pros- pects of the country through which this line will pass, will justify the extension to Rejaf, as soon as, or even before, the transport facilities in the region of the Great Lakes to which this paper specially refers, have been provided. When the railway has reached Rejaf, the distance still to be traversed in order to reach Sennar will be 600 miles. If the line were carried to this point the only section between Capetown and Cairo which would still be without railway communication would be from to , a distance of about 225 miles (Fig. 4). At this point, in regard both to the limited question of the utility of an extension to Rejaf, and, more generally, to the value of the central north-and-south trunk system as a whole, I regret to find myself in complete disagreement with the views which have been expressed in a recent paper by so high an authority as Sir Alfred Sharpe (G.y., Sept. I918), who said:

" It has been suggested that Nimule and Rejaf might be connected by railway, so as to continue the line of communication north and south. It appears to me, however, that the only reason for doing this would be one of sentiment. W7hen the Mombasa railway has been continued through Uganda, these northern Uganda districts will have a comparatively cheap route by which to send their products to the . Whether the construction of a railway from Nimule to Rejaf would cheapen these rates and enable northern Uganda, instead of sending trade to Mombasa, to forward it at a slightly lower rate by the Nile to the north, is a question; it might turn out that it would prove cheaper for the Upper Sudan to send its goods through Uganda to Mombasa. But in either case the difference could not be great, and the money necessary for making a railway from Nimule to Rejaf would be better spent in the construction of some really vitally needed line.

" I know that there have been proposals for the construction of a section of a Cape-to-Cairo railway along the backbone of Africa, more or less following the line of the rift lakes, and finally either descending to the Nile at Rejaf and following the eastern edge of the swamps, or else running along the Nile-Congo watershed to El-Fasher. Every one interested in Africa would be glad to see such a railway in existence; but as a commercial undertaking at the present date, its construction would not be justified when there are so many districts crying for an outlet to the nearest seaport.

," It is only when Lake Albert is passed and the Upper Nile reached that a railway running north will be the best line of communication, and even in this case most goods for the Upper Nile will not go via , Cairo or , but by the existing railway from Port Sudan on the western shores of the Red Sea.

. ., . . . . . **

" The line from Port Sudan (and Egypt) running to Khartoum and beyond has already reached its limit of usefulness as a north-and-south railway at Kosti, whence it turns to the west to El Obeid. As I have already mentioned,

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 95

this line will probably soon be carried on to El Fasher in Darfur; and here we have the possibility of a transcontinental railway east and west.

,, ...... , ., "We have heard so much of the Cape-to-Cairo railway, a phrase which, after all when the question is carefully studied, does not mean much. There is little doubt that some day in the future it will be possible to travel by rail from the Cape to Cairo, but for some years to come the needs of Africa in the matter of railways are not so much a north-and-south through line as railways going from the interior to the east and west coasts."

With regard to north-to-south versus east-to-west routes in Africa, I think that the requirements of the country tend to vary from time to time, and it must be frankly admitted that hitherto there has been no effective demand for through communication by land from north to south of the continent. There has, moreover, been no force of sentiment, outside Rhodesia, to incline the balance whenever commercially possible in favour of north-to-south communication. Even in Rhodesia a halt in northern extension towards Tanganyika has been called for more than ten years at Rhodesia Broken Hill, for want of sufficient local encourage- ment to extend further, in just the same fashion as Sir A. Sharpe rightly points out, the southern impetus has faded away at Kosti. Looking into the future, I am ready to admit that the Cape-to-Cairo system, when constructed, will never be a through route for goods on any considerable scale. It is obvious that sea freight must always be cheaper than railway freight over such a long distance as 4000 miles, and also over much shorter distances. This fact, however, does not necessarily mean that the commercial usefulness of the various sections of the central system may not be very considerable, nor does it furnish any conclusive reason why the system as a whole should not pay. On the other hand, I question very much whether, in what I have called the central zone, there would be any distinct commercial advantage in providing additional means of communication with the east coast. In this region, all the ports are disadvantageously situated for trade with Europe, because the traffic has to pass vi~ either the Suez Canal or the Cape of Good Hope. In its southern half a longer haul by land and lake to Beira may quite conceivably be more economical than a shorter journey by land to say Dar-es-Salaam or Kilindini. Similarly in the northern section it is quite possible that, notwithstand- ing the longer haul, a considerable portion of the Uganda traffic will be able to be dealt with more economically vim Port Sudan than by Kilindini. Again, on the west coast, I doubt whether any considerable traffic from the regions which I have been discussing will ever flow to any west-coast ports to the north of Boma and Lobito Bay, because the geographical position of any port situated further to the north will always handicap it very heavily in comparison with the nearer eastern ports. It may be laid down as an invariable general rule in these matters, that, while passenger

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 96 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES traffic will, other things being equal, tend to follow the quickest route, goods traffic will flow by the cheapest. So far, then, as I can see at present, the only connections with the sea for which there is likely to be any real demand in the immediate future are concerned with the com- pletion of the Lobito Bay line, and of the programme for linking up the centres of the Belgian Congo with Boma. In these circumstances, I demur altogether to the view that such money as may be available for railway construction in Africa should be devoted to railways running from the interior to the coasts, and that the building of further north-and-south connections is a matter which can stand over. On the contrary, so far as the east coast is concerned, I think it would be a mistaken policy to expend money in multiplying competitive connections between eastern ports and the central trunk system when, as will soon be the case, each of the three great lakes has been provided with direct and efficient rail communication with the sea. In my view, the most crying need of Africa, and a need which must soon be recognized, is for through north-to-south communication, not so much for goods as for passenger traffic. Some new impulse, however, will be needed to accelerate the construction of the last links in the chain of communications which I have described to you, and I believe that this impulse will come from a pressing demand, which will soon be apparent, for better facilities for passenger traffic in the northern zone, and from a wide- spread feeling that apathy in these regions should not be permitted to retard the progress of central and southern Africa when they have done their share in the common work of construction. The capital expenditure required for the last sections will be relatively so moderate that in my judgment it would be well worth the while of all the Governments in Africa to co-operate, if need be, in providing for it. When only the last links remain to be constructed, the benefits both direct and indirect which all these Governments will reap from their completion will be so obvious, and the demand for them will become so insistent, that I have no doubt as to the result. Considerable use will certainly be made by passengers of a through route at all seasons of the year, but there is one aspect of the possibilities of this route which I think deserves special attention. A rapidly growing demand has sprung up in recent years for winter resorts, where people of sufficient means and leisure can be certain of finding warmth, health, recreation, and interest without having to face a long and inclement journey by sea. If improved facilities for travelling can be provided, all these requirements can be met to the full in Africa. At one point or another practically every variety of climate can be found there, with the exception only of that which is met with in the frozen seas of the Arctic and Antarctic Circles. Hot springs, with valuable medicinal qualities, occur in many places between Capetown and Cairo in delightful surround- ings. The facilities for sport of all descriptions are in many respects

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 97

unique. Natural spectacles of wonderful beauty and impressiveness are to be found in constant succession. Geographers, archaeologists, ethno- logists, botanists, and scientific men generally can find the widest fields for study, while persons of more commercial tastes cannot fail to be both interested and impressed by the mining and other resources of the Dark Continent, and by the methods which are being employed to develop them.

TIME TABLE OF TR&NSCONTINENTAL JOURNEY FROM CAPETOWN TO CAIRO BY THE PRINCIPAL ALTERNATIVE ROUTES BETWEEN BULAWAYO AND KHARTOUM.

I. Capetown to Bulawayo: Rail, I362 miles; 2 d. 21 h. 2. Bulawayo to Khartoum : Route A, vidc Bukama, Kabalo, Stanleyville, Irumu, E1 Fasher, Sennar: Railway constructed ...... I963 miles; 6 d. 2I h. ,, projected ...... 2450 ,, 8 d. I2 h. Route B, vid Salisbury, Beira, Malindi, Kigoma, Tabora, Kampala, Sennar: Railway constructed ...... II98 miles; 3 d. I7~ h. ,, projected...... I635 ,, 5 d. I6 h. Train ferry ,,...... 840 ,, 2 d. I6~ h. Route C, vid Rhodesia, Broken Hill, Abercorn, Kigoma, Tabora, Kampala, Sennar: Railway constructed ...... Io060 miles; 3 d. Io0 h. ,, projected ...... 1705 ,, 5d.22h. Train ferry ,, ... .. 500o ,, I d. I4 h. 3. Khartoum to Cairo: Railway constructed ...... Io034 miles; I d. 23 h. Steamer ...... 208 ,, I d. I8 h. Totals: Route A ...... 70I7 ,, 2I d. 23J h. Bg ...... 6277 ,, I8 d. 16 h. C ... ,... .. 5869 ,, 17 d. 13 h.

Tlhe average speed is taken as I2 m.p.h. on new railways, and I3 m.p.h. on train ferries.

I can conceive of no more delightful excursion than a tour through Africa extending over four or five months, which would range fromr north to south of the continent, and possibly include a return journey by an alternative route. During such a tour time would be of no special object; it would be permissible to break one's journey and remain as long as one desired at any place where fancy or interest inclined, experiencing during the journeys, which would be made in comfort, the charm and beauty of African surroundings, the exhilaration of the high-veld climate, and the glories of the African nights. These varied attractions will, in my view, ensure a constantly increasing use of this most beneficent avenue for human intercourse, and help to bring the inhabitants of Great Britain into closer and more sympathetic touch with the denizens of every race who are playing their several parts in the building up of the territories of the Crown in Africa. Its success will, I believe, be assured, and I hope that I may live to see the day when the Cape-to-Cairo Railway can no longer

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms 98 THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES be described as a dream, a phrase, or an ideal, and it has become an accomplished fact.

APPENDICES

NOTES BY LIEUT.-COLONEL MURRAY, C.B., C.M.G., D.SO., UPON POSSIBLE ROUTES FOR RAILWAY CONSTRUCTION IN THE REGIONS IN THE VICINITY OF THE NORTH END OF LAKE NYASA AND THE SOUTHI END OF LAKE TANGANYIKA. z. Rhodesia Broken Hill to Abercorn, Northern Rhodesia. A motor road runs fiom Kashitu north of Broken Hill via Serenje Mpika to Kasama and thence to Abercorn, a distance of approximately 500 miles. The country south of Kasama is perfectly flat, and the Chambezi river is the only large river to be bridged. To give some idea of how simple railway con- struction is in this country, I may state that the motor road was started in the latter end of January I916, and by May I916 motor transport was running between Kasama and Kashitu--a distance approximately 5o00 miles. This work was carried out during the wet season. From Kasama to Abercorn (distance ICo miles) the railway construction would run along the watershed; there are no rivers to be bridged, and a good well-graded road already exists. The country along the whole route is well wooded. Labour is also plentiful. The country is most uninteresting, being thickly covered with forest; rather like what you get in parts of Matabeleland. It improves rather when you get to Kasama; you then get a few hills from there on to Abercorn, but still the same bush. You cross on the road to Abercorn the source of the Chambezi River. This, I think, is always accepted as the headwaters of the Congo. Abercorn itself is most beautifully situated on the high ground above Lake Chila. The lake is about three-quarters of a mile away from the township. It has a most perfect climate-rather like England, but plenty of sunshine. Coffee and all sorts of things do well there. The soil is deep and red in parts. It is a healthy part of the country for Europeans. From Abercorn the railway would connect up with Kitutu Bay on Tanganyika, a distance approximately of I5 miles, thence another I5 miles by water to Bismarckburg.

2. North End of Lake Nyasa. The Bay with Old Langenburg on the eastern shore, and Mwaya on the western shore, is pretty correct as shown on the map; there is, however, no outlet inland from Old Langenburg except over the Livingstone Range by a goat track. At Mwaya you have plenty of space for buildings, etc.

3. Lake Nyasa to Lake Tanganyika. The country between Tanganyika and Nyasa is well watered and wooded, undulating, with small mountain ranges and isolated hills along the whole route, fertile valleys, and in other parts open park-like country. The Chomba Mountains (i.e. escarpment) overlooking Karonga have a drop of 3500 feet in 40 miles; the present motor roa i is a series of hairpin bends down the range. Certainly Karonga would never make a suitable harbour. There are at present two parallel routes where roads already exist in the vicinity of Ikomba, only a few miles apart. Route I starts from Bismarckburg and goes over the Igale Pass to the foot of the Rungwe Mountain v7ia Neu Lankenburg to Mlwaya in the north-west corner of Lake Nyasa. Route 2 leaves Kitutu Bay vid Abercorn, thence along the Tanganyika plateau vid Fife

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES 99 to the top of the escarpment west of Karonga, Lake Nyasa. Along this route the trans-continental telegraph line runs. Route X is the most suitable for railway construction. It runs along the north face of the watershed, a well- graded road already exists, and the only disadvantage is that there are a few rivers to be bridged. Bismarckburg would make an excellent port with deep water right up to the shore. The crossing of the mountains vid the Igale Pass to Neu Langenburg thence to Mwaya is no difficult feat. At Mwaya it would be easy to get a suitable harbour; the shipping also would be sheltered from the storms. Route 2, Kitutu Bay ,id Abercorn to Karonga, has only the advantage that there are no rivers to be bridged. From Kitutu Bay to Abercorn the grading is fairly stiff in parts: a motor road already exists. From Abercorn along the Stephenson Road there is no difficulty until the top of the escarpment over- looking Nyasa is reached, but it would be a difficult feat to get a line down the escarpment. The harbour at Karonga is also bad. The shipping lies some distance from the shore, and runs south to Vua immediately bad weather threatens. This disadvantage does not exist at Mwaya. The approximate mileage of both routes is about 200 miles. The advantages certainly rest with the Bismarckburg-Mwaya route. A well-graded motor road runs from Neu Langenburg to Mwaya, and you also have a wide choice of harbours. The Livingstone range of mountains runs right down to the lake, quite close to this point, which gives deep water and protection from storms. The establishment of railway communication between Tanganyika and Nyasa would open out a fertile part of Central Africa and add to the world's food supplies in both beef and cereals. It would give a new country, that has hitherto been untapped, and also a part of country in which Europeans could live in comfort. The climate of the Tanganyika plateau from Neu Langenburg east to Njombe Lupembe is very mild. The rainfall from the records is high, but this does not appear to make the country particularly unhealthy. The Ufipa country between Tanganyika and Rukwa is governed by the British Administration from Bismarckburg under the nominal sovereignty of a sultana. The inhabitants own large herds of cattle and are a purely pastoral people. A high state of agricultural efficiency has been brought about under the French Fathers at the Mwaza Mission; large crops of maize, beans, potatoes, etc., have been produced in this area, To the east of Lake Rukwa you again get into a great cattle country peopled by the Ussangu tribe under a sultan named Merere, who, previous to British occupation, was a source of anxiety to the Germans owing to his pro- British leanings. This tribe own great number of cattle of the "hump" variety. Crossing the Igale pass you get into the Neu Langenburg area. The soil hereabouts is volcanic and very fertile. Water is very plentiful, the whole land being intersected by small streams which run all the year through. This part has been described as the Germany of German East Africa. It is inhabited by the Wankondi tribe, a purely agricultural people who produce wheat, potatoes, etc. The climate is very similar to that of England, and English fruits such as currants, etc., do well here. Further to the south-east towards Mwaya the country is more tropical, the motor road passing through many miles of banana groves to the shores of Lake Nyasa at Mwaya. 4. South End of Lake Tanganyika. In most parts the mountains run right down to the lake-shore; there is, however, the open harbour of Kitutu Bay at the most southern point, about I5

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms oo THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES miles from Abercorn, which can be reached by motor. This harbour is in Northern Rhodesia. About 15 miles from Kitutu Bay is the old German post, Bismarckburg, in German East Africa, native name, " Kasanga "; this is a first-rate anchorage, and deep water right up to the shore.

NOTES BY CAPTAIN J. E. PHILLIPS, M.C. (written during the war), UPON THE REGIONS OF CENTRAL AFRICA IN THE VICINITY OF THE NORTH END OF LAKE TANGANYIKA, CONSIDERED FROM THE STAND- POINT OF THE ROUTES BY WHICH COMMUNICATION MAY BE ESTABLISHED BETWEEN THE CAPE AND CAIRO BY MECHAN[C4L TRANSPORT.

General. West of Tanganyika. The South African Railway system may be considered already linked with the Belgian line of communication from Tanganyika to the Atlantic. Thence the shortest projected line to establish contact is from the Congo bend at Stanleyville to Mahagi on Lake Albert. Although from Mahagi advantage can be taken of Uganda steamers to Dufile, there is still Io4 miles unspanned thence to Rejaf. However, from Rejaf south-westward there is under construction a motor road of I30 miles to Aba in Belgian territory. This should be com- pleted within the year. It is anticipated that traffic from the new goldfields (at Kilo) may make use of this route. The Belgian Government have recently been anxious to push forward the Stanleyville-Mahagi project, but the difficulties of output and transport of material must postpone it until after the conclusion of hostilities, and, even when completed, the Dufile-Rejaf link is still missing between the two systems. An extension of the Rejaf-Aba road further into the heart of the gold-producing area, or its conversion into a light railway for establishing an outlet via Port Sudan, is economically likely to be the earliest development, failing effective mechanical transport from the Congo.

Imperial. East of Tanganyika. Northward development on this side of Africa, with a view to junction through Uganda, as originally planned by Rhodes, who had, however, no means of access to knowledge of the country concerned, requires British con- trol over: (I) the eastern shores of Lake Tanganyika: from Abercorn to Ujiji; (2) the section Kigoma-Tabora of the German Central Railway; (3) the pro- vince of Ruanda. These three essentials have now been under Belgian pro- visional administration for exactly a year. Any connection from the Cape system to Tanganyika through British territory would probably indicate the revival of Rhodes' scheme of rail from Broken Hill to Abercorn, which was deflected after his death, when Katanga copper began to attract Central African rails like a magnet. From Tanganyika northwards no use can be made of the valleys or lakes in the Central rift Kivu-Albert. The Rusisi River rushes down a deep gorge closed in by an enormous complication of volcanic mountains stretching both east and west to Lake Victoria and the Congo River respec- tively. Northward, Kivu is cut off from lakes Edward and Albert by a range of active volcanoes, running up to I4,oo000 feet and thrown laterally across the divide. From Tanganyika, use will have to be made of the central line from Kigoma to Tabora. From Tabora the Germans had projected and surveyed a railway via St. Michael (Msalala) and Mariahilf (Ushirombo) to the junction of the rivers Nyawarongo and Kagera at Kaseke. From Kaseke water trans- port on the Nyawarongo to Mogo ferry, five hours from , the administra- tive capital of Ruanda, was designed to carry off the enormous traffic in hides.

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms THE CAPE-TO-CAIRO RAILWAY AND TRAIN FERRIES IoI

This railway from Tabora is frequently referred to erroneously as the Mwanza Railway. It is known to the Germans as Kagera Railway, and was never intended to touch on Lake Victoria, which was to have been tapped by an extension of the Usambara line from Moshi to Speke Gulf, through Arusha and the goldfields. Of the Kagera Railway, 40 kilometres had been con- structed from Tabora northward during 19t4-I5. The rails were removed by the Belgians at the beginning of this year for the completion of the Cape- Katanga line to the Lualaba. In German maps the Kagera is shown as flowing from west to north to Massekes passing Kigari, its upper waters only being termed Nyawa-Rongo, whereas it is our custom to show the whole length of this river as Nyawarongo, flowing into the main Kagera (or Alexandra Nile) at Kaseke; while, higher up the main stream, the German Ruvuvu is termed by us Kagera, rising within 30 miles of Tanganyika, with the Ruvuvu as a small tributary to its upper waters. From Kaseke, downstream to its mouth in Victoria Nyanza, the Kagera is infested with tsetse fly, mainly Glossina palpalis. There are roughly 3,700,oo000 head of native cattle in Ruanda. The hide traffic now passes from Kikali to Bukoba to Port Florence to Mombasa. The Kagera Railway was designed to divert this vid Dar-es-Salaam. For many years it has been definitely laid down by authorities that in the rectangle formed by: (North) the southern frontier of Uganda; (south) the German Central Railway; (east) the eastern shore of Lake Victoria; (west) from Lake Kivu to Ujiji on Tanganyika, it will not be possible to construct a line to connect the Central rail with the railway system of Uganda. This is not correct. It is true that the country is very mountainous, showing extreme volcanic complications over the whole area indicated, with no clear systems of ranges or valleys. No European has hitherto had an opportunity of spending any time in this area with the exception of the few German officials. In Ruandi and Urundi there are no white traders or settlers. Before the war, the country was a closed area from the time the Germans opened political stations in 191o-12. I have been for five months engaged there on a political mission up to March I9I7. It will be noticed on every existing map that the country north' ward from the terminus (Kaseke) of the projected railway remains, by accident or design, quite uncharted. British maps are recent compilations from German sources; also from the northern edge of this area up to the Uganda frontier the map is in many cases curiously incorrect. I was able, when my duties and the unsettled state of the country permitted, to . . . confirm the existence in the uncharted portion lying parallel to the course of the Kagera, of a remarkable valley which provides an unexpected avenue ,f communication between Kaseke and the Kakitumba valley. Thence access is assured to the Rushenyi plains of Uganda and the Fort Portal system of rail. Thence, also, north-westward through the Ntungwe valley, if required, to Lake Edward (and also to Stanleyville on the main Congo communications, via the Obo-Chodo divide). This I consider to be the key to the development and junction of the central and northern railway systems of Africa, east of the great divide.

Befoxe the paper the PRESIDENT said: It is twenty-three years ago since I had a most pleasant breakfast at Fort Salisbury in Rhodesia with Mr. Wilson Fox, when I was up there as Times correspondent, stopping with Mr. Cecil Rhodes. We were then discussing this great scheme, first of the trans- continental telegraph, and then of the Cape-to-Cairo railway. Mr. Wilson Fox

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms r -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~3 20 10 O

I / /t~~~~~~l _L t>/. - :/ . .,s t , . U======/ /o a o /-"^ / /"Tet ^^^ k; ^ 0?wcS%C^7^~^ ^. J^ ^^ , ^^v>lg01' Conte99fw 1'~L?u. r- 30 -- / li ^^ v -^T _ t, ,G+ B~~~~~P ?^K, O | E ConBsaX n I: A/~ ae*^-, O \ , / M G51llad j / ? 9d.-^ _ f ^ ^ 5 / X wa>o D t t

,F~~~~~~~~~~~1

D XI~~~~~~~~2at 4w b 'E-<~ol~h' | e

6t; --. _ /S ,r// ^, ^<'ta&.i \vv / gfmB&h oWr'l,

/ ? A . r4 |r e I c ;ue 2 rg D ^ - " ~- t ? ^ S & ^^ ? -j S V S T R t T 0 f 1 ' I ~e?~ . . . .:- ,::

/ ~~~~/ ""---Jy/-^.--c^sns .^ML^L-^. 0^^^1^^'^ ''t* ^ ^x9~bl' j ^QvUa^ ^(*y^3a.?3^^' ^ s".^ .,-Bfr?tr ^ I ' 0 ^^^S^ii^^ -^^K /-wy^^ -^e^~~~~~~~~~~~10 Bt)a ^tnu, // S 45 O /'\ o "N,fiywt l,,//'^ ,' ~ ~ ~ ~ ~ ~ ~ ~ ~ i on W '3eA o<^y K^m / CJ^l7sdl T ^' L ' A N T"^* ?-5I x v C 1'^

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms Ir

0 de

O eq

a

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms "", I C:_ff._r'A I"f I I)rNlA I Rc. f l."n U

----- ~ ~~~~~~~ ~ ~ ~ ~~~~~~~~~~~~ ?- --- _ u a, w r -- v v v r e s -

50 6O 60~~~~-

3, 5 I A

"9

O

X N D A NT This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms I \I \

I A'etta Cama'? - L A N T I C A T May,^.Jba l~O a.D

Loan

Lan

b

e~ ,.scrn,. (Br)

] LO d 0 C E A

Mosiaas

r eSfgYr- r;< 2^sh tHr\

.M. OSS~--- '-- L--~~~~~~~~~~~~~~~~~~~- I~~~~~

Map to illustrate the paper on 20 I THE (APE TO CAIRO RAILWAY & lr T'RAIN FERRIES ty H.WILSON FOX, M.P.

Scale 1:20,000,000 or lInah = 315-6 Stat. Miles o100 0 100 200 300 400 500 600

eferenc e _ij'laCys wns tnict ec...... ? prpo-e projected...... i Tru/.~/ *erm-es ...g...... S Heights ir ?eez;

20 10 0 I I ZerithaLt ~Eau.- a ..e .P:oje j

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms O

.~~ ~ ~~~~~~~~~~~v :~?:~ i:3!.>i:"

!:::i:~, :::::::::::::::::::::::::::: , :i:::::::::

i'

I~~~~~~~~~~~~~~~~

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms I N D A ~~~~ .?

Aaxx&Qte I'. * CBr.) fB^) :,*

0 C Ec xN

50 60 i1 i i _ _ 6

CAPE TO CAIRO RAILWAY ~qlson FQx.

This content downloaded from 128.197.26.12 on Mon, 27 Jun 2016 09:58:11 UTC All use subject to http://about.jstor.org/terms