it often happens that in prosperous public enterprises the applause of the nation and the rewards of the sovereign are bestowed on those whose offices are splendid and whose duties have been dramatic. others whose labours were no less difficult, responsible, and vital to success are unnoticed. if this be true of men, it is also true of things. in a tale of war the reader's mind is filled with the fighting. the battle—with its vivid scenes, its moving incidents, its plain and tremendous results—excites imagination and commands attention. the eye is fixed on the fighting brigades as they move amid the smoke; on the swarming figures of the enemy; on the general, serene and determined, mounted in the middle of his staff. the long trailing line of communications is unnoticed. the fierce glory that plays on red, triumphant bayonets dazzles the observer; nor does he care to look behind to where, along a thousand miles of rail, road, and river, the convoys are crawling to the front in uninterrupted succession. victory is the beautiful, bright-coloured flower. transport is the stem without which it could never have blossomed. yet even the military student, in his zeal to master the fascinating combinations of the actual conflict, often forgets the far more intricate complications of supply.

it cannot be denied that a battle, the climax to which all military operations tend, is an event which is not controlled by strategy or organisation. the scheme may be well planned, the troops well fed, the ammunition plentiful, and the enemy entangled, famished, or numerically inferior. the glorious uncertainties of the field can yet reverse everything. the human element—in defiance of experience and probability—may produce a wholly irrational result, and a starving, out-manoeuvred army win food, safety, and honour by their bravery. but such considerations apply with greater force to wars where both sides are equal in equipment and discipline. in savage warfare in a flat country the power of modern machinery is such that flesh and blood can scarcely prevail, and the chances of battle are reduced to a minimum. fighting the dervishes was primarily a matter of transport. the khalifa was conquered on the railway.

hitherto, as the operations have progressed, it has been convenient to speak of the railway in a general manner as having been laid or extended to various points, and merely to indicate the direction of the lines of communication. the reader is now invited to take a closer view. this chapter is concerned with boats, railways, and pack animals, but particularly with railways.

throughout the dongola campaign in 1896 the nile was the main channel of communication between the expeditionary force and its base in egypt. all supplies were brought to the front as far as possible by water transport. wherever the nile was navigable, it was used. other means of conveyance—by railways and pack animals—though essential, were merely supplementary. boats carry more and cost less than any other form of transport. the service is not so liable to interruption; the plant needs only simple repair; the waterway is ready-made. but the nile is not always available. frequent cataracts obstruct its course for many miles. other long reaches are only navigable when the river is in flood. to join the navigable reaches, and thus preserve the continuity of the communications, a complex system of railways and caravans was necessary.

in the expedition to dongola a line of railway was required to connect the two navigable reaches of the nile which extend from assuan to wady halfa, and from kerma to merawi. before the capture of dongola, however, this distance was shortened by the fact that the river at high nile is navigable between the third cataract and kerma. in consequence it was at first only necessary to construct the stretch of 108 miles between wady halfa and kosheh. during the years when wady halfa was the southernmost garrison of the egyptian forces a strong post had been maintained at sarras. in the nile expeditions of 1885 the railway from halfa had been completed through sarras and as far as akasha, a distance of eighty-six miles. after the abandonment of the soudan the dervishes destroyed the line as far north as sarras. the old embankments were still standing, but the sleepers had been burnt and the rails torn up, and in many cases bent or twisted. the position in 1896 may, in fact, be summed up as follows: the section of thirty-three miles from wady halfa to sarras was immediately available and in working order. the section of fifty-three miles from sarras to akasha required partial reconstruction. the section of thirty-two miles from akasha to kosheh must, with the exception of ten miles of embankment completed in 1885, at once be newly made. and, finally, the section from kosheh to kerma must be completed before the nile flood subsided.

the first duty, therefore, which the engineer officers had to perform was the reconstruction of the line from sarras to akasha. no trained staff or skilled workmen were available. the lack of men with technical knowledge was doubtfully supplied by the enlistment of a 'railway battalion' 800 strong. these men were drawn from many tribes and classes. their only qualification was capacity and willingness for work. they presented a motley appearance. dervish prisoners, released but still wearing their jibbas, assisted stalwart egyptians in unloading rails and sleepers. dinkas, shillooks, jaalin, and barabras shovelled contentedly together at the embankments. one hundred civilian soudanese—chiefly time-expired soldiers—were also employed; and these, since they were trustworthy and took an especial pride in their work, soon learned the arts of spiking rails and sleepers, fishing rails together, and straightening. to direct and control the labours of these men of varied race and language, but of equal inexperience, some civilian foremen platelayers were obtained at high rates of pay from lower egypt. these, however, with very few exceptions were not satisfactory, and they were gradually replaced by intelligent men of the 'railway battalion,' who had learned their trade as the line progressed. the projection, direction, and execution of the whole work were entrusted to a few subalterns of engineers, of whom the best-known was edouard girouard.

work was begun south of sarras at the latter end of march. at first the efforts of so many unskilled workmen, instructed by few experienced officers, were productive of results ridiculous rather than important. gradually, however, the knowledge and energy of the young director and the intelligence and devotion of his still more youthful subordinates began to take effect. the pace of construction increased, and the labour was lightened by the contrivances of experience and skill.

as the line grew longer, native officers and non-commissioned officers from the active and reserve lists of the egyptian army were appointed station-masters. intelligent non-commissioned officers and men were converted into shunters, guards, and pointsmen. traffic was controlled by telephone. to work the telephone, men were discovered who could read and write—very often who could read and write only their own names, and even that with such difficulty that they usually preferred a seal. they developed into clerks by a simple process of selection. to improve their education, and to train a staff in the office work of a railway, two schools were instituted at halfa. in these establishments, which were formed by the shade of two palm-trees, twenty pupils received the beginnings of knowledge. the simplicity of the instruction was aided by the zeal of the students, and learning grew beneath the palm-trees more quickly perhaps than in the magnificent schools of civilisation.

the rolling stock of the halfa-sarras line was in good order and sufficient quantity, but the eight locomotives were out of all repair, and had to be patched up again and again with painful repetition. the regularity of their break-downs prevented the regularity of the road, and the soudan military railway gained a doubtful reputation during the dongola expedition and in its early days. nor were there wanting those who employed their wits in scoffing at the undertaking and in pouring thoughtless indignation on the engineers. nevertheless the work went on continually.

the initial difficulties of the task were aggravated by an unexpected calamity. on the 26th of august the violent cyclonic rain-storm of which some account has been given in the last chapter broke over the dongola province.

a writer on the earlier phases of the war [a. hilliard atteridge, towards freedom.] has forcibly explained why the consequences were so serious:

'in a country where rain is an ordinary event the engineer lays his railway line, not in the bottom of a valley, but at a higher level on one slope or the other. where he passes across branching side valleys, he takes care to leave in all his embankments large culverts to carry off flood-water. but here, in what was thought to be the rainless soudan, the line south of sarras followed for mile after mile the bottom of the long valley of khor ahrusa, and no provision had been made, or had been thought necessary, for culverts in the embankments where minor hollows were crossed. thus, when the flood came, it was not merely that the railway was cut through here and there by the rushing deluge. it was covered deep in water, the ballast was swept away, and some of the banks so destroyed that in places rails and sleepers were left hanging in the air across a wide gap.'

nearly fourteen miles of track were destroyed. the camp of the construction gangs was wrecked and flooded. some of the rifles of the escort—for the conditions of war were never absent—were afterwards recovered from a depth of three feet of sand. in one place, where the embankment had partly withstood the deluge, a great lake several miles square appeared. by extraordinary exertions the damage was repaired in a week.

as soon as the line as far as kosheh was completed, the advance towards dongola began. after the army had been victorious at hafir the whole province was cleared of dervishes, and the egyptian forces pushed on to merawi. here they were dependent on river transport. but the nile was falling rapidly, and the army were soon in danger of being stranded by the interruption of river traffic between the third cataract and kenna. the extension of the line from kosheh to kerma was therefore of vital importance. the survey was at once undertaken, and a suitable route was chosen through the newly acquired and unmapped territory. of the ninety-five miles of extended track, fifty-six were through the desert, and the constructors here gained the experience which was afterwards of value on the great desert railway from wady halfa to the atbara. battalions of troops were distributed along the line and ordered to begin to make the embankments. track-laying commenced south of kosheh on the 9th of october, and the whole work was carried forward with feverish energy. as it progressed, and before it was completed, the reach of the river from the third cataract to kenna ceased to be navigable. the army were now dependent for their existence on the partly finished railway, from the head of which supplies were conveyed by an elaborate system of camel transport. every week the line grew, railhead moved forward, and the strain upon the pack animals diminished. but the problem of feeding the field army without interfering with the railway construction was one of extraordinary intricacy and difficulty. the carrying capacity of the line was strictly limited. the worn-out engines frequently broke down. on many occasions only three were in working order, and the other five undergoing 'heavy repairs' which might secure them another short span of usefulness. three times the construction had to be suspended to allow the army to be revictualled. every difficulty was, however, overcome. by the beginning of may the line to kenna was finished, and the whole of the railway battalion, its subalterns and its director, turned their attention to a greater enterprise.

in the first week in december the sirdar returned from england with instructions or permission to continue the advance towards khartoum, and the momentous question of the route to be followed arose. it may at first seem that the plain course was to continue to work along the nile, connecting its navigable reaches by sections of railway. but from merawi to abu hamed the river is broken by continual cataracts, and the broken ground of both banks made a railway nearly an impossibility. the movements of the french expeditions towards the upper nile counselled speed. the poverty of egypt compelled economy. the nile route, though sure, would be slow and very expensive. a short cut must be found. three daring and ambitious schemes presented themselves: (1) the line followed by the desert column in 1884 from korti to metemma; (2) the celebrated, if not notorious, route from suakin to berber; (3) across the nubian desert from korosko or wady halfa to abu hamed.

the question involved the whole strategy of the war. no more important decision was ever taken by sir herbert kitchener, whether in office or in action. the request for a british division, the attack on mahmud's zeriba, the great left wheel towards omdurman during that battle, the treatment of the marchand expedition, were matters of lesser resolve than the selection of the line of advance. the known strength of the khalifa made it evident that a powerful force would be required for the destruction of his army and the capture of his capital. the use of railway transport to some point on the nile whence there was a clear waterway was therefore imperative. berber and metemma were known, and abu hamed was believed, to fulfil this condition. but both berber and metemma were important strategic points. it was improbable that the dervishes would abandon these keys to khartoum and the soudan without severe resistance. it seemed likely, indeed, that the khalifa would strongly reinforce both towns, and desperately contest their possession. the deserts between korti and metemma, and between suakin and berber, contained scattered wells, and small raiding parties might have cut the railway and perhaps have starved the army at its head. it was therefore too dangerous to project the railway towards either berber or metemma until they were actually in our hands. the argument is circular. the towns could not be taken without a strong force; so strong a force could not advance until the railway was made; and the railway could not be made till the towns were taken.

both the korti-metemma and the suakin-berber routes were therefore rejected. the resolution to exclude the latter was further strengthened by the fact that the labour of building a railway over the hills behind suakin would have been very great.

the route via abu hamed was selected by the exclusion of the alternatives. but it had distinct and apparent advantages. abu hamed was within striking distance of the army at merawi. it was not a point essential to the dervish defences, and not, therefore, likely to be so strongly garrisoned as berber or metemma. it might, therefore, be captured by a column marching along the river, and sufficiently small to be equipped with only camel transport. the deserts through which the railway to abu hamed would pass contain few wells, and therefore it would be difficult for small raiding parties to cut the line or attack the construction gangs; and before the line got within reach of the dervish garrison at abu hamed, that garrison would be dislodged and the place seized.

the plan was perfect, and the argument in its favour conclusive. it turned, however, on one point: was the desert railway a possibility? with this question the general was now confronted. he appealed to expert opinion. eminent railway engineers in england were consulted. they replied with unanimity that, having due regard to the circumstances, and remembering the conditions of war under which the work must be executed, it was impossible to construct such a line. distinguished soldiers were approached on the subject. they replied that the scheme was not only impossible, but absurd. many other persons who were not consulted volunteered the opinion that the whole idea was that of a lunatic, and predicted ruin and disaster to the expedition. having received this advice, and reflected on it duly, the sirdar ordered the railway to be constructed without more delay.

a further question immediately arose: should the railway to abu hamed start from korosko or from wady halfa? there were arguments on both sides. the adoption of the korosko line would reduce the river stage from assuan by forty-eight hours up stream. the old caravan route, by which general gordon had travelled to khartoum on his last journey, had been from korosko via murat wells to abu hamed. on the other hand, many workshops and appliances for construction were already existing at wady halfa. it was the northern terminus of the dongola railway. this was an enormous advantage. both routes were reconnoitred: that from wady halfa was selected. the decision having been taken, the enterprise was at once begun.

lieutenant girouard, to whom everything was entrusted, was told to make the necessary estimates. sitting in his hut at wady halfa, he drew up a comprehensive list. nothing was forgotten. every want was provided for; every difficulty was foreseen; every requisite was noted. the questions to be decided were numerous and involved. how much carrying capacity was required? how much rolling stock? how many engines? what spare parts? how much oil? how many lathes? how many cutters? how many punching and shearing machines? what arrangements of signals would be necessary? how many lamps? how many points? how many trolleys? what amount of coal should be ordered? how much water would be wanted? how should it be carried? to what extent would its carriage affect the hauling power and influence all previous calculations? how much railway plant was needed? how many miles of rail? how many thousand sleepers? where could they be procured at such short notice? how many fishplates were necessary? what tools would be required? what appliances? what machinery? how much skilled labour was wanted? how much of the class of labour available? how were the workmen to be fed and watered? how much food would they want? how many trains a day must be run to feed them and their escort? how many must be run to carry plant? how did these requirements affect the estimate for rolling stock? the answers to all these questions, and to many others with which i will not inflict the reader, were set forth by lieutenant girouard in a ponderous volume several inches thick; and such was the comprehensive accuracy of the estimate that the working parties were never delayed by the want even of a piece of brass wire.

in any circumstances the task would have been enormous. it was, however, complicated by five important considerations: it had to be executed with military precautions. there was apparently no water along the line. the feeding of 2,000 platelayers in a barren desert was a problem in itself. the work had to be completed before the winter. and, finally, the money voted was not to be outrun. the sirdar attended to the last condition.

girouard was sent to england to buy the plant and rolling stock. fifteen new engines and two hundred trucks were ordered. the necessary new workshops were commenced at halfa. experienced mechanics were procured to direct them. fifteen hundred additional men were enlisted in the railway battalion and trained. then the water question was dealt with. the reconnoitring surveys had reported that though the line was certainly 'good and easy' for 110 miles—and, according to arab accounts, for the remaining 120 miles—no drop of water was to be found, and only two likely spots for wells were noted. camel transport was, of course, out of the question. each engine must first of all haul enough water to carry it to railhead and back, besides a reserve against accidents. it was evident that the quantity of water required by any locomotive would continually increase as the work progressed and the distance grew greater, until finally the material trains would have one-third of their carrying power absorbed in transporting the water for their own consumption. the amount of water necessary is largely dependent on the grades of the line. the 'flat desert' proved to be a steady slope up to a height of 1,600 feet above halfa, and the calculations were further complicated. the difficulty had, however, to be faced, and a hundred 1,500-gallon tanks were procured. these were mounted on trucks and connected by hose; and the most striking characteristic of the trains of the soudan military railway was the long succession of enormous boxes on wheels, on which the motive power of the engine and the lives of the passengers depended.

the first spadeful of sand of the desert railway was turned on the first day of 1897; but until may, when the line to kerma was finished, no great efforts were made, and only forty miles of track had been laid. in the meanwhile the men of the new railway battalion were being trained; the plant was steadily accumulating; engines, rolling stock, and material of all sorts had arrived from england. from the growing workshops at wady halfa the continual clatter and clang of hammers and the black smoke of manufacture rose to the african sky. the malodorous incense of civilisation was offered to the startled gods of egypt. all this was preparation; nor was it until the 8th of may that track-laying into the desert was begun in earnest. the whole of the construction gangs and railroad staff were brought from kerma to wady halfa, and the daring pioneers of modern war started on their long march through the wilderness, dragging their railway behind them—safe and sure road which infantry, cavalry, guns, and gunboats might follow with speed and convenience.

it is scarcely within the power of words to describe the savage desolation of the regions into which the line and its constructors plunged. a smooth ocean of bright-coloured sand spread far and wide to distant horizons. the tropical sun beat with senseless perseverance upon the level surface until it could scarcely be touched with a naked hand, and the filmy air glittered and shimmered as over a furnace. here and there huge masses of crumbling rock rose from the plain, like islands of cinders in a sea of fire. alone in this vast expanse stood railhead—a canvas town of 2,500 inhabitants, complete with station, stores, post-office, telegraph-office, and canteen, and only connected with the living world of men and ideas by two parallel iron streaks, three feet six inches apart, growing dim and narrower in a long perspective until they were twisted and blurred by the mirage and vanished in the indefinite distance.

every morning in the remote nothingness there appeared a black speck growing larger and clearer, until with a whistle and a welcome clatter, amid the aching silence of ages, the 'material' train arrived, carrying its own water and 2,500 yards of rails, sleepers, and accessories. at noon came another speck, developing in a similar manner into a supply train, also carrying its own water, food and water for the half-battalion of the escort and the 2,000 artificers and platelayers, and the letters, newspapers, sausages, jam, whisky, soda-water, and cigarettes which enable the briton to conquer the world without discomfort. and presently the empty trains would depart, reversing the process of their arrival, and vanishing gradually along a line which appeared at last to turn up into the air and run at a tangent into an unreal world.

the life of the strange and lonely town was characterised by a machine-like regularity, born perhaps of the iron road from which it derived its nourishment. daily at three o'clock in the morning the 'camp-engine' started with the 'bank parties.' with the dawn the 'material' train arrived, the platelaying gangs swarmed over it like clusters of flies, and were carried to the extreme limit of the track. every man knew his task, and knew, too, that he would return to camp when it was finished, and not before. forthwith they set busily to work without the necessity of an order. a hundred yards of material was unloaded. the sleepers were arranged in a long succession. the rails were spiked to every alternate sleeper, and then the great 80-ton engine moved cautiously forward along the unballasted track, like an elephant trying a doubtful bridge. the operation was repeated continually through the hours of the burning day. behind the train there followed other gangs of platelayers, who completed the spiking and ballasting process; and when the sun sank beneath the sands of the western horizon, and the engine pushed the empty trucks and the weary men home to the railhead camp, it came back over a finished and permanent line. there was a brief interval while the camp-fires twinkled in the waste, like the lights of a liner in mid-ocean, while the officers and men chatted over their evening meal, and then the darkness and silence of the desert was unbroken till morning brought the glare and toil of another long day.

so, week in, week out, the work went on. every few days saw a further advance into the wilderness. the scene changed and remained unaltered—'another, yet the same.' as wady halfa became more remote and abu hamed grew near, an element of danger, the more appalling since it was peculiar, was added to the strange conditions under which the inhabitants of railhead lived. what if the dervishes should cut the line behind them? they had three days' reserve of water. after that, unless the obstruction were removed and traffic restored, all must wither and die in the sand, and only their bones and their cooking-pots would attest the folly of their undertaking.

by the 20th of july a hundred and thirty miles of line had been finished, and it became too dangerous to advance further until abu hamed had been cleared of the dervish force. they were still a hundred miles away, but camels travel fast and far, and the resources of the enemy were uncertain. it appeared that progress would be checked, but on the 7th of august general hunter, marching from merawi along the river bank, attacked and took abu hamed—an operation which will be described hereafter. work was at once resumed with renewed energy. the pace of construction now became remarkable. as much as 5,300 yards of track was surveyed, embanked, and laid in a single day. on the 1st of november abu hamed was reached, and by the banks of the nile the men who had fought their way across the desert joined hands with those who had fought their way along the river.

the strain and hardship had not, however, been without effect on the constructors. two of the engineer subalterns—polwhele and cator—out of the eight concerned in the laying of the dongola and the desert railways had died. their places were eagerly filled by others.

the completion of the line was accelerated by nearly a month through the fortunate discovery of water. at the beginning of july a well was sunk in what was thought to be a likely place at 'no. 4 station,' seventy-seven miles from halfa. after five weeks' work water was found in abundance at a depth of 90 feet. a steam-pump was erected, and the well yielded a continual supply. in october a second well was sunk at 'no. 6 station,' fifty-five miles further on, whence water was obtained in still greater quantity. these discoveries modified, though they did not solve, the water question. they substantially increased the carrying capacity of the line, and reduced the danger to which the construction gangs were exposed. the sinking of the wells, an enterprise at which the friendly arabs scoffed, was begun on the sirdar's personal initiative; but the chronicler must impartially observe that the success was won by luck as much as by calculation, for, since the first two wells were made, eight others of greater depth have been bored and in no case has water been obtained.

as the railway had been made, the telegraph-wire had, of course, followed it. every consignment of rails and sleepers had been accompanied by its proportion of telegraph-poles, insulators, and wire. another subaltern of engineers, lieutenant manifold, who managed this part of the military operations against the arabs, had also laid a line from merawi to abu hamed, so that immediate correspondence was effected round the entire circle of rail and river.

the labours of the railway battalion and its officers did not end with the completion of the line to abu hamed. the desert railway was made. it had now to be maintained, worked, and rapidly extended. the terminus at halfa had become a busy town. a mud village was transformed into a miniature crewe. the great workshops that had grown with the line were equipped with diverse and elaborate machines. plant of all kinds purchased in cairo or requisitioned from england, with odds and ends collected from ishmail's scrap heaps, filled the depots with an extraordinary variety of stores. foundries, lathes, dynamos, steam-hammers, hydraulic presses, cupola furnaces, screw-cutting machines, and drills had been set up and were in continual work. they needed constant attention. every appliance for repairing each must be provided. to haul the tonnage necessary to supply the army and extend the line nearly forty engines were eventually required. purchased at different times and from different countries, they included ten distinct patterns; each pattern needed a special reserve of spare parts. the permutations and combinations of the stores were multiplied. some of the engines were old and already worn out. these broke down periodically. the frictional parts of all were affected by the desert sand, and needed ceaseless attention and repair. the workshops were busy night and day for seven days a week.

to the complication of machinery was added the confusion of tongues. natives of various races were employed as operatives. foremen had been obtained from europe. no fewer than seven separate languages were spoken in the shops. wady halfa became a second babel. yet the undertaking prospered. the engineer officers displayed qualities of tact and temper: their director was cool and indefatigable. over all the sirdar exercised a regular control. usually ungracious, rarely impatient, never unreasonable, he moved among the workshops and about the line, satisfying himself that all was proceeding with economy and despatch. the sympathy of common labour won him the affection of the subalterns. nowhere in the soudan was he better known than on the railroad. nowhere was he so ardently believed in.

it is now necessary to anticipate the course of events. as soon as the railway reached abu hamed, general hunter's force, which was holding that place, dropped its slender camel communications with merawi and drew its supplies along the new line direct from wady halfa. after the completion of the desert line there was still left seventeen miles of material for construction, and the railway was consequently at once extended to dakhesh, sixteen miles south of abu hamed. meanwhile berber was seized, and military considerations compelled the concentration of a larger force to maintain that town. the four battalions which had remained at merawi were floated down stream to kerma, and, there entraining, were carried by halfa and abu hamed to dakhesh—a journey of 450 miles.

when the railway had been begun across the desert, it was believed that the nile was always navigable above abu hamed. in former campaigns it had been reconnoitred and the waterway declared clear. but as the river fell it became evident that this was untrue. with the subsidence of the waters cataracts began to appear, and to avoid these it became necessary first of all to extend the railway to bashtinab, later on to abadia, and finally to the atbara. to do this more money had to be obtained, and the usual financial difficulties presented themselves. finally, however, the matter was settled, and the extension began at the rate of about a mile a day. the character of the country varies considerably between abu hamed and the atbara river. for the first sixty miles the line ran beside the nile, at the edge of the riparian belt. on the right was the cultivable though mostly uncultivated strip, long neglected and silted up with fine sand drifted into dunes, from which scattered, scraggy dom palms and prickly mimosa bushes grew. between the branches of these sombre trees the river gleamed, a cool and attractive flood. on the left was the desert, here broken by frequent rocks and dry watercourses. from bashtinab to abadia another desert section of fifty miles was necessary to avoid some very difficult ground by the nile bank. from abadia to the atbara the last stretch of the line runs across a broad alluvial expanse from whose surface plane-trees of mean appearance, but affording welcome shade, rise, watered by the autumn rains. the fact that the railway was approaching regions where rain is not an almost unknown phenomenon increased the labour of construction. to prevent the embankments from being washed away in the watercourses, ten bridges and sixty culverts had to be made; and this involved the transport over the railway of more than 1,000 tons of material in addition to the ordinary plant.

by the arrival of the reinforcements at berber the fighting force at the front was doubled: doubled also was the business of supply. the task of providing the food of an army in a desert, a thousand miles from their base, and with no apparent means of subsistence at the end of the day's march, is less picturesque, though not less important, than the building of railways along which that nourishment is drawn to the front. supply and transport stand or fall together; history depends on both; and in order to explain the commissariat aspect of the river war, i must again both repeat and anticipate the account. the sirdar exercised a direct and personal supervision over the whole department of supply, but his action was restricted almost entirely to the distribution of the rations. their accumulation and regular supply were the task of colonel rogers, and this officer, by three years of exact calculation and unfailing allowance for the unforeseen, has well deserved his high reputation as a feeder of armies.

the first military necessity of the war was, as has been described, to place the bulk of the egyptian army at akasha. in ordinary circumstances this would not have been a serious commissariat problem. the frontier reserves of food were calculated to meet such an emergency. but in 1895 the crops in egypt had been much below the average. at the beginning of 1896 there was a great scarcity of grain. when the order for the advance was issued, the frontier grain stores were nearly exhausted. the new crops could not be garnered until the end of april. thus while the world regarded egypt as a vast granary, her soldiers were obliged to purchase 4,000 tons of doura and 1,000 tons of barley from india and russia on which to begin the campaign.

the chief item of a soldier's diet in most armies is bread. in several of our wars the health, and consequently the efficiency, of the troops has been impaired by bad bread or by the too frequent substitution of hard biscuit. for more than a year the army up the river ate 20 tons of flour daily, and it is easy to imagine how bitter amid ordinary circumstances would have been the battle between the commissariat officers, whose duty it was to insist on proper quality, and the contractors—often, i fear, meriting the epithet 'rascally'—intent only upon profit. but in the well-managed egyptian service no such difficulties arose. the war department had in 1892 converted one of ismail pasha's gun factories near cairo into a victualling-yard. here were set up their own mills for grinding flour, machinery for manufacturing biscuit to the extent of 60,000 rations daily, and even for making soap. three great advantages sprang from this wise arrangement. firstly, the good quality of the supply was assured. complaints about bread and biscuit were practically unknown, and the soap—since the soldier, in contrast to the mixture of rubble and grease with which the contractors had formerly furnished him, could actually wash himself and his clothes with it—was greatly prized. secondly, all risk of contractors failing to deliver in time was avoided. lastly, the funds resulting from the economy had been utilised to form a useful corps of 150 bakers. and thus, although the purchase of foreign grain added to the expense, the beginning of the war found the commissariat of the egyptian army in a thoroughly efficient state.

vast reserves of stores were quickly accumulated at assuan. from these not an ounce of food was issued without the sirdar's direct sanction. at the subsidiary depot, formed at wady halfa, the same rule prevailed. the man who was responsible to no one took all the responsibility; and the system whereby a chief of the staff is subjected to the continual bombardment of heads of departments was happily avoided. sufficient supplies having been accumulated at akasha to allow of a forward movement, firket was fought. after firket the situation became difficult, and the problem of the supply officers was to keep the troops alive without delaying the progress of the railway with the carriage of their food. a small quantity of provisions was painfully dragged, with an average loss of 50 per cent from theft and water damage, up the succession of cataracts which obstruct the river-way from halfa to kosheh. camel convoys from railhead carried the rest. but until the line reached kosheh the resources of the transport were terribly strained, and at one time it was even necessary to send the mounted troops north to avoid actual famine. the apparent inadequacy of the means to the end reached a climax when the army moved southward from dulgo. the marches and halts to dongola were estimated to take ten days, which was the utmost capacity of camel and steam transport, a few boat-loads of grain might be captured; a few handfuls of dates might be plucked; but scarcely any local supplies would be available. the sailing-boats, which were the only regular means of transport, were all delayed by the adverse winds. fortune returned at the critical moment. by good luck on the first day of the march the north wind began to blow, and twelve days' supplies, over and above those moved by camel and steamer, reached dongola with the troops. with this reserve in hand, the occupation of the province was completed, and although the army only existed from hand to mouth until the railway reached kerma, no further serious difficulty was experienced in supplying them.

the account of the commissariat is now complete to the end of the dongola expedition; but it may conveniently be carried forward with the railway construction. in the abu hamed phase the supplies were so regulated that a convoy travelling from murat wells along the caravan route arrived the day after the fight; and thereafter communications were opened with merawi. the unexpected occupation of berber, following abu hamed, created the most difficult situation of the war. until the railway was forced on to berber a peculiarly inconvenient line of supply had to be used; and strings of camels, scattering never less than 30 per cent of their loads, meandered through the rough and thorny country between merawi and abu hamed. this line was strengthened by other convoys from murat and the approaching railhead, and a system of boats and camel portages filtered the supplies to their destination.

even when the railway had reached dakhesh the tension was only slightly relaxed. the necessity of supplying the large force at berber, 108 miles from railhead, still required the maintenance of a huge and complicated system of boat and camel transport. of course, as the railway advanced, it absorbed stage after stage of river and portage, and the difficulties decreased. but the reader may gain some idea of their magnitude by following the progress of a box of biscuits from cairo to berber in the month of december 1897. the route was as follows: from cairo to nagh hamadi (340 miles) by rail; from nagh hamadi to assuan (205 miles) by boat; from assuan to shellal (6 miles) by rail; from shellal to halfa (226 miles) by boat; from halfa to dakhesh (railhead)—248 miles—by military railway; from dakhesh to shereik (45 miles) by boat; from shereik by camel (13 miles) round a cataract to bashtinab; from bashtinab by boat (25 miles) to omsheyo; from omsheyo round another impracticable reach (11 miles) by camel to geneinetti, and thence (22 miles) to berber by boat. the road taken by this box of biscuits was followed by every ton of supplies required by 10,000 men in the field. the uninterrupted working of the long and varied chain was vital to the welfare of the army and the success of the war. it could only be maintained if every section was adequately supplied and none were either choked or starved. this problem had to be solved correctly every day by the transport officers, in spite of uncertain winds that retarded the boats, of camels that grew sick or died, and of engines that repeatedly broke down. in the face of every difficulty a regular supply was maintained. the construction of the railway was not delayed, nor the food of the troops reduced.

the line continued to grow rapidly, and as it grew the difficulties of supply decreased. the weight was shifted from the backs of the camels and the bottoms of the sailing-boats to the trucks of the iron road. the strong hands of steam were directed to the prosecution of the war, and the swiftness of the train replaced the toilsome plodding of the caravan. the advance of the dervishes towards berber checked the progress of the railway. military precautions were imperative. construction was delayed by the passage of the 1st british brigade from cairo to the front, and by the consequently increased volume of daily supplies. by the 10th of march, however, the line was completed to bashtinab. on the 5th of may it had reached abadia. on the 3rd of july the whole railway from wady halfa to the atbara was finished, and the southern terminus was established in the great entrenched camp at the confluence of the rivers. the question of supply was then settled once and for all. in less than a week stores sufficient for three months were poured along the line, and the exhausting labours of the commissariat officers ended. their relief and achievement were merged in the greater triumph of the railway staff. the director and his subalterns had laboured long, and their efforts were crowned with complete success. on the day that the first troop train steamed into the fortified camp at the confluence of the nile and the atbara rivers the doom of the dervishes was sealed. it had now become possible with convenience and speed to send into the heart of the soudan great armies independent of the season of the year and of the resources of the country; to supply them not only with abundant food and ammunition, but with all the varied paraphernalia of scientific war; and to support their action on land by a powerful flotilla of gunboats, which could dominate the river and command the banks, and could at any moment make their way past khartoum even to sennar, fashoda, or sobat. though the battle was not yet fought, the victory was won. the khalifa, his capital, and his army were now within the sirdar's reach. it remained only to pluck the fruit in the most convenient hour, with the least trouble and at the smallest cost.