To Arms

by Hew Strachan

  The possibility that the forces of South Africa would be dispatched to East Africa once South-West Africa had been overrun had been considered by Lettow and Schnee in June and July. A descent on Dar es Salaam or Bagamoyo, a landward thrust following the railway under construction from Voi, or a combination of the two—all contributed to Lettow’s concerns for protection to the north and east. But by October he had convinced himself that the South Africans would go to the western front or to Gallipoli. His illusions were shattered on the very day he ordered the support for Kasigao. The district commissioner of Lindi reported that the press of Portuguese East Africa and of South Africa had announced that Smuts and up to 25,000 men were bound for the East African theatre in order to launch an attack early in the new year. Moreover, Portugal, German East Africa’s southern neighbour, seemed bound to enter the war. A week later a letter captured at Tsavo, originating from London, confirmed the intelligence.

  Lettow abandoned his offensive. He was still uncertain whether the attack would come from the sea or from Voi. But the coastal climate, and its unsuitability for white mounted troops, suggested the Kilimanjaro area as more likely. British reconnaissance activity corroborated Lettow’s analysis. From mid-January the thrust of British strategy was manifest. Longido, a hill north-west of Kilimanjaro, was occupied on 15 January, Mbuyuni to the south-west was taken on the 22nd, and Serengeti on the 24th. The Germans abandoned Kasigao. On 12 February a major British attack on Salaita Hill (known to the Germans as Oldorobo), which barred the way to Taveta, was repulsed. Lettow concentrated almost half his total forces, 800 Europeans and 5,200 askaris, together with forty-seven machine-guns and ten field guns, in the Kilimanjaro area.199 His indirect bulwark, South-West Africa, had collapsed; the fight for German East Africa was about to begin.

  EAST AFRICA, 1916–1918

  The conquest, rather than the neutralization, of German East Africa, had already entered the minds of the CID subcommittee when it issued its orders to Aitken in October 1914. In February 1915 Wapshare recommended the construction of the railway line from Voi to Taveta, an essential preliminary to an offensive action from British East Africa. Thus Kitchener’s insistence on defence, sustained throughout 1915, smacked of procrastination. Not even his own director of military operations, Sir Charles Callwell, had much enthusiasm for the policy.200

  Nonetheless, British soldiers in East Africa concluded—rightly—that their needs and preoccupations did not attract much attention in the War Office. The London department much more concerned by the fights at Jasin and Saisi was the Colonial Office. The German raids across the frontier into Northern Rhodesia had created a lawless strip, whence colonialism had retreated and where a scorched-earth policy to create a neutral buffer appeared to be the only viable option. In British East Africa, the tribes around Jasin had fallen back northwards to escape German rule, and inland British prestige was being eroded by the attacks on the Uganda railway.

  On 23 January 1915 John Chilembwe, an American-educated black missionary, had led an attack directed against the employment practices of white-owned estates in Nyasaland. Chilembwe’s rebellion was limited and easily contained. But a number of elements gave cause for concern. Millenarianism, anticipating that the war would eliminate the colonial powers and would enable the black elect to enter the New Jerusalem, found an audience among the African educated elite. Traditional, tribal divisions had been overcome, even if only in a limited and specific way. Secondly, economic pressures, already evident before the war, had been compounded by the loss of labour through the recruitment of porters, and threatened to give a mass appeal to a minority movement. Thirdly, Chilembwe challenged the basis on which the war was being fought. In a mixture of Christian pacifism and natural law, he repudiated the notion that Africans should fight for white men’s rivalries, not least when their lack of property and of civic rights should have relieved them of military obligations. Chilembwe’s death did not prevent his spectre haunting colonial minds thereafter. The withdrawal of white administrators for military service, the preoccupation of those that remained with wartime needs, and the progressive Africanization of Christian missions all served to reinforce the bases of Chilembwe’s original appeal.201

  The Conservative leader Bonar Law, who succeeded Harcourt at the Colonial Office with the formation of the coalition government in May 1915, was soon convinced that the need to restore British prestige in East Africa demanded the conquest of German East Africa ‘once and for all.202 What he needed was a sufficiently large body of troops with which to do it. Tombeur’s preparation of the Belgians’ Force publique, far from solving that problem, added urgency to Colonial Office considerations: the brutality of the Belgian askaris had not contributed to good order in Northern Rhodesia, and a successful Belgian invasion around Lake Tanganyika, if independently conducted, would weaken Britain’s relative status yet further.

  As in the case of South-West Africa, London’s short-term needs conveniently merged with Pretoria’s long-term ambitions. The implicit agenda in South Africa’s act of Union was its extension to the line of the Zambezi. The motivations were at once elevated, economic, and national. Smuts identified himself with Cecil Rhodes, and saw South Africa’s task as ‘the progress of European civilization on the African continent’. Commercially, the adjacent ports for the Transvaal were not Capetown and Walvis Bay but Delagoa Bay and Beira, both in Portuguese hands. Thus, the development of Pretoria and of Afrikaner interests was stunted by the geographical configuration of the Union. The addition of South-West Africa would increase the weighting in favour of Capetown and the English population, not diminish it. The solution suggested in April and May 1915 by Harcourt and by South Africa’s governor-general, Lord Buxton, was to persuade Portugal to swap Mozambique for South-West Africa. Smuts, however, recognized that the deal would be too poor to commend itself to the Portuguese. His idea was to conquer German East Africa, and then allocate its northern territory to Britain and its southern to Portugal. In exchange, Portugal would be asked to give the southern part of its existing colony, including Delagoa Bay and Beira, to South Africa. To achieve this the South Africans were prepared to provide troops for the East African campaign, initially at imperial expense, but—if the deal worked— eventually at South Africa’s.203

  The obstacle remained the War Office. The case for allocating South African troops, Europeans of proven military worth, not to East Africa but to the western front or at least to Gallipoli was supported by Kitchener’s own determination that the main fronts were European. However, two political factors weighed against the strategic argument. First, the use of Boers in the conquest of South-West Africa, despite being an area of immediate South African interest, had aroused anti-imperial sentiment; thus, their deployment outwith the African continent was likely to be even more provocative. Secondly, imperial rates of remuneration were two-thirds less those paid to South African troops. To reward South Africans in Europe at South African rates promised ill-feeling between them and the British troops, and would create friction between London and Pretoria as to who should pay the difference. Sending the South Africans to East Africa, therefore, sidestepped the pay issue, albeit in part and only temporarily.204 The War Office could console itself with the prevalent notion that East Africa was ideal country for the operations of Boer commandos. It was not; but then nor was the western front.

  Even so, War Office agreement was not secured without subterfuge. Kitchener was absent from London, visiting Gallipoli, when the CID subcommittee reported to the War Council on 12 November 1915. The subcommittee recommended that 10,000 troops be sent to East Africa with a view to commencing operations before the April rains and to conquering the German colony with as little delay as possible. Kitchener was not at all happy when he returned. Both Sir Archibald Murray, the chief of the imperial general staff, and Callwell had colluded with Bonar Law to force his hand. Kitchener’s riposte was to ensure that no British brigade was sent as part of the 10,000 and to withhold the proper compleme
nt of supporting arms, including artillery and engineers. During December the staff of the East African force planned their campaign in daily anticipation of its cancellation.205

  Sir Horace Smith-Dorrien, the hero of Le Cateau but subsequently Sir John French’s scapegoat, was appointed to the command. Smith-Dorrien’s principle was ‘more haste less speed’. Having built up a staff of East African and colonial experience, he was convinced that nothing should be attempted until training was complete, lines of communication secure, and the rains over. He anticipated doing no more by March than drive the Germans in on Taveta; the main offensive would not be launched until June, but would then be sudden, complete, and inexorable. Kitchener had never formally sanctioned the campaign. Now Smith-Dorrien forfeited South African and Colonial Office approval as well: both were anxious for an attack before the rains. They got it. Smith-Dorrien became ill en route for Mombasa and was replaced by Smuts.206

  Bonar Law had wanted Smuts for the job all along, but the latter had initially pleaded the state of the Union’s domestic politics as reason against his going. In every other major British command of the war, professional experience with the regular army counted for more than the qualities of intellect and personality possessed by many amateurs. But Bonar Law was convinced by reports from France that ‘we are suffering from the want of brains in the higher command’. So anxious was he not to appoint a soldier to the job that he (if nobody else) was prepared to take seriously Churchill’s request to go as governor-general and commander-in-chief, equipped with a fleet of armoured cars.207 The fact that Law got his way is indicative both of the War Office’s indifference to East Africa and of the divided counsels emanating from the War Office during Kitchener’s absence. Callwell, the British army’s leading authority on colonial operations, supported Smuts. In reality, Smuts’s military experience was almost entirely irrelevant to the task that now faced him. In the Boer War he had led a commando of 300–400 men with minimal logistical needs in a defensive campaign in a moderate climate over familiar terrain. In East Africa he was responsible for a ration strength of 73,300 men, committed to the conquest of a tropical colony, much of it barely mapped. His South African experience made him ‘a remarkable soldier’, often to the front and admired by his men; but it also rendered him ‘a bad tactician and strategist’ and ‘an indifferent general’.208

  The obvious foil to such a commander was his staff. But Smuts, albeit gently, dismantled the body created by Smith-Dorrien. Thus, not one officer in Smuts’s field headquarters had ‘ever previously filled an appointment on the General Staff with troops’.209 Hoskins, whose local knowledge was to have found sensible employment as chief of staff, was shunted out to a division; J.J. Collyer, his replacement, was an ex-ranker of entirely South African experience. British commanders—Tighe, Malleson, and Stewart—were, with good reason, removed, but their places were taken by South Africans of experience comparable to Smuts’s own—Deventer, Brits, and Enslin. Divisions therefore fought their own battles, failing to report back to a staff that lacked the authority to exercise initiative. Smuts followed the procedure ordained in Field Service Regulations, but modified in practice in France, and divided his headquarters in two, with himself and a small group at his advanced headquarters, and the heads of the administrative services at base. Thus, field command and logistics were separated in a theatre of operations where their mutual dependence was paramount. Supply was initially in the hands of an Indian army colonel with a tendency to over-centralization and peacetime economies, and communications in those of a former chief of East African police, who knew the area but was highly strung and fearful of giving offence. At the end of January a veteran of the western front and a rare survivor from Smith-Dorrien’s appointees, P. O. Hazelton, took charge of transport. But lack of existing records prevented him from determining what resources units already had, and lack of time forestalled his remedying any deficits. Smuts’s continued, if paradoxical, reliance on the higher formations of European warfare, divisions and brigades, increased the logistical burden, and contrasted with Germans’ preference for the more flexible and self-contained field company. Never resolved but constantly disputed was the chain of command— between supply, transport, and communications; between the rear and the front; and between the operational and administrative branches of Smuts’s headquarters.210

  Smuts justified his practice of placing himself well forward by reference to the difficulties of communication in the bush. But although this impressed his troops, it militated against effective command. Close liaison with the heads of his administrative services was further jeopardized. Wireless was unreliable, visual signalling impossible. Cable was therefore vital. Laying it was another job for the porters; in places it had to be raised 8 metres to avoid damage by giraffes; telephones were superimposed on a single line also operating as a telegraph circuit. The entire service was described by Collyer as composed ‘of men of different nationalities—of different training—speaking different languages, with equipment of varying patterns thrown together without any coordinated training to carry out an important operation in unknown country’.211 None of the confusions generated by his polyglot force, logistic or linguistic, was resolved before Smuts advanced.

  During the course of 1915 Meinertzhagen had taken the intelligence services in hand. By a variety of methods, including the recovery of German orders that had been used as lavatory paper, he built up a picture of Lettow’s order of battle.212 But there was still a tendency to exaggerate the Germans’ strength: Smith-Dorrien put it at 2,200 whites and up to 25,000 blacks.213 Moreover, the use of captured German maps created a false security. Plausible because they ‘were presented in a form which [commanders] associated with accuracy’, in reality they omitted much. Roads built since 1914 were frequently unmarked; duplications and difficulties with place-names were not resolved; marching distances proved much greater than cartographic distances. Thus, orders based on maps proved consistently over-optimistic.214 Reconnaissance could rarely fill the gap. Thick bush obscured the ground from aerial observation and tsetse fly limited the value of mounted troops.

  Smith-Dorrien’s plan had been to attack around Kilimanjaro, while the Belgians thrust in from the north-west and a third attack from Northern Rhodesia entered in the south-west. When Lettow had been forced to commit his forces to the west, a brigade was to be landed at Dar es Salaam or Tanga, supported by four cruisers. The decisive thrust would thus have rested on secure and short lines of communications. Neither Smuts nor his staff was sea-minded. No landing at Dar es Salaam or Tanga was attempted. The main blow ran across rather than along the two main land lines, the northern and central railways. Smuts’s invasion, therefore, played to the strengths already bestowed on the German defence by the nature of the terrain. The available axes of approach were limited by the mountain ranges, and the valleys were covered with bush. As he pushed on, his line of communications lengthened. The halts to allow his supply services to catch up gave the Germans the opportunity to regroup. Thus, his onset lacked the momentum that his dispatches suggested. Throughout 1916 Smuts’s rate of advance failed to match his strategic conceptions.

  Moreover, for all his talk of battle, Smuts’s aim was to manoeuvre rather than to fight. Lettow’s avoidance of a decisive engagement throughout 1916 can be attributed to Smuts’s supply difficulties, the consequent loss of operational flexibility, and the German desire to avoid fighting in order to preserve lives and ammunition. But it can also be seen in the context of Smuts’s political preoccupations. By the end of 1916 the conduct of the campaign would be the focus of public outcry in South Africa. Smuts was constantly reminded from Pretoria of the difficulties of raising men; he dreaded returning to the Union dubbed a butcher.215 The fighting in South-West Africa suggested that indeed wars could be won by sweat rather than by blood, by mobility rather than by hard fighting. Neglecting or even abandoning lines of communication had been made possible by the speed of envelopments conducted by mounted troops. And so Smu
ts planned a sequence of envelopment battles, where success eluded him because of Lettow’s refusal to fight. In reality, East Africa was—particularly thanks to the tsetse fly—not appropriate for mounted troops. His flank attacks, partly no doubt because of his supply problems, never extended sufficiently far to master German communications. Instead, he would probe towards the German wing without fixing the enemy frontally, so dispersing his troops and enabling the numerically inferior Germans to use the bush to break the battle up into a series of isolated fire-fights. Forced to deploy on ground of Lettow’s choosing, Smuts would have to wait for the arrival of heavier weaponry; no attempt would be made to retain contact as night fell. Thus Lettow could escape because he had never been gripped.216

  Plans for the initial attack east and west of Kilimanjaro had already been drawn up when Smuts arrived. The main German concentration was in the sector bounded by Salaita-Moshi-Kahe. Lettow had left only weak forces west of Kilimanjaro. His main concern was for his line of retreat down the northern railway, and he therefore watched with anxiety the Ngulu Gap through the Pare Mountains, opposite Lembeni. The original British intention was to launch their mounted troops not here but west of Kilimanjaro, past Longido, on to Moshi, in order to cut off the Germans protecting Taveta.

  Smuts amended this plan. He recognized the strength of the German defences at Salaita, but calculated that Lettow had too few troops for his area of concentration. He therefore brought the major thrust east of Kilimanjaro, directed not at Salaita itself but in a flanking move to its north. On the night of 7/8 March the South African brigades under Deventer marched on Chala, north of Salaita, and on the morning of the 8th the 2nd division moved into positions in front of Salaita. Despite the British patrols sent towards the Ngulu Gap, Lettow realized the true direction of the advance only on the 8th. When the 2nd division launched its attack on Salaita on 9 March it found the German trenches untenanted.