by Hew Strachan
On 16 August the two western columns debouched from the Ngurus and at Dakawa ran into German fire from across the Wami. Sheppard’s column had already crossed the Wami at Kipera, and moving up the right bank threatened to strike the German flank on the 17th. The Germans fought a delaying action and then fell back on the night of 17 August, Lettow and the main force retiring on Morogoro and Kraut moving south-westwards towards Mahenge.
On 16 August Smuts began to realize that Lettow did not propose to oblige him by fighting it out on the central railway.229 He nonetheless hoped to be able to pin the Germans against the Uluguru Mountains, a range rising in places to 3,000 metres, and whose northern tip extended to Morogoro and the railway. Deventer reached Kilosa on 22 August. But his division was exhausted, and he reported that he could not continue unless the Germans facing him were threatened from the rear. Smuts therefore directed Enslin’s 2nd mounted brigade to the south-west, to strike the central railway at Mkata. Deventer and Enslin, with Enslin leading, were then to push south and west of the Ulugurus to Mlali, while Smuts’s main body moved to the east of the range. Smuts did not know whether Lettow planned to retire to Mahenge or to the Rufiji river: his hunch in favour of the latter explained the greater weight attached to his eastern hook. But once again he ducked a frontal assault. Arguing that the hills which screened Morogoro to the north would make a direct advance from the Wami a costly and protracted operation, he put all his weight into the flanks. Enslin reached Mlali on 23 August; Mikese, 32 kilometres east of Morogoro, was gained on the 26th. Thus, when Morogoro itself was occupied on the same day, no German troops remained to contest Smuts’s entry. Abandoned German supplies bore testimony to the speed of the British advance, but by now Smuts’s operational style was too manifest for there to be any likelihood of Lettow waiting to fight. Lettow’s orders of 26 August established the main German positions at the southern end of the Ulugurus, on the further bank of the River Mgeta. By deploying his troops in depth to the north, he hoped to render them proof against British turning movements, and also to enable them to strike blows against isolated components of Smuts’s forces, scattered by design and divided by the terrain.
On 5 September Brits’s 3rd division, which had taken up the western movement from Deventer’s shattered 2nd division, was approaching Kisaki. Lettow had 600,000 kilos of supplies dumped here, and the poverty of the region south of the Mgeta encouraged him to fight for their protection.230 On Brits’s left, plunging through the Ulugurus themselves, was Nussey’s 1st mounted brigade, detached from Deventer’s division and now without its horses. Both commanders had wireless, and on 7 September, the day Brits attacked, each was within 8 kilometres of the other. But ‘once off the road ... even the sky overhead was invisible, and one could not see one’s next-door neighbour three feet away’.231 Not even the sound of Brits’s rifle fire was audible to Nussey. His wireless was lost over a precipice. Lettow was able to deal with each in turn, smashing Brits on 7 September and checking Nussey on the 8th.
To the east the pattern was repeated. Delayed by the need to repair roads and bridges, Hoskins’s 1st division did not reach Tulo until 9 September. The Germans, in entrenched positions on the Dunthumi river, could therefore be reinforced by the victors of Kisaki. The Germans fought a stubborn action from the 10th to the 12th, but in their wish to counter-attack were as hampered by the terrain as the British were in their advances. On 14 September Lettow abandoned Kisaki, withdrawing across the Mgeta and establishing a new base at Beho-Beho.
Smuts now paused. Most of his staff thought he should have done so at Morogoro. Deventer’s 2nd division was 650 kilometres from its railhead, Brits’s 3rd division 390 kilometres, and Hoskins’s 1st division 360 kilometres. All three were on half rations; Hoskins’s had no forward dumps and depended on a daily lift of 17,000 to 20,000 pounds.232 Smuts used lorries as far as he could. But, although formally speaking the rains had not yet begun, rain was falling and within a day tracks across the black cotton soil were turned into a sea of ‘sticky, black mud’.233 Baggage animals could not make good the deficiency. Of 54,000 mules, donkeys, horses, and oxen put to work on the supply lines around and south of the central railway between June and September 1916, all but 600 fell prey to the tsetse fly.234
The health of the humans was little better. By May 1916, within three months of their arrival, most South African units had already lost half their strength to disease. The 9th South African Infantry mustered 1,135 all ranks on 14 February; on 25 October its parade state stood at 116. The 2nd Rhodesia Regiment deployed 1,038 all ranks between March 1915 and its departure from the theatre in January 1917. Thirty-six were killed in action, eighty-four wounded, and thirty-two died of disease. But there were 10,626 cases of sickness, including 3,127 of malaria, and 2,272 of these resulted in hospital admission.235 Some at least were the product of the punishing rates of march, and of malnutrition caused by insufficient food or inadequate cooking. Thus, ‘what Smuts saves on the battlefield he loses in hospital, for it is Africa and its climate we are really fighting not the Germans’.236 An early battle, a frontal assault when lines of communication were short, might ultimately have proved less costly than long-range but combat-free manoeuvring.
The means by which to improve the supply position were available. At the beginning of September 1916 Smuts’s base was still Mombasa. And yet Tanga was taken on 7 July. Smuts, convinced that complete victory was imminent, saw its restoration and use as of no value. The British entered Dar es Salaam on 3 September. Little energy was put into its reopening. Smuts’s administrative staff remained at Tanga to the north, while his shipping was securing the ports of Kilwa, Lindi, Sudi, and Mikindani to the south. The damage to the central railway was not as severe as it might have been. The Germans had destroyed the bridges and removed much of the rolling stock to its western end, around Tabora. But the track was largely intact. By converting Ford cars to run on the rails, the British were able to use the permanent way almost immediately. During November locomotives once again linked Dar es Salaam and Morogoro. If Smuts had paused on the central railway, and re-established his communications through Dar es Salaam before pushing on into the Ulugurus, he might have been able to terminate the campaign north of the Mgeta, or at least of the Rufiji.
However, the advance to the central railway had inflicted at least one major blow on the Schütztruppen. Lettow’s theatre of operations was bisected. In the west Wahle and his men were out of contact, left to conduct their own operations. The campaign which he and Tombeur waged was comparable in range and significance with that in the east. It served neither British colonial interests, nor the post-war publicity accorded to Smuts and Lettow, to acknowledge the fact. But Tabora, Wahle’s headquarters, was the largest town in German East Africa, centrally positioned in the colony. According to a British estimate of 1 April 1916 as many as 373 European troops and 7,650 askaris were deployed in the square Kivu-Mwanza-Tabora-Ujiji.237 In themselves these figures confirm how exaggerated were Meinertzhagen’s estimates of Lettow’s total strength. Wahle’s disposable force actually numbered about 2,000. But the calculation reveals how dependent Smuts’s thrusts in the east were on Tombeur’s simultaneous commitment in the west. The troops concentrated under Lettow’s direct command in the north only just exceeded half the Schütztruppen’s total force.
The campaign in the west also showed how unsympathetic to guerrilla operations and how wedded to colonial stability were many Germans. Both at Bukoba and at Mwanza the principles of German peacetime administration persisted for the first two years of the war. The local economy was nurtured, porters only served within their own regions, and productivity was sustained. Loyalty to Germany enabled auxiliaries to be raised, and even prompted rebellion across the Ugandan frontier at Kigezi. The basis existed for a sustained popular defence of Ruanda and Urundi. But instead the people were prepared for the arrival of the British and told to co-operate with them. The commander of the Bukoba area’s final words were that the Germans ‘wished to f
ind the country in the same condition as they left it when they returned in three months time after the approaching German victory in Europe’. A preference for German colonialism, not the prosecution of the European war, led to the rejection of revolutionary methods.238
The first phase of the Entente’s offensive in the west was bizarre, even in a theatre of operations dominated by larger-than-life personalities. Lieutenant-Commander G. B. Spicer-Simson was one of the Royal Navy’s less distinguished officers. In command of a destroyer he had sunk a liberty boat, and he had been ashore entertaining some ladies when a gunboat anchored under his orders had been torpedoed. Nonetheless, Spicer-Simson was selected to command two gunboats that were hauled overland from Capetown via Elizabethville to Lake Tanganyika. Most of the journey was by rail, but the significant sections were not. The expedition sailed from England on 15 June 1915; on Boxing Day Spicer-Simson’s two craft put out into the waters of Lake Tanganyika. By February two German gunboats were accounted for, but a third, the Graf von Götzen, a much bigger vessel, only recently completed and mounting one of the Königsberg’s 10.5 cm guns, remained at large. Spicer-Simson, who had now taken to wearing a skirt with his commander’s jacket, refused Belgian pleas to engage the Graf von Götzen. Instead, he set off to the southern end of the lake to support the British forces on the Northern Rhodesian frontier. The Germans’ control of the lake was sufficiently dented to allow the Belgians to transport goods along its western coast, but the navigation of the eastern bank remained in their hands. Thus, both sides were able to use the lake route to feed their forces in Ruanda and Urundi.239
For the Belgians, supply was a major headache. In August 1914 the Force publique concentrated 1,395 men along the Congo; in May 1916 Tombeur’s strength was 719 Europeans and 11,698 blacks (a much lower ratio of Europeans to blacks than the other powers thought advisable). They were deployed in three groups: a brigade north of Lake Kivu close to the Ugandan frontier (Molitor’s); a brigade on the Russissi (Olsen’s); and a defensive group on the western shore of Lake Tanganyika. No food reserves had been formed before the offensive. The Belgians hoped to live off the land. But the 1916 harvest was not yet in, and the 1915 crop was either eaten or destroyed.240 Dispersing to requisition, the Belgian askaris developed a fearsome reputation. Tombeur had stressed that the Belgians’ war was with the troops of Germany, not with the inhabitants of Ruanda and Urundi;241 in reality, the lack of sufficient European officers to ensure adequate supervision undermined the hopes of local collaboration. Each of the groups needed 7,000–8,000 porters for its own needs in addition to the porters for the lines of communication. The former were locally impressed. Tombeur tried to recruit the latter from the interior of the Congo, but eventually had to ask the British for 5,000 porters and 100 ox-wagons. The effect of this British contribution was temporarily to rob the troops in Uganda, Lakeforce (commanded by Sir Charles Crewe), of their offensive capacity.
Tombeur, like Smuts, was more interested in the conquest of territory than in the defeat of the Schütztruppen. Moreover, the examples of Tanga and Kahe convinced him that his opponents were masters of the defensive battle. Although sixteen 70 mm howitzers, originally ordered by the Mexican government, had been promised him by the French, they had yet to arrive. Therefore, the Force publique was instructed to shun attacks on trenches and strong-points. The Germans would thus be denied the opportunity of crushing one Belgian column before turning against the other, and so would be robbed of any advantage to be derived from operating on interior lines. Tombeur wanted to combine tactical conservatism—the holding of ground once gained, the avoidance of defeat, the keeping of casualties to a minimum—with a strategy of manoeuvre. And yet the porter problem, the lack of intelligence, and the absence of effective communications between the brigades all constrained such operations. The best that Tombeur could do was to set down his general principles and to agree a common date on which Molitor and Olsen were to advance.242
On 5 April Wintgens, the officer commanding the three Schütztruppen companies in Ruanda, received a letter from the Belgians offering an armistice. Tombeur said that the war in East Africa had been begun by Germany, not Belgium, and therefore asked Schnee for compensation. Lettow, his mind focused on the eastern theatre, was not unenthusiastic; Schnee saw it as a Belgian ruse to lower Wahle’s guard, and said he would have to refer the whole matter to Berlin. In fact Tombeur was behaving with remarkable consistency. An armistice on such terms would have fulfilled Belgian objectives to the letter.243
On 12 April the Belgians, without replying to Schnee, opened fire on the Russissi front. Tombeur imagined that the Germans would fall back to a line flanked by Kigali in the north and Nyanza in the south. The northern brigade demonstrated against German positions on the River Sebea while directing its main body towards Kigali. Wintgens’s communications with Wahle were cut on 22 April, and on 2 May he ordered the evacuation of Kigali. Further south, Olsen got across the Russissi and took Nyanza on 21 May. The Belgians had won Ruanda by manoeuvre alone.
So far Molitor’s and Olsen’s columns had pursued convergent lines of march. But now annexationism and operational necessity created divergent objectives. Olsen thrust south at the end of May towards Usumbura and Kitega (entered on 17 June), in order to secure Urundi. Molitor moved east towards the south-western corner of Lake Victoria in order to collaborate with Crewe’s Lakeforce. Thus the Belgians fanned out, their force-to-space ratio diminishing and their units increasingly pursuing independent objectives. The Belgian government’s priority was the seizure of Ujiji and the domination of Lake Tanganyika; Olsen, therefore, had the principal role. But for the British Molitor was more important. They wanted him to combine with Crewe in the capture of Mwanza. By the end of June both objectives had been adopted. On Crewe’s suggestion Mwanza, once secured, was to be the base for a thrust on Tabora. Olsen would support by moving eastwards from Ujiji along the central railway.
In practice, Olsen’s progress proved more rapid than Molitor’s: thus Belgian aims prevailed over British. Wintgens had reckoned that the rains would cause Tombeur to halt on the Kigali-Nyanza line. In reality, the lack of porters proved a greater impediment to pursuit. Those recruited in Ruanda were reluctant to follow Olsen into Urundi. The southern brigade overcame its immediate problems by shifting its base to Usumbura on Lake Tanganyika. It then occupied Ujiji without opposition, Wahle having decided to concentrate his forces north of Tabora. The British took Mwanza on 14 July. By 19 September 40,000 loads were accumulated at Mwanza. But its value as a base was minimized by lack of porters to carry them forward. The area south of Mwanza was waterless, its resources stripped, and its population shifted by the Germans. In the Belgian column seven out of every twelve loads were needed to feed the porters themselves. Crewe’s force had 10,000 porters for 2,800 combatants.244
By the beginning of August Molitor reckoned Wahle’s major positions were at Kahama, south of Whimo, with a supporting group opposite Crewe at Shinyanga. Crewe concluded that the Shinyanga positions were stronger. Each, therefore, saw himself as having the principal task and requested his ally to support him. For Tombeur, Molitor’s job was to continue due south, so relieving the pressure on Olsen as soon as possible. If Crewe was right, then Molitor’s troops would turn the Germans at Shinyanga without their having to lose time by marching to the east. Crewe, nonetheless, insisted on Molitor’s direct support. He was proved wrong, but not until 17 August was Molitor able to resume the original direction of his advance.245 On 28 August Crewe took the Shinyanga position without fighting.
In Tombeur’s mind he was commanding a massive concentric attack on Tabora, with Olsen’s southern brigade leading while Molitor’s northern brigade fixed the Germans around Whimo. In reality, Tombeur was a spectator. By early September his forces described a quarter-circle of nearly 200 kilometres, each brigade’s base 400 kilometres distant from the other. On 2 September an intercepted German signal reported a German victory over Olsen’s brigade the previous day at Ma
bama. This was Tombeur’s first indication of Olsen’s current position. The check to Olsen was minor, only one battalion having been engaged at Mabama, but Tombeur now decided that the northern brigade should take up the running while the southern held its ground. However, the supply problems of Molitor’s brigade made his advance sporadic and slow. Each daily stage of his route could accommodate only two battalions, and his troops carried sufficient ammunition for a single day’s fighting. On 9 September he took up defensive positions in order to reconcentrate. The southern brigade, not the northern, made the offensive efforts of the next four days. Even when direct contact between the two brigades was established on the 14th, the northern brigade remained on the defensive. Olsen’s push on Lulanguru on 16 September was thus unsupported.
Wahle knew of Tombeur’s intentions from radio intercepts and therefore established his main positions north of Tabora at Itanga. The proximity of Itanga and Lulanguru, particular given the railway line, enabled him to move troops from one to the other. But his force had now fallen to 1,100 rifles. He disbanded the naval unit under his command in order to redistribute its Europeans among the field companies. Nonetheless, desertions multiplied. Many of the askari came from the Tabora region and had no intention of leaving it. The Germans no longer had the men to defend the town’s perimeter. On 16 September a captured letter, intended by Crewe for Molitor, revealed that the main blow from the north would be delivered on the 19th. Wahle divided his command into three components, two to go east along the central railway before turning south to Kiromo, and one to go due south from Tabora to Sikonge. The Germans abandoned their positions under cover of dark on the 18th.246
