by Hew Strachan
The pursuit of the Schütztruppen was half-hearted and limited. The occupation of Tabora on 19 September marked the limit of Belgian territorial ambitions. Logistic constraints as well as political directives tied Tombeur to the town. Now that the Force publique was reconcentrated its demands for food exceeded the capacity of the local supplies. The plight of Lakeforce, outside Tabora to the east, and 390 kilometres from its base, was even worse. On 3 October Crewe’s command was dissolved. Had military priorities carried more weight than colonial rivalries in allied counsels, Olsen’s advance from the west would have been held back. Crewe might then have hit the railway to the east before the attack from the west had made itself felt. As it was, Olsen’s outpacing of Molitor and Crewe meant that Wahle’s path to the south and east lay open. His three columns made off towards Iringa, in a bid to reunite with Lettow and the main force.
The Germans’ retreat south of the central railway gave the British advance from Northern Rhodesia more than local significance. In origin, however, the Colonial Office’s decision to switch from defence to attack related not to the East African campaign as a whole but to the increasing instability in Nyasaland and Northern Rhodesia. The colonization of the northern parts of both territories was not complete in 1914. The district staff in Northern Rhodesia fell from 102 in 1914 to seventy-six by the end of 1915. And yet the strains on the colony multiplied. The German incursion across the frontier—and the demand for labour to support the troops needed to counter it—threw the value of British protection into question and caused economic crisis. Famine was evident by late 1915. Brigadier-General Edward Northey, the only British general in East Africa for whom the Germans confessed admiration, was appointed to the command, and his force boosted to 2,500 men. Having established four posts from Karonga to Abercorn, he took the offensive on 25 May 1916.247
But Northey’s attack exacerbated many of the colonial and economic problems it was designed to mitigate. The railhead for his force was at Ndola, 960 kilometres from Abercorn and 730 kilometres from his main base at Kasama. The road cut from Kashitu, south of Ndola, to Kisama and Abercorn in 1915 was not suitable for wheeled traffic, particularly in the rains; an experiment in the use of oxen in 1914 fell victim to tsetse fly. So porters were vital. The demands on the adult male population were enormous. In Nyasaland alone, perhaps 200,000 men, or 83 per cent of the total available, were employed as labourers in Northey’s campaign. In north-east Rhodesia 92,337 carriers were engaged between April 1915 and March 1916, and 138,930 between April 1916 and March 1917. At a local level Abercorn, with a population of 8,500 taxable males, contributed 5,000 carriers and 800 road-builders by August 1916; Luwingu, with 7,000 taxable males, drew in sufficient labour from outside to account for 12,786 engagements between March 1915 and March 1916. The effect of such massive new employment was to jeopardize local cultivation. The British South Africa Company could not introduce measures to encourage African food production without threatening the preferential arrangements for white settlers, already alienated by rising labour costs generated through the army’s demands and rates of pay. Food production, therefore, fell. Increasingly, the porters had to carry their own supplies as well as those of the troops. Sir C. A. Wallace, the administrator of Northern Rhodesia, reckoned that if the porters carried their own food only 1/27 of their load from Ndola would reach its destination at Kasama. To deliver 1 ton of food a day (enough for 1,000 men) from Ndola to Kasama needed 2,250 carriers if food was available en route and 23,300 if not. The exponential effects of the lack of local produce raised these numbers to 3,000 and 71,000 respectively, if delivery was to Abercorn. At the end of 1914 about one load in twenty was actually reaching its destination. The decision to move from defence to attack increased the numbers of troops to be supplied, extended the lines of communication yet further, and took the porters away from their own territories. The major threat to Northey’s advance was less the Germans than administrative collapse, and even rebellion.248
Northey developed two palliatives to ease his supply problem. The first was to use boats to cross the swamps in Belgian territory, between Ndola and Kasama. More than half the total food delivered in the build-up to Northey’s attack came by this route, and in 1916–17 12,000 paddlers using 2,000 canoes delivered 2,500 tons. Secondly, as Northey advanced so he established bases at the head of Lake Nyasa, and was able to use the route from Beira to Chinde, thence up the Zambezi, and finally to Fort Johnston.249
By 4 June Northey had cleared the frontiers of Northern Rhodesia and Nyasaland, and had established his headquarters at Neu Langenburg. A thrust north towards Tabora and Kilimatinde would have taken Northey’s force across barren country and extended his line of communications. Therefore, as the main German forces were pushed in on the central railway, Northey inclined north-east towards Iringa. For Lettow, Northey’s attack was both a threat and an opportunity. Its impact on his plan to withdraw to the south was immediate, robbing him of the cultivated areas around Neu Langenburg and endangering Mahenge. On the other hand, Northey’s command, kept small by its supply problems, and divided into separate columns spread over a broad front, created the chance of local German successes. Lettow boosted the forces around Iringa to five companies, drawing troops from Dodoma and Dar es Salaam to do so. However, on 24 July Braunschweig, the local German commander, was defeated at Malangali, south-west of Iringa, by two British columns. The British were scarcely superior in numbers, but their converging movements and Braunschweig’s undue concern about his single but useless howitzer unhinged the German conduct of the battle from the outset.
Smuts now asked Northey to shun Iringa, and instead to aim south for Lupembe. He was worried that Northey’s column would drive the Germans off the central railway before the main attack could trap them there. But on 27 August, with the Germans south of the central railway and concentrated around Morogoro, Smuts changed tack, telling Northey to take Iringa in order to discourage Lettow from any attempt to move in that direction. The combination of Northey’s move to Iringa and of Deventer’s to Kilosa kept Kraut—after the German retreat from the line of the Wami—to the south-east and Mahenge.
Northey was increasingly worried. His small command was spread over a 160-kilometre front, between Lupembe and Iringa, and his line of communications stretched back to Mwaya. Lettow’s main forces were moving south and would be anxious to secure the harvest around Mahenge. Northey wanted Deventer to relieve his northern flank at Iringa, but the 2nd division was exhausted and in any case was required further east. Instead, Smuts urged Northey on, sensing the need to deny the Germans the use of the Mahenge area before the rains. By the end of September Northey’s Lupembe column was 96 kilometres from its base on the Ruhuje, and a further force had been detached south to Songea. Although his total strength was raised to 3,800 men, his front was now 320 kilometres long, and his supply dependent on the 17,000 reluctant carriers north of the frontier.250
Facing him, Kraut had concentrated 2,450 rifles in the Mahenge area. Kraut’s efforts to turn defence into local counter-attacks were hampered in part by the mountainous terrain and the extent of his front. The Germans had simultaneously to face west to the Ruhuje, between Lupembe and Mahenge, and north on the Ruaha. But Kraut’s problems were not eased by the offensive instincts of his superior. Lettow’s interventions, including direct communications with Kraut’s subordinates, added to the difficulties of co-ordination.251
Onto this fragile and delicately balanced scene broke Wahle’s force from Tabora. After setting a frantic pace for the first ten days in order to escape any Belgian pursuit, Wahle had found himself deep in unexplored and uninhabited terrain. Neither water nor food was readily available. A supply dump had been prepared at Malongwe, on the central railway east of Tabora. But the porters had run off, and with only 600 Indian prisoners of war as carriers Wahle had to leave most of the food, to concentrate on carrying munitions and medical supplies. After six days such food as he had was exhausted, and his men had to forage, and eve
n harvest and thresh, as well as march. The Wahehe, whose territory he entered in November, secreted their food stocks and supported the British. An initial strength of 5,000 men was whittled away: 786 askaris were captured, 146 were reported as missing, 916 deserted, and 300 were left behind. Morale, even among the Europeans, slumped. But Wahle kept his command active. He knew from intercepts that Northey had captured Iringa and that Lettow was moving south from Morogoro. He therefore determined to attack Iringa, so threatening Northey’s lines of communications and forcing the British to turn west away from Lettow. But his supply problems forced him to remain divided in three columns, and by 24 October, two days before the day appointed for the attack on Iringa, he had still not regrouped. The British, alerted by a German assault on Igominyi, were already turning back to cover Iringa. Wahle therefore abandoned the attack on Iringa, and instead pressed on towards Mahenge and Lupembe, believing he would effect a junction with Lettow. Northey’s movements in response to Wahle’s relieved Kraut, but also confused him, as he was unaware of Wahle’s approach. On 9 November Wahle made contact with a patrol sent out by Kraut. On the 26th one of Wahle’s columns was surrounded and surrendered, but the other two, a total of 750 men, four guns, and fifteen machine-guns, completed their junction with Kraut. At the end of November Wahle superseded Kraut in command of the Mahenge area. His force totalled 350 Europeans, 3,000 askaris, nine guns, and thirty-nine machine-guns, spread over a front of 450 kilometres. His task was to protect the area under cultivation and, if possible, extend it westwards to Ubena (or Nyombe).252
Thus, throughout November Northey’s force was exposed to the dangers of defeat in detail, and was engaged in a series of small fights to its rear and to its front. In December Deventer took over responsibility for Iringa. Northey’s sector now ran from Songea to Lupembe, his line of communications snaking across the Livingstone Mountains east of Lake Nyasa to Mwaya.
The Germans’ withdrawal towards the south-east corner of their colony increased the potential significance of Portugal’s entry to the war in March 1916. In practice, however, unlike Northey’s advance, Portugal’s impact remained local and limited. The prime reason was Portuguese incompetence; but in addition, British interests no more than German would be served by Portugal extending its claims across the Ruvuma river.
In August 1914 reports that Portugal had declared war on Germany and that Portuguese agitators were at work among the native population north of the Ruvuma led to clashes and casualties. Schnee established the truth in September, and apologized to the Portuguese. But in German eyes Portuguese neutrality seemed increasingly fictional.253 Lisbon allowed the passage of British troops through Mozambique and it authorized its Angolan administrations to co-operate with Pretoria in mopping up after the conquest of South-West Africa. In the autumn of 1915 the postal service via Portuguese territory was disrupted and then suppressed. More significantly, an expeditionary force of 1,527 men arrived in November 1914 from Lisbon to boost the existing garrison of twelve poorly equipped native companies. It was relieved by a second, comparably sized force the following year. The immediate task of this second force was to create a network of posts along the German frontier. But the real aim of the government was the recovery of the Kionga triangle at the mouth of the Ruvuma, seized from Portugal by Germany in 1894. Portugal continued to pursue its imperial ambitions against the background of European war, but until March 1916 contrived to do so without committing itself to the larger conflict: even after 1914 clashes in the colonies did not inevitably lead to belligerence in Europe. When at last it did formally enter the war against the Central Powers, its military actions remained confined to Africa for the time being. In April 1916 Portuguese forces occupied the Kionga triangle. The Germans retaliated by seizing sixteen Portuguese posts on the Ruvuma, largely without opposition.254
The apparent success of Smuts’s campaign gave urgency to Portugal’s ambitions. A British conquest of German East Africa without Portuguese participation would diminish its claims to territory. But, once south of the central railway, Smuts’s own ambivalence on Portuguese co-operation became evident. If the Portuguese gained by their own efforts that which Smuts wished to give them, the case for getting them to hand over to South Africa Delagoa Bay and its adjacent territories would be considerably weakened. On the other hand, a Portuguese advance on Lindi would, militarily, be a considerable contribution. The successful voyage of the Marie clearly demonstrated how important to the Germans the continued possession of the coastline might be. Therefore, when a third Portuguese expeditionary force, mustering 4,642 men, arrived at Palma on 5 July, and its commander, General Jose Cesar Ferreira Gil, proposed to move up the coast, Smuts responded by saying that the British could do that. Instead, he asked Gil to move into the German food-producing areas inland around Masasi, and pointed to Liwale as an ultimate objective. No roads existed to aid such a march.255 It was alleged that Gil himself had been appointed because of his political credentials (he was a good republican), rather than his professional aptitude. He lacked experience of Africa; his exercise of command was lackadaisical. He was reported to be far to the rear playing cards while his forward units pressed on without proper reconnaissance.256 His men showed no interest in the rudiments of tropical hygiene; they would not take quinine; they were reluctant to drink boiled water; tuberculosis, syphilis, and malaria were rampant. By the beginning of September Gil’s effective strength was 2,700 rifles. Smuts’s proposed line of advance could only reduce such a force to total ineffectiveness.
Gil crossed the Ruvuma and on 26 October occupied Newala. The local population on the Makonde plateau welcomed the Portuguese. Lettow had been so anxious to concentrate all his forces in the north that in 1915 he had appealed to Berlin in a bid to override Schnee’s wish to keep a company at Lindi.257 Looff, whose Känigsberg crew formed the kernel of the German garrison in the south-east in the autumn of 1916, agreed with Schnee. Despite having inferior numbers (a total of 840 men), Looff checked Gil’s advance and encircled him in Newala. On 28 November, after six days’ siege, Gil abandoned Newala and retired across the Ruvuma, leaving four mountain guns, six machine-guns, and 100,000 rounds. Portuguese credit among the tribes of the Makonde withered. Looff’s resolute action secured the food-producing areas of the south and made possible Lettow’s continued resistance a year later.
Smuts’s support for Gil was indirect, delayed, and, ultimately, thwarted by Lettow. Forced to pause on the Mgeta in September, he at last began to exploit his naval supremacy to secure the coastline. Kilwa was seized in September, and during October and November the whole of Hoskins’s division was transported thither. Smuts’s intention was not so much amphibious envelopment—Kilwa was too far from the Mgeta positions for such a move to have reciprocal effects. Instead, Hoskins was to advance inland on Liwale, so converging with the Portuguese and laying waste the crops on which Lettow might rely in 1917. By using a division, where originally he had intended a brigade, Smuts prevented Hoskins being ready to move until 29 November. The tracks inland were barely passable, sand alternating with black cotton soil. The rains came early and the area was thick with fly. The whole scheme proved ridiculously ambitious, making more sense on the map than it did on the ground.258
Moreover, Lettow was not fixed to the Mgeta position by offensive action. The South African artillery plotted the German trenches and lines of communication, and kept them under well-directed harassing fire. But Lettow was still free to divide his force, switching his Schwerpunkt from the river to the coast. The first British troops to move inland from Kilwa had occupied Kibata to the north-west, as a guard for their right flank. Lettow saw the opportunity to inflict a defeat on a portion of the enemy forces. Increasing his strength to ten companies, he advanced on Kibata on 5 December. The battle that developed was one of the most desperate and sustained of the campaign, fought in incessant rain between positions often only 80 yards apart, and dominated by artillery and grenades. By 21 December each side had fought the other to a s
tandstill. The British had progressively reinforced Kibata during the battle, so leading Lettow to claim the strategic victory as he had deflected the danger to Liwale. But Lettow had not overwhelmed the British post, and protracted, attritional combat made little tactical sense. By the end of the year he could muster only 1,100 Europeans and 7,300 askaris fit for service.259
Lettow’s concentration at Kibata had left only 1,000 men on the Mgeta. Smuts therefore embraced the plan which earlier he had spurned. The Kilwa force was to push north-west on Ngarambi, so pinning Lettow, while his main body crossed the Mgeta and then the Rufiji. On 1 January 1917 Smuts attacked the Mgeta positions, using his main force to envelop from the west. The British were on the Rufiji by the 3rd. But the Kilwa force had not been able to hold Lettow. He had already anticipated Smuts’s men and placed himself in a central position between the two attacks, at Lake Utungi. The British crossed the Rufiji on 17 January, taking Mkindu and Kipongo. But German resistance was now hardening. Smuts discounted the rains, of whose effects he had received ample warning. They had never really stopped in December, and now began in earnest. The Rufiji turned into a torrent, hundreds of yards across, its current too strong for any of the available boats. As the Germans had already discovered, the area between the Mgeta and the Rufiji, afforested and uninhabited, was devoid of food. It was becoming a swamp. The nearest railhead was at Mikese, 255 kilometres distant. Although Dar es Salaam was open and the central railway functioning, the supply services had still not recovered from their punishing marches earlier in the year. ‘The transport was used up; the mechanical transport broken down, and in need of thorough overhaul and reconditioning; the animal transport mostly dead, and the porters worn out and debilitated.’260 The British advance again ground to a halt.