To Arms

by Hew Strachan

  This time Smuts would not be able to renew it. On 27 December 1916 he heard that he was to go to London, to represent South Africa at the Imperial War Cabinet. He left on 20 January. He had continuously anticipated a short campaign; his cables had consistently reported great successes; he had avoided mentioning the effects of the weather; now he described the war in East Africa as all but finished. At one level he was right. South African war aims were territorial. Four-fifths of the colony and nine-tenths of its infrastructure had been overrun. This was conquest, even if it was not ultimate victory. Smuts had served the cause of imperialism rather than that of the Entente’s war effort. In doing so he had behaved as a politician rather than a general, boosting his own reputation and aiming to hallow the integration of the Union with battlefield triumph.

  Where propaganda turned to fiction was in the depiction of a great South African feat of arms. On their arrival both Smuts and his fellow countrymen had dubbed the German askaris ‘damned Kaffirs’.261 But they had learnt that blacks could outfight and outwit whites. Privately their respect had grown; publicly they could not admit it. The deadliness of the East African climate was exaggerated rather than acknowledge the effectiveness of the Schütztruppen. ’Hospitals’, Meinertzhagen commented on 8 October, ‘are full to overflowing with strong healthy men suffering from cold feet or an excess of patriotism.’262 South African medical officers colluded. Invalids were sent to recuperate in South Africa, and then did not return. When convalescent camps were established in Kenya, wastage rates were cut by about half.263 Carried to its logical conclusion, the South African arguments about health had to acknowledge that the British, like the Germans, should rely on black, not white, troops. In October 1916 Smuts himself eventually concluded as much. But, by saying that after January 1917 only policing duties remained, he simultaneously protected the amour propre of the white South Africans.264

  The Africanization of the campaign began with the completion of the conquest of the Cameroons. In July 1916 the Gold Coast Regiment arrived, and it bore the brunt of the fighting at Kibata. But Lugard at first opposed the use of the major component of the West African Frontier Force, the four battalions of the Nigeria Regiment. Their losses in the Cameroons had been greater than those of the Gold Coast Regiment. About 1,000 men had been discharged when the campaign ended, and the remainder were anticipating a period with their families. Lugard asked the War Office for forty-two senior officers, but was told that, in view of the demands of the western front, he could have only fifteen subalterns. The advice that he enlist from the local European population, although eventually followed, filled him with apprehension for the internal security of the colony. Over a third of his pre-war administrators were absent on army duties. A rebellion among the Oujo in November 1916, prompted by the chiefs’ methods of recruiting porters, lent credence to his fears. Nonetheless, the Nigerians sailed in the same month, their establishment of 5,000 filled by voluntary re-enlistment. Not until 1917 would an indirect form of conscription be necessary. They entered the line on the Mgeta, and occupied the Rufiji valley during the rains.265

  In April 1916 Hoskins had prevailed on Smuts to sanction a moderate increase in the King’s African Rifles. Four new battalions were raised, doubling the strength to 8,000 men. Smuts remained sceptical, bound by the idea that most tribes of the British East African possessions were not warlike, and that therefore the recruiting capacity of the colonies was restricted. Nonetheless, by January 1917 the King’s African Rifles had thirteen battalions, and in February a target of twenty was set. By November 1918 the establishment was twenty-two battalions, and its total strength 35,424 all ranks.266 One regiment, the 6th, raised its three battalions from former German askaris; elsewhere, recently pacified tribesmen enlisted in order to recover their traditional vocations. But generally the notion of ‘martial races’, imported from the Indian army, was abandoned, and the recruiting base accordingly broadened.267

  The expanded King’s African Rifles was designed ultimately for imperial, not simply East African, service. Smuts, after all, was predicting a rapid end to the campaign, and the conventional wisdom was that the formation and training of each new battalion would take a year. Sensing a fresh source of manpower for other theatres, and thinking particularly of Palestine, which took priority over East Africa from February 1917, the War Office took a more benevolent interest than might otherwise have been the case. In June 1916 the battalions adopted the four-company structure, thus aligning themselves with European norms. A year later the allocation of machine-guns for each battalion was increased to four. Smuts had argued that the efficiency of the Schütztruppen was due to their high proportion of European NCOs; the Germans agreed. From a ratio of one white for every 35.5 blacks in January, the King’s African Rifles moved to one for 9.25 by the end of the war. However, not all the experienced campaigners shipped out from Britain proved valuable, some being selected because shell-shock had made them unfit for further service in France.268 In the event, the King’s African Rifles’ only other service than in German East Africa was on the northern frontiers of Kenya and Uganda, against Somalis, Swazis, and Abyssinians. Their contribution to the major theatres of war was therefore indirect. Henceforth, Lettow was not tying down white troops that could be deployed in France and Flanders. A major strategic rationale for his campaign had been eroded.

  The Africanization of the British effort in East Africa did not resolve its logistical problems. On 28 November 1916 715 of 980 other ranks in the Gold Coast Regiment were on the sick list; ironically, the British officers, with nineteen out of thirty-six unfit, were rather more robust.269 The supply arrangements reached their nadir on the Rufiji in February 1917. Half rations were ordered. The daily food allocation averaged 17 ounces, predominantly of maize. Carriers ate roots and berries, and consequently died of alkaloidal poisoning.270 Labour was at a premium: 12,000 porters were needed for every 3,000 soldiers, and the British forces on the Rufiji required 135,000, and eventually 175,000.271 A. R. Hoskins, Smuts’s successor, had taken over an army robbed of offensive capacity. Only his West African units could be deemed reliable; it would be some months before the new battalions of the King’s African Rifles were sufficiently trained to face the Schütztruppen.

  Lettow’s problems were worse. The area into which he had now retreated was the focus of the Maji-Maji rebellion: to the south the Makonde had welcomed the Portuguese, to the west some of the Wangeni in Songea had joined Northey. The rains had failed in 1913/14 and 1914/15, and famine ensued. Two armies had entered a region already made destitute. Lettow had reckoned on having 450,000 kilos of corn stored between the Rufiji and the Mbemkuru. On 26 January 1917 he learned he had only 350,000. There were a further 150,000 kilos in the Lindi area, but to bring them north would have required porters, and his ration strength was already a quarter above what he had anticipated. Lettow dispensed with all useless mouths, handing over to the British those not fit to fight. Each European was reduced to five porters, each company to 150, and those on the lines of communication were sent home: a total of 8,000 carriers were thus discharged. Rations, which might otherwise have been cut to a quarter, were set at a third. The askari (including his wife or boy) got between 600 and 700 grammes of meal a day, the Europeans about 6 kilos of food for twenty-eight days. Maize was eaten before it ripened. Sickness rates reached 80 per cent among the Europeans, and dysentery ravaged the blacks.272

  Lettow’s objective was to hold the ground south of the Rufiji until the harvest ripened in March and April. Even sowing the crops was complicated, given the loss of labour and the lack of pre-existing German administrative control in the area. But the rains were heavier and more prolonged than for some years. The harvest was good, although the water flooded the fields and impeded its distribution.273

  On the Rufiji, Hoskins’s paralysis saved Lettow. To the north-west, Deventer’s line of communications from Iringa to Dodoma was sundered by the flooded Ruaha, which reached a width of 26 kilometres. His push on Lukeg
ata was reduced to three battalions and a squadron, and then halted entirely. Only Northey remained active. His forces were healthier, not so much because half of them were blacks but because they were operating on the higher ground away from the coast. In January a captured German message revealed that the detachments under Kraut and Wintgens were dependent on supplies in the area north of Songea. Northey’s columns began to converge on Iringa from Lupembe and Songea.

  Wahle’s main concentration was now to Northey’s north, threatening Hoskins’s right flank while converging on Lettow. The food available to him was not sufficient to support Kraut’s and Wintgens’s men, and on 29 January he ordered them to move south and north, to feed off the enemy lines of communication. Kraut entered the area between the Ruvuma and Lake Nyasa, marching first south-east and then north-west, before finally turning east along the Ruvuma to Tunduru.

  Kraut told Wintgens to follow him. Wintgens refused and took his 524 men north-west towards Tundala, and thence along the Northern Rhodesian border towards Tabora. Wintgens fed off the land, causing mayhem on the supply routes stretched across German East Africa. The pursuit, conducted initially by Northey’s Lupembe column, traversed a land already laid bare.

  As the Germans approached Bismarckburg early in March, the British asked the Belgians if they could use the western end of the central railway, within the Belgian area of occupation; they also requested permission to recruit. Two months previously, anxious to limit Belgium’s gains, Britain had told it that its assistance was no longer required. Only 2,000 of the Force publique remained in German East Africa, others had been demobilized, many of the whites had returned to Europe, and the porters had dispersed or entered British service. By handing Tabora over to the British the Belgians had left the protection of Urundi and Ruanda against any fresh German offensive to their allies. Nonetheless, they preferred to offer the British troops rather than indirect support. At the beginning of April Hoskins and Huyghé, Tombeur’s successor, agreed on a Belgian contribution of 6,600 askaris, 600 Europeans, and 18,000 porters. In reality, only 456 Europeans were available, and therefore junior officers occupied senior posts; 5,000 of the 18,000 porters were already in British service and could not be released. The Hoskins-Huyghé plan, to deploy 4,000 rifles in Wintgens’s path, was unrealizable.274 Without porters, the Belgians could not move with sufficient speed.

  On 1 May the British element in the pursuit was changed. Northey’s column fell back south, and a new force of 1,700 men, including a King’s African Rifles battalion still under training, took up the running. The Germans were approaching Tabora. On 21 May Wintgens himself, sick with typhus, surrendered. Heinrich Naumann, his successor, planned to rejoin Wahle by marching south-east, but finding his path barred he moved north-east, across the central railway, towards Mwanza and Lake Victoria. Responsibility for this area had been passed over to the Colonial Office, and friction between the army and the civilian administration now added to the problems of British and Belgian co-ordination. The lack of a united command meant that the pursuit was devoid of consistency or purpose. Early in July Naumann, now operating in the north around Ikoma, flirted with the idea of a raid on Nairobi. But instead he decided to thrust south in an effort to reunite his troops with Lettow’s. By August he realized that he could never get through, and divided his command into three sections, each to go in divergent directions with the aim of drawing as many British troops from the main theatre as possible. The section sent to the south-east surrendered on 2 September. But the two northerly sections held out for a further month, Naumann himself raiding Kahe on 29 August. Naumann remained a reluctant guerrilla. At the end his intention, once again, was to link with Lettow, rather than maximize the effects of dispersion. Nonetheless, he had conducted a classic guerrilla operation.275 His men had marched almost 3,200 kilometres since February; they had found a population that was passively, if not actively, supportive; they had drawn the attention of up to 6,000 men away from the main battle.

  Wintgens’s and Naumann’s marches served to deepen London’s frustration with Hoskins’s failure to complete a campaign already pronounced victorious by Smuts. Hoskins, however, was the first, and last, British commander-in-chief in East Africa who appreciated the constraints under which he was operating. His knowledge of local conditions, his awareness of the training needs of the African troops, led him to avoid the hyperbole of Smutsian advances. From February to May the British forces in East Africa were rebuilt. Their demand for porters—they needed 160,000, and a further 15,000 a month to cover wastage—so exceeded supply (Hoskins had about 40,000 when he took over) that lorries became an increasingly vital component. Hoskins reckoned that one lorry was equivalent to thirty porters, and wanted 400 of them. But lorries were unusable in the rains, and therefore necessitated a pause in operations until May.276

  Hoskins’s demands embarrassed both Smuts and his new colleagues in the War Cabinet. Warnings against the presumption of speedy or easy victory chimed ill with the South African’s claim that the campaign was effectively over. If the spoils of East Africa were to go to Pretoria, Smuts had to argue that his South Africans had done the job. If they had not, then the credibility of white South African citizen soldiers would be forfeit to the subsequent achievements of blacks. ‘Military training of the native’ in Central Africa would thereby be endorsed, and, he warned in May 1917, would eventually present ‘a danger to civilization’.277 Personal pique coincided with the territorial needs and the racial policies of the Union.

  Hoskins recognized that Lettow might well cross the Ruvuma into Portuguese East Africa. His plan, therefore, rested not on continued pushes southward to clear territory, but on a concerted effort to trap the Germans. The British agreement with the Belgians included provision for a column to join Deventer’s forces south of Dodoma: their task, once they had dealt with Naumann, was to advance on Mahenge. The Nigerians on the Rufiji would also strike towards Mahenge. Liwale would be approached from the west by Northey and from the east by the 1st division at Kilwa. The brigade at Lindi, which had been occupied on 16 September 1916, was to break free of Looff’s attentions and aim for Masasi, so cutting German communications to the south.

  The chances of trapping Lettow seemed good. The main German forces were deployed along the coast rather than inland, presumably in the hope that a third supply ship might break the blockade. In April Lettow concentrated south of Kilwa. On 3 June Wahle, his junction with Lettow complete, was given overall authority around Lindi. But on 29 May, before the British offensive could get under way, Hoskins was relieved of his command. Smuts’s lobbying had convinced Sir William Robertson, the chief of the imperial general staff, that Hoskins was losing his grip, and that his successor should be ‘Jaap’ van Deventer.278 Once again a British regular had been replaced by an Afrikaner amateur. Deventer’s instructions were to end the campaign as soon as possible; with merchant tonnage losses soaring in the Atlantic, the objective was to save shipping. He was given no territorial or operational objectives.

  Lettow’s strengths against the British converging movements remained interior lines and deployment in depth. For neither side was the apparent front line, the Rufiji, of major concern. The key battles of the second half of 1917 were fought around Kilwa and Lindi. The foundation of Lettow’s strategy was the containment of the British attempt to break out from Lindi, thus keeping open the Germans’ route to the south. In June Lettow reinforced Wahle at the expense of the Kilwa front.

  Like Hoskins, Deventer was aware that Lettow intended ultimately to move into Portuguese territory. He therefore remained loyal to Hoskins’s plan, consolidating his position at Lindi so as to cut off the Germans’ line of retreat. But Kilwa had the better harbour, and his main initial effort came from there. On 19 July the three converging British columns from Kilwa were fought to a standstill by a German force of 945 men at Narungombe. The Germans, reduced to five rounds of ammunition per man and unaware that Lettow was marching north to reinforce them the next day, fell bac
k to Nahungu, on the Mbemkuru.279 But Deventer’s progress on the Kilwa front was halted until September. On the 19th of that month, his communications extended to enable the next advance, Deventer’s Kilwa force moved on Nahungu. In the next eighteen days the Germans counted thirty-seven separate engagements, many of them battles for the control of water supplies. The thick bush impeded not only the British aerial reconnaissance but also the Germans’ co-ordination on the ground. The British were again held on 27 September, but the Germans fell back once more. They were running low on smokeless ammunition, and by 1 October were having to rely on the 1871-model carbine, whose bullets used black powder, so providing targets for British guns.280 Deventer now decided to put his weight on the Lindi sector. He ordered the Nigerian brigade to detach itself from the Kilwa force and march on Nyangoa, so converging with the Lindi force attacking from the east.

  However, British intelligence had lost track of Lettow himself. The German commander moved between fronts, his attention during the August lull increasingly drawn south to Masasi and Tunduru, one of Northey’s columns having begun operations along the Portuguese frontier. Whether the Nigerian brigade’s task was to envelope Wahle as he faced the Lindi force, or to prevent a junction between Wahle and Lettow is therefore not entirely clear. Lettow himself saw the opportunity to strike a decisive blow, using his interior lines to effect a concentration in a way that had eluded him at Narungombe and Nahungu. The Nigerians’ eight-day march, dogged by lack of water, halved their effective strength to 1,000 men. Rather than envelop the Germans, they themselves were encircled. Lettow had concentrated a total of eighteen out of twenty-five available companies. In a fierce four-day battle at Mahiwa, beginning on 15 October and fought at close quarters, ground was won and lost up to six times. The Lindi force found itself endeavouring to break through Wahle and so extricate the Nigerians from Lettow’s clutches. British losses totalled 2,700 out of 4,900 engaged. But German casualties, though ostensibly light (about 600), were relatively more serious. By the second day the number of wounded exceeded the number of porters to carry them, and men with three or four injuries continued to fight. Wahle’s command lost nearly 30 per cent of its combat strength, and two field companies were disbanded. Moreover, all their smokeless ammunition (500,000 rounds) was expended, machine-guns had to be destroyed, and only twenty-five rounds remained for each of the older-pattern rifles.281 Mahiwa was the first sustained battle of the entire campaign. It confirmed that the avoidance of combat had been the strategy not of Lettow but of Smuts; it also demonstrated that such a strategy had served Germany’s interests rather better than Britain’s.282