by Hew Strachan
The third major German group, in addition to those round Kilwa and Lindi, was the twelve companies in the Mahenge area, from 9 May under the command of Theodor Tafel. Deventer’s plan was for the Belgians to advance on Mahenge from Kilosa. But the Mahenge region had been stripped bare by the Germans, and consequently the Belgians could not live off the land to the extent that they had done in the Tabora campaign. Too few porters were collected in time to allow them to take part in the first stages of the July offensive. Carriers recruited from the Belgian-occupied areas of East Africa proved useless in the pursuit of Naumann; the British therefore allowed the Belgians to requisition up to 6,000 porters in their zone of occupation to replace the 5,000 previously handed to the British by the Belgians; a further 10,000 were raised in the Congo between July and November 1917. The Belgians were finally ready to move in September, and on 9 October occupied Mahenge. Tafel withdrew to the south-east. Further supply problems delayed the Belgian pursuit for eight days. Then the rains came and the road from Kilosa to Mahenge was rendered impracticable. Two Belgian battalions, the most that could be fed, were left at Mahenge, and the rest pulled back to the central railway for redeployment to Kilwa and Lindi.283
Tafel made good his escape to the south-east. Northey’s columns, reorganized in July with bases at Lupembe, Songea, and Fort Johnston, entered Liwale on 29 October. But his force was now up to 480 kilometres from its Lake Nyasa bases and the striking powers of each component limited. Tafel broke through Northey’s screen on 15 November and made for Newala, hoping to link with Lettow.
Lettow had gone. Between April and September 1917 a detachment of Kraut’s force, 400 strong, had entered Portuguese territory, and reached as far south as Lurio and Fort Johnston. This preliminary reconnaissance suggested that the local population would be friendly and the country fertile. At a conference at Lukuledi on 24 October the German leaders debated their next step. Schnee, the defeat of his colony complete, advocated surrender. Lettow answered with an argument that drew its inspiration as much from Schnee’s own creed, that of German colonialism, as it did from the needs of war in Europe. With German territory forfeit, Germany’s claim to be an African power resided in the Schütztruppen themselves.284 They must carry on the war across the Ruvuma, in Portuguese East Africa, so maintaining German presence in Africa until the peace. Schnee agreed.
Practical as well as political considerations shaped the German decision. The area between the Rufiji and the Ruvuma was on the brink of famine. The harvest was not due until March, and in the event the rains failed in November. Lettow had sufficient food for six weeks. His stocks of quinine would last a month. He could not fight in order to capture supplies because the ammunition situation made him reliant on the 1871-pattern carbines and two-thirds of his force were equipped with the 1898-models or captured British equivalents. The artillery ammunition, so painstakingly dried and reassembled in 1915–16, had got wet again: in 1917 the fuses in particular required refabrication. He now had only enough shells for two mountain guns, and the last of the Künigsberg’s guns was destroyed. Lettow therefore shaped his force according to his resources. In July his rifle strength had been 800 Europeans and 5,500 askaris. On 25 November, when he crossed the Ruvuma, he took 300 Europeans and 1,700 askaris. One thousand fit askaris were left behind, as well as over 1,500 Europeans, mostly the sick and wounded, and women and children. At least 3,000 blacks—wives, porters, and boys—accompanied the Schütztruppen; their families’ presence was a major factor in the continuing loyalty of the askari, and only a small number responded to British appeals to surrender.285
Mahiwa, for all its self-inflicted damage to the Germans, enabled them to break contact with the British. The Kilwa force, its line of communications now 200 kilometres long, could not open a shorter connection through Lindi until the Lindi force was ready to resume its movement. Deventer planned to round up Lettow’s forces in the area of Chiwata. He asked the Portuguese to demonstrate north of the Ruvuma in the hope of blocking the Germans to the south, and so encouraging them to hold Newala. But the junction of the Kilwa and Lindi columns was not effected until 11 November. By then the major threat seemed not to be Lettow but Tafel. His command, a ration strength of 5,471, including 181 Europeans and 1,558 askaris, still had 262,000 smokeless rounds. On 20 November three out of four British columns were directed against Tafel. Uncertain of the whereabouts of the German western forces, Deventer also lost track of Lettow’s lines of march, covered by the thick bush of the Makonde plateau from aerial observation. Deventer got Tafel. However, his success was the result, not of manoeuvre nor of battle, but of a loss of nerve on the part of the German commander. Tafel crossed the Ruvuma, failed to find either food or Lettow, and then returned into German East Africa, surrendering on 28 November.
Mindful of the experiences of a year previously, Deventer did not at first follow Lettow into Portuguese territory. Fear of the rains (which in practice proved far less heavy than on the Rufiji), and the need to regroup, brought the British to a halt. Deventer issued a somewhat optimistic summons to Schnee to surrender. It was ignored. He saw the move into Portuguese territory as short-lived, reckoning that the Germans would recross the Ruvuma once the harvest was ripe in the Songea region. His immediate operational objectives were consequently defensive—to stop the Germans breaking back into German East Africa and to guard Nyasaland. He reduced his forces, sending the Nigerians home and keeping only the Gold Coast Regiment and the King’s African Rifles.
Nor was Lettow under any pressure from the Portuguese. A fourth expeditionary force, mustering 5,277 men, had arrived from Lisbon in 1917, and a further 4,509 reinforcements were dispatched to make good the losses to the 1916 expedition. But Portuguese strength in numbers and equipment was betrayed by the poverty of command and morale. Tomas de Sousa Rosa, a cavalryman in tsetse country, who had never been to Africa before, succeeded Gil in September 1917; his tenure ‘went beneath the lowest levels of insignificance’.286
Portugal’s preoccupations in 1917 were not with the Germans but with their own internal order. The Portuguese Makonde, south of the Ruvuma, had never been properly pacified. The tasks of the 1917 expeditionary force were the systematic reduction of tribal resistance in the area, and the construction of a road from the coast inland. By the time of the German invasion the first objective was almost fulfilled but the second was not; the road did not reach Chomba until 30 July 1918.
Concentration in the north weakened Portuguese presence in the south, while at the same time increasing labour demands. The Portuguese drafted their carriers, and neither paid nor fed them. In March 1917 the Makombe rebellion broke out in Portuguese Zambezia. The Portuguese themselves were defeated and besieged in Tete. The revolt derived its strength from traditional elites, who briefly overcame ethnic divisions in the rejection of Portuguese colonialism. The Portuguese held Sena, but their strategy for reconquering an area 800 kilometres broad, and embracing up to 20,000 rebels, was confused. The army and the Mozambique Company found themselves at loggerheads over how best to proceed. The British, aware that the Makombe saw their rule in Rhodesia in a favourable light, refused their allies troops, Nyasaland instead presenting 200 rifles and Southern Rhodesia somewhat belatedly contributing two obsolete machine-guns and 200 drill rifles. In the end African divisions, not imperial co-operation, determined the fate of the rebellion. Between 10,000 and 15,000 Ngoni were called in as auxiliaries, and were promised all the booty they could carry away, including women and children. Thus, by condoning terrorism and slavery, the Portuguese broke the back of the rebellion by the end of 1917.287
Neither the Makonde nor the Makombe responded in any obvious way to the possibility of German support. The north-east of the Portuguese colony remained settled throughout the German invasion, an indirect tribute to the Portuguese army’s work, and Lettow never penetrated Zambezia. But almost wherever they went the Germans were well received. The Yao in the north-west had accommodated those fugitives from the 1915 Chilembwe reb
ellion who had escaped into Portuguese East Africa, and had also welcomed the 1917 German expedition. The Lomue, south of the Lurio, and the Angoche, along the coast beneath Mozambique, proved equally hospitable. The Germans, for all that their paper money was worthless, did at least pay for their goods rather than seize them. German doctors attended to the sick. But Lettow did not take the opportunity to turn opposition to Portuguese rule into revolution. The Germans neutralized the African population rather than armed it. Thus, even in 1917–18, and even in Lettow’s own hands, the anxiety to buttress collective European colonial rule outweighed immediate military advantage. Lettow still rejected a true strategy of revolutionary warfare.
Nonetheless, with his entry into Portuguese territory Lettow’s actual style of operations for the first time conformed to that of a guerrilla leader. His supply position had forced him to reduce his fighting strength. Thus he had to quell his predilection for the offensive. He fought to feed, and to feed he had to keep moving to fresh sources of supply. Mobility, not striking power, was henceforth his major asset. When he crossed the Ruvuma his column was 30 kilometres long, the main body separated by one day’s march from its advanced guard and two days’ from its rear. Confined to jungle paths, frequently crossing precipitous terrain, the Germans trudged in single file. In such circumstances rapid concentration for battle was impossible. Lettow reorganized his forces into three columns, his own, Wahle’s, and Kohl’s, each with its own supply train and field hospital. By following parallel routes the columns overcame the worst dangers of dispersion. The Germans marched for six hours a day, with a half-hour halt every two hours, and aimed to cover 24 to 32 kilometres a day. It was a considerable achievement. The porters frequently carried additional loads weighing up to 30 kilos; the askaris bore iron rations for fifteen days as well as their rifles and ammunition; their wives on occasion gave birth on the line of march and within hours had rejoined the column.288
The greatest potential impediment to mobility, the bush apart, was ill-health. In this respect Lettow’s force began its trek with several advantages: only the fittest were selected, thirteen doctors were among them, and the small European complement was adjusted to the available supply of quinine. Plunder made good many deficits in medical supplies. But the Portuguese had done little to eradicate disease within their colony. Locally recruited porters and prolonged residence in native settlements introduced new sicknesses. Smallpox appeared in February and July 1918. In August 1918 pneumonia (not, the German doctors were sure, influenza) struck 250 of the force and killed at least twenty-two. By then only eighty sick could be carried. Periodically they would be collected into a hospital and left, together with a doctor, for the British. By the end of the war Lettow had only six doctors remaining.289
The area between the Ruvuma and the Lujenda did not prove as rich as Lettow had hoped. Game formed much of the diet, but the thick, tall vegetation made stalking and shooting difficult. However, now the Germans, effectively for the first time since Tanga, could plunder. The Portuguese frontier forts along the Ruvuma provided arms, ammunition, and European food. At Ngomano, on 25 November 1917, the Germans surprised 1,200 Portuguese troops and captured 600 rifles and 250,000 rounds. Three more forts were taken in December, and the Schütztruppen thus re-equipped themselves with Portuguese rifles and almost a million rounds.
Lettow commenced his march south before the rains ceased, so as to maintain his lead over Deventer. Two British battalions from Fort Johnston began to advance on Mwembe in January 1918, and Lettow concentrated around Nanungu at the end of February. As the rains eased the Germans were able to rig up a wireless. In late March they heard the news of the German victories in France, and of the imminent capture of Amiens. Their purpose in maintaining a German presence in Africa reaffirmed, Lettow briefly flirted once again with offensive options. His central position seemed to give him the chance to strike enemy forces in isolation, particularly those with longer lines of communication coming from the west. In late April and early May Lettow placed Kohl at Medo to guard him from the east, while he concentrated five companies for a blow to the west. The action at Kireka mountain on 5 May cost him 27 per cent of those engaged.
Deventer’s concern was still to stop the Germans going north. He therefore planned to create a line of posts from Port Amelia to Fort Johnston, via Medo and Mahua. The conception was ridiculous: he never possessed the resources to create an impermeable barrier 560 kilometres long. His main base was still at Dar es Salaam. Therefore goods from Britain proceeded via the Cape and Dar, before being transshipped and routed south again to Port Amelia. The conviction that Lettow was about to be defeated had not, despite all the evidence to the contrary, dissipated. Thus, when Port Amelia and later Mozambique were established as intermediate bases, provision was made for 12,000 men to be fed up to 320 kilometres from the coast. Ultimately 33,000 men were dependent on the two ports.290
Deventer’s other major obstacle was the Portuguese. The presence of Portuguese troops did little more than create supply dumps from which Lettow could replenish his food and ammunition needs. They also antagonized the local population, making it increasingly hard for the British to recruit porters. The British paid the hut tax of those Africans who enlisted as carriers. The effect, however, was to antagonize the local administrators, who were in the habit of appropriating a percentage of the tax for themselves. Ultimately the King’s African Rifles lived largely off the land, an expedient which slowed their pursuit as they foraged over areas through which the Germans had already passed. The campaign was fought on Portuguese territory but increasingly without Portuguese participation. In July 1918 Sousa Rosa was recalled to Lisbon and arrested.
On 22 May Lettow got his major battle. Kohl’s column, now effectively his rearguard, was nearly trapped by the British forces from Port Amelia at Korewa, near Maketi. He managed to extricate himself, but at the price of losing all his supplies, including 70,000 rounds, 30,000 rupees, and all Schnee’s official documents; Schnee himself was lucky to escape. The British completed their junction from east and west, and the Germans marched south.
On 16 June Lettow captured a Portuguese map showing the area as far as Lugella and Quelimane. Anxious to capitalize on what little intelligence he could garner, he pressed on, aware that the Lugella Company’s base might provide rich pickings. He found them at Namakura on 1–3 July. The defences, 3,000 yards in extent, proved too great for the combined Portuguese-British garrison of 1,030 men to hold. Neither ally fought well; in their efforts to escape along the Namakura river many were drowned or eaten by crocodiles; the Germans lost nine men, the British and Portuguese 200 dead and 543 captured. Far more importantly, the booty included ten machine-guns, 350 rifles, 350 tonnes of food, and large quantities of ammunition. The Germans were almost entirely re-equipped with modern British and Portuguese rifles, and had a stock of 813,800 rounds. While the pursuing forces pressed on to Quelimane, anticipating that the next German move would be to strike the harbour there and then go up the Zambezi, the Germans rested at Namakura, drinking the abundant quantities of schnapps which they had looted, and then doubled back to the north-east.
Across the River Namirrue Lettow attacked an isolated British force on the night of 22/3 July, and then captured Namirrue itself. On 28 July he paused at Chalaua, recruiting and training 310 porters as askaris. A captured letter alerted Lettow to the next British move and on 8 August he quitted Chalaua. His direction at first was north-east, but then, having deceived his opponent, he switched to due west, reaching Numarroe on 24 August. The British hoped that Lettow would attack Regone, but their intention of enveloping him while he did so was known to Lettow. He marched north to Lioma. Here he was hard hit by three battalions of the King’s African Rifles, losing 48,000 rounds and large quantities of stores. His total casualties between 27 August and 6 September included thirty-nine Europeans, 184 askaris, and 317 porters (242 were reported as missing). His intention now was to aim through the Livingstone Mountains, around the northern
end of Lake Nyasa, and then turn west. He calculated that Deventer would rein in his pursuit and reconcentrate on the central railway to cover Tabora. He was only partly right. Intelligence acquired on 22 September suggested that there were no troops on the Ruvuma, but that there were major concentrations in Nyasaland. Morale was slipping. Lettow’s refrain, that their efforts were tying down 30,000 enemy troops, was less persuasive as the Schütztruppen’s strength dwindled, their supply and health problems multiplying. On 28 September the Germans recrossed the Ruvuma; they again speculated about a push to the north, aiming to get beyond the Ruaha before mid-December and the advent of the rains.
The projects discussed on 29 September bordered on the fantastical. Some favoured the northern thrust, even as far as Abyssinia, in the hope that it might be pro-German. Others suggested taking ship for Afghanistan (presumably via Persia). But opinion veered once again towards the west, and a march to Angola.291 The morale of the askaris recovered as they regained their own territory. At Songea, the local population welcomed and resupplied them. At Ubena on 18 October Lettow found papers alerting him to the situation in Europe; thereafter, a number of reports confirmed that Germany was seeking an armistice. The British prepared their defences at Mahenge, Iringa, and Tabora, and, forewarned by the example of Wintgens’s raid to the north, anticipated a dash across the central railway. But round Dodoma the requisitioning of grain and livestock in 1915–16 had driven those of the population not taken as porters into the forests and mountains. Cultivation had declined, and the lateness of the rains in 1918–19 meant that shortages turned into famine.292 Lettow struck west towards Fife, not north. His intention was to raid the depots on the Kasama-Fife road, working along Northey’s line of communications, and then push west between lakes Bangweulu and Mweru.