By Spring 1917 Britain was running out of money to fund the war and the unrestricted U-boat campaign was increasingly successful in throttling its seaborne trade. In the first four months of 1917, British losses to submarines had risen from 109,954 tons (thirty-five ships) in January to 516,394 tons (155 ships) in April. At this rate, the country would soon be starved into submission, as replacement ships could not be built fast enough to offset the losses. Given the gravity of the situation, it was decided to concentrate shipping on the North Atlantic routes at the cost of reductions in Indian and Australian waters. At the same time, the Russian military defeat and the mutinies in the French Army meant that Britain would have to carry the strain of military operations in Europe, leaving little for peripheral theatres. At home, declining civilian morale was becoming a major problem and Lloyd George continued to look for a striking victory that would sustain faith in the war effort. With all the important parts of German East Africa in Allied hands, there seemed little prospect of further achievements in that theatre, while the continued existence of the Schutztruppe meant that the campaign had to carry on. The War Cabinet now appreciated that Lettow had to be defeated militarily and as quickly as possible.
On his arrival in East Africa on 29 May, the new commander-in-chief was given his instructions. The CIGS came immediately to the point:
In view of requirements other theatres and of the fact that it is essential to release at earliest possible moment vessels absorbed by supply and maintenance of your force, His Majesty’s Government attach great importance to early termination of campaign. I must also impress upon you importance during remainder of campaign of limiting demands for tonnage to minimum.
Van Deventer was instructed to try to prevent the enemy from leaving German East Africa and entering Portuguese East Africa. He was authorised to operate in the latter territory if necessary, although the Portuguese themselves were expected to give little useful assistance. In contrast, considerable hopes were placed on the Belgians, who would provide a column in the continued pursuit of Abt Naumann as well as building up their main striking force in the Iringa-Kidatu area. There was no explicit mention of any other campaign objectives, but it was clear that they implied the destruction of the enemy force and its ability to fight.
The new commander-in-chief was very different to his predecessors. Now a lieutenant general, van Deventer had been a loyal member of the Botha/Smuts axis since the second Anglo-Boer war. He had begun his military career as a gunner in the Transvaal Artillery in 1896, rising rapidly through the ranks, reaching battery commander by the outbreak of war in 1899. He fought with distinction throughout the conflict and showed himself to be an able independent commander. Promoted to combat general in 1902, he served as Smuts’ second-in-command in the Cape Colony. He was recalled to service as a colonel in 1914 to participate in the South-West Africa expedition, but helped to suppress the Afrikaner rebellion first. Having completed that mission, he was promoted brigadier in general command of the Upington column in the advance against German South-West Africa, finishing as a divisional commander.
After a few months at home, he followed Smuts to East Africa where he rapidly rose from brigade to divisional commander. Van Deventer returned with the bulk of the South African troops in January 1917 and resumed farming until being asked to assume the chief command in May. If the new commander-in-chief was a Smuts loyalist, he was very different in outlook and character. He was renowned for his sense of discipline, a quality often lacking in Boer formations, and did not shrink from battle. Although he was most at home leading mounted columns, he realised that the infantry would be the mainstay of the campaign. This did not preclude the use of mounted troops, but only in areas where the tsetse fly was not present.
He had learned from bitter experience the futility of trying to outmanoeuvre the Germans in such difficult terrain and grasped that they had to be defeated in battle. Equally important was his understanding that operational plans had to be based on sound planning and administration. This meant that his place was at GHQ, where he could plan and direct operations without becoming drawn into the details of the tactical battle. It also ensured that the various staff branches could co-ordinate their work much more effectively. Finally, he kept on the talented and highly professional Brigadier General S. H. Sheppard as his Chief of General Staff. Together, they were to form a close-knit team.
Both men understood that the Schutztruppe had two major vulnerabilities; its dependence on European officers and NCOs, all of whom were irreplaceable, and its need to live off the land. This led to a policy of hard-hitting wherever possible, with the aim of inflicting the maximum casualties while lowering morale and the will to continue. While the firepower of the machine gun and German mobility meant that the attacker would generally suffer higher casualties than the defender, this was not a decisive disadvantage. For restricted as British manpower and equipment reserves were, they were still vastly superior to their opponents and could be replaced, albeit with difficulty. But this aggressive approach would be complemented by a campaign against the enemy lines of communication, and in particular eliminating their food-growing areas.
Nevertheless, van Deventer appreciated the formidable difficulties facing him. Although Hoskins had done much to reorganise and re-equip the force, the exertions and overwork of the preceding year had taken a very heavy toll. Furthermore, the Germans still retained interior lines in a territory some 480 km square and difficult to penetrate. For example, Mahenge to Liwale was seven days’ march for the Germans while for the British to have moved a force from Mahenge to either Kilwa or Lindi, still some distance from Liwale, would have taken the same number of weeks. None of this was eased by the lack of accurate maps and lack of roads or railways in the area of operations. Equally, the problems of disease remained enormous and could only be minimised through careful planning and strict attention to health.
PLANS FOR THE GENERAL OFFENSIVE OF 1917
The general offensive for 1917 could not begin before the end of the rains and the subsequent drying out of the countryside, which would not occur until late June at the earliest. Van Deventer had very little time to settle into his new command, but Hoskins had set the necessary planning, reorganisation and regrouping in train. Five possible lines of advance had been identified. The first was the Dodoma-Iringa-Mahenge route in the west. While it was the healthiest, it was also the longest and passed through the Ulanga and Great Ruaha valleys which were unsuitable for motor transport and highly vulnerable to flooding in the wet season. The second option was slightly further east along the line Kilossa-Kidatu-Mahenge. It was shorter and more practical for vehicles than the first approach, but it too suffered from its vulnerability to the rains. The third approach followed Smuts’ route across the middle Rufiji going via Mikesse-Kibambawe-Mahenge. Plunging as it did through the Mgeta and Rufiji valleys, it was by far the most physically difficult to support. The nightmarish conditions of the last rainy season were still all too evident, while the country south of Kibambawe was a wild tangle of bush and almost totally impracticable for motors. The fourth option was an advance on the line Kilwa-Liwale. It had the unhappy reputation of having the unhealthiest climate of all, although it did offer the advantages of being close to a large proportion of the enemy’s forces and to a first-class harbour. The final choice was the approach Lindi-Massassi. Its chief advantage was that it offered a direct line of advance that could potentially cut off any retreat towards the Portuguese border, although the country was very difficult. Health, too, was a problem, while Lindi harbour had considerable limitations and could only be used by a proportion of the available shipping.
The attractiveness of operations launched across the river basins, which had never been high, now waned significantly. Apart from the unhealthiness of the climate, the difficulty of the countryside made supply very difficult and there was insufficient transport – either human or vehicular – to support concurrent major advances from the interior and the coast. As well, Lettow’s gradual redeployment to the south near Kilwa made the coastal operations more attractive as they were easier to supply and reinforce.
General Northey considered that the Germans would only move into Portuguese East Africa if pushed there, and, until the situation east of Mahenge was cleared up, there was the potential for another raiding force to break westward. He proposed converging simultaneous thrusts against the enemy, but, conscious of the time needed to get Murray’s column back and suitably rested, he expected that his advance would start after the others. Van Deventer agreed, realising the importance of having Northey attack Mpepo and Mtarika as it would draw off enemy forces on the Ruipa line prior to the arrival of the Belgians. Once they were ready, Northey could then shorten his line by handing over the Ruhuje sector to them. By that stage, Murray’s column would be rested and ready to participate in the general offensive.
This would leave the Songea column at about 1,000 rifles with sixteen machine guns and two mountain guns, and the Fort Johnston column with a similar amount. While these columns would be ready to start at the end of June, Murray’s column would need extra time to reach Songea and would not be ready to advance before 7 July. This would provide another 1,000 rifles, giving him over 3,000 men advancing on two axes. He also noted that two battalions’ worth of reinforcements would be arriving in July, but neither would be ready for the field before August. Likuju would also form an important staging base, as both a wireless station and small airfield were established there.
The Portuguese were given a small role to play in these plans. Co-operation and the exchange of information had been greatly improved by the appointment of the former consul general, Errol MacDonell, as military liaison officer to the Portuguese Headquarters in mid-March 1917. This was reciprocated several weeks later with a Portuguese officer, Major Azumbuju Martins, being attached to the British GHQ. He had East African experience, having been chief of staff to the ill-fated General Gil. Hitherto, neither side had any detailed information on each other’s forces, intentions or capabilities and Major MacDonell’s first task was to ascertain where the Portuguese troops were located and whether they intended to defend the Rovuma River boundary.
However, MacDonell had spent many years in Moçambique and knew that his job would not be an easy one in light of Portuguese sensitivities and pride:
A very large number of Portuguese Officers and men are fully alive to the fact that they have up to the present made a hopeless fiasco of the German East campaign, and though this has been stated to me in private, if any foreigner were to make a similar statement he would incur the obloquy of the Military, Press and public…I shall be looked upon with suspicion by the military authorities.
The change of command also changed some anxiety with the Portuguese, as they had heard rumours that van Deventer was taking over from Hoskins. They had enjoyed working with Hoskins and feared that the South Africans harboured anti-Portuguese sentiments. Furthermore, they had a major task in the suppression of the Barue rebellion and were not pleased by the lack of British support.
On 10 June 1917, van Deventer signalled his plan back to London. It reflected the changed strategic situation as well as the imperatives given in his own instructions during the previous month. It was as follows: first he wanted to catch and destroy Abt Naumann as quickly as possible in order to rest and redeploy the forces involved; secondly a preliminary and limited assault would be launched at Lindi to secure the high ground surrounding the town and to secure a better water supply. At the same time, he would attack the Ruipa River position from Iringa in order to deny its rich food-growing area to the Germans. Finally, he would push southward from Kilwa as soon as sufficient reinforcements could be brought up. Concurrently, Northey was to concentrate his force at Songea, having got Colonel Murray’s column back, and then to advance eastwards as soon as he was ready. The commander-in-chief understood that Lettow would attempt to avoid encirclement and escape south:
The establishment of considerable food depots in the MASSASSI (GGF. F. 8.c.) area, together with the presence of a considerable enemy force in PORTUGUESE EAST AFRICA, appears to point to his eventual retirement by this line. One of my chief aims must therefore to be to prevent his main force breaking through into PORTUGUESE EAST AFRICA, as that might necessitate a new campaign.
Van Deventer and Huyghé met at the Belgian Grand Quartier Général (GQG) at Dodoma on 18 June to agree final details. However, before the main operational objectives could be reached, there were several preliminary operations that had to be completed beforehand. The most important of which was the clearance of Tafel’s forward troops from the Ulanga Valley between those two settlements between Mpepo and Malinje which would also cover the move forward of the Belgian main body from Kilossa. Anxious to get going, van Deventer had already asked Huyghé to provide a column of 1,200 rifles by mid-June, but this could not be readied until the end of July owing to lack of transport. However, Huyghé showed some flexibility over British concerns that another German raiding force might break out to the west. He agreed to send 500 rifles to fill the gap around Mpanga, despite breaking the agreed principles of a concentrated Belgian deployment.
For the main push, the Belgians were to drive Abt Tafel out of the Mahenge area with the Brigade Sud establishing itself between Colonel Tytler’s column at Iringa and Norforce at Lupembe. The Brigade Nord would assume sole responsibility for the pursuit and destruction of Abt Naumann with all British columns being withdrawn into reserve. Two of its battalions would continue the chase while the remainder continued their concentration on the railway. Once the brigade came into action in late July it would advance on Mahenge while the detached troops to the north would take over more of the British sector as soon as Naumann had been dealt with. Effectively, this would turn the whole of the Mahenge operations over to the Belgians. Lastly, in the vital matter of supplies, it was agreed that the British would provide carriers and vehicles to bring Belgian provisions to their supply base at Dodoma. Equally, the Belgian pioneers would be responsible for road and bridge building in the forward areas.
General van Deventer issued his final timetable for the offensive on 27 June. A preliminary operation in Portuguese East Africa would begin on the same day with Norforce sending a column against Mwembe. The main phase would commence on 2 July, with the Iringa Column attacking the Ruipa position supported by 250 rifles from Norforce moving from Lupembe against Mpepo. The principal attack would follow on 4 July when Linforce moved on Mtua, with an assault planned there on the following morning. Also on 5 July, Hanforce would assault the Nguara River position and Norforce would advance Likuju towards Liwale. Finally, on the middle Rufiji, a battalion-sized force from the Nigerian Brigade would leave Kibambawe for an attack on Msswega on 9 or 10 July. Furthermore, the Belgians would send a column of 500 rifles from Kilossa towards Kidodi on 8 July, although the remainder of their forces would take some time to get into position.
THE OPPOSING FORCES AT THE END OF JUNE 1917
The Allied forces had used most of June to prepare themselves for the coming offensive. As ever, the question of effective numbers remained a major concern, for despite the improvements to supply and hygiene, sickness constantly depleted the units and many of those in the forward units were still weak. Reinforcements in the form of two newly raised KAR battalions were due to be fielded by the end of June, with a number of other battalions in various stages of formation, but it would be a considerable period before they could be considered remotely ready for operations. Further help was also due from South Africa with two new reformed battalions scheduled to arrive at the end of the same month. Although these units had fought through the 1916 campaign, in reality they were brand new and only partially trained. It is interesting to note that these battalions could not be raised for Hoskins, but were made available to van Deventer.
Only the Nigerian Brigade remained on the Rufiji front, based on Kibambawe, while further east at Iringa there was a small two-battalion column commanded by Colonel Tytler who was to prevent any breakouts and hold the line until the arrival of the Belgians who were still concentrating. Their GQG moved from Ujiji to Dodoma on 29 June, but Brigade Sud remained at Kigoma, on Lake Tanganyika. Brigade Nord had two battalions chasing Abt Naumann, with another still guarding the rear areas south-west of Tabora and the fourth en route for Iringa. The artillery was moving forward to join the brigades while the pioneers were preparing to start work on the new base at Dodoma. The rear services operated on the existing line Stanleyville-Albertville with new stages of Dodoma-Iringa and Kilossa-Ruaha. Slowly but surely, the Belgians were moving into position for the next phase of the campaign.
To the south, Norforce had moved its headquarters from Fort Johnston to Songea on 24 June, arriving a week later. In the north, the main body of Northey’s troops were in their jumping-off positions around Kitanda and Likuju ready to strike north-east. Further to the west, the 1,000-strong Murray column was continuing its long and wearying journey back while a separate striking force of 500 under Colonel Shorthose prepared itself to clear Abt von Stuemer from Portuguese East Africa. In principle General Northey had nearly 3,400 soldiers, but many of these were insufficiently trained recruits and sickness forced him to relegate a number of units to the lines of communication.
On the coast at Kilwa, Hanforce disposed of two substantial columns that numbered 1,100 and 500 strong respectively, although they too were short of officers and trained troops. In light of the potential opposition, the commander-in-chief had already agreed to reinforce them with one South African and up to two KAR battalions before beginning the advance. Finally, in the south, Linforce had about 2,500 rifles concentrated in the immediate area of Lindi although a considerable proportion were not fit for field duties.
Overall, the situation was hardly good. The Indian Army battalions were largely worn out with average strengths being less than 400 men and two barely able to muster 100 men each. Despite the less demanding duties on lines of communications, the Imperial Service troops were also suffering heavily from disease. The pursuit of Naumann had drawn off many troops, leaving only a single battalion in general reserve, although the eventual return of Murray would help. More than ever, the Belgians were needed to help fill the gaps in the centre of the line. They were arriving in the Kilossa area, with several battalions moving forward to Kidodi, although it would be several weeks before the planned brigade was complete. Far to the west and north, the equivalent of a brigade of Belgians was pursuing Naumann and there was a small central reserve at Morogoro. The Portuguese had landed their latest expeditionary force of 4,500 at the port of Palma, but it was far from being ready to take the field.
On the German side, their troops remained in two major groupings, with the Osttruppen under Lettow, based at Liwale, and the Westtruppen commanded by Tafel from Mahenge. They numbered an estimated 6,200 and 2,500 respectively, giving effective combat strengths of 4,100 and 1,500. Lettow’s force was split into four main detachments; in the far south, Abt Looff had six companies and two guns between Lindi and the Portuguese border;Abt Göring, with seven companies and a gun, held the ground near Mpotora against Hanforce; Abt Otto had seven companies and a gun on the Rufiji; while Lettow retained direct control of six companies and four guns under his direct control, chiefly in the Mpotora-Liwale area.
Tafel had three substantial Abteilungen radiating some 112 km west, north and east of Mahenge down to Likuju; Abt von Brandis with three companies was near Kidodi;Abt Aumann with three companies was along the Ruhudje River;Abt Lincke with five companies was facing the main body of Norforce. The marauding Abt Naumann of six companies had been long out of contact and was effectively an independent force.