The Road to Zeilah by David Pentland. (P)
British Somaliland, 1940. Italian Carro Amato M11/39 tanks of the ‘Compania Speciale Carri M’ advance on the British held town of Zeilah. These where only one of two medium tank companies of 12 tanks each which were deployed in Africa Orientale (Ethiopia, Eritrea, and Somaliland) by the Italian Army.
Italian invasion of British Somaliland.
When Wavell had visited British Somaliland in January 1940 he concluded that, whilst this had great potential for halting any Italian advance, there were about four miles of ground to be defended and at least three strongpoints would be needed, set about a mile apart from each other. Another visit was made two months later by a junior officer from the Royal Artillery who viewed at first hand this strategically significant feature, which he measured at about 8,000 yards in length. On his return to Cairo he submitted a detailed appreciation highlighting how the advantages offered by the local terrain could be used to best effect. This also recommended that the smallest possible defending force needed was a full battalion of troops with, potentially, a second in reserve, supported by eight 25-pounder guns split into two groups. With a maximum range of 13,000 yards, these were vital, and it was recommended that ideally they should be situated at the foot of Castle Hill, one of the position’s small flat-topped hills, from where there was good visibility over the entire gap. The report noted the need for additional fixed defences to be prepared, but by this stage the lack of available men and money had slowed down work and prospects for anything being finished already looked bleak. The officer’s comment at the very end of his report was that ‘whatever the type and number of guns employed they are sure to be inadequate for the demands which will be made upon them’. The clear suggestion was that Tug Argan could not be held.
Despite such a pessimistic conclusion, four defended localities were formed and manned by the recently arrived North Rhodesians and a machine gun company of the Camel Corps. Consisting of well-prepared defences with barbed wire and concrete posts, they were dug in on what were named Black Hill, Knobbly Hill, Mill Hill and Observation Hill, with Castle Hill being used as the headquarters. Elsewhere, Indian and Nigerian troops occupied other key positions across the territory, defending routes that offered access either to the coast road or the mountain passes. From an entirely desperate position just a few months before, when the Italian attack eventually began there were in fact a total of 4,507 British and Commonwealth troops scattered across the protectorate, and whilst 75 per cent of these were African infantrymen and Somali irregulars, this was still an eightfold increase on the available strength at the beginning of the year. Although this represented less than 1 per cent of the British and Commonwealth troops then available for active duty across the British Empire, in many respects it was a remarkable outcome, particularly in light of the worsening situation in Europe and the calls for manpower that were being made from elsewhere.
There remained, however, significant deficiencies. Other than a light battery on Knobbly Hill manned by Kenyans, and despite the recommendations that had been made, crucially there were still no significant anti-tank or anti-aircraft guns. This failure to provide adequate firepower, when it was known that the Italians had tanks and armoured cars, undermined any claims that there were plans to conduct a credible defence. Some of the shortages were made up by the homemade mortars that had been produced in the railway workshops of Nairobi, but when they were eventually used, lacking proper dial sites as they did, there was no way of achieving precision when firing. There were other key problems: lack of transport meant that the defenders lacked mobility, despite a strong mobile reserve being thought absolutely critical. The positions were widely separated but the Italians were known to have mules, which would allow them to cross the rough terrain and outflank the defenders. The differences that still existed between the senior political and military figures in Berbera about what strategy to follow only made the situation worse. Whilst preparations had been made at the Sheikh Pass to completely destroy the road and make its repair a very long operation, Glenday continued to refuse to allow explosives to be put in place for fear of unsettling public opinion. He had proved a grave obstacle throughout the year and it was only in July, as the situation dramatically worsened, that the governor finally adopted a new outlook, proposing that he now leave as ‘there was not much left for him to do as the military had everything more or less in their hands’. Chater chose to encourage him to stay while also asking Wavell’s headquarters what he should do if London’s senior political representative did depart.
The lack of clear decision-making was, for once, not mirrored on the Italian side. Despite his previous orders from Rome, the viceroy of Italian East Africa, the Duke of Aosta still wanted to launch attacks on Jibuti in order to secure control of the coast and prevent his opponent from using the excellent port to land additional forces. As he assumed that the troops in British Somaliland would interfere, Aosta proposed to march on Berbera at the same time. His plan was once again submitted to Mussolini on 18 June but it was not until early the following month that he was finally given authority to proceed. Aosta had used the intervening period to study how an invasion might be conducted, and the viceroy and his deputy, General Guglielmo Nasi, who had arrived in Italian East Africa the previous May and who was one of Italy’s most capable military officers, had produced an accurate estimate of likely enemy forces and an appreciation which outlined the campaign’s objectives. On 25 July detailed instructions were issued to his troops by Lieutenant-General Carlo De Simone, who was in command of what would be the main advancing column. He controlled the bulk of Italian fighting power, including the reinforced Harrar Division, with its three colonial brigades comprising eleven infantry battalions supported by plentiful artillery and even some tanks and armoured cars, and a further two Blackshirt battalions. The priority still appeared to be to keep apart the French and British forces and prevent any landings which might lead to a counter-offensive against Harrar. A subsequent British analysis also identified two main objectives, but these were to occupy British Somaliland and to destroy the defending garrison. De Simone had chosen to interpret his main task as being to pass through Hargeisa and Sheikh in order ‘to annihilate the enemy and occupy Berbera’.
The timetable for the advance was set and East Africa’s period of relative calm was about to come to an end. There were very few aircraft available to the defending garrison, just three Blenheims and a number of antiquated Gladiators, but air reconnaissance on 3 August confirmed that about 400 Italian troops had crossed the border at Biyad. The next morning, a Sunday, additional reports identified De Simone’s column moving towards Hargeisa. Along with the road to Odweina further to the south, these were the only practical routes that could be taken to Berbera. Having been held up by the Camel Corps, just after 10 a.m. on 5 August the Italians attacked the protectorate’s major inland town with a mixed bombardment from mortars and light and heavy artillery supported by aircraft flying over the position. Three hours later, twelve light tanks advanced in line and, although three were disabled by anti-tank rifles, the decision was taken to withdraw the company of Rhodesians who had blocked the Italians’ progress. Moving in three columns but separated by a considerable distance, Nasi relied upon wireless and aircraft to communicate with De Simone he as manoeuvred into a position to attack Tug Argan from the front. The other two columns tried to mislead their opponent and potentially exploit any weakness as it appeared.
Elsewhere troops led by Lieutenant-General Sisto Bertoldi had occupied the port of Zeilah on the invasion’s first day, which removed any possibility of help for the British coming from French Somaliland. The Italian commander failed to exploit the opening presented to him and proceeded cautiously south-east along the coast, managing only to occupy the small village of Bulhar. This was possibly because the local defences had been thought to be much stronger, and thus more able to hold out longer, than was the case. However, had the defenders, operating about 150 miles from Berbera, tried to fight in such an isolated position they would have run the risk of being surrounded and destroyed, and it was sensible that they withdrew. Whilst the opportunity had not been taken to seize this potentially open road to the port, there were now scant defences blocking the advance towards Tug Argan and both sides appeared to recognise, as Wavell had anticipated, that this was the critical point in the coming battle.
Despite British and Commonwealth troops falling back at every point, the initial media reports showed an apparent lack of concern, the suggestion being that the invasion was merely ‘a “face-saving” tactic designed to strengthen morale in Italy’. Several referred to it as having been expected for some time, and that the Camel Corps, ‘an excellent and capable body of men’ with knowledge of the local terrain, was well equipped to use guerrilla tactics and act as a mobile defence that would cause casualties and delays. Great emphasis was also placed on the challenging nature of the terrain, with the Italians forced to conduct long marches over the mountainous Golis ranges, 10,000-feet-high peaks across which mechanised troops could not travel and where British aircraft could easily find targets. There was also the climate: August was the start of the dry season during which a constant burning wind and temperatures in excess of 120°F made conditions almost unbearable. Another cause for optimism were the local nomadic groups who were said to ‘both dislike and despise’ the Italians and could be counted on to fight for the Empire. These kinds of themes were commonly repeated throughout the campaign’s initial days, during which a narrative was developed for the largely ignorant readership in Britain and elsewhere about just how difficult it would be for the attack to succeed. However, this failed to grasp that virtually all of the factors enumerated as slowing down the Italians were also challenges for the garrison. Within only a few days of the fall of Zeilah and Hargeisa there was a subtle change in the media’s tone, with references to how much more difficult the defence was due to the collapse of France and the removal of any chance of support from Jibuti. Such reports even accepted that the possible loss of ‘the wretched tract’ of British Somaliland might have to be considered, and it seemed clear that the British public were being readied for worse to come.
Enemy aircraft had initially flown over Tug Argan late on 6 August 1940, by which point the Italians had already secured local air superiority and the RAF had practically lost the battle. What few aircraft there were had been put in the air immediately following the invasion but there were insufficient fighters to provide any protection, and of the only three bombers one was lost within hours. With no radar and little or no anti-aircraft defence, the decision had been quickly taken to withdraw from the temporary landing strips that had been built at Berbera and fall back on the small permanent base at Aden more than 200 miles from the battle area. Two days later, six Italian aircraft carried out a first raid on the gap, killing an askari and three Somali refugees. At the same time reports were received that a small column of tanks and infantry were moving from Hargeisa, and were finding the roadblocks and homemade landmines in front of the hills to be no barrier to their advance. The troops in the forward trenches were consequently withdrawn just after midday on 10 August and pulled back to the main prepared defences. Here they faced an intensive artillery bombardment followed by an advance of troops, both colonials and some of the paramilitary Blackshirts whom the defenders considered to be the equivalent of second-line infantry. The assault continued into the night and, after a brief pause, resumed the next morning shortly after first light as every defended post along the British line was attacked. Although the wire was reached in each case, only one of them fell. Even though by this stage it was clear that the defending forces were handicapped by a lack of adequate provisions, Chater believed that ‘my present feeling is that troops will stick it out but [I] do not think we shall be out of the wood for some days yet’.
With the battle under way it was very difficult to provide additional support to the embattled British and Commonwealth forces. A few attacks were launched from Aden by the remaining aircraft, and HMS Kimberley, Auckland, Carlisle, Ceres and HMAS Hobart, all of which were patrolling the coast, bombarded shore targets. The last of these vessels, a light cruiser under the command of Captain H.L. Howden, Royal Australian Navy, had sailed in October 1939 for the northern Arabian Sea and, from its new base in Aden, its role during the intervening months had been to escort troopships carrying reinforcements to Berbera, and then, as the battle progressed, helping to evacuate refugees. With the invasion now well under way, and in response to an attack by three Italian fighters, the ship’s single Walrus seaplane was launched against Zeilah and what was believed to be the newly created Italian military headquarters. The lone aircraft machine-gunned trucks and staff cars and two 112-pound bombs were also dropped within forty yards of the target. When the Walrus returned to the ship it had two bullet holes to show for its efforts but little actual damage had been done; the British media, looking for any positive story at this stage, nonetheless reported it as a great triumph.
Back at Tug Argan itself, one of the war’s most gallant defences was being fought. Even according to Glenday, writing a few days later in a letter sent prior to his evacuation, this ranked ‘amongst the historic actions of the British Army’ – a conclusion which seems to have been widely held. The defenders managed to hold out for three days and nights and only stopped firing when all their ammunition was exhausted. As the various military visitors who had made the journey to the protectorate beforehand had warned, it was the lack of artillery that proved decisive: it meant the defenders had little meaningful firepower to halt such a large advancing body. As the Italians neared the gap an urgent cable was received in Cairo pleading for anti-tank guns to be sent by fast ship; while four gunners were immediately flown out, it was more difficult to move equipment, which did not make it in time. When the guns did eventually arrive they were instead sent to the Sudan, later forming an important element of a mobile anti-tank gun troop. The absence of anti-aircraft defences was also considered acute and the headquarters in Cairo issued instructions that the guns helping defend Port Said should be removed and sent at once. This would leave another important element of the imperial network unguarded, and two days later the orders were quietly rescinded. All that did arrive to aid the defenders, driven in on a truck at first light on 10 August, was a three-pounder Hotchkiss gun from HMAS Hobart along with three Australian naval volunteers dressed in soldier’s uniform to act as its crew. Even this gun, however, had to be dismantled to load rounds, which meant it could only achieve a rate of fire of one shell every five minutes. There were in any case only thirty-two rounds of high explosive and the same of steel shell, hardly a devastating counter to the Italian tanks and aircraft. Although more guns were on the way, it was, however, too little and too late.
Over the days that followed, near-constant attacks by waves of Italian and colonial troops progressively moved closer to the defenders’ positions. Mill Hill had been abandoned on 12 August and, eventually, the attackers managed to reach within a few yards of the centre of the entire line. This caused the rest of it to also fall back and, with a counter-attack failing to recover the lost positions, it was only a matter of time before a withdrawal would need to be made. The situation was now reviewed by the local British commander in the knowledge that every man had already fought for seventy-two hours without rest and little ammunition remained. What proved to be the determining factor was the loss of the few available artillery pieces. There had been four guns of the 1st East African Light Battery deployed with the Rhodesians, two on Knobbly Hill and two more on Mill Hill; these either ran out of ammunition or, when the Italians had closed within their minimum range, they could no longer fire, even over open sights.
Observation Hill held out until the evening of 15 August and, even before it fell, the order had been given to pull back towards Berbera. This part of the plan was carried out without any real Italian opposition and the remaining British and Commonwealth troops withdrew in the direction of Knobbly Hill. They left behind Acting Captain Eric Wilson, who had been responsible for directing the fire from the Camel Corps’ machine guns on the last of the hills still held by the defenders. Hit by artillery fire on 11 August, which severely injured his right shoulder and left eye, he repaired his weapon and continued to fire on the advancing Italians until his position was finally overrun four days later; by then all of those around him, including his dog, were dead. As the garrison fell back on Berbera his capture went unnoticed, but his actions on Observation Hill were not: he was awarded the first Victoria Cross of the war in Africa, though he was presumed dead. (In fact, Wilson died in 2008, by which point he was the oldest surviving wartime recipient of this medal.) His formal citation, gazetted on 11 October 1940, opened with the words ‘For most conspicuous gallantry on active service in Somaliland’ and ended ‘The enemy finally overran the post at 5pm when Captain Wilson, fighting to the last, was killed’. Wilson’s father’s comment on hearing the news was that ‘he has died, I suppose, as every soldier would wish to die – fighting. It’s a great and terrible loss to us, but we know that he did his duty.’ The more sensational media wrote of ‘Another Rorke’s Drift’.