Page 165

Chapter 9: The First Encounters on the Borders of Italian East Africa

See Map 10

Map 10

Map 10. Italian East Africa and surrounding countries

THE UNPREPAREDNESS of Italy to face a long war was nowhere more acutely felt than in Italian East Africa. The situation of Somalia, Eritrea, and Ethiopia was strategically weak because the sea communications could easily be cut, and even the air communications were liable to be interrupted, by an enemy based in the Middle East. On land the nearest Italian post and landing ground, the oasis of Uweinat, was nearly a thousand miles away across the Sudan. It was not as if Italian East Africa had the industries, the natural resources, or the accumulated reserves to offset a prolonged interruption of the normal routes of supply. To make things worse, Ethiopia was a military liability even in peace time, because the first and ever present need was to be able to enforce internal order. It was for this, and not for operating against an external enemy, that the forces were organized.

The Duke of Aosta, Viceroy of Ethiopia and Supreme Commander of all armed forces in Italian East Africa, was so uneasy about the state of these forces that in April 1940, when Italy’s intervention in the war was becoming likely, he went to Rome to represent his deficiencies to the Duce and the Ministers concerned.1 As a result. he received an allotment of 900 million lire and a promise of reinforcements, weapons, and warlike stores. But it had been left too late, for by the time that events elsewhere had forced Italy into the war only a few officers and specialists, one company of light and one of medium tanks, 48 field guns, and some machine-guns and mortars had reached East Africa out of this consignment. Other ships were on their way, but were either recalled or were intercepted by the Royal Navy.

On May 30th the Viceroy received instructions from Marshal Badoglio that his forces were to be brought to full battle order by

Page 166

June 5th, and that their role was to be strictly defensive. This was, no doubt, in accordance with the Duce’s belief that the impending collapse of the Allies in Europe would make unnecessary any serious fighting overseas, but the Viceroy at once enquired whether the orders precluded the operations against Jibuti which had formed an important part of the current Italian war plan, and if so whether it would be correct for him, while refraining from any large enterprises, to undertake local frontier actions. Marshal Badoglio replied that the Viceroy’s main object must be to guarantee the integrity of the Empire: he was to maintain a strictly defensive attitude but was to be ready to react swiftly and strongly to any attack; later he might be asked to study certain offensive plans. Having been informed on June 9th of the hour at which war would begin, the Viceroy asked if he might then attack the British by sea and air, in order to anticipate them. He received a prompt reply ordering him to take no offensive action.

Within Ethiopia the needs of internal security had led to the establishment of many scattered garrisons, reasonably accessible by road, and each containing a substantial force of colonial troops, stiffened here and there by one or more Blackshirt battalions. In the colonial battalions the officers and many of the NCOs were Italians. The native troops varied greatly in fighting quality, and broadly speaking were armed and trained only for ‘tribal’ warfare; they were unaccustomed to manoeuvre on any but the smallest scale. Distributed along the frontiers were numerous groups of small irregular units or ‘bande’, more lightly armed than the colonial troops and tactically rather more mobile.

But although the land forces in general lacked cohesion and flexibility, they were certainly strong in numbers. They consisted in the main of twenty-nine colonial brigades, mostly of three or four battalions of infantry and two batteries of pack artillery each. In addition there were seventeen independent colonial battalions, sixteen Italian battalions, and ten Italian artillery groups. Natives of all types and in all the Services, including the police, accounted for about 70% of the whole. The main reserves of troops were located at Addis Ababa, Dessie and Adigrat, the chief formations in reserve being one regular Italian division (the Savoia) and one of locally enlisted Italian nationals (the Africa). There were more than a hundred armoured cars manned by Italians and—from just before the start of hostilities—about 60 medium and light tanks. The 24 bomber and 4 fighter units, with a total of 183 aircraft and 61 in reserve—there were a further 81 under repair—had been accustomed to co-operate by reconnaissance and liaison in the work of policing. The ground organizations were accordingly scattered over wide distances, just as the army garrisons were.

Page 167

Besides the dispersion imposed by the internal situation, a further handicap lay in the weakness of the administrative backing, which existed only on a scale suitable for the task of maintaining order and not for the support of operations against much resistance or at great distances. The shortages which caused particular concern were of anti-tank and anti-aircraft guns, ammunition of all kinds, motor and aviation fuel, aircraft and aircraft spares, and, above all, motor tyres. In short, the Duke of Aosta was by no means satisfied with his logistic situation but realized that there was little prospect of its being improved now that physical communication with Italy was cut. He resigned himself to a policy of imposing all possible economies, and of making do.

These administrative anxieties did not however prevent him from taking a balanced and sensible view of the strategical situation. He did not accept the exaggerated estimates of the British strength which were submitted to him, and did not expect to be attacked immediately. But he realized that there was a great deal of internal unrest, and in reviewing the position on June 18th he reported that in the past seven weeks there had been no less than five major disturbances, which he believed were intended to dissipate his forces and interrupt internal communications. He concluded that they were instigated by the British and supported by British propaganda and money, and he was especially concerned over the activities of the eight thousand or so Ethiopian exiles who were now scattered round the frontiers of the Empire, mostly in Kenya. The news of the Emperor’s arrival in the Sudan on July 3rd reached him from Italy three weeks later and naturally increased his expectation of trouble. He .thought it probable that Ethiopia would ultimately be attacked from both Kenya and the Sudan, in, conjunction with, an Ethiopian rising in the Gojjam and Shoa districts, which, if successful, would have the effect of cutting the country in half. But as the first two months of war passed he noted that the collapse of the French, the conquest of British Somaliland, and the various successful actions on the frontier seemed to have had a sobering effect on Ethiopian opinion. Indeed, he began to feel that the rebel elements had decided to adopt a waiting policy.

During July the Viceroy had become more and more irritated by the failure to clear up the situation in Jibuti. He had always recognized that French Somaliland presented a strategic threat to Ethiopia, but the defensive attitude imposed upon, him by the Duce precluded any action being taken to reduce it. He therefore welcomed the prospect offered by the Franco-Italian armistice that the threat would be speedily removed, and viewed with increasing annoyance the successful evasions of General Legentilhomme, whom he regarded as ‘an incubus sent to us by the British’. The Duce and Marshal

Page 168

Badoglio shared his annoyance and did all they could to hasten matters, but it was not until July 27th that the Viceroy was able to report that the crisis had been overcome and that General Germain was in control. Even then the application of the armistice terms did not remove all his suspicions about French intentions, and he remained uneasy lest French Somaliland might yet become the bridgehead for a British invasion.

Throughout the months of June and July the Duce remained unwilling to authorize the Viceroy to undertake any offensive other than the limited attacks on Kassala and Gallabat in the Sudan, and on Moyale on the Kenya border. But he authorized the study of operations against Zeila and Berbera and plans were made to begin on any date after July 22nd. The entanglement over Jibuti, however, caused a postponement, and not until it had been unravelled was the Viceroy free to embark upon his only offensive operation of any size.

Prior to the outbreak of war with Italy an increase had been noticed in the Italian forces in Eritrea, and there were indications of concentrations towards the Sudanese frontier. It was realized that these might be defensive precautions, but they might equally well be the first signs of an intended invasion of the Sudan. In this case the choice of objectives obviously lay between Khartoum, Atbara, and Port Sudan. Khartoum was the centre of political and military control; Atbara was at the important junction of the railways to Khartoum and Port Sudan, and contained the only heavy workshops in the country; Port Sudan was the sole useful port. The distances to these objectives from the Eritrean border were very great; over 200 miles to Atbara and nearly 300 to Khartoum. This part of the country was devoid of military resources, mostly arid, and with no metalled roads, but it was traversable almost everywhere by motor transport in dry weather. With the beginning of the rains about the end of June or early July movement was liable to be severely restricted over periods of two days at a time until about the end of September. At no time, therefore, would invasion to any great distance be particularly easy.

The G.O.C. Troops in the Sudan, Major-General W. Platt, had under his command three British battalions; his only other regulars were the units of the Sudan Defence Force.2 This Corps had been formed in 1925 from existing irregular units and certain Sudanese battalions of the Egyptian Army, and consisted of natives of the

Page 169

Sudan led by British and Sudanese officers. Until 1936 its task had been to maintain internal security in the Sudan, and it was organized and lightly equipped accordingly, The conquest of Ethiopia by the Italians made it advisable to increase the scope of the force and a reorganization was begun. In June 1940 it comprised twenty-one companies, or 4,500 men in all, the most modernized units being five (later six) Motor Machine-Gun Companies—small mobile units consisting of light machine-guns carried in vans and trucks—and a number of locally constructed armoured cars. The force had as yet no artillery, though the Sudan Horse was in process of conversion into a battery armed with 3.7-inch howitzers.

General Platt’s plan was to hold the three vital centres initially with his British battalions; 2nd Battalion The West Yorkshire Regiment at Khartoum; 1st Battalion The Essex Regiment at Atbara; and 1st Battalion The Worcestershire Regiment at Gebeit and Port Sudan. Upon the frontier were stationed units of the Sudan Defence Force, provincial police, and sundry irregular scouts, with the object of observing, harassing, and delaying the enemy. If a hostile thrust was clearly disclosed General Platt intended to concentrate the greatest possible force against it, relying upon distance, poor communications, and lack of supplies to absorb much of the impetus of the advance.

The tasks of the Air Force in the Sudan were the protection of shipping in the Red Sea, including anti-submarine patrols; the air defence of Port Sudan, Atbara, and Khartoum; and the close support of the Army and of the Ethiopian patriots. For these tasks there were only three bomber squadrons (Nos. 14, 47, and 223) armed with obsolescent Wellesleys, later reinforced by No. 45 (Blenheim II) Squadron from Egypt. For Port Sudan there were six Gladiator fighters of No. 112 Squadron, known as K Flight. For army co-operation No. 430 (Vincent) Flight was formed from No. 47 Squadron. These units formed No. 254 Wing, with headquarters at Erkowit, conveniently placed for controlling operations in the Port Sudan and Red Sea areas, but not those at Kassala and Gedaref or across the border. Accordingly H.Q. No. 203 Group, under the command of Air Commodore L. H. Slatter, was formed on August 17th at Khartoum, where the Air Officer Commanding was in close touch with the General Officer Commanding and in a position to supervise the Sudan sector of the Takoradi air route. No.1 (Fighter) Squadron S.A.A.F., recently rearmed with Gladiators in Egypt, arrived at Khartoum early in August.

The first three weeks of war passed without any signs that the Italians intended to make full use of the short remaining period of dry weather, though their air forces were active and made frequent reconnaissances: bombing attacks were directed chiefly upon Kassala,

Page 170

Port Sudan, Atbara, Kurmuk and Gedaref. The sole ground defence was by small-arms fire, and civilian morale suffered accordingly, but there was no military damage of any consequence. British air operations were directed chiefly against warships based on Massawa and against airfields in Eritrea. On land the Italians made no move. The Sudan Defence Force, on the other hand, was as active as possible and frequently patrolled and raided across the frontier, especially in the neighbourhood of Kassala and Gallabat. Casualties were inflicted upon the enemy, a few prisoners were taken, and the Sudanese soldiers gained in confidence. Early in July the enemy’s attitude changed, and the expected attacks on the frontier posts began. Karora, Kassala, Gallabat, and Kurmuk were those chosen.

Kassala, on the long eastern loop of the Sudan Railway, was a provincial town of some importance. It was situated on the river Gash, about 20 miles from the point on the frontier to which ran the one good road—the Via Imperiale—from Asmara, the Eritrean capital. Kassala was an obvious first step in an advance towards either Atbara or Khartoum. At the beginning of July a concentration of the enemy was discovered near Tessenei, about twelve miles across the border, and a plan was made for attacking it on July 3rd. A failure in wireless communication led to a postponement. On July 4th the enemy advanced on Kassala from the east and along both sides of the Gash—three columns in all. Captured documents show that the force employed was two colonial brigades, four cavalry squadrons, about twenty-four light and medium tanks and armoured cars, and ten batteries of artillery of various calibres. To oppose this considerable array were two companies of the Sudan Defence Force, No. 5 Motor Machine-Gun Company and No. 6 Mounted Infantry Company, reinforced (at mid-day) by No. 3 Motor Machine-Gun Company from Butana Bridge. The action began with the bombing of the town and by 6.30 a.m. the enemy’s cavalry were in contact with the defenders. The extreme disparity in numbers and weapons could lead to only one result, but though Kassala passed into enemy hands the conduct of the Sudan Defence Force gave great promise for the future. In their difficult task of causing as much loss as they could without becoming deeply involved they did in fact have 10 casualties, while the Duke of Aosta reported the Italian losses to have been 117. The three companies reassembled successfully at Butana Bridge.

On the same day an Italian colonial battalion and a Banda (irregular) group drove the platoon of No. 3 Company Eastern Arab Corps from Gallabat, while Karora was occupied by the enemy after the police party there had been withdrawn. On July 7th an Italian colonial battalion and irregulars, with artillery and air support, dispersed the 60 Sudanese police at Kurmuk after an hour’s fight. From these successes the Italians had undoubtedly made

Page 171

valuable gains with which to impress the native population, to say the least. Militarily they had secured an important entrance to the Sudan at Kassala, and by capturing Gallabat they made it harder for the British to make contact with the patriots in Gojjam. The loss of Kurmuk had a bad effect on the local population, many of whom turned bandit. In contrast, and in spite of Italian boasts, the inhabitants of Kassala remained loyal throughout the six months of Italian occupation and helped the British very considerably by collecting information.

From General Platt’s point of view there was no alternative to the policy of observing and harassing the enemy, while retaining the power to concentrate against renewed and deeper thrusts. In the Kassala and Gallabat areas the Sudan Defence Force continued to patrol actively. On July 5th a company of 2nd Bn. West Yorkshire Regiment was moved up to Gedaref to act as a reserve to the Sudan Defence Force. Learning that the report of this move had reached the enemy greatly exaggerated, General Platt proceeded to adopt every form of ruse to give the impression of far greater strength than he really possessed. There were good reasons for believing that these efforts were being successful, and an Italian General Staff map, of 25th July 1940, captured later, shows some 20,000 troops in or near the Kassala province of the Sudan. The situation was nevertheless not one for complacency, and when, on August 2nd, the destination of the 5th Indian Division was changed from Iraq to the Middle East, General Wavell at once ordered the leading brigade group to be disembarked at Port Sudan.

At the Anglo-French meeting at Aden in May 1939 it had been agreed that French Somaliland was of great strategic importance to the Allies. The French had been reasonably confident in their ability to hold the direct lines of approach from Ethiopia, but they had been apprehensive of an Italian advance through British Somaliland culminating in an attack on Jibuti from the direction of Zeila. It had been agreed that the French should be enabled to make the best possible tactical dispositions, moving forward, if necessary, into British Somaliland. As regards air support, the primary aim of the British air forces must be to reduce the scale of attack on Aden and the shipping. Operations to this end would be concerted from. Aden and the Sudan. But if a critical situation arose in Somaliland the maximum of air support would be given to the troops, consistent with the security of any convoy that might be in the Red Sea at the time.

It had been the intention to abandon British Somaliland if invaded, but in December 1939 the Chiefs of Staff decided on General

Page 172

Wavell’s recommendation that in principle the policy should be to defend the territory and, in the last resort, Berbera. The troops were to come under General Wavell’s Command,3 but he was not to add to their number without the permission of the British Government. On visiting the country General Wavell came to the conclusion that a comparatively small addition to the force would make it possible to hold British as well as French Somaliland, which was very desirable for reasons of prestige and would be useful when the time came to attack the Italians. The fact that the Colonial Office, the Foreign Office and the War Office were all involved in this proposal did not make for its rapid acceptance; in fact it was March before permission to bring in one battalion was given, and it was May before the shipping was available. Financial approval for much of the intended work on defences was never given, and on June 1st General Headquarters, Middle East, took over administrative control from the Colonial Office.

On the outbreak of war with Italy the troops in British Somaliland were under the immediate command of Lieut.-Col. A. R. Chater, Royal Marines, hitherto the Commanding Officer of the Somaliland Camel Corps. In the last week of July the collapse of the opposition in French Somaliland to the terms of the armistice released large Italian forces for use against British Somaliland, but General Wavell still thought that to defend the approaches to Berbera was preferable to a withdrawal without fighting, a view which was shared by the Air Officer Commanding-in-Chief and the Commander-in-Chief East Indies. It did not seem that if the Italians occupied British Somaliland they would necessarily intensify their air attacks on Aden and on shipping, because even from well-established bases in Eritrea they had made a very poor showing.

The real issue was—to fight or not to fight; and General Wavell agreed with Lieut.-Col. Chater that nothing less than five battalions would have a reasonable chance of holding on, and these he considered could be made available. This was not to say that they could be sent at once or that Lieut.-Col. Chater could make firm plans for a force of this size, nor was it certain that any medium or field artillery would be available. By the beginning of August Brigadier Chater, who had just been promoted, had under his command 1st Bn. Northern Rhodesia Regiment, 2nd Bn. King’s African Rifles and 1st East African Light Battery (four 3.7-inch howitzers) from Kenya; 1/2nd Punjab Regiment from Aden; 3/15th Punjab Regiment diverted from Aden; and the lightly armed Camel Corps, which had on mobilization received a valuable reinforcement of seventeen officers and twenty NCOs from the Southern Rhodesia

Page 173

Regiment. 2nd Bn. The Black Watch, which had been sent to Aden from Egypt on July 1st in response to General Legentilhomme’s request for a battalion to be ready to support him in French Somaliland, joined the force on August 8th, making the fifth battalion. The force therefore amounted to a last minute assemblage of units of four distinct races, each requiring different treatment and food. It lacked an organized base, a proper headquarters, and many other essentials, especially artillery, transport, and signal equipment. Air support could be provided only from the force at Aden, already committed both directly and indirectly to the protection of convoys and to the air defence of that port. Two 3-inch A.A. guns of 23rd Battery, Hong Kong and Singapore Brigade, R.A., were sent from the Aden defences. This was all that could be spared from other Middle East commitments.

The objective of any invasion would obviously be Berbera—the capital and the only port of any size, but not a suitable one for supporting a military expedition, for all unloading had to be done by lighters and a 3,000-ton ship normally took ten days to discharge. The frontier with Ethiopia was long and open, and offered no suitable positions on which to defend the main approaches to Berbera, of which there were three: one through Zeila, near the French Somaliland border, and thence eastward along the coast road; the second through Hargeisa; and the third through Burao. The main physical feature of the country is the range of rugged hills which runs parallel to the coast and about 50 miles or more from it. It was impassable by wheeled or tracked vehicles except by the roads to Hargeisa and Burao. The former and more direct road crossed the hills at a wide gap known as the Tug Argan. The road from Burao passed through a much narrower defile at the Sheikh pass. Between these hills and Berbera the ground was generally flat and open, and offered no natural position on which a smaller force could delay a determined advance for long. Brigadier Chater allotted initially two battalions and his light battery to Tug Argan and one to the other main approaches, and kept one battalion in reserve. When the fifth battalion (The Black Watch) arrived it went into reserve in place of 3/15th Punjab Regiment, which was moved up to strengthen the Tug Argan position. The Camel Corps formed a thin advanced screen for observing and delaying the enemy, and were supplemented by patrols of the Illalos, a small force of tribal levies whose normal task was rudimentary police work in the frontier areas; they showed great courage and loyalty throughout the campaign. The Camel Corps and Illalos made several bold and successful raids against enemy detachments near the frontier.

The Italian plan, as has been seen, was prevented by the ingenuity

Page 174

of General Legentilhomme from being put into effect sooner. The operation was under the command of Lieut.-General Nasi, General Officer Commanding Eastern Sector, who had the equivalent of an army corps made up of three Italian and twenty-three colonial battalions, three or four field howitzer batteries in addition to the pack artillery with each colonial brigade, making twenty-one batteries in all, a half-company of medium and a squadron of light tanks, some armoured cars, five groups of irregulars, and fifty-seven aircraft in support. The main column, under Major-General De Simone was to seize Hargeisa, establish a base, and then advance on Berbera. The right, or eastern column of irregulars under Brigadier-General Bertello was to move to Odweina to protect De Simone’s flank, and be prepared to join him if necessary. On the extreme left Lieut.-General Bertoldi’s column was to seize Zeila, seal off French Somaliland, and then send small forces eastwards. The Duke of Aosta seems to have been reasonably well informed of the British strength and dispositions, and in a written appreciation of 14th July expressed the view that the decisive encounter would take place between the Karrim and Jerato passes and that if the enemy held his ground the Italians would be able to manoeuvre so as to envelop the flanks.

The Italians crossed the frontier on August 3rd. The forward patrols of the Camel Corps fell back in touch with the advancing columns, keeping them under observation and striking when possible. On August 5th General Bertoldi occupied Zeila and began to push eastward. These movements, which seemed to threaten communications between the British force and its base, but which were not in reality vigorous, were harassed from the air and sea and by the patrols of the Camel Corps. On the eastern flank General Bertello reached Odweina on the 6th and then turned, not to Burao, but north-west towards Adadle, that is to say towards the line of advance of the centre column. General De Simone’s initial advance on Hargeisa was held up by the Camel Corps, but on August 5th it was renewed in greater strength with the aid of light tanks, and the Camel Corps and a company of the Northern Rhodesia Regiment operating with them were obliged to withdraw. There was then a few days pause, except for some air reconnaissance and bombing by both sides, while the enemy stocked the Hargeisa area in preparation for a further advance. In response to the Duke of Aosta’s urgings for speed General Nasi pointed out that his only road was in a very bad state, thanks to the unusual traffic and heavy rain. The advance was resumed on the 8th, and the two following days were spent in making contact with the main position and in preparing to attack it. It was during this phase that the presence of Italian medium tanks was first reported. Because there were no anti-tank guns the Captain

Map 11

Map 11. The Battle of Tug Argan, August 1940

Page 175

of H.M.A.S. Hobart offered a 3-pdr saluting gun with 30 rounds of ammunition—an offer which was gladly accepted.4

When it was clear to General Wavell that the Italians were indeed intending an invasion in strength, he decided not to leave the British force entirely without field or anti-tank artillery, and ordered one field regiment (less one battery) from 4th Indian Division in the Western Desert, and one section of two anti-tank guns, to be sent by special convoy to Berbera. At the same time India was asked to load the first flight of the 5th Indian Division so that a battalion, a field battery, and a field company could be disembarked at Berbera. These reinforcements did not reach British Somaliland in time, and were sent instead to the Sudan.

It seemed to General Wavell that with all these troops under orders for Somaliland it would be appropriate to appoint a Major-General to command. Major-General A. R. Godwin-Austen, who had been temporarily commanding the troops in Palestine, had just handed over his duties and was on his way to East Africa to take command of the 2nd African Division. He was sent instead to Berbera, with instructions to prevent the Italians from advancing beyond the main position, but nevertheless to prepare, in secret, plans for evacuating the force in case this should become necessary. General Godwin-Austen arrived at Berbera on August 11th and assumed command that evening.

The same day marked the opening of the battle for the Tug Argan gap, so called from the name of a dry sandy river-bed or ‘tug’ which ran across the front.5 This is the point at which the road from Hargeisa crosses the line of the Assa range and immediately turns east to pass between it and the confused mass of flat-topped hills and tugs which lie to the north. The gap consists of a flat stony floor, thinly sprinkled with thorn bush, intersected by numerous tugs and dotted with a few isolated rocky hills, 2,000 to 2,500 yards apart. These hills formed the defended localities which are named on Map 11 from north to south Black, Knobbly, Mill, and Observation Hills, and Castle Hill about two miles east of Mill Hill. Each had been prepared to some extent with machine-gun posts, and with a weak barbed-wire obstacle. In general, the position was a good one in that the enemy would be obliged to attack it, and it afforded good observation, although this could not be made full use of owing to the complete lack of any field artillery. (Two Bofors guns, for anti-tank use, arrived on the 13th.) The gap itself was very wide for the troops available, and the defended localities were far enough apart for infiltration to be possible between them; and except for Castle Hill

Page 176

they were all in the foremost line, so that most of the position had no depth. Finally—and as it turned out, this was a point of some importance—the Assa hills to the east of the bend in the Berbera road were not impassable to infantry, and there were several camel tracks such as those which crossed by the Mirgo and Jerato passes, which had the effect of widening the front still further; from the Mirgo pass to Black Hill was about eight miles.

On August 11th the enemy shelled and then heavily attacked the part of the westerly end of the Assa hills known as Punjab Ridge, employing a force estimated at not less than a brigade, which drove off the company of 3/15th Punjab Regiment holding it, and repulsed the ensuing counter-attack. Attacks on Mill Hill and Knobbly Hill failed. The next day all the principal defended localities were attacked. Towards evening the weakest, Mill Hill, was lost after severe fighting and two of the 1st East African Light Battery’s invaluable howitzers had to be abandoned, having first been made useless. More serious, however, than the loss of Mill Hill was the fact that the enemy succeeded in improving his hold in the Assa hills and by nightfall was in possession of high ground dominating the southern part of the gap. The 13th saw another attack on Knobbly Hill repulsed, but small parties of the enemy began to filter down from the Mirgo pass and during the night ambushed, but failed to halt, a convoy of water and ammunition which was coming forward to Castle Hill. On the 4th Castle Hill and Observation Hill were bombed and heavily shelled and another attack on the latter was beaten off.6 Meanwhile, in an attempt to relieve the threat to the road, a counter-attack was made towards the Mirgo pass by two companies of 2nd King’s African Rifles, but after a temporary success it was thrown back.

It was now clear to General Godwin-Austen that not only was the enemy almost in a position to cut the road along which all supplies and water had to come, but that his marked preponderance in artillery, which outranged the light howitzers and so was able to fire without interference, meant that he could concentrate against each defended locality in turn. The regiment of field artillery, still on its way from Egypt, would have made all the difference in this respect. After four days of fighting the troops were becoming tired. Should Tug Argan fall there was no position behind it that could not easily be outflanked, and it would be useless simply to try to hold Berbera itself. On the 14th he therefore informed General Wilson, who was acting as Commander-in-Chief during General Wavell’s absence in England, that if the present position fell he saw no alternative to evacuation, which would mean perhaps 70% of the force being

Page 177

saved. He was of course prepared to fight to the end if this were required. After the failure to recapture the Mirgo pass it was evident that, he had not the troops both for stabilizing the forward area and for covering a withdrawal, should this be necessary. The choice was therefore between fighting it out, with the loss of the whole force, and evacuating. He communicated this view early on the 15th and shortly after noon received General Wilson’s order to evacuate. His plan was to hold two rearguard positions, one at Barkasan, and one at Nasiyeh, 33 and 17 miles from Berbera respectively. During three successive nights first the civilians and then the troops would be embarked. The spread over three nights was necessary because shipping had to be collected and at this season the monsoon conditions normally prevented embarkation by boat during the night and forenoon.

After being shelled for most of the day of the 15th, Observation Hill at last fell in the late afternoon to a strong attack; for extreme gallantry in its defence Captain E. C. T. Wilson, The East Surrey Regiment, attached Somaliland Camel Corps, was awarded the Victoria Cross. After dark the Northern Rhodesia Regiment was withdrawn from Black, Knobbly, and Castle Hills. The Black Watch, joined by two companies of 2nd King’s African Rifles and elements of 1/2nd Punjab Regiment, took up a rearguard position at Barkasan, while another composite force occupied a second position at Nasiyeh. Through these positions the troops from Tug Argan withdrew. The Italians followed up slowly and not until August 17th did they attack at Barkasan. The brunt of the action was borne by The Black Watch, who carried out several successful counter-attacks during the day. The enemy was firmly held, though it was evident that his numbers would allow him in time to outflank the rearguard. Meanwhile embarkation had been going on with unexpected speed, so that General Godwin-Austen was able to accelerate his programme. He accordingly abandoned the Nasiyeh position and withdrew the Barkasan rearguard after dark on the 17th. Embarkation was complete by 2 p.m. on the 18th, though H.M.A.S. Hobart with Force Headquarters remained at Berbera until the following morning before sailing for Aden. The Italians resumed their advance on the 19th and entered Berbera that evening.

Throughout the short campaign the Royal Air Force were at a great disadvantage in that the airfields at Berbera and Laferug soon became unusable from enemy bombing; two of No. 94 Squadron’s Gladiators were in fact destroyed on the ground. These airfields had only small-arm defence, since the primary task of the two 3-inch anti-aircraft guns was the defence of the port area. The bomber squadrons were therefore forced to operate from Aden, which involved a flight of 200 miles to and from the target area. Lacking

Page 178

the protection of short-range fighters they were freely harassed by enemy fighters and suffered accordingly. The long distance also reduced the bombing effort and made it impossible to keep in touch with the army’s situation. Nevertheless, in addition to distant reconnaissance, attacks were made on troops and transport, and also on enemy bases, including Addis Ababa, with the object of drawing off the fighters. Blenheim aircraft in the fighter role maintained standing patrols over Berbera for 13 days and secured for the port a reasonable freedom from air bombardment, especially during the evacuation. Air losses during the fortnight were relatively severe, seven aircraft being lost and ten others badly damaged. The Blenheim squadrons were Nos. 8, 11, and 39, all Mark I, and No. 203 (G.R.) Squadron Mark IV, recently converted to four-gun fighters. On August 14th Aden was reinforced by No. 223 Squadron from the Sudan and a flight of No. 84 from Iraq, both Blenheim Mark I’s.

The Royal Navy’s activities took many forms. HMS Kimberly, Auckland, Carlisle, Ceres and H.M.A.S. Hobart co-operated by patrolling the coast, bombarding shore targets, and maintaining touch with and finally rescuing outlying parties of troops. In the end the Army had every reason to be grateful for the skill and speed with which 7,000 people, including civilians, were embarked.

The defence of British Somaliland cost 260 casualties, of which the majority were in the Northern Rhodesia Regiment and the machine-gunners of the Camel Corps, who together bore the brunt of the fighting in the Gap. A quantity of stores and equipment was lost because in the circumstances it could not be embarked. In spite of the safe withdrawal of the force there is no doubt that this latest setback, and at the hands of the Italians, came as a shock to British public opinion. Yet General Wavell strongly endorsed the action that had been taken in his absence; General Godwin-Austen, he said, had judged the situation correctly and General Wilson had given a sound decision. The conduct of the troops, in very testing circumstances, had been excellent.

It may seem in retrospect that the British force had been set a task beyond its powers. On the other hand it must be remembered that in war nothing is certain, unless it be that the enemy will do what is least expected of him and that no results will ever be achieved if no risks are taken. It was not certain that the Italians intended to increase the spread of their forces still further; their internal situation was imperfectly known and might have seemed much worse to them than it did to us; their policy for all we knew might have been to conserve their reserves rigorously, realizing that until communication with Libya was opened up they could replace nothing—in itself a good reason for our not letting them have Berbera without paying for it; their colonial troops might have proved unreliable; their maintenance

Page 179

services might have been unequal to supporting a long advance. All these unknowns, coupled with the reflection that wars are not won by looking on but that battle experience has to be gained in battle, no doubt weighed with General Wavell in deciding that here was a risk that ought to be taken without endangering the decisive front in Egypt. In the event, the Italians used ample forces; they made a good plan, which exploited their advantages; they had supply difficulties and overcame them, at the cost of much wear and tear on their transport; and having won the battle they were unable to follow up in sufficient strength to interfere with the embarkation. They paid for their success with 2,052 casualties and an expenditure of material that they could ill afford.

The situation in Kenya in June 1940 bore many resemblances to that in the Sudan. There were very few troops for the defence of an enormous expanse of territory, but the vital areas—which were broadly those served by the Mombasa-Nairobi-Uganda railway—were a long way from the frontier. Between the escarpment marking the southern limit of Ethiopia and the foothills of the Kenya highlands stretches a flat grey landscape of 200 miles of sun-baked earth and leafless thorn trees, with occasional hills and very little water. None of the tracks across this waste remained passable during the rains, which in this district occur from April to June and from mid-October to mid-December.

Kenya’s great asset was its well-equipped deep-water port of Mombasa, already referred to as a potential back door to Egypt, to which it was connected by a combination of rail, road and river; not an easy or rapid line of communication, but nevertheless a slight insurance against the closing of the Red Sea. When the Italians failed to do this the need for the back door receded at—any rate for a time—and Kenya took on as its primary task the organization of a base for an advance into Italian East Africa.

Ever since the Italian conquest of Ethiopia in 1936 the Inspector-General of the African Colonial Forces, Major-General G. J. Giffard had worked hard for an increase in the size and scope of these forces. but with only partial success as the necessary hinds were not forthcoming. But he succeeded in so organizing them as to allow of efficient expansion in emergency, and, in the event, considerable headway was made with their growth and training during the eight months while Italy was hesitating. In peace time they consisted almost entirely of infantry, the men being voluntarily enlisted, with officers and senior NCOs seconded from British regiments. The Royal West African Frontier Force consisted of eight battalions: five in the Nigeria Regiment, two in the Gold Coast Regiment, one

Page 180

Sierra Leone Battalion, and one Gambia Company. In East Africa there were the King’s African Rifles, consisting of two brigades, with three battalions from Kenya and Uganda and three from Tanganyika Territory and Nyasaland; and the Northern Rhodesia Regiment of one battalion.

On the outbreak of war with Italy Major-General D. P. Dickinson, who had succeeded General Giffard as Inspector-General of the African Colonial Forces, was appointed General Officer Commanding East Africa Force, comprising only two East African brigades, an East African reconnaissance regiment and a light battery, and the 22nd Mountain Battery R.A. from India. The task given to him by General Wavell was to defend Kenya and without compromising that defence to contain as many Italians as possible on his front. He decided to hold a coastal area in front of Mombasa; to deny the enemy access to the river Tana and to the water at Wajir; and to station detachments at Marsabit, at Moyale on the frontier, and in Turkana by Lake Rudolf, which meant that the force was stretched over an arc of 850 miles. The enemy was believed to have troops at the principal centres Kismayu, Mogadishu, Dolo, Moyale and Yavello; these were in fact colonial brigades and bande in each case, and a further force of about two brigades near Jimma could reinforce Moyale or strike south towards Lake Rudolf and so into Uganda.7

The policy of the Union of South Africa in providing troops and air forces has already been referred to, and among the most valuable of their early contributions to the defence of Kenya was the 1st South African Anti-Aircraft Brigade, whose arrival enabled protection to be given to Mombasa. By the outbreak of war with Italy the air force units in Kenya comprised No. 237 Rhodesian Squadron, the Kenya Auxiliary Air Unit, Nos. 11 and 12 Bomber Squadrons S.A.A.F., armed at first with Hartbeest and German Junkers 86 aircraft respectively, followed shortly afterwards by No. 40 Squadron which took over the Hartbeests and became an Army Co-operation unit; No. 11 Squadron was then rearmed with Fairey Battles.

At first the air forces of both sides devoted most of their attention to the forward areas, while patrols of the King’s African Rifles had several brushes with the enemy at points on the frontier of Italian. Somaliland and at Moyale. It was only here that the Italians dis- played any marked activity. Situated on the edge of the Ethiopian escarpment Moyale covered the junction of the tracks leading to Wajir and Marsabit and faced a very likely line of approach from Ethiopia. The British did not intend to hold this exposed position to the last, nor yet abandon it without a fight. The garrison of one company of the 1st King’s African Rifles repulsed an attack on July 1st, and, as a precaution, reinforcements were moved up to

Page 181

within supporting distance. On July 10th, after considerable shelling, a much heavier attack began for which the enemy appears to have allotted four battalions. Alter three days of intermittent fighting the Brigade Commander judged that to hold Moyale any longer would absorb too much of his strength. The withdrawal was successfully achieved and there was no follow up, though in due course the Italians advanced to the water-holes at Dobel and Buna. With this measure of success they appeared to be content.

At the end of June the West African contingent, which by now had been formed into two brigade groups, one from Nigeria, one from the Cold Coast, arrived by sea. A step forward in organization was taken by combining the East and West African brigades into two weak divisions, the 1st and 2nd African Divisions, later to be called the 11th and 12th. The 1st, formed of the Nigerian and 1st East African Brigade Groups, took over the coastal and Tana sector of the defences. The 2nd, formed of the Gold Coast and 2nd East African Brigade Groups, became responsible for the northern sector. At about the same time the 1st South African Brigade Group, forerunner of the 1st South African Division, arrived from the Union by sea and continued its training in the area to the north-west of Nairobi.

With an eventual offensive in view, and with many problems of expansion and organization to be solved—the supply and training of British leaders for African troops, to mention only one—General Dickinson had a formidable task indeed. His instructions from General Wavell were that he was not to assume the offensive until he was quite certain that his administrative arrangements were adequate. In 1939 East Africa had been quite unready for war, and the Bon Voisinage Agreement set a severe limit on the steps that could be taken before Italy made her intentions clear.8 Even then, the needs of Europe and Egypt left little for East Africa, and General Dickinson, with a modest staff and practically no administrative services or tradesmen or military stores or reserve of equipment of any kind, was obliged to fend very largely for himself. He was helped in every way by the Colonial Government Departments; India came to his rescue over some of the most urgent requirements, notably ammunition; but it was on South Africa that he soon came to rely for help at every turn. The Union Government were naturally interested in the country to which their forces were being sent; they understood the difficulties, and they provided invaluable equipment and skilled men of all kinds. Best of all, they supplied transport, the key to the whole problem. From early August a stream of vehicles, men and stores began to wend its way along the earthen track, or ‘Great North Road’, from Broken Hill in Northern Rhodesia, to which point they had been brought by rail from the Union. By the

Page 182

end of the year some 9,000 vehicles, mostly 1- and 3-ton trucks, many driven by inexperienced Africans, were to make the 1,500 mile trek to Nairobi over sand, stones, cotton soil, desert, and bush, up steep grades and over passes. This picturesque exploit proved to be a useful preliminary to the drive to Addis Ababa a few months later. Its importance at the time was that it met a great need without using any precious shipping space.

At the Anglo-French meeting held at Aden in May 1939 the possibility of a rebel rising in Ethiopia had been examined. There were districts where the local chieftains had never been subdued by the Italians, notably in Gojjam and to the east of Lake Tana, and it should be possible to pass arms and ammunition to them across the Sudan frontier between Kassala and the Blue Nile. It had been extremely difficult, however, to learn the true facts. The normal diplomatic channels of information had been narrowed when the consulate at Dangila, a town to the south of Lake Tana, had been closed in 1934; two others at Gorei and Maji were closed on the outbreak of the Italo-Ethiopian war; and after the Italian victory the Legation at Addis Ababa was reduced to a Consulate-General. Moreover, the Italians had been taking pains to mislead observers and conceal anything that might lessen the impression of their complete and beneficent conquest. Consequently the British and French agreed that it was essential to acquire more contacts in the country. Encouragement would not be given to native uprisings before the prospect of their success could be gauged and until they could be supported by more effective means than by propaganda or the use of inadequate forces. It was recognized that premature risings would be likely to fail in view of the military superiority of the Italians; they would do more harm than good.

In July 1939, shortly after his arrival in the Middle East, General Wavell instructed his staff to study the problem of fostering rebellion in Ethiopia, but there was still not sufficient information on which to base a sound plan. In September Colonel D. A. Sandford, who had formerly settled in Ethiopia until forced out by the Italians, was appointed to direct the work. Information received during the winter from an emissary of the Emperor and from French sources indicated that the seeds of a national rebellion existed and that there was a desire for the Emperor to return to his country to unite and deliver his people. Owing to the Bon Voisinage Agreement the work was delicate and severely restricted, but Sandford prepared a plan for fostering and supporting revolt in the event of war with Italy, and this formed the basis of General Wavell’s future policy.

In March 1940 Major R. E. Cheesman, formerly Consul at

Page 183

Dangila, was appointed to deal with Ethiopian intelligence matters at Khartoum. The names of eleven Ethiopian chieftains to be approached in the event of war were chosen, and letters were prepared containing promises of arms and money if men were sent to fetch them. In May began the discreet forming of the Frontier Battalion, Sudan Defence Force, whose role would be to garrison the small bases at which these issues would be made. On June 11th, the day after the declaration of war, the letters, signed by General Platt, were sent off on their long journey. Within a fortnight the first chief arrived with his, men and thereafter the arrival of similar bands was an almost daily occurrence at one point or another on the frontier.

In an instruction on policy issued on June 10th General Wavell laid down that the existing unrest was to be encouraged to spread as widely as possible, in order to force the Italians to expend their resources on internal security. The early activities were to be confined to isolated attacks by only small parties of patriots, culminating later in larger operations against important objectives, which would depend on the success already achieved and the extent to which arms and money could be introduced. He attached great importance to this step-by-step method, and insisted that the patriots were not to be led into enterprises beyond their powers by false hopes that they would be supported by troops. A small military mission (No. 101) was to be sent to give them advice and co-ordinate their efforts; its organization was entrusted to Colonel Sandford.

The Emperor, who was in England, had as yet no part in these plans On June 18th the War Cabinet agreed to facilitate his return and six days later he was flown to Egypt in great secrecy and haste, for now that France seemed lost any delay might make the flight too dangerous. His sudden arrival incognito in Khartoum on July 3rd was the cause of some embarrassment, since the Middle East had been given no guidance on the Government’s policy regarding him but now incurred a heavy responsibility for his safety and fitting treatment. The plans for assisting the revolt were explained to him with the suggestion that he should remain in Khartoum until the situation in Ethiopia could be more accurately judged. Although the Emperor was clearly disappointed, particularly at the little material help that could be offered to him, he accepted the position with dignity and understanding.

Closely connected with the Emperor’s arrival was the question of the departure of 101 Mission. General Platt had directed on June 21st that it was, if possible, to be established by August 1st, and that it was an no account to allow itself to be destroyed or captured, for a bad start would be disastrous. Colonel Sandford was eager to set out, but it now had to be decided whether or not the Emperor should be

Page 184

invited to accompany him. Sandford urged that he should, but there were doubts whether the moment was in all ways the right one for His Majesty to re-enter his country, and the matter was referred to General Wavell, who decided that a part of the Mission was to go into Ethiopia forthwith to reconnoitre and report on the situation in the frontier provinces to the east of Gallabat. The Mission was to stimulate the first phase of the revolt—the isolation of outlying Italian garrisons by small patriot parties—and was to influence the operations in such a way as to secure an area under patriot control safe for the entry of the Emperor. On August 12th Colonel Sandford crossed the frontier on the difficult and hazardous journey to Gojjam. On the 31st Lieutenant A. Weinholt, soon to die at the hands of Italian bande, followed him. On September 18th Major Count Bentinck led another party of the Mission across the frontier towards the north of Lake Tana.

All this time Khartoum was the centre of much activity. An invitation had been issued to all Ethiopian refugees and exiles who wanted to fight for their country to assemble in Khartoum, where they arrived in hundreds, requiring a great deal of sorting out and organizing. Eventually four battalions, with other units, were organized, requiring a large number of British officers and NCOs. On September 8th General Wavell issued another instruction outlining the steps to be taken in anticipation of a favourable report by 101 Mission. Until this report was received he would be unable to decide whether to try to accelerate the revolt at once or whether to await the moment when British troops could carry the offensive into Italian territory, thereby heightening British prestige and patriot morale. In the meantime all possible preparations were to be made so that no time would be lost.

Meanwhile in Addis Ababa the Italians appear to have thought that some British enterprise was afoot at Sarako, to the north-west of Lake Tana, but it did not prevent the Viceroy from concluding that his successes in British Somaliland and at Kassala and Moyale had ‘paralysed the possibility of an extension of internal rebellion’. By the end of September the Italians had heard that some Europeans were on their side of the frontier; and in October suspicion plays round a ‘Mr. Rhoms’, said to have once been British consul at Dangila. On December 4th it fastens upon ‘a certain Colonel Sandford’.