Vichy French officials did not capitulate on the island of Madagascar until November 1942, the same month the Allied landings during Operation Torch were conducted in North Africa. Here, British troops land at Diego Suarez during the effort to gain control of the port city’s facilities.
Vichy French forces on Madagascar included both French soldiers and colonial troops from Senegal and other locations. In this photo, colonial soldiers fire a canon under the watchful eye of a French officer in 1942.
British operations on Madagascar encountered stiff resistance from the Vichy French, including a series of reinforced strongpoints that were assaulted during a period of heavy fighting.
The control of Diego Suarez was the crucial part of the operation against Madagascar. It was the most important place from the strategic point of view, since holding it denied the port to the Japanese fleet, and to its possible use by German or Italian submarines. It was also the first step in the campaign to suppress the Vichy administration of Madagascar. Some may well have seen it as all that was necessary, given the relative naval unimportance of the rest of the island by comparison. It was also, of course, hoped that Governor-General Paul Annet might crumble and change sides. Certainly there was no danger of the British in Diego Suarez being dislodged by the local Vichy forces, though a distant threat from the Japanese fleet might still exist – but without eliminating Somerville’s Eastern Fleet first, the stretch right across the Indian Ocean was too much, even for Nagumo’s ships.
The 5th Division was now urgently required in India, for Burma had been almost completely conquered by the Japanese army and an invasion of India looked probable. Thirteen Brigade, with sickness beginning in the troops, sailed on 20 May. Five days later Madagascar was transferred to the East Africa Command, under Lieutenant-General Sir William Platt, who had very competently commanded the liberation of Ethiopia the year before. His first task was to get 17 Brigade also on its way to India; he sent two battalions of the King’s African Rifles across to the island from Mombasa, and 17 Brigade left on 20 June. In the meantime the area under British control in northern Madagascar was expanded, partly for security reasons, in case the Vichy forces in the rest of the island tried a counter-attack, unlikely though this was, and partly to control the area from which Antsirane could be fed.
The overall purpose of the operation had been to pre-empt any Japanese seizure of Diego Suarez. While it was going on the Vichy Premier, Pierre Laval, had asked the Japanese to occupy the island in order to exclude the British, a neat reversal of the former Vichy idea of asking the US for help to exclude Japan, but even before that it had been made clear to Governor-General Annet that he should allow Japanese submarines to use Diego Suarez. This was in late April 1942, before the French anywhere knew of the British expedition: it was, that is, a gratuitous offer clearly founded on hostility to Britain, and presumably to the United States as well. (This offer was made in the aftermath of the raids on Boulogne-Billancourt and St Nazaire, when Laval believed he had some popular support for his collaborationist policies – certainly there were some anti-British demonstrations in the wake of the landings.)
The British capture of Diego Suarez harbour was thus fully justified since it was about to become a hostile base. On 29 May the Japanese submarine I-10 sent her aircraft to fly over Diego Suarez to locate the British naval force; the next night two midget submarines from I-16 and I-20 went into the bay and attacked the battleship Ramillies and the tanker British Loyalty. Ramillies was badly damaged and the tanker was sunk. (Two Japanese sailors were captured a couple of days later, and papers on them confirmed their mode of attack.) The submarine sent its aircraft over the harbour again the next day to check the results. None of these air reconnaissances seem to have been noticed, still less intercepted, by the British aircraft.
The Japanese attacks were not a response to the appeal by Laval, but their presence may well have been to investigate the possibility of French hospitality at the port. The submarines involved had been touring the Indian Ocean looking for the British Eastern Fleet, though they never did find it. They had investigated the entire East African coast from Djibouti to Durban and even Simonstown in South Africa. Five submarines were involved in the search, and between them they sank twenty-two ships. For the British, however, the attack on the Ramillies was confirmation that Madagascar would have been vulnerable to a Japanese landing.
At the same time it was obvious that such an event as the Japanese occupation of Diego Suarez would hardly take place out of the blue. Madagascar was scarcely the first Indian Ocean target for Japanese forces, and the patrols by hostile submarines or even by German auxiliary cruisers were by no means new or unexpected in the Indian Ocean. If an expedition aimed at Madagascar set out it would in all probability be noticed, and to reach its target it would need to sink the Eastern Fleet first, by which time it would itself be damaged. On the other hand, if it did get through, now that there was a British military presence on the island, and a British naval presence at Diego Suarez, it would probably be welcomed by the Vichy authorities at some other port on the island, especially given the earlier instruction for Vichy to be hospitable.
So, while there was no particular urgency about further conquests on the island, it would clearly be helpful to the war effort for the British to have control of the whole island in a fairly short time. This also raised the issue of who should control the island once the Vichy authorities had succumbed. The British had no real wish to keep it, but to hand it over to the Gaullist Free French was only marginally more acceptable than to leave the Vichy government in charge, and might well stimulate a stronger Vichy resistance; not that either of the French groups believed British protestations of a lack of interest in the island in the long term; de Gaulle was, or claimed to be, convinced that part of Britain’s war aims was to take over the French Empire; Vichy said the same thing, but with more fervour in that it was Vichy’s Empire which was being demolished. If Annet and his officials could be brought to agree to accept some sort of detailed British supervision, that would be acceptable. Free France was not going to be given the island on a plate, considering the long difficulties and arguments which had resulted from the similar situation in Syria and Lebanon over the past year. So the British wanted Gaullist concessions on Syria before handing over Madagascar – that is, Madagascar was dangled as a carrot before the Free French: accept the British terms on Syria and they would get the great island; the stick was that without concessions they would gain neither Madagascar nor Syria.
But there was still another consideration. Free France had been kept out of the planning and execution of the Madagascar expedition because of its bad security. This had in fact proved to be a sensible decision, and the arrival of British forces and their landings at Diego Suarez had come as a complete surprise to the Vichy regimes in both France and Madagascar. (At Vichy the government learned of the British landing in a message from President Roosevelt.) And now an even greater and more important expedition and landing in a different Vichy territory was in its final planning stages. This was Operation Torch, the Anglo-American landings in French North Africa. It was of infinitely greater importance to have this expedition kept secret and made successful than that of Madagascar.
This consideration had its effect also on the situation in Madagascar. If the Vichy French regime in France thought that their administration in Madagascar would be maintained, in however subordinate a position to the British conquerors, they might be induced not to oppose the North African landings very strongly. If they saw, on the other hand, that the British quickly handed over the island to the Free French and its traitor leader de Gaulle, and dismissing or even interning Vichy’s faithful officials, they might be so incensed that their opposition to the North African landings could be intensified. So until the issue in North Africa was determined, Free French control of Madagascar would need to be delayed. At the same time the Free French could not be told any of this, because of their past security lapses. So the Syria-Madagascar linkage proved very useful, not just in hopefully promoting a deal over Syria, but in distracting the Free French away from North Africa. It also meant that there was no urgency in Madagascar, for so long as fighting was going on the British had a good excuse to delay any political decisions.
In Madagascar the linkage with Syria was not visible, any more than was the prospect of landings in North Africa. Contacts between the Vichy administration in Tananarive and the British commanders in Diego Suarez began soon after the landing had succeeded. The intermediaries were Captain Fauché, Governor-General Annet’s aide and military intelligence officer, and Leslie Barnett, the representative of the Vacuum Oil Company of South Africa in Tananarive, who was presumably in the city at the time of the invasion. Annet was intent on preserving as much of the island under his control as possible, and on maintaining his control over its administration, and so he appeared to be offering a quasi-acceptance of the British position; the British commanders did not really wish to embark on a conquest with the relatively weak force they had on hand after 5th Division and the big ships had left. So both sides thought they were playing for time, and stringing the other along, while blaming the other for doing so. The War Cabinet in London was quite content with the stalemated situation, though eventually it was the British intention to deliver the island, all of it, to the Free French, once its usefulness as a bargaining chip and distraction was ended. Meanwhile no Free French representatives could get anywhere near the island because the British controlled their transportation.
The changeover of the British forces at Diego Suarez took place over a period of two months, for the British were really in no hurry, and partly because of the shortage of shipping. The campaign now became a largely African affair. Apart from the British 29 Independent Brigade, the specialists in opposed landings, and 5 Commando, the rest of the British forces present were the 27 Brigade of the King’s African Rifles, with battalions from Kenya, Tanganyika, and Nyasaland, the 7 South African Brigade, recruited mainly from the Transvaal, and assorted artillery, engineering, and other units. On 11 August the overall commander, General Sir William Platt, was given permission from London to begin a campaign to conquer the rest of the island. Again speed was hardly of the essence, but thorough planning was. Late in August 29 Brigade was taken to Kenya for further training. The 1 City Regiment of the 7 South African Brigade (the ‘city’ was Pretoria) was also given rudimentary training in landing from the sea at the island of Nosi Mitsio off the north-west coast, beginning on 4 September. They had to use dhows, not the most convenient vehicles for the purpose, but all that was available.
The night of 9 September was designated for the next forward movement. A new brigade, 27 Northern Rhodesian, had arrived in a convoy at Diego Suarez late in August. That same convoy was now to be used to collect the East African Brigade and bring it to the landing place, with the hope that the enemy, whom it was reasonably assumed had good sources of information in Diego Suarez, would think that this was a process of routinely exchanging brigades. The Eastern Fleet once more provided a substantial covering force, including the carrier Illustrious, the cruisers Birmingham and Gambia, and the Dutch cruiser Jacob van Heemskerck, plus three British, one Australian, and two Dutch destroyers.
No less than five separate operations were to begin at the same time on 9 September. In the north the 1 City Regiment began its march southwards on a rough road from Diego Suarez along the west coast, while one company of the regiment made the landing they had practiced for at Antanambao in advance of the main body. Eight armoured cars of the Pretoria Highlander Regiment, a field battery, and part of the 88 Field Company (engineers), were with it. This set of forces – armoured cars, some guns, infantry, and some engineers, was to be the norm for any force which set out to campaign in Madagascar. It took the force two days to move along the road and join the landing force at Antanambao. The road was basically of sand, and went through mangrove swamps at times. Physical progress was therefore slow and laborious. There was only occasional opposition from Vichy forces, but those forces did carefully destroy every bridge along the road, and planted roadblocks as well. The movement of the northern force therefore depended mainly on the speed with which 88 Field Company could lay its one box-girder bridge over a waterway where the original bridge had been broken, get everyone across, then pick up the bridge and move it on to overcome the next obstacle the infantry had encountered. Occasional snipers were the other real obstacle – apart, of course, from the active and numerous mosquitoes and the high sickness rates these produced among the white soldiers.
At the same time a company of the regiment moved by land across to the east coast, where there was a road of sorts, rather better than that on the west coast, connecting the coastal towns and villages. Progress was reasonably good for the first two days during which a hundred miles was covered as far as the village of Vohemar. But the road deteriorated, and from then on culverts and bridges were regularly broken. It took another nine days to go the next hundred miles to Sahambava. After that just one more village was to be reached, but this campaign was not going to win the war.
The island of Nosi Bé, off the north coast, was attacked before dawn on 9 September, preceded by a bombardment by the minelayer Manxman. Then the landing by part of the Pretoria Highlanders and some Royal Marines captured the local town of Hellville. The island was in British control by noon, with the few uncaptured Malagasy soldiers coming in to surrender voluntarily.
This was the first of five landings at different places which took place on 9 and 10 September. The main landing was to take place at Majunga, 200 miles south of the operations at Nosi Bé and Antanambao. This was an important port at the mouth of the island’s main river, the Betsiboka, and from the town a relatively good road ran through to the capital Tananarive. Majunga also had an airfield, and when this was taken there would be no Vichy air capability north of the capital. The force to be used in the landing was, of course, 29 Brigade, coming directly from its training in Kenya, together with 5 Commando.
Landings were made at three places, one of them some miles north of Majunga and one in Majunga itself. The third was to take place south of the harbour, where it was thought there was a coastal battery; 5 Commando carried out this part of the operation, but there was no battery. The commandos went on inland to secure a bridge thirty miles along the road to block the arrival of any force which might come from inland to interfere. None did.
The main landing was the northern one, a few miles from the town, undertaken by the East Lancashires and the Welch Fusiliers. There was little resistance and by daylight on 10 September they had moved inland and had reached the road which led to the town. The landing in the town itself was by the South Lancashires and again they met some resistance which ceased when the local garrison commander was captured and immediately ordered his men to cease fire; he then toured the town with a British officer to ensure that various separated groups of his men stopped fighting. The East Lancashires captured the airfield, and had been about to attack a Vichy position north of the town when the potential defenders were alerted that the fighting had ended.
The fifth landing was by a single troop of 5 Commando, who landed by boat from the destroyer Napier at the small port of Morondava almost 400 miles south of Majunga. This was also the terminus of another road from the capital to a port, but the object of the landing – which was by only a few men, after all – was to cause a distraction, like Hermione’s antics at the battle of Diego Suarez. They landed in daylight, met no opposition, occupied the town, and sent a party inland supposedly to mark out the billets for a larger force, meanwhile carefully being careless to ‘reveal’ that a larger force was due to arrive. The absence of opposition seems to have been accompanied by an absence of local alarm at the attack, so they themselves had to telephone the capital to report the landing of a large British force. Hoping to have distracted the government in the capital and to have caused the dispatch of a force which might have gone to Majunga down the road to Morondava, they then withdrew. It seems unlikely that anything was achieved, for the official in Tananarive who answered the phone had said that they could do nothing to help.
The 29 Brigade was used just for the initial landings at Majunga, and not even all the men had been landed by the time the fighting in and around the town ended. In the rest of the convoy was the East African Brigade, who were landed over the next few days while 29 Brigade was withdrawn. The change having been completed, a curiously constituted army began to advance along the road from Majunga to the capital, over 250 miles away. A squadron of South African armoured cars manned by Afrikaners from the Pretoria region of the Transvaal was accompanied by successive battalions of the infantry of the King’s African Rifles, recruited from various parts of East Africa. The infantry was intended to be moved in trucks where possible, but this turned out to be very optimistic. The first objectives were two bridges, over the Kamoro and the Betsiboka Rivers. The first, ninety miles along the road, was reached and crossed by 4.00pm, but on the succeeding part of the route they met delays in the form of plenty of roadblocks, so the Betsiboka Bridge was not actually reached that day. When the advance resumed on the morning of 11 September they found that the bridge cables had been cut and the bridge itself had been tumbled into the river, though it proved to be relatively easy to get across the next day.
The same tactics had thus been employed as in the north: breaking down the bridges, occasional snipers opposing the advance, and frequent roadblocks, which were clearly rapidly improvised. They could also be fairly quickly cleared by recruiting local Malagasies, who were often actually the same people who had blocked the road in the first place on French orders, but it always meant that the soldiers had to disembark and deploy. Often they had to drive out the snipers, and sometimes had to cross the rivers under fire, before the bridges could be restored. This was all somewhat annoying and had held up the advance considerably, as it was intended to. At the same time it was not clear whether this was all a process of drawing the British forces ever deeper into the interior as a prelude to mounting a more determined resistance, perhaps by a series of ambushes at the broken bridges or at particularly large roadblocks when the British had outrun their supplies and support. It was therefore necessary to advance with some care. It was slow and laborious, as in the north, but progress was maintained.