Légionnaire, Madagascar, 1900 The soldier illustrated is serving with one of the companies which returned to Madagascar to put down prolonged rebellions at the turn of the century. In fact the only way in which he differs from those involved in the original invasion is in the gaiters; the short, laced black leather model was first issued in 1900, prior to which the trousers were simply gathered at the ankle and tied, or allowed to fall free over the ankle-boot. The humidity and heat of this region led to a general adoption of shirt-sleeve order in the field, although the steel blue-grey capote is carried on the pack, as are pots and pans, spare boots and bivouac poles. The Negrier pouch is still worn, together with the two leather belt pouches for Lebel ammunition. The white shirt and trousers of coarse linen or canvas are standard Legion fatigue issue.

Other units had fared far worse—the 200th Infantry Regiment, formed of conscripts, most of whom had been drawn by lot from twelve designated metropolitan infantry regiments, had virtually ceased to exist. Of 800 Chasseurs à Pied recruited in part among French Alpine troops, only 350 were able to hold a rifle, barely. Yanked from their worksite and thrown into an attack against Hova positions on June 30, a great many men simply fell exhausted on the ground. “Half of the unit had to be evacuated,” wrote Reibell, “and the rest is anemic, spent, can never, with the exception of a few men, press on.” Engineering units, used to build, or rather attempt to build, bridges over the swamps had been destroyed—when Duchesne saw ten soldiers led by three officers marching back from a bridge worksite, he is alleged to have shouted, “That’s a lot of officers for so few men!” That, however, was what remained of a company of engineers. Only 20 of the original 150-man contingent of Chasseurs d’Afrique were able to mount a horse. The 13th Marine Infantry Regiment, 2,400 strong when they disembarked at Majunga in March, was reduced to 1,500 men.

As the Dahomey campaign had shown, by far the most resistant troops were the native formations. The ragged but resistant régiment colonial formed of Hausas and Malagasies and volunteers from the island of La Reunion counted only 600 casualties, most of them Réunionais. The prize for resistance, however, went to the Algerian tirailleurs, who had lost only 450 of their original 1,600 men.28 Reibell reported that the general staff was singing the praises of the Algerians. It was a pity that they had not included more of them in their plans. And despite the praise lavished upon the Algerians, Lentonnet complained that the Legion was repeatedly favored over his Algerians in the distribution of supplies, which “produces a bad impression in our ranks.”

However, if the Algerian tirailleurs were winning the praises of the staff, the Algerians recruited to work as mule and wagon drivers soon found themselves bearing the brunt of the blame for the poor conditions on the march. This was in part a logical extension of their association with the Lefèbvre wagons, which the men had come to see as the symbol of their misery and the cause of so many deaths among their comrades. Langlois’s legionnaires nicknamed them “La Fièvre” (fever) wagons, and greeted their arrival in camp with “the most unanimous and energetic shouts and catcalls. We hate them, these miserable vehicles. . . . Some soldiers even spit on them, as if [the wagons] could understand and suffer from the abuse with which they are covered.”

They also began to hate those who looked after them, a tendency that was encouraged, according to Reibell, by some staff officers who cited deliberate sabotage by the Kabyle conductors for the expedition’s difficulties rather than their poor planning and organization. The Kabyles were accused of stealing supplies destined for the front and even of making the return trip of the evacuated soldiers piled in the bins of the Lefèbvre wagons so miserable that many died. No doubt some of these accusations were true. However, it was obvious that the logistical planning of the expedition had been so haphazard as to be almost criminally negligent. As demonstrated in Tonkin and Dahomey, porterage was a crucial element in the success of a campaign, as critical as the combative qualities of the troops. In Dahomey, Dodds had assembled 1,858 porters for a force of 1,366 men, and given strict orders that they “be treated without brutality,” be paid regularly and carry no more than thirty kilos. Each soldier was assigned a porter who carried his pack. And while these orders were not always followed to the letter, especially as the campaign became more difficult and more porters attempted flight, at least they sustained the Dahomey column well in the opening weeks, when it virtually destroyed the Dahoman army.

By contrast, in Madagascar the requirement for porters and drivers had been desperately underestimated, so that, even with the Lefebvre wagons, soldiers had to carry their own packs, which added to their fatigue. Furthermore, planners had placed the Algerian porters and drivers low on their priority list—even on the ship out from Algeria, Reibell discovered that desperately inadequate arrangements had been made to feed the Kabyles. Once on shore, they had been expected to work all day and then virtually fend for themselves. “The Kabyles, who have been worn out and abused, are at the end of their tether,” Reibell recorded in September. “The hospitals refuse to receive them on the pretext that there is no room. No one looks after them even though they have the worst job to do. They leave before first light and arrive at the staging hut after midnight, battered, lacking food, sores on their feet and legs.” “Among these poor, ragged, pitiful devils, I was amazed to see a child perhaps ten years old at the most,” Legion Lieutenant Langlois wrote in August. “Like his comrades, he bravely led his mule, stretching out his little legs to follow the animal.”

Even Duchesne began to realize that, despite the arrival of reinforcements, his force was wasting away to the point that soon it would be incapable of combat. According to his report to the French parliament, on August 4 he took the decision to create a light column to press forward to Tananarive. Nevertheless, this column could not be created, he believed, until the road had reached Andriba, which would serve as the staging area for the projected offensive. In early September, the road, such as it was, was declared complete, and the light column began to form at Mangasoarina on the plain of Andriba. Three hundred fifty tons of supplies were collected there, and reinforcements, including 150 legionnaires, joined the force: “The most impressive of all these relief detachments was that of the Legion,” Reibell, who had seen them disembark at Majunga in mid-August, wrote. When, on September 12, Langlois witnessed the review of the men, including a battalion of the Legion, whom Duchesne had selected for the light column, he was less impressed. There, surrounded by the naked red mountains of Madagascar, 1,500 men stood to attention,

so dejected, so depressed, so pale, that one would have believed them more dead than alive. Their clothes were in rags, their boots in pieces, their helmets, too large for their emaciated heads, fell to their shoulders, covering almost entirely their yellow faces where only eyes the color of fever seemed to exist. And they seemed so pathetic, so poor, so miserable, that unconsciously tears came to the eyes.

On September 14, the light column left Andriba. In the van was the régiment d’Algérie, made up of two battalions of Algerian tirailleurs and a battalion of the Legion. A second group had at its nucleus the aptly named régiment mixte, which contained a battalion of the 13th Marine Infantry Regiment and some Malagasies and Senegalese. In the rear, the remnants of the 200th Infantry Regiment and some more Africans and marines guarded the convoy. In all, 4,013 soldiers led by 237 officers and followed by 1,515 mule drivers and 2,809 mules left for Tananarive. Some pessimists were already calling this march “the suicide of General Duchesne,” but Langlois’s legionnaires were already dreaming of the riches to be found in the capital. Nevertheless, it was apparent as the column made its way off the plateau of Andriba and up the narrow valley of the Mamokomita that the countryside of rocky, naked slopes rising to jagged summits would be difficult enough to march through, much less fight over if a determined enemy chose to defend it. Therefore, it was with great relief that the column, after a march of seven miles, arrived at the foot of Mount Tafofo at the end of the Mamokomita defile to find it undefended.

The view from the summit was glorious—three large mountain massifs separated by valleys whose floors, made green by marshes, cacti and mango trees, contrasted with the stark, boulder-strewn slopes. However, hardly a mile and a half away a Hova army was in the process of throwing up fortifications along a ridge line near the village of Tsmainondry, which blocked the column’s projected line of march up the Firingalava valley. At dawn on the following morning, Duchesne launched his troops into the attack. The task of the Legion was to carry out a frontal assault, wading through the marshes that covered the valley floor and attacking up the slopes, while other units flanked the positions. However, hardly were they through the marshes than the Hova artillery opened up. Fortunately, few of their shells exploded. The legionnaires began to fire on the offending batteries from over two thousand yards away, inaccurate from such a great distance, but enough to send the Hova troops dressed in their white lambas, a shawl-like garment, scattering like the flocks of aquatic birds that the legionnaires had raised while wading through the marsh. “One of the defenders, probably believing that his resistance had been sufficient, got up, looked around, and bolted off with no self respect,” wrote Langlois.

The entire trench, worried, became agitated. Two or three other men got up, looked about, and…, like their comrade, took French leave. That proved too much for the delicate morale of the intrepid warriors of Ranavalo. As one man, like those jack-in-the-boxes pushed out on their springs, they suddenly surged out of the back of the parapets and disappeared rapidly down the thousands of ravines which furrowed the terrain, throwing away their arms so as to run faster. We saluted this grotesque flight with the most energetic catcalls.

Yet the undignified retreat of the Hovas before Tsmainondry, while gratifying, simply allowed the French to march on to the next, and many believed the most formidable, obstacle of the campaign, the Ambohimenas mountains. The route of the column climbed the valley of the Firingalava river, wading through torrents that spilled down mountainsides so steep that only with difficulty could the mules be kept from tumbling into the water below. The fatigue of the men was obvious, and a wake of stragglers trailed behind the advancing French who, on the 17th, could see the great mass of the Ambohimenas, which lay across their path in the distance. Duchesne made camp to allow the remainder of the plodding column to catch up, and on the 18th ordered a night march to approach the Hova defense lines. The men ate their meager rations and then bent into a path that led straight up the mountainside. “The men . . . advanced with great difficulty,” according to Langlois. “We were obliged to use clubs to make these poor feverish men march. We struck them with a heavy heart, but absolutely convinced that our duty and their interest commanded it. All who remained behind are lost men.”

When dawn broke, the Hova fortifications were visible atop the high ridge to the front. The régiment d’Algérie was given the task, once again, of mounting a frontal assault, while the régiment mixte climbed the mountain to turn the Hova positions on the flanks. As the legionnaires approached, puffs of black powder smoke appeared from the heights, but the defenders were firing from a distance too great to cause any damage. The Hova artillery soon joined in, but it shot badly. Even though the legionnaires and tirailleurs were still over a mile away, the line of Hovas in their white lambas lower down the slopes rose up and began grappling their way toward the fortifications on the summit, sowing panic. Duchesne ordered his legionnaires and tirailleurs forward toward groups of men milling around some of the fortifications that had yet to be abandoned, when the appearance of the marines and Malagasy tirailleurs, recruited among Sakalave tribesmen, traditional enemies of the Hovas, on the ridge to their flank precipitated a complete panic. “Despite the fact that our troops were beginning to become accustomed to the speed with which the Hovas ordinarily break off the combat,” Duchesne reported, “they had the profound surprise to see them abandon completely their formidable positions and beat a retreat along the entire front.”

The French now stood on the threshold of the Hova heartland. From the naked heights of the Ambohimenas, they could look out over a spread of plateaus covered by a checkerboard of rice paddies and small villages surrounded by hedges of cactus. All that lay before them now was a few days of hard marching. According to E. F. Knight, the Malagasy commander-in-chief had told his prime minister that “I can do nothing. My men will not stand. They run away as soon as they perceive that two or three of their friends have been killed. Nothing will stop them.” An English officer trying to put some backbone into the Hova army told him that, in their panic to get away from the French attack on the Ambohimenas, fully 300 soldiers had tumbled over a precipice and been killed. Soldiers had bribed their officers to run away “so as to have an excuse for saving their own skins,” which had reduced the army, which had numbered around 7,000 on September 12, to 1,313 starving men by September 23. To be fair to the Hovas, on the Ambohimenas their soldiers probably had shot up what little ammunition they had been given while the French were still well out of range. The Hova government was calling up miserable peasants and even prisoners, who passed through the streets of Tananarive shackled by the neck, to carry out a last-ditch defense of the capital.

However, the ability of the French to take advantage of Hova confusion and demoralization was impaired by their own condition, which verged upon utter collapse—the light column was barely in a fit state to lurch forward, much less administer a knockout punch. Langlois believed that the mere effort to climb the Ambohimenas had cost the attackers a tenth of their strength. Patrols sent out returned only with corpses already stiff with death, victims of the night cold of the mountains, or exhausted by the intense heat of the day. Several had obviously committed suicide. Indeed, suicide had already manifested itself in the Legion—on September 3, Langlois’s company had marched out of camp past the body of one of their soldiers who had hanged himself from a mango tree. In the wee hours of the morning of September 21, a legionnaire shot himself through the head with a bullet that then pierced the colonel’s tent, narrowly missing him. On the following day, a second man in his company shot himself just before dawn: “Our excellent paymaster … is of the opinion that the men are intentionally killing themselves just to force him to write out the death certificates, a work which he claims to detest,” wrote Langlois. Lentonnet noted on July 27, even before the light column set out, that “Suicides are very rare in the regiment of Algerian tirailleurs. But they are more frequent in the Legion. In three days, there have been no less than six legionnaires who have killed themselves. Yesterday, one hanged himself, another blew his brains out.” On September 21 he again recorded that “suicides are more and more frequent in the Legion, which is becoming demoralized.”

Langlois agreed that morale was seriously low: “In the silent camp, the men seem discouraged and without energy,” he wrote. “It is the first time since the beginning of the campaign, that I have noticed such an utter demoralization.” Low morale, a great danger in any unit, was thought especially serious in the Legion. Legionnaires, because of their pasts, were believed to be psychologically fragile, apt to give themselves over to expressions of despair, of which suicide was one. It is possible, as some believed, that for many legionnaires enlistment had been an alternative to suicide. Raimond Premschwitz and Flutsch both admitted that they had considered it, and that the Legion, for them, had been an option of last resort. Frederic Martyn also recommended it for those who were contemplating suicide, because “it may possibly introduce you to a zest of life that you have never felt, and in any case you can commit suicide just as well in Algeria, you know, as you can in London.” But service as an antidote to suicide might not work in every case. Jacques Weygand, who commanded a Legion cavalry squadron in the 1930s, believed legionnaires particularly susceptible to “crises, whims and depression.” However, the theory, in any case, was that the active life of the Legion kept these under control. Suicide, like desertion, hit its peak in calm garrisons where “they must look themselves in the mirror, and it’s not happy.” For this reason, Legion officers were always careful to keep their men busy and occupied, to prevent unconstructive brooding over the past, “to distract their men, to tear them from the mortal prostration which grips the best ones.”

Therefore, two perceptions seem prevalent about suicide in the Legion. The first is that the Legion had a relatively high suicide rate, at least higher than that of other corps. The second is that suicide appeared to occur more frequently in garrison because legionnaires, predisposed to depression in conditions of enforced boredom and inactivity, rose to the occasion when confronted with the challenges of a campaign. Neither of these is possible to resolve with any certainty. French historian Bernard Savelli concluded in his thesis on the Legion in Tonkin before World War I that suicides in the Legion at that time averaged only three or four a year, and the rate was no higher than those of other corps, including the Algerians. But as has been seen, Indochina, at least after 1885, became a garrison of choice, with enough distractions to adjourn thoughts of suicide except among those most inclined toward self-destruction. No comparative statistics exist for Algeria, which was a far less attractive garrison. However, Savelli seemed to agree with the second perception, that suicides declined on campaign, citing as his evidence that only one legionnaire committed suicide during the grueling, and apparently hopeless, defense of Tuyen Quang. Likewise, the regimental diaries for Tonkin in 1885 and Dahomey in 1892 make no mention of suicides, nor do the memoirs mention any.

The Legion battalion in the light column counted six suicides, and it appears that eleven of the sixteen suicides that occurred in the Algerian regiment during the campaign were in the Legion. And while this is certainly not catastrophic, other suicides might have passed unnoticed among the thirty-three legionnaires listed as “missing” from the light column. There are perhaps three explanations of why Madagascar appears to have produced the exception to the general rule that suicides in the Legion declined on campaign. One, of course, is that the evidence is too fragmentary to draw a firm conclusion, that there were suicides in Tonkin in 1885 and during the Dahomey expedition of 1892 that were not recorded. However, the common sense explanation would suggest that suicides in those two campaigns were too rare to be considered worthy of mention.