by Hew Strachan
The effects of Messimy’s initiative at the local level were virtually instantaneous. On 15 August itself an iron foundry at Renage in the south-east was visited by an officer from the Intendance Militaire based at Lyons: within two days the firm had received an order for 200,000 pickaxes, and within a week its workforce had increased from fifty to sixty-five, including four soldiers released from the army. Prompted by his anxiety concerning employment for those not mobilized, the president of the Paris chamber of commerce had already issued a national call, urging firms to compete for army contracts. Departmental committees to maintain employment, ranging over the entire field of economic activity and embracing the heads of local business, deputies, and trade unionists, sprang into life. That for the Isère was formed on 19 August. The principal members were the town maires, and their significance lay not least in their marriage of business and government at the local level.235
This combination—the fusion of national with local responses, and the incorporation of private enterprise in the affairs of the state—was confirmed on 20 September. Millerand, who had replaced Messimy at the end of August, summoned a meeting of the representatives of heavy industry at the government’s new headquarters, Bordeaux. He set a production target of 100,000 shells a day. To achieve this, he recalled General Louis Baquet from the front to be director of artillery. Baquet was a designer of guns, and in particular was responsible for the 75 mm high-explosive shell which had so proved its worth at the Marne. France was divided into twelve regions, each regional head being responsible for the allocation of resources and contracts. The Comité des Forges profited from its tight organization and its early involvement: under its secretary-general, Robert Pinot, it dominated the regional structure, manipulating resources, prices, and orders for the advantage of its own members. Millerand effectively bound himself over to the abilities of heavy industry to manage its own affairs. He met the heads of sectors every eight days (at least at first), but accepted their assurances rather than testing them. He promised the release of skilled labour from the army: this was ordered on 11 October. In November a central labour exchange was established to recruit for the munitions industry. By June 1916 287,000 workers had been returned from the army, and by the end of 1915 the workforce engaged in the metal industries was fully up to its pre-war level; that of the chemical industry had recovered to 93 per cent of its pre-war level. Renault’s payroll, which had peaked at 6,300 in 1914, reached 12,800 in 1915.236
At the Bordeaux meeting the representatives of industry told Millerand that they could deliver 40,000 shells per day by the beginning of December. On the following day, buoyed by their bullishness, Millerand informed Joffre that he could expect 30,000 rounds per day in three to four weeks time. But output in August had averaged 6,000 shells a day, and it had first exceeded 10,000 a day only six days previously, on 15 September. Eight days later Millerand revised his estimate downwards, informing the commander-in-chief that 20,000 rounds per day would be the norm until the beginning of November, but that the rate would then rise to 33,300 rounds. Even on 3 November Joffre was being assured that the daily rate of production would jump from 13,000 rounds to 23,000 by the middle of the month. But on the 14th he was told to expect only 18,000 shells per day. In the event, output for November remained stubbornly fixed at 13,000. The state-owned workshops were meeting their mobilization targets, but private industry could not yet fulfil its. The latter’s total deliveries on 10 November were 2,000 rounds, against the 7,000 that had been promised. Its orders for the coming six months were ten times a full year’s peacetime production. On 28 September Bouchayer and Viallet in the Isère were told to install a forge to produce 1,000 shells a day: that was their total output by the year’s end. In October the Societé des Constructions Mécaniques undertook to produce 1.2 million shells in plants at Paris and Lyons. But by the end of March 1915 it was over 800,000 rounds in arrears, and the Lyons factory had yet to deliver any.
Baquet reckoned that the corner was turned in December, others somewhat later. He calculated that the new factories produced 32,500 shells a day in the last month of 1914, as against their target of 50,000. In addition, the existing major munitions firms delivered 7,500 shells a day and the state workshops 16,000. Other accounts claim that daily production did not reach 50,000 rounds until March (when in fact it averaged 62,173), and that the production target set for January 1915, 80,000 shells per day, was not attained until July. By then the army was abundantly supplied with 75 mm shells. On 10 August 1915 the stocks under Joffre’s own control (excluding those already in the hands of army commanders) totalled 2,681,486 shells. The number had almost doubled during December and January, from 301,361 to 573,696; more than doubled in February, March, and April, to 1,202,670; and more than doubled again during the next three months.237 As in Germany, the Ministry of War, by setting objectives which were not met on schedule, undersold its own achievement.
By the summer of 1915 the problem was no longer quantity but quality. The state factories apart, in 1914 only Schneider-Creusot and St-Chamond possessed the hydraulic presses required to forge shells. Production methods, therefore, had to change in order to convert other plants to the needs of war industry. On 23 August 1914 Renault told Messimy that automobile factories could manufacture shells by boring out the interior of steel ingots, a process which would require a turning-lathe rather than a hydraulic press. But forged steel was stronger: a shell made before the war could take a water pressure of 1,400 kilos per square centimetre, whereas one made in the straitened circumstances of 1914–15 could only cope with 400 kilos per square centimetre. Moreover, the turning process could not fashion the nose-cone of the shell, and the shell therefore had to be made in two parts. Renault’s shells were nicknamed ‘bi-blocs’, as opposed to the ‘monoblocs’ produced in the state arsenals. Techniques were developed to reinforce the more delicate parts of the shell, but they mitigated the problems rather than removed them. Inspection standards were inevitably lowered in the rush to boost output. Problems with fuses had caused many French shells to misfire even at the outset of the war, when production had been the monopoly of properly constituted arms factories. At Muhlhouse the Germans found pits which the French gunners had dug to bury their defective shells. In January 1915 one German observer calculated that 50 per cent of French rounds fired in a single day were duds. Some had problems with the charges, others had fuses but no charges, and a few were even training rounds filled with sawdust.238 Moreover, the French lost over 600 guns in 1915 through premature explosions. Before the war one gun burst for every 500,000 rounds; in the spring of 1915 one gun burst for every 3,000 rounds. On 1 March 1915, with shell production now in full swing, the decision was taken to abandon the ‘bi-bloc’, and on 15 April the pre-war standard for forged steel was reapplied. However, not until late 1916 would the quality of shell production equal that of peacetime.239
Nor were guns and their crews the only casualties of the drive to increase shell production. The complexities of shrapnel were beyond the capabilities of the firms suddenly drafted into the ordnance business. The pre-war ratio of production, 15/26 shrapnel to 11/26 high explosive, had altered by December 1914 to 1/7 shrapnel and 6/7 high explosive. The construction of trenches had, of course, increased the need for high explosive beyond pre-war expectations, but in the French case the constraints of munitions production pushed the pendulum too far in the other direction: the gunners were left with insufficient shrapnel for interdiction and harassing fire.240
Most of these problems were ironed out in the course of 1915. Citroën expressed an interest in the production of shrapnel in January, and by the end of September private industry was delivering 10,000 rounds per day. Renault devised a system for manufacturing the cone of the shell under heat. Car and bicycle factories proved well suited to the production of gaines (the metal casings for shell fuses), and textile firms made fuses. Copper shortages, which threatened to curtail the output of the driving rings for shells, were mastered in part by th
e recovery and reworking of those sent back from the front. The rate of premature explosions was halved in the summer of 1915, and the problem was more fully mastered in 1916, as hydraulic presses become widespread in private firms. By the end of the war France, the biggest single producer of shells in the Entente, had 375 businesses engaged in their production.241
In France, as opposed to Germany, the pace of shell manufacture was never determined by the output of powder. As a detonator, France used a solution of gun cotton in a mixture of alcohol and ether, called powder B. The plan on mobilization was to increase the daily peacetime production of 15 tonnes to 25 tonnes by the end of the second month of the war. But the problems of shell fabrication meant that at first powder output exceeded the army’s requirements. Demand was also suppressed by a reduction in the individual charge from 0.665 kilograms to 0.620 as part of the bid to avoid premature explosions.242 Not until early 1915 did the need for powder run ahead of domestic capacity. In January France produced 50 tonnes a day, and by October 100 tonnes. The rate of shell production meant that 80 tonnes a day was needed in January and 150 in October. The deficit was made good by purchases from the United States, and by favouring shell production over that of small-arms ammunition.
More immediate was the pressure to expand explosives output, as a result of the switch from shrapnel. The French chemical industry grew at a rate of almost 5 per cent per annum between 1895 and 1913; one of its strengths was the sulphuric acid works of St Gobain. But in 1914 domestic production of explosives still remained small. Stocks, primarily melinite or picric acid, totalled 2,400 tonnes at the war’s outbreak. On 18 August Messimy was told that these stocks had been prepared for an output of 8,000 shells a day, and that there were only enough for three more weeks. Schneiderite, used by Schneider-Creusot for its exports, was employed from October 1914, but it was derived from ammonium nitrate, whose output was itself limited, and it was disliked by gunners as the colourless gas emitted on impact made fire hard to observe. By January total explosive stocks had dwindled to 700 tonnes, and the daily output of melinite was 16 tonnes, enough to load 20,000 75 mm shells. These worries were stoked by Joffre, who—as an insurance measure—pitched his expectations ahead of his likely consumption. But they were also fuelled by the nature of the business: its close links with Germany before the war generated suspicions, and they were not allayed by the high prices secured as a result of government demand. However, the government saw in the war an opportunity to foster the growth of an indigenous chemical industry, free of German influence and control. Thus, the concentration of the business in fewer hands was encouraged by the state. So too was the emphasis on sulphuric acid, whose output rose from 1.16 million tonnes in 1913–14 to 1.611 million in 1916–17.243 All in all, the production of explosives proved relatively easy to set in train. In 1914 France imported its benzol from Britain and Germany; after 1915 it was manufacturing between 35 and 40 tonnes a day. In 1914 its output of phenol was 1 tonne a day; at first this too came from Britain, but by December 1915 it was producing 60 tonnes a day. The production of chlorates, which was still minimal in December 1914, reached 40 tonnes in May 1915 and 100 in March 1916. In July 1917 France was producing 985 tonnes of explosives a month, thereby exceeding its target of 900 tonnes.244
By the summer of 1915 efforts to boost the production of 75 mm shells had proved almost too successful. Sufficient shells highlighted the lack of sufficient guns from which to fire them. Burst barrels as a result of premature explosions compounded the losses incurred in the initial retreat. Between 30 November 1914 and 1 May 1915 the number of 75 mm guns available fell by a further 468, to leave a total of 3,071. In mid-April Joffre reckoned he needed 805 new field guns; even if he had had them he would still have been short of over 700 guns compared with the number at his disposal on mobilization. Not least because the scale of loss through premature explosions was not evident until the new year, Baquet had been reluctant to place fresh orders in the first six months of the war. Schneider-Creusot and St-Chamond each received contracts for eighty guns after the Marne, but both firms accepted that the work would compete with their efforts to increase shell deliveries, and in November Schneider-Creusot was persuaded to withdraw from a contract for a further 160 guns because of its commitment to the production of heavy artillery. The 75 mm manufactured by the private firms was regarded as inferior to that produced in the state ordnance factories. But by the time the problem became urgent, in February 1915, the state workshops were largely taken up with repair work, replacing barrels on guns damaged by premature explosions. Thus, private industry received the lion’s share of the contracts for new guns, 500 to 100, but the state factories also repaired a further 458 guns. The situation at the front did not improve until the autumn. By 1 October 1915 the armies had 3,524 guns, and production in the last quarter of the year for the first time exceeded its pre-war level of 500 guns to reach 700. In the first quarter of 1916 1,000 guns were completed, almost as many as in the whole of 1915.245
In the first year of the war the position in regard to heavy artillery was even more desperate.246 In 1911 the Conseil Supérieur de Guerre favoured long 105 mms in the field howitzer role, as they were mobile and ranged 10 kilometres. But the gunners, Baquet among them, protested that at that range, with a light shell, the fall of shot would be at such a low angle as to be hard to observe and correct. They favoured the 155 mm, whose shell could contain 10 kilos of high explosive. The tendency in the debate was therefore for heavy artillery designed for the field either to be pulled in a direction where its performance was sufficiently akin to the 75 mm as to make the latter a satisfactory substitute, or to align it with siege artillery and so jeopardize its mobility. When the fighting began the tactical arguments at least spawned routes out of the impasse, particularly in the employment of the 75 mm—an example of ideas as alternatives to material inferiority. But the immediate result of deep thought for actual armament was stasis. In 1914 France’s heavy guns were almost all antiquated and of limited range; few were quick-firers. Only about 300 were ready to take the field. Joffre was able to circumvent the immediate crisis by plundering fortresses, coastal defences, and the navy. This was a slow process: their carriages had to be adapted, harness had to be procured, and teams of horses—most heavy guns required eight and the heaviest ten—assembled. Even by 15 November 1914 only 724 heavy guns had been put into the field. Few were howitzers. Numerically the most significant was the long 120 mm Bange gun, made between 1877 and 1882, of which there were ultimately 2,400 available. Also present in large numbers was another siege gun, the long 155 mm: there were 1,400 of these. Having been designed for static warfare, they remained ponderous in their deployment and could take the best part of a day to bring into action. Lacking manoeuvrability, punch, or high rates of fire, many in the French army remained sceptical of heavy artillery while the war remained mobile. Of the short guns, there were only 200 of the 120 mm and 400 of the 155 mm. The latter included 100 quick-firing guns (the Rimailho model of 1904). Orders had been placed for 220 105 mms, due for delivery in 1914–15, and 120 long 155 mms, due in 1916–17.
The shells crisis, followed by the pressure to rebuild the stocks of field guns, had knock-on effects in the procurement of heavy artillery. When GQG felt able to move its attention from the former to the latter, in April 1915, it opted to concentrate on the types of gun already in use so as to maximize existing machine tools. The effect was to postpone any move to serial production until 1917. In the first year of the war the only new heavy guns received by the army were 110 long 105 mms from Schneider-Creusot. In the second half of 1915 Schneider-Creusot delivered a further seventy 105 mms, and St-Chamond added forty short 155 mms: the total output in all calibres was 140. Even in 1916 deliveries only reached 245 in the first half of the year and 345 in the second.
The lack of heavy artillery under Joffre’s command in 1914 delayed the onset of shell shortage for these larger calibres. Although consumption was high during the battle of the Marne, t
he stocks in the eastern fortifications averted crisis. But most of these were cast iron, as this had been deemed sufficient for defensive purposes in the 1880s. They carried, at most, a third of the high explosive of a steel shell and, with a weaker charge, they generated less smoke, making the fall of shot hard to observe. Total daily production of steel shells for all calibres of heavy artillery was about 1,000 rounds for the first three months of the war, rising to 2,000 in December. By then Joffre wanted 3,000 shells a day for the 155 mms alone, when the planned production on mobilization was 465 and actual output in December 400. In January the problems with the 75 mm increased the army’s reliance on the 90 mm and 95 mm guns as substitutes, and as a result Joffre wanted 6,000 rounds a day for the former. The expedients used in the production of 75 mm ammunition were not appropriate for shells of larger calibres, and Baquet’s short-term response was to revert to cast iron. Although massive orders for steel shells were placed, 660,000 rounds for calibres between 90 and 105 mm, 340,000 for 120 mm and 155 mm, and 50,500 for 220 mm and above, they were way beyond the immediate capacity of the contractors. The retention of steel for heavy artillery shells restricted their manufacture to the major armaments works already equipped with hydraulic presses; it also prevented the same plant from being used for 75 mm shells, and so indirectly increased the number of premature explosions in field guns. A process to harden cast iron with steel was developed late in 1914, and production of shells made in this way began in March 1915. On 5 July 1915 Joffre suggested production targets of 3,000 shells per day for the 220 mm, 12,000 for the 155 mm, 10,000 for the 120 mm, 4,600 for the 105 mm, 10,000 for the 95 mm and 12,000 for the 90 mm. But the previous month’s output for each of these calibres had been respectively 100, 2,500, 3,000, 1,500, 1,500, and 1,000. Joffre’s figures would not be reached until March 1916, and then only in part.247