I. A Premature Engagement

Exactly at noon on Saturday, September 5, the divisions of General Lamaze, constituting the right (save for elements connecting it with the British) of the French 6th Army, came under fire from advanced posts of General Schwerin’s IV Corps of Reserve, hidden on the wooded hills just beyond the highroad from Dammartin to Meaux. A surprise for both sides; and with this began the battle of the Ourcq.

The battlefield—a rough quadrilateral, extending from the Dammartin road eastward to the deep ditch occupied by the Ourcq and its canal, and bounded on the north by the Nanteuil–Betz highway, on the south by the looping course of the Marne—consists of open, rolling beet- and corn-fields where some part of the crops were still standing. A soldier would call it an ideal battlefield, its many and good roads helping the movement of troops, its wooded bottoms and the stone walls of its farmsteads and hamlets giving sufficient cover, its hills good artillery emplacements. The eastern and higher part of the plateau is crossed from south-east to north-west by three ridges, against which the French offensive beat in successive waves. The northernmost rises to 300 feet above the Ourcq, from near May-en-Multien, along the little river Gergoyne, by Etavigny and Acy, to Bouillancy; the central ridge, that of the Therouanne, runs from opposite Lizy-sur-Ourcq, by Trocy and Etrepilly, to Marcilly; the southernmost from Penchard, through Monthyon and MontgÉ, to Dammartin. The combat, as we shall see, began in the last-named area, its centre of gravity then moving northward. The Germans had the better of the hill positions, with forward parties well spread out; and, as in Lorraine and the Ardennes, directly they were threatened they entrenched themselves, though not continuously or deeply. Caught in full movement toward the Marne, Kluck’s rearguard at once protected itself as it had been taught to do. The position was an awkward one, in the angle of two river-courses. But the German communications necessarily traversed the Ourcq, and hereabouts the west bank rises high above the eastern, covering the passage and commanding the country for miles around.

Starting out in the morning from the hamlet of Thieux, 3 miles south of Dammartin, Lamaze’s columns were directed as follows: de Dartein’s Division, the 56th Reserve, on the left, toward St. Soupplets, by way of Juilly and MontgÉ; the 55th, under General Leguay, toward Monthyon, by Nantouillet; the Moroccan Infantry Brigade of General Ditte, toward Neufmontiers. After tramping nearly a hundred miles in three days and nights, with scanty food and sleep, and frequent rear actions, Lamaze’s Corps had spent a whole day at rest, and, though far from its full strength, was a little recovered from the pains of the retreat. The sight of Paris near at hand, and the feeling that the supreme crisis was reached, set up a higher spirit, and prepared the men for the stirring appeal of the Generalissimo. They were now to need all their recovered confidence and courage.

The 5th battalion (276th regiment) of the 55th Division was settling down to its midday meal in face of the hamlet of Villeroy, when it was surprised by a storm of shells from three of Schwerin’s batteries, masked by the trees on the heights of Monthyon and Penchard. A French 3-inch battery in front of the battalion, and another brought up toward Plessy-l’Eveque, at once returned this fire, as it was afterward found, with good effect. But the heavier German field-guns, stationed 8 or 9 miles away in the loop of the Marne, at Germigny and GuÉ-À-Tresmes, and farther north behind Trocy, were far out of range of the French pieces, and were worked with impunity until near the end of the battle. Between Monthyon and Penchard, the enemy had three groups of machine-guns, which kept up a deadly rain of bullets. In two and a half hours, the 5th battalion, just referred to, lost 250 men out of a short thousand; in course of the day, there fell of the 19th company all the chief officers, including the brilliant young writer, Lieut. Charles Peguy, and 100 men.56 Nevertheless, the line jerked itself forward by short bounds past Plessis and Iverny toward the MontgÉ–Penchard ridge. Neufmontiers was the first village carried by assault; and, generally, the Moroccan chasseurs made the most rapid progress—their officers, with swords uplifted in gloved hands, leading them through the cornfields and orchards—until they reached the stronghold of Telegraph Hill, by Penchard, where they were thrice repulsed during the afternoon. By 6 p.m., the enemy being reinforced, all the captured ground was lost. The 55th Division, before Monthyon, and the 56th, on its left, were also at once arrested; but, having administered this check, Von Schwerin proceeded to abandon his advanced position, from Neufmontiers northward. On the left, a patrol of the 56th Division found St. Soupplets evacuated, at 9 p.m. In the evening, while the 7th Corps was coming in on its left, from the highroad between Plessis Belleville and Nanteuil-le-Haudouin, Lamaze’s front was drawn back lightly to the line MontgÉ–Cuisy–Plessy-l’Eveque–Iverny–Charny. Night brought a lull in the battle, a snatch of broken sleep for some of the rank and file at least. A harvest moon shone red through the smoke of flaming hayricks and farmhouses.

This was far from being what General Joffre had counted upon in ordering the 6th Army to be in a position on the morning of the 6th, as an essential part of the general offensive, to pass the Ourcq and march upon ChÂteau-Thierry. Maunoury was still 9 miles from the Ourcq at Lizy, with no prospect of an easy passage. “Some one had blundered.” It is clear that Maunoury’s reconnaissance service was gravely at fault. But there is more than that. In determining to precipitate the intended movement of the 6th Army, the Generalissimo depended upon the telephonic representations made to him by Gallieni. Knowing that, from his starting points on the morning of the 5th, Maunoury had 12 or 14 miles to make to reach the Ourcq, the Governor of Paris must have assumed that no opposition would be encountered—a rash conclusion in face of a commander like Kluck.57 Lamaze’s force was too small to sweep aside any substantial rearguard, too large to come into action without giving the alarm. Why was the 7th Corps not in line with it? Everything must depend upon the efficacy of this flank blow. When the enemy was discovered on the hills of Monthyon and Penchard, should contact have been broken till the attack could be made in full force? Suppose that it did not then succeed, after the loss of precious hours? Cruel dilemma! The decision was to go ahead; and the result came near being the abortion of the whole plan of battle.

The morning of September 6 gave Lamaze an easy success on his left, offset by grievous difficulties on his right. The 56th Division, having occupied St. Soupplets at daybreak, rapidly reached the Therouanne at Gesvres, Forfry, and Oissery; and Marcilly was taken in the afternoon. The 55th, checked for a time at the central height of Monthyon, next met a more determined resistance before Barcy and Chambry. The former village was lost twice, and taken a third time, at the cost of many lives. Ditte’s brigade, strengthened by Zouaves from the 45th Division, reoccupied Neufmontiers, and took Penchard and Chambry, but failed before the Vareddes ridge. Everywhere it was the same tale; though served with the utmost courage, the bayonet is no match for the machine-gun. Before retreating toward the loop of the Marne, the Germans burned down, by means of hand grenades, the village of Chauconin, with its household goods and farm implements. It is curious that the large town of Meaux altogether escaped damage during the battle.

All possibility of surprise was now past; and an average gain of about 5 miles had been dearly bought. Kluck, just installed at Coulommiers, 14 miles away, had been instantly sobered by the news from his rear, and with a speed and judgment worthy of his repute had taken measures to meet the danger.58 The French left, the 7th Corps, had no sooner come into action on this morning of the 6th than two enemy columns were signalled as having reached the Ourcq about Vareddes and Lizy. By the middle of the afternoon, when Lamaze was facing the hills beside Etrepilly, and General Vautier’s two divisions, which had easily attained the line Villers St. Genest–BrÉgy, were striking out from the first to the second line of heights, from Bouillancy to Puisieux, with the prospect of turning the right of the German IV Reserve Corps, they found this new adversary before them. It was a part of the II Corps, withdrawn from the British front by a hard night march, and now thrown adroitly against Maunoury’s left wing.

II. The British Manoeuvre

To understand how this withdrawal, so big with results, was possible, and to do justice to Sir John French’s command in regard to it, we must leave Lamaze and Vautier at grips with the two German corps on the Ourcq, and turn for a moment to the situation south of the Marne.

On September 3, the British Army lay just south of Meaux, from Lagny to Signy Signets, having destroyed the Marne bridges behind it at General Joffre’s request. Kluck, as we have seen, was then approaching the river from the north-west, coming on at a great pace. Several of his Staff officers, pelting eastward from Meaux in an armoured automobile at nightfall, did not see that the last arch of the Trilport bridge was broken, pitched over, and were drowned. A little study of the map will show that Kluck’s rapid movement—his pontoon corps established bridges of boats across the Marne on the night of the 3rd, and the next day his patrols were beyond the Petit Morin and on the Grand Morin—required not simply a farther retreat, but a different direction of retreat, of the British force. To throw it up against the neighbouring French columns, those of the 5th Army (commanded by General Franchet d’EspÉrey since the evening of September 3) was exactly what Kluck was aiming at. To avoid such a calamity, and perhaps to tempt the rash commander farther south, Joffre asked Sir John French to retire some 12 miles farther, drawing his right south-westward, pivoting on his left. This manoeuvre, which to the British commander could only seem the natural pursuance of the French Army Orders of September 2, by him received on the following day, was carried out on September 4. The Expeditionary Force, as it was called, had been on the Continent for hardly three weeks, had fought in that time two great battles and many smaller engagements, and had retreated 160 miles in twelve days, losing much material and nearly a fifth of its original strength, about 15,000 officers and men. Behind the Forest of CrÉcy, close to the railway junctions south of Paris, it was able, on the night of September 4 and during the 5th, to pick up much-needed reinforcements, bringing its effective strength up to five divisions and five cavalry brigades, with guns and supplies.

At midday on September 5, when the battle of the Ourcq was beginning, the I German Army had reached the following positions:—Marwitz’s IX Cavalry Division was north of CrÉcy, the II near Coulommiers. Richthofen’s V Cavalry Division was at Choisy, south-west of La FertÉ-Gaucher, the Guard Cavalry a little farther east, near Chartronges. The II Corps was extended from the Marne near Montceaux to the Grand Morin west of Coulommiers. The IV Corps was on the latter river about La FertÉ-Gaucher. The III Corps was on the great highroad about Sancy and Montceaux-les-Provins; and the IX north of Esternay. The general strategical significance of these dispositions will presently appear; for the moment, we are concerned with them specially in relation to Maunoury’s and the British Armies. Twelve hours later, Kluck’s front was advanced a little farther, extending from near CrÉcy-en-Brie, along the Grand Morin, by Coulommiers and La FertÉ-Gaucher, to Esternay, with the cavalry of Marwitz before the centre and left. The bulk of this force was aimed at the 5th French Army; but the II and part of the IV Active Corps faced the British. Such was the position at the moment when Kluck, informed of the danger to his rearguard, decided to send back to the Ourcq his II Corps, bringing the western wing of the invasion to a sudden and humiliating end.

Neither at French nor at British Headquarters were these dispositions exactly known; still less could the German commander’s intentions be known. The last stage of the British retirement, asked for by General Joffre, had taken the body of Sir John French’s troops out of direct contact with the enemy. They had to embody newly-arrived men and guns, and then to return over this ground. Joffre’s order of September 4 had named as the British line for the evening of the following day “the front Changis–Coulommiers, facing east, ready to attack in the general direction of Montmirail”—due east, that is to say, not north-east. It is evident, from this instruction, that the Generalissimo (1) did not anticipate any serious resistance west or south of Coulommiers, for the British could not be fighting on their north flank while marching due east, and they could not start from Coulommiers when the enemy was 8 miles farther south; and (2) did not anticipate a sudden withdrawal of Kluck northward, which would require the British to turn thither in aid of Maunoury. When Joffre and French met at Melun on September 5, the instruction was modified, but not radically; it was now, in Sir John’s words, “to effect a change of front to my right—my left resting on the Marne, and my right on the 5th Army, to fill the gap between that army and the 6th.” The right of the 5th Army, however, was not at Coulommiers—both Changis and Coulommiers were in the hands of the enemy—but Courtacon, 12 miles farther to the south-east; and to join the 6th and 5th Armies implied a north-easterly, not an easterly frontage. Joffre so far recognised the difficulty of filling this wide space with five divisions as to instruct Gallieni to send across the Marne the 8th Division of the French 4th Corps; and this came in, with prompt effect, between Meaux and Villiers-sur-Morin, 5 miles farther south, beside the British 3rd Corps, at 9 a.m. on September 6. There then still remained a space of over 20 miles between the 6th and 5th Armies, and it is, therefore, idle to suggest, as some zealous partisans of Gallieni have done,59 that the British commander was needlessly nervous as to the continuity of the line, when it became evident that considerable bodies of the enemy were spread across his path.

It was not till September 7 that any need appeared to help Maunoury. But, as we now know, Kluck ordered the withdrawal of his II Corps to the Ourcq at 3.30 a.m. on September 6—2½ hours before the beginning of the Allied offensive. The withdrawal was well covered, and was not observed for twenty-four hours. The change of direction of the British advance toward the north could not be effected with the instancy that paper strategists have imagined; and the necessity of keeping touch with d’EspÉrey continued. The question whether the British advance was timid and halting must be judged in the light of the facts not as we now know them, but as they revealed themselves from day to day; and in the light not of Gallieni’s desires or needs only, but of the whole battle, and particularly of the instructions given to the British Army by General Joffre, who alone was responsible for the whole battle. That Maunoury would be seriously engaged with Kluck’s rearguard on the afternoon of the 5th was not anticipated by the French; it could not, then, be anticipated by the British. Since criticisms are raised as to one side of a converging movement, it must be pointed out that, if the French attack on the Ourcq had been delayed for twelve hours, and had not anticipated the general offensive, all would have been well. Kluck would have been unable to evade one assailant in order to throw all his force upon the other; and the tasks of Maunoury and the British would have been more advantageously divided. We are here, apparently, in face of one of those failures of information and agreement which are liable to occur, even under the best leadership, between armies of different nationality when plans are suddenly changed. It may now be recognised that the battle of the Marne would have yielded a completer, cheaper, and speedier victory if the rectangular movement of the French 6th and British Armies had been more exactly designed and timed to a strict simultaneity. There was a lack of assimilation. Perhaps the British were slow in getting under weigh; it is much more certain that Gallieni was precipitate.

The front of the British 3rd (incomplete), 2nd, and 1st Corps at the opening of the offensive lay, then, from Villiers-sur-Morin, across the edge of the Forest of CrÉcy, by Mortcerf, Lumigny, Rozoy, and Gastins, to near the Forest of Jouy, where Conneau’s Cavalry Corps connected with the infantry of the 5th Army. The battle here opened with an enemy attack. To mask its withdrawal to the Ourcq, a part of the German II Corps had delivered, early on the morning of September 6, a blow at the British right, and fighting was sharp till noon over the farmlands of the Brie plateau between Hautefeuille and Vaudoy—that is, 8 miles south-west of Coulommiers. “At this time,” says Field-Marshal French, “I did not know that a retreat had really set in, or how the various German corps and divisions were placed.” Columns of the IV Active Corps were still farther south, to the east of Vaudoy, on the Provins road, with large forces of cavalry and the III Corps on their left. It was a delicate part of the front, the space between the British and 5th French Armies. During the afternoon, while the khaki line slowly progressed over the stubble fields and broken forest around the villages of Lumigny, Pezarches, and Touquin, unmistakable evidence began to come in that the German foreguard had become a rearguard, and that the body of the II Corps had been in retreat all day. The charred walls of the hamlets of Courchamps and Courtacon, destroyed with deliberate ferocity, marked the most southerly points of the invasion in the western field.60 To the Allied soldiers who knew not Maunoury, it must have seemed that their offensive was commencing magically well. About 10 a.m., the British left and centre—the 4th Division and the 2nd Corps—had been surprised to find the pressure on their front suddenly relieved. On their right, the 1st Corps soon saw its way free, and strode northward. At 6.30 p.m., the IV Active Corps received orders to follow the II Corps back to the Ourcq. Thus, by evening on September 6, Sir John French was able to reach the Grand Morin, from CrÉcy-en-Brie eastward, with scouts beyond the stream at Maisoncelles. Coulommiers, where Kluck had had his headquarters, was occupied during the night.

The Allied plan was now fully revealed. Instead of presenting on the Grand Morin an ironclad face, safe in flank and rear, the I German Army had been suddenly thrown on to a rectangular defensive on a front of 50 miles between Betz and Courtacon, against attacks converging from the west, south-west, and south. That evening, at Joffre’s request, the British line was directed more to the north, thus emphasising the effect of Maunoury’s move. From this moment, the withdrawal of the whole of Kluck’s forces over the Marne must have been envisaged. On the following day, September 7, in fact, the III and IX Corps (west of Montmirail), were preparing to follow the IV Active Corps across the Marne; but the Allies were then aware of what was happening. Marwitz’s Cavalry Corps covered the movement along the Grand Morin, with one division to the west, one to the east, and one 4 miles north of Coulommiers, while Richthofen’s Divisions operated farther east, all available artillery supporting them. The task was fulfilled with much resource and energy; but the position was not one that could be long maintained, for the British 3rd Corps was at Maisoncelles, 4 miles beyond the Grand Morin, and the French 8th Division threatened the German flank at double this distance northward by occupying St. Fiacre and Villemareuil. At noon, Marwitz gave way, falling back to the Petit Morin, from La FertÉ-sous-Jouarre south-eastward. By evening, the British 3rd and 2nd Corps were beyond the Grand Morin at La Haute Maison and Aulnoy; the 1st was held back somewhat from Chailly to near La FertÉ-Gaucher, in touch with the French 5th Army. General de Lisle’s Cavalry Brigade, with the 9th Lancers and the 18th Hussars, showed especial vigour. The men were full of cheer, and ready for anything; but Sir John French was a careful commander. The measure of the enemy’s retreat could not be immediately taken through the curtain of cavalry and artillery—aviation was in its infancy in those days. All the strength available was in line; and it was so thin a line as to tempt surprise. The Field-Marshal considered the alternative of sending direct help round to Maunoury, but concluded that the best aid would be to drive rapidly to and across the Marne.61

III. A Race of Reinforcements

On the Ourcq, each adversary was bringing up reserves, and was trying to turn the other by the north, with a slight advantage in time on the French, but a superiority of speed on the German, side. We left the centre of the 6th Army, on September 6, practically stationary about Marcilly and Barcy; while, moving from BrÉgy and Bouillancy, the 7th Corps gained Puisieux and Acy during the afternoon, and the 8th Division, thrown across the Marne, drove some enemy contingents into the woods of the river loop east of Meaux. Maunoury decided to attack frontally the three plateaux of Vareddes, Trocy-Vincy, and Etavigny, throwing picked columns into the valleys between, that of the Therouanne at Etrepilly and the Gergoyne ravine at Acy-en-Multien, in the hope of turning the hill positions. His field batteries were now in force at Bouillancy, Fosse-Martin, La RamÉe, Marcilly, and Penchard; but he had no heavy artillery. Worse, from September 5, when his only aviator was brought down at Vareddes, to September 9, when Captain Pellegrin found a machine and discovered the nest of German mortars in the gullies by Trocy, he had no air scouts, so that, almost throughout the battle, the German gunners dominated the field.

On September 7, Schwerin’s IV Reserve Corps, strengthened during the day by a part of the IV Active Corps, rallied against Lamaze’s harassed men, who, still untutored to spade work, suffered heavily, but did not give way. Ditte’s Moroccan Brigade commenced at dawn a new move toward Vareddes, was beaten off, spent the afternoon in a fearful hand to hand struggle on Hill 107, won it, but was finally driven back to Chambry. The Algerian troops of General Drude, the 45th Division, had come in on the right-centre; and they were able, during the morning, to make a long stride forward east of Marcilly. Beyond Barcy, however, they were immediately stopped; repeated charges were broken, many officers and men being left on the ground. During the night, under a brilliant moon, the north wing of the division cut its way into the village of Etrepilly, but could not carry the cemetery, 300 yards beyond, and had to fall back.

The 7th Corps was no more fortunate. After taking Etavigny and the hillsides above Acy with a rush, it was suddenly overwhelmed by a massive counter-attack of the newly arrived II Corps, and had to abandon both villages, re-forming before Bouillancy and Puisieux. Many units had lost nearly all their officers. A panic was threatened. At a moment when it seemed that the left of the army could not be saved, Colonel Nivelle, with five field batteries of the 5th artillery regiment, gave a first exhibition of the qualities which, two years later, were to secure the defence of Verdun, and to bring him to the chief command. Carrying forward through the wavering ranks of the infantry a group of his field-guns, he set them firing at their utmost speed upon the close-packed columns of the enemy. The “75” is a murderous instrument in such circumstances; and those greycoats who remained afoot broke in disorder. It was an hour’s relief; but manifestly this wild situation could not long continue. The enfeebled lines approached the extreme limit of endurance. And still the tide of slaughter swayed to and fro. Nogeon, Poligny, and Champfleury Farms—the first north, the others south, of Puisieux, large stone buildings topping the plateaux—were the scenes of most bloody and obstinate encounters. Nogeon, the largest of them, was stormed and lost three times at intervals during the day. Under sustained fire from Trocy, its massive walls were broken; the corn barns took fire and blazed across the expanse of the battlefield.

In the evening, Von Schwerin drew back his lines a little from the edge of the plateau, and the ruined farms and hamlets gave the French a precarious shelter. At the same time, a reciprocal attempt at envelopment by the north began to design itself. The 61st Reserve Division had just been brought up from Pointose; and Maunoury decided to throw it, with the 1st Cavalry Corps, out to his extreme left, the former at Villers St. Genest, the latter beyond Betz. Almost simultaneously, new German detachments reached the Ourcq, and were set to prolong to the north the front of the II and IV Corps, while a Landwehr Brigade acting as line of communication troops was summoned urgently from Senlis. There was now no question of the 6th Army fulfilling its original task; the utmost hope was that it might hold till the British came up, across the enemy’s rear. Maunoury had to cope with an equal mass in better positions—three strong corps, the IV, the II, and the IV Reserve, with the IV Cavalry Division—against Lamaze’s two Reserve Divisions, Drude’s Division, the 7th Corps, the 61st R.D., and the Cavalry Corps. Only the III and IX Corps and Marwitz’s Cavalry remained beyond the Marne; and, though the British pressure was increasing, the enemy’s withdrawal had not been seriously disturbed. Kluck’s boldness, skill, and decision were undeniable. It was evident that he had recovered from the first shock, and meant, if possible, to overwhelm its authors. Exhausted, and tormented by thirst, it was with sinking hearts that the Army of Paris looked up to the smoking hills.

Viewed from French General Headquarters, however, the prospect was more favourable. The retreat of the I German Army was gravely compromising the position of its neighbour, the II; and its effects were beginning to show farther to the east. For three days, these two forces were moving in opposite directions—Kluck to the north-west, BÜlow to the south-east. The task of exploiting the dislocation thus produced fell to the British and d’EspÉrey’s Armies. The rÔle of the 6th Army was thus radically changed by the development of events; but it remained as important as ever in the whole design. If Gallieni and Maunoury could have reviewed the field from the Ourcq to Verdun, they would have been well satisfied.

IV. The Paris Taxi-Cabs

I spent September 7 among the rear columns of the 6th Army. In the morning, the little town of Gagny, half-way between Paris and Claye (Maunoury’s headquarters), and the last point one could reach by rail from the city, was full of men of the 103rd and 104th regiments, belonging to the 4th Corps (General BoËlle), just arrived from Sarrail’s front. They sat in and before the cafÉs, lay on the grass of the villa gardens, lounged in the school playground, where their rifles were stacked and their knapsacks piled. Some had managed to get their wives and children to them, and were telling great tales of the first month of war. I went out into the deserted countryside toward the front, passing marching columns, motor-wagons, dispatch-riders, here a flock of sheep in charge of uniformed shepherds, there a woodland bivouac, and in the evening returned to Gagny. More regiments had arrived; the town was boiling from end to end. In the main street, a battalion was already marching out to extend Vautier’s left, a thin file of country folk watching them, waving handkerchiefs, the girls running beside the ranks to give some handsome lad a flower. Up the side roads, other columns waited their turn, standing at ease, or sitting on the edge of the pavement; a few men lay asleep, curled up against the houses. But the most curious thing was a long queue of Parisian taxi-cabs, a thousand or more of them, stretching through by-roads out of sight. The watchful and energetic Gallieni had discovered, at the cost of us boulevardiers, a new means of rushing reinforcements to the point where they were direly needed.

It was the idea of his chief of staff, General Clergerie. Joffre had at this moment only one remaining unit of the Regulars to give to Maunoury—a half of the 4th Corps, which had been brought round by rail from the Verdun front, and of which we have seen the 8th Division in action south of Meaux. The 7th Division had detrained in Paris during the afternoon of September 7. It was to be sent to Maunoury’s extreme left, near Betz, 40 miles away. Everything might now depend upon speed. It was found that only about a half of the infantry could be carried quickly by train. Clergerie62 thought the remaining 6000 men might be got out by means of taxi-cabs. The Military Government of Paris already had 100 of the “red boxes” at its disposal; 500 more were requisitioned within an hour, and at 6 p.m. they were lined up, to our astonishment, beside the Gagny railhead. Each cab was to carry five men, and to do the journey twice, by separate outward and return routes. Measures were taken in case of accident, but none occurred. This first considerable experiment in motor transport of men was a complete success; and at dawn the 7th Division was in its place on the battlefield.

On September 8, the British Expeditionary Force, steadily gathering momentum, reached, and in part crossed, the Petit Morin, taking their first considerable number of prisoners, to the general exhilaration. On the left, the 3rd Corps advanced rapidly to the junction of that river with the Marne; but the enemy had broken the bridges at La FertÉ-sous-Jouarre, and held stubbornly to their barricades on the north bank through this night and the following day. Farther up the deep and thickly-wooded valley, the 2nd Corps had some trouble between Jouarre and Orly; while, on the right, the 1st Corps, after routing the German rearguards at La TrÉtoire and SablonniÈres, made the passage with the aid of a turning movement by some cavalry and two Guards battalions of the 1st Division.

The orders of this day for the 6th Army were to attack on the two wings—Drude’s Algerian Division (relieving the exhausted 56th Reserve and the Moroccan Brigade), with the 55th Reserve, on the right, towards Etrepilly and Vareddes; the 61st Reserve Division, General BoËlle’s 7th Division, and SordÊt’s cavalry, on the left—while the 7th Corps stood firm at the centre, and, south of the Marne, the 8th Division pressed on from Villemareuil toward Trilport, in touch with the British. For neither side was the violence of the struggle rewarded with any decisive success. On the French right, the Germans had more seriously entrenched themselves, and had much strengthened their artillery. Lombard’s Division of the 7th Corps was heavily engaged all day at Acy; at night the enemy still held the hamlet, while the chasseurs faced them in the small wood overlooking it. On the left, the 7th Division of the 4th Corps had no sooner come into action than it had to meet a formidable assault by the IV Active Corps. This was repulsed; but the Cavalry Corps seems to have been unable to take an effective share in the battle. During the afternoon, German troops occupied Thury-en-Valois and Betz. Reinforcements were continually reaching them. At nightfall, although BoËlle’s divisions were resisting heroically, and even progressing, the outlook on the French extreme left, bent back between Bouillancy and Nantheuil-le-Haudouin, had become alarming. Maunoury, however, obtained from General Gallieni the last substantial unit left in the entrenched camp of Paris, the 62nd Division of Reserve, and gave it instructions to organise, between Plessy-Belleville and Monthyon, a position to which the 6th Army could fall back in case of necessity. In course of the night, Gallieni sent out from Paris by motor-cars a detachment of Zouaves to make a raid toward Creil and Senlis. It was a mere excursion; but the alarm caused is very comprehensible when the extreme attenuation of the supply lines of the German I Army is remembered. Marching 25 miles a day, and sometimes more, it had far overrun the normal methods of provisioning. During the advance, meat and wine had been found in plenty, vegetables and fruit to some extent, bread seldom; here, in the Valois, the army could not feed itself on the country, and convoys arrived slowly from the north. Artillery ammunition was rapidly running out. Hunger quickly deepens doubt to fear.

But Maunoury’s men were at the end of their strength. On the morning of September 9, a determined attack by the IV Active Corps, supported by the right of the II, was delivered from Betz and Anthilly. The 8th Division had been summoned back from the Marne, to be thrown to the French left. Apparently it could not be brought effectively into this action; and the 61st and 7th Divisions and the 7th Corps failing to stand, Nanteuil and Villers St. Genest were lost, the front being re-formed before Silly-le-Long. “A troop which can no longer advance must at any cost hold the ground won, and be slain rather than give way.” Such a summons can only be repeated by a much-trusted chief. Maunoury repeated it in other words. Thousands of men, grimy, ragged, with empty bellies and tongues parched by the torrid heat, had already gone down, willingly accepting the dire sentence. Few of them could hear or suppose that the enemy was in yet extremer plight. So it was. Early in the morning, Vareddes and Etrepilly had been found abandoned; greater news had been coming in for hours to Headquarters—some of it from the enemy himself by way of the Eiffel Tower in Paris, where the French “wireless” operators were picking up the conversations of the German commanders. Marwitz was particularly frank and insistent; his men were asleep in their saddles, his horses broken with overwork. He was apparently too much pressed to wait for his message to be properly coded. By such and other means, it was known that Kluck and BÜlow were at loggerheads, that, even on the order of Berlin, the former would not submit himself to his colleague, and that, in consequence, BÜlow in turn had begun to retreat before Franchet d’EspÉrey.