(PLATES 51 TO 66.)

I have been able to traverse several times since the war the great stretch of country in Picardy which is generally spoken of at home as "the Somme"—country over which much of our hardest fighting took place in 1916 and 1918, and where thousands of our brave men are now lying. We became only too familiar with the names of places within it, which might have peacefully remained for centuries more in the happy oblivion in which they had rested for centuries past, had not the war waves broken upon them and destroyed them while making them immortal. Much of the country had been so completely devastated that there was nothing in it or on it to show in a picture—nothing beyond an irregular expanse of ground broken everywhere into shell-holes and covered over with an untidy wild herbage of rank weeds. But the interest of this country to all of us at home—and "at home" in this case more than ever includes the homes overseas—is so close and so poignant that it is probably worth while to add here some little description of the characteristics of the great area which we call "the Somme," and the positions of the places which we fought over.

In the thirty miles from Amiens to PÉronne the Somme runs from east to west in a narrow valley, with eight immense double bends round spurs which project alternately from the higher country (some 200 feet above the river-level) on the north and south. The main road eastwards from Amiens lies south of the river, and rises gradually to the higher level at Villers Bretonneux, about ten miles from the city, and then continues dead straight and nearly level, till it drops again at Brie (twenty-nine miles from Amiens), to the Somme valley, after the river has taken its sharp bend to the south at PÉronne, which is four miles north of Brie. As one goes eastwards from Amiens the route becomes more and more war-worn. At first the ordinary avenues of trees still stand, farther on the trees become fewer and fewer, and finally disappear entirely (Plate 51), and a region of total destruction is reached, where only rough indications remain of the sites of the villages.

But in 1919 I found German prisoners at work filling up shell-holes (the French and ourselves did not make prisoners dig front-line trenches), levelling the ground, and clearing up generally, and some reoccupation of land had already started, peasants and "store"-keepers living in such temporary bungalows as they could construct. Somehow or other the owners of different strips of land along the road seemed to have discovered which particular strip belonged to each one, ploughing was already going on, and cultivation had been started in quite a number of places.

The river itself lies always too low down in its valley to be visible from the road, from which the view to the north looks right over to the high ground between the Somme and the Ancre. The Avre, coming from Montdidier and Moreuil in the south, falls into the Somme close to Amiens, and the Ancre, coming from Albert in the north, joins the main river at Corbie, four miles north of Villers Bretonneux.

Albert is about eighteen miles north-east of Amiens by a straight road through Pont Noyelles, which continues to Bapaume, eleven miles farther. On the high ground between the Ancre and the Albert-Bapaume road stood Thiepval and the German redoubts, and on the Bapaume road itself PoziÈres, Courcelette, and Warlencourt.

In the angle between the Albert-Bapaume road and the northern bank of the Somme every village and every wood became part of a tragic history—Mametz, Contalmaison, Longueval, Guillemont, Combles, with Trones Wood, Delville Wood, and the others. South of the Somme and between the river and the road lie Hamel and Chuignes, while on the main road itself, east of Villers Bretonneux, once stood the villages of WarfusÉe (Lamotte), EstrÉes, Villers Carbonnel, and others, while the town of PÉronne—protected by Mont St. Quentin on the north and by the Somme and a tributary on the other three sides—lies just at the bend. South of the road and in the triangle between it and the Avre lie the uplands on which, very generally, the French were fighting to the right (south) of the British, and in which the village names are therefore less familiar to us than those farther north.

---

The first battle of the Somme commenced on the 1st of July, 1916, and lasted, with more or less quiescent intervals, until the late autumn. British and French were fighting side by side—the British on the northern half, from the Upper Ancre to the Albert-PÉronne road; the French to their right, facing PÉronne, crossing the Somme, and extending southward as far as Chaulnes.

The German defences in the north, which had been under construction for more than a year, were enormously strong, the "first line" alone being a maze of trenches half a mile wide.

Their position stretched southward from below Arras to GommÉcourt, and covered Beaumont-Hamel and the heights east of the Ancre valley, crossed the Albert-Bapaume road two miles north of Albert, passed eastwards through Fricourt, crossed the Somme some miles short of PÉronne, and then ran southwards west of Chaulnes. The story of the fighting, both French and British, as it can be read even in Lord Haig's official despatches, and still more in the unofficial accounts, is a continuous record of episodes every one of which would have been called Homeric in any other war, but which in this gigantic struggle seem to have become ordinary events. The ordinary civilian of common life from workshop or warehouse or office or studio turned out to be in essence exactly the same being as the noble and adventurous heroes of the stories and histories of our youth. And while he would grumble seriously about his baths or his meat or his shaving facilities, he would yet go into action cheerfully without hesitation, although he knew well enough the horror of his own work and the great chance that he might never return.

The village of Mametz (Plate 53), like nearly all the villages in "the Somme," has disappeared; it was among those taken by the British on the first day of the battle. It was in this attack that East Surrey men are said to have gone forward dribbling footballs, some of which they recovered in German trenches, in front of them.[22]

The French on our right got within a mile or two of PÉronne, but its defences were too strong and it was not actually captured until 1918.

Trones Wood (Plate 54) was cleared early in July. Here a small body of 170 men of the Royal West Kents and Queens[23] held out all night, completely surrounded, until relieved the next morning. Delville Wood (Plate 55) was captured the next day. The fighting in these woods has left nothing of them but churned-up ground and a few bare stems, although the rank undergrowth makes some of them appear quite green from a distance. Combles (Plate 56) was not taken until the 26th of September, when the British and French entered the town simultaneously (from the north and the south respectively), and captured a company which had not been able to get clear away in time. The little town is not so entirely wrecked as many other places, but the house which is shown under reconstruction in the photograph is perhaps one of the least damaged.

Thiepval and the redoubts on the Thiepval plateau were not finally secured until November. The Germans had said beforehand that we "would bite granite" in trying to take them. We did bite granite, but our teeth proved the harder.

Along the Albert-Bapaume road the villages of PoziÈres and Courcelette have disappeared altogether. Sometimes a big iron gate, or half a gate, or a stone gatepost, shows where an entrance once existed to some more or less pretentious mansion, but the building itself has gone entirely, and its site is grown over with rank herbage, which hides every indication even of where the house once stood. The whole Thiepval plateau is now a wilderness of weedy vegetation, and the weeds seem to have swallowed up the redoubts altogether, as well as Thiepval itself.

The defences on the Upper Ancre still barred the way to Bapaume along the road by Le Sars and the Butte de Warlencourt (Plate 57).[24] The Butte was the centre of the German position, as strongly protected by trenches and wire as even the Thiepval plateau itself. Fierce attacks in October and November, 1916, failed to secure it, and the chalky hillock was only finally taken in February, 1917. It now carries five crosses erected in memory of the units which fought there.

The mud, our chief enemy, made active operations impossible for a time. It was an even worse enemy than the Germans. General Haig says[25] that the trenches were channels of deep mud and the roads almost impassable, making all problems of supply most serious. General Maurice calls it, later on, a "morass of stinking mud." We were, in fact, at that time—and at other times as well—fighting the elements as well as the Germans. On the 17th of March, 1917, however (after the Warlencourt Ridge had been carried), Bapaume itself, which had been systematically destroyed by the Germans before they evacuated it, was at last entered.

Bapaume in 1919 was, like Albert, being rapidly reinhabited, and the new buildings (perhaps due to their being closer to the main road) were more in evidence than in most other places.

The villages north of Bapaume on the Arras road (Behagnies, Ervillers, and others) are, like those nearer the Somme, practically wiped out. But here, also, peasants and small shopkeepers were returning "home," and sheltering themselves as best they could in some sort of hutments.

On the 17th of March, 1917, also, the Germans having just commenced their great retirement, Mont St. Quentin was taken, and the next day PÉronne itself. Plate 58 shows the dry bed of the Nord Canal where the road crosses it just at the rise on the back (north) of Mont St. Quentin. Plate 59 shows the ruin of the Church of St. Jean at PÉronne. The little town itself, originally of about 5,000 inhabitants, was in parts systematically burnt and destroyed by mines by the Germans before they evacuated it in 1917, and further damaged by Franco-British shell-fire in 1918. On the spot I was told that the great church had been among the buildings deliberately burnt by the Germans; in any case it is now, like the rest of the town, a mere ruin. The outrages perpetrated by the Germans in their masterly retreat in 1917 extended across the whole area of the retirement (see Plate 119), and have been sufficiently described, so far as it has been possible in any decent paper to describe them. But the burnt and shattered houses were not the matters, bad as they were, which caused the intense feeling of loathing in addition to anger among the French, when they were at last able to return to their desecrated homes.

For a year after March, 1917, the Somme area ceased to be fought over, as the German retirement in 1917 had removed them far to the east. A year later the tables were turned, when on the 21st of March Ludendorff's great attack, cleverly directed against our weakest spot, began to drive us back from St. Quentin towards Amiens, and succeeded so rapidly that on the 23rd the Germans were at PÉronne and on the 25th near EstrÉes, three miles east of the Somme on the Amiens road. On the 25th our Allies, on our right, had been compelled to fall back as far as Noyon. At this critical moment there was got together surely the most remarkable auxiliary force that a British General has ever had under his command. General Haig says:

The line on which this "mixed force" was placed passed through WarfusÉe (Plate 60). Some of the men collected were Engineer civilians with no previous training, and no knowledge of rifle-shooting. I have been told that they were pronounced most plucky, "but somewhat dangerous"! In the result, however, they did yeoman service in helping to hold back the onslaught until the distant reserves could arrive and until the attackers had eventually exhausted themselves.

On the next day came the historic conference at Doullens, which resulted in the appointment of General Foch in supreme control of the united forces (p. 3). But General Haig found it necessary to withdraw his troops still farther, and the German advance was finally checked only at WarfusÉe-Ablancourt, some ruins of which appear in Plate 60. The enemy never succeeded in reaching the crest of the high ground from which he could so completely have commanded Amiens (p. 38), although he was able to hold Villers Bretonneux, after a new attack on the 24th of April, for a few hours, after which he was turned out by the Anzacs and never got back. It was in this attack that British tanks met German tanks and beat them.

It was not until the 8th of August, 1918, after Foch had carried on his successful attacks on the Marne Salient for three weeks, that the great counter-attack on the Somme was fully started, although before that day there had been some important gains. Especially had a notable combination of Australians, Americans, and Tanks had a great success on the 4th of July, after a heavy barrage, in capturing Hamel, a village on the Somme just north of WarfusÉe. Both the Australians and the Tank Corps have given picturesque accounts of this fighting, with a somewhat amusing preference, in each case, for the service with which the writer is connected. It was here that the Australians are credited with having pronounced their new colleagues from across the Atlantic to be "good lads, but too rough"!

Most elaborate (and successful) precautions[27] had been taken to make sure that the attack of the 8th of August (Ludendorff's "black day") should be a surprise. In spite of all these precautions, some anxiety may have been felt by those who knew that a sergeant, who was well acquainted with everything that was on foot, had been taken prisoner by the Germans a few days before. Oddly enough, the minutes of the cross-examination of this N.C.O. were afterwards captured, and it was found that, like a plucky Englishman, he had given nothing whatever away.[28]

Villers Bretonneux had been throughout in our possession, but only its ruins were standing, the one "hotel" which I found there in 1919 being a tarpaulin-covered shed (Plate 61) calling itself the "HÔtel des Trois Moineaux," and bearing a cryptic message from "Toto" which I am unable to explain. The first day's advance swept far beyond WarfusÉe, just south of which the village of Marcelcave was captured by a tank whose Lieutenant demanded—and obtained—a receipt from the Australians before he would hand over his spoils to them. Abreast of Foucaucourt (Plate 52) and between it and the Somme lies Chuignes, where the Australian advance captured a 380-mm. gun on an elaborate emplacement, which had been put in position, but I believe too late to be used, for the purpose of a long-range bombardment of Amiens. The gun was dismantled before we reached it, and lies on the ground shorn of some 10 feet of its muzzle end, which had been cut off by its captors to send home as a "souvenir."

The Chipilly spur (Plate 62), north of WarfusÉe on the north side of the river, caused some heavy fighting, but was taken by the Londoners on the second day of the advance, with the help of two companies of Americans who are said to have lost touch with their own division and to have been quite ready to lend a hand in any fighting that was going on. The photograph gives some idea of the river itself at a place where it is navigable over a great breadth. Cappy (Plate 63) is a little farther upstream, where the river has divided itself into various channels, the particular one seen being the navigable canal, the rest of the river spreading over a quarter of a mile of marsh land to the north bank.

The land beside the Amiens-PÉronne road becomes more and more ruined as one goes eastwards. Plate 51 shows something of what the actual road looks like, but no picture can indicate the state of the land itself, the country that was once fertile fields and farms. On my last visit (early in 1920) it was pleasant, but pathetic, to see that many peasants had somehow been able to find out which strip had been theirs before the war, and had built themselves hutments—they could hardly be called houses—in which they could at any rate live beside the land which they loved and which they are trying once more to cultivate. Towards Villers Carbonnel the countryside shows itself as more and more destroyed. Plate 64 (an officially taken photograph) indicates the appearance of that village immediately after we passed through it in 1918 and before the clearing-up work had commenced. A little later the broken woodwork would be collected for firing and the bricks from the fallen walls (if enough were left) would be trimmed and stacked ready for use again in making such dwellings as will anyhow make it possible for the peasants to get back again. In France they do not wait for trade union permissions, or "skilled" labour, or the sanitary (and other) regulations of County Councils, but go straight ahead and build. It seems certainly the best way of getting houses.

General Haig tells us how in March, 1917, when we were trying to keep up with the retreating Germans along this road, the part of it between Villers Carbonnel and the Somme at Brie was almost knee-deep in mud, so that it took the troops sixteen hours to cover the last four and a half miles. The difficulty of this crossing can be well understood by everyone who has seen the breadth and character of the flat marshy ground which, over a great part of the distance from Amiens, represents the bottom of the Somme valley. Some indication of the difficulty of troops crossing the river can be gathered from Plate 65, which is taken from below ClÉry, close to the point at which the Australians crossed the river on the 31st of August, 1918, and made the magnificent attack on Mont St. Quentin, which resulted in the capture of PÉronne the next day, and earned such warm praise from the Commander-in-Chief.[29]

The ChÂteau of Brie (Plate 66) lies on the Somme only half a mile south of the crossing of the main road from Amiens to St. Quentin, and therefore some four miles south of the great bend of the river at PÉronne. On the 27th of September, 1918, it was the scene of a wonderful dress rehearsal for the crossing of the Hindenburg Line at the St. Quentin Canal two days later (see p. 58). Rafts, collapsible boats and life-lines, and some of the 3,000 life-belts which had been hurried up from the coast, were all tested, to make sure that there should be neither hesitation nor failure in their use in the attack on the "absolutely impregnable" section. And, as everyone knows, there was neither hesitation nor failure; the St. Quentin Canal was carried by the Terriers on the appointed day, and with this success, and the crossing of the "Kriemhilde" Line by the Americans in the middle of October, the last standing places for the retreating German armies vanished.

On the road east of the Somme from Brie to PÉronne one saw a curious phenomenon which I seldom saw elsewhere, and cannot explain. In some way the trees in the felled avenue had been able to reassert their life, and for a considerable distance the road was lined in an unsightly fashion with what looked like gigantic bushes growing out of the stumps of the once tall and beautiful trees.

I have said nothing in this section as to Amiens itself; it had serious enough troubles, although it was never in the fighting zone, having been evacuated by the Germans after only ten days' occupation in September, 1914. That time, however, was sufficient for a requisition of half a million francs to be enforced, and for a number of civilians to be deported.

Some parts of the city, including the railway-station, were seriously damaged by bombing and by heavy shells, and the city suffered much from April to June in 1918. The civilian inhabitants left it early in April. Several shells hit the cathedral, and houses within a few yards of it are entirely wrecked, but happily very little damage was done to the structure itself, from which the stained glass had been safely removed.