The design which I had formed after the battle of August 8th of driving the enemy completely out of the bend of the Somme—but which I was obliged to abandon for the time being because of the decision of the Fourth Army to thrust in a south-easterly direction—was now about to be realized. The effect of the battle of Chuignes, following so closely upon the advance of the Third Army two days before, made it probable that the enemy would decide upon a definite withdrawal to the line of the Somme.

It now became my object to ensure, if he should attempt to do so, firstly, that his withdrawal should be more precipitate than would be agreeable to him, and, secondly, that when he reached that line he should be accorded no breathing time to establish upon it a firm defence from which he could hold us at bay for the remainder of the fine weather.

The French Army took over from me on the night of the 23rd August the whole of that portion of my front which still extended south of Lihons. General Nollet, Commander of the 36th French Corps (34th and 35th French Divisions), became my southern neighbour, displacing my Fourth Division, and also a Canadian Division, for whose sector I had become responsible since the departure of General Currie, a few days before.

During these redispositions, probably induced to do so by evidences patent to him that large troop movements were in progress, the enemy carried out a very heavy gas bombardment and maintained it for some hours over the whole of the front which was being taken over by the French.

The wind blowing from the south, the gas, which was unusually dense, drifted over the whole areas both of the Fourth Australian and the 32nd British Divisions, and caused a large number of gas casualties, which weakened the available garrisons of these sectors.

The Second and Fifth Divisions were brought up on the night of August 26th to relieve the First Division, which had worthily earned a rest, and by these redispositions my whole frontage, which, in spite of the reduction effected, still exceeded nine miles, was organized to be held by four Divisions, counting from south to north as follows: 32nd Division, Fifth Division, Second Division and Third Division, the latter lying north of the River Somme.

The First and Fourth Divisions were each sent back, the former to a pleasant reach of the Somme near Chipilly, and the latter to the neighbourhood of Amiens, there to have a long rest and to recuperate after their strenuous labours. These two Divisions were, I had resolved, to be kept in reserve for any tour de force, the need for which might arise later. This disposition was based on intuition rather than on reasoning; but the event proved that it was a fortunate decision; for, at a juncture, three weeks later, when a great opportunity presented itself, these two Divisions, then fully rested, proved of priceless value.

The Third Division held my front north of the Somme, and their presence there ensured my unchallenged tactical control of that important river valley. Numerous crossings had been systematically destroyed by the enemy, as he was being driven back from bend to bend, and as systematically repaired by my indefatigable engineer and pioneer services, as fast as the ground passed under our control.

Reconstruction of bridges and culverts is as tedious a business as their demolition is expeditious. A charge of gun-cotton, placed in the right spot, a primer, a short length of fuse, or an electric lead to a press button are all that are needed, and a single sapper standing by with a match, to be lighted at the last moment, can do all that is necessary to provide three days' work for a whole Company of Engineers.

Nevertheless, the control of the river valley was of inestimable advantage, for it enabled me to carry out a policy of continuous and rapid repair. Consequently, during the whole of our subsequent advance, every means of traversing the valley from south to north, which had been tampered with, was soon restored, as fast as my infantry had made good their advance beyond the ruined crossing.

This facility was to have an important bearing upon my freedom of action, not many days later, when the Corps came head on to the north and south stretch of the Somme, and found every bridge gone. That circumstance alone would have proved an irretrievable misfortune, if I had not had already available numerous restored crossings upon the east and west reach of the river. For by that means, my ability to pass troops and guns rapidly from one bank of the Somme to the other remained unimpaired.

Before leaving the line, the First Division had captured Cappy and advanced its line on the right to the western outskirts of Foucaucourt, while the Third Division had possessed itself of Suzanne. This was the situation when, on the night of August 26th, the Second and Fifth Divisions came into the line. Conferences with the four line Divisions were held both on the 25th and 26th August, in order to ensure co-ordinate action for the process of hustling the enemy across the Somme.

I was, at this stage, sorely perplexed by the uncertain attitude of the Fourth Army. I was all for pushing on energetically, and received General Rawlinson's approval to do so on August 24th; but on the very next day he enunciated a diametrically opposite policy, which greatly embarrassed me.

The gist of the Army attitude on the 25th may be thus expressed. The presence of a new German Division, the 41st, of whom we had taken many prisoners in Cappy, pointed to an intention on the part of the enemy to reinforce. This negatived any intention to undertake a withdrawal. This conclusion justified a revision of the Fourth Army policy. The Army had done its fair share; it had drawn in upon its front all the loose German reserves. Its resources in Tanks had been depleted, and it would take a month to replace them. Other Armies would now take up the burden, and the Fourth Army would now mark time, and await events elsewhere. There was no object in hastening the enemy's evacuation of the bad ground in the bend of the Somme, or in our taking possession of it. There was a possibility of the French taking over more frontage from us, and the Australian Corps front might in consequence be reduced to a three-Division front, with three Divisions in Corps Reserve.

The course of events, in the next seven days, convinced me that the results which were then achieved were totally unexpected by the Fourth Army, and very vitally influenced the whole subsequent course of the campaign. In point of fact, Lord Rawlinson quite frankly conceded to me as much in express terms a week later. The appreciation made at the time was doubtless an intentionally conservative one, but it did not take into account the reserve of striking power which remained in the Australian Corps, even after the past eighteen days of continuous fighting, and even without the assistance of the Tanks.

There was only one saving clause in the Army attitude, and this fortunately gave all the loophole necessary for the continued activity which I desired to pursue. It was this: "Touch must be kept with the enemy." This was of course a mere formality of tactics, and was intended as no more than such. But it was sufficient to justify an aggressive policy on my part.

As the result of my redispositions, completed by the night of August 27th, and of my conferences with the line Divisions, each Division stood on that morning on a single Brigade front, with its two remaining Brigades arranged in depth behind it. My orders were that in the event of the enemy giving way, the line Brigade was to push on energetically, and was to be kept in the line until it had reached the limits of its endurance. The other two Brigades were to follow up more leisurely, but to be prepared, each in turn, to relieve the line Brigade.

I had calculated that, by this method, each Brigade should be able to function for at least two days on the frontage allotted; and that, therefore, the present line Divisions could continue for at least six days; and if the stress upon the troops had not been severe, they could carry out a second rotation of Brigades for a second tour of six days. The calculation was, in general terms, fully realized; and all of the four line Divisions of that day did actually carry on for twelve days, and two of them for an additional six days.

The Artillery resources of the Corps were throughout the whole of this period fully maintained at the standard of the early days of August. I still had at my disposal eighteen Brigades of Field Artillery; and so was able to allot four Brigades of Artillery to each line Division, while keeping two in Corps Reserve.

Early on the morning of August 27th, a policy of vigorous patrolling all along our front was initiated. At several points, enemy posts which were known to have been strongly held the night before were found to be now unoccupied. Although reports varied along my front, they so fully confirmed my anticipations, that without waiting to make any reference to the Army, I ordered an immediate general advance along my whole front.

There followed a merry and exciting three days of pursuit; for the enemy was really on the run, and by nightfall on August 29th, not a German who was not a prisoner remained west of the Somme between PÉronne and Brie.

In previous years, during the enemy's retreat from Bapaume to the Hindenburg Line, we had had experience of his methods of withdrawal. Then they were deliberate, and his rearguards so methodically and resolutely held up the British advance, that the enemy had been able not only to remove from the evacuated area every particle of his warlike stores, which were of any value, but also to carry out a systematic devastation of the whole area, even to the felling of all the fruit trees, and the tearing up of all the railways for miles.

The present withdrawal was of a very different character. To begin with, it had been forced upon him by the battle of Chuignes, and he had to undertake it precipitately and without adequate preparation. Secondly, he had an impassable river behind him, which could be crossed only at three points, Brie, Eterpigny, and PÉronne. Thirdly, he had in front of him a Corps flushed with its recent victories, while he had been suffering a succession of defeats and heavy losses.

Nevertheless, he put up a good fight, and employed well-considered tactics. The German Machine Gun Corps was much the best of all his services. The manner in which the machine gunners stood their ground, serving their guns to the very last, and defying even the Juggernaut menace of the Tanks, won the unstinted admiration of our men. During these three days of retreat the enemy used his machine guns to the best advantage, and they constituted the only obstacle to our rapid advance.

These tactics were not unexpected by me, and I had an answer ready. Defying the whole traditions of Artillery tactics in open warfare, I insisted upon two somewhat startling innovations. The first was to break up battery control, by detaching even sections (two guns), to come under the direct orders of Infantry Commanders for the purpose of engaging with direct fire any machine-gun nest which was holding them up.

The second was to insist that all batteries should carry 20 per cent. of smoke shell. This elicited a storm of protest from the gunners. Every shell carried which was not a high explosive or shrapnel shell meant a shell less of destructive power, and, therefore, a shell wasted. That had been the Gunnery School doctrine. But I imagine that the test made at this epoch of the liberal use of smoke shell against machine guns will lead to a revision of that doctrine.

Smoke shell proved of inestimable value in blinding the German machine gunners. A few rounds judiciously placed screened the approach of our Infantry, and many a machine-gun post was thereby rushed by us from the flanks or even from the rear. General Hobbs (Fifth Division) and General Rosenthal (Second Division), both of whom had formerly been gunners, proved the strongest advocates for these smoke tactics.

By such means an energetic and successful pursuit was launched and maintained. By the night of August 27th, our line already lay to the east of the villages of Vermandovillers, Foucaucourt (on the main road) and Fontaine. We also mastered the whole of the Cappy bend, including the crossings of the Somme at Eclusier. The Fifth Division had a particularly hard fight at Foucaucourt, which did not fall to us until we had subjected it to a considerable bombardment. Tivoli Wood was the chief obstacle encountered that day by the Second Division. The advance of the 32nd Division also progressed smoothly.

During August 28th our advance was continued methodically, and by that night the Corps front had reached the line GÉnermont—Berry-en-Santerre—EstrÉes—Frise.

On August 29th the line of the Somme was reached, and all three Divisions south of the Somme stood upon the high ground sloping down to the Somme, with the river in sight from opposite ClÉry, past PÉronne and as far south as St. Christ.

In the meantime the Third Division north of the Somme had marched forward, in sympathetic step with the southern advance, successively seizing Suzanne, Vaux, Curlu, Hem and ClÉry. The Third Corps on my left had followed up the general advance, though always lagging a little in rear, thus keeping my left flank secure; and beyond the Third Corps, the Third Army was approaching the line of the Canal du Nord, which lay, as explained, in prolongation of the south-north course of the Somme.

The war correspondents of this time were given to representing the progress of the Australian Corps during these three days as a leisurely advance, regulated in its pace by the speed of the retiring enemy. But it was nothing of the kind.

On the contrary, it was his withdrawal which was regulated by the speed of our advance. There was not a foot of ground which was not contested by all the effort which the enemy was able to put forth. It is quite true that his withdrawal was intentional; but it is not true that it was conducted at the deliberate rate which was necessary to enable him to withdraw in good order.

He was compelled to fight all the time and to withdraw in disorder. He was forced to abandon guns and huge quantities of stores. The amount of derelict artillery ammunition found scattered over the whole of this considerable area alone reached hundreds of thousands of rounds, distributed in hundreds of dumps and depots, as well as scores of tons of empty artillery cartridge-cases, the brass of which had become of priceless value to the enemy.

Regimental and even Divisional Headquarters were abandoned as they stood, with all their furniture and mess equipment left intact. Signal wire and telephone equipment remained installed in all directions, hospitals and dressing-stations were left to their fate. The advance yielded to us over 600 prisoners, some half-dozen field-guns, and large numbers of smaller weapons.

The last two days of the advance led us across a maze of trenches and the debris of the 1916 campaign. The weather was unfavourable, there was much rain and an entire absence of any kind of shelter. As a result the line Brigades had to put forth all their powers of endurance and reached the Somme in a very tired condition.

In the meantime my air squadron had an exceptionally busy time. Contact patrols were maintained throughout every hour of daylight. Difficult as it was to identify the positions reached by our leading troops during an organized battle, where their approximate positions and ultimate objective lines were known beforehand, it was doubly so when no guide whatever existed as to the probable extent of each day's advance, or as to the amount of resistance likely to be encountered at different parts of the front.

Yet it was just under these circumstances that rapid and reliable information as to the progress of the various elements of our front line troops was more important than ever, and no means for obtaining such information was so expeditious as the Contact Aeroplane.

To assist the air observer in identifying our troops, the latter were provided with flares, of colours which were varied from time to time in order to minimize the risk of imitation by the enemy. The method of their employment, whether singly or in pairs, or three at a time, was also frequently varied.

These flares on being lit gave out a dense cloud of coloured smoke, easily distinguishable from a moderate height. The contact plane, which would carry coloured streamers so that the infantry could identify it as flying on that particular duty, would, when ready to observe, blow its horn and thereupon the foremost infantry would light their flares.

It was a method of inter-communication between air and ground, which, after a little practice, came to be well understood and intelligently carried out. By its means a Divisional or Brigade Commander was kept accurately informed, with great promptitude, of the progress of each of his front line units, in relation to the various woods, ruined mills, and other obstacles which lay spread across their path.

But the Air Force had another interesting duty, which was to watch the roads leading back from the enemy's front line to his rear areas. During tranquil times little movement could ever be seen on the enemy's roads in the hours of daylight, for the very good reason that he took care to carry out all his transportation to and from his front zone under cover of darkness.

Now, however, his needs pressed sorely upon him; and our air reports, from this time onwards, became almost monotonous in their iteration of the fact that large columns of transport were to be seen moving back in an easterly direction. These were his retiring batteries or his convoys of wagons carrying such stores as he was able to salve.

Occasionally, too, came reports of convoys, which looked like motor lorries or buses, moving hurriedly westward towards the German front. These were generally diagnosed by us as reinforcements which were being continually hurried forward to replace his human wastage, which was considerable both by direct losses from death, wounds and capture and by reason of the fatigue of such a strenuous and nerve-racking retreat.

All this movement in the enemy's rearward areas was a legitimate object of interest to my Artillery. But, unfortunately, most of it lay well beyond the range of my lighter Ordnance. The mobile Field Artillery was effective at no greater range than about four miles. The longer range 60-pounders found it a formidable task to traverse such broken country, while the still heavier tractor-drawn 6-inch guns found it quite impossible.

The latter, and all the Heavy and Super-Heavy guns and howitzers were tied down to the roads, and it proved a tremendous business to advance them in sufficient time and numbers to make their influence felt upon the present situation. I have nothing but praise for the admirable manner in which Brigadier-General Fraser and his Heavy Artillery Headquarters carried out the forward moves of the whole of his extensive Artillery equipment and organization from August 8th onwards to August 23rd. But the rapid advance of the battle line during the last week of August left the great bulk of Heavy Artillery far behind.

This was not entirely or even appreciably a question of the rate of movement of the great lumbering steam or motor-drawn heavy guns. They could quite easily march their eight or ten miles a day if they could have a clear road upon which to do it. But it was this question of roads that dominated the whole situation during this period, and subsequently until the end of the campaign of the Corps.

The construction and upkeep of roads throughout the Corps area had been, even in the days of stationary warfare, a difficult problem. At a time like the present, when the battle was moving forward from day to day, it became one of the first magnitude.

The rate of our advance was controlled almost as much by the speed with which main and secondary roads could be made practicable for traffic as by the degree of resistance offered by the enemy. Obstacles had to be removed, the debris of war cleared to one side, shell holes solidly filled in, craters of mine explosions bridged or circumvented, culverts repaired and drains freed of obstructions.

The road surfaces, speedily deteriorating under the strain and wear of heavy motor lorry traffic, had to be kept constantly under repair. The transportation of the necessary road stone for this purpose alone, imposed a heavy burden upon the roads and impeded other urgent traffic. The amount of road construction and reconstruction actually in hand within the Corps area, at any one time, far exceeded that normally required in peace time for any great city district.

The traffic on the roads was always of the most dense and varied character. For the proper maintenance and supply of a large Army Corps at least three good main roads, leading back to our sources of supply, would have been no more than adequate; but I seldom had at my disposal more than one such main road, which had often to be shared with an adjoining Corps.

There was ever an endless stream of traffic, labouring slowly along in both directions. On such a road as that leading east from Amiens towards the battle front, the congestion was always extreme. Ammunition lorries, regimental horsed transport, motor dispatch riders, marching infantry, long strings of horses and mules going to and from water, traction engines, convoy after convoy of motor buses, supply wagons, mess carts, signal motor tenders, complete batteries of Artillery, motor tractors, tanks, Staff motor cars and gangs of coolie labourers surged steadily forward, in an amazing jumble, with never a moment's pause.

Such were some of the difficulties with which I was beset in the rear of my battle line. They were negligible compared with those which now loomed in front of it.

The reach of the Somme which runs northerly from Ham past Brie to PÉronne and there turns westerly, differs entirely in its topographical features from that picturesque Somme Valley along both of whose banks the Corps had been fighting its way forward. The steep banks have disappeared, and for a mile or so on either side the ground slopes gently towards the river bed.

The river itself is not less than 1,000 yards wide, being, in fact, a broad marsh, studded with islets which are overgrown with rushes, while the stream of the river threads its way in numerous channels between them. The marsh itself is no more than waist-deep, but the flowing water is too deep to be waded.

Along the western side of this marsh runs the canalized river, or, as it is here known, the Somme Canal, flowing between masonry-lined banks. The construction of a crossing of such a marsh was, even in peace time, a troublesome business. It meant, to begin with, a causeway solidly founded upon a firm masonry bed sunk deep into the mud of the valley bed. The canal itself and each rivulet required its separate bridge, in spans varying from thirty to sixty feet.

What, therefore, came to be known as the Brie Bridge, situated on the line of the main road from Amiens to St. Quentin, really consisted of no less than eight separate bridges disposed at irregular intervals along the line of the causeway, between the western and eastern banks of the valley. The demolition of even the smallest of these eight bridges would render the whole causeway unusable, and would prohibit all traffic.

There exists an almost exactly similar arrangement of bridges at St. Christ, about two miles to the south of Brie, but no other traffic crossing to the north of Brie until PÉronne is reached. There, both the main road and the railway, which cross side by side, are provided with large span lattice girder bridges, over the main canal, while the marsh has been reclaimed where the town has encroached upon it. The river overflow is led through the town in several smaller canals or drains, all of them liberally bridged where crossed by roads and streets.

The PÉronne bridges are, therefore, no less indispensable, and no less easily rendered useless than those at Brie. Should such crossings be denied to me, it would be just possible to pass infantry across the valley, by night, by wading and swimming, or by the use of rafts, always provided that no opposition were to be met with. But to pass tanks or heavy guns, or even vehicles of the lightest description across the marsh, would have been quite impossible.

The Somme threatened, therefore, to be a most formidable obstacle to my further advance. It was incumbent upon me to assume that at the very least one of each series of bridges would be demolished by the enemy in his retreat. It would have been criminal folly on his part were it to have been otherwise; and I had had previous evidence of the efficiency of his engineer services.

Reconnaissances pushed out on the night of August 29th speedily verified the assumption that some at least of the bridges had been wrecked. It was ultimately ascertained that every single bridge in every one of the crossings named had been methodically and systematically blown to pieces.

There was only one tactical method by which such an obstacle could be forced by a frontal operation. By bringing up sufficient Artillery to dominate the enemy's defences on the east bank of the river valley, it might have been possible to pass across sufficient infantry to establish a wide bridge-head, behind which the ruined crossings could be restored, probably under enemy Artillery fire.

But it would have been a costly enterprise, and fraught with every prospect of failure, should the enemy be prepared to put up any sort of a fight to prevent it.

The value to me of the possession of the whole of the Somme Valley from ClÉry westwards, and the rapid repair of the bridges therein which I had been able to effect, will now become apparent. For it permitted the crystallizing into action of a project for dealing with the present situation, which had been vaguely forming in my mind ever since the day when I took over the Chipilly spur.

This was the plan of turning the line of the Somme from the north, instead of forcing it by direct assault from the west.

It may be argued that such a plan would have been equally practicable, even if the left flank of the Australian Corps had hitherto remained and now still lay south of the Somme, instead of well to the north of it. In that case other Corps on the north would have carried out that identical plan, which ultimately did achieve this important and decisive result.

I very much doubt it.

I had also had some experience of the futility of relying too much upon the sympathetic action of flank Corps, who usually had their hands full enough with their own problems, and had little time to devote to the needs of their neighbours. It would, moreover, have been disagreeable and inexpedient in the extreme to seek a right of way through the territory over which another Corps held jurisdiction. Corps Commanders were inclined to be jealous of any encroachment upon their frontiers, or upon the tactical problems in front of them.

Moreover, I wanted, more than anything else, that this should be an exclusively Australian achievement.

The situation being as it was, I possessed freedom of action, elbow room, and control not only of all the territory which I should require to use, but also of all the Somme crossings west of ClÉry.

Final Instructions to the Platoon—an incident of the battle of August 8th, 1918. The platoon is waiting to advance to Phase B of the battle.