The only course that remained open to me was to initiate immediate measures for taking the fullest advantage of the enemy's demoralization by exploiting the success obtained to the utmost possible extent. No later than on the afternoon of the battle of Hamel itself, orders were issued to all three line Divisions to commence most vigorous offensive patrolling all along the Corps front, with a view not merely to prevent the enemy from re-establishing an organized defensive system, but also ourselves to penetrate the enemy's ground by the establishment therein of isolated posts, as a nucleus for subsequent more effective occupation.

Enterprise of such a nature appeals strongly to the sporting instinct of the Australian soldier. Divisions, Brigades and Battalions vied with each other in predatory expeditions, even in broad daylight, into the enemy's ground, and a steady stream of prisoners and machine guns flowed in. On the nights of July 5th and 6th, the Fifth Division, now in the sector between the Ancre and the Somme, possessed themselves with very little effort of a strip of some three hundred acres of hostile positions, bringing our front line so near to Morlancourt as to make that village no longer tenable by the enemy.

On the same nights, and again on July 8th and 9th, the Second and Fourth Divisions advanced their lines by an average of two hundred to three hundred yards along their respective fronts, and this advance was, in the case of the Second Division, particularly valuable in carrying our front line over the crest of the plateau of Hill 104, and giving us clear and unbroken observation far into the enemy's country, in the directions of Warfusee and Marcelcave.

It was a period replete with instances of individual enterprise and daring adventure. One incident, characteristic of the varied efforts of these days, was the capture, single-handed, and in broad daylight, by Corporal W. Brown, V.C., of the 20th Battalion, Second Division, of an officer and eleven men of the German Army, whom he stalked as they lay skulking in a trench dug-out not far from his observation post, and terrorized into submission by the threat of throwing a bomb at them.

But perhaps the best testimony of the successful activities of my troops during this period, and of the serious impression which they made upon the enemy, can be gathered by extracts from his own documents, a number of which were captured during this and subsequent fighting. Of these, the following, issued by the Second German Army Headquarters (Von der Marwitz), are among the more interesting:

"The enemy has in his minor enterprises again taken prisoner a complete front line battalion and part of a support battalion. The reason is our faulty leadership."

"The enemy penetrated the forward zone of the 108th Division by means of large patrols at midnight, on July 8th, 1918, without any artillery preparation, and again on the same night at 11 p.m., with artillery preparation, astride of the Marcelcave—Villers-Bretonneux railway. He occupied the trenches where our most advanced outposts lay, and took the occupants, comprising fifteen men, prisoner. The larger part of the forward zone has been lost."

"In the case of the present trench Division, it has often happened that complete picquets have disappeared from the forward zone without a trace."

All the above refers to the period between July 4th and 12th. We read again under date July 13th:

"During the last few days the Australians have succeeded in penetrating, or taking prisoner, single posts or picquets. They have gradually—sometimes even in daylight—succeeded in getting possession of the majority of the forward zone of a whole Division."

"Troops must fight. They must not give way at every opportunity and seek to avoid fighting, otherwise they will get the feeling that the enemy are superior to them."

One last extract from these interesting papers:

"The best way to make the enemy more careful in his attempt to drive us bit by bit out of the outpost line and forward zone is to do active reconnaissance and carry out patrol encounters oneself. In this respect absolutely nothing seems to have been done. If the enemy can succeed in scoring a success without any special support by artillery or assistance from special troops, we must be in a position to do the same."

Our line in front of Villers-Bretonneux had for months run very close to the eastern outskirts of that town, a circumstance which cramped and embarrassed our defence of it. The enemy could peer into its streets and sweep them with machine guns. He had held in strength a locality known as Monument Wood, the ruins of a once prosperous orchard, and his possession of it had been a source of annoyance both to us and to the French, for it lay just opposite the international boundary posts.

The time seemed opportune for a set-piece operation designed to advance our line opposite the town by 1,000 yards, on a broad front, to dislodge the enemy from Monument Wood, gain valuable elbow room, and obtain mastery of the remainder of the plateau on which the town was built. I had actually completed the draft of a plan for such an operation, and had held a preliminary conference with my Staff to discuss it, when it became apparent that the nightly encroachments which the Second Division were effecting in this region would, in the course of a few days, achieve the capture of the whole of this territory without any special organized effort at all.

And so it proved; for before the middle of July, Rosenthal had succeeded in possessing himself, by such a process of "peaceful penetration," of the whole of the coveted area. It was a further evidence of the serious demoralization which our aggressive attitude of the preceding months had wrought among the German forces opposed to us.

The era of minor aggression by the Australian Corps was, however, about to draw to a close, and the situation was rapidly beginning to shape itself for greater events.

CHAPTER IV
TURNING THE TIDE

The course of events during June and July pointed to the conclusions, firstly, that the enemy contemplated no further offensive operations in the Somme Valley, and, secondly, that the condition of the whole German Second Army, astride of the Somme, offered every temptation to us to seize the initiative against it.

So far as the Australian Corps was concerned, however, my total frontage, which had been increased (as the result of our exploitation) to over eleven miles, precluded the possibility, with only four Divisions at my disposal, of maintaining, even if I could succeed in initiating, an ambitious offensive. The time was nevertheless ripe for action on a scale far more decisive than had become orthodox in the British Army in the past. Efforts on that method had been confined to a thrust, limited in point both of distance and of time, and followed by a period of inaction; they had often given the enemy ample leisure to recover, and to reorganize his order of battle.

To maintain an offensive, day after day, indefinitely, would require sufficient resources, particularly in infantry, to allow Divisions to be used alternatingly. Only in such a way, by having rested Divisions always available to alternate with tired Divisions, could a continuous pressure be maintained.

I took every opportunity of pressing these views upon the Army Commander, and expressed the readiness of the Australian Corps to undertake and maintain a long sustained offensive, provided that arrangements could be made to shorten my frontage from a three to a two-Division battle front, and to increase my resources, from the present four, to five or even six Divisions. It was further essential that in any advances attempted by us, other Corps must co-operate on both flanks.

It would be bad tactics to drive into the enemy's front a salient with a narrow base, for such a salient would make our situation worse instead of better, affording to the enemy the opportunity of artillery attack upon it from both its flanks as well as from its front. The salient must therefore be broad based in relation to its depth, and the base must ever widen as the head of the salient advances.

This principle implied that a large-scale operation of such a nature must be begun on a whole Army front, and that, even at its inception, at least three Corps must co-operate, to be aided by the entry of additional Corps on the outer flanks as the central depth developed. In other words, it was a project implying a large commitment of resources, and the urgent question was whether the time was yet ripe for taking the risks involved.

The matter, however, now became a subject at least worthy of practical discussion, and, during the days which followed Hamel, the Staffs of both the Corps and Army were kept busy with the investigation of data, maps, and information, while the availability of additional resources in guns, tanks and aeroplanes became the subject of anxious inquiry.

A circumstance which troubled me sorely was the fact that my Corps stood on the flank of the British Army, and that the troops on my right belonged to the French Army. The relations between the Australian troops and the Tirailleurs and Zouaves of the 31st French Corps (General Toulorge) had always been the very friendliest, and the joint "international" posts had been the scenes of hearty fraternization and of the evolution of a strange common vernacular.

This comradeship of "poilu" with "digger" did not, however, lessen the difficulties incidental to the joint conduct of a major Operation of War by two Corps of different nationalities, speaking different languages, with diverse tactical conceptions, and, above all, of substantially divergent temperaments. The French are irresistible in attack as they are dogged in defence, but whether they will attack or defend depends greatly on their temperament of the moment. In this they are totally unlike the British or Australian soldier who will at any time philosophically accept either rôle that may be prescribed for him.

In short, it was not possible to hope for an effective co-ordination of effort, controlled particularly by the minute observance of a time-table, on the part of the Australian and its adjacent French Corps, and I felt quite unprepared to count upon it. It was for this reason that I expressed to the Army Commander the hope that a British Corps might be obtainable to operate on my right flank in any undertaking that should be decided upon. Understanding that the greater part of the Canadian Corps was then unemployed, resting in a back area, I ventured to hope that this Corps might be made available, in the event of a decision that the proposal should be proceeded with.

My hesitation to accept the French as colleagues in such a battle was based not altogether on theoretical or sentimental grounds. The steady progress in mopping up enemy territory to the east of Villers-Bretonneux, which had been made by my south flank Division (the Second) as the aftermath of Hamel, soon produced a contortion of the Allied front line at this point which bade fair to prove just as troublesome to me as had been the great re-entrant opposite Hamel, which that battle had been specially undertaken to eliminate.

No persuasions on my part, or on that of my flank Division, could induce the adjacent French Division to extend any co-operation in these advances or to adopt any measures to flatten out the re-entrant which, growing deeper every day, threatened to expose my right flank. I am convinced that such hesitation was based upon no timidity, but was the result wholly of an entirely different outlook and policy from those which the Australian Corps was doing its best to interpret. But the experience of it made the prospect of punctual co-operation on their part in much more serious undertakings distinctly less encouraging.

The proposed offensive involved, therefore, far-reaching redispositions, comprising a substantial displacement southwards of the inter-Allied boundary, a lengthening by several miles of the whole British Western front, and an entire rearrangement of the respective fronts of the Third and Fourth British Armies. It is not surprising that a decision was deferred, while the project was being critically investigated from every point of view.

Then, suddenly, a new situation arose. On July 15th, the enemy opened a fresh attack against the French in the south. The scale on which he undertook it immediately made it patent to all students of the situation that he was probably employing his whole remaining reserves of fit, rested Divisions; that he meant this to be his decisive blow; and that whether he gained a decision or not, it would be his last effort on the grand scale.

It did not succeed; for just as he had once again reached the line of the Marne and had on July 17th achieved his "furthest south" at Château-Thierry, a beautifully timed counter-stroke by the French and Americans upon the western face of the salient, extending from Soissons to the Marne, resulted on July 18th in the capture by the Allies on that day alone, of 15,000 prisoners and 200 guns.

It was the end of German offensive in the war. Their mobile reserves were exhausted, and they were compelled slowly to recede from the Château-Thierry salient. The appropriate moment, for which Foch and Haig had doubtless been waiting for months, had at last arrived to begin an Allied counter offensive, and it was only a question of deciding at what point along the Franco-British front the effort should be made, and on what date it should open.

Doubtless influenced by the reasons already discussed, the choice fell upon that portion of the front of the Fourth Army which lay south of the Somme; in other words upon the southern portion of the Australian Corps front. The date remained undecided, but the requisite redisposition of Armies and Corps was so extensive that no time was to be lost in making a beginning.

It was on July 21st that General Rawlinson first called together the Corps Commanders who were to be entrusted with this portentous task. The strictest secrecy was enjoined, and never was a secret better kept; with the exception of the Field Marshal and his Army Commanders, none outside of the Fourth Army had any inkling of what was afoot until the actual moment for action had arrived.

Yet an observant enemy agent, if any such there had been in the vicinity, might well have drawn a shrewd conclusion that some mischief was brewing, had he happened along the main street of the prettily-situated village of Flexicourt, on the Somme, on that bright summer afternoon, and had observed in front of a pretentious white mansion, over which floated the black and red flag of an Army Commander, a quite unusual procession of motor-cars, ostentatiously flying the Canadian and Australian flags and the red-and-white pennants of two other Corps Commanders.

There were present at that conference, General Currie, the Canadian, General Butler, of the Third Corps, General Kavanagh, of the Cavalry Corps, and myself, while senior representatives of the Tanks and Air Force also attended. Rawlinson unfolded the outline of the whole Army plan, and details were discussed at great length in the light of the views held by each Corps Commander as to the tasks which he was prepared to undertake with the resources in his hands or promised to him.

The conditions which I had sought in my previous negotiations with the Army Commander were, I found, conceded to me almost to the full extent. My battle front was to be reduced from eleven miles to a little over 7,000 yards. It would, in fact, extend from the Somme, as the northern, to the main Péronne railway, as the southern flank. And—what was equally important, and profoundly welcome—the First Australian Division was shortly to be relieved in Flanders, and would at last join my Corps, thus for the first time in the war bringing all Australian field units in France under one command.

The Canadians were to operate on my right, and further south again the First French Army (Debenay) was to supply a Corps to form a defensive flank for the Canadians. The Third British Corps was to carry out for me a similar function on my northern flank. Thus, four Corps in line were to operate, the two central Corps carrying out the main advance, while the two outer flank Corps would be employed further to broaden the base of the great salient which the operation would create.

The Cavalry Corps would appear in the battle area also, with all preparations made for a rapid exploitation of any success achieved. The utility of the Cavalry in modern war, at any rate in a European theatre, has been the subject of endless controversy. It is one into which I do not propose to enter. There is no doubt that, given suitable ground and an absence of wire entanglements, Cavalry can move rapidly, and undertake important turning or enveloping movements. Yet it has been argued that the rarity of such suitable conditions negatives any justification for superimposing so unwieldy a burden as a large body of Cavalry—on the bare chance that it might be useful—upon already overpopulated areas, billets, watering places and roads.

I may, however, anticipate the event by saying that the First Cavalry Brigade was duly allotted to me, and did its best to prove its utility; but I am bound to say that the results achieved, in what proved to be very unsuitable country beyond the range of the Infantry advance, did not justify the effort expended either by this gallant Brigade or by the other arms and services upon whom the very presence of the Cavalry proved an added burden.

For the full understanding of subsequent developments both during and after the battle it becomes of special importance to consider the proposed rôle of the Third Corps in relation to my left flank. It is to be remembered that the Fourth Army decided that the River Somme was to be the tactical boundary between the two Northern Corps. It was not competent for me to criticize this decision at the time, but I am free now to say that I believed such a boundary to have been unsuitable, and the event speedily proved that it was.

It is always, in my opinion, undesirable to select any bold natural or artificial feature—such as a river, ravine, ridge, road or railway—as a boundary. It creates, at once, a divided responsibility, and necessitates between two independent commanders, and at a critical point, a degree of effective co-operation which can rarely be hoped for. It is much better boldly to place a unit, however large or small, astride of such a feature, so that both sides of it may come under the control of one and the same Commander.

This was especially the case in this part of the Somme Valley which is broad, and has an ill-defined central line, tortuous, and with the slopes on either side tactically interdependent; but most of all because, as I have already described, the high plateau on the north completely overlooks the relatively lower flats on the south of the river. The point I am trying to make should be borne in mind, for I believe it has been fully borne out by subsequent events.

The decision standing, however, as it did, it fell to the task of the Third Corps to make an assault (concurrently with that of the Australian Corps south of the river) for the capture of the whole of that reach of the river known as the Chipilly Bend, and of all the high ground on the spur which that bend enfolds. The object was to deprive the enemy of all ground from which he could look down upon my advancing left flank, or from which he could bring rifle or artillery fire to bear upon it.

The Third Corps was to operate on the front of one Division, the 58th, which, pivoting its left upon the Corbie—Bray road, was to advance its right—in sympathy with the advance of the left of the Australian Corps—until it rested upon the river about one mile downstream from Etinehem. It was a movement the success of which was rendered promising by the nature of the ground and the disorganized condition of the enemy between the Ancre and the Somme.

As regards my right flank, this was to rest as stated upon the main railway. The Canadian Corps, of four Divisions, would take over from the French a frontage of about 6,000 yards and deliver a thrust parallel to and south of the railway, in the direction of Caix and Beaucourt, and would aim at the seizure of the important Hill 102, immediately to the west of the latter locality. At no time did any question of the security of my right flank furnish me with any cause for anxiety; the prowess of the Canadian Corps was well known to all Australians, and I knew that, to use his own expressive vernacular, it was General Currie's invariable habit to "deliver the goods."

The comprehensive project thus outlined at the conference of July 21st involved, as a preliminary step, a far-reaching redisposition of very large bodies of troops over a very wide front. With the readjustment of the boundaries between the Third and Fourth British Armies we are not particularly concerned, because this affected a region, north of the Ancre, which lay well outside of the battle area. Nor did the internal readjustment of the northern part of the Fourth Army front present any difficulty, as it meant nothing more than a routine "relief" by the 58th Division of the Fifth Australian Division which was at this juncture holding that part of my Corps sector which lay between the Somme and the Ancre.

But the southern half was a very different matter. The First French Army was to give up to the British a section of about four miles, extending from Villers-Bretonneux to Thennes. This was ultimately to be taken over by the Canadian Corps as a battle front, but that Corps still had two of its Divisions in the line in the neighbourhood of Arras.

Moreover, it was of the utmost importance to conceal from the enemy until the last possible moment any change in our dispositions. This meant concealing them from our own troops also, because the loss by us of a single talkative prisoner would have been sufficient to disclose to the enemy at least the suspicion, if not the certainty, that an attack was in preparation.

After examining the problem and discussing several alternative solutions, it was ultimately decided at this conference that, five or six days before the date fixed for the attack, the French would be relieved in this sector by a Division, not of Canadians, but of Australians; that under cover of and behind this Australian Division, the Canadian Corps would come in from the north, and would proceed to carry out its battle preparations; and finally that the actual appearance of Canadian troops in the front line would not ensue until three days before the battle.

During the preceding two days, the Australian troops would be gradually withdrawn from the sector, leaving only one Brigade in occupation of the line, to be backed up by the incoming Canadians in the unexpected contingency of an attack by the enemy. This last Brigade would quietly melt away, leaving the Canadians in full possession of the field.

It was hoped that, during the days of the temporary Australian occupation of the sector, nothing would happen which might disclose to the enemy that the French had left it; and even if we were to have the misfortune to lose from this sector any Australian prisoners to the enemy, it was further hoped that, if kept in total ignorance of the inflow of Canadians, such prisoners would be unable to make any embarrassing disclosures. The dénouement, which will be told later, showed that this judgment of possibilities was a shrewd one, and that such precautions were not taken in vain.

At this period of the war, large numbers of Americans had already arrived in France, but only few of them were yet ready to take their places in the line of battle. The time had not yet arrived, therefore, when, by taking over large sections of the Western front they could help to shorten the French and British frontages. The British front was, therefore, still so extended that the mobile reserve Divisions at the disposal of the Field Marshal were few.

This consideration made the contemplated reliefs and interchanges of Corps and Divisions, and their transference from one part of our front to another a matter of great complexity, and one which required time to execute. Each stage of the process was contingent upon the due completion of a previous stage. It is, moreover, a process which cannot be unduly hastened, without serious discomfort and fatigue to the troops and animals concerned.

Troops destined for battle must be kept in the highest physical condition. This means good feeding, comfortable housing, and adequate rest. A couple of weary days and sleepless nights spent in crowded railway trains, with cold food and little exercise, are sufficient to play havoc with the fighting trim of even a crack battalion. So, the daily stages of the journey must be short, and comfortable billets must be in readiness for each night's halt. The day's supplies must arrive punctually and at the right railhead, to ensure hot, well-cooked meals.

With the very limited number of serviceable railway lines which remained available behind the British front—and with the congestion of traffic resulting from the daily transportation of many thousands of tons of artillery ammunition and other war stores—it was not surprising that as the result of the deliberations of the conference it was resolved to advise the Commander-in-Chief that it would take not less than five days to rearrange our order of battle on the lines decided upon, and another five days, after Corps and Divisions had taken over their battle fronts, to enable them to complete their preparations.

Thus, the Fourth Army could be ready at ten days' notice, and the conference broke up, pledged to secrecy and complete inaction, until formal approval had been given to the proposals and a date fixed for their realization.

The remainder of July passed with no very startling occurrences. In the south the German withdrawal from the Soissons salient and the Marne continued steadily, with the French and Americans on their heels; but it was a methodical retreat, which would bring about a substantial shortening of the German line, and so release Divisions to rest and refit, which might conceivably become available for a fresh assault elsewhere.

But there was still no sign of any such design upon that always tender spot, the Allied junction at Villers-Bretonneux. On the contrary, my second Division still continued to make free with the enemy's advanced patrols, and in a very brilliant little infantry operation by the 7th Brigade captured the "Mound," a long spoilbank beside the railway at a point about a mile east of the town, which dominated the landscape in every direction. The ardour of his troops was only enhanced when they heard that General Rosenthal himself, while reconnoitring from the Mound, had been sniped at and had received a nasty wound in the arm.

The enemy attempted nothing in the way of infantry retaliation. But whenever he had been thoroughly angered, he treated my front to a liberal drenching of mustard gas, fired by his artillery. His supplies of mustard gas shell seemed inexhaustible, and he would frequently expend as many as 10,000 of them in a single night upon the half-ruined town of Villers-Bretonneux or on the Bois l'Abbé and other woods which he suspected were sheltering my reserve infantry.

These gas attacks were annoying and troublesome, in the extreme. During the actual bombardments, troops wore their gas masks as a matter of course, but doffed them when the characteristic smell of the gas had disappeared. But it was warm weather, and as the sun rose, the poisonous liquid, which had spattered the ground over immense areas, would volatilize, and rise in sufficient volume still to attack all whose business took them to and fro across this ground. In this way hundreds of our men became incapacitated; although there were a few serious cases, most of the men would be fit to rejoin in two or three weeks. But this form of attack, and the constant dread of it, made life in the forward areas anything but endurable.

I was beset by quite another trepidation also. Prisoners captured during the German withdrawal from the Marne, which was then in progress, told tales of contemplated withdrawals on other fronts, and some even asserted that a withdrawal opposite my own front was being talked of. Judged by subsequent events, it is more than probable that these stories were stimulated by the many articles which were at that time appearing in the German newspapers from the pens of press strategists, who, in order to allay public anxiety, were representing these withdrawals as deliberate, and as a masterpiece of strategy, compelling the Allies to a costly pursuit over difficult and worthless ground.

Opposite Albert, signs that such a withdrawal was actually in progress also began to appear, although it subsequently transpired that, in its early stages, this procedure was merely prompted by a purely local consideration, namely, the desire of the enemy to improve his tactical position by abandoning the outposts, which he had been maintaining in the valley of the Ancre, and transferring them to the higher and better ground on the east of that river.

It was only natural that those of us who knew of the impending attack, and of the immense effort which its preparation would involve, felt nervous lest the enemy might forestall us by withdrawing his whole line to some methodically prepared position of defence in the rear, just as he had done once before in 1917 on so large a scale in the Bapaume region. It would probably have been a sound measure of military policy, but it would assuredly, at that juncture, have had as disastrous an effect upon the moral of the German people as his enforced withdrawal, which was soon to begin, actually produced not long after.

The order to prepare the attack, and fixing the date of it for August 8th, came in the closing days of July, and at once all was bustle and excitement in the Australian Corps. Commanders, Staff Officers, and Intelligence Service, the Artillery, the Corps Flying Squadron, the map and photography sections spent busy days in reconnaissance, and toilsome nights in office work. The vast extent of the detailed work involved, particularly upon the administrative services, can only be appreciated by a study of the plan for the battle, which it fell to my lot, as Corps Commander, first to formulate, and then to expound to a series of conferences which were held at Bertangles on July 30th, and on August 2nd and 4th.

It is, therefore, perhaps appropriate that I should now attempt to repeat, in non-technical language, an exposition of the outlines of that plan.

CHAPTER V
THE BATTLE PLAN

My plan for the impending battle involved the employment of four Divisions in the actual assault, with one Division in reserve. The Reserve Division was to be available for use in one of two ways; either as a reserve of fresh troops to exploit any successes gained upon the first day, or else to take over and hold defensively the ground won, if the assaulting Divisions should have become too exhausted to be relied upon for successful resistance to a counter-attack in force.

The frontage allotted to the Corps was 7,000 yards, and this extent of front accommodated itself naturally to the employment of two first-line Divisions, each on a 3,500 yard front, each Division having two Brigades in the front line, with one Brigade in reserve.

As four Divisions were available to me for immediate use in the battle, I decided to undertake, for the first time in the war, on so comprehensive a scale, the tactical expedient of a "leapfrog" by Divisions over each other.

This term had, long before, passed into the homely phraseology of the war, in order to describe a procedure by which one body of troops, having reached its objective, was there halted, as at a completed task, while a second body of troops, of similar order of importance, but under an entirely separate Commander, advanced over the ground won, reached the foremost battle line, took over the tactical responsibility for the fighting front, and after a prescribed interval of time continued the advance to a further and more distant objective.

This conception of an advance by a process of "leapfrog" had been evolved early in 1917 in connection with a method of assault on successive lines of trenches. It was intended at the outset to be applied only to very small bodies of infantry, such as platoons. A normal battle plan for a company of infantry of four platoons was for the first two platoons to capture and hold the front line trench, while the next two following platoons would leap over this trench and over the troops who had gained it, and then pass beyond to the capture of the second, or support trench. The method was used, for the first time, on such a modest scale, at the battle of Messines, in June, 1917, and later on in the same year was adopted for bodies as large even as Battalions, in the fighting for the Broodseinde and Passchendaele heights.

But on no previous occasion had such a principle been applied to whole Divisions. It is true that at the battle of Messines, the Fourth Australian Division passed through the New Zealand Division after the latter had completed the capture of the main Messines ridge, but this was really exploitation, undertaken in order to take advantage of the temporary confusion of the enemy, and for the purpose of gaining ground upon the eastern slopes of the captured ridge. It was not a movement which was really part of the main assault, and it was confined to a single Division.

On the present occasion my purpose was to carry out a clear and definite process of "leapfrogging," not only simultaneously by two Divisions side by side, but also as an essential part of the time-table programme for the main battle, and before the exploitation stage of the fighting was timed to be reached. It was, undeniably, a daring proposal, involving very definite risks, enormously increasing the labour of preparation and the mass of detailed precautions which had to be undertaken in order to obviate the possibility of great confusion.

The preparations necessary for a single Division proposing to advance alone, to a prescribed distance, over country much of which was usually visible to us from our front line, are sufficiently complex, relating as they do, not only to the establishment of numerous protected headquarters for Brigades and Battalions, of miles upon miles of buried and ground cables, of dumps of all kinds of supplies, and of dressing stations and medical aid posts; but also to the disposition, in concealed positions, of all the assaulting units, down to the smallest of them, of Infantry Engineers and Pioneers. All these preparations assume a tenfold complexity when a second Division has to make arrangements exactly similar in character, variety and extent, using exactly the same territory for the purpose and at the same time, and planning to advance over more distant country, entirely beyond visual range and preliminary reconnaissance.

The project also involved a much greater crowding of troops into the areas immediately behind our line of departure, and, therefore, enormously increased the risk of premature detection by the enemy, both from ground and from air observation, of unusual movement and of other symptoms which presaged the possibility of an attack by us. The plan also necessitated the closest possible co-ordination of effort, and mutual sympathy and understanding, between the Commanders and Staffs of the twin Divisions having a common jurisdiction over one and the same area of preparation, and one and the same battle front. This was a degree of co-operation which could not have been looked for unless the personnel concerned had already established, from long and close association with each other, the most cordial personal relations. And dominating all other difficulties were those involved in the proposal to execute this difficult and untried operation of a Divisional leapfrog, not singly but in a duplex manner, necessitating the assurance of exactly similar simultaneous action, similarly timed in every stage, both before and during battle, by each of two separate pairs of Divisions.

These threatening difficulties were surely formidable enough, but I knew that I could rely upon the goodwill of the Divisions towards each other, and upon the loyal support of them all. This seemed to me to justify the attempt, and to minimize the risks; having regard above all else to the results which I stood to gain if the operation could be executed as planned.

On no previous occasion in the war had an attempt ever been made to effect a penetration into the enemy's defences at the first blow, and on the first day, greater than a mile or two. Rarely had any previous set-piece attack succeeded in reaching the enemy's line of field-guns. The result had been that the bulk of his Artillery had been withdrawn at his leisure, and his losses had been confined to a few hundred acres of shattered territory. But the task I had set myself was not only to reach, at the first onslaught, the whole of the enemy's Artillery positions, but greatly to overrun them with a view to obliterating, by destruction or capture, the whole of his defensive organizations and the whole of the fighting resources which they contained, along the full extent of my Corps front.

To achieve this object I prepared my plans upon the basis of a total advance, on the first day, of not less than 9,000 yards. This was to be divided into three separate stages, as follows:

The opening phase involved no novel or unusual features so far as the infantry were concerned, and was conceived on lines with which the fighting of 1917 had familiarized me, modified further by the accumulated experience gained from earlier mistakes in the technical details of such an enterprise. The recent battle of Hamel became the model for this phase, the conditions of that battle being now reproduced on a much enlarged scale.

But there was one very important feature which distinguished the present undertaking from the battles of Messines and Broodseinde, and that was in regard to the frontage allotted for attack to a single Division. At Messines, the Divisional battle front was 2,000 yards; in the third battle of Ypres it differed but little from the same standard. For the present battle, I adopted a battle front of two miles for each assaulting Division, or a mile for each of the four assaulting Brigades.

This innovation seemed to me to be justified by four principal factors. The first of these was that the weather, which was dry, and the state of the ground, which was hard, made the "going" easy and the stress upon the infantry comparatively light. Next, the condition of the enemy's defensive works was undeveloped and stagnant, as clearly disclosed by the air photographs which the Corps Air Squadron produced in great numbers on every fine day. No doubt this was due to the encroachments we had made on his forward works during the fighting at Hamel and in the remaining weeks of July. Thirdly, the powerful assistance anticipated from a contingent of four Battalions of Tanks which General Rawlinson had arranged to place under my orders led me to estimate that I might greatly reduce the number of men per yard of front. Lastly, the plan was justified by the known distribution of the enemy's infantry and guns along the frontage under attack. For all these reasons, I felt prepared to impose on the infantry a task which, computed solely upon the factor of frontage, was more than twice that demanded by me on any previous occasion.

At the same time, so extended a frontage involved the employment of a much higher ratio of barrage artillery to the number of battalions of Infantry actually engaged. Success depended more upon the efficiency of the fire power of the barrage than upon any other factor, and I could not afford to incur any risk by weakening the density of the barrage. For this reason, I adhered to the standard which previous experience of several major battles and many minor raids had shown to be adequate for covering the assaulting infantry, and for keeping down the enemy's fire. This standard never fluctuated widely from one field-gun per twenty yards of front, and involved the employment, on this occasion, of some 432 field-guns in the barrage alone. This result could not have been achieved if the Fourth Army authorities had not seen their way to place at my disposal five additional Brigades of Field Artillery over and above the thirteen Australian Brigades which formed a permanent part of the whole Artillery of the Corps.

Phase A, as already stated, involved a penetration of 3,000 yards, and the objective line for this phase, which came to be known as the "green" line (from the colour employed to delineate it upon all the fighting maps propounded by the Corps), was chosen, after an exhaustive study of all aeroplane photographs, and of the results of numerous observations, by many diverse means, of the locations of the enemy's Artillery, so as to make certain that during this phase the whole mass of the enemy's forward Artillery would be overrun, and captured or put out of action.

The green line was, in fact, located along the crest of the spur running north-easterly from Lamotte-en-Santerre in the direction of Cerisy-Gailly, with the object of carrying the battle well to the east of the Cerisy valley, in which large numbers of the enemy's guns had been definitely located. This would give us, by the capture of this valley, suitable concealed positions in which the Infantry destined for Phase B could rest for a short "breather;" and would land the Infantry of the original assault in a position from which they could detect and forestall any attempt on the part of the enemy to launch a counter-attack before the time for the opening of Phase B had arrived.

The task of executing Phase A of the battle fell to the Second and Third Australian Divisions, in that order from south to north, the southern flank of the Second Division resting upon the main railway line from Amiens to Péronne, and being there in contact with the Canadian Corps, under General Currie. The northern flank of the Third Division rested on the River Somme, and was there in contact with the Third British Corps under General Butler, while the inter-divisional boundary was at the southern edge of the Bois-d'Accroche.

These two Divisions were the line Divisions during the period immediately preceding the battle, and had been holding the line each with two Brigades in line and one Brigade in support. Three days prior to the battle, however, it was arranged that each Division should hold its front with only one Brigade, thereby making available two Brigades each for the actual carrying out of Phase A of the attack. These assaulting Brigades were the 7th, 5th, 9th and 11th, in that order from south to north, each Brigade having its due allotment of Tanks and machine guns, etc.

The total estimated time for the completion of Phase A was to be 143 minutes after the opening of the barrage at "zero" hour; and there was then to be a pause of 100 minutes to allow time for the advance and deployment into battle order of the succeeding two Divisions, who were to carry out the process of "leapfrogging" and to execute Phases B and C of the battle.

The planning of Phase B, or the advance from the "green" to the "red" line, involved a totally different tactical conception and the adoption of a type of warfare which had almost entirely disappeared from the Western theatre of war since those far-off days in the late autumn of 1914, when the German Army first dug itself in, in France and Belgium, and committed both combatants to the prolonged agony of over three years of stationary warfare. I allude to the moving battle, or as it is called in text-book language, "open warfare;" a type of fighting in which few of the British Forces formed since the original Expeditionary Force had any experience except on the manœuvre ground under peace conditions—a disability which applied equally to the Australian troops. Confident, however, in their adaptability and in their power of initiative under novel conditions, I did not hesitate to prescribe, for this second phase of the battle, the adoption of the principles and methods of open warfare.

In two very important respects in particular, this type of fighting involved conditions to which the troops had not been accustomed, and under which they had no previous experience in battle. In trench warfare, and in a deliberate attack on entrenched defences, the positions of all headquarters, medical aid posts, supply dumps and signal stations remained fixed and immovable. The whole of the internal communications by telegraph and telephone could, therefore, be completely installed beforehand, down to the last detail, and the transmission of all messages, reports, orders and instructions, during the course of the battle, was rapid and assured. But in a moving battle no such comprehensive or stable signalling arrangements are possible, and reliance must be placed upon the much slower and much more uncertain methods of transmission by flag and lamp signalling, by dispatch riders, pigeons and runners.

Divisional Headquarters would, therefore, almost as soon as the battle commenced, fall out of touch with Brigades, and they in turn with their Battalions; information as to the actual situation at the fighting front would travel slowly, and would reach those responsible for making consequential decisions often long after an entire alteration in the situation had removed the need for action. Thus, a greatly enhanced responsibility would come to be imposed upon subordinate leaders to decide for themselves, without waiting for guidance or orders from higher authority, and to grasp the initiative by taking all possible action on the spot in the light of the circumstances and situation of the moment.

Again, the nature of the Artillery action is, in the moving battle, fundamentally different from that which prevails during trench warfare. To begin with, only that portion of the Artillery which is in the strictest sense mobile can participate to any extent in open warfare. The employment of Artillery is, therefore, confined to a few and to the smaller natures of Ordnance, namely, the 18-pounder field-gun, the 4½-inch field howitzer and the 60-pounder, which are all horse drawn and which are capable of being moved off the roads and across all but the most broken country. Heavier guns, from 6-inch upwards, are in practice confined to roads, and are too slow and cumbersome to keep pace with the Infantry. The Artillery fire action is also intrinsically different, because the guns can be sighted directly upon their targets, while in trench warfare they are always laid by indirect methods, with the use of the map and compass, and without observation, at any rate by the crew of the gun, of the objects fired at.

The decision which I had to take of carrying out the second phase of this great battle on the principles of open warfare was, therefore, one which also involved a certain element of risk. But it was a risk which I felt justified in taking, in spite of the fact that the German High Command had more than once expressed itself in contemptuous terms of the capacity of any British troops successfully to undertake any operation of open warfare. My justification lay primarily in my confidence in the ability of the subordinate commanders and troops to work satisfactorily under these novel conditions—a confidence which the event abundantly justified. But I was placed in the position of having either to accept this risk, or else abandon altogether the project of a quite unprecedented penetration of enemy country to be completed on the first day. It would have been clearly impossible to continue the advance beyond the green line without an interval of at least forty-eight hours, which would have been necessary to enable the Artillery to be redisposed for barrage fire in forward positions and provided with the necessary supplies of ammunition for such a purpose.

The Divisions which were told off to carry out the "leapfrog" enterprise and to execute Phase B of the battle were the Fifth Australian Division on the south and the Fourth Australian Division on the north, the outer flanks of the attack remaining as before, i.e., the Péronne Railway on the south and the River Somme on the north. Each of these Divisions was directed to deploy, on its own frontage, two Infantry Brigades. Its third Brigade was to be kept intact and to advance during Phase B at some distance behind, as a support to the fighting line, and to be employed in the subsequent phase, if it were found that Phase B could be completed without calling upon this spare Brigade. The actual dispositions of the Brigades finally proposed by the respective Divisional Commanders and approved by me brought about the arrangement that the four first-line mobile Infantry Brigades were successively, from south to north, the 15th, 8th, 12th and 4th, while the 14th and 1st Brigades followed as supports in a second line.

To each of these Infantry Brigades I allotted a Brigade of Field Artillery, to be employed under the direct orders of the Infantry Brigade Commander, and, in addition, three Artillery Brigades as well as one Battery of 60-pounders, to each Divisional Commander. As my resources in Artillery were not unlimited, the twelve Artillery Brigades, so disposed of, were necessarily drawn from the original eighteen Brigades which were to fire the covering Artillery barrage for Phase A of the battle. The orders to that portion of the Field Artillery which was to become mobile in pursuance of this plan, accordingly, were that immediately upon the completion of their original tasks, by the capture of the green line, they were to "pull out of the barrage."

This meant, in effect, that all the teams, limbers, battery wagons, and ammunition wagons of these twelve Brigades, waiting in their wagon lines far in rear, fully harnessed up and hooked in at the opening of the battle, had to advance during the progress of the first phase, so as to reach their guns just at the right time, but no earlier, to enable these guns to be limbered up, and the batteries to become completely mobile in order to join and advance with the Infantry of the second phase.

This was an operation which required the greatest nicety in timing, and the greatest accuracy in execution. No Australian Artillery had ever previously undertaken such an operation, except perhaps on the manœuvre ground, and then only on the very limited scale of a Brigade or two at a time. That this rapid transition from the completely stationary to the completely mobile battle was carried out, during the very crisis of a great engagement, without the slightest hitch, and with only the trifling loss of two or three gun horse teams from shell fire, reflects the very highest credit upon every officer and man of the Australian Field Artillery.

The open warfare Infantry Brigades were also to be provided, out of their own divisional resources, each with a Company of Engineers, a Company of Machine Guns, a Field Ambulance, and a detachment of Pioneers, so that, in the most complete sense, they became a Brigade Group of all arms, capable of dealing, out of their own resources and on their own ground, with any situation that might arise during their advance of nearly three miles from the green to the red line. A detachment of nine tanks completed the fighting equipment of each of the four front line Brigades destined to capture the red line.

I must now briefly describe the nature of Phase C, the third and last stage in this ambitious and complex battle programme. This phase was to consist of "exploitation," which implies that it was a provisional preparation, which was to be carried out only if complete success attended the two preceding phases. The objective of Phase C was the "blue" line, which I had located about one mile to the east of the red line, along a system of old French trenches extending from the river at a point near Méricourt, and running southerly to the railway at a point a little to the south-east of Harbonnières. This line gave promise of furnishing a good defensive position in which to deal with any possible counter-attack. It also gave a good line of departure for subsequent operations, and provided ideal artillery positions in a series of valleys, running parallel and a little to the west of the line itself.

The troops earmarked for this Exploitation Phase were the two second line Brigades of the two Divisions which were to capture the red line, namely, the 14th and 1st Brigades, and the orders to the Divisional Commanders were that if the red line was reached without mishap, without undue loss of time, and without involving the Reserve Brigades, but not otherwise, these Reserve Brigades were to push on with the utmost determination to secure and hold the blue line until such time as they could be reinforced.

Each of these exploitation Brigades was equipped similarly to the red line Brigades in all respects except that they were provided with a special contingent of 18 Mark V. (Star) Tanks of the very latest design. These differed from the Mark V. Tank employed at Hamel and in the other stages of the present operation, in that they were longer and had sufficient internal space to carry, as passengers, over and above their own crews, two complete infantry Lewis gun detachments each. It was expected that this infantry fire power, added to the fire power from the machine guns carried by these 36 Tanks themselves and operated by the Tank crews, would go far to compensate for the somewhat attenuated line of probably tired Infantry spread in two Brigades over an ultimate frontage of over 10,000 yards.

No definite time-table was laid down for the closing phases of the battle, except for the regulation of the times when our Heavy Artillery should "lift off" designated targets—such as villages, farms, and known gun positions—and lengthen its range so as not to obstruct the further advance of our own Infantry. But it was estimated that, from the opening of the battle, the green line would be reached in two and a half hours, the red line in six hours, and the blue line in eight hours. As the battle was to open at the first streak of dawn, it would, if all went well, be completed according to plan by about midday.

In every battle plan, whether great or small, it is necessary first of all to map out the whole of the intended action of the Infantry, at any rate on the general lines indicated above. When that has been done the next step is to work backwards, and to test the feasibility of each body of infantry being able to reach its allotted point of departure, punctually, without undue stress on the troops, and without crossing or impeding the line of movement of any other body of infantry. It is often necessary to test minutely, by reference to calculations of time and space, more than one alternative plan for marshalling the Infantry prior to battle, and for the successive movements, day by day, and from point to point, of every battalion engaged.

The present case was no exception, and, indeed, presented quite special difficulties. The whole of the area for a depth of many thousands of yards behind our then front line was open rolling country, devoid of any cover, and (except in the actual valley of the Somme) with every village, hamlet, farmhouse, factory and wood obliterated. The plan involved the assembly, in this confined area, fully exposed by day to the view of any inquisitive enemy aircraft, of no less than 45 Infantry Battalions, with all their paraphernalia of war; not to speak of our 600 guns of all calibres, their wagon lines, horse lines and motor parks, together with Engineers, Pioneers, Tanks, Medical and Supply Units amounting to tens of thousands of men and animals.

A new factor which, however, ultimately controlled the final decision which I had to make as to the nature of the dispositions prior to battle, lay in the consideration of the maximum distances which would have to be covered by the foot soldiers in such a far-flung battle. I had little difficulty in coming to the conclusion that the obvious and normal arrangement was on this occasion a totally wrong arrangement. If the assaulting Brigades had been arranged, from front to rear, in their assembly areas prior to battle, in the same order as that in which they would have to come into action, this would have involved that the individual man, who was to be required to march and fight his way furthest into enemy country, and, therefore, was to be the last to enter the fight, would also be called upon to march furthest from his rearmost position of assembly before even reaching the battle zone. The maximum distance to be traversed on the day of battle by infantry would have amounted, according to such a plan, to over ten miles. While this is an easy day's march on a good road, under tranquil conditions, it would have been an altogether unreasonable demand upon any infantryman during the stress and nervous excitement of battle. It would have been courting a breakdown from over-fatigue, among the very troops upon whom I had to rely most to defend the captured territory against any serious enemy reaction.

I therefore adopted the not very obvious course of completely reversing the normal procedure, and of disposing the Brigades in depth, from front to rear, in exactly the reverse of the order in which, in point of time, they would enter the battle.

The following represents, diagrammatically, the disposition of all twelve Brigades after having been fully deployed in the actual course of the battle: