Chapter VI.
From Egypt to France: Trench Warfare at Armentières.

The New Zealand and Australian Division was now one of the Divisions available for the defence of the Suez Canal, but being in reserve had only a very few active duties to perform. After supplying guards for the reservoir, railway bridge, and aerodrome, the Division was free to devote the whole of its energies to re-organization and re-equipment (badly needed after the rigours of the Gallipoli campaign) and training. In this way January and February passed.

Meanwhile, arrangements were being completed for the establishment of a New Zealand Division, and the 2nd Infantry Brigade came into being on March 1st. One of the new battalions created for the new brigade was the 2nd Canterbury Battalion, commanded by Major H. Stewart, who was now promoted to the rank of Lieutenant-Colonel. He brought with him from the 1st Battalion Major G. C. Griffiths (second in command of the new battalion), Lieutenants A. L. Ford, D. P. Fraser, N. F. Shepherd, F. Starnes, and N. R. Wilson, and Second Lieutenants H. Campbell, M. W. Duncan, H. S. Gabites, and W. J. Marriott, and thirty-seven non-commissioned officers. The whole of the Canterbury draft of the 7th Reinforcements (which had been delayed in Cairo on guard duty over Turkish prisoners) was posted to the new battalion; and this and other drafts, including a number of officers^* and men from the Mounted Rifles Regiments brought up the strength of the battalion to thirty officers and seven hundred and thirty-five other ranks on March 21st.

Definite orders that the Division was soon to go to France were received in March. On the 21st, His Royal Highness the Prince of Wales visited the Canal area, and saw the Division at work; but there was no ceremonial parade in his honour. The last parade of this nature held in Egypt was on April 3rd, when General Sir Archibald Murray, Commander-in-Chief of the Egyptian Expeditionary Force, inspected the Division in full marching order.

On April 6th, the 1st Canterbury Battalion (with the exception of its first line transport personnel and horses) entrained at Ismailia, and on arrival at Port Said the same day, embarked on the Franconia. The first line transport went by train to Alexandria and embarked there on the Cestrian, the whole operation being completed in one day—the 6th. The 2nd Battalion entrained at Ismailia on the 7th, and embarked the following day on the Canada (thirteen officers and five hundred and five other ranks), Ascania (ten officers and one hundred and fifty other ranks), and Haverford (four officers and one hundred and fifty other ranks).

The transports did not sail in a single fleet. Thus the Franconia sailed on April 8th and arrived at Marseilles on the 11th; and the 1st Battalion disembarked the following day. The 2nd Battalion's transports sailed later, and arrived at Marseilles between April 14th and 16th.

On arrival, the troops of both battalions entrained immediately for Steenbecque, three miles south-west of Hazebrouck. The journey was a long one, the troops being in the train for between sixty and seventy hours; but its tediousness was relieved by the beauty of the French countryside, which was all the more welcome to men who had spent so many months on the edge of the desert. The 1st Canterbury Battalion on arrival at Steenbecque marched to a camp at Morbecque, two page 83 miles from Hazebrouck; and it did not exchange its tents for billets in the same village till April 19th. Part of the 2nd Battalion detrained at Steenbecque on April 17th, and the remainder detrained at Hazebrouck on the 20th; and both had a very long and tiring march to their billets at Roquetoire, west of the St. Omer-Aire road and three miles north-west of the latter town. The transport personnel and animals of both battalions went direct by train to Abbeville, where harness vehicles, and all necessary equipment were issued to them. The experience of the transport is an interesting illustration of the good organization of the Army Ordnance at Abbeville. On a given date the number of horses was made up to the establishment of an infantry battalion, and harness for all the horses was issued. The transport officers were then instructed to report at another place, at a certain time on another day, with all their animals ready harnessed. On arrival there, they found all the limbers, G.S.^* waggons, field kitchens, and other vehicles on the establishment of an infantry battalion ranged up in proper order, with all equipment in its proper place, and loaded with the rations required for man and beast. After being asked to satisfy themselves that everything they should have was in its place, the officers were asked to sign for what they had received, and then were given in succession the orders to yoke up, to mount, and to march off on trek to join their battalions.

While the Division was occupied in this fashion, its equipment had been completed, so that at the end of April it was ready to take its turn in the front line. On April 30th, orders were received that the Division was to remain in reserve for the meantime, but would go into the line about May 20th. It was warned at the same time, however, that a gas attack by the enemy was expected before that date; and that when the attack was made, it was highly likely that the Division would be required to move up at very short notice, to support the troops attacked. It may be noted here that this attack did actually take place; but that the British anti-gas protection proved so efficient that the losses inflicted on the troops in the line were very slight, and consequently the New Zealand Division was not called upon to move.

Early in May, the Division prepared for its move to the line, by sending the 2nd Brigade to the Doulieu area and the 3rd Brigade to the Estaires area (both south-west of Armen-tieres), the 1st Brigade remaining for the meantime near Morbecque. In consequence of these moves, the 2nd Canterbury Battalion (2nd Brigade) marched from Roquetoire, on May 1st, to billets at Neuf Berquin—a distance of twenty miles; but the 1st Battalion (1st Brigade) remained at Morbecque till the 9th.

The I Anzac Army Corps at this time consisted very largely of troops who had never been under fire; and while every battalion of the 1st and 2nd New Zealand Brigades had a considerable stiffening of officers, non-commissioned officers, and men who had seen much active service on Gallipoli, none had any experience of the different conditions of trench warfare in France. Two battalions of the 3rd (Rifle) Brigade had had a small experience of guerilla warfare, against the Senussi in north-west Egypt; but the remainder of this brigade had seen no active service at all. The two Australian Divisions, which made up the remainder of the Corps, were in a similar state. page 85 The Corps had therefore been assigned what was then a very quiet sector of the line—part of the Armentières Salient.

The New Zealand Division's portion of the Corps sector extended from a point about two miles due west of Pérenchies and a mile south-west of Chappelle d'Armentières (south of the Armentières-Lille Railway) to the left (or northern) flank of the Corps sector on the River Lys, midway between Houplines and Frélinghien. The 2nd Australian Division was in the portion of the line to the immediate right of the New Zealand Division.

The distance between the enemy's lines and ours here averaged about three hundred yards; but at certain points, notably on the Division's left flank, and at Pont Ballot in its centre. No-Man's-Land was only a hundred yards in width. The whole of the front line was not to be garrisoned: various lengths of line, known as "localities," were held strongly, and between these were "gaps" (of an average length of two hundred yards with a straight parapet and dummy parados) which were not held at all; but were patrolled at frequent intervals by the garrisons of the adjoining localities, and were enfiladed by the flank trenches of these localities.

The support trenches (usually about two hundred yards behind the front line) were also held on the locality system, though the trenches were capable of being manned and defended throughout their whole length, and were not broken by gaps, as was the front line. Behind the line at various points, determined by the natural lie of the ground, were constructed "strong-points"—small earth-work forts which were capable of holding out for some hours, in the event of the enemy overwhelming the garrisons of the localities; or which could, by their commanding positions, render our trenches untenable, in the event of the enemy gaining a footing there. About six hundred yards behind the support trenches there was another continuous system of defences, known as the "subsidiary line."

The style of the trenches themselves was quite different from that to which the "old hands" had been used on Gallip-oli. From the high ground seven miles north of Armentières, on which stood the village of Messines, and for many miles away to the south, stretched the Plain of Flanders; which was page 86 not only practically level, but owing to the heaviness of the annual rainfall, was water-logged nearly all the year round. Experience had proved that trenches below ground-level were uninhabitable: the only thing to be done was to build a continuous breast-work for a parapet, and a smaller breast-work behind for a parados.

Such was the nature of the trenches which the New Zealand Division was ordered to hold: they had been built many months before, and had not been kept in good repair. In addition to the great amount of labour required to put the trenches in proper condition, there was much work to be done before the wire in front of the trenches could be considered an effective protection against a sudden rush by the enemy.

The New Zealand Division received orders early in May that it was to relieve the 17th Division in the line, and that the relief was to be completed by the 20th. The 2nd Brigade received orders on the 10th to relieve the 52nd Brigade, on the nights of May 14th/15th and 15th/16th, on a four-battalion frontage in the left (or northern) portion of the New Zealand Division's sector.

After moving to Neuf Berquin the 2nd Canterbury Battalion had continued its training, while all its officers in turn visited the sector which the battalion was to take over. Marching out of the billets at Neuf Berquin on May 14th, the battalion reached Armentières in the evening, and went into billets there. The following day, advance parties went up to the trenches and took over the signal stations and observation posts by daylight. In the evening the rest of the battalion relieved the 9th Battalion Northumberland Fusiliers in the sections of the front line trench numbered 79 to 81 inclusive, and called the "Pont Ballot" Sector, after the farm-house of that name in the German lines opposite.^* This was the right centre portion of the 2nd Brigade's sector, which was held by all four battalions, each distributed over the front, support, and subsidiary lines in its own sub-sector.

Since leaving Morbecque on May 9th, the 1st Battalion had been billeted at Estaires, where it was inspected on the 12th by General Birdwood, the General Officer Commanding the I Anzac Army Corps; and it left there to march to Armentières on the 13th. It was not until the night of May 20th/21st that the battalion went into the line, when it relieved the 1st Wellington Battalion on the extreme right of the New Zealand Division's sector, in trenches numbered 67 to 72 inclusive. This sub-sector included a salient known as "the Mushroom," of evil memory, of which more will be heard later.

It is not proposed to follow in detail the various reliefs of the two battalions: the records are incomplete, but so far as they exist, they will be found tabulated in Appendix "B."

At first the general system of holding the line was that a battalion should hold a short sector in depth, i.e., a small portion of the front line, with the corresponding length of support and subsidiary lines. The 1st Brigade held two such seetors; and the two battalions which were not in the line formed a brigade reserve, and lived in billets at Armentières, spending their time either in training or in improving the defences. On the left of the 1st Brigade, the 2nd Brigade held four sectors, all four battalions therefore being in the line, and its only reserve being the New Zealand Pioneer Battalion, which was attached to the brigade for this purpose. The 3rd Brigade, in Divisional reserve, lived in Armentières and in Houplines, the eastern suburb of the town. Reliefs took place every eight days, when the line battalions of the 1st Brigade exchanged places with the reserves in Armentières; and the 2nd and 3rd Brigades also changed places.

This system was found unsatisfactory, not only on account of the obvious unsoundness of having either the 2nd or the 3rd Brigade spread over a long line, with only slight reserves of its own to meet any sudden emergency, but also because the whole of the 1st Brigade was never out of the front line at one time. Accordingly, the method of holding the line was changed on June 5th, when the Divisional sector was divided into two equal portions, each of which was held by one brigade; while the remaining brigade was held in Divisional reserve.

The frontage of each brigade in the line was again divided into two battalion frontages, extending as far back as, and including, the support line. Another battalion held the whole of the subsidiary line on the brigade frontage, while the fourth battalion was kept as a brigade reserve, and was housed in billets behind the subsidiary line. In consequence, a new inter-brigade boundary was established immediately north of L 'Epin-ette; and the 2nd Canterbury Battalion took over the right half of the 2nd Brigade's sector.

The following memorandum, issued by Divisional Headquarters at the time of the change, shows the distribution of the battalions of the two brigades holding the front line;—

Distribution of Battalions in Front System.

It has been stated that the trenches and wire were in a very had condition: there was unlimited work in sight, and there began for the Division a period of unremitting toil, such as later reinforcements never dreamed of, except those who helped in the preparations for the Battle of Messines. When the Division left this sector at the beginning of August, 1916, it was able to hand over to the relieving Division trenches in perfect repair, amply protected by belts of wire of its own erecting. Nor must it be supposed that the brigade in reserve had a much easier time than those in the line; for large working parties were provided every day and night, many as far forward as the front line. Preparations for gas attacks on the enemy involved the carrying of many hundreds of heavy cylinders to the front line, and this work fell to the share of the reserve troops.

So much for its defensive work. The more important side of a soldier's life—his offensive against the enemy—received even more thought, which bore fruit in ceaseless activity. The Division found the sector a quiet one when it arrived, but it page 90 was not in the character of the General Officer Commanding the Division to continue a "live and let live" policy. The name of No-Man's-Land must be made true, so far at any rate as the enemy was concerned; and to ensure this, a system of vigorous patrolling was instituted at once. Every means possible was used to annoy the enemy—shell-fire, trench mortar bombs, rifle grenades, sniping, and machine-gun, Lewis gun, and rifle fire—and he naturally retaliated; with the result that the sector rapidly became anything but a peaceful one. The British attack at the Somme on July 1st, and the unsuccessful attack of the 5th Australian Division on July 19th, on the right of the New Zealand Division, naturally increased the enemy's activity on the latter's front.

The daily routine of trench warfare, familiar as it is to those who have experienced it, may be described here for the benefit of other readers. The routine varied in detail, according to the personal views of battalion commanders; but in essentials it was much the same throughout the Division.

The two periods of the day most favoured for making attacks, and accordingly the two periods when the garrison of a trench needs to be best prepared to meet the enemy without delay, are from just before dawn till broad daylight, and from dusk till after dark. It is therefore laid down in all modern armies that, when in the face of the enemy, all troops must stand to their arms at these two periods at least during the day. In the British Army, these periods (called "stand-to") are fixed at an hour each, but are subject to extension at the discretion of the responsible officer on the spot, in the event of a fog or other circumstances making a longer "stand-to" advisable. The actual times when the morning and evening "stand-to" began and ended, varied, of course, according to the season of the year. On these periods, when every officer and man in the trench had to be in his fighting station, with all his equipment on and his bayonet fixed, the daily routine was built up.

The main considerations on which the routine depended were; first, the necessity of rendering it impossible that the enemy should at any time of day or night surprise the garrison; second, the necessity of ensuring that every man should get enough sleep to enable him to carry on his duties properly; page 91 third, the proper feeding of the garrison; fourth, the amount of work required to be done to keep the trenches in fightable order.

In describing the routine of a day of trench-warfare, it is better to begin with the evening, as in this form of warfare the night is the most important period of the twenty-four hours. During this period precautions against surprise obviously need to be much more elaborate than during daylight. They consist firstly of the evening "stand-to," when everyone of the garrison is in his battle position. In the middle of 1916, the front-line garrisons were much larger than they were later on, and the trenches were unduly crowded. It was often necessary to put the whole of a section into one bay of the trench; and it was the duty of the section commander, being responsible for the protection of the men under him, to see that sentries were posted, and personally to change the sentries at the end of their watches. Later on, when experience had shown that trenches were best defended by a deep series of comparatively weak lines, rather than by two or three stronger ones, a section usually occupied two bays: but the same precautions against surprise were taken in each occupied bay.

It was laid down that there must always be two sentries on duty at night, but only one of these need be looking out over the parapet; the other, however, had to be so near the man on the look-out, that the latter could attract his attention by touching him, and without speaking to him. The sentries were relieved, as a rule, every two hours; but in very cold weather reliefs took place every hour. The remaining men of the garrison of the bay were allowed to sleep if they could, but, of course, were forbidden to remove either boots or equipment. So the night passed, till the whole garrison turned out at morning "stand-to." At the end of "stand-to," if there was a ration of rum on issue it was distributed to the men; and breakfast was, as a rule, ready immediately after "stand-down."

Further precautions against surprise by night were effected by means of listening posts and patrols. A listening post usually was garrisoned by six men under a non-commissioned officer; their post was in No-Man's-Land, and was any distance from ten yards upwards in front of our trenches, according page 92 to the width of No-Man's-Land and the activity of the enemy. The post was generally a shell-hole, and was sometimes protected by a little wire; but as concealment was essential, it could not be made an elaborate work. Two men only occupied it at a time, one of whom was on the look-out, while the other sat down. The rest of the garrison of the post were accommodated in a bay of the front line trench, and usually kept in touch with their sentries in No-Man's-Land by means of a piece of cord, over which signals were passed by the sentry making an agreed number of pulls. The non-commissioned officer in charge visited the post and changed the sentries—hourly as a rule. This work was particularly trying in cold and wet weather, as practically no movement on the part of the sentry was possible.

Patrols were employed both in our own trenches and in No-Man's-Land. In our own trenches, besides the officers and non-commissioned officers on trench duty, who visited the posts to see that all the sentries were alert, patrols were also sent out at frequent intervals during the night, from the posts on the flank of each company in the front line to the flank posts of the two adjoining companies. Also, where the line was held on the "locality" system, and there was a "gap" between two companies, one of the two was definitely made responsible for the defence of the gap; and its duty was to supply a patrol to report to its neighbour's post on the other side of the gap. at frequent intervals.

No-Man's-Land patrols were also of two kinds, fighting patrols and patrols sent out merely to watch the enemy. At Armentières, in 1916, where our object was to drive the enemy out of No-Man's-Land, fighting patrols were common; their duty was to seek out and attack the enemy, and their strength of numbers was large accordingly. As a rule, however, a patrol was sent out with instructions to obtain some definite piece of information about the enemy's doings, and in order that it should be as inconspicuous as possible, its numbers rarely exceeded three men. Such a patrol had instructions that it was not to get itself involved in a fight, unless it was absolutely impossible to do anything else. Three patrols of this kind were usually sent out on each company's front every night, and the page 93 period between the evening "stand-down" and the morning "stand-to" was divided equally between them; so that at no period of the night was No-Man's-Land free of patrols. The officer on trench duty was responsible for seeing that the patrols went out at the proper times and by the proper routes, and for keeping the sentries in the trench informed of their probable movements.

The meal-times of the garrison were usually—breakfast immediately after morning "stand-down," lunch at twelve, and tea just before "stand-to." In the long cold winter nights (after the Division had returned from the Somme) soup or cocoa was issued at midnight.

It was difficult to prevent the hours of sleep and work clashing, except in the summer months, when the day was a very long one. If the work was within a few yards of the front line, it was possible to set the garrison of each bay a definite task to be done during the night. The men were only too glad of some occupation to pass the time; for sleep was not easy, and was broken at every relief. If the work had to be done in the day-time, and each man had to do six hours' good work every day (the standard task when the Division first went to Armentières), the conflicting claims of work and sleep were hard to adjust, and the day's routine varied from time to time.

Besides the daily task of work allotted to each battalion, there were other things which had to be done, but which did not count as "work" from the military point of view. For instance, rations had to be carried by the trench garrison from the point beyond which the horse transport was unable to come. The distance over which rations had to be man-handled usually depended upon the quietness of the sector; but in some sectors the saps leading from the nearest tracks available for limbers were very long ones, and the labour of bringing up rations was increased accordingly.

It was the rule that a man must shave every morning; and though a few men actually found this a hardship, and a rather larger number affected to do so, there can be no doubt that the rule was a good one. For two reasons: firstly, an unshaven man loses his self-respect, and with it no small portion of his page 94 morale; and secondly, when a man has shaved he almost invariably has the wash which, but for the act of shaving, might easily slip his memory. In the wet season, another detail of the day's routine was rubbing one's feet with whale-oil, and changing one's socks. These were highly important precautions against trench feet, and were the subject of frequent enquiry by the Commanding Officers of battalions when on their daily rounds of inspection—another feature of the day's routine.

Among the distinctive features of trench warfare was the minor operation known as a "raid," which was carried out by a party of selected officers and other ranks. A party of this kind, after special training, would make an attack on a short portion of the enemy trenches opposite its own; and after killing or capturing the garrison, return to its own trenches. The intelligence branch of General Headquarters did all it could to encourage troops in all parts of the line to initiate minor operations of this nature. It is obvious that, if we were in the habit of taking prisoners frequently in every part of the line, we could keep ourselves informed of the movement of enemy divisions from one part of the line to another, or of the arrival of troops from another front. To move large bodies of troops is a troublesome operation, and it is seldom done without a pretty strong reason; so it follows that our intelligence service could generally rely on deductions which they had made from the movement of the enemy's divisions.

The main object of a raid was, therefore, to get prisoners: raiding parties usually tried to capture documents as well; but documents might be sometimes misleading, while it was very seldom indeed that some detail of uniform, badges, pay-book, identity-disc or even marks on boots or clothing did not show with certainty the unit to which a prisoner belonged. Also, prisoners frequently were willing to talk: if they would not talk to our intelligence officers, they sometimes unthinkingly gave away valuable information to our "agents," disguised as other prisoners.

The first raid in which troops of the New Zealand Division were engaged took place on the night of June 16th/17th, when a party consisting of four officers and eighty-three other ranks, drawn from all four battalions of the 2nd Brigade, raided a page 95 new enemy work known as "the Breakwater." This work (as its name suggests) was a sap which curved out from the enemy's front trenches, west of "Les 4 Hallots" Farm, which lay just inside the enemy's lines. The party from the 2nd Canterbury Battalion consisted of 2nd Lieutenant H. G. de F. Garland and sixteen other ranks. As all subsequent raids differed from this one only in matters of detail (though, as a rule, a raiding party was not so large as on this occasion), it may be described rather more fully than its actual importance warrants.

The enemy's new work was about two hundred and fifty yards long; and this work, together with thirty-five yards of his original trench on each side of its junction with the Breakwater, formed the objective of the raid. The main object of the raid was to find out why the enemy was building this work—whether merely to cut off a re-entrant in his line, or to provide a position of assembly from which he might attack our line—but another important object was to identify the enemy unit opposed to the Division, preferably by the capture of prisoners. The raid also had the minor aims of causing loss to the enemy and lowering his morale. The party had carefully rehearsed the operation over a model of No-Man's-Land and the enemy's trenches.

The force was divided into several smaller parties, each of which had its own definite task allotted to it. Thus, some parties were to cut gaps in our own wire, other parties were to act as scouts, to protect the movement of the rest across No-Man's-Land; and there were larger parties whose task was to assault the new work, and after capturing it, to protect other parties engaged in searching for identifications. There was no intention to hold the enemy position after it had been captured, or even to remain there for more than a few minutes: there was, therefore, no need for the members of the party to be encumbered by heavy clothing, equipment, or ammunition. All that anyone carried was a rifle, with bayonet fixed, and with an electric torch firmly bound to the stock, so as to illuminate, if need be, any object at which the bayonet was pointed. A "knobkerry" (or club), hung from the wrist by a lanyard, was designed to serve as an emergency weapon, in the event of a man losing his rifle. Faces and hands were blackened, all page 96 marks of identification were removed from clothing, and no papers of any kind were carried.

The night of the 16th was a calm one, and the sky was clear. The wire-cutting parties left the trenches at 11 p.m., and had cut the necessary gaps in our own wire before 11.15 p.m., when our artillery began to bombard the objective and the enemy's wire in front of it. Under cover of the bombardment, which lasted twenty minutes, the remaining parties passed through our wire, and formed up in No-Man's-Land, as near to the enemy's trenches as our artillery fire would allow. At 11.35 p.m., the artillery lifted from the enemy's wire and the Breakwater, and came down on the main trench behind; while certain guns still continued to fire on the flanks of the objective, with the result that the latter was isolated in a semi-circle of fire—known in technical language as a "box-barrage."

The scouting parties then dashed forward to find whether the enemy's wire had been cut by the artillery fire; and on their reporting that the way was clear, the main parties assaulted the Breakwater in the centre, and began to work their way along the sap towards either end. The work was found to be unfinished, and evidently only lightly manned. There were only two of the enemy found there, and these were bayoneted by the left assaulting party; but identifications were obtained from their bodies. The sappers attached to this party demolished with gun-cotton a listening post, which was found at the forward end of the sap. The right assaulting party, working down the sap, found it effectually blocked at a point fifteen yards short of its junction with the main trench. As the time fixed for rallying had almost arrived, this party did not attempt to enter the main trench by going across country.

The raid was a successful one, and set a standard for the many raids which followed it. Its good results were due, in great measure, to the artillery, which not only thoroughly cut the enemy's wire and barraged the raiding party against infantry attacks, but also did such good counter-battery work that the enemy's retaliation was very weak. Indeed, nearly all the losses of the party—one officer and one other rank killed and three officers and five other ranks wounded—were caused by the attacking troops getting too close to our own barrage.

Lieut.-Colonel A. E. Loach

Lieut.-Colonel D. Macbean Stewart.

The enemy was also fully aware of the possibilities of raiding, and on July 3rd unsuccessfully raided the 1st Auckland Battalion. The 1st Canterbury Battalion's turn came on the 8th: it had come into the line on the 3rd, and in the interval the enemy had been busy "registering" our trenches—i.e., ascertaining, by observing the fire of single guns, the exact elevation and direction necessary to hit the target, and recording the information for further use. This "registration" had borne fruit in heavy bombardments of our positions, on the nights of the 4th/5th and 6th/7th, by artillery and trench-mortars (minnen-werfer)^* of all calibres.

At 9.15 p.m. on July 8th, the enemy opened a still heavier bombardment on all our positions, but concentrated particularly on the centre of the 1st Canterbury Battalion's front line trenches. On the right centre (known as No. 2 Locality, with the strong point called "the Mushroom" in advance of the front line) was part of the 1st Company, the remainder of which was in No. 1 Locality (on the right flank of the battalion) and in the support trenches of the two localities. Part of the 12th Company was in No. 3 Locality, in the left centre.

In the light of subsequent experience, all the British front line trenches were at this time far too strongly garrisoned, and it was impossible for any shell to land in a trench without causing several casualties. The Mushroom had a garrison of one officer and forty other ranks, of which the bombardment which began at 9.15 p.m. killed the officer, his platoon sergeant, and five men, besides severely wounding several others.

When the bombardment lifted off the Mushroom, about fifty of the enemy attacked the strong-point. The survivors of the garrison, under Sergeant S. G. Brister, repelled this attack, but were immediately attacked from both flanks by enemy bombing parties. Fighting desperately, the garrison on the right was driven back up the communication-trench leading back to No. 2 Locality. This party was commanded by Sergeant Brister, who, though wounded, refused to surrender, and established a block in the communication trench, which he held till our counter-attack had been delivered. On the left, the remainder of the garrison was forced to fall back across country.

Meanwhile, the garrison of No. 2 Locality had suffered even more severely from the bombardment than the platoon in the Mushroom, and almost all had been either killed or wounded. In No. 3 Locality, things were not so desperate, but the officer commanding the garrison had been killed. Here, Lieutenant E. H. T. Kibblewhite, of the 1st Machine-Gun Company, took charge of the position, and organized a counter-attack which he led against the Mushroom. He met with no opposition; on reaching the strong-point he found that the enemy had removed his dead and wounded, and had abandoned the trenches. Apparently our dead had not been searched; no wounded had been taken prisoner, nor was there anything missing from the dugouts.

On the night of August 7th/8th the 1st Battalion completed its last tour of duty in the front line in this sector, and the 2nd Battalion was relieved the following night by the 2nd Auckland Battalion. Both Canterbury Battalions moved to the subsidiary line.

The 1st and 2nd Canterbury Battalions were relieved by the 4th and 1st/5th Battalions respectively of the Gordon Highlanders on August 15th, and moved to billets in Armentières. Marching out with their respective brigades on the 16th, the two battalions entrained at Steenwerck and left train at Ebling-hem, midway between Hazebrouck and St. Omer. The 1st Battalion then marched to billets at Wardrecques (three miles away), while the 2nd Battalion marched to Blaringhem, about the same distance from the railway, and four miles south-east of the 1st Battalion's billets. Both battalions remained at these villages for three days, carrying on training, and while there received a draft of reinforcements.

Leaving its billets on August 21st, the Division entrained at Arques, two miles south-east of St. Omer. The 1st Canterbury Battalion left the train at Abbeville, near the mouth of the Somme, on the 21st, and arrived at Merelessart on the evening of the same day, after a march of twelve and a half miles—a long march for men who had so recently left the trenches. The 2nd Brigade detrained the same day at Pont Remy (five miles southeast of Abbeville), and the 2nd Canterbury Battalion marched to Allery, seven miles due south of Pont Remy.

Total for Regiment—4 officers and 98 other ranks killed; 10 officers and 356 other ranks wounded.

The Division now temporarily ceased to belong to the I Anzac Army Corps, and came under the orders of the General Officer Commanding the X Corps of the First Army. The men needed much training to bring them into good physical condition again after their long stay in the trenches; and this kept everyone very busy during the following ten days, in spite of the wet weather which prevailed at the end of the month.

Owing to the progress which had already been made in the battle, the new methods of attack which had been introduced in its first stages had even now been considerably modified. While the ground gained by the more recent attacks was considerable, it could not compare in extent with that wrested from the enemy in the early stages of the battle. Conditions had settled down into what had been given the name of "semi-trench-warfare."

On a front of twenty miles the enemy had lost his elaborate first and second defensive system, on which he had worked for nearly two years and which he considered impregnable.^* The further defences on which he had now fallen back were not nearly so formidable: in certain places where our attacks had met with the greatest success, his front line consisted merely of a series of lines of shell holes, hastily joined together to form trenches. Under cover of these the enemy was working hard on defensive systems further to his rear; but the time at his disposal alone ensured the impossibility of his building any defences of which the strength could in any way compare with that of his old front line.

Up to now, the chief novelty of the new warfare was the use of a "creeping" artillery barrage. The idea of a stationary barrage was borrowed from the French; but the "creeping" barrage was a British idea, and in the first battle of the Somme had been elaborated to an extent which had not hitherto been thought practicable. In former British offensives, the artillery had been used to bombard the enemy trenches before the attack; but it had not had at its disposal the unlimited ammunition upon which it could now rely. On many occasions, too, the enemy had evaded our bombardment by sheltering in shell-holes in No-Man's-Land, and had then surprised our infantry as it moved to the assault.

The new method of using artillery was to combine a stationary barrage of heavy artillery and 4-5-inch field-howitzers with a "creeping" barrage of 18-pounder field-guns. The stationary barrage rested on the enemy's trenches. till the attacking troops had advanced as close to it as safety allowed, when it was lifted back to the next trench towards the rear. The "creeping" barrage (as its name implies) moved forward in front of the infantry, beginning just in front of their assembly positions, so that it thoroughly swept all the ground over which they had to advance. The rate at which the infantry could be expected to advance governed the rate of the "creep" of the barrage.

It should be explained, however, that the word "creeping" is not a very good one to describe the action of the artillery: unless a gun is practically new, it is almost impossible to lengthen its range by less than a hundred yards at a time. When a "creeping" barrage was in progress, each gun therefore fired on a given line for a period of three or four minutes, or even a longer time (according to the rate at which the infantry were expected to move), and at the end of that period the range was lengthened a hundred yards. The infantry then moved forward at a walking pace, as far as they could go without coming under our own shells, and then halted and knelt down; and so the process was repeated till the infantry had captured the objective which they had been ordered to take.

It must be borne in mind that the area swept by the "creeping" barrage was a deep one, so that the guns engaged were not page 102 by any means all firing on a line directly in advance of the infantry, but that the majority were sweeping a broad belt of country well in front of the assaulting troops. It was therefore not unduly complicating the task of the gunners to arrange lifts of fifty yards only, by making the lines on which half the barrage guns were to fire fifty yards in advance of what would normally have been their targets. By lifting half the guns at a time, at the same time halving the rate of lift, the "creep" of the barrage could be reduced to fifty yards at a time; and, of course, the "creep" could be further reduced by more elaborate arrangement of targets and times of lifts. Where the enemy positions were very strong, and concrete shelters protected his troops, the "creep" of the barrage was reduced in this way. As a general rule, however, fifty yards lifts were found to be the most suitable.

The "limited objective" was a second new feature of the Somme fighting. The range of field guns is limited, and also barrage firing is a great strain on the gunners: the infantry advance was therefore made by definite stages, from one line to another, strictly according to time-table. As a result, even if the infantry saw there was no enemy in front of them, they were forbidden to advance beyond the spot on which the time-table showed they should be. The necessity for this is clear; for where a great number of scattered batteries of artillery are working in unison, a considerable time must elapse before information from the firing line can reach them all, in its final shape of orders from the various divisions under which they are working. Infantry units were therefore forbidden to advance beyond the objectives assigned to them, since they would thereby come under the fire of our own guns when the attack was continued.

The training of the New Zealand Division was mainly devoted to the study of the new methods of fighting; and in particular advancing under a barrage was continually practised, the limit of the ground on which our shells were imagined to be falling being represented by lines of men waving flags, and running forward by stages of a hundred yards every three or four minutes. No opportunity was lost by the senior and junior officers of getting in touch with officers who had been engaged page 103 in the fighting, and the valuable information so gained was circulated as widely as possible.

Standing orders during an offensive operation now issued by General Headquarters laid down that no more than twenty officers and six hundred and eighty other ranks of a battalion were to go into action. The remainder of the battalion was called "the B team": it included the second in command of the battalion, either the commander or second in command of each company, all other officers in excess of twenty strong, and a certain number of non-commissioned officers and of the specialists—experienced Lewis gunners, signallers, bombers, and so on. "When the New Zealand Division went into the battle, the "B teams" from all the battalions were retained in a special camp near Fricourt, and were not allowed to be recalled to their units, except by permission of the brigadier under whose command their particular unit came.

The Division left its training area on September 2nd, when the 1st Battalion left Merelessart, and moving with the remainder of the 1st Brigade, marched to Airaines, five miles to the east. After spending the night in billets there, it marched to La Chaussée (on the main Abbeville-Amiens road, seven miles west of Amiens), where it remained in billets until the 7th. On that day the battalion marched to Coisy (fifteen miles north of Amiens), spent the night there, and on the 8th, after marching east fifteen miles to Dernancourt (south of Albert) bivouacked at "Area 'A'" near there, from that afternoon till the morning of September 10th. It then marched three and a half miles north-east to another bivouac area near Fricourt, where it remained till it was ordered into the line, in the meantime supplying working parties and carrying parties for the front line and also digging assembly trenches for the 3rd (Rifle) Brigade's attack of September 15th.

The 2nd Brigade also left its training area on September 2nd, and on that day the 2nd Canterbury Battalion marched out from its billets at Allery, passed through Airaines, and was billeted for the night in Le Mesge, two and a half miles south of Hangest-sur-Somme. Moving on next morning, the battalion the same day reached Picquigny (on the opposite bank of the Somme from La Chaussée), and remained in billets there for the following three page 104 days. The time was spent in training, and a brigade attack was practised. As the method adopted was the same as was afterwards used against the enemy, a description may be found interesting. The final practice took place in the presence of the General Officer Commanding the Division, who expressed his approval of the way in which the operation was carried out.

The scheme of the attack was that the brigade, being part of the garrison of a system of trenches, with a No-Man's-Land three hundred yards in width between it and the enemy's trench system, had received orders to attack and hold seven hundred yards' frontage of the enemy's trench system, which consisted of firing-line, support, and reserve trenches.

The attack was divided into three phases, each of which ended at a limited objective—the first objective being the enemy's firing-trench, the second his supports, and the third (and final) his subsidiary, or reserve, trenches. Every part of the advance was to be supported by stationary and creeping barrages, which were represented by officers with flags. The creeping barrage moved forward at the rate of fifty yards per minute—a rate quite suitable for practice on good ground, but too rapid for an advance over country which has been badly shelled. Two battalions only, the 2nd Otago and 2nd "Wellington, were to take part in the assault and to take all the objectives, the 2nd Auckland Battalion being in brigade reserve, while the 2nd Canterbury Battalion was detailed to carry out special duties and to await orders.

The orders for the artillery ran thus:—"The bombardment of heavy artillery and howitzers, and the double barrage of shrapnel will be as under:—

0.00^* Bombardment lifts from enemy's front line to 2nd objective.

First shrapnel barrage continues on enemy's front line.

Second shrapnel barrage opens 50 yards in front of our 1st line.

Meanwhile, the remaining companies of the assaulting battalions had moved out of our support trenches, and were extended across No-Man's Land, each in two waves; and on the capture of the first objective the third company of each battalion at once crossed the captured trench without entering it, and "hugged" the barrage preparatory to attacking the enemy's support line. It followed up the barrage in the same way as the original leading waves had done; and on its capturing the support trenches, the last remaining company, which had been close on its heels, crossed these trenches, and following the barrage closely, captured the enemy's reserve trenches. This system of attack was called "leap-frogging."

On referring to the barrage time-table, the reader will see that at the time fixed for the capture of the third objective (zero plus 17), the infantry came within fifty yards of both the first and second shrapnel barrages; and that at zero plus 18 the barrages moved on another fifty yards. Here the barrages remained, protecting the leading infantry while it dug in; and when ample time had been given for this to be done, the barrages gradually died away. But. in accordance with the rule that every body of troops, no matter how large or how small, is responsible for its own protection against surprise, the leading company of each battalion, directly it reached the final objective, pushed forward Lewis guns to protect the troops engaged in digging, and sent out patrols to traverse the area between the captured trench and the barrage and to watch for enemy counterattacks.

A feature of the practice was co-operation with aeroplanes. which were supplied by the 3rd Squadron of the Royal Flying Corps. Co-operation of this nature was another successful experiment of the Somme Battle: low-flying aeroplanes, recognizable by a distinctive mark, were ordered to fly over the objectives at defined times; and it was then the duty of the front-line troops to light flares to mark the limit of their advance. In this way, reports of the progress of the attack could be sent to the page 107 commanders of brigades, divisions, and higher formations much earlier and with more certainty of arrival than by any system of runners or signallers.

The 2nd Brigade resumed its march on September 7th: the 2nd Canterbury Battalion spent that night in billets at Cardonnette, two miles south-east of Coisy. Next day the battalion marched viâ Querrieu and along the Amiens-Albert road to a camp south of the road and west of Dernancourt. At this camp the battalion remained till early in the afternoon of the 10th, and then marched through Dernancourt, Méaulte, and Fricourt, and bivouacked in Fricourt Wood.