Appendix 40: "The Coming of the Creeping Barrage," 1931
By Major A. F. Becke, late R.F.A
Source: Journal of the Royal Artillery 58:1

IT has often been said that the creeping barrage was first used in 1916 by the XV Corps (Fourth Army) during the attack on the 15th September, sometimes called the Battle of Flers- Courcelette. Lately the writer has had to read through the mass of papers: war diaries, orders, instructions, reports, etc., which narrate and deal with the opening operations in the Somme Battle, between the 1st and 14th July, and in the files examined many documents have been found that throw further information on this step in the evolution of artillery fire-tactics. Before the references are lost it seems wise to place on record what has been discovered. Many officers who were concerned with this innovation have died, and with the passage of years it has already become difficult to say definitely when this change first came into use. Consequently for the assistance of the regimental historian of the future the writer has given here such information on the subject as he has been able to collect from his recent study of the con-temporary documents, supplemented by statements from various officers.

There is no doubt that the idea of affording protection to advancing infantry is very old. To mention only four cases from past history: in the first case, on the 25th August 1346, the day before the Battle of Crecy, Edward the IIIrd's archers loosed off a protective barrage of arrows to cover the successful crossing of the Somme ford at Blanchetache (near Saigneville). Another well-known case occurred at the Battle of Friedland, 14th June 1807, when Senarmont prepared the final advance of Dupont's Division of Victor's Corps by a sustained barrage from 30 guns, moving his guns forward twice when it became necessary to lift, and from his last position, when preparing the Russian infantry for Dupont's onslaught, the guns fired case shot at a range of 150 yards. The third instance occurred on the 31st August 1813, in the successful assault of San Sebastian, when, after a temporary failure, the stormers held on to the lower part of the breach, whilst the concentrated fire of 50 heavy British siege guns was turned on the high curtain. As one eye witness wrote at 7 a.m. on the 1st September: "So exact was the fire of our batteries that our shot passing over the heads of our men, then lying on the demi-bastion, carried away the heads of the enemy who fired from the ruins of the rampart of the curtain...." Consequently when a number of powder barrels and live shells suddenly blew up just behind the curtain, where they had been accumulated to assist the French in the defence, the British infantry, who had been lying out under the barrage, were able to deliver an immediate assault under cover of the smoke and general confusion and before any defensive measures could be taken, and the town was stormed. This exemplifies the very closest co-operation between assaulting British infantry and its covering artillery. The fourth instance occurred some years before the Great War, and it was an experiment that was carried out at the School of Musketry, Hythe, in which machine guns were used to give protection to advancing infantry by firing over their heads. Unfortunately, as it turned out, the targets representing the advancing lines were rather badly peppered, and not unnaturally this form of co-operation was not received with much enthusiasm by the infantry officers who were present.

To return to the coming of the Creeping Barrage in the Great War. At the outset it may be stated that the documents do not altogether support the oft-repeated story of the introduction of this barrage; but, as might be expected, they show that the innovation was of gradual growth and that it was being tried in more corps than one and at the same time. It is well known that normally a change would spring from the fighting troops, gradually, and generally it would be adopted, then in due course G.H.Q. would hear of it and in a circular memorandum give it an official blessing. Thus it happened in the case of the creeping barrage.

In this study it will be necessary to approach the subject historically, that is to glance through some of the earlier battles, so as to ascertain if the handling of the artillery in them can throw any light on the solution of this particular problem, and in doing so we can note, for the assistance of the regimental historian, some of the milestones in the progress of artillery tactics.

The artillery tactics of the pre-war training manuals were undoubtedly modified by the siege warfare conditions that soon prevailed, and possibly, so far as the creeping barrage was concerned, "as it takes two to make a liaison, both artillery and infantry are jointly responsible for its introduction." To begin with, it is an acknowledged fact that the "lifting barrage" had been foreseen before the war and had been alluded to in the Field Artillery Training of January 1912, and in the Field Artillery Training of April 1914 , for in both these manuals it is suggested [that] artillery is covering an attack and "it is no longer possible to continue firing at the immediate objective of the infantry the range of the artillery should be increased so as to search the rear of the position, but no considerable amount of ammunition should be devoted to this object," thus the 1914 issue; and in the 1912 book the instruction ran: "As a rule, therefore, the range of the artillery should be increased as soon as the infantry starts to assault, so as to search the rear of the position, but no considerable amount of ammunition should be devoted to this object." Here is a definite lift, but it is still a far cry from the procedure enshrined in these two manuals to the creeping barrage of 1917.

In the Battle of Neuve Chapelle, 10th March 1915, the first definite attack mounted by the British after the stabilization of siege warfare conditions, several milestones can be noticed. In this battle the artillery time-table was introduced, and the three phases given in the time-table were called: "the preliminary bombardment," "the assault of the enemy's first line of trenches," and "the assault of the Village of Neuve Chapelle." Also we do find the word 'barrage' used in the true sense of the word—a barrier, or dam, to prevent the arrival of reinforcements—and it appears in the First Army's Report on the operations: ...Artillery Bombardment. "... Other 18-pdr. batteries and 13-pdr. batteries were given the task of creating a barrage of shrapnel fire, so as to prevent reinforcements reaching that area of the enemy's position selected for attack...." Further before the Battle of Neuve Chapelle the R.F.C. photographed the German trenches that were opposite to the front of attack, and these trenches were then carefully plotted onto large scale maps, which proved of the greatest assistance both to the artillery and to the infantry during the progress of the attack, an innovation that was tremend-ously developed during the War, playing a particularly important part in mapping forward areas, enemy trench systems, battery positions, etc.

In the assault of the German position at Neuve Chapelle the guns covering the British made two main lifts, from trench line to trench line, and then beyond the final objective all 13-pdr. and 18-pdr. batteries were ordered to "establish a belt of fire round the front of the position," in other words to put down a curtain, or protective barrage. Thus Neuve Chapelle saw considerable evolution in the tactics of artillery.

In the Battle of Festubert, 15th-27th May 1915, corps artillery command was for the first time exercised in the field and this led to the appointment of a G.O.C., R.A., Corps, who in March 1916 was converted into a B.G., R.A. In this same battle the opposing trenches on the front selected for attack at dawn were between 80 and 200 yards apart, and, to cover the assault, the final phase of the artillery preparation was a H.E. bombardment by 18-pdrs. Under cover of this fire the assaulting infantry climbed out of their trenches and lay down in No Man's Land to wait for the moment of assault. Despite the dim light and the proximity of the opposing lines, no casualties from our own fire were reported, and in places the initial assault was successful. Possibly this epi-sode was recalled when the introduction of something akin to a creeping barrage was being considered. Before the opening of the Battle of Loos (25th September-13th October 1915) there was the first prolonged, systematic bombardment, lasting 4 days, but neither the number of guns nor the quantity of ammunition available were sufficient to make it really formidable. At Loos, too, Corps artillery organization came into being; and "the hour of zero" appears in First Army Operation Order, No. 95, and is possibly the first use of this term. But for the purpose of this study the most interesting feature of the battle was that just before the actual assault an 18-pdr. barrage was fired by the 15th Divisional Artillery, searching forward 50 yards at a time, the guns firing a certain number of rounds at each move. But though it was not an attack barrage closely followed by the infantry, yet it may have occurred to somebody that if only the assaulting infantry had kept close up behind this barrage as it swept slowly forward, then the attacking troops would probably have been saved from incurring such heavy losses. Generally, however, on the 25th September the assault itself was supported by lifting barrages, though doubt-less where the number of trench lines was considerable and close together something approaching the nature of a creeping barrage was fortuitously employed, as the field artillery fire progressed in rapid lifts from trench line to trench line firing a certain number of rounds at each move. The lack of success that attended the earlier attacks supported in this manner was sometimes due to the fact that at that time the infantry had not been sufficiently trained to get close up to the barrage, and on other occasions the barrage advanced at too fast a pace for the infantry to keep pace with it. Occasionally the last-named was due to the infantry misjudging the pace at which it would advance and asking for the lifts to take place too quickly, and then attempting to advance on the objective some time after the barrage had lifted off it, and the guns were serenely firing perhaps a thousand yards beyond the line where their fire was needed. In addition the Battle of Loos sounded the death-knell of the H.A.R. Group system, by emphasizing that a Corps Commander must control all the guns in his own area.

The next big battle, or series of battles, The Somme, which raged between the 1st July and the 18th November 1916, defin-itely saw the introduction of the creeping barrage, which was the most practical way of dealing with the enemy's plan of placing riflemen and putting machine guns in shell-holes outside his knocked-about trenches. Whether this innovation originally came from the French Army and was adopted by the British Fourth Army is still rather a matter of doubt, and some doubt also exists about the date of the introduction of the first creeping barrage in the German Army. To take the last-named case first. In 1916 the only attack made by the Germans on the Western Front was at Verdun (21st February-31st August), and in the orders for the original attack on Verdun, given in Part 1 of the official mono-graph, Die Tragödie von Verdun (pp. 258, 259) there is no mention of any barrage, but only of a bombardment, and in the text the only word used is an "artillery curtain" (Vorhang), and this, presumably, protective barrage had already been employed by the British at Neuve Chapelle.

But the Germans do claim that just before the opening of the Battle of the Narotsch Lake (18th March-30th April) the "Fire roller" (Feuerwalze) was successfully used by the 86th Division for the recapture of a hill near the Lake. It is also claimed that in a later stage of the fighting at Verdun a "rolling barrage" was used in May in the attack of Thiaumont by the Jäger Brigade of the Alpine Corps. On the other hand, in Colonel Bruchmuller's book, Die Artillerie beim Angriff im Stellungskrieg, several artillery orders are given as specimens, and on the Eastern Front, even as late as November 1916, these orders only speak of lifts of 100 metres. If an examination of the German documents in the possession of the Reichsarchiv shows that the creeping barrage was used by the German artillery at Narotsch in March, and later at Thiaumont in May 1916, then this innovation certainly occurred at an earlier date in the German Artillery than it did in ours. But as in the case of the leather guns of the XVIIth century, which were evolved about the same time and quite independently in Sweden and in England, so the creeping barrage, as far as the evidence available at present goes, probably was evolved inde-pendently on both sides of No Man's Land in 1916.

As for the French, no trace has so far been found in our documents of any proof that we are indebted to our Allies for this particular form of barrage, nor is it at all certain that a creeping barrage (barrage mobile) was being used by the French in the Somme battle. The French official volume dealing with this period has not yet been published, and there is no mention of such a barrage in the brief official report on their July operations that has been furnished to the Historical Section.

We now come to the examination of the procedure followed by the British artillery at the opening of the Battle of the Somme. It is unnecessary and it would be wearisome to examine and quote from the Diaries of all the 18 British divisions of the 6 Corps which took part in the attack on the 1st July, and it will suffice for our purpose to draw attention to examples of those changes in artillery co-operation which show traces, all along the front of attack, of close relationship to the procedure of the creeping barrage. At this time in France there were with the B.E.F. 3,304 field guns and howitzers and 714 heavy guns and howitzers, as well as 114 A.-A. guns; and on the 30,000 yards of the Somme battlefront from Maricourt in the south to beyond Gommecourt in the north the British employed 1,434 field guns and howitzers and 527 heavy guns and howitzers, a total of 1,961 pieces, giving an average of 1 field piece to every 20.9 yards, and 1 heavy piece to every 56.9 yards, or 1 piece to every 15.3 yards. This is a very different proportion to the artillery available at Loos, although the front of attack on the 1st July was fully twice as wide as it was on the 25th September 1915; and if on the 28th June zero had not been postponed for forty-eight hours, and the preliminary bombardment had not been extended for this reason for two extra days there would probably have been ample ammunition for every task.

To begin with, at a Fourth Army conference, held on the 16th April, in dealing with Artillery co-operation in the offensive which was then under consideration and with the experience of Loos fresh in his mind, the Army Commander stated: "... The lifts of the artillery time-tables must conform to the advance of the infantry. The infantry must be given plenty of time. The guns must 'arrose' each objective just before the infantry assault it. Timing is a matter of most careful consideration...." In a later conference held on the 17th May the Army Commander emphasized that on former occasions the lifts had always been too quick and not enough time had been allowed for the artillery to cover the advance of the infantry. But neither allusion to the artillery lifts lays stress on the barrage creeping forward by short lifts, and neither does much more than give vent to the pious hope that the barrage shall not run away from the infantry. In addition, in May 1916, the Fourth Army issued a brochure called Tactical Notes, and under the heading "Co-operation of Artillery with Infantry" (paras. 66-69) it was clearly laid down that the ideal was for the artillery to keep its fire immediately in front of the infantry as the latter advanced, but it also pointed out that this ideal was very difficult to attain. It was also made clear that experience had shown that the only safe method of artillery support during an advance was a fixed time-table of lifts to which both arms must conform; and the Notes stated that it was better for the infantry to have to wait in front of an objective until the artillery lifted off it, than it was for the artillery fire to be too far ahead of the infantry. It was laid down that if the lift happened to be a long one it should be carried out by stages, as it would thus be able to protect the infantry advance more effectively. This last sentence is most important for it emphasizes what is the germ of the whole idea.

The Notes are interesting for they show the trend of thought of artillery officers of light and learning whilst the great attack was being mounted, and even though they do not definitely prescribe the use of a creeping barrage, yet they show that the idea of short lifts, just keeping ahead of the infantry advance, had already engaged the serious attention of those responsible for issuing tact-ical instructions to the troops about to attack in the coming battle. But the Notes do make quite clear that in May 1916 the creeping barrage had not so far received official sanction nor had the name come into use. The closest artillery co-operation was essential, for the German position that faced the Fourth Army on the 1st July was far more formidable than any which the B.E.F. had previously assaulted. The Front Line, or system, was very strong, and behind it there were intermediate positions, numerous strong points and redoubts, and eleven strongly defended villages. Behind again was a well-sited, strongly wired and constructed Second Line or position.

Now we can probe lower down and see what procedure was actually ordered in some of the Corps and Divisions engaged on the British Front on the 1st July. In the XIII Corps, on the British right, the Corps Plan of Operations, issued on the l5th June, laid down in para. 11 (d) that "... the advance of the infantry will be covered by a heavy barrage from all natures of guns and mortars. The heavy artillery barrage will lift direct from one line onto the next. The field artillery barrage will creep back by short lifts. Both will work strictly according to time-table. The lifts have been timed so as to allow the infantry plenty of time for the advance from one objective to the next, on the principle that it is preferable that the infantry should wait for the barrage to lift than that the latter should lift prematurely and thus allow the enemy to man their parapets. The infantry will follow as close behind the barrage as safety admits...." In the foregoing extract, written before the attack on the 1st July, we notice the mention of the word "creep," and so far as the writer has been able to ascertain this, and its employment in the VIII Corps scheme issued on the 11th June, is the first time the word has been used to describe the procedure for advancing a barrage. In the two front-line divisions of this corps, the 30th on the right issued a barrage map, showing 6 lifts, but in the Divisional Operation Order No. 19 of the 21st June the word creep does not occur, nor is there any allusion to short lifts; on the other hand it was laid down that "... the Divisional artillery, firing only H.E., was to remain on each barrage line for one minute longer than the heavy artillery...."

In Operation Order No. 1, issued on the 19th June by the B.G., R.A., of the 18th Division, it was laid down in paragraph 14 that "... The bombardment to support the assault will consist of a concentrated bombardment from -65 to zero hour, followed by a succession of lifts timed to conform with Corps Artillery lifts and the forward movement of attacking infantry...." So as to cover the last thousand yards of the advance, after the formidable trench system on the forward slope in front of Montauban had been captured, the Artillery Orders laid down that "the Shrapnel barrage would move in front of the infantry by increments of range until the final barrage was established" to protect the infantry who would then be overlooking Caterpillar Valley. An amend-ment issued on the 27th June made it clear that during this thousand yards advance:—"Batteries ... will increase their range by increments of 50 yards every 1_ minutes." Reading the two together it will be seen that, in this part of the advance, a real creeping barrage was intended.

There is in a diary a description of what occurred on the 1st July: "... The general plan for the bombardment was an intense bombardment of the front line system for 65 minutes, followed by successive lifts from trench to trench. Then concen-tration of fire on certain trench systems and strong points, and finally a general barrage along the divisional front preceding the advance of the infantry by increments of range of 50 yards every 1_ minutes. The above plan was considered most adapted to the general configuration of the ground. The bombardment was carried out in accordance with the pre-arranged time-table, and the general timing of lifts appeared satisfactory ... as a whole the lifts appear to have suited the time of the infantry assault...." In passing it may be mentioned that the attack of the XIII Corps was by far the most successful of all those that were delivered by the British on the 1st July.

Now we can consider the artillery arrangements that made for the lifting of the barrage in the next Corps, the XV. In the instructions issued by the B.G., R.A. of the Corps on the 14th June it was laid down in paragraph 4: "...When lifting, 18-pdrs. should search back by increasing their range, but Howitzers and Heavy Guns must lift directly onto their next objectives..." Also accompanying the Corps documents there is an elaborate barrage map showing the proposed 6 lifts from the German front line back to Caterpillar Valley, and the west side of Mametz Wood. The barrage maps for the attack on the 1st July appear to be the earliest barrage maps issued in the B.E.F., as only bombardment maps were issued in 1915 before the British attacks at Neuve Chapelle, Aubers Ridge, Festubert, and Loos.

In the 7th Division of the XV Corps in Operation Order No. 11, issued by the B.G., R.A., at 7 p.m. on the 18th June 1916, there occurs the following paragraph which states very clearly the procedure that this divisional artillery was to follow in the attack on the 1st July. It ran as follows :—"... 3 ... (b) During the advance of the infantry a barrage of artillery fire will be formed in front of the infantry according to the timings shown on the tracings issued to those concerned. The lines shown on the tracings indicate the nearest points on which guns will fire up to the hour indicated. At the times shown heavy guns will lift their fire direct to the next barrage line. The divisional artillery will move their fire progressively at the rate of 50 yards a minute. Should the infantry arrive at any point before the time fixed for the barrage to lift, they will wait under the best cover available and be prepared to assault directly the lift takes place...." In this same divi-sion, the 7th, the Instructions for the forthcoming operations, "No. 3, Action of artillery and artillery co-operation" were issued on the 14th June, and in para. 1 (b) the divisional artillery was ordered to "move their fire progressively at the rate of 50 yards a minute," and the same procedure is given among the papers of the 22nd and 91st Infantry Brigades. Also in an appreciation of the task allotted to the 91st Infantry Brigade on the 1st July (attacking to the east of Mametz) there is a confirmatory reference "... the barrages of the Field Artillery move ahead of the infantry as it advances...," and in the Instructions for the forthcoming operations, No. 2, issued on the 19th June by the 20th Infantry Brigade, attacking on the left of the 91st, there is the paragraph "... Method of Assault. 10. The assault will be carried out steadily behind the artillery barrage. At the hour named for the barrage to lift the leading line will be as close to the hostile position as possible, and on the barrage lifting will at once move forward steadily, keeping touch, and only halt and lie down when next compelled to do so by awaiting the lift of the artillery barrage." As a result of his experiences on the 1st July, the Brigadier of the 20th Brigade was convinced that it was essential the objective should be entered immediately the barrage lifted off it. The attack of the 7th Division on the 1st July was very successful.

In the left division of the Corps, the 21st, "Instructions for Offence, No. 3," were issued on the 15th June by the B.G., R.A., and in para. 12 it was directed that "... Batteries will search back to the next barrage, in order that the whole ground may be covered by fire immediately before our infantry advance over it..." and the same order is found repeated in the papers of the 62nd Infantry Brigade. Further light is shed on the question by the documents of the 64th Infantry Brigade, the left Brigade, which attacked well to the north of Fricourt. In Operation Order No. 6 of the left battalion of this brigade, 10/K.O.Y.L.I., for the attack on the 1st July, it is stated in para. 6, "Barrage ... IT IS TO BE REMEMBERED, [sic] however, that this barrage will be lifted gradually, sweeping the entire ground before it, and the Troops must move forward as it advances so as to take all advantages of its protection...." Further, in Omissions in Operation Order No. 25 which were issued on the 25th June by the left supporting battalion of the brigade, the 15/Durham Light Infantry, there is a very lucid and important explanation: "Barrages. The Barrages will not exactly lift from one point and be put onto another, they will gradually drift forwards, leaving certain lines at certain hours (which may be changed). The line of the barrage must be con-stantly watched by the infantry, whose front lines must keep close up to it...." The attack of this brigade too, was remarkably successful. Thus both in the artillery and the infantry of this division there was a clear conception of a field artillery barrage that would "drift forward" just in front of the advancing infantry, and in the extracts already quoted from the documents of the units of this Corps there is probably enough material to account for the belief that the creeping barrage did originate on the Somme and in the XV Corps.

But there is a different story to tell of the attack of the 50th Infantry Brigade, attached on the 1st July to the 21st Division. It was specially detailed for the capture of Fricourt Village so as to connect the two front line divisions of the XV Corps. Both in the Brigade Operation Order No. 76 of the 22nd June and in the battalion orders of the 7/Yorkshire Regiment (Green Howards), issued on the 25th June, it is stated that after a thirty-minutes preliminary bombardment the barrage would lift 500 yards back at zero, and fifteen minutes later the barrage would again lift 250 yards further back, and continue on this line until 1 hour and 45 minutes after zero. The attack delivered by the 50th Brigade on Fricourt Village was repulsed with heavy loss.

In the III Corps documents there is an elaborate map showing the eight major lifts of the artillery barrage that were to be put down on the III Corps front on the 1st July, but actually on this day the barrage on this Corps front seems to have been made to jump from one trench system to the next, and jumping in this way the barrage never touched the intermediate shell-holes which held riflemen and machine guns. Possibly the costly failures in front of La Boisselle and Ovillers may be partly traced to this procedure. But even so in the 34th Divisional Artillery O.O. (No. 10), issued on the 18th June 1916, it is clearly stated that "... 3 (g) ... Lifts are timed to commence at the same time as Heavy Artillery. But instead of lifting straight back onto the next line, Divisional Artillery will rake back gradually to the next line.... The speed at which this rake goes back is shown in Appendix V...." In Appendix V the rake is shown to be either one short lift of 50 yards (in one case), or 100 yards (in three cases), or 150 yards (in two cases), and then lifting direct to the next barrage line.

In the 34th Divisional O.O. (No. 16) issued on the 15th June, there is a further explanation :—"... Artillery. 3... The speed at which this [Field Artillery] rake goes back to the next line will be calculated so that the shrapnel barrage moves back faster than the infantry can advance...." If there was any doubt before whether this rake was a creeping barrage in embryonic form, this explanation shows that the real idea underlying the employment of a creeping barrage is absent from "the rake" as employed by the 34th Division on the 1st July.

In the X Corps, in the attack of the Leipzig Salient, an interest-ing expedient was tried by the brigadier (Br.-General J. B. Jardine) commanding the right brigade (97th) of the 32nd Division. Recalling his experience of the successful combination in the attack of the Japanese artillery and infantry, in the battles of 1904/5, General Jardine insisted that the front lines of the attacking infantry of his brigade should creep out into No Man's Land before zero and lie down within 30 or 40 yards of the barrage that was then on the German front line, so that they could rush the German trench directly the barrage lifted off it. Acting in this way General Jardine's right battalion succeeded in forcing its way into the Leipzig Salient. Unfortunately there seems to have been no idea of any further advance close behind the barrage, indeed the map in the R.A. June Diary, which shows 10 distinct lifts, makes it quite clear that the lifts were from trench to trench, and in no sense was it a creeping barrage. Consequently, with the original penetration that had been achieved at zero, the advance of the 97th Infantry Brigade came to a standstill. Later on the Brigadier heard that the advance of his infantry had been checked in the Leipzig Salient but that the barrage was carrying on and lifting according to plan, so General Jardine gave immediate orders that two batteries should be taken out of the barrage and switched onto the defence of those men who were still maintaining a precarious hold on their gains in the Salient. This timely artillery co-operation undoubtedly went far towards assisting the hard-pressed infantry to hold what they had gained at the first onrush.

The projected attack of the 3rd Objective on this front was to have been entrusted to the reserve brigade, the 14th, and, accomp-anying the July documents of the 19th Lancashire Fusiliers of this Brigade, there is a sketch with a table showing the distinct lifts which were to be made by the covering 18-pdr. batteries when on this front the 14th Brigade attacked the German Second Line. The sketch shows that the barrage was not creeping forward, but jumping from one trench line to the next. In the 36th Division, on the left of the 32nd, the true creeping barrage was not employed, but the lifts were from one trench to another, whilst some 18-pdrs. and 4.5" hows. "walked up" the communication trenches. South of the Ancre, the first onslaught of the 36th was extremely successful and a deep penetration was achieved, though the gains could not be maintained.

In the VIII Corps to the north of the Ancre, a Scheme for the Offensive (69 typed foolscap pages) was issued on the 11th June. In this detailed scheme, if anyone ever lived long enough to read it through, there was a subsection (No. 13) dealing with Artillery Lifts which laid down that "times marked [on a certain map issued with the Scheme] are for Divisional Artillery Lifts. Heavy Artillery will lift, in all cases, five minutes beforehand. At the commencement of each infantry attack, the Divisional Artillery will lift 100 yards, and continue lifting at the rate of 50 yards a minute to the objective, firing 3 rounds of gunfire at each step. The Heavy Artillery will lift straight onto the objective...." Later on in this subsection there was a subdivision which furnished Further Notes on Artillery Lifts, and relevant extracts from them read as follows:—"... The rate of advance of the infantry has been calculated at 50 yards per minute.... It is the intention of the Divisional Artillery to assist the infantry forward by lifting very slowly 50 yards each minute.... The Infantry therefore must make their pace conform to the rate of Artillery lifts. ... The success or otherwise of the assault largely depends on the Infantry thoroughly understanding this 'creeping' method of the Artillery.... The Infantry must be taught to realise that it is better to be occasionally checked rather than our artillery should lift off the objectives too soon, thereby allowing the enemy to get their heads up and open rifle and machine-gun fire on the assaulting lines...." It will be noticed that these instructions for the lifts of the barrage in the VIII Corps were very complete and up-to-date, and even the word "creeping" was actually employed. The employment of this phrase was intentional, for in some further instructions, with regard to the advance on the 1st July from the 2nd to the 3rd objectives, the 4th Divisional Artillery was ordered "to creep forward" towards the Puisieux-Grandcourt road from the wire running north and south between the Beaucourt-Serre Road and the Puisieux road; and the 31st Divisional Artillery was ordered "to creep up the village [of Serre] at 50 yards per minute...."

In the Divisions which attacked on the VIII Corps front the right division, the 29th, on whose front the Hawthorn Redoubt Mine was fired, arranged for the guns to make short lifts. In the instructions issued about the 20th June by the E.G., R.A., para. 6 reads: "When lifting off the front line at 0000 fire will be brought onto the support line, but in no case will the lift be less than 100 yards. Successive lifts will be made every two minutes.... 100 yards at each lift...." In the 4th Division, attacking to the north of Beaumont Hamel and on the left of the 29th, in the Divisional Artillery Orders issued on the 22nd June the para-graph dealing with Artillery Lifts stated: "... The rate of advance of the infantry has been calculated at 100 yards per 2 minutes. The Divisional Artillery will assist the infantry forward by lifting very slowly at the rate of 100 yards every 2 minutes...." This method of lifting is described as "lifting by steps." These orders also definitely connect the rate of progress of the barrage with the advance of the infantry, and they give still another name to the barrage.

In the 31st Division, attacking on the left of the Corps front, in the Divisional Instructions for the assault, issued on the 26th June, and repeated in the Divisional Artillery Instructions the batteries are ordered in one case "to creep forward by 100 yards every two minutes," so that all through this Corps the creeping arrangement was well recognized. Yet on the 1st July the attack of the VIII Corps was a tragic failure. Possibly this was partly accounted for by the fact that although zero was at 7.30 a.m. yet the Corps heavy artillery lifted at 7.20 a.m. off the German front line to the reserve trenches, and at 7.25 a.m. they were joined by the howitzers which had been firing on the support trenches, whilst the 15" and 12" hows. pounded Beaumont Hamel and Serre. If any further warning was needed by the Germans that the assault was coming the Hawthorn Redoubt Mine (40,000 lbs.), on the right of the Corps front and in front of Beaumont Hamel, was fired at 7.20 a.m. Thus the up-to-date artillery arrangements of the Corps Offensive Scheme were nullified.

To the northward again on the 1st July the VII Corps of the Third Army made a subsidiary attack which was intended to pinch out Gommecourt, the most westerly point of the German line on the Western Front. For this attack the artillery issued a very detailed barrage map, but so far as can be ascertained in the 56th Division the field artillery seem to have lifted from barrage line to the next whereas in the 46th Division it was laid down that the batteries on lifting off the first barrage line would search and sweep back to the second line, and in a few cases thereafter the same procedure would be followed. Actually the attack of the 56th Division, delivered with great dash, met with considerable success at first, although the assailants eventually found it impossible to maintain the ground that they had gained at the first onset. The attack of the 46th Division failed badly.

The foregoing examination of the orders makes clear that along the front of attack on the 1st July there were several definite indications of the use of some form of barrage that might be looked on as a direct forbear of the creeping barrage; and both in the XIII and XV Corps, the two corps in which the greatest success was achieved, there were very distinct attempts made to advance the barrage by short lifts, putting it down in front of the attacking infantry as the men pushed forward.

We must now briefly consider the next operations on the Somme. With the capture, or occupation, of Fricourt on the 2nd June the first phase, the so-called "Battle of Albert," came to an end. To the north of Ovillers the attack was temporarily closed down, whilst to the southward a slow advance by siege methods had to be adopted, so as to gain a suitable jumping-off place for the second phase: the assault on the German Second Line which ran in front of Maurepas, Longueval, and the Bazentins, and behind Pozieres. In addenda to Operation Order No. 65 (X.4/1488/15) of the 9th Division, issued on the 7th July for the attack on Trones Wood, there is one paragraph which shows that the XIII Corps had not abandoned the idea of artillery fire creeping forward:—"... 3. The Lifts of the XIII Corps Heavy Artillery Group are as under:—Until zero minus 5, gas shells will be fired into Trones Wood. At zero, heavy howitzers will be firing at a point in Trones Alley 200 yards east of Bernafay Wood and will creep eastwards along Trones Alley." The 9th Division held Bernafay Wood, and Trones Alley ran between the southern ends of Bernafay Wood and Trones Wood, a distance of about 500 yards. Two days later another indication is to be found of this innovation in the 7th Divisional Artillery O.O. No. 19A, issued on the 9th July for the attack on Mametz Wood which was to be delivered on the 10th; and in this order the paragraphs (6 and 7) which give the lifts were amended to read as follows:—"... Barrages...At 4.15 a.m. all guns will search back or inwards (i.e. away from advancing infantry) by short quick lifts of about 50 yards every minute...."

On this same day (the 9th) the 25th Divisional G.S. Diary records a change that was soon to become almost the normal procedure:— "... From this date onwards the Division worked as a whole, less Divisional Artillery which only once later came under the Division. It became evident thus early that the Divisional Artillery would cease to form an integral part of the Division." The 25th and 32nd Divisions at this time held part of the X Corps (Reserve Army) front from La Boisselle northwards to Authuille Wood, and whilst the front of the 25th Division was covered by the 12th Divisional Artillery that of the 32nd Division was covered by the 25th Divisional Artillery. There is little doubt that it soon became evident that the creeping barrage, or "rake," or whatever it might be called locally, was of little use unless it was of considerable density, hence as each division was taken out of the line it left its artillery in action. Also all through these operations any apparent lack of communi-cation between the artillery and infantry was most probably due to the communications being incessantly cut.

On the 10th July, in the attack of Contalmaison by the 23rd Division (III Corps), the lifts were not carried out as arranged in the original programme. During the attack the artillery observers reported that the infantry were running into the artillery barrage, an enfilade one, and the B.G., R.A., promptly took the respon-sibility of advancing the time of the last lifts. Thus in this early creeping barrage the time of certain lifts were actually altered successfully whilst the attack was in progress.

So we come to the dawn attack on the German Second Line from Guillemont to the Bazentins. In the Fourth Army Operation Order (32/3/35G), issued on the 10th July (when the attack was planned to take place on the 13th) it is stated that "... the bombardment ... will be as intense as possible ... for a short period prior to zero, when it will lift as arranged by Corps...."

There is nothing in this order that suggests that at Zero the barrage will lift and creep forward closely followed by the assaulting infantry.

In this attack the front allotted to the XIII Corps, on the right of the Fourth Army, ran from Trones Wood to the eastern edge of Bazentin le Grand Wood. In the left division of this Corps, the 3rd, the artillery orders, issued on the 13th July, show no trace of the employment of a creeping barrage; at zero the 18 pdrs. lifted back onto the support line, the next lift was one of 200 yards, the next of 100 yards, and the last lift was one of 300 yards. But in the XV Corps, on the left of the XIII, an indication of the creeping barrage is found. In this attack on the 14th July the XV Corps front lay between the eastern edge of Bazentin le Grand Wood and the western edge of Bazentin le Petit Wood. The Corps Instructions, issued on the 12th July, for the final bombardment and barrages on Z Day, were accompanied by a Barrage Map which showed 5 lifts, "B" to "F" , timed to take place at 1 hour, 2 hours, 2.20 hours, 2.40 hours and 3.10 hours after zero, and extending back from the middle of the Bazentin Woods to the middle of Bazentin le Petit village, but the orders accom-panying the map made clear that "... the guns will search back quickly to the next line by very short lifts..." and the entry in the Corps G.S. Diary for the 14th July records: "... The barrages worked well, 7th Division infantry, in particular, advancing close under the barrage, reached their objectives with but slight [missing text]. In the 7th Divisional Artillery O.O. (No. 20), issued on July, it is laid down: "... 4... 18-pdrs. will always search back by short quick lifts to the next line..." and in an amendment to this same order further detail is given: "... 2 ... At 0.9 all guns and howitzers will lift back to the 'Red Line' in short lifts, 50 yards a minute...." In this same division for its attack on High Wood on the 20th July the Artillery Orders (No. 23, issued on the 19th) prescribed: "... 5 ... The first lift will take place at 0 and the second at 10 minutes. The lifts will be made in short jumps of 50 yards at a time...."

Thus we see that the change sprang from the line in the normal way, and it was gradually adopted by the attacking divisions, so all that remained to be done was for G.H.Q. to give the innovation its official blessing, and this was duly given on the 16th July; in a circular memorandum (O.A. 256) issued by the Chief of the General Staff from Advanced G.H.Q. at Beauquesne. It ran as follows: —

"One of the outstanding artillery lessons of the recent fighting has been the great assistance afforded by a well-directed field-artillery barrage maintained close in front of the advancing infantry. It is beyond dispute that on several occasions where the field artillery has made a considerable 'lift', that is to say has outstripped the infantry advance, the enemy has been able to man his parapets with rifles and machine guns. It is therefore of first importance that in all cases infantry should be instructed to advance right under the field artillery barrage, which should not uncover the first objective until the infantry are close up to it (even within 50 or 60 yards). ... An infantry brigadier, whose command has met with conspicuous success, ascribes it largely to the fact that his men have insisted on advancing close under the field artillery fire ... and he has stated that on more than one occasion his men were thus enabled to gain an enemy's trench almost without loss, and in time to meet the defenders hand-to-hand as they emerged from their dug-outs and before they could mount their machine guns.
(sd.) L. E. Kiggell, Lieut.-Gen., Chief of the General Staff."

A month later in a Fourth Army Operation Order (No. 32/3/98G) for the attack on the 18th August against Guillemont and Delville and High Woods, there is this clear indication: — "... 3... At zero the infantry will leave their assembly trenches and advance to the assault covered by the Field Artillery Curtain, which will be regulated by Corps in accordance with the distance of the assembly trenches from the first objective. The infantry will follow as close behind the curtain as possible till they reach their objectives...." The innovation was receiving official approval in the highest quarters. Closely following on the above, further recognition of this innovation was given in a brochure, "Preliminary Notes on the Tactical Lessons of the Recent Opera-tions" (S.S. 119) that was issued in September 1916. This brochure emphasizes that, "... the infantry must conform absolutely to the time-table of the various lifts, and must advance as close to the Barrage as possible before it lifts.... [Under bombardment] the Germans ... are in the habit of ... taking up positions in shell holes in front and in rear [of their trenches]. A well directed field artillery barrage maintained close in front of the advancing infantry with very gradual lifts will enable such situations to be effectively dealt with...." Official blessing having been given to the creeping barrage, we should not expect to have to wait long for its universal introduction, for a name to be given to it, and for it to appear under that name in Artillery Orders; and in the XV Corps Artillery Operation Order (No. 47), issued on the 12th September in connection with the arrangements for the attack on Flers on the 15th (sometimes called the Battle of Flers-Courcelette), there is this paragraph: "... 4. The 18-pdr. Barrage will be divided into two parts: (a) Stationary, (b) Creeping. Two or three batteries per Divisional Artillery will be employed in search-ing approaches and the more distant communication trenches, headquarters, and roads. Of the remainder, three-quarters will be employed in the Stationary Barrage, and the remaining quarter in the Creeping Barrage." In paragraph 6 (a) of the same order it was explained that "the creeping barrage, from 6 minutes after zero, was to search back by lifts of 50 yards every minute." Naturally one point, and an important one, that had to be settled was how slowly the barrage must move so as to search out the shell holes thoroughly and also for it not to run away from the heavily loaded infantry who were unable to move very rapidly across the shell-torn ground. An artillery officer walked over the ground after the attack on the 15th September, in which a creeping barrage was fired from Pozieres to Courcelette at the rate of 100 yards in 3 minutes, and he said: "There was no doubt that a barrage at this rate fired by shrapnel was effective. The machine gunners were killed in their shell holes, and the shell cases covered the ground in a regular pattern." After more than a year of experi-ment and gradual evolution the creeping barrage had come to stay, and from now onwards constant use developed its application.

NOTE.—It is hoped that any officers in possession of further information upon this subject will send it to the writer at the Historical Section, Audit House, Victoria Embank-ment, E.C.4, for the use of the Official Historian.