Chapter X: The “hush” Operation and the Cockcroft

I.

The result of the first day’s fighting at Ypres was on the whole satisfactory. The left of the attack had been quite successful. The French, Guards, and 51st Division had carried their final objectives, and the French had even gone beyond theirs. But from St Julien southwards things had not gone so well. Repeated counter-attacks during the afternoon and night drove our infantry out of St Julien itself, and away at Hooge the net result of twenty-four hours’ terrible fighting was a very small gain indeed. The 2nd and 3rd Tank Brigades, operating in this southern sector, suffered heavy casualties, although the average time during which their machines were in action was comparatively short--five hours in the case of the 2nd Brigade and seven in that of the 3rd. The same average in G Battalion, the only representative of the 1st Brigade, worked out at sixteen hours. The two southern brigades sent in between them 94 fighting and 13 supply, signalling, and cable-laying tanks--a total of 107. Thirty-three were disabled by shell-fire, a similar number remained ditched at the end of the day, and 22 had mechanical breakdowns. This last figure appears extraordinarily high. Nine officers and 27 men were killed, 29 officers and 168 men were wounded, and 8 men were reported missing--in all, 241 casualties out of some 800 of all ranks who went into action. In many cases valuable help was given to the infantry, especially by the 3rd Brigade (C and F Battalions), working with the 19th Corps against Spree and Bank Farms, Beck House and Capricorn, and Frezenberg. The 2nd Brigade was hampered by the appalling state of the ground in the Hooge area. Most of its tanks were late in coming up, and the infantry in the meanwhile having been checked, machine after machine was hit in the narrow defiles by which they endeavoured to get forward. Our counter-battery work, thorough as it appeared to be, failed completely in this region to keep down the German artillery fire. The two tank battalions, A and B, lost more tanks and men than the other two brigades together.

Individual tank actions during this day presented, one and all, similar features, and I shall not consider them here. The machines dragged themselves laboriously through the mud, became ditched and extricated themselves, and then probably became ditched again. Crews, while in the open during unditching operations, were much harassed by machine-gun fire from low-flying aeroplanes. There were fifty-one recorded cases in which tanks assisted the infantry. Sixteen of these were credited to our two companies on the left, and there can be no doubt that G Battalion, in its first action, carried off such honours as the Heavy Branch gained in that dismal battle. Whether this performance was due to the ground in our sector being a trifle less atrocious than elsewhere (although in all truth it was bad enough), or whether, as one likes to think, it was due rather to the excellent training and spirit of the battalion, must remain a matter of personal opinion. It is certain, however, that our discussions with the infantry beforehand led to a standard of co-operation in the field which does not seem to have been reached in other cases. On several occasions the infantry directed our tanks on to objectives, or called them up from a considerable distance; for we had brought their officers to understand that the view from a tank is so restricted that without such help half its value will be wasted.

‘Weekly Tank Notes,’ a concise history of the corps issued recently from the Headquarters in London, sets out the main lessons to be drawn from this day’s fighting, as follows: “The unsuitability of the Mark IV. tank to swamp warfare; the danger of attempting to move tanks through defiles which are swept by hostile artillery fire; the necessity for immediate infantry co-operation whenever the presence of a tank forces an opening, and the continued moral effect of the tank on both the enemy and our own troops.” This last was exemplified very strikingly in the case of one of 21 Company’s tanks, which led the 51st Division to the Steenbeek, and helped it to establish itself there, without firing a shot. Successive German posts surrendered as soon as the tank came within range. As for “swamp warfare,” the unsuitability for this of an experimental machine weighing thirty tons had been manifest to every one in the corps itself long before. But to infantry commanders, regular British officers with the conservative outlook and the prejudices common to their type, this consideration carried no weight. They had never liked this newfangled weapon. It was absurd, in their eyes, to attempt to introduce the tactics of naval warfare on land, and they seized upon the inevitable failures at Ypres as a happy confirmation of all their doubts. The three corps commanders concerned supplied the 5th Army H.Q. with their opinions on the work of tanks during the battle. Major Williams-Ellis, in his book, deals with this incident so effectively that I cannot do better than quote him.

“The three summaries were agreed that the courage and
perseverance shown by tank personnel had been admirable. One
corps, however, had given way thoroughly to the spirit of the
time. They practically reported that tanks had been of no
use to any one, and moreover that they were never likely to
be. With the 30th Division they had been unable to deal with
certain machine-gun emplacements; with the 24th they had been
late--they always drew enemy shell-fire; and with the 8th
Division one tank had even lost direction and been reported as
firing on our own men. Another corps had found tanks helpful,
and said all they could for them. Tanks had greatly assisted
the Gordons and Black Watch at Frezenberg; they had dealt
effectively with concrete dug-outs; with the 55th Division they
had broken the wave of an enemy counter-attack at Winnipeg, and
everywhere their moral effect on the enemy had been of great
assistance. Twenty-four tanks had been put out of action by bad
going or shell-fire.

“A third corps, with fewer machines, had in many cases reached
their objective without being held up. The tanks had in these
cases merely followed the infantry, but they reported that
without tanks the capture of the strongly-wired position of
Alberta would have cost the 39th Division dear, and that on
the Steenbeek near Ferdinand Farm the enemy, who had bolted at
the mere sight of a tank, had been ‘dealt with’ at ease with
a machine-gun by infantry of the 51st Division. Upon these
summaries and upon later failures the commander of the 5th Army
was subsequently to base a generally unfavourable report upon
tanks.

“The report may be condensed into a simple syllogism:--

1. Tanks were unable to negotiate bad ground.
2. The ground on a battlefield will always be bad.
3. Therefore tanks are no good on a battlefield.

“He added to this, that being no longer a surprise to the
enemy, he considered that tanks had lost their moral effect,
and had no value used in masses. This report was not officially
presented for some weeks, but the Higher Tank Command must
early have perceived the drift of affairs.”[17]

The prospects of the Tank Corps, in fact, were become suddenly very gloomy. I cannot say how near the corps came to extinction, or at least to drastic limitation, at this period; but by all accounts it came very near. All the prophets of evil were up and about and busy, saying everywhere, “I told you so!” The labour of three years, the sacrifice of many lives, seemed likely to come to nothing after all. A weapon which, properly applied, could alter the whole face of the war, was in imminent danger of abandonment because it had been used, in defiance of all authoritative opinion, amid conditions for which it was wholly unfitted. By a parity of reasoning the cavalry might have been threatened with abolition; for their patrols were supposed to be at hand to cross the Steenbeek, and not a trooper got even as far as the canal. And, when one comes to think of it, a state of mind which can hope that cavalry will be of any use whatever at such a stage of such a battle must be capable of the most astounding misjudgments. It was in this state of mind that too many senior officers, even after three years of shattering surprises and disillusionment, looked upon tanks or any other new thing. Faced by so obstinate a prejudice, now fanned into activity by the events of the 31st July, the position of the new arm obviously was precarious. It is possible also that the Treasury, always alarmed by the growth of a very expensive organisation, and encouraged by the fact that the Tank Corps, unlike the R.A.F., had few friends and no publicity agents, was backing up any proposals made at home for economical reform. In short, it looked very like becoming a question whether the corps would have another chance of justifying itself before its fate was taken out of its hands. For, in the meantime, at Ypres the rain still fell, the whole offensive came virtually to a standstill, and the mere idea of further tank operations in that dissolving world seemed ludicrous. One such attempt, indeed, was made, and ended as an inglorious but inevitable fiasco.

Before we come to this, however, and to the surprising success which, following immediately after, re-endued the Tank Corps with life and hope, another and unique enterprise falls to be noticed.

II.

The most interesting phase of the third battle of Ypres, as originally planned, was never fought. This was the landing on the Flanders coast. Tanks were to have played an important and indeed essential part in the undertaking.

The landing was designed to take place in the neighbourhood of Middelkerke, a little _plage_ midway between Nieuport and Ostend. The sea-front along the whole of this flat coast is protected by a high concrete wall, on the top of which runs the usual esplanade and roadway. On the other side of the road are the villas and cafés and _pensions_ where, in happier days, visitors crowded during the summer and autumn. The whole esplanade now was all but hidden by barbed wire, and there was more wire along the beach; batteries were emplaced at close intervals along the wall; and the ground-floors and cellars of the lodging-houses were become concrete machine-gun posts. A landing here presented extraordinary difficulties, but the main problem before the assailants (assuming that all else went well) was how to haul field-guns and supplies rapidly over the concrete wall under fire. Infantry might be got ashore under cover of a smoke-screen and a bombardment from the sea; but infantry alone, even if it could capture that fortified esplanade, could not hope to hold it. Guns were essential, and, if possible, some other factor which would upset the enemy’s _moral_ and prevent him recovering from the initial surprise. For this purpose tanks, of all things, were to be employed.

The attacking force consisted of the 1st Division, several batteries of field-guns and 4.5 howitzers, a company or two of engineers, and nine Mark IV. tanks. To throw this considerable force ashore three immense pontoons or barges were constructed, each 700 feet long. These were taken over to Dunkirk harbour toward the end of July. On the night of the attack (which, as I have said before, was to follow the capture of the Passchendaele Ridge, provided that was effected within three or four days) these barges were to be pushed on to the beach by pairs of monitors, lashed together, one barge at Middelkerke itself and one on either side of that place. Each barge carried a brigade of infantry, several batteries, three tanks, and some motor-lorries. The landing was to take place just before dawn, covered by a dense smoke and ordinary bombardment from warships supplied by the Dover Patrol, which was in charge of the naval side of the enterprise.

The wall itself was a very serious obstacle. It rose to a height of some thirty feet, sloping up from the beach at an easy angle, but ending at the top in a vertical coping with a lip or overhang. A tank, fitted with special shoes on the tracks, could climb the slope easily enough, but the vertical coping was too high to be surmounted without additional appliances. To each tank, therefore, was fitted in front a massive detachable steel ramp, slightly wider than the machine, supported on lattice-work girders carrying a pair of wheels, and so constructed that it sloped upward from the inclined part of the sea-wall to the top of the overhang. When the tank reached the foot of the incline, the ramp would be lowered by tackle until its wheels rested on the concrete, pushed up against the coping, and there detached. The tank could then climb over it on to the esplanade. Afterwards this highly ingenious contrivance was to be used for the haulage of guns and motor-lorries over the obstacle, the tanks being fitted with special cables and attachments for this purpose. Another clever device was a system of wooden sledges carrying ammunition-boxes, the sledges being simply jammed on to a rope and hauled up by the tanks, releasing themselves automatically as they cleared the top of the overhang. The actual disembarkation from the huge barges to the beach was effected by strong hinged gangways, which fell over the bows.

The troops detailed for this hazardous attempt were trained very thoroughly. The infantry, segregated for weeks in a camp near Dunkirk, practised repeatedly the whole attack with the aid of a concrete model of a section of the sea-wall. A similar model was built at Merlimont, where the special tank detachment, commanded by Major the Hon. J. D. Y. Bingham, and made up of volunteers from the original four battalions, climbed up and down with their machines, towed up guns and lorries, and demonstrated conclusively that this essential part of the enterprise was perfectly feasible. The ramps worked admirably. Major Bingham informs me that tanks made as many as 200 ascents of the model at Merlimont.

Had this attack taken place, it would have been profoundly interesting. Obviously it was an immense gamble. It bore many points of resemblance to the successful raid at Zeebrugge, with this great difference: it was intended not as a raid but as a permanent landing. Once troops and tanks were ashore, it is difficult to see how they could have been got off again if things went wrong. The great barges would have been blown to matchwood if they remained on the beach after daybreak, unless an extensive clearance of the sea-front was effected at the very start. The roadway above the sea-wall was none too wide, and there were doubts as to whether the tanks would have room to tow the guns, to say nothing of the lorries, up the narrow ramps. It was found afterwards that the lower storeys of the houses across the road were veritable fortresses of concrete. There were a hundred other obstacles to success which might arise, such as bad weather, premature discovery, or even a wind sufficient to dissipate the smoke-screen. And it seems probable that, well as the secret was kept in England, the Germans had some uncertain information about it. I believe that the camp near Dunkirk was bombed with quite exceptional thoroughness. The chances, in short, seem to have been very greatly against success, and failure would have meant a ghastly slaughter on the beach and along that narrow esplanade. On the other hand, whatever the Germans suspected, it does not seem likely that they possessed any exact information. Some sort of a landing somewhere on the Flanders coast was always a bogey to them. And after so prolonged a crying of “Wolf!” the prospects of a genuine surprise, if one could be effected, were very hopeful. Above all, if the German machine-gunners, buried in their concrete basements and discouraged by news of a successful advance at Passchendaele, threatening their whole coast-line, were to be confronted suddenly with tanks where no rational human being could expect tanks to be, a widespread panic was possible. One can imagine these men, deluged with gas and high explosive and blinded by smoke, peering anxiously out of their loopholes for human targets, and seeing instead, huge and distorted in the gloom, tanks which seemed to have arisen from the sea....

But this great experiment was never attempted. By our sanguine calculations, it was due for the 3rd or 4th of August. It was admittedly useless, however, to land a force at Middelkerke unless the armies from Ypres were in a position to do their share of the work; and at the end of the month, and for many weary weeks after, the armies from Ypres were still floundering in a morass a bare mile or two from where they started. Passchendaele and its insignificant ridge were not ours until November. And with the ruin of its parent enterprise, the conception of the Flanders landing, the great “Hush” operation, necessarily came also to nothing. Long before we had even a footing on the ridge, the amphibious attack had been abandoned. The infantry dispersed to more normal fields of action, the great barges lay derelict, like Napoleon’s, in Dunkirk harbour, and the special tank detachment was disbanded. The scheme remains one of those unfulfilled but deeply interesting adventures over which any person of a speculative mind could write pages of inconclusive arguments.

III.

We will return from these cursory surveys of the whole field of operations to the small world of Oosthoek Wood and the camp by Lovie Château.

The morning after I rejoined from Frascati, the tank commanders left in our two companies--eight out of the twenty-three had become casualties--were interviewed by the brigade commander and his Intelligence officer, that same Major (then Captain) Williams-Ellis whose book I have quoted. For the next two or three days most of us were suffering from the reaction after the strain and fatigue of the approach-marches and the battle itself. For all tank personnel (and this applied to every other arm) the fighting at Ypres was by far the most exhausting and nerve-wracking ordeal undergone throughout the war. The dreadful conditions of ground and weather, on top of all the ordinary trials of warfare, brought every one to the limit of endurance very quickly. The hard labour involved in taking and fighting tanks anywhere was multiplied tenfold in the Salient; and after a severe action there, it was a physical impossibility for any crew to go in again until it had been given a complete rest. On other fronts the same tanks and men often were fit enough for a renewed march and attack after a few hours’ sleep. In our case there was now little or nothing for some of us to do, for the two companies were left for the moment virtually without tanks. Of the twenty-three which went into action, one only returned the same day to the neighbourhood of Frascati. Four, which had received direct hits, were lying in the new No Man’s Land around St Julien; others, like G 46, were derelict in positions where salvage work at present was out of the question; and the rest, half engulfed or mechanically crippled, or both, lay scattered about the west slope of the Pilkem Ridge. During the next fortnight most of these nearer machines were brought back to a spot known as Ghent Cottages, near Brielen, where the battalion had established an advanced base.

In the meantime it rained. The camp at Lovie, its leaky tents immersed in dripping shrubs and undergrowth, and surrounded by sodden parapets of sandbags as a protection against the persistent bombing raids, grew always more evil-smelling, steamy, and unhealthy, and those of us who had little to do became more melancholy every day. There was a time, some four days after the battle, when the reaction was at its worst, and when, personally, I felt I could cut my throat for twopence. An expedition up the line to retrieve my six Lewis guns from Maelor Jones’s tank was a diversion, but not a strikingly attractive one. For once, it is true, it was not raining; I do not know why. We went by lorry to Salvation Corner, a miserable road-junction near the Dead End of Ypres, and thence walked once more over the familiar route by Bridge 4, Frascati, and La Brique. At Frascati, under the trees where our tanks had lain during Y-Z day, 9.2 howitzers were being “assembled”--if one may use this expression with regard to such imposing weapons. A few duckboard tracks had been laid over the tumbled waste of mud and water beyond, by which we proceeded with comparative ease for a part of the way. I was able for the first time to examine at leisure the ground over which we had advanced, and it appeared to me to be rather astonishing that tanks had got over it at all. A good number of dead were still lying about, among them a whole batch of Highlanders, I remember, near Hurst Park, and everywhere hung the horrible sweet odour of decomposing animal matter. Away on the top of the ridge the Germans, with a curious futile industry, were shelling with heavy metal the big concrete mebus in front of Kitchener’s Wood.

Our directions for finding the tank were not too explicit, and even so large an object was easily missed in that distorted wilderness, where everything was mud-coloured. When at length we discovered it, we found it had been hit on one track. The guns were lying rusting inside amid a litter of tools, petrol-tins, and rubbish, and mud was everywhere. I had intended to go on to G 46, if the situation seemed propitious, to collect Kessel’s cherished flag and my binoculars, which I had left somewhere under water, but the auguries were not favourable. The enemy began to burst heavy black shrapnel overhead. As no one else was within sight, I concluded that my little party probably was under observation, and the epidemic of high explosive on the crest appearing to spread at the same time, we made for home with all the speed of which we were capable, hampered as we were by half a dozen Lewis guns. Johnstone, who was not in good health, presently became so ill that I took his burden from him. It was while I was staggering along under this double load, gingerly circumnavigating craters and cursing the very name of the American Colonel who invented the weapons, that I stepped literally into an inglorious predicament. I tried foolishly to cross a well-churned artillery track, went into the soft mud up to my knees, and there stuck, absolutely helpless, with the wretched guns occupying both my hands and weighing me down, and at the same time totally unable to move my legs. It required the cautious assistance of most of my crew, helped by some providential gunners who brought bits of board, to get me out.

There ensued another week of rain, squalor, and depression at Lovie Camp. Our doctor was despatching officers and men by twos and threes to a rest-camp at Wormhoudt, just over the Belgian frontier, and I managed to get Johnstone away with one of these parties. He was not over strong, and he had never recovered from the strain of our thirty-six hours’ vigil on the ridge following on that of driving for the best part of three successive nights and a part of the fourth day. Neither of my other drivers had his knowledge or skill, and during the latter stages of the approach-march, when he should have been resting, I had been obliged to call upon him. As it happened, he was still at Wormhoudt when the sudden call came for the Cockcroft operation. In the meantime the general outlook, from what we could deduce, was not encouraging. The great offensive, launched with so vast a preparation, had come to a sudden standstill at the very start. Fighting continued every day: there were minor attacks and counter-attacks all round the Salient, and we had blown our way once more into St Julien and into various other posts captured and then lost again on the 31st; but in essentials our line remained where it had been on the evening of that day. Already more than a week had passed since, by the original time-table, Passchendaele should have fallen, and now it was tacitly confessed that the whole ambitious plan (persistently described as a “residue” of something even greater, although before the end almost every division in the British Army was involved and decimated in it) had dwindled to a struggle, amid conditions indescribably ghastly, for that single ridge, which at one time was to have been but the starting-point for infinitely bigger things. Into this struggle the Tank Corps was drawn virtually to the last tank and the last man. We have seen what happened on the 31st, and how hopeless it was, even then, to expect decisive results from these machines in such a country. A fortnight later, when the ground was a mere slimy pulp, 20 Company of G Battalion moved up to take part, if possible, in a second serious attempt to advance.

This attack was fixed for the 16th August. 20 Company, which had been in corps reserve on 31st July, had since moved forward to our old park at Halfway House. On the 13th the twelve tanks started for Frascati. The ground everywhere was now in so appalling a state that one machine actually became bellied in Rum Road, a few hundred yards from the start. As she could not be got out that night, I was detailed on the following day to take my crew down and see what I could do, while her own people came back to camp for the short rest. We worked the whole afternoon, digging round the tank, which was sunk to the sponsons, and building a small causeway of timber and brick-rubble in front; and when her crew returned in the evening they were able, with the help of the unditching beam, to pull her out at once. But if this sort of thing happened on Rum Road, what would happen in the shelled zone? In fact, the progress of the company across the old trench systems and up the slope toward St Julien, during the dark hours of the 14th and 15th, was one long nightmare of ditching and unditching under shell-fire and gas. Four only got anywhere near the front line, and at length it was clear that even these could not arrive in time. When the infantry attacked, these four were still a mile behind. So far as the 1st Tank Brigade was concerned, this was the last attempt made to move tanks across country in the Salient.

The infantry assault was, in results, very similar to that of the 31st. The left, where Langemarck was carried by eight o’clock in the morning, was successful, but from St Julien southwards little or no progress was made. In this part of the field we had now come up against the main system of fortified farms and other strong field forts already mentioned. A cluster of these forts, just beyond St Julien, checked every attempt to debouch from the village up the Poelcapelle road; and until they could be carried no further general advance was possible, as our line now bulged out into an awkward little salient in the north. General Maxse, commanding the 18th Corps, was told that an attack on these forts would cost him 600 to 1000 casualties, without any guarantee that we should capture them in the end. The Lieutenant-General, who had more faith in tanks than some of his colleagues, for they had served him well on the 31st, consulted with the Brigadier of the 1st Tank Brigade. The position of the forts, disposed about a triangle of roads, was such that tanks could get close to them, and in some cases absolutely alongside, without necessarily venturing into the slough of mud on either side of the _pavé_. Details of the operation were accordingly hurriedly worked out. One company of tanks was to be used, and G Battalion, straitened for machines as it was, once more was called upon. This, at least, was a notable compliment, for D Battalion, one of the old originals, as it never failed to remind us by word and behaviour, had not yet been into action at all.

IV.

About midday on 17th August, Kessel, I, and several others were standing together in the marquee which served us for a company mess at Lovie, drinking the usual sherry and bitters, or something of the kind, before lunch. To us entered Paisley, second in command of the company. He held a sheet of paper, from which he proceeded to read. “The following officers will take over tanks this afternoon, fill up with petrol, oil, and ammunition, and be prepared to move at six o’clock.” The officers named were Coutts, Claughton, Chaddock, Bubb, and myself. It transpired further that four tanks from 20 Company, and three from 19 also were to move: that we were to trek that night to Bellevue, on the old German front line, and again the night following to St Julien; and that we were going into action on the morning of the 19th. The charge of this composite detachment had been given to Major Broome, who had just arrived from the depot at Erin to replace Torbett, invalided home, in command of my own company. There were to be three section commanders, Kessel, Guy of 20 Company, and Wright of 19.

Sudden orders of this kind, when one has been contemplating a quiet afternoon with a book, are always somewhat disturbing. There ensued now the inevitable bustle, for there was no time to lose, as we were leaving by lorry at 1.30. My first task was to make up my crew. I had lost one man on the 31st, and now Johnstone was away; and it was essential to get a good first driver in place of the latter. I went immediately to find Sergeant Meyrick, the section sergeant (already mentioned as Merchant’s N.C.O.), and told him what I wanted. He volunteered on the spot to come himself as an ordinary gunner under my own corporal. Meyrick was, I think, one of those rare people who positively enjoy a battle, and, what was more valuable, he was a peculiarly intelligent and competent regular soldier, who had come out with his battery in the original Expeditionary Force. For a first driver he recommended Waller, of his own crew, who volunteered as promptly. I was extremely fortunate in obtaining these two men. Leaving Meyrick and Corporal Mitchell to arrange about rations and other details, I reported, with the rest of the tank commanders, to the Major for further instructions. The first thing to be settled was the allotment of tanks to one or two of us who had none. My own, of course, was still lying by Kitchener’s Wood; but by this time some six or eight of the others had been got back to Ghent Cottages, where they were parked under a hedge in various stages of unreadiness or actual disrepair. We could just about manage to provide five dependable machines. John Alden being away at the Wormhoudt rest-camp, I was told to take over his tank, G 47, or “Gitana,” my second eponym from the variety stage. The general idea and tactics of the proposed attack were then discussed. At this time our outpost line just skirted St Julien on the north and east. A thousand yards beyond the village two important roads cross one another--the St Julien-Poelcapelle road, pointing almost due north at this point, and the Zonnebeke-Langemarck road, running north-west. A triangle, of which these thoroughfares form two sides, has for its base a third road, running eastward from St Julien and cutting the Langemarck road at Winnipeg. A fourth road, whose direction is also roughly east and west, bisects the triangle 250 yards from its apex. Around the sides, or within this symmetrical boundary, lay all, save two, of our objectives. Of the two exceptions, one, called the Maison du Hibou, was situate west of the triangle, and some 300 yards along the fourth or bisecting road; the other, the Cockcroft, was due north of this, a quarter of a mile from the apex up the Langemarck road. In the original scheme, four tanks were to devote themselves to these two objectives, while the remaining eight moved in line ahead along the roads right round the triangle, dealing with the other strong-points as they passed, and beginning and ending at St Julien. Very unfortunately, as it turned out, this plan was afterwards modified considerably, for what reason I do not know. The eastern side and base of the triangle were left for another day. Our work was confined to the two outlying forts, the Poelcapelle road, and a couple of farms at the apex. It was, in any case, an entirely novel operation in every respect, much resembling a naval action where a squadron of ships passes by and reduces a line of batteries. The tanks, if possible, were to keep strictly to the roads, as the neighbouring ground, watered by one or two small tributaries of the Steenbeek, was now reduced by rain and shell-fire to a swamp notable even in the Salient. The infantry was to follow on either side of the roads, one platoon to each tank 250 yards behind it, and was not to attempt to occupy the objectives until the tanks signalled that all was clear, or that they wanted help. The signal in either case was a shovel waved out of the manhole in the roof. There was to be no artillery preparation: the affair was intended as a genuine surprise; but at zero, which was 4.45 or dawn, an ordinary creeping barrage would come down, together with a heavy smoke barrage on the German third-line trenches, which lay behind the posts to be attacked. This was devised to blind the enemy’s observation of what was happening. A large number of extra field batteries were brought up to supply this smoke-screen, much to the disgust of the weary gunners who manned them, and who had just gone back to rest. The infantry concerned were the 48th and 11th Divisions, the former a Midland unit. Our own arrangements for the attack and the particular objectives allotted to tanks will be better left until I come to deal with the Divisional conference held on the 18th, when we were told of the modification already referred to, and when all final details were decided.

Having obtained this rough outline of the scheme, we ate a hasty and uninspiring lunch; collected such articles as we thought we might require; were loaded in addition with maps, periscopes, compasses, and clocks; and departed by lorry at 1.30. Very fortunately the weather was fine, and remained so, by some miracle, over the next two days. On arriving at Ghent Cottages--a row of melancholy ruins on the Brielen road--I found that I had a great deal of work to get through in a very short time. G 47 having suffered from some mechanical trouble, it had taken Alden and his crew several days and nights to repair her temporarily and drive her back from Mousetrap Farm, where he had left her on the 31st, to her present quarters. Since then she had been in the hands of our workshops personnel. Although very dirty, she appeared now to be in good order mechanically, but she had to be greased up and filled with water, oil, and petrol. She was short of about a hundred Lewis drums (Alden must have used a fair amount of ammunition in capturing Alberta and engaging the German train), and many tools and spare parts were missing. This last calamity we considered, perhaps unjustly, to be inevitable when workshops had been about. I raided a supply-tank and replaced most of the deficiencies, while my crew cleared up inside, greased the rollers and stauffers, and cleaned the six Lewis guns we had brought with us from store. We had time for tea, and were ready to start when the first tank moved off just after six o’clock. I should have mentioned that there were only eleven fit tanks at Ghent Cottages. Our twelfth machine--a doubtful starter belonging to 19 Company--was on the other side of the canal, where she had lain since the 31st with a twisted extension-shaft and badly ditched into the bargain. To such an impoverished state had the Salient already reduced us!

The first stage of the night’s journey was a short one. As we did not wish to cross the canal before dusk, we pulled up before seven o’clock at Murat Farm, close to Marengo Causeway. I never saw any vestige of a farm at this place; but there was a row of tall splintered trees hung with camouflage netting, a few muddy enclosures bounded by overgrown hedgerows, and the usual squalid dug-outs and elephant-iron shelters. Behind one of the hedges was a 9.2 naval gun on a railway mounting. The brigade had formed a petrol dump here, and we took in a final supply of the spirit. Although we had come such a short distance the water in my radiator was boiling furiously, and for some time we could not discover the cause, until, with the help of Paramor, the engineer of 20 Company, who was ministering to the detachment, we found that the workshops people had unscrewed a large baffle-plate from one side of the copper envelopes and omitted to replace it. The plate was lying on the floor beneath an accumulation of oil-drums and water-tins. We readjusted this by 7.30, when the eleven tanks moved off again.

We now began one of the most wearisome and maddening treks I have ever undertaken with tanks. At first everything went well. The evening was fine: the German artillery was inoffensive, although our own was very noisy and persistent; and we crossed the causeway and the bad ground beyond without a hitch, for there was still light to see by. We passed 19 Company’s ditched tank, still immobile and plainly in no condition to attempt even to catch us up later. It was just getting dark when we swung out of the wire and shell-holes on to Boundary Road, with the trees of Frascati to our left. Boundary Road, as I have mentioned in the last chapter, eventually crosses the old front lines, passes the north end of Kitchener’s Wood, and then descends to the Steenbeek. Our route now followed it as far as our old acquaintance Hammond’s Corner, and thence by Buffs Road to the main St Julien highway. Both Buffs and Boundary Roads, presumably having been (at a prehistoric era) ordinary country lanes, had gone utterly to pieces under the impact of high explosive, and long stretches of them since our advance had been timbered. A tank on a timbered road can do a great deal of damage, especially if the driver is inexpert and swings too often. During this trek I had the misfortune to meet the R.E. officer responsible for the upkeep of Buffs Road at the moment when G 47 was swinging unavoidably round a bend, and when, to the accompaniment of loud rending and crashing noises, all the beautifully laid and clamped baulks of timber on our off or turning side were opening upward and outward like a fan. It was not my fault, as I hope he realised after a period of reflection. At the moment his language and despairing gestures seemed fully adequate to the situation, and I was too much occupied with my own troubles to spare much sympathy for his. For long before this we had begun to experience some of the vicissitudes which attend the movement of tanks by road at night anywhere near the front line. Our first encounter after getting on to Boundary Road was with one of our own machines on its way home, in charge of a young officer, Wayne, who had joined the battalion only the day before. To acclimatise him without delay, Wayne had been detailed immediately to bring in this tank from some unhealthy spot near St Julien, where it had been lying in the mud undergoing repairs since the last battle. By this time it was growing dark rapidly, and we now began to encounter the mules. I have never seen so many mules as we met that night. I did not know there were so many in the world, or that any modern quadruped could be so stupid. We had collided, of course, with the usual nightly ammunition trains, going up loaded to the forward battery positions and returning empty. Even a mule could not travel at night across country in the Salient, and there were only three or four roads for all this vast traffic. Buffs Road and its continuations rearward formed one of the most important lines of supply. It was well choked with opposing streams of mules and limbers, when into it turned our eleven tanks. The result, I hope, can be imagined: it was something beyond the power of polite language to describe. The night fell intensely black; one could see only a hundred yards or so ahead, and the world seemed to be full of mules. My tank was one of the last, so that the beasts were thoroughly disorganised by the time I reached them or they reached me. The latter state of affairs seemed the more normal of the two, for I was forced to stop continually for five or ten minutes at a time while limbers or pack-mules were pulled and pushed and cajoled past G 47. Occasionally there would supervene a thoroughly satisfactory block, when the road in front was filled from edge to edge, and, as far ahead as one could see (which was not far), by a seemingly insoluble congestion of animals and transport--the former heading to every point of the compass, some standing rooted as if stricken with paralysis, some bucking and kicking, some deliberately getting entangled in their own and others’ harness, and a number wandering casually off into the mud on either side. I was sorry for their drivers, who must have had a perfectly abominable evening, but I was still more sorry for myself. I soon lost touch with the tanks in front. They were swallowed up in the nightmare ahead, wrestling with their own problems. I was expecting every minute the whistle of a shell, and I shall never understand why the German gunners, who were very prodigal of ammunition at that time, and who possessed, in the 5.9, the most accurate of created guns, did not devote themselves every night and all night to such arteries as the St Julien road. They would have found it far more profitable than the blind “area” shooting to which they were so addicted. I am speaking, of course, in the judicial spirit that comes with comfortable surroundings where 5.9’s are not. There were nights and to spare when the Germans did shell the roads and turned them into shambles, and I shall be the last to complain because they omitted to do so that evening. Wedged for hours amid the stream of demoralised animals, the expectancy was bad enough. We were infinitely fortunate that it was not realised. For our progress up that road was in itself like some fiendish torture invented for animal haters by the S.P.C.A. Hour after hour went by, the night grew blacker and blacker, dawn, even, was drawing near, and still we were embedded in mules. Under the circumstances, no better obstacle to tanks could have been devised. We were entirely helpless. We could not drive over the mules, if only because they were controlled, in theory, by human beings, who (when they were not falling off) were blaspheming and raving in their saddles. We could not drive round them, because in the dark we did not dare to leave the road. We could only wait their pleasure. Like ships struggling against a violent current, we forged slowly, very slowly, ahead. I do not wish to appear to be exaggerating, and a trivial arithmetical computation will prove in a few lines what has taken me several pages to explain in words. From the point where we turned on to Boundary Road to our destination at Bellevue is roughly a mile and a half: it is actually 2500 yards. (I have just measured it on a large-scale map with a simple but ingenious instrument like a speedometer.) Judging by the light, it cannot have been later than 8.30 P.M. when I struck the road. When I turned off the road again at Bellevue, it was two o’clock in the morning. It had taken me five and a half hours to cover those 2500 yards--on a road. With no obstructions, the tank could have done it in thirty minutes.

If I have written flippantly about this journey, it did not strike me as being in the least amusing at the time. I walked outside, of course, throughout, and nothing is in itself more fatiguing than this sort of enforced crawl with frequent stoppages. Added to it was a consuming anxiety about shells and other things peculiar to an approach-march. Waller, who drove most of the way, must have been even wearier than I was. I do not remember how I discovered Bellevue, which was nothing but a name on the map, but I think I overtook the tank in front, thanks to some unexpected hiatus in the torrent of mules. Kessel was there, however, to guide me to a vacant space behind an exiguous hedge. The place struck me as being extraordinarily noisy. We were, in fact, surrounded by field batteries, all of which were very busily in action. The gun-flashes supplied our only illumination, and this was insufficient, as I discovered when I had fallen three times into a cesspool under the nose of my tank while helping to rig up the camouflage net. After this we lighted primus stoves, brewed ourselves some cocoa, wound up with a tot of rum, and then crawled inside with an optimistic notion of obtaining a little sleep. It was just on three o’clock.