In October 1917, during the Passchendaele attacks, there occurred the most celebrated missing-persons case of World War I.
The story of the lost company of Celtic Wood surfaces as a brief paragraph and a couple of footnotes near the end of the 967 pages of C.E.W. Bean’s 1917 volume for The Official History of Australia in the War of 1914-1918. The circumstances of the mass disappearance in Flanders, as Bean sets them down, have the classic simplicity that seems to go with the most unyielding mysteries.
At 5:20 a.m. on October 9, 1917, a day of general attack against Germans dug in on the slopes below Passchendaele, a patrol of 85 men of the 1st Australian Division were sent to raid Celtic Wood, “a large, broken copse containing many pillboxes.” The idea was to create a diversion, but it was a costly failure. Only 14 of the raiders got back unwounded, some not until the next day. “The missing,” Bean writes, “were never heard of again. Their names were not in any list of prisoners received during the war. The Graves Commission found no trace of their bodies after it.” Can we pick a moment when the episode crosses the line from bare fact to legend? Surely it is its reappearance in Leon Wolff’s popular 1958 account of the Passchendaele campaign, In Flanders Fields–a tale that is as moving and full of anger as its history is imperfect. What is so memorable about his version of the events in Celtic Wood is not that he embroiders on Bean–his only source, after all–but the context in which he places them. The seven-and-a-half-mile long offensive of October 9 stalled in the rain and mud, with mostly insignificant gains; it would be another month before British Empire troops reached the smear of brick that had once been Passchendaele. “It was simply mud that defeated us,” said the British commander in chief, Field Marshal Sir Douglas Haig; Wolff called his chapter on the October 9 affair “The Slough of Despond.”
For the preeminent Haig basher of his time, Celtic Wood was the epitome of an offensive that never should have taken place. The clear implication was that the raiders had been sucked down by a ponderous maelstrom of glutinous clay or had been vaporized by artillery fire (perhaps a little of both ). Far more than Bean–whose six volumes of official history are among the best and most thorough works written on the First World War–Leon Wolff may be responsible for making Celtic Wood the scene of the most celebrated missing persons case of World War I.
It was Wolff’s account that originally led a British amateur historian named Tony Spagnoly to seek answers to Celtic Wood. He may not yet have turned up anything conclusive; only the discovery of skeletal remains, which can be examined by forensic experts, will accomplish that. But he has brought us within hailing distance of a solution .
Spagnoly recently published his findings in The Anatomy of a Raid: Australia at Celtic Wood 9th October 1917. His little book, privately printed in England, not only reconstructs the raid and the circumstances surrounding it but deals with matters that have never been satisfactorily accounted for. How many of the raiders actually disappeared in Celtic Wood? How many returned, or in some other way were accounted for? What went wrong? And, most important, what seems most likely to have happened to the missing men?
Spagnoly has talked to the Belgian farmers who live in the immediate area. He has reproduced raid orders and after-action reports. He has followed leads in Australia. The “intolerably nameless names” (Siegfried Sassoon’s bitter phrase about the missing of World War I) begin to take on, if only for a moment in that cold gray dawn, the shapes of humanity.
Even the landscape conspires with the mystery. The first time I visited Ypres, I spent a frustrating hour trying to locate the spot where I reckoned Celtic Wood to be. I couldn’t find it–and with good reason. After the war, what was left of the former copse was plowed under. Today you can see nothing but fields sloping into an inconspicuous depression. Celtic Wood has disappeared as completely as those Australian raiders who never returned from it.
The tragedy of the event must be seen against the larger tragedy of Passchendaele. In just a few days in early October 1917, what had promised to be the most important Allied victory since the congealing of the Western Front three years earlier dissolved into a sour draw in the mud. Until then, Haig’s patient step-by-step attacks had seemed to be working, even as defense-in-depth tactics seemed to be failing the Germans. There is a section of the high ground that curves around Ypres known as the Broodseinde Ridge, named after a crossroads hamlet less than a mile from Passchendaele itself. On October 4 Australian troops spilled over the top of the ridge. They hit the enemy just as the Germans were forming up for a strike of their own–an unplanned for, preemptive attack that caught the Germans off balance and devastated them. British divisions also made significant gains. That morning German casualties approached 30,000. Their reserves were stretched perilously thin; morale was crumbling. “The black day of October 4,” the German Official History called it. Their high command was even contemplating the abandonment not just of their long-held position around the Ypres salient but of the entire Belgian North Sea coast. By midday, when the attack ceased, a decision that was truly strategic at last hung in the air. Full of optimism, Haig moved the resumption of his offensive forward a day, to October 9.
Then, half an hour after noon, a barely perceptible rain began to fall. The drizzle continued through the next day though in photos taken on October 5, the churned-up earth along the crest of Broodseinde Ridge still looks firm enough. In one, Australian soldiers in their washbasin helmets glance indifferently at the camera, or pause in the puff of a cigarette. They are packed in narrow trenches, obviously meant for temporary occupation. The Diggers (as the Australians were called) were, with the Canadians, the shock troops of the empire, spare rangy men inured to the hardships of the outdoors. They had already suffered appalling losses that autumn, but they were counted on for the next show. The machine-gun post in the picture is on a forward slope; behind it the land inclines downward toward battered scatterings of trees. One of them must be Celtic Wood.
The name Celtic Wood is one of convenience given by British army cartographers–Cyclops Wood and China Wood were among its near neighbors. Celtic wasn’t the Flemish name, of course, which is as lost today as the wood itself. Besides, what amounted to a woodlot (with, no doubt, a sprinkling of huntable pheasants) hardly deserves a name. It can’t have taken up much more than an acre.
On October 6, the constant drizzle turned to constant showers. Already the endless lines of mules and packhorses carrying up supplies for the next show were making the tracks to the front all but impassable. In the middle of that afternoon, an order came down to one of the forward battalions of the Australian 1st Division for a first raid on Celtic Wood. The object was to find out the identity and the strength of the opposing unit. The plan called for three officers and 67 men to move out into no-man’s-land just before 11:00 P.M., wait for the covering barrage, and then dash across the intervening 100 yards. But after the men were already in position, a hurried message arrived: The barrage would be delayed for an hour. When the covering fire did begin, it was, according to an after-action report, “so feeble as to be negligible.” At exactly two minutes after midnight, the men rose and started across ground that they reported was “very heavy going.”
But the results of this first raid were as spectacular as they were humiliating to the enemy. (That fact, surely, would cause problems for the next raid.) The whole business took less than 20 minutes; in that violent interval, the Australians surprised troops of the German 448th Regiment in their funk holes (we would call them foxholes) and dugouts, killing 20–could they tell in the dark? They captured a machine gun and 15 men, whom they herded back in the rain. Australian casualties amounted to two men wounded.
One report written after the return of the raiders not only gives a vivid picture of the fighting but also is the only description we have of the wood itself:
We did not observe much wire or front trench. Three dug-outs were bombed. These were fairly deep underground and had staircases…. One man reported seeing a field gun fire at us through a concrete emplacement concealed in an old farm building…. It fired through the rafters and was protected by machine guns.
So the place did not contain the “many pillboxes” that Bean assigned it. The report concludes by remarking that “The ground was very wet and swampy, and the wood itself badly knocked about.”
Now we begin to enter the Twilight Zone. Or do we?
October 7, wrote Bean (who was there as a war correspondent), was a day of ” bitter, drenching squalls.” Still, preparations continued for Haig’s next ” hopover” (as the Diggers called going over the top). But now his methodical plan began to flounder in the deepening mud. Batteries could not reach their intended positions, and even when they did, records one Australian unit diary, ‘”The guns sank lower in the mud with every shot.” Horses packing in ammunition “sank down out of sight, the driver just keeping the [horse’s] head up until assistance arrived.” In this sort of operation, remember, artillery preparation and support could be decisive; the lack of them would be an important element both in the failure of the coming offensive and in the episode of Celtic Wood.
The skies cleared on the morning of October 8, and what was described hope fully as a “drying” wind blew in from the west. Half a world away from home, young men like Lieutenant Frank Scott of Gawler, South Australia; Private John Murphy of Broken Hill, New South Wales; and Private Glen Bates of Hog Bay, Kangaroo Island, South Australia, spent the time helping engineers extend duckwalks and lay signal cables–or perhaps just waiting in slimy reserve trenches on the back side of the ridge. They belonged to Company C, 10th Battalion, 3rd Brigade of the Australian 1st Division and most were going into action for the first time. They were also spending the last full day of their lives.
At four in the afternoon, the rain returned–in torrents. Haig’s meteorological experts could promise no real break in the bad weather. It was at this point that the field marshal made what Bean has called “the most questioned decision of his career.” There was still time to call off the attack. He refused to do so.
An inky darkness came with the tempest. Men prepared to make the final trek to the jump-off points. A journey that normally required little more than an hour would in cases take 12, and whole units would arrive late for the attack. Exhausted, the men would slog forward anyway, with perhaps only a pause for a cigarette.
At 6:00 p.m. the commander of the 10th Battalion, Lieutenant Colonel Michael Wilder-Neligan, sent runners to summon his company commanders. Wilder-Neligan, a much-respected officer, had instructions to launch a new raid on Celtic Wood, and it was set to go with the general attack at 5:20 the next morning. Although the rest of the Australian 1st Division would not take part, the raid (it was hoped) would fool the enemy into holding back reserves and spreading his artillery fire, instead of concentrating it on the main attack. Company C, led by Lieutenant Scott, would make the raid, while two other companies would provide support. Wilder-Neligan asked for four raiding parties, each composed of two officers and 25 other ranks; but sickness (mainly trench foot), exhaustion, and the ordinary run of casualties had taken their toll. Company C could muster only five officers and 80 men, and it would make the raid as a single party.
Where, or from whom higher up in the command chain, did the order for a raid on Celtic Wood originate? Was there argument that evening at Wilder Neligan’s battalion headquarters dugout about the need for it? How many honestly believed the Germans would regard a raid by 85 men as a significant demonstration? And what about the battalion commander’s decision to use men who had mostly seen no action before? We’ll probably never know.
Sometime during the hours before dawn, Scott led the men of Company C over the crest of Broodseinde Ridge to the positions they would occupy in front of Celtic Wood. In the one photograph we have of Scott, his face, at twenty three, is still a boy’s face. How much, by that night, had it begun to age? The rain had ended but the going remained treacherous. A chilling wind took over. Perhaps Company C could take more comfort in the knowledge that when the raid was done-in the “Zero plus 30 minutes” allotted to it–the survivors could expect a hearty breakfast.
As Spagnoly reconstructs the raid, calamity loomed from the beginning. Once again the supporting barrage hardly seemed to exist and was lost in the din of the main bombardment to the north. Even that was no great shakes. It had proved impossible to manhandle enough guns forward in time. The Germans, said the same report, “seemed quite unconcerned as the raiders drew near.” In the two days since its midnight humiliation, the 448th Regiment had been reinforced–and it may have outnumbered the raiders of Company C by more than two to one. The wonder was that the Australians penetrated as deep as they did. Although much of the struggle was a confusion of gun flashes in the semi darkness, at least one survivor remembered seeing Scott urge his men forward; he succeeded in working behind the foremost defenders, almost reaching the other side of the wood.
Once the Germans recognized that this was an isolated attack, and that there was no danger of being overrun from the flanks, resistance stiffened. Now machine gunners began to search out the raiders, and entrenched riflemen fired at anything that moved. One of the few who made it back said that the wood was “teeming with snipers and sharpshooters.” The Australians managed to hurl bombs down some dugouts, but defenders emerged from others that had been overlooked and caught the raiders from behind. There was hand-to-hand fighting. The main raiding party fragmented into small groups, and the German infantry began to hunt them down. Scott was hit and killed–a survivor verified that. Now a German counter-barrage came down, cutting off the escape route across no-man’s-land. Zero plus 30 minutes came and went. As the sky grew light, a silence fell over the wood.
A handful of raiders did eventually straggle back, most of them having faked death amid the real dead as they awaited the cover of darkness. The next day, stretcher-bearers searching for wounded tried to enter Celtic Wood. Although they carried improvised red-cross flags, they were shot at and forced to retire.
Under the circumstances, the best that could be hoped for was that the missing men of Company C were prisoners. But as months passed, none of their names turned up on prisoner lists delivered to the International Red Cross. Yet when graves registration units worked through Celtic Wood a year later, after the Germans had retreated for good, they found neither graves nor scatterings of human remains. The records of the 448th Infantry Regiment contain no mention of the raid. For all anyone knew, two-thirds of Company C had simply vanished.
This brings us to the question of casualties, which is at the heart of the Celtic Wood story. Spagnoly’s patient research has yielded what seems a definitive breakdown. Twenty-six, by his accounting, were not among the missing. These include the 14 who returned unscathed, those known to be wounded (five died soon after), and a final five whose bodies were found, and positively identified, after the war. (Two privates in Company C, Gerald and Willie Campion , were brothers; one died of wounds, and the other never returned.) What of the remaining 59 members of the raiding party? Only 22 names, including that of Lieutenant Scott, appear on the walls of the Menin Gate Memorial to the Missing in Ypres. That leaves us with 37 men who are not merely presumed to be missing but who have not so much as a name.
The Great War was long over before historians even recognized the mystery of Celtic Wood. The reality of the Western Front missing was too overwhelming. On 27 British memorial walls in Belgium and France are engraved the names of 314,234 missing men. At the Menin Gate, after chiseling in 54,360 names, the war graves people ran out of space. Close to half the British and empire dead of the Western Front have no known grave. And if Celtic Wood is a fair indication, the true number may be higher. (By way of comparison, the number of U.S. MIAs from the Vietnam War is 2,273–a total that does not begin to defy political chutzpah.)
What, then, did happen to the missing 59 men of Company C? We can dispense with any and all hints of supernatural intervention . Nor do explanations we might dub “Western Front romantic” ring true. Fifty-nine bodies could not have sunk so deeply into the mud as to stay undetected for 75 years: Bones, like stones, tend to work their way to the surface–as the area’s farmers can testify. Besides, though the mud of Broodseinde Ridge unpleasantly hindered movement, it was nothing like the famously deep bouillabaisse on the plain below. The notion that so many bodies could be pulverized by artillery is also absurd, especially considering the difficulty the Australian guns had in mustering support fire.
John Laffin, the Australian military historian who has done significant work in battlefield archaeology on the Western Front, has spent much time roaming the fields where Celtic Wood once stood and has questioned the farmers who live there. “Even the oldest,” he says, “have no memory of human remains or military equipment having been dug up.” Elsewhere, bones from the Great War are still uncovered. In 1981, near Ovillers on the Somme, men digging a silo hole came upon 51 skeletons near the site of a 1916 casualty clearing station.
Since no remains have been found at Celtic Wood, it stands to reason that they must be somewhere else. One bit of negative evidence is conspicuous: the failure of the 448th Regiment to acknowledge the raid, especially since it must have resulted in German casualties rather beyond the normal daily wastage of trench warfare. Too, the Germans must have taken prisoners, a fact that unit diaries invariably note. If so, the Germans would have marched the captured remnants of Company C to the rear-but, as we know , the men never reached prisoner pens. Spagnoly hints that the prisoners may have been killed on the way. Laffin goes beyond hinting. Although he concedes that no proof has turned up so far, he has “no doubt that the men who survived the raid were massacred.” Were the victims then thrown into an unmarked grave? A few shell holes or an unoccupied deep dugout would have sufficed. Since we know that the fighting was confined to Celtic Wood, the farther from it that bodies may eventually be found, the greater the presumption of foul play.
Years ago, in London, I had a landlord who had been an officer on the Somme, and a decorated one. British troops, he told me, adhered to an unwritten rule about taking prisoners. If someone put up his hands without a struggle, he would be spared. If someone resisted and then tried to surrender, he would probably be killed. But the Germans (in that war at least) tended to treat all who surrendered with equal restraint. They were more likely to shoot down their own men whom they saw giving themselves up. But Celtic Wood may have been the exception–which might explain the suspicious (and perhaps guilty) silence in the records of the 448th.
This would seem to be the place to take a look at the character of the 233rd Division, of which the 448th was a part. The 233rd was a new division, having been formed only in January 1917. Many of those who filled it were 17-year-olds, barely of conscription age, and they were thrown into the line with a minimum of training. The division’s first extended experience of combat on the Ypres front that summer had been traumatic. An official report was hardly complimentary: “These men are too young to be able to furnish prolonged resistance and to have great endurance in a critical situation.” The division “is not suited for trench warfare.” After the unusually heavy losses it suffered, mostly from British artillery in the weeks leading up to Haig’s offensive, the 233rd was withdrawn and sent to a quiet sector of France to recuperate. There it distinguished itself with a spree of pillaging in the town of Saint Quentin. Discipline in the German army, customarily so rigid, was beginning to slacken dangerously.
The 448th, as we know, had suffered a humiliation on October 7–and the losses of the second Celtic Wood raid, given the nature of the fighting, must have been gallingly high. Did something snap? The sniping at the stretcher-bearers is further evidence of a breakdown in discipline. Maybe, too, the 448th felt it had something to hide. Veterans on either side respect one another; it is newer troops, like the frightened young men of My Lai, who are apt to commit atrocities. The 448th Regiment, and its division, fits the profile.
Every time I return to Ypres, I find that a little more of the old salient has succumbed to a creeping barrage of new building. Americans know that pattern of suburban sprawl too well. Is it unreasonable to think that someday, before too long, a contractor excavating a house foundation may come on a hideous discovery–but one that will at last solve the mystery of Celtic Wood?
ROBERT COWLEY was a former editor of MHQ.
This article originally appeared in the Summer 1992 issue (Vol. 4, No. 4) of MHQ—The Quarterly Journal of Military History with the headline: What Happened at Celtic Woods?
Want to have the lavishly illustrated, premium-quality print edition of MHQ delivered directly to you four times a year? Subscribe now at special savings!