Introduction

This is the second story in an examination of how the Battle of Crete was shaped, fought and ultimately lost.

The first story, Crete Was Lost Before Maleme, stepped back from the fighting itself and asked a broader question: whether the conditions that made defeat possible had already been created before the first German parachute opened on 20 May 1941. It examined the decision to defend Crete, the consequences of the campaign in mainland Greece, shortages of equipment and transport, the limitations of Allied air power, intelligence, communications and the difficulty of defending several widely separated objectives without a strong mobile reserve.

This second story moves from preparation to battle. The strategic decisions have been made, Freyberg’s forces are in position, and at about eight o’clock on the morning of 20 May the Luftwaffe appears over western Crete. From that point there is no longer time to improve the island’s defences or correct the decisions of the preceding months. The question becomes much more immediate: could the defenders stop Operation Merkur with the forces they actually had?

What happened during the following twelve hours is sometimes obscured by our knowledge of the eventual German victory. The image of thousands of parachutists descending over Crete, followed by the capture of Maleme and the arrival of German mountain troops, can make the outcome appear inevitable. It was not. Across the island German airborne formations suffered devastating losses, several of their principal objectives remained in Allied hands, and by the evening of 20 May the invasion was in serious difficulty.

Maleme was the exception, but even there the Germans had not yet secured the victory they needed. West of the airfield, surviving airborne troops had established themselves around the Tavronitis and were beginning to consolidate, while on Hill 107 Lieutenant-Colonel Leslie Andrew and 22nd New Zealand Battalion were trying to understand a battlefield increasingly obscured by broken communications, incomplete reports and German air superiority.

The first story asked how the Allies arrived at Maleme in such a vulnerable position. This one asks what happened when the Germans finally came—and how an invasion that came dangerously close to failure during its first day survived long enough to become a victory.

That takes us back to the morning of 20 May 1941.

At about eight o’clock on the morning of 20 May 1941, the argument ended.

For weeks the defenders of Crete had watched German reconnaissance aircraft pass overhead and endured increasingly heavy bombing. They knew an invasion was coming. Major-General Bernard Freyberg had received intelligence indicating both airborne attack and reinforcement by sea, units had been placed around the principal airfields, and men had spent days digging positions into the hard Cretan ground.

Then the sky over western Crete filled with aircraft.

The bombardment around Maleme intensified shortly after eight. Dust and smoke obscured parts of the battlefield as fighters and bombers attacked defensive positions around the airfield and Hill 107. Gliders appeared low over the ground and Junkers Ju 52 transport aircraft followed, releasing German paratroopers across the western end of the island.

Operation Merkur had begun.

We know how it ended, and that knowledge distorts how we see its beginning. Germany captured Crete, the British and Commonwealth forces eventually withdrew across the White Mountains, and the island remained under German occupation until 1945. From that outcome it is easy to imagine the invasion of 20 May as an overwhelming demonstration of German airborne power.

It was nothing of the sort.

By the evening of the first day, the German operation was in serious difficulty. At Rethymno and Heraklion the airfields remained in Allied hands. Around Chania the attackers had failed to achieve their principal objectives. German casualties were severe, formations were scattered and commanders in Athens had reason to fear that the invasion might fail. The New Zealand official account describes the German foothold at the end of the day as extremely tenuous.

The battle had not yet been won by Germany.

It had very nearly been lost.

Before Breakfast

The morning had begun almost routinely for the New Zealanders around Maleme.

Air raids were no longer unusual. German aircraft had been attacking the airfield repeatedly during the preceding days, and at first the activity on 20 May appeared to be another bombardment. The war diary of the 28th (Māori) Battalion records air attacks beginning at 0620. By 0800 the area around Maleme was being bombed and machine-gunned so intensely that clouds of dust obscured visibility. At approximately 0830, in the middle of that confusion, the parachutists began to descend.

The scale of what followed must have been extraordinary.

Transport aircraft crossed the coast in formation. Gliders detached from their towing aircraft and came down at low level. Parachutes opened behind them in large numbers, some almost directly above prepared Allied positions.

For airborne troops, those first seconds on the ground were particularly dangerous. A parachutist descending beneath his canopy could do little to protect himself, and once down he had to release his harness and recover his weapons. German airborne doctrine meant that many heavier weapons were dropped separately in containers, forcing men to locate them before they could fight effectively.

Where German troops landed among prepared defenders, the consequences were devastating.

New Zealand rifles and machine guns opened fire on men still descending or attempting to organise themselves after landing. German units that had looked formidable on an operations map became scattered groups of individuals trying to find their comrades, commanders and equipment.

Around Maleme, however, geography produced very different results within a remarkably small area.

Tavronitis

This is where riding the battlefield becomes particularly useful. Stand on Hill 107 and look west and the Tavronitis appears immediately below. The riverbed is broad and, for much of the year, largely dry. Beyond it the ground stretches towards Tavronitis village and the coast.

On a map, it looks exposed. On the ground, it is more complicated.

The riverbed, banks, vegetation and folds in the surrounding terrain provide dead ground and concealment. German planners had identified the western side of Maleme as critical to the assault, and gliders carrying elements of Generalmajor Eugen Meindl’s Luftlande-Sturm-Regiment came down in and around the Tavronitis.

About fifty gliders landed around Maleme, many along the riverbed. The experience of the airborne troops depended heavily upon where they came down. Those dropping south and east of the airfield landed among New Zealand positions and suffered severely. West of Maleme the situation was different: gliders and parachutists came down in ground less effectively covered by the defenders, giving survivors an opportunity to assemble and begin moving against the airfield. That difference would prove crucial.

When we rode from Hill 107 towards the Tavronitis, the relationship between the two features became much clearer than it appears on a small-scale map. The defenders possessed commanding high ground, but commanding ground does not mean that every metre beneath it can be observed or covered by fire.

The Germans had found space in which to survive.

For the moment, survival was enough.

22 Battalion

The immediate defence of Maleme Airfield and Hill 107 rested principally with Lieutenant-Colonel Leslie Andrew VC and 22nd New Zealand Battalion, supported by machine-gunners and artillery. Elsewhere within Brigadier James Hargest’s 5th New Zealand Brigade, 23rd Battalion lay east of Maleme, the under-strength 21st Battalion farther south-east, and the 28th (Māori) Battalion around Platanias.

The arrangement looks reasonably coherent when reproduced on a map.

Once the attack began, it became much less so.

German bombing and ground fighting disrupted communications. Telephone lines were cut. Wireless communications were limited. Company commanders saw only the battle immediately around them, while battalion and brigade headquarters attempted to build a picture from incomplete reports.

That is particularly important in the case of 22 Battalion.

Andrew’s forward companies west and north of Hill 107 became heavily engaged, but their ability to communicate with battalion headquarters deteriorated. Men were fighting and, in places, fighting successfully, while commanders elsewhere could not be certain that they were still there.

The battlefield was beginning to divide into separate realities.

Killing the Parachutists

East of Maleme, the German experience was often disastrous.Parachutists landed among 23rd Battalion and the New Zealand engineers. The official New Zealand account records between 100 and 150 descending near the western flank of 7th Field Company; many were killed while landing or were subsequently hunted down. At the Field Punishment Centre south of the engineers, German troops again dropped directly into defended ground and suffered heavily.

The Māori Battalion’s war diary gives an unusually immediate impression of the confusion. German equipment containers fell into New Zealand positions, sometimes providing weapons and ammunition to men who had previously been short of them. Patrols were sent out against isolated parachutists, while groups of Māori soldiers and headquarters personnel moved through creek beds and broken ground clearing German pockets. By early evening the battalion believed most of the parachutists dropped into one of its company areas had been destroyed.

Cretan civilians also entered the fighting. Some attacked German soldiers with whatever weapons were available, an involvement for which communities would later pay terribly during the occupation.

None of this resembled a German force sweeping irresistibly across Crete. In many places, the invasion was being broken apart as it landed.

The Germans West of Maleme

West of the airfield, however, the surviving German troops were beginning to organise. This was the part of the battlefield that mattered most. Men who had survived the landing found weapons containers, assembled into groups and began working towards their objectives. Officers gathered whatever troops they could find, sometimes irrespective of their original unit. The German airborne system had been designed with this sort of confusion in mind, but the losses among officers and men still made the process extraordinarily difficult.

The airfield itself remained under Allied fire. German aircraft circling overhead could not simply land reinforcements onto a secure runway, because no such secure runway yet existed.

Lieutenant Haddon Donald of 22 Battalion later recalled German aircraft circling the airfield without landing because conditions remained unsafe. Only later did aircraft begin attempting to come in, while New Zealand fire continued against the landing ground. That is worth remembering when looking down on Maleme today. During much of 20 May the Germans did not possess the airfield in any meaningful operational sense.

They were fighting for the opportunity to use it.

Andrew’s Problem

By afternoon Lieutenant-Colonel Andrew knew his position was becoming dangerous, but he did not know precisely how dangerous. This distinction is important.

From our position eighty-five years later, we can consult New Zealand and German records simultaneously. We can see where the German troops landed, which units suffered catastrophic casualties, which New Zealand companies remained intact and where German groups were assembling.

Andrew could see none of that.

He knew that German troops were pressing around the western side of the airfield. He had lost effective communication with some of his forward companies and feared they had been overwhelmed. His battalion reserve was small, while the Luftwaffe continued to dominate the battlefield above him.

He requested assistance from 23rd Battalion to the east.

Brigadier Hargest refused the request because he believed 23rd Battalion was itself heavily engaged. In reality, the battalion had spent much of the day dealing with scattered German troops and was in considerably better condition than Hargest appreciated.

Here, once again, the problem was not simply lack of troops. It was lack of information. A potentially useful force existed relatively nearby, but the brigade commander did not possess an accurate enough picture of the battlefield to release it.

Two Tanks

Andrew still attempted to restore the situation himself.

His reserve amounted to an infantry platoon and two tanks. The tanks have acquired almost symbolic importance in accounts of Maleme because they demonstrate the fragility of the resources available at the decisive point.

They were sent towards the western side of the airfield with infantry support.

The attempt failed.

Mechanical problems and the difficult conditions brought the tanks’ contribution to an end, and the small counter-attack could not drive the Germans back. There was no substantial armoured reserve waiting behind Hill 107, no fresh brigade ready to move through and clear the Tavronitis, and no Allied aircraft overhead able to attack German assembly areas.

The defenders were fighting with what they had. That had been sufficient to inflict enormous casualties. Whether it was sufficient to finish the job was another question.

Elsewhere, the Invasion Was Failing

The importance of Maleme becomes clearer if we leave it briefly.

The second major wave of German airborne troops attacked Rethymno and Heraklion later on 20 May. Neither airfield was captured. German units suffered severe losses and were left isolated around their landing areas.

Around Chania and the Prison Valley the position was also confused. German forces established footholds, but their principal objectives had not been achieved. Proposals for Allied counter-action were considered, but commanders remained concerned about further airborne attacks and the expected threat from the sea.

By the evening, none of Crete’s principal airfields was securely in German hands. This was not what Operation Merkur required.

The German plan depended upon capturing usable airfields quickly enough to bring in follow-on troops. Luftwaffe planning explicitly treated seizure of Maleme, Rethymno and Heraklion as necessary to the subsequent air-landed phase.

Without an airfield, the Germans had a serious problem.

Parachutists could be supplied from the air for a time, but an airborne force could not indefinitely sustain a major ground campaign using only what had descended beneath parachutes and in containers. Somewhere, an airfield had to be opened.

Athens

The view from German headquarters was therefore very different from the confident picture later created by victory.

Reports arriving from Crete were fragmentary and disturbing. Casualties had been enormous, commanders were dead or wounded, and the expected rapid capture of the airfields had not occurred.

The German command had to decide what to do with its remaining reserves.

Disperse them in an attempt to rescue several failing attacks, or concentrate everything upon the one place where some progress had been made? General Kurt Student chose Maleme.

That decision is one of the most important of the battle. On 21 May the remaining German effort would be concentrated in the western sector. If Maleme could be secured, mountain troops could be flown directly onto Crete. If it could not, the invasion faced the possibility of failure. The New Zealand historical account records that German commanders in Athens feared precisely that outcome at the end of 20 May. The Germans therefore ended the first day with a plan.

What they did not yet realise was that the New Zealand defence above Maleme was beginning to move.

Hill 107 After Dark

This is where the first day becomes inseparable from the night that followed.

Andrew remained unable to communicate reliably with his forward companies. He believed parts of his battalion had been destroyed or were in danger of being cut off. His attempt to improve the position with the tanks had failed, and assistance from 23rd Battalion had not arrived.He spoke to Hargest.The brigade commander did not order him to remain on Hill 107 regardless of circumstances. Nor did Hargest order an immediate brigade counter-attack to clear the western side of Maleme. When Andrew indicated that he might have to withdraw, Hargest gave the response that has followed both men through the historiography of the battle: if Andrew considered withdrawal necessary, the decision was his.

The subsequent controversy is understandable, but hindsight makes it dangerously easy to convert uncertainty into incompetence.

Andrew did not know that German commanders were themselves uncertain about the outcome of the invasion. He did not know that other German attacks across Crete had largely failed. Most importantly, he did not know that some of his own forward companies were still fighting.

That last failure of communication would have enormous consequences.

Two Battles on the Same Hill

There were now effectively two versions of the Battle of Maleme.

In one, known to Andrew, his battalion had been badly hit, communications had collapsed, German forces threatened the airfield, his small counter-attack had failed and forward companies might have been destroyed.

In the other, which we can reconstruct afterwards, those forward companies had survived, German airborne troops had suffered devastating losses and the enemy’s entire invasion plan remained precarious.

Both versions existed at the same time. Only one was available to the commander making the decision.

That is why reducing the loss of Maleme to one man’s failure has never seemed satisfactory to me. Andrew made a decision whose consequences proved disastrous, and it is entirely legitimate to examine whether he should have acted differently. Hargest’s decisions and the wider brigade command arrangements require the same scrutiny. But the battlefield asks a larger question. Why did the command system fail to tell them what was actually happening?

The Ground Between Them

When I ride from Hill 107 towards the Tavronitis and then back across the battlefield towards Xamoudochori, that question becomes physical. The distances are not enormous. That is almost the point. Units could be relatively close to one another while being separated by folds in the ground, vegetation, riverbeds and broken communications. A commander on one feature might hear intense firing from another without knowing who was winning. A company could continue fighting after battalion headquarters had effectively lost contact with it.

Today I can stop the bike, look at a GPS screen, check my precise location and know how far I am from the next position. In May 1941 there was no such luxury. Information travelled by field telephone until the wire was cut, by wireless where sets were available and functioning, by runner, motorcycle or officer physically moving between headquarters. Add bombing, dust, smoke, casualties and German troops moving through the same ground, and the battlefield begins to fragment.

The map shows one battle.

The men on the ground experienced dozens of them.

What Nearly Happened

By nightfall on 20 May, the Germans had paid an extraordinary price and had failed to achieve most of their immediate objectives. Rethymno had not fallen. Heraklion had not fallen. Suda Bay had not fallen. Maleme Airfield was still contested and remained under Allied fire.

German airborne troops were scattered across several areas, and the force west of Maleme had survived rather than achieved the clean seizure upon which the operation had depended. The German command had reached the point where it intended to concentrate its remaining resources upon the only sector offering a realistic prospect of success.

Had the New Zealand position around Maleme remained intact through the night, the Germans would have faced a profoundly difficult situation on the morning of 21 May.

That does not mean German defeat was certain. The Luftwaffe still controlled the sky, German troops remained on the island, and the Allies had serious weaknesses of their own. But Operation Merkur was in danger. The Germans knew it.

The Battle That Nearly Failed

This is the part of the Battle of Crete that subsequent events tend to conceal.

German victory makes the invasion look stronger than it was.

The photographs of parachutes filling the Cretan sky suggest overwhelming power. The later arrival of mountain troops at Maleme reinforces the impression.

Knowing that the British eventually retreated to Sfakia encourages us to read inevitability backwards into the first morning.

The evidence does not support that interpretation.

On 20 May the defenders inflicted enormous damage upon the German airborne force. In several sectors they did exactly what an anti-airborne defence was supposed to do: they engaged the enemy while he was dispersed, killed men before they could organise and prevented the immediate capture of the airfields.

The German plan survived because enough men survived west of Maleme to create a foothold.

Then Allied command and communications failures prevented the defenders from recognising just how vulnerable that foothold remained. As darkness fell, both sides were therefore closer to defeat than either fully understood. The Germans believed their invasion might be failing. The New Zealand commanders around Maleme believed their own position might be collapsing.

The tragedy of the next few hours was that the Germans were wrong first.

The Next Night

Standing on Hill 107 today, the morning of 20 May is relatively easy to reconstruct. The Tavronitis Bridge is still there, the airfield remains immediately below, and much of the surrounding terrain can still be read against the wartime maps.

The night is harder.

There are no parachutes to follow and no great formations of aircraft overhead.

Instead there are telephone calls, missing reports, assumptions, requests for reinforcements, two failed tanks, companies fighting without contact with their battalion headquarters and commanders trying to make decisions from fragments of information.

Yet those quiet hours after the spectacular airborne assault were arguably more important than the assault itself.

The German invasion had survived the day.

It had not yet won the battle.

What happened on Hill 107 during the night of 20–21 May would give it the opportunity.

That is the next story.

Sources and Further Reading

Dan Davin, Crete, Official History of New Zealand in the Second World War 1939–45 (War History Branch, Department of Internal Affairs, Wellington, 1953).

Christopher Buckley, Greece and Crete 1941 (HMSO, 1952).

I. S. O. Playfair et al., The Mediterranean and Middle East, Volume II: The Germans Come to the Help of Their Ally (1941) (HMSO, 1956).

Callum MacDonald, The Lost Battle: Crete 1941 (Macmillan, 1993).

Antony Beevor, Crete: The Battle and the Resistance (John Murray).

Manatū Taonga — New Zealand Ministry for Culture and Heritage, The Battle for Crete, including the Maleme battlefield maps, campaign narrative and oral histories of New Zealand participants.

28th Māori Battalion, May 1941 war diary and associated veterans’ accounts.

Air Historical Branch, The Campaign in Crete, May 1941.

Air Historical Branch, Report on RAF Operations in Crete, 17 April–31 May 1941.

Air Historical Branch, translated German operational material relating to the invasion of Crete.

Hellenic Army General Staff, Battle of Crete histories and battlefield maps.