Chapter 8: Preparing for PUGILIST
The Terrain
SOUTH and south-east of Tripoli an escarpment, almost a range of hills, trends from the coast near Homs in a long crescent that swings in a half-circle to the west and ends just south of Gabes. This escarpment forms the southern limit of a coastal plain, some 80 miles at its widest, between Nalut and the sea. From Foum Tatahouine northwards the range is known as Monts des Ksour, popularly called the Matmata Hills, the north-western end terminating at Djebel Melab. Between the Matmata Hills and the sea the north-western end of the coastal plain steadily narrows, and in the hills and across the plain ran the Mareth Line from Toujane to the sea.
The Matmata Hills run generally north and south. Behind them to the west a stretch of desert country, known as the Dahar, runs parallel to the hills from near Nalut, with the village of Ksar Rhilane towards the northern end. Farther west lies a stretch of impassable sand desert, the Grand Erg Oriental. The northern end of the Dahar merges into a long series of salt marshes known as Chotts, the most easterly of which is entitled Chott Djerid, with an eastern extension named Chott el Fedjadj. South of and parallel to Chott el Fedjadj is a long ridge lying east and west – the Djebel Tebaga – which continues northwards as a low watershed separating the Chotts from the coastal strip. This strip, about 15 miles from Gabes, was the Gabes Gap, better known to Eighth Army as the Wadi Akarit position, from the wadi flowing north-eastwards between the watershed and the sea.
Between Djebel Tebaga and Djebel Melab a low pass runs from the Dahar to the coastal plain south-west of Gabes. This is the Tebaga Gap.1 It is about four miles long in its narrowest section and is passable for infantry and tanks over a width of about three and a half miles between the two djebels, although wheeled vehicles, generally speaking, must use the tracks. A force entering the Dahar from the coastal plain farther south could find its way back to the coast through this pass.
The Outflanking Problem
The problem now confronting Eighth Army was to force the Mareth Line and advance on Sfax, the next useable port, and the next area with useful landing grounds. The line had been built by the French in pre-war years to meet possible Italian threats from Tripolitania. In the coastal plain, here from ten to 15 miles wide, it ran immediately behind the line of Wadi Zigzaou from Zarat to Touati, the wadi being a natural anti-tank obstacle with sheer banks reaching in places a height of seventy feet. This stretch was moreover covered along the whole length by concrete and steel pillboxes and emplacements; and these defences, which were in existence before the war, had in recent months been strengthened by anti-tank ditches, wire and minefields. From Touati the line swung south-west to a point just south of Toujane, and then north-west through the Matmata Hills towards Djebel Melab. In the hills the nature of the country was relied on for defence, and there were few artificial aids.
In pre-war days it was considered that the Mareth Line could not be outflanked, because the Dahar was thought to be impassable for mechanical transport. In 1938 a French lorried force carried out an exercise to determine whether such an operation was possible, and came to the conclusion that it was not. But since 1938 the motor vehicle had improved enormously, particularly in the introduction of four-wheel drive; and most of the MT in Eighth Army was of this type. Tracked vehicles would also make light of difficulties that had stopped the pre-war lorry. Moreover, Eighth Army was by now expert in desert movement, so that there was every justification for the belief at Army Headquarters that the Dahar was passable and an outflanking move a possibility.
Rommel was never in any doubt about this, and on 10 February in an appreciation prepared for the Fuehrer, indicated clearly that it would be difficult, if not impossible, to hold off both a frontal attack and a large-scale outflanking attack via Tebaga Gap, particularly if these were combined with a southwards thrust from Gafsa. Hence he much preferred a position in the Gabes Gap, where the flanks rested on either the impassable Chotts or the sea. He had hammered away at this ever since the retreat from Alamein, but without avail, and had to be content with getting his troops back into the Mareth Line. He was no longer in command when the line was attacked, but all turned out as he had foretold, and the position in the Gabes Gap was taken up too late and was stormed by the Eighth Army within a week of making contact.
More than this, Rommel regarded the Gabes Gap position merely as a better alternative than Mareth for the defence of southern Tunisia. He had insisted until as late as 3 March, when the proposal was rejected in both Rome and Berlin, that the only chance the Axis had to retain a front in Africa was to offer limited delays at both Mareth and Gabes Gap, and to concentrate for protracted defence in the Enfidaville line. From this area the best troops, at least, could be evacuated to Europe. When he visited Hitler on 10 March he renewed his argument without success, but managed to convince Hitler that the Gabes Gap was a sounder defensive position than Mareth. Kesselring was consequently ordered to move the Spezia and Pistoia Divisions to the Gabes Gap for work there, to relieve 164 Light Division about Matmata by Centauro Division, the former to stiffen the Italian units in the Mareth Line, and to employ Trieste Division to watch for movement from Gafsa. But Kesselring did not agree! The Comando Supremo was not informed, so that although Kesselring had passed the orders on to von Arnim, who tried to implement them, Messe refused to comply without instructions from Rome. He believed that this was just a back-door method of forcing the Axis troops back to Enfidaville. However, under pressure, Messe agreed on 14 March to release Spezia and Pistoia Divisions, but by the 16th Kesselring had prevailed upon Hitler to change his mind and these divisions were ordered back to their former positions in the Mareth Line. It was a process that could not fail to aggravate the tensions that already existed between the Axis partners.
But within limits, Rommel did his best to make things difficult for Eighth Army. For as long as he could he defended the passes into the Dahar from the area round Foum Tatahouine, but the Axis forces were scanty and in the outcome were easily driven away by 4 Light Armoured Brigade. From there they went to swell the forces holding the Tebaga Gap, where some attempt had been made to prepare defences. The work had started some time before, indeed as early as 1941, but from air photographs the defences seemed to be patchy, and to include only short lengths of anti-tank ditch, a few weapon pits and emplacements, and some stretches of wire. The line was not continuous, had no depth and could not impose more than some slight delay on an attacker.
Steps were taken also to defend other crossings of the Matmata Hills from the Dahar back to the coast via the road through Ksar el Hallouf, and via the road through Tamezred and Matmata. Nervousness about these roads persisted for some days after the battle had started, with some reason, for consideration had been given to using the French to open these passes from the west, and eventually 4 Indian Division was directed on Matmata from the east.
The outflanking operation had been in Montgomery’s mind for some months, but before making final plans it was necessary to discover a practicable route into the Dahar at a point well clear of the Mareth Line, and whether in fact the Dahar was passable. While at Marble Arch in December 1942, the LRDG was instructed to reconnoitre the area early in January, and T1 Patrol, under Captain Wilder ,2 crossed into Tunisia south of Nalut on 12 January, the first troops of Eighth Army to do so. About 30 miles south-west of Foum Tatahouine they found a pass through the hills into the Dahar, later known as Wilder’s Gap. Wilder’s reconnaissance did not penetrate very far, but a later reconnaissance under Lieutenant Tinker3 went north on 27 January to Djebel Tebaga and examined the Tebaga Gap, confirming that the going throughout was suitable for a force of all arms.4 During his reconnaissance the LRDG base camp at Ksar Rhilane was shot up by enemy aircraft and considerable damage was done. This action was the first of a number which showed the enemy’s nervousness about operations over this route.
But the last doubts had been dispersed, and it was now known that the route was a practicable one, and that any force in the Dahar could be supplied from Medenine by way of Wilder’s Gap. Eighth Army had in 2 NZ Division a formation already well trained in long desert moves.
Montgomery had moved Leclerc’s Force (now known as ‘L’ Force) forward from Nalut to Ksar Rhilane, where it operated during the Battle of Medenine.5 Here on 10 March it was suddenly attacked by an enemy force of armoured cars, artillery and aircraft. ‘L’ Force stood firm, and helped by the Desert Air Force drove off the attackers and inflicted severe losses on them. It was a spirited performance and ensured protection to the Dahar south of Ksar Rhilane, but showed again the enemy’s sensitivity.
2 NZ Division Prepares
The Army plan, issued on 26 February, prescribed that an enlarged Division, entitled New Zealand Corps, would make a turning movement via Nalut and Ksar Rhilane. Since then, however, 2 NZ Division had moved forward to Medenine and on 10 March was still there.6 The route via Nalut had been prescribed on the assumption that the Division would start from Tripoli and would move along the southern edge of the coastal plain. There was obviously no point now in going back as far as Nalut; but some rearward movement was necessary, for the direct road from Medenine to Foum Tatahouine was likely to be open to enemy ground observation. In addition there was always the possibility that rearward movement would mislead the enemy. It was decided therefore that NZ Corps should go back to Ben Gardane in daylight, and move at night on the road from Ben Gardane to Foum Tatahouine.
On 5 March, the day before the enemy attack at Medenine, a party from 6 Field Company reconnoitred this route and from Foum Tatahouine went south for 30 miles as far as, and indeed beyond, the turn-off to Wilder’s Gap. All the roads were found to be suitable for all types of traffic, although needing some repairs. On 8 March detachments from 5 Field Park Company and 6 Field Company began clearing and improving the road from Ben Gardane to Foum Tatahouine and on to Wilder’s Gap, and marking a track through the Gap into an assembly area in the Dahar, ten miles north-west of the Gap and 35 miles south-west of Foum Tatahouine. The track was, as usual, marked with the black diamond sign.
The engineers at this time also made a plaster model of the Tebaga Gap area, used by the GOC at many of his conferences, and later by brigade commanders. There were as usual varied opinions about the usefulness of the model, but it appears that it was genuinely helpful during the planning stages, although in no way taking the place of ground reconnaissance.
On 10 March General Freyberg held a conference to discuss the move to the assembly area. At this conference he compared the strength of the future NZ Corps with any enemy forces that might be met, and at this stage he considered only the German Africa Corps7 which, disregarding any Italian forces, could be expected to oppose an outflanking movement. The 164th Light Division had been identified in the Mareth Line, where the joint operations of 10 and 30 Corps would probably retain it, and 10 Panzer Division was known to be north of Gabes.
The GOC concluded that NZ Corps would be stronger than Africa Corps in troops, about 20,000 (actually 25,600) to 19,300; in field and medium artillery, 112 to 55; in anti-tank guns, 172 to 120, and in tanks, 150 (excluding Divisional Cavalry and ‘L’ Force) to 70. Estimates of enemy strengths prepared after this conference, to 19 March, lifted the number of troops to 21,500, the field and medium artillery to 100 and the tank strength to 120. But whatever discrepancies between the estimated strength and the actual, for which no reliable figures are available, General Freyberg ‘s comment at his conference on the 10th, that the operation was not as rash as might appear when considering the map, seems valid enough.
During the days from 7 March onwards 2 NZ Division stocked up with supplies ready for a move which was to start about 11 March – six days’ rations and water, the latter at half a gallon per man per day, and petrol for 300 miles. As a security measure while in the sparsely populated coastal area, fernleaf signs were obliterated from all vehicles, and shoulder titles and hat badges removed. A press release was made that 2 NZ Division was holding part of the Mareth Line. Whether for these or for other reasons the Germans were slow in identifying the Division when operations resumed and reported attacks by formations that were nowhere in the area.
The 8th Armoured Brigade from 7 Armoured Division came under command on 10 March. It consisted of 3 Royal Tanks, Notts Yeomanry, and Staffordshire Yeomanry, all heavy tank regiments, and 1 Buffs, 111 Field Regiment, RA, and ancillary units. The number of tanks held varied a little from day to day, but just before the campaign started their tank state was as follows:
3 RTR Notts Yeo Staffs Yeo HQ Total
Shermans 25 23 28 76
Grants 4 4 3 2 13
Crusaders 22 19 19 2 62
Total 151
Armoured Cars 8 6 7 21
It was normal in the brigade to allocate companies of the Buffs to armoured regiments, so forming ‘Armoured Regimental Groups’.
‘A’ Company thus moved under command of Notts Yeo, ‘B’ Company under 3 RTR, and ‘C’ Company under Staffs Yeo. Only in a special case, where for instance some sector was to be held as a firm base, did 1 Buffs operate as a battalion.
On 11 March the additional platoons for Petrol Company, approved while the Division was at Bardia,8 duly joined up, so increasing the company from two to five platoons and increasing the reserves of petrol it was possible to have immediately available.
At midnight on 11–12 March 2 NZ Division passed from the command of 30 Corps to that of NZ Corps, the constitution of the latter being:
2nd NZ Division – Lieutenant-General Sir Bernard Freyberg
8th Armoured Brigade – Brigadier C. B. C. Harvey
King’s Dragoon Guards (armoured cars)
64th Medium Regiment, RA
57th Anti-Tank Regiment, RA, less one battery
One battery 53 Light Anti-Aircraft Regiment, RA
‘L’ Force – General Leclerc9
FFF Column
At this time ‘L’ Force consisted of the following:
Two troops armoured cars
One squadron self-propelled guns
Eleven other guns of various types
Anti-tank regiment
Anti-aircraft guns
Two reconnaissance companies
Five lorried companies
One Greek squadron
The Free French Flying Column consisted of:
Two armoured car squadrons
One tank company (11 Crusaders and 2 Shermans)
Two platoons infantry
The whole French force numbered about 3500 men and 900 vehicles, and was self-contained for long periods, longer even than 2 NZ Division. A proportion of its supporting artillery was British. The French force joined NZ Corps in situ, remaining in its existing area round Ksar Rhilane.
General Freyberg commanded both the Corps and the Division, and did not form a separate Corps Headquarters nor was there any corps echelon of administrative troops. Strictly speaking, therefore, the force was a much augmented division and not a true corps. On the tactical side, the absence of a separate Corps Headquarters was not queried before the battle, but at a conference held in Tripoli towards the end of February the administrative staff of the Eighth Army expressed some concern at difficulties that might arise owing to the absence of the supply echelon normally interposed between a division and the army roadhead.10 But the administrative staff of 2 NZ Division and the CRASC were confident that they could compete with the task and did not want additional staff. All that was wanted was additional RASC units, and these were duly provided. At no point during the operations was there any administrative restriction or delay.
The initial movement of the Corps, less the French Group, to the assembly area began at a starting point a few miles east of Medenine, and was to be via a staging area about halfway between Ben Gardane and Foum Tatahouine. Here dumps of petrol had been arranged so that units could replenish. Group movements were as follows:
To Staging Area To Assembly Area
6 Infantry Brigade Group 8 a.m., 11 March Night 11–12 March
5 Infantry Brigade Group 8 a.m., 12 March Night 12–13 March
Headquarters 11.30 a.m., 12 March Night 12–13 March
ASC Group Night 13–14 March
Artillery Group 8 a.m., 14 March Night 14–15 March
Reserve Group 10 a.m., 14 March Night 14–15 March
8 Armoured Brigade Group 2 p.m., 14 March Night 15–16 March
The distance from starting point to staging area was about 60 miles, and from there to assembly area about seventy. The move back to Ben Gardane was in daylight, as there was no objection to enemy aircraft spotting an apparent withdrawal; but for most groups the move thence to the staging area was also in daylight, probably on the grounds that this location could have served as an assembly area for troops moving into the Mareth Line. The move forward to the assembly area was by night. Administrative Group, which had been in the Ben Gardane area during the Medenine operations, was to move off at 11 a.m. on 17 March and would thus be last in the column.
Tracked vehicles of all groups were to be moved on transporters from the Medenine area, leaving on 15 March for the staging area, and moving on the night 16–17 March to an unloading point south of Foum Tatahouine. Vehicles would be unloaded before daylight and lie camouflaged during 17 March, moving after dark to rejoin their units in the assembly area.
Except in cases of operational necessity, wireless silence was to be observed, all wireless traffic for NZ Corps being routed as for Rear Headquarters ‘L’ Force near the assembly area, through a set manned by Royal Signals operators. After convoys left the staging area no enemy aircraft were to be engaged unless they made a direct attack. Once in the assembly area movement was to be at a speed that would not raise dust, and there were to be no fires or lights during darkness.
NZ Corps Moves Forward
The moves to the staging area passed off without incident. From there to the assembly area there were some delays owing to the opening and closing of columns in the darkness and without headlights; but again Groups reached their areas in reasonable time. Vehicles were at once dispersed facing north in an attempt to reduce shadow and prevent reflection of sunlight from windscreens, and camouflage nets were spread. Troops were dug in. But none of these precautions could prevent an enemy reconnaissance aircraft flying over the area in the evening of 12 March, when 6 Brigade Group was already there. It was at an estimated height of 10,000 feet, and evoked much speculation. German air reconnaissance in fact failed to detect the assembly area until the 16th, and not positively until the 18th.
The NZASC Group arrived in the assembly area early on 14 March, dumped its second-line holdings and, with all available transport, moved to an Army Roadhead at Dehibat, 50 miles southeast of Wilder’s Gap. Two RASC companies – one Petrol and one Ammunition – had already come under command to help in the formation of a Field Maintenance Centre11 at Bir Amir, just short of the Gap. On 17 March, NZASC was further augmented by three General Transport Companies and two Water Tank Companies from RASC, all to provide third-line transport for NZ Corps between the roadhead and the Field Maintenance Centre. Some water was found in a well at Bir Amir and was issued from 14 March onwards.
On 17 March Administrative Group arrived in the assembly area and the Corps’ concentration was completed, formations being disposed along the axis over a distance of some six miles, with Divisional Cavalry in front, and Administrative Group in the rear. The Corps spent the time resting and training, the latter including short route marches during the early hours of darkness. The perimeter of the area was patrolled, and in the interests of security all troops on patrol were searched for documents before commencing duty. No contact was made with the enemy, but Arabs were troublesome, nearly a hundred lamps disappearing from the Corps axis. Roving patrols had to be used to check this.
PUGILIST was explained to all officers on 14 March, and later to NCOs and the rank and file. Some very good air photographs of the enemy defences at Tebaga Gap were issued and were examined and discussed, particularly by the artillery and by 6 Brigade Group, which was to be in the lead.
The French Group was instructed on 14 March to maintain patrols north and north-east of Ksar Rhilane to prevent enemy ground observation of the Corps’ assembly. That there was some justification for this precaution was made evident on the night 15–16 March, when considerable movement of motor transport with headlights was seen immediately to the east of the El Outid feature. Again, on the morning of 16 March scattered vehicles were seen south-east of Ksar Tarcine.
A party of representatives from KDG, LRDG and 2 NZ Divisional Engineers left on 14 March to reconnoitre the going to the next staging area, some 25 miles farther on. Armoured cars from KDG provided protection. This reconnaissance was uneventful, but a party from 6 Field Company that ventured still farther ahead on 16 March to investigate the crossing of Wadi bel Krecheb (north-east of Ksar Rhilane) was unable to reach its objective owing to enemy fire from El Outid.
NZ Corps Operation Order No. 1
On 16 March NZ Corps issued Operation Order No. 1. This gave briefly the Eighth Army plan – and the ‘Intention’ paragraph which reads: ‘NZ Corps will capture the airfields West of SFAX destroying any enemy forces encountered.’
The groupings and order of march were:
2nd NZ Divisional Cavalry, less B Echelon transport
KDG plus one troop artillery
8th Armoured Brigade Group, less B2 Echelon transport
Gun Group (4th NZ Field Regiment, 64th Medium Regiment, and those anti-tank and light anti-aircraft units not with other groups)
6th NZ Brigade Group – normal, plus an extra anti-tank battery and machine-gun company B Echelon
Group NZ Corps Headquarters and Signals Reserve Group – portions of 27 (MG) Battalion and other subunits not allocated elsewhere
5th NZ Brigade Group – normal, but with 1 NZ Ammunition Company under command for the march
Administrative Group, including NZASC units not with other groups
The advance would be in three stages. Stage I was a march of some 20 to 30 miles, on the night 19–20 March, commencing at 7 p.m. This would bring the leading elements of the Corps just short of Wadi bel Krecheb.
Stage II, a further advance of 40 miles, was to be carried out on 20–21 March with the same timings, but less Administrative Group, which would not move. The B Echelon Group would move as part of 6 Brigade Group in order to have protection with the closer approach to the enemy. At the end of this stage the head of the column would be some ten miles short of Tebaga Gap.
Prior to Stage II the French Group would capture El Outid and Bir Soltane, and maintain active patrolling to the north and northeast, while KDG provided flank protection along the right flank. There are one or two references in the order to the need for watching this flank, which was a long one.
All vehicles were to refuel at Stage II and be prepared to move forward at first light on 21 March for Stage III, ‘with the object of penetrating the Eastern flank of the enemy defences … and capturing the objective PLUM’, which was the entrance to the Tebaga Gap. If this was not captured immediately it would be taken as soon after first light as possible. Divisional Cavalry and 8 Armoured Brigade were entrusted with the initial penetration of PLUM.
After the capture of PLUM the Corps was to advance on El Hamma (PEACH) and finally to a line of hills overlooking the coast road just north-west of Gabes (GRAPE). In addition, Montgomery and Freyberg agreed on a possible alternative advance from PLUM which turned more to the east and bypassed El Hamma well to the south, thereafter heading towards Gabes. This alternative would be acted upon on receipt of its codename, SIDEWINDOWS, from Montgomery. (It will be remembered that 30 Corps was to break in through the Mareth Line and advance along the main road and capture Gabes.)
After securing GRAPE, NZ Corps ‘ next objective was the landing grounds west of Sfax, an operation not assigned to the Corps in the Army outline plan, but which the GOC explained at his conference would be carried out by continuing the outflanking move, with the French giving flank protection and 10 Corps, with two armoured divisions, driving on Sfax itself.
This plan, resulting from many discussions subsequent to the issue of the Eighth Army general plan on 26 February 1943, set the tasks awaiting NZ Corps.
The order contained instructions about wireless silence and recognition of ‘own troops’, and laid down ‘ground to air’ and ‘target marking’ signals, pointing out that the operation would be closely supported by the Desert Air Force.
The part to be played by NZ Corps, within the Eighth Army plan, was that by the night 20–21 March, the night of the 30 Corps attack on the main Mareth position, the Corps would have bumped the enemy at the Tebaga switch line, made evident the seriousness of the threat from this flank and so have attracted the uncommitted German reserve, and by further vigorous activity would prevent a counter-attack against the 30th Corps break in the line. Thirtieth Corps, protected on its open flank by 10th Corps, which was to begin operations in the Matmata Hills, would then start ‘rolling up’ the Mareth position from east to west. Continuing its advance, NZ Corps would establish itself on the objective northwest of Gabes, commanding the coastal road, which by then would have become the only withdrawal route for the Axis forces not already destroyed or escaped. Tenth Corps, supported by NZ Corps, would then exploit through Gabes to Sfax, for it was hoped that the complete defeat of the enemy, followed by rapid exploitation, would prevent a delay at Wadi Akarit. In the terms of Montgomery’s general plan published on 26 February, the final objective for PUGILIST was Sfax. ‘Once operations have begun on night 20–21 March they will be conducted relentlessly until Sfax has been reached.’
Administrative Instructions
All first-line units were to leave the assembly area with six days’ rations and water, petrol for a minimum of 300 miles and ammunition up to scale. Second-line vehicles would have four days’ rations and water, and petrol for 100 miles for all vehicles. One day’s rations and water and petrol to top up vehicles would be available at the end of Stage I, and thereafter replenishment would be as the tactical situation permitted. Local water supplies, high in magnesium content, were not to be used for drinking. A special ‘Golden List’ was prepared for essential administrative personnel, and distinctive labels provided for the windscreens of their vehicles, so as to ensure their high priority in the advance. The list included personnel for the airfields round Gabes and for Sfax port, and for certain NZASC units which were to control supplies at Gabes and Sfax.
New Zealand Corps was to have priority during PUGILIST for the evacuation of wounded by air, three aircraft being made available. Two would each carry six lying and two sitting patients, and the third would carry eleven lying or twenty-four sitting. It was hoped to use suitable landing grounds near the Main Dressing Station, but otherwise one Field Ambulance would take patients to the landing ground and superintend evacuation by air.