EEIE-nav-960
The Saint-Jean-de Luz Passage
The "Classic" Comète Crossing of the Bidassoa into Spain
By Geoff Warren and Philippe Connart
This is an abbreviated and slightly reworded version of an article that Philippe and Geoff have produced and I strongly recommend that you click HERE to refer to the original version on the Reseau Comete website, both for its extra detail and for the photographs and numerous maps that will enable you to follow the route more closely.
The De Greef family house in Anglet was the first of the safe houses used by Comète to hide evaders when the men arrived in the south after their train trip from Paris, but after the Gestapo visited the family in the spring of 1942, the house could no longer be used. The German investigation was linked to Frederic De Jongh having to move to Paris when the presence of a Belgian Gestapo agent (Victor Demets) was discovered in their midst. In fact, Demets had infiltrated the network in May 1941, even before Comète's "formal" existence.
Housing arrangements in Anglet were changed with the 21 May 1942 arrival of evader Lawrence Carr in Bayonne. From that time, a variety of temporary Bed & Breakfast type, and other "safe houses" were used around Bayonne and Anglet, This was the final stop before the evaders were handed over to Basque guides, the mugalari or smugglers, who would take them across the Pyrenees and the final part of their long journey to freedom.
Arnold Deppé had lived and worked in Saint-Jean-de-Luz for ten years before the war and it was he who initially arranged contact with the mugalari in June 1941. At Saint-Jean-de-Luz, the venues for the helper/evader rendezvous were concentrated around the railway station or bus depot. Then, a short walk across the guarded bridge to Ciboure, over the waters of the La Nivelle river, led them towards the first hills, and so-called "neutral" Spain.
Arnold Deppé had his flat at 3 Rue de la Baleine, less than 100m from the town hall and the "Château de l'Infante". Many Basque refugees from Spain had arrived in France and stayed at Hotel Eskualduna (now a residential building) near the market and less than 150m from the railway station. Catherine Dubarbège (widow of Muruaga) owned the hotel. Catherine Lamothe, soon Mrs Aguirre and better known as Katalina, had worked at the Eskualduna from the age of sixteen and Alejandro Elissalde was living just across the road from the hotel.
Ambrosio San Vicente lived at 7 Rue Salagoity (less than 10 minutes from the station) and his flat was regularly used to house evaders from 6 June 1942 until 13 January 1943. This corresponds with the loss of the first Comète guide, Manuel Iturrioz, who was arrested in Spain by Franco's police on 22 April 1942. He escaped two days later but Florentino Goikoetxea took over his job, continuing the crossings with Tomás Anabitarte, who had been working with Iturrioz from the beginning.
After the arrests of 15 January 1943 at Urrugne (of Andrée de Jongh and RAF evaders Stan Hope, George Ross and Bill Greaves), Jean-Francois Nothomb resumed Comète's activities in April or May 1943 but by then, most crossings were by routes other than across the Bidassoa River (ie. Bidarray, Laressorre and Souraïde). Arriving on bicycles from the Larre restaurant of Marthe Mendiarra in Sutar (used from mid-1943), evaders sometimes stayed at the house of Katalina Aguirre (née Lamothe) at 58 (now number 20) Rue du Docteur Mice in the Socoa quarter of Ciboure. Florentino is said to have been lodged with Manuel Cestona at the house Mamutguia of Etiennette Halzuet in the Bordagain quarter of Ciboure.
Except for some minor details found in various British and American E&E reports, there are very few and only very rough descriptions of the "classic" Comète route. The first known is that written by Gérard Waucquez in his famous "Brichamart report", dated 03 January 1942, after his arrival in London from Spain. We then have the diary of Paul Henry de la Lindi that he left in London after his crossing to Spain in early February 1942, and which was later published posthumously in the book "Une Mission Très Secrète".
More accurate is the autobiography of Manuel Iturrioz written in the 1980s. The first Basque guide of the network, Manuel names some of the landmarks according to the local topography. Evader Dennis Hornsey also gives some additional hints in his 1945 book "The Pilot Walked Home". Andrée De Jongh and Jean-Francois Nothomb retraced the route a couple of years after the war but left no known written trace or complete reports with the exception of a published interview with Nothomb by Juan Carlos Jimenéz in 1991.
This "classic" Comète route is often called the Saint-Jean passage, or sometimes the coastal route by the locals. The route presents three major advantages when compared to the routes across the central Pyrenees, or nearer to the Mediterranean coast from Perpignan. First, the hills to be crossed are not so high; about 500m at Erlaitz and Castillo del Inglés in Spain. Second, the winter season was not usually a serious obstacle in these relatively low hills, and the crossing of the border was easier due to lighter snowfalls. Third, the German controls in the occupied and restricted coastal zone were much less vigorous than those of the Vichy French "Milice" in non-occupied France further east.
Gérard Waucquez left Anglet by tramway to Biarritz in early December 1941 and then continued to Saint-Jean by bus. There, a meal was taken at some square (most probably by the mairie) while they waited for sunset. Their guide was waiting at the bridge to Ciboure. After walking for two hours, they arrived at a farm, got a bowl of warm milk and had a rest for 30 minutes while the guides checked the location of the border patrols.
Paul Henry de Lindi made an identical trip to Saint-Jean in early February 1942 and had a meal at some "pension de famille" (Eskualduna?). They crossed the town and the Ciboure bridge, followed the N10 for an hour before turning left, and then followed a muddy path for another hour to a farm. They had the usual bowl of milk at the same farm as Waucquez.
Dennis Hornsey (late November 1943) went via Dax and the Larre restaurant in Sutar. He was among the few at the time who then followed the "classic" Saint-Jean route for entering Spain. He arrived at Saint-Jean very early by bicycle and went to the house of Katalina Aguirre in Ciboure. He and his party left the house with Jean-Francois Nothomb and met their guides around the Bordagain quarter of Ciboure. After a two hour walk, they reached the farm at Urrugne.
As already noted, the classic route leaves from Saint-Jean-de-Luz (Doniban Lohitzun in Basque), crosses the Ciboure bridge (or later from Ciboure itself), then along the N10 road to a "relay" farm. The journey was two hours long and during successive periods, Comète used three different relay farms in Urrugne before the actual frontier crossing. The evaders were given a last bowl of milk and provided with rope soled espadrilles, walking sticks (indispensable for the coming trek) and blue serge workingmen's clothing (from Brussels after Easter 1942). Only Bidegain berri, the relay farm where Andrée De Jongh was arrested on 15 January 1943, is still remembered and celebrated.
THE THREE RELAY FARMS OF URRUGNE
The first farm, found by Manuel Iturrioz, is  Tomásénéa, the home of Francoise "Frantxiska" Halzuet, spouse of Dominique Irastorza, a PoW. Juan Manuel Larburu of Bidegain Berri, at first refused to help because of the danger. Donato Errazti, a young Spanish exile lad living at Tomásénéa, did agree to help but only as far as the Bidassoa River. In June 1942, Frantxiska's brother was wounded by the Germans while smuggling mules and she asked not to be involved anymore for a while. The guides at the time were Iturrioz (about 40 years old) and the young Tomás Anabitarte. As previously noted, Iturrioz was arrested in Spain on 22 April 1942, escaped on the 24th, but had to go into hiding and could not therefore resume the regular passages. One of his last descriptions corresponds anyway to the passage of 21 July 1942. It should be remembered that Tomás Anabitarte was the one who pioneered the route. Very tall and strong for a Basque, he was the one who helped the others and knew the Bidassoa best. The grand-daughter of Frantxiska Irastorza-Halzuet still lives there at Tomásénéa.
The second farm, Bidegain Berri, was the home of Francoise "Frantxia" Halzuet, the widow of Philippe Usandizaga, but was farmed by Juan Larburu. It began operating as a Comète relay farm in July 1942, coinciding with the Florentino Goikoetxea period and following the arrest, escape and clandestine life of Manuel Iturrioz. The young farm helper at Tomásénéa, Donato Errazti, continued doing his same job as far as the Bidassoa River, bringing back the used blue serge clothes, until the 15 January 1943 arrests at Bidegain Berri.
The third farm was Yatxu Baïta which was owned and farmed by Joseph Larretche - who also took care of neighbouring Bidegain Berri from 15 January 1943 onwards. The French escape line Margot also used this farm, and later OSS intelligence networks like Nana and Démocratie. Twelve children lived there and one of them, young Maialen (Madeleine in French), fed Donato who was then hiding in the nearby woods until he vanished into Spain after liberation of the area by the Allied armies. This farm has changed a lot since WWII but Maialen is (in 2011) still leaving nearby.
ITINERARY SAINT-JEAN-DE-LUZ VIA SAN MIGUEL AND ERGOIEN
The first part of the route from the three relay farms is a two hour cross-country march uphill through meadows and along brooks and torrents to avoid search dogs (so writes Gérard Waucquez) leading to "open heights where they can see illuminated Spanish cities from these vast and naked meadows". Iturrioz clearly cites the Mont du Calvaire (277m) and Xoldokogaina (486m) and he asks Juan Manuel Larburu or Donato Errazti to check ahead for the location of potential border patrols. There, the lights of Fuentarabia, Irún and San Sebastian are clearly visible and announce the nearby freedom. The Bidassoa can be heard (so writes Paul Henry de Lindi) and seen or imagined along the path. Hence, the evaders are on the southern flank of the hills facing Spain and not along the reservoir on the northern, French facing flank of the hills. Waucquez describes a march of two hours along a ridge.
Then begins the descent to the crossroads at the Col des Poiriers (316m): Iturrioz says that this was the most dangerous part of the route before the Bidassoa. There was a serious risk of encountering border patrols (Germans or French gendarmes) there because it was the crossroads of the well-used footpaths between the Ventas d'Ibardin, Biriatou and the Bidassoa. Juan Larburu and/or Donato Errasti, earlier having gone to San Miguel on reconnaissance, would wait there to give the green light. The crossroads was also the beginning of the first planned (but never used) route to Endarlatsa via the road to Ibardin.
From the crossroads, the descent to the Bidassoa, along the ravine of the Lantzetta Erreka (stream/creek) is quite easy to follow. This descent took an hour to San Miguel. During the crossing by Waucquez in early December 1941, after a failed attempt to cross the Bidassoa at San Miguel due to overflowing waters from heavy rains, Tomás Anabitarte found a shortcut to Endarlatsa along the way. This shortened route reduced both the distance, the time it took (reduced to about two hours more than the San Miguel route instead of four) and the danger of encountering border patrols compared to using the longer road to Ibardin route.
At San Miguel, some evaders describe a pasture before the river but very few mention the railway station building, sometimes described as the post of the Spanish guards. Once out of the water, they first had to cross the narrow tracks of the "Bidassoa Ferrocarril" railway, called the "Txikito Tren" (Little Train) by the locals, and then the Irún-Pamplona road. It seems logical that the now abandoned train station was used at the time as a barracks for the Spanish Guarda Civil and patrols and was thus to be avoided at all costs. Iturrioz once heard a guard snoring in the "farm" (San Miguel station? or it might be the farm slightly south of it, where the first remembrance treks crossed). It is also logical for the evaders to have crossed the Bidassoa approximately in front of a ravine on the Spanish side, 350m downstream from San Miguel or 600m upstream; rather than alongside the so-called "Txikito Tren" station at San Miguel. This was not a passenger station as one might imagine. Rather, the existence and construction of this desolate station in the 19th century was only justified and used for the loading of mineral ore onto bulk train carriages until passenger services were added in 1916 for the very few people living in the area.
After crossing the river, the railway and the road, the next step is the slope upwards to Erlaitz, very steep and exhausting. The bottom of the ravine, 350m from San Miguel, is the streambed of a seasonal brook. The evaders entered the cover of a pine wood right after crossing the road (Carretera Irún-Pamplona) and climbed to the former Carlist tower of Erlaitz (the 1891, never-completed Fuerte de Erlaitz). The uphill route they followed is parallel to the former 1.8km long tracks of two ancient rectilinear narrow mining railway lines running along the ridge from the heights of Erlaitz/Pagogana down to San Miguel station.
At Erlaitz, Iturrioz says the evaders would go "in front of the barrack blocks used by workers maintaining the road Pikoketa-Oiartzun" (today's GI-3454). He sometimes took a longer but safer road to avoid the barracks, occupied by Carabineros of the Spanish army, then rejoined the road to Pikoketa and Oyartzun. Here at Erlaitz, the infrequently used alternate evader route from Endarlaza joins the traditional route from San Miguel. Of the alternate route, Paul Henry de Lindi says they "had a rest in an old farm on the peak, one hour march after climbing the cliff at Endarlaza. This peak was the highest in the vicinity and the farm was the only one to be seen on the whole horizon. In all directions, he could only see steep mountains and deep ravines, shrouded in a deadly silence and covered with snow". He was told there were two more hours by foot until the next station where they would take a train to San Sebastian. Waucquez writes, "eventually, around 6 o'clock in the morning (after crossing at 4 o'clock in Endarlaza) they arrived at an abandoned farm or a water tower and rested. One of the guides went to Irún and alerted Aracama in San Sebastian by phone". This was in December 1941.
In Iturrioz' autobiography, today's GI-3454 road is described as a former ox carriage road going around Peña de Aya (Three Crowns): a "rather flat road until Pikoketa and from there about an hour of descent to Xagu" (Sarobe). We have learned that the first 5 kms from Pagogana was the former "narrow track" mining railway called Las Tres Coronas. Hornsey recalls having followed a railway tunnel before Sarobe, and Nothomb confirms in his 1991 interview that "there was then a railway for the mines, with narrow tracks if I'm not mistaken. One followed it for three or four kilometres and, at a certain point, one went through a rather long and dark tunnel. A little further, one arrived at the farm".
The route described can be either the Three Crowns railroad, located on the high ground between Erlaitz and Castillo del Inglés (but no tunnel seems to have existed there and the tracks were removed in 1932), or the Pasaia-Arditurri mine railway alongside the GI-3420 (the rail line is now a bicycle path called Arditurri bidea). Located in the valley below Pena de Aya (Three Crowns) and Castillo del Inglés, there are several tunnels along what is now the bicycle path and they are all located less than 2 kms of easy walking from Xagu (Sarobe). There is some evidence that the evaders route sometimes followed the Arditurri mines railway in the valley to Sarobe farm. It was the only railway still existing in WWII. Another route, very close to the present remembrance trek runs roughly parallel to it, along the flank of the ridge from Penãs de Aya to Oyartzun.
Surprisingly, the three most detailed accounts of the "Saint-Jean" route are from three evaders who actually used the rather exceptional detour via Endarlaza rather than the "classic" Bidassoa crossing at San Miguel station: Waucquez, Henry and Hornsey. Even though the name of Endarlaza is mentioned, it should be noted that the "detour" crossings of the Bidassoa actually happened using a suspension footbridge at the Endara electrical plant, about 2 kms upstream from Endarlaza. This alternate route starts near the Col des Poiriers crossroads and Tomás Anabitarte says it was four extra hours of marching compared to the San Miguel route.
These five special and exceptional crossings are :
10 Dec 1941 : Jack Newton (RAF) Gérard Waucquez (Belge) Hillary Birk (RAAF) and Howard Carroll (RAF). The only detailed, recorded evidence of this crossing comes from the Waucquez report. This crossing gave the name Endarlatsa (or Endarlaza the Basque language has no "z"). Waucquez says it was a 2 hour trek along the ridge, then 2 kms upstream. Guided by Andrée De Jongh, Manuel Iturrioz, Tomás Anabitarte and Donato Errazti.
08 Feb 1942 : Paul Henry de Lindi (Belge) Georges Osselaer (Belge) and Norman Hogan (RASC). A rather precise description of the crossing was recorded by Henry de la Lindi. Guided by De Jongh, Manuel, Tomás and Elvire Morelle (to learn the route - she broke her ankle on the way back).
10 May 1943 : Bernard Marion (RCAF) John Whitley (RAF) David Sibbald (RNZAF) William Laws (RAF) and Gordon Brownhill (Cdn/RAF) crossed "by the West of Mount Alecor" (the origin of this detail is unknown). This crossing is with J-F Nothomb and Jacques Tinel. The latter, a Parisian, was chosen to replace Albert Johnson and become Nothomb's assistant, but was soon arrested. After crossing the suspension bridge at Endara, they followed a tunnel, crossed a stream, and climbed. Brownhill became separated from the group and the time spent searching for him may explain why Nothomb remembers this trip as being very long.
28 Sep 1943 : Edward Bridge (RCAF) Arthur Bowlby (RCAF) Elmer Dungey (RCAF) James Allison (RAF) George Baker (RAF) and George Duffee (RAF). Guided by Nothomb and Florentino.
On 28 August 1943, Halifax JD368 of 10 Squadron RAF was shot down 12 kms southeast of Mons, Belgium. The pilot was George Baker and Geoff Warren's uncle, Sgt George R.M. Warren, the 19 year old rear-gunner, and the only Canadian on board, perished that night. The 7 remaining crew members baled out safely : 6 evaders (4 Comète/2 Bourgogne), 1 PoW.
27 Nov 1943 : Dennis Hornsey (RAF) George Gineikis (USAAF) Leon MacDonald (USAAF) Geoffrey Madgett (RAF) and J.J. Greter (Dutch). Guided by Nothomb and Florentino plus, according to Gineikis/MacDonald, a Basque refugee from the Spanish civil war named "Monterro". The US aviators refer to Monterro as a "guide".
When it became necessary to use the suspension bridge route at the Endara electric facility for the first time in early December 1941, Manuel Iturrioz planned to use the well guarded, and therefore dangerous path from Ibardin. This was the occasion of the Wauquez failed crossing of the Bidassoa at San Miguel, which Andrée De Jongh demanded be attempted, notwithstanding the water levels being too high after recent heavy rainfall. The crossing was aborted and a second attempt was made the next night using the "detour" route. Fortunately, Tomás Anabitarte and Donato Errazti had found a shortcut to avoid the Ibardin "road" route and had placed two trail markers on the path to assist the men in following the correct route to Endara and the suspension footbridge.
Hornsey describes crossing a tributary of the Bidassoa (Lantzetta or Lizarlan erreka/stream?).
Iturrioz descended to the Bidassoa immediately after the second trail marker and reached the suspension bridge where a guard barrack (now demolished) was seen in the direction of Irún, on the right hand bank of the river. They crossed together in single file in a crouched position.
Hornsey followed a road leading to a little village/hamlet (that was the "old" road to Vera de Bidassoa) where the (swinging) bridge hung over the river.
Henry de la Lindi crossed the Bidassoa where the power plant is located on the other side, and then followed a railway (Txikito train) that entered a tunnel after 100m. He crossed a 6m wide brook (at the Ermita?) and then climbed a "200m wall". This first tunnel is about 200m long, and dark, and wet - even today. Further towards Endarlaza, there is another arch and a third tunnel 50m long.
Hornsey climbed directly up a terrible slope at sunrise and similarly says in his book that they passed alongside an abandoned tower (Torreón de Pika) near the border station at Endarlaza. They leave the Dutchman Greter at some "hamlet" (Fuerte de Pagogana, Erlaitz, the Pikoketa road workers barracks?)
Hornsey crossed a railway and a road and followed the railway through a tunnel. Is this the railway from the mines of Arditurri or the Txikito railway? (thus, immediately after the suspension bridge). He then says that "we came to the old abandoned tower" meaning he crossed at Endara and somehow followed the Bidassoa.
FROM ENDARA POWER PLANT TO PAGOGAÑA - THE DILEMMA
To get from the power station to Pagogaña without being detected by the Guardia Civil at the Endarlaza bridge and barracks, there are two options:
OPTION 1 : Follow the railway to the Guardia Civil barracks.
Advantages : The railway route to Ermita cliff is flat and easy to follow.
Disadvantages : The presence of Guardia Civil around barracks/bridge.
Mitigating Factors : Timing, noise and lighting.
Timing : The evaders arrived at the barracks building before or just at sunrise when the guards were least likely to have been awake, alert or interested. As anyone who has "stood a watch" during a night shift knows, the hours between 0300 to just before sunrise are the least pleasant, most boring, and most likely to induce sleep or lethargy.
Noise : The flow of the river over the nearby upstream dam, and the subsequent "rapids" between the dam and the bridge/barracks, would have muffled the sounds of the men passing.
Lighting : It is logical that any bright lighting used to detect intruders would have been mounted at the bridge; not at the barracks. Any guards on duty at the bridge (± 100m from the barracks) would have their sight severely impaired beyond the limits of the bright lights. One even wonders if the guards would want to be "illuminated" at night when a disgruntled veteran on the losing side of the civil war could easily, at fairly short range from the heights across the river, take shots at the on-duty Guardia Civil members (or any others outside the nearby barracks) and make his escape back into the hills in the darkness.
Any lights outside the barracks, if any, would likely have been quite dim and provided ample "shadows" for the evaders to pass the barracks undetected if sufficient caution was taken. Plus, the Ermita ravine (likely in total darkness) begins only steps from the Guardia Civil building.
The men had just crossed the suspension footbridge at the electric plant which was reportedly "brightly lit" at all times and supposedly "guarded". The evaders had to pass within a few feet and along two sides of the power plant building en route to the first train tunnel, yet managed to do so without detection. That said, the noise from the electric machinery would likely have provided more protection than the sound of the river at the barracks downstream.
OPTION 2 : Leave the railroad tracks and climb the hills before reaching the Guardia Civil barracks.
Advantages : Reduces considerably the risk of being detected by the Guardia Civil at Endarlaza.
Disadvantages : In darkness, having to climb the steep slope above the railway, and then traverse at least one extra ridge before reaching the Ermita ravine and the cliff to Pagogana. Those who have been there know it is a task almost beyond human capability.
Mitigating Factors : Terrain, which is rather a determining factor.
Terrain : If the men in this scenario left the railroad tracks after negotiating the tunnels and the rock "cut" (the latter was described as"like walking between two walls"), there are two spots where this was feasable. They were then left with only one, two at the most, ridges to cross before descending into the Ermita ravine. In fact, as noted below, it was possible to avoid the ridge summits entirely.
It appears from personal observation, Google and topo maps that it was possible, after climbing uphill a short distance from the railway, to reach the Ermita ravine by traversing the flank of the ridges on a route roughly parallel to the river and thereby avoid the ridge summits themselves. There also seems to be a few farm paths/tracks along the way which could have served to lighten their load even more. The distance from the climb upwards from the railway track to the descent into the Ermita ravine is between 500m and 800m (or slightly more), depending at which point they decided to begin their ascent from the tracks.
Bearing in mind that the Ermita streambed (Rio Endara) averages about 10m wide in the immediate area and would generally decrease in width the further upstream the men walked, it seems possible they could avoid a treacherous water crossing by simply walking a few hundred metres up the ravine before attempting the steep slope to the Pagogana plateau.
TRIVIA/SPECULATION :
Oiartzabaleko Borda : Immediately upon reaching the summit of the steep slope/cliff above the Ermita ravine, the men would arrive at a sloping grass meadow plateau immediately below the Pagogaña ruins. Located on this plateau today are what appear to be two farms; one apparently called Oiartzabaleko Borda. Is it one of these farms (or a predecessor) described by de Lindi as the place they rested in "an old farm on the peak" or by Waucquez when he mentions arriving "at an abandoned farm or water tower"?
Torreón Pika : If the men climbed the Ermita slope near the Pika tower while en route to the Pagogaña plateau, and then followed the same ridge on which the tower is built to reach the summit, it appears to be a reasonably benign route to have taken. Following that ridge would also bring them directly to the Oiartzabaleko Borda then, after 600m slightly uphill, to Pagogaña ruins.
CONCLUSIONS :
None! It's unlikely to ever be known conclusively whether the "Endarlaza detour" evaders crept past the Guardia Civil barracks on the tracks or bypassed the location using the ridges above. That 500m to 800m of travel is the only significant missing link in the story and that isn't too bad considering the distances (horizontal and vertical) covered by the men from the time they left Saint-Jean-de-Luz and their relay farm.