Mysteries of P.C. Cadix and its evacuation in 1942/43

From 1940 to 1942 a group of Polish codebreakers, authors of the initial Enigma breakthrough, had been working at the P.C. Cadix, a secret codebreaking center situated in the unoccupied France. Their work in this period, and in particular the circumstances of their evacuation after the German/Italian occupation of Vichy France, are shrouded in mystery. This paper represents an attempt to put information known so far into a broader context, revealing in the process possible distortions and omissions in the accounts of the participants of the events.


What did we know so far?
In June 1940 several officers of the pre-war French intelligence services, gathered in the Bon Encontre seminar near Agen, had decided to continue their fight against Germany in spite of their country's defeat. Major Gustave Bertrand, pre-war commanding officer of Section D of the French Service de Renseignement, was in their number. By manipulating the truth, he managed to retain his control over two groups of foreign codebreakers, Poles and Spaniards, constituting the entire assets of his service. Having temporarily secured their members in the French Africa, he started work on reorganizing his service. In November 1940 a country house near Uzès in the non-occupied zone of France, was ready to accept new residents. This is how the clandestine codebreaking center known as P.C. Cadix started its operation.
Clandestine organizations usually preserve a minimal documentation of their activities. P.C. Cadix was no exception. Even declassified in 2015 parts of Bertrand's private archive provide surprisingly limited information on its operations. In this situation any reconstruction of the P.C. Cadix' story must be based on the available personal recollections of its former staff members. Bertrand had published his memories in 1973, unveiling for the first time the information about breaking the Enigma cipher. Marian Rejewski had deposited his memories with the Institute of Military History in 1967, but they were published for the first time only in 2011 (Rejewski 2011). After reaching Great Britain in 1943 Wiktor Michałowski submitted his report from the activities of P.C. Cadix, or rather Ekspozytura 300, as the center was known to the headquarters of the Polish intelligence service in London. Lt Colonel Gwido Langer submitted his report only after his return from the internment camp, in 1945.
Reconstruction of real activities of P.C. Cadix based on those sources is difficult, if possible, at all. P.C. Cadix represented a slightly schizophrenic organization, staking several layers of external and internal conspiracy one upon the other. Its existence and activity were obviously hidden from the Germans, as contradictory to the terms of the armistice treaty. They were kept secret, to the possible extent, also from the Vichy government and the HQ of its army, both pretending to respect the terms of the armistice. The purpose of center's work and its position within the structure of the clandestine French secret services were kept hidden from the teams of Polish and Spanish codebreakers. Both teams were working, and to some extent living, in the strict isolation from each other. Finally, Polish team was working in the conditions of double subordination.
It was administratively subordinated to Bertrand, and through him to the French intelligence service, but operationally directly to London HQ of the Polish intelligencefact well hidden from Bertrand and the French hosts. In the conditions described it is obvious that no member of the team could learn and report a credible story of the P.C. Cadix. Moreover, reports and memories of the team members representing different groups do not sum up to a consistent picture, and include obvious gaps and contradictions.
What did we know then about center's activities, or rather what we believed to know? P.C. Cadix represented a part of Travaux Ruraux, created by Paul Paillole immediately after the collapse of France, on 1 st July 1940 (Pailolle, undated). Travaux Ruraux represented a branch of the secret services of the Vichy France masqueraded as a commercial organization specializing in the agricultural works. As a successor to the former codebreaking section of the French military intelligence service, P.C. Cadix was focusing on the signals intelligence and the codebreaking, and consisted of three teams: 'L'équipe D' -group of Spanish republicans dealing mostly with Italian and Spanish ciphers, 'L'équipe Z' -team of Poles breaking the German and Soviet ciphers; finally the French component, responsible for purely administrative functions. Maj. Gustave Bertrand acted as center's commanding officer, with Lt Col Gwido Langer commanding L'équipe Z, and Antonio Camazón L'équipe D. Both intelligence services were aware that the possible imprisonment of any team member by the Germans was a mortal threat not only directly to him, but above all to the secret of breaking the Enigma code.
The real command structure of P.C. Cadix was complicated by the fact that, unbeknown to Bertrand, L'équipe Z was operationally subordinated directly to the London HQ of Polish intelligence service. In that role it was known as "Ekspozytura 300", and its existence seemed to serve a double purpose. First, it permitted keeping the pre-war team of the Cipher Bureau, for political reasons alienated within the structures of the service, far from London. Second, its presence in the southern France offered a comfortable relay service for the messages exchanged with the highly efficient Polish intelligence network in North Africa, "Rygor", too distant to communicate directly with London.
Requirements of the "Rygor" network were probably one of the reasons for the creation of P.C. Cadix's branch station in Algiers, known as "Kouba". Maj. Maksymilian Ciężki, deputy commander of L'équipe Z, was spending in 1941-1942 most of his time at "Kouba", combining the functions of its commander with that of "Rygor's" signals officer. We may only guess that the presence of the Polish codebreakers in Algiers had one more, and very important reason. Marian Rejewski and his colleagues attempted to continue their work on Enigma ciphers, but were suffering from the lack of the cipher material. French interception service, Groupement des contrôles radioélectriques (GCR), delivering cipher material to P.C. Cadix, did not intercept Enigma traffic due to the technical problems, and, as we shall learn later, different priorities of its work. Poles could continue their work on Enigma only taking care of interception themselves. Shifting their wireless sets to Algiers, closer to the battlefields of Libya and Egypt, they could hope to provide their codebreakers with the current cipher material. We will probably never know how Różycki and Zygalski (who were usually manning the Algiers outpost) managed to break Enigma keys during 1941 and 1942, without access to equipment used at that time at Bletchley Park. All we know from Bertrand's later book is that they managed to read around 4.000 Enigma messages, including several of high importance. Poles were able to break the Enigma keys used by the German forces fighting in the eastern front. Some of them represented the earliest information about the atrocities committed by the Einsatzgruppen following the Wehrmacht in Russia. End of their adventure with Enigma was marked on 9 January 1942by the catastrophe of the passenger liner Lamoricière, taking the lives of three Polish codebreakers, including Jerzy Różycki, and two copies of Enigma machine returning with them to France.
While Różycki was working in Algiers on Enigma messages, his colleagues in P.C. Cadix were attacking other German ciphers. In his memories Rejewski describes two ciphers being solved by the Polish team. Antoni Palluth, nominally the team's engineer, and not the codebreaker, focused his attention on the first one, representing a columnar transposition with blanks. For the reasons we shall cover later, his solutions must have been extremely precious for the French hosts. This cipher was in use by the German agents working in the unoccupied part of France. Rejewski describes one of Palluth's solutions leading to the discovery of agents' secret meeting in a hotel at Marseille. Their summary arrest was facilitated by the fact, that all of them appeared at the meeting place carrying their radio sets in identical suitcases.
The other cipher represented a standard Playfair, so its breaking was a rather routine job, but at some time its solution proved to be of a critical importance for entire team of P.C. Cadix. One of the most spectacular achievements of the French service was tapping the telegraph lines used to coordinate the network of German radio monitoring stations. Those stations used to exchange via teletype information on the underground transmitters, their working frequencies and fixed positions. Using the cables believed to be under German control, network stations did not bother to use a more secure cipher. Tapping the cables by the French underground, and breaking the intercepted messages by the Polish codebreakers must have saved a number of Allied wireless operators, whose sets were mentioned as aushebereif (ready to be captured). In one of the most spectacular twists of the history, in September and October 1942 messages broken by the Polish codebreakers started to contain references to their own radio station. Finally, in late October two German cars with visible direction-finding antennas appeared in the close vicinity of the villa hosting P.C. Cadix. Officers commanding the center were informed about coming Allied landings in North Africa and realized the danger of German occupation of Vichy France; it was obvious that the existence and activities of P.C. Cadix are coming to an end.
Events of next few months represent one of the most mysterious periods in the Enigma history. The effects of actions undertaken by the participants of the events diverge from their declared intentions to an extent, that suggests caution in their interpretation. Before we compare these declarations with several new sources, external to the participants of the events, let us recall the story in the version resulting from the accounts of their direct actors.
Communicating directly with London SIS HQ, Maj. Bertrand had been pre-warned about the danger to P.C. Cadix resulting from the planned Allied operations in North Africa. London was to signal the imminent start of the operation with a pre-agreed message: "the harvest is bountiful". The SIS and Polish intelligence headquarters assumed that this signal would immediately cause the evacuation of a group of Polish codebreakers out of the zone where they would be in danger of falling into German hands. British intelligence offered help in the evacuation aboard a submarine that could pick up the Poles from the French coast. The evacuation of General Giroud aboard a British submarine on November 4 confirms the possibility of such an operation. Polish intelligence networks in France also had secure evacuation routes that could be used to evacuate the team to Spain and then via Gibraltar to the UK. Both services were aware that the possible capture of any team member by the Germans represented a deadly threat not only directly to him, but above all to the secret of breaking the Enigma cipher.
In October 1942 Bertrand and Langer decided to pay a visit together to the head of Polish intelligence network "F", to investigate its potential role in the evacuation of the Polish team. It is difficult to understand the need for this visit. More or less at the same time Bertrand had decided to use means controlled by his own service to evacuate to Algiers the Spanish codebreakers, members of "l'équipe D". Their evacuation must have been smooth, as they were soon able to continue their work in North Africa, this time under American control. When asked by Langer why not to use the same route to evacuate the Polish team, Bertrand considered this option too risky, as North Africa, in his view, would soon turn into a battlefield.
During Bertrand's absence at P.C. Cadix, an expected message from London was received, but not read, having been enciphered using Bertrand's private key. It was only after his return that he deciphered the text informing that the "harvest is already very bountiful". Finally, Bertrand's dilemmas were resolved by the previously described visit of German cars with direction finding equipment. Over the next few hours crucial equipment was hidden in previously prepared caches in the walls, center's archive was secured in Bertrand's mother's house at Grasse, the villa had been abandoned and the entire team of P.C Cadix went into hiding, preparing for the evacuation from the danger zone. On a personal level, the same danger applied to all members of the organization within which the center operated. In his memoirs Bertrand (1973) suggests that he expected his superiors to organize the evacuation of the team. The time was ripe for a decision, and the Germans provided a few days for its implementation. "Fall Anton", German and Italian occupation of Vichy France, started only in the evening of 10 November, three days after Allied landings in North Africa. Therefore, Bertrand suggests to be disappointed after his superiors departed for Africa from the air base at Istres, leaving behind him and entire team. His disappointment was all the greater because the team of P.C. Cadix could reach the airfield in about an hour, and there were still plenty of empty seats on board the flights to Africa. Immediately after the German/Italian occupation of Vichy France Bertrand transferred Èquipe Z to Italian zone, considering it safer. At the same time, he divided the entire team into smaller groups, which made it easier to find hiding places and reduced the risk of exposure. A side effect, however, was making the Polish team completely dependent on Bertrand's concepts and actions. And these changed frequently. For some time, the plan was to evacuate the Poles on board of a submarine, which was to pick them up from one of the secluded bays on the Côte d'Azur. This plan failed, reportedly due to the planned embarkation point being manned by Italian troops. Next, the possibility of evacuation via Switzerland was analyzed. However, this would entail the inevitable internment of Poles, which did not generate enthusiasm either on the part of those concerned or the British, with whom Bertrand consulted this option. As the days and weeks passed, the crisis of trust between Bertrand and the officers commanding the Polish team deepened. The Poles accused Bertrand of sabotaging the evacuation of their team and thereby exposing the secret of Enigma. Bertrand responded accusing Poles of alcohol abuse and rejecting his subsequent concepts without justification. Ultimately, both sides agreed that the only viable, albeit risky, option was to cross the green border to Spain, from where Poles would be evacuated to Great Britain via Red Cross channels. From the distance of time, it is hard to tell whether the news that Bertrand would not be accompanying them on the road, caused more anxiety or relief. To supervise the operation Bertrand designated his deputy from P.C. Cadix, certain Captain Louis (whose talents, as later recollections show, he did not highly value).
Poles, still divided into smaller groups, were moved near the Spanish border, to territory controlled by German troops. There, they waited for the possibility of crossing the border in the hideouts in Toulouse, Perpignan and Narbonne. Two groups, consisting mainly of the codebreakers, managed to cross the border. In particular, Rejewski and Zygalski made it to Spain on the night of January 29/30, 1943. And although on the way they were thoroughly robbed by the guide, and on the Spanish side they were arrested by the gendarmerie, they were safe. The fate of group led by Maj. Wiktor Michałowski was similar.
The fate of the officers, commanders of the Polish team, was different. A group including Col. Gwido Langer and Maj. Maksymilian Cieszki, on the first attempt to cross the border, was stopped by the French gendarmerie, and the officers were imprisoned in the Perpignan citadel. The French unofficially expressed regret over the incident, implying that they had not received the customary in this situation warning about the need to turn a blind eye to a group of travelers. The sympathetic French gendarme also noted that certain Monsieur Gomez, the smuggler whom the French handlers had hired for the task, was suspected of collaborating with the Germans. Finally, after paying the necessary bribes, the Poles were released and in March they were able to make another attempt to cross the border.
In the circumstances described above it was natural for the officers, once they landed again in the border zone, to try to avoid contact with Monsieur Gomez, who botched the previous attempt. Following the suggestion of a French gendarme, they agreed terms of service with another smuggler group. However, on the eve of the planned border crossing, a local representative of Monsieur Gomez found them and forced them back into cooperation. Gomez must have sensed a lack of trust from his clients. To calm them down, he took out a bill, ordered Langer to sign it, then tore it halfway, keeping one half, and handing the other to the officer instructing him to give half of the note to the guide only after safely reaching Spain; the guide would be paid only after presenting the correct half of the banknote. On the night of March 10-11 the escapees had covered only a few kilometers when they were surrounded by a group of German soldiers on motorcycles. All were detained, except for the guide, who was allowed to leave unhindered by the Germans.

Greater picture in the context
Functioning, organization, and activities of P.C. Cadix were so far analyzed mostly from the point of view of inter-allied cooperation in the codebreaking. This point of view does not permit to explain and clarify ambiguities or even contradictions in the available sources. However, there exists an alternative point of view that has not been sufficiently exploited by historians of the subject so far. P.C. Cadix represented a part of the secret services of Vichy France, whose functioning was, and is, the subject somewhat overlooked by historians. Their activities, however, left important source materials that permit to place the activities of Polish codebreakers in the appropriate context and find answers to questions that have caused problems in previous attempts to analyze the subject. This paper represents an attempt to place the known facts in a new context and to offer the reader the resulting conclusions. It is focused on two particular questions; P.C. Cadix's scope of activity and the circumstances of codebreakers' evacuation after German/Italian occupation of Vichy France.

P.C. Cadix -scope of activity
The mainstream of Enigma history research treats the operation of the P.C. Cadix as a natural continuation of the activities of the P.C. Bruno, under slightly more complicated conditions. In this version, Polish codebreakers were to continue breaking Enigma, with significantly lower efficiency than before, due to the lack of technical equipment. In fact, breaking Enigma ciphers was only a marginal part of their activity during this period. The lack of technical equipment was, of course, an important factor, but it was the completely different priorities set for the center by the French principals that were decisive. P.C. Cadix represented a part of the structure of the French organization Travaux Ruraux (TR), one of the secret services of Vichy France disguised as a commercial company. TR was created on 1 July 1940, by Capt. Paul Paillole, before the defeat of France the adjutant of the head of the French military counterintelligence, Lt Col. Guy Schlesser (Paillole, undated). In accordance with the previous experience of its founder, TR was an organization of a clearly counter-intelligence character. In this role, it enjoyed a significant level of autonomy. Bureau des Menées anti-Nationales (BMA), the structure, which was supposed to act as the supervisor of Vichy France's special services, was not established until August 25, 1940. From the outbreak of war to the defeat of France, the 5ème Bureau (Col. Louis Rivet) provided some level of coordination of the intelligence activities of the services. After the defeat of France, this structure was not recreated, so the counterintelligence, as well as the intelligence services of the army, navy and air force, regained almost complete independence 1 .
As a result of the described changes, the codebreakers' team, previously subordinated to the command of the French intelligence service, was transferred entirely to the domain of counterintelligence. Enigma messages the codebreakers had been previously providing were of marginal importance to their current supervisors. TR's main task was to identify and neutralize agents of foreign powers operating in territories subordinated to the Vichy authorities. Neither the agents, nor their handlers were enciphering their messages with Enigma, so the main asset of the Polish team lost its significance in the eyes of their present French superiors.
Continuation of work on Enigma was also difficult due to the lack of cipher material. Before the defeat of France, the signals intelligence services of France and Great Britain kept on exchanging intercepted messages. The defeat of France disorganized her intercept service and cut off the codebreakers from British sources. From Rejewski's memoirs (1967), we know that Bertrand tried to keep Poles busy by delivering a package of intercepted Swiss messages, encrypted with a commercial Enigma (which the Poles easily broke). In the meantime, the French attempted to rebuild their own intercept service, which was not a simple task given the limitations of the Armistice Treaty. Groupement des Contrôles Radioélectriques (GCR) started its operation on 10 August 1940 at the Château des Cours in Hauterive. GCR, organization created and led by Gabriel Romon, formally represented a part of the French Post Office, less formally providing valuable information to Vichy secret services. Its early activities are poorly documented; the earliest GCR documents preserved in French archives date back to 1942 (GCR). Fortunately, its history has been largely reconstructed by the son of its founder, François Romon, and described in his book (2017). Romon suggests that 'German Enigma messages deciphered by Gustave Bertrand at P.C. Cadix mostly came from GCR's intercepts'. Cooperation between GCR and P.C. Cadix was greatly facilitated by the fact, that one of the GCR's interception centers was located in the vicinity of Uzès, at Bouillargues. However, Romon's assumption is not confirmed in Rejewski's memoirs. He mentioned that the Polish team had to dedicate two out of four radio sets and several operators only to intercept Enigma messages. Romon's claim may refer to the work of Polish codebreakers in Algiers, where they had at their disposal only one radio set and had to rely on the French intercepts. On the other hand, Romon himself points out that 'GCR's intercepts concerned mainly the messages of German agents in non-occupied zone. Content of some messages, among the others, permitted to discover and annihilate their entire network, comprising 10 agents and 6 radio sets'. Romon refers clearly to an episode described also by Rejewski, when the messages broken by Antoni Palluth (who specialized in double transposition cipher used by the agents) permitted to surprise a group of agents during their meeting in a hotel in Marseille. Paillole estimated that between October 1940 and November 1942 TR was able to eliminate over 1.000 enemy agents (Paillole, Chronologie). We will never know how many of them had been identified by the messages broken by P.C. Cadix. Moreover, Paillole's definition of the "enemy agent" had been a bit fuzzy… Anyway, it is clear that TR's priority was not intercepting and breaking Enigma, but the hand ciphers used by the agents working in Vichy France.
The second area of activity of GCR, and consequently Polish codebreakers, was the interception and decrypting of German teletype communications. The operation, which resulted in the acquisition of the cipher material, referred to within the GCR as Source K, began in October 1941. In March 1942, the French rented a house located on the cable route connecting Paris with Metz. On April 18, they simulated a cable failure, and during the "repair" tapped the lines used by the Germans between Paris and Berlin. With one exception, no data is available on how this source of information was used. This only exception are messages exchanged between German signals intelligence stations cooperating in the direction finding of Allied underground transmitters in the occupied countries. Since the messages were exchanged over a cable line that the Germans considered safe, they were secured by a relatively low-level cipher, i.e. Playfair. In this form, they were transferred to P.C. Cadix, where Polish codebreakers kept on breaking them without much difficulty. In his memoirs, Rejewski describes the satisfaction of being able to warn underground station operators against exposure. He did not record his own reaction, however, when the broken messages referenced the P.C. Cadix's own station, indicating impending danger. Later on the tap was discovered by the German security service, and the line was secured using machine cipher. Sadly, some members of the GCR were imprisoned and executed, including Gabriel Romon.
Summing up, breaking by P.C. Cadix of about 4,000 Enigma messages was an astonishing success of a Polish codebreakers' team, who did not have at their disposal any of the equipment used at the same time at Bletchley Park. However, it is equally obvious that breaking Enigma keys was merely a sideline of the team's activities. P.C. Cadix functioned as part of the French counterintelligence, whose priority was to decipher messages, on the one hand, enabling the identification of foreign agents in Vichy France, and on the other, protecting its own structures against identification by the enemy.

P.C. Cadix -mystery of evacuation
The evacuation of the Polish team from the south of France as a result of its occupation by the German and Italian forces represents one of the most mysterious episodes in the history of Enigma. Considering the facts presented in the previous section, during the period of work at P.C. Cadix, Polish codebreakers only marginally dealt with problems in which they had the greatest experience. The value of their work in the south of France to the Allied cause was highly debatable. The occupation of the south of the country by the Axis forces made it not so much worthless as dangerous.  , p 14). Paillole still in 1942 represented an opinion that although Germany was 'enemy number one', Britain was 'enemy number two' (Kitson, 2008). In line with this mutual distrust, at the turn of 1941 and 1942 the British handed over to P.C. Cadix several keys to the Enigma cipher, taking care of the security of their own operation; they only shared several keys captured in North Africa, excluding those broken cryptanalytically (Borowiak, Grajek, 2022; note for CSS, April 9, 1942, TNA, HW 65/7).
The British made efforts to urgently evacuate the Polish team from France. However, there was little they could do but make the French aware of the importance of the matter and offer help. Quoting Paillole: 'on the evening of November 5, Colonel Rivet, head of the special services, called me urgently to inform me of the content of a dispatch from the IS, signaling the imminence of an Allied landing in North Africa. The dispatch also asked to expressly withdraw to Algiers the precious Polish decryption personnel of the PC "Cadix"' (Paillole, 1975). On November 6, Rivet convened the final meeting of the body known as the Little Chancellery, composed of the heads of Vichy's secret services. He wanted to obtain information about the intentions of individual services in the face of the upcoming Vichy occupation. Paillole requested the navy's assistance in evacuating the TR's archives, weighing some 40 tons, waiting on the quay in Toulon. Chief of Naval Intelligence, Capt. Sanson, after consulting his superiors, conveyed their answer stating that 'no, there is no reason why those archives should fall into the hands of the British any more than into those of the Germans.' (Paillole, Résume...). The climate of those days was clearly not conducive to cooperation with former (and future) allies.
The complexity of the situation is compounded by the fact that an Allied operation caught the Vichy special services in the process of reorganization. The Germans, having a good understanding of the actions of the French, forced the Vichy government to dissolve the BMA and its subordinate agencies as early as April 1942. The French delayed the implementation of their orders, but the takeover of the reins of government by the pro-German Laval precluded ignoring them for a long time. In August 1942, the BMA was disbanded and replaced by the Service de Sécurité Militaire (SSM), under the command of ... Paillol. At the end of September, the new SSM chief organized a secret meeting of the commanders of intelligence services of the army, navy and air force. The subjects discussed included plans for action in the event of the German occupation of Vichy. It was established that in such a case TR structures would stay in the occupied territories and continue its counterintelligence activities.
Despite this conclusion, on November 9, just after the first news about the Allies' landings, 'Bonnefous is sent to Istres to find out about possible departures. He returns on 10th at 10 p.m.: the disorder and congestion on the airfield are such that a quick connection to and from Algiers is unimaginable.' (Paillole, Résume). The same chaos and congestion did not hinder the planned departure to Algiers of the entire team of the air force intelligence service, with its commander, Colonel Ronin. It was to this episode that Bertrand referred to in his memoirs, when he expressed his disappointment about the order of his superiors, that did not allow him to place a Polish team on board as well. Bertrand's unidentified superior was to decide that only officers, presumably French, were to be evacuated (les officiers d'abord). In fact, the evacuation covered a slightly wider group: 'Colonel Ronin succeeded, when the time came, in transferring Air Intelligence en bloc by flying it from Istres to Algiers. With him, incidentally, he took other members of the Special Services who were particularly sought after by the Germans' (Stead 1959).
Bertrand was undoubtedly in a predicament. On the one hand, he was aware of the need to evacuate the codebreakers and knew about the request of the British. On the other, the order to stay in France along with other TR structures concerned not only him personally, but also the team he led. The chaos accompanying the evacuation of the Vichy services to Africa probably meant that there was no one to take care of giving Bertrand instructions to make an exception for the Poles. On a strictly personal level, Bertrand must have also been in trouble.
After the Spanish team had been evacuated to Africa, the Poles were his raison d'être in the structures of the French special services. To let them evacuate meant to lose his position in the only environment that mattered to him. Anyone who knew Bertrand, either directly or through what he did or said, could not doubt his decision: he let events take their own course.
After the opportunity to evacuate the team to Africa by air before and after the Allied landings in Africa had been wasted; as the period of poor surveillance of the coast just after the occupation had been missed, the situation became significantly more complicated. The heads of the Polish team pressed for evacuation, and its members were waiting, dispersed in the Italian occupation zone. Bertrand signaled the impending evacuation aboard the submarine several times, but each time the operation was canceled at the last minute.
It is interesting to learn why the evacuation by submarine, vigorously rejected in the period preceding German occupation of Vichy France, was accepted now? The answer seems to be very simple; this time it was supposed to be a French, and not British submarine. The descriptions of the evacuation plans in Bertrand's and Rejewski's memoirs correspond exactly to the fragments of Paillole's report, presenting the operations he was planning in Algiers with the participation of the submarine Casabianca, one of the French ships that managed to leave Toulon. Casabianca under command of Capitaine Jean L'Herminier was to deliver men and equipment for the French underground to the landing site between Cannes and Nice, and pick up evacuees from there. However, the original plans had to be delayed and modified: 'two TR officers reconnoitered the point selected: it was occupied by Italian troops and the whole coast between Cannes and Nice was bristling with defense-works and guard posts' (Stead 1959).
It will be understandable for the reader to ask how Paillole could plan these operations in Algiers, since we said goodbye to him in France, where he declared to remain in the occupied country? Quoting his own report: 'on 17 [November] in the morning, departure for the Pyrenees. While the details of the clandestine border crossing are worked out by the local TR post, Paillole goes to Toulouse'. And then: 'on the night of November 28 to 29, 1942 the border is crossed at Puigcerda in the company of Villeneuve and Poniatowski' (Paillole, Résume). From (Stead, 1959) we learn more details of his escape: 'travelers dined at an inn at La-Tour-de-Carol (…), they slipped into a friend's house 100 yards from the frontier. A guide was waiting for them and he led them without undue incident to Puigcerda'. This evacuation route was managed by Ramonatxo brothers from Perpignan, and its functioning was presented in the book written by one of them (Ramonatxo, 1955). From Puigcerda Paillole got to Barcelona, Gibraltar, and further on through London to Algiers, to secure his position in the power struggle between the supporters of General de Gaulle and General Giroud.
In his later book Paillol clearly refers to the evacuation of the Polish team and the route he had used himself: 'Poles, left to their own devices, resorted, in circumstances unknown to me, to a network of Pyrenean smugglers controlled by the enemy. It is a serious fault of our house which had absolutely safe channels of passage to Spain' (Paillole, 1959). This statement does not fully reflect the facts. The details of the evacuation described in Rejewski's memoirs fully correspond to the realities of the route covered also by Paillol. As far as we know, the team under the command of Wiktor Michałowski followed the same trail. Due to Antoni Palluth's death in a concentration camp, we will never know the circumstances of his, and his companions' capture. The controversy mainly concerns the circumstances of the capture of the group including both officers commanding the team of Polish codebreakers.
According to Bertrand's (and Paillole's) account, after the fiasco of the first attempt, the Poles rejected the guides provided by TR and were betrayed by members of the other smuggler group they had chosen on their own. Langer and Ciężki's version has been reported in the last sentences of section 1 above. Which version is closer to the truth is decided by a detail in Bertrand's later book and Langer's post-war memories. Bertrand tried to put the blame for the failure of the evacuation on Langer, arguing that the Poles had rejected the services of smugglers provided by his organization. At the same time, both in his book and in a post-war conversation with Langer, he confirmed having paid the smugglers' price after they delivered the other half of the bill signed by Langer (confiscated by Germans after the officer was captured). This lie, rather unprofessional for an intelligence service officer, confirms that Langer and his companions were led into the ambush by a guide hired by Bertrand, or by someone supervising their evacuation on behalf of TR.
In the described circumstances, the natural question is whether the failed evacuation of a significant part of the Polish team was caused (according to Paillole's assessment) by tragic negligence on the part of the French services, or was it a conscious settlement of scores with a partner who was no longer useful? 'Thou shalt not kill, but need'st not strive officiously to keep alive' 2 . Ciężki's aggressive attitude towards his French fellow prisoners during his internment at the Eisenberg castle and the unambiguous note in Langer's post-war letter to his wife (at the author's disposal) clearly confirm that they both pointed, not surprisingly, to the latter possibility. The fact that the capture of Langer and Ciężki did not lead to the disclosure of the Enigma secret, the Allies owe only to the tactics of both officers, adopted during the interrogations in March 1944 (Grajek 2017).

Somewhat
paradoxically, if Bertrand intentionally or negligently handed over the heads of the Polish team into the German hands, he had the right to assume that he was not exposing the Enigma secret itself. When in an earlier exchange with London he insisted that the current cipher keys be sent to P.C. Cadix, Denniston instructed SIS to reply that this was impossible, as BP was currently unable to recover them. As a consequence, Bertrand may have believed that he was only jeopardizing his disliked partners, but not a secret vital to the Allied cause.

Appendix
The description of the difficult relationship between Bertrand and the group of Polish codebreakers is a good opportunity to draw attention to its aspect with long-term consequences. It is natural to ask why haven't been Rejewski and Zygalski invited to join Bletchley Park after their arrival to London in mid-1943? According to one of the possible explanations the secret of Enigma could not be entrusted to people arriving from territory controlled by the enemy. Another explanation was that the British counter-intelligence believed Bertrand's organization to be infiltrated or even inspired by the Germans. The sum of information 2 Arthur Hugh Clough, The Latest Decalogue originating from two independent and unrelated sources might indicate that such speculations were not completely unfounded.
In his memoirs, Bertrand describes his contacts with an unidentified representative of the German embassy in occupied Paris, who was willing to share some useful information with the French underground. He referred to his interlocutor as "Max". In his memoirs, Oskar Reile, head of the Abwehr's post in Paris, mentions an attempt to infiltrate the French underground through two Germans publicly demonstrating their disillusionment with their own country's policies. He refers to his honeypots as "Max" and "Moritz" (Reile, p. 216). If "Max" from the memoirs of both officers represented the same person, which is probable, Reile was dangerously close to discovering the Enigma secret, and was prevented from the success only by Allied landings in Africa, which forced Bertrand into hiding.

Summary
Functioning of P.C. Cadix was an exemplification of the complicated British-French-Polish relations after the defeat of France. None of the partners involved appears to have been fully sincere, and acting in good faith. The British simulated cooperation with Bertrand, trying to prevent the transfer of the Enigma secret into German hands. Bertrand was given by his Vichy supervisors unprecedented permission to contact SIS not to spy for the British, but on them. Polish HQ tried not so much keep a group of codebreakers in France, as simply away from London.
Work carried out in such conditions (not to mention the lack of technical equipment) could not bring spectacular results. The prospects for evacuation in the face of Allied landings in Africa were not much better. In a lecture given to a group of TR officers on June 6, 1942, Paillole described the priorities of his service as follows: 'Germany is danger number one [and] England is danger number two' (SHAT 1942). The safe evacuation of the Polish codebreakers to London was in conflict with both the political priorities of the service that was supposed to assure it and the personal interests of Bertrand, who was supposed to oversee it directly. In the described conditions, the successful evacuation of at least the professional core of the group should be considered a success of the operation.