Between Silk and Cyanide (19 page)

Heffer had once asked me to define a good security risk and I'd replied, 'Someone who knows whom it's safe to be indiscreet to.' If there was slightly more truth in this than in most pat responses, then a bad security risk was somebody likely to confide in the wrong 'safe someone'. None of us knew whom the coders talked to in their off-duty hours.

An idea occurred to me in mid-swing: there was something which might remind them for the rest of their coding lives that they must talk to no one, and the more I considered it, the more promising it seemed.

I realized that my next-door neighbours were watching me from the balcony, and that if they were the feather-donors I owed the idea to them. I waved my gratitude without falling in.

The Dutch section was more determined than ever to bring Jambroes back to London for consultations.

Four agents now in the final stages of training were going to parachute into Holland in the February moon period to help him to cross the Spanish escape route into France and Belgium. Their code-name was to be Golf, N section having exhausted its supply of vegetables. I was due to brief the Golf team within the next fortnight.

Messages continued to arrive from Holland reporting the steady success of Jambroes's organization and the build-up of the Secret Army. The encoding was perfect.

Signals directorate's night-light returned to Baker Street and Heffer was at the head of the long queue waiting to be guided by it.

Nicholls occupied a small office in Norgeby House which I hoped was temporary. He took my report from me before I could say good morning and at once began reading it. He was interrupted a few mintes later by a call from CD asking to see him immediately. He promised to finish the report by the end of the day and to send for me a soon as he had considered it.

Three days later I still hadn't heard from him. It seemed a very long time for the jury to be out. When he finally sent for me on the morning of the fourth day Heffer was sitting opposite him in watchful trance. My report, now covered in red-ink annotations, was open on the desk. There was a large grey folder beside it with Most Secret printed on its cover in block capitals. I couldn't remember seeig one like it in SOE. Next to the folder was a map (which I saw to be of Holland) with coloured pins stuck all over it.

Nicholls was studying this map with his eyes closed. Heffer was studing Nicholls. I edged forward and tried to read the red-ink notations. They looked like a spider bleeding to death. I had an immediate rapport with it. It was hard to believe that I was watching a highly trained Signals mind at work, but I knew that this stylistic violence was Nicholls at his most productive.

Nicholls opened his eyes and shot his first question at me:

Had I established whether the Dutch WT operators were transmitting from their usual operating posts or were they sending their messages from new sites? If so, which were they?

It hadn't even occurred to me to make direction-finding enquiries and I undertook never again to criticize my Signals colleagues for insularity. Heffer nodded his approval.

Nicholls picked up the grey folder without comment and read out a list of the WT operators who'd previously been transmitting from The Hague, Rotterdam and Amsterdam but who were now sending their messages from Eindhoven, Utrecht and Arnhem. He quoted the districts in these towns from which the traffic was being sent as well as the date of each transmission.

These precise readings could not possibly have come from Grendon, whose direction-finding facilities were rudimentary, and I was even more curious about the source of that folder.

Nicholls now proceeded to minimize the significance of his discovery.

He had read all the Dutch traffic and was satisfied that the transmissions from Eindhoven, Utrecht and Arnhem coincided with the agents' accounts of their movements around Holland. The new transmission sites were not therefore grounds for suspicion. He would confirm this with the Dutch section at a general discussion on the WT situation in Holland.

It almost cost me a lip not to interject that the 'accounts of movements' around Holland might be coming from the Germans and that the new operating sites strongly reinforced the grounds for suspicion, but a glance from Heffer warned me to let Nicholls finish.

He agreed that the linking together of three circuits of agents was appalling security, especially in WT terms, but doubted if this practice was confined to the Dutch section. However, the Signals directorate had no jurisdiction over the way that a country section dispersed its agents as decisions of this kind were essentially operational. But he was going to warn the country sections of the dangers of WT operators being in a position to compromise the Signals security of other WT circuits. He would make particular reference to Ebenezer, Boni and Trumpet when he had his general discussion with the Dutch.

This was the second time he'd spoken about a 'general discussion', and it worried me. There was nothing more specific than a lack of indecipherables, yet it hadn't once been mentioned. A glance from Heffer warned me to be patient.

Nicholls next fastened with a technician's relish on the plain language messages broadcast to Holland over the BBC's Radio Oranje.

How and why did SOE use these messages? Was there a standard procedure?

I said that all country sections had 'Voice of Freedom' facilities from BBC to broadcast plain language code-phrases to their respective territories. The rival French sections were the most prolific broadcasters and shared a BBC programme called 'Les Français Parlent Français', which was about all that they did share. Other country sections used similar programmes for the same purposes. If an agent such as Peter Churchill, for example, wanted to borrow a large sum money in the field and the prospective lender doubted whether the debt would be honoured by the British government when the war was over, Peter could invite the prospective lender to make up a sentence known only to the two of them, which the BBC would subsequently broadcast on 'Les Français Parlent aux Français'. This usually the only verification which the lender required. Plain language phrases were also used by the country sections for last minute confirmations or cancellations of impending operations. My anxiety about these phrases was that their wording was arranged in peom-codes. If the Germans recorded them and matched them with code-groups which they intercepted, it would make the poem-code much easier to break.

Nicholls pointed out that there were a great many plain language messages and that the Germans would have to be very skilful indeed pinpoint which messages to anagram. I asked if there were any reason to suppose that they lacked this skill. He replied that there was't, but there was every reason to suppose that they lacked the man-power. He conceded the security risk and I said that the real answer was to change the poem-code.

He reminded me sharply that this was not the object of the present session and I apologized for its irrelevance.

He then glanced at my report and, with the persistence of a steam engine capable of flight, asked why I had put an asterisk against the Dutch section's en clair message to Vinus (I had forgotten to erase it)

I explained that this message was intended to confirm to Vinus that he could introduce Akkie to the Committee of Resistance. If the Germans were reading the traffic and Akkie was taken to the Committe of Resistance, how far behind him would the Germans be?

Nicholls stressed that these were operational matters for the Dutch section to decide upon and begged the question as to whether the traffic was being read. The Signals directorate couldn't interfere with any country section's right to broadcast messages to its agents via the BBC.

I replied that I wasn't concerned with the country sections' right to broadcast but with the agents' to survive.

He looked at me thoughtfully and I tried to make a draw of it. We both knew that the moment had come for the main event and I was fascinated by the way he prepared for it.

He read and then re-read the last page of my report, which included the statistics of other country sections' indecipherables. He looked at the pins spread across the map and seemed to be searching for one in particular. He referred to his folder and then looked again at the map.

He wasn't playing for time. To all of us in Signals, he was time. But from the very start of the meeting I'd sensed a private conflict of some kind in Nicholls. I also sensed that Heffer knew what it was. I glanced at Heffer. He was deep in Nicholls's thoughts.

Nicholls closed the report, the folder and his eyes. His next question opened mine: When an agent in the field sent an indecipherable message, did I regard this as proof that he wasn't caught?

I assure him that I didn't.

'In that case, why are you trying to prove the converse?'

I respectfully suggested that this question was a soriticism.

'A what?'

I quoted the classic example of a
reductio ad absurdum:
'If a man with one hair is bald, then a man with ten thousand hairs must be ten thousand times as bald.'

'You haven't answered the colonel's question,' said Heffer, his first contribution to date.

Before answering it, I tried to imagine how it felt to be in a prison cell in Holland hoping that someone in London was awake. I put to Nicholls that, when an agent sent an indecipherable, I didn't take it to mean that he was in good health but that his coding temperature was normal. But the Dutch were behaving abnormally by any known standards of code conduct, and I'd left it very late in the day to spot it. If Nicholls had lost confidence in me, then I must accept the consequences. But if he still retained some, then he must accept that there was something grievously wrong in Holland for not one Dutch agent to have made a mistake in his coding. So wrong that only the normal or the purblind could continue to ignore its implications.

It is not begging the question as to whether the Germans were reading the Dutch traffic. I was begging for answers which would satisfy me that they weren't. Until those answers were forthcoming, I would continue to believe that the Germans were sending at least half of the traffic and nothing was going to budge me from that position, least of all a polemic.

I couldn't tell from Nicholls's expression whether I had said too much or too little. Heffer was studying the map.

Nicholls lifted his telephone so sharply that I wondered if I was about to leave SOE the way I'd arrived, under armed escort. He ordered tea for three. It would be undrinkable but it didn't seem opportune to offer him Mother's.

Heffer decided that it was time to join the party and (typically) asked the most perceptive question yet: if the Germans were running raffle, they were doing it very skilfully—so surely it would have occurred to them that they must send us some indecipherables to alay any suspicions which London might have?

I replied that perhaps we weren't the only ones who made mistakes.

How did I know the calibre of the Germans we were up against in land or even who they were.

As casually as if he were telling us the date, Nicholls said that we were likely to be up against Giskes, a most experienced Intelligence officer. I asked whether Herr Giskes was experienced in the ways of poem-code. Nicholls assured me that he was, and invited me to explain why he hadn't sent us any indecipherables.

I said that Herr Giskes and his assistants must have decided that any signals organization which was stupid enough to use the poem-code wasn't likely to be bright enough to notice the absence of indecipherables and was even less likely to draw the proper conclusions from their absence. That put a stop to what was becoming a chat. Clasping both hands behind his head and looking like a radio mast about to earn its keep, Nicholls delivered his summation.

My report had drawn his attention to what was potentially a disastrous WT situation in Holland, if it weren't one already, and had given him a great deal to think about. He was most concerned about the lack of indecipherables and could offer no explanation for it—but still couldn't decide if I were lending too much weight to it.

However, it certainly could not be ignored and he was most concerned about what practical steps could now be taken. We still had no positive proof that any agents had been caught. Nor could I even say how many agents I believed were caught or specify their names. I could only put question marks against some of them, and question marks were not enough for an organization which hadn't been educated in WT security, a deficiency which he would shortly remedy.

The immediate problem was that if my theory had substance, all the Dutch agents might be caught. What was he to recommend to the Dutch section? That it should cancel all its operations or only some of them? And which ones? And for how long? Until more tangible evidence could be found by the Signals directorate? It just wasn't on. The most he could do until this evidence was found was stress the Signals directorate's concern about the general WT situation in Holland and the overloading of operators such as Trumpet, Boni and Ebenezer. We would also press the Dutch section to explain what steps they took to confirm that their agents weren't blown. But if I or anyone else in Signals found anything which could be regarded as proof positive, he would raise it at the highest level.

To me he was the highest level, but it wasn't getting me very far. I told him that I very much wanted to examine the early codegroups which had been received and decoded by C's wireless station before June '42 so that I could establish the length of the keys the agents had chosen and see if there'd been any significant changes in their coding style.

He promised to do his best and made a long note on his pad which seemed out of proportion to the request. He then proceeded to develop his ideas about educating the country section in Signals security but I was no longer listening.

Nicholls had paid us the compliment of thinking aloud and I responded to it by thinking about him. The breadth of his knowledge was extraordinary. He knew where to get information about the location of WT sites, he knew the latest direction-finding techniques, and he knew Giskes's name. I admired him more than ever and he probably knew that too. Yet almost all his observations had been slanted from the viewpoint of a wireless expert, with codes a poor second. And I was convinced that there was some problem to do with the Dutch which was greatly worrying him but which he wasn't going to mention in front of me.