The Man Who Broke Napoleon's Codes (35 page)

Wellington and the staff went back into Salamanca, where heavy guns were set up, and steps were prepared for the siege of the San Vicente fortified convent. For the general, there was an opportunity to receive his secret correspondent, Father Patrick Curtis, and show him the respect due to such an invaluable spy. Some officers questioned the wisdom of this, for once the hysteria of Salamanca's liberation calmed down, it became apparent that a pro-French party still existed among its people and they would surely get to hear of Curtis's visit. Many of the staff simply collapsed into bed after days on the San Christoval position with, as one of them complained, never more than four hours' sleep a night.

While Marmont had come forward to challenge his enemy, Don Julian's guerrillas had once more made their way around the flanks and rear of the French army. More captured messages were brought into town.

A few days in Salamanca had allowed Scovell to push his calculations forward considerably. He was working from dozens of dispatches, some of them up to twenty pages, and it was most likely in some smoky billet
in this Spanish university town that he was able to draw together, for the first time, all of the discoveries he had jotted on scraps of paper and the deciphering table.

Records suggest that the assistant quarter master general had entered values for about 450 of the 1,400 cipher numbers by the end of June. Even though a small number of Scovell's suppositions were actually wrong, this was solid progress, and represented rather more than one-third of the code, since there were a few score numbers that had never been filled in by the French on their enciphering tables, particularly in the series 1,201–1,400.

With his knowledge of the earlier ciphers he had worked on, Scovell would have known that a code number that did not appear in any of the messages was not really worth worrying about. There were perhaps one hundred entries in the cipher standing for places or terms irrelevant in Peninsular warfare that fell into this category. Vacant code numbers did find their way into some of the messages, but Scovell would already have spotted many of them.

But viewed from the perspective of someone who was not a specialist decipherer, and who was a perfectionist to boot, Scovell's fold-out crib showing his decoding efforts contained a disturbing number of blank spaces and crossings out. Wellington resolved that he must obtain the services of the best decipherers in Britain to join the attack on the Great Paris Cipher. Wellington felt it was time to share Scovell's early work with London. Accordingly he wrote off to London on 25 June, enclosing Scovell's early results and copies of three letters in the
Grand Chiffre.
The general felt that much of the
Grand Chiffre
still remained a mystery to him and he thought that Whitehall would also find it useful, since “it is the same cipher used by the Emperor's ministers, and the discovery of the key, therefore, may be important for other objects as well as for our operations.” In this last suggestion, the general showed his limited knowledge of French secret writing, for the
Grands Chiffres
being used by Napoleon in eastern Europe, for example, were based on sheets of the 1750 diplomatic table and made up with completely different ciphers.

Since the commander of forces had enclosed Scovell's work to date, we have an excellent insight into what the staff were able to make of the dispatches brought into Salamanca at the end of that second week of the
campaign. A tiny scrap of paper concealed in the clothing of a Spanish messenger revealed Marmont's thoughts as he marched from the San Christoval position:

Using the codes already discovered by Scovell, this became:

The first unknown construction
e
53
es
is not that difficult to guess, the code 53 stands for
gal
and forms the word
equal
in English. Similarly 1252 seems to be the commander of the Army of the North. Although the last sentence contains some unknowns, the sense of this passage emerges: Marmont was telling his superiors that he would avoid battle until he was reinforced. The fragmentary nature of these partial decipherings may have irked Wellington, but he was sufficiently confident of
what had been discovered to tell London on 30 June, “I know from intercepted letters that Marshal Marmont expects to be joined by a division of the Army of the North.”

The stalemate between Marmont and Wellington was broken on the twenty-seventh when the defenders of the San Vicente convent finally capitulated. The British had started firing red-hot shot at the roof of the massive building and succeeded in setting it on fire. Even though they had survived one attempted storm and the British siege guns were making little impact on their walls, the defenders had no choice but to surrender. With the forts fallen, the Army of Portugal no longer had any reason to be lingering in such a dangerous and exposed location as the plains of the Tormes, particularly as Wellington would now be free to bring his entire armament into action.

So Marmont ordered a withdrawal about forty miles north to the line of the River Douro (which, in this higher part of its course is called Duero by the Spanish). Wellington's troops swiftly followed them.

The landscape, with its tall fields of billowing corn, vineyards and orchards, seemed lush and wonderful to British soldiers who knew only the barren moorland of the Beira Baixa frontier. There were other pleasant surprises too. The local Villa Verde wine, made from a green grape rather like the whites of northern Portugal, was most refreshing. Officers riding to and fro with dispatches or on reconnaissance discovered that most villages had a café where they could buy a bottle of wine and soothe their parched mouths with lemon ice cream.

For the rank and file, unable to fall out of their column of march to go in search of such delicacies, the heat became a severe trial. Most men felt the backs of their necks, their noses and foreheads burn and blister. The hot dry breeze chapped and broke their lips. In order to try to save his men from the effects of marching all day in this heat, Wellington set predawn reveille even earlier. Major William Warre wrote to his father about the consequences:

The Army of Portugal's staff noticed a new energy about their commander as they fell back to the Duero line. While it was evident that this great natural barrier would give his army a measure of security, many also assumed he knew something about possible reinforcements. General Foy had a different explanation: he believed the strain on his commander was becoming increasingly obvious. While close to Wellington, the marshal had been “cold and apathetic,” but as soon as they put some distance between the two armies he was once more “ardent and enterprising.” The divisional commander wrote of Marmont, “He did not want to deal with problems and always looked for ways around them.” The marshal's tendency to switch between introspection and bravado had its reflection in the way he ciphered his letters: proud defiance was broadcast
en clair,
real insights into his strategy were locked in code.

As he fell back toward his new field headquarters on the north bank of the Duero at Tordesillas, Marmont fired off letters to King Joseph and General Louis Marie Caffarelli, the new commander of the Army of the North (Dorsenne, his health broken, had gone home to France, where he died of tetanus). Marmont stressed the absolute need for reinforcements, particularly of cavalry. One of these messages, from Marmont to Joseph and dated 1 July, was captured. In taking up his new defensive line, Marmont explained:

Looking at this message with Scovell's discoveries about the
Grand Chiffre:

The first passage of code makes up
reinforcements
and the code 922 that eluded Scovell stood for
au
which joined to the next code (311, which he appears to have known) made
auront,
or
will.
It is quite likely that this context would have allowed a good guess at 1345, which meant
cavalry.
Marmont had in fact also asked for five or six thousand infantry, but this is where a previous deduction led Scovell into a mistake, albeit not one of enormous significance. The information sent to London on 25 June showed he thought the code number 135 meant
six,
when in fact it meant
seven.

Notwithstanding the gaps in the decipherer's knowledge, this message would have given Wellington a clear sense of what kind of reinforcement Marmont needed to receive before he resumed the offensive. It was not entirely obvious from the letter whether by thousands of extra infantry he meant Bonnet's long-expected division, but this was the assumption in headquarters.

There were two other most important pieces of information in the letter. The first was a complaint by Marmont that not one soldier of the Army of the North's promised reinforcement had arrived in Valladolid, to his rear, by 28 June. The other was this humble sentence: “I dare to beseech your Majesty if you can, send me your cavalry from the Tagus.” This was most significant because it allowed Wellington to begin unlocking the secret of where both Joseph's Army of the Center and D'Erlon's corps to the south might be, and what their orders were. If Marmont succeeded in drawing up the former to unite with his troops, then D'Erlon would be left too weak to mount a major strike against central Portugal.

Marmont's letter of 1 July also presented almost comically the
French commander's desire to say one thing to the British who read his mail and another to his masters in Madrid. It contained the passage:

Scovell's grasp of the cipher revealed the coded passage to mean “if I receive the reinforcements that I have asked for.”

By 5 July, Wellington's headquarters had settled into Rueda, a pleasant town in the wine country a few miles south of the Duero. The British put scouts close to the river but kept their divisions a few miles back from it, denying the French a chance to study their deployment. Marmont's forces, on the other hand, were deployed in strength, ready to receive anyone who attempted to cross this watery line.

Once again, Scovell had a chance to stay put for a few days and set his mind to attacking the cipher. His every waking hour consisted of trying to please his commander and he knew that if he could deliver some breakthrough in deciphering, Wellington would be gratitude itself.

Mulling over the information extracted from Marmont's coded letter of 1 July, the general must indeed have been satisfied, because it had given him new insight into French dispositions. Marmont's complaint about the nonappearance of any troops from the Army of the North helped convince Wellington, along with other reconnaissance reports, that there would not be any substantial reinforcement from this quarter. Some cavalry had been sent off by Caffarelli, and some intercepted uncoded messages helped the British keep track of its progress. Since Marshal Soult would not help either, this left only the force under Madrid's direct control, the Army of the Center, to provide substantial succor to Marmont. Although Wellington had received reports that some of these troops had started marching, he did not know where. They might be intended to join with the Count D'Erlon in some diversionary raid. Now he had clear evidence of Marmont trying to draw the Army of the Center's cavalry into his own field of operations in front of the British main army.

As the hot July days passed on the Duero line, the outposts kept a vigilant watch on French positions and one false alarm followed another of an impending French assault. Wellington could only hope that the guerrillas operating on the other side of the river would warn him of the arrival of any new French forces. Better still they might bring him some dispatch that might tell him how long he had to bring Marmont to battle before that marshal's constant pleas for help finally produced an army that was larger than his own and he would have to turn back toward the Portuguese frontier.