The Moses Stone

'But Margaret O'Connor took pictures of an argument in the souk. Couldn't the killers have wanted to silence her for that reason, and stole the camera to remove the hard evidence?'

'That might well be a part of it,' Bronson conceded. 'It would explain why the camera and memory stick weren't found in the wreckage. But unless Margaret O'Connor threw away the clay tablet before they left Rabat, somebody took that as well.'

'And you don't think she just chucked it away?'

'No. Kirsty told me her mother was going back to the souk the next morning to return the tablet to the man – the Moroccan – who'd dropped it, and if she couldn't find him she was going to take it back home with her as a souvenir of their holiday. She put all that in the email she sent to Kirsty the evening before she and Ralph left the hotel. But by then the Moroccan was lying dead outside the medina with a stab wound in his chest – Kirsty got a final message from her mother the following morning, telling her she'd actually seen the dead man. Talabani's confirmed he was one of the people Margaret O'Connor photographed.'

'Margaret didn't say what she was going to do with the tablet, though?'

'No. Her last message was very short, just a couple of lines, probably sent while her husband was paying the hotel bill or getting the car or something.' Bronson paused and leant forward. 'Now, the tablet. What did you manage to find out about it?'

'As I told you on the phone,' Angela replied, 'it's a lump of clay of almost no value. The writing is Aramaic, but Baverstock told me he could only translate one line. And I think he was probably being honest in that at least, because he knows I can read a little Aramaic. If he was trying to mislead me, all I'd have to do to check that would be to compare his translation with the original.'

'And have you?' Bronson asked.

'Yes. I looked at a couple of the lines on the photograph, and I came up with the same words.'

'OK,' Bronson said grudgingly, 'for the moment, let's assume he is being accurate. Tell me what he said.'

'On that single line of text, the words are clear but they don't make sense. I've got a translation of that line and another couple of words written out for you.'

'Is there anything special about the tablet? I mean, anything that would make it worth stealing, let alone killing someone because of it?'

'Nothing. Baverstock found a part of a word that might refer to the Essene community at Qumran, but even that's not conclusive.'

'Qumran? That's where they found the Dead Sea Scrolls, isn't it?'

'Yes, but that's probably irrelevant. As far as Baverstock could tell, the tablet didn't originate at Qumran, but simply mentions the place. What's interesting is that one of the few other words he translated was "cubit".'

'And a cubit was what?' Bronson asked.

'It was a unit of measurement equivalent to the length of a man's forearm, so it was pretty variable – there were at least a dozen different sizes, ranging from the Roman cubit of about seventeen inches up to the biggest, the Arabic Hashimi cubit of nearly twenty-six inches. But the fact that there's a mention of a cubit could mean that the tablet is written in a type of code, and it might be indicating the location of something that's been hidden. Maybe that's why it's important.'

'Let's face it,' Bronson said, 'if Baverstock was being accurate, the inscription has to be a code of some sort. Nothing else makes sense.'

'I agree. Here' – Angela opened her handbag and fished around inside it – 'this is the translation of the Aramaic.'

Bronson took the single sheet of A4 paper from her and quickly scanned the list of about half a dozen words.

'I see what you mean,' he said, looking at the text more carefully. 'Did Baverstock think this might be encrypted?'

'No, but his field of expertise is ancient languages, not ancient codes, and that's something I do know about. The good news is that we're dealing with an object that dates back around two millennia. And that's good because although there are very few known examples of codes and ciphers from that period of history, those that we do know about are very, very simple. The best known was probably the Caesar Cipher, which was allegedly used by Julius Caesar in the first century BC to communicate with his generals. It's a really basic monoalphabetic substitution cipher.'

Bronson sighed. He knew that Angela had done some research on cryptology as part of a project at the museum. 'Don't forget I'm just a simple copper. You're the one with the brains.'

Angela laughed. 'Now why don't I quite believe you?' She took a deep breath. 'To use a Caesar Cipher, you write out the message as a plaintext, apply whatever shift you've selected to the alphabet, and then transcribe the enciphered message.'

Bronson still looked blank, so Angela moved her empty plate to one side and took a piece of paper and ballpoint from her handbag.

'Let me give you an example. Say your message is "move forward",' she said, writing the words in capital letters on the paper, 'and the shift is left three. You write out the alphabet, then write it out again underneath, but this time you move each letter three places to the left, a socalled left rotation of three. So you'd find "A" directly above "D", "B" above "E" and so on. In this case, the enciphered message "move forward" would read "pryh iruzdug". The obvious problem with this method is that every time a particular letter appears in the plaintext, the same enciphered letter will be in the coded message. So in this example, which is only two words in length, two letters – the "R" and "U" – are repeated, and somebody trying to decrypt the message can use frequency analysis to crack it.'

She looked hopefully at Bronson, who shook his head. 'Sorry, you'll need to explain that as well.'

'Right,' Angela said. 'Frequency analysis is a simple method of cracking a basic code. The twelve most common letters in the English language, in order, are "E", "T", "A", "O", "I", "N", "S", "H", "R", "D", "L" and "U". I remember it as two words – "ETAOIN SHRDLU". And you probably already know the most famous example of a Caesar Cipher.'

'I do?' Bronson looked blank and shook his head. 'Help me out here.'

'2001,' Angela said, and sat back in her chair. '2001 – A Space Odyssey. The sci-fi film,' she added.

Bronson frowned, then his expression cleared. 'Got you,' he said. 'The film-makers didn't want to use the acronym "IBM" for the computer on the spaceship, so they came up with the name "HAL", which, if I've understood you correctly, is a Caesar Cipher with a right rotation of one.'

'Exactly. There's another slightly bizarre example,' Angela said. 'The French "oui" becomes the English "yes" if you apply a left rotation of ten.'

'Do you think anything like that is probable in this case?'

'No,' Angela replied, 'and for one very simple reason: we can read the Aramaic words on the tablet. One of the obvious problems with a Caesar Cipher is that every word of the enciphered text is invariably gibberish, which is the biggest clue that the message is encrypted. That definitely isn't the case here.'

'What about other kinds of ciphers?' Bronson asked.

'You've got the same problem with all of them. If the individual words are encrypted, they cease to be recognizable as words and end up as collections of letters. The Aramaic words on this tablet' – she tapped the paper in front of Bronson – 'aren't encrypted. But that doesn't mean there isn't some kind of message hidden in the text.'

'You need to explain that,' he said, 'but wait until we're back on the road.'

'Just wait in here a second,' Bronson said as they reached the door of the restaurant. 'I want to check there's nobody waiting for us out there. Then I'll bring the car over.'