I tried to deploy the technological breadth of the internet as best I knew. I set up a small website with a plea for information, and linked it with several larger, busier sites. I wrote an email appeal containing the few facts that I possessed and bulk emailed it to all my friends and acquaintances, asking them to send it on to all their friends and acquaintances. The message snowballed out from my computer in a real-time effort to test the theory that everyone in the world was now only six degrees of separation away from everyone else. One of my friends replied that he was sending it on immediately to his address book of 5,500 people. At low moments in my search, I would console myself that my message was being forwarded through cyberspace and would eventually end up on the screen of someone who had known Maurice or his mother.

A few weeks later I received an email from an American academic querying whether the whole thing was a mathematical hoax.

What is given are ages (13), standard rebus clues – décédé, deceased – which is DCD, 900 in bad Latin numerals, and names: Maurice and Christine (which makes me think of Christine de Pizan, or Christine de Suède, or Maurice de Saxe). The rest of the main message is a triple oxymoron, reminiscent of surrealist poetry: you can’t ‘extinguish rest’, ‘rest’ can’t be ‘tireless’, and ‘outstretched arms’ aren’t at ‘rest’. So seeing as there’s nothing else to speak of, the whole thing looks like a puzzle or a test or a brainteaser to me. It is not a moving message, it is a totally implausible piece of French put together from probable allusions, quotations and fixed phrases. I haven’t cracked the code. But I will eat my hat if this is not part of a plot of some kind.

All this seemed unconvincing to me – I thought 900 would be CM in Latin numerals; I didn’t see what a fourteenth-century poet, an eighteenth-century general and a seventeenth-century Catholic convert would have in common – and if we were looking for name associations, why not Maurice Chevalier and Christina Aguilera? However, I was open to any possible source of help so I immediately contacted the American and asked whether I could send him the whole message. He agreed and, once he had studied the complete letter, wrote to retract his original analysis. ‘Sorry for my previous suspicions – this is no coded prank.’ His new analysis agreed with mine, which was reassuring but also a pity, since the rebusian triple oxymoron had held out the promise of an interesting intellectual challenge.

Meanwhile I redoubled my internet efforts. I found the French equivalent of Friends Reunited, Les amis sont réunis, and asked them to put out an appeal. The automatic response asked me for my school and years of attendance.

I trawled the net for sites referring to letters in bottles, and was amazed at the differences between them, differences that seemed to epitomize the disparity in international cultures. The British site, www.letterinabottle.com, was very down-to-earth.