1996: 7

1995: 3

1994: 9

1993: 7

1992: 7

1991: 8

She also told me that in a study carried out in 1995, 90 per cent of drownings happened during the months of July or August, and 86.1 per cent occurred in the sea. She referred me to another government statistical institution that, she assured me, could provide more details.

In America, drowning is the fourth leading cause of accidental death, claiming 4,000 lives each year, one-third of whom are children under the age of fourteen; and for children under the age of fifteen, drowning claims the second-highest number of lives. That works out at 1,333 children drowned each year, which is proportionately far more than the French figures I had been given. Something seemed out of kilter here.

I had no way of telling whether Maurice had drowned in France, or even within the last decade, but I tried not to think about that. I was encouraged by the woman’s response to my plea for information. Although she obviously thought the case hopeless, she was touched by the story. And if a government bureaucrat’s heart could be stirred to spend several hours on the search, then it was not just my friend the dog-walker and me, two sentimental and over-imaginative mothers, who found the letter moving. The call to INED had been the very first step in my quest, and although I was in reality no further forward, the official’s sympathy had provided me with a much-needed moment of encouragement before embarking on an uncharted sea of unforgiving statistics. The story of Maurice was not just a private obsession, it had a more universal resonance, and if this woman’s reaction was symptomatic of what I would come across in my search, then others would be equally helpful. The case, though undeniably tricky, was well on its way to being solved.

The next institution, INSERM (National Institute of Health and Medical Research), dealt in more detail with the same statistics. However, it seemed that the INED figures for drownings of young boys – those aged between ten and fourteen – were only those which were classified as e984 (‘Undetermined’); in other words, those deaths that could not be classed as e910 (‘Accidental drowning and submersion’), e954 (‘Suicide and self-inflicted injury by submersion’) or e964 (‘Assault by submersion’). The total number of drownings per region was in fact much higher, although it was still substantially lower than in America. The INSERM official kindly sent me the figures for all drownings, whether accidental, self-inflicted or by assault, which added up to between seven and fifty-one per year, and also broke the figures down into regional totals, which provided overall totals of between one and thirty-nine per year for each region of France. None of the figures from one graph seemed to tally with the others, and I was engulfed in a flood of statistics. But the overall message was clear. There were a reasonably large number of boys drowned each year in France and when I added up the numbers over the past twenty years, this amounted to a significant figure.