Counting the devil's waves


When I went to French Guiana a week or two ago to watch the launch of Meteosat7, I brought with me a copy of Pa- pillon, by Henri Charriere. The book tells the tale of Charriere's escape from Devil's Island, the former French penal colony, which lies just a few miles off the South American coast near the Ariane launching site at Kourou. And as it happened, during my stay I found myself one day upon that now deserted Devil's Island. The Ile du Diable is a beautiful but a most inhospitable spot, and I can vouch for Papillon's evocative description: "The sun blazed down at noon; a tropical sun that made your brains feel they were boiling in your skull; a sun that shrivelled every plant that had not grown big enough to stand it; a sun that made the air dance and tremble; and a sun that dried out the shallow pools of sea water in a few hours, leaving nothing but a white film of salt."

But other assertions in the book are less reliable. Planning his escape and looking at the sea, Papillon notices something odd about the waves: "The great roller, twice as high as the other waves, only came once in every seven waves. From noon until sunset I watched to see if this always happened, to see if there was no sudden change and therefore an alteration in the rhythm and the shape of this huge seventh wave. But not once did the great roller come before its turn or after." This was once a common belief - that there was a pattern to ocean waves whereby individuals were of increasing height until a maximum was reached, and then the series all began again. One tradition had it that each tenth wave was the biggest; Charriere had obviously heard the story with a seven. But the development of waves is complex. Each is the result of the interaction of perhaps several separate trains of waves, often approaching a spot from different directions, sometimes reinforcing each other to create a wave of quite exceptional height and at other times cancelling each other out, leaving the surface of the water momentarily undisturbed. The process is unlikely, even for a short period, to produce a regular pattern of larger waves that could be distinctly identified as one in 10 or one in seven.

Statistical calculations, however, suggest that in a chaotic sea one wave in every 20 may be about twice the average height; one in every thousand three times the average height; and that one in every 300,000 waves may be a monster four times as big as usual.