Investigates the aftermath of the massive eruption in 1985 of the Colombian volcano Nevado del Ruiz, which resulted in the deaths of over 20,000 people in the nearby town of Armero and led to an intensified search to devise a reliable method of prediction. Discusses two widely divergent theories: one devised by Bernard Chouet, who believes that the key is the study of seismic signals called long period events made by molten magma resonating; the other, devised by the American volcanologist Stanley Williams, involves climbing into craters and measuring the amount of gas being emitted. In 1993, Williams led a group of volcanologists attending a conference at the foot of another Colombian volcano, Galeras, on a trip into its crater. When the volcano erupted suddenly, six scientists and three tourists were killed. Chouet's methods were used in Mexico in December 2000 to predict an eruption of Popocatépetl in time to evacuate tens of thousands of people safely. Other contributors include: David Harlow (US Geological Survey 1971-1994); Dr Marta Calvache and Fernando Gil (Colombian Institute of Geoscience); Prof Bill McGuire (Benfield Greig Hazard Research Centre, UCL); Dr Norman Banks (US Geological Survey 1967-1999); Dr Larry Malinconico (Lafayette College, Pennsylvania); Dr John Stix (McGill University, Montreal); Dr Andrew MacFarlane (Florida International University); and Dr Carlos Valdes (National Centre for Prevention of Disasters, Mexico).