The secondary nucleation hypothesis predicts that nucleate boiling benefits from bubbles bursting from a liquid film on the heating surface. This possibility was examined experimentally. Water was boiled at near atmospheric pressure on a steam heated horizontal copper plate with the heat flux varied up to 100 kW/sq m. The head of liquid was varied and heads up to 40 mm were investigated. Heat transfer coefficients increased and sometimes doubled as liquid head was decreased below a threshold which increased with heat flux. The threshold increased from 10 mm at 30 kW/sq m to 30 mm at 80 kW/sq m. The increase in heat transfer coefficient was greater at higher values of heat flux. Foam occurred at the higher heat flux values to bring a liquid-vapor interface much closer to the copper plate than indicated by the liquid head. This allowed secondary nucleatlon to operate closer to the heating surface than without the foam. At reduced liquid heads the proportion of liquid in the foam decreased as the heat transfer coefficient increased to enhance nucleate boiling. Although the secondary nucleatlon hypothesis has provided a useful prediction it remains a hypothesis.