Claims of ’excess heat’ from measurements of the heat of electrolysis at several watts of power are largely based on use of poorly characterized, isoperibol, heat-conduction calorimeters with single-point temperature sensors. This paper describes construction, testing, and calibration of a calorimeter of similar design. Heat-conduction calorimeters with single-point temperature sensing and inadequate mixing are subject to large systematic errors resulting from non-uniform heat distribution within the system. Confirmation of electric-heater calibration by a chemical reaction with a well-known enthalpy change is a minimum requirement to insure accuracy. Improper or incomplete calibration is a probable cause for many claims of ’excess heat’ in ’cold fusion’ experiments.