The melting point of relatively pure iron was determined at pressures up to 40 kb. When the experimental data are fitted with a straight line of slope 2.85 deg/kb most of the scatter is less than 4°. If observed temperatures are corrected according to Hanneman and Strong for the effect of pressure on thermocouples, the slope of the melting curve becomes 3.8 deg/kb. Some limits upon the melting temperature of iron at higher pressures are obtained by consideration of zero‐pressure and low‐temperature data, as well as high‐pressure data, for iron and its alloys. Because of the considerable ignorance concerning the course of fusion curves at high pressure, it seems unwise to extrapolate the present data in order to estimate temperatures near the center of the earth.