The amount of heat flowing from the earth is a quantity of considerable geophysical importance. For some years there has been speculation about the possible consequences of the earth’s crust warming up, on the assumption that the amount of heat generated by the radioactive content of the crust is greater than that flowing to the surface (Joly 1925). Although a number of different workers have made measurements of the radioactive content of rocks from various parts of the world (Jeffreys 1936), until the present time the heat flow received little attention. If the amount of heat flowing from the earth in unit time per unit area is denoted by H , then | H | = kp , (1) where k and p are the thermal conductivity and temperature gradient in the earth’s crust respectively, and it is therefore apparent that two quantities must be measured before a value for H can be obtained. For a great many years experiments have been made to determine the thermal conductivity of rocks, and for a long time temperatures have been measured in wells, bore-holes and tunnels, but the writer knows of only two places, the Simplon Tunnel and the Lake Superior Copper Mining District, where both have been measured, and the values of the heat flow deduced from these data are rather uncertain (Anderson 1938; Ingersoll and Hotchkiss 1934; Peirce 1903). As the value of k for different rocks varies through a factor of twenty or more, and as p in one region may differ by a factor of six from that in another district (neglecting hot spring and volcanic areas), the practice of estimating the heat flow by taking ‘‘average” values of p and k must lead to a very uncertain value of H . In the past, H has been taken by various writers at different times to lie anywhere between 3.45 x 10 -6 and 1.6 x 10 -6 cal./sq. cm.-sec., and it will be shown below and in a companion paper (Bullard 1939) that at least for Britain and South Africa even the second estimate is too high. It must be added that in general there are practical difficulties in the work. Temperatures measured in mines are often unreliable, and for various reasons there are not many “dry” bores in this country which have been left unplugged for a sufficiently long time for reliable temperature measurements to be made, and from which a well-preserved selection of representative rock samples is available of a nature suitable for making rock specimens for conductivity purposes.

Terrestrial heat flow in Great Britain. (1939). Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 173(955), 428–450.