A new and relatively simple rncthod is described for. the mcasirremeni of the heat changes which accompany magnetization processes in fields of the order of a few hundred oersteds. The adiabatic changes in temperature of a rod which occur when the latter is taken step-by-step through a hysteresis cycle are measured by a series of thermocouples of which the " hot " junctions are directly attached to theq rod. The temperature-measuring system is capable of dctecting changes in temperature of the order of deg. c., and its sensitivity is such that 1 mm. scale deflection corresponds to the absorption or liberation of about 330 ergs per C.C. in the specimcn, deflqctions of 0 . 1 rhm. being easy to read.Experiments made with pure annealed and hard-drawn nickel arid with certain nickel-iron alloys, strained and unstrained, are described, and mapy grpphs are given to show the observed temperature changes as functions of the iritensity of magnetization and of the applied field. T h e results provide a very effective proof of M'arburg's law, but their main importance is that they enable a detailed analysis of the energy changes accompanying magnetization to be made. They show that Becker's views on magnetization processes in ordinary ferromagnetic materials arc essentially correct, but do not support Preisach's conception of the formation of demagnetization nuclei in the casc of extrcmely soft ferromagnetic materials. $ 1 . I N T R O D U C T I O NH E study of the magnetocaloric effect or t h e reversible changes in temperature which accompany magnetization has added much to our knowledge of t h e T behaviour of the spontaneous magnetization of ferromagnetics. It is well known that on placing a paramagnetic or a ferromagnetic substance in a very strong field a n adiabatic rise in temperature is observed, and a corresponding fall in temperature when the substance is removed from t h e field. T h e fields employed in such experiments are so intense that ordinary hysteresis phenomena, which are confined to fields of some few hundred oersteds, d o not play a part.It is correct to say that verv much more is known about tcrnperature changes accompanying niagnctization processes in very strong fields than about those in 15 cm. long and 1.5 cm. in diameter. , H e found that (I) HdI correctly represented the total energy dissipated when 1 c.c-. was taken through a complete cycle, although the difference between the liberated energy calculated from the ( I , H ) curve and that measured directly was sometimes as much as 10 per cent. Similar experiments were carried out by Constant (1928) with rings of " I(. S. Magnet Steel " (' 4 11 J r I I Adiubutic temperuture chunges uccompunying magnetization