Abstract Probably from the beginning of the rubber industry no process has changed so little as that for softening raw rubber, its socalled mastication. Although Hancock's machine, the masticator (1826), consisted of an internal rotating roll and a stationary shell, and modern machines have two rotating rolls, the same principle has always held true, viz., a compression and consequent stretching of the rubber by means of great friction. Even enclosed mixers are essentially only a development of the Hancock principle, the value of which naturally will not be discussed here. In spite of such long experience, the process of mastication is not clearly understood, and it is remarkable that there has apparently been little interest in this problem until recently. Only within the last few years has it received somewhat more attention, and attention should be called to the work of Cotton, in which there is a compilation of the literature. As a result of this lack of knowledge, the conditions under which the most rational processes of mastication can be carried out have not been established, and even today the manufacturers of machines produce for the most part mills which have the most widely varying constants, without any satisfactory reasons for the particular constructions. The following discussion aims therefore to contribute toward the establishment of these conditions, and at the same time to view critically the various theories of the process of mastication. In this discussion plasticity values obtained by the method of Marzetti have been used as a measure of the degree of mastication since, as I have been able to show earlier and has recently been confirmed by others, these plasticity values approach most nearly to practical conditions.