The Nature of the

The available experimental data and steady-state calculations make it difficult to explain the M discontinuity beneath both oceans and continents on the basis of the same phase change. The oceanic M discontinuity may be a chemical discontinuity between basalt and peridotite , and a similar chemical discontinuity ma,y thus be expected b�neath the conti­ nents. Since ayailable experimental data place the basalt-eclogite phase change at about thE' same depth as the continental M discontinuity, intersections may exist between a zone of chemical discontinuity and a phase transition zone, the transition being either basalt-ecloO"ite or feldspathic peridotite-garnet peridotite. Detection of the latter transition by seismic t�'h­ niques ma,y be difficult. The M discontinuity could therefore represent the basalt-eclogite phase change in some localities (e.g. mountain belts) and the chemical discontinuity in others (e.g. oceans and continental shields). Variations in the depth to the chemical discontinuity and in the positions of geoisotherms produce great flexibility in oro genetic models. Interse;­ tions between the two zones at depth could be reflected at the surface by major fault zones separating large structural blocks of different elevations. Chemical discontinuity. The conyentional view that the Moho discontinuity at the base of the crust is caused by a chemical change from basaltic rock to peridotite has been expounded recently by Hess [1955], Wager [1958], and Harris and Rowell [1960]. Petrological and geo­ chemical concepts related to the composition of the earth's mantle and problems of heat dis­ tribution have been discussed by Bowen [1928J, Buddington [1943], Ross et al. [1954], Ver­ hoogen [1954J, Birch [1958], Lovering [1958, 1959], Kuno [1959], MacDonald [1959a, 1959b], Harris and Rowell [1960], and many others. A conclusion that the upper mantle is composed of material having the composition of a feld­ spathic peridotite is in reasonable agreement with the hypothesis that the oyer-all composi­ tion of the mantle is the same as that of the silicate fraction of chondritic meteorites. Al­ though there are divided opinions about the validity of meteorite analogies, the chondritic model is favored in most recent reviews [Mac­ Donald, 1959b; Harris and Rowell, 1960; Vino� gradov, 1961; Ringwood, 1962a; Mason, 1962]. MacDonald [1959b], after summarizing argu­ ments in favor of the chondritic model, con­ cluded that material having radioactiyity in1 Contribution 62-81 from the College of Mineral Industries. termediate between dunite and basalt is qualitatively reasonable for the mantle. Ringwood [1962a, c] adopted a specific model compounded from the concepts advanced and developed by many petrologists and geochem­ ists, and he concluded that to a first approxima­ tion the mean chemical composition of the upper mantle and crust over any e::l.iensive region of the earth is similar to a mixture of 4 parts of dunite to approximately 1 part of basalt. Ring­ wood called this material 'pyrolite,' but I prefer to retain the terms 'feldspathic peridotite' and 'ga,rnet periciotite' (for the same material in the eclogite facies) which were used long ago by