Mare basalts, which are believed to form by partial melting at considerable depths in the lunar interior, are capable of providing a wealth of information concerning the compositions of their source regions. Conversely, any acceptable estimate of the lunar bulk composition must in principle be able to provide source regions capable of yielding mare basalts. A wide range of lunar bulk compositions has been proposed in the recent literature. These differ principally in the proportions of involatile elements, e.g. Ca, A1, to elements of moderate volatility, e.g. Mg, Si, Fe. A detailed experimental investigation has been made of the capacity of the Taylor—Jakes compositional model (8.2 % A12O3) to provide source regions for mare basalts. It is demonstrated that this composition is much too rich in alumina to be acceptable. Other lunar bulk compositions even richer inA12O3 such as those advocated by Ganapathy & Anders, Wanke and co-workers and Anderson can likewise be rejected. In order to produce mare basalts, particularly the least fractionated varieties represented by some Apollo 12 and 15 basalts, lunar bulk compositions containing only about 4 % of A12O3 appear to be required. This is similar to the alumina content of the Earth’s mantle. The relative abundances of many other involatile elements, e.g. Ga, U, Ti, r.e.e., Zr, Ba, Sr, may likewise be similar in the Moon and in the Earth’s mantle. These relationships point towards a common origin for the Moon and for the Earth’s mantle.
Solid state convection models of the lunar internal temperature
