Creator:
Date Created:October 20, 1958
Place Created:Northern Ireland
Keywords:mantle transition zone
Context:article from Geochimica et Cosmochimica Acta
**************************************************
Geochimica et Cosmochimica Acta, 1959, Vol. 16, pp. 192 to 193. Pergamon Press Ltd. Printed in Northern Ireland
Constitution of the mantle—a revision*
a	(Received 20 October 1958)
In some papers recently published (Ringwood, 1958, Parts I, II, III) the nature of the phase transition zone in the mantle between 2-400 km and 900 km has been extensively discussed. The large width of the transition zone was attributed primarily to the presence of a high temperature gradient in this region combined Avith a high pressure gradient dPjdT for the transitions, and secondarily to solid solution effects among the participating phases. Subsequent data have shown that the relative importance of these factors should be reversed.
The slope dP/dT of the olivine-spinel transition in Mg2Si04 was estimated very tentatively as 100 bars/degree. This implies an entropy change of 10 cals/mole per degree. Meijering and Rooymans (hi press) have pointed out that the entropy change would probably be substantially less than this, which means that dP/dT would probably be proportionally smaller. Roy and Roy (1954) have determined the initial slope dPjdT for the olivine-spinel transition in Mg2Ge04 as 40 bars/degree, and the corresponding value for Mg2Si04 may well be close to this. At higher temperatures and pressures it would probably be somewhat larger. This is supported by the later work on the transition in Mg2Si04 at 600°C, which was presented as an appendix to Part II. This showed that the transition pressure was substantially higher than the calculated value at 600°C. This result was obtained too late for incorporation of its implications in the manuscript.
As a result of the smaller gradient, the width imparted to the transition region by any plausible temperature distribution is less than previously assumed.
On the other hand, an investigation by Meijering and Rooymans (in press) has shown that solid solution effects are more important than previously supposed in spreading the transition region. They made a thermodynamic study of the olivine-spinel transition as a function of pressure at constant temperature, in the ternary system Mg0-Fe0-Si02 and concluded that solid solution effects could spread the transition over a range of several hundred kilometres.
A further contribution to the transition range, which has not previously been considered, is made by solid solution between diopside and enstatite which occurs at temperatures above 1200°C. A small amount of diopside (which is probably present) would extend upwards the pressure required for solid solution of MgSi03 in Mg2 Si04 spinel, or for decomposition of MgSiOs into Mg2 Si04 spinel and Si02 (coesite).
In view of these considerations it seems reasonable to attribute the transition range in the mantle largely to solid solution effects. This does not exclude the possibility of a high temperature gradient in this region, which may be supported by other evidence ►	(Tozer, in manuscript). The influence of the temperature gradient in spreading the
transition region would still be important, but not dominant.
* Paper No. 172 published under the auspices of the Committee of Experimental Geology and Geophysics and the Division of Geological Sciences at Harvard University.
192
GEOCHEMICAL NOTE
Geochemical note
A diagram illustrating the revised interpretation is presented (Fig. 1).
Fig. 1. Possible relationship between transition zone and temperature distribution in the mantle. Letters on temperature depth curve refer to phase assemblages which may be present (Ringwood, Part III), 1958.
A.	Olivine (jp^Q	enstatite	small amounts of diopside-jadeite solid
solution, pyrope rich garnet.
B.	Olivine, enstatite-diopside-jadeite solid solution (henceforth called "pyroxene"),
minor garnet.
C.	Olivine, pyroxene, iron-rich spinel.
/MgO \
D.	Spinel I~5), pyroxene.
E.	Disordered, non stoichiometric spinel, containing (MgFe)Si03 in solid solution, minor
diopsidic pyroxene.
F.	Homogeneous, highly disordered, spinel solid solution.
Notes:
(I) Disappearance of garnet between B and C is controlled by solubility- of A1203 and Cr203 in spinel and pyroxene. All other components present in garnet are readily soluble in these phases. In regions where A1203 is locally concentrated, garnet may persist as a phase,
to greater depths.
(II) Principal transition between C and D is the olivine-spinel inversion, which takes place over a substantial depth range due to Mg—Fe solid solution. (ill) Principal equilibria involved in spinel-pyroxene solid solution between D and E are: 2(MgFe)SiOa -* (MgFe)8Si04 (spinel) + Si02 (Coesite) (MgFe)2Si04 (Spinel) + Si02(coesite) —* non-stoichiometric defect spinel.
Acknowledgement—The author is indebted to Drs. Meijering and Rooymans for a pre-print of their paper.
A. E. Ringwood
Department of Geophysics, Australian National University Canberra, A. C. T., Australia
References
Meijering J. L. and Rooymans C. J. M. (1958) On the Olivine-Spinel transition in the earth's
mantle. Proc. Kon. Ned. Akad. (in press).	I
Ringwood A. E. (1958) Constitution of the mantle: Parts I, II and III. Geochim. et
Cosmoschim. Acta 13, 303-321; 15, 18-29 and 195-212. Roy D. M. and Roy R. (1954) Amer. Min. 39, 957. Tozer D. In preparation.
13
193