170363a0Nature1704322195208303633640028-0836195210.1038/170363a0ukNatureNatureNATUREnatureNature is a weekly international journal publishing the finest peer-reviewed research in all fields of science and technology on the basis of its originality, importance, interdisciplinary interest, timeliness, accessibility, elegance and surprising conclusions. Nature also provides rapid, authoritative, insightful and arresting news and interpretation of topical and coming trends affecting science, scientists and the wider public./nature/journal/v170/n4322issueJournal homeArchiveCurrent issueAdvance online publicationPrivacy policySubscribeNature Publishing GroupCurrent issue170363a0Cores of Terrestrial Planets
AU  - BULLEN, K. E.University of Sydney, N.S.W. May 26.ONE feature of Prof. H. C. Urey's new theory1 of the development of the planets is that it leads him to conclude on several grounds that the earth's central core has an iron composition. This conclusion is in conflict with Ramsey's proposal2 that the density throughout the earth's deep interior, at least below a depth of 1,000 km., is essentially determined by pressure right through to the earth's centre rather than by change of chemical composition. The conclusion as stated is also in conflict, though to a less degree, with my proposals, which differ from Ramsey's through finding a balance of evidence3 in favour of a separate chemical composition, such as nickel-iron with possibly some denser material, for the earth's inner core, but agree with Ramsey's in postulating a modified ultrabasic rock composition for the outer part of the core, that is, between depths of 2,900 km. and 4,980 km.[mdash]the region that I have referred to4 as /`E/'; the latter postulate will be referred to as [alpha]. On the further hypothesis, [beta] say, that the terrestrial planets have a common primitive composition, Ramsey's theory implies that neither Venus, Mars nor Mercury has an inner core, whereas my work favours the view that all three have inner cores.The purpose of this communication is to show that evidence can be brought to bear to qualify a number of the arguments by which Urey infers an iron composition for the region E. At the same time, it is interesting that strong new support is found for the suggestion that the inner core is chemically distinct from the rest of the earth. Four main points will be considered, as follow:
(1) Urey appears to accept Elsasser's conclusion5, based on Bridgman's experimental results and calculations of Feynman, Metropolis and Teller6, using a Thomas-Fermi-Dirac model, that a silicate composition for the region E "appears to be excluded". I have recently shown, however, that Elsasser's calculations are unsatisfactory in an important particular, a criticism which has been accepted by Elsasser (in a personal communication). My investigation leads me to the view that, on balance, the data of Bridgman and of Feynman, Metropolis and Teller tend slightly to favour a silicate composition for the region E, although they do not provide a crucial test. The details, which are intricate, are being published elsewhere.
(2) Urey considers that the mass and radius of Venus are so similar to those for the earth "that they can be made to fit any theory that accounts for the Earth". Actually, calculations by Jeffreys7 on Venus show that, on the assumption p, the hypothesis that the region E is chemically distinct from the mantle is untenable; the hypothesis yields 1: 5-6 for the ratio of the mass of the central core of Venus to the total mass of Venus, as against a ratio of 1: 4 1 for the earth. This discrepancy is in significant contrast to the measure of agreement obtained with observational data on Venus by Ramsey and myself on the opposite hypothesis a.
(3) In respect of Mars, Urey directs attention to the discrepancy found by Ramsey in fitting this planet to the hypotheses a and p. But I have shown3b that if use is made of the earth model B 8, constructed from the hypothesis9 that compressibility is a smooth function of pressure through the earth's deep interior, then a and p fit the observed radius, mass and ellipticity of Mars within the standard errors of observation. I am repeating these calculations using revised observational data quoted by Urey, but unless the revision brings to light unexpected changes, Mars must be regarded as giving good support to a and p, the earth model B having been constructed without any ad hoc adjustment to fit planetary data.
(4) Observational data on Mercury prior to 1950 were considered to be too imprecise to be used as a test of a and p. Rabe's recent calculations10, based on the perturbations of Eros, give the mass of Mercury as 0-0543 times that of the earth, an increase of order 20 per cent from the previous estimate. Estimates of the value of Mercury's diameter are still very uncertain, and, taken along with Rabe's value of the mass, lead to estimates of the mean density ranging from 4-6 to 5-2 gm./cm.3. Such values are incompatible with the possibility that Mercury may have the same mean density at zero pressure as other planets, and Urey interprets this as a further test against the presence of a silicate core in the earth.
If Rabe's revision is accepted, it is necessary to accept with fairly high probability that, in relation to the hypothesis p, Mercury is an exceptional planet in some important respect; only a considerable future upward revision of estimates of Mercury's diameter could affect this conclusion. But Mercury may still be exceptional in this respect and yet at the same time compatible with a and p.
An obvious exceptional feature of Mercury is that it is the nearest planet to the sun, and it now turns the same face to the sun. According to radiometric measurements, the present temperature of the sub-solar point at perihelion is 685  K., and the molecular weight of escaping molecules at this point would be about 70. This Would permit the escape of some of the ingredients of an initial rocky mantle provided they could be volatilized sufficiently readily. Various mechanisms could be suggested for bringing about the temperature necessary for such volatilization; of these perhaps the most likely is that, earlier in the history of the solar system, the orbit of Mercury Was much more eccentric than now and that Mercury was once much nearer the sun at perihelion. The suggestion then emerges that Mercury may well have lost not merely a normal atmosphere, but also an appreciable proportion of an initial rocky mantle, by volatilization and subsequent escape, and that Mercury was thus once a considerably larger planet.
On this suggestion, Mercury could still have the same primitive composition as the earth, Venus and Mars, and the region E of the earth could still be modified silicate rock. In fact, on this set of assumptions, it should be possible (subject, of course, to a number of uncertainties, including the uncertainty of the value of Mercury's diameter) to estimate the mass of the primitive Mercury; it is hoped to present this estimate in a later paper.
The suggestion here made may possibly bear some relation to the mechanism which Urey proposes, in his theory, for the fractionation of silicate and iron phases in various planets, taken in conjunction with the statement in his main thesis "that the primitive planets consisted of a grossly homogeneous mixture of silicate and iron phases". The statement quoted is consistent with hypothesis p and is consistent with ideas I have favoured. But the possible significance of Mercury's proximity to the sun leads me to question whether the present mean density of Mercury can properly be used for discriminating in favour of a predominance of the iron phase in the region E of the earth.
On the other hand, it would seem that the probability is now increased against that part of Ramsey's view which holds that the earth's inner core is essentially a second pressure modification of ultra-basic rock. On Ramsey's view, Mercury has no core, and Urey's criticism cannot be so readily met by assuming a substantial loss of material in the course of Mercury's evolution. A substantial loss ould imply the presence of a modified-silicate core in the primitive Mercury, and, in the absence of a chemically distinct inner core, loss of the core would lower the mean density appreciably.
On the question of the difference between the views of Ramsey and myself, it may be stated that all available astronomical evidence now favours a distinct chemical composition for the earth's inner core. The observational data on all four of (a) the mass and diameter of Venus, (6) the mass and diameter of Mars, (c) the figure of Mars, and (d) the mass and diameter of Mercury, independently favour the presence of inner cores in these planets.
It has already been suggested by Jeffreys11 that Urey's stimulating theory, which rests in large part on arguments from physical chemistry, may need some modifications in order to meet arguments from other sources; and it will be of interest to know how far hypothesis a can be fitted into Urey's theory of the origin of the planets without disturbing other important conclusions.Urey, , H. C., Geochimica et Cosmochimica Acta, 1, 209 (1951); [ldquo]The Planets, their Origin and Development[rdquo] (Yale Univ. Press, 1952).ArticleISIChemPortRamsey, , W. H., Mon. Not. Roy. Ast. Soc., 108, 406 (1948).ISIChemPortBullen, , K. E., Mon. Not. Roy. Ast. Soc.: (a) 109, 457 (1949); (b) 109, 688 (1949); (c) 110, 256 (1950).ISIBullen, , K. E., [ldquo]Theory of Seismology[rdquo] (Camb. Univ. Press, 1947).Elsasser, , W. M., Science, 113, 105 (1951).ISIChemPortFeynman, , R. P., Metropolis, , N., and Teller, , E., Phys. Rev., 75, 1561 (1949).ArticleISIChemPortJeffreys, , H., Mon. Not. Roy. Ast. Soc., Geophys. Supp., 4, 62 (1937). 355 (1949); 6, 50 (1950).Bullen, , K. E., Mon. Not. Roy. Ast. Soc., Geophys. Supp., 5, 355 (1949); 6, 50 (1950).Bullen, , K. E., Nature, 157, 405 (1946).ISIRabe, , E., Ast. J., 55, 112 (1950).ISIJeffreys, , H., Nature, 169, 260 (1952).Article