3 o SCIENCE PROGRESS 



quake records received at Gottingen during the years 1904- 

 191 1, and as a consequence they have been able to distinguish, 

 not two zones alone, but four, separated by marked physical 

 discontinuity at depths of approximately 750, 1,060, and 1,530 

 miles. 1 The Wiechert law of density must be correspondingly 

 modified. Beneath the outer crust, which has a mean density 

 of about 2 - 8, are four zones, of which the outer one has 

 probably a density of 34; the inner being as before about 

 7 , 8-8 , o. The two intermediate zones, which are of less bulk 

 than the others, are presumably characterised by intermediate 

 densities. This conclusion is more likely to be in accordance 

 with the facts than Wiechert's, for it implies a gradual down- 

 ward increase of density in place of a relatively sudden change. 



It is manifestly impossible ever to know directly the 

 chemical constitution of the earth's interior. However, we may 

 study in the laboratory the disrupted fragments of some other 

 world, 2 for it is now believed that meteorites were once 

 arranged according to their densities as parts of a cosmic body. 

 If this be true it seems highly probable that the constitution of 

 the meteoritic parent body thus determinable was essentially 

 similar to that of the earth. Let us briefly review the evidence 

 in favour of these suggestions. 



Meteorites may be divided into five well-marked classes 

 according to the relative proportions of their metallic and stony 

 constituents. The iron meteorites, which are known as 

 holosiderites, consist almost entirely of a coarsely crystalline 

 nickel-iron alloy (average density 78). The dimensions and 

 uniformity of structure of the giant metallic crystals of many 

 iron meteorites indicate that they crystallised very slowly from 

 a magma which remained for a long period at a nearly uniform 

 temperature not far below the fusion point. This in turn 

 suggests that most of the holosiderites were formed in the 

 deep interior of the parent body, where the pressure would 

 be high. 



It has been conjectured that the well-known Widmanstatten 

 figures represent an internal structure which may have been 

 due either to (a) very slow crystallisation, (b) sudden chilling 



1 Gessellwiss Gottengen Nachr. Nath. Phys. Klasse, 2, pp. 121-206, 1912; 6, 

 pp. 625-75, 1912. 



* Chamberlin, Journ. Geo/., vol. ix. p. 369, 1901. See also a valuable series 

 of papers by Farrington in the same volume. 



