120 GEOLOGY. 



that could not be retained in the gaseous state by a small body, the whole may be 

 taken to represent the material of the full-grown earth. This addition would 

 probably not remove the total average earth-material very far from the average 

 of the meteoritic matter, and so, with some qualification, meteoritic matter may 

 be taken as a rough, but somewhat too basic and metallic, standard of the total 

 earth-material. 1 It is easy, however, to gain a false impression of the average 

 nature of meteoritic material, for the stony meteorites are less readily detected 

 than the iron ones, and hence a less proportion are recognized. Besides this, 

 the stony meteorites usually disintegrate rapidly and soon disappear, and hence, 

 if not found promptly, are not found at all. If, therefore, the meteorites that are 

 averaged include simply those that are found, the result will be much too high 

 in metallic material. The true average should include only those seen to fall. 

 On this basis, Farrington concludes that the true average is represented essen- 

 tially by stony material of the basic silicate type, with only a subordinate amount 

 of metallic material, and that the average specific gravity is about 3.69. 2 



Volcanic differentiation. — It will be recalled that from an early stage, molten 

 material is supposed to have been poured out on the surface at intervals, and 

 buried by the inf ailing planetesimals. At the same time, the material was 

 slightly weathered on the land-areas, and such of the material as was dissolved 

 was carried into the water-areas. Later, when the surface had been sufficiently 

 built up by accessions, this same material was exposed to a second possible vol- 

 canic experience, with further selective fusion and further surface differentiation. 

 Still later, the process may have been repeated, perhaps again and again, for 

 the first volcanic surfaces, may possibly have been as much as 1200 or 1500 miles 

 below the present surface, and the number of possible repetitions of the volcanic 

 round is large. A wide differentiation from the original state was therefore pos- 

 sible. As the last series of extrusions came after the cessation of appreciable 

 accessions from without, practically nothing but the various varieties of the differen- 

 tiated product would appear at the surface. That these should differ widely from the 

 original average rock is not strange ; but why they should differ in precisely the 

 ways in which the surface rocks differ from the supposed average rock remains 

 to be worked out, and the solution of this problem will, in some sense, be a test 

 of the validity of the hypothesis. It is to be borne in mind that the result is 

 not a mere selection based on the melting-points of the minerals found in the 

 surface rocks, but is rather the outcome of a complex process of mutual solution 

 and precipitation in which the modifying conditions of temperature, pressure, 

 ratio of admixture, stress-differences, and specific gravities, play essential parts. 



Differentiation in suboceanic versus continental ssgments. — It is reasonable 

 to assume that the lavas which arose from the more basic segments that underlie 

 the deep-sea basins would be more basic than those that arose from the more 

 acidic land segments, and this seems to be supported by the greater prevalence of 



1 For composition of meteorites, with references to the literature, see Farrington, 

 Jour, of Geol., Vol. IX, 1901, pp. 393 and 522. 



2 Jour, of Geol., Vol. V, 1897, p. 130. 



