28 



A. P. VINOGRADOV 



tion of Stony meteorites or chondrites differs considerably from that of the ultra- 

 basic formations within the Earth, exemplified by dunites, in having a high 

 content of chemical elements which would appear to be easily mobile in molten 

 silicates (Table 2). 



Table 2 



The contents of some chemical elements in chondrites^ dunites, basalts 

 and granites (% %) 



Consider what happens when you heat a bar made of the material of a stony 

 meteorite towards its melting point in such a way that only a very narrow zone 

 of it is actually melted and that you then move the source of heat gradually along 

 the whole length of the bar. This can be repeated many times. You will then have 

 a liquid fraction which will travel repeatedly through the bar of chondrite being 

 so treated. As a result of this zone melting you will get, collected at one end of 

 the bar, those substances which lower the temperature of crystalUzation, while 

 at the other end there will be those which raise the temperature of crystallization 

 of the silicate melt. I have carried out this experiment with chondrite and con- 

 vinced myself that there actually does occur a process of enrichment in a definite 

 direction (the direction, as it were, of driving out of alkahs and other elements). 

 This method will obviously permit us, in future, to obtain a large number of 

 relevant quantitative data. It seems to me that it also provides us with the possi- 

 bility of giving an experimental explanation of the mechanism of the differen- 

 tiation of the substance of the Earth. These experiments have enabled me, though 

 only speculatively, to take a first step and to suggest that the dunites forming the 

 mantle of the Earth are the residues left by melting out the terrestrial material which 

 had a composition similar to that of the chondrites. If this is true, an estimation of 

 the amount of the easily transferred elements in dunites, the crust of the Earth 

 and chondrites should give us a measure of the thickness of the mantle of the 

 Earth which was involved in the formation of its crust (Table 2). As yet we have 

 no accurate extraction coefficients for the behaviour of each element mentioned 

 in the Table. However, if we use for our calculation the figures for chondrites 

 as representing the original composition and those for dunites as representing the 



