188 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



Average Composition of Stony Meteorites Compared ivith Terrestrial Rocks. 



1 Average of 46 dets. 



2 Average of 42 dets. 

 ' Average of 50 dets. 

 * Average of 41 dets. 

 5 Average of 19 dets. 



6 Average of 6 dets. 

 ' Average of 33 dets. 



8 Average of 49 dets. 



9 Average of 44 dets. 

 1" Average of 15 dets. 



11 Average of 44 dets. 



12 Average of 51 dets. 



13 Average of 16 dets. 

 K Average of 8 dets. 

 i!" Average of 5 dets. 



Now with our present knowledge regarding the composition of the 

 earth's crust, of the relative proportional abundance of the basic and 

 acidic rocks, I feel that it is self-evident that no amount of chemical 

 differentiation of such a magma as that presented by the meteorites 

 could bring into existence such a body as that of our earth so fa?' as 

 its composition is yet revealed. It is to be noted, however, that we 

 have no direct evidence as to the chemical nature of meteoric matter 

 that may have come to the earth in past ages. Incidentally, I will 

 mention one fact that has often impressed itself upon me. Those 

 who affirm that the salt of the ocean is wholly secondary, have no 

 difficulty in finding a source of the sodium through the ordinary 

 atmospheric decomposition of sodium bearing silicates. Not so, how- 

 ever, with the chlorine, and as with the exception of comparatively 

 small amounts of this element carried by such minerals as sodalite 

 and some of the apatites, there are known no original chlorine-bear- 

 ing minerals, it is difficult to account for its relative abundance in 

 oceanic waters. In the comparatively abundant meteoric mineral 

 lawrencite we have at least a suggested source, and that too, in a 

 form easily broken up when exposed to atmospheric conditions. 



