184 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1917. 



lite, and troilite are largely, if not wholly, unknown among terres- 

 trial rocks, while nickel-iron is rare. On the other hand such com- 

 mon terrestrial minerals as free quartz and the compounds of silica 

 with aluminum, calcium, and the alkalies, as orthoclase, albite, nephe- 

 line, the micas and amphiboles are rarely, if ever, found in meteorites. 

 Such secondary minerals as serpentine after olivine, talc, chlorite or 

 hornblende after pyroxenes, and indeed any minerals requiring the 

 agency of water or the vapor of water for their production, as zeo- 

 lites, the micas, tourmalines, etc., are also lacking. It is evident 

 that the meteorites were formed under conditions of a limited sup- 

 ply of oxygen and that they have since their formation been sub- 

 jected to high temperatures and the reducing power of gases. 



All meteorites may be traced unmistakably to an igneous origin, 

 and are of a basic nature, related closely to the basalts, pyroxenites, 

 and peridotites among terrestrial forms. None have as yet been 

 discovered which can be considered as sedimentary, or metamorphic 

 as the word is commonly used, although many of the tuf aceous forms 

 have undergone certain changes that may be ascribed to the high 

 temperatures and reducing vapors already referred to. But there 

 is among them nothing comparable with our sand and limestone 

 or argillites, and absolutely nothing of a fossil nature or necessarily 

 indicative of any form of animal or vegetable life, although in years 

 past some of the peculiar radiating and branched forms such as I 

 have shown have been mistaken for and even described and figured as 

 fossil corals and crinoids. 



That meteorites as they come to our earth are plainly fragments 

 of preexisting masses has been already stated. Some of these, like 

 those of Stannern and Shergotty, owe their internal structures to 

 direct crystallization from a molten magma. Nevertheless, the struc- 

 tures are by no means similar to those found in terrestrial rocks. 

 As a whole, they show evident signs of hasty crystallization. The 

 Shergotty stone, to be sure, has a somewhat familiar aspect, but it 

 belongs to a type standing almost wholly by itself. So far as I can 

 recall, no structure is found among terrestrial rocks more nearly 

 approaching that of the chondritic meteorite than that of the orbicu- 

 lar gabbro of Davie County, North Carolina, or the kugel granite 

 of Sweden. The resemblance is, however, merely suggestive, and 

 disappears the moment the rock is submitted to critical study. In a 

 large proportion of the kugel chondritic stones the structure of the 

 ground is quite obscure, though the chondrules contain well-devel- 

 oped, as well as radiating and acicular forms, which result from cool- 

 ing of molten material. The absence of a glassy base from the 

 ground which bears the chondrules is antagonistic to the idea of 

 the origin of both portions through the same agencies and under the 



