COMPOSITION AND STRUCTURE OF METEORITES 57 



my observation are those described ^^ in the stones of Parnallee and 

 Tennasilm. Even here, however, the resemblance is sHght, consisting 

 in the case of the latter in an outer zone of crystals surrounding a 

 granular interior. The kugel gabbros as those of Slattmossa, Norway, 

 and Dehesa, Calif., show a variable granular nucleus surrounded 

 by more or less radiating and elongated minerals of the same nature 

 as those in the main rock mass. This holds true also of the kugel 

 diorites of Corsica and Davie County, N. C. The structure and 

 origin of these is so obviously different from those of the true chondrule 

 as to render discussion unnecessary. 



A further striking and decisive distinction between the meteoric 

 and terrestrial forms lies in the fact that in the last named the 

 spherules in any rock mass are invariably of a single type only. In 

 meteorites, two or more types, varying in mineral composition and 

 degrees of crystallization usually occur in the same stone and often 

 in close juxtaposition. 



Finall}", and so far as relates to the kugelchen, or spherulitic 

 chondrites there is a difi'erence in that the meteoric chondrules are 

 in all cases foreign to their hosts. Those of the terrestrial rocks, 

 with the exception of the Thetford conglomerate, are formed in 

 place. 



METHODS OF ANALYSES OF STONY METEORITES 



The making of satisfactory chemical analyses of stony meteorites is 

 attended with difficulties little appreciated by one accustomed only 

 to ordinary silicate work. This is due to several causes, the chief of 

 which is the presence of iron in its several forms (as metal alloyed with 

 nickel and cobalt, as phosphide, sulphide, and carbide and also in 

 ferrous and ferric combinations in the sihcates) and the desirabiUty 

 of determining each of these in its individual percentage amount. 

 On first thought the magnetic separation of the iron might seem in 

 all cases most feasible, but long experience has shown the impossi- 

 bility of complete separation and many analysts have resorted to the 

 expedient of solvents, as digestion in mercuric chloride. The follow- 

 ing actual analysis illustrates the method adopted by Museum chem- 

 ists in the analysis of the Florence, Tex., stone. '^ It is in substantial 

 agreement with that used by Prior of the British Museum and J. E. 

 Whitfield of Philadelphia. 



A piece of the stone weighing 19.08 grams was crushed to pass 

 80 mesh. The larger particles of metal which would not pass the 

 sieve were pounded flat to free them from silicate as far as possible 

 and reserved. The powdered material was then worked over thor- 



'• On Chondrules and Chondritic Structure. Proc. Nat. Acad. Sci., vol. 6, 1920. 

 " J. T. Lonsdale, American Mineralogist, Vol. 12, no. 11, November, 1927. 



