ANNIYERSAET ADDEESS OF THE PEESIDENT. Ixvii 



has the same mineral composition as these bodies, or, indeed, may be 

 nothing else than an aggregation of meteorites*. 



Eor the purpose of comparison with reference to the substances 

 of which they are composed, the meteorites may be divided broadly 

 into two great classes, the irons and the stones. The former admit of 

 three subdivisions : — 1st, the metallic iron, without any stony ad- 

 mixture ; 2nd, the iron containing grains or globules of peridote 

 (chrysolite or olivine) ; 3rd, iron associated with silicates, pyroxene, 

 and peridote (as that of the Sierra de Chaco). 



The meteoric stones, on the other hand, for the most part contain 

 native iron only in minute grains, disseminated in the midst of 

 silicates, the bases of which are chiefly magnesia and protoxide of 

 iron, bringing a large portion of them into the category of peridote. 

 Others contain no native iron, but are formed chiefly of peridote 

 (Chassigny), or of a less basic silicate (Bishopville, 1863) — or, again, 

 of a granular compound of anorthite and pyroxene, like many of the 

 lavas (Juvinas, Jonzac, and Stannern). 



M. Daubree first describes his experiments on the imitation of the 

 meteoric irons. The most characteristic feature of these masses is 

 the crystalline pattern (Widmanstatten's figures) which is brought 

 to view on a poHshed siu-face by the action of an acid. Simple 

 fusion of the meteorite of Caille (Yar) in a hrasque of alumina (to 

 avoid the contact of carbon, which would have combined with the 

 iron), was insufiicient to reproduce the appearance, although the 

 resulting substance was certainly crystalline. Further experiments, 

 in which soft iron was associated with some of the other substances 

 that commonly accompany meteoric iron, such as nickel and proto- 

 sulphide of iron and silicon, yielded a highly crystalline result, but 

 not yet of the true character. If, however, to the soft iron was added 

 phosphide of iron, in the proportion of from two to five or ten per cent., 

 and, still better, if there was introduced at the same time nickel, and 

 if a mass of as much as two kilogrammes in weight was operated on, 

 there appeared, when the cooled lump was polished and etched, in the 

 midst of dendritic patterns of great regularity, lines of a brillia,nt 

 material dispersed in a reticulated form, 



A third mode of attempting the imitation was that of melting 

 down certain terrestrial rock- substances, as peridote, Iherzolitef, 

 hypersthene, basalts, and melaphyres. By this means specimens of 

 iron were obtained which, both in composition and structure, bore 

 strong resemblance to many of the siderolites. Especially was this 

 notable in the metal obtained from the Iherzolite of Prades 

 (Eastern Pyrenees). These artificial irons were then found, like 

 the natural meteoric ones, to contain nickel, chromium, and phos- 

 phide of iron, the latter in long needles, recalling the appearance of 

 the natural patterns. 



Imitation of the Meteoric Stones. — Contrary to what might have 

 been expected from the appearance of the black vitrified crust on the 



* PoggendorflF's ' Annalen,' vol. cv. pp. 560-562. 



t Lherzolite (so called from Lherz, in the Pyrenees) is a rock composed of 

 peridote, enstatite, and pyroxene (augite). 



e2 



