Ixviii PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



surface, the substance produced by the melting down of meteorites 

 obtained from above thirty different falls, was in every case highly 

 crystalline. Those of the common type present a group of metallic 

 granules, disseminated in a stony mixture of peridote (Mg^Si) and 

 enstatite (Mg Si), the former generally on the surface as a thin crys- 

 talline pellicle, the latter in the interior as long acicular crystals. 

 A notable contrast was pelded by the aluminous meteorites, such as 

 those of Juvinas, Jonzac, and Stannern, which produced, instead of 

 a crystalline, a vitreous mass. 



But perhaps the more remarkable results were those obtained 

 synthetically by melting down pieces of rock characterized by the 

 minerals peridote and enstatite. For this purpose peridote (olivine), 

 from the basalt of Langeac (Haute Loire), and Iherzohte, from 

 Yicdessos and Prades, were fused in earthen crucibles. They 

 melted easily and yielded crystalline substances, the latter especially 

 closely resembling the original rock. The proportion of enstatite (the 

 bisihcate of magnesia) was found to be increased by the addition of 

 silica. 



When similar mineral substances were melted in presence of a 

 reducing agent, the iron (which in the other case remained combined 

 in the silicate) segregated itself in grains of various sizes, sepa- 

 rable by the magnet. Thus a perfect analogy was established be- 

 tween the above rocks and the meteorites, as well in their stony 

 minerals as in the iron, which always contained nickel. . 



Furthermore some remarkable characters in the structure of 

 the stony meteorites were found to have been imitated, especially 

 the dehcate parallel lines, attributable to cleavage, which are visible 

 when a thin shoe is examined under the microscope, and the glo- 

 bular structure where the little spherules are sometimes smooth 

 at the surface, at others drusy, or roughened vdth the points of 

 minute projecting crystals, like the meteorite of Sigena, ]S"ovember 

 17, 1773. 



When hydrogen was employed as the reducing agent the results 

 were very similar, and the reaction would take place at a tempera- 

 ture not exceeding red heat. 



Again, another method of imitation, the reverse of the foregoing, 

 was by oxidation. Prom silicide of iron, heated in a hrasque of 

 magnesia by the gas blowpipe, a substance was obtained extremely 

 similar to the common type of meteorite. The iron was separated 

 partly as native iron, partly as a silicate, forming peridote, some 

 of it in the crystallized state. Further details of resemblance were 

 attained by heating a mixture of silica, magnesia, nickeliferous iron, 

 and phosphide and sulphide of iron. The stony gangue of the 

 melted product was found to be free from the latter three substances ; 

 and instead of the simple phosphide introduced in the experiment, 

 there was observable the triple phosphide of iron, nickel, and mag- 

 nesium, first noticed by Berzelius in meteoric irons. 



The preceding experiments suggest some important deductions on 

 the condition of the planetary matter from which the meteorites have 

 been diverted to our own globe. M. Daubree makes no attempt to 



