EXPERIMENTS RELATIVE TO METEORITES. 331 



©f tlu'ir arrival in our atmosphero, for it is not in tlie latter that these volatile 

 compountls can have been incorporated with them. 



CHEMICAL CONSTITUTION AXD MODE OF FORMATIOiV. 



After the experiments which we have detailed, the very characteristic natiu'o 

 of the masses from which meteorites proceed may be readily explained, and that 

 in a two-fold manner, according as we have recourse to the experiments of reduc- 

 tion or to those of oxidation. 



It has been seen that the characters of meteorites are reproduced, even to the 

 intimate details of their structure, in the reduction of basic silicated rocks bj 

 means of charcoal. Let us not conclude, however, that the meteorites have been 

 formed In* this process ; for if this were so, the carbon would doubtless have car- 

 buretted the iron in a very considerable degree, as in steel or in casting, which 

 is not ordinarily the case. Moreover, there is room for asking, in the case in 

 which the formation of meteorites may have been accompanied by a reductive 

 action, whether it would not be necessary to attribute it rather to a hydrogenated 

 atmosphere.* The ingenious experiment by which M. Graham has veriiied the 

 presence of hydrogen in a state of occlusion in the meteonc iron of the specimen 

 of Lenarto, would serve as a confirmation of this idea, which had been announced 

 previously to the discovery of the eminent English chemist, [Comptcs llcndus, 

 19th February, 1SG6, t. Ixii). 



In j)lace of considering- the cosmical bodies with whic^h we are occupied as the 

 result of a reduction of silicated rocks, perhaps it is more simple and conclusive 

 to have recourse to the idea of an oxidation analogous to those whicdi we have 

 artificially realized. Let us su})[)ose, as has been done for our globe, that the 

 silicium and metals of the meteorites have not always been combined with oxygen, 

 as they are at present for the most part, and this, perhaps, because the initial 

 temperatiu'e of these bodies was sufficiently high to prevent them from entering 

 into combination, or because, being at first at a distance, they did not approach 

 one another. 



If, in consecpience of refrigeration, or from any other cause, such as an approx- 

 imatit>n of these bodies, the oxygen comes to act suddenly, it will unite itself 

 with the more oxidable elements. The silicium and magnesium will burn before 

 the iron and nickel ; and if the burning gas is not sufficiently abundant to oxi- 

 dize the whole, or if it does not act during a sufficient time, it will leave a metal- 

 lic residue com])osed of the less oxidable bodies ; these metals, the iron and nickel, 

 must remain disseminated in a gangue of silicates while retaining their metallic 

 state, exactly as wo observe it in the m(;teorites. ]\Ioreover, there will l)e thus 

 formed a silicate of magnesium more or less rich in protoxide of ii'on, having the 

 composition of peridot. 



As is now seen, if we su[)pose the oxidation })ashed successively to dilierent 

 degrees, the preceding experinu'uts exjjlain not only the formation of meteorites 

 of the common type, but also that of the group of syssiderous and of the sub- 

 group of j)olysiderous meteorites. These bodies are, therefore, to be assimilated 

 to products by the dry way. This mode of formation appears not to apply so 

 well to the meteorites pertaining to the group of crt/ptosklcrous, and particularly 

 to those of the type of Juvinas, of Stannern and of Jonzac. It has been seen 

 how close an analogy allies them with certain aluminous lavas formed of pyrox- 



* It these meteorites have been thus t'oriiied, there iiuist have been water produced ou the 

 surface of the bodies of which they made a part, lint those bodies may uot have been able 

 to preserve the water by reason ot their small dimensions. Further, the reduction, if it lias 

 taken place, could only have been partial ; for, in f^eneral, the iron is but reduced in part, 

 whether in a metallic slate or in the state of a suljjhurnt or plmsphuret ; another part of this 

 same nutal is ordinarily combined, as protoxide, in a silicate, and also iu the state of chio- 

 uiated iron, (chromitc of protoxide of iron.) 



