A E R

A E R

( 9 )

sequent ignition, hence this oxide contains 70per cent, of metal, but by the concurrence ofProust, Davy, Thomson, Berzelius and all thebest authorities, the per-oxide of iron containsabout 52 per cent, of metal, and the protoxide73 per cent. ; hence it is evident that Mr.Howard’s oxide was a mixture of the two, andwas not sufficiently calcined. Secondly, allow-ing with Mr. H. that the real quantity of ironwas 10.5 grains, this when combined with sul-phur in the only two proportions in which itis capable of combining, would produce 16.5grains of magnetic pyrites, or 21.5 of the com-mon kind. Again, there is the distinct testi-mony of Count Bournon that the pyrites wasnot magnetic, and the equally distinct testi-mony of Mr. H. that it was decomposable bymuriatic acid, and that a little sulphuretted hy-drogen was at the same time evolved. But ofthe two known kinds of pyrites the magneticalone is decomposable by muriatic acid withevolution of sulphuretted hydrogen. We musttherefore either suppose some error in the ana-lysis, or that the pyrites in question differs es-sentially from the only two that chemistry asyet acknowledges. In either case all that canbe concluded from the analysis is the presenceof much iron, of a little sulphur, and of lessnickel.

After examining the pyrites Mr. H. under-took the analysis of the metallic iron of theaerolite, which being treated in the way alreadydescribed in the article Iron , appeared to be analloy of about 17 parts iron and 6 nickel.

The globular concretions were then analysedby fusion with potash in the usual way andafforded 50 silex

15 magnesia34 oxide of iron

2.5 oxide of nickel

101.5

Lastly, the earthy matter, which served as acement for the three other ingredients, wasexamined in the same manner as the globularconcretions were, and yielded48 silex18 magnesia '

34 oxide of ironoxide of nickel

102.5

The excess in both these analyses is no doubtto be attributed to the difference of oxydationof the iron as existing in the substance and asobtained by analysis.

The component parts of the stones fromSienna, from Yorkshire, and from Bohemia ,being in too small grains to allow of their se-paration with sufficient exactness, a small por-tion of each specimen was pulverized and ex-amined in the usual way. They all affordedsilex, magnesia, the oxides of iron and of nickel,and nothing else; thus exhibiting a perfectanalogy with each other and with the aerolitefrom Benares.

It deserves to be remarked how striking aresemblance as to their component parts existsbetween the earthy parts of these aerolites, andthe granular peridot or olivine which is almostpeculiar to basalt •, this substance being com-posed according to Klaproth of

50 silex38 magnesia12 oxide of iron

It is further worthy of notice that the exter-nal characters of the small transparent globulefound in the stone from Sienna for the mostpart agree with those of the transparent globulesthat occur in the cells of the supposed meteoriciron from Siberia ; and that these are alsostrikingly similar to those of olivine.

It remains to take some notice of a few otheraerolites of still more recent origin, and whichhave exhibited certain peculiarities in theiranalysis.

8. On a the 12th of March, 1798, the airbeing calm and without clouds, the inhabitantsof Valence, in the Lyonnais,'and its neighbour-hood, were alarmed with the appearance of ameteor, accompanied by a singular noise, andexhibiting partial explosions. It was seen tofall in a vineyard, and on the day after, whenthe terror of the peasants had in a degree sub-sided, an officer of the police being present,search was made in the vineyard, a recent ir-regular hole of the depth of about 20 inchespresented itself, at the bottom of which wasfound a black stone weighing above 20 lbs. Onbeing broken it appeared to be composed of agranular ash-grey earthy matter, in which wereimbedded grains of metallic iron, and of palereddish yellow pyrites, with globular concre-tions of a grey colour, and some small irregularmasses of a substance resembling olive coloured

VOL. HI.

Journ. de Phys. 1802.

»