108 Tlie American Geologist. August, i89i 
Hence, the centesimal composition — of the metallic portion of 
the spherules, would be : 
Iron 97.79 
Manganese 0.57 
Nickel 0.11 
Cobalt 0.23 
Copper 0. 10 
Sulphur 0.13 
Phosphorus 1.07 
Carbon ? 
100.00 
and that of the insoluble, non-metallic, residue ; 
Silica 93.95 
Alumina 1.13 
Ferric oxide 1.02 
Lime 0.62 
Magnesia 0.31 
. Loss?» 2.97 
100.00 
The limonite did not contain any, so to sa}-, denuded nuclei, so 
that supposing it to have resulted from the weathering of the me- 
tallic spherules, the change had in no instance proceeded so far 
as to effect the removal of the whole of the metallic covering. 
Some of the nuclei were of microscopic minuteness, hence it 
ma}- be infeiTcd that even the smallest of the metallic spherules 
contained a siliceous nucleus. The limonite gave no reaction for 
chlorine. 
From what has preceded it will be seen that the metallic look- 
ing spherules which were disseminated through the limonite, con- 
sisted of nuclei of silica coated with a humus-like substance which 
in turn, was overlain hs a metallic layer, the latter containing all 
the elements most frequentl}' met with in meteoric iron. The 
amount of phosphorus is relativeh' large, that of the nickel, how- 
ever, very small, consequently if it be regarded as present in the 
form of any of the varieties of nickel-iron or schreibersite, it 
would represent but coraparativel}- trilling amounts of either of 
these bodies. 
Owing to the fact that all meteoric iron contains nickel, the 
presence of that element in a native iron has generalh' been re- 
garded as evidence of its extra-terrestrial origin and for this rea- 
son the irons of Chotzen and Petropaulwosk although sometimes 
*The analysis was conducted upon a very small amount of material — 
the silica may, not improbably, be a little too low. 
