THE FELIX 3IETE0RJTE— MERRILL. 
197 
From these analvses the total chemical composition of the entire 
stone was calculated as follows: 
Fe 2. 59 
Ni 0. 36 
Co 0. 08 
Cu 0. 01 
SiO, 33.57 
AlA ---- 3.24 
CrA 0.80 
FeO 26.22 
FeS 4. 76 
MnO 0. 68 
XiOctCoO 1.01 
CaO 5.45 
MgO 19.74 
K^O 0.14 
N a.,0 0. 62 
C (Graphite) 0.36 
H.,Oatll0° 0.16J 
Metallic portion =3. 04 per cent. 
Stony portion =96. 96 per cent. 
99.79 
Specific gravity, at 30° c, as determined by Mr. Tassin, 3.78. 
The mineralogical composition may therefore be given as follows: 
Per cent. 
Metal 3. 04 
Troilite 4. 76 
Chromite 1.17 
Graphite 0. 36 
Soluble silicate ( olivine in part ) 72. 60 
Insoluble silicate (enstatite and augite in part) 18. 07 
100. 00 
There are certain points of these analyses which I am unable, at 
present, to satisfactorily explain. The insoluble portion ma}^ be con- 
sidered as essentially enstatite and an aluminous monoclinic pyroxene, 
and the soluble portion as largely olivine. But the high per cent of 
iron protoxide (FeO) as well as the lime and alumina in this latter 
portion, are not easily accounted for. It is possible that the last two 
elements may be constituents of the colorless, undetermined mineral 
referred to, but the source of the iron protoxide is for the present 
unexplainable. 
The case is, however, not without precedent, J. Lawrence Smith 
reporting^ similar conditions in the Warrenton, Warren County, Mis- 
souri, stone, which, hoAvever, he allows to pass without comment. 
^ Original Researches in Mineralogy and Chemistry, p. 532. 
