264 A. Schrauf and E. 8. Dana—Thermo-electrical 
investigated, all of which showed the same abnormal behavior. 
The outer shell, 2mm. in thickness, on all the planes, was neg- 
ative, while the kernel was always positive. If 2mm. of the 
exterior be filed away, the — shell passes gradually into the + 
sign. Such a case especially in the orthorhombic system does 
not allow of hemihedrism being assumed as an explanation. 
Taking into consideration the great variation in the specific 
example, gave approximately 195 per cent Co. The varia- 
tons are consequently, in all probability, caused by the ele- 
ments Fe, S, as in danaite. 
It was a matter of great regret to us that we were unable to 
obtain for investigation any of the glaucodot from Chili. The 
dent from the specific gravity that the analysis * of the glauco- 
dot from Hakansbé was made from a fragment containing 
both shell and kernel. 
n_ observation of Tschermak + adds plausibility to this 
hypothesis of a variation in the amount of Fe. He describes 
crystals of cobaltite imbedded in the outer portion of a crystal 
of glaucodot. In the transition from glaucodot to cobaltite, 
the elements Fe and Co are alone involved, and then only in 
their relative proportions, thus: 
As Co Fe 
Glaucodot 440 19°8 16:0 19°3 Ludwig 1. ¢. 
Cobaltite 43°4 20°8 33°1 3-2 Stromeyer f 
E. Galenite. 
Kobellite 
Sardinia | i: Pribram PbeBi.Sb28i2 
granular. England. crystals, 0 
Shetia 7-428 7575 
; An analysis by v. Kobell agrees completely with the former 
Tschermak, Sitzungsb, d.k. Akad. Wain i bv, 429: 1067 
| Stromeyer Soke jt E Abed. Vis ae ee ; 
‘ 
