THE GENESIS OF ORE DEPOSITS 495 
Of the first type we have an excellent example in the marginal 
phase of the gabbro of Carrock Fell, as described by Harker. 
Here a well-marked segregation of titaniferous iron ore is found 
along both steeply inclined margins of a laccolith of considerable 
size, believed to be still more or less in the same position as when 
intruded. From the figures and description given by Harker it 
is clear that gravity is excluded, and the only possible explanation 
is some kind of diffusion to the cooling surfaces during crystalliza- 
tion. The concentration is not very strongly marked, as the most 
concentrated type of ilmenite-gabbro has only about 25 per cent 
of iron ore, but at any rate the principle is clear. A very peculiar 
type is the great mass of titaniferous iron ore at Taberg in Sweden, 
which forms the central portion of a boss of ultrabasic rock, usually 
described as hyperite, that is, olivine-norite in modern terms. 
The reason for this reversal of the normal sequence is unknown. 
It will doubtless be generally conceded that it must be difficult 
on field evidence alone to distinguish between cases of heterogeneity 
due to gravity-sinking and to immiscibility in the liquid state. In 
both cases, on complete solidification the heavier rock will be found 
below, and owing to the high viscosity of fused silicates no very 
sharp line of demarcation is likely to be seen. It is moreover 
highly probable that in some instances both causes have been 
operative. Bowen has brought forward much evidence in support 
of differentiation by gravity-sinking, and Vogt long ago showed 
that fused silicates and sulphides possess very strictly limited 
miscibility... In the well-known instances of Sudbury and Insizwa 
we have thick intrusions ranging from acid or intermediate at the 
top to basic or ultrabasic below, with a well-marked layer of 
sulphides at the bottom. Here it may be suggested that the 
gradation in the silicate rock is due to gravity with an immiscible 
separation of sulphide at the base. It seems pretty clear from 
Daly’s latest observations that the concentration of magnetite and 
apatite at Kiruna is due to gravity-settling of rather large units of 
differentiation, since many blocks of ore have been caught and 
fixed at higher levels by increasing viscosity. The occurrence of 
limited miscibility in silicate solutions is a much disputed point, 
1 Vogt, Die Silikatschmelzlésungen, Part I (Kristiania, 1903), p. 96; ‘‘Die Sulfid- 
Silikatschmelzlésungen,” Norsk geologisk tidsskrift, 1917. 
