702 STEINAR FOSLIE 
It may be noted here that these percentages are not propor- 
tional to the areal distribution of the rocks at the surface, because 
the quartz-norite and the normal norite without doubt are concen- 
trically arranged throughout the field, so that any section through 
the center of the field will give the same image. Accordingly, 
their relative volume is with the greatest approximation calculated 
here after the formula for concentric spheres. 
After the results given above, stating the quantitative propor- 
tion and the chemical composition of the main groups of rocks, we 
are in the rare and happy position to be able to state—with a fair 
amount of accuracy—the mean composition of the whole field, 
representing the original, undifferentiated eruptive magma. This 
composition is given in column o, in the table of analyses. The 
most remarkable feature is the very high magnesia content in 
connection with the relatively high silica content. It is an ideal 
norite magma. 
The regular, stepwise development of the differentiation products ° 
is best seen from the norm calculations. Not only the chemical 
composition changes along definite lines, but also the composition 
of the individual minerals, especially the plagioclase and the 
rhombic pyroxene. Very remarkable is the regular change in the 
ratio MgO™!:FeO™! in the silicates, decreasing markedly toward 
the salic rocks. The reason for this appears from what follows. 
THE DIFFERENTIATION PROCESS 
Starting with a study of the known forsterite-diopside-silica and 
the forsterite-anorthite-silica diagrams,’ we learn that in the mean, 
undifferentiated magma of the Raana field, nearly without olivine 
in the norm, olivine will anyway be the first mineral to crystallize— 
beginning at the border toward the surrounding schists—and it 
will continue to separate out until the eutectical line olivine- 
pyroxene isreached. By further cooling, rhombic pyroxene crystal- 
lizes out at the same time as the already formed olivine crystals 
are resorbed. 
If this process continued undisturbed, the final result would be 
an olivine-free rock, corresponding to the norm. However, we find 
*See N. L. Bowen, Jour. Geol., Vol. XXIII, suppl., pp. 20 and 29. 
