220 PROCEEDINGS OF SECTION C. 
supposing movements to have taken place in the magma 
after the process of differentiation had become wholly or 
partially complete. The gradual passing over of gabbro to 
pure pyrrhotite is not, however, so frequently observed as 
is the case with the oxide concentrations. This is probably 
due in part to the low fusibility of pyrrhotite. Microscopi- 
cal research has proved that whereas the titanic iron ore 
and the chromite have solidified before the accompanying 
silicates, pyrrhotite has been the last to solidify. It is also 
highly probable that the fluid moleculesof the sulphide ores, 
as soon as they are formed, have a tendency to run together 
and separate from the gabbro magma in the same way that 
matt separates itself from molten slag. Any movements in 
the magma, from whatever cause arising, before the differen- 
tiated products had solidified might produce intrusions of 
ore into the gabbro or into the adjoining country-rock. The 
nickel pyrrhotite deposits are essentially a marginal phe- 
nomenon. In the great majority of cases they are found 
directly at the contact of the gabbro mass. Sometimes, 
however, they occur within the gabbro, but usually, even 
then, they are nearer the periphery than the centre. The 
intrusions into the neighbouring country-rock are seldom 
found more than a few hundred feet from the contact. One 
or two deposits of nickeliferous pyrrhotite are known in 
Tasmania, but so far have not been worked extensively. 
As further examples of metallic concentrations which are 
believed to be caused by magmatic differentiation, may be 
mentioned, the native iron in basalt on Disko Island on the 
west coast of Greenland; the nickel-iron alloy awaruite 
in pyroxenite at Awarua, New Zealand; the native 
platinum in olivine gabbro in the Ural Mountains. Gold 
has been recorded as an original constituent of granite 
from a great number of localities, as also native copper in 
basaltic rocks. In the two latter cases, however, the pri- 
mary nature of the metal can hardly be regarded as 
absolutely certain. 
The cause of magmatic differentiation is a subject of great 
complexity, and a very large number of theories have been 
put forward by eminent geologists. So far, however, these 
efforts have failed to supply an adequate explanation of the 
facts as recorded. Vogt discusses the various theories at 
length in a recent number of the Zeitschrift fur practische 
Geologie, and though he finds them all wanting, he comes 
to the conclusion that the chemical action of water in the 
magma must have been one of*the principal factors in 
bringing about magmatic differentiation. This is a very 
interesting conclusion, for if this becomes established, there 
