MAGMATIC ALTERATION OF CERTAIN MINERALS 269 
rupting force of a surrounding moving magma. In most cases, 
however, the mass will be somewhat incoherent, so that it will 
be broken up by the action of the current, yielding the rounded 
forms so generally observed, and leading to other consequences 
of importance. 
It will also be seen that there is nothing in the theory incom- 
patible with a true resorption or solution of the hornblende or 
biotite substance if it should be enveloped in a part of the 
magma in which it is soluble under the conditions prevailing. In 
this case, however, we would look for embayments and other 
true corrosion phenomena, with a disappearance of the dissolved 
substance, while the pseudomorphism process could go on at 
another, or the same, time. 
It may seem, at first sight, that the starting of the alteration 
at the surface and its gradual progress inwards is a_ serious 
objection to this theory, in accordance with which we might 
think that the change should take place all at once through the 
whole crystal. But when we come to examine other cases of 
paramorphism such as the change of monoclinic to rhombic sul- 
phur, or wrought to cast iron, we find that the changes here take 
place from certain definite points or surfaces (generally the 
latter), so that this uniform progress of alteration need not sur- 
prise us, especially in view of the fact that the alteration evi- 
dently requires time, as in all similar changes. 
It is possible also that this method of progress is dependent 
on the hydrogen or fluorine content of the two minerals, the 
attraction between their molecules and those of the other con- 
stituents being lessened by diminished pressure at the tempera- 
ture of the liquid magma, so that they tend to dissociate, thus 
causing a molecular disintegration and starting the molecular 
rearrangement. Such an action in their case would naturally 
start at the surface or along cracks, the walls of which need 
not be in actual contact with the magma. This action of the 
gaseous contents of the minerals would also explain the cases 
of non-alteration, in elsewhere altered crystals, where abutting 
against other phenocrysts. 
