MAGMATIC ALTERATION OF CERTAIN MINERALS 261 
crystallize out of the magmatic solution and take on again its 
original outlines, or at least the crystallographic outlines of horn- 
blende, with changed molecular and molar structure; while the 
magmatic solvent, much of which is also certainly capable of 
crystallization under the same conditions, abandons the horn- 
blende and solidifies as groundmass crystals or basis. 
On the “magmatic” hypothesis we should expect the fre- 
quent occurrence of glass basis or groundmass among the grains, 
an expectation in which we are universally disappointed. It is 
true that we find other minerals than the two principal ones 
among the alteration products, but their formation is easily 
understood on other grounds. 
Conditions of alteration —Let us now examine briefly the con- 
ditions under which the alteration does or does not take place ; 
which may be broadly divided into two classes—chemical and 
physical. 
The chemical conditions relate to the magma in which the 
crystals are formed and in which they float, and for reasons to be 
seen presently we shall confine ourselves in discussing them to 
the volcanic rocks. 
On examining the whole range of volcanic rocks the most 
striking fact that we notice is that the alteration of hornblende 
and biotite is chiefly confined to the intermediate group—those 
with from 55 to 65 per cent. of silica—the trachytes, andesites, 
phonolites and tephrites (both leucitic and nephelinic). It is 
of comparatively rare occurrence in the two extreme groups, the 
acid and the basic rocks, though more frequent in the latter than 
in the former. 
The alteration among intermediate rocks, especially the ande- 
sites, is so extremely common that it is the exception rather than 
the rule to find specimens which show no, or even very few, altered 
crystals. Among these rocks we, expect beforehand to find a 
greater or less amount of alteration of the hornblende and biotite 
(excepting ina few structural types), and are seldom if ever 
disappointed in our expectations. Indeed in many regions’ the 
*OEBBEKE, Phillipines. Neu. Jahrb., B. Bd. 1, 460; SIEMIRADZKI, Ecuador, ditto, B. 
