MAGMATIC ALTERATION OF CERTAIN MINERALS 275 
to the groundmass augites, and further that it is unscientific to 
connect closely a hornblende-free augite-andesite with a horn- 
blende-andesite because part of the augite of the former could 
have been possibly in the condition of hornblende at an earlier 
magmatic period. 
While not wishing to be in any degree polemical I must differ 
in this matter from Professor Zirkel, as it seems to me to be an 
inevitable conclusion from the facts that many of the ground- 
mass augites in certain andesites have been so derived from 
altered hornblende substances. Muigge’s example of the Azores 
trachytes already cited shows also that such derivative grains 
may attain the size of phenocrysts and hence be reckoned 
as such, through accretion of crystallizing augitic substances. 
Such a derivative origin must not, however, be attributed to all 
phenocrysts nor to the well formed augite-microlites and ground- 
mass crystals. Nor is such an attribution necessary, as I have 
already pointed out that hand in hand with the alteration of 
hornblende may go on the crystallization of augite out of the 
magma. It may be mentioned here that Zirkel™ speaking of the 
hornblendes in pyroxene-andesites says that they give the impres- 
sion that they had nothing to do with those conditions of solid- 
fication in which the rock-mass as it exists at present was pro- 
duced, from which the inference is to be drawn that they have 
been formed at a much earlier period. 
But when Zirkel brings the charge of being unscientific 
against the idea of the derivation of augite-andesites from horn- 
blendic magmas, he touches upon what seems to me to be a vital 
and important point in the question. If it were sweepingly 
asserted that all, or even the most, of the augite-andesites were to 
be connected with hornblendic magmas? his reproach would be 
justified, as being an unwarranted and too broad a generalization 
from the facts. But if it can be shown that certain augite-ande- 
sites are of independent origin while others are derivative from 
a previously hornblendic magma, then we have advanced a step 
*ZIRKEL, Lehrb. II, 811. 
7 As VON LASAULX perhaps seems to do. 
