Origin of the ArcJican Igneous Rocks. — WinclieU. 301 
alkaline magma, it is seen at once to consist in the presence of 
potassium, which is wanting in the normal ferro-magnesian 
magma. Another characteristic is the prevalence of free 
quartz. Let it be supposed that the crenitic operation began 
to act on a doleryte, which Hunt supposes was probably the 
original rock, it may be asked, how could there be leeched 
out of doleryte a large amount of potash? It may alsO' be 
asked how, while the potash was being extracted, and also the 
necessary amounts of lime, sodium and of aluminum, there 
could be obtained from the same rock an excess of silica 
which would give rise to a large amount of free quartz in 
the resulting product. If it be said that under exceptional 
conditions such a result might be produced locally, it is not 
sufficient, for exceptional and local conditions are not those 
demanded by the hypothesis. The extraction of the char- 
acteristic elements of the acid-alkaline magma from the ferro- 
magnesian is to be considered the usual and normal operation, 
for there is no reason in assigning the origin of one of the 
greatest rock masses of the globe to the operation of abnormal 
and exceptional conditions. It may be confidently asserted 
that no potash could be obtained from a rock or from a 
magma in which none existed, and therefore that the chief 
characteristic of the alkaline magma must have had some ex- 
traneous origin. 
It is not to be inferred from this that Hunt overlooked the 
necessity of accounting for the origin of the alkaline silicates. 
It is necessary, therefore, to look a little more closely into his 
facts and his inferences from them. This, however, may be 
confined to his discussion of the secondary origin of orthoclase 
in connection with the diabase rocks of the Keweenawan, as in 
this mineral are contained all the elements of the investiga- 
tion. It is plain that if orthoclase is found as a secondary 
mineral, as a result of change in a diabase, even in small 
amount, it is presumptive evidence of the extraction of potash 
from the associated diabase. 
It is true that, prior to Hunt's announcement of the cre- 
nitic hypothesis, it had been shown by J. D. Whitney that 
orthoclase occurs in veins and sometimes in vesicular cavi- 
ties in the trap rocks of the Keweenawan,* and this had also 
*Am. Jour. Sci., (2), XXVIII, 16, 1859. 
