3o6 The American Geologist. May, 1900 
them allotriomorphic. And the third case combined the other 
two, overtaxing and the consequent change from the devel- 
opment of the phenocrysts with wide spacing to the develop- 
ment of the groundmass with narrow spacing having super- 
vened as increasing viscosity retarded molecular flow. 
The contrast of phenocrysts and groundmass has been re- 
garded as proving sudden changes in the physical conditions. 
Apart from this contrast we have no evidence that such 
changes occur ; and the probabilities of the case are entirely 
against them. But of course we cannot fairly or logically use 
the abrupt changes to explain the porphyritic texture in the 
absence of independent evidence that such changes are pos- 
sible. 
It ma)- occur to some to inquire why, if this theory be true, 
all igneous rocks, save the deeper plutonics, are not porphy- 
ritic; although this query would be almost equally pertinent 
under the old theory, for all intrusive and effusive types. As 
expressing an objection to the theory now under considera- 
tion it would seem to be especially applicable to the finer 
grained holocrystalline rocks, such as the fine grained gran- 
ites, diabases, etc. If the rate of cooling was such as to de- 
termine close spacing of the primary centers of crystallization, 
why did it not later, through increased viscosity, lead to a 
still closer spacing and the dilTerentiation of phenocrysts and 
groundmass? In reply it may be noted, first, that the over- 
taxing of molecular flow is, in the nature of the case, less likely 
to occur where the distances through which it must act are 
short than where they are long, especially if we may assume 
that the energy of the crystallizing force varies inversely as 
the square of the distance; and, second, that although fineness 
of grain indicates relatively rapid crystallization, due to rapid 
cooling, and tending directly to increased viscosity, it tends 
indirectly to diminished viscosity through a rapid evolution of 
the latent heat of crystallization and a constantly increasing 
proportion of water and other mineralizers in the magma res- 
iduum. In the slow crystallization of the coarse-grained rocks 
the latent heat and mineralizers are likely to be dissipated 
as fast as evolved; but they are not essential here to uniformity 
of grain; and in the fine-grained rocks, where they are es- 
.sential to uniformity of grain, their efficient cooperation is in- 
