316 
Syenite and granite. — The syenitic magmas are denser 
and generally solidify at a higher temperature than do the 
granitic ones. The contact relations between a syenite and a 
granite belonging to one cycle of igneous activity will, there- 
fore, depend upon the same kinds of factors as are mentioned 
above for gabbro and syenite, and the granite will usually seem 
to be younger than or contemporaneous with the syenite, what- 
ever the original order of intrusion may have been. 
After the coming to place of the magmas, crystallization 
by reason of the difference as to chemical composition between 
crystals and magma, will involve ‘secondary differentiation 
processes’. ‘These processes will tend generally to accentuate 
the evidence that granite is younger than syenite, and the 
residual magma from the crystallization of the latter rock may 
give rise to the formation of quartz bearing dykes or veins 
within it. 
Summary. — From the above discussion it will be seen 
that the sequence of eruption of associated igneous rocks may 
most safely be ascertained from the succession of lava flows, or 
from the intersections of their dykes and minor intrusions. Studies 
of this kind have been pursued even since the beginning of the 
science of field petrology, and their importance has been em- 
phasized by у. Ricurnoren, Ippines, and many others. Of late 
years attention has been paid to abyssal rocks and the pheno- 
mena of their mutual contacts have been observed. For these 
rocks a sequence corresponding to decreasing basicity has 
proved to be the usual rule, and it has been assumed that 
this rule indicates the sequence of intrusion, and with it per- 
haps also the sequence of differentiation. The interpretation, 
however, of the contact relations of large abyssal rock bodies 
involves special difficulties which appear to have been to some 
extent overlooked. For lavas and dyke-rocks sequence of con- 
solidation of course is identical with sequence of protrusion, 
but with abyssal rocks no such connection needs necessarily 
