Different parts of the same continuous rock-mass frequently display chemical and mineralogical variations, resulting from a process of differentiation, or segregation, of the magma. How this is brought about, is far from certain, but there can be little or no doubt as to the fact. "When large areas of eruptive rocks are carefully investigated, it is found that there is a perfect and gradual transition of one kind into another - all intermediate varieties existing - and that quantitatively no special part of the series is universally predominant, although there are often immense masses of nearly uniform character, and there may be smaller bodies of quite variable composition." (Iddings.) It is further found that in a given volcanic district, or petrographi-cal province, the rocks erupted at a particular geological period have certain peculiarities which distinguish them from those of other provinces. It thus appears probable that the igneous rocks of such a province were derived from the same magma, and the relationship between the various kinds of rocks of the province is called consanguinity.

The existence of these petrographical provinces does not imply that the rocks of each one differ from those of all the others. In fact, similar or identical groups of rocks are found in many parts of the world, but each province differs more or less from the surrounding ones. Thus, rocks which from the genetic point of view are closely related, are, by any scheme of chemical or mineralogical classification, often placed in widely separated groups.