4. Summary: In this paper economic mineral deposits are regarded as rocks, a view which makes unavoidable the inference that ore genesis is a part of petrology, and that any scientific system of genetic classification that applies to rocks should apply also to economic mineral deposits.

A consideration of fundamental principles shows that petrography has its scientific basis in dynamical geology, and that any genetic scheme of classification for rocks should also be a geological scheme. An appeal to historical considerations likewise shows that the requirements of science, as regards the classification of rocks, seem to lie in the direction of a scheme based on what may be called genetic-geological principles.

A scientific classification in accordance with these principles presupposes the possibility of a geological classification of formative agents and processes. These processes fall into certain natural groups and sub-groups according to the way in which they operate on and modify the rocks of the earth's crust. Having decided on a definite geological grouping of formative processes in this way, the grouping of rocks with reference to the processes that have determined their type characteristics becomes possible, the rock-groups corresponding to process-groups.

The genetic classification of rocks and economic mineral deposits is thus essentially a classification of formative processes, and a correlation of these processes with the rock types the intrinsic characteristics of which they have developed. The use of intrinsic data and other non-genetic data in petrography finds its place in the definitions of rock types, but these definitions should be controlled by genetic and geological considerations.

Petrologists have for the sake of convenience, but with lack of wisdom, refused to regard most economic mineral deposits as rocks. This attitude on their part made possible, at a comparatively early date, an approximation to a sound genetic-geological scheme of grouping for rocks. The adoption of such a scheme for ore deposits was prevented by uncertainty as to the origin of subterranean solution deposits which fill an important place in ore genetics.

It has been known for a long time among petrologists that certain metalliferous vein deposits are genetically connected with igneous intrusions. The studies of recent years in ore genesis have largely extended the significance of solutions of deep-seated origin in connexion with ore deposition. There are, nevertheless, certain vein types which are known to arise from solutions of surface origin ; and although the criteria for determining the origin of vein deposits are not fully established for all types, it seems permissible to adopt a genetic scheme in which processes are grouped geodynamically in two broad groups according to their internal or external origin in relation to the earth's crust. Such a scheme of division into endogenetic and exogenetic processes is adopted in this paper, and the subdivision is made in accordance with the geological action of the processes concerned.

The corresponding scheme for rocks may claim to be scientific. It is based on a well-defined genetic method. It comprehends all rockforming processes, and finds a place for all types of rocks. The fact that uncertainty exists as to the nature of certain processes and the origin of certain types should be no bar to the adoption of such a scheme up to the limits of known facts and well-established theories, since it is one that will adapt itself to the progress of petrology.