These descriptive matters disposed of, we proceeded to discuss the various classes of surface features with a view to explaining the precise actions which appear to us to have led to their formation. Naturally the craters first demanded our attention. We pointed out the reasons for regarding the great majority of the circular formations of the moon as craters, as truly volcanic as those of which we have examples, modified by obvious causes, upon the earth; and, tracing the causative phenomena of terrestrial volcanoes, we showed how the explanations which have been offered to account for them scarcely apply to those of the moon: and thus, driven to other hypotheses, we endeavoured to demonstrate the probability of the lunar craters having been produced by eruptive force, generated by that pre-solidifying expansion of successive portions of the moon’s molten interior, which we enunciated in our third chapter. The precise course of phenomena which resulted in the production of a crater of the normal lunar type, with or without the significant central cone, were then illustrated by a series of step-by-step diagrams with accompanying descriptive paragraphs. And after treating of craters of the normal type we pointed out and explained some variations thereupon that are here and there to be met with, and likewise those curious complications of arrangement which exhibit craters superimposed one upon another and intermingled in strange confusion. From craters manifestly volcanic we passed to the consideration of those circular formations which, from their vastness of size, scarcely admit of satisfactory explanation by a volcanic hypothesis. We summarized several proffered theories of their origin, and pointed out what we considered might be a possible key to the solution of the selenological enigma which they constitute, without, however, expressing ourselves entirely satisfied with the validity of our suggestion. The less mysterious features presented by peaks and mountain ranges were then discussed to the extent that we considered requisite, viewing their comparatively simple character and the secondary position they occupy in point of numerical importance upon the moon. At greater length we dealt with the cracks and chasms and the allied phenomena of radiating streaks, pointing out with regard to these latter the strikingly beautiful correspondence in effect (and therefore presumably in cause) between them and crack-systems of a glass globe “starred” by an expanding internal medium.