Starting with this theory of evolution as a basis, the structure of certain families was studied and they were taken as types with which other related families were compared. After a classification of all known flowering plants into gymnosperms and angiosperms, and subdividing the latter into monocotyledons and dicotyledons, the lily family was considered as typical of monocotyledons. It and its related families afforded a simple means of demonstrating the problems under consideration. Members of this family were found to be characterized by having an endogenous stem, usually parallel veined leaves, six-parted perianth free from a three-celled superior ovary, and six stamens. The allied families were shown to agree with the type in the internal or fundamental characters, such as the number of carpels and cells of the ovary, but were found to differ in the more external or environmental characters, such as the arrangement of the parts of the perianth. After studying the relations between the various groups of endogens, the trees and weeds of the apetalous division of exogens were next considered, and through _Ranunculaceæ_ connected with polypetalous dicotyledons. These latter were classified according to whether the parts of the flower were hypogenous, perigynous, or epigynous. These terms signify, respectively, under the pistil, around the pistil, and on the pistil. In this group the rose family presented several modifications of the pistil, according to which it was divided into tribes. When the group of _Gamopetalæ_ was studied, _Solanaceæ_, the nightshade family, with its regular flower, and _Labiatæ_, or mint family, with irregular flower, were taken as types with superior ovaries. Various modifications from these types were found in several families. _Ericaceæ_, the heath family, presented, in its suborders of _Ericineæ_, _Pyroleæ_, and _Monotropeæ_, which had superior ovaries, and _Vacciniæ_, which had inferior ovaries, an intermediate order between the preceding _superæ_ and following _inferæ_, of which latter group _Campanulaceæ_ was considered a type. The relations between many families were traced, and the _Compositæ_ were lastly considered, this family showing the greatest differentiation with its coalescence of circles, adnation of different circles, reduction in parts, and number of individuals brought together. The greatest deviation from a simple flower and a complexity of structure were here presented. Through the co-operation of parts these flowers were of high physiological efficiency. Throughout the course, families of medicinal or other economic value, or such as presented evidences of adaptation for cross-fertilization, dissemination of seed, life in desert regions, or contained examples of parasiticism or many poisonous genera, were incidentally considered.