PDF - Wallace Online

462 TROPICAL NATUREshapes of the primary wing-feathers, in the relative length ofthe hind toe, or in habits of life, as in roosting and buildingon trees. But the above objection shows how completelythe principleof selection has been misunderstood. It is notlikely that characters selected by the caprice of man shouldresemble differences preserved under natural conditions, eitherfrom being of direct service to each species, or from standingin correlation with other modified and serviceable structures.Until man selects birds differing in the relative length of thewing-feathers or toes, etc., no sensible change in these partsshould be expected.. . . With respect to the domestic racesnot roosting or building in trees, it is obvious that fancierswould never attend to or select such changes in habits."Studies of Cultivated and mid PlantsStill more remarkable, perhaps, is the collection of factsafforded by plants, which can be so much more easily cultivatedand experimented upon than animals, while the generalphenomena they present are strikingly accordant in the twokingdoms. As an example of the great mass of facts affordedby horticulture, he records that three hundred distinct varietieswere produced, in the course of fifty years, from a singlewild rose (Eosa spinosissima). We find in these volumesenormous collections of facts on bud- variation, or the occurrenceof changes in the flower or leaf -buds of full-grownplants, from which new varieties can be and often are produced;and, after a most full and interesting discussion ofthe cases, it is shown that some are probably due to reversionto an ancestral form, others to reversion to one parent whenthe plant has been derived from a cross, and others, again, tothat spontaneous variability which seems to be the universalcharacteristic of all living organisms.Three very interesting chapters are then devoted to thesubject of inheritance, and a host of strange and heretoforeinexplicable facts are brought together, compared, and classified,and shown to be in accordance with a few general principles.Then follow five chapters on crossing and hybridism,perhaps the most important in the whole work, since theyafford the clue to so much of the varied structure and complexrelations of animals and plants. Notwithstanding the