The "philosophic naturalist" plainly requires just so much philosophy as is implied in keeping his eyes open, and, indeed, so long as species were believed to be separately created, and organic characters could be only correlatively and not genetically explained, there was nothing else for him to do. Natural History before Darwin was like Natural Philosophy before Newton; its inductions were incomplete, and the deductive procedure which could alone raise its constituent groups into sciences was impossible. It was at this stage in the development of Natural History that Mr. Bain took up its method, and set about applying it to the "Feelings." Its power in the hands of a keen and dispassionate observer is indisputable, and the two instructive volumes which contain Mr. Bain's systematic exposition are at once a treasure-house of observations of priceless value, and such a compendious generalization of mental facts of all orders into laws as doubtless marks the climax of the method. But it is fundamentally unscientific. If it be true that the higher forms of life and mind have been evolved out of the lower, then the most resolute introspection, and the most cutting analysis, with the help of stray observations of children, and some patient experimenting on animals, will go no appreciable distance in discovering mental constituents which may have had their origin in an indefinitely remote past. That this is not only a necessary result of the "natural history method," but that it has in point of fact resulted in Mr. Bain's treatise, it may be well to make clear. To keep the analogy in view, we again quote from Dr. Carpenter. "The naturalist," he says—
Now, such an endeavor to simplify, by the adoption of easily-recognized external characters as the basis of his classification, is a feature prominent in the fore-front of Mr. Bain's work. The mode of diffusion of an emotion, the institutions it generates, and its peculiarities as a state of consciousness—all of them the most manifest characters of the emotions—are avowedly adopted as bases of classification.[3] That easily-recognized external characters are not always "indicative of