If this study is more satisfactory in its results than that of other fossil remains of animals, it is also beset with more numerous difficulties. Fossil shells usually present themselves in an entire state, and with all the characters requisite for comparing them with their analogous species, preserved in the collections or figured in the works of naturalists. Even fishes present their skeleton more or less entire; the general form of their body is almost always distinguishable, and most commonly, also, their generic and specific characters, which are drawn from their solid parts. In quadrupeds, on the contrary, even should the skeleton be found entire, it would be difficult to apply to it characters derived, for the most part, from the hair, the colours, and other marks which have disappeared previous to their incrustation. It is even excessively rare to find a fossil skeleton approaching in any considerable degree to a complete state. The strata, for the most part, only contain separate bones, scattered confusedly, and almost always broken, and reduced to fragments; and these constitute the only resources of knowledge to the naturalist in this department. It may also be stated, that most observers, deterred by these difficulties, have passed slightly over the fossil bones of quadrupeds; have classed them in a vague manner, according to superficial resemblances, or have not even ventured to assign them a name; so that this part of the history of fossil remains, although the most important and most instructive of all, is, at the same time, that which has been the least cultivated[81]. _Principle by which this determination is effected._ Fortunately, comparative anatomy possesses a principle, which, when properly developed, enables us to surmount all the obstacles. This principle consists in the mutual relation of forms in organised beings, by means of which, each species may be determined, with perfect certainty, by any fragment of any of its parts. Every organised being forms a whole,--a peculiar system of its own, the parts of which mutually correspond, and concur in producing the same definitive action, by a reciprocal reaction. None of these parts can change in form, without the others also changing; and consequently, each of them, taken separately, indicates and ascertains all the others. Thus, if the intestines of an animal are so organised as to be fitted for the digestion of flesh only, and that flesh recent, it is necessary that its jaws be so constructed as to fit them for devouring live prey; its claws for seizing and tearing it; its teeth for cutting and dividing it; the whole system of its organs of motion, for pursuing and overtaking it; and its organs of sense for discovering it at a distance. It is even requisite that nature have placed in its brain the instinct necessary for teaching it to conceal itself, and to lay snares for its victims.