The places where the developing embryos depart from the common mode show where the several divisions took leave of one another in their evolution,--a point that comes out with great clearness when the facts of mammalian development are broadly compared. The embryos of carnivora and rodents and hoofed animals are alike in their earlier development, and their agreement means a community of origin. At a certain point the cat and dog depart from the common mode, but they remain alike up to a far later stage than the one in which they are similar to the embryos of rats and sheep. The rat and squirrel and rabbit, on their part, remain together until long after they take leave of the carnivora and ungulates; while the sheep and cattle and pigs have their own branch line, which they follow in company after leaving the embryos of the other orders. The reasons for these facts seem to be that the members of the three orders exemplified have evolved from the same stock, which accounts for their embryonic similarity for a long time after they collectively come to differ from amphibia and reptiles, while the members in each order became differentiated only later, wherefore their embryonic paths coincide for a longer period. Thus the degree of adult resemblance which indicates the closeness of relationship corresponds with the degree of embryonic agreement; that is, the cat and dog are much alike and their modes of development are essentially the same to the latest stages, while the cat and horse agree only during the earliest and middle stages, and their lines diverge before those of the cat and dog on the one hand, or those of the horse and pig on the other. * * * * * Like the fundamental principle of comparative anatomy in its sphere, the Law of Recapitulation, formulated as a summary description of the foregoing and similar facts, is one that holds true throughout the entire range of embryology and for every division of the animal series, however large or small. We have discussed its broader application, and now we may take up some of the more or less special cases mentioned in the earlier section of the present chapter, to see how it may work in detail. The flounder was noted as a variant of the fish theme which seemed to be a descendant of a symmetrical ancestor because its structural plan was like that of other bony fishes. If this be true, and if in its development a flounder must review its mode of evolution as a species, the young fish ought to be symmetrical; and it actually is. The grotesque skate and hammerhead shark were demonstrated to be derivatives of a simpler type of shark; their embryos are practically indistinguishable from those of ordinary dogfish and sharks.