The rays of the caudal fin articulate with the extremity of the backbone, but this portion of the fish’s anatomy undergoes such remarkable changes that we must devote a few words to it. It is probably well known to our readers that the tails of fishes exhibit three distinct forms. The first of these is a simple fringe formed by the union of unaltered dorsal and ventral fins; the second is the unsymmetrical or unequally lobed tail so characteristic of sharks, dogfishes, and rays; and the third is the broad symmetrical tail fin, often distinctly forked or bi-lobed, such as we meet with in the majority of our bony fishes. These three kinds are known respectively as the _diphycercal_, _heterocercal_, and the _homocercal_ tails. [Illustration: FIG. 228.--THE SKELETON OF A FISH (PERCH) _d_, dorsal fin; _p_, pectoral fin; _v_, pelvic fin; _t_, tail fin; _a_, anal fin] Now, it is an interesting fact that the most ancient fishes of our globe possessed tails of the first type; and that these gradually gave place to the heterocercal form; while the higher fishes of the present day nearly all possess the homocercal tail. Thus as time advanced the heterocercal tail was gradually evolved from the diphycercal, and the homocercal from the heterocercal. Further, if we watch the development of one of the highest fishes of the present day from its embryo, we find that similar changes take place in the individual. At first its tail is a simple fringe round the extremity of the backbone, the latter being straight, or nearly so, to the end, so that the embryo fish, as yet still in the egg, reflects a characteristic of its very early ancestors. Then the end of the vertebral column turns upward, and strong fin-rays are developed on its ventral side, so that the tail becomes a heterocercal one like that of the less remote ancestors of a later geological period. Next, the upward-bending portion of the vertebral column is slowly absorbed, till nothing of it remains except a small upturned bony spine, while, at the same time, the ventral lobe expands on the upper side until the tail fin is once more of a symmetrical form. [Illustration: FIG. 229.--THE INTERNAL ORGANS OF THE HERRING _a_, œsophagus; _bc_, stomach; _e_, intestine; _l_, duct of swimming bladder; _k_, air-bladder; _h_, ovary] Following these interesting changes, it becomes evident that the symmetry of the tail fin of the bony fishes is really a false one, the whole of it having been formed from the ventral lobe of a heterocercal tail; and although the backbone seems to terminate abruptly exactly opposite the middle of the fin, it still contains the remnant of the raised extremity of the backbone that ran to the tip of the dorsal lobe when the tail was of the heterocercal type.