the changes which occur in the development of an individual, the successive forms of geological times, and the series of living forms, lower and higher, or more generalized and more specialized.
In the smallest gar here described, and presumably in still younger examples, the axis of the body, represented by the notochord or primitive vertebral column, is nearly horizontal, about midway between the upper and the lower borders of the tail. This is likewise the case with the lowest known vertebrate, Amphioxus; with the forms next above, the hag-fishes (Myxine and Bdellostoma) and lamprey-eels (Petromyzon); with the larvæ (tadpoles) of frogs and toads; and with the adults of the aquatic and tadpole-like salamanders, Menopoma and Menobranchus.
Finally, such a tail exists in the Dipnoans, or mud-fishes, of Africa, South America, and Australia (Frotopterus, Lepidosiren, and Ceratodus), which have some striking affinities with Batrachians, but are usually regarded as fishes, and are, perhaps, the best illustration of generalized forms.
To this variety of tail, Cope has applied the name isocercal; Huxley calls it diphycercal, and gives as an example Polyhyterus, where, however (as in Calamoichthys), the "end of the notochord is hardly at all bent up." Wyman, finding this kind of tail in the embryo of a skate, called it protocercal, and, on some accounts, this seems the more suitable name.
As the gars grow older, the relative length of the filament and the infra-caudal lobe constantly changes. At first the former is the longer; in a specimen 108 millimetres long, their tips coincide; in one 142 millimetres long, the lobe projects beyond the filament; and in a third, 300 millimetres long, the filament is much the shorter, is ragged and attenuated, and during life was rarely employed. This second stage, or rather series of stages, has several counterparts among living Selachians and Ganoids. The most accurate resemblance is presented by the shovel-nosed sturgeon of the Mississippi River (Scaphyrhynchus). The filament is excessively elongated in Chimera, and exaggerated as to both length and breadth in the thrashing-shark (Alopias). But, with many sharks, the common sturgeons, and the spoonbill (Polyodon), the size of the infra-caudal lobe is so nearly that of the filament as to give the whole tail a nearly symmetrical outline, and lead zoologists to speak of the "upper lobe," whereas it is really the bent-up end of the body. This kind of tail is called heterocercal.
The gars above mentioned are supposed to be the young of the Lepidosteus osseus. Just at what size the filament wholly disappears in that species is not known. But with the smaller and proportionally shorter species, L. platystomus, there is no sign of the filament when eighteen inches in length. The tail might then be thought, at first sight, to be symmetrical. But the longest rays are a little above the
