584 SYSTEMATIC RELATIONS OF FISHES. 
always attached to the cranium, on the other hand, among true 
fishes, for in the eels it is quite as in the sharks, and the spinous- 
finned Mastacembelus presents the same features. 
The characters presented by the pelvic bones and limbs seem to 
be of higher import. Thus all the bony fishes and sturgeons lack 
all the pelvic elements. In the sharks and rays they are also want- 
ing; but two elements on each side appear in the Holocephali, 
(Chimera) according to Leydig and Gegenbaur. In Lepidosiren 
a large median pelvic cartilage exists, but which element it repre- 
sents is unknown. This is evidently a character of high signifi- 
cance. As to the limbs, the peculiarities of Polypterus have been 
pointed out above. They mean nothing less than the develop- 
ment of the elements of the arm and leg of the higher vertebrata 
which intervene between the point of articulation and the distal 
segments, in Polypterus and the sharks and rays. In the former 
the distal segments are articulated exclusively to the extremities 
of the proximal pieces, which thus resemble, as well as represent, 
humerus and femur, and render the limb pedunculated. The prox- 
imal pieces are not continued distally, however, into the represen- 
tatives of the main axis, which, as demonstrated by the admirable 
studies of Gegenbaur, consist after humerus, of radius, tarsals 
and metatarsals, and thumb ; in the hind limb, of the line of the 
tibia and inner toe. This continuation is observed in the Elasmo- 
branchi, where, however, the divergent segments extend along the 
sides of the proximal pieces to near, in some Rajide quite, to the 
articulation with the scapular arch. In the true fishes, including 
some of the old ganoids already considered, the divergent rays 
always reach this articulation, while the number of proximal or 
basal pieces is diminished. These pieces have been called by 
Gegenbaur the metapterygium (humerus), mesopterygium, and 
propterygium ; the first being axial, the second and third being 
divergent from it. In Polypterus the propterygium and mesop- 
terygium are largely developed ; in sharks and rays the proptery- 
gium is sometimes small, sometimes wanting, while in the true 
fishes the propteryginm and mesopterygium are both wanting, ex- 
cepting in Amia, Lepidosteus, and the sturgeons, where a cartil- 
aginous mesopterygium exists, according to Gegenbaur. This 
author finds it rudimental in young Salmonide and Siluride. 
Lastly, in the true fishes the distal elements of the axis of the 
imb are wanting, just as in Polypterus. 
