to its apparent homologue, the radius. Fig. _e_, a slim light bone, which narrows and thickens in the centre, and flattens and broadens at each end, was probably a scapula or shoulder-blade,—a bone which in most fishes _splices_ on, as a sailor would say, by squamose jointings, to the coracoidian bone at the one end, and the super-scapular bone at the other. As indicated by its size, it must have belonged to a small individual: it is, however, twice as long, and about six times as bulky, as the scapula of a large cod. [Illustration: Fig. 41. INTERNAL BONES OF ASTEROLEPIS. (One third nat. size, linear.)] [Illustration: Fig. 42. ISCHIUM OF ASTEROLEPIS. (One half nat. size, linear.)] Of the bone represented in fig. 42, I have determined, from a Cromarty specimen, the place and use: it formed the interior base to which one of the ventral fins was attached. In all fishes the bones of the hinder extremities are inadequately represented: in none do we find the pelvic arch complete; and to that nether portion of it which we do find represented, and which Professor Owen regards as the homologue of the _os ischium_ or hip-bone, the homologues of the metatarsal and toe-bones are attached, to the exclusion of the bones of the thigh and leg. In the Abdominales,—fishes such as the salmon and carp,—that have the ventrals placed behind the abdomen, in the position analogous to that in which the hinder legs of the reptiles and mammals occur, the ischiatic bones generally exist as flat triangular plates, with their heads either turned _inwards_ and downwards, as in the herring, or _outwards_ and downwards, as in the pike; whereas in some of the cartilaginous fishes, such as the Rays and Sharks, they exist as an undivided cartilaginous band, stretched transversely from ventral to ventral. And such, with but an upward direction, appears to have been their position in the _Asterolepis_. They seem to have united at the narrow neck A, over the middle of the lower portion of the abdomen; and to the notches of the flat expansion B,—notches which exactly resemble those of the immensely developed carpal bones of the Ray,—five metatarsal bones were attached, from which the fin expanded. It is interesting to find the number in this ancient representative of the vertebrata restricted to five,—a number greatly exceeded in most of the existing fishes, but which is the true normal number of the vertebrate sub-kingdom as shown in all the higher examples such as man, the _quadrumana_, and in most of the _carnaria_. The form of this bone somewhat resembles that of the analogous bone in those fishes, such as the perch and gurnard, cod and haddock, which have their ventrals suspended to the scapular belt; but its position in the Cromarty specimen, and that of the ventrals in the various specimens of the Cœlacanth family in which their place is still shown, forbids the supposition that _it_ was so suspended,—a circumstance in keeping with