546 TRANSACTIONS OF SECTION D. 
was thus associated with such extinct forms as Osteolepis, Glyptopomus, and 
Holoptychius, a position in which it has been left by subsequent writers, 
But comparing Polypterus with these fossil fish, we find that the resemblance 
is by no means close. The similarity in the disposition of the surface bones of the 
skull is only general, and such as may be found in most primitive teleostomes. 
The scales, though like those of osteolepids in outward appearance, are really of 
very different structure: the latter are covered with an outer layer of cosmine, 
and do not grow by the addition of complete layers of bone below and ganoine 
above. The scales of Polypterus, on the other hand, are of the true ganoid type, 
resembling closely those-of the palzoniscids, and grow like these by the addition 
of concentric layers of bone and ganoine. 
That the tail is not truly diphycercal is shown by its internal structure. A 
heterocercal fin has a well-developed ventral or hypochordal lobe, which is always 
supported by direct prolongations of the hemal arches, whereas other median fins 
have always separate distal radial elements. Now in a diphycercal caudal fin, 
such as the dipnoan or crossopterygian, both the ventral and the dorsal lobes 
have such separate radials jointed on to the vertebral arches. If we examine the 
tail of Polypterus we find that the notochord does not nearly reach the extremity ; 
that its extreme end is, in the young at all events, slightly twisted upwards (as 
shown by Budgett); and, lastly, that the hypochordal lobe is supported by direct 
prolongations of the hzemal arches (as described long ago by Kolliker). The 
evidence clearly points to the caudal fin of Polypterus being of a modified hetero- 
cercal type. 
Coming now to the gular plates, Polypterus alone among living fish has a large 
ventral pair. Such plates are found among early extinct Dipnot and Crosso- 
pterygu: in these the large pair appears not to belong to the lateral series, but 
occupies the ventral region between them. In the Actinopterygii there may 
occur a median ventral plate with two lateral series as well (Ama). Now two 
of the anterior paired plates may be more or less enlarged (Paleoniscide) ; and it 
therefore becomes a question whether, after all, the paired plates of Polypterus are 
not the homologues of these plates of the Actinopterygii, instead of the homologues 
of the more median plates of the Crossopterygii. This suggestion is supported by 
the discovery of a median gular fold between the plates of Polypterus. 
Already some years ago ' it was pointed out that the skeleton of the pelvic fin 
and girdle of Polypterus is much more actinopterygian than crossopterygian in 
structure. With regard to the more important pectoral fins, the resemblance to 
the lobate fins of the Crossopterygii has been shown by Budgett ? to be superficial 
only. For in these extinct forms the endoskeleton possessed a jointed well-defined 
axis, with probably a series of radials on each side of it ; while in Polypterus the 
radials are grouped together in fan like arrangement, with an ill-defined posterior 
axis, much as in the lower Actinopterygii. 
The evidence in favour of a close affinity with the extinct Crossopterygit is 
therefore neither convincing nor even strong. On the other hand, the theory that 
Polypterus is nearer to the Actinopterygii is supported by a comparison of the 
structure of the scales and of the fins, by the presence of large solid otoliths in the 
ear and of double dorsal nostrils on each side of the snout. The study of the 
anatomy of the soft organs, though of little use in a comparison with fossils, on 
the whole strengthens the theory; the brain, the alimentary canal with its pyloric 
cecum, the kidneys and testis, the separate anus and urino-genital aperture, are 
all consistent with the view that Polypterus is related to the Actinopterygii. 
5. Lhe Thickness of the Skull in Mammatia. 
By Professor Ricuarp J, Anprerson, W.A., UD. 
Light shines through the orbital roofs in the young chimpanzee. The coronal 
and sagittal regions are opaque. 
The occipital fosse, and roof and sides of the skull are translucent in the ox, 
1 Quart. Jour. Mier, Sci., vol. xlv. 1901. 2 Trans. Zool, Soc., vol. xvi. 1903. 
