388 RAMSAY H. TRAQUAIR ON THE 
bital or “ lachrymal” becomes developed to fill up the space between it and the 
anterior frontal. 
This type of skull, which reaches its most extreme development in Platy- 
somus, is strongly marked even in Hurynotus, but not so much so in Benedenius, 
although the Platysomid nature of the last mentioned genus is sufficiently indi- 
cated by the backward position of the orbit and. the vertical direction of the 
hyomandibular bone. 
3. While the Paleoniscidee, with very few exceptions (Gonatodus, Micro- 
conodus), possess acutely conical teeth of different sizes, that form of tooth is 
in the Platysomide either absent or very rarely seen.. But the remarkable 
differences in the external shape of the teeth displayed by the various genera 
of Platysomide themselves amply show that the shape of the teeth is here of 
very little systematic value. 
4, In the majority of Platysomide the scales of the body have the keel of 
the internal surface, which passes above into the articular spine, coincident — 
with, or close to the anterior margin. ‘This is, however, not the case in 
Eurynotus. 
5. The dorsal fin has an elongated base, and the anal fin tends to assume a — 
similar form, though it is short-based in Benedenius and Hurynotus. 
6. The ventral fins are in some Platysomide very small (Platysomus), or 
possibly absent, as in Cheirodus.. 
Weighing these points of resemblance and difference together, it is quite — 
obvious that the latter are of a much more superficial nature than the former ; 
in other words, the Platysomid type is simply a modification of the Palzonis- — 
coid one. The Platysomide are specialised Paleoniscide. 
Stray glimpses of the progress of this specialisation are also, in fact, exhibited | 
to us in contemplating the series of Platysomid genera. Benedenius, though 
in my opinion standing upon the Platysomid side of the boundary, is consider- 
ably more Palzoniscoid in aspect than the rest of the family. The head of 
Eurynotus is decidedly, indeed strongly, Platysomid in structure, but from the 
body it would, in spite of its long dorsal fin, be hard to refuse it a place in the 
Paleeoniscidee. The anal fin in Benedenius and EHurynotus is short-based, like 
that of Hlonichthys; that of Mesolepis has already considerably extended in 
length, while in Cheirodus and Platysomus the anal has become nearly as long — 
as the dorsal. 
The genera of Platysomidee do not, however, form a straight line. Cheir- 
odus branches off in a rather different direction from Platysomus, and itis — 
abundantly evident that, although the relationship of the Platysomid genera to 
each other and to the Paleoniscide favours the doctrine of Evolution, many 
genera haye yet to be discovered before the line of descent can be satisfactorily 
exhibited. i 4 
