DR. A. GUNTHER’S DESCRIPTION OR CERATODUS. 
553 
shall be obliged either to abandon the subclass Dipnoi altogether, or, if we maintain it, 
to materially alter its definition, so as to comprise Ceratodus as well as Lepidosiren. 
It appears to me that of these two courses the former ought to be adopted, and that the 
Dipnoi should be united with the Ganoids, among which they may form a well-marked 
subdivision, like the Ilolocephala among the Chondropterygians. 
In order to justify this view, I will compare the chief characters of the Dipnoi with 
those of Ganoid and Chondropterygian fishes. The presence of a pulsating conus arte- 
riosus is common to all three divisions ; a notochordal skeleton is found in the Acipen- 
sers* and Chimseras ; a cartilaginous skeleton of the paired fins is fully developed in 
the Chondropterygians, being only foreshadowed by the single row of cartilaginous 
rods in the fins of Polypterus ; the cellular air-bladder of many Ganoids, which in 
Polypterus communicates with the ventral side of the oesophagus, is by all anatomists 
admitted to be a very close advance towards the Dipnoous lung ; the gills of Ceratodus , 
which extend and are fixed to the walls of the branchial cavity, represent an arrange- 
ment indicating the first step towards the fixed gills of the Sharks and Rays ; accessory 
external gills have hitherto been found in certain examples of Protopterus and Poly- 
pterus f ; the presence of a spiral valve in the intestine, again, is common to Ganoids 
and Chondropterygians ; finally, the convoluted oviducts with the addition of peritoneal 
openings, and the numerous ova of small size, remind us unmistakably of similar 
conditions among Ganoids, for instance Acipenser and Lepidosteus%. 
The only remaining absolute characters by which we can distinguish the Dipnoi 
from the Ganoids and Chondropterygians are the position of the nostrils within the 
mouth, the dentition consisting of two pairs of molars and one of vomerine teeth, and 
the lobate fins or paddles, supported by an axial cartilaginous skeleton. 
These three points cannot be considered to constitute the characters of a subclass 
equivalent to Teleostei , Ganoidei, &c. ; and therefore the Dipnoi are better united with 
the Ganoids. 
But it appears to me that also the Ganoids and Chondropterygians are much more 
closely allied to each other than either of them to the Teleosteans. It would be beyond 
the limits of this paper to enter into all the various points of organization which have to 
be considered with regard to this question. But I may urge, as the most important fact 
in favour of this view, the treble partition of the heart of all these fishes — the bulbus aortee 
of the Teleosteans being, as Muller has shown, simply the thickened origin of the aorta, 
separated from the pulsating heart by the (almost always) double valve. A heart with 
a true conus arteriosus is always accompanied by a spiral valve of the intestine (which 
only exceptionally, as in Lepidosteus, remains in a rudimentary condition), and by non- 
~ The similarity between vertical sections of the cartilaginous brain-capsule of Acipenser and Ceratodus is 
surprising. 
T In the latter genus only recently discovered by Steendachner ; they may be found also in young Ceratodus 
and other Ganoids. 
^ The male organs of Lepidosiren are not known, and those of Ganoids only imperfectly ; Plagiostomes have 
a deferent duct, but it is in direct communication with the sperm-gland. 
4 g 2 
