DR. A. GUNTHER’S DESCRIPTION OE CERATODUS. 
559 
been able to find any trace of it preserved in the fossil remains, as it is entirely destroyed 
like the remainder of the cartilaginous portions. In the fringed fins of the second type, 
termed “ obtuse ” or “ subacutely lobate,” the scaly covering is limited to the base ; the 
fin-rays are very distinct, comparatively few in number, and joined to a simple transverse 
series of cartilaginous rods. Such are the fins of Polypterus , of the Ccelacanths as 
restricted by Huxley, and probably also of the Saurodipterini. Having drawn attention 
to this, as I believe, important diversity in the only character on which the suborder 
Crossopterygii is founded, I leave the bearing of this fact upon the further classification 
of these fossils to be developed by those who are more intimately acquainted with the 
details of their structure than myself. 
I should be found wanting in respect towards the founder of Paloeichthyology, were I to 
pass over in silence the opinion expressed by him with regard to the affinities of Ceratodus. 
In a letter addressed to Sir Philip Egerton (‘ Nature,’ iii. No. 61, p. 166), Professor 
Agassiz states that this fish is clearly a member of the family named by him Coelacan- 
thini. I am not aware how far Agassiz has modified the characters and limits of this 
family since the publication of the ‘ Recherches ’ (vol. ii. p. 168) and 4 Yieux Gres 
Rouge’ (p. 64); at that time he pointed out as the principal characters the hollow con- 
dition of the fin-rays and bones (as in birds), the presence of interspinous bones in the 
caudal fin, the continuation of the vertebral column between the two lobes of that fin, 
and the prolongation of the caudal extremity beyond it as a filamentary appendage. 
Arapaima ( Sudis ) gif/as was stated to be one of the living representatives of this family. 
Now it is scarcely correct to describe the fin-rays and bones of these fishes as hollow like 
“ those of birds before they underwent the alterations during the process of decom- 
position and fossilization, they were thoroughly solid, without hollow space in the inte- 
rior, like the fin-rays and bones of Ceratodus and other notochordal fishes, with or with- 
out a bony covering. With regard to the second and third characters, they are common 
to so many other fishes not reckoned among the Ccelacanths, that no safe conclusion 
can be drawn from them regarding the affinities of a fish. Finally, the prolongation of 
the caudal extremity as a filamentary appendage is not observed in Ceratodus. No 
student of recent ichthyology has followed Agassiz in placing Arapaima among the 
Coelacanths, or indeed among the Ganoids ; Muller has shown it to be a true Teleos- 
tean ; and the degree of affinity between this genus and Ceratodus is not greater than 
that between a Salamander and a Lizard*. 
Tiiiolliere and Huxley have independently come to the conclusion that Agassiz’s 
family of Coelacanths comprises too many heterogeneous forms to allow us to regard 
the affinities of a fish as determined by its being referred to it. Both have limited 
the term to the genus Ccelacanthus as type, and a few other forms closely allied to it. 
The family thus restricted appears to me even more remote from the Dipnoi than the 
* Botanical science is in this respect more advanced than Zoological : no botanist would allow himself to be 
influenced by merely external similarity ; and a system in which the African Euphorbia were associated with 
the American Cactus would be repudiated by all. 
MDCCCLXXI. 4 H 
