E. T. NEWTON ON FISHEs' TAILS. 91 



in the Devonian rocks, where it is represented by the genera Dlp- 

 terus and Conchodus. The tail of Dipterus is heterocercal, and 

 resembles stage 4 or 5 in the development of the flounder. Ctenodus, 

 from the Coal Measures, probably belongs to this group, but its tail 

 is not known. Teeth of Ceratodus have been found in the Trias and 

 Oolite, but from that time nothing more is known of this group 

 of fishes until the present day, and now we find it represented by 

 Lepidosiren, Protopterus, and Ceratodus, all these three having 

 diphycercal tails. Looked at, therefore, in the light of embryological 

 development, the Devonian Dipterus is more highly developed than 

 its recent representatives. 



Elasmobranchii. — This group of the sharks and rays is first 

 found fossil in the Upper Ludlow Rocks, where it is represented by 

 the spiny defences called Onchus, but no other part of the fish is 

 known. From the Ludlow onwards to the present day, teeth and 

 spines of sharks are met with in all marine deposits, but the forms 

 of their bodies are unknown, so that we cannot say what kind of 

 tail they had. 



Ganoidei. — By far the larger number of the fishes from the 

 Palaeozoic and Mesozoic formations belong to this group, and they 

 present so many different forms that it has been found necessary to 

 divide it into seven sub-orders, four of which have living representa- 

 tives, and the others are extinct. It will be well to take each group 

 separately. 



1. Amiadce, The Amia is the only form in this sub-order, and is 

 not found fossilized. 



2. Lepidosteidce. Most of the genera from the Devonian and 

 Carboniferous Rocks, which were at one time placed in this family, 

 are now referred by Dr. Traquair to the Palceoniscidte, a sub-division 

 of the Acipenseroidei {inde Pala3ontographical Society, 1877). 



It is in the Secondary Rocks that the Lepidosteidas have their 

 greatest development, and present us with the widest variety of 

 form, for although we find no diphycercal tail, and the extreme hete- 

 rocercal is the lowest type, yet, on the other hand, we get the 

 highest type known among the Ganoids, so far as we can judge by 

 outward appearance, for several genera have apparently homocercal 

 tails. 



Eugnathus, from the Lias, and Ophiopsis, from the Purbeck, show 

 the lowest type of tail in the Secondary Rocks ; they are heterocercal, 

 and agree with stage 4 or 5 of the flounder. 



