174 M. Agassiz on the Relations between Animals 



marine^ Petrostoma proper being both marine and fluviatile : the 

 higher type of Ammocoetes (for we must consider Ammoccetes as 

 higher, inasmuch as the division of the lips indicates a tendency 

 towards a formation of a distinct upper and lower jaw) is exclu- 

 sively fluviatile. The Goniodonts, which from their affinities to 

 Sturgeons rank higher than the Siluridse, are exclusively fluvia- 

 tilcj whilst there are some marine types among the latter. Among 

 Percoids we find in fresh water a larger number of those in which 

 the two dorsals are distinct, a character making them eminently 

 superior to the forms with undivided fins. For the same reason 

 we should consider the Sparoids inferior to the Percoids, their 

 dorsals being not only generally undivided, but even covered with 

 scales. Among the Eels, those destitute of all fins are exclusively 

 marine, those without pectorals also exclusively marine, and we 

 may fairly consider the freshwater Eels as the higher type of the 

 family on this ground. If there is any natural connection, as I 

 have attempted elsewhere to show that there is, between Scom- 

 broids and Scomberesoces, and Esoces proper, it becomes plain at 

 once that the latter are the higher from the abdominal position 

 of their ventrals, and they are a fluviatile family. Even taking 

 the Cycloids as a whole, we find among them the lower families 

 of Thoracici and Jugulares, as the families of Cod and Scom- 

 brides, chiefly marine, whilst the families of Salmonidse and Cy- 

 prinidse are chiefly fluviatile. Among the Gadoids we have those 

 with many vertical fins, as the true Cod, marine, while those in 

 which the dorsals and anals are reduced, such as the genus Lota, 

 are fluviatile. Even among the Salmonidse in the widest exten- 

 sion which this family had formerly, we find the Scopelidse with 

 the inferior structure of their jaws chiefly marine, while the Co- 

 racini and true Salmonidse are chiefly fluviatile. Everywhere, in 

 fact, in each minor group, the fluviatile representatives show cha- 

 racters indicating their superiority over their marine represen- 

 tatives. Whatever exceptions might be found to this law, which 

 in the outset appears so general, I have no doubt will lead at 

 some future time to the discovery of some other principle as yet 

 unknown. 



The class of Reptiles is one of the most interesting in the point 

 of view under consideration, and each of their types exemplifies 

 in itself the law of the intimate connection between animal types 

 and the media in which they live in the most striking mannei*, 

 inasmuch as here the gradation, which might be inferred from 

 structural and embryological evidence, agrees most fully with the 

 gradation of the elements in which they live. Among Batra- 

 chians we have chiefly fluviatile and terrestrial families. The 

 Ichthyodes, or Batrachians with permanent branchiae, are all 

 aquatic, and acknowledged the lowest in the class. Some of then* 



