MODEKN SYSTEMS.

299

the Ganoids should be also separated from the ordinaryFishes . This, however, cannot be decided until theirembryological development has been thoroughly investi-gated, though I have already collected data which favourthis view of the case. Should this expectation be rea-lized, the branch of Yertebrata would contain the follow-ing classes :—

1st Class : Myzontes; with two orders, Myxinoidsand Cyclostomes .

2nd Class: Fishes proper; with two orders, Ctenoidsand Cycloids. 1

3rd Class : Ganoids; with three orders, Ccelacanths,Acipenseroids, and Sauroids; and doubtful, the Siluroids,Plectognaths, and Lophobranches. 2

4th Class: Selachians; with three orders, Chimoerm,Galeodes, and Batides.

5th Class : Amphibians ; with three orders, Csecilioe,Ichthyodi, and Anura.

6th Class: Reptiles; with four orders, Serpentes, Saurii,Bhizodontes, and Testudinata.

7th Class: Birds; with four orders, Natatores, Grallse,Basores, and Insessores (including Scansores and Acci-pitres.)

8th Class: Mammalia ; with three orders, Marsupialia ,Herbivora , and Carnivora. 3

1 I am satisfied that this subdivi-sion of the Fishes proper requiresmodifications; but I fear it wouldlead me too far, were I to discusshere the reasons for the changes Ipropose to introduce into it.

2 I have observed a very curious

and peculiar mode of locomotion inall the Lophobranches, Sclerodermsand Gymnodontes, which I have seenalive. They do not progress by thelateral motions of the vertebral co-

lumns, as other fishes do; but chieflyby an undulatory movement of theirvertical fins, resembling very muchthe mode of action of the vibratilemembranes. In this they resemblethe young Lepidosteus; and I consi-der this fact as a new argument infavour of their association with thetrue Ganoids. Comp, also the re-marks respecting the limits of theGanoids, p. 242.

3 Since this chapter was written