AND ON THE CLASSIFICATION OF FISH. 555 
to the Celacanthi, I should consider it as proved that the true 
osseous fishes of the present world, contrary to all the earlier 
deductions of Agassiz, extend into the oldest formations of the 
former world. In his new monograph, Agassiz has differently 
characterized the Celacanthi, with the addition of some fishes 
from other families, as Ganoids with round imbricate scales and 
plaited teeth. These round scales are only distinguished from 
the scales of the osseous fishes by their enamel. But Sudis has 
neither the enamel of the scales nor the teeth of the Celacanthi. 
Accurately speaking, our whole knowledge of these Ceelacanthi, 
which are generally considered as peculiar to the Old World, is 
uncertain and scanty. The ganoid nature of the true Celacanthi, 
in my opinion, depends upon the absence in them of the verte- 
bral body. [Undina in Graf. Minster’s Beitr. v. taf. 2, also 
examined by me.] On the other hand, we have a decisive 
knowledge of the nature of Amia and Sudis, and their organiza- 
tion, which is quite different in the two, can hardly serve to ex- 
emplify the doubtful family of Celacanthi. As I long ago ex- 
amined Sudis anatomically in the specimen in spirit, and the 
skeleton sent by Rich. Schomburgk, in every respect, I can 
positively assert that it does not differ in a single point from the 
general type and plan of all our common osseous fishes of the 
present world. It approximates moreover by Osteoglossum to 
Megalops and Notopterus, and by these to Chatoéssus and Clu- 
pea. In my former treatise I have shown that they do not in 
any instance differ in the pseudo-branchie, but form a continu- 
ous series. 
As Amia agrees with the other Ganoids in those relations of 
its structure which have hitherto been examined, in the muscu- 
lar covering of the arterial trunk, the numerous rows of valves, 
and in the spiral valve in the intestine, it may with greater cer- 
tainty be predicted that it also has a chiasma of the optic nerves, 
a thymus gland, and an uncleft retina; and as it does not pos- 
sess any opercular accessory gills, it may be supposed that, like 
Polypterus and Spatularia, it has the branch of the gill-artery 
to the operculum as the equivalent of the opercular gill. Com- 
‘parative anatomy in its perfect form leads to such necessary 
consequences, that expressions may be obtained for organizations 
which resemble the expression of an equation. Only let these 
expressions be found for any given case, then the unknown 
quantities may be calculated from those that are known, just as 
