554. MULLER ON THE STRUCTURE OF THE GANOIDS, 
naked and scaly amphibium. As certainly as all naked amphibia 
agree in having an aortic heart, and all scaly amphibia in not 
possessing this heart, so do the Ganoids and the Osseous Fishes 
differ decidedly in this absolute character. The position of 
Sudis and Osteoglossum is determined with certainty by their 
structure, which I have detailed, and equally well is the position 
of Amia as a Ganoid decided by the observation of Vogt. 
Hitherto Hsox, Belone and Lepisosteus have been considered 
as so similar and related, that on account of their form they 
were placed in the same genus. After Lepisosteus was sepa- 
rated, the genera Esow and Belone at least appeared to be in- 
separable; this affinity has been overthrown by anatomy, so 
that there can no longer be a question of it*. And in what does 
the intimate affinity of Amia with Sudis and Osteoglossum now 
consist? and with Erythrinus? which, according to Vogt, may 
also perhaps be Ganoid, although it is decidedly proved that it 
belongs to the family of Characini.} Amia, Sudis and Osteoglossum 
are fishes with soft fins, abdominal ventral fins, and a hard scale- 
less head, large cheek-bones, long dorsal and anal fins, and whose 
upper jaw lies external to the intermaxilla. They agree in these 
points, which in the present question is not of the slightest im- 
portance ; innumerable fishes of the most different divisions have 
a hard scaleless head and large cheek-bones, and it is as little 
extraordinary in Sudis as in Erythrinus, Xiphoramphus and Xi- 
phostoma, and many other Characini. The scales of Sudis and 
Amia are entirely dissimilar. Those of Sudis (Arapaima), Hete- 
rotis and Osteoglossum resemble mosaic, granulated on the sur- 
face; the scales of Osteoglossum, as in other osseous fishes, are 
concentrically striped; the scales of Amia are not compound, 
and on the surface have engraved lines running parallel to their 
length. 
I still less comprehend why Agassiz, in the third part of his 
Poissons Fossiles du vieux grés rouge, places Sudis in the family 
of the Celacanthi among the fossil Ganoids. He figures in that 
work the skeleton of a Sudis to illustrate the Celacanthi. The 
latter, according to Agassiz, are fishes remarkable in their bones 
and fin-rays being hollow. In Celacanthus the ossa interspi- 
nosa adhere to the processus spinosi, and the fin-rays do not 
ramify. All this does not apply to Sudis. If Sudis were related 
* See the Abhandlung ueber die Naturlichen Familien der Fische. Archiv. 
f. Naturges,, 1848, i. 
