520 MULLER ON THE STRUCTURE OF THE GANOIDS, 
the Ganoids is frequently osseous, still the unossified condition 
of the central portion, where the apophyses are barely ossified, is 
an important sign where any part of the above most important 
characters are absent. ‘The mere rhomboidal form of the scales 
without any peculiar enamel, without articulations, with fulcra 
on the margins of the fin, without heterocercy, with an ossified 
vertebral column, and still more so where the ventral fins are 
wanting, or not placed on the abdomen, would be doubtful, as 
is the case with Balistes. ‘But even when other characters are 
wanting, if the scales be articulated, as in Gyrodus, there appears 
no room for doubt. With respect to several Ganoids, Agassiz 
does not bring forward complete proofs why they are Ganoids. 
A long acquaintance with his work however creates great con- 
fidence in his experience on this point. We are satisfied as re- 
gards the Celacanthi, when we see them with their round, im- 
bricate scales figure among the Ganoids, as soon as we find that 
the apophyses only of their spine, not its central portion, is os- 
sified, as is met with so distinctly in Undina. The age of the 
formation may also be taken into account, to a certain extent, in 
arranging a fish among the Ganoids. But this indeed is already 
a petitio principii. 
The osseous substance of the scales of Lepisosteus and Poly- 
pterus, when microscopically examined, exhibits the radiating os- 
seous corpuscles as they exist in the bony plates of other fishes, 
but which do not usually occur in the ordinary scales of the 
osseous fishes. In very large scales, however, even of osseous 
fishes, a lowermost layer containing osseous corpuscles is some- 
times met with. I also find them in the scales of Sudis, which 
do not otherwise differ from those of other osseous fishes. In the 
genera Megalurus and Leptolepis, from the upper member of the 
Jura formation,—the lithographic slate,—we are reduced in their 
determination to the consideration, that their round, imbricate 
scales, without osseous corpuscles, which appear similar to the 
scales of the osseous fishes, are covered with enamel, and that 
they belong to the Jura formation. On a microscopic exami- 
nation of these scales, I find the concentric lines as in the scales 
of the osseous fishes, but they are really covered with a thin vi- 
treous layer of enamel, so that they usually leave no impression 
of the lines on the stone. I am uncertain what position should 
be assigned to this fish. 
As there exist among the living Ganoids some which are 
