AND ON THE CLASSIFICATION OF FISH. 557 
structed according to the principles of the order or family. No 
value can be attached under any circumstances to the presence 
of the swimming-bladder, but its structure, when present, is 
subjected to immutable laws, which we learn as soon as we be- 
come acquainted with the true orders and families of fishes. 
According to this law, among all the Physostomi abdominales 
and apodes it is furnished with an air-passage when present, 
and deprived of the air-passage in all the Anacanthi (subbrachii 
and apodes), all Acanthopteri and all Pharyngognathi with 
prickly or soft fins. According to this law of the relative ana- 
tomical characters, the swimming-bladder in the Cyprinoidei and 
Characini is transversely divided, and when present in the Cy- 
prinoidei, Characini and Siluroidei, is without exception com- 
bined with the organ of hearing by a chain of small auditory 
bones. 
All this led me to an exactly opposite result from that which 
_ Vogt deduces from his observations, and with which he con- 
cludes his remarks ; and I thus prove, that the anatomical cha- 
racters in a certain sequence of the sections, orders and families 
are exclusive; that with them alone the classification of fishes 
can be undertaken ; and it may also be expected, that the com- 
parative embryology of fishes, far from yielding facts of a dis- 
 crepant nature, will confirm that only which comparative ana- 
tomy has taught us, as is even now evident as regards the 
embryology of the osseous fishes and Plagiostomi. 
Among the external characters there are some similar to those 
which we designated as relative anatomical characters, and which 
belong to the more important—for instance, the fulcra are not 
_ peculiar to all the Ganoids, but are absent, without exception, 
in the osseous fishes. When they occur, they indicate with cer- 
tainty Ganoids and their complete internal structure ; other- 
wise the external characters are mostly of subordinate import- 
ance. No value can ever be placed in the scales, thoracic shield, 
and such like ; they seldom enter into consideration in families, 
and mostly only in some few genera of families. As the enamel 
on the scales of the Ganoids is still frequently spoken of, I shall 
erely observe that Amia has no enamel on the scales; the 
_ kind of enamel, which consists of elevated lines of a substance 
different from the body of the scales, occurs in most ot the osseous 
fishes: and, again, osseous fishes occur with dachryoid enamel, 
