:3o 
ERIK A : SON STENSIO 
of the two last-mentioned fins are also characterized by being always more or less 
strongly bifurcated in their proximal part, while that of the anterior dorsal fin is generally 
rounded and triangular. In exceptional cases, however (Scleracanthus), it is bifurcated 
in about the same way as the other two but not so much. It has generally been considered 
that the plates in question have been formed by fusion of the proximal segments of the 
endoskeletal radials, and it is probably pretty certain that this is true of those of the 
posterior dorsal fin and the anal fin. With regard to that of the anterior dorsal fin, on 
the other hand, it seems most probable that this comprises not only the proximal but 
also the middle segments of the endoskeletal radials. 
Both the posterior dorsal fin and the anal fin are weakly lobed. The anterior 
dorsal fin, on the contrary, is not lobed and in consequence has a straight basal line. 
In all three fins the lepidotrichia usually grow narrower, in most cases continuously, 
towards the distal ends. In certain Mesozoic forms, however, they expand first in distal 
direction, decreasing then towards the ends. A longer or shorter proximal part is most 
often unjointed. All the lepidotrichia of the anterior dorsal fin are furnished in Mesozoic 
forms with tuberculation or spines on the lateral surfaces. 
The caudal fin generally consists of three lobes, a dorsal one, a middle one and 
a ventral one, and has nearly a diphycercal shape. It is however practically never 
quite symmetrical, but the ventral lobe has generally a few more lepidotrichia than the 
dorsal one. The middle lobe, the so-called supplementary caudal fin (called «Pinselflosse» 
in German works), is developed nearly as a small independent fin, reaching more or 
less far backwards caudally of the two other lobes. It may, however, also be quite 
reduced as is the case in Macropoma and Heptanema. 
The lepidotrichs of the caudal fin show the same conditions as to shape, joints, 
etc., as the ones of the other unpaired fins. Like those of the anterior dorsal fin they 
are furnished in the Mesozoic forms with spines or tubercles on the lateral surfaces. 
The only exceptions in this case are, as I have hitherto found, the so-called Coelacanthus 
africanus Broom and C. gracilis Agassiz, where instead of tuberculations they have some 
fine longitudinal striae of a shining ganoine-like substance. z ) 
For our views on the inner skeleton of the caudal fin Schmalhaussen’s investigations, 
published a few years ago (igx 3 a) are of great importance and I may consequently be 
allowed in this connection to give some of his most important results. 
After Schmalhaussen by means of embryological investigations (1912) came to the 
conclusion that the supporting endoskeleton of the unpaired fins had arisen phylogeneti- 
cally without connection with the axial skeleton he set to work to study the Dipnoans, 
Coelacanthids and Cyclostomes, all of which had previously been considered as presenting 
evidence in favour of the view of the endoskeleton of the unpaired fins having originally 
been developed from the axial skeleton. 
The endoskeleton of the caudal fin in Dipnoi has its «Anlage», as Schmalhaussen 
has verified, in connection with the axial skeleton, and because of this he looks upon 
it as only being distal parts of the neural and hemal spines. The anatomy etc. 
of the muscle shows moreover that the caudal fin is progressively developed forward 
x ) This is clearly shown by the specimens in British Museum. 
