i94 
ERIK A : SON STENSIO 
Together with the pelvic plate and the endoskeletal radials, now treated, a larger 
or several smaller cartilaginous elements must also have been present in the endoskeleton 
of the ventral fins. For the io or 12 radials that did not articulate against the pelvic 
plate occupy such a position that their proximal ends must have articulated with one 
or several cartilaginous elements situated behind this plate. In other words, there must 
have been one or, according to my view, most probably several cartilaginous meta- 
pterygeal elements. In my reconstruction of the endoskeleton of the ventral fins in 
text fig. 68 B I have drawn a number of such elements (Ml) but of course in this 
respect I cannot claim to have hit on the correct number , or shape. In such a case as 
this a reconstruction must always be more or less arbitrary. 
Guided chiefly by the conditions in Coccolepis, Regan (1904, p. 334) and to some 
extent Woodward (1895 a, p. 7) have expressly stated that the Palaeoniscids must have 
been developed to some extent in the same way as the recent sturgeons with regard to 
the endoskeleton of the ventral fins. The conditions in Birgeria mougeoti now show that 
this view was partly correct. For it is easy to see that in the pelvic plate in this species 
we have an element that in certain respects resembles the plate in the sturgeons that 
v. Rautenfeld (1882) calls the «Basalplatte» and Wiedersheim (1892, p. 63, text fig. 11) 
«vorderes (proximales) Basale», and, further, that the metapterygoid elements in the 
same species had a position in relation to the surrounding elements corresponding to 
that of the so-called «basalsegmente» (v. Rautenfeld, 1882) or «distalen basalia» (Wieders¬ 
heim, 1892) in the sturgeons. Common to both B. mougeoti and the sturgeons is the sub¬ 
division of the inner radials into two segments, of which the proximal one may show 
an ossification in the sturgeons as well. 
On the other hand the present material also seems to make clear that the endo¬ 
skeleton of the ventral fins shows, at least in certain Palaeoniscids, several deviations 
from that of the sturgeons. Not only has B. mougeoti, for instance, ossifications in a 
greater number of elements than the sturgeons, but the ossifications in these elements 
are more strongly developed and mainly of an endochrondal nature, and further, the pelvic 
plate is without any indications of subdivisions in the lateral parts, which, on the other 
hand, is often the case in the sturgeons in their so-called «Basalplatte». 
As Goodrich pointed out (1909, p. 3 o 3 ), the endoskeleton of the ventral fins of the 
recent sturgeon fishes resembles to some extent that of the Cladodontids. And the same 
may be said with equal justification about that of the Palaeoniscids, according to what 
we have seen in B. mougeoti. We can even say that the skeleton of the ventral fins 
possesses a certain, at least a habitual, agreement with that of the Elasmobranchii in general. 
If from these g-eneral comparisons we now pass on to attempt to discover the 
homologues of the different elements in the endoskeleton of the ventral fins in B. mougeoti, 
we shall find considerable difficulties. In order to have a general view of the questions 
that emerge in connection with this attempt at homologization it is, however, first 
necessary to give a brief historical account of the different views that have been put 
forward with regard to the inner skeleton of the ventral fins in the Actinopterygii in general. 
Gegenbaur was really the first to start to make a comparative anatomical investi¬ 
gation of the skeletons of the extremities in the lower vertebrates — and especially in the 
sharks. In this he paid attention preferably to the phylogenetic points of view, and his 
