140 
ERIK A : SON STENSIO 
maintened that this is true of the development of the paired fins as well. The reductions 
which according- to a view of this sort have taken place in the skeleton of these fins 
are caused, according- to Abel (1919, p. 185) by the adaption of the Actinopterygians 
for rapid swimming at a very early phylogenetic' stage. 
The pectoral fins of the Coelacanthids have a certain interest in this connection, 
as their skeletons, as Woodward (1895 a, pp. 3 — 4 ) and Wellburn (1901, pp. 71—72) have 
emphasized, resemble that of the Actinopterygians to a certain extent, a fact that seems 
to support the view that a developing of the paired fins of the Crossopterygians in the 
direction towards the Actinopterygian type could probably take place fairly easily. 
In the continuation of this work I shall have a further opportunity to enter on a 
number of theoretical questions with regard to the skeleton of the paired fins and so 
I confine myself here to what has been said (cf. my description below of the skeleton 
of the ventral fins in Birgeria mougeoti and Saurichthys ornatus ). 
Coelacanthids and Tetrapods. 
According to the view now prevailing the Tetrapods are supposed as we know, 
to descend more or less directly from some Crossopterygian forms or at any rate 
forms that were closely allied to the Crossopterygians (Baur, 1896; Birks, 1916; 
Broom, 1913c, d; Dollo, 1895; Gregory, igi 3 , 1915; Modie, 1908; Pollard, 1892; Smith, 
1912; Thevenin, 1910; Watson, 1912a; igi 3 b; Woodward, 1898a, pp. 123 —125, etc.). 
Among the known Crossopterygii the Rhipidistids are justly considered to show the 
closest relationship to the Tetrapods. We have seen above that the Coelacanthids, on 
account of their specializations, have been thought to differ greatly from the Rhipidistids 
and thus from the ancestral forms of the Tetrapods as well. The Spitzbergen Coelacanthids 
described above have, however, helped in this case to give us to some extent another 
view; it is certain that in spite of their specialization the Coelacanthids have preserved 
a number of primitive characters and that because of this they show obvious resemblances 
to the primitive Tetrapods. 
It is not my intention within the limits of this work to embark on an exhaustive 
account of the relations between Coelacanthids and Tetrapods and I have accordingly 
confined the following exposition merely to certain details with regard to the structure 
of the head. I have taken the Stegocephalians as types of the Tetrapods and among 
them especially the two sub-orders Tetnnospondyli and Stereospondyli. 
The Coelacanthids and also the Rhipidistids have, with regard to the general 
shape of the primordial neurocranium, resembled the Tetrapod type more closely than 
other known fishes. Tis is due, in the first place, to absence of a real postorbital process, 
but in addition a number of other common features, although less prominent and at 
the same time less important, can also be traced with regard to the mutual relations 
and the shape of the different regions. Further there is often a certain habitual agreement 
between the Coelacanthids on the hand one and certain Temnospondyli and Stereospondyli on 
the other with regard to the distribution of the ossified and cartilaginous (possibly also 
membranous) parts of the primordial neurocranium. But in this case it is, at least to 
some extent, not a question of any primary resemblance, but a secondary one, due to 
