TRIASSIC FISHES FROM SPITZBERGEN 
137 
Before leaving- the Dipnoans there is, however, one more matter that is worthy of 
attention. As now it seems to be practically proved that not onty Coelacanthids but 
also certain of the Rhipidistids had an articulation between the palatoquadrate and the 
posterior part of the orbitotemporal region, this fact removes, in my opinion, one of the 
most important differences between the Dipnoi on the one hand and the Rhipidistids and 
Coelacanthids on the other. We have in these Crossopterygian forms simply the original 
type from which the autostyly of the Dipnoi must have arisen, and not only is this true, but, 
if Watson’s statement 1912 (1912 a, p. 9) with regard to Megalichthys proves to be correct 
there are also forms among the Crossopterygians which bear a considerable resemblance 
to the Dipnoi inasmuch as they have a rather firm connection between the palatoquadrate 
and the neurocranium. 
The study of the Coelacanthids thus seems also to be of help in throwing fresh 
light on the relationship of the Dipnoi to the Crossopterygians. According to our present 
knowledge (cf. Dollo, 1895, pp. 105— no; Watson and Day, 1916, pp. 41—45) the latter 
or forms very nearly allied to them must be considered as the ancestors of Dipnoans, 
and must consequently be grouped closely together with them from a systematic point 
of view (cf. Dean, 1900, p. 29). There is thus no justification at all for separating the 
Dipnoi from the other Teleostomes, as has hitherto often been done (cf. Dollo, 1895, 
p. 1 13 ). 
Coelacanthids and Actinopterygians. 
In the primordial neurocranium of primitive Actinopterygians we find, as will be 
seen from the description of Birgeria mougeoti (Agassiz) given below in this work, at 
least the following bones: a basioccipital, a paired prootico-opisthotic, comprising also 
the homologue of the epiotic, a paired autosphenotic and an unpaired sphenoid. 
Of these the two prootico-ophisthotics are certainly on the whole considerably 
larger and stronger than is the case in the Coelacanthids, but as in these, they only extend 
forwards to the posterior boundary of the exit for nervus facialis. 1 ) The sphenoid is also 
large and corresponds in its extension rather closely to the similarly termed bone in Dictyo- 
nosteus (Stensio, 1918 c) having, as in this form, its pars basisphenoidea extending into 
the labyrinth region behind the fossa hyphophyseos and thus belonging with this part 
to the chordal region of the cranium. Basipterygoid processes are present, but are situated 
low, and there is no connection between them and the palatoquadrates. As far as one 
can see from the facts mentioned, the primitive Actinopterygii thus show, in spite of 
great differences, a decided approach to the Crossopterygian type. 
As I have tried to show*, the Coelacanthids, like the higher Actinopterygians, seem 
to have had a myodome developed, and, if my opinion is correct, this, although it has 
arisen quite independently of that in the Actinopterygians, has still been developed in 
close relation to the so-called canalis transversus, which is pierced by v. pituitaria 
(Gegenbauer, 1872, pp. 75—79; Sagemehl, 1884, pp. 215—217; 1885, pp. 85—87; 1891, 
pp. 574—575; Allis 1909 a, pp. i 83 —208; 1914 a, pp. 225—253; 1918 a, pp. 241—246). 
T ) In higher Actinopterygians the prootic generally extends, as is known, forward to the exit of the trigeminus 
and often surrounds the foramen prooticum. 
Stensio, Triassic Fishes from Spitzbergen. 18 
