266 
ERIK A : SON STENSIO 
which have only one serration farthest ventrally and on the tail even this one seems 
to be absent. The ventral scales also generally have only one or two serrations. 
The exposed surface of the scales is covered by a layer of shining ganoine, which 
is smooth on the whole. Small furrows are visible on this really only close to the 
posterior margin, where they are arranged so that each of them reaches this margin 
in the small sinus between two neighbouring serrations. This character is most striking 
in the abdominal region (text fig. 84 A), and is gradually lost backwards. The scales 
on the neck immediately behind the extrascalpulars sometimes show a tubercular sculpture 
of the same type as that on the membrane bones of the cranial roof. As can easily be 
seen, P. woodivardi shows a great resemblance to P. altolepis (Deecke) with regard to the 
development of the scales. 
The lateral canals of the body. 
The main lateral canal of the body can be followed in P. woodwardi for a consi¬ 
derable distance. It pierces the supracleithral as usual and then runs backwards and 
somewhat downwards, first penetrating the uppermost high scales of the anterior vertical 
rows (/, PI. 34, figs. 1, 2). It opens with narrow vertical pores, which are found on every 
third or fourth, sometimes on every fifth scale. 
Another shorter lateral line runs on either side higher up from behind the extra¬ 
scapulars backwards towards the dorsal fin. It has thus the same position and extension 
as in P. altolepis (cf. p. 256 above) and Lepidotus (Woodward 1895 b, p. 95, fig. 22). 
* 
Remarks. — Perleidus woodwardi was described for the first time (1912) by Woodward 
under the name of Colobodus altilepis and I myself later on (Andersson 1916 a) placed it 
under this name. Recently Stolley (1920, pp. 38 — 43 ; 84—85) has emphasized that it 
cannot belong to the genus Colobodus, but that, like the so-called Colobodus ornatus 
(Agassiz), and Colobodus latus (Agassiz), it is most closely related to Lepidotus instead. 
He is doubtful, however, whether to group it with these two Alpine species in the genus 
Paralepidotus established by him, or whether to place it in closer relation to Heterolepidotus 
dorsalis. According to his opinion, Perleidus woodwardi would thus even belong to quite 
another family than Colobodus. 
The present material has, however, completely proved that the so-called Colobodus 
altilepis A. S. Woodward shows the closest resemblance to the Palaeoniscids with regard 
to the skeleton of the head and the development of the sensory canals, and that it 
must undoubtedly belong to the same large group as Dollopterus, Perleidus and Collobodus 
s. str. The evidence, seems also to show that among these genera it is most closely 
related to Perleidus, and, as pointed out above, I considered that I ought, at least for 
the present, to include it in this genus. 
As under these conditions there would be in the genus Perleidus two species with 
practically identical names, I considered it necessary to suggest the'new name woodwardi 
for the Spitzbergen species, after the eminent palaeontologist Dr. A. S. Woodward, espe¬ 
cially as he was the first to describe the species. 
