254 
ERIK A : SON STENSlO 
fishes of Lombardy placed his new genus Perleidus, based on the so-called Semionotus 
altolepis, in the family Catopteridae. 
The majority of the modern authors, such as Schellwien (1901, loc. cit.-.), Eastman 
(1904, pp. 95—too; 1911, pp. 46—57; 1914a, pp. i 3 g—146; 1917, pp. 279—280), Gregory 
(1907, p. 448), Goodrich (1909, p. 3 i 3 ), de Allessandri (1910, loc. cit.), Koken (1911, p. 102), 
Schlosser (1918, p. 100), etc. have, like Woodward, looked upon the Catopterids as an 
independent family. Recently, however, Abel in his large work «Die Stamme der Wirbel- 
tiere< (1919,. pp. 187, 189) has grouped them with the Palaeoniscids. 
In the definition that the family Catopteridae has had up to new, it has only 
comprised, as has been shown, the genera Catopterus, Dictyopyge and Perleidus. 
In 1916 I was able myself to observe a considerable agreement between the genus 
Colobodus Agassiz on the one hand and Palaeoniscids and Platysomids on the other 
( Andersson 1916 a, pp. 17—25). I accordingly assigned to Colobodus a close relationship 
with the Catopterids, although under the conditions then present I did not consider that 
I could refer it to these, but established a new family Colobodontidae for it. With this 
family I then incorporated provisionally the new genus Meridensia. 
In a work on the Ganoids of the German Muschelkalk, published in 1920, Stolley 
(pp. 25—86) has grouped the genera Colobodus s. str., Nephrotus H. v. MeVer, Crenilepis Dames 
and Dollopterus Abel into one family, which he also calls Colobodontidae. 1 ) This family, accord¬ 
ing to his opinion, would form (p. 45) «einen zusammengehorigen Formenkreis, der ganz 
vorwiegend den germanischen Muschelkalk charakterisiert und in keine der bisher unter- 
schiedenen Familien hineinpaflt. Durch die Ausbildung der Kopfknochen einerseits an 
Palaeonisciden und Platysomiden erinnernd, zeigt er andererseits auch vorgeschrittenen, 
den Orthoganoiden ahnlichen Charakter und verdient zu einer besonderen Familie zu- 
sammengefafit zu werden, die fiiglich den Namen der Colobodontidae zu tragen hat. Diese 
Familie kann im geltenden System nicht mehr den Heterocerci als jiingste und vor- 
geschrittenste Familie eingefiigt werden, da sie eben keine heterozerke Schwanzflosse 
mehr besitzt, sondern mu8 den Orthoganoiden als erste und primitivste, die Verknupfung 
mit den Heterocerci herstellende Familie eingereiht werden, gefolgt von den Semionotiden 
und den alteren wie jiingeren Lepidotiden im hier prazisierten Sinne dieser letzteren 
Familie (pp. 76—79).» 
He feels very doubtful as to whether the genus Perleidus really belongs to the 
family Catopteridae or Colobodontidae and in this connection writes as follows on the rela¬ 
tions between these two families (p. 65): «Wenn Catopterus und Dictyopyge-in der Aus¬ 
bildung der mafigebenden Kopfknochen entsprechend beschaffen sein sollten, wie Perlei¬ 
dus altolepis und ? P. letticus, so diirfte fur alle diese Formen der Familienname Catopte- 
riden berechtigt sein und die Stellung dieser Familie wiirde dann entfernt von den 
Semionotiden und Lepidotiden in der Nahe der Colobodontiden zu wahlen sein, zwischen 
diesen und den altertiimlicheren Palaeonisciden der Heterocerci.® 
I have dealt here in detail with Stolley’s views because I shall have to criticize it 
below on the basis of new facts. As we shall see, it seems as if we are no longer 
justified in maintaining the family Colobodontidae, but Colobodus 1 ) s. str. and Dollopterus and 
r ) At least for the present Crenilepis ought, according to my opinion, to be regarded as a synonym of 
Colobodus (cf. Stoi-ley 1920, p. 44). 
