4 RAY LANKESTER, FOSSIL FISHES OF SriTZBERGEN. 



It appears then that so far as its fish fauna is concerned the so-called Devonian 

 system readily falls inte two divisions, a lower which is naturally associated with the 

 Silurian system — and an upper which equally naturally attachés itself to the Carboni- 

 ferous. The lower division includes what is in England called »Lower Devonian» viz. 

 the lower beds of South Devon, the Herefordshire Cornstones and the lower old Red 

 Sandstone of Scotland — in all these rocks Cephalaspida; are found but no Holopty- 

 chians or other Glyptodipterini. With them oceur Acanthodida3 and Ctenodipterini and 

 fi^agments indicative of distinct but as yet unknoAvn families. The upper division 

 includes the middle and upper Devonian limestones of Devonshire, the middle old Red 

 sandstone of Scotland (with Holoptychius nobilissimus), the Coccosteus- & Ptericthys- 

 bearing flagstones of Caithness and Orkney and the upper »3'ellow» old red sandstones 

 of Dura Den, of Ireland and of the edge of the Forest of Dean. No Cephalaspids occur in 

 this upper division — whiist Glyptodipterini and a number of Crossopterygians simi- 

 lar to Carboniferous forms are present. Characteristic and confined to this horizon 

 are the Placodermata (Coccosteus and Ptericthys). 



The icthyological evidence in favour of splitting the Devonian system into two 

 parts and assigning the lower division to the Silurian System and the upper division 

 to the Carboniferous — is strengthened by the evidence from other classes of organic 

 remains. The essentially Silurian Eurypterina (Pterygotus) are confined to the lower 

 of the two divisions. 'Ihe flora of the upper division is on the other hand essentially 

 Carboniferous — i. e. consists of Carboniferous genera very largely. The Brachiopoda 

 of the lower division are Silurian in facies — so are its Trilobites. — The represen- 

 tatives of the same groups in the upper division are essentialh^ Carboniferous. 



Without pursuing this subject any further or endeavouring to establish here on 

 palasontological evidence the illogical and fallacious character of the classification of 

 the palasozoic strata, which recognizes a »Devonian system» instead of assigning the 

 beds there included to the systems above and below it — enough has been said to 

 shew the importance of the question which is presented by the Spitzbergen Fish Re- 

 mains. Should the Mimers dal siiales which contain Holoptychian fish remains prove 

 to be conformable with and not widely remote vertically from the red sandstones of 

 Dickson Bay — then we have a much nearer association of the Cephalaspidian and 

 Glyptodipterine fish-fauna than has hitherto been known. On the other hand the 

 proba-bilities are that there is a very coiisiderable gap, concerning which the field 

 geologists may be able to furnish some direct evidence, betAveen the red Sandstones 

 of Dickson Bay and the Mimers dal shales — the former being true Lower Devonian 

 (Silurian) the latter being true upper Devonian (Carboniferous). 



II. Description of the specimeiis. 



From amongst a large number of specimens more or less fragmentary I have 

 selected those which are sufficiently chai-acteristic either to make it worth while to have 

 them draAvn in the plates for future comparison and reference or to enable me to assign 



