of Palaeospondylus, Traquair. 299 



means an extravagant demand looking to the great variety in the 

 teeth of Dipnoi, both in different species of the group, and in 

 different stages of the life history of the same species) to give an 

 appearance which would be extremely like that of Palaeospondylus. 

 Finally the Dipnoi possess structures which might in a fossil well 

 give rise to appearances suggestive of cirrhi round the mouth 

 opening — in the cartilaginous bars of the roof of the nasal capsule, 

 and in those other cartilaginous processes, whether representing 

 merely labial cartilages or processes of the original lower jaw, 

 which project upwards in the region of the lower lip in Lepidosiren. 



In one of Traquair's figures 1 a pair of plates are shown con- 

 verging forwards at an angle beneath the hind portion of the 

 skull. These recall strongly the two moieties of the strongly 

 developed Hyoid arch in Lepidosiren. 



Dr Traquair says that it is only the ventral side of the skull 

 which is exposed in all his specimens. There seems no very 

 special reason why this should be so in a Cyclostome, while in 

 a young Dipnoan on the other hand, the cranial roof being still 

 incomplete, it is obvious that the dorsal side would be firmly 

 imbedded in the matrix and would not afford a plane along which 

 the stone would readily split. 



It is to be noted that Dipnoans were abundant about the 

 times in which Palaeospoudylus lived {Dipterus is one of the most 

 frequently occurring genera in the rocks of the quarry from 

 which Palaeospoudylus is obtained and the Arthrodiran Coccosteus 

 and Homosteus also occur, Traquair), while of undoubted Cyclo- 

 stomes beyond Palaeospondylus there is no trace. 



A point apparently against the Dipnoan nature of Palaeo- 

 spondylus lies in the highly-developed Vertebrae, though this 

 objection tells obviously equally against its being a Cyclostome. 

 If however the tubular centra are really, as Dr Traquair suggests, 

 formed in the sheath of the notochord their presence would tell 

 much more strongly against the latter view than against the former, 

 for, as Dr Gadow 2 has well accentuated, while the Dipnoi are poten- 

 tially chordo-centrous, the Cyclostomes are essentially arcocentrous. 



I have now referred to the main points which have struck me 

 of resemblance between Palaeospoudylus and a young Dipnoan. 

 In regard to the structure of Palaeospondylus I have preferred 

 to depend entirely upon the data set forth by the unrivalled skill 

 in this department of investigation by Dr Traquair himself. 

 Whether the resemblances are superficial or not they seem to 

 me at least sufficiently striking to deserve notice, as suggesting 

 the possibility of Palaeospondylus having been either a young 

 Dipnoan, or the young of some form closely allied to the Dipnoi. 



1 (3) Plate ix. Fig. 2. 



2 Phil. Trans. Vol. 186 (1895), B, p. 191. 



