Dr R. H. Traquair on Chondrosteus acipenseroides. 301 



chain, and where the post-temporal is movably articulated, 

 — whereas in Aeipenser the frontals are entirely separated 

 by an intercalated plate, and the post-temporals and two of 

 the median body-plates are immovably joined to, and form 

 a part of the cranial buckler itself. Especially palaeoniscoid 

 is, however, the aspect of the opercular and branchiostegal 

 apparatus, as will be seen if the reader will compare the 

 restored drawing (Fig. 5) with the figures of the heads of 

 Palceoniscus and Nematoptychiits given in my account of the 

 Palseoniscidae, although in Chondrosteus there is no prseoper- 

 culum and the series of branchiostegal rays of the two sides 

 may not have met in the middle. The special resemblance 

 of the shoulder-girdle to that of Falmoniscus is also very 

 striking, especially in the form of the post-temporal and 

 supra-clavicular bones. 



In my already quoted essay on the structure of the Pala^- 

 oniscidse, I pointed out certain strange and previously unre- 

 cognised resemblances which Palceoniscus bore to Folyodon, 

 especially in the internal skeleton, the shoulder-girdle, and 

 the jaws and palato-quadrate apparatus (even although there 

 are premaxillary bones and there is no evidence that the 

 palato-quadrate elements met in the middle line in front). 

 I also remarked that the resemblances between Palceoniscus 

 and Acipenser are of course much less prominent. Here, 

 however, is a form which in many of its features presents 

 strong Palaeoniscoid resemblances, but whose afiinities are, 

 nevertheless, more with Acipenser than with Polyodon ! 



The af&nities of Chondrosteus seem, therefore, to radiate in 

 three directions, towards Acipenser, towards Polyodon, and 

 towards the PahTeoniscidse, and certainly, of all the three 

 directions, the distance towards Acipenser is the least. 



