Classification of the Order Scleroparei. 183 



the pelvic fins, the pelvic bones not articulating* with the 

 cleithia, and tlie development of a pair of dermal plates, 

 ectocoracoids, ankjlosed with the hypocoracoids in the alult 

 fish. 



Suborder 3. Dactylopteroidea. 



Mesethmoid unossified ; nasals ankylosed to form a large 

 median plate; no opisthotic; second suborbital a small bone 

 movably articulated with first and with prseoperculum 

 (fig. 1, A)*. Post-temporal a very large plate suturally united 

 to epiotic, ptorotic, and exoccipital ; supra-cleithrurn wedged 

 in and rigidly united to upper end of cleithrum ; pelvic bones 

 directly attached to cleithra. First three vertebra? elonorate, 

 united by suture ; first attached by suture to basioccipital. 

 Pectoral base subhorizontal. 



In addition to these diagnostic characters may be men- 

 tioned : the parasi)henoid sends up a pair of wings which 

 meet the frontals and alisphenoids in the interorbital region, 

 the third suborbital emits a broad subocular lamina, the 

 hyo-palatine and oj)ercu]ar bones are all present, tliere are 

 three pairs of dentigerous upper pharyngeals, tiie pectoral 

 radials number four, the lower elongate hourglass-shaped, 

 and there are no ribs, but a series of epipleurals on the free 

 pisecaudal and anterior caudal vertebrae. 



This suborder is a remarkably well-defined one, quite 

 dissimilar to the other Scleroparei. Gill has placed the 

 Dactylopteridaj with tiie Triglidaj to form the order Craniomi, 

 on account of supposed resemblances in the structure of the 

 post-temporal and supra-cleithrum t, but has recognized their 

 divergence in other characters. 



Family Dactylopteridae. 



I have examined skeletons of Dactylopterus volitans and 

 D actiflcplena orientalis. In both there are 22 vertebrae 

 (3 + 6+13) ; these are compressed and from the sixth to the 

 nineteenth bear paired processes directed obliquely upwards 



* This small bone, the " pontinal" of Gill, may be the sejjmented off 

 distal end of the second suborbital, in which case the first may be fused 

 with the pr^orbital ; the intei-pretation given above is more probably 

 correct. By this device the strong spine at the angle of the prasoper- 

 culum acquires great freedom of movement without sacrificing the 

 support of the suborbital stay. 



t The supra-cleithrum of the Dactylopterids was evidently mistaken 

 for part of the cleithrum, and one or more bony scales that adhere to the 

 lower part of the posterior end of the post-temporal may have been takea 

 for the supra-cleithrum. 



