582 H. H. SWrNNKRTON. 



those forms in which it is constantly present its relationships 

 are extremely variable. Nevertheless it cannot be a mere 

 coincidence that all the forms belonging to the Scomber- 

 esocine series, including so lowly a form as Belone, should 

 be lacking in this bone. It must therefore be lacking also, 

 or much reduced, in their immediate ancestors and allies. 



Turning now to the palfeontological side of the question, 

 we find the Thoracostei well represented by forms differing 

 but little from those of the present day as early as the 

 Eocene. 



There seems at present much more to be said against than 

 for Kingsley's suggestion concerning the relationship between 

 these forms and the extinct Dercetiformes or Hoplopleurides 

 (Pictet, 54, p. 215). Smith Woodward, to whom I am 

 indebted for tlie opportunity of examining a number of the 

 original fossils, places Belonorhynchus, with considerable 

 show of reason, among the Ohondrostei (95), and the re- 

 maining forms in two families, the Dercetidae and Bncho- 

 dontidse (01). Despite the presence of rows of bony scutes 

 along dorsal, lateral, and ventral lines, the condition of the 

 cranial roof and mandibular suspensorium alone make it 

 very improbable that they have anything to do with the 

 Thoracostei. 



One other group of bony fishes — Zanclidas, Acronuridae 

 (Teuthididse), and Plectognathi — demands a reference be- 

 cause it possesses the acrartete condition. Jordan and 

 Evermann (96) place the first two families together with 

 the Ephippidge and CliEetodontidas in one sub-order, the 

 Squamipinnes, and say that "The Teuthididae and Balis- 

 tidae are as nearly related to each other as the Ephippidae 

 are to the Chastodontidae ; " and again : '^ There can be no 

 doubt of the common origin of the Balistidae and Teuthididae, 

 and that the divergence is comparatively recent." For this 

 reason they subordinate the Plectognathi to the Acantho- 

 pterygii, and place them as an offshoot of the Squamipinnes. 



What has already been said concerning this ethmoid 

 and suspensorial region in Teuthis and Balistes tends to 



