1904.] OSTEOLOGY OP THE ELOPID.E AND ALBULIDtE. 63 



is no such cavity. The swim-bladder diveiticulum in Hyodon and 

 Notopter%is is of an entirely difierent character. It is large in 

 size, its outer wall is of fibrous tissue, and its inner wall is con- 

 stituted by the exoccipital and basioccipital in the former, and 

 the opisthotic and basioccipital in the latter genus. The want of 

 uniformity in the relations of the air-vesicles in the above- 

 mentioned fishes, coupled with the occurrence of such vesicles 

 in Mormyroids, in which the relations are yet again difierent, 

 points to the conclusion that such adaptive features cannot be 

 relied upon to any large extent in determining whether any two 

 fishes are closely or distantly related. 



The opisthotic or intercalary bone does not appear to be at all 

 constant in bony fishes, as has already been pointed out by Yrolik 

 (Nied. Arch. f. Zool. i. 3, 1873), Klein (Jahresh. Yer. vaterl. 

 Naturk. Wurtt. xxxv. 1879), and Sagemehl (Morphl. Jahrb. ix. 

 2, 1883). It must here be borne in mind, however, that what 

 Gill calls the opisthotic is not the bone that is now genei'ally 

 known as the opisthotic, but the squamosal, which is of invariable 

 occurrence. (See Proc. U.S. Nat. Mus. xiii. 1890 (1891), pi. 30. 

 figs. 2-4, and pi. 31. figs. 2, 4, & 5.) In this matter he appears 

 to be following the now obsolete terminology of Huxley (see 

 Glinther, ' Study of Fishes,' 1880, p, 60). Except in the Gadoid 

 fishes, the opisthotic is never of very large size. It is well 

 developed in Hyodon and Megalops, and is moderately large in 

 Notopterus, Gonorhynchus, Osteoglossum, Arapaima, a.nd Heterotis ; 

 in most of the Clupeidas it is small ; it is absent in Engraulis and 

 Coilia, in the Mormyridse and in Alepocephahos. 



As a general rule, in the forms under consideration, the lateral 

 wing of the parasphenoid that passes up along the anterior 

 edge of the pro-otic is of very small extent. In Osteoglossum 

 leichardti, however, it rises so high as to meet the alisphenoid, and 

 in Osteoglossum bicirrhosum, Osteoglosswin fovmosum, and in 

 Gonorhynchus it enters into relation with both alisphenoid and 

 postfrontal. This is not exactly comparable with what occurs in 

 Am.%a, for in that genus there is one long process of the para- 

 sphenoid to the endosteal postfrontal (sphenotic), and a separate 

 shorter one to the alisphenoid. It may here be noted that 

 Swinnerton (Quart. Journ. Micro. Sci. xlv. 4, 1902, p. 532), in 

 mentioning the union of processes of the parasphenoid and f lontal 

 immediately in front of the postfrontal bone in Gastrosteus, quotes 

 Klein as having recorded a similar relation obtaining in the case 

 of Lophius. The junction, however, in Lophius occurs in front of 

 the optic foramen, and is in no way comparable with the above. 



Although the view of Leydig (Zeitschr. f. wiss. Zool. v. 1854), 

 Hertwig (Arch. f. mikr. Anat. xi., Suppl. 1874 ; also Morph. 

 Jahrb. ii. 1876), Sagemehl (Morph. Jahrb. ix. 2, 1883), and 

 Klaatsch (Morph. Jahrb. xxi. 2, 1894), that such bones as the 

 vomer and pai'asphenoid have arisen by the coalescence of tooth- 

 bases, is not shared by Walther (Jen. Zeitschr. xvi. 1882) and 

 Carlsson (Zool. Jahrb. viii. 2, 1894), who claim that tooth-bearing 



